2016 ASCB Meeting-Oral Abstracts

ascb.org/2016meeting

peter walter, president • jonathan weissman, program chair

san francisco, california dec 3-7

abstracts: oral presentations

download the ascb meeting appsearch for “ascb”

ORAL PRESENTATIONS‐ SUNDAY

Oral Presentations‐Sunday, December 4

Symposium 1: Mechanical Forces in Cell Biology

S1 Rapid cell migration through dense tissues ‐ it is easier for a camel to go through the eye of a needle. M. Piel1; 1UMR144/IPGG, Institut Curie, Paris, France The quest to understand how the microenvironment of cells impacts their behavior and fate has been a major force driving the development of new technologies in cell biology research, in order to bridge the gap between the classical and simple cell culture plate and the biological reality of actual tissues. This question is particularly relevant to dendritic cells (DCs) of the immune system that patrol the body seeking out foreign invaders, such as bacteria or viruses. To achieve this function, DCs have to cross various types of tissues by migrating between other cells or squeezing through dense extracellular matrix. Guided by chemokines and rapidly maneuvering through complex environments, DCs experience large deformations of the cell body and the nucleus. Their final destination is the closest lymph node, where they will trigger an adaptive immune response. Following a DC through its journey from the skin to the lymph node, I will present recent work we performed to understand what it takes for this important cell to achieve its function and alert the immune system efficiently. We used various in vitro systems aimed at decomposing the complexity of the migration environment into a series of elementary constraints, including 2D confinement, microchannels, collagen gels, and tissue explants. I will show how cytoskeletal actin filaments contribute in various manners, and in various locations in the cell, to specific stages of DC’s journey. I will also discuss whether some of these findings can extend to other types of migrating cells. I will propose key elements to start building an integrated image of migratory behaviors of cells by considering sub‐regions of a generic phase diagram reflecting states of the acto‐myosin cytoskeleton. Finally, I will discuss whether it is possible to predict features of large‐scale cell trajectories based on these general considerations.

S2 Under pressure: epithelial cell population control. J. Rosenblatt1; 1Oncological Sciences, Huntsman Cancer Institute, University of Utah, Salt Lake City, UT All of our organs are encased and protected by simple epithelium. Despite their collective function as a barrier, epithelial cells turnover at the fastest rates in the body through cell death and division. Little was previously known about what controls epithelial cell death or division or how the two processes are linked to maintain constant cell densities. We discovered that mechanical tension governs both processes: when cells become too numerous, crowding triggers cells to squeeze out of the layer and die, when they become too sparse, stretch activates rapid cell division. Through identification of the signaling and mechanisms that control the fundamental processes controlling cell numbers, we have identified new etiologies for diseases like asthma and pancreatic cancers. These discoveries may now lead to better treatments for these diseases.


S3 Forcing Form and Function. V.M. Weaver1; 1University of California, San Francisco, San Francisco, CA Cells experience force and possess mechanotransduction machinery to detect physical cues from their microenvironment and to transduce and biochemically amplify these signals to modulate their differentiation, growth, survival, migration and morphogenesis. Tumors show increased tissue level forces and transformed cells exhibit a perturbed mechanophenotype. We have been studying how cells transduce mechanical cues to regulate their fate and how altered force compromises tissue homeostasis to drive malignancy and metastasis. We found that the ECM in breast, pancreas, skin and glioblastomas is remodeled during malignancy such that the ECM stiffens as a function of tumor progression and aggression. A stiffened ECM compromises tissue differentiation by promoting integrin focal adhesion (FA) assembly that potentiates transmembrane receptor signaling and induces cytoskeletal remodeling and actomyosin contractility. Our findings indicate that high mechanosignaling collaborates with oncogenes including Ras and ErbB2 and reduces the levels of key tumor suppressors including PTEN and BRCA1 to drive transformation. By this means an increase in ECM stiffness promotes the malignant transformation of pre malignant cells and drives tumor cell metastasis, and inhibiting ECM stiffening restores tumor suppressor function and reduces oncogenic signaling to prevent malignancy. We determined that these effects are linked to the cells ability to "tune" their actomyosin tension to the stiffness of their surrounding microenvironment which induces focal adhesion assembly through a vicious feed‐forward mechanism. The elevated cell contractility and increase in focal adhesions modifies cell signaling through RhoGTPases, modify ERK, PI3 kinase, Myc and Wnt signaling to amplify proliferation and survival, and invasion, to increase the level of factors such as VEGF that drives angiogenesis, to stimulate cytokine and chemokine expression that fosters inflammation and to induce hypoxic reprogramming to drive tumor metastasis and compromise therapeutic efficacy. Interestingly, we showed that oncogenic transformation and the loss of key tumor suppressors per se also elevate cell intrinsic tension and drive tumor progression and aggression by inducing ECM remodeling and stiffening to elevate mechanosignaling. We could show that the cancer cells genotype "tunes" the stromal response of the tissue and influences tumor aggression in part by modifying tissue tension. We will present evidence to support a role for both intrinsic and extrinsic mechanical force in solid tumors and discuss how these findings could provide insight to guide and improve cancer therapy (Supported by DOD BCRP BC122990 to VMW, NIH NCI R01CA192914, U54CA163155‐01, R01CA174929, R33 CA183685‐01 and 1U01CA202241‐01 to VM

Symposium 2: Organelle Organization

S4 The behavior of mitochondria. J. Nunnari1; 1Department of Molecular and Cellular Biology, College of Biological Sciences, University of California, Davis, Davis, CA Mitochondria are double membrane‐bounded organelles that perform a myriad of diverse and essential functions dependent on the collective behavior of the organelle within cells. We have characterized the separate, yet somehow connected features of the machines that control mitochondrial behavior. Our work has addressed the physiological functions and mechanisms of mitochondrial division and fusion,


which are important determinants of overall mitochondrial shape and distribution. We have uncovered contact sites that intimately link mitochondria with the ER and determined their roles in mitochondrial positioning, dynamics and mitochondrial genome transmission. We have also addressed the fundamental question of how mitochondrial membranes are sub‐compartmentalized to reveal how the complex internal architecture of the organelle is generated. We are currently focused on how mitochondrial behaviors are integrated with one another and physiologically regulated within cells and organisms.

S5 The Cell Biology of Lipid Storage. T.C. Walther1,2,3; 1Cell Biology, Harvard Medical School, Boston, MA, 2Genetics and Complex Diseases, Harvard School of Public Health, Boston, MA, 3HHMI, Boston, United States All organisms face fluctuations in the availability and need for metabolic energy. To buffer these fluctuations, cells use neutral lipids, such as triglycerides, as energy stores. We study how lipids are stored as neutral lipids in cytosolic lipid droplet organelles. Specifically, we will present our current understanding on the molecular processes that govern the synthesis of energy storage lipids as well as their storage in and mobilization from lipid droplets.

Minisymposium 1: Bacterial Mechanics, Development, Division, and Polymers

M1 Mechano‐microbiology: How physical forces impact bacterial pathogenesis. Z. Gitai1; 1Molecular Biology, Princeton University, Princeton, NJ Bacteria, like all organisms, organize their subcellular components to efficiently grow and adapt to constantly changing environments. Our lab uses imaging‐based approaches to study how bacterial cells use cytoskeletal polymers to establish robust shapes, compartmentalize metabolic networks, and form complex multicellular communities. Recently we discovered that some pathogens, like Pseudomonas aeruginosa, a major cause of hospital‐acquired infections and cystic fibrosis complications, develop virulence in response to sensing mechanical cues from their hosts. Historically, single‐cell organisms were thought to only respond to chemical cues like quorum sensing or nutrient availability, but our data indicate that bacteria also “feel” the presence of their hosts and integrate this information into their developmental response. We have characterized two distinct mechanosensation pathways: a short‐time‐scale response mediated by Type IV Pili, and a long‐time‐scale response mediated by PilY1. P. aeruginosa can infect virtually every host cell type assayed, so mechanosensitive pathways enable the bacteria to regulate the energetically costly process of virulence using host‐nonspecific environmental cues. We also find that mechanical environments can dramatically impact bacterial colonization, as species like P. aeruginosa that have polarized motility structures use the regulated retraction of adhesive appendages to migrate upstream against fluid flows. By moving upstream, these "bacterial salmon" colonize niches like catheters and mucosa that flush away most bacterial species. By revealing fitness benefits that are not obvious in the absence of flow, we can thus use microfluidics to both better model the naturalistic environments of bacteria and to understand the selective benefits of features such as bacterial polarity.


M2 A Diffusion Trap at the Caulobacter Cell Poles Leads to Spatially Resolved Transcription. K. Lasker1, A. von Diezmann2, D.G. Ahrens1, T.H. Mann1, W.E. Moerner2, L. Shapiro1; 1Developmental Biology, Stanford University, Stanford, CA, 2Chemistry, Stanford University, Stanford, CA Caulobacter undergoes an asymmetric division during each cell cycle yielding daughter cells that differ in both structure and function. Prior to division, different complements of signaling proteins are localized to each cell pole. CtrA, an essential transcription factor that mediates differential gene expression, is activated by phosphotransfer from a polarly‐localized histidine kinase and a cytoplasmic phosphotransfer protein. We have demonstrated both theoretically and in vivo that a polar volume‐filling polymeric matrix functions as a diffusion trap for proteins in the CtrA activation signaling pathway. Because Caulobacter’s single circular genome is aligned with the long axis of the cell, a consequence of this selective diffusion trap is that a gradient of the activated CtrA transcription factor links the genomic position of CtrA‐ regulated genes with their expression level. The spatial regulation of gene expression was explored first by measuring the contribution of genomic position to expression levels of CtrA regulated genes, and second by constructing a reaction‐diffusion model to test the ability of different distributions of CtrA and its upstream partners to generate the observed spatially distinct gene expression profiles. We concluded that spatial regulation of gene expression could only be achieved if diffusion of proteins in the phospho‐signaling pathway activating CtrA was slowed down in the polar polymeric matrix microdomain, where the signal is generated. We then tracked the movement of members of the phospho‐signaling pathway relative to the polar matrix in vivo using two‐color single‐molecule tracking and super‐resolution imaging and observed that indeed the polar polymeric matrix creates a zone of restricted diffusion, which facilitates both the efficient phospho‐transfer reactions within the polar domain and a polarized distribution of the activated CtrA transcription factor. Finally, we demonstrated that the size of the diffusion trap alters the gene expression profile of CtrA activated genes.

M3 A Conserved Switch Regulates Phosphatase Activity to Initiate Cell‐Specific Transcription in Response to Spatial Cues. N. Bradshaw1, V.M. Levdikov2, A.J. Wlikinson2, C.M. Zimanyi1, R. Gaudet1, R. Losick1; 1Molecular and Cellular Biology, Harvard University, Cambridge, MA, 2Structural Biology Laboratory, Department of Chemistry, University of York, York, United Kingdom To initiate cell‐specific transcription during spore formation by B. subtilis, the PP2C phosphatase SpoIIE is compartmentalized and activated in small cells resulting from asymmetric division. We identified self‐association of SpoIIE as a key molecular event that links the small cell‐specific localization, stabilization and activation of SpoIIE, but the mechanism by which self‐association relays regulatory information to control phosphatase activity remained undefined. We determined the x‐ray crystal structure of the SpoIIE phosphatase domain with a portion of its regulatory domain. In this structure, SpoIIE forms a dimer that is templated in part by the regulatory domain. Systematic mutagenesis of the dimerization interface suggests that dimerization is required for localization and stabilization of SpoIIE in small cells. This dimer additionally repositions two helices at the base of the catalytic center relative to their position in the structure of the isolated SpoIIE phosphatase domain. Genetic and biochemical evidence suggests that this repositioning switches the phosphatase from the inactive to the active conformation to communicate spatial and temporal information from the regulatory domain. From combined structural and genetic evidence, we propose that this switch is conserved throughout the PP2C family,


and is modularly accessed by diverse regulatory domains to control varied stress responses and developmental programs in all kingdoms of life.

M4 Mycomembrane Dynamics and V‐Snapping: Growth and Division of the Cell Envelope in Corynebacterium glutamicum. X. Zhou1, F.P. Rodriguez‐Rivera2, C.R. Bertozzi2,3, J.A. Theriot1,3,4; 1Biochemistry, Stanford University School of Medicine, Stanford, CA, 2Chemistry, Stanford University, Stanford, CA, 3Howard Hughes Medical Institute, Stanford, United States, 4Microbiology and Immunology, Stanford University School of Medicine, Stanford, CA Based on the results of Gram staining, bacteria are generally classified into two major groups depending on their cell envelope architecture. One fascinating exception to this classification is Corynebacteria and related acid‐fast species including the human pathogen Mycobacterium tuberculosis which do not retain the Gram stain but are structurally Gram‐positive with a thick layer of peptidoglycan and a mycolic acid‐based “outer‐membrane like” mycomembrane. In addition, Corynebacteria and Mycobacteria are known to elongate at the poles and undergo V‐snapping during cell division. We discovered that V‐snapping in Corynebacterium glutamicum occurs within milliseconds through a turgor pressure‐driven mechanical fracture mechanism, similar to the “popping” in Staphylococcus aureus . How is the complex cell envelope assembled in C. glutamicum to accommodate the polar growth and V‐snapping? To illuminate the cell envelope assembly strategy, we combined fluorescent chemical reporters engineered for specific incorporation into different layers of the cell envelope and quantitative microscopy to follow C. glutamicum cell cycle. In particular, the fluorescent trehalose reporters exploit the activity of the extracellular mycotransferases and are incorporated into the extractable/free trehalose mycolates (which mainly constitute the outer leaflet of the mycomembrane). We found that the free trehalose mycolates are relatively mobile in C. glutamicum (with an apparent diffusion coefficient of 10‐2~10‐1 µm2/s revealed by FRAP analysis), but in contrast are rather immobile in Mycobacterium smegmatis. Interestingly, despite the high mobility in C. glutamicum, trehalose mycolates are excluded from the septum during septation, and only migrate into the septum immediately prior to V‐snapping. Finally, we observed a correlation between the appearance of trehalose mycolates at the septum with the perforations formed at the peripheral ring observed by SEM, suggesting that the peptidoglycan layer at the periphery is likely responsible for the diffusion barrier for the free trehalose mycolates. Overall, our approach revealed the differential assembly of the cell envelope in C. glutamicum.

M5 Mechanistic interplay between coupled morphogenetic events in Caulobacter. E.L. Meier1, E.D. Goley1; 1Biological Chemistry, Johns Hopkins University School of Medicine, Baltimore, MD In the dimorphic bacterium Caulobacter crescentus, a series of morphogenetic events are coordinated over its cell cycle. Cytokinesis, for example, separates the cell into two, but also forms the new cell poles upon which an appendage called the stalk is later formed. In our laboratory, we characterize the mechanisms of cytokinesis and its relationship with other morphogenetic events. Here, we have characterized the function of the cell division ATPase FtsE which, with its transmembrane binding partner FtsX, forms an ABC transporter family complex. In other organisms, FtsEX activates cell wall hydrolysis to promote cell separation at the end of division. We hypothesized that FtsEX might also


serve as a membrane anchor for polymers of the tubulin‐like GTPase and central orchestrator of division, FtsZ, through interaction of FtsE with both FtsZ and FtsX. To address its function in Caulobacter, we deleted ftsE and characterized the growth, morphology, and FtsZ‐ring structure in the resulting cells. FtsE was required both for efficient cell separation and for assembly of a focused Z‐ring at midcell. Additionally, deletion of ftsE was synthetic lethal with deletion of the cell wall hydrolytic factor, dipM, and was synthetic sick with loss of the amidase AmiC or of the Z‐ring focusing factor, ZapA. We used bacterial two hybrid and genetic analyses to delineate the pathway from FtsEX to cell wall hydrolytic factors. These observations indicate that, indeed, FtsEX plays an important role in cell division by anchoring FtsZ to membranes and signaling to cell wall hydrolytic enzymes for cell separation. Surprisingly, loss of ftsE function led to cell chaining in a manner distinct from that observed in other organisms: chained cells were connected by thin, unresolved constrictions with envelope topology and dimensions similar to the Caulobacter stalk (~50‐100 nm in diameter and up to several µm long). This is in contrast to the thick, fully compartmentalizing septa observed for similar mutants in Escherichia coli. We hypothesized that (1) proteins that localize to the stalk might also localize to the thin constrictions of ftsE mutant cells, if their localization cue is stalk geometry and (2) that stalk morphogenesis in Caulobacter may be mechanistically related to the formation of these extended constrictions. Preliminary data support these possibilities: the stalk protein StpX is robustly recruited to the stalk‐like constrictions connecting chains of ftsE mutant cells and, without FtsE, the proteinaceous stalk bands that compartmentalize Caulobacter stalks do not appear to form properly. These data reveal an unanticipated mechanistic interplay between coupled morphogenetic processes in Caulobacter, namely cytokinesis and stalk development.

M6 Mechanistic insights of the Min Oscillator via cell‐free reconstitution and imaging. A.G. Vecchiarelli1, K. Mizuuchi1; 1National Institutes of Health, Bethesda, MD In E.coli, the Min system forms a standing wave that oscillates from pole to pole on the inner membrane to prevent division at the cell poles. Using purified and fluorescent‐labeled MinD and MinE, we find these proteins can self‐organize into a spectrum of patterns on a flat bilayer. Comparing and contrasting these modes of patterning in vitro has allowed us to probe the mechanism behind Min oscillation in vivo. We find that periodic depletion of MinD from the cytoplasm is essential for the standing wave. Also, we find that the role of MinE in the patterning mechanism goes much farther than simply stimulating MinD release from the membrane as subscribed by previous models. Rather, our data support the proposal that MinE successively recruits, stabilizes, releases, and inhibits MinD interaction with the membrane. These events are temporally regulated by the MinD‐to‐MinE ratio on the bilayer, which acts as a toggle switch between membrane‐binding and ‐release by MinD, which ultimately drives the Min oscillator.

M7 Along for the ride: directional motion of FtsZ filaments controls cytokinesis in B. subtilis. A. Bisson1, G.R. Squyres1, Y. Hsu2,3, E. Kuru2,3, M. Van Nieuwenhze2, Y. Brun3, E.C. Garner1; 1Molecular and cellular Biology, Harvard University, Cambridge, MA, 2Department of Chemistry, Indiana University, Bloomington, IN, 3Department of Biology, Indiana University, Bloomington, IN


Cell division is a process present in every living organism, and whereas multiple components of different bacterial division machineries have already been identified, considerably little less is known about the mechanismtic details controlling this process. Towards addressing this end, wWwe studied the dynamics of some of the key division components in Bacillus subtilis. FtsZ, a prokaryotic homolog of tubulin, self‐associates into filaments, to forming a ring‐like structure (Z ring) that to establishes the site of future cell division. We showed by TIRF microscopy that FtsZ filaments present directional, circumferential motion both inside and outside the Z ring. Using Employing a number of genetic and chemical perturbations, we demonstrated that this motion is not driven by the cell‐wall synthesis machinery, but by the filaments’ own polymerization dynamics. By iInvestigating the dynamics of the monomers subunits within comprising these filaments, we discovered that they motion is are powered by treadmilling rather than translocation. Strikingly, we observed that individual molecules of Pbp2B, a putative D‐D‐transpeptidase enzyme essential for cell division, move directionally around the septa, and Pbp2B their dynamics are coupled with the motion of FtsZ treadmilling. filaments. Finally, we demonstrated that it is possible to tightly modulate control cytokinesis rates by applying systematically altering perturbations to in creasing or decreasing FtsZ polymerization dynamics. that increase or decrease filament velocities. Collectively, these results reveal dynamic Z rings comprising FtsZ filaments circumferentially traversing treadmilling around the division site, pulling the cell wall synthetic enzymes. This and suggest that this behavior is may be responsible for distributing enzymes that ultimately coordinate organized septum synthesis.

M8 Architecture of a lipid transport system for the bacterial outer membrane. G. Bhabha1, G. Greenan1, S. Ovchinnikov2, J.S. Cox1, R.D. Vale1, D.C. Ekiert1; 1University of California, San Francisco, San Francisco, CA, 2University of Washington, Seattle, WA How phospholipids are trafficked between the bacterial inner and outer membranes through the intervening hydrophilic space of the periplasm is not known. Here we report that members of the mammalian cell entry (MCE) protein family, which were previously implicated in outer membrane function, form hexameric rings with a central hydrophobic channel that can mediate phospholipid transport. The E. coli MCE protein MlaD forms part of a large multi‐protein complex in the inner membrane that includes an unconventional ABC transporter. MlaC, a soluble lipid‐binding protein, interacts with MlaD and an outer membrane protein complex, suggesting that it ferries lipids between the two membranes. In contrast, the syringe‐like architecture of a second E. coli MCE protein, PqiB, creates a continuous channel of sufficient length to potentially span the entire periplasmic space. This work reveals a system of highly conserved protein‐based channels that translocate lipids between the inner and outer membranes of bacteria and some eukaryotic organelles.

M9 Structural studies of actin‐like filament a (AlfA) and Alf‐A driven plasmid segregation. G.D. Usluer1, J.M. Kollman1, E.J. Charles2; 1Biochemistry, University of Washington, Seattle, WA, 2 Department of Cellular and Molecular Pharmacology, University of California , San Francisco, CA Structural studies of actin‐like filament a (AlfA) and Alf‐A driven plasmid segregation Most bacteria possess plasmids that encode genes beneficial for the host, such as antibiotic resistance, toxins and virulence factors (1). To ensure their equal distribution into daughter cells, low copy number plasmids are equipped with various endogenous partition systems (1,2). Type II partition system includes an actin homologue that drives plasmid movement towards opposite poles of a dividing cell, a


DNA binding protein to serve as an adaptor between the actin‐like filament and plasmid DNA and a centromere‐like DNA sequence for adaptor protein to bind during partition (3,4). Recent structural studies showed that each bacterial actin possess a unique filament architecture, while biochemical studies revealed their differences in polymerization dynamics (3). Actin‐like filament A (AlfA) is found in low copy plasmids in selected Bacillus strains, and has been shown previously to have unique dynamic properties as well as novel filament architecture (5). Here, we present the AlfA structure generated by cryo‐electron microscopy to sub‐nanometer resolution. This structure emphasizes distinct features of AlfA filament architecture amongst bacterial actins. A subdomain deletion within the AlfA monomer is likely to be the cause of highly twisted filament architecture and the bundles formed through lateral contacts between the individual AlfA filaments are thought to be an important component to establish a bi‐oriented partitioning system. The link between the AlfA, DNA binding protein AlfB and DNA sequence ParN provides insight to nucleation and polymerization dynamics of AlfA filaments and allows us to propose a partition mechanism based on structure and function relations. References 1. Couturier, M., et al. (1988). "Identification and classification of bacterial plasmids." Microbiol Rev 52(3): 375‐395. 2. Austin, S. J. (1988). "Plasmid partition." Plasmid 20(1): 1‐9. 3. Ozyamak, E., et al. (2013). "Bacterial actins and their diversity." Biochemistry 52(40): 6928‐6939. 4. Jensen, R. B. and K. Gerdes (1997). "Partitioning of plasmid R1. The ParM protein exhibits ATPase activity and interacts with the centromere‐like ParR‐parC complex." J Mol Biol 269(4): 505‐513. 5. Polka, J. K., et al. (2009). "The structure and assembly dynamics of plasmid actin AlfA imply a novel mechanism of DNA segregation." J Bacteriol 191(20): 6219‐6230.

M10 Electron Cryo‐Tomography Visualizes the Spindle Apparatus and Nucleus‐like Structures Formed During Infection by ΦKZ Family Bacteriophages. A.F. Brilot1, V. Chaikeeratisak2, K. Khanna2, E. Villa2, J. Pogliano2, D.A. Agard1; 1Biochemistry and Biophysics, University of California, San Franscisco/HHMI, San Francisco, CA, 2Department of Chemistry Biochemistry, University of California, San Diego, La Jolla, United States ΦKZ‐like bacteriophages are a genus of virulent phages with very large genomes (210‐320 kbp) which infect Pseudomonads. These phages encode the tubulin gene PhuZ which forms a spindle apparatus used to center phage DNA in infected bacteria. To extend in vivo light microscopic observations, we have used cryo‐FIB milling of infected cells with electron cryo‐tomography to visualize the infection process and PhuZ spindle at high resolution. Tomograms reveal the structure of PhuZ filaments in cells extending from the bacterial poles. They further reveal that phage DNA is segregated from the cytoplasm by a protein shell in a nucleus‐like structure. These structures form immediately after infection and grow to about 0.7 µm in diameter as the phage DNA replicates. While they have planar edges reminiscent of an icosahedral structure, they can grow and fuse during the infection. We further visualize the assembly of phage capsids near the cell periphery followed by DNA encapsidation at the surface of the nucleus‐like structure, suggestive of a transport process between the nucleus and the capsid.


Minisymposium 2: Cell Cycle Regulation and Decisions

M11 Towards an integrated mesoscale model of budding yeast cell size control. K.M. Schmoller1, D. Chandler‐Brown1, Y. Winetraub1, J.M. Skotheim1; 1Department of Biology, Stanford University, Stanford, CA While it has long been clear that cells actively regulate their size, the molecular mechanisms underlying this regulation have remained unclear. In budding yeast, it is known that size modulates division rate, primarily by determining the timing of the G1/S transition in the cell division cycle. We have recently shown that this is achieved by dilution of the cell cycle inhibitor Whi5 during cell growth in G1. However, recent phenomenological studies in yeast and bacteria have shown that cells increase approximately a fixed volume during their entire cell cycle, independent of what size they are born. These results seem to be in conflict, as the phenomenological studies suggest that cells measure the amount they grow, rather than their size, and that size control acts over the whole cell cycle, rather than specifically in G1. Here, we propose an integrated model that unifies the adder phenomenology with the molecular mechanism of G1/S cell size control. Using single cell microscopy combined with analytical and computational models, we analyze the implications of our Whi5 dilution model on population cell size distributions. We calculate cell size distributions by combining our G1/S model with measured correlations for cell growth and S/G2/M duration. We find that our model predicts the size‐independent volume increase during the full cycle. This suggests that the adder phenomenon is an emerging property of several aspects of the budding yeast cell cycle rather than a causal consequence of an underlying molecular mechanism sensing and actuating a fixed volume increase. More broadly, modeling of the cell cycle has tended either towards integrating nearly all known molecular mechanisms or towards remaining phenomenological and excluding molecular mechanisms. The problem with highly mechanistic models is that too many parameters are unknown; whereas the problem with phenomenological models is that they do not easily integrate genetic data to provide any molecular insight. Here, we take a third path to produce a mesoscale model that integrates well known molecular information with phenomenological modeling. The success of our approach to provide novel insight into budding yeast size control demonstrates the power of this mesoscale modeling approach, which may be applicable to other cellular phenomena.

M12 Cell cycle exit to a quiescent state involves alternative isoform use. M. Mithun1,2, E.L. Johnson3, D.C. Corney1,2,3, D.G. Robinson4, D. Jelinek1,2, A. Ambrus1,2, D.G. Taylor1,2, W. Wang4, S. Batista5, H.A. Coller1,2; 1Molecular, Cell and Developmental Biology, UCLA, Los Angeles, CA, 2Biological Chemistry, David Geffen School of Medicine, Los Angeles, CA, 3Molecular Biology, Princeton University, Princeton, NJ, 4Lewis‐Sigler Institute, Princeton University, Princeton, NJ, 5Computer Science, Princeton University, Princeton, NJ In response to a wound, quiescent fibroblasts are activated to migrate toward the wound, proliferate, and participate in the wound healing response. We hypothesized that the transition between quiescence and proliferation reflects not only changes in gene expression, but also changes in pre‐mRNA processing events. Using RNA‐Seq, we discovered widespread transcriptome reprogramming that led to reduced expression of genes involved in mRNA processing, including splicing and cleavage and polyadenylation factors. RNA‐Seq revealed 1975 significant examples of changes in the expression of


gene isoforms in proliferating versus quiescent fibroblasts. Introns were more likely to be retained in quiescent fibroblasts, which may reflect a decrease in the levels of some accessory splicing factors. To investigate specific changes in the selection of polyadenylation sites, we used polyadenylation site‐enriched RNA‐Seq to discover 454 instances of significant alternative polyadenylation with quiescence. Compared with proliferating fibroblasts, quiescent fibroblasts tend to use more distal polyadenylation sites. The synthesis of longer transcripts from more distal polyadenylation sites was associated with higher expression levels of these transcripts and a stabilization of the transcripts in the quiescent state based on genomewide transcript decay analysis. Three cleavage and polyadenylation factors (CstF64, CFIm25, and CPSF73) were expressed at higher levels in proliferating compared with quiescent fibroblasts, and were also present at higher levels in proliferating fibroblasts adjacent to wounds than in fibroblasts in unwounded skin. Knockdown of cleavage and polyadenylation factor CstF‐64 resulted in reduced use of more proximal polyadenylation sites, reduced proliferation, and decreased ability to migrate into a denuded area, thus recapitulating the characteristics observed in quiescent fibroblasts. Our findings support alternative polyadenylation as a regulatory mechanism that contributes to the transition to cell division and migration in response to proliferative signals. Taken together, our results underscore the importance of mRNA processing events for cellular quiescence and its associated phenotypes.

M13 Examining chromatin structure in different states of G0. Y. Ma1, D.J. Mckay2, L. Buttitta1; 1Molecular, Cellular and Developmental Biology, University of Michigan, Ann Arbor, MI, 2Dept. of Biology and Dept. of Genetics, University of North Carolina, Chapel Hill, NC States of cellular withdrawal from the cell cycle or G0 can range from readily reversible to permanently postmitotic. We are interested in how different states of G0 are controlled during development and why some are more reversible than others. Emerging evidence suggests a close relationship between a repressive chromatin environment and the silencing of cell cycle genes during the postmitotic state, but how the chromatin state differs between reversible and permanent G0 remains unclear. We focused our studies on the Drosophila pupal wing at a stage where the cells transition from active proliferation to a postmitotic state. We find there are two stages of G0 in this tissue, a flexible G0 period where cells can be induced to re‐enter the cell cycle under specific genetic manipulations and a state we call “robust”, where cells become strongly refractory to cell cycle re‐entry. We examined chromatin morphology changes in these different states of G0 by Fluorescent in situ Hybridization and found that chromatin becomes more compact in robust G0. To examine changes in the accessibility of regulatory elements during different states of G0, we carried out Formaldehyde‐Assisted Isolation of Regulatory Elements (FAIRE)‐seq during stages of proliferation, flexible G0 and robust G0 in Drosophila wings. We find that upon robust G0, the enhancers of specific key cell cycle genes such as Cyclin E and cdc25c become occupied by nucleosomes, blocking their accessibility to transcription factors. This suggests a potential role for nucleosome remodeling complexes in establishing and maintaining the silencing of cell cycle genes in G0. Consistent with this, in a screen for chromatin remodelers essential for proper G0, we identified Mi‐2, the ATPase component of the Nucleosome Remodeling and Deacetylase (NuRD) complex, as essential for proper robust G0.


M14 Neurogenic lncRNAs mutated in human neurodevelopmental disorders. C. Ang1, O. Wapinski2, Q. Ma2, S. Fan3, R.A. Flynn2, B. Coe4, M. Onoguchi2, B.T. Do2, J. Xu2, Q. Lee1, U. Elling5, J. Penninger5, E.E. Eichler4, A.K. Srivastava3, H. Chang2, M. Wernig1; 1ISCBRM, Stanford University, Stanford, CA, 2Dermatology, Stanford Universtiy, Stanford, CA, 3Greenwood Genetic Center, J.C. Self Research Institute of Human Genetics, Greenwood, SC, 4Howard Hughes Medical Institute and Department of Genome Sciences, University of Washington, Seattle, Seattle, WA, 5Institute of Molecular Biotechnology of the Austrian Academy of Science (IMBA, Vienna Biocenter, Vienna, Austria Long non‐coding RNAs (lncRNAs) are important in the establishment and maintenance of cellular identity and have increasingly been found to be involved in disease progression. Despite that, little is known about the roles of lncRNAs in neurogenesis. Adult cerebral cortex comprises of two main populations of neurons: excitatory projection neurons and inhibitory interneurons. Most cortical interneurons originate from progenitors residing primarily in the medial and caudal ganglionic eminences in ventral telencephalon. Those progenitors migrate tangentially to reach their nal positions in the developing cortex. Given the role of GABA as the primary inhibitory neurotransmitter in the cortex, deviations from the usual ratio of excitatory and inhibitory neuron is often suspected to be the cause of many neuropsychiatric diseases. Thus, studies on the development and function of inhibitory neurons are highly warranted. To investigate how coding and lncRNAs change their expressions during tangential migration from ganglionic eminences into the cortex, we decided to perform RNAseq for inhibitory progenitor and migratory cells at E13.5. Combining publicly curated lncRNAs lists and a novel lncRNA discovery pipeline reveals that there are 11102 dierentially expressed lncRNAs among the three populations (MGE/LGE/Cortex‐GFP) and 327 of them are within 50kb of neuronal transcription factors. Integrating excitatory neuronal datasets from Calegari and Arlotta group revealed that 170 of those are enriched in the inhibitory progenitors and migratory cells. Next, we were interested in seeing whether the same subsets of lncRNAs are enriched in broblast to neuron direct reprogramming. We discovered that direct lineage reprogramming of broblasts (MEF) to induced neu‐ ronal (iN) cells upregulates a subset of lncRNAs which are conserved in mouse and human brain development. Overexpression of those lncRNA candidates in MEF is sucient to promote iN maturation. Overlapping the candidates with the database of copy number variation morbidity map of patients with neurodevelopmental delay showed a recurrent focal genomic mutations aecting one lncRNA candidate (ncE). Finally, we located a triad family where ncE is the only transcript disrupted and only the father and the son inherited the truncated ncE display neurocognitive decits. Summing up, this shows how integration of candidates obtained from suciency screen done on the induced neuron platform with human genomic and mouse RNAseq data provides an exquisite insight into the eect of lncRNA in human neurogenesis.


M15 An mRNA competition mechanism regulates localized phase separation of liquid‐like P granules in C. elegans embryo. S. Saha1, C.A. Weber2, M. Nousch3, C. Hoege1, O. Adame‐Arana2, M. Hein4, E. Osborne‐Nishimura5, J. Mahamid4, M. Jahnel1, L.M. Jawerth1,2, A. Pozniakovski1, C.R. Eckmann3, F. Julicher2, A.A. Hyman1; 1Max Planck Institute of Molecular Cell Biology and Genetics, Dresden, Germany, 2Max Planck Institute for the Physics of Complex Systems, Dresden, Germany, 3Martin Luther University, Halle (Saale), Germany, 4Max Planck Institute of Biochemistry, Martinsried, Germany, 5Colorado State University, Fort Collins, CO The asymmetric segregation of components into daughter cells during cell division is a crucial problem in development. The early embryos in C. elegans asymmetrically segregate non‐membrane‐bound RNA‐protein compartments called P granules to the cells of the germline, where P granules contribute to maintenance of the germline cell fate. P granules are initially distributed throughout the one‐cell embryo, but segregate to the posterior following polarity establishment. It is thought that gradients of polarity proteins such as the mRNA‐binding protein MEX‐5 facilitate P granule segregation by driving localized phase separation of liquid‐like P granules selectively at the posterior of the embryo. However, the mechanisms by which polarity proteins regulate phase separation of P granules remain unknown. To address this, we reconstituted P granule‐like droplets in vitro using a single protein PGL‐3. We show that competition between PGL‐3 and MEX‐5 for binding to mRNA can regulate the formation of PGL‐3 droplets. Using theory we show that in a MEX‐5 gradient, this mRNA competition mechanism can drive a gradient of P granule assembly with similar spatial and temporal characteristics to P granule assembly in vivo. We conclude that gradients of polarity proteins can position RNP granules during development, by using RNA competition to regulate local phase separation.

M16 Cyclin‐dependent kinase control of motile ciliogenesis. E.K. Vladar1, M.B. Stratton2, T. Stearns2,3, J.D. Axelrod1; 1Pathology, Stanford University School of Medicine, Stanford, CA, 2Biology, Stanford University, Stanford, CA, 3Genetics, Stanford University School of Medicine, Stanford, CA The multiciliated cells (MCCs) of the airway epithelium each contain hundreds of motile cilia that beat in concert to propel inhaled contaminants out of the lungs, which is essential for respiratory function. Cilia are comprised of a centriole or basal body, which anchors it to the cell surface and the axoneme, which projects from the cell. Cells that adopt the MCC fate exit the cell cycle, launch a MCC‐specific transcriptional program to express ciliary components, then assemble and target centrioles to the surface to elongate the axoneme. Transcriptional regulators include both MCC‐specific factors and well‐known cell cycle‐related proteins. In cycling cells, centrioles duplicate only during S‐phase of the cell cycle, under common timing and regulation with DNA replication. Based on the presence of common regulators, we propose that post‐mitotic MCCs enter a unique cell cycle state (S‐phase), which permits centriole generation uncoupled from DNA replication. Cyclin‐dependent kinases (Cdks) are the principal regulators of cell cycle progression, and DNA and centriole duplication in S‐phase depend on Cdk2 coupled with Cyclins E or A2. To elucidate the nature of S‐phase in MCCs, we asked whether Cdk2 activity is required for centriole assembly. We found that in vitro differentiated primary mouse airway epithelial cells treated with Cdk2 inhibitors or expressing dominant‐negative Cdk2 fail to undergo motile ciliogenesis. We show that these cells still acquire the MCC fate, but do not activate the gene expression program or generate centrioles, indicating that Cdk2 is an early regulator of motile ciliogenesis. Cdk2


activity in MCCs appears to depend largely on CyclinA1, a cyclin known to activate Cdk2 in meiotic, but not normal somatic cell cycles. CyclinA1 is strongly upregulated during ciliogenesis, and cells lacking CyclinA1 have reduced numbers of MCCs. Cdk2 and CyclinA1 localize to the nucleus during early ciliogenesis, as well as to nascent centrioles, suggesting that Cdk2 controls the pathway both at the level of gene expression and centriole assembly. We propose that Cdk2 and other cell cycle regulators establish a permissive environment for centriole assembly in post‐mitotic MCCs, and MCC‐specific regulators such as CyclinA1 enable centriole amplification and the uncoupling of DNA and centriole duplication.

M17 Myosin phosphatase targeting subunit 1 (MYPT1) mediates the antagonistic actions of Cyclin A/Cdk1 and Plk1 on k‐MT attachment stability to promote efficient error correction in early mitosis. A.G. Dumitru1, S.F. Rusin1, A. Kettenbach1,2, D. Compton1,2; 1Biochemistry Cell Biology, Dartmouth Geisel School of Medicine, Hanover, NH, 2Norris Cotton Cancer Center, Dartmouth Geisel School of Medicine, Lebanon, NH The fidelity of chromosome segregation in mitosis is safeguarded by the precise regulation of kinetochore microtubule (k‐MT) attachment stability. During prometaphase, microtubules must detach frequently from kinetochores to ensure the efficient correction of abundant errors in k‐MT attachment in early stages of mitosis. We have previously demonstrated that cyclin A/Cdk1 promotes the rapid detachment of microtubules from kinetochores during prometaphase, but the mitotic substrates of cyclin A/Cdk1 are not known. Here, we used stable isotope labeling of amino acids in cell culture (SILAC) and mass spectrometry to identify a cyclin A/Cdk1 phosphoproteome during prometaphase. This strategy identified over 17,000 phosphopeptides with the minimal Cdk1 consensus motif, of which 252 independent phosphosites on 124 protein substrates with known mitotic functions were found to exhibit significant reduction in phosphorylation upon loss of cyclin A/Cdk1 in prometaphase. These proteins have demonstrated functional roles in spindle organization, at centrosomes, and at kinetochores. Targeted experiments show that one of these candidates, myosin phosphatase targeting subunit 1 (MYPT1), regulates k‐MT attachment stability. Loss of MYPT1 activity significantly reduces the rate of microtubule detachment from kinetochores in prometaphase leading to the hyperstabilization of k‐MT attachments. Cyclin A/Cdk1 activity is necessary for MYPT1 localization to kinetochores, and MYPT1 negatively regulates the quantity and activity of Plk1 at kinetochores. This work reveals an antagonistic relationship between two key mitotic regulators: cyclin A/Cdk1 and Plk1. In prometaphase, high levels of cyclin A work through MYPT1 to dampen Plk1 activity at kinetochores. This ensures the frequent detachment of microtubules to promote efficient error correction.

M18 An ectopic, kinetochore‐independent checkpoint activator reveals the biochemical design of the mitotic checkpoint. A.P. Joglekar1,2, C. Chen1, P. Sekhri1, I.M. Cheeseman3, I. Whitney3; 1Cell Developmental Biology, University of Michigan Medical School, Ann Arbor, MI, 2Biophysics, University of Michigan, Ann Arbor, MI, 3Biology, MIT, Boston, MA Precise control of the duration of mitosis, from prophase to anaphase onset, is critical for accurate chromosome segregation and timely cell division. The dividing cell controls this duration using a signaling cascade known as the Spindle Assembly Checkpoint (SAC). The biochemical basis of SAC


signaling is known. The SAC is activated when the Mps1 kinase phosphorylates the kinetochore protein KNL1. This key event stimulates a cascade of biochemical reactions and results in the recruitment of a number of SAC signaling proteins to the kinetochore. These kinetochore‐localized proteins interact with one another to generate a diffusible ‘wait‐anaphase’ signal. Despite this thorough understanding of SAC biochemistry, fundamental questions remain regarding its physiological operation. How does the nanoscopic kinetochore, with its limited signaling capacity, inhibit cell division rapidly and over the entire cell volume? Furthermore, how does cell limit the cumulative signaling output of many kinetochores to avert unnecessarily delayed cell division? The main difficulty in answering these questions has been the localization of key steps of the SAC signaling cascades within the crowded environment of the kinetochore. Therefore, to define the biochemical design and physiological tuning of the SAC, we engineered an ectopic, kinetochore‐independent SAC activator (eSAC) for human cells. The eSAC, which activates the SAC through the controlled dimerization of the phosphodomain of KNL1 and the kinase domain of Mps1, made the SAC amenable to in vivo biochemical analysis. Its dose‐response characteristics revealed two fundamental features of the SAC signaling cascade: non‐linear output through cooperative biochemistry and automatic gain modulation through the limiting abundance of signaling proteins. These characteristics elucidate how the SAC signaling cascade minimizes chromosome loss as well as the duration of mitosis.

M19 Large‐scale analysis of CRISPR cell cycle knockouts reveals that spindle assembly checkpoint function is dispensable for viability. K.L. McKinley1, I.M. Cheeseman1; 1Department of Biology, Whitehead Institute/MIT, Cambridge, MA Defining the genes that are essential for cellular proliferation is critical for understanding organismal development and identifying high‐value targets for disease therapies. However, the requirements for cell cycle progression in human cells are incompletely understood. To elucidate the consequences of both acute and chronic elimination of cell cycle proteins, we generated inducible CRISPR/Cas9 knockout human cell lines targeting 208 representative genes. We performed single‐cell microscopic analyses to establish the effects of the knockouts on subcellular architecture. By analyzing knockouts across different cell lines, we demonstrate that p53 is a major determinant of the phenotypic severity of cell cycle defects. Surprisingly, we find that a functional spindle assembly checkpoint is dispensable for proliferation in culture, and that checkpoint defects confer resistance to microtubule poisons, which are common cancer chemotherapeutics. Together, our work reveals new insights into cell cycle function and demonstrates the power of large‐scale CRISPR‐based approaches for cell biological analyses.

M20 A Cdc20 Switch at the Kinetochore Reveals a Duality in Control of Anaphase Onset. T. Kim1, P. Lara Gonzalez1, K. Oegema1, A.B. Desai1; 1Dept. of Cellular Molecular Medicine, UCSD, Ludwig Institute for Cancer Research, SAN DIEGO, CA The anaphase‐promoting complex/cyclosome (APC/C), together with its essential coactivator Cdc20, triggers the metaphase‐anaphase transition by targeting cyclin B and other substrates for degradation. Kinetochores, when not attached to the spindle, are known to restrict APC/C activation via the spindle checkpoint pathway that targets Cdc20. Employing one‐cell C. elegans embryos as a model, we have found that kinetochores also promote APC/C activation via a mechanism that also targets Cdc20. In prior work, we showed that kinetochore‐localized Bub1/Bub3 complex promotes anaphase onset


independently of its functions in spindle checkpoint signaling and chromosome alignment (Kim et al. J. Cell Biol. 2015; 209:507‐17). Here, by monitoring cyclin B1 levels, we show that Bub1 depletion delays APC/C‐Cdc20 activation. A fluorescent fusion of Cdc20 exhibits robust kinetochore localization, which is abolished by mutating a conserved motif in Bub1 that directly interacts with Cdc20. Notably, preventing kinetochore localization of Cdc20 phenocopies both effects of Bub1 depletion: it delays anaphase onset while also abolishing spindle checkpoint signaling. To address how kinetochore localization of Cdc20 promotes anaphase onset, we focused on its phosphoregulation. Cdc20 is phosphorylated on its N‐terminus by Cdks and this phosphorylation reduces binding affinity for the APC/C. A non‐phosphorylatable form of Cdc20 accelerated progression into anaphase and suppressed delayed anaphase onset induced by Bub1 depletion or by the Bub1 mutant that abolishes kinetochore localization of Cdc20. The non‐phosphorylatable form of Cdc20 also suppressed the checkpoint‐independent anaphase onset delay induced by preventing docking of protein phosphatase 1 (PP1) onto the kinetochore scaffold protein Knl1, but retained the ability to function in the spindle checkpoint. These results lead to a model in which Cdc20 at the kinetochore can adopt two fates: be dephosphorylated by Knl1‐docked PP1 to promote APC/C activation or become part of the spindle checkpoint pathway to inhibit APC/C. As PP1 localization is observed after chromosome alignment, we propose that the choice of Cdc20 fate is controlled by microtubule attachment status: at unattached kinetochores, Cdc20 is incorporated into the checkpoint pathway to inhibit APC/C whereas following attachment and PP1 recruitment, its dephosphorylation promotes APC/C activity. We propose that this duality contributes to the tightly controlled switch‐like transition from metaphase to anaphase.

Minisymposium 3: Intermediate Filaments from Cytoplasm to Nucleus

M21 Polarization of the intermediate filament network during directed cell migration. S. Etienne‐Manneville1, C. Leduc1; 1Cell Polarity and Infection, Institut Pasteur ‐ CNRS, Paris, France Cell migration requires dramatic and coordinated reorganization of the different elements of the cytoplasmic cytoskeleton. The cytoskeleton is mainly composed of three types of filamentous structures; actin microfilaments, microtubules and intermediate filaments (IFs). Actin structures are crucial for membrane protrusion, and cell adhesion and contractility and play a well‐characterized role in cell motility. Evidence has also emerged demonstrating the involvement of microtubules in cell polarization and migration. However, little is known about the role and regulation of the IFs network during cell migration. Qualitative and quantitative description of cytoskeletal rearrangements are required to decipher the molecular mechanisms that specifically control the changes in actin, microtubule and IF organization in migrating cells. Using primary astrocytes, we have previously shown that IFs contribute to cell migration by affecting both cell protrusion, and nucleokinesis. Wound‐healing assay allows to induce cell polarization and to initiate directed migration under controlled conditions. Wounding of an astrocyte monolayer induces profound rearrangements of IFs in the wound edge cells. IFs elongate along microtubules in the direction of migration. IF rearrangements rely on actin dynamics, microtubule‐driven transport and microtubule associated proteins. Using the combination of cutting edge quantitative microscopy techniques, we show that a wound‐induced signalling cascade locally induces the differential regulation of anterograde and retrograde transport to promote the polarization of the IF network along the axis of migration.


M22 Interplay between Rho‐GEF Solo and keratin filaments is crucial for mechanotransduction in epithelial cells. S. Fujiwara1,2, H. Abiko2, K. Ohashi2, K. Mizuno2; 1Department of Mechanical Science and Bioengineering, Graduate School of Engineering Science, Osaka University, Osaka, Japan, 2Department of BioMolecular Sciences, Graduate School of Life Sciences, Tohoku University, Sendai, Japan Mechanical forces play fundamental roles in various cell activities and functions. Epithelial cells sense and respond to external mechanical forces primarily through cell‐substrate and cell‐cell adhesions, resulting in reinforcement of actin stress fibers (SFs) and intermediate filament (IF) networks. Keratin filaments are major IFs in epithelia, anchored to desmosomes and hemidesmosomes, and these keratin‐filaments‐associated cell adhesions have important roles in cellular mechanosensing. Rho family small GTPases are activated by Rho‐guanine nucleotide exchange factors (Rho‐GEFs) and essential for cytoskeletal reorganization. However, the mechanisms underlying mechanical‐force‐induced activation of Rho family GTPases remain elusive. In the present study, we screened shRNAs targeting 63 Dbl‐related Rho‐GEFs and identified 11 Rho‐GEFs (Abr, Alsin, ARHGEF10, Bcr, GEF‐H1, LARG, p190RhoGEF, PLEKHG1, P‐REX2, Solo/ARHGEF40 and α‐PIX) that are required for the cyclic‐stretch‐induced perpendicular reorientation of human umbilical endothelial cells (HUVECs). Of these Rho‐GEFs, we focused on Solo (the GEF for RhoA and RhoC) and found that Solo binds to keratin‐8/keratin‐18 (K8/K18) IFs through multiple sites. Using the Madin‐Darby canine kidney (MDCK) epithelial cells, we provided evidence that Solo is required for the proper organization of actin SFs and K8/K18 networks. Overexpression of Solo promoted the formation of thick SFs and keratin bundles, whereas knockdown of Solo or expression of a GEF‐inactive mutant of Solo suppressed SF formation and reduced keratin networks, especially at cell‐substrate and cell‐cell adhesion sites. Solo knockdown markedly reduced the localization of plakoglobin (a component of desmosomes and adherens junctions) at cell‐cell adhesion sites, but has no apparent effect on the localization of β‐catenin (a component of adherens junctions) at these sites. Time‐lapse analysis revealed that knockdown of Solo or K18 or overexpression of GEF‐inactive or deletion mutants of Solo suppressed tensile‐force‐induced SF formation. We also showed that knockdown of Solo or K18 suppressed tensile‐force‐induced RhoA activation by biochemical methods. Together these results suggest that K8/K18 filaments serve as the upstream regulator as well as the downstream effector of Solo and that the interplay between Solo and K8/K18 IFs plays a crucial role in mechanical force‐induced RhoA activation and consequent SF reinforcement in epithelial cells.

M23 Drafting the intermediate filament proteome. M. Wiking1, C. AitBlal1, L. Björk1, A. Bäckström1, F. Danielsson1, J. Fall1, C. Gnann1, M. Hjelmare1, D. Mahdessian1, R. Schutten1, M. Skogs1, C. Stadler1, D.P. Sullivan1, P. Thul1, C.F. Winsnes1, L. Åkesson1, M. Uhlén1, E. Lundberg1; 1Affinity Proteomics, Science for Life Laboratories (KTH), Solna, Sweden Intermediate filaments are key players in providing mechanical support and structure to the cell on a subcellular level, and may also act in other cellular processes such as adhesion, migration and tumor invasion. However, of the approximately 20 000 human genes, only around 70 genes have been described to code for proteins localizing to the intermediate filaments. The Cell Atlas, part of the Human Protein Atlas (HPA) project, is a Swedish research project aiming to localize all human proteins on a subcellular level, making the data freely available in the Protein Atlas


database (www.proteinatlas.org). In the Cell Atlas we have detected more than 100 proteins localizing to the intermediate filaments, including several new possible candidates for intermediate filament proteins, such as MCOLN2. Here, we present data for previously uncharacterized genes, which we have experimentally shown to localize to intermediate filaments using antibody based protein profiling. High‐resolution confocal images showing the location of these proteins provide the groundwork of further analysis, and a subset of the proteins have been validated using siRNA knockdown or staining of cell lines stably expressing the protein of interest coupled to a fluorescent protein. As intermediate filaments are differentially expressed between cell types, it is crucial to study them in expressing cells. Here, HPA provides a valuable resource as it contains gene expression data for all human genes for more than 40 different human cell lines, which lineages covers primary cells, myeloid and lymphoid cells as well as several cell lines originating from solid tumors. In summary, we here present a draft of the human intermediate filament proteome, including proteins previously not localized to intermediate filaments.

M24 The desmosome/intermediate filament linkage regulates cell mechanics. J.A. Broussard1,2, R. Yang3, C. Huang3, S. P. Nathamgari3, A.M. Beese3, L.M. Godsel1, S. Lee1, F. Zhou3, N.J. Sniadecki4, H.D. Espinosa3,5, K.J. Green1,2; 1Department of Pathology, Northwestern University, Chicago, IL, 2Department of Dermatology, Northwestern University, Chicago, IL, 3Department of Mechanical Engineering, Northwestern University, Evanston, IL, 4Departments of Mechanical Engineering and Bioengineering, University of Washington, Seattle, WA, 5Theoretical and Applied Mechanics Program, Northwestern University, Evanston, IL Mechanotransduction, the process by which mechanical forces are translated into biochemical signals, is critical for regulating normal physiological processes during tissue development and homeostasis and also contributes to pathological conditions, including cardiovascular disease and cancer progression. While actin‐plasma membrane connections are well known to contribute to this process, the role of intermediate filament (IF) connections is not understood. The major IF connections to the plasma membrane in tissues such as skin and heart that undergo large amounts of mechanical strain are desmosomes. In addition to their essential role in ensuring tissue integrity, desmosomes act as signaling scaffolds regulating physiological processes including epidermal differentiation. We addressed the importance of desmosome‐IF interactions in regulating cell mechanics by uncoupling or enhancing the interaction between desmosomes and the IF cytoskeleton using mutant forms of the IF anchoring molecule desmoplakin (DP). DPNTP, an IF‐binding domain mutant that uncouples IF from desmosomes, or DP.S2849G, a point mutant that enhances DP’s association with IF by interfering with GSK3beta‐dependent processive phosphorylation, were expressed in doxycycline‐inducible keratinocyte cell lines. Micropillar arrays and atomic force microscopy were used to measure the effects of wild type and mutant DP on cell‐substrate and cell‐cell forces as well as cell stiffness. Uncoupling IF from desmosomes or siRNA‐mediated knockdown of DP led to a significant decrease in the average traction and intercellular forces in cell pairs and a significant decrease in cell stiffness of cell pairs and semiconfluent sheets of cells. Enhancing the desmosome‐IF interaction led to a significant increase in the average traction and intercellular forces in cell pairs as well as a significant increase in average cell stiffness in cell pairs and cell sheets. The cell stiffness increase was most notable in 6‐day confluent cell sheets, by which time desmosomes develop a strong form of calcium‐independent adhesion. DP‐mediated effects on cell mechanics were abrogated by the actin depolymerizing drug cytochalasin D. These data suggest that strengthening the desmosome‐IF network could increase the resistive capacity of the system, allowing for more robust actomyosin‐generated tension and increased cell forces/stiffness. The known tissue and differentiation‐specific diversity in desmosome/IF composition


provides an opportunity to differentially regulate tissue mechanics by balancing and tuning forces among cytoskeletal systems.

M25 Desmin modulates Nkx2.5 expression in cardiac stem cells by entering the nucleus and participating in transcription factor complexes that interact with the nkx2.5 gene. C. Fuchs1, S. Gawlas1, P. Heher1, S. Nikouli2, H. Paar1, M. Ivankovic1, M. Schultheis1, J. Klammer1, T. Gottschamel1, Y. Capetanaki2, G. Weitzer1; 1Department of Medical Biochemistry, Max F. Perutz Laboratories, Vienna Biocenter, Medical University of Vienna, Vienna, Austria, 2Biomedical Research Foundation Academy of Athens, Athens, Greece Progression of cardiomyogenesis during embryonic development is regulated by a cascade of transcription factor genes. The transcription factor Nkx2.5 and the intermediate filament protein desmin are simultaneously expressed in cardiac progenitor cells during commitment of primitive mesoderm to the cardiomyogenic lineage. Up‐regulation of Nkx2.5 expression in desmin over‐expressing cardiac progenitor cells previously suggested that desmin may contribute to cardiogenic commitment and myocardial differentiation by directly influencing the transcription of the nkx2.5 gene. Here, we demonstrate that desmin activates transcription of nkx2.5 reporter genes, rescues nkx2.5 haploinsufficiency in cardiac progenitor cells, and is responsible for the proper expression of Nkx2.5 in adult cardiac side population stem cells. These effects are consistent with the temporary presence of desmin in the nuclei of differentiating cardiac progenitor cells and its physical interaction with transcription factor complexes bound to the enhancer and promoter elements of the nkx2.5 gene. These findings introduce desmin as a new and unexpected player in the regulatory network guiding cardiomyogenesis in cardiac stem cells. In contrast to the hierarchically structured interaction of transcription factor genes during embryonic cardiomyogenesis, a more interwoven network allowing additional feedback mechanisms in cardiac stem cells may sustain cardiac stem cell homeostasis in the adult heart.

M26 Tissue mechanics regulate heterochromatin formation in breast epithelial cells. J. Northcott1, R. Bainer1, V.M. Weaver1; 1Department of Surgery, University of California San Francisco, San Francisco, CA Alterations to breast tissue architecture during normal development and cancer progression are known to be accompanied by a plethora of transcriptional changes and a modified nuclear organization. We showed that the extracellular matrix (ECM) stiffens progressively and in concert with disrupted tissue organization in breast cancer, and that a stiffened ECM compromises breast tissue architecture. Whether ECM rigidity disrupts tissue architecture to induce gene expression changes and whether this occurs at the level of chromatin organization remains largely unknown. Using the normal human breast epithelial cell lines HMT‐3522 S1 and MCF10A, we observed that the formation of polarized 3D multicellular structures is associated with an increase in heterochromatin (marked by H3K9me2 and H3K27me3) and a reduction in total RNA content, as compared to cells grown in 2D. In contrast, increased ECM stiffness alone was sufficient to cause a loss of heterochromatin in MCF10A cells. Analysis of the spatial coherence of gene expression changes across chromosomes demonstrated excess coincidence of differentially expressed genes and lamina associated domains (LADs), where genes localized inside LADs were more likely to have altered levels of expression in response to changes in


ECM stiffness or multi‐cellularity. Furthermore, ChIP‐seq experiments revealed a negative association between gene expression levels and enrichment for the LAD‐associated H3K9me2 mark. Taken together, our data indicate that cellular architecture and tissue stiffness have a direct effect on epigenetic regulation of gene expression by regulating heterochromatin formation. Ongoing studies are focused on determining whether chromatin remodeling occurs in response to mechanosignalling through signal transduction pathways and/or direct forces via the linker of the nucleoskeleton and cytoskeleton (LINC) complex. This work supported by: NIH/NCI R01CA192914, 1U01CA202241‐01 and DOD W81XWH‐13‐1‐0216.

M27 Lamins modulate nuclear mechanics, migration efficiency, and nuclear envelope integrity during cancer cell migration in confined 3‐D environments. C.M. Denais1, R.M. Gilbert2, P. Isermann1,2, B. Weigelin3,4, A.L. McGregor1,2, E.S. Bell1,2, M. te Lindert3, P. Friedl3,4, K. Wolf3, J. Lammerding1,2; 1Weill Institute for Cell and Molecular Biology, Cornell University, Ithaca, NY, 2Meinig School of Biomedical Engineernig, Cornell University, Ithaca, NY, 3Department of Cell Biology, Radboud University Medical Center, Nijmegen, Netherlands, 4Department of Genitourinary Medical Oncology, The University of Texas MD Anderson Cancer Center, Houston, TX Cell motility plays a crucial role in many physiological and pathological settings, ranging from wound healing to cancer metastasis. The physical challenges cells face when moving in three‐dimensional (3‐D) environments are only now emerging. We and others previously found that the deformability of the large and relatively rigid nucleus can constitute a rate‐limiting factor during migration through tight constrictions, such as when invading cancer cells encounter interstitial spaces and dense extracellular matrix networks. Using microfluidic devices that mimic physiological environments, along with migration in collagen matrices and intravital imaging of cancer cell invasion in mouse xenografts, we investigated the biomechanical factors that govern cell migration through tight spaces, as well as the biological consequences of the associated mechanical stress on the cell nucleus. We found that cell migration through tight spaces in vitro and in vivo frequently results in transient loss of nuclear envelope (NE) integrity (‘NE rupture’), characterized by the uncontrolled exchange of nuclear and cytoplasmic proteins, protrusion of chromatin across the NE, DNA damage, and, in severe cases, nuclear fragmentation. NE rupture was preceded by the formation of nuclear membrane blebs at sites of local defects in the nuclear lamin network. Nuclear membrane blebbing was often followed by herniation of chromatin through the nuclear lamina. Accordingly, depletion of lamins significantly increased the probability of NE rupture. The incidence of NE rupture increased significantly with increasing cell confinement. In breast cancer cells, for example, NE rupture dramatically increased when pore sizes were decreased to 10 µm2 or less in cross section. Cells migrating through small pores also had significantly higher rates of nuclear fragmentation and more DNA double‐strand breaks, identified by 53BP1 and gamma‐H2AX foci, than cells in unconfined conditions. Loss of nuclear/cytoplasmic compartmentalization persisted for 30‐90 minutes. NE repair required components of the endosomal sorting complexes required for transport (ESCRT) machinery, and depletion of the ESCRT proteins CHMP2A and CHMP7 significantly delayed NE repair. While the vast majority (>90%) of cells survived even repetitive nuclear rupture, combined inhibition of NE repair and DNA damage repair pathways resulted in substantial increase in cell death after NE rupture. In conclusion, transient loss of NE integrity during 3‐D migration may provide an additional mechanism for genomic instability in (cancer) cells, while inhibiting nuclear membrane repair may present a novel approach to specifically target invasive cancer cells.


M28 New insight into the interplay of actomyosin contractility and LINC complex in regulation of nuclear shape and high‐speed tumor cell motility using precisely sized collagen I/fibronectin1D nanofibers. V.P. Sharma1,2, J. Williams3, E. Leung1, J. Sanders3, R.J. Eddy1, J. Castracane3, J.S. Condeelis1,2; 1Anatomry and Structural Biology, Albert Einstein College of Medicine, Bronx, NY, 2Gruss Lipper Biophotonics Center, Albert Einstein College of Medicine, Bronx, NY, 3Colleges of Nanoscale Science and Engineering, SUNY Polytechnic Institute, Albany, NY In vivo, breast tumor cells metastasize by traveling on linear extracellular matrix (ECM) fibers towards blood vessels in a directional migration mode known as tumor cell streaming. Streaming is characterized by tumor cell migration at high speed and directional persistence on 1D fibers. Tumor cell directionality is the result of chemotaxis to HGF secreted by endothelial cells in blood vessels (Leung et al. 2016), but the relationship of 1D geometry of ECM fibers and the underlying mechanotransduction mechanism regulating high‐speed migration of tumor cells is not well understood. Here, we analyzed in vivo ECM architecture that streaming cells use by second harmonic intravital imaging and found a narrow peak of diameters falling in the range 2‐3 µm. The fibers were found to be composed of collagen I and fibronectin. Based on this in vivo ECM characterization, we developed a high fidelity in vitro nanofiber system to study the molecular mechanisms underlying tumor cell streaming migration. Breast carcinoma cells plated on 2 µm ECM‐coated fibers showed enhanced motility matching in vivo velocities averaging 1.2 µm/min. Employing both nanofibers and micro‐patterned 1D stripes, we varied 1D diameter (0.7‐20 µm) and found that tumor cells move the fastest with highest persistence on smaller 1D fibers and stripes within a narrow range of diameters from 0.7‐3 µm. High tumor cell speeds correlated with enhanced alignments of F‐actin fibers and focal adhesions along the fibers axial dimension. Unexpectedly, we also observed nuclear deformation during carcinoma cell migration on narrow fibers in vitro, similar to nuclear deformation observed in vivo. We were surprised because nuclear deformation in vivo was assumed to be caused by squeezing through ECM pores. Thus we hypothesized that actomyosin contractility not only regulates cell motility parameters, but also nuclear deformation independently of ECM pore size. Testing this hypothesis we found that disruption of the nucleo‐cytoskeleton linkage by SUN1 and SUN2 KD led to increases in nuclear elongation and cell motility parameters, through F‐actin upregulation at the leading edge. These results indicate that in carcinoma cells, F‐actin associated forces are shared between the leading edge (to maintain cell speed) and the nucleus (to dynamically regulate nuclear shape). LINC complex disruption releases F‐actin forces acting on the nucleus to the cell front, leading to higher tumor cell motility speeds. In summary, our results have led to new insight into the interplay between actomyosin contractility and the LINC complex in the regulation of nuclear shape and high‐speed tumor cell motility during metastasis.

M29 A detailed assembly mechanism of nuclear lamin A and its use to characterize laminopathic mutant variants. H. Herrmann1,2, U. Aebi2; 1Molecular Genetics, German Cancer Research Center, Heidelberg, Germany, 2Biozentrum, University of Basel, Basel, Switzerland Lamins A/C, B1 and B2 are sequence‐related intermediate filament proteins exhibiting an extended central alpha‐helical “rod” flanked by non‐alpha‐helical “head” and “tail” domains. In cells, all four proteins form independent but interconnected filament systems. Here we investigated, starting from


coiled‐coil dimers, the assembly mechanism of human lamin A with regard to the minimal requirements for assembly into filaments and higher order filament arrays. Therefore, we progressively trimmed the head and the tail thereby revealing what parts of the unstructured end domains are essential for longitudinal as well as lateral organization of dimers into higher order assemblies. We found that for longitudinal assembly only the two short segments representing CDK1 phosphorylation motifs adjoining the rod on either side are needed. By glycerol spraying / rotary metal shadowing, we demonstrate that lamin A [22‐395] still forms dimeric head‐to‐tail polymers, however, it does not form proper filaments under standard filament assembly conditions (250 mM NaCl, 25 mM Mes, pH 6.5). In contrast, with just 7 amino acids of the tail lamin A [22‐402] forms regular filaments under this assembly regime. Lamin A [31‐395], the rod, is not able to elongate at all. Next, we developed an assembly schedule to examine these truncation variants as well as laminopathic mutants under near physiological buffer conditions, i.e. between 200 and 100 mM NaCl buffered to pH 7.5. Notably we found that – starting from dimers at 250 mM NaCl, pH 7.5 – the first structures found when dialyzed against lower salt (200 mM) were short, open structures of approximately 100 by 25 nm exhibiting a 50 nm axial repeat. By lowering the NaCl concentration stepwise, we obtained progressively longer paracrystalline fibers that became considerably wider when the NaCl concentration reached 100 mM. Concerning lateral assembly, mutant lamin A [1‐416] formed long filaments of somewhat regular diameter, but did not yield highly ordered paracrystaline fibers. In contrast, lamin A [1‐437] exhibited both highly regular paracrystalline assemblies as well as open criss‐cross arrays of fibers that appear to be on their way to laterally associate, indicating that segment 417‐437 is engaged in higher order lateral organization of lamin filaments. These experiments were all conducted in the absence of divalent cations, which force even the rod into paracrystals. In stark contrast, a number of point mutated lamin A progeria variants yielded very different assemblies. For instance, the mutant lamin A [T10I] formed huge paracrystalline networks already at 150 mM salt. Similarly, also the other mutant proteins deviated distinctly from the wild‐type appearance indicating that they harbor an inbuilt structural deficiency.

M30 The structural organization of lamins in the nuclear lamina. O. Medalia1, Y. Turgay1, R.D. Goldman2, M. Eibauer1, T. Sapra1; 1Biochemistry, University of Zurich, Zurich, Switzerland, 2Department of Cell and Molecular Biology, Northweastern university, Chicago, IL Lamins are nuclear intermediate filaments (IFs) of metazoan cells. They assemble into fibrous structures that are positioned between the inner nuclear membrane and the peripheral chromatin, although a small fraction of lamins is present in the nucleoplasm. Lamins are required to maintain nuclear structure and, together with many interaction partners, are involved in most nuclear activities. Mutations in lamins cause a group of >14 distinct diseases called laminopathies, it is not clear how lamins are organized in vivo and how these mutations affect lamin functions. Understanding how lamins are assembled, and how mutations in lamins and lamin binding proteins affect lamin filament assembly and cellular localization is essential for understanding the basic mechanisms of laminopathic diseases. Taking the advantage of recent advances in cryo‐EM, we reconstructed the molecular organization of the nuclear lamina in situ, within somatic cells. Individual lamin filaments resolved a globular‐decorated fiber appearance and showed that A‐ and B‐type lamins assemble into tetrameric 3.5 nm thick filaments. Thus, lamins exhibit a structure that is remarkably different than the other canonical cytoskeletal elements. Finally, a scaled model of the lamin was generated.


Minisymposium 4: Membrane Traffic Control By Lipids, Cargos and Motors M31 Spatial regulation of myosin V transport via the PAK kinase, Cla4. S. Wong1, R.G. Yau1, L.S. Weisman1; 1Cell and Developmental Biology, Cell and Molecular Biology Training Program, Life Sciences Institute, University of Michigan, Ann Arbor, MI

Movement of organelles to their correct locations is essential for cellular function. In most cells, long‐range movement of organelles occurs on microtubules via kinesin and dynein motors. Some cargoes are then transferred to myosin V motors, which provide short‐range transport or tethering to the final destination. In these cases, spatial and/or temporal regulation of the detachment of myosin V from its cargoes likely specifies the destination. However, for most cargoes, it is largely unknown how this regulation is achieved. In the budding yeast, Saccharomyces cerevisiae, most intracellular organelles are moved solely by the myosin V motors, Myo2 and Myo4. The yeast vacuole is one of six cargoes moved by Myo2. Early in the cell cycle, Myo2 binds the vacuole specific adaptor, Vac17, and transports a portion of the vacuole into the bud. However, later in the cell‐cycle, the vacuole detaches from Myo2 and Vac17 is degraded. At that point, the vacuole is observed near the center of the bud. The landmark that initiates the detachment of Myo2 from the vacuole was not was not known. Here we report that in small‐budded cells, the bud cortex is the landmark for delivery of the vacuole and that the p21‐activated kinase, Cla4, initiates the termination of vacuole transport at this location. We show that as a first step in the termination of vacuole inheritance, the vacuole extends to the bud cortex where the vacuole, Vac17, and Cla4 first colocalize. Cla4 then moves to and persists on the vacuole and Vac17 is degraded. We show that Cla4 directly phosphorylates Vac17‐S222, and that this phosphorylation is required for Vac17 ubiquitylation and degradation thereby detaching the vacuole from Myo2. These results indicate that Cla4 signals the detachment of the vacuole from myosin V and plays a role in mediating its correct intracellular location. Notably, Cla4 is a key player in cell polarity and septin ring formation, events that occur prior to vacuole transport. This additional role for Cla4 in detachment of the vacuole suggests that there may be a direct molecular link between cell polarity, and the positioning of the vacuole in the bud.

M32 Budding yeast has a minimal endomembrane system. K.J. Day1, B.S. Glick1; 1Department of Molecular Genetics and Cell Biology, University of Chicago, Chicago, IL The trans‐Golgi network (TGN) communicates closely with the endocytic pathway, sorting internalized cargos back to the cell surface and delivering newly synthesized proteins to endosomes. We observed in the yeast Saccharomyces cerevisiae that the bulk membrane endocytic dye FM 4‐64 localized predominantly to the TGN shortly after endocytosis. To understand the extent of TGN involvement in the endocytic pathway, we characterized the compartments and pathways that comprise the yeast endosomal network. A pulse of FM 4‐64 was used to track the progression of the endocytic pathway. The dye localized primarily to the TGN first, and later to peri‐vacuolar puncta called pre‐vacuolar endosomes (PVEs), before ending up at the vacuole. We observed that the TGN and PVE have distinct components and dynamics. The PVE is a long‐lived structure, whereas the TGN undergoes rapid turnover. The organization of yeast endosomes is generally assumed to parallel that of higher eukaryotes, which have an additional endosomal compartment, the early endosome. To determine whether we had missed


the passage of FM 4‐64 through such a third endosomal compartment, we assessed the localization of several potential markers for a yeast early endosome. Surprisingly, each marker localized to either the TGN or PVE, not to a distinct third compartment. A mutant of the TGN‐localized phosphatidylinositol 4‐kinase Pik1 blocked FM 4‐64 trafficking to the vacuole, suggesting that the dye travels through the TGN en route to the PVE and vacuole. We conclude that budding yeast possess only two types of endosomes and that the TGN fulfills the role of an early endosome by receiving newly internalized cargo. These findings support a simplified circuit diagram in which the endomembrane system in yeast has fewer compartments and more direct pathways than in mammalian cells. With this insight, the yeast system is poised to provide a clearer picture of the endocytic pathway, and to reveal how mammals have elaborated a minimal endomembrane system for their more complex requirements.

M33 Ezrin activation by LOK phosphorylation involves a PIP2‐dependent coincidence detection mechanism in a multi‐step reaction. T. Pelaseyed1, R. Viswanatha1, J.J. Filter1, M.L. Goldberg1, A.P. Bretscher1; 1Department of Molecular Biology and Genetics, Cornell University, Ithaca, NY All nucleated cells are polarized, meaning they form distinct cellular domains with asymmetric protein compositions and morphology. Epithelial cells display highly polarized apical and basolateral membrane domains, which allow the cell to undertake vital biological tasks necessary for homeostasis of the organism. One hallmark of epithelial polarization is the presence of apical actin‐based microvilli, which increase cell surface area and concentrate proteins necessary for apical domain function. Ezrin, a member of the ezrin, radixin and moesin (ERM) proteins, functions to support the microvilli as a crosslinker between the underlying actin cytoskeleton and the plasma membrane. Ezrin cycles between a closed, inactive cytoplasmic state and an open, active phosphorylated form restricted to the apical membrane. Inactive ezrin is characterized by an intramolecular head‐to‐tail interaction between the N‐terminal 4.1 ERM (FERM) domain and the C‐terminal domain (CTD) that are connected via a α‐helical coiled‐coil loop. We have shown that lymphocyte‐oriented kinase (LOK) and the related Ste20‐like kinase (SLK) are the major epithelial kinases that activate ezrin upon phosphorylation (Viswanatha et al., JCB 2012). LOK/SLK phosphorylate ezrin at position T567 to generate an electrostatic repulsion that dissociates the CTD from the FERM domain, allowing open ezrin to function as a crosslinker. However, T567 is masked in the closed structure, so how the kinases access their substrate has not been resolved. Earlier work has indicated an involvement of PIP2, which binds to ezrin, but its mechanism of involvement has not been elucidated. Here we present the in vitro reconstitution of ezrin phosphorylation by LOK in a PIP2‐dependent manner. This has allowed us to dissect the mechanism of unmasking and phosphorylation of ezrin T567. Our results indicate that this process involves a novel coincidence detection mechanism requiring both PIP2 and LOK and involving multiple steps: 1) specific priming of ezrin by PIP2 to induce a conformational change that renders ezrin accessible to LOK, 2) an interaction between the regulatory domain of LOK and ezrin to unmask T567, 3) an interaction with ezrin to activate LOK, and (4) a final phosphorylation step involving binding of the kinase to a region in ezrin CTD distal to T567 permitting recognition of the classical kinase consensus sequence flanking T567. This intricate mechanism ensures highly specific and local phosphorylation of ezrin by LOK, thereby contributing to the polarized function of epithelial cells. This work was supported by grants from the Wenner‐Gren Foundations (TP) and by NIGMS grants GM036552 (RV, APB) and GM048430 (JJF, MLG).


M34 Sphingomyelin is sorted at the trans Golgi network into a distinct class of secretory vesicle. Y. Deng1, F.E. Rivera‐Molina1, D.K. Toomre1, C.G. Burd1; 1Cell Biology, Yale University School of Medicine, New Haven, CT One of the principal functions of the trans Golgi network (TGN) is to sort proteins into distinct vesicular transport carriers that mediate secretion and inter‐organelle trafficking. Are lipids also sorted into distinct TGN‐derived carriers? The Golgi is the principal site of synthesis of sphingomyelin, an abundant sphingolipid that is transported to the plasma membrane, where it resides exclusively in the outer leaflet. Most previous investigations of sphingolipid sorting employed surrogate sphingolipid probes, such as synthetic fluorescent ceramides, but these have substantial chemical/structural differences compared to native sphingolipids that compromise their utility. Thus, to address the specificity of native sphingomyelin transport to the plasma membrane, we engineered a natural sphingomyelin‐binding pore forming toxin, equinatoxin‐II, into a non‐toxic reporter protein, called “Eqt‐SM,” and used it to monitor intracellular trafficking of sphingomyelin. Quantitative fluorescence live cell imaging shows that Eqt‐SM is enriched in a subset of trans Golgi network‐derived secretory vesicles that are also enriched in a glycophosphatidylinositol anchored reporter protein. In contrast, a secreted integral membrane reporter protein, CD8α, is not enriched in these carriers. To elucidate the native protein secretome of these sphingomyelin‐enriched carriers, and therefore the physiological roles of this arm of the secretory pathway, we developed a method to generate Golgi‐derived vesicles from permeabilized HeLa cells that express Eqt‐SM fused to an unspecific biotinylation enzyme, APEX2. Proximity‐based biotinylation and mass spectrometry proteomics facilitated the identification of proteins that are co‐packaged with Eqt‐SM into secretory vesicles. Co‐sorted cargo includes native GPI‐anchored proteins, proteins that cycle between the Golgi and plasma membrane (i.e., organelles composed of sphingomyelin‐rich membranes), and proteins involved in lipid metabolism. The results demonstrate the sorting of native sphingomyelin at the TGN and its transport to the plasma membrane by a specific carrier that also contains a subset of secreted proteins.

M35 TMEM24, a lipid transporter at ER‐PM contacts, regulates pulsatile insulin secretion. M. Messa1,2,3,4, J.A. Lees2, E.W. Sun1,2,3,4, H. Wheeler1,2,3,4, F. Torta5, M.R. Wenk5, K.M. Reinisch2, P. De Camilli1,2,3,4,6; 1Department of Neuroscience, Yale University School of Medicine, New Haven, CT, 2Department of Cell Biology, Yale University School of Medicine, New Haven, CT, 3Howard Hughes Medical Institute, Yale University School of Medicine, New Haven, CT, 4Program in Cellular Neuroscience, Neurodegeneration and Repair, Yale University School of Medicine, New Haven, CT, 5Department of Biochemistry, Yong Loo Lin School of Medicine, National University of Singapore, Singapore, Singapore, 6Kavli Institute for Neuroscience, Yale University School of Medicine, New Haven, CT The endoplasmic reticulum (ER) is the cell compartment where many cellular components, including most lipids, are synthesized. In addition to be functionally connected to other membranes by vesicular transport, the ER also forms juxtapositions (10‐30nm) with other cell membranes (membrane contacts sites) including the plasma membrane, where protein tethers bridge the two bilayers. At least some of these tethers function as lipid transport proteins that mediate lipid exchange between the two adjacent membranes, independently of fusion and fission reactions. Several such tethers have been identified and characterized, but the full picture is far from complete. Here, we show that TMEM24, a protein enriched in neurons and neuroendocrine cells, is an ER‐anchored protein that accumulates at ER‐PM


contact sites where it binds the PM in trans via its basic, unstructured C‐terminal region. TMEM24 harbors an SMP domain (a module of the TULIP superfamily) that can nest phospholipids with a preference for PtdIns and transport them between bilayers. As we observed with experiments in insulinoma cells and in primary neuronal cultures, TMEM24 accumulation at ER‐PM contact sites is sensitive to cytosolic Ca2+, since it redistributes from contact sites to the reticular ER upon cytosolic Ca2+ elevations as it occurs in neurons upon depolarization, NMDA receptor stimulation or Ca2+ uncaging. In insulinoma cells, TMEM24 association with the PM undergoes cycles of association/dissociation that parallel with, but have opposite phase to, Ca2+ oscillations induced by glucose stimulation. This dynamic redistribution of TMEM24 depends on the phosphorylation/dephosphorylation of its polybasic C‐terminus by PKC and PP2B/calcineurin, respectively. Functional results indicate that TMEM24 is implicated in the coordination of PM phosphoinositide dynamics with cytosolic Ca2+ changes and insulinoma cells lacking TMEM24 display a robust reduction of Ca2+ oscillatory events and evoked insulin secretion. Thus, TMEM24 represents a new link between lipid dynamics and Ca2+ signaling with a role in the control of regulated secretion.

M36 Cargo Mediated Regulation of Clathrin Mediated Endocytosis. P. Sengupta1, J. Lippincott‐Schwartz1, E. Betzig1; 1Janelia Research Campus, HHMI, Ashburn, VA Clathrin‐mediated endocytosis (CME) of diverse membrane cargos plays a critical role in numerous physiological processes including nutrient uptake and receptor‐mediated signaling. Recent data support an emerging viewpoint that cargo molecules actively modulate the dynamics and cargo content of CCPs and consequently contribute to compositional and functional specialization of CCPs. However, an outstanding challenge is to delineate the mechanisms of such regulation. In this study, we have used dSTORM and TIRF‐SIM to reveal a potential mechanism for cargo‐mediated regulation of CME. We investigated the spatio‐temporal properties of CCPs internalizing two cargo molecules with differently sized extracellular domains: CD4, a glycoprotein with a large extracellular portion, and, transferrin receptor (TfR) with a smaller extracellular domain. TIRF imaging showed that CD4 and TfR are present in distinct subsets of CCPs. Multicolor superresolution images and kinetic analysis revealed that CD4‐containing CCPs were larger in size and have significantly longer lifetime than those containing TfR. A mutated CD4, lacking the extracellular domains, internalized in CCPs with pit‐size and lifetime similar to TfR‐containing pits. These results indicate that increased size and maturation time of CD4‐containing CCPs is a direct consequence of its exoplasmic domain. dSTORM experiments further showed that CD4, but not TfR, is excluded from the high‐curvature neck region of the CCPs. Live cell imaging further revealed that the positioning of TfR at the neck of the pits allowed TfR molecules to diffuse out of long‐lived CD4‐containing CCPs, resulting in virtual absence of TfR in CD4 pits at the time of scission. Taken together, our data provides evidence for a novel mechanism by which cargo molecules with large extracellular domains are able to orchestrate the formation of specialized CCPs. Bulky extracellular domains of CD4 would experience significant steric pressure as they are accommodated in the concave face of the nascent CCP. The steric pressure is relieved by accommodating the extracellular domains in larger CCPs, which take longer to mature. Steric crowding further restricts the spatial distribution of cargo within the forming CCP, precluding them from the high‐curvature neck region while allowing cargo molecules with smaller exoplasmic domains to diffuse out through the neck. The prolonged surface residence of CCPs along with kinetic filtering of cargo molecules provide a mechanism for forming distinct subpopulation of CCPs with specialized cargo molecules. Our results also underline the utility of superresolution microscopy in teasing out the mechanism of cargo‐mediated regulation of CCP dynamics that operate at nanoscopic spatial scales.


M37 Cryo‐CLEM of clathrin coats reveals the architecture of the clathrin lattice during maturation. O. Avinoam1, M. Kaksonen2, J.A. Briggs1,3; 1Structural and Computational Biology Unit , The European Molecular Biology Laboratory (EMBL), Heidelberg, Germany, 2Department of Biochemistry, University of Geneva, Geneva, Switzerland, 3Cell Biology and Biophysics Unit , The European Molecular Biology Laboratory (EMBL), Heidelberg, Germany Clathrin mediated endocytosis (CME) is a basic cellular function playing essential roles in nutrient uptake, membrane recycling, signal transduction, synaptic transmission and viral infection. During CME, clathrin coated vesicles (CCVs) form from the plasma membrane in a sequence of events that begins with cargo recognition and concentration, proceeds with membrane bending and culminates in vesicle scission. We have recently shown using correlative light and electron microscopy (CLEM) that clathrin is already recruited to the membrane early in endocytosis, before any significant membrane bending has occurred, and then rearranges as the membrane bends to wrap around the forming vesicle. Furthermore, we showed by fluorescence microscopy (FM) that clathrin undergoes rapid exchange with the cytoplasmic pool at sites of endocytosis, providing insights into the mechanism driving coat rearrangement during CCV maturation. While numerous EM studies have documented a population of flat hexagonal clathrin lattices in cells it remains unclear whether these are the true precursors of CCVs. In order for assembled flat lattices to evolve into vesicles they must reorganize to acquire pentagons, which would allow the structure to curve. An alternative hypothesis is that CCVs actually emerge from unassembled clathrin patches on the membrane, which subsequently polymerize, assembling the final curved lattice directly from a high local concentration of amorphous clathrin. To distinguish between these two possibilities we performed correlated cryo ‐FM and ‐EM (Cryo‐CLEM) on intact genome edited mammalian cells expressing fluorescently tagged clathrin and AP‐2 at physiological levels. Cryo‐FM was used to identify bona‐fide sites of endocytosis in which both markers are present while cryo‐EM was used to obtain structural information on the organization of clathrin at those sites. To obtain 3D information we used cryo electron tomography. We found that sites labeled by both markers corresponded to CCVs and clathrin‐coated pits (CCPs) at different stages of maturation ranging from nearly flat to highly curved invaginations. Clathrin was organized in a defined lattice at all stages, suggesting that clathrin first assembles as planar lattice, which only subsequently rearranges and bends. To determine the structure of the clathrin coat in situ and understand how lattice rearrangement and membrane bending are triggered, we applied sub‐tomogram averaging to CCPs at different stages of maturation. These results highlight the usefulness of cryo‐CLEM for structural level characterization of biological machines in their native environment.

M38 TWO ALTERNATIVE MODES OF DYNEIN REGULATION BY LIS1. M.E. DeSantis1, Z. Htet1, M. Cianfrocco1, P.T. Tran1, A.E. Leschziner1, S.L. Reck‐Peterson1; 1CMM, UCSD, La Jolla, CA Dynein is the primary minus‐end‐directed, microtubule‐based molecular motor in animals and fungi. By coupling ATP‐hydrolysis to force generation, dynein is able to walk along microtubules, providing the basis for its roles in cargo transport, organelle positioning, and mitotic spindle pole alignment. Recently, it has become clear that dynein rarely acts in isolation. Instead a vast library of regulator and adaptor proteins is required to modulate dynein’s activity and its ability to interact with cargo. Our lab is particularly interested in Lis1, a ubiquitous dynein regulator that is crucial for many of dynein's


functions. Previous work in our lab has illustrated that Lis1 is required for transport initiation and dynein localization at the microtubule plus‐end. We have also shown that Lis1 binds dynein’s motor domain and slows the velocity of dynein movement on microtubules, presumably by increasing dynein’s microtubule binding affinity. In the work presented here, we show that the mechanism of Lis1 activity is far more complex than previously thought. We have observed that Lis1 can both increase and decrease dynein’s microtubule binding affinity, depending on the nucleotide state of dynein. Interestingly, these “high” and “low” affinity dynein states bind Lis1 with different stoichiometry. Finally, a Cryo‐EM structure of the low affinity state reveals a novel and highly conserved Lis1 binding site that is distinct from the previously characterized binding site. We suggest that the dual functionality of Lis1 may be required for its roles in transport initiation and dynein localization.

M39 Novel motor‐independent function of the dynein light chain in clathrin mediated endocytosis. K.B. Farrell1, S. McDonald1, C. Worcester1, O.B. Peersen1, S.M. Di Pietro1; 1Department of Biochemistry and Molecular Biology, Colorado State University, Fort Collins, CO Clathrin‐ and actin‐mediated endocytosis is essential in eukaryotic cells. Using yeast, we demonstrate Tda2 is a novel protein of the endocytic machinery necessary for normal internalization of native cargo. Tda2 has not been previously classified in any protein family. Remarkably, the three‐dimensional structure of Tda2 determined by X‐ray crystallography revealed Tda2 is a dynein light chain, the first of the TcTex1 type identified in yeast. Tda2 works independently of the dynein motor complex, consistent with the lack of function of microtubules in endocytosis. Similar to actin‐associated proteins, Tda2‐GFP localizes to endocytic sites in a filamentous actin dependent manner and arrives during late stages of the endocytosis process. Biochemical analysis and live cell fluorescent microscopy demonstrated Tda2 forms a tight complex with Aim21 and it depends on Aim21 for localization to endocytic sites. Tda2 and Aim21 regulate the recruitment of actin capping protein, a key factor of the actin cytoskeleton at endocytic sites. This is the first report of a dynein light chain working as a component and regulator of the clathrin‐ and actin‐mediated endocytic machinery.

M40 Proximity‐specific ribosome profiling reveals efficient co‐translational ER targeting in the absence of SRP. E.A. Costa1, J.S. Weissman1; 1Cellular and Molecular Pharmacology, University of California San Francisco, San Francisco, CA The signal recognition particle (SRP) has been known for over three decades to be an integral component of the machinery responsible for delivering secretory proteins to the surface of the endoplasmic reticulum (ER) for translocation. However, the precise biological role of SRP including its substrate range and the consequences of loss of SRP in vivo remain poorly defined. For example, many proteins do not require SRP to translocate across the ER membrane. There are several SRP independent pathways that can mediate targeting of these substrates, which has generally been thought to be post‐translational. Recent work has shown that in vivo many SRP independent substrates are targeted to the ER in a co‐translational manner, indicating that coupling of translation and targeting might be tighter than previously appreciated for these substrates. It is unclear whether these proteins may still utilize SRP under normal growth conditions to achieve co‐translational ER localization. To comprehensively understand which proteins rely on SRP for efficient ER targeting, we have applied proximity‐specific ribosome profiling coupled with the plant‐based auxin inducible degron system, which


allows for rapid degradation of SRP without additional perturbations. We find that upon loss of SRP, the majority of proteins that are predicted to be SRP independent are still efficiently co‐translationally targeted to the ER, while we see depletion at the ER of known and predicted SRP substrates. We additionally find that secretory proteins without a hydrophobic domain within the first 60 amino acids show rapid loss of ER targeting, indicating a role for SRP in the targeting of this previously unclassified group of proteins. Remarkably, we find that upon loss of SRP, secretory proteins are aberrantly targeted to mitochondria. Our results help elucidate the essential role for SRP in maintaining proper localization of secretory proteins.

Minisymposium 5: Microtubule Dynamics

M41 Microtubules grow in vivo through pathways that use bent tubulin protofilaments. J.R. McIntosh1, E.A. O'Toole1, J. Austin2, R. Ding1, E.V. Ulyanov3, F.I. Ataullakhanov3,4, N.B. Gudimchuk3,4; 1Molec., Cell., and Devel. Biol., University of Colorado, Boulder, CO, 2Molec. Gen. and Cell Biology, University of Chicago, Chicago, IL, 3Dept. Physics, Lemonosov Moscow State Univ. , Moscow, Russia, 4Cent. Theoret. Probs. Physicochem. Pharmacol. RAS, Moscow, Russia Textbook knowledge, based on cryoEM of purified protein components, asserts that microtubules (MTs) grow by the addition of straight GTP‐tubulin onto almost straight protofilaments (PFs) that then associate laterally to form sheets; these fold to form a patent cylinder. This view is questioned by atomic resolution studies of GTP‐tubulin, which show that this dimeric protein is as bent as GDP‐tubulin (Ravelli et al.,2004; Buey et al., 2006; Rice et al., 2008). We have examined the shapes of MTs elongating in vivo and found shapes that differ significantly from the textbook images of tubulin polymerization but are consistent with the atomic structures of GTP‐ and GDP‐tubulin. Interpolar MTs in the interzone of mitotic spindles in anaphase B are elongating by the addition of tubulin at their plus ends. We have used electron tomography of rapidly frozen, freeze‐substitution fixed material to examine these MT ends in PtK1 cells, fission yeasts, budding yeasts, and meristematic cells from the root tips of Arabadopsis. The PFs on these MT ends display significant curvature; they form a unimodal distribution with a mean curvature of 9.4 +/‐ 4.2 degrees per tubulin monomer. All these PFs appear to be independent of their neighbors. A few PFs extend straight from the MT wall, but these are associated with at least one near neighbor. Interestingly, however, these extending PFs commonly have a curling end that diverges from the neighboring PF. Sheets are never seen. Since these results are at odds with descriptions of MTs elongating in vitro, we have examined the possibility that MT tip structure in vivo is modified by the high pressure commonly used to freeze cells without ice crystal damage. In plunge‐frozen S. pombe, growing MTs display flaring PFs whose curvatures are quantitatively indistinguishable from those of MTs in high‐pressure frozen cells, dismissing this explanation. We have used computer modeling of tubulin polymerization, based on important molecular properties of MT subunits such as flexural rigidity and the strengths of their lateral and longitudinal bonds, to seek values that predict MT growth intermediates that resemble our observations of MT end structures in vivo. This analysis suggests a plausible mechanism of MT growth via a pathway that uses bent PFs. Additionally, it provides constraints on the way MT‐associated proteins, like the TOG‐domain proteins known to catalyze MT growth, should modulate effective molecular properties of tubulin in vivo. Plausible reasons for the discrepancy between our results and textbook “knowledge” will be discussed.


M42 Tubulin isoform composition tunes microtubule dynamics. A. Vemu1, J. Atherton2, J.O. Spector1, C.A. Moores2, A. Roll‐Mecak1; 1NINDS, National Institutes of Health, Bethesda, MD, 2Department of Biological Sciences, Birbeck College, University of London, London, United Kingdom Microtubules are essential non‐covalent polymers composed of αβ‐tubulin subunits. They cycle stochastically between polymerization and depolymerization i.e., they exhibit “dynamic instability”. This behavior is crucial for cell division, motility and differentiation. Microtubules are diversified through the use of multiple tubulin isoforms and abundant posttranslational modifications that mark microtubule subpopulations with distinct dynamics. Eukaryotes have multiple tubulin genes (humans have eight α‐ and eight β‐tubulin isoforms) with tissue specific distributions. While studies in the last decades have made fundamental breakthroughs in our understanding of how cellular effectors modulate microtubule dynamics, analysis of how tubulin isoform usage in various cell types affects microtubule dynamics has been held back by a lack of in vitro dynamics measurements with homogenous or well‐defined tubulin species. The majority of in vitro dynamics studies presently performed use highly heterogenous mosaic brain microtubules with isoform composition and posttranslational modifications different from those found in vivo, for example in an epithelial cell or the axonal or dendritic compartment of a neuron. Here we present the 4.2 Å cryo‐EM structure and in vitro dynamic parameters of microtubules assembled from tubulin purified from a human embryonic kidney cell line. Tubulin isolated from this cell line consists of mainly α1B/βI and βIVb isoforms. In vitro dynamics assays show that these microtubules polymerize twice as fast and catastrophe ten times less frequently compared to hetergenous brain microtubules. Furthermore, we show that the dynamics of these non‐neuronal microtubules can be modulated by the addition of recombinant neuronal α1A/βIII tubulin, an isoform expressed in non‐neuronal tissues only during tumorigenesis. Thus, in addition to the potential isoform specific recruitment of microtubule regulators, different microtubule dynamics in cells can be elicited by modulating the relative expression levels of tubulin isoforms. Future studies of in vitro microtubule dynamics with tubulin purified from various tissues as well as drug resistant tumors should shed light on how cells balance their tubulin isoforms to regulate microtubule dynamics.

M43 Processive motility of multi‐motor kinesin‐14 teams underlies regulation of microtubule minus‐end dynamics. S.R. Norris1, C.E. Strothman1, R. Ohi1, M. Zanic*1; 1Cell and Developmental Biology, Vanderbilt University, Nashville, TN *These authors contributed equally to this work. Microtubules (MTs) are dynamic polymers with two structurally distinct ends: fast‐growing (“plus‐”) and slow‐growing (“minus‐”). Regulated assembly of both ends contributes to the function and formation of MT arrays in cells, but our understanding of minus‐end regulation is limited. This is particularly true during cell division, as current evidence suggests that the known minus‐end regulators – the CAMSAPs – are displaced from the mitotic spindle during cell division. Using time‐lapse total‐internal‐reflection fluorescence (TIRF) microscopy, we demonstrate that the mitotic minus end‐directed kinesin‐14 HSET/KIFC1 protects growing MT minus‐ends by increasing their lifetime and decreasing their catastrophe frequency in a dose‐dependent fashion. Under these conditions, we also observe a strong accumulation of HSET at the growing minus‐ends of MTs. This is surprising because most kinesin‐14 motors are known to be non‐processive as single molecules, and thus decorate the entire MT lattice.


Using single‐molecule TIRF microscopy, we show that HSET indeed undergoes bidirectional diffusion on the MT lattice in a manner driven by its non‐motor tail domain. However, we observe that the motor initiates processive, unidirectional minus end‐directed motility in the presence of free tubulin. We use a quantitative fluorescence intensity analysis to show that each moving particle contains an average of 3‐4 dimeric HSET motors and 12 αβ tubulin dimers. Finally, we show that the non‐motor tail domain of HSET is necessary and sufficient for binding soluble tubulin with nanomolar affinity. From these data, we propose that HSET protects the minus end of MTs through a mechanism where: 1) HSET interacts with its cargo – soluble tubulin – via its non‐motor tail domain, 2) HSET undergoes a cargo‐induced team assembly, and 3) HSET transports these clusters of tubulin to MT minus‐ends which are then protected from catastrophe. To our knowledge, this work represents the first in vitro description of a mitotic minus‐end protector and provides the first characterization of a molecular motor being activated by a tubulin‐induced multimerization.

M44 Electron cryo‐Tomography of Drosophila S2 centrosomes provides novel insights into centriole to centrosome conversion. G. Greenan1,2, B. Keszthelyi1, R.D. Vale2, D.A. Agard1; 1Department of Biochemistry Biophysics, University of California, San Francisco, San Francisco, CA, 2Department of Cellular Molecular Pharmacology, University of California, San Francisco, San Francisco, CA The centrosome is the major microtubule‐organizing center of most animal cells, and plays central roles in cell polarization, intracellular trafficking and cell division. Centrosomes nucleate microtubules from the ill‐defined pericentriolar material (PCM) that nucleates around the centriole, itself a microtubule‐based structure with 9‐fold symmetry. Recent light‐microscopy studies have identified a subset of centrosome proteins that are in close proximity to the centriole wall. However, which proteins attach directly to the centriole wall and the nature of their connection is unclear. Centrosomes purified from Drosophila S2 cells are an excellent system to address these questions; they are comparatively small in terms of centriole length, PCM amounts and centrosome microtubule numbers. Electron cryo‐Tomography (cryo ET) of centrosomes allowed us to generate tomograms in which many features, including the central hub, radial spokes and doublet microtubules (DMT) of the centriole were extremely well resolved. Strikingly we were able to see fibrous brush‐like structures that extended outwards from the cytoplasmic face of the DMT with an apparent 9‐fold symmetry. Using ~10,000 unique subunits, we performed sub‐tomogram alignment and averaging of the DMT and resolved well‐defined densities binding to both the luminal and cytoplasmic‐facing surfaces of the A‐tubule. These densities bind along specific A‐tubule protofilaments, but are distinct from previously identified flagellar and cilia densities. Furthermore, preliminary data indicate remarkably similar decoration of mammalian centrioles, suggesting a conserved decoration of the A‐tubule in centrioles. These data are consistent with a model whereby luminal A‐tubule densities induce conformational changes along a protofilament, potentially licensing it to perform functions like accumulate PCM, or bind inner‐arm and outer‐arm dyneins. Our current work is combining cryoET with specific labeling of candidate proteins, in order to identify which proteins constitute the A‐tubule densities we observed. In this way we hope to use cryoET to bridge the gap between sub‐diffraction light‐microscopy and X‐ray crystallography data of individual components, and assess their contributions in a whole‐centrosome context.


M45 Mechanisms of microtubule nucleation and size in spindles. F. Decker1,2, E. Rieckhoff1,2, J. Brugués1,2; 1Molecular Cell Biology and Genetics, Max Planck Institute, Dresden, Germany, 2Physics of Complex Systems, Max Planck Institute, Dresden, Germany The spindle is the protein machinery responsible for segregating the genetic material into the daughter cells. We now have identified the key molecular players that contribute to spindle structure and function, including motor proteins, cross‐linking proteins, and proteins that regulate microtubule nucleation and polymerization. However, it is still not known how the constituent proteins of the spindle self‐organize into a into a proper size and shape. Furthermore, microtubules—the main building blocks of the spindle—have a lifetime of 20 sec while the entire spindle remains for minutes or even up to hours. Thus, maintenance of the spindle requires constant creation of new microtubules through microtubule nucleation. The mechanism of microtubule nucleation in spindles, however, is still largely unknown because we currently have no means of measuring microtubule nucleation in spindles. Here, we used laser ablation to measure the minus ends of monopolar spindles—where minus ends remain static—assembled in Xenopus leaevis egg extract as a proxy to microtubule nucleation. We found that bulk nucleation from a RanGTP‐mediated gradient of nucleators alone cannot account for the spatial profile of microtubule nucleation in monopoles. Instead, microtubule dependent microtubule nucleation (branching) regulated by the RanGTP gradient explains the nucleation profile and microtubule density in these structures. To further test the microtubule dependent nucleation model in spindles, we perturbed the microtubule lifetime of spindles using MCAK antibodies. This depletion leads to a 5‐fold increase on the microtubule lifetime and a corresponding change in the measured microtubule nucleation profile that can be predicted by the model. These results suggest that the amount of nucleators activated in chromosomes sets the amount of microtubules in spindles and therefore determine their mass and size. This nucleation mechanism could account for the scaling of spindles with cell volume as observed in early embryogenesis or spindles encapsulated in extract, and provides an alternative prediction to previous models based on microtubule dynamics (via a limiting pool of tubulin or MAPs). To test whether microtubule dynamics or microtubule nucleation are responsible for the scaling of spindles, we performed measurements of microtubule dynamics and nucleation during the early rounds of cell division in Zebrafish embryos. Remarkably, all our data on the scaling of microtubule‐lifetime, polymerization and depolymerization dynamics, microtubule length and minus ends distributions, is consistent with the depletion of nucleators as determining the size of spindles.

M46 Dual nucleotide recognition underlies tip‐binding specificity of mammalian EB proteins. D. Roth1, B.P. Fitton1, A. Straube1; 1Centre for Mechanochemical Cell Biology, University of Warwick, Coventry, United Kingdom EB proteins track the ends of growing microtubules and regulate microtubule dynamics both directly and by acting as the hub of the tip‐tracking network. Mammalian cells express cell type‐specific combinations of three EB proteins with different cellular roles. Here we compare the plus end‐tracking behaviour of EB1, EB2 and EB3 in cells and in vitro. We find that the signal from all three EBs correlates strongly to the microtubule assembly rate, but the three signals differ in their maxima and the position from the microtubule tip. Using microtubules built with nucleotide analogues, simulations and site‐directed mutagenesis, we explore the molecular basis for these differences. Our experiments strongly support the idea that EBs bind between 2 protofilaments and sense the nucleotide state of both flanking


β‐tubulins. Different specificities are conferred by amino acid substitutions at the right hand side interface of the EB microtubule‐binding domain with tubulin.

M47 A novel function of kinesin‐1: changing microtubule conformation that accelerates successive kinesin binding. T. Shima1,2, Y. Okada1,2; 1RIKEN QBiC, Osaka, Japan, 2Graduate School of Science, The University of Tokyo, Tokyo, Japan KIF5C, a member of the kinesin‐1 family, transports a wide variety of cargoes along microtubules. Although there are many microtubules in the cytoplasm, it has been reported that KIF5C often binds to only a subset of them. In neurons, for instance, KIF5C accumulates in microtubules in axons preferentially over many other neurites. This selectivity has been explained by existence of different types of microtubules, but its cause is not well understood. Here, we show evidence that positive feedback in KIF5C binding may explain the selectivity. In vitro single molecule imaging showed that the binding of KIF5C accelerates the successive binding of KIF5C to the microtubule, making the binding reaction highly cooperative. In response, a subset of microtubules exclusively bound with a number of KIF5C molecules, while other microtubules bound to far fewer, suggesting that KIF5C itself regulates its affinity with microtubules. These results prompted us to further test the idea that KIF5C induces conformational changes of the microtubule to a high‐affinity form. Consistent with this hypothesis, structural analysis, including the X‐ray fiber diffraction, cryo‐electron microscopy and optical microscopy, demonstrated that the binding of KIF5C induces elongation of the axial tubulin pitch of the microtubule. This elongated form was very similar to the conformation of GMPCPP‐stabilized microtubules, which show high affinity to KIF5C. These results indicate KIF5C can change the microtubule conformation into a high‐affinity form. Moreover, same as the binding of KIF5C and microtubules, the KIF5C‐induced elongation of the tubulin‐pitch occurred in a highly cooperative manner that depended on the KIF5C concentration. These results suggest positive feedback between KIF5C binding and microtubule conformation that regulates the affinity. Furthermore, microtubules differed in their stability, suggesting that KIF5C plays an important role in controlling the dynamics of microtubule network in cells. Collectively, our results overturn the conventional view that kinesin‐1 is only a transporter along microtubules, and unveil a new role of kinesin in controlling the conformation and dynamics of microtubules.

M48 SPEF1 Imparts Mechanical Stability To The Central Pair Microtubules In Cilia. M. Guha1, Y. Jiang1, K.K. Vasudevan2, J. Gaertig1; 1Department of Cellular Biology , The University of Georgia , Athens, GA, 2Department of Biology , Stanford University , Stanford, CA Axonemes of beating cilia experience strong mechanical stresses. Little is known about the factors that preserve the structural integrity of microtubules in motile cilia. SPEF1 is a protein that is conserved in species with motile cilia. SPEF1 contains an N‐terminal calponin homology (CH) microtubule‐binding domain, and its size and predicted tertiary structure resembles that of the End‐Binding (EB) proteins. Tetrahymena cells lacking SPEF1 swam more slowly and ultrastructural studies revealed abnormal 9+0 and 9+1 axoneme cross‐sections, indicating defects in the central microtubules. The C1 central microtubule, positioned on the outside of the ciliary bend, was frequently truncated near the proximal (minus) end. SR‐SIM microscopy of mNeonGreen‐SPEF1 revealed its presence near both the plus and


minus ends, and to a lesser extent along the length of isolated axonemes. Immunogold labeling revealed GFP‐SPEF1 on both the central and outer microtubules. TIRF microscopy and FRAP experiments showed that mNeonGreen‐ SPEF1 diffuses along the cilium and rapidly turns over at the ciliary tip. Homologs of SPEF1 are present only in ciliated organisms but are absent in species that have only 9+0 cilia that lack the central pair. This suggests that SPEF1 has co‐evolved with the central apparatus. To evaluate each of the two central microtubules, we fluorescently tagged either PF16 or Hydin, proteins associated with projections on the C1 and C2 central microtubules, respectively. The loss of SPEF1 caused both truncations near the ends, as well as internal breaks in the central microtubules. C1 was more strongly affected than C2, which correlates with the ultrastructural observations. The loss of SPEF1 made the central microtubules more prone to breakage under an artificial mechanical stress. Importantly, inhibition of ciliary motility rescued the structural defects in the central microtubules caused by the loss of SPEF1. Thus, SPEF1 could protect the central microtubules against damage caused by ciliary motion. While SPEF1 appears to have co‐evolved with the 9+2 cilia, in both Tetrahymena and animal models, SPEF1 also associates with the non‐ciliary microtubules, including the wide diameter microtubules in the pillar cells in the inner ear (Dougherty et al., Cell Motility Cytosk. 2005). We propose that SPEF1 imparts mechanical stability to microtubules that function under high force.

M49 A dynein‐based mechanism, re‐purposed from neuronal migration, establishes and preserves microtubule organization in the axon. A.N. Rao1, A. Falnikar1, M.M. Black1, P.W. Baas1; 1Neurobiology and Anatomy, Drexel University College of Medicine, Philadelphia, PA In addition to its capacity to transport organelles toward minus‐ends of microtubules (MTs), cytoplasmic dynein is capable of sliding MTs in a polarity‐sorting manner. We propose that such forces manifest in different ways at different stages of neuronal life, depending on the capacity of MTs to slide. MT sliding is limited by the attachment of MTs to the centrosome, as well as their length, with longer MTs being less able to slide. In migratory neurons, it has long been believed that all functionally relevant MTs are attached to the centrosome, which would render them unable to slide relative to one another. However, we were able to document a limited amount of MT sliding in the leading process of cultured cerebellar migratory neurons, with treatment with a drug that inhibits MT sliding resulting in an impairment of the trajectory of migration. By knocking down in cultured rat cerebellar granule neurons the expression of ninein, a centrosomal protein responsible for anchoring MTs to the centrosome, we increased the number of centrosome‐unattached MTs and observed an increase in MT sliding. Concomitantly, migration was severely compromised and the leading process became longer, reminiscent of the phenotype of an axon‐bearing post‐migratory neuron. Next, we pursued the idea that in a bona fide axon where MTs appear in a variety of lengths that dynein‐based forces are repurposed to establish and preserve the nearly uniform plus‐end‐out polarity pattern of the MT array. Exposure of cultured rat sympathetic neurons to the Ciliobrevin D, a small molecule dynein inhibitor, for various windows of time resulted in a dose‐dependent appearance of greater numbers of minus‐end‐distal MTs in the axon. In addition, MT transport events were reduced in frequency in both the anterograde and retrograde directions. The MT movements that did occur in the presence of the drug showed an increase in abrupt pausing of movements and directional changes in the movements, as well as abnormal transport rates. These effects are consistent with a mechanism whereby dynein‐based forces propel MTs with plus‐end‐out anterogradely down the axon, while clearing MTs with minus‐end‐out back to the cell body. These two sets of data, on neurons of different stages of development,


demonstrate how dynein‐driven forces that slide MTs can be re‐purposed to orchestrate different neuronal phenotypes.

M50 Fidgetin‐like 2, a novel microtubule regulator, can be targeted in vitro and in vivo to enhance axon regeneration. L.A. Baker1, R. Charaffedine1, M.T. Tar2, P. Nacharaju 1, S.O. Suadicani2, J.M. Friedman1, K.P. Davies1,2, D.J. Sharp1; 1Physiology and Biophysics, Albert Einstein College of Medicine, Bronx, NY, 2Urology, Albert Einstein College of Medicine, Bronx, NY We recently identified the microtubule severing enzyme fidgetin‐like 2 (FL2) to be a potent regulator of cell migration. FL2 suppresses cell migration by selectively severing dynamic MTs at the leading edge of migratory cells. Knockdown of FL2 was found to enhance the rate of cell migration twofold, while application of nanoparticle‐encapsulated FL2 siRNA to rodent cutaneous wound sites significantly increased the rate of wound closure. Observing that FL2 accumulates in mammalian neuronal growth cones, we hypothesized that the severing enzyme might also be a negative regulator of growth cone motility and could therefore be a promising therapeutic target for enhancing axon regeneration after nerve injury. To test this, we analyzed the growth rate of dissociated dorsal root ganglion (DRG) axons with and without FL2 knockdown. We found that neurons expressing FL2 shRNA grew significantly longer processes than neurons expressing scrambled shRNA. To test if the effect on axon growth was mediated by an increase in dynamic MTs, neurons were fixed and double labeled for βIII tubulin (present in all neuronal MTs) and polyglutamylated (pGlu) tubulin (a modified form of tubulin associated with stable MTs). With FL2 knockdown, the ratio of pGlu to βIII tubulin decreased in the distal ends of processes, supporting the hypothesis that FL2 suppresses axon growth by selectively paring down dynamic MTs. In vivo, application of FL2 siRNA promoted nerve regeneration in two rodent nerve injury models: a spinal cord compression model and a cavernous nerve injury model. After moderate‐to‐severe spinal cord crush, treatment of the injury site with nanoparticle‐encapsulated FL2 siRNA led to improved recovery of locomotor and urogenital function. Targeted depletion of FL2 after cavernous nerve transection similarly led to significantly improved functional recovery as early as two weeks post injury (functional recovery was assessed by measuring intracorporal blood pressure in response to electrical stimulation of the nerve). The presence of regenerated axons distal to the injury site was confirmed by histomorphological analysis. We conclude that FL2 as a novel regulator of axon growth and a promising therapeutic target for enhancing nerve regeneration after injury in both the adult central and peripheral nervous systems.

Minisymposium 6: Technological and Biological Frontiers in Signaling and Differentiation

M51 Tuning in to the right station: using optogenetics to probe and rewire MAP kinase signaling. H.E. Johnson1, M. Wilson1, N. Pannucci1, J.E. Toettcher1; 1Molecular Biology, Princeton University, Princeton, NJ Every cell – whether in a microbial or multicellular context – exists in a complex and changing environment. To deal with their complex surroundings, cells have evolved diverse systems to sense external cues (such as nutrients, stresses, or signals from neighboring cells) and create an internal


representation of this information. But we are still largely in the dark about how external information is stored in patterns of protein activity, and how this information is decoded into specific cell fate decisions. I will talk about our efforts to overcome these challenges using cellular optogenetics: the delivery of precise spatial and temporal activity patterns to a signaling protein of interest. We have developed a suite of optogenetic tools to precisely control MAP kinase (MAPK) signaling. We have found that direct stimulation of Ras bypasses intracellular feedback, enabling us to precisely control MAPK activity over time regardless of the history of previous stimuli. Applying light‐induced Ras in vivo revealed specific developmental stages that are sensitive to changes in the dose, duration, or spatial range of MAPK activity. Coupling this system to reporters of MAPK target genes in vitro further enables us to dissect how MAPK dynamics are decoded at the transcript and protein levels.

M52 The role of GPCR localization in signal transduction. N.G. Tsvetanova1, M. von Zastrow1; 1Psychiatry, University of California, San Francisco, San Francisco, CA G protein‐coupled receptors (GPCRs) comprise the largest family of signaling receptors and a very important class of therapeutic targets. GPCRs are well known to signal via cyclic AMP (cAMP) production at the plasma membrane, but it is now clear that various receptors also signal after internalization. This endosome‐derived signal is already known to modify the temporal profile of downstream responsiveness but it remains unknown, particularly for signals transduced by diffusible second messengers, if there is any functional significance to the spatial segregation of initiation sites. Using the beta2‐adrenoceptor (β2‐AR) as a model, we show that endocytosis is required for the full repertoire of downstream cAMP‐dependent transcriptional control. Next, we describe an orthogonal optogenetic approach to definitively establish that the location of cAMP production is indeed the critical variable determining the transcriptional response. Further, we demonstrate that the endocytosis‐dependent transcriptional signal is highly sensitive and saturable, effectively limiting the downstream cellular response to a uniform level. These properties allow the endosome to function as a ‘noise filter’, reducing variability of the downstream response and thus enhance reliability of cognate ligand detection. In summary, our findings reveal a discrete principle for achieving cellular signaling specificity, based on endosome‐mediated spatial encoding of intracellular second messenger production and 'location aware' downstream transcriptional control.

M53 Localization and regulation of mTORC2 activity in cells. M. Ebner1,2, B. Sinkovics2, I. Yudushkin1,3; 1Dept. of Medical Biochemistry, Medical University of Vienna, Vienna, Austria, 2Center for Molecular Biology, University of Vienna, Vienna, Austria, 3Dept. of Structural and Computational Biology, Max F. Perutz Laboratories, Vienna, Austria Mammalian mechanistic target of rapamycin (mTOR) is a critical control of cellular anabolic functions, growth, survival and proliferation. In eukaryotic cells, mTOR exists in multisubunit complexes with distinct functions, mTORC1 and mTORC2. To integrate extracellular signals with intracellular nutrient availability, the activity and regulation of mTOR must be tightly coupled to its intracellular localization. However, unlike mTORC1, the precise localization of mTORC2 in cells and how growth factors affect its activity is currently unknown.


To examine the localization of the endogenous mTORC2, we developed LocaTOR2 – an intracellular reporter of mTORC2 activity, which could be recruited to specific subcellular compartments. In agreement with localization of the obligate mTORC2 component mSin1, examined using structured illumination microscopy, we demonstrate that the plasma membrane, outer mitochondrial membranes and a subpopulation of early‐ and late endosomes, but not ER nor recycling endosomes, display an endogenous mTOR‐dependent, rapamycin‐independent mTORC2 activity. No phosphorylation of the reporter was observed in mSin1‐/‐ cells, demonstrating its specificity for mTORC2. Using LocaTOR2, we show that, in contrast to a previous model (Liu et al, 2015), neither growth factors nor PI3K inhibition affected the localization or activity of the endogenous mTORC2 at the plasma membrane, demonstrating presence of membrane‐bound PI3K‐independent constitutive mTORC2 activity. Consistent with that notion, recruitment of mTORC2 substrate Akt was sufficient for its phosphorylation, independent of growth factors and PI3K activity. Unlike plasma‐ and mitochondrial membranes, mTORC2 activity on endosomes was PI3K‐dependent, suggesting the existence of the distinct mTORC2 populations in cells. Furthermore, we show that regulated phosphorylation of mSin1 affected the localization, but not assembly of mTORC2 in cells. We demonstrate that mSin1 phosphorylation and localization as well as the activity of the endogenous mTORC2 correlated with the cell cycle. Based on these results, we propose that mTORC2 activity towards its substrates is regulated through its intracellular targeting. Collectively, our results demonstrate that, similar to mTORC1, mTORC2 activity is limited to cellular membrane compartments and suggest a new model of mTORC2 regulation based on its localization inside cells. Our study highlights the role of subcellular partitioning as a general mechanism of regulating signalling efficiency inside cells.

M54 OPTOGENETIC INVESTIGATION OF EFFECTS OF BETA‐CATENIN SIGNALING DYNAMICS ON NEURAL STEM CELL DIFFERENTIATION. A. Rosenbloom1, L.J. Bugaj1, M. Tarczyński1, R. Kane2, D. Schaffer1,3,4; 1Bioengineering, University of California, Berkeley, Berkeley, CA, 2School of Chemical Biomolecular Engineering, Georgia Institute of Technology, Atlanta, GA, 3Chemical and Biomolecular Engineering, University of California, Berkeley, Berkeley, CA, 4Helen Wills Neuroscience Institute, University of California, Berkeley, Berkeley, CA Adult neural stem cells (NSC) continuously generate new neurons in specific regions of the brain throughout adulthood to modulate learning and memory, and represent therapeutic targets due to their potential roles in the pathology of diseases and promise in cell replacement therapies. The NSC niche is likely highly dynamic and presents signaling molecules that vary in concentration and duration at timescales ranging from electrophysiological activity, to circadian rhythm, to organismal behavior. Wnt signaling activates β‐catenin to induce neuronal differentiation of adult hippocampal NSCs in vivo and in vitro. To address the question of how dynamics in signaling impacts cell function, our lab developed a tunable optogenetic system to modulate β‐catenin signaling via Cry2 oligomerization of the LRP6 intracellular domain. Similar to responses observed upon Wnt3a ligand addition, blue light illumination of NSCs expressing Cry2‐LRP6c induces neuronal differentiation. This outcome enables one to pose questions related to signaling dynamics: do stem cells differentiate when the integral of signaling during a given temporal window exceeds a key threshold, or do dynamics in signal presentation matter? Continuous illumination at different light intensities in vitro resulted in a saturable dose response in which neuronal differentiation progressively increased from 5% to 60%. However, variation in signaling intensity over time yielded different results. Specifically, we observed that light initial pulses of variable


duration, or oscilllating illumination at frequencies >12 hours, yielded considerably less neuronal differentiation than in cells that received the same overall signal dosage but with continuous illumination. Furthermore, this signal stimulation followed by signal loss led to increased apoptosis, indicating that exposure to this signal rendered cells dependent upon it not only for differentiation but also for survival, potentially offering a general mechanism for removal of incompletely or poorly differentiated stem cells from tissue. Whole transcriptome sequencing (RNA‐seq) revealed several candidate genes, and overexpression of one cell cycle regulatory factor, whose endogenous expression is lost upon Wnt signal withdrawal, significantly decreased apoptosis and rescued differentiation upon signal loss. These results indicate potentially critical roles for both cell cycle exit and prevention of aberrant cell cycle re‐entry upon loss of β‐catenin activity. In sum, we harness optogenetics to demonstrate that not only the overall dosage of a signal but temporal dynamics in its presentation can exert a strong impact on stem cell behavior, work that offers further insights into the biology and potential translational potential of adult neurogenesis.

M55 A DNA based T cell receptor reveals the mechanistic role of spatial organization in ligand discrimination. M.J. Taylor1,2,3, K. Husain2, Z.J. Gartner4, S. Mayor2,3, R.D. Vale1,3; 1Molecular and Cellular Pharmacology, University of California San Francisco, San Francisco, CA, 2Cellular Organization and Signalling, National Centre for Biological Sciences, Bangalore, India, 3HHMI Summer Institute, Woods Hole, MA, 4Dept. of Pharmaceutical Chemistry, University of California San Francisco, San Francisco, CA The adaptive immune response requires that T cell lymphocytes respond to foreign antigens but not self‐peptides. This process begins with an interaction between the T cell receptor (TCR) on the T cell and a peptide‐loaded MHC receptor (pMHC) on the antigen‐presenting cell. The lifetime of TCR‐pMHC bond is believed to be the key determinant in discriminating between activating and non‐activating pMHC molecules. However, the mechanism by which receptor‐ligand lifetime is converted into signals for T cell activation is poorly understood. Here, we have developed a synthetic receptor‐ligand system in which the extracellular domains of the TCR and pMHC are replaced with two hybridizing strands of DNA. Remarkably, the T cell signaling response can discriminate between DNA ligands that differ by a single base pair. Single molecule imaging studies reveals that a long receptor‐ligand lifetime, by itself, does not robustly trigger the phosphorylation of the TCR and recruitment of the downstream kinase ZAP70. Rather, we find that ZAP70 is most efficiently recruited to small clusters containing 2‐4 ligated receptors, which form more readily with longer receptor‐ligand bound times and at higher ligand densities. Receptor clusters form initially through an increased ligand binding rate near pre‐existing receptor‐ligand complexes and later through an additional mechanism of the diffusion‐based collision and coalesce of ligated receptors. This phenomenon of increased ligand binding rate potentially arises from the closer physical proximity of the two membranes established by an initial receptor‐ligand interaction. Collectively these results reveal how T cells convert the extracellular binding energy from a receptor‐ligand interaction into spatial organization and intracellular signal transduction.


M56 Cancer‐associated Ras/Erk signal misperception can drive hyper‐proliferation. L.J. Bugaj1, A. Sabnis2,3, A. Mitchell4, J. Garbarino1, T. Bivona2,5, J.E. Toettcher6, W. Lim1,7,8; 1Department of Cellular and Molecular Pharmacology, University of California, San Francisco, San Francisco, CA, 2Helen Diller Family Comprehensive Cancer Center , University of California, San Francisco, San Franciso, CA, 3Department of Pediatrics, University of California, San Francisco, San Francsicso, CA, 4Program in Systems Biology, University of Massachusetts Medical School, Worcester, MA, 5Division of Hematology and Oncology, University of California, San Francisco, San Francisco, CA, 6Department of Molecular Biology, Princeton University, Princeton, NJ, 7Center for Systems and Synthetic Biology, University of California, San Francisco, San Francisco, CA, 8Howard Hughes Medical Institute, San Francisco, CA Introduction Within cancer, it is believed that oncogenes drive disease through constitutively high levels of pathway signaling. An alternative and untested hypothesis is that cancer mutations might ‘rewire’ a cell’s perception of external signals, so that a cell might errantly interpret non‐proliferative signals as proliferative. Building upon recent work using optogenetics to quantitatively characterize signal filtering properties within living cells, we tested the hypotheses that signal filtering within the Ras/Erk pathway might be altered in cancer cells, and that these changes in input/output responses could contribute to disease phenotypes like hyperproliferation. Methods Optogenetic activation of Ras signaling was accomplished using inducible membrane recruitment of a Ras activator (optoSOS). Optogenetic experiments were conducted both through 1) live cell fluorescence microscopy and 2) newly developed custom devices for controllable illumination of 96‐ and 384‐ well plates followed by single cell immunofluorescence. Results We measured how a panel of Ras/Erk mutant non‐small cell lung cancer lines responded to dynamic Ras inputs, and we found a BRAF mutant with significantly extended kinetics of Erk activation (~5‐fold) and inactivation (~10 fold) compared to non‐transformed cell types. We showed that these changes were a function of the mutant BRAF, and a further search uncovered related BRAF‐mutant cell lines that showed similar behavior. We found that cancer‐associated paradox‐activating BRAF inhibitors also extend Erk kinetics within non‐transformed cells, and we used these as a model to study whether changes in Erk kinetics and signal filtering can result in cellular misperception. We found that while wild‐type and rewired cells responded similarly to constant signals, rewired cells were unable to discriminate dynamic signals from constant ones, leading to improper accumulation of downstream immediate early genes and CyclinD1. By measuring proliferative response to a systematic scan of dynamic Ras inputs, we further demonstrated that, while wild‐type and rewired cells both proliferated under high levels of Ras, rewired cells also proliferated across a range of weaker Ras input patterns that the wild‐type cells ignored. Conclusions This study represents a new paradigm for understanding cancer through a functional lens and highlights the unique ability of optogenetics to quantitatively probe the disease state of signaling networks in living cells. With this approach, we found that as a result of oncogenic mutations, the cell’s internal computation of external inputs can be fundamentally changed in a manner contributing to disease.


M57 Cell‐fate determination by ubiquitin‐dependent regulation of translation. A. Werner1, S. Iwasaki1, C.M. McGourty1, S. Medina‐Ruiz1, N. Teerkorpi1, N.T. Ingolia1, M. Rape1,2; 1Molecular and Cell Biology, UC Berkeley , Berkeley, CA, 2Howard Hughes Medical Institute, Chevy Chase, MD Metazoan development depends on the accurate execution of differentiation programs that allow pluripotent stem cells to adopt specific fates. Differentiation requires changes to chromatin architecture and transcriptional networks, yet whether other regulatory events support cell‐fate determination is less well understood. We have recently uncovered ubiquitin‐dependent regulation of translation as an important feature of cell‐fate determination. The ubiquitin ligase CUL3 in complex with its vertebrate‐specific substrate adaptor KBTBD8 (CUL3KBTBD8) is an essential regulator of human and Xenopus tropicalis neural crest specification. CUL3KBTBD8 monoubiquitylates the ribosome biogenesis regulator NOLC1 and its paralog TCOF1, the mutation of which underlies the neurocristopathy Treacher Collins syndrome. Ubiquitylation drives formation of a TCOF1–NOLC1 platform that connects RNA polymerase I with ribosome modification enzymes and remodels the translational program of differentiating cells in favor of neural crest specification. Here, we make the intriguing finding that β‐arrestin proteins regulate TCOF1/NOLC1 ubiquitylation and the assembly of the ribosome biogenesis platform and we provide evidence that CUL3KBTBD8 regulates translation through cis‐regulatory elements of target transcripts. Taken together, our findings demonstrate an important function of regulated translation during development and suggest a molecular framework for how the dynamic regulation of ribosome function is integrated into cellular differentiation programs.

M58 The roles of IFT and the transition zone in the assembly of the ciliary necklace. S.K. Dutcher1, H. Lin1, U.W. Goodenough2, R. Roth3; 1Genetics, Washington University, St Louis, MO, 2Biology, Washington University , St Louis, MO, 3Cell Biology and Physiology , Washington University, St Louis, MO The ciliary necklace is a collection of highly organized intramembranous proteins (IMPs) in the membrane that surrounds the transition zone of cilia. A ciliary necklace is found in most cilia, but its role remains unclear. We used quick‐freeze deep‐etch electron microscopy to visualize the ciliary necklace in Chlamydomonas cells, and found that the intraflagellar transport machinery is needed to assemble and to maintain the ciliary necklace. Mutations in IFT172, IFT144, IFT140, IFT121 and IFT52 have a reduced number of disorganized IMPs. A temperature‐sensitive mutant in the cytoplasmic dynein heavy chain (fla24) reduces the number of IMPs, but IMPs remain organized. NPHP4 and RPGRIP1L localize to the transition zone, and nph4 and rpg1 mutants also have a reduced number of disorganized IMPs. To begin to understand the role of the ciliary necklace, we asked if the ciliary necklace provides a barrier to the entry of cytoplasmic proteins, a transport/retention function for ciliary proteins, or functions in the production of ectosomes in the gametic phase of the life cycle. We used mass spectroscopy to compare the protein composition of isolated flagella from wild‐type and four mutants with disrupted ciliary necklaces (fla11 (IFT172), fla24, nph4, and rpg1). If the ciliary necklace serves as a barrier between the cytoplasm and cilia, the isolated flagella would contain proteins that are not normally present in the flagella. If it serves in protein transport, known flagellar proteins would be absent. Only one nonflagellar protein (FOX1) is present in all four mutant flagella. Flagella from the rpg1 mutant have more than 50 proteins that are not normally present in the flagella; these include the complete TRiC/CCT chaperonin complex. Since the rpg1 mutant fails to localize NPHP4, we expected, and found, that the nph4 mutant


is missing a subset of these proteins. These 50 proteins are excluded from fla11 and fla24 flagella. In addition, no common set of proteins are missing in the ciliary necklace mutants. We suggest that the ciliary necklace itself is not the barrier between the cytoplasm and the flagellum, but that proteins in the transition zone act as a barrier to the entry of a subset of cytoplasmic proteins. Using the tagged ciliary membrane signaling protein, SAG, we found that loss of the ciliary necklace does not prevent the production of ectosomes containing this molecule. The transmembrane protein TMEM67, which is localized to the transition zone plays a role in the assembly of the ciliary necklace as well as in the assembly of the cilium. We are investigating its localization and role in the ciliary necklace.

M59 CD2AP is a novel regulator of the mechanosensitive adhesion receptor ICAM‐1 and Rac1 activity to control leukocyte transmigration. A. Schaefer1,2, T. van Duijn3, J. Majolee3, K. Burridge1,2, P.L. Hordijk3,4; 1Dept.of Cell Biology and Physiology, University of North Carolina at Chapel Hill, Chapel Hill, NC, 2Lineberger Comprehensive Cancer Center, University of North Carolina at Chapel Hill , Chapel Hill, NC, 3Dept. of Molecular Cell Biology, Sanquin Research and Landsteiner Laboratory, University of Amsterdam, Amsterdam, Netherlands, 4Dept. of Physiology, VU University Medical Center , Amsterdam, Netherlands Inflammation is driven by leukocytes migrating from the blood circulation across the barrier of endothelial cells, lining the inner face of blood vessels. Leukocyte β2‐integrin binding to the endothelial adhesion receptor ICAM‐1 induces ICAM‐1 clustering to mediate leukocyte adhesion, crawling and transmigration into the underlying tissue. Similar to integrins, ICAM‐1 is a mechanosensitive receptor that responds to pulling forces of adherent leukocytes. Although several proteins have been identified to be required for ICAM‐1 function and leukocyte adhesion, much less is known about mechanosensitive negative ICAM‐1 modulators which limit the receptor function to ensure a tight and balanced regulation of the receptor. The majority of neutrophils transmigrate through endothelial cell‐cell junctions (paracellular) and only a minority transmigrate through the endothelial cell body (transcellular). Live‐cell imaging studies show that CD2AP depletion in inflamed primary human endothelial cells promotes neutrophil transmigration. Of note, neutrophils prefer to use the transcellular way because their crawling distance, speed and the time until they transmigrate is reduced. Using live‐cell confocal microscopy, biochemical binding assays and mechanobiological experiments, we identify the endothelial actin‐binding protein CD2AP as a novel interaction partner of the mechanosensitive receptor ICAM‐1 which is required for force‐dependent downstream signaling such as Cortactin and F‐actin recruitment, activation of the RhoGTPase Rac1 and PI3K signaling. CD2AP depletion increases the dynamics of ICAM‐1 clustering and the ICAM‐1 mobility in the clustered adhesion complex, as shown in quantitative confocal microscopy studies. Mechanical force applied on clustered ICAM‐1 impairs CD2AP binding, indicative for a force‐dependent feedback loop. In summary, we identify CD2AP as a negative force‐sensitive modulator of ICAM‐1 clustering which limits the formation of ICAM‐1 adhesion complexes to prevent uncontrolled transcellular leukocyte transmigration. Finally, our findings show that the adhesive function of the mechanoreceptor ICAM‐1 is tightly regulated by negative and positive modulators.


M60 Ciliary Hedgehog signaling controls fatty degeneration of skeletal muscle. D. Kopinke1, J.F. Reiter1; 1Biochemistry, UCSF, San Francisco, CA The Hedgehog pathway signals through the primary cilium to regulate development. The roles of the ciliary Hedgehog signaling in mature tissues is less well understood. Skeletal muscle has a remarkable capacity to regenerate. However, in chronic diseases such as Duchenne muscular dystrophy or with age, healthy muscle is replaced by fibrotic scar tissue and adipocytes. We discovered that mesenchymal progenitors called fibro/adipogenic progenitors (FAPs) in skeletal muscle dynamically upregulate ciliation during regeneration and give rise to adipocytes after injury. Removing cilia from FAPs impairs their ability to differentiate into adipocytes by altering expression of Timp3, an inhibitor of adipogenesis. Removing FAP cilia in a mouse model of Duchenne muscular dystrophy decreases fatty degeneration of the muscle. Pharmacologically enhancing Timp3 activity in vivo represses the adipogenic conversion of FAPs, pointing to a strategy to combat fatty degeneration of skeletal muscle in aging or disease. These results reveal how ciliary signaling controls the response to muscle injury.

Education Minisymposium: Evidence‐Based Education

M61 Scientist Spotlight Homework Assignments Shift Students’ Stereotypes of Scientists and Enhance Science Identity in a Diverse Introductory Biology Class. J. Schinske1, H. Perkins2, A. Snyder1, M. Wyer2; 1Biology, De Anza College, Cupertino, CA, 2Psychology, NC State, Raleigh, NC Research into science identity (1), stereotype threat (2), and possible selves (3) suggests a lack of diverse representations of scientists could impede traditionally underserved students from succeeding in science. However, faculty might refrain from adopting activities directly addressing issues of diversity due to a lack of training on approaching the topic or due to concerns that such activities might impact content coverage. We developed a series of metacognitive assignments, Scientist Spotlights (SS), to feature counter‐stereotypical examples of scientists in a biology class at a diverse community college. Each week students reviewed biographical information about and scientific contributions by diverse scientists matching course topics and responded to metacognitive prompts as homework. We used a quasi‐experimental approach to evaluate four hypotheses surrounding implementation of SS: 1) completion of SS will correlate with use of non‐stereotypical descriptions of scientists, 2) completion of SS will correlate with ability to personally relate to scientists, and 3) shifts in stereotypes/relatability will correlate with interest in science and 4) course grades. Students completed beginning‐ and end‐of‐course essays and quantitative surveys during five SS (n=338 students) and two quasi‐control (n=126 students) courses to assess scientist stereotypes, relatability, and science interest. SS students additionally completed surveys six months after the courses. We coded essays to characterize scientist descriptions and identified prominent themes regarding scientist relatability. Data from coding essays were used to evaluate our hypotheses using RM ANCOVAs, controlling for student demographics. Students’ essays shifted from including 19% counter‐stereotypical descriptions of scientists to 54% after completing SS (p<.001) and conveyed an enhanced ability to relate to scientists (from 2.1 to 3.0 mean relatability ratings, p=.004). These shifts were not apparent in quasi‐control settings. Longitudinal data suggested these shifts were maintained six months after the courses. We further observed positive correlations between these shifts, interest in science, and course grades. As SS require little class time


and complement existing curricula, they represent an easily applicable tool for enhancing science identity, shifting stereotypes, and connecting content to issues of equity and diversity in a broad range of science classrooms. 1) Brickhouse, et al. (2000). Journal of Research in Science Teaching, 37(5), 441‐458 2) Steele & Aronson (1995). Journal of Personality and Social Psychology, 69(5), 797 3) Steinke, et al. (2009). Journal of Women and Minorities in Science and Engineering, 15(4)

M62 Biomedical Sciences Enrichment Program: A model for increasing undergraduate research among underrepresented minority STEM students. D.L. Horton1, T. Hasson1; 1Undergraduate Research Center ‐ Sciences, University of California, Los Angeles, Los Angeles, CA According to the NSF National Center for Science and Engineering Statistics 2012 report, of the 11,288,386 science and engineering students enrolled fulltime at a four year institution, only 14% were Hispanic, 13% African‐American, and 1% American Indian or Alaskan Native. In order to remain globally competitive, the United States will need to recruit, mentor, and train STEM researchers from all segments of the population. Studies have demonstrated that early engagement in undergraduate research experiences helps underrepresented minorities (URM) confidently identify as scientists and develop more focused career paths. To that end, the NIH‐funded Biomedical Sciences Enrichment Program (BISEP) was developed to increase the number of URM students participating in early undergraduate research experiences at the University of California, Los Angeles (UCLA). BISEP is an unpaid, rigorous 6‐week introductory biomedical research training program in which students participate in an abbreviated Molecular Biology Lecture, a Biotechnology Laboratory, and a faculty‐lead Journal Club during the summer following their freshman year. The specific aim of was to have at least 60% of BISEP participants engage in independent research. BISEP students also participated in a variety of workshops and seminars that introduced them to reading, writing and presenting science as well as careers in science. Between 2002 and 20015, 66% of BISEP participants who completed the program (191 out of 288) engaged in independent, faculty‐mentored research, surpassing the original aim of at least 60% engagement. Of the 191 students who participated in research, 120 (63%) received competitive funding to support their independent research projects. Additionally, students reported increases in perceived research‐related knowledge in areas such as molecular biotechniques, scientific literature, and bioinformatics. The rigorous curriculum that BISEP provided enhanced the appreciation of research as a desirable career path amongst URM STEM students at UCLA. Future directions include the addition of an inquiry‐based laboratory component in which students will discover and characterize bacterial mutations that confer antibiotic resistance.

M63 A summer bridge program helps students to maximize their active learning experiences. K.M. Cooper1, M. Ashley1, S.E. Brownell1; 1School of Life Sciences, Arizona State University, Tempe, AZ National calls to improve student academic success in STEM have sparked the development of summer bridge programs designed to help students transition from high school to college (Wischusen et al., 2010; Stolle‐McAllister, 2011; Tomasko et al., 2013). Academic success during the first year in college is often a focus of biology‐centered bridge programs because of the high rate of student attrition from the major during the first year (e.g., Buck, 1985; Ackermann, 1990; Garcia, 1991; Wischusen and Wischusen,


2007; Walpole et al., 2008; Wischusen et al., 2010; Cabrera et al., 2013). However, it is unclear whether these bridge programs are taking into account that many introductory biology courses are being transitioned from traditional lecture to active learning. We designed a two‐week summer bridge program to teach first‐year students introductory biology content and taught the program in an active learning way. We evaluated our program through a set of exploratory interviews of a subset of students who participated in the bridge program. We identified that Bridge students seemed to have a more sophisticated conception of active learning than most introductory biology students. We further explored whether the bridge program positively influenced student approaches to active learning and conducted an additional set of semi‐structured interviews focused on active learning and compared interviews of 26 Bridge students to 8 comparison Non‐bridge students who had been eligible but did not participate in the program. We used grounded theory and content analysis to identify themes from the interview transcripts. We found that Bridge students perceive that they benefitted more from active learning in introductory biology than Non‐Bridge students. Bridge students also described using more strategies than Non‐bridge students to maximize their experiences in active learning. Strategies that only Bridge students identified included: deeply engaging in active learning (reported by 81% of Bridge students), encouraging others to participate in active learning to benefit the other student (65%), being open minded/optimistic when approaching active learning (50%), and intentionally sharing their thoughts with others (54%). Lastly, in stark contrast to Non‐bridge students, Bridge students indicated that they take an equitable approach to groupwork, viewing part of their role in the group to help other students participate. These findings suggest that we may be able to prime students to maximize their own experiences, as well as others’ experiences in active learning classrooms. This has implications for helping us create inclusive classroom communities where students support the learning of each other.

M64 Preparation of Underrepresented males for Scientific Careers: A Study of the Dr. John H. Hopps Jr. Defense Scholars Program at Morehouse College. J.K. Haynes1, R.C. Thompson1, M.R. Moore1; 1Division of Science and Mathematics, Morehouse College, Atlanta, GA The objective of this study was to determine if participants in the Hopps Scholars Program, a research intensive undergraduate experience for STEM students at Morehouse College, performed better academically, and gained entrance to graduate school and research intensive graduate institutions at a higher rate than a comparable group of Morehouse STEM students. Hopps is a highly structured enrichment program aimed at increasing participation of African‐American males in STEM fields. This study uses a two‐to one nearest neighbor matched comparison group design. Morehouse institutional records, Hopps Program records and National Student Clearinghouse data were used to examine differences between Hopps and non‐Hopps STEM graduates of Morehouse. Two‐ way samples t‐and Chi‐square tests revealed significant differences in academic achievement, likelihood of STEM degree pursuit and the classification of graduate institutions attended by Hopps versus non‐Hopps students. After matching, Hopps Scholars earned significantly higher grade point averages at gradation (GPA=3.36 vs 3.21), were 2.5 times more likely to enroll in STEM graduate programs, and 14 times more likely than there non‐Hopps STEM peers to attend doctoral research institutions with higher research activity than those with lower research activity. The Hopps Program’s approach to training African‐American male students for scientific careers is a model of success for other programs committed to increasing the number of underrepresented group males pursuing advanced degrees in STEM.


M65 Decibel Analysis for Research on Teaching (DART): Measuring Classroom Decibel Levels to Quantify Active Learning. M.T. Owens1, S.B. Seidel2, M. Wong3, J. Schinske4, K.D. Tanner1; 1Biology, San Francisco State University, San Francisco, CA, 2Biology, Pacific Lutheran University, Tacoma, WA, 3Center for Computing for Life Sciences, San Francisco State University, San Francisco, CA, 4Biology, DeAnza College, Cupertino, CA Although evidence supports using active teaching strategies in university classrooms, measuring the extent to which such strategies are used remains difficult. Currently available classroom observation tools provide detailed descriptions of pedagogies instructors use but are difficult to deploy at large scales, so often only a handful of class sessions per course are analyzed. To fill the need for a method that can efficiently assess whether large numbers of faculty consistently use non‐lecture strategies, we created Decibel Analysis for Research on Teaching (DART). We hypothesized we could measure deviations from lecture by analyzing patterns in decibel levels of classroom noise associated with lecture and non‐lecture teaching strategies. To develop DART, >50 hours of classroom recording were annotated and compared to the corresponding classroom decibel levels, showing a high correlation between certain decibel patterns and teaching strategies such as pair‐discussion and minute papers. These human‐curated annotations were used to train a computer algorithm, DART, to distinguish Single‐Voice (e.g. lecture), Multi‐Voice (e.g. pair‐discussion), and No‐Voice (e.g. minute paper) modes. By comparing human‐ and computer‐generated annotations, we found DART correctly determined the type of activity 89.5% of the time, estimated with 10‐fold stratified cross‐validation. To what extent can DART reveal differences in patterns of active learning across large numbers of courses? In collaboration with 49 biology instructors from 15 community colleges and a public comprehensive university, DART was applied to 1720 hours of recordings of 67 entire courses, ranging in size from 4‐287 students. We quantified the percentage of time spent in Single‐, Multi‐, and No‐Voice modes. We found that courses taught by different instructors varied considerably, with per‐course Single‐Voice percentages ranging from 69‐100%, demonstrating DART can identify differences in teaching strategy use. Also, some instructors had high variability between class sessions: in one course, the percentage of time spent in Single‐Voice ranged from 15‐90% in different class sessions. Furthermore, while only 31% of courses had Multi‐ or No‐Voice in every class session, 88% had Multi‐ or No‐Voice in at least half of their class sessions, emphasizing the need for an instrument that can analyze each session of an entire course. While DART cannot assess the quality of active learning, our results demonstrate that DART is valuable for high‐throughput, cost‐effective, and comprehensive analyses of the extent of non‐lecture pedagogies in biology courses. Future DART studies may be able to inventory the presence of non‐lecture activities across courses, departments, or even entire universities over time.

M66 Measuring student course preparation and the effect on exam performance in a partially flipped class. K.B. Shannon1; 1Biological Sciences, Missouri ST, Rolla, MO I have flipped one day a week of my Cell Biology course. For the flipped day, students watch online videos before class. For the other two days of the week, students are assigned a textbook reading before


the lecture. Although there are incentives for both types of pre‐class preparation, studies have shown that students rarely read the textbook before class. I hypothesized that more students would watch the videos than read the textbook, and that reading the textbook and watching online videos would positively correlate with exam performance. To examine the effect of student engagement outside of the classroom, I used student surveys and data automatically collected by the video streaming software. Extra credit questions at the end of each exam on student reading and study habits were used to give a reading or studying rating. For student video viewing analysis, the percentage of total minutes of video watched was determined. I have collected data for four semesters with four exams each semester. Analysis of the data so far shows that exam performance correlates significantly to reading and to watching the online videos in a single variable analysis. Multivariate analysis for each semester to control for student GPA was also performed on one year’s data. Reading has the strongest effect, with video viewing having a smaller effect, and time spent studying having a negative effect. Attitude surveys showed that students reported liking videos more than reading the textbook, and data analysis shows that watching videos with little reading occurs more frequently than reading the textbook without watching videos. In conclusion, although students prefer to watch videos, their reading of the textbook has a larger effect on their exam scores. The negative effects of study time on exam scores may be due to “cramming.” Implications for teaching Cell Biology as a flipped course will be discussed.

M67 Research intensive courses: An alternative to faculty‐mentored independent research for undergraduate students. B.M. Fenner1, M.E. Fenner2; 1Biology, King's College, Wilke‐Barre, PA, 2Fenner Training and Consulting LLC, Forty Fort, PA Faculty‐mentored independent research (IR) provides applied exposure to experimental design, protocol development, troubleshooting, data analysis, and student collaboration while developing broadly applicable skills. King’s biology department has traditionally required all majors to complete at least one semester of independent research. While many students benefit from IR, there is a subset of students who lack motivation or aptitude for a meaningful IR experience. Therefore, we re‐evaluated our IR requirement and pursued alternative tracks that would also meet our learning objectives. We implemented research intensive courses (RICs) that are five‐credits for one‐semester with an integrated lecture and laboratory component. Through a RIC, students learn the theory behind their research topic, become trained in relevant techniques, design a research project, write a research proposal, conduct their proposed research, participate in a journal club, analyze and synthesize data, and present their research as a poster and a mock journal article. Here, we report data from 4 semesters of two RICs offered by the same professor: Molecular Neuroscience and Cellular Biology. We quantify the value of RICs, compared to IR, by assessing student performance on four measures: research proposal, laboratory proficiency, poster design and presentation, and final article. Our assessment goal is for 80% of students to surpass 75% on each measure. Preliminary data suggest that the RICs met our goal. The differences in mean grade for students in RICs and IR over all measures were not statistically significant (p > 0.05, n=32). Interestingly, 87.5% RIC students and 79% IR students met the final paper goal. The difference might be due to additional writing assignments required in RICs. Currently, we are performing data analysis: (1) to identify predictors of student success (or risk) in IR and (2) to determine outcome differences between RICs versus IR. We conclude that RICs provide a meaningful research experience for students of all aptitudes, meet the learning objectives of our research requirement, are useful as a bridge to independent research, and are an attractive alternative for programs with limited resources.


M68 Developing a curriculum‐wide "Pipeline" CURE that connects PUIs with R1s. T. Lee1, B. Carpenter1, C. May2, D.J. Katz1, K.L. Schmeichel2; 1Cell Biology, Emory University, Atlanta, GA, 2Biology, Oglethorpe University, Atlanta, GA Course‐based undergraduate research experiences (CUREs) improve student learning, engagement, and retention in STEM disciplines. Most CUREs are implemented during a single semester, and thus may have limited benefits among an entire undergraduate cohort. We reasoned that embedding a curriculum‐wide CURE would allow students to acquire concept sophistication and confidence in laboratory skills in a deliberately developmental manner throughout their undergraduate career. Such an approach is particularly feasible at small teaching‐focused colleges where students tend to take more than one course with an instructor while completing a major. Moreover, if such a CURE was developed in collaboration with a lab at an R1 research institution, it could provide an opportunity to explore “high risk” ideas that could spin‐off into fruitful projects for other trainees. A collaboration between an investigator at Emory University, and a nearby small liberal arts college, Oglethorpe University, resulted in the design of such a mutualistic strategy, the “Pipeline” CURE. Students at Oglethorpe are incrementally trained, over the course of a four‐year curriculum, through laboratory activities designed around the R1 laboratory’s research focus: the epigenetic basis of germline reprogramming in C. elegans. Students are introduced to nematode husbandry and behavior in Introductory Biology. As sophomores, they hone their skills in a C. elegans in a linkage project as part of their Genetics course requirements. In an upper‐level class, students apply their accrued skills in performing novel genetic screens germane to the R1 laboratory’s research agenda. While a fully‐trained cohort will not emerge from the pipeline until late in 2017, we introduced the first modules throughout the curriculum during the 2015‐2016 academic year. Positive signs of this program’s impact are already evident. Four student groups trained in Introductory Biology in the fall 2015, unprompted by their instructors, chose to use C. elegans for their year‐end projects in spring 2016. In an upper level Developmental Biology class the majority of lab work was devoted to a class‐wide RNAi screen for enhancers/suppressors of a sterility defect in C. elegans. A pre‐/post‐style survey of this class of 19 (four of whom had prior experience in Genetics) indicated that on average, students self‐reported improvements in both their laboratory skills and their confidence in lab work. Qualitative assessments indicated that working on this project helped them understand the hard work that it takes to make scientific discoveries. Once this strategy is fully streamlined, especially with respect to logistics, the “Pipeline” CURE approach may prove applicable to other teaching‐focused/research‐based academic alliances.

M69 Developing course‐based undergraduate research experiences (CUREs) through long‐term mentorship: The ASCB Mentoring in Active Learning and Teaching (MALT) program. M.J. Wolyniak1, A.J. Prunuske2, M.J. Dobro3, S.M. Wick4; 1Biology, Hampden‐Sydney, Hampden‐Sydney, VA, 2Microbiology and Molecular Genetics, Medical College of Wisconsin, Wausau, WI, 3Human Biology, Hampshire, Amherst, MA, 4College of Biological Sciences, University of Minnesota, St. Paul, MN The life science education community has responded to the recommendations of the American Association for the Advancement of Science (AAAS) Vision and Change document with several initiatives designed to improve the way in which undergraduates learn science. These initiatives have often taken the form of one‐time workshops that generate awareness of and interest in active learning techniques


among participants. However, they have been less successful with respect to generating the sustainable change necessary to bring real reform to undergraduate science education. To create sustainable change, long‐term faculty development initiatives focused on mentorship are needed so that instructors seasoned in active learning can convey their experiences to mentees interested in using these pedagogical techniques as the centerpiece of their own teaching. The American Society for Cell Biology (ASCB) Education Committee created the Mentoring in Active Learning and Teaching (MALT) program to provide a means for members with an interest in developing improved and sustainable active learning techniques to gain experience in this style of teaching through close, long‐term interaction with a veteran teaching mentor. While originally conceived as a means to mentor new student‐centered life science instructors in the undergraduate classroom, MALT now focuses on the development of instructors who wish to develop a dynamic course‐based undergraduate research experience (CURE). Current and future life science instructors pair themselves up with seasoned veterans of CURE development and work with them and their students over the course of a semester or longer to develop a CURE that will allow the mentee to bring authentic research into his or her classes. In our pilot studies, we collected qualitative and quantitative data based on participant interviews and coding videos of student and instructor actions during classroom activity (Smith et al., 2013), respectively, that suggest that MALT mentorships have made positive gains in promoting sustainable active learning techniques among participants. Going forward, we wish to use instruments like the Laboratory Course Assessment Survey (Corwin et al., 2016) and Experimental Design Ability Test (Sirum and Humburg 2011) to assess the effectiveness of the MALT laboratory intervention. MALT provides an excellent means to equip current and future faculty with the proven pedagogical techniques needed to effectively train the next generation of scientists. In 2017, the ASCB will financially support a limited number of MALT Fellows to travel to their mentor's institution and/or obtain laboratory supplies for CURE development. Consult http://www.ascb.org/mentoring‐in‐active‐learning‐and‐teaching‐malt/ for more information.

M70 Exploring innovation, assessment, and evaluation beyond ASCB 2016. A.J. Prunuske1, S.B. Seidel2; 1Medical College of Wisconsin, Wausau, WI, 2Pacific Lutheran University, Tacoma, WA This part of the session will actively engaged participants in reflecting on topics from the first set of the presenters. It will include a structured discussion with the first set of speakers and will be lead by session co‐chairs.

Microsymposium 1: Autophagy/ESCRT

E1 Autophagy maintains metabolism and functional activity of a subset of aged hematopoietic stem cells. T.T. Ho1, M.R. Warr1, E. Adelman2, E.V. Verovskaya1, O.M. Lansinger1, M.E. Figueroa2, E. Passegue1; 1Medicine, University of California, San Francisco, San Francisco, CA, 2Pathology, University of Michigan, Ann Arbor, MI With age, hematopoietic stem cells (HSCs) lose their ability to produce all blood cells, resulting in a decline in immune responses and increased rates of blood diseases in the elderly. We previously showed


that autophagy is crucial for protecting HSCs from metabolic stress and that old HSCs have increased autophagy levels. Subsequent analyses of aged GFP‐LC3 autophagy‐reporter mice revealed a striking heterogeneity in old HSCs, with only ~ 30% having high autophagy levels. To understand how autophagy regulates HSC function, we generated mice with conditional deletion of the essential autophagy gene Atg12 (Atg12cKO) in the blood system. We observed premature aging phenotypes in Atg12cKO HSCs that resemble old HSCs, including similar differentially methylated regions, enhanced myeloid differentiation, rapid loss of self‐renewal potential and HSC depletion under regenerative challenges such as transplantation. Analyses of Atg12cKO HSCs by electron microscopy also revealed an excess of mitochondria, similar to the ~ 70% old HSCs that do not show high autophagy levels. In both cases, this resulted in increased OXPHOS, ROS production, cell cycle activation, and precocious myeloid differentiation. In addition, only old HSCs with low levels of autophagy showed functional exhaustion after transplantation, while old HSCs with high levels of autophagy resembled young healthy HSCs and displayed robust long‐term regenerative capacity. Our results demonstrate that HSCs require autophagy to maintain a low metabolic, quiescent state that preserves their regenerative capacity, and that during aging HSCs become increasingly dependent on autophagy activation for their functional maintenance. Old HSCs with low autophagy levels are severely impaired, likely causing age‐associated declines in regenerative potential and blood production. This suggests that increasing autophagy in old HSCs could help promote HSC fitness and rejuvenate the aging blood system.

E2 Proximity‐based biotinylation identifies a new role for autophagy in ribonucleoprotein loading into extracellular microvesicles. A.M. Leidal1, H. Huang2, J. Ye1, T. Solvik1, J.Y. Liu1, F. Kai3, J. Goldsmith1, M. Stanley1, T. Marsh1, A. Wiita2, J. Debnath1; 1Department of Pathology, University of California, San Francisco, San Francisco , CA, 2Department of Laboratory Medicine, University of California, San Francisco, San Francisco, CA, 3Department of Surgery, University of California, San Francisco, San Francisco, CA Autophagy traditionally involves the sequestration of cytoplasmic material into vesicles and their delivery to the lysosome for degradation. Predominantly viewed as a catabolic mechanism, we are beginning to appreciate that the autophagy pathway also promotes protein secretion; it is genetically required for the unconventional secretion of IL1beta and HMGB1, two proteins lacking N‐terminal leader sequences. However, the broader functions of autophagy in unconventional secretion and precise mechanisms remain poorly defined. To further address the role of autophagy in secretion, we developed a novel strategy using proximity‐specific biotinylation to label proteins that engage the autophagy machinery, and subsequently, are secreted outside of the cell. Specifically, we fused the promiscuous mutant BirA biotin ligase to the autophagy protein MAP1LC3B (ATG8), and expressed this recombinant probe within cells. Addition of exogenous biotin to BirA‐LC3‐expressing cells promotes the robust biotinylation of known LC3‐interacting proteins (e.g. p62/SQSTM1). More importantly, the conditioned media of biotin‐pulsed BirA‐LC3 cells contains numerous biotinylated unconventionally secreted proteins that are distinct from established intracellular LC3 interaction partners. Quantitative proteomics reveals the BirA‐LC3 secretome as a highly interconnected network enriched in ribonucleoproteins (RNPs) and components of extracellular microvesicles (EMVs, exosomes). We have validated multiple new candidate substrates of secretory autophagy that biochemically interact with LC3 and genetically require autophagy for their extracellular release. The profound enrichment of EMV cargo proteins in the BirA*‐LC3‐labelled secretome suggests that specific RNPs are loaded into microvesicles in an autophagy‐dependent manner. Focusing on two candidate RNPs for detailed mechanistic study,


HNRNPK and SAFB, we demonstrate these molecules interact with LC3B and localize within LC3+ vesicles. Moreover, secreted HNRNPK and SAFB co‐fractionate with markers of EMVs in a form that is protease‐resistant. EMVs harboring these proteins are also highly enriched with lipidated LC3, but devoid of the unmodified form. Remarkably, the packaging of EMVs containing LC3, HNRNPK and SAFB is completely ablated upon genetic deletion of ATG7 and ATG12. We are presently testing the genetic requirement for ESCRT/nSMase pathway components in the autophagy‐dependent secretion of these RNPs, and interrogating if autophagy regulates the packaging and release of discrete RNAs via EMVs. Collectively, our data demonstrates a new role for the autophagy pathway in specifying RNPs that are packaged into EMVs and highlights a novel mechanism by which autophagy controls the secretion of RNPs and RNAs outside of the cell.

E3 The rapamycin‐insensitive TOR Complex 2 network signals through mitochondria and the ER To regulate autophagy during amino acid limited growth conditions. T. Powers1, A. Vlahakis1, N.L. Muniozguren1; 1Molecular and Cellular Biology, University of California, Davis, Davis, CA The conserved target of rapamycin (TOR) kinase is a central regulator of growth in eukaryotic cells that couples nutrient availability to cellular behavior. TOR forms two distinct membrane‐associated protein complexes, termed TORC1 and TORC2, wherein TORC1 is uniquely inhibited by the immunosuppressant drug rapamycin. TORC1 is an established negative regulator of autophagy, a catabolic process required for recycling of cytoplasmic contents under conditions of cellular starvation and other forms of stress. Thus, in response to rapamycin treatment or nitrogen starvation, conditions that inhibit TORC1 activity, the process of autophagy becomes induced significantly. By contrast, we have demonstrated that TORC2 functions as an independent and positive regulator of autophagy, specifically under conditions of amino acid‐restricted growth. Under these conditions, TORC2 signals to its downstream target kinase, Ypk1, and inhibits the calcium/calmodulin‐dependent phosphatase calcinuerin, which in turn allows for activation of the general amino acid control (GAAC) response and induction of autophagy. Mitochondria have also been shown to regulate autophagy in response to amino acid availability; however, the relationship between mitochondria and TORC2 signaling has remained unexamined. We have now determined that mitochondrial respiration is influenced by TORC2‐Ypk1 signaling and that mitochondria function upstream of both calcineurin as well as the GAAC response to regulate autophagy. Moreover, we also find that that mitochondrial‐derived ROS, in response to TORC2‐Ypk1 signaling, regulates autophagy by controlling the activity of the ER‐localized transmembrane protein Mid1 that is involved in calcium homeostasis. Remarkably, accumulation of ROS requires components involved in formation of ER‐mitochondrial contacts, emphasizing the role of organelle contact sites in the regulation of autophagy. Together, these findings define a novel pathway involving TORC2, mitochondria, the ER, and autophagy.

E4 Membrane binding by CHMP7 coordinates ESCRT‐III dependent nuclear envelope reformation. J.G. Carlton1, Y. Olmos1, A. Perdrix‐Rosell1; 1Cancer Studies, KCL, London, United Kingdom In addition to its role in membrane abscission during cytokinesis, viral budding, endosomal sorting and plasma membrane repair, the Endosomal Sorting Complex Required for Transport‐III (ESCRT‐III) machinery has recently been shown to seal holes in the reforming nuclear envelope (NE) during mitotic


exit. ESCRT‐III also acts during interphase to repair the NE upon migration‐induced rupture, highlighting its key role as an orchestrator of membrane integrity at this organelle. Whilst NE localisation of ESCRT‐III is dependent upon the ESCRT‐III component CHMP7, it is unclear how this complex is able to engage nuclear membranes. Here, we show that the N‐terminus of CHMP7 acts as a novel membrane‐binding module. This membrane‐binding ability allows CHMP7 to bind to the Endoplasmic Reticulum (ER), an organelle continuous with the NE, and provides a platform to direct NE‐recruitment of ESCRT‐III during mitotic exit. CHMP7’s N‐terminus comprises tandem Winged‐Helix domains and by using homology modelling and structure‐function analysis, we identify point mutations that disrupt membrane‐binding and prevent both ER‐localisation of CHMP7 and its subsequent enrichment at the reforming NE. These mutations also prevent assembly of downstream ESCRT‐III components at the reforming NE and proper establishment of post‐mitotic nucleo‐cytoplasmic compartmentalisation. These data identify a novel membrane‐binding activity within an ESCRT‐III subunit that is essential for post‐mitotic nuclear regeneration.

E5 Autophagy regulates myelin compaction in the final stages of CNS myelination. A.N. Bankston1,2, M.D. Forston2,3, A.M. Smith1,2, R.M. Howard1,2, S.R. Whittemore1,2; 1Department of Neurological Surgery, University of Louisville, Louisville, KY, 2Kentucky Spinal Cord Injury Research Center, University of Louisville, Louisville, KY, 3Department of Anatomical Sciences and Neurobiology, University of Louisville, Louisville, KY Myelination, the concentric wrapping of highly specialized membrane around axons, is critical for fast action potential conduction in vertebrates, and lack of myelin causes severe neurological disorders. Development of myelinating oligodendrocytes (OL), the sole providers of central nervous system (CNS) myelin, from proliferating oligodendrocyte precursor cells (OPCs) requires migration to target axons and defined morphological changes prior to myelination. During myelination, the extension of an OL process that transitions into a flattened myelin sheath, initiation of axon ensheathment, wrapping of the myelin sheath around the axon, and removal of cytoplasm to form compact myelin is tightly controlled. Although removal of cytoplasm through myelin compaction is critical to form mature, functional myelin, this final stage of myelination is poorly understood. Here, we present evidence that autophagy, the targeted isolation of cytoplasm and organelles by the double‐membrane autophagosome for lysosomal degradation, regulates myelin formation in mature OL. A marked increase in autophagic activity coincides with OL differentiation, with OL processes having the greatest increase in autophagic flux. Multiple lines of evidence indicate that autophagosomes form in developing myelin sheathes before trafficking from myelin to the OL soma. Pharmacological autophagy inhibition blocked myelination, producing OL surrounded by many short processes. Conversely, autophagy stimulation enhanced myelination while reducing the overall number of OL processes. Mice with OL‐specific genetic deletion of the essential autophagy gene Atg5 develop a rapid tremor and die around postnatal day 12. Further analysis revealed nearly complete loss of CNS OPCs and reduced myelination. Intriguingly, the surviving Atg5‐null OL produced myelin wraps that failed to compact. Mass spectrometric analysis of autophagosomes purified from differentiating OL identified actin and 2’,3’‐cyclic‐nucleotide 3’‐phosphodiesterase, proteins shown by others to be involved in myelin compaction. These results implicate autophagy as a key regulator of OPC survival initially and myelin compaction during the terminal stages of myelination. Autophagy may provide an attractive target to both promote OL survival and subsequent myelin repair after injury. Supported by NS073584, GM103507, Norton Healthcare, and the Kentucky Spinal Cord and Head Injury Research Trust.


E6 RILP is a stress‐activated dynein recruitment factor required for mammalian autophagosome biogenesis, maturation and transport. N.V. Khobrekar1,2, R.B. Vallee2; 1Department of Biological Sciences, Columbia University, New York, NY, 2Department of Pathology and Cell Biology, Columbia University, New York, NY Autophagy has emerged as a critically important mechanism for clearance of damaged and degraded proteins and organelles, and for more general responses to cell stress and starvation. Defects in autophagosome formation and behavior lead to a variety of neurodevelopmental and neurodegenerative diseases. However, the cellular mechanisms controlling autophagosome biogenesis, maturation and transport are only partially understood. RILP (Rab Interacting Lysosomal Protein) is a α–helical coiled‐coil protein, which binds dynein and dynactin through its N‐terminal, and vesicular cargoes through its C‐terminal domains. In studies to define the function of RILP in motor protein‐mediated vesicular transport in neurons, we have identified an important role for RILP as a novel adaptor protein for autophagosome (AP) transport, with a surprising role in biogenesis as well. We found that RILP and LC3 exhibit clear punctate colocalization in cortical neurons, the extent of which is significantly increased by small molecule mediated induction of autophagy. GFP‐RILP expression alone also dramatically stimulated AP formation, with clear colocalization and comigration of most RILP‐ and LC3 puncta. In contrast, the RILP C‐terminal cargo‐binding domain decorated APs but completely arrested their transport. Together, these results suggest potential roles for RILP in stress‐induced AP biogenesis and transport. To test these possibilities further, we identified and mutated LC3‐interacting sites in RILP, which abrogated LC3 binding and colocalization and blocked autophagosome transport. These mutations, however, had no effect on RILP targeting to Rab7‐positive late endosomes/lysosomes (LLEs), indicating that the Rab7 and LC3 interactions with RILP are mutually exclusive. To test further the role of RILP in AP biogenesis, we looked at several early AP markers, and found that, indeed, RILP associates with omegasome‐like structures on the ER as well as with early isolation membranes. Collectively, our findings identify RILP as a novel autophagosomal ‘intelligent scaffold’ required not only for AP retrograde transport, but also for stress‐induced biogenesis and maturation. Supp. by NIH GM102347.

E7 The nuclear splicing factor Acinus moonlights as a cytosolic activator of autophagy. N. Nandi1, L.K. Tyra1, H. Kramer1; 1Dept. of Neuroscience, University of Texas Southwestern Medical Center, Dallas, TX To remove cellular debris and maintain homeostasis neurons employ basal autophagy, a process that is positively regulated by Acinus [1]. In Drosophila and mammalian cells, Acinus is a primarily nuclear protein. The ASAP complex formed by Acinus together with Sap18 and RNPS1 binds to spliceosomes and regulates a subset of alternative splicing [2, 3], suggesting that Acinus may stimulate autophagy through its function in alternative splicing. However, comparison of the transcriptomes of wild type and acinus larval fat bodies did not uncover any support for this hypothesis. Furthermore, AcinusASAP , a mutant that interrupts its binding interface with Sap18 and RNPS1, disrupts its function in alternative splicing as predicted. However, when expressed at endogenous levels in flies, AcinusASAP is primarily cytosolic and enhances autophagy indicating that this mutant separates the regulatory functions of Acinus in splicing and autophagy. Further support for a splicing‐independent, non‐nuclear role of Acinus came from expression of a myristoylated Acinus that also displayed enhanced autophagy. Because wild‐type Acinus


is degraded in the cytoplasm [1], we reasoned that modifications that stabilize Acinus would promote its function in autophagy. From a targeted screen, we found that Cdk5 kinase phosphorylates Acinus at serine 437. Knocking down Cdk5, or its required cofactor p35, drastically reduces Acinus phosphorylation at serine 437. Physiological relevance for this modification was confirmed by a phosphomimetic S437D mutation, which stabilized Acinus protein, enhanced autophagy and improved life span in wild type or p35 mutant flies. Nandi, N., L.K. Tyra, D. Stenesen and H. Krämer, J Cell Biol, 2014. 207: 253‐68. Malone, C.D., et al., Genes Dev, 2014. 28: 1786‐99. Hayashi, R., D. Handler, D. Ish‐Horowicz and J. Brennecke, Genes Dev, 2014. 28: 1772‐85.

E8 LAMP proteins bind cholesterol and contribute to NPC1‐mediated cholesterol export from lysosomes. J. Li1, S.R. Pfeffer1; 1Department of Biochemistry, Stanford University School of Medicine, Stanford, CA LAMP1 and LAMP2 proteins are highly abundant, ubiquitous, mammalian proteins that line the lysosome limiting membrane and protect it from lysosomal hydrolase action. LAMP2 deficiency causes Danon’s disease, an X‐linked hypertrophic cardiomyopathy. LAMP2 is needed for fusion of phagosomes with lysosomes and chaperone‐mediated autophagy, and its expression improves tissue function in models of aging. We have shown that LAMP1 and LAMP2 bind cholesterol in a manner that buries the cholesterol 3β‐hydroxyl group: binding is competed by 25‐hydroxycholesterol and 7‐ketocholesterol but not lanosterol or cholesterol sulfate. A secreted LAMP2 protein carries bound cholesterol as determined by thin layer chromatography and mass spectrometry of material released from the purified protein. LAMP2 also binds tightly to NPC1 and NPC2 proteins that export LDL‐derived cholesterol from lysosomes, in vitro and in cells. In cyclodextrin treated cells, less LAMP2 associates with NPC1, and more of it oligomerizes; this process may signal cholesterol availability to the cytoplasm. Quantitation of cellular LAMP2 and NPC1 protein levels shows that LAMP proteins are more abundant than NPC1 and represent a significant cholesterol binding site at the lysosome limiting membrane. Functional rescue experiments in embryonic fibroblast cells from mice deficient in LAMP1 and LAMP2 proteins show that the ability of LAMP2 to facilitate cholesterol export from lysosomes relies on its ability to bind cholesterol directly. We propose that NPC2 protein solubilizes cholesterol from accumulated membranes within lysosomes; this cholesterol is then transferred to NPC1 (if available) or LAMP proteins, which can drive the overall export process by temporarily holding on to the cholesterol before NPC2 transfers it to NPC1 for export. Experiments are in progress to test this model.


E9 Vps4 Induces Dynamic Instability in ESCRT‐III Polymers. B. Mierzwa*1, N. Chiaruttini*2, L. Redondo‐Morata3, J. Moser von Filseck2, J. König4,5, I. Poser6, T. Müller‐Reichert4, S. Scheuring3, A. Roux2,7, D.W. Gerlich1; 1Institute of Molecular Biotechnology of the Austrian Academy of Sciences (IMBA), Vienna, Austria, 2Department of Biochemistry, University of Geneva, Geneva, Switzerland, 3U1006 INSERM, Aix‐Marseille Université, Marseille, France, 4Medical Theoretical Center, Dresden University of Technology, Dresden, Germany, 5Electron Microscopy Unit, Francis Crick Institute, London, United Kingdom, 6Max Planck Institute for Molecular Cell Biology and Genetics, Dresden, Germany, 7Swiss National Centre for Competence in Research Programme Chemical Biology, Geneva, Switzerland *Equal contributions The Endosomal Sorting Complex Required for Transport (ESCRT)‐III forms polymers that mediate membrane constriction and fission during various fundamental cellular processes, such as cytokinetic abscission, multivesicular body formation, and virus budding. Prevailing models propose that ESCRT‐III forms persistent filaments that deform membranes for fission. However, whether ESCRT‐III polymers exchange their subunits with soluble cytoplasmic pools – like other force‐generating filament systems such as actin and tubulin – is not known. We therefore investigated ESCRT‐III polymer dynamics in live human cells and in vitro. Using fluorescence recovery after photobleaching, we observed a rapid turnover of growing ESCRT‐III polymers at cytokinetic abscission sites – almost two orders of magnitude faster than net accumulation rates of ESCRT‐III. The ATPase Vps4, previously thought to function as a disassembly factor only, accumulated simultaneously with ESCRT‐III during abscission, and was required for ESCRT‐III turnover and membrane constriction. To study how individual ESCRT‐III components contribute to polymer dynamics, we reconstituted polymerization of budding yeast ESCRT‐III subunits on supported lipid membranes. The scaffold subunit Snf7 alone polymerized into stable patches consisting of single‐stranded filament spirals. Subsequent addition of the ESCRT‐III subunits Vps2 and Vps24 doubled the filament width and suppressed further polymerization of Snf7. Unexpectedly, this growth inhibition was alleviated when Vps4 was added, resulting in a gradual increase of ESCRT‐III polymers that rapidly exchanged with cytoplasmic pools. High‐speed atomic force microscopy revealed that under these conditions, membrane‐bound ESCRT‐III filaments continuously grow and shrink in highly dynamic spiral arrays. Our study reveals that Vps4 induces dynamic instability in ESCRT‐III polymers. This rapid and continuous ESCRT‐III filament reorganization might leverage the amount of energy available for membrane deformation, and it might facilitate morphological adaptation of ESCRT‐III polymers to variable membrane geometries, thereby enabling ESCRT‐III to function in a large diversity of cellular processes.

Microsymposium 2: Genome Replication, Gene Regulation, and Gene Editing

E10 Rif1 controls the development of late replication at the Drosophila Mid‐Blastula Transition. C.A. Seller1, P.H. O'Farrell1; 1Biochemistry and Biophysics, University of California, San Francisco, San Francisco, CA Late replication of heterochromatin is a highly conserved feature of the eukaryotic S phase, but the rapid cell cycles of many early embryos deviate from this norm. Embryogenesis in Drosophila melanogaster begins with remarkably short S phases (3.5 min) because of the early replication of the


normally late satellite repeats. Beginning at cell cycle 14, known as the Mid‐Blastula Transition (MBT), the cell cycle lengthens dramatically due to the emergence of late replication. By unknown mechanisms replication of each satellite repeat is delayed by differing amounts producing a protracted replication schedule. Developmental onset of late replication follows and requires the downregulation of Cdk1, and the speed of early S phases requires high interphase Cdk1 activity. Here we examined the Drosophila homolog of the yeast Rif1 protein based on suggestions that one of its activities is to control replication timing. Using a live imaging approach, we show that Rif1 localizes to satellite sequences as the embryonic S phase lengthens, and that the recruitment of Rif1 to these repeats depends upon pre‐RC formation. By tracking an individual satellite in real time we show that the dissociation of Rif1 from this repeat is coupled to the initiation of replication. Experimental manipulations of Cdk1 activity demonstrate that localization of Rif1 to foci requires the downregulation of Cdk1. Rif1 is phosphorylated in the early embryo, and a mutant of Rif1 lacking predicted CDK and DDK phosphorylation sites does not disperse from satellite foci, and its expression blocks the completion of replication. Rif1 interacts with protein phosphatase 1 (PP1) in the embryo, and Rif1 function depends upon its PP1 interaction motifs. Embryos lacking Rif1 show early replication of satellite repeats, a fast S phase 14, and a partially penetrant extra mitotic division. Thus, Rif1 is required for the onset of late replication at the MBT, and we propose that Rif1 is a part of a Cdk1 centered developmental program orchestrating the progressive lengthening of embryonic S phases.

E11 Can a steroid hormone receptor induce DNA re‐replication? S.A. Gerbi1, J. Urban1, Y. Yamamoto1, L. Kadota1, A. Lee1, J. Leung1, J. Bliss1, M.S. Foulk2; 1BioMed Division; Department of Molecular Biology, Cell Biology and Biochemistry, Brown University, Providence, RI, 2Department of Biology, Mercyhurst University, Erie, PA Normally DNA replication origins are activated no more than once per S phase to ensure complete duplication of the genome. One of only two known biological systems of intrachromosomal re‐replication that overcome this regulation are the nine major and nine minor DNA puffs that occur in late larval life in the salivary gland polytene chromosomes of the lower dipteran fly, Sciara coprophila . We have extensively studied one DNA puff (II/9A). We have now identified many of the other amplification loci by first assembling the Sciara genome from embryos followed by identifying loci in Sciara puff‐stage salivary glands that have increasing copy number throughout the developmental window of DNA amplification. How do these rogue DNA puff amplification origins defy the genomic controls that normally regulate origin activation to no more than once per S phase? For the largest DNA puff (II/9A), DNA amplification precedes a burst of transcription at this locus that is coupled with morphological puffing. DNA puffing can be induced prematurely by injection of the steroid hormone ecdysone, and transcription at DNA puff II/9A is regulated by ecdysone, but is this also the trigger for the preceding DNA amplification? We have cloned the two isoforms of the Sciara ecdysone receptor (EcR); EcR‐A is the predominant form in Sciara salivary glands. Immunofluorescence using an antibody raised against EcR‐A reveals its localization at polytene chromosome loci while they are undergoing re‐replication before they subsequently form DNA puffs. At the molecular level we have shown that injection of ecdysone into young larvae prematurely induces amplification of DNA puff II/9A, and we have now extended these experiments to other DNA puffs. The DNA puff II/9A amplification origin resides 2.5 kb upstream of gene II/9‐1, as shown by 2D gels, 3D gels and PCR analysis of nascent strand abundance. We have developed the method of Replication Initiation Point (RIP) mapping and have used this method to identify the start site of DNA replication at the nucleotide level for DNA puff II/9A. Furthermore, we found that the Origin Recognition Complex (ORC) binds at this start site. Inspection of the DNA puff II/9A sequence showed matches to the consensus EcRE, including one that is directly adjacent to the ORC


binding site and has the ability to bind EcR‐A as determined by gel shifts. We have developed a method for Sciara transformation with site‐specific replacement of a large fragment of DNA and are now testing whether the ORC‐adjacent EcRE is required for amplification of DNA puff II/9A. We are eager to see if EcR, a transcription factor that is the master regulator of insect development, can also serve as a replication factor and if it is the long sought after amplification inducing factor.

E12 Female‐biased embryonic death from DNA replication stress‐induced inflammation. A.J. McNairn1, C. Chuang2, M. Wallace3, J.C. Schimenti1; 1Biomedical Sciences, Cornell University, Ithaca, NY, 2Genetics, Stanford University, Stanford, CA, 3Medicine, University of Oxford, Oxford, United Kingdom DNA replication stress (RS) can trigger genomic instability, checkpoint activation, senescence, and apoptosis. Though extensively studied in cell culture and cancer paradigms, little is known about the impact of RS during embryonic development, a period of rapid cellular proliferation. To study the effect of genetically‐encoded RS on development, we utilized a previously characterized hypomorphic allele in mice, Mcm4 Chaos3 (mini‐chromosome maintenance‐deficient 4 homolog), that causes genomic instability linked to MCM complex destabilization and pan‐reduction of Mcm2‐7 mRNA and protein levels. Further genetic reductions of Mcm genes caused synthetic phenotypes including embryonic lethality. We have found that female mouse embryos with a high level of RS due to MCM‐deficiencies are dramatically more susceptible than males to embryonic lethality. This gender bias was not attributable to differences in Mcm levels or defects in X ‐inactivation, but rather differential sensitivity to RS‐induced inflammation. Female embryos with high intrinsic RS could be rescued by testosterone or NSAID (ibuprofen) treatment during gestation, as well as transgenic sex reversal, and maternal genotype. These results demonstrate that the response to DNA replication stress differs dramatically between the sexes and has profound effects on development and viability.

E13 Redundant GA‐binding early transcription factors regulate the Drosophila histone locus body. L.E. Rieder1, K. Boltz2, B. Duronio2, E.N. Larschan1; 1Molecular and Cellular Biology, Brown University, Providence, RI, 2Genetics and Molecular Biology, University of North Carolina Chapel Hill , Chapel Hill, NC Within the complex environment of the nucleus, specific transcriptional and regulatory processes are coordinated within membraneless domains known as nuclear bodies. These bodies are critical for the precise spatial and temporal regulation of the genome and include the conserved histone locus body (HLB), which regulates the cell cycle transcription and unique processing of the core histone genes. The histone genes are arrayed in ~100 tandem copies at the histone locus, around which the HLB forms during embryonic development. While multiple factors are present at the HLB, the molecular mechanism by which this essential nuclear body is established during embryogenesis remains unknown. Here we show that the conserved GA dinucleotide repeats in the 300bp Histone3‐Histone4 bidirectional promoter (H3‐H4p) are required to establish the HLB in Drosophila. There are two known Drosophila GA‐binding transcription factors: 1) Chromatin Linked Adaptor for MSL Complex (CLAMP), which is known to organize the dosage compensated male X‐chromosome nuclear body and 2) the well‐studied GAGA Factor (GAF), which opens chromatin and modulates transcriptional pausing. However, immunostaining in multiple tissues reveals that only CLAMP is present at the HLB in wild‐type situations. ChIP‐seq experiments demonstrate that CLAMP precisely binds to the GA repeats in the H3‐H4p where it


regulates local chromatin accessibility and histone gene transcription, while GAF is specifically excluded. Therefore, CLAMP regulates two different nuclear bodies: the HLB and the male X‐chromosome. Animals lacking CLAMP expression die at the larval stages with intact HLBs, likely because GAF re‐localizes to the histone locus in the absence of CLAMP. Overall, we have identified the first DNA‐binding motif that regulates HLB organization and two redundant transcription factors that are recruited by this motif to coordinate transcription of histone genes. We have identified CLAMP as a new common regulator of nuclear bodies because it regulates both the HLB and the male X‐chromosome.

E14 Identification and characterization of active enhancers as potential biomarkers of aggressive colorectal cancer. R. Wu1, C.R. Jerde1, R.B. Diasio1, S.M. Offer1; 1Department of Molecular Pharmacology and Experimental Therapeutics, Mayo Clinic, Rochester, MN Colorectal cancer is the third most common cancer and second leading cause of cancer‐related deaths in the United States. Each year there are 1.2 million new cases and 600,000 deaths globally attributed to colorectal cancer. Tumor metastasis is the primary cause of mortality; more than one third of colorectal tumors progress to the metastatic state. Understanding the molecular changes that occur within cancer cells during metastasis is key to developing targeted therapies and identifying early‐stage cancer markers that may be predictive of later tumor aggressiveness. To identify these changes, we developed a cellular model of metastasis by exposing the pre‐metastatic SW‐480 colorectal cancer cell line to multiple rounds of selection for the capacity to invade through a synthetic extra‐cellular matrix. Following repeated selection, cell invasion capacity was enhanced by greater than 10‐fold. This change was accompanied by the loss of epithelial‐cell markers, such as E‐cadherin, and the up‐regulation of markers commonly associated with tumor cells that have undergone epithelial‐mesenchymal transition, such as N‐cadherin and vimentin. Likewise, invasive cells showed increased adherence to Laminin I, Collagen I, and Collagen IV. Thousands of variant enhancers that correlated with cellular invasiveness were identified by histone H3K27ac and H3K4me1 ChIP‐seq. We found that some enhancers lost H3K27ac prior to down regulation of H3K4me1 and correlated gene expression, which was a potential biomarker to identify aggressive colorectal cancer. Collectively, our data support the use of this in vitro cancer metastasis model for the study of the molecular mechanisms that contribute to cancer progression.

E15 A "toggle" in the RNase H domain of Prp8 correlates splicing fidelity and catalytic efficiency. M. Mayerle1, S. Ledoux1, A. Price1,2, M. Raghavan3, E. Moehle1,4, H. Hadjivassiliou1, S. Mendoza1, D. Velazquez1, N. Stepankiw3, A. de Bruyn Kops1, K. Patrick1,5, M. Dinglasan1, Y. He1, J. Abelson1, J. Pleiss3, C. Guthrie1; 1Biochemistry and Biophysics, University of California, San Francisco, San Francisco, CA, 2Physics, University of Colorado, Boulder, Boulder, CO, 3Molecular Biology and Genetics, Cornell University, Ithaca, NY, 4Molecular and Cell Biology, University of California, Berkeley, Berkeley, CA, 5Microbial Pathogenesis and Immunology, Texas AM Health Science Center, Byron, TX The spliceosome, which catalyzes pre‐mRNA splicing, must balance the need for high fidelity splicing with the need for rapid, efficient splicing. We propose that the Prp8 RH domain contributes to this balance by toggling between transitional and catalytic conformations throughout the splicing cycle. Using a set of previously published prp8 mutants, we link the transitional conformation of RH to high


fidelity, low efficiency splicing, and the catalytic to low fidelity, high efficiency splicing. Specifically, our data show that 1) transitional RH alleles splice reporter pre‐mRNAs less efficiently than wild type or catalytic RH alleles, 2) transitional RH alleles are deficient in promoting the 1st step of splicing, 3) catalytic RH alleles exhibit increased ability to splice mutant ACT1‐CUP1 pre‐mRNA reporters whereas splicing is decreased by transitional RH alleles, and 4) that genome‐wide, splicing is more error‐prone in catalytic RH alleles and hyperaccurate in transitional RH alleles. These data support a model where the spliceosome conformation is similar at both catalytic steps and directly link splicing fidelity and efficiency.

E16 Single Molecule Tracking of Ace1p In Saccharomyces cerevisiae defines a characteristic residence time for non‐specific interactions of transcription factors with chromatin. G.D. Mehta1, D.A. Ball1, R. Salomon‐Kent1, D. Mazza2, T. Morisaki3, F. Mueller4, J.G. McNally5, T.S. Karpova1; 1Laboratory of Receptor Biology and Gene Expression, National Institutes of Health, Bethesda, MD, 2Centro di Imaging Sperimentale e Università Vita‐Salute San Raffaele, Istituto Scientifico Ospedale San Raffaele, Milano, Italy, 3Department of Biochemistry and Molecular Biology, Colorado State University, Colorado, CO, 4Computation Imaging and Modeling Unit, CNRS, Institut Pasteur, Paris, France, 5Helmholtz Center Berlin, Berlin, Institute for Soft Matter and Functional Materials, Berlin, Germany In vivo Single Molecule Tracking (SMT) has recently developed into a powerful technique for measuring and understanding the transient interactions of transcription factors (TF) with their chromatin response elements. However, this method still lacks a solid foundation for distinguishing between specific and non‐specific interactions. To address this issue, we took advantage of the power of molecular genetics of yeast. Yeast TF Ace1p has only five specific sites in the genome and thus serves as a benchmark to distinguish specific from non‐specific binding. Here we show that the estimated residence time of the short‐residence molecules is essentially the same for Hht1p, Ace1p and Hsf1p, equalling 0.12‐ 0.32 s. These three DNA‐binding proteins are very different in their structure, function and intracellular concentration. This suggests that (a) short‐residence molecules are bound to DNA non‐specifically, and (b) that non‐specific binding shares common characteristics between vastly different DNA‐bound proteins and thus may have a common underlying mechanism. We develop new and robust procedure for evaluation of adverse effects of labeling, and new quantitative analysis procedures that significantly improve residence time measurements by accounting for fluorophore blinking. Our results provide a framework for the reliable performance and analysis of single molecule TF experiments in yeast.

E17 Peptide based non‐viral delivery of CRISPR‐Cas9 for efficient genome engineering in mammalian cells. G. Divita1, N. Desai1; 1Aadigen LLC, Pacific Palisades, CA The recent discovery of CRISPR‐Cas9 has provided a highly efficient tool for site‐directed genome engineering in eukaryotic cells, opening new perspective for innovative gene therapy. However, the cellular delivery of the RNA‐guided nuclease Cas9 and of the associated guide RNA (gRNA) remains a limitation to the overall potency and for the therapeutic application of the CRISPR‐Cas9 system. Here we report a new delivery platform that can potently deliver active Cas9/gRNA ribonucleoprotein complexes in a variety of hard‐to‐transfect mammalian cell lines, including primary and T cells. We have designed


the ADGN‐technology based on short amphipathic peptides that form stable neutral nanoparticles with Cas9/gRNA ribonucleoprotein complexes through non‐covalent electrostatic and hydrophobic interactions. Self‐assembly of these peptides around Cas9/gRNA leads to peptide/protein/nucleic acid nanoparticles that remain stable over time in serum. ADGN‐nanoparticles mediate functional targeted‐delivery of Cas9:gRNA complexes into a wide variety of cell lines, allowing simultaneous multiple‐gene editing via both Non‐Homologous End Joining (NHEJ) and Homology Directed Repair (HDR). The ADGN‐technology promoted delivery of active Cas9/gRNA targeting the eGFP exogenous gene in different GFP‐expressing cell lines (U2OS, HEK, K562, JURKAT), without inducing any significant toxicity. FACS analysis revealed that the level of GFP positive cells was reduced by >85% using a 1 nM concentration of Cas9:gRNA. In addition, ADGN mediated delivery of three different Cas9:gRNA complexes targeting three endogenous EMX1, AAVS and HPRT genes led to simultaneous robust editing of the three genes in hard‐to‐transfect human neuronal and Embryonic stem cells, with mean indel frequencies of >70% for the three genes. In contrast, no or negligible gene editing were observed on ES cell lines using a control lipid‐based delivery system, RNAiMAX. Furthermore, ADGN was used for the delivery of a ternary Cas9/gRNA/ssDNA complex for genome engineering using precise HDR. In the ternary complex, Cas9/gRNA targeting EMX1 gene was associated to an ssDNA donor allowing the insertion of 10 nucleotides containing HindIII and NheI restriction enzyme sites into the EMX1 gene by HDR in U2OS and JURKAT cells, with high efficiency (up to 47%, >3 fold higher than RNAiMAX) of HDR in both cell lines. Given the robustness of the biological response achieved through this approach and the absence of associated toxicity, we are further exploring the ADGN‐technology for therapeutic application of CRISPR‐Cas9 based genome editing along with non‐viral gene delivery (e.g. to generate CAR‐T cells) which has been previously demonstrated with the ADGN peptides.

E18 Conformational dynamics of Cas9 during DNA binding. Y.S. Dagdas1, J.S. Chen2, S.H. Sternberg3, J.A. Doudna2,3,4,5, A. Yildiz2,6; 1Biophysics Graduate Group, University of California, Berkeley, Berkeley, CA, 2Department of Molecular and Cell Biology, University of California, Berkeley, Berkeley, CA, 3Department of Chemistry, University of California, Berkeley, Berkeley, CA, 4Howard Hughes Medical Institute, University of California, Berkeley, Berkeley, CA, 5Physical Biosciences Division, Lawrence Berkeley National Laboratory, Berkeley, CA, 6Department of Physics, University of California, Berkeley, Berkeley, CA Cas9 is an RNA‐guided endonuclease that cleaves double‐stranded DNA as part of the bacterial adaptive immune system. By programming its single‐guide RNA, Cas9 is also widely utilized as a genome editing tool. DNA cleavage activity of Cas9 is controlled directly by the conformational state of the HNH nuclease domain, but Cas9 conformational dynamics during DNA binding remain unclear. Using single‐molecule FRET assays, we identified a long‐lived intermediate state of S. pyogenes Cas9. Upon DNA binding, the HNH nuclease reversibly transitions from the RNA‐bound to the intermediate state before it docks irreversibly into its catalytically active conformation for DNA cleavage. The intermediate state becomes more populated as mutations at the protospacer adjacent motif (PAM) distal site are increased, preventing transitions from the intermediate to the docked conformation. We show that Cas9 requires divalent cations to lock into an “on” switch upon cleavage of a complementary target, and provide a structural basis for the loss of DNA cleavage activity when Cas9 binds to off target sites.


Microsymposium 3: Cellular Interactions and Disease

E19 Rapid phenotypic assessment for mutations involved in human diseases: Uncovering a novel model for cardiac arrhythmia in the nematode C. elegans. Y.M. Clovis1, A. Webb1, C. Turner1, K.E. McCormick1, S.R. Lockery1; 1RD, NemaMetrix Inc., Eugene, OR Humans have over 70 potassium channel genes, but only some of these have been linked to disease. In this respect, the KCNQ family of potassium channels is exceptional: mutations in four out of five KCNQ genes underlie a range of diseases including cardiac arrhythmias, deafness, and epilepsy. For example, mutations in KCNQ1 are associated with heart arrhythmias such as long‐QT syndrome (LQTS), in which the cardiac action potential is prolonged. Mutations in KCNQ2 and KCNQ3 are associated with benign neonatal epilepsy. Homologs of KCNQ genes are found in a wide range of model organisms, including flies and mice, where they often have functions analogous to those in humans, but less is known about the functions of these potassium channels in the nematode C. elegans. We investigated the effect of mutations in C. elegans KCNQ‐like genes on the electrical excitability of the pharynx, a rhythmic muscular pump involved in feeding. One such gene, kqt‐1, is orthologous to the subfamily of human KCNQ genes that comprises KCNQ2 to KCNQ5, and is mainly expressed in the muscles of the pharynx. kqt‐3 is orthologous to the human gene KCNQ1. Although not expressed in the pharynx, kqt‐3 is present in mechanosensory and chemosensory neurons, which can regulate feeding behavior. We hypothesized that mutations in kqt‐1 and kqt‐3 induce abnormalities in pharyngeal pumping that are reminiscent of cardiac arrhythmias. Taking advantage of a new microfluidic system for C. elegans electrophysiology (NemaMetrix ScreenChip(TM)), we tested this hypothesis by measuring the duration of pharyngeal action potentials and related parameters of pharyngeal pumping. We present evidence that kqt‐1 and kqt‐3 are necessary for normal pharyngeal pumping, consistent with the well‐known role of KCNQ potassium channel mutations in generating cardiac arrhythmias in humans and model organisms. Both genes are required for normal pumping frequency and pump duration, whereas kqt‐1 is also required for normal inter‐pump intervals. We propose that the strains tested, as well as the recording methodology used, could be the basis of future studies to identify pharmacological agents that mitigate certain arrhythmias.

E20 Developing 3D Molding Technique to Study the Effect of Geometry on Protein Expression. J.L. Flournoy1, S. Park1, Y. Chen1; 1Mechanical Engineering, Johns Hopkins University, Baltimore, MD To gain insights into the dynamic process of tumor progression in breast cancer, it is very useful to study the development of breast cancer from the onset in a longitudinal manner. However, longitudinal in vivo studies are difficult to perform in animals and not practical in cancer patients. Therefore, in vitro models are commonly adopted for such studies. However, the current prevalent in vitro model system, namely 3D cell culture, does not fully represent the biophysical characteristics of the mammary gland. The mammary gland exhibits a branched morphology comprising of ducts and acini. The recent finding that the geometry governs the expression level of many breast cancer‐related proteins through mechanosignaling pathways, such discrepancy is especially profound. The difficulty in modeling the ductal structure of the mammary gland is the necessary resolution. Naturally occurring mammary ducts can be less than a hundred microns in diameter. To study the differences in a ductal structure compared


to spheroids and 2D cell culture, we develop a novel 3D molding technique in order to create wells to mimic partial duct structure. A micron size wire is embedded in a gelatin scaffold and cells are then injected and grown to confluence. The expression of Vimentin and E‐Cadherin are measured using immunostaining. The proteins were analyzed in varying geometries. We found that the Vimentin expression level is 1.77 fold higher in spheroids compared to its ductal counterparts; and E‐Cadherin expression is 1.87 fold higher in spheroids than ducts. When compared to traditional 2D cell culture Vimentin expression was 1.07 fold higher in spheroids and E‐Cadherin expression was 7.44 fold higher. The results indeed demonstrate tissue geometry is critical in protein expression regulation. By comparing expression of proteins in different geometries, the most accurate in vitro model can be identified, we expect such more realistic model system can facilitate more precise drug screening, as the biophysical environment will be more closely represented, hopefully leading to less drug failures between the in vitro tests to clinical trials. Studying breast cancer with our model system may also contribute to improved accuracy of early diagnosis by characterizing the changes in cellular morphologies and protein expression in a quantifiable manner. Furthermore, our novel molding techniques can be modified in the future to form other glandular structures, where various diseases can be studied in vitro but in a more realistic microenvironment.

E21 Safe dose of bisphenol A to mothers during gestation and lactation increases proliferative rate and neoplastic lesions in mammary gland of female gerbil (Meriones unguiculatus) offspring. E.R. Leonel1, S.P. Campos1, D. Leonel1, L. Falleiros‐Junior1, S. Taboga1; 1Departament of Biology, São Paulo State University (IBILCE/UNESP), São José do Rio Preto, Brazil The established oral reference dose (RfD) of the xenoestrogen bisphenol A (BPA) may promote development of neoplastic lesions. However, its effects in offspring after gestation and lactation treatment are less known. This work aimed to check the presence of morphological alterations and proliferative pattern in mammary gland of female gerbil offspring after gestational and lactational treatment with two different doses of BPA. Pregnant gerbils were divided in three groups (n=6): G1 (50 µg/kg BPA), G2 (5000 g/kg BPA) and C (control group). The pregnant animals from G1 and G2 were submitted to daily gavage with BPA until weaning. Female animals from each offspring were used. They were submitted to a carcinogenic dose of N‐methyl‐N‐nitrosurea (MNU) (50mg/kg) at three months age and euthanized at 6 months age. Mammary glands were removed and processed for light microscopy. Analysis comprised immunohistochemistry for quantification of proliferative epithelial cells with PCNA and phospho‐histone‐H3 (P‐H‐H3) and α‐actin in myoepithelial cells. Stroma surrounding the epithelial structures was identified by Picrosirius staining and quantified by stereology. Statistical analysis was performed for all parameters. The proliferation rates were higher in G1 for both immunohistochemical analysis. P‐H‐H3 showed an increase of at least three fold in the proliferation rate of mammary gland epithelium in G1 (9.14%) in comparison to C (2.15%) and G2 showed an intermediate rate (4.36%). Although there was no statistical difference between the groups for PCNA the results showed higher percentages for G1 (69.42%) and G2 (62.29%) in comparison to C (57.75%). The α‐actin quantification showed no statistical differences between the groups, despite the percentage was higher in C group (6.56%) in comparison to G1 (4.76%) and G2 (5.59%). The quantification of collagen type I from connective tissue around epithelial structures did not show statistical differences between the groups either. Nevertheless, there was more collagen in C group (21.26%) and G1 (20.06%) in comparison to G2 (13.53%). The morphological analysis of the glands showed neoplastic lesions mainly in G1, not observed in C, suggesting carcinogenic alterations in these animals. Few morphological alterations were observed in C group including alveolar hyperplasia in four of 6 animals evaluated. In conclusion, the daily RfD


treatment with BPA during embryonic and lactational phases may cause worse injuries in the mammary gland than higher doses of this compound. According to the findings, it may facilitate the development of neoplastic lesions in the gerbil mammary gland. Approved by UNESP Ethics Committee (113/2015). Financial support: FAPESP, CNPq, CAPES.

E22 Mesenchymal stem cells respond to matrix stiffness to promote mammary carcinoma proliferation via prosaposin secretion. S. Ishihara1, D.R. Inman1, W. Li2, S.M. Ponik1, P.J. Keely1; 1Department of Cell and Regenerative Biology, University of Wisconsin‐Madison, Madison, WI, 2Departments of Orthopedics and Rehabilitation Biomedical Engineering, University of Wisconsin‐Madison, Madison, WI The tumor microenvironment contains cancer cells, non‐cancerous cells, and extracellular components such as extracellular matrix (ECM). Previous studies found that interactions between the cells and the microenvironment contribute to cancer progression via chemical and mechanical stimuli. Recently, it has been reported that mesenchymal stem cells (MSCs) differentiate into cancer associated fibroblasts (CAFs) in response to chemical stimuli from cancer cells and thereby promote cancer progression. However, the contribution of mechanical stimuli, such as stiffness of the ECM, to MSCs in cancer is poorly understood. In this study, we revealed that MSCs showed CAF phenotypes and promote mammary cancer progression in response to mechanical stimuli. On a stiff substrate, MSCs treated with conditioned media from cancer cell culture expressed increased levels of alpha smooth muscle actin (alpha‐SMA), a marker of CAFs, compared to the MSCs cultured on a soft substrate. MSCs grown on a stiff substrate displayed higher expression and activity of YAP and increased phosphorylation of myosin light chain (MLC) compared to a MSCs grown on a soft substrate. In addition, knockdown of YAP by shRNA decreased expression of alpha‐SMA and phosphorylation of MLC in MSCs on a stiff substrate. Pharmacological inhibition of MLC phosphorylation by H1152 treatment also reduced expression of alpha‐SMA and activity of YAP in MSCs. Cell‐cell communication between MSCs and carcinoma cells was bi‐directional, as conditioned medium from MSCs cultured on a stiff substrate, but not a soft substrate, increased proliferation of mammary carcinoma cells. The soluble factor prosaposin was highly secreted by the MSCs on a stiff substrate, and the addition of recombinant prosaposin increased proliferation of mammary carcinoma cells. Furthermore, secretion of prosaposin was promoted in YAP‐overexpressed MSCs on a soft substrate. These results suggest that increased stiffness of the ECM in the tumor microenvironment induces differentiation of MSCs to CAFs via YAP and actomyosin contractility, which may trigger progression of mammary cancer via secretion of prosaposin.

E23 Disruption of cell polarity, but not adhesion, alters Notch‐mediated proliferation and differentiation in endometrial cancer. A.B. Gladden1, E.R. Williams1, R.R. Broaddus2; 1Genetics, The University of Texas MD Anderson Cancer Center, Houston, TX, 2Pathology, The University of Texas MD Anderson Cancer Center, Houston, TX Apicobasal polarity is required for the proper formation and organization of specialized epithelial tissues. Loss of polarity is observed in most epithelial cancers, but whether this is an early or late event is not known. The epithelium of the endometrium, the primary cell type that is thought to give rise to endometrial cancer, is shed and re‐established after each menstrual cycle. This continual repopulation


requires constant reorganization of the epithelium. We investigated the status of apicobasal polarity in low‐grade endometrial cancer to assess the role of polarity in early endometrial tumor development. In normal human endometrial glandular epithelial cells, the apical polarity protein Par3 and the apical marker Ezrin localize to the apical region of the cell, but this localization is lost in low‐grade endometrial cancer, suggesting that apicobasal polarity is disrupted. This loss of polarity in early endometrial cancer correlates with increased proliferation and decreased differentiation. Interestingly, localization of the adherens junction marker, E‐cadherin, is not altered in low‐grade endometrial cancer, suggesting that loss of apicobasal polarity precedes loss of adhesion.Moreover, whereas Par3 protein levels are decreased in the majority of endometrial cancer cell lines, levels of E‐cadherin remained unchanged. Reintroduction of Par3 resulted in decreased proliferation as well as increased expression of markers associated with differentiation. As changes in cell polarity can alter overall cell membrane and membrane receptor localization, we examined membrane receptor signaling pathways and found that Notch downstream signaling was decreased in cells lacking Par3. Reintroduction of Par3 promoted membrane localization of Notch, which was lost in tumor cells with reduced Par3 levels. Surprisingly, in low‐grade endometrial cancer we observed a significant decrease in Notch target gene expression as well as mislocalization of Notch receptors compared to normal endometrium. These findings indicate that loss of polarity in endometrial tumors disrupts Notch receptor localization and signaling thus promoting early tumor development. Our work now reveals a role for apicobasal polarity in endometrial development, homeostasis and tumorigenesis mediated through the regulation of signaling pathways that maintain normal endometrial proliferation and differentiation.

E24 A plasmid‐based MMP14‐FAP biosensor for monitoring and measuring MMP14 activity in real‐time. A. Braun1, K.A. Myers1; 1Biology, University of the Sciences in Philadelphia, Philadelphia, PA Matrix metalloproteases (MMPs) function to guide endothelial cell (EC) migration by locally degrading the extracellular matrix (ECM) to promote directional cell movement. MMP14 is a unique metalloprotease that is trans‐membrane and has been implicated in cell invasion in cancers. Engagement with and modification of the ECM is crucial for directed cell migration. However, it is currently not possible to monitor MMP14 activity in living cells in real time. We have developed a Fluoragen‐Activated Protease (FAP)‐based biosensor that contains a MMP14‐specific substrate cleavage sequence. Once cleaved by MMP14, the biosensor undergoes dimerization that allows it to sequester a fluorescent dye. In this way, biosensor cleavage results in a fluorescent signal that can be used to measure the activity of MMP14 with high spatial and temporal resolution. Experiments designed to evaluate the biosensor as an imaging tool have revealed that the trans‐membrane biosensor travels through normal vesicular exocytic and endocytic pathways, and that dye fluorescence requires biosensor dimerization. Additionally, live‐cell imaging measurements of the activated/cleaved MMP14 biosensor have identified that the trafficking and delivery of MMP14 is modulated in response to the density of the extracellular matrix. Moreover, experimental data show that MMP14‐induced cleavage of the biosensor is spatially restricted to the leading edge and branched protrusions of ECs, where dye binding of the biosensor is co‐localized with growing MT plus‐ends. Live‐cell monitoring of active Rac1 revealed an accumulation of biosensor dye binding specifically within Rac1‐active subcellular protrusions, suggesting that MMP14 activity is increased in regions of the cell where Rac1 activity and MT growth is increased. Thus, we have designed a biologically active MMP14‐FAP biosensor that is effective in monitoring MMP14 activity with high spatial and temporal resolution in living cells.


E25 Direct visualization of phagocytes acting as Trojan horses and their contribution to brain invasion by the environmental yeast Cryptococcus neoformans. F.H. Santiago‐Tirado1, M.D. Onken2, R.S. Klein3, J.A. Cooper2, T.L. Doering1; 1Molecular Microbiology, Washington University School of Medicine, St. Louis, MO, 2Biochemistry Molecular Biophysics, Washington University School of Medicine, St. Louis, MO, 3Medicine, Washington University School of Medicine, St. Louis, MO The blood‐brain barrier (BBB) is a formidable cellular layer that maintains an optimal neural microenvironment by restricting the passage of molecules to and from the brain and, consequently, also keeps microorganisms and toxins in the blood from entering the brain. The environmental pathogen Cryptococcus neoformans can, however, breach this barrier, causing a meningoencephalitis that is responsible for over 600,000 deaths annually. C. neoformans enters the body via the lungs and evidence strongly suggests an important role played by phagocytes in dissemination towards the brain, although how they do this is not yet clear. One confounding factor in determining the mechanism is that multiple routes of cryptococcal brain entry have been experimentally supported, including transcellular crossing by sequentially entering and exiting brain endothelial cells; paracellular crossing between these cells; and Trojan horse crossing, as a passenger within host phagocytes. Evidence for these routes of transit has also derived from a variety of studies, including in vitro and in vivo experiments, making direct mechanistic comparisons difficult. Our goal in this study was to examine all the different pathways under the same in vitro conditions with special focus on Trojan horse crossing. By optimizing a flow cytometry strategy to load and isolate singly‐infected phagocytes, we were able to directly compare the transmigration efficiency of free fungi and fungi‐loaded phagocytes across model BBBs, and show that Trojan horse crossing can contribute significantly to such migration. We found that this pathway responds to host physiological signals and also provides a mechanism of invasion for fungal mutants impaired in entry as independent cells. We also used still and video microscopy to demonstrate that C. neoformans remains inside the host cell while it traverses brain endothelia and record additional interactions between infected phagocytes and the brain endothelia that might contribute to brain infection. Our work will allow further dissection of the mechanisms used by this versatile pathogen to invade the brain, and offers experimental approaches that can be applied to other neuropathogens.

E26 Oxidative stress polarizes junctions and the cytoskeleton to drive embryonic wound healing. M.V. Hunter1,2, R. Fernandez‐Gonzalez1,2,3,4; 1Cell and Systems Biology, University of Toronto, Toronto, ON, 2Ted Rogers Centre for Heart Research, University of Toronto, Toronto, ON, 3Institute of Biomaterials and Biomedical Engineering, University of Toronto, Toronto, ON, 4Developmental and Stem Cell Biology Program, Hospital for Sick Children, Toronto, ON Embryos display an outstanding ability to rapidly heal wounds, without inflammation or scarring. Embryonic wound repair is a conserved process, driven by cytoskeletal and junctional rearrangements in the cells around the wound. Immediately after wounding, both F‐actin and the molecular motor non‐muscle myosin II are polarized in the cells adjacent to the wound. Actin and myosin accumulate at the interface with the wounded cells, forming a supracellular cable around the wound whose contraction acts as a “purse string” to draw the wound closed. We recently showed that polarized endocytosis of the junctional protein E‐cadherin from the wound margin is a prerequisite for actomyosin purse string


assembly and rapid wound healing. However, the upstream signal that triggers polarized endocytosis around the wound is unknown. We used quantitative time‐lapse microscopy in Drosophila embryos and measured a burst of reactive oxygen species (ROS) production immediately after wounding, exclusively in the cells that are eventually enclosed by the purse string. Blocking ROS production severely impaired junctional redistribution and actomyosin accumulation around the wound, and significantly inhibited wound healing. Using fluorescence recovery after photobleaching, we found that ROS are necessary for E‐cadherin turnover at cell‐cell junctions. Our results suggest that ROS production and diffusion trigger rapid wound healing, possibly by promoting E‐cadherin trafficking and subsequent purse string formation. Consistent with this, we found that increasing ROS production resulted in larger wounds. ROS can post‐translationally modify proteins through oxidation of electron‐rich cysteine residues. We have identified several redox‐sensitive proteins that are required for wound healing and that may be activated by ROS to trigger the wound response, including the kinases Src, JNK, and p38. Pharmacological inhibition of the three kinases resulted in severe wound closure phenotypes. We are currently investigating the mechanisms by which ROS may activate Src, JNK, and p38, as well as how these proteins promote wound closure. We propose that ROS act as a critical wound signal, orchestrating redistribution of junctions and the cytoskeleton to drive rapid wound healing.

E27 An optimized image analysis‐based approach to quantification of liver pathology. J. Abad1, X. Liang2, M. Calvert1; 1Histology and Light Microscopy Core, Gladstone Institutes, San Francisco, CA, 2Bioengineering, UCSF, San Francisco, CA Fatty liver disease is highly prevalent among middle‐aged Americans, with over 3 million cases annually in the U.S. alone. To date, there is no known cure. Patients must therefore rely on changes in diet and exercise in order to reduce symptoms such as fatigue and abdominal pain. Advanced liver pathology studies are thus essential for developing effective fatty liver disease treatments – especially for patients concurrently diagnosed with type 2 diabetes who are at higher risk for progression into cirrhosis. ImageJ is a widely‐known, free image analysis software capable of measuring a wide array of parameters. However, specific protocols for multi‐parametric Periodic‐Acid Schiff (PAS) and Oil Red O (ORO) analysis of liver tissue are thus far lacking. Using these stains, we sought to develop an efficient method for quantification of glycogen and lipid accumulation that can be easily utilized in future liver pathology experiments. For this study, mouse liver cryosections were stained with either PAS or ORO and digital brightfield color images were acquired. For PAS samples, quantification of glycogen levels was achieved through intensity‐based thresholding. For ORO samples, lipid accumulation was calculated using area‐based measurements. A color‐calibration system was used to maintain consistency between samples. As a result, we successfully developed an efficient, inexpensive analysis protocol for scoring liver pathology based on the quantification of both PAS and ORO‐stained tissues. For PAS samples, we found that utilizing the default thresholding algorithm was sufficient for glycogen quantification. For the ORO samples, we created a hue‐based protocol that is more advantageous than current 8‐bit binary methods due to its compatibility with counterstaining. This protocol can thus be utilized for multi‐parametric data acquisition unlike previously published methods. Our newly‐developed ImageJ analysis protocol can be easily applied to future liver pathology studies in order to obtain quantitative liver disease data in a timely, cost‐efficient, and reproducible manner.


Microsymposium 4: Organelles

E28 A novel regulator of mitochondrial DNA copy number identified by an unbiased genetic screen. A. Göke1,2, C. Osman1,2, R.E. Diaz1,2,3, P. Walter1,2; 1Howard Hughes Medical Institute, San Francisco, CA, 2Department of Biochemistry and Biophysics, University of California at San Francisco, San Francisco, CA, 3Undergraduate Neuroscience Program, University of Miami, Miami, FL Mitochondria contain their own genome (mtDNA), which is present in multiple copies in every cell. mtDNA encodes proteins essential for oxidative phosphorylation and respiratory growth, and mtDNA copy number needs to be maintained at adequate levels to meet cellular demands. The cellular mechanisms that regulate mtDNA copy number, however, remain poorly understood. We developed a genetic screen in the budding yeast S. cerevisiae to identify cellular pathways that regulate mtDNA levels. Through colony blot hybridization, we quantified and compared mtDNA levels normalized to nuclear DNA of the mutants in the yeast deletion library and comprehensively identified genes that positively or negatively affect mtDNA levels. This screen revealed an uncharacterized mitochondrial protein, whose deletion results in a robust doubling of mtDNA levels. The identified protein forms a complex with a conserved component engaged in mitochondrial protein quality control, and super‐resolution microscopy analysis supports a role of it in the segregation of mitochondrial chromosomes. These results provide the first mechanistic link between mitochondrial protein quality control and mtDNA maintenance, suggesting a multilayered system for mtDNA copy number control. Elucidating these molecular mechanisms promise to advance our understanding of pathologies arising from abnormalities in mtDNA copy number and further the development of new therapeutic tools.

E29 Membrane‐associated localization and interactomic study of primary cilia. P. Kohli1,2, M. Rinschen1,2,3, M. Hoehne1,2,3, T. Benzing1,2,3, B. Schermer1,2,3; 1Department II of Internal Medicine and Center for Molecular Medicine, University of Cologne, Cologne, Germany, 2Cologne Excellence Cluster on Cellular Stress Responses in Aging‐Associated Diseases, University of Cologne, Cologne, Germany, 3Systems Biology of Ageing Cologne, University of Cologne, Cologne, Germany Cilia have recently emerged as sensory organelles present in many mammalian tissues involved in a wide variety of cellular processes such as fluid flow, cell movement as well as multiple signaling pathways. Over the last years an increasing number of human diseases have been linked to defects in cilia with autosomal‐dominant polycystic kidney disease (ADPKD) perhaps being the most prominent example. However, very little is known about the protein composition as well as the time‐resolved dynamic processes within this important organelle. To characterize the ciliary membrane‐associated proteome, we used the genetically‐engineered enzyme ascorbate peroxidase (APEX) and targeted it to the ciliary membrane (cmAPEX). With STED microscopy, we could confirm the subciliary localization of the tag specifically at the membrane of the cilium. Upon sequential addition of biotin phenol and H2O2 to the cells, APEX generates free biotin radicals that label proteins within the closest proximity, in particular at the ciliary membrane. Subsequently, cells are lysed, and biotinylated proteins are enriched with streptavidin beads and identified by mass spectrometry. We identified numerous known ciliary proteins, in addition to novel ciliary membrane proteins and several actin modifying proteins. Inhibiting these actin‐modulating proteins regulated ciliary length, reaffirming the importance of actin dynamics in


ciliogenesis. The results also identified ciliary proteins that are differentially regulated upon inhibition of actin polymerization and functional assays characterized their effect on ciliogenesis. cmAPEX is a versatile tool to study perturbations in this small and dynamic organelle. This approach provides quantitative analysis of the proteomic composition of the cilium and at the same time eliminates the need to isolate ciliary factions. We believe that this work will provide further insights into the cilium biology and suggest potential therapeutic targets for cilia‐associated genetic disorders.

E30 Molecular motors and cytoskeleton dynamics control the maturation and secretion of a lysosome‐related organelle. L. Ripoll1, X. Heiligenstein1, L. Domingues1, I. Hurbain1,2, M.K. Dennis3,4, M.S. Marks3,4, E. Coudrier5, G. Raposo1,2, C. Delevoye1,2; 1Structure and Membrane Compartments, Institut Curie, PSL Research University, CNRS, UMR144, Paris, France, 2Cell and Tissue Imaging Facility (PICTIBiSA), Institut Curie, PSL Research University, CNRS, UMR144, Paris, France, 3Department of Pathology and Laboratory Medicine, Children Hospital of Philadelphia, Philadelphia, PA, 4Department of Pathology and Laboratory Medicine and Department of Physiology, University of Pennsylvania, Philadelphia, PA, 5Molecular Mechanisms of Intracellular Transport, Institut Curie, PSL Research University, CNRS, UMR144, Paris, France Skin melanocytes are specialized cell types that generate lysosome‐related organelles, called melanosomes, in which melanin pigments are synthesized and stored. In epidermis, pigmented melanosomes are then transferred to neighboring keratinocytes allowing skin pigmentation and photoprotection against solar radiations. While early events leading to the biogenesis of melanosomes are extensively studied, the trafficking pathways and molecular mechanisms required for the maturation and transfer of melanosomes are still poorly understood. We recently showed that melanosome maturation requires the recycling of a melanosome‐associated SNARE, VAMP7, in tubular carriers that emerge from melanosomes (Dennis et al., J Cell Biol, 2016). Here we show that a myosin actin‐based motor contributes to the sorting and trafficking of those cargoes out of melanosomes. Using cutting‐edge microscopy approaches (correlative light and electron microscopy and electron tomography), we find that impairing the function of myosin or actin dynamics generates long‐lasting tubules that emerge from melanosomes but are not released. Our data highlight that myosin, actin and associated machineries cooperate to remodel melanosomal membranes and release transport intermediates. Importantly, this novel recycling route is functionally required for melanosome secretion and subsequent transfer of melanin to keratinocytes. We are now investigating whether such a pathway might be the target of diseases characterized by hypopigmentation.

E31 Actin‐induced calcium dynamics trigger key events in mitochondrial division. R. CHAKRABARTI1, W. Ji1, H.N. Higgs1; 1Biochemistry, Geisel School of Medicine, Dartmouth College, Hanover, NH Mitochondria are highly dynamic organelles, capable of migration, fusion and division. Division is necessary to create new mitochondria, but it also contributes to quality control by enabling the removal of damaged mitochondria, as well as participating in apoptosis during cellular stress. A key protein to mitochondrial division is the dynamin family GTPase Drp1, which oligomerizes at the division site and constricts the outer mitochondrial membrane. Drp1 recruitment and mitochondrial division are significantly stimulated by close contact between mitochondria and endoplasmic reticulum (ER), in a


process called ERMD. We previously reported that actin polymerization through the ER‐bound formin, INF2, contributes to ERMD in mammalian cells (Korobova et al (2013) Science). We also showed that ionomycin, a calcium ionophore, stimulates Drp1 oligomerization and mitochondrial division through a transient “burst” of INF2‐mediated actin polymerization, with a t1/2 of 28 sec and a return to baseline within 200 sec. Inhibiting actin polymerization, myosin IIA, or INF2 reduces both ionomycin‐induced Drp1 accumulation and mitochondrial fission (Ji et al (2015) eLife). Here, we investigate the sequence of events in ionomycin‐induced division more precisely, focusing on the role of calcium. Similar to previous studies (Abramov & Duchen, 2003), ionomycin stimulation causes a rapid and reversible rise in both cytosolic and mitochondrial calcium ions, with t1/2 of 10 sec and 25 sec respectively and both returning to baseline within 250 sec. Chelation of extracellular calcium eliminates the cytoplasmic and mitochondrial calcium changes, as well as the actin burst and Drp1 accumulation. The mitochondrial calcium increase occurs after the actin burst, and is dependent on actin polymerization through ER‐bound INF2, as well as myosin IIA. We show that this actin‐dependent mitochondrial calcium increase also requires the MCU calcium channel. Finally, we show that this calcium increase is required for mitochondrial division, but is not a component of the actin‐stimulated Drp1 oligomerization response. In total, we show that actin polymerization at the ER‐mitochondrial interface induces two independent responses that stimulate mitochondrial fission: an increase in Drp1 oligomerization, and an increase in mitochondrial calcium ion.

E32 Heterogeneous mechanisms of primary cilia disassembly favor high speed dynamics. M. Mirvis1, W.J. Nelson1,2; 1Molecular and Cellular Physiology, Stanford University, Stanford, CA, 2Biology, Stanford University, Stanford, CA Primary cilia are ubiquitous signaling and sensory organelles that are assembled in quiescent cells and disassembled upon cell cycle re‐entry prior to mitosis. Misregulation of this ciliary cycle is associated with proliferative defects in ciliopathies and cancer. While primary cilia assembly has been well characterized, very few studies have focused on mechanisms of disassembly, particularly in mammalian systems. A common hypothesis from previous work in the Chlamydomonas is that intraflagellar transport (IFT), a ciliary transport complex, mediates export of axoneme tubulin from the ciliary compartment, driving ciliary disassembly by resorption into the cell body. To test this, and other hypotheses, mouse IMCD3 cells expressing SSTR3‐GFP marking the ciliary membrane and PACT‐mCherry marking the basal body were subjected to serum starvation and subsequent stimulation in order to synchronize ciliary disassembly. Cilia were imaged by immunofluorescence and high resolution confocal live imaging, revealing multiple, distinct mechanisms for disassembly, with dynamics ranging from seconds to several hours. Surprisingly, the most prevalent method of cilia disassembly (~70% of cases) was a super‐fast behavior occurring on the order of seconds (2.8 +/‐ 1.3 um/min), which is several orders of magnitude faster than would be expected for an IFT‐based mechanism (hours). Trapping of cellular debris from cell culture media following cilia disassembly revealed the presence of SSTR3‐GFP–positive, but PACT‐mCherry‐negative structures consistent with the size and shape of cilia, indicating that this high‐speed mechanism of cilia disassembly is due to severing and/or shedding of cilia. These results offer new insight into a sequence of events during primary cilia disassembly, and show that that disassembly is an unexpectedly complex, heterogeneous, and stochastic process that favors a very fast mechanism.


E33 Endosome‐mitochondria interactions are modulated by iron release from transferrin. A. Das1, M.M. Barroso1; 1Molecular Cellular Physiology, Albany Medical College, Albany, NY Transient “kiss and run" interactions between endosomes containing iron‐bound‐transferrin (Tf) and mitochondria have been shown to facilitate direct iron transfer in erythroid cells. For the first time using super‐resolution 3D dSTORM imaging, we have furnished compelling evidence of Tf‐endosomes directly interacting with mitochondria in MDCK epithelial cells. Employing live‐cell time‐lapse fluorescence microscopy followed by 3D rendering, object tracking, and distance transformation (DT) algorithm, we have tracked individual Tf‐endosomes and characterized the dynamics of their interaction with mitochondria. Subsequently, the direct interaction‐mediated functional iron transfer from Tf‐endosome to mitochondria was confirmed by the quenching of a fluorescent iron sensor labelled mitochondria. The motility of Tf‐endosomes was found to significantly reduce upon interaction with mitochondria. To further assess the functional role of iron in the ability of Tf‐endosomes to interact with mitochondria, we blocked endosomal iron release by using a Tf K206E/K534A mutant. Blocking intra‐endosomal iron release led to significantly increased motility of Tf‐endosomes and increased duration of endosome‐mitochondria interactions. Utilizing different fluorescence microscopic techniques coupled with in‐depth, semi‐automated, and unbiased image quantitation protocols and minimal perturbation of the cellular physiology, we have provided visual evidence and detailed characterization of Tf‐endosome‐mitochondria direct interactions in non‐erythroid cells, thus enriching the field of organelle‐organelle interactions. Our findings indicate a role of intra‐endosomal iron in regulating the kinetics of the interactions between Tf‐endosomes and mitochondria in epithelial cells. Ongoing work includes investigation and characterization of endosome‐mitochondria interactions in human non‐erythroid cell lines.

E34 Illuminating mitophagy in living mt‐Keima mouse tissues via STED super‐resolution microscopy. D. Malide1, N. Sun2, T. Finkel2; 1Light microscopy core, NIH\NHLBI, Bethesda, MD, 2Center for molecular medicine, NIH\NHLBI, Bethesda, MD Alterations in mitophagy have been increasingly linked to aging and age‐related diseases. There are, however, no convenient methods to analyze mitochondrial turnover and mitophagy in vivo. To this end we recently reported (Sun et al, Molecular Cell, 60, 2015) a transgenic mouse model in which we expressed a mitochondrial targeted form of the fluorescent reporter Keima (mt‐Keima). Keima is a coral‐derived protein that exhibits both pH‐dependent excitation and resistance to lysosomal proteases. This reporter mouse containing the fluorescent reporter mitochondrial targeted mt‐Keima allows for the in vivo assessment of mitophagy in various tissues under normal conditions using confocal microscopy. Furthermore on another scale of resolution we demonstrated that mt‐Keima is suitable for imaging via super‐resolution STED microscopy allowing observation of biological phenomena never before seen. Fluorescence nanoscopy, or super‐resolution microscopy has become an important tool for cell biologists with several methods achieving sub‐100‐nm resolution by taking advantage of reversible photo‐physical switching properties of fluorescent markers. Thus using the mt‐Keima mouse, we can assess at nanoscale resolution in living tissues including skeletal muscle, heart, liver, adipose tissue, and kidney, how tissues mitophagy is altered following changes in diet, oxygen availability, genetic perturbations, or aging. Dual and multi‐color experiments showed 3D characterization of intracellular


compartments: lysosomes, endoplasmic reticulum, nucleus, plasma membrane, lipid droplets. Moreover, we also achieved STED imaging using resonant‐scanning fast‐time acquisition. In conclusion we demonstrated that mt‐Keima mouse is a versatile tool for investigating mitophagy via fluorescence nanoscopy at 50‐nm resolution.

E35 Reeling in the cilium: chytrid fungi as a model for ciliary growth and resorption. K. Vasudevan1, C.M. Baumer1, L. Tracy1, E. Turk1, J.E. Stajich2, T. Stearns3; 1Department of Biology, Stanford University, Stanford, CA, 2Department of Plant Pathology Microbiology and Center for Plant Cell Biology, Institute for Integrative Genome Biology, University of California‐Riverside, Riverside, CA, 3Departments of Biology and Genetics, Stanford University, Stanford, CA Cilia are highly conserved microtubule‐based structures that arise from the cell surface and are required for a variety of functions including development, cell motility and signaling. In general, cilia are resorbed or excised before mitosis, and regrow after cell division is complete. Resorption in mammals is controlled in part by Aurora‐A kinase, HDAC6 and heat shock, but little is known about how it occurs, or how it is regulated. We have chosen to address this question using a simple model system, a chytrid fungus, that has particularly striking features with respect to ciliary resorption. Chytrids are basal fungi that, unlike higher fungi, have retained centrioles and cilia. Chytrids are a diverse group, including both plant and animal pathogens. Many chytrid species have a simple life cycle, alternating between a zoospore that swims by means of a single posterior cilium, and a sporangium, in which mitotic divisions generate ~100 zoospores. We chose one species of chytrid, Rhizoclosmatium globosum (Rg) for study, based on its useful properties in the lab. Rg grows on simple agar medium at room temperature, with an effective doubling time of about 3 hours. The zoospores of Rg are about 3 µm in diameter and have a single, long posterior cilium (~25 µm) with a pair of centrioles at its base. Plating zoospores on solid medium causes them to rapidly resorb their cilium. Using a combination of live‐imaging, immunofluorescence and electron microscopy, we found that the axoneme is “reeled” into the cell body at a rate of about 0.5 µm/second, accompanied by simultaneous cytoplasmic rotation. This results in the intact axoneme being wrapped about 2.5 times around the periphery of the cell. The ciliary membrane appears to be discarded, as a large vesicle. We also observed disassembly of the axoneme inside the cell body within 30 minutes after resorption. We are working to identify factors responsible for driving the ciliary resorption process and its regulation. The Rg genome sequence was obtained as part of this work, and preliminary analysis indicates that most of the microtubule, cilium and centriole proteins show a high degree of homology with those of humans. We are currently developing genetic and molecular biological techniques for Rg. Our use of this model is predicated on the notion that availability of a genome sequence and modern molecular tools, combined with ready lab culture, is all that is required to exploit the advantages of many understudied organisms. We propose that Rg is an excellent model system to study ciliary resorption, and other centriole and cilium processes.


E36 Automated High‐throughput correlative light and electron microscopy. A.M. Steyer1, J.M. Serra Lleti1, N.L. Schieber1, C. Tischer1,2, V. Hilsenstein1,2, B. Neumann1,2, R. Kirmse3, D. Unrau4, J.M. Lucocq5, R. Pepperkok1,2, Y. Schwab1; 1Cellbiology Biophysics, EMBL , Heidelberg, Germany, 2Advanced Light Microscopy Facility, EMBL , Heidelberg, Germany, 3Carl Zeiss Microscopy GmbH, Jena, Germany, 4Fibics, Ottawa, Canada, 5University of St. Andrews, Saint Andrews, United Kingdom The need for quantitative analysis is crucial when studying fundamental mechanisms in cell biology. Common assays consisting in interfering with a system via protein knockdowns or drug treatments very often exhibit significant response variability that is generally addressed by analyzing large populations. Whilst the throughput of light microscopy imaging is high enough for such large screens, electron microscopy still lags behind and is not adapted to collect many data from highly heterogeneous cell populations. Nevertheless, EM is the only technique offering high resolution imaging of the entire subcellular context. Investigation of rare events or heterogeneous populations has been made possible by combining light and electron microscopy (CLEM). Our goal is to develop new strategies in CLEM. More specifically we aim at automatizing the processes of screening large cell populations (living cells or pre‐fixed) by identifying the sub‐populations of interest, targeting these by EM and measuring the key components of the subcellular organization. New 3D‐EM techniques such as focused ion beam ‐ scanning electron microscopy (FIB‐SEM) enable a high degree of automation for the acquisition of high‐resolution, full‐cell datasets. So far this has only been applied to individual target volumes, often isotropic and has not been designed to acquire multiple regions of interest. The ability to acquire full cells at a voxel size of 5 nm x 5 nm x 5 nm (x, y referring to pixel size, z referring to slice thickness), leads to the accumulation of large datasets. Their analysis involves tedious manual segmentation or ad‐hoc development of automated segmentation algorithms. To enable the analysis and quantification of increased numbers of cells, we decided to explore the potential of sampling based stereology protocols in combination with the automated acquisition in the FIB‐SEM. Instead of isotropic datasets, a few evenly spaced sections are used to quantify subcellular structures. Our strategy therefore combines CLEM, 3D‐EM and stereology to collect and analyze large numbers of cells selected based on their phenotype as visible by fluorescence microscopy. We demonstrate the power of the approach in a systematic screen of the Golgi apparatus morphology upon alteration of the expression of 12 proteins.

Microsymposium 5: Chromatin and Intranuclear Organization

E37 Chromatin organization and replication in mature striated muscle rely on mechanical coupling through intact LINC‐complex. S. Wang1, U. C. P.1, K. Fridman1, B. Markus2, T. Volk1; 1Molecular Genetics, Weizmann Institute of Science, Rehovot, Israel, 2G‐INCPM / Mantoux Institute for Bioinformatics, Weizmann Institute of Science, Rehovot, Israel Mature skeletal muscle fibers are multinucleated cells in which the numerous nuclei display relative equal size. Abnormal nuclear size and shape are characteristic of muscular disorders, however the molecular link between aberrant nuclear size and muscle function remains unclear. Recently we found that nuclear mechanical constraints coordinate endoreplication events in the terminally differentiated Drosophila larval muscle cells. LINC‐complex mutants with variable nuclear size were analyzed by high‐resolution light and electron microscopy in combination with genomic analysis of RNA Polymerase II


profiling (using a targeted DamID approach). We found that the perturbation of mechanical coupling between cytoskeletal stress through LINC‐complex not only modifies the nuclear architecture, but also actively determines epigenetic modifications which lead to structural change of chromosomal architecture, DNA‐replication arrests as well as expression reduction of many essential genes involved in muscle structure and function. Therefore, our findings may facilitate understanding the molecular link between muscular disorders and nuclear mechanics in human body.

E38 Reversible liquid droplet formation by HP1α suggests role for phase separation in heterochromatin function. A. Larson1, D. Elnatan1, M.J. Trnka1, J.B. Johnston 1, A. Burlingame1, G.J. Narlikar1; 1Biochemistry, University of California San Francisco, SAN FRANCISCO, CA The segregation of genetic material into actively transcribed (euchromatin) and heritably repressed (heterochromatin) domains is essential for the maintenance of cell identity. In most cell types, the euchromatic and heterochromatic domains often occupy distinct nuclear territories. HP1 proteins are core constituents of late replicating and gene‐poor heterochromatin that is marked by histone H3 lysine 9 methylation (H3K9me). The gene silencing function of heterochromatin is proposed to arise in part from the ability of HP1 proteins to promote the physical compaction of chromatin and recruit repressive activities. These HP1 proteins have a stereotypical domain architecture consisting of a Chromodomain that binds the histone H3 lysine 9 methyl (H3K9me) mark, a Chromoshadow domain that dimerizes and provides an interface for recruiting diverse ligand proteins, a hinge region connecting the two, as well as unstructured N‐and C‐ termini. Here we identify a new property of the major human HP1 protein that is dependent on the unstructured HP1 regions: the ability to form phase‐separated droplets. We find that while unmodified HP1α is soluble and primarily exists as a dimer, phosphorylation of the N‐terminal extension (NTE) of HP1α drives higher‐order oligomerization and the formation of phase‐separated droplets. Oligomerization correlates with a switch from a compact conformation of the HP1α dimer to an extended conformation. The phase‐separated droplets can be rapidly dissolved by phosphatase action and their formation can be promoted or inhibited by specific HP1α ligands including the lamin B Receptor and shugoshin protein. Nucleosomes and DNA are included in these droplets but a non‐interacting protein is excluded. This specificity for chromatin suggests the possibility for some degree of organizational capabilities of phase separated domains in the nuclear context. These findings suggest a new way to conceptualize heterochromatin in which gene silencing is enabled by sequestration of chromatin in phase‐separated compartments that can be dissolved or formed by specific ligands based on their nuclear environment. We propose that this physical property helps explain: (i) why HP1α has different biological roles than its highly conserved paralogs HP1β and HP1γ, (ii) how heterochromatin sequesters large regions of the genome from the transcription machinery.


E39 A novel role for the essential DNA topoisomerase, Top2, in nuclear morphology and chromatin organization. R.A. Meseroll1, N. Holay1, C. Caridi2, J. Bachant2, O. Cohen‐Fix1; 1National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK), National Institutes of Health, Bethesda, MD, 2Graduate Program in Cell and Molecular Biology, University of California, Riverside, Riverside, CA It has long been observed that nuclei of a given cell type have a stereotypical size and shape. Abnormal nuclear morphology arises in many cancers and is a cellular hallmark of the premature aging syndrome progeria. Yet, the mechanisms governing nuclear geometry and the relationship between nuclear structure and function are not well understood. We are using the budding yeast, Saccharomyces cerevisiae, as a model for investigating the pathways that govern nuclear morphology. The budding yeast nucleus is round during interphase, with a crescent‐shaped nucleolus at the nuclear periphery. To identify proteins that affect nuclear morphology we generated a temperature‐sensitive (ts) mutant collection and screened it for mutants that exhibited abnormal nuclear shape at the non‐permissive temperature. One of the identified mutants encodes for a point mutation in the Type II topoisomerase Top2. Top2 catalyzes a conserved strand passage reaction that plays an essential role in resolving entanglements between replicated chromosomes as well as in other, less defined, aspects of DNA topology. Unlike previously identified nuclear abnormalities that form under conditions such as mitotic arrest and are confined to the nuclear envelope adjacent to the nucleolus, more than half of the nuclei in this top2 mutant (top2‐mf) form multiple extensions away from the nucleolus. Examining a panel of historical and newly‐generated top2 mutants, and depleting Top2 levels via a Top2 degron, revealed that the penetrance of the nuclear morphology defect varies widely among mutants, with top2‐mf being the most penetrant, and suggests the TOPRIM and DNA‐binding domains of Top2 contribute to nuclear shape. Furthermore, the multiflare phenotype does not require passage through the cell cycle, as top2‐mf cells shifted to the non‐permissive temperature during an S phase arrest exhibit the same nuclear morphology defect as temperature‐shifted asynchronous cells. Notably, chromatin delocalizes from the nuclear periphery in the vast majority of top2‐mf nuclear extensions, suggesting chromatin structure is highly aberrant in this mutant. This is further supported by diffuse staining with the minor‐groove intercalator DAPI and elevated DNA double‐strand breakage in the top2‐mf mutant. We hypothesize that nuclear morphology relies on chromatin structure and that defective chromatin conformation in the top2‐mf mutant—in particular the abnormal detachment of chromatin from the nuclear envelope—leads to nuclear shape abnormalities. We are currently testing the role of chromatin structure in top2 and other multiflare mutants from our screen, as well as assessing topoisomerase activity in the top2‐mf mutant to determine how Top2 functions in nuclear morphology.

E40 Super‐resolution microscopy reveals the three‐dimensional organization of meiotic chromosome axes in intact C. elegans tissue. M. Wojcik1, S. Köhler2, A.F. Dernburg2, K. Xu1; 1Chemistry, UC Berkeley, Berkeley, CA, 2Molecular and Cell Biology, UC Berkeley, Berkeley, CA During the beginning of meiosis, chromosomes reorganize into linear arrays of densely packed DNA loops around a proteinaceous central axis. Meiotic chromosome axes are the platform for the assembly of the synaptonemal complex, which bridges paired homologous chromosomes, and they are required for double‐strand break and crossover formation by inter‐homolog recombination. Genetic experiments


have revealed that proteins of the chromosome axes hold a central role in all key aspects of meiosis. Despite this fact, little is known about the three‐dimensional organization of the many components within the chromosome axes, which consist of meiosis‐specific proteins and cohesin complexes. This knowledge gap can be attributed to the sub‐diffraction scale of synapsed homologous chromosomes in meiosis, with synapsed axes being separated by only 100 nm, which is below the resolution limit of conventional fluorescence microscopy. Super‐resolution microscopy has emerged as a powerful tool that has pushed the resolution of fluorescence microscopy to tens of nanometers. We here report the first super‐resolution study of the molecular organization of meiotic chromosome axes in whole, intact C. elegans gonads. Using a fluorescent protein tagged to one component (HIM‐3) of the paired complex alongside antibodies recognizing epitope tags inserted at distinct sites in other components of the complex by CRISPR/Cas9 genome editing, we perform 2‐color combined PALM and STORM microscopies on synapsed meiotic chromosome axes. Utilizing HIM‐3 as a biologically faithful internal standard to reference the positions of all other components of the chromosome axes, we perform extensive three‐dimensional averaging across multiple synapsed axes to determine the 3D organization of axial components to a precision of 2‐5 nanometers. We demonstrate that the two major components of the axis, meiosis specific HORMA‐domain proteins and DNA‐organizing cohesin complexes, are each clustered together at distinct regions of the chromosome axes, with HORMA‐domain proteins falling around 55 nm off‐center and cohesin complexes falling around 70 nm off center. Our 3D data further shows that the cohesin complex, which spans around 60 nm in diameter, is not oriented randomly, but falls in a defined orientation relative to the chromosome axis. We confirm the dependence of this orientation on REC‐8, a critical component of the cohesin complex, by mutagenic deletion of rec‐8, resulting in a loss of defined cohesin‐complex orientation.

E41 The p150N domain of Chromatin Assembly Factor‐1 regulates chromosome organization throughout the cell cycle via Ki‐67. T.D. Matheson1, P.D. Kaufman1; 1Molecular, Cell and Cancer Biology, U. Massachusetts Medical School, Worcester, MA We will present data implicating two proteins, CAF‐1 p150 and Ki‐67, in the coordination of interphase and mitotic chromosome structures. Chromatin Assembly Factor 1 (CAF‐1) is a conserved three‐subunit protein complex which deposits histones onto replicating DNA during S‐phase [1]. The largest subunit of CAF‐1, p150, also contains a functionally separable N‐terminal domain (p150N) that is dispensable for histone deposition [2]. Our earlier studies showed that p150N instead regulates the localization of specific genomic loci (Nucleolar‐Associated Domains, or “NADs” [3] and several proteins to the nucleolus during interphase [4]. One of these p150N‐regulated proteins is the proliferation marker protein Ki‐67. Ki‐67 regulates heterochromatin localization at nucleoli during interphase, and during mitosis is fundamental component of the perichromosomal layer (PCL) [5, 6], a sheath of proteins surrounding condensed mitotic chromosomes [7]. In the PCL, Ki‐67 prevents aggregation of mitotic chromosomes after condensation [8]. After mitotic exit, Ki‐67 transiently occurs in small foci coincident with repetitive DNAs; these foci then accumulate at the reforming nucleolus in G1 phase [9]. Thus, Ki‐67 links interphase and mitotic genome organization. Here, we will present new data demonstrating that p150N coordinates the three‐dimensional localization of Ki‐67 not only during interphase but also on mitotic chromosomes. We show that during mitosis, a subset of p150 localizes to the PCL, and also to the small Ki‐67 foci in early G1 cells. Depletion experiments indicate that p150N is required for normal levels of Ki‐67 accumulation on the PCL. This activity requires the Sumoylation Interacting Motif (SIM) within p150N, which is the same motif required


for the nucleolar localization of NADs and Ki‐67 during interphase [4]. We will present additional data outlining post‐translational regulation of Ki‐67 by p150N. In sum, a separable domain of a DNA‐replication‐linked chromatin assembly complex governs both interphase and mitotic localization of Ki‐67, a cornerstone of nuclear organization. 1. Smith, S. and B. Stillman, Cell, 1989. 58(1): p. 15‐25. 2. Kaufman, P.D., et al., Cell, 1995. 81(7): p. 1105‐14. 3. Matheson, T.D. and P.D. Kaufman, Chromosoma, 2016. 125(3): p. 361‐71. 4. Smith, C.L., et al., Mol Biol Cell, 2014. 25(18): p. 2866‐81. 5. Booth, D.G., et al., Elife, 2014. 3: e01641. 6. Sobecki, M., et al., Elife, 2016. 5: e13722. 7. Van Hooser, A.A., P. Yuh, and R. Heald, Chromosoma, 2005. 114(6): p. 377‐88. 8. Cuylen, S., et al., Nature, 2016. 535(7611): p. 308‐12. 9. Bridger, J.M., I.R. Kill, and P. Lichter, Chromosome Res, 1998. 6(1): p. 13‐24.

E42 Vertebrate transvection: evidence for chromosome X coalescence in human disease. A.I. Laskowski1, I.M. Konieczna2, J. He2, D.S. Neems1, J.M. Mathew2, J.R. Leventhal2, R. Ramsey‐Goldman3, S.T. Kosak1; 1Department of Cell and Molecular Biology, Feinberg School of Medicine, Northwestern University , Chicago, IL, 2Department of Surgery, Feinberg School of Medicine, Northwestern University , Chicago, IL, 3Department of Medicine – Rheumatology, Feinberg School of Medicine, Northwestern University , Chicago, IL The three‐dimensional organization of the genome within the nucleus is intrinsically linked to its function, in particular gene regulation. Dynamic changes to nuclear organization coincide with dramatic changes in cellular identity such as during differentiation and transformation. Whole chromosomes themselves display organizational patterns based upon gene density and coordinate gene regulation. A prime example of the importance of whole chromosome organization is X inactivation. During early stages of female embryogenesis, differentiating embryonic stem cells silence one X chromosome for gene dosage compensation. X chromosomes coalesce (co‐localize) and Xist, a long non‐coding RNA, is expressed from the X chromosome randomly selected for inactivation. We have observed an interesting phenomenon in which X chromosome coalescence is occurring outside of early development, with varying levels of frequency depending on cell type. A subset of T lymphocytes, regulatory T cells (Tregs), has demonstrated the highest frequency of chromosome X coalescence thus far. We have also observed evidence of transcriptional silencing spreading from the active X (Xa) to the inactive X (Xi) chromosome. Tregs function to suppress autoimmunity and proper levels of Foxp3 expression is required for differentiation and development. FOXP3 is located on the X chromosome. In addition, individuals suffering from active Lupus, an autoimmune disease, have decreased levels of Foxp3+ Tregs. Lupus is nine times more common in women than men, suggesting a genetic predisposition linked to chromosome X diploidy. We hypothesize that higher levels of chromosome X coalescence in female lymphocytes may lead to inappropriate silencing of X linked genes important for lymphocyte function, thereby increasing susceptibility to autoimmunity.


E43 Regulation of Rvb1/Rvb2 by a region within the INO80 chromatin remodeler implicates the yeast Rvbs as protein assembly chaperones. C.Y. Zhou1, C.I. Stoddard1, J.B. Johnston 2, M.J. Trnka2, I. Echeverria3, A. Burlingame2, Y. Cheng1,4, A. Sali2,3, G.J. Narlikar1; 1Biochemistry and Biophysics, UCSF, San Francisco, CA, 2Pharmaceutical Chemistry, UCSF, San Francisco, CA, 3Bioengineering and Therapeutic Sciences, UCSF, San Francisco, CA, 4Howard Hughes Medical Institute, San Francisco, CA Rvb1 and Rvb2 (Rvbs) are well‐conserved AAA+ ATPases that are involved in a diverse range of cellular processes including transcription, cell differentiation, signaling, mitosis, snoRNP assembly, and DNA damage repair. Despite being essential for these processes, the role of the Rvbs in biology is unknown. In vitro, Rvb1 and Rvb2 interact with each other to form hexameric rings. In some cases, a dodecameric species is also observed, though the role of the two oligomeric states is unclear. Due to their stable association with purified protein complexes, including telomerase, the R2TP complex, and the INO80 family of ATP‐dependent chromatin remodeling complexes, it was recently proposed that the Rvbs may act as assembly chaperones for these multi‐subunit complexes. However, evidence for this hypothesis has been lacking. In this study, we provide the first mechanistic insight into how the Rvbs may act as chaperones. We find that a small domain within the ATPase subunit of the INO80 complex (Ino80INS) is sufficient to stably interact with the Rvb1/Rvb2 complex and stimulates the ATPase activity of the Rvbs by 16‐fold, providing the first evidence of a robust activator for these enzymes. Crosslinking‐MS of the Rvb1/Rvb2/Ino80INS complex reveals that the majority of the interactions between Rvbs and the Ino80INS are mediated through the conformationally dynamic DII domains of the Rvbs. Based on native mass spectrometry and SEC‐MALS results, the Rvb1/Rvb2/Ino80INS complex is predominantly dodecameric, consistent with our finding that the DII domains affect higher‐order oligomerization of Rvb1/Rvb2. Using a combination of single particle cryo‐EM and integrative modeling, we find that the Ino80INS likely interacts near the central channel of the Rvb1/Rvb2 dodecamer, partially enclosed by the DII domains that form the dodecameric interface. This structure is reminiscent of other ATP‐dependent chaperones such as GroEL and Hsp100. Strikingly, the addition of ATP to the Rvb1/Rvb2/Ino80INS complex causes the complex to dissociate into hexamers, suggesting that oligomeric state of the Rvbs is coupled to nucleotide state. Taking these results together, we propose that the dodecameric state of Rvb1/Rvb2 represents an activated intermediate toward assembly of multi‐subunit complexes such as INO80. This model offers a potential explanation for the role of the Rvb1/Rvb2 dodecamer, and provides a unifying mechanism for the diverse roles the Rvbs play in the cell.

E44 Fascin regulates nuclear actin during Drosophila oogenesis. D.J. Kelpsch1, C.M. Groen2, T.N. Fagan1, S. Sudhir1, T.L. Tootle1; 1Anatomy and Cell Biology, University of Iowa, Iowa City, IA, 2Regenerative Neurobiology Laboratory, Mayo Clinic, Rochester, MN Over 40 years of controversy surround the nuclear localization of actin. Now, it is well understood that actin localization to the nucleus is highly regulated, suggesting actin has critical nuclear functions. Studies have demonstrated that nuclear actin has roles in regulating transcription, chromatin remodeling, nuclear organization, and DNA damage repair. However, much remains unknown about nuclear actin, including its structure in the nucleus and its functions during development. Here we present our novel finding that Drosophila oogenesis, or follicle development, is a model for studying the


structure, regulation, and function of nuclear actin. Follicles are composed of roughly 1000 somatic follicle cells and 16 germline cells, including 15 nurse or support cells and a single oocyte, and progress through a series of 14 morphological stages, from the germanium to Stage 14. During Stages 5‐9, nuclear actin levels are high in the oocyte and exhibit variation within the nurse cells. Known regulators of actin nucleocytoplasmic transport, Cofilin and Profilin, also localize to the nuclei during follicle development. Further, germline expression of GFP‐tagged Actin results in nuclear actin rod formation in nurse cells during Stages 5‐9. These nuclear actin rods are static, label with both phalloidin and Cofilin. These findings suggest that nuclear actin is tightly regulated during follicle development. As we have previously determined that Fascin, an actin bundling protein, localizes to nurse cell nuclei, we postulated that Fascin may be a novel regulator of nuclear actin. Overexpression of Fascin enhances the frequency and length of GFP‐Actin rods, and Fascin colocalizes with these rods. Additionally, the loss of Fascin reduces, while the overexpression of Fascin increases, the frequency of nurse cells with high levels of endogenous nuclear actin. However, genetic manipulation of Fascin does not alter the overall nuclear level of actin within the ovary. Thus, these data suggest Fascin regulates the ability of specific cells to accumulate nuclear actin. The loss of Fascin reduces nuclear Cofilin suggesting that Fascin positively regulates nuclear actin through Cofilin, a key factor in the nuclear import of actin. Indeed, Fascin and Cofilin genetically interact, as double heterozygotes exhibit a reduction in the number of nurse cells with high nuclear actin levels. In conclusion, we have established Drosophila oogenesis as a new system to uncover the physiologically important functions and regulation of nuclear actin. Further, we have identified Fascin as a novel regulator of nuclear actin. The nuclear localization of Fascin and Actin is highly conserved among systems, indicating these findings are applicable far beyond Drosophila follicle development.

E45 Understanding the role of actin filaments in DNA double‐strand break repair. B.R. Schrank1, T.A. Casado1, J.J. Gautier2; 1 Institute for Cancer Genetics, Columbia University Medical Center, New York, NY, 2Genetics and Development, Columbia University Medical Center, New York, NY Double‐strand breaks (DSBs) are among the most lethal forms of DNA damage, and their repair requires a highly efficient, coordinated effort to defend against genome instability. DSB repair is carried out by the sequential recruitment of an array of kinases and enzymes, which process and repair DSB ends. Recently, we performed a proteomic screen using liquid chromatography‐tandem mass spectrometry to identify novel proteins recruited to damaged chromatin in Xenopus laevis cell‐free extracts. Surprisingly, we uncovered significant enrichment of beta‐actin, actin filament capping proteins, and subunits of the arp2/3 complex in DSB‐containing chromatin. In mammalian cells, we found that inactivation of the arp2/3 complex or the actin filament nucleator N‐WASp, results in marked defects in homologous recombination and single‐strand annealing repair pathways. Notably, we have observed that the migration of DSB foci into discrete sub‐nuclear clusters is abrogated by arp2/3 complex inhibition, and that essential repair events such as Rad51 foci formation and phosphorylated histone H2AX foci resolution are subsequently impaired. Finally, we observed a decrease in the number of surviving clones when PARYlation and actin polymerization are inhibited, consistent with a requirement for both processes in the homologous recombination pathway to resolve endogenous DNA damage. Together, our findings establish actin polymerization as a process required for faithful resolution of double‐stranded DNA lesions.


Microsymposium 06: Cytoskeletal Molecular Dynamics

E46 Nano‐clustering of ligands on synthetic APCs influences T‐cell membrane and actin organization. E. BENARD1, P. Dillard1,2, F. Pi1,3, L. Limozin2, K. Sengupta1; 1STNO, CINaM, Marseille, France, 2LAI, Marseille, France, 3current adress : School of food science of Jiangnan University , Jiangsu , China

The interface between an Antigen Presenting Cell (APC) and a T‐lymphocyte (T cell) plays a key role in the sensitivity and precision of antigen recognition by T‐cells. At the molecular level, this recognition results from the interaction between the antigens on the surface of the APC and the T‐cell receptors (TCR) present on the T‐cell membrane. In in‐vitro experiments the APC is often replace by a fluid supported lipid bilayer (SLB) which is homogeneously functionalized with ligands of TCR, as well as ICAM1 the ligand of T‐cell integrin LFA1.However it has recently been shown that the TCR ligands, on the membrane of APCs are present in the form of submicronic clusters [1]. We have developed patterned substrates to mimic this situation which present an array of sub‐micrometric dots, of ligands against the TCR complex, surrounded either by a sea of polymers (type I) [2,3] or by a SLB (type II) [3]. Using type I substrates, we showed that global cell parameters like adhesion area (quantified from reflection Interface Contrast Microscopy) depends only on the average ligand density and not on the details of the pattern [4]. However, localy , TCR/ZAP‐70 organization, the actin distribution and membrane topography (measured using RICM and TIRFM) are severely altered on pattern as compared to homogeneously distributed ligands. In the patterned but not in the homogeneous case, the TCR/ZAP‐70 and actin are present in the form of dots. This result is reproduced in the type II substrates. In both cases, the pattern induces a characteristic topography of the adhering cell membrane where the cells adhere mostly on the dots. Interestingly, in experiments realized with type II substrates, the addition of ICAM1on the SLB does not appreciably perturb the organization of TCR or ZAP‐70, which remain in form of organized clusters. However, the actin organization changes to a peripheral ring, resembling the classical actin distribution seen on homogeneous substrates. Additionally, the adhesion area increases and the membrane flattens at the surface. References: [1] Lu X et al, Proc Natl Acad Sci U S A. (2012) [2] Pi, F., et al.,Nano Lett,. 13(7): p. 3372‐8, (2013) [3] Pi F., Dillard P., Alameddine R., Benard E., et al. Nano Lett 15 (8): p. 5178–5184 (2015) [4] Dillard,P et al Integr Biol . 14;8 (3):287‐301 (2016)

E47 Microtubules acquire resistance from mechanical breakage through intralumenal acetylation. Z. Xu1,2,3, M.V. Nachury2, Z. Werb1, P. Marinkovich3, D. Portran2, L. Schaedel4, M. Théry4, J. Gaillard4; 1Anatomy, University of California, San Francisco/UCSF, San Francisco, CA, 2Molecular and Cellular Physiology, Stanford University , School of Medicine, Stanford, CA, 3Dermatology, Stanford University , School of Medicine, Stanford, CA, 4Laboratoire de Physiologie Cellulaire et Végétale, Institut de Recherche en Technologie , Paris, France Microtubules are essential cellular compression‐bearing structural elements in eukaryotic cells. Despite the fact they carry the signature of dynamic instability, a subset of long‐lived microtubules with half‐lives over hours exists to provide mechanical framework for organelle transport, cell migration and


division. Intriguingly, long‐lived microtubules are acetylated inside their lumen on lysine 40 of α‐tubulin (αK40). Yet the causal link between αK40 acetylation and stability remains unsolved. Here we find that tubulin acetylation is required for the continued presence of long‐lived microtubules in mammalian cells. Surprisingly, it is not dynamic stability but rather mechanical resistance that is compromised in the absence of αK40 acetylation. First, cells lacking of αK40 Tubulin AcetylTransferase (TAT1) show a significant increase in the frequency of microtubule breakage. Second, long‐lived microtubules lost upon depletion of αK40 acetylation can largely be restored by either pharmacological or physical removal of compressive forces. Finally, using a microfluidics system that recapitulates microtubule breakage in vitro , we find that αK40 acetylation protects microtubules from mechanical breakage. Since acetylation concentrates in curved segments of long‐lived microtubules, we propose an adaptive model of microtubule mechanical stability where TAT1 enters into the microtubule lumen through lattice fenestration or stress‐induced breathing events and localized acetylation increases the mechanical resilience of microtubules to ensures the persistence of long‐lived microtubules in the face of pervasive mechanical stresses.

E48 Mechanosensitive inhibition of formin facilitates contractile ring assembly in fission yeast. D. Zimmermann1, K.E. Homa1, G.M. Hocky2, L.W. Pollard3, G.A. Voth2, M.J. Lord3, K.M. Trybus3, D.R. Kovar1,4; 1Molecular Genetics and Cell Biology, University of Chicago, Chicago, IL, 2Department of Chemistry, James Franck Inst. for Biophysical Dynamics and Computation Inst., University of Chicago, Chicago, IL, 3Molecular Physiology and Biophysics, University of Vermont, Burlington, VT, 4Biochemistry and Molecular Biology, University of Chicago, Chicago, IL Cytokinesis is a highly conserved process where through the formation of a contractile actomyosin ring cells physically separate into two. A popular model system to seek mechanistic insights about cytokinesis is the fission yeast Schizosaccharomyces pombe, where the contractile ring has been hypothesized to assemble via the Search‐Capture‐Pull‐Release model from membrane‐bound ring precursors that contain the type‐II myosin Myo2 and the formin Cdc12. However, the underlying molecular mechanisms by which formin Cdc12 and myosin Myo2 collaborate to drive effective ring assembly remain largely elusive. By reconstituting Search‐Capture‐Pull from purified components in vitro, we discovered that the contractile ring formin Cdc12 acts as mechanosensor whose ability to elongate actin filaments becomes reversibly suppressed once the ring myosin Myo2 exerts sub‐piconewton pulling forces on the formin‐bound actin filament. Furthermore, using mathematical modeling and live cell imaging approaches we found that the mechanosensitive behavior of formin Cdc12 is required to ensure proper ring assembly from cytokinetic ring precursors in vivo.

E49 Profilin directly regulates microtubule dynamics and microtubule‐actin crosstalk using residues mutated in ALS. J.L. Henty‐Ridilla1, B.L. Goode1; 1Biological Sciences, Brandeis University, Waltham, MA Cells require tight coordination of microtubule (MT) and actin dynamics for directed cell migration, cell adhesion, cell division, and other fundamental processes. Most of the proteins known to function at the MT‐actin interface (e.g., CLIP‐170, EB1, Formin, Tau, Spectraplakin, IQGAP, and many more) interact directly or indirectly through higher‐order complexes to bridge each polymer system. Here, we report a


distinct mechanism of MT‐actin dynamics crosstalk that tunes tubulin and actin monomer levels. We show that the actin monomer‐binding protein profilin binds to tubulin and MTs to stimulate MT polymerization. Using TIRF microscopy, we show that purified budding yeast, Drosophila, and human homologs of profilin each increase the elongation rate of MTs 3‐5 fold. In addition, human profilin lowers the critical concentration for tubulin polymerization. These profilin activities were unaffected by previously characterized mutations in profilin that disrupt its canonical interactions with G‐actin and polyproline sequences. However, several point mutations in human profilin linked to Amyotrophic Lateral Sclerosis (ALS) were impaired in promoting MT growth and binding MTs. Consistent with these biochemical results, MT growth rates were increased in neuroblastoma (N2A) cells overexpressing wildtype human profilin but not by the ALS‐mutants. Finally, adding increasing concentrations of monomeric actin abrogated the effect of profilin on MT growth in vitro. Together, these observations suggest a additional cellular role for profilin as a molecular sensor of actin monomer levels that coordinates MT and actin growth dynamics, and a new cellular mechanism underlying neuronal defects in ALS.

E50 The mammalian dynein/dynactin complex is a strong opponent to kinesin in a tug‐of‐war competition. A. Yildiz1; 1Department of Physics, Department of Molecular and Cell Biology, University of California Berkeley, Berkeley, CA Kinesin and dynein motors transport intracellular cargos bidirectionally by pulling them in opposite directions along microtubules, through a process frequently described as a ‘tug of war’. While kinesin produces a 6 pN force, mammalian dynein was found to be a surprisingly weak motor (0.5‐1.5 pN) in vitro, suggesting many dyneins are required to counteract the pull of a single kinesin. Mammalian dynein’s association with dynactin and Bicaudal‐D2 (BICD2) activates its processive motility, but how this affects dynein’s force output remained unknown. Here, we show that formation of the dynein‐dynactin‐BICD2 (DDB) complex increases human dynein’s force production to 4.3 pN. An in vitro tug‐of‐war assay revealed that a single DDB successfully resists a single kinesin. Contrary to previous reports, the clustering of many dyneins is not required to win the tug‐of‐war. Our work reveals the key role of dynactin and a cargo adaptor protein in shifting the balance of forces between dynein and kinesin motors during intracellular transport.

E51 High‐resolution structural characterization of the myosin VI‐F‐actin interface during the force generation cycle. P.S. Gurel1, L.Y. Kim1, T. Omabegho2, Z. Bryant2, G.M. Alushin1; 1Cell Biology and Physiology Center, National Heart, Lung, and Blood Institute, Bethesda, MD, 2Department of Bioengineering and Department of Structural Biology, Stanford University School of Medicine, Stanford, CA The large superfamily of myosin motor proteins is responsible for movement and force generation at multiple scales of biology, ranging from muscle contraction to intracellular transport. While the mechanism of actomyosin has been the subject of extensive biophysical, biochemical, and structural characterization, high‐resolution insight into the myosin‐F‐actin interface has been limited. Using cryo‐electron microscopy (cryo‐EM) and an adapted protocol for Iterative Helical Real‐Space Reconstruction


(IHRSR) which implements refinement and reconstruction of independent‐half datasets to minimize noise bias in resolution estimation and alignment, we present reconstructions of myosin VI in nucleotide‐free (rigor, 4.5 Å average resolution) and MgADP (strong‐bound ADP, 5.5 Å average resolution) states bound to F‐actin, as well as corresponding atomistic models. We observe rearrangements in both the myosin motor domain and actin relative to structures of these components in isolation, suggesting a reciprocal relationship between actin and myosin conformation during force generation. Furthermore, comparison to a recent high‐resolution structure of the myosin IIC‐F‐actin interface demonstrates that while the regions of contact are similar, the specific residues involved and the chemical nature of the interactions differ between the two myosins. This highlights the need to structurally characterize diverse myosins in complex with F‐actin to gain insight into motor‐specific properties. In addition to providing mechanistic insights specific to myosin VI, the reconstruction and modeling approaches we have developed are applicable to visualizing diverse actomyosin complexes in multiple states at near‐atomic resolution.

E52 FUNCTIONS of the TOG domain proteins, FAP256 and CHE‐12, at the ciliary tip. P. Louka1, M. Guha1, K. Vasudevan2, D. Włoga3, W.L. Dentler4, J. Gaertig1; 1Cellular Biology, University of Georgia , Athens, GA, 2Biology, Stanford University , Stanford, CA, 3Cell Biology , Nencki Institute of Experimental Biology , Warszawa, Poland, 4Molecular Biosciences , University of Kansas, Lawrence, KS TOG domain‐containing proteins (e.g. XMAP215) bind to the extreme plus ends of the cytoplasmic microtubules, where they promote tubulin addition and regulate the polymer dynamics. FAP256 and CHE‐12/crescerin are two conserved TOG domain proteins that localize to cilia but their function is not well understood. In particular, it is unclear how the ciliary TOG domain proteins affect the plus ends of microtubules of motile cilia, that are permanently terminated by the capping structures: the distal filaments (DFs) at the ends of the A‐tubules and the central microtubule cap (CMC). We used Tetrahymena thermophila to address the role of the ciliary TOG proteins for cilia assembly and motility. Using SR‐SIM, FAP256‐GFP was detected as a pair of dots near the ciliary tip. The distance between the two dots suggests that FAP256 is associated with the ends of the B‐tubules and the ends of A‐tubules and the central pair microtubules. A knockout of FAP256 reduced the number of cilia per cell (without affecting their length) and decreased the beat frequency. At the ultrastructural level, the loss of FAP256 led to a more frequent loss of the CMC (the DFs remained intact), and a distal shortening of the B‐tubules. GFP‐CHE‐12 localized to the tips of assembling but not steady‐state cilia. Loss of CHE‐12 resulted in shortening of the B‐tubules, but had no effect on the caps. A loss of both FAP256 and CHE‐12 produced a strong synthetic phenotype of slow cell multiplication and motility. The cilia of the double knockout cells had B‐tubules that were shorter than those in any of the single knockout cells, while the overall length of cilia remained unaffected. Consequently, the distal segment of cilia in cells lacking both FAP256 and CHE‐12 is longer as compared to the wild type (1.6 μm versus 0.6 μm in the wild type). We conclude that FAP256 and CHE‐12 are both important for proper organization of the ciliary tip compartment. While FAP256 has a unique function in stabilizing the CMC, both FAP256 and CHE‐12 affect the length of the B‐tubules and in this way regulate the length of the distal segment.


E53 The +TIP, TACC3, is an important regulator of microtubule dynamics, axon outgrowth and guidance. B. Erdogan1, G. Cammarata1, A. Francl1, E. Lee1, B. Pratt1, J. Tiber1, L.A. Lowery1; 1Biology, Boston College, Boston, MA Precise neuronal connection requires proper axon guidance. Microtubules (MTs) of the growth cone are the driving force to navigate the growing ends of axons. Pioneering microtubules and their plus‐end resident proteins, +TIPs, play integrative roles during this navigation. Recently, we introduced the protein TACC3 as a member of the +TIP family regulating microtubule dynamics in Xenopus laevis growth cones and showed manipulation of TACC3 levels affects axon outgrowth by regulating axon outgrowth velocity and the frequency of axon retraction. Here, we examine the impact of the highly conserved domains of TACC3 on MT dynamics regulation and axon outgrowth. We find that deletion of the TACC domain, the domain that ensures plus‐end localization, significantly reduces both MT and axon growth length. Additionally, we show that over expressing TACC3 mitigates Nocodazole‐induced reduction in MT dynamics parameters, such as MT growth speed, length and lifetime. While this mitigation could result from increased MT stability, immunocytochemical analysis of growth cones for stable (de‐tyrosinated tubulin) and dynamic (tyrosinated tubulin) MTs demonstrates that neither TACC3 knockdown nor its overexpression have impact on the levels of dynamic versus stable MTs, suggesting TACC3 antagonizes Nocodazole‐induced reduced MT dynamics by a different mechanism. We had previously shown that TACC3 co‐localizes with its well‐known partner XMAP215 at the extreme plus‐ends of MTs in a co‐dependent manner (Nwagbara et al 2014). Our epistasis analysis demonstrates that TACC3 and XMAP215 cooperate to promote axon outgrowth and rescue axon growth defects. Moreover, we demonstrate that manipulation of TACC3 levels interferes with the growth cone response to the axon guidance cue Slit2 ex vivo. We also show that ablation of TACC3 causes pathfinding defects in axons of developing spinal cord motor neurons and retinal ganglion cells in Xenopus laevis in vivo. Together, our results suggest that by regulating MT behavior, the +TIP TACC3 is involved in axon outgrowth and pathfinding decisions of neurons during embryonic development.

E54 High‐throughput detection of cytoskeletal contributions to cellular viscoelastic properties with a new microfluidic paradigm and analytical framework. L. Guillou1, J.B. Dahl2, J.G. Lin3,4, A.I. Barakat1, J. Husson1, S. Muller2, S. Kumar2,3,4; 1Mechanics, Ecole Polytechnique, Palaiseau, France, 2Chemical and Biomolecular Engineering, UC‐Berkeley, erkeley, CA, 3The UC Berkeley‐UCSF Graduate Program in Bioengineering, Berkeley, CA, 4Bioengineering, UC‐Berkeley, Berkeley, CA The quantification of cellular mechanical properties is of tremendous interest in biology and medicine, yet single‐cell tools are often low‐throughput, labor‐intensive, and require specialized equipment. Microfluidic technologies show enormous promise for overcoming these barriers by rapidly inferring rheological properties from analysis of flow‐induced deformations. Nonetheless, it remains extremely challenging to extract mechanical properties from current microfluidic systems due to channel confinement effects and the use of high strain rates, which introduce complex inertial forces. Moreover, many microfluidic platforms require deformation of cells into highly distended and aphysiologic shapes, clouding the interpretation of the measurements. Here, we address both of these challenges by introducing a simple microfluidic cross‐slot platform featuring low strain rate and channel confinement to modestly elongate single cells near the stagnation point of a planar extensional flow. This simple


cross‐slot platform consists of two orthogonal channels in which we infuse a high‐viscosity cell suspension on opposing ends of the same channel. At the intersection of the two channels, the two opposing fluid flows create planar extensional flow that elongates the cell from a spherical to an elliptical shape. To convert these deformations into viscoelastic properties, we also developed a new and fully analytical theory that enables us to extract cellular stiffness and fluidity while accounting for cross‐slot geometry, fluid viscosity, and applied flow rate. We first validated our theory by characterizing crosslinked dextran hydrogel beads and obtained elastic moduli in close agreement with both our own micropipette aspiration measurements and reported suspension‐based values. We next used our system to measure the stiffness and fluidity of 3T3 fibroblasts and primary glioblastoma tumor initiating cells. For both cell types, our system produced stiffness values in order‐of‐magnitude agreement with atomic force microscopy but at much higher throughput. Importantly, we were also able to detect cellular softening and stiffening in response to cytoskeletal depolymerizing and crosslinking agents, confirming that our approach offers a biologically relevant dynamic range. The ability to detect mechanical effects of cytoskeletal perturbations represents a significant advance relative to past microfluidic cross‐slot systems, which we ascribe to our use of high‐viscosity medium and significantly reduced strain rates. We envision that our approach will prove valuable for the rapid rheological characterization of cells, thereby accelerating fundamental studies of cellular mechanics and establishing a platform for future diagnostic technologies.

Keith R. Porter Lecture

A1 Visualizing microtubule structure and interactions. E. Nogales1,2,3; 1MCB, UC Berkeley, Berkeley, CA, 2Molecular Biophysics and Integrated Bio‐Imaging, Lawrence Berkeley National Laboratory, Berkeley, CA, 3Howard Hughes Medical Institute, Berkeley, CA Cell division is a complex and highly regulated process in which the microtubule (MT) cytoskeleton plays a central role, serving as energy source for dramatic movements requiring complex mechano‐chemistry, as well as acting as a scaffold that facilitates molecular encounters at the right time and place within the cell. Essential for MT function during cell division is dynamic instability, a property inherent to tubulin but that can be both regulated and utilized by a myriad of cellular factors. The structural characterization of microtubules and their interaction with their associated factors has been the realm of three‐dimensional electron microscopy and thus has evolved hand in hand with the progress of this technique. While the first atomic structure of tubulin was obtained from electron crystallographic studies of zinc‐induced 2D sheets, an aberrant polymer of tubulin, modern cryo‐electron microscopy (cryo‐EM) is today allowing atomic‐detail visualization of microtubules themselves as well as their interacting partners. We have investigated the structural basis of MT dynamic instability using cryo‐EM, by characterizing, at atomic resolution, the conformational changes in MTs that accompany GTP hydrolysis. To this end we obtained cryo‐EM structures at ~3.5 Å resolution of MTs bound to GMPCPP, GTPγS, or GDP, either decorated with kinesin or copolymerized with the +TIP protein EB3. This resolution has allowed us to derive atomic models directly from the density maps and to shed unique light on several key aspects of MT structure and dynamics and its regulation by EB proteins. Our studies show that through the GTP hydrolysis cycle, subtle changes around the E‐site nucleotide trigger conformational changes in α‐tubulin and leading to global lattice compaction and strain generation. The structures also revealed, for the first time, the atomic details of the native lateral contacts between protofilaments and showed that they do not change significantly with GTP hydrolysis, except at the


seam. We have described the atomic details of the interaction of EB3 the microtubule, and been able to provide a structure‐based model of both the nucleotide state preference of this protein and its effect on microtubule dynamics. Our main efforts now are to decipher the mode of binding and action of other cellular factors that regulate microtubule assembly, dynamics and organization in the cell, with a special emphasis on proteins critical for the mitotic process.

E. E. Just Lecture

A2 Th cell biology of synapses and behavior. D.A. Colón‐Ramos1; 1Cell Biology and Neuroscience, Yale University, New Haven, CT A central unsolved question in neurobiology is how synapses are precisely assembled to build the neuronal architecture that underlies behavior. To address this, we developed tools in the thermotaxis circuit of C. elegans. Our system enables unbiased genetic screens to identify novel pathways that instruct synaptogenesis in vivo, and single‐cell manipulation of these pathways to understand how they influence behavior. Using this system we identified and characterized two unexpected roles for glia in governing synaptic positions within neural circuits. These roles, although mediated by the same specific glia cell, depend on different molecular pathways and are required at distinct life stages of the organism: 1) During development, glia‐derived Netrin controls synaptic partner choice by specifying the site of synaptogenesis through a novel, but conserved mechanism; 2) During post‐developmental growth, synaptic positions are maintained by glia location, which is determined by epidermal cells via a novel transporter and the FGF receptor. To achieve this, we developed multidisciplinary approaches for imaging and manipulating synapses in vivo and with single‐cell resolution to identify key molecular players, determine how they regulate information flow (by calcium imaging) and quantify their role on locomotory strategies during thermotaxis.

Bruce Alberts Award for Excellence in Science Education

A3 Investigations of undergraduate research experiences and science learning: How findings drive new questions. D. Lopatto1; 1Center for Teaching, Learning, and Assessment, Grinnell College, Grinnell, IA In an effort to sustain and increase the community of scientists, science teachers and mentors often produced successors by folding students into copies of themselves. The basis for this approach was the recollection of the scientist’s own personal development, a life path that typically included an undergraduate research experience. A study of these scientists in the 1950s found that an undergraduate research experience under the guidance of a talented mentor was the common catalyst for a science career. Undergraduate research was clearly a significant factor in the development of scientific interest, but the extent to which it had its effect, and indeed what constituted an authentic research experience, were questions that deserved investigation. This investigation has propelled my work for nearly two decades. An initial attempt to answer the questions about the essential features and benefits of undergraduate research included a survey called ROLE that recruited information from student researchers at 4 liberal


arts institutions. The survey focused on the relatively intense form of undergraduate research, the 10‐week summer research program in which the student was an apprentice to a faculty mentor. The results revealed that the range of benefits to student researchers extended beyond training as a scientist to include increased self‐confidence, tolerance for obstacles, and understanding science. A broader effort to evaluate student research led to the SURE survey, currently used by approximately 3000 students per year. The sharing of SURE results provoked discussions of situating the essence of the research experience within a scheduled course. The investigation of these “research‐like” experiences inspired the creation of the CURE, a classroom version of the SURE survey. Subsequently the extension of the research to interdisciplinary science courses led to the creation of the RISC survey, an instrument that captures interdisciplinary learning. Going forward, it is likely that our research questions will shift from the focus on survey instruments to a broader view, including defining the reference population for this research and the criterion for success of undergraduate research programs.

ORAL PRESENTATIONS‐ MONDAY

Oral Presentations‐Monday, December 5 Symposium 3: Disease Informing Cell Biology

S6 Cell biological and genetic basis of neurodevelopmental disease. J.G. Gleeson1; 1Neuroscience, HHMI, University of California, San Diego, San Diego, CA Disease can inform cell biology, and is key to understanding mechanisms and developing new treatments. Studying disease is a powerful approach to understanding cellular processes because humans collectively have mutations in every gene—if not every codon—of the genome, providing a rich opportunity to connect cellular processes with function. Recent sequencing approaches have vastly expanded our understanding of the genetics of these diseases, uncovering literally thousands of new causes of disease. Nowhere is this more apparent than in the study of neurodevelopmental disease, in part because of the complexity of human brain development, and in part because of the rich diversity of gene expression and cell types in the developing brain. For most of these novel genes, almost nothing is known why these mutations produce strong neurological phenotypes. These mutations largely fall into three categories: recessive, de novo and somatic (post‐zygotic). Recent examples include homozygous loss of function mutations in actin regulators and somatic gain of function mutations in mTOR pathway both leading to defects in neuronal migration, revealing convergence of signaling and cell biological pathways. The next generation of treatments will require integration of cell biology within knowledge of human phenotypes and genetics.

S7 Mitochondrial parts, pathways, and pathogenesis. V.K. Mootha1,2,3; 1Molecular Biology, Massachusetts General Hospital, Boston, MA, 2Systems Biology, Harvard Medical School, Boston, MA, 3Investigator, Howard Hughes Medical Institute, Boston, MA Mitochondria are ancient organelles that serve as cellular hubs for energy metabolism, signaling, and growth. Mitochondria contain their own genome (mtDNA) that encodes 13 proteins, while all remaining 1000+ proteins are encoded in the nuclear genome and imported. There is growing interest in mitochondrial biology since virtually all age‐associated diseases are associated with a decline in mitochondrial function. Our laboratory focuses on monogenic mitochondrial disorders – a large collection of inborn errors of metabolism, which can present in infancy or in adulthood, impacting virtually any organ system. Although next‐generation sequencing has revolutionized the genetic diagnosis of these disorders, we still know precious little about pathogenesis, and we do not have a single proven therapy. Here I will share our latest efforts that combine human physiology, genetic screens, and functional genomic profiling to better understand mitochondrial pathogenesis. Our work is providing fundamentally new insights into mitochondrial homeostasis with potential implications for the therapy of these disorders.

Symposium 4: Quality Control

S8


Selective inhibition of a phosphatase to correct protein quality control failures and treat neurodegenerative diseases. I. Das1, A. Krzyzosiak1, A. Sigurdardottir1, K. Schneider1, A. Bertolotti1; 1MRC Laboratory of Molecular Biology, Cambridge, United Kingdom The deposition of misfolded proteins is a defining feature of many age‐dependent human diseases, including the increasingly prevalent neurodegenerative diseases. Why misfolding‐prone proteins accumulate in aged cells remains largely unclear. Cells normally strive to ensure that proteins get correctly folded and have powerful and sophisticated protein quality control mechanisms to maintain protein homeostasis under adverse conditions. However, with age, the cellular defence systems against misfolded proteins gradually fail, leading to the accumulation of misfolded proteins with devastating consequences for cells and organisms. In principle, improving the cells’ ability to deal with misfolded proteins should represent a generic approach to reduce pathology in diverse protein misfolding diseases. My lab has identified powerful strategies to help cells survive when protein quality control fails and implemented one of such strategies in mice. Exploiting the current knowledge on protein quality control systems, we have identified a small drug‐like molecule that safely boosts the natural defence system against misfolded proteins. When given to mice, the small molecule completely prevents two unrelated neurodegenerative diseases, without any deleterious side effects. The small‐molecule was named Sephin1 because it is a selective inhibitor of a holophosphatase. Our work demonstrates that generic approaches aimed at helping cells to survive protein quality control failures can be useful to prevent protein misfolding diseases, including the devastating neurodegenerative diseases. Moreover, whilst phosphatases were previously thought to be undruggable, our work establishes that phosphatases can be selectively inhibited by targeting their regulatory subunits. This finding has broad relevance because there are hundreds of phosphatases that could in principle be inhibited using the same paradigm. This opens up a broad range of possibilities to manipulate cellular function for therapeutic benefit.

S9 Stressed Out: A Novel Approach to Cancer Immunotherapy. L.H. Glimcher1; 1Dana‐Farber Cancer Institute, Boston, MA Cancer cells induce a set of adaptive response pathways to survive in the face of stressors due to inadequate vascularization1 . One such adaptive pathway is the unfolded protein (UPR) or endoplasmic reticulum (ER) stress response mediated in part by the ER‐localized transmembrane sensor IRE1 and its substrate XBP1. We have shown that the transcription factor XBP1 promotes intrinsic tumor growth directly in the setting of triple negative breast cancer, and now have established that this signaling pathway also regulates the host anti‐tumor immune response. Dendritic cells (DCs) are required to initiate and sustain T cell‐dependent anti‐cancer immunity. However, tumors often evade immune control by crippling normal DC function. Constitutive activation of XBP1 in tumor‐associated DCs (tDCs) drives ovarian cancer (OvCa) progression by blunting anti‐tumor immunity. XBP1 activation, fueled by lipid peroxidation byproducts, induced a triglyceride biosynthetic program in tDCs leading to abnormal lipid accumulation and subsequent inhibition of tDC capacity to support anti‐tumor T cells. Accordingly, DC‐specific XBP1 deletion or selective nanoparticle‐mediated XBP1 silencing in tDCs restored their immunostimulatory activity in situ and extended survival by evoking protective Type 1 anti‐tumor


responses. Targeting the ER stress response should concomitantly inhibit tumor growth and enhance anti‐cancer immunity, thus offering a unique approach to cancer immunotherapy.

S10 Mechanisms of nascent protein quality control in the cytosol. R.S. Hegde1; 1MRC Laboratory of Molecular Biology, Cambridge, United Kingdom The biogenesis of proteins involves a complex series of steps that include synthesis, targeting, modification, folding, and assembly. Failure at any of these steps results in the production of an aberrant product that must be degraded by the cell. An inability to promptly recognize or degrade aberrant proteins can lead to their accumulation, cell dysfunction, and in many cases, disease. Our group’s research efforts have focused on determining how the cell monitors each step in protein biosynthesis to identify failed products that are selectively targeted for degradation by the ubiquitin proteasome system. Using a primarily biochemical approach, we have been able to reconstitute several of these ubiquitination pathways including stalled translation products on the ribosome, mis‐targeted membrane proteins, and orphan subunits of multi‐protein complexes. Our studies are beginning to reveal the molecular logic of how nascent proteins are recognized by various cellular factors, and how the nature of these interactions ultimately leads to a triage decision that segregates polypeptides between biosynthetic and degradative fates. These decisions are crucial for maintaining overall cellular homeostasis, and it is our hope that a molecular understanding of these pathways will provide new insights into the diseases caused by protein misfolding.

Minisymposium 7: Actin Dynamics

M71 Competitive and cooperative interactions between actin binding proteins drive their sorting to different actin cytoskeleton networks. J.R. Christensen1, K.E. Homa1, M.E. OConnell1, A.N. Morganthaler1, D.R. Kovar1; 1Molecular Genetics and Cell Biology, The University of Chicago, Chicago, IL

Multiple diverse actin filament networks are assembled within a crowded common cytoplasm to assist in a variety of cellular functions such as cytokinesis, endocytosis, and polarization. Each of these F‐actin networks possesses a unique set of actin binding proteins (ABPs) that give the network its defined architecture and dynamics. Though much is understood about the properties and mechanics of individual ABPs, the mechanisms by which a set of ABPs localizes to one F‐actin network rather than another are unclear. We hypothesize that competition between ABPs for association with actin filaments may be a driving force in establishing this sorting. Using a combination of multi‐color TIRF microscopy and live cell imaging, we have dissected a web of competitive interactions between ABPs in fission yeast. An initial survey in fission yeast determined ABPs that have competitive interactions, and led us to focus on three ABPs: fimbrin, α‐actinin, and tropomyosin. Fimbrin and α‐actinin are F‐actin bundling proteins that sort to different F‐actin networks—fimbrin associates predominantly with actin patches and α‐actinin associates with the contractile ring. In vitro multi‐color TIRF microscopy experiments show that fimbrin and α‐actinin self‐sort to distinct two‐filament bundles, providing a potential mechanism for their segregation to different F‐actin networks in vivo. In addition, altering the concentration and dynamics of α‐actinin and fimbrin in fission yeast cells results in their relocalization to


the opposing actin network, demonstrating their competitive interactions with each other. Additionally, we show that side‐binding ABP tropomyosin behaves differently on fimbrin‐ versus α‐actinin‐mediated actin bundles. Tropomyosin is excluded from fimbrin‐mediated actin bundles. Conversely, tropomyosin is not excluded from α‐actinin‐mediated bundles, and in fact, enhances the bundling ability of α‐actinin. These results suggest that tropomyosin and α‐actinin work together to compete with fimbrin for association with the contractile ring. Overall, our findings show that a delicate balance of competitors dictates the sorting of ABPs to different actin networks.

M72 An actin filament stress sensor based on utrophin. A.R. Harris1, P. Jreij2, D.A. Fletcher2; 1QB3, University of California Berkeley, Berkeley, CA, 2Department of Bioengineering and Biophysics Program, University of California Berkeley, Berkeley, CA The actin cytoskeleton of metazoan cells is organized by actin‐binding proteins into distinct networks of actin filaments with diverse functions. Force transmission through these actin filament networks is central to their roles in cell movements, shape change, and internal reorganization. However, tracking actin cytoskeletal forces during dynamic cellular processes has not been possible due to the lack of appropriate measurement methods. Here we demonstrate a method for directly quantifying forces on the actin cytoskeleton in live cells that takes advantage of force‐dependent protein binding to actin filaments. We modify the actin‐binding domain of utrophin, a commonly‐used marker of actin filaments in live cells, to create a mutant utrophin whose affinity to filamentous actin scales with tension on the filament. We show that the mutant utrophin binds preferentially to filaments under tension in actin filament gels and single filaments in vitro. By comparing wild type and mutant utrophin binding in live cells, we are able to generate stress maps of actin cytoskeleton forces during cell spreading, motility, and blebbing. We find that this force‐dependent binding to actin is not unique to mutant utrophin, suggesting that filament tension may serve as a more general organizing principle for actin filament networks. The actin filament force sensor presented here provides a new measurement tool that may be useful for exploring mechanical regulation of actin‐associated cellular processes.

M73 Comprehensive analysis of formin localization at cell‐cell junctions and cytokinetic contractile rings in epithelial cells. T. Higashi1, A.L. Miller1; 1MCDB, University of Michigan, Ann Arbor, MI Reorganization of the actomyosin cytoskeleton regulates dynamic cellular processes including cytokinesis and remodeling of cell‐cell junctions. Although formin proteins are known to assemble linear actin filaments, it is not fully understood which formin is responsible for formation or remodeling of specific actin structures, such as the actomyosin bundle connected to cell‐cell junction or the cytokinetic contractile ring. Here, we show that the formin inhibitor SMIFH2 perturbed the F‐actin bundle at cell‐cell junctions in gastrula‐stage Xenopus laevis embryos, but did not cause detectable cytokinesis defects. To determine which formin(s) were responsible, we cloned all 15 X. laevis formins and investigated the expression and localization of each formin in a comprehensive analysis using a developing vertebrate model organism. We measured mRNA levels of all formins in early X. laevis developmental stages using semi‐quantitative RT‐PCR. We also examined cellular localization of all formins by expressing 3xGFP‐tagged formins in animal cap epithelial cells at gastrula stage. Dia1, Dia2,


Fhod1 and Fhod3 were strongly localized at cell‐cell junctions. Overexpression of Dia3, which weakly localized at cell‐cell junctions, induced formation of abnormal filopodial and lamellipodial structures at cell‐cell junctions, and Dia3 was localized at the tip of filopodial structures and the surface of lamellipodial structures. In dividing cells, Dia3 exhibited strong localization at the contractile ring, whereas Dia1 and Dia2 were weakly localized at contractile ring. Additionally, Fmn1 (Formin‐I) was localized on microtubules and Fmn2 showed nuclear localization. Altogether, our results provide a systematic analysis of the expression and localization of all vertebrate formins in the same organism and identify new information about which formins may be functionally important for regulating actin structures in cytokinesis and at cell‐cell junctions.

M74 Interplay between filopodia and focal adhesions through the formin FMNL3. L.E. Young1, E.G. Heimsath2, H.N. Higgs1; 1Biochemistry, Dartmouth College, Hanover, NH, 2Cell Biology and Physiology, UNC, Chapel Hill, NC Filopodia are finger‐like, actin dependent structures that protrude from the plasma membrane. Although they are of fundamental importance to cellular physiology, the mechanisms governing filopodia assembly are still in question. One model, called convergent elongation, proposes that filopodia arise from Arp2/3 complex‐nucleated dendritic actin networks, with factors such as formins elongating these filaments into filopodia. Another model, called tip nucleation, proposes that formins both nucleate and elongate the filaments used in filopodia. Our previous work had shown that expression of constitutively active construct of the formin FMNL3 induces potent filopodia assembly in a variety of mammalian cells (Harris et al (2010) Cytoskeleton). We subsequently found that FMNL3 has cell‐type specific mechanisms of filopodia assembly. In suspension cells, there is an absolute requirement for Arp2/3 complex in filopodia assembly. In adherent cells, however, a sub‐set of filopodia remains after Arp2/3 complex inhibition. Interestingly, the formin mDia1 also contributes to filopodial assembly in adherent cells, yet does not play a role in suspension cells. Furthermore, we find a disproportionate number of filopodia associated with focal adhesions in adherent cells (Young et al (2015), MBoC). We now find that this is also true for full‐length FMNL3, and have analyzed in detail the roles of FMNL3, mDia1, and VASP in the assembly and properties of filopodia and three types of focal adhesion‐associated actin‐based structure: dorsal stress fibers, ventral stress fibers and transverse arcs. From these results, we propose an extension of the existing models for filopodial assembly whereby any cluster of actin filament barbed ends in proximity to the plasma membrane, either Arp2/3 complex‐dependent or ‐independent, can initiate filopodial assembly by specific formins. Further, these findings outline a close inter‐relationship between dorsal stress fibers and filopodia, with both structures utilizing the same actin filament “reservoir”. Focal adhesions may be the origin for this reservoir, and decisions to assemble dorsal stress fibers or filopodia may be mediated by the interplay between FMNL3, mDia1 and VASP.

M75 Actin and microtubule crosstalk mediate persistent polarized growth. S. Wu1, M. Bezanilla1; 1Biology, University of Massachusetts, Amherst, MA How organisms impart geometric information ultimately leading to cell and tissue morphogenesis is not well understood. Eukaryotes often control this by building a wide variety of extracellular structures, which are patterned over macroscopic scales. Yet, individual cells are responsible for depositing


extracellular matrix. In plants a complex extracellular matrix constrains cells and dictates tissue morphogenesis. Thus in plants, cell shape becomes a read out for extracellular matrix patterning. As a model, we study how the highly polarized tip growing cells in mosses are patterned. A dynamic actin cytoskeleton is essential in these cells. Without actin, cells swell at their apexes and growth is severely inhibited. During growth, wild type cells have an accumulation of actin just below the cell tip. While this apical actin accumulation is persistent over long time periods, imaging with high temporal resolution reveals that this population of actin filaments is highly dynamic. In the absence of microtubules, the apical actin accumulation loses the longer temporal persistence and instead stochastically forms and reforms in various regions of the cell, often leading to cell expansion at other sites. In wild type, microtubules are focused with their plus ends just below the apical actin accumulation. We find that in a mutant lacking all five myosin VIII genes, the microtubule focus dissipates and the actin accumulation loses the long term persistent apical accumulation similar to loss of all microtubules. Interestingly directional persistence is impaired in this mutant. We also find that myosin VIII localizes at the junction between the focused microtubule plus ends and the apical actin accumulation. These data suggest that microtubules are required to stabilize apical actin and myosin VIII is the physical link between the two cytoskeletons. To gain insights into the molecular basis of this interaction, we used TIRF microscopy to analyze the interaction between actin, myosin VIII, and end binding protein (EB) 1 at the cell cortex. We found that a population of myosin VIII particles moves with a speed similar to the growth rate of microtubule plus ends. This group of myosin VIII also colocalized with microtubule plus ends, and their movements paused when encountering a cluster of actin filaments. These findings suggest that myosin VIII serves as a molecular break between microtubule plus ends and actin structures. Thus, at the cell apex, microtubules may deliver factors that promote spatial restriction of actin polymerization and actin turnover leading to the generation of the persistent apical actin accumulation, ultimately leading to cell expansion proximal to this site.

M76 Actin cable arrays controlling nuclear position in Drosophila nurse cells require microtubules, EB1, APC1, APC2, and formins for their assembly. R. Jaiswal1, O. Molinar2, B.L. Goode1, B.M. McCartney2; 1Department of Biology, Brandeis University, Waltham, MA, 2Department of Biological Sciences, Carnegie Mellon University, Pittsburgh, PA Striking arrays of colossal actin cables (~30 µm long, ~25 filaments thick) develop in nurse cells late in Drosophila oogenesis. The 15 nurse cells are connected to each other and to the oocyte through actin‐based tunnels called ring canals, through which nurse cells send essential mRNAs, proteins, and organelles into the developing oocyte. Actin cables grow inward in an array from the nurse cell cortex to contact the nucleus, pushing it away from the ring canals and thus protecting them from nuclear blockade. These cable arrays develop with tight spatial and temporal control that mirrors their requirement in oogenesis. We have discovered that the formation of actin cable arrays requires two Adenomatous polyposis coli proteins (APC1 and APC2), the formin Diaphanous (Dia), the kinase GSK3ß, and microtubules (MTs). Biochemically, APC1 and Dia collaborate to nucleate actin assembly, and our genetic observations show that both APC1 and Dia are required for actin cable formation. This is the first compelling evidence that APC1 is an actin assembly factor in the living animal. In contrast, APC2 restrains premature Dia‐dependent cable assembly through a mechanism regulated in part by GSK3ß phosphorylation. Further, the C‐terminal half of APC2 inhibits APC1‐ and APC1‐Dia‐mediated actin assembly in vitro, and a phosphomimetic APC2 mutant exhibits much stronger inhibitory activity. Later, at the proper developmental stage, APC2 promotes cable formation, suggesting that APC2 has both


positive and negative roles in actin assembly, which are switched by GSK3ß phosphorylation. Finally and unexpectedly, we found that cytoplasmic MTs strikingly co‐align with elongating actin cables and are required their formation, and further the MT +Tip protein EB1 suppresses premature and spatially ectopic cable assembly and directly inhibits APC1‐Dia‐mediated actin nucleation in vitro. Together, our combined “top‐down” genetic and “bottom‐up” biochemical approach has revealed a complex network of regulatory factors including MTs that mediate the formation of actin cable arrays critical for oogenesis.

M77 How does actin disassembly induce myelin wrapping? J. Zuchero1, B. Barres1; 1Neurobiology, Stanford University, Stanford, CA Myelin, the electrical insulator surrounding neuronal axons, is required in vertebrates for rapid nerve conduction. Myelin is built in the CNS by the spiral wrapping of an oligodendrocyte’s terminal process around the axon and simultaneous compaction of the membrane to squeeze out cytoplasm. During myelination oligodendrocytes increase their membrane surface area by many thousand fold, yet the mechanisms of membrane generation, trafficking, and protrusion remain fundamental, unanswered questions. It had been proposed that myelin wrapping is mechanistically similar to the protrusion of a motile cell’s lamellipodium, powered by the dynamic assembly of an actin cytoskeleton. Surprisingly, we discovered that myelin wrapping coincides with the dramatic upregulation of actin severing proteins in oligodendrocytes, whose actin cytoskeleton is rapidly disassembled at the start of wrapping. In contrast to the extension of lamellipodia, myelin wrapping does not require actin assembly by the Arp2/3 complex and is accelerated in vivo by inducing actin disassembly. Thus myelin wrapping represents a novel form of cellular protrusion that is independent of actin filaments. We proposed a “blebbing model” of myelin wrapping in which membrane compaction and insertion of new lipids power processive growth. To test this model, we have developed DeActs, a genetically encoded toolkit for perturbing the actin cytoskeleton in vivo in a cell type of interest. We validated DeActs in tissue culture cells and in vivo in the worm and mouse nervous system, and showed that they can specifically and inducible ablate the actin cytoskeleton in single cells. Expressing DeActs in oligodendrocytes in culture phenocopies treatment with latrunculin, triggering loss of the actin cytoskeleton and expansion of the myelin membrane. We will now use DeActs in vivo to test the hypothesis that oligodendrocyte‐autonomous actin disassembly is a key signal to initiate myelin wrapping.

M78 Intermediate filament and Plakin family function in the morphogenesis of actin‐based microridges in zebrafish. Y. Inaba1, V. Chauhan1, A. Sagasti1; 1Department of Molecular, Cell and Developmental Biology, , University of California, Los Angeles, Los Angeles, CA Microridges are beautiful actin‐based structures that protrude from the apical surfaces of mucosal epithelial cells and form fingerprint‐like patterns. Microridges appear on many mucosal cell surfaces, including fish skin and the corneas of many species. Defects in the mucosal layer of the human cornea cause dry eye diseases, so defects in microridges may contribute to water retention on mucosal cell surfaces. We study microridge morphogenesis in the larval zebrafish skin, which consists of two layers, an outer layer called the periderm, which has microridges, and an inner layer of basal cells, which do


not. F‐actin accumulates, on the apical surface of periderm cells from 16 hours post fertilization. The distribution of F‐actin dramatically changes during microridges morphogenesis: at early stages, F‐actin forms finger‐like protrusions reminiscent of microvilli; later these microvilli‐like structures combine to form microridges. The molecular mechanism of microridge morphogenesis is not well understood. Previous studies in our lab determined expression profile of zebrafish skin cells with RNA‐seq. We found several genes encoding cytoskeletal proteins that were highly expressed in the periderm, but not in basal cells, including Envoplakin, Periplakin and Keratin17. Envoplakin and Periplakin are members of the Plakin family of proteins, which cross‐link cytoskeletal elements and junctional complexes. To determine the localization of Envoplakin, Periplakin and Keratin17, we made Envoplakin::GFP, Periplakin::GFP and Keratin17::GFP reporter fusions in bacterial artificial chromosomes (BACs) and determined their distributions in transgenic zebrafish. Envoplakin, Periplakin and Keratin17 all localized to microridges. Strinkingly, Keratin17 formed filament patterns at the base of microridges. We observed the same localization patterns using homologous recombination with CRISPR/Cas9 to tag endogenous Periplakin and Keratin17 with GFP. We therefore hypothesize that Envoplakin and Periplakin link Keratin17 with F‐actin during microridge morphogenesis. To begin to determine the function of Envoplakin and Periplakin in microridge morphogenesis, we overexpressed them. Surprisingly, microridges were longer when Envoplakin and Periplakin were overexpressed together, but not when each was overexpressed alone. These results suggest that Envoplakin and Periplakin interact in periderm cells and play a role in microridge morphogenesis. We are currently testing which domains of Envoplakin and Periplakin are required to cause longer microridges and creating genetic knockouts of these proteins with CRISPR/Cas9.These studies will help reveal the function of Envoplakin and Periplakin in microridge morphogenesis.

M79 Form and Function: Distinct Actin Architectures and Nucleators for the Epithelial Plasticity Programs of Embryonic Stem Cell Differentiation and Epithelial to Mesenchymal Transition. F.M. Aloisio1, M. Rana1, D.L. Barber1; 1Department of Cell and Tissue Biology, University of California, San Francisco, San Francisco, CA Distinct actin architectures with different properties drive myriad processes, including bud site selection for yeast cell division, the immune synapse in activated T cells and leading‐edge membrane protrusions in a migrating cells. We addressed questions on actin filament architectures and functions for two epithelial plasticity programs, the differentiation of naïve mouse embryonic stem cells (mESCs) and epithelial to mesenchymal transition (EMT). Despite extensive knowledge of the transcriptional regulation for these two programs, little is known about how actin nucleators and filament dynamics contribute to the extensive morphological changes that occur with both processes. Both programs initiate from similar compact cuboidal epithelial cells with a cortical ring of unbranched actin filaments. Also, both programs result in differentiated cells with similar morphologies of elongated cell shapes but with distinct actin architectures. We found that differentiated mESCs have predominantly dense, branched actin filaments within motile cell protrusions, which had not previously been described. In contrast, as previously described, mesenchymal cells with EMT have abundant unbranched contractile actin filaments traversing the cell body. We found that for mESCs, blocking activity of the Arp2/3 complex with the pharmacological inhibitors CK666 and CK869 but not activity of formins with the inhibitor SMIFH2 prevents actin remodeling and inhibits differentiation, as determined by suppressed expression of epiblast‐specific Fgf5, Branchyury and miR‐302 clusters, as well as retention of embryoid body formation capability. Our data suggest that Arp2/3‐dependent mESC differentiation is mediated by nuclear translocation of MRTF, a co‐activator of transcription factor SRF that regulates a miRNA gene cluster in stem cells. In contrast, inhibiting formins but not Arp2/3 complex activity blocks EMT of


alvelolar, renal and mammary epithelial cells in response to TGF‐β, including preventing actin filament remodeling to contractile stress fibers, increased phosphorylation of MLC and expression of the transcription factor Snail, as well as retaining expression of the epithelial adherens junction protein E‐cadherin. Our data also indicate for the first time that formin but not Arp2/3 complex activity is necessary for proximal TGF‐β signaling and phosphorylation of SMAD2. Using shRNA for selective formin family members we found that DIAPH1 and 3 but not other formins are necessary for TGF‐β phosphorylation of SMAD2 and EMT, which has broad significance for the many cell behaviors regulated by this cytokine. Collectively, our findings reveal the critical role and mechanistic actions of distinct actin nucleators and architectures for two epithelial differentiation programs.

M80 Investigating the role of ERK‐mediated actin phosphorylation in ciliary dynamics. S. Dutta1, P. Avasthi1,2; 1Anatomy and Cell Biology, University of Kansas Medical Center, Kansas City, KS, 2Ophthalmology, University of Kansas Medical Center, Kansas City, KS Eukaryotic cilia and flagella are evolutionary conserved microtubule‐based organelles that play essential roles in diverse physiological processes including cell motility, signal transduction and sensory perception. Understanding the mechanism involved in ciliary dynamics are crucial for identifying sources of cilia‐related diseases such as polycystic kidney disease, primary cilia dyskinesia and retinal degeneration. Using Chlamydomonas as a model organism, we have found that a mitogen activated protein (MAP) kinase phosphatase‐3 inhibitor, BCI [(E)‐2‐benzylidene‐3‐(cyclohexylamino)‐2, 3‐dihydro‐1H‐inden‐1‐one] shortens the length of flagella, activates extracellular regulated kinase (ERK) and leads to increased actin phosphorylation. Based on previous studies demonstrating ERK regulates actin filament stability and our previously identified relationship between actin dynamics and flagellar dynamics, we propose that ERK mediated phosphorylation of actin in response to BCI treatment contributes actin filament reorganization and affects flagellar length. Our current findings reveal that ERK phosphorylation as well as flagellar shortening started within 15 minutes of BCI treatment and continued up to 120 minutes. Phosphorylated actin appeared within 30 minutes and then decreased gradually in a time‐dependent manner. Surprisingly, mutants lacking actin made the BCI‐mediated flagellar shortening more severe. This suggests that the role of actin modification may be to resist the disassembly rather than mediate it. Interestingly, the MEK inhibitor U0126 [1,4‐diamino‐2,3‐dicyano‐1,4‐bis(2‐aminophenyl thio) buta‐diene] slightly lengthens the flagella and blocks ERK phosphorylation. However, ERK phosphorylation recovers and strongly increases by 120 minutes. This suggests that U0126 activates a feedback loop that initially suppresses but ultimately upregulates ERK phosphorylation. This work is ultimately expected to connect an essential mitogenic pathway with actin‐mediated ciliary dynamics and provide a broader context for understanding the basic biology of the cilia‐related disorders.


Minisymposium 8: Cell Biology of the Nucleus

M81 Super‐resolution study of nuclear envelope transmembrane protein transport in live cells. K.C. Mudumbi1, E. Schirmer2, W. Yang1; 1Department of Biology, Temple University, Philadelphia, PA, 2Wellcome Trust Center for Cell Biology, University of Edinburgh, Edinburgh, United Kingdom The double membrane of the nuclear envelope (NE), consisting of the outer and inner nuclear membranes (ONM and INM), separates the cytoplasm and the nucleus in eukaryotic cells. Transport of nuclear envelope transmembrane proteins (NETs) from the ONM to the INM after synthesis in the endoplasmic reticulum plays a critical role in the maintenance and organization of the nucleus as well as in gene expression regulation. Dysfunctions in this transport process have been associated to many human diseases. However, multiple transport mechanisms of NETs across the NE have been postulated and remain in debate. To refine the mechanism, we have developed and utilized a new high‐speed super‐resolution microscopy approach to study the real‐time transport of various NETs into the INM in live cells. Here we will show the new findings obtained by the newly developed techniques.

M82 Super‐resolution imaging reveals epigenome organization and its spatial coordination with transcription factories in single mammalian cells. J. Xu1, H. Ma1, J. Jin1, Y. Huang2, N. Davidson3, Y. Liu1; 1University of Pittsburgh Cancer Institute, School of Medicine , Pittsburgh, PA, 2Magee‐Women Research Institute, University of Pittsburgh, Pittsburgh, PA, 3University of Pittsburgh Cancer Institute, UPMC Cancer Centers, Pittsburgh, PA Eukaryotic genomes are packaged into nucleosome, a basic subunit of chromatin, which contains an octamer of histone proteins. The spatial organization of epigenome (e.g., histone modification) has a profound impact in regulating transcription and gene expression. However, existing methods to study the interaction between histone modification and transcription heavily rely on biochemical assays that are performed on pooled cell populations and obscure the spatial information at a single cell level. Recent development of super‐resolution nanoscopy has transformed our way to understand biology, as it offers new possibilities of visualizing the previous invisible spatial information at a molecular scale. Here, we present the use of super‐resolution imaging of genome‐wide histone modifications and their functional relationship to transcription regulation. By using stochastic optical reconstruction microscopy (STORM), we first visualized nanoscale chromatin organization at a single cell level in which different histone modification markers were used to identify the functionally distinct heterochromatin and euchromatin. Transcription factories were visualized by imaging the transcribed regions of active genes via phosphorylated RNA polymerase II. To investigate the relationship between chromatin organization and transcription activity, suberoylanilide hydroxamic acid (SAHA) and 5,6‐dichloro‐1‐beta‐D‐ribofuranosyl‐benzimidazole (DRB) were used either to de‐condense chromatin or inhibit the transcription activity. Our STORM images reveal that spatial organization of nucleosomes enriched by functionally distinct histone modifications (i.e., H3K9me3, H3K27me3, H3K4me3, H3K36me3, H3K9Ac, Acetyl‐H4) and transcription factories marked by RNA polymerase II are assembled in heterogeneous discrete clusters of varying sizes, in a sharp contrast of DNA which spread throughout the nucleus. Increased size and density of PolII cluster units correlates with increased transcription activities. Two‐color STORM imaging reveals that the heterochromatin and euchromatin clusters in general are spatially


exclusive from each other, and euchromatin clusters show a higher level of co‐localization with Pol II clusters compared to heterochromatin clusters. When treated with SAHA, an increased PolII cluster size is seen corresponding to increased transcription activities and distinct de‐condensation effect in heterochromatin and euchromatin organization. However, reciprocal disruption by inhibiting transcription only affects the euchromatin organization but not the heterochromatin. The transcription elongation regions marked by H3K36me3 exhibit more co‐localized with Pol II clusters, compared to the promoter regions (marked by H3K4me3 and H3K9ac), indicating more Pol II molecules recru

M83 Hunchback and Krüppel mediate early cell fate competence correlated with distinct actin‐dependent intranuclear re‐localization of gene loci in Drosophila neuroblasts. P. Chikte1, S. Jansen1, N. Dinges2, J. Urban1; 1Biology, Institute of Genetics, Mainz, Germany, Mainz, Germany, 2Institute of Molecular Biology, Mainz, Germany Neural stem cells produce specific cell types in a time dependent manner passing through different competence windows. However, how these competence windows are regulated is largely unknown. Here we use embryonic Drosophila neuroblasts (NB) as a model system to investigate this question. These NBs produce specific cell lineages stereo‐typically and are characterized by sequential expression of transcription factor cascade Hb/Kr>Kr> Pdm>Pdm/Cas>Cas. It has been shown earlier that the competence to produce Hb‐dependent early cell fates can be extended if Hb expression is prolonged. Similar results were observed in Cortex development of vertebrates where the Hb homologue Ikaros permits early cell fate competence to neural stem cells of this region. Here we show that in Drosophila continuous activation of the Kr‐locus which normally gets silenced after termination of its expression is necessary for extension of NB competence window. Intriguingly this silencing of Kr gene locus is correlated with its unusual re‐localization from nuclear periphery, a normally repressive subnuclear compartment, towards the interior and can be inhibited by continuous Hb activity. Strikingly, co‐expression of hb and Kr outside of the early competence window induces a reversion of Kr‐locus position towards nuclear periphery correlating with ectopic production of early neuronal cell types at this late time of development. An opposite and Hb‐independent re‐localization pattern from nuclear interior towards periphery has been described earlier for the hb‐locus during early competence window. However, less is known about possible mechanisms required for such chromatin rearrangements and there are hints in vertebrate system that nuclear ß‐actin might play role. Here we found evidence that oligomerization of nuclear ß‐actin is involved in such events. This has been deduced from over‐expression experiments using dominant negative non‐oligomerizing NLS‐GFP‐actin which resulted in deceleration of locus re‐localizations whereas wild‐type NLS‐GFP‐actin showed acceleration. Furthermore, over‐expressing Exportin 6, a specific transporter of ß‐actin out of the nucleus, leads to a block of hb and kr locus re‐localization. Most importantly, late overexpression of Hb in this situation reestablishes the ability to induce early cell fates showing a causal relationship between gene locus re‐localization and NB competence. We propose that within the early competence window there are multiple actin dependent gene loci re‐localization in different directions suggesting ongoing epigenetic silencing which can be partly inhibited or even reverted by the combined activity of Hb and Kr.


M84 A self‐extinguishing signaling circuit within the synaptonemal complex regulates meiotic recombination. A.F. Dernburg1,2,3,4, O. Rog1,2,3, S. Köhler1,2,3, L. Zhang1,2,3; 1Howard Hughes Medical Institute, Chevy Chase, MD, 2Molecular and Cell Biology, University of California, Berkeley , Berkeley, CA, 3California Institute for Quantitative Biosciences (QB3), Berkeley, CA, 4Life Sciences Division, Lawrence Berkeley National Laboratory, Berkeley, CA During meiosis, homologous chromosomes pair, synapse, and undergo crossover recombination, enabling them to segregate from each other. The regulation of crossover recombination remains mysterious. Most eukaryotes display "crossover assurance," meaning that crossovers are tightly limited in number and distributed in a widely spaced pattern. This has suggested that an inhibitory signal acts to prevent subsequent crossovers along the lengths of paired chromosomes. Evidence has suggested that the synaptonemal complex (SC), a unique polymer that forms between homologous chromosomes, may contribute to crossover regulation. Our work has analyzed meiotic chromosome dynamics in Caenorhabditis elegans through in vivo imaging. Several observations have led us to conclude the that the SC is not a conventional polymer, but rather a liquid‐like compartment that assembles through phase separation. We find that structural proteins of the SC are highly mobile within assembled complexes, and that polycomplexes, 3D lattices of SC proteins, behave as liquid droplets in vivo. Additionally, both SCs and polycomplexes are rapidly and reversibly dissolved by aliphatic alcohols such as hexanediols, indicating that their structural integrity depends on weak hydrophobic interactions, consistent with a liquid phase. Coupled with the ordered arrangememt of proteins within this complex, these observations indicate that the SC is a liquid crystal. We speculated that the SC compartment might regulate crossovers by concentrating recombination factors between homologs, and/or might act as a conduit for a crossover interference signal. Consistent with these ideas, we have elucidated a regulatory circuit that acts within the SC to designate crossovers, and self‐extinguishes to limit their occurrence to one per chromosome/compartment. Remarkably, we find that this circuit also signals autonomously within polycomplexes in the absence of recombination intermediates, in response to spatially regulated cell cycle signals. Taken together, these findings suggest a new model in which fluid dynamics within the SC mediate crossover regulation to ensure faithful meiotic chromosome segregation.

M85 Structure and Function of the Nuclear Pore Complex Cytoplasmic mRNA Export Platform. J. Fernandez‐Martinez1, S. Kim2, Y. Shi3, P. Upla4, R. Pellarin2,5, M. Gagnon6, I.E. Chemmama2, J. Wang3, I. Nudelman1, W. Zhang3, R. Williams1, W.J. Rice4, D.L. Stokes4, D. Zenklusen6, B.T. Chait3, A. Sali2, M.P. Rout1; 1Laboratory of Cellular and Structural Biology, The Rockefeller University, New York, NY, 2Department of Bioengineering and Therapeutic Sciences, Department of Pharmaceutical Chemistry, and California Institute for Quantitative Biosciences, University of California San Francisco, San Francisco, CA, 3Laboratory of Mass Spectrometry and Gaseous Ion Chemistry, The Rockefeller University, New York, NY, 4Simons Electron Microscopy Center, New York Structural Biology Center, New York, NY, 5Structural Bioinformatics Unit, Institut Pasteur, Paris, France, 6Département de Biochimie et Médecine Moléculaire, University of Montreal, Montreal, QC


The penultimate steps in mRNA export are performed by the Nup82 complex, a large conserved assembly at the cytoplasmic face of the nuclear pore complex (NPC), but little is known about its architecture and how it coordinates mRNA export and remodeling. By integrating diverse structural data, we have determined the molecular architecture of the native Nup82 complex at subnanometer precision. The complex consists of two compositionally identical multiprotein subunits that, unusually, adopt different configurations. The Nup82 complex fits into the NPC through the outer ring Nup84 complex. Surprisingly, our map shows that this entire 14 MDa Nup82‐Nup84 complex assembly positions the cytoplasmic mRNA export factor docking sites and mRNP remodeling machinery right over the NPC’s central channel, rather than on distal cytoplasmic filaments as previously supposed. We suggest that this configuration efficiently captures and remodels exporting mRNP particles immediately upon reaching the cytoplasmic side of the NPC.

M86 The molecular architecture of lamins in somatic cells. Y. Turgay1, M. Eibauer1, A.E. Goldman2, T. Shimi2, M. Khayat3, K. Ben‐Harush4, K. Sapra1, R.D. Goldman2, O. Medalia1,3; 1Department of Biochemistry, University of Zurich, Zurich, Switzerland, 2Department of Cell Biology, Northwestern University Feinberg School of Medicine, Chicago, IL, 3Department of Life Sviences and the National Institut for Biotechnology in the Negev, Ben Gurion University, Beer‐Sheva, Israel, 4Department of Chemical Engineering, Shamoon College of Engineering, Ashdod, Israel The nuclear lamina is a fundamental constituent of metazoan nuclei. It is composed mainly of lamins that assemble into a filamentous meshwork bridging the nuclear envelope and chromatin. Besides providing structural stability to the nucleus, the lamina is involved in many nuclear activities, e.g. chromatin organization, transcription and replication. However, the structural organization of the nuclear lamina is poorly understood. Here, we use cryo‐electron tomography to obtain a detailed view of the organization of the lamin meshwork within the lamina. Data analysis of individual lamin filaments resolved a globular‐decorated fiber appearance and showed that A‐ and B‐type lamins assemble into tetrameric 3.5 nm thick filaments. Thus, lamins exhibit a structure that is remarkably different than the other canonical cytoskeletal elements. Our findings examine the architecture of the nuclear lamina at molecular resolution, providing insights into its scaffolding and mechanosensory function.

M87 Tracking multiple genomic elements using correlative CRISPR imaging and sequential DNA FISH. J. Guan1, H. Liu2, X. Shi1, B. Huang1; 1Pharmaceutical Chemistry, University of California San Francisco, San Francisco, CA, 2Bioengineering, University of California San Francisco, San Francisco, CA Live imaging of genome has offered important insights into the dynamics of the genome organization and gene expression. The demand to image simultaneously multiple genomic loci has prompted a flurry of exciting advances on multi‐color CRISPR‐imaging methods using Cas9 protein orthologues or modified single‐guide RNA (sgRNA) scaffolds that recruit different RNA‐binding proteins. Here we introduce an approach to achieve highly multiplexed live imaging. After live imaging of multiple genomic loci using one‐color CRISPR system of one Cas9 protein and multiple sgRNAs, we perform rapid sequential rounds of DNA‐FISH to resolve loci identities. The staining is essentially complete in 1.5 minutes and signal is detectable in as short as 0.5 min with optimized DNA‐FISH protocol, contrary to the common practice of overnight incubation. We resolve 7 genomic loci including telomere and centromere in rapid sequential


rounds and demonstrate consistent and reversible staining and washing in up to 20 rounds. We have developed a correlation‐based algorithm to faithfully register between live images and FISH images and successfully resolve genomic loci identity with sequential DNA‐FISH after multiplex live imaging of four genomic loci with one‐color CRISPR system. The method frees up other color channels in live imaging. Furthermore, the method to register faithfully between live and fixed imaging is directly transferrable to other systems such as multiplex RNA imaging with RNA‐FISH and multiplex protein imaging with antibody‐staining.

M88 Phase separation drives heterochromatin domain formation. A.R. Strom1,2, A. Emelyanov3, D. Fyodorov 3, G.H. Karpen1,2; 1Biological Systems and Engineering, Lawrence Berkeley National Laboratory, Berkeley, CA, 2Molecular and Cell Biology, University of California, Berkeley, Berkeley, CA, 3Department of Cell Biology, Albert Einstein College of Medicine, Bronx, NY Recent studies revealed how biological phase separation can create distinct cellular compartments without bounding membranes. Here we use Drosophila and mammalian cells, and in vivo and in vitro approaches, to test the hypothesis that an essential chromatin compartment (heterochromatin) is organized by phase separation. We observe that, similar to other phase separated systems, heterochromatin domains 1) nucleate, grow and fuse in a liquid‐like fashion, 2) display phase‐like molecular dynamics, including reduced diffusion and coordinated movement at the domain boundary, and 3) the domain boundary excludes inert probes (selective permeability). Further, purified Drosophila Heterochromatin Protein 1a (HP1a) can form phase droplets in vitro. We conclude that heterochromatin domains form via phase separation, and that accessibility is predominantly determined by phase interactions. We propose that emergent biophysical properties associated with phase separated systems are important to understand how heterochromatin, and perhaps other chromatin domains, form and regulate essential nuclear functions.

M89 Resolving the molecular architecture of the NPC with 3D super‐resolution fluorescence microscopy. V. Jimenez‐Sabinina1, M. Bates2, A.B. Szymborska3, B. Nijmeijer1, S. Mosalaganti4, M. Beck4, S.W. Hell2, J. Ellenberg1; 1Cell Biology and Biophysics Unit, European Molecular Biology Laboratory, Heidelberg, Germany, 2NanoBiophotonics Department, Max Planck Institute for Biophysical Chemistry, Göttingen, Germany, 3Max Delbrück Center for Molecular Medicine, Berlin, Germany, 4Structural and Computational Biology Unit, European Molecular Biology Laboratory, Heidelberg, Germany Nuclear pore complexes (NPCs) are large macromolecular machines that mediate all the traffic between the nucleus and the cytoplasm. In vertebrates, each NPC consists of several hundred proteins, termed nucleoporins (Nups), and has a mass of over 110 MDa. Over the past years, the integration of multiple techniques including cryo‐electron tomography, X‐ray crystallography, mass spectrometry, and recently also super‐resolution microscopy1 has revolutionized our understanding of the NPC structure. Nevertheless, many details of the structure still remain elusive due to the enormous size, complexity and conformational dynamics of the NPC. Here, we present an approach for unraveling the three‐dimensional (3D) architecture of the NPC by combining 3D super‐resolution imaging of stochastically switched fluorophores with a 4Pi detection scheme and computational 3D single particle averaging,


which can resolve the structure of the NPC with molecular specificity and nano‐scale resolution. This method will allow the reconstruction of a 3D molecular map of the NPC by interrogating the spatial positioning of nucleoporins endogenously fluorescence tagged by genome editing. This approach allows us to address also dynamic parts of the NPC that have been inaccessible for atomic resolution methods and opens the exciting possibility to observe NPC structure in different functional or assembly states. 1. Szymborska, A. et al. Nuclear pore scaffold structure analyzed by super‐resolution microscopy and

particle averaging. Science 341, 655–8 (2013)

M90 Live cell imaging reveals dynamic sister chromatid linkage at individual genomic loci. R. Stanyte1, J. Nuebler2, R. Stocsits3, A. Schleiffer3, L. Mirny2, J. Peters3, D.W. Gerlich1; 1Institute of Molecular Biotechnology of the Austrian Academy of Sciences (IMBA), Vienna Biocenter (VBC), Vienna, Austria, 2Institute for Medical Engineering and Sciences, and Department of Physics , Massachusetts Institute of Technology (MIT), Cambridge, MA, 3Research Institute of Molecular Pathology (IMP), Vienna Biocenter (VBC), Vienna, Austria Faithful chromosome segregation requires physical linkage of sister chromatids from the time point of DNA replication until anaphase onset. The linkage between sister chromatids is mediated by the cohesin complex, which forms a ring that topologically embraces two DNA strands. Cohesin enriches at specific genomic sites, where it contributes to the formation of intra‐chromatid DNA loops that govern transcriptional gene regulation. To which extent the genomic enrichment sites of cohesin also determine the positions of sister chromatid linkage is not known. The cohesin ring is capable of sliding along DNA laterally, which might lead to a pool of cohesin that is not reproducibly positioned within a cell population and thus undetectable by conventional chromatin immunoprecipitation and sequencing (ChIP‐seq) approaches. To determine the sites of sister chromatid cohesion and to study its dynamics, we generated a collection of human cell lines with fluorescently labeled genomic loci, using EGFP‐tagged dCas9 targeted by different sgRNAs. Using automated live‐cell microscopy, we found that individual loci of sister chromatids were not persistently tethered from DNA replication until mitosis, but rather transiently split and merged multiple times. Unexpectedly, the linkage probability of the labeled genomic loci did not correlate with their distance from genomic cohesin enrichment sites previously determined by ChIP‐seq. Nevertheless, RNAi experiments confirmed that sister chromatid linkage strictly depended on the presence of cohesin. Our data suggest that persistent cohesion of whole sister chromatids emerges from multiple dynamic linkage sites.

Minisymposium 9: Chromosome Segregation Mechanisms M91 Epigenetic specification of rapidly evolving centromeres. L. Rosin1, J. Palladino1, A. Chavan1, B. Mellone1,2; 1Molecular and Cell Biology, University of Connecticut, Storrs, CT, 2Institute for Systems Genomics, University of Connecticut, Storrs, CT Centromeric chromatin represents a paradigm of epigenetic regulation of a genomic locus because the presence of the centromeric histone CENP‐A is both necessary and sufficient to confer centromere activity. Centromeres mediate the critical and conserved function of segregating chromosomes, yet centromeric DNA and CENP‐A are rapidly evolving. The observation that loop 1 (L1) of CENP‐A is under positive selection in Drosophila led to the proposal that this positive selection modulates CENP‐A DNA‐


binding preferences to neutralize the transmission advantage resulting from centromeric satellite expansion (centromere drive). Consistent with this model, CENP‐A from D. bipectinata fails to localize to D. melanogaster centromeres due to amino acid differences within L1. We show that the co‐expression of CENP‐A and its assembly factor CAL1 from D. bipectinata into D. melanogaster cells restores centromeric localization. Through a series of chimeric constructs, we identify the first 40 residues of CAL1 as critical for the interaction with rapidly evolving L1 of CENP‐A, demonstrating that so long as the CENP‐A/CAL1 pair is compatible, CENP‐A from any species can be deposited at mel centromeres. Collectively, our data show that L1 mediates CAL1 recognition and that CAL1 evolves adaptively with CENP‐A.

M92 The CENP‐A nucleosome bound by CENP‐C and CENP‐N is the fundamental unit for maintaining centromere identity. L.Y. Guo1, L. Zandarashvili1, K.L. McKinley2, N. Sekulic1, D. Fachinetti3, D.W. Cleveland3, I.M. Cheeseman2, B.E. Black1; 1Biochemistry and Biophysics, University of Pennsylvania, Philadelphia, PA, 2Whitehead Institute for Biomedical Research and Department of Biology, Massachusetts Institute of Technology, Cambridge, MA, 3Ludwig Institute for Cancer Research and Department of Cellular and Molecular Medicine, University of California, San Diego, La Jolla, CA Maintaining centromere identity relies upon the persistence of the epigenetic mark provided by the histone H3 variant, CENP‐A, but the molecular mechanisms that underlie its remarkable stability remain unclear. Here, we define the contributions of each of the three candidate CENP‐A nucleosome‐binding domains (two on CENP‐C and one on CENP‐N) to CENP‐A stability using gene replacement and rapid protein degradation strategies. Surprisingly, the most conserved domain, the CENP‐C motif, is completely dispensable. Instead, the stability is conferred by the unfolded central domain of CENP‐C and the folded N‐terminal domain of CENP‐N that becomes rigidified upon engaging the CENP‐A nucleosome. Disrupting the ‘arginine anchor’ on CENP‐C for the nucleosomal acidic patch disrupts the CENP‐A nucleosome structural transition and removes CENP‐A nucleosomes from centromeres. CENP‐A nucleosome retention at centromeres requires an intact Core Centromeric Nucleosome Complex (CCNC) where CENP‐C clamps down a stable nucleosome conformation and CENP‐N tethers CENP‐A to the DNA.

M93 Elucidating the regulation of kinetochore assembly using a de novo kinetochore assay. J. Lang1, A. Barber1, S. Biggins1; 1Division of Basic Sciences, Fred Hutchinson Cancer Research Center, Seattle, WA The kinetochore is a megadalton structure that assembles on centromeric chromatin to mediate chromosome segregation. It contains multiple copies of greater than 35 unique components that assemble into a complex that can bind to microtubules and control the cell cycle via the spindle checkpoint. Although the kinetochore components are known, the mechanism and regulation by which it assembles into a functional kinetochore is not well understood. To elucidate the mechanisms that control kinetochore assembly, we have developed a cell‐free method for kinetochore assembly. Budding yeast centromeric DNA is attached to beads and incubated in whole cell extracts to allow kinetochore assembly. This method generates kinetochores that contain all kinetochore components and exhibit Ndc80‐dependent MT‐binding activity. The assembly of the outer kinetochore depends on inner kinetochore components. Furthermore, assembly is dependent on the centromeric histone


chaperone. To determine the role of mitotic kinases in kinetochore assembly, we analyzed assembly in extracts prepared from various kinase mutants. We found that Aurora B‐mediated phosphorylation of Dsn1 promotes kinetochore assembly, and there are additional unknown roles for Aurora B phosphorylation in kinetochore assembly. We have also identified a previously unknown role for the Psh1 ubiquitin ligase that targets the centromeric histone variant in regulating kinetochore assembly. Taken together, this assay has revealed previously unknown regulatory events required to assemble the budding yeast kinetochore.

M94 Sex‐specific regulation of kinetochores distinguishes sperm meiotic chromosome segregation. V. Cota1, J. Wu2, G. Fabig3, J.A. Powers4, T. Cintra1, L. Mateo1, L. Quintanilla1, M.A. Monroy1, T. Müller‐Reichert3, D.S. Chu1; 1Biology, San Francisco State Univeristy, San Francisco, CA, 2Department of Clinical Laboratory Sciences and Medical Biotechnology, National Taiwan University, Taipei, Taiwan, 3Center for Regenerative Therapies, TU Dresden, Dresden, Germany, 4University of Indiana, Bloomington, IN Though chromosome segregation is a fundamental feature of cell division, different cell types adapt core segregation machinery to produce specialized cells. During meiosis, for example, spermatocytes divide symmetrically twice to produce four haploid sperm using centrosome‐organized spindles whereas oocytes divide asymmetrically twice to generate one haploid cell using acentriolar spindles. Though meiosis is a fundamental process, sex‐specific regulators that distinguish chromosome segregation to produce sperm or oocytes are largely unknown. Using C. elegans, we find that the kinetochore, which links microtubules to chromosomes, has sperm‐specific composition and localization during sperm meiosis compared to oocyte meiosis or mitosis. Further, though kinetochore proteins are disassembled and dispensable for oocyte meiotic anaphase, outer kinetochore proteins, like KNL‐1 and NDC‐80, are retained on chromosomes during sperm anaphase. Other kinetochore‐associated components, like the microtubule stabilizer CLASP (CLS‐2), that transition to the central spindle region to help separate chromosomes during oocyte meiotic anaphase and mitosis, remain chromosome associated in sperm meiotic anaphase. Live‐imaging of chromosome and spindle pole dynamics reveals that sperm chromosome movement is powered both by movement towards spindle poles as well as via pole‐pole separation. Taken together, our data suggest that, unlike oocyte meiosis, sperm meiotic chromosome separation relies primarily on microtubule attachment to kinetochore proteins to pull chromosomes apart. We have identified two sperm‐specific phosphatases, GSP‐3/4, as sex‐specific regulators of chromosome‐associated kinetochore localization dynamics. GSP‐3/4 localize to sperm meiotic chromosomes in a pattern similar to that of the kinetochore. GSP‐3/4 are required during sperm meiosis for timely completion of anaphase 1 and the separation of sister chromatids in meiosis II. Importantly, the loss of function of GSP‐3/4 results in CLASP mislocalization from chromosomes to the midzone during sperm meiotic anaphase I and loss of CLASP during meiosis II. This suggests that GSP‐3/4 are required for the sperm‐specific localization dynamics of kinetochore‐associated proteins between meiotic divisions. Taken together, these results define sex‐specific molecular features of meiotic chromosome segregation and identify the GSP‐3/4 sperm‐specific phosphatases as key regulators that distinguish chromosome segregation in sperm from that of other cell types.


M95 Kif18A coordinates the alignment and attachment of chromosomes during cell division. H. Kim1, C. Fonseca1, J.K. Stumpff1; 1Molecular Physiology and Biophysics, University of Vermont, Burlington, VT Chromosomes are spatially confined at the spindle equator just prior to chromosome segregation. During metaphase chromosome alignment, the attachments between kinetochores and dynamic spindle microtubules also increase. Chromosome alignment and stabilization of kinetochore‐microtubule attachments must be temporally coordinated to ensure proper chromosome segregation, but the molecular mechanisms that link these two processes are not understood. Previous studies have shown a motor protein within the kinesin‐8 family, Kif18A, to be absolutely essential for alignment of all chromosomes during metaphase (Zhu et al . Mol Biol. Cell 2005, Mayr et al. Curr. Biol. 2007, Stumpff et al . Mol Cell, 2008). Interestingly, in addition to chromosome alignment defects, we observe that certain cell types, including germline precursor cells (Czechanski, et al. Dev Biol. 2015), HeLa (Zhu et al , Mayr et al , Stumpff et al.), and triple negative breast cancer (TNBC) (unpublished) cells, display mitotic arrest when Kif18A is depleted. Therefore, we hypothesized that Kif18A may play a role in regulating kinetochore‐microtubule attachments. Our data indicate that Kif18A promotes dephosphorylation of Hec1, which is a part of the Ndc80 complex that directly contacts the microtubule lattice within the spindle. Furthermore, we find that a direct interaction between the Kif18A C‐terminus and protein phosphatase 1(PP1) is required for Kif18A’s role in Hec1 desphosphorylation. However, we find that the Kif18A‐PP1 interaction is not required for Kif18A’s well characterized role in chromosome alignment, suggesting that chromosome alignment and spindle attachment can be separated. Additionally, we see that mitotic arrest can be rescued by expressing a Hec1 variant that mimics a low‐phosphorylation state, supporting the notion that Kif18A‐dependent Hec1 dephosphorylation is necessary for efficient mitotic progression. These data support a model in which Kif18A has two independent functions at kinetochore microtubule plus‐ends that allow the motor to temporally coordinate chromosome alignment and kinetochore‐microtubule attachment.

M96 The mammalian kinetochore independently regulates its passive and active interfaces with microtubules. A.F. Long1, D.B. Udy1, S. Dumont1,2; 1Cell and Tissue Biology, University of California San Francisco, San Francisco, CA, 2Cell and Molecular Pharmacology, University of California San Francisco, San Francisco, CA The kinetochore links chromosomes to dynamic spindle microtubules and drives chromosome segregation at cell division. Although we know many of the mammalian kinetochore’s key molecules, our understanding of its mechanics is limited. We do not know what interfaces it uses to hold on to growing and shrinking microtubules and how these interfaces are regulated to support different mitotic functions. Using laser ablation to trigger cellular pulling forces on kinetochores, we find that phosphorylation of the load‐bearing protein Hec1 controls both the magnitude and timescale of a sister kinetochore pair’s mechanical responsiveness to spindle forces. Analysis of both sisters’ responses, and microtubule photomarking under normal spindle forces, reveal that Hec1 phosphorylation tunes passive friction along polymerizing microtubules. In contrast, we show that Hec1 phosphorylation does not affect the kinetochore’s grip on depolymerizing microtubules, or its ability to switch to this active force generating state. Together, these data suggest that different interfaces engage with growing and


shrinking microtubules, and that passive friction can be regulated without disrupting active force generation – giving the cell independent control of these two key functions during mitosis.

M97 Mechanism of Cdt1‐mediated control of kinetochore microtubule attachment stabilization by the Ndc80 Complex. S. Agarwal1, S. Seshadrinathan1, J.G. Cook2, R.J. McKenney3, D. Varma1; 1Cell and Molecular Biology, Feinberg School of Medicine, Northwestern University, Chicago, IL, 2Biochemistry and Biophysics, UNC Chapel Hill School of Medicine, Chapel Hill, IL, 3Molecular and Cellular Biology, University of California at Davis, Davis, CA We have previously demonstrated that the DNA replication licensing protein, Cdt1, has a novel function during mitosis in chromosome alignment and segregation. Cdt1 is recruited to mitotic kinetochores by the Ndc80 complex with the help of its loop domain, where it facilitates the function of the Ndc80 complex in stabilizing kinetochore microtubule (kMT) attachments. High‐resolution microscopy studies supported a scenario where Cdt1 bound to the Ndc80 complex enabled an extended conformation of the complex in the presence of microtubules (MTs) which in turn could aid stronger kMT attachment formation by the complex. Here we provide data that explains the molecular basis of Cdt1’s role in kMT attachment stabilization during mitosis by demonstrating that Cdt1 is a novel microtubule‐binding protein. In HeLa cells, ectopically expressed Cdt1 localized to and bundled MTs in metaphase and interphase, respectively. In vitro MT co‐pelleting assays show that purified Cdt1 binds to MTs with an apparent kD of ~ 1.0 µM. Single molecule TIRF microscopy assays and TEM studies substantiated a direct interaction between Cdt1 and MTs. Deletion mapping revealed that the C‐terminal region of Cdt1 encompassing a winged‐helix domain but not the N‐terminus binds to MTs; however, the entire protein was found to be required for efficient binding. Recently, the Ska complex was also shown to utilize a winged‐helix domain to bind to MTs, which suggests the emergence of this domain, traditionally considered to be a DNA‐binding domain, as a novel MT‐binding domain. Subcellular localization studies reveal that the Ska complex has an important role in Cdt1 recruitment to kinetochores, but Cdt1 on the other hand is not required for Ska recruitment. Previous work has revealed that Cdt1 is extensively phosphorylated during mitosis by proline‐directed kinases (MAPKs and CDKs). We performed rescue of function experiments in stable cell lines expressing siRNA‐resistant wild type, non‐phosphorylatable and phosphomimetic versions of Cdt1 after the knockdown of endogenous Cdt1. We find that cells expressing phosphomimetic Cdt1 exhibited a prominent mitotic arrest with marked loss of spindle MT‐stability. Current studies in the lab are focused on understanding how mitotic phosphorylation influences the function of Cdt1 in kMT attachments. In summary, our work suggests that Cdt1 bound to the Ndc80 loop domain provides an orthogonal MT‐binding site for the Ndc80 complex in addition to the established MT‐binding site at the N‐terminus of the Hec1 subunit. We hypothesize that the Ndc80‐Cdt1 complex likely has a more stable attachment to MTs than either protein alone, facilitating stable kMT attachments during mitosis.

M98 The KMN Network has Essential Post‐Mitotic Functions During Embryo Morphogenesis. D.K. Cheerambathur1, B. Prevo1, S. Wang1, R.A. Green1, K. Oegema1, A.B. Desai1; 1Cellular and Molecular Medicine, Ludwig Institute for Cancer Research/UCSD, La Jolla, CA The conserved 10‐subunit KMN network, comprised of the Knl1, Mis12 and Ndc80 complexes, constitutes the primary microtubule attachment module at the kinetochore during cell division. The


Ndc80 subunit of the Ndc80 complex harbors the direct microtubule binding activity that is critical to form the dynamic, load‐bearing interface with spindle microtubules that segregates chromosomes. While investigating Ndc80 function in C. elegans embryos, we unexpectedly identified essential post‐mitotic roles for this protein network in morphogenesis during late embryonic development. This discovery initiated from analysis of a mutant form of Ndc80 lacking its basic N‐terminal tail, which contributes substantially to microtubule‐binding affinity in vitro. Surprisingly, no chromosome segregation or kinetochore‐microtubule attachment defect is observed for the Ndc80 N‐tail deletion mutant in vivo, in both direct imaging assays and a sensitive reporter assay that monitors low rates of chromosome loss. However, the N‐tail deletion failed to rescue an ndc‐80 null mutant and exhibited a variety of phenotypes consistent with defective morphogenesis. Endogenous locus tagging of subunits of all 3 subcomplexes of the KMN network revealed robust localization in multiple post‐mitotic tissues, including dendritic extensions of sensory neurons. To investigate potential functions of the KMN network in morphogenesis, we focused on the nervous system, employing a recently developed means to degrade GFP fusions in a tissue‐specific post‐mitotic manner in C. elegans (Wang et al. eLife 4:e08649). Using this approach to target Knl1, Ndc80 or the Mis12 complex revealed significant defects in neuronal cell body positioning as well as phenotypes on the organismal level such as egg‐laying and behavioral defects that indicate a contribution of the KMN network in proper patterning and function of the nervous system. The localization of KMN network constituents is transient during embryonic morphogenesis, and overlaps with microtubule arrays suggesting that these components may act in these distinct contexts in a similar capacity as at kinetochores during cell division. Current experiments are extending analysis in this non‐mitotic context to other kinetochore components, assessing the effect of KMN network inhibitions in other post‐mitotic tissues, as well as determining the effect of mechanism‐guided mutations in KMN subunits that we have already characterized extensively in chromosome segregation. Overall, our results highlight an essential non‐division role for the KMN network and provide a striking example of the repurposing of evolutionarily ancient and essential cellular machinery to a new context in metazoan development.

M99 A SUMO‐dependent protein network regulates meiotic chromosome congression in C. elegans oocytes. F.G. Pelisch1, T. Tammsalu1, B. Wang1, A. Gartner1, R.T. Hay1; 1Centre for Gene Regulation and Expression, University of Dundee, Dundee, United Kingdom The process of meiosis is at the core of reproductive biology as it is responsible for the production of haploid gametes. In meiosis, a single round of DNA replication is followed by two consecutive segregation steps. Unlike mitotic spindles, little is known about the assembly of the meiotic spindle and how this relates to chromosome dynamics. During Caenorhabditis elegans oocyte meiosis, lateral microtubule bundles promote chromosome congression through the action of a protein complex localised between homologous chromosomes in a ring pattern (ring complex, RC). This RC contains the kinases AIR‐2/Aurora B and BUB‐1, and the kinesin KLP‐19, which provides the plus‐end‐directed force required for chromosome congression. While some RC components are known as well as their localisation interdependence, the mechanism(s) that govern RC assembly and regulation have remained elusive. We show here that assembly of the RC starts with the recruitment of the chromosomal passenger complex (CPC) components AIR‐2/Aurora B and ICP‐1/INCENP during oocyte maturation. A wave of SUMO modification is triggered during fertilisation and is required for CPC function and BUB‐1


recruitment to the RC. The SUMO E3 ligase GEI‐17 is part of the RC and triggers the sumoylation and recruitment of the kinesin KLP‐19. A remarkable feature is the degree to which SUMO modification is constrained within space and time: within a 30 minute period there are two waves of SUMO modification and deconjugation of the RING complex which parallels it assembly and disassembly. During this regulation, the neighboring kinetochores are not affected. Thus, our results highlight the relevance of sumoylation and non‐covalent SUMO interaction networks in regulating highly dynamic and specific protein complex assembly, in the context of female meiosis in C. elegans.

M100 An actin spindle protects mammalian oocytes against chromosome segregation errors. B. Mogessie1, M. Schuh1; 1Meiosis, Max Planck Institute for Biophysical Chemistry, Göttingen, Germany The accurate segregation of chromosomes during female meiosis is a prerequisite for the formation of a healthy embryo upon fertilization. However, chromosome segregation in human oocytes is highly error‐prone and a leading cause of miscarriages, infertility and Down’s Syndrome. The mechanisms that safeguard accurate chromosome segregation in mammalian oocytes are only poorly understood. Here, we reveal an unexpected actin‐dependent mechanism that protects oocytes against aneuploidy. We found that an actin spindle permeates the acentrosomal microtubule spindle in oocytes of several mammalian species, including humans. We demonstrate that actin promotes the accurate alignment and segregation of chromosomes. Oocytes in which the actin spindle in disrupted display a significantly increased number of misaligned chromosomes as well as lagging chromosomes during meiosis I and II. We show that actin safeguards chromosome alignment and segregation by promoting the formation of kinetochore fibres, which mediate the alignment and segregation of chromosomes. These findings establish an essential role for actin in meiotic chromosome segregation and suggest a new principle of how kinetochore fibres are established.

Minisymposium 10: Connecting Cells to Tissues

M101 Increasing β‐catenin/Wnt3A Activity Levels Drive Mechanical Strain‐Induced Epithelial Cell Cycle Progression through Mitosis. B.W. Benham‐Pyle1, J. Sim2, K.C. Hart3, B.L. Pruitt2, W.J. Nelson3; 1Cancer Biology, Stanford University, Stanford, CA, 2Mechanical Engineering, Stanford University, Stanford, CA, 3Biology, Stanford University, Stanford, CA Activation of β‐catenin transcriptional activity is a downstream response to mechanical force and Wnt signaling that promotes cell proliferation and tissue expansion. However, it is not known whether mechanical force and Wnt signaling act independently or can synergize to activate β‐catenin signaling and cell division. Here, we show that mechanical strain induced a Src‐dependent phosphorylation of Y654 β‐catenin and β‐catenin accumulation in the cytoplasm and nucleus, resulting in increased β‐catenin transcriptional activity. Under these conditions, however, cells accumulated in S/G2 (in the absence of DNA damage) but did not enter mitosis. Blocking cytoplasmic β‐catenin degradation by inhibition of Casein Kinase I or addition of Wnt3A increased β‐catenin‐mediated transcriptional activity and the accumulation of cells in S/G2 following mechanical strain. Significantly, only the combination of both mechanical strain and Wnt/β‐catenin pathway activation triggered those cells in S/G2 to transition


into mitosis. These results indicate that mechanical strain‐induced Src phosphorylation of β‐catenin and Wnt pathway‐dependent stabilization of cytoplasmic β‐catenin synergize to increase β‐catenin transcriptional activity to a level that is required to drive cell cycle progression through mitosis. Thus local Wnt signaling may fine‐tune the effects of global mechanical strain on β‐catenin signaling to restrict cell divisions in order to maintain tissue homeostasis.

M102 Imaging How Cells Choose their Fate, Shape and Position in the Early Mouse Embryo. N. Plachta1; 1Institute of Molecular and cell Biology, A*STAR , Singapore, Singapore We recently established single‐cell imaging technologies to show how the early mouse embryo forms. We applied new fluorescence correlation spectroscopy (FCS) methods to demonstrate how transcription factors bind to DNA in live embryos. We discovered that differences in the binding of Sox2 to its specific DNA sites 1) appear as early as the 4‐cell stage, 2) are regulated by histone modifications and 3) predict cell fate in vivo. We also found that as cells choose their distinct fates, they extend long filopodia protrusions to pull their neighbor cells closer, revealing a new mechanism for embryo compaction. Finally, we designed new membrane segmentation and laser ablation methods to show how differences in the action of mechanical tensile forces drive some cells inside the embryo to form the pluripotent inner mass. Our findings reveal dynamic mechanisms controlling how mammalian cells function in real time and in vivo. Recent references: White et al (2016) Cell Samarage et al (2015) Dev. Cell Fierro‐Gonzalez et al (2013) Nat. Cell Biol.

M103 The spatiotemporal limits of developmental Erk signaling. H.E. Johnson1, Y. Goyal2, N. Pannucci1, G.M. Schupbach1, S.Y. Shvartsman2, J.E. Toettcher1; 1Molecular Biology, Princeton University, Princeton, NJ, 2Chemical Engineering, Princeton University, Princeton, NJ Animal development is characterized by a series of signaling events that occur at precise locations and times within the embryo. It has proven a longstanding challenge of developmental biology to determine when and where this precision is required or dispensable for proper development. Here, we describe a light‐inducible system to activate Ras in the early Drosophila embryo that can be used to tune Erk activity with subcellular spatial and minutes‐timescale precision to levels not achievable with any known mutant. We find that proper development is remarkably robust to perturbations in the level and spatial range of Erk activity, except from 1 to 4 hours post fertilization where the embryo is robust to variations in the level but not spatial range of Erk. Our findings suggest that information is not encoded in the precise level of Erk activation, implying that during tissue patterning Erk is more likely to function as a switch than a rheostat. This work highlights the power of optogenetics for revealing how signaling pathway activity is interpreted during development.


M104 Motile stem cells exhibit tissue‐level spatial order during homeostasis but not growth of the adult Drosophila midgut. X. Du1, J. Martin1, S. Balachandra1, S. Ngo1, I. Riedel‐Kruse1, L.E. O'Brien1; 1Molecular and Cellular Physiology, Stanford University, Stanford, CA Many self‐renewing, solid organs contain stem cell populations that are dispersed throughout the expanse of the tissue. Spatial dispersal ensures that each tissue field contains stem cells. How dispersed stem cells maintain proper spacing during organ renewal is unknown. Here we combine quantitative spatial analyses with long‐term live imaging to investigate the underlying basis of stem cell spacing. Comprehensive statistical maps of stem cell‐stem cell distances reveal that stem cells are spatially ordered during steady‐state organ renewal, but not during adaptive organ growth. Overnight imaging of midguts within living animals, together with tissue‐wide cell tracking and spatio‐temporal analysis of individual cell displacements, shows that stem cells are autonomously motile. Individual cells, propelled by actin‐rich protrusions, flatten and slither beneath the basal epithelium. At the population level, stem cell motility appears stochastic in frequency and direction. Progenitor‐specific depletion of motility factors disrupts the spatial ordering of stem cells, implying that motility is needed for tissue‐level dispersal. Our findings show that spatial order is a hallmark of steady‐state but not expanding stem cell populations and suggest that autonomous motility is involved in stem cell distribution during tissue homeostasis.

M105 Asymmetric microtubule activity facilitates differential epigenetic inheritance in Drosophila male germline stem cell. R. Ranjan 1, X. Chen1; 1Department of Biology, Johns Hopkins University, Baltimore, MD Due to the crucial role that epigenetic mechanisms play in regulating cell identity and function, it has been a long‐standing question whether and how stem cells maintain their epigenetic memory through many cell divisions. We found that during asymmetric cell division of Drosophila male germline stem cell (GSC), histone H3 (H3) becomes asymmetrically segregated — the “old” H3 is retained in the GSC while the “new” H3 is enriched in the differentiating daughter cell. Recently, we also found that randomized H3 segregation pattern correlates with both GSC loss and progenitor germline tumor phenotypes, suggesting that asymmetric H3 inheritance is required for both GSC maintenance and proper differentiation. We propose that old and new H3 are asymmetrically deposited to sister chromatids, and mitotic machinery recognizes this asymmetry for differential segregation. To understand this, we are using high‐resolution time‐lapse movie to visualize spatial and temporal regulation of sister chromatids segregation in GSCs. We have obtained promising data showing asymmetric microtubule activities in GSC, which first interact with the nuclear membrane at the GSC side during G2 phase, and then start “poke‐in” and make a hole to interact with chromatin during G2/M phase. Interestingly, during the G2 phase, centromeres cluster near the nuclear membrane at the GSC side, which is likely to secure attachment between “old” H3‐enriched sister chromatids and microtubules emanating from mother centrosome at the GSC side. This occurs prior to the attachment of “new” H3‐enriched sister chromatids to microtubules from the daughter centrosome at the differentiating daughter cell side. Coincidentally, we also observed that sister chromatids enriched with old vs. new H3 are condensed in a temporally separable manner: the old H3‐enriched sister chromatids condense prior to new H3‐enriched sister chromatids. Together, asymmetric microtubule activity, centromere cluster, and differential sister


chromatids condensation act together to ensure their preferential attachment and segregation. By contrast, such an asymmetry in the microtubule activity, centromere cluster or differential sister chromatid condensation is not observed in symmetrically dividing progenitor germ cells. We propose that the GSC microtubule has a novel function to ensure that the old H3‐enriched set of chromatids is preferentially attached by the microtubule at the stem cell side. The new H3‐enriched chromatids are attached at a later time by microtubules emanating from the centrosome at the differentiating daughter cell side. This proposed function provides an unappreciated role of the microtubule in selectively recognizing and segregating sister chromatids based on their differential epigenetic information.

M106 Tissue‐level forces regulate Wnt/β‐catenin‐dependent mesoderm differentiation within self‐organized regions of human embryonic stem cell colonies. J.M. Muncie1,2, L. Przybyla1, J.N. Lakins1, R. Sunyer3, X. Trepat3, V.M. Weaver1,4; 1Center for Bioengineering and Tissue Regeneration, Department of Surgery, University of California San Francisco , San Francisco, CA, 2Joint Graduate Group in Bioengineering, University of California San Francisco and University of California Berkeley, San Francisco, CA, 3Institute for Bioengineering of Catelonia (IBEC), Barcelona, Spain, 4Department of Anatomy, Department of Bioengineering and Therapeutic Sciences, Eli and Edythe Broad Center of Regenerative Medicine and Stem Cell Research, Helen Diller Family Comprehensive Cancer Center, University of California San Francisco, San Francisco, CA Mechanics regulate cell differentiation in culture and tissue development in vivo. Molecularly, we and others showed that force modulates cell and tissue differentiation by activating Wnt/β‐catenin signaling. Nevertheless, how mechanics activates Wnt in the context of a multi‐cellular tissue to direct cell fate remains poorly understood. During embryonic development, morphogen‐directed gastrulation orchestrates Wnt‐dependent mesoderm differentiation as cells of the epiblast ingress through the primitive streak. Using geometrical patterning of human embryonic stem cells (hESCs) on polyacrylamide hydrogels to model the microenvironment and tissue geometry of gastrulation‐stage embryos, we observed that stabilization of adherens junctions on soft substrates (400 Pa) stimulates Wnt secretion and nuclear localization of β‐catenin to promote mesoderm differentiation. Conversely, culturing hESCs on a stiff substrate (60 kPa) inhibited Wnts, caused β‐catenin degradation, and limited mesoderm differentiation. Interestingly, we noted that the mesoderm differentiation of the hESCs on the soft substrates proceeded through primitive streak‐like regions that emerged in discrete regions localized within the peripheral margin of the hESC colonies. The primitive streak regions expressed T (brachyury), confirming their gastrulation‐like phenotype. Intriguingly, we quantified elevated cell‐cell forces within the circular primitive streak‐like regions at the periphery of the hESC colonies. Our findings revealed that upon induction of differentiation with defined morphogens, these high intercellular forces preceded the disassembly of E‐cadherin junctions and that the hESCs within these regions underwent an epithelial‐to‐mesenchymal transition to differentiate into mesoderm. Consistently, we observed intense Wnt expression and nuclear β‐catenin within the cells localized at the regions. Altering the geometry of the hESC colonies ablated localized force development and compromised mesoderm differentiation efficacy, emphasizing a critical role for tissue architecture in early development. Our data also illustrate how cell context can foster the self‐organization behavior of tissues to drive developmental processes including gastrulation. Specifically, we showed that hESC colonies with a defined disc architecture and specified diameter spontaneously organize to generate localized regions of high cell‐cell force that enhances mesoderm differentiation. Ongoing experimental are currently exploring how tissue geometry manifests


localized cell‐cell force generation using our in vitro hESC model as well as an in vivo model of chick gastrulation. Funded by CIRM RB507409 to VMW.

M107 Mouse intestinal crypt morphogenesis. K. Sumigray1, T.H. Lechler1; 1Dermatology, Duke University, Durham, NC The adult intestinal stem cell niche, termed crypts, are cup‐like invaginations of epithelia surrounded by mesenchyme. These structures are absent in neonates and develop within the first weeks of life. Our aim is to describe crypt morphogenesis, identify key players that drive the process, and ultimately determine whether the crypt architecture is an important component of the intestinal stem cell niche. We have performed a quantitative morphometric analysis of crypt development. We find that crypt progenitor cells first apically constrict, followed several days later by crypt invagination and lengthening. Using a genetic loss of function approach we show that type II nonmuscle myosins are essential for the initiation of crypt invagination. In the absence of myosin IIA/C, cells did not undergo crypt morphogenesis. To further identify candidate regulators of this process, we performed a kinetic RNA‐seq analysis through this morphogenesis. Several pathways were found to be active in crypt progenitors cells, including planar cell polarity and Rac GTPase signaling. Genetic disruptions to either of these pathways suggest that they indeed play key roles in the patterning, invagination and elongation stages of crypt morphogenesis. With these new tools, we can now assess whether the crypt architecture is a key component of the intestinal stem cell niche.

M108 The balance between cells with distinct spindle dynamic behaviors is maintained by negative feedback during airway tube morphogenesis. Z. Tang1, Y. Hu2, K. Jiang1, C. Zhan1, W.F. Marshall3, N. Tang1; 1 National Institute of Biological Sciences, Beijing, Beijing, China, 2Zhou Pei‐yuan Center for Applied Mathematics, Tsinghua University, Beijing, Beijing, China, 3Department of Biochemistry and Biophysics, University of California, San Francisco, San Francisco, CA The orientation of cell division plays a key role in controlling organogenesis. However, little is known about the regulation of cell division orientation at the whole organ level. We previously showed that lung tube morphogenesis is regulated by the ratio of two cell populations with different distributions of division axes. Here, by combining four‐dimensional imaging analysis and mathematical modeling, we show that these two populations correspond to two epithelial cell populations with distinct spindle dynamic behaviors during mouse lung development. One population of cells followed the classic “long‐axis rule” and quickly fixed their spindles during early metaphase. In contrast, the other cell population did not follow the “long‐axis rule”; rather, they kept randomly rotating their spindles throughout metaphase. We showed that mechanical signals play a role in regulating spindle dynamic behaviors and the organization of astral microtubules. Interestingly, we observed that the ratio of these two cell populations is under negative feedback regulation. This form of regulation maintains the balance between the two cell populations for ensuring a normal mitotic spindle angle distribution and thus a normal lung tube. Our findings demonstrated that there are multi‐level cellular processes controlling cells to coordinate their division orientations during airway tube morphogenesis.


M109 Cells help each other remember the past through cell‐cell communication and by collectively backing‐up information. D. Gomez‐Alvarez1, H. Youk1; 1Department of Bionanoscience, Kavli Institute of Nanoscience, Delft University of Technology, Delft, Netherlands Cellular memory, in which a transient stimulus triggers a lasting cellular response (e.g., lasting gene expression), has primarily been attributed to intracellular mechanisms like genetic circuits with positive feedbacks, and epigenetic and genetic modifications. But instead of being autonomous individuals, cells often exist as collectives. For example, cells inside metazoan tissues and microbial biofilms communicate with each other. Moreover, these cells must maintain their differentiated identities indefinitely after the disappearance of the original differentiation cue. If and how multicellular systems can use cell‐cell communication for cellular memory have been relatively underexplored. To reveal such multicellular mechanisms for cellular memory, we engineered yeasts that secrete a pheromone due to a transient presence of sugars. Sensing the pheromone activates the expression of a target gene. After the sugars disappear, the cells “remind” each other to permanently maintain the gene expression through extracellularly interlocked positive and negative feedback loops that sustain the pheromone secretion. Even if every cell forgets, the pheromone acts as an “extracellular backup” that reminds everyone. A mathematical model explains these features and limitations of unicellular memory that this multicellular memory resolves. This versatile mechanism for collective memory suggests how cells within biofilms and tissues might control their extracellular environment to permanently maintain their identities without de‐differentiating.

M110 Tissue origami: Directed folding of tissues by programmed cell contractility networks. A.J. Hughes1, M.C. Coyle1, J. Zhang1, Z.J. Gartner1; 1Pharmaceutical Chemistry, University of California, San Francisco, San Francisco, CA A major barrier to the engineering of tissues to study development is the difficulty in controlling tissue structure across a wide range of length scales. Further, proper function of synthetic biological tissues is hampered by mass transport limitations, and the lack of extracellular matrix (ECM) physicochemical anisotropies that guide appropriate cell behaviors in space and time. In biology, tissues are typically built over a range of length scales via self‐organized morphogenesis processes in which extracellular matrix remodeling occurs at the same time as cell division, migration, and fate specification. For example, the specification of endoderm cell fate at crypts in the gut is governed by cell‐cell signaling networks that are influenced by the folding state of tissue layers. We take a biomimetic approach to building tissues by engineering the spatial organization of sheets of ECM gels via contractile networks of cells. Such contractile networks are co‐patterned with “passenger” cell types that can be guided into prescribed 3D topologies that instruct cell self‐organization processes. Thus, the folding trajectory and form is combined with control over the type and placement of different cell types on the sheet, enabling control over 3D tissue architecture in concert with cellular interfaces. We build arrays of contractile tissues at the upper and lower surface of 250 micron‐thick matrigel‐collagen ECM‐mimetic gels using DNA‐patterned assembly of cells. The type and location of cells on each surface of the gel prescribe the final folded geometry of the ECM. Regular arrays of contractile mammary epithelial cells placed on one side of circular ECM gels yield cap‐shaped objects with maximum curvatures that increase over time. The curvature at a given timepoint


increases monotonically as the spacing between tissues on the sheet is decreased (and more tissues are added), scales linearly with cell contractility, and is partially blocked by blebbistatin. We found that anisotropic grids of fibroblast tissues produce consistent anisotropic folds along controlled axes, allowing us to fabricate a range of 3D topographies and predict them with a biophysical tensegrity model. For example, we constructed the Miura‐ori fold, which resembles folds of the gut epithelium in development. We report engineering control over the 3D shape of biomimetic ECM scaffolds by leveraging tissues as mechanical actuators. Our ability to design specific fold geometries in concert with cell compositions and spatial arrangement in ECM sheets could enable studies of the necessary conditions for topological transitions in animal development.

Minisymposium 11: Organelle Contact Sites and Biogenesis

M111 GPS2 regulates the expression of nuclear‐encoded mitochondrial genes through mitochondria retrograde signaling. M.D. Cardamone1, B. Tanasa2, C.T. Cederquist1, J. Huang1, K. Mahdaviani3,4, J. Orofino1, C. Lentucci1, M.G. Rosenfeld5, M. Liesa4, V. Perissi1; 1Biochemistry , Boston University, Boston, MA, 2Pediatrics, Stanford University, Stanford, CA, 3Medicine, Boston University, Boston, MA, 4Medicine, UCLA, Los Angeles, CA, 5Medicine, UCSD, La Jolla, CA A key step in the evolution of eukaryotic cells has been the acquisition of oxidative metabolism through the symbiotic incorporation of an aerobic prokaryote. As this primordial organism evolved into an intracellular organelle, now known as mitochondrion, most of its genetic information was transferred to the nuclear genome. While this implies that mitochondrial biogenesis, health and functionality, ultimately depend on the nucleus being informed of the metabolic needs of the cell and the status of existing mitochondria, much is still unknown about direct mitochondria‐to‐nucleus communication strategies, particularly in mammalian cells. Here, we have identified a key mediator of the mitochondria retrograde signaling pathway in the transcriptional cofactor G‐Protein Suppressor 2 (GPS2). Overall, our data reveals that GPS2 translocates from mitochondria to nucleus in a regulated manner to promote the activation of the nuclear‐encoded mitochondrial gene program through the establishment of a chromatin environment permissive for RNA polymerase II (POL2)‐mediated transcriptional activation. We will present in vitro data characterizing GPS2 function during adipocyte differentiation and in response to mitochondrial stress, as well as discuss its role in the regulation of brown adipose tissue metabolism in vivo. Together, our findings shed new light on the strategies evolved in eukaryotic cells to maintain efficient mitochondria:nuclear inter‐organellar communication, with important implications for our understanding of the molecular mechanisms underlying the regulation of mitochondria biogenesis and metabolic flexibility in both homeostasis and disease.

M112 ER‐mitochondria contacts couple mtDNA synthesis with mitochondrial division in human cells. S.C. Lewis1, L.F. Uchiyama1, J. Nunnari1; 1Molecular and Cellular Biology, University of California, Davis, Davis, CA Mitochondrial DNA (mtDNA) encodes RNAs and proteins critical for cell function. In human cells, hundreds to thousands of mtDNA copies are replicated asynchronously, packaged into protein‐DNA nucleoids, and distributed within a dynamic mitochondrial network. The mechanisms that govern how


nucleoids are chosen for replication and distribution are not understood. Mitochondrial distribution depends on division, which occurs at endoplasmic reticulum (ER)–mitochondria contact sites. These sites were spatially linked to a subset of nucleoids selectively marked by mtDNA polymerase and engaged in mtDNA synthesis—events that occurred upstream of mitochondrial constriction and division machine assembly. Our data suggest that ER tubules proximal to nucleoids are necessary but not sufficient for mtDNA synthesis. Thus, ER‐mitochondria contacts coordinate licensing of mtDNA synthesis with division to distribute newly replicated nucleoids to daughter mitochondria.

M113 Unraveling the mechanism of ER‐associated organelle fission. G. Voeltz1, M. Phillips1, P.J. Chitwood1; 1MCD Biology, University of Colorado, Boulder, CO ER membrane contact sites (MCSs) define the position where early and late endosomes undergo constriction and fission. ER tubules are recruited to the divide between cargo sorting domains on endosomes that will then constrict and undergo fission. Our aim was to purify ER factors that regulate MCS formation and fission in animal cells. Based on the identity of the candidate proteins, we hope to be able to test hypotheses about the mechanisms used by the ER to drive organelle constriction and division at MCSs. Our strategy to identify MCS proteins was to target a promiscuous biotin ligase (BirA) to ER‐associated fission sites. BirA has a modification reach to approximately 30nm, which is ideal for bridging membrane contacts (which have a typical spacer distance of ~5‐15nm). Our hybrid construct localizes as predicted and biotinylates ER proteins at MCSs in mammalian cells that we could purify and then identify by mass spec. This strategy has identified ER proteins that regulate ER‐associated endosome fission.

M114 Endosome‐ER contacts control actin nucleation and retromer function through VAP‐dependent regulation of PI4P. R. Dong1,2,3,4, Y. Saheki1,2,3,4, S. Swarup4, L. Lucast1,2,3,4, J.W. Harper5, P. De Camilli1,2,3,4,6; 1Department of Neuroscience, Yale University School of Medicine, New Haven, CT, 2Department of Cell Biology, Yale University School of Medicine, New Haven, CT, 3Howard Hughes Medical Institute, Yale University School of Medicine, New Haven, CT, 4Program in Cellular Neuroscience, Neurodegeneration and Repair, Yale University School of Medicine, New Haven, CT, 5Department of Cell Biology, Harvard Medical School, Boston, MA, 6Kavli Institute for Neurosciences, Yale University School of Medicine, New Haven, CT VAP (VAPA and VAPB in humans) is an evolutionarily conserved ER‐anchored protein that functions as an anchor for a multiplicity of cytosolic proteins at the ER membrane. One of its functions is to participate in the generation of membrane contact sites between the ER and other membranes, via its interaction with lipid transfer proteins that binds "in trans" adjacent membranes. These membrane contact sites have an important role in the control of lipid homeostasis within cells. Mutation in VAPB that disrupt binding with these lipid transfer proteins results in a dominant form of familial amyotrophic lateral sclerosis. Elucidating the impact of VAP loss of function may thus illuminate mechanisms of disease in addition to providing insight into fundamental aspects of lipid dynamics and cell physiology. Here we report that by regulating PI4P levels on endosomes, VAP affects WASH‐dependent actin nucleation on these organelles and the function of the retromer, a protein coat responsible for endosome‐to‐Golgi traffic. VAP is recruited to retromer budding sites on endosomes via an interaction


with the retromer SNX2 subunit. Cells lacking VAP accumulate high levels of PI4P, actin comets and trans‐Golgi proteins on endosomes. Such defects are mimicked by downregulation of OSBP, a VAP interactor and PI4P transporter that participates in VAP‐dependent ER‐endosomes tethers. These results reveal a role of PI4P in retromer/WASH dependent budding from endosomes. Collectively, our data shows how the ER can control budding dynamics and association with the cytoskeleton of another membrane by direct contacts leading to bilayer lipid modifications.

M115 RASSF4 regulates PIP2 and formation of ER‐plasma membrane junctions. Y. Chen1, C. Chang1, J. Liou1; 1Physiology, UT southwestern , DALLAS, TX RASSF4 (RAS association domain family 4) is involved in tumorigenesis and regulates the Hippo pathway. Here, we identify RASSF4 as a modulator of phosphatidylinositol 4,5‐ bisphosphate (PIP2) at the plasma membrane (PM). We first discovered that RASSF4 regulates store‐operated Ca2+ entry (SOCE) by affecting the translocation of the endoplasmic reticulum (ER) Ca2+ sensor STIM1 to ER‐PM junctions. It was further revealed that RASSF4 regulates the formation of ER‐PM junctions and the ER‐PM tethering function of extended synaptotagmins E‐Syt2 and E‐Syt3. Remarkably, knockdown of RASSF4 greatly reduced steady‐state PM PIP2 levels, whereas overexpression of RASSF4 significantly increased PM PIP2. An acute increase of PM PIP2 rescued the defects in ER‐PM junctions, STIM1 translocation, and SOCE in RASSF4 knockdown cells. Overall, our study reveals an important role of RASSF4 in maintaining PM PIP2 levels, pivotal for ER‐PM junctions, SOCE, and a plethora of physiological processes.

M116 The biogenesis of a mitochondria‐plasma membrane anchor. L.M. Kraft1, H.A. Ping1, W. Chen1, L.L. Lackner1; 1Molecular Biosciences, Northwestern University, Evanston, IL The mitochondria‐ER‐cortex anchor (MECA) is required for proper mitochondrial distribution and inheritance and functions by tethering mitochondria to the plasma membrane. The core component of MECA is the multi‐domain protein, Num1, which assembles into clusters at the cell cortex. We show Num1 adopts a highly extended, polarized conformation in which its N‐ and C‐termini can be spatially resolved by light microscopy. Its N‐terminal coiled‐coil domain (Num1CC) is proximal to mitochondria and the C‐terminal pleckstrin homology domain is associated with the plasma membrane. We find that Num1CC interacts directly with the mitochondrial membrane and displays a strong preference for the mitochondria‐specific phospholipid, cardiolipin. This direct membrane interaction is critical for MECA function. Thus, mitochondrial anchoring is mediated by a protein that interacts directly with two different membranes through lipid‐specific binding domains, suggesting a general mechanism for inter‐organelle tethering. The assembly of Num1 is critical for mitochondrial anchoring and likely functions to increase the avidity of Num1’s membrane binding domains for their target membranes. We find that the assembly of Num1 is regulated and dependent on an interaction with the mitochondrial membrane. Once assembled, the structures formed are stable and persistent. The assembly of Num1 is also critical for its role in anchoring dynein to the plasma membrane, where dynein‐mediated microtubule sliding functions to pull the nucleus into growing daughter cells. We find that dynein is anchored to the cell cortex by Num1 clusters that have been assembled by mitochondria. The mitochondrial‐dependent assembly of a dual‐


function cortical anchor provides a mechanism to integrate the positioning and inheritance of two essential organelles and expands the function of organelle contact sites.

M117 Identifying and characterizing contact sites between organelles. N. Shai1, S.G. Chuartzman1, L. Zada1, E. Zalckvar1, M. Schuldiner1; 1Molecular Genetics, Weizmann Institute of Science, Rehovot, Israel In order to optimize cellular functions, organelles must collaborate and communicate with the surrounding intercellular environment including other organelles. A common way of communication between organelles is through membrane contact sites where membranes of two organelles are tethered, facilitating close‐range interactions such as exchange of lipids, small molecules and intracellular signaling. In addition, contact sites are important for controlling processes such as metabolism, organelle trafficking, inheritance and division. The full extent of all organelle contact sites, as well as how different organelles rely on contact sites for their various cellular activities are far from being fully characterized. We created a split fluorescence reporter for contact sites in yeast by which one part of a fluorophore is fused to the outer membrane of one organelle while the other is fused to the outer membrane of another organelle. We have performed a wide variety of controls to ensure that only when the organelles are in very close proximity, such as that occurring at contact sites, a fluorescent signal is emitted and that the signal represents bona fide contact sites and not a random proximity between organelles. By using this method we created all pairwise combinations of organelles and were able to detect known contacts as well as several novel ones. Many of the new contacts that we identified occurred between peroxisomes and other organelles. We further focused on the peroxisome‐mitochondria (PerMit) contact, aiming to uncover its molecular tethers and regulators. To this end we used high content screens in which we examined how the PerMit reporter is affected by deletion or over expression of thousands of genes. We now further characterize the proteins identified as well as study the physiological role of the PerMit contact.

M118 ACBD5 and VAPB mediate membrane associations between peroxisomes and the ER. J.L. Costello1, I.G. Castro1, C. Hacker1, T.A. Schrader1, J. Metz1, D. Zeuschner2, A.S. Azadi1, L.F. Godinho1, V. Costina3, P. Findeisen3, A. Manner4, M. Islinger4, M. Schrader1; 1Biosciences, University of Exeter, Exeter, United Kingdom, 2Max‐Planck‐Institute for Molecular Biomedicine, Munster, Germany, 3Institute for Clinical Chemistry, University of Heidelberg, Mannheim, Germany, 4Institute of Neuroanatomy, University of Heidelberg, Mannheim, Germany Peroxisomes and the endoplasmic reticulum (ER) cooperate in cellular lipid metabolism and form tight structural associations, which were first observed in ultrastructural studies decades ago. ER‐peroxisomal associations have been suggested to impact on a diverse number of physiological processes including lipid metabolism, phospholipid exchange, metabolite transport, signalling, peroxisome biogenesis, and pexophagy. Despite the decades that have passed since peroxisome‐ER associations were first observed, and despite their fundamental importance to cell metabolism we still know very little about the formation, structure, and regulation of these contact sites . In yeast, proteins of the EPCON (ER‐PO contact) complex act as a molecular tether between ER and PO, but orthologues in higher eukaryotes have not been found. Here, we identify the peroxisomal membrane protein, acyl‐CoA binding protein 5 (ACBD5) as a binding partner for the resident ER protein vesicle‐associated membrane protein‐


associated protein–B (VAPB). Both proteins are anchored in the respective organelle membrane via a C‐terminal membrane‐spanning domain with their N‐termini projecting into the cytoplasm. We show that the N‐terminal cytosolic parts of ACBD5 and VAPB directly interact with each other via a FFAT‐like motif and that ACBD5‐VAPB interaction regulates ER‐peroxisome associations. Moreover, we demonstrate that disruption of the ACBD5‐VAPB complex results in loss of peroxisome‐ER association which perturbs peroxisomal membrane expansion and increases peroxisomal movement. Our findings reveal the first molecular mechanism for establishing peroxisome‐ER associations in mammalian cells and report a new function for ACBD5 in peroxisome‐ER tethering.

M119 A family of membrane‐shaping proteins generates ER subdomains that mediate pre‐peroxisomal vesicle biogenesis. A.S. Joshi1, X. Huang2, V. Choudhary1, T. Levine3, J. Hu2, W. Prinz1; 1NIDDK, National Institutes of Health, Bethesda, MD, 2Department of Genetics and Cell Biology, Nankai University and Tianjin Key Laboratory of Protein Sciences, Tianjin, China, 3Institute of Opthalmology, UCL, London, United Kingdom S. cerevisiae contains three conserved reticulon and reticulon‐like proteins that help maintain ER structure by stabilizing high membrane curvature in ER tubules and the edges of ER sheets. A mutant lacking all three proteins has dramatically altered ER morphology. We found that ER shape is restored in this mutant when Pex30p or its homologue Pex31p is overexpressed. Pex30p can tubulate membranes both in cells and when reconstituted into proteoliposomes, indicating that Pex30p is a novel ER‐shaping protein. Similar to reticulons, these proteins are localized to tubules of peripheral ER and to the edges of ER sheets. In contrast to the reticulons, Pex30p and Pex31p are low abundance and we found that they localize to subdomains in the ER. Interestingly, Pex30p localizes to regions of the peripheral ER that are nearly devoid of reticulons. This suggests that Pex30p is the primary ER‐shaping protein in these ER subdomains. We show that these ER subdomains are the sites where most pre‐peroxisomal vesicles (PPVs) are generated. In addition, overproduction or deletion of Pex30p or Pex31p alters the size, shape, and number of PPVs. The number of Pex30p punctae is more than the number of PPVs per cell, suggesting that Pex30p‐containing domains might have additional functions that remain to be investigated. Thus, our findings suggest that Pex30p and Pex31p help shape and generate ER subdomains for biogenesis of PPVs and possibly other organelles.

M120 Systematic fluorescent protein tagging of endogenous genes in human cells. D. Kamiyama1, S. Sekine1, M.D. Leonetti2, S. Feng3, V. Pessino4, J.S. Weissman2, B. Huang1,5; 1Department of Pharmaceutical Chemistry, University of California, San Francisco, San Francisco, CA, 2Department of Molecular and Cellular Pharmacology, University of California, San Francisco, San Francisco, CA, 3Graduate Program of Bioengineering, University of California, San Francisco, San Francisco, CA, 4Graduate Program of Biophysics, University of California, San Francisco, San Francisco, CA, 5Department of Biochemistry and Biophysics, University of California, San Francisco, San Francisco, CA To systematically study the products of gene expression in a native cellular background, we sought to create libraries of cell lines expressing proteins tagged with a functional sequence at their endogenous loci. Taking advantage of the split fluorescent protein (FP) system and CRISPR/Cas9‐mediated gene editing, we have developed a scalable method for the robust, scarless, and specific tagging of


endogenous human genes with FPs. In particular, the split‐GFP system allows the use of a short peptide of 16 amino acid residues, GFP11, to labeling the protein, which is derived from the 11th beta strand of an engineered super‐folder GFP. We have also engineered FP11 tags from other fluorescent proteins for multicolor labeling. Tandem arrangement of the FP11 tags further enables the proportional amplification of fluorescent signal, which facilitates the detection of proteins with low expression level. Our approach requires no molecular cloning and allows a large number of cell lines to be processed in parallel. We demonstrate the scalability of our method by successful GFP labeling of 30 human genes for fluorescence microscopy as well as for biochemical isolation. Using this mini‐library, we have analyzed the spatial organization of a series of endoplasmic reticulum shaping proteins. Our method paves the way for the large‐scale generation of endogenously tagged human cell lines for the proteome‐wide analysis of protein localization and interaction networks in a native cellular context.

Minisymposium 12: Post‐transcriptional Gene Regulation

M121 Uncovering mechanisms of lncRNA function through comprehensive identification of direct lncRNA‐interacting proteins. C.A. McHugh1, M. Guttman1; 1Biology and Biological Engineering, California Institute of Technology, Pasadena, CA Mammalian cells transcribe many long non‐coding RNAs (lncRNAs) that play important roles in diverse biological processes. However, the mechanisms by which lncRNAs act remain unknown, in large part because we still do not know the identities of their interacting proteins in living cells. To address this, we adapted our RNA antisense purification (RAP) method to enable comprehensive identification of direct lncRNA‐interacting proteins by quantitative mass spectrometry (RAP‐MS). We applied the RAP‐MS method to identify novel binding partners of the Xist lncRNA during initiation of X chromosome inactivation. Xist is a well‐studied lncRNA that orchestrates X chromosome inactivation in female cells by spreading across the future inactive X, silencing transcription by excluding RNA Polymerase II, and repositioning active genes into a silenced nuclear compartment. Using RAP‐MS, we identified several novel direct Xist‐interacting proteins that are essential for Xist‐mediated silencing during the initiation of XCI. Specifically, we show that these proteins are required for several aspects of Xist‐mediated silencing including Xist localization, RNA Polymerase II exclusion and repositioning of active genes into the silenced compartment. These results identify novel mechanisms for Xist‐mediated transcriptional regulation and explain how Xist acts to silence transcription and recruit chromatin regulatory proteins across the inactive X chromosome. Our results highlight the power of the RAP‐MS method to identify novel lncRNA mechanisms that have proven elusive.

M122 Regulation of m6A mRNA Modifications in Mouse Embryonic Stem Cells by Gsk‐3. K.J. Faulds1, J. Egelston1, L.J. Sedivy1, C.J. Phiel1; 1Integrative Biology, University of Colorado Denver, Denver, CO Messenger RNA (mRNA) is the key ubiquitous mediator through which genetic information stored in DNA is converted into proteins. In addition to standard nucleotides, mRNAs contain chemically modified bases, further broadening their functional repertoire. One such modification, methylation of adenosine bases at the C6 position (referred to as m6A), had been discovered 40 years ago, but the functional significance of m6A remained elusive due to a lack of a fundamental understanding of how this mark is


established. Recently, significant strides have been made in understanding the m6A modification of mRNA. Enzymes that add the m6A tags (Mettl3), remove the m6 A tags (FTO and Alkbh5), and a protein that recognizes and binds to the m6A tags (YTHDF2) have all recently been identified, allowing for studies to be performed to determine the biological relevance of m6A tags on mRNA; current evidence suggests that the m6A tag appears to be a mechanism controlling mRNA degradation. Deletion or knockdown of the mRNA m6A methyltransferase Mettl3 in ESCs reduces levels of m6A mRNA, and renders the ESCs unable to differentiate, i.e., the cells remain pluripotent. These data suggest that the m6A tag plays an important role in stem cell differentiation. Interestingly, the inability to differentiate is the same phenotype found in ESCs in which the genes Gsk‐3α and Gsk‐3β are genetically deleted. Gsk‐3α and Gsk‐3β are kinases whose activity is regulated by different signaling pathways. Based on the similarity in phenotypes between Mettl3 and Gsk‐3α / Gsk‐3β double knockout (DKO) ESCs, we hypothesized that the process of mRNA m6A modification is regulated by Gsk‐3α/Gsk‐3β activity. Using a variety of approaches, we provide evidence that m6A mRNA levels are reduced 50% in Gsk‐3α / Gsk‐3β DKO ESCs compared to WT ESCs. Furthermore, we reveal that levels of the RNA demethylase FTO is markedly elevated in Gsk‐3α / Gsk‐3β DKO ESCs compared to WT ESCs, suggesting a mechanism for the reduced m6A levels seen in Gsk‐3 DKO ESCs. In addition, we show that FTO is phosphorylated in WT ESCs, but this phosphorylation is significantly reduced in Gsk‐3α / Gsk‐3β DKO ESCs. Taken together, we provide a molecular basis for the regulation of the methylation of mRNA in ESCs. Our findings thus raise the possibility that signal transduction pathways regulated by Gsk‐3 activity could have a direct effect on the m6A modification of mRNA in mammalian cells, and sheds light on the basis for how Gsk‐3 inhibition or genetic deletion promotes pluripotency in ESCs.

M123 Alteranative Polyadenylation of RECK Regulates Cell Migration and Invasion. H. Lee1, M. Mithun2,3, D.C. Corney2,3,4, E.L. Johnson4, O. Bosompra2, H.A. Coller1,2,3; 1Molecular Biology Interdepartmental Doctoral Program, University of California, Los Angeles, Los Angeles, CA, 2Department of Molecular, Cell and Developmental Biology, University of California, Los Angeles, Los Angeles, CA, 3Department of Biological Chemistry, David Geffen School of Medicine, University of California, Los Angeles, Los Angeles, CA, 4Department of Molecular Biology, Princeton University, Princeton, NJ Cell migration is a highly conserved process involving cytoskeletal reorganization and the restructuring of intracellular communications between cells and the surrounding extracellular matrix. We demonstrate that fibroblast migration and invasion, an essential step for wound healing, can be regulated by alternative polyadenylation (APA) of REversion‐inducing‐Cysteine‐rich protein with Kazal motifs (RECK) protein. RECK encodes a GPI‐anchored protein that inhibits extracellular matrix metalloproteinases (MMPs) through its Kazal motifs. RECK has been found in the perinuclear region and the cell surface. Previous studies found that RECK‐/‐ MEFs show slower migration than wild type as a result of defective focal adhesion formation, increased levels of GTP‐bound Rac1 and Cdc42 activity, and lower levels of detyrosinated tubulin. We observed that RECK undergoes coding region APA, resulting in two distinct proteins in human fibroblasts. Proliferative signals resulted in increased expression of the short RECK isoform. In contrast, entry into quiescence increased expression of the long RECK isoform. We discovered that knockdown of the shorter isoform of RECK results in reduced migration and invasion, while knockdown of the canonical longer isoform of the RECK protein results in more rapid fibroblast migration and invasion, implicating the short RECK isoform as a new player in fibroblast migration. We hypothesize that the short RECK isoform suppresses the activity of the long RECK isoform


through protein‐protein interactions. Co‐immunoprecipitation (Co‐IP) of the short and long RECK isoforms revealed interaction between short RECK and long RECK when the proteins are overexpressed and when they are expressed at endogenous levels. For further study, we generated RECK‐/‐ 293T cells by CRISPR/Cas9 and overeexpressed RECK isoforms. Treating the cells with phosphatidylinositol‐specific phospholipase C, which cleaves GPI‐anchors, released both long RECK (containing GPI anchor) and short RECK (no GPI anchor) from the cell surface supporting an interaction between the two isoforms on the cell surface. When we visualized short RECK–long RECK interaction through bimolecular fluorescence complementation assays, we observed fluorescence signal on the cell surface and in the perinuclear region. Co‐IP using domain‐specific deletions of the long RECK isoform revealed that the Kazal motif domain, important for suppressing MMPs activity, is also critical for interaction with the short RECK isoform. This suggests that MMPs compete with short RECK on the cell surface for the long RECK Kazal motif. Taken together, our results support a model in which cell migration and invasion are inhibited by long RECK, but can be reactivated by expression of and interaction with short RECK.

M124 RanBP2, a giant nuclear pore protein, regulates the translation of diverse sets of mRNAs which encode secretory proteins. A.F. Palazzo1, M. Truong1, K. Kulendra1, H. Zhang1; 1Biochemistry, University of Toronto, Toronto, ON Previously, we discovered that RanBP2 (also known as Nup358), a protein that sits on the cytoplasmic face of the nuclear pore, is required to efficiently translate newly synthesized mRNAs that code for secretory proteins (Mahadevan et al., PLoS Biology 2013). RanBP2 contains zinc finger repeats that directly recognize unique features within the signal sequence coding region (SSCR) of these mRNAs, and these zinc finger are required to regulate the translation of secretory mRNAs. Our data thus suggested that mRNAs that encode secretory proteins directly interact with RanBP2 and this interaction helps to reorganize the packaging of the mRNAs into messenger ribonucleoprotein (mRNP) complexes, which ultimately promotes their translation into secretory proteins. We recently discovered a second region of RanBP2 (the N‐terminal domain) that directly interacts with SSCR RNAs. This region is also required to potentiate the translation of secretory mRNAs. Interestingly, mutations in the N‐terminal domain of the human RanBP2 are associated with Acute Necrotizing Encephalopathy 1 (ANE1), a condition where patients experience a massive stimulation of cytokines in response to influenza infection, which often leads to coma and death. We found that these mutations alter RanBP2’s RNA binding activity. We also show that the SSCRs of mRNAs coding for ANE1‐associated cytokines, such as IL6, lack the features that are normally associated with most SSCRs. Finally we demonstrate that the depletion of RanBP2, in both tissue culture cells and lung epithelial cells, upregulates the translation of the IL6 mRNA. Using an in vivo reconstitution system, we are now investigating how other nuclear pore proteins, such as Nup88 and Nup214, regulate RanBP2’s ability to bind to RNA and thus affect mRNA translation. We are also identifying the cis‐elements in IL6 mRNA that are responsible for its downregulation by RanBP2. Overall, our data suggests that upon the completion of nuclear export, RanBP2 can directly associate with mRNAs and sorts these into one of three classes: the majority of SSCR‐containing mRNAs whore translation is upregulated, cytokine mRNAs whose translation is downregulated, and other mRNAs who are not affected by RanBP2. Finally, our work provides insights into the basic processes that regulate the cellular response to infection, thereby gaining a deeper understanding of the inflammatory process. Ultimately, this work will help us to develop new therapies useful in treating ANE1 and other hyperinflammatory conditions.


M125 Activation of the general translation initiation factor eIF2B antagonizes the integrated stress response and enhances cognition. J.C. Tsai1,2, C. Sidrauski1,2, A. Anand1,2, M. Kampmann2,3, B.R. Hearn4,5, P. Vedantham4,5, P. Jaishankar4,5, M. Sokabe6, A.S. Mendez1,2, C.S. Fraser6, J.S. Weissman2,3, A.R. Renslo4,5, P. Walter1,2; 1Biochemistry and Biophysics, University of California San Francisco, San Francisco, CA, 2Howard Hughes Medical Institute, San Francisco, CA, 3Cellular and Molecular Pharmacology, University of California San Francisco, San Francisco, CA, 4Pharmaceutical Chemistry, University of California San Francisco, San Francisco, CA, 5Small Molecule Discovery Center, University of California San Francisco, San Francisco, CA, 6Molecular and Cellular Biology, University of California Davis, Davis, CA The mammalian integrated stress response (ISR) is a major translational control point activated in response to diverse cellular stress, such as unfolded proteins in the ER, viral infection, heme‐deficiency and amino‐acid starvation. The ISR is mediated by four kinases that upon stress activate and phosphorylate the alpha subunit of eukaryotic translation initiation factor 2 (eIF2‐α). This phosphorylation event renders eIF2 an inhibitor of its dedicated guanine exchange factor (GEF), eIF2B, by potently reducing its off‐rate during nucleotide exchange. This leads to the attenuation of global protein synthesis and selective translation of uORF‐containing mRNAs, e.g. the stress‐responsive transcription factor ATF4. Using a cell‐based screen, our lab discovered the small molecule ISRIB (integrated stress response inhibitor) which reverses the effects of eIF2α‐phosphorylation, restores protein synthesis and remarkably enhances cognition in mice. In this work a reporter‐based shRNA screen identified eIF2B as a candidate molecular target of ISRIB. We found that ISRIB stabilizes an eIF2B heterodecamer, and enhances the GEF activity of eIF2B independent of eIF2 phosphorylation. We also designed new analogs of ISRIB that improve its EC50 to 600pM in cell culture. Furthermore, we have established a robust recombinant system for eIF2B to gain further mechanistic and structural insight into ISRIB activity. Pharmacological modulation of eIF2B promises insight into human health and potential for novel cognitive therapies.

M126 The role of translation elongation in differential gene expression during stress. B.M. Zid1, A. Harvey1, H. Yang1; 1Chemistry and Biochemistry, University of California San Diego, La Jolla, CA A common response to adverse environmental conditions is the suppression of overall translation and the formation of RNA‐protein granules. Translation initiation has traditionally been thought of as the key control step of translation across a variety of stresses. Surprisingly, we found that during glucose starvation in yeast, translation elongation is a crucial mechanism that allows preferential translation of select stress genes, dependent on mRNA localization throughout the cytoplasm. For a number of growth mRNAs that become sequestered to processing bodies, we found that ribosomes become arrested along the transcript, but that these ribosomes can resume translation upon glucose readdition. We also found that transcriptionally upregulated mRNAs that are sequestered to stress granules also have poor translation elongation even though they have high rates of translation initiation. Though these mRNAs have low translation rates during glucose starvation, they are primed for translation upon carbon source readdition. Finally, transcriptionally upregulated mRNAs that are excluded from RNA‐protein granules have high rates of translation initiation and elongation, leading to robust protein synthesis during starvation. We are currently exploring the role that various factors have in regulating translation


elongation during glucose starvation with the use of ribosome profiling and SILAC mass spectrometry. We argue that regulating protein synthesis through translation elongation allows cells to more dynamically control gene expression during fluctuating environmental conditions.

M127 The translational landscape during human neuronal differentiation at transcript isoform resolution. S.N. Floor1,2, J. Blair1, H. Bateup1, D. Hockemeyer1, J.A. Doudna1,2; 1Molecular and Cell Biology, University of California, Berkeley, Berkeley, CA, 2Howard Hughes Medical Institute, Berkeley, CA All steps of gene expression are subject to regulation including transcription, translation, and degradation. Strict control over protein translation is essential in many biological contexts, while dysregulated translation is a hallmark of numerous cancers and developmental disorders. Eukaryotic genes usually produce more than one RNA product, or transcript isoform, through the combined actions of alternative transcription, splicing, and polyadenylation. In principle, each of these transcript isoforms may contain different features that control translation even if the open reading frame is identical. Ribosome profiling is frequently used to measure global translation. However, ribosome profiling is blind to many transcript isoform differences, such as alternative 3′ UTRs, due to the inherent RNase digestion. We recently developed an approach termed TrIP‐seq to measure the translational potential of all cellular transcript isoforms using a polysome profile and deep sequencing. We found that long 3′ UTRs are negatively associated with translation in HEK 293T cells. We also found surprisingly high translation of thousands of retained intron and antisense transcripts. Large transcriptome changes occur during cellular differentiation, but it is unknown how these changes affect translation. Furthermore, neurons express remarkably long 3′ UTRs of up to 20 kilobases and are subject to exquisite translational control. We therefore measured global translation using TrIP‐seq and ribosome profiling in human embryonic stem cells and differentiated neural progenitor cells and neurons. We find that key pluripotency factors such as NANOG express multiple transcript isoforms that are differentially translated and that 3′ UTRs repress translation to a stronger extent as cells differentiate. We identify thousands of transcripts whose translation changes as cells differentiate, including genes critical for neuronal function. We then use ENCODE RNA binding protein and other data to identify candidate factors driving this difference. Taken together, our work defines the translation changes that occur as human stem cells differentiate into neurons, and nominates RNA binding proteins that control these translation changes.

M128 The DEAD‐box helicase Dhh1p couples mRNA decay and translation by monitoring codon optimality. A. Radhakrishnan1, Y. Chen2, S. Martin2, N. Alhusaini2, R. Green1, J. Coller2; 1Program in Molecular Biophysics, Johns Hopkins University, Baltimore, MD, 2Center for RNA Molecular Biology, Case Western Reserve University, Cleveland, OH mRNA decay plays a critical role in the regulation of gene expression. Mechanisms accounting for the disparate nature of mRNA half‐lives between individual transcripts have remained elusive. Our labs recently found that codon optimality is a major determinant of mRNA stability. Codon optimality is a term coined to discuss the non‐ uniform recognition of each of the 61 codons by the ribosome based on supply and demand of tRNA. Codons that are evolutionarily enriched in highly translated mRNA


transcripts are often optimal codons (i.e. triplets that are decoded by tRNAs of relatively higher abundance), whereas codons that exhibit no such selective bias are typically non‐optimal and are decoded by tRNAs of relatively lower abundance. Overall, codon optimality influences translational elongation rate ‐ transcripts of high codon optimality translate faster than those of lower codon optimality. Surprisingly, we found that the rate of ribosome transit on a message (as determined by codon content) dramatically influences mRNA stability. The subject of our current work focuses on how translation elongation is coupled to mRNA decay. Here we establish that the DEAD‐box helicase Dhh1p is a sensor of codon optimality that targets an mRNA for decay. First, we find mRNAs whose translation elongation rate is slowed by inclusion of non‐optimal codons are specifically degraded in a Dhh1p‐dependent manner. Biochemical experiments show Dhh1p is preferentially associated with mRNAs with suboptimal codon choice. We find these effects on mRNA decay are sensitive to the number of slow moving ribosomes on an mRNA. Moreover, we find Dhh1p overexpression leads to the accumulation of ribosomes specifically on mRNAs (and even codons) of low codon optimality. Lastly, Dhh1p physically interacts with ribosomes in vivo. Together, these data argue that Dhh1p is a sensor for ribosome speed, targeting an mRNA for repression and subsequent decay.

M129 Dual interaction of Zuo1, an Hsp70 J‐protein co‐chaperone, with both 60S and 40S subunits of the ribosome, suggesting a role in coordinating protein folding and translation. K. Lee1, R. Sharma1, O.K. Shrestha1, C.A. Bingman1, E.A. Craig1; 1Biochemistry, University of Wisconsin‐Madison, Madison, WI Generating properly folded nascent chains is a challenge for all cells, because the cellular environment is crowded. For optimal cellular function, the rate of protein synthesis at the decoding center needs to be optimized, both for folding of proteins at the ribosomal polypeptide exit and for translational fidelity. In yeast, the Hsp70 Ssb promotes folding of nascent chains. Zuo1, a J‐protein co‐chaperone of Ssb binds near the exit site, allowing interaction between Ssb and nascent chains. Recent cryo‐electron microscopy (EM) analysis indicated that Zuo1 binds not only the 60S, but also the 40S, subunit of the ribosome1 . Although our understanding of chaperone function on the ribosome is growing, two of the major questions remaining are: (1) How Zuo1 interacts with the 60S? (2) What is the importance of Zuo1’s interaction with the 40S? To address question 1, we solved, by crystallographic analysis, the structure of the Zuotin homology domain (ZHD), which is functionally important for association of Zuo1 with the 60S. The ZHD forms an alpha helical bundle of 3 helices. Using this new structural information, we conducted in vivo site‐specific protein cross‐linking and rRNA deletion experiments. Our results demonstrated cross‐linking of helix 3 of the ZHD to ribosomal protein eL31 and importance of rRNA helix 24 (H24) for Zuo1 ribosome association. These results, in conjunction with the previous cryo‐EM data, allowed us to dock Zuo1’s ZHD on the 60S, positioning the critical residues R247 and R251 at the tip of helix 3 close to H24 near the exit site. Cryo‐EM results1 suggested that the C‐terminus of Zuo1 interacts with the 40S by binding 18S rRNA expansion segment 12 (ES12) of helix 44, which stems from the decoding center. We confirmed this positioning by demonstrating that deletions in either the C‐terminus of Zuo1 or ES12 affected Zuo1 ribosome association. This raises a question as to the importance of 40S interaction to Zuo1’s known effect on translational fidelity. We demonstrated that normal stop codon readthrough and ‐1 frameshifting requires the interaction between C‐terminus of Zuo1 and ES12.


In summary, our study offers insight into how dual interaction of Zuo1 with the 60S and 40S subunits of the ribosome may play a unique role in protein homeostasis. Our working model is that Zuo1 monitors protein folding at the exit site and coordinates it with protein synthesis at the decoding center2. 1. Zhang, Y. et al. Structural basis for interaction of a cotranslational chaperone with the eukaryotic

ribosome. Nat Struct Mol Biol 21, 1042‐6 (2014). 2. Lee, Sharma and Shrestha contributed equally to this work.

M130 Role of C‐terminal Alanine and Threonine Extensions (CAT tails) in Nascent Polypeptide Degradation. K. Kostova1, J.S. Weissman1; 1CMP , University of California, San Francisco , San Francisco, CA Protein biosynthesis is by far the most energy‐consuming process during cellular proliferation. Therefore, any event that interferes with protein production can jeopardize the viability of the cell. Of particular concern is ribosome stalling ‐ a phenomenon in which the ribosome gets trapped on a message and can neither proceed with translation, nor get released. To counter the threat to cell viability posed by such stalled ribosomes, cells have evolved quality control pathways that degrade the messenger RNA, release and recycle the trapped ribosome, and target the partially synthesized polypeptide for proteasomal degradation. The ribosome quality control complex (RQC) engages the peptide‐bound 60S ribosome subunit, and facilitates the ubiquitination of the nascent polypeptide, and its extraction from the exit tunnel. In addition, a poorly characterized protein (RQC2) binds near the transfer RNA binding site, and recruits charged alanine and threonine tRNAs, which subsequently allow the addition of Ala and Thr to the C terminus of the nascent polypeptide. Although this unique type of mRNA‐independent protein synthesis has been observed both in vivo and in vitro, the significance of these Ala and Thr C‐terminal extensions (CAT tails) is poorly understood. Our research aims to determine the function of the CAT tails in vivo. We have evidence that the addition of Ala and Thr extensions to a partially synthesized polypeptide serves as a fail‐safe mechanism for protein degradation. Ribosome stalling is a stochastic process, and therefore the topology or amino acid composition of the nascent peptide is not always optimal for ubiquitination, extraction from the exit tunnel, or targeting to the proteasome. In such cases, addition of CAT tails can alleviate steric hindrance, and provide appropriate context for ubiquitination. To test this hypothesis, we utilized a mutant of RQC2 that facilitates the assembly of the complex on the 60S subunit, but does not support the recruitment of tRNAs, and therefore addition of CAT tails to the nascent polypeptide. This mutant allowed us to identify RQC substrates that are degraded in a CAT‐tail‐dependent manner and explore how the addition of CAT tails facilitate degradation. All members of the RQC are conserved up to humans, suggesting that the complex serves similar function in higher eukaryotes. Furthermore, mutations in RQC components, including RQC2, have been previously linked to neurodegeneration. Therefore, understanding the cellular function of CAT tail synthesis can further our understanding of human disease and open new therapeutic windows.


Microsymposium 7: Cell Adhesion and Migration

E55 Proper actin network architecture is essential for restricting polarization to a single leading edge in neutrophils. B.R. Graziano1, A. Diz‐Muñoz2, O.D. Weiner1; 1Cardiovascular Research Institute, UCSF, San Francisco, CA, 2Cell Biology and Biophysics, EMBL, Heidelberg, Germany Efficient cell migration requires that actin assembly events are spatially restricted to form a single leading edge in response to external stimuli. In neutrophils, cells of the innate immune system that rely on chemotaxis to track down pathogens, the protrusive force of actin polymerization leads to increases in membrane tension. This tension increase, in turn, serves to funnel subsequent actin assembly events into the dominant protrusion (i.e., the leading edge) to simultaneously maintain its growth while blocking the formation of additional protrusion sites. Previous work has uncovered a direct relationship between cellular actin polymer levels and the amount of tension experienced by the plasma membrane. However, it remains less clear if different types of actin networks are capable of generating similar increases in membrane tension in response to polarity cues. A multitude of actin nucleators and nucleation‐promoting factors (NPFs) function alongside one another to drive cell migration. In particular, the WAVE regulatory complex (WRC), itself an NPF, plays a central role in chemotaxis by organizing the assembly of highly‐branched actin networks at the leading edge of migrating cells. Here, we assess the role that actin network architecture plays in generating plasma membrane tension by using CRISPR‐mediated genome editing to knockout the HEM1 subunit of WRC in neutrophil‐like HL‐60 cells. Our ‘HEM1null’ cells, despite lacking a major actin assembly pathway, produced significantly higher amounts of actin polymer than wildtype cells. However, the ability of HEM1null cells to increase membrane tension when stimulated with chemoattractant was dramatically impaired, indicating that actin polymer level and membrane tension do not always correlate with each other. Consistent with this membrane tension defect—and in striking contrast to both wildtype and HEM1 knockdown cells—stimulated HEM1null cells failed to restrict actin assembly to a single leading edge. They instead formed multiple dynamic needle‐like protrusions in a completely non‐polarized fashion. Together, our observations suggest that not all actin assembly pathways can produce filament networks capable of generating increases in membrane tension that enforce a single axis of polarity.

E56 Biomechanical Alterations of Endothelial Cells Induced by Listeria monocytogenes. E.E. Bastounis1, J.A. Theriot1,2,3; 1Biochemistry, Stanford University, School of Medicine, Stanford, CA, 2Microbiology and Immunology , Stanford University, School of Medicine, Stanford, CA, 3Howard Hughes Medical Institute, Stanford University, School of Medicine, Stanford, CA Endothelial cells (ECs) line the inner lumen of vessels forming a protective barrier separating blood from tissue compartments. Interaction of bacteria with such barriers is critical for the course of invasive infections. When some pathogenic bacterial species infect ECs, they utilize the cells’ actin cytoskeleton to invade and spread to adjacent ECs by producing actin tails that assist them in propelling forward. As a result of infection, the ECs’ actin cytoskeleton needs to dynamically re‐arrange in a way that depends on the infecting species. Changes in the actin architecture of cells are long known to have a critical impact on how cell mechanics is regulated. Although EC mechanics and response to shear have been studied


extensively, how bacterial infection progresses in ECs residing on different mechanical environments and the effect of infection on vascular biomechanics are still largely unknown. We are exploiting Listeria monocytogenes (LM), a facultative intracellular bacterial pathogen to determine the biophysical relationships between ECs, the stiffness of their extracellular matrix and infection. We developed a multi‐well format platform where we build polyacrylamide hydrogels of varying stiffness and on which we seed ECs, which we subsequently infect with LM. Our platform is compatible with flow cytometry, immunostaining and traction force microscopy. We use flow cytometry and microscopy to evaluate the degree of bacterial adhesion, invasion and cell‐ to‐cell spread depending on the stiffness of the matrix where ECs reside and find that both adhesion and invasion of LM on ECs increase with increasing matrix stiffness. We also find that that ECM‐stiffness dependent efficiency of EC infection is abrogated when ECs are treated with ROCK and FAK inhibitors, suggesting that these kinases are implicated both in stiffness sensing of ECs and LM internalization. To determine the effect of infection on EC mechanics we use traction force microscopy and simultaneously examine EC monolayers infected by wild‐type LM but also strains lacking specific proteins necessary for the production of actin tails or virulence factors. We find that infection of ECs with LM leads to an increase of the traction forces that infected ECs exert on their ECM at short times post‐infection, underlying a possible change in their biomechanical properties. Our long‐term goal is to understand the physical and mechanical interactions between ECs, their extracellular matrix and infectious bacteria and therefore lay important groundwork for the understanding of the pathogenesis of vascular bacterial infections and possibly for developing therapeutic interventions.

E57 Two distinct actin networks mediate traction oscillations to confer mechanosensitivity of focal adhesions. Z. Wu1, S.V. Plotnikov2, A.Y. Moalim2, C.M. Waterman1, J. Liu1; 1NHLBI, NIH, Bethesda, MD, 2Department of Cell Systems Biology, University of Toronto, Toronto, ON Focal adhesions (FAs) are integrin‐based transmembrane assemblies. They serve as mechanosensors through which cells sense the mechanical stiffness of their extracellular matrix (ECM) by exerting actin cytoskeleton‐mediated traction force. Interestingly, FA itself is a dynamic structure that adapts its area and composition in response to mechanical force. How the cell manages this FA plasticity to accurately sense ECM stiffness remains an open question. Strikingly, FA‐localized traction forces oscillate in time and space and govern the cell mechanosensing of ECM stiffness. Precisely how and why the FA traction oscillates is unknown. Combining theory and experiment, we develop a model of FA growth that integrates coordinated contributions of a branched actin network and stress fibers (SF) in the process. We show that while promoting the proximal growth of the FA, the mechanical engagement with retrograde flux of branched actin network contributes to a traction peak near the FA distal tip. This temporally results in a traction gradient within the growing FA that favors SF formation near the FA proximal end. The SF‐mediated actomyosin contractility stabilizes the FA and generates a second traction peak near the FA center. Formin‐mediated SF elongation negatively feeds back with actomyosin contractility, resulting in the central traction peak oscillation. This underpins the observed FA traction oscillation, and importantly, broadens the ECM stiffness range, over which FAs could accurately adapt with traction force generation. The process that sets the stage for FA traction oscillation therefore reflects a fundamental mechanism that couples the actin‐mediated FA growth with efficient FA mechanosensing.


E58 Actin retrograde flow actively aligns and orients ligand‐engaged integrins in focal adhesions. V. Swaminathan1, J.K. Mathew2, S. Mehta3, P. Nordenfelt4, T.I. Moore5, K. Nobuyasu6, D. Baker6, R. Oldenbourg3, T. Tani3, S. Mayor2, T.A. Springer5, C.M. Waterman1; 1Cell Biology and Physiology Center, NHLBI/NIH, Bethesda, MD, 2National Center for Biological Sciences, Bangalore, India, 3Marine Biological Sciences, Woods Hole, MA, 4Division of Infection Medicine, Lund University, Lund, MD, 5Program in Cellular and Molecular Medicine, Harvard University, Boston, MA, 6Department of Biochemistry, University of Washington , Seattle, WA Integrins are transmembrane receptors that, upon activation, bind extracellular matrix (ECM) or cell surface ligands and link them to the actin cytoskeleton to mediate cell adhesion and migration. One model for the structural transitions mediating integrin activation termed “the cytoskeletal force hypothesis” posits that force transmitted from the cytoskeleton to ligand‐bound integrins acts as an allosteric stabilizer of the extended‐open, high‐affinity state. Since cytoskeletal forces in migrating cells are generated by centripetal “retrograde flow” of F‐actin from the cell leading edge, where integrin‐based adhesions are initiated, this model predicts that F‐actin flow should align and orient activated, ligand‐bound integrins in integrin‐based adhesions. Here, polarization‐sensitive fluorescence microscopy of GFP‐αVβ3 integrin chimeras in migrating fibroblasts shows that integrins are aligned with respect to the axis of FAs and the direction of F‐actin flow, and this alignment requires binding immobilized ligand and talin‐mediated linkage to a flowing cytoskeleton. Polarization imaging and Rosetta modelling of chimeras engineered to orient GFP differentially with respect to the integrin headpiece suggest that ligand‐bound αVβ3 integrin may be markedly tilted by the force of F‐actin flow. These results show that actin cytoskeletal forces actively sculpt an anisotropic molecular scaffold in FAs that may underlie the ability of cells to sense directional ECM and physical cues. Additionally, these results are the first demonstration of “active ordering” of transmembrane receptors and highlight a novel role for physical forces in subcellular and membrane organization.

E59 Novel Role of Septin2 in Endothelial Cell Podosome Formation and Matrix Degradation. K.B. Collins1, H. Kang1, J. Klomp1, A.V. Karginov1, A.B. Malik1; 1Molecular and Cellular Pharmacology, University of Illinois at Chicago, Chicago, IL In the initial phases of angiogenesis endothelial cells must degrade the extracellular matrix before migrating to alternative sites to form nascent vessels. The degradation of the matrix is performed by podosomes, actin rich matrix degrading structures that are regulated by Src family kinases. Here we report that the filamentous GTPase Septin2 is phosphorylated upon activation of Src and plays a critical role in Src‐mediated podosome formation and matrix degradation. We have identified that activation of Src induces phosphorylation of Septin2 and its translocation to the podosome structures in endothelial cells. Superresolution microscopy analysis further revealed that Septin2 localizes to the borders of podosome rosettes on the basal side of the cell. We have found that downregulation of Septin2 expression in endothelial cells inhibits Src‐mediated podosome formation as well as matrix degradation. Additionally, we have shown that Septin2 expression is necessary for endothelial cell invasion into a 3D collagen matrix, indicating a possible role for Septin2 in angiogenesis via regulation of podosome function. We are currently investigating the role of Septin2 in angiogenesis as well as the functional significance of Septin2 phosphorylation secondary to Src activation.


E60 Actomyosin‐dependent structural alterations to tight junctions result in loss of barrier function in MDCK cells upon acute mechanical strain. N. Chavez1, M.A. Garcia1, T. Yasumura2, J. Moeller3, B.L. Pruitt3,4,5, J.E. Rash2, R.G. Johnson6, W.J. Nelson1,5; 1Department of Biology, Stanford University , Stanford, CA, 2Department of Biomedical Sciences, Colorado State University , Fort Collins, CO, 3Department of Mechanical Engineering, Stanford University, Stanford, CA, 4Department of Bioengineering, Stanford University, Stanford, CA, 5Department of Molecular and Cellular Physiology, Stanford University , Stanford , CA, 6Department of Genetics, Cell Biology, and Development, University of Minnesota, Minneapolis, MN The tight junction (TJ) forms a permeability barrier comprised of size‐ and charge‐selective pores that regulates the paracellular movement of ions and solutes across tissues. TJs are composed of transmembrane proteins (claudins and occludin) and adaptor proteins (e.g. ZO‐1) bound to the actin cytoskeleton, which together regulate paracellular permeability, as well as cell proliferation and differentiation. Mechanical strain is theorized to regulate these and other critical cellular processes, but the effect on TJ structure and function in epithelial cells is poorly understood. We analyzed TJ function, structure and molecular organization in MDCK epithelial monolayers strained for 1 hour (8‐13% biaxial strain), using a combination of thin‐section transmission electron microscopy (tsTEM), freeze fracture TEM (FF), and immunofluorescence microscopy (IF). Studies with ruthenium red, a tracer added to the apical surface of cells to assay for paracellular diffusion, by tsTEM analysis revealed that paracellular permeability increased with increasing strain, indicating a decrease in TJ integrity. Following mechanical strain, FF revealed discontinuities in TJ strands, decreased strand complexity and topographical changes with increased beaded appearance of the strands. TsTEM revealed a decrease in the electron density of membranous and sub‐membranous elements of TJs, further suggesting strain‐induced alterations to the molecular organization of TJs. These functional and structural changes following mechanical strain were correlated with decreased levels of claudin‐1 and ZO‐1 at TJs, as detected by IF. To test roles of actomyosin contractility in modulating TJ structure/function under strain, cells were treated with ML‐7 before and during strain. Cells strained in the presence of ML‐7 appeared more similar to unstrained control cells by their lack of significant ruthenium red diffusion across the TJ, and retention of normal levels of claudin‐1 and ZO‐1 at TJs. Altogether, our data indicate the structure and molecular organization of TJs are altered upon acute mechanical strain, in an actomyosin contractility‐dependent manner, suggesting that mechanical strain has a role in regulating TJ structural integrity and barrier properties.

E61 Traction forces and three‐dimensional shape of the cell edge mediate organization of cell‐edge activity during polarization. A. Bornert1,2, N. Piacentini1,3, F. Raynaud1,4, A.B. Verkhovsky1,2; 1Lab. of Cell Biophysics, Ecole Polytechnique Fédérale de Lausanne, Lausanne, Switzerland, 2Lab. of Physics of Living Matter, Ecole Polytechnique Fédérale de Lausanne, Lausanne, Switzerland, 3Intel Semiconductor AG, EPFL Innovation Park, Lausanne, Switzerland, 4Computational Systems Oncology, University of Lausanne, Lausanne, Switzerland How cells break symmetry and organize their activity to move directionally is a fundamental question in cell biology. Models of cell polarization commonly rely on front‐to‐back gradients of regulatory factors to control cell‐edge activity, but it is not clear how the front‐back axis is defined in the first place. We


have recently proposed a novel and simple principle of self‐organizing cell activity, in which local cell‐edge dynamics depends on the distance from the cell center, but not on the orientation with respect to the front–back axis (Raynaud et al., Nat. Physics, 12:367‐373, 2016). Stochastic modeling demonstrated that distance‐sensitivity is sufficient for the symmetry breaking and the emergence of persistent cell motion and stable shape. However, physical mechanisms of distance‐sensitivity remain unclear. Here we show that traction forces exerted on the substrate by polarizing and migrating cells generally increase with the distance from the cell center, that maximal forces are associated with protrusion‐retraction switches at the cell edge, and that in the absence of endogenous cell traction, protrusion‐retraction switches could be induced by the application of external force. These results suggest that traction forces could mediate distance‐sensitivity and organize cell activity during symmetry breaking. We also show that three‐dimensional shape of the cell‐substrate interface depends on the distance and influences the distribution of protrusion‐retraction switches, suggesting that it may be another factor involved in distance‐sensitivity. Supported by Swiss National Science Foundation Grant 31003A‐135661.

E62 Molecular Mechanism of Fascin Regulated Cancer Cell Invasion. E. Pham1, S.P. George1, A. Ahrorov1, S. Patnaik1, J.C. Conrad2, S. Khurana1,3; 1Biology and Biochemistry, University of Houston, Houston, TX, 2Chemical and Biomolecular Engineering, University of Houston, Houston, TX, 3Baylor College of Medicine, Houston, TX Invasion and metastasis represent the primary sources of morbidity and mortality in most cancers. Defining and validating therapeutic targets to eliminate these malignant mechanisms remains of critical importance in cancer biology. Fascin is an actin crosslinking protein that organizes F‐actin into tightly packed bundles. Consequently, it is typically localized to actin‐based structures such as filopodia. In humans, fascin is normally expressed throughout development of the nervous system, and in adult vascular endothelial cells and mesenchymal cells. However, fascin is frequently aberrantly expressed in human cancers of epithelial origin, where it is normally not expressed. Though fascin expression is correlated with invasion, metastasis, and poor prognosis in many cancers, the mechanism underlying its contribution to cancer progression is not well defined and requires further investigation. Here, we uncovered the mechanism of Fascin in promoting tumor invasion. We show that cells overexpressing fascin are able to generate long and thin cell‐cell connections called tunneling nanotubes. Furthermore, fascin expression causes defects in cell‐cell adhesions in both 2D monolayer and 3D tumor sphere cultures. These defects coincide with localization of β‐catenin away from cell‐cell adhesions, and into nuclei. Additionally, we show that fascin promotes activation of the developmental and stem cell Wnt/β‐catenin signaling pathway, and expression of proliferation, transformation, and invasion genes. Pharmacologic inhibition of fascin disrupts the interaction between fascin and β‐catenin, and is able to block 3D invasion of tumor cells by preventing Wnt/β‐catenin signaling at the tumor edge. Colorectal cancers have shown to be particularly addicted to oncogenic Wnt/β‐catenin signaling, but drugs targeting the pathway have not reached the clinic, possibly due to the pathway’s fundamental importance in maintenance of adult stem cells. In this study, we provide evidence that pharmacologic inhibition of fascin may be a promising strategy to mitigate oncogenic Wnt/β‐catenin signaling to prevent malignant progression of cancers.


E63 The two‐dimensional nature of leading‐edge lamellipodia is intrinsic and does not depend on surface interactions. L. Fritz‐Laylin1, M. Mehan1, B. Chen2, T. Goddard1, T. Ferrin1, E. Betzig2, G. Johnson1, R.D. Mullins1; 1Molecular and Cellular Pharmacology, University of California San Francisco, San Francisco, CA, 2HHMI Janelia Farm Research Campus, Ashburn, VA Single‐celled amoebae, immune cells and other amoeboid cells change shape as they move, forming actin‐filled pseudopods that advance and retreat from their leading edges. Although conventional light microscopy reveals the architecture and dynamics of thin lamellipods made by slow‐moving cells on flat surfaces, it lacks the time and spatial resolution required to track complex pseudopods of ecells moving in three‐dimensions. We therefore turned to lattice light‐sheet microscopy (LLSm) to perform three‐dimensional, time‐lapse imaging of neutrophil‐like, HL‐60 cells crawling through collagen matrices. To analyze 3D pseudopods we: (1) developed fluorescent probe combinations that distinguish cortical actin from dynamic, pseudopod‐forming networks, and (2) adapted molecular visualization tools from structural biology to render complex cell surfaces. Surprisingly, 3D pseudopods turn out to be thin (<0.75 µm), flat sheets that sometimes interleave to form rosettes. Their sheet‐like nature is not templated by an external surface, but likely reflects the linear arrangement of underlying signaling molecules. Although dispensable for 3D locomotion, loss of pseudopods dramatically decreases the frequency of cell turning, and pseudopod dynamics increase when cells change direction, indicating an important role in pathfinding.

Microsymposium 8: Regulation of the Trafficking Machinery

E64 The adaptor protein APPL1 regulates cancer cell migration through α5β1 trafficking and Rac activation. N. Diggins1, D.J. Webb1,2,3; 1Biological Sciences, Vanderbilt University, Nashville, TN, 2Kennedy Center for Research on Human Development, Vanderbilt University, Nasvhille, TN, 3Cancer Biology, Vanderbilt University, Nasvhille, TN Cell migration is a tightly, spatiotemporally regulated process that requires the coordination of many molecular components. Because adaptor proteins can serve as integrators of various cellular events, they are being increasingly studied as regulators of cell migration. The adaptor protein containing a pleckstrin‐homology (PH) domain, phosphotyrosine binding (PTB) domain, and leucine zipper motif 1 (APPL1) is a 709‐amino acid endosomal protein that is known to play a role in cell proliferation and survival as well as endosomal trafficking and signaling. However, its function in modulating cell migration is currently poorly understood. Here, we show that APPL1‐GFP expression decreases migration speed, compared to GFP controls, in an α5β1 integrin‐dependent manner, as plating cells in the presence of α5 integrin‐blocking antibody negates the APPL1‐mediated effect on migration. Converesely, knockout of APPL1, using the CRISPR/Cas9 system, increases migration speeds compared to non‐targeting controls. Using internalization assays, we demonstrate that β1 integrin is internalized slower in APPL1‐GFP‐expressing cells. The effect of APPL1 on internalization is specific to integrin trafficking, because Transferrin Receptor trafficking is unaltered in APPL1‐GFP‐expressing cells. APPL1 interacts with Rab5, a small GTPase found on early endosomes, and Rab5 has been implicated in promoting cell migration. We therefore hypothesized that APPL1 regulates integrin trafficking through its interaction with Rab5. Indeed, a mutation rendering APPL1 deficient in Rab5 binding (APPL1‐N308D)


has no effect on integrin trafficking and migration. Rab5 has been shown to activate Rac on endosomes to promote migration, and our data show that APPL1‐GFP expression leads to a decrease in active Rac. Additionally, expression of APPL1‐GFP, but not APPL1‐N308D‐GFP, decreases levels of active p21‐activated kinase (PAK), a Rac effector. Moreover, knockout of APPL1 led to an increase in active PAK. These data indicate a role for APPL1 in regulating migration by modulating integrin trafficking as well as Rac and PAK activity.

E65 The structure of the COPI coat determined within the cell. Y.S. Bykov1, M. Schaffer2, S.O. Dodonova1, S. Albert2, J.M. Plitzko2, W. Baumeister2, B.D. Engel2, J.A. Briggs1; 1Structural and Computational Biology Unit, European Molecular Biology Laboratory, Heidelberg, Germany, 2Max Planck Institute of Biochemistry, Martinsried, Germany Vesicular trafficking is a fundamental mechanism for maintaining compartmentalization and protein secretion in eukaryotic cells. Protein cargos are sorted from a donor compartment into coated vesicular carriers, which are selectively delivered to the target compartment. Eukaryotic cells employ three main protein coat complexes to mediate cargo sorting and vesicle formation. COPI is responsible for Golgi to ER and intra‐Golgi transport while COPII is responsible for ER to Golgi transport. The clathrin coat mediates transport from the plasma membrane, endosomes, and from trans Golgi. These coats consist of multiple subunits that can be classified as belonging to either an outer coat or an adaptor sub‐complex. While it has been previously shown that the adaptor and outer coat sub‐complexes of COPII and clathrin form two distinct layers, each with a different arrangement, the arrangement of COPI remained a mystery. We recently published a pseudo‐atomic model of the COPI coat derived from an in vitro reconstituted system and found that it differs from COPII and clathrin: the outer coat and adaptor subunits in COPI do not form distinct layers but rather intertwined trimeric assemblies (triads). Here, we aimed to determine the structure of the COPI coat in situ , within cells. We applied cryo‐FIB (focused ion beam) milling to prepare thin lamella from vitrified Clamydomonas reinhardtii cells and imaged them using cryo‐electron tomography, targeting regions in the vicinity of the Golgi. We found COPI buds on both curved rims of flat citsternae and on trans‐Golgi structures with much lower curvature. We also identified COPI coated vesicles. To resolve the in situ structure of the COPI coat complex on Golgi buds and associated vesicles, we applied subtomogram averaging using a reference‐free workflow. The structure and arrangement of the coat within the cells matched the structure previously obtained in vitro , suggesting a common budding mechanism in vitro and in vivo . Interestingly, within the cells we were able to visualize an additional density corresponding to cargo, bound to the N‐terminal β‐propeller of β’ COPI subunit. In ongoing analyses, we are assessing whether the structure of the coat and associated cargo differs depending on its location within the Golgi or on the stage of budding.

E66 Regulation of COPII vesicle trafficking by O‐GlcNAcylation. B.J. Bisnett1, N.J. Cox1,2, B.M. Condon1,2, T.R. Meister1, T.J. Smith1, M. Boyce1; 1Biochemistry , Duke University , Durham, NC, 2Pharmacology and Cancer Biology , Duke Univeristy , Durham, NC Coat protein complex II (COPII) traffics vesicles in the early secretory pathway from the endoplasmic reticulum (ER) to the Golgi apparatus. Consisting of five essential components, COPII vesicle formation is


initiated by the recruitment of Sar1, a small GTPase, followed by the formation of an inner coat consisting of Sec23/Sec24 heterodimers, and lastly an outer coat of Sec13/Sec31 heterotetramers. While the mechanism and structure of the COPII cage have been extensively studied, little is known about the regulation of COPII vesicle trafficking in response to nutrient status, stress or signal transduction. Recently, we and others have found that many COPII components, including Sec23A, Sec31A and several Sec24 family members, are reversibly modified by O‐linked beta‐N‐acetylglucosamine (O‐GlcNAc). O‐GlcNAc is a single sugar modification that, like phosphorylation, is dynamically regulated on serine and threonine residues of intracellular proteins. To understand the effects of O‐GlcNAc on COPII vesicle trafficking, we examined Sec24C as a model protein. Using mass spectrometry analysis, we discovered nine novel O‐GlcNAc sites on human Sec24C, six of which are located in the long unstructured N‐terminal domain, which has no known function. We also used a chemical biology approach to show that O‐GlcNAc mediates protein‐protein interactions among human Sec24 proteins, including Sec24C. Finally, through reconstitution of unglycosylatable Sec24C O‐GlcNAc site mutants in Sec24C CRISPR‐deleted human cells, we are evaluating the functional significance of each O‐GlcNAc site in membrane association, stress responses, nutrient availability and cargo trafficking. Our work addresses a long‐standing gap of knowledge in the regulation of the early secretory pathway by defining a role for dynamic glycosylation in governing Sec24C function.

E67 LTK is an ER‐resident receptor tyrosine kinase that sense the load of glycoproteins and regulates COPII trafficking. F.G. Centonze1, C. Behrends2, H. Farhan1; 1Institute of Basic Medical Sciences, University of Oslo, Oslo, Norway, 2Institute of Biochemistry II, University of Frankfurt, Frankfurt, Germany Our understanding of the regulation of ER export by signaling is only beginning to emerge. In previous work we and others showed that environmental signaling (mainly by mitogen activated protein kinases) regulates the formation of ER exit sites (ERES). It remains unknown whether and how ER export is regulated by signaling from the ER itself. Here, we focus on Leukocyte Tyrosine Kinase (LTK), a receptor tyrosine kinase that we have uncovered previously to play a role in determining the number of ERES. Strikingly, we find that LTK localizes mainly to the ER, forming tyrosine phosphorylated dimers in ER microdomains. Silencing of LTK reduces the number of ERES and the rate of ER‐to‐Golgi transport. By mapping the interactome of LTK we found that it interacts with lectin‐type cargo receptor ERGIC‐53. Overexpression of ERGIC‐53 dependent cargo resulted in activation of LTK, while other types of cargo did not active LTK. Furthermore, we identify Sec12, the exchange factor for the small GTPase Sar1 as a substrate for LTK. Accordingly, overexpression of LTK changed the dynamics of Sar1 on ERES, as determined using FRAP microscopy. Altogether, our findings indicate that LTK is an ER‐resident receptor tyrosine kinase that signals locally at the ER. By interacting with ERGIC‐53, LTK senses the load of glycoproteins and transmits the signal to ERES via phosphorylation of Sec12.


E68 Vps4 is required for clathrin‐mediated endocytosis mutant viability and cargo trafficking from the plasma membrane. K. Hoban1, S. Lux1, J. Poprawski1, Y. Zhang1, D.C. Prosser1, C. Norris1, B. Wendland1; 1Department of Biology, Johns Hopkins University, Baltimore, MD Endocytic trafficking from the plasma membrane (PM) regulates many processes, including lipid and protein turnover, signaling, and nutrient uptake. During clathrin‐mediated endocytosis (CME), the predominant internalization pathway in budding yeast, transmembrane cargo proteins are bound by endocytic adaptors. Many of these adaptors are directly involved in the recruitment of clathrin. The clathrin‐associated sorting protein family of adaptors includes the yeast epsins, Ent1/2, and AP180/PICALM homologues, Yap1801/2. Previous work has shown that cells lacking these four adaptor‐encoding genes are inviable; however, in a quadruple adaptor deletion background, the presence of a single ENTH domain from one of the epsins produces viable cells with severe endocytic defects such as cargo accumulation at the PM. This CME‐deficient, quadruple adaptor mutant strain provides a sensitized background ideal for revealing cellular components that interact with clathrin adaptors. A mutagenic screen was performed to identify alleles that are lethal when combined with the quadruple adaptor mutant using a colony‐sectoring reporter assay. To determine the causative mutation, a genomic library was utilized to complement these mutants and restore viability. For one complementation group, multiple plasmids were observed to contain distinct, yet overlapping, genomic intervals that included the gene VPS4, which encodes an AAA‐type ATPase necessary for ESCRT‐III disassembly. Upon sequencing, members of this complementation group possessed mutations in VPS4, including a missense mutation in the large ATPase domain. Moreover, deletion of VPS4 in the adaptor mutant background replicated the synthetic lethality phenotype from the reporter assay. Interestingly, loss of Vps4 in an otherwise WT background resulted in aberrant accumulation of the cargo Ste3 at the PM. These findings reveal a previously unappreciated genetic interaction between VPS4 and CME adaptors.

E69 Molecular Investigations Into the Presynaptic Functions of Synucleins. K.J. Vargas1,2, N. Schrod3, T. Davis1,2, R. Fernandez‐Busnadiego2,4,5, Y.V. Taguchi2,4, U. Laugks3, V. Lucic3, S.S. Chandra1,2,6; 1Department of Neurology, Yale University , New Haven, CT, 2Program in Cellular Neuroscience, Neurodegeneration and Repair, Yale University , New Haven, CT, 3Max Planck Institute of Biochemistry, Martinsried, CT, 4Department of Cell Biology, Yale University , New Haven, CT, 5Howard Hughes Medical Institute, Yale University , New Haven, CT, 6Department of Neuroscience, Yale University , New Haven, CT Parkinson’s disease (PD) is a neurodegenerative disease characterized by the accumulation of Lewy bodies. The main component of these pathological aggregates is α‐synuclein, a presynaptic protein. Furthermore, point mutations, duplication and triplication of the α‐synuclein gene cause familial PD. Hence, there is great interest in elucidating the physiological function(s) of α‐synuclein and how they impact synaptic structure and function. We and others have previously shown that synucleins can sense and generate membrane curvature, properties consistent with roles in synaptic vesicle exo‐ and endocytosis. Recently, using αβγ‐synuclein triple knock out (TKO) mice, we showed that deletion of synucleins leads to a deficit in synaptic vesicle endocytosis (SVE). Here, we explore the molecular roles of α‐synuclein in SVE using several independent approaches. Using a membrane recruitment assay, we


discovered that synucleins are necessary for the normal recruitment of a subset of endocytic proteins to synaptic membranes. These proteins include those that function in exo‐endocytosis coupling as well as early coat assembly proteins. Strikingly, we also observed an increase in the recruitment of clathrin to synaptic membranes in the absence of synucleins. Hence, we are currently testing if synucleins function to limit the size of clathrin patches, allowing for normal clathrin coated pit maturation. In a set of biochemical experiments, we observed changes in the phosphorylation of several key endocytic proteins in TKO synaptosomes, which is an important means of regulating SVE. In addition, using cryoEM and conventional EM, we showed that synucleins are important regulators of synapse size as well as synaptic vesicle distribution. Therefore, we conclude that α‐synuclein maintains synaptic vesicle pools through modulation of SVE, endocytic protein recruitment, and synaptic vesicle localization. Thus, long‐term malfunction of α‐synuclein, as in PD, is likely to cause changes in SVE and synaptic vesicle organization, leading to progressive synaptic deficits and concomitant neurodegeneration.

E70 Structure, inhibition, and regulation of a vacuolar two‐pore channel TPC1. A.F. Kintzer1, R.M. Stroud1; 1Biochemistry and Biophysics, University of California, San Francisco, San Francisco, CA Membrane transport serves vital functions in the cell, providing nutrients for growth, relaying electrical signals, evading pathogens, and maintaining homeostasis. Voltage‐ and ligand‐gated ion channels propagate cellular electrical signals by coupling changes in membrane potential and the binding of molecules to channel opening and selective passage of ions through the membrane barrier. Two‐pore channels (TPCs) are intracellular ion channels that integrate changes in membrane potential, second messengers, and phosphorylation to control endolysosomal trafficking, autophagy, cellular ion and amino acid homeostasis, and ultra‐long action potential‐like signals. They broadly impact human diseases related to trafficking including filoviral infections, Parkinson’s disease, obesity, fatty liver disease, and Alzheimer’s disease. The response of TPCs to multiple cellular inputs suggests a multi‐state gating mechanism. Nevertheless, the mechanisms that govern cycles of activation and deactivation or ‘gating’ in the channel remain poorly understood. To understand the bases for ion permeation, channel activation, the location of voltage‐sensing domains and regulatory ion‐binding sites, and phosphoregulation, we determined the crystal structure of TPC1 from Arabidopsis thaliana to 2.87Å resolution. This reveals for the first time how TPC channels assemble as ‘quasi‐tetramers’ from two non‐equivalent tandem pore‐forming subunits. We determined sites of phosphorylation in the N‐terminal and C‐terminal domains that are positioned to allosterically modulate channel activation by cytoplasmic calcium. One of the two voltage sensing domains (VSD2) encodes voltage sensitivity and inhibition by luminal calcium locks VSD2 in a ‘resting’ conformation, distinct from the activated VSDs observed in structures of other voltage‐gated ion channels. The structure shows how potent pharmacophore trans‐Ned‐19 allosterically acts to inhibit channel opening. In animals, trans‐Ned‐19 prevents infection by Ebola virus and Filoviruses by blocking fusion of the viral and endolysosomal membranes, thereby preventing delivery of their RNA into the host cytoplasm. The structure of TPC1 paves the way for understanding the complex function of these channels and may aid the development of antiviral compounds.


E71 To bind or not to bind: a reciprocal inhibition of SH3‐PRD domain interaction between BIN1 and Dynamin‐2. J. Laiman1, Y. Liu1; 1Institute of Molecular Medicine, National Taiwan University, Taipei, Taiwan Accurate expression and post‐translational modification are critical for proteins to function properly in maintaining normal structure and cellular processes in living organisms. Many proteins consist of several domains that could either regulate their own activities or that of others through protein‐protein interaction. One of such examples is SH3‐PRD domain interaction. BIN1/Amphiphysin 2, a membrane sculpting protein, contains N‐Bin/Amphiphysin/Rvs (BAR) domain capable of generating membrane curvature and binds to dynamin‐2 through its SH3 domain. Dynamin is a GTPase that mediates plasma membrane fission during endocytosis in eukaryotic cells. The muscle‐specific form of BIN1 is found on muscle T‐tubules and is important for its biogenesis. In contrast, we have previously discovered that dynamin‐2 is involved in the scission of T‐tubules. We analyzed membrane fission in vitro and found that BIN1 inhibits GTP‐dependent membrane scission of dynamin. Using fluorescence microscopy, we observed membrane tubulation by BIN1 that is absence when BIN1 is incubated with dynamin‐2. Interestingly, truncated BIN1 containing N‐BAR without SH3 domain promotes dynamin‐mediated membrane fission instead of inhibiting it; membrane tubulation by BIN1 N‐BAR domain could also be observed even in the presence of dynamin‐2, which indicates that the inhibition of both dynamin fission activity and BIN1 membrane tubulation is a result of SH3‐PRD domain interaction. This reciprocal inhibition is not due to membrane binding defect as shown by liposome binding assay. Rather, BIN1 affect the assembly efficiency of dynamin‐2 on the membrane. We then try to analyze mechanism to relieve SH3‐PRD interaction and found a potential phosphorylation site in PRD domain that would affect SH3 binding. These findings reveal the interplay between BIN1 and dynamin‐2 and their significant functions in our body.

E72 Differential sorting of planar cell polarity signaling receptors, Frizzled6 and Vangl2, at the trans Golgi Network. Y. Guo1, T. Ma1, R. Wang1, P. Lau1, Y. Huang1, R.W. Schekman2; 1Division of Life Science, Hong Kong University of Science and Technology, Hong Kong, Hong Kong, 2Department of Molecular and Cell Biology, University of California, Berkeley, Berkeley, CA Planar cell polarity (PCP) is mediated by a group of conserved signaling receptors that localize asymmetrically on opposing cellular boundaries. We have examined sorting of two asymmetrically localized PCP signaling proteins, Vangl2 and Frizzled6 (Fzd6), at the trans Golgi network (TGN). Previously, we demonstrated that the GTP‐binding protein, Arfrp1, and the clathrin adaptor complex 1 (AP‐1) are required for Vangl2 transport from the TGN. In contrast, TGN export of Fzd6 does not depend on Arfrp1 or AP‐1. Here, we reconstituted the vesicular release of Vangl2 and Fzd6 from the TGN in vitro, and the behavior of the two proteins in this assay suggests that they exit the TGN in separate vesicles. We also find that a TGN‐localized clathrin adaptor, epsinR, regulates TGN export of Fzd6 but not Vangl2. EpsinR forms a stable complex with clathrin heavy chain (CHC) and this complex directly interacts with Fzd6 C‐terminal cytosolic domain. Further, TGN export of Fzd6 depends on a conserved polybasic motif in Fzd6 C‐terminal cytosolic domain and depleting this motif inhibits binding of Fzd6 to the epsinR/CHC complex. Using super‐resolution imaging, we found Vangl2 in association with AP‐1‐decorated structures and Fzd6 in separate epsinR‐decorated structures as they exit the TGN. Our results


suggest that Vangl2 and Fzd6 are packaged into separate vesicles and use different clathrin adaptors, which may contribute to their asymmetric localizations.

Microsymposium 9: New Insights Into the Cell Division Mechanisms

E73 Yeast metabolic cycle: the relationship between metabolic state and the cell cycle at single‐cell resolution. A.J. Burnetti1,2,3, N.E. Buchler3,4; 1University Program in Genetics and Genomics, Duke University, Durham, NC, 2Cell and Molecular Biology Program, Duke University, Durham, NC, 3Center for Genomic and Computational Biology, Duke University, Durham, NC, 4Biology, Duke University, Durham, NC The budding yeast Saccharomyces cerevisiae undergoes synchronous metabolic oscillations between storage carbohydrate catabolism and respiration with periods between 45 minutes and several hours when grown in chemostat conditions. This yeast metabolic cycle (YMC) consists of rhythmic expression of large fractions of the transcriptome in addition to an oscillation in energy metabolism. The cell division cycle (CDC) also couples strongly to this oscillation, with all cells that divide during a given metabolic cycle doing so during a single pulse a fixed period after the switch from low to high oxygen consumption. The interaction of these coupled oscillators can reveal principles of how biological clocks coordinate with each other within cells despite varied periods. To date, all studies of the YMC have been performed at the population level, via readouts of oxygen consumption and metabolite production and bulk measurements of cell division. While this reveals important principles of the relationship between these two co‐existing oscillators, all coupling ultimately occurs within individual cells. We developed techniques to study the coupling of the YMC and CDC, using ratiometric pHluorin pH‐sensitive GFP and fluorescent cell cycle reporters. Via flow cytometry and live‐cell microscopy we observe the cell cycle state and metabolic state (as indicated by cytosolic pH) in individual cells. Our results reveal a population structure in which a ‘one‐to‐some’ coupling dominates, with all cells undergoing a metabolic shift each metabolic cycle, but with distinct differences in behavior between cells based on replicative age and cell mass.

E74 The Mad1/Mad2 spindle checkpoint complex is repurposed in development to promote cell cycle progression. P. Lara Gonzalez1, M. Moyle1, K. Oegema1, A.B. Desai1; 1Dept. of Cellular Molecular Medicine, UCSD, Ludwig Cancer Research, San Diego, CA The anaphase‐promoting complex / cyclosome (APC/C) is the essential ubiquitin ligase that, together with its coactivator Cdc20, promotes anaphase by targeting Cyclin B and other substrates for degradation. During mitosis, APC/C‐Cdc20 activity is controlled by the spindle checkpoint that acts to prevent chromosome loss by restraining anaphase onset in the presence of unattached kinetochores. Components of the spindle checkpoint include the evolutionarily conserved Mad and Bub, proteins, which catalyze formation of an APC/C inhibitor at kinetochores that targets the coactivator Cdc20 to restrain APC/C activation. Surprisingly, we found that the severe embryonic and post‐embryonic developmental defects associated with deletion of the genes encoding Mad1 and Mad2 in C. elegans are not due to loss of the spindle checkpoint, as deleting the gene encoding Mad3 or selectively removing Mad1 and Mad2 complex from unattached kinetochores does not give rise to these defects despite


abrogating checkpoint signaling.. Thus, Mad1 and Mad2 have an essential role in organismal development that is independent of canonical checkpoint signaling at unattached kinetochores. A comparison of the phenotypic profile of Mad2 removal to the profile of known essential genes in one of its target tissues, the C. elegans germline, revealed similarity to components that are required for mitotic entry, such as Cyclin B, Cdk1 and Cdc25. In agreement with this, Mad2 inhibition greatly reduced the rate of proliferation of germ cells, as well as a subset of embryonic lineages. These data suggest that the Mad1/Mad2 complex is activated by specific developmental cues to promote cell cycle progression independently of its established function at unattached kinetochores. Interestingly, this function still requires formation of the Mad1/Mad2 complex, the ability of Mad2 to undergo a conformational conversion following asymmetric dimerization, and inhibition of APC/C‐Cdc20, all of which are key features of the spindle checkpoint pathway. We propose that the ability of the Mad1/Mad2 complex to inhibit APC/C‐Cdc20 has been adapted in a kinetochore‐independent developmental context to promote progression into mitosis, most likely by allowing accumulation of cyclin B to a level sufficient for mitotic entry. Current experiments are directly testing the effect of Mad1/Mad2 complex inhibition of cyclin B accumulation and analyzing Mad1/Mad2 activation in specific developmental contexts. Overall, this effort has revealed how a core module of the spindle checkpoint pathway has been repurposed to provide an essential function in organismal development, and should influence interpretation of phenotypes associated with checkpoint component inhibitions in metazoans as well as therapeutic approaches desi

E75 Golgi‐derived microtubules play a major role in the disassembly of nuclear envelope during mitosis in human cells. K. Frye1, M. Fomicheva1, D. Yampolsky1, E. Kolobova2, V. Magidson3, J.R. Goldenring2, X. Zhu1, A. Khodjakov3, I. Kaverina1; 1Cell and Developmental Biology, Vanderbilt University, Nashville, TN, 2Surgery, Vanderbilt University School of Medicine, Nashville, TN, 3Wadsworth Center, New York State Department of Health, Albany, NY Entry into mitosis in vertebrate cells requires a rapid breakdown of the nuclear envelope (NEB). This process is mediated by microtubules and microtubule‐based motor proteins. However, the exact mechanism(s) of spatial and temporal controls of NEB remain obscure. Here we demonstrate that reorganization of the Golgi apparatus and a change in the microtubule‐organizing activity of this organelle enhances the efficiency of NEB. During late G2, the single Golgi complex starts to fragment, and the fragments relocate from the vicinity of the centrosome towards the nuclear envelope. This relocation is mediated by the interaction between the Golgi‐derived microtubules attached to the Golgi membranes and cytoplasmic dynein associated with the nuclear envelope. Inhibition of dynein anchoring at the nuclear envelope or depletion of Golgi‐associated MT nucleation factors prevents relocation of the Golgi towards the nuclear envelope. Interestingly, reorganization of the Golgi during prophase is independent of centrosome separation and/or location. Thus, centrosomal microtubules known to drive the assembly and positioning of a common Golgi complex during G1 and S periods of the cell cycle do not contribute to Golgi organization in late G2. As a result of the dynein‐mediated movement, the Golgi forms a characteristic symmetric distribution, encircling a large part of the nucleus. NEB is initiated at sites where Golgi membranes are tightly associated with the nuclear membrane. Electron microscopy analyses reveal a partial fusion between the Golgi and nuclear membranes. We propose that this fusion facilitates rapid disassembly of the nuclear envelope. Our data support the view that fragmentation of the Golgi membranes is necessary for timely and correct cell division and they


offer a new insight into the mechanism by which the Golgi regulates mitosis. Supported by NIH GM078373 and GM059363.

E76 HMMR/RHAMM balances motor forces needed to complete the spindle assembly checkpoint. H. Chen1, C.A. Maxwell1; 1Paediatrics, University of British Columbia, Vancouver, BC Mitotic progression to anaphase is restricted by the spindle assembly checkpoint until all chromosomes reach bipolar attachment with the mitotic spindle. Stable chromosome attachment requires a balance between outward forces directed by kinesins, Eg5 (kinesin‐5) and Kif15 (kinesin‐12 / human kinesin‐like protein 2), and inward forces directed by dynein. Motor activities are regulated, in part, by non‐motor adaptor proteins, such as targeting protein for XKlp2 (TPX2), and dysregulated motor activities induce aberrant spindle figures and delay mitotic progression. The receptor for hyaluronan mediated motility (HMMR/RHAMM) crosslinks microtubules to maintain spindle pole integrity, and interacts with TPX2 to promote mitotic spindle assembly. Here, we find that murine embryonic fibroblasts deficient for HMMR/RHAMM are retarded as they progress through metaphase. Using HeLa cells, we find that RHAMM is needed for progression into anaphase and its depletion results in sustained activation of the spindle assembly checkpoint as evidenced by a significant delay in mitotic progression, delayed Cyclin B degradation and reduced tension at kinetochores. Eg5 inhibition by STLC, but not depletion of Kif15, was able to alleviate mitotic delay and restore kinetochore tension, suggesting excess Eg5 activity in RHAMM‐silenced cells. Indeed, formation of the TPX2 and Eg5 inhibitory complex was significantly reduced in the absence of RHAMM, while TPX2 and Kif15 complex formation was not affected. In the absence of RHAMM, metaphase spindles appeared unfocused with reduced TPX2, but not Eg5 or Kif15, localized at spindle poles. Re‐expression of GFP tagged RHAMM variants that encoded a carboxy‐terminal domain known to locate TPX2 to spindle poles was sufficient to rescue mitotic kinetics and kinetochore tension. In a minor proportion of RHAMM‐silenced cells, prolonged metaphase delay was observed and resisted Eg5 inhibition but resolved through cohesion fatigue. Collectively, our findings demonstrate a novel role for RHAMM in regulating the TPX2‐Eg5 complex along the mitotic spindle, which balances motor forces needed for chromosome attachment, alignment and segregation.

E77 Defining a cell's internal compass: Focal adhesions control cleavage furrow shape and spindle tilt during mitosis. N. Taneja1, A.M. Fenix1, L. Rathbun2, B.A. Millis1, M.J. Tyska1, H. Hehnly2, D.T. Burnette1; 1Cell and Developmental Biology, Vanderbilt University, Nashville, TN, 2Cell and Developmental Biology, SUNY Upstate Medical University, Syracuse, NY Cytokinesis, the process by which two daughter cells physically separate from each other, is driven by contraction of an actomyosin ring. Epithelial cells dividing in vivo often show asymmetric furrowing of the cleavage furrow, with ingression proceeding primarily from the basolateral domain. This asymmetric shape of the cleavage furrow is essential for normal differentiation of progenitors within the mouse neuroepithelium and epidermis. How asymmetric shapes of the cleavage furrow arise is poorly understood. The three dimensional shape of a cell is ultimately determined by a balance of cellular forces. In crawling cells during interphase, a dynamic balance between actomyosin contraction and integrin‐based adhesion to the substrate effects shape changes that facilitate cell movement. It has been proposed that a similar balance causes asymmetric furrowing in adhesive cells in culture. The


bottom (substrate bound side) of the cell ingresses to a lesser extent than the top of the cell, suggesting that substrate attachment may oppose ingression of the contractile ring. Substrate adhesion during mitosis is primarily thought to occur through retraction fibers that form as the cell rounds up during mitotic entry. Here we report the presence of large focal adhesions (FAs) underneath the cell body, which we term mitotic FAs, that are spatially and temporally distinct from retraction fibers. Using both high and super resolution imaging approaches, we show that mitotic FAs drive asymmetric furrowing during cytokinesis in single cells. Our results also unexpectedly reveal that mitotic FAs can direct furrow ingression in epithelial cells in co‐ordination with cell‐cell contacts in both two and three‐dimensional systems. Furthermore, contrary to the present view that integrin mediated adhesion maintains the spindle parallel to the substrate, our results show that presence of larger mitotic FAs correlates with higher spindle tilt. Finally, we show the older spindle pole displays a higher propensity to be closer to the substrate and is localized to the daughter cell that makes more mitotic FAs. These results therefore reveal substrate adhesion as a key mechanism contributing to cell shape, polarity and fate during mitosis.

E78 The Timing of Midzone Stabilization during Cytokinesis Depends on Myosin II Activity and an Interaction between INCENP and Actin. J.E. Landino1, R. Ohi1; 1Cell and Developmental Biology, Vanderbilt University, Nashville, TN The final steps of cell division are tightly coordinated in space and time, but whether mechanisms exist to couple the actin and microtubule (MT) cytoskeletons during anaphase and cytokinesis (C phase) is largely unknown. During anaphase, MTs are incorporated into an anti‐parallel array termed the spindle midzone (midzone MTs), whereas F‐actin and non‐muscle myosin II, together with other factors, organize into the cleavage furrow. Previous studies in somatic cells have shown that midzone MTs become highly stable after furrows have begun ingression, indicating that furrow‐to‐MT communication may occur. Midzone formation is also inhibited in fly spermatocytes that fail to form a cleavage furrow and during monopolar cytokinesis when myosin contractility is blocked by blebbistatin. We show here that midzone MT stabilization is dependent on actomyosin contraction, suggesting that there is active coordination between furrow ingression and microtubule dynamics. Midzone microtubule stabilization also depends on the kinase activity of Aurora B, the catalytic subunit of the chromosomal passenger complex (CPC), uncovering a feedback mechanism that couples furrowing with microtubule dynamics. We further show that the CPC scaffolding protein INCENP (inner centromere protein) binds actin, an interaction that is important for cytokinesis and for midzone MT stabilization following furrow ingression. Stabilization of midzone MTs with low amounts of Taxol rescues cytokinesis in INCENP actin‐binding mutant‐expressing cells. Collectively, our work demonstrates that the actin and microtubule cytoskeletons are coordinated during cytokinesis and suggests that the CPC is integral for coupling furrow ingression with midzone microtubule stabilization.


E79 Investigating the role of Condensins in mitotic chromosome architecture using quantitative super‐resolution microscopy. N. Walther1, B. Koch1, M. Kueblbeck1, Y. Cai1, M.J. Hossain1, A.Z. Politi1, M. Lampe2, J. Ries1, J. Ellenberg1; 1Cell Biology Biophysics Unit, European Molecular Biology Laboratory (EMBL), Heidelberg, Germany, 2Advanced Light Microscopy Facility, European Molecular Biology Laboratory (EMBL), Heidelberg, Germany A fundamental structural and functional change of the human genome is the compaction of replicated interphase chromatin into rod‐shaped mitotic chromosomes. This process, referred to as mitotic chromosome condensation, is essential for faithful genome partitioning and involves two structural maintenance of chromosomes (SMC) complexes, Condensin I and II. The two Condensin isoforms localize to chromosomes at different cell cycle stages and reportedly occupy distinct regions along the chromosome axis. They have been assumed to promote different aspects of compaction and are part of a non‐histone scaffold of mitotic chromosomes. However, if a higher order Condensin structure along the mitotic chromosome axis exists, its molecular architecture as well as how it relates to Condensins’ essential function remains unclear. To address this gap in our knowledge, we are investigating the organization of Condensin complexes along the chromosome axis throughout mitosis by quantitative super‐resolution microscopy. We have established 3D stimulated emission depletion (STED) and stochastic optical reconstruction microscopy (STORM) imaging pipelines for Condensins, using genome‐edited human cell lines in which the endogenous genes for shared and isoform‐specific Condensin subunits are homozygously fluorescently tagged. In addition, we have determined the absolute number of Condensin subunits at different mitotic stages using fluorescence correlation spectroscopy (FCS)‐calibrated confocal time‐lapse imaging. Integration of our quantitative dynamic and 3D super‐resolution microscopy data allows us to provide, to our knowledge, the first quantitative 3D map of Condensins in single dividing cells and propose models for how they organize mitotic chromosomes.

E80 Rho GTPases signaling: from cytoskeleton regulation to nuclear DNA damage response and repair. F.L. Forti1, Y.T. Magalhaes1; 1Department of Biochemistry, Institute of Chemistry, University of Sao Paulo, Sao Paulo, Brazil Rho GTPases are overexpressed/overactivated in tumorigenic cells and associated with poor cancer prognosis, because are involved in different and essential steps of cancer cell progression including cell invasion, metastasis and uncontrolled proliferation. This work compares the participation of the three typical Rho GTPases, RhoA, Rac1 and Cdc42, in the responses of two tumor cell lines after DNA damage caused by ultraviolet radiation (UV). The modulation of enzymatic activity for the three GTPases was achieved in HeLa and MeWo cells by transduction with retroviral vectors containing the dominant negative (RhoA‐N19, Rac1‐N17, Cdc42‐N17) or constitutively active (RhoA‐V14, Rac1‐V12, Cdc42‐V12) mutants, or by the use of pharmacological inhibitors (C3 toxin for RhoA; ML141 or ZCL278 for Cdc42; NSC23766 for Rac1). Then cells were subjected to UV treatments and analyzed for cell survival, DNA repair and regulation of DNA damage response (DDR) pathways. We constructed interactomes for the three GTPases, before and after the cellular and molecular experiments, by using the Cytoscape platform and plug‐ins for the identification of cluster, hubs and analysis of topological and centrality parameters in the networks. Inhibition of RhoA or Rac1 led to decrease in cell proliferation and survival,


an increase in DNA damage and a delay/defect in DNA repair in both cell lines, while the activation of DNA damage response pathways (phosphorylation of H2AX, Chk1/2 and p53 proteins) was strongly compromised with the inhibition of both GTPases compared with normal cells for this enzyme. On the other hand, the constitutively active mutant Cdc42‐V12 showed higher radiosensitivity to UV radiation, with delay/defect in DNA repair and reduced phosphorylation of H2AX and Chk1 proteins. Analysis of interaction networks for the three GTPases showed a considerable change of its protein partners and the suggestion of potential indirect mechanisms of action of these enzymes in the regulation of DDR proteins. Thus, the modulation of these GTPase’s activity can affect the cellular responses to UV radiation leading to a delay or inefficiency of DNA repair mechanisms and consequent reduced cell survival. Additionally, we identified promising interactions of these enzymes with other proteins through interactomes, enabling us the construction of hypotheses for possible indirect mechanisms of action of RhoA, Rac1 and Cdc42 in the maintenance of genomic stability. (Supported by FAPESP, Capes, CNPq).

E81 Diaphanous formin mDia2 bridges small GTPase signaling with nuclear environment to regulate stable CENP‐A loading at the centromere. C. Liu1, Y. Mao1; 1Pathology and Cell Biology, Columbia University, NEW YORK, NY Accurate chromosome segregation in mammalian cells is guided by the centromere, a specialized chromosome region epigenetically defined by the histone H3 variant centromere protein A (CENP‐A). In order to maintain centromere identity against CENP‐A dilution caused by S phase genome replication, new CENP‐A molecules are deposited at preexisting centromeres in G1 phase of the cell cycle. Although several important stages and molecular players have been identified in CENP‐A replenishment, a complete understanding is missing regarding how new CENP‐A proteins become stably incorporated into centromeric nucleosomes. We have previously found that diaphanous formin mDia2 is required for newly synthesized CENP‐A to be stably deposited at G1 centromeres (Liu and Mao, JCB 2016). Here we provide temporal evidence that mDia2 functions downstream of the MgcRacGAP‐dependent small GTPase molecular switch during early G1 phase. Quantitative ratiometric live cell imaging revealed the temporal requirement of the small GTPase pathway, and that ectopic activation of endogenous mDia2 in the absence of MgcRacGAP is sufficient to maintain stable CENP‐A loading. This novel function of mDia2 depends on its nuclear localization and its ability to polymerize actin. Dynamic and short actin filaments were observed in early G1 nuclei in an mDia2 dependent manner. Indeed, centromeric CENP‐A level was reduced in cells depleted of IPO9 and expressing a nonpolymerizable actin mutant. To further examine the physical role of mDia2 mediated nuclear actin during CENP‐A replenishment, single particle tracking of centromere movement was performed in early G1 nuclei over the time scale of initial CENP‐A loading. Quantitative analysis revealed subdiffusive behaviors where normal G1 centromeres movements are relatively confined. The confinement of centromere motion, however, is significantly impaired upon mDia2 knockdown. Thus, our findings suggest the diaphanous formin mDia2 forms a link between the upstream small GTPase signaling and the downstream confined viscoelastic nuclear environment, and therefore regulates the stable assembly of new CENP‐A containing nucleosomes to mark centromere’s epigenetic identity.


Microsymposium 10: Spatial Organization of the Cell

E82 Amyloid‐like Self‐Assembly of a Cellular Compartment. E. Boke1, M. Ruer2, M. Wuhr1,3, M. Coughlin1, R. Lemaitre2, S.P. Gygi3, S. Alberti2, D.N. Drechsel2, A.A. Hyman2, T.J. Mitchison1; 1Department of Systems Biology, Harvard Medical School, Boston, MA, 2Max Planck Institute of Molecular Cell Biology and Genetics, Dresden, Germany, 3Department of Cell Biology, Harvard Medical School, Boston, MA A key characteristic of female germ cells, oocytes, is that their complement of mitochondria and RNA is kept intact for decades before fertilization. However, we have little idea how these are protected for such long periods of time. One prominent feature of dormant oocytes is the Balbiani body, which is conserved from fish to humans. These enigmatic structures, which can be thought of as a non‐membranous compartment, or perhaps super‐organelle, are prominent in the cytoplasm of early oocytes, and are packed with mitochondria, ER, Golgi and RNA. Very little is known about their organization and structure. We have studied the organization of the Balbiani body in Xenopus, the large size of which makes it amenable to biochemistry. We show by quantitative mass spectrometry that the most enriched protein in the Balbiani body of Xenopus is Xvelo, a homolog of the germ line protein Bucky ball in zebrafish. Xvelo has a prion‐like domain in its N terminus, which is sufficient and necessary to target to the Balbiani body. We show that Xvelo forms a mitochondria‐embedding, amyloid‐like matrix pervading the entire volume of the Balbiani body. Moreover, recombinant Xvelo forms micron‐sized networks in vitro that can cluster mitochondria on its own in a cell free system in a prion‐like domain dependent manner, thus reconstituting aspects of a Balbiani body in vitro. We propose that the Balbiani body forms by amyloid‐like aggregation of Xvelo. Because the prion‐like domain of Xvelo is conserved in different germ plasm‐related proteins in other species, Balbiani body formation by amyloid‐like assembly could be a conserved feature in evolution for maintaining the immortal character of germ cells.

E83 Multicolor Super‐Resolution Microscopy And Correlation Analysis Reveal 3D Structure of Ciliary Transition Zone. X. Shi1,2, G. Garcia2, J.F. Reiter2, B. Huang1,2; 1Pharmaceutical Chemistry, University of California ‐ San Francisco, San Francisco, CA, 2Biochemistry and Biophysics, University of California ‐ San Francisco, San Francisco, CA Ciliary transition zone (TZ) is a small (~300 nm) barrel‐shaped structure between the cilium and the basal body that regulates ciliogenesis and signaling pathways important for development. However, the molecular architecture of TZ and the spatial relationship between TZ components and intraflagellar transporters are still largely unknown. Utilizing a ratiometric approach for multicolor Stochastic Optical Reconstruction Microscopy (STORM), we determined the localizations of TZ proteins, intraflagellar transporters (IFT) and a hedgehog signaling receptor Smoothened in wild type and mutants. The super resolution images revealed more complex symmetries of TZ protein Nphp1, Tmem231 and B9d1 different from the 9‐fold symmetry of axoneme and transition fiber. Multicolor images captured the rotational offsets between Tmem231, Nphp1, Rpgrip1L, B9d1 and IFT88 at the cross sections of cilia. Quantitative information of the symmetries and rotational offsets was obtained by cross‐correlation


analysis. Super resolution imaging combined with quantitative analysis opens a door to many interesting structure‐function questions about ciliary gating and intraflagellar transporting mechanisms.

E84 Nanoscopic compartmentalization of membrane protein motion at the axon initial segment. D. Albrecht1, C.M. Winterflood2, M. Sadeghi3, F. Noe3, H. Ewers1; 1Chemistry and Biochemistry, Free University Berlin, Berlin, Germany, 2Randall Division, King's College London, London, United Kingdom, 3Department of Mathematics and Computer Science, Free University Berlin, Berlin, Germany The axon initial segment (AIS) is enriched in specific adaptor, cytoskeletal and transmembrane molecules. During AIS establishment, a membrane diffusion barrier is formed between the axon and the somatodendritic domain. Recently, an axonal periodic pattern of actin, spectrin and ankyrin forming 190 nm distanced, ring‐like structures has been discovered. However, whether this structure is related to the diffusion barrier function is unclear. Here, we performed single‐particle tracking timecourse experiments on hippocampal neurons during AIS development. We analyzed the mobility of lipid‐anchored molecules by high‐speed single‐particle tracking and correlated positions of membrane molecules with the nanoscopic organization of the AIS cytoskeleton. We observe a strong reduction in mobility early in AIS development. Membrane protein motion in the AIS plasma membrane is confined to a repetitive pattern of ~190 nm spaced segments along the AIS axis as early as DIV4 and this pattern alternates with actin rings. Mathematical modeling shows that diffusion barriers between the segments significantly reduce lateral diffusion along the axon. Our data provide a new mechanism for the AIS diffusion barrier.

E85 Quantitative analysis of the 3D spatial organization of cells and organelles. J. Chen1, D. Dedham1, A. Walter1, R. Wercberger1, J. Kuhn1, M.A. Le Gros1,2, A. Basbaum1, C.A. Larabell1,2; 1Anatomy, University of California San Francisco, San Francisco, CA, 2Molecular Biophysics Integrated Bioimaging, Lawrence Berkeley National Laboratory, Berkeley, CA Regulation of the spatial organization of cells and organelles remains poorly understood. One course of action required to elucidate this process is to visualize and quantify all organelles in the cell in three dimensions and, to account for cell heterogeneity, to examine a large number of cells. We are doing this using soft x‐ray tomography, a high‐resolution (better than 50 nm), quantitative imaging technology that generates high‐contrast views of organelles in intact cells in the near‐native state. Image contrast is generated by the absorption of x‐rays by cellular constituents, and absorption is linear with thickness and concentration of organic molecules in each voxel. Therefore each voxel contains a value, the linear absorption coefficient (LAC), that reveals the concentration of bio‐organic molecules in that voxel (3D pixel). We can use these values to identify and segment individual organelles. For example, mitochondria can be identified by histogram using their LAC values (between 0.31‐0.35 µm‐1) and computationally segmented, making identification rapid and more objective. We have used this technique to examine three cell types, fibrosarcoma cells, mouse lymphocytes, and microglial cells. Preliminary studies of ten of each cell type show that nuclear volume of ranges from 102 µm3 to 450 µm3, and mitochondrial volume ranges from 5.5 µm3 to 26.2 µm3 and quantitation of other organelles is in progress. In summary, we show that soft x‐ray tomography reveals both spatial information about the distribution of cell organelles and quantitative information about the numbers and volumes of cellular organelles.


E86 Using optogenetics to uncover principles governing cell size scaling relationships in giant yeast. C. Allard1,2, F. Decker1,3, O.D. Weiner1,4, B.R. Graziano1,4, J.E. Toettcher1,5; 1Marine Biological Laboratory, Woods Hole, MA, 2Department of Biochemistry, Dartmouth Medical School, Hanover, NH, 3Max Planck Institute of Molecular Cell Biology and Genetics, Dresden, Germany, 4Cardiovascular Research Institute, UCSF, San Francisco, CA, 5Department of Molecular Biology, Princeton University, Princeton, NJ Among the most basic aspects of cell physiology that must be regulated is size. Precise control of cell size is evident across nearly all forms of life, and many cellular properties, such as DNA content and organelle volume, display a strong correlation with cell size. A variety of models have been proposed describing how cells govern their volumes between successive rounds of division, including ‘sizer’ (division is triggered when cell growth reaches a particular volume), ‘incremental’ (cells add a constant volume between division events), and ‘timer’ (the duration of the cell cycle is specified independently of cell volume). A major challenge in discriminating between such models arises from the tendency of cells to undergo only small changes in volume—often less than twofold—as they traverse the cell cycle. Overcoming this obstacle requires either the development of sophisticated tools that can measure cell volume more accurately and precisely or perturbations that increase the range over which cell volumes are distributed. While both approaches have been employed in past studies, introducing size‐altering perturbations has typically required the use of genetic mutations (e.g., temperature‐sensitive alleles) where cells must be subjected to an environmental stress to induce changes in size. As such stresses have broad impacts on cell physiology, it can be difficult to determine whether any size control mechanisms observed under such conditions also apply to unstressed cells. Here, we employed an optogenetics‐based system to acutely and reversibly block polarity establishment in the budding yeast S. cerevisiae. Importantly, as our system relies only on red/IR light for activation/deactivation, we can maintain constant environmental conditions throughout our experiments and remove them as convoluting factors. While blocked, cells continued to grow isotropically with their growth rate remaining proportional to their surface area, despite undergoing a 30‐fold increase in volume. Upon releasing the polarization block, these ‘giant’ yeast cells initiated bud formation, growth, and cytokinesis. Remarkably, the duration between bud emergence and cytokinesis remained nearly constant, regardless of the size of the mother. Furthermore, we observed a strong correlation between cell volume at birth vs. cell volume added between budding events, which is inconsistent with an ‘incremental’ size regulation model. Our observations thus support a model where yeast cell size is largely specified by a ‘timer’ mechanism.

E87 Predicting the targeting of tail‐anchored proteins to subcellular compartments in mammalian cells. J.L. Costello1, I. Castro1, F. Camões2, T.A. Schrader1, D. McNeall3, S. Gomes2, E. Giannopoulou4, V. Pogenberg4, N. Bonekamp2, D. Ribeiro2, M. Wilmanns4, M. Schrader1, M. Islinger5; 1Biosciences, University of Exeter, Exeter, United Kingdom, 2Centre for Cell Biology, University of Aveiro, Aveiro, Portugal, 3Met Office, Hadley Center, Exeter, United Kingdom, 4c/o DESY, EMBL Hamburg, Hamburg, Germany, 5Institute of Neuroanatomy, Medical Faculty Mannheim, University of Heidelberg, Mannheim, Germany Tail‐anchored (TA) proteins contain a single transmembrane domain (TMD) at the C‐terminus, anchoring them to organelle membranes where they mediate a variety of critical cellular processes. Mutations in


individual TA proteins cause a number of severe inherited disorders. In order to function correctly the positioning of TA proteins in the appropriate organelle membrane is essential, however, the molecular mechanisms and signals facilitating proper TA protein targeting are not fully understood, in particular in mammals. In recent years biochemical parameters impacting targeting, such as TMD hydrophobicity, have been identified and constituents of the targeting machinery, such as the GET proteins which target TA proteins to the ER, have been discovered. We, and others, have also shown that several TA proteins localise to multiple different cellular membranes, in particular mitochondria and peroxisomes, adding another layer of complexity. Wider, integrated studies investigating how targeting is coordinated to direct TA proteins to individual or multiple organelles are currently lacking. Here, we identify additional TA proteins at peroxisomes or shared by multiple organelles in mammals and reveal that a combination of TMD hydrophobicity and tail charge determines targeting to distinct organelles. We demonstrate that subtle alterations in those physicochemical parameters can shift TA protein targeting between organelles, explaining why peroxisomes and mitochondria share many TA proteins. Focussing on peroxisomes, we demonstrate that high affinity for Pex19 leads to peroxisomal targeting, whilst a stepwise reduction in Pex19 affinity, correlating with an increase in Get3 activity, leads to mitochondrial and finally ER targeting. This suggests a simple evolutionary mutation principle to switch localisation between different organelles. Our analyses enabled us to allocate characteristic physicochemical parameters to different organelle groups. This classification allows for the first time, successful prediction of the location of uncharacterized TA proteins.

E88 Ready, Set, Miro! Miro1‐mediated mitochondrial positioning at the leading edge accelerates cell movement. MH. Schuler1, B. Cunniff1,2,3, A. Lewandowska1, T. Kirchhausen2,3, J.M. Shaw1; 1Department of Biochemistry, University of Utah, Salt Lake City, UT, 2Department of Cell Biology, Harvard Medical School, Bostson, MA, 3Program in Cell and Molecular Medicine, Boston Children’s Hospital, Boston, MA *Co‐communicating authors. In eukaryotic cells, mitochondria are strategically positioned to meet localized energy requirements and participate in calcium buffering and reactive oxygen species signaling. Polarized cell migration depends upon the coordinated protrusion of a leading edge, formation of focal adhesions and retraction of the cell body. The biochemical processes underlying the iterative reorganization of cell shape and anchorage consume considerable amounts of energy and are regulated by a multitude of signaling cascades. We used mouse embryonic fibroblasts (MEFs) deficient for the mitochondrial Rho‐GTPase 1 (Miro1) to investigate the role of mitochondrial positioning during cell movement. Miro1 is a central component of the multi‐protein complex that mediates mitochondrial transport along microtubule tracks. Deletion of Miro1 restricts the mitochondrial network to the perinuclear area in MEFs without affecting function of the organelle. We show that Miro1‐/‐ MEFs migrate slower than wild‐type controls in the context of collective and single cell migration. Actin dynamics including actin ruffling and protrusion of the lamellipodium are perturbed in Miro1‐/‐ MEFs compared to Miro1+/+ MEFs. Moreover, impaired transport of mitochondria to the leading edge decreases focal adhesion stability in Miro1‐/‐ MEFs, leading to the rapid turnover of newly formed adhesions. Re‐expression of Miro1 redistributes the mitochondrial network to the cell periphery and rescues cell migration defects to levels of wild‐type cells. Together, our data indicate that Miro1‐mediated mitochondrial positioning at the leading edge supports efficient cell migration by promoting remodeling of the actin cytoskeleton and focal adhesion stability.


E89 PhotoGate Microscopy for Tracking Single Molecules in Crowded Environments. V. Belyy1,2, S. Shih3, J. Bandaria3, Y. Huang1, R. Lawrence4, R. Zoncu4, A. Yildiz1,3,4; 1Biophysics, University of California, Berkeley, Berkeley, CA, 2Biophysics, University of California, San Francisco, San Francisco, CA, 3Department of Physics, University of California, Berkeley, Berkeley, CA, 4Department of Molecular Cell Biology, University of California, Berkeley, Berkeley, CA Tracking single molecules inside cells reveals the dynamics of biological processes, including receptor trafficking, signaling and cargo transport. However, individual molecules often cannot be resolved inside cells due to their high density. We developed the PhotoGate technique that controls the number of fluorescent particles in a region of interest by repeatedly photobleaching its boundary with a steerable focused laser beam. PhotoGate bypasses the requirement of photoactivation to track single particles at surface densities two orders of magnitude greater than the single‐molecule detection limit. Using this method, we observed ligand‐induced dimerization of a receptor tyrosine kinase at the cell surface and directly measured binding and dissociation of signaling molecules from early endosomes in a dense cytoplasm with single molecule resolution. We additionally developed a numerical simulation of PhotoGate that models diffusion and photobleaching from first principles and allows users to optimize experimental parameters prior to performing the experiment.

E90 Mapping the spatial and functional interactions of transition zone proteins and nucleoporins during ciliary gating. D. Takao1, L. Wang1, A. Boss1, T.L. Blasius1, K.J. Verhey1; 1Cell and Developmental Biology, University of Michigan Medical School, Ann Arbor, MI As the cell’s antennae, cilia have a unique protein and lipid composition that is generated and maintained by gating mechanisms that exist in a specialized region at the cilium base, the transition zone (TZ). Previous work has implicated both resident TZ proteins and nucleoporins in gated entry of ciliary proteins. However, the structural and functional details of ciliary gating, including the exact location and dynamics of gating components, have yet to be determined. To establish a spatial and functional map of protein‐protein interactions during ciliary gating, we utilized the bimolecular fluorescence complementation (BiFC) assay. We find that nucleoporins interact with cytosolic proteins and with membrane proteins containing long cytosolic tails, nephronophthisis (NPHP) proteins show varying interactions depending on the location of the split YFP tag, and MKS module proteins show strongest interactions with ciliary membrane proteins. These results provide strong support for the functional roles of nucleoporin, NPHP, and MKS module proteins during ciliary gating and begin to define the orientation of gating components within the transition zone. Using fluorescence recovery after photobleaching (FRAP), we find that nucleoporins and NPHP proteins show relatively fast turnover at the base of the cilium (t½ ranging from 2‐10 min) whereas MKS module proteins show little to no turnover. These results begin to define the spatial architecture of the transition zone as consisting of an outer ring of membrane‐associated MKS module components coupled to an inner ring of dynamic NPHP‐ and nucleoporin‐containing components. Taken together, these results defining the interactions and dynamics of transition zone components provide important clues to understand ciliary gating mechanisms.


Microsymposium 11: Development, Regeneration and Wound Healing

E91 Multicellular context contributes to cell type specific protection against cytokinesis failure. T. Davies1, N. Romano Spica1, Y. Zhuravlev1,2, M. Shirasu‐Hiza2, J. Dumont3, J.C. Canman1; 1Department of Pathology and Cell Biology, Columbia University, New York, NY, 2Department of Genetics and Development, Columbia University, New York, NY, 3Institut Jacques Monod, Paris, France Accurate cytokinesis, the physical division of one cell into two, is essential for the development of all multicellular organisms. In animal cells, cytokinesis is driven by constriction of an actomyosin contractile ring. Conventionally, the molecular mechanisms of cytokinesis are thought to be the same in all dividing animal cells. However, it is becoming clear that cytokinesis has more plasticity, or cell‐based variation in functional regulatory pathways, than previously appreciated. For example, human genetics has revealed that some cell types are highly sensitive to mutation of cytokinesis genes, while others are naturally programed to fail in cytokinesis. Indeed, cells receive multiple environmental cues that likely play key roles in regulating division. Thus to fully understand cytokinesis, we need to determine how it is differentially regulated in different cell types. To address this question, we used fast‐acting (~20s), temperature sensitive (ts) cytokinesis‐defective mutants in C. elegans to study division in the 4‐to‐8 cell embryo. At this early stage of development, each individual cell already has a specific cell identity leading to a distinct cell lineage that is controlled by cell‐cell contacts, cell polarity, and conserved cell fate‐signalling molecules (e.g. Notch, Wnt, Src). For each individual cell, we probed the requirement for two key contractile ring components: diaphanous formin (CYK‐1) and the motor myosin‐II (NMY‐2), which respectively assemble and constrict f‐actin in the contractile ring. By rapidly shifting ts mutant embryos to restrictive temperature, we determined the cell specificity of protein function required for cytokinesis 1) across a thermal range and 2) throughout the cell cycle. Both analyses revealed cell type specific requirements for formin, but not myosin‐II activity. In formin(ts) mutants, a shift to restrictive temperature led to cytokinesis failure in two of the four cells, as expected. However, in the same embryo, two of the four cells (EMS and P2) were often protected from cytokinesis failure and successfully divided, even with little or no formin activity and dramatically reduced contractile ring f‐actin. This cell type specific protection is lost upon physical isolation of these cells from the context of the embryo in formin(ts) mutants. This result indicates that non‐cell autonomous mechanisms provide cytokinesis protection of specific cell types. Together, our results show that the requirements for a successful contractile ring vary between cells and that this depends on multicellular context. We are currently investigating which features of a multicellular environment (e.g. cell contacts, adhesions, and signalling) affect the molecular requirements for cytokinesis.

E92 TRPV4 activity modulates development of the fetal lung. J.T. Morgan1, P.A. Sariano1, W.G. Stewart1, J.P. Gleghorn1; 1Biomedical Engineering, University of Delaware, Newark, DE Embryonic development of the lung requires the coordinated morphogenesis of multiple tissue compartments, including the epithelial lined airways, the pulmonary vasculature, and smooth muscle. The spatial and temporal organization of these tissue compartments is essential for proper signaling among tissues and formation of the correct lung architecture, which is critical for respiratory function and survival. Clinical observations as well as in vivo and ex vivo studies suggest that the dynamic


regulation of pressure and smooth muscle mediated fluid motion throughout the developing airways is essential for overall pulmonary development. However, it remains unknown how these mechanical forces on the airway epithelium are transduced into molecular and transcriptional changes within the cells. In this study, we demonstrate a previously unknown function of a well characterized mechanosensitive Ca2+ channel, TRPV4, in lung development. TRPV4 expression was confirmed in embryonic day (E) 12.5 and E13.5 at the protein and transcript levels and whole mount immunofluorescence revealed strong TRPV4 localization to the epithelium, subepithelial mesenchyme, and major pulmonary vessels. E12.5 lung explants were cultured at the air‐fluid interface in media supplemented with a TRPV4 agonist (GSK1016790A), inhibitor (GSK205), or vehicle control. Timelapse imaging at 1 Hz was used to assess airway smooth muscle contractility and airway branching dynamics. Lung explants were subsequently fixed and whole mount immunofluorescence was used to visualize the airway (E‐cadherin) and vasculature (PECAM). Pharmacological modulation of TRPV4 activity regulated the morphogenesis of all three major tissue compartments in the developing lung: epithelium, smooth muscle, and vasculature. TRPV4 agonism demonstrated significant increase in airway smooth muscle contraction frequency and moderate increases in airway branching and vasculogenesis. TRPV4 inhibition reduced airway smooth muscle peristalsis contraction frequency by ~55% and branching was reduced by ~25%. In the vascular compartment, TRPV4 inhibition resulted in a near complete loss of the vasculature. This work demonstrates that TRPV4 is broadly expressed in the developing lung and TRPV4 activity has profound impact on the coordinated morphogenesis of three essential tissue compartments in the lung. These studies identify TRPV4 as a potential mechanosensor involved in transducing mechanical strain on the airways to molecular and transcriptional events that regulate development. Further studies will focus on elucidating pathways downstream of TRPV4 responsible for signaling between tissue compartments and more broadly the critical role of Ca2+ signaling in lung development.

E93 Monitoring Dendrite Regeneration after Injury in vivo. K.L. Thompson‐Peer1,2, L. DeVault1,2, T. Li1,2, L.Y. Jan1,2, Y.N. Jan1,2; 1Physiology, University of California, San Francisco, CA, 2Howard Hughes Medical Institute, San Francisco, CA Neurons receive information along dendrites, and send electrical signals along axons to synaptic contacts. Traumatic brain injury, stroke, and other clinical conditions result in damage to dendrites. The factors that control axon regeneration have been examined in many systems, but research into dendrite regeneration has been neglected. Ramón y Cajal observed that dendrites in the cortex fail to regenerate after severing, yet observed dendrite sprouting in the spinal cord after contusion. The extensive literature on normal dendrite patterning of class IV da sensory neurons in the Drosophila peripheral nervous system allows us to explore fundamental features of dendrite regeneration, specifically looking at the ways in which dendrite growth after injury is similar to or diferent from normal dendrite outgrowth. We observe robust injury‐induced growth which occurs after full dendrite removal. In many ways regenerated dendrites are similar to uninjured dendrites, such as the trafficking of sensory ion channels and the expression of cell‐type specific transcription factors. However the growth and morphology of injury‐induced dendrites is significantly different from uninjured dendrites. Unlike normal dendrites, regenerating dendrites fail to avoid crossing over other branches from the same neuron. Regenerated dendrites are also pruned differently from uninjured branches. Injuring neurons at different developmental stages highlights the variable patterning of regrowing dendrites. Together our work describes essential features of dendrites grown in response to acute injury.


E94 Cellular Neighborhood Influences the Equilibria of Phenotypic States in a Heterogeneous Population. C. Krishnamurthy1, A. Paulson2, Z.J. Gartner1; 1Pharmaceutical Chemistry, University of California, San Francisco, San Francisco, CA, 2Biomedical Sciences Program, University of California, San Francisco, San Francisco, CA Tumor heterogeneity poses serious challenges in the diagnosis and treatment of cancer, and derives from both genetic and non‐genetic sources. Recent evidence suggests that steady state distributions of heterogeneous cell states in tumors and tumor cell lines can arise from stochastic state transitions within a plastic tumor cell population. However, the role of the local microenvironment in determining the probability of state transitions, and thus the overall distribution of cell types, has not been explored. To investigate the role of the microenvironment on phenotype distributions, we utilize the transformed epithelial cell line MCF10AT. MCF10AT cells have a stable distribution of epithelial and mesenchymal phenotypes, which can be distinguished by their differential expression of cell surface markers CD49f and CD24. Enriched populations of these cells relax to the same steady state population‐level distribution as the parental cell line over time. Surprisingly, single cell clones established from the parental cell line are homogeneous in their morphology, span a range of EMT phenotypes, and do not return to the parental steady state phenotype distribution. Transcriptional analysis of the stable epithelial and mesenchymal clones revealed differential expression of ECM molecules, secreted factors, and cell‐adhesion molecules, suggesting a role for local cell‐cell and cell‐ECM communication in reinforcing cellular phenotypes. Consistent with this notion, we found laminin and fibronectin biased the parental population towards the epithelial and mesenchymal phenotypes, respectively. Additionally, when populations of epithelial‐ or mesenchymal‐like cells purified by CD49f and CD24 were co‐cultured with stable clones, the distribution cell states was altered. Thus, we hypothesize that local feedback from the extracellular matrix, neighboring cells or secreted factors can counterintuitively both maintain a heterogeneous population or stabilize homogeneous cell states in a differentiation hierarchy. To better understand the interactions between heterogeneous cell states within a population, we generate a cellular Potts model to allow us to explore the impact of differential growth rates and probabilistic cell‐fate changes, together with cell‐cell interactions that locally modify these probabilities. Moreover, we investigate how perturbing local microenvironmental cues can influence cellular phenotypic transitions. These results may have significant implications for our understanding of tumor heterogeneity and its relationship with the local microenvironment.

E95 Flares of active RhoA locally reinforce tight junctions. R.E. Stephenson1, T. Higashi1, B. Coy1, T.R. Arnold1, A.L. Miller1; 1Molecular, Cellular, and Developmental Biology, University of Michigan, Ann Arbor, MI The small GTPase RhoA is an important regulator of cell‐cell junctions. While it is well appreciated that active RhoA promotes actomysosin contractility in epithelial tissues at the cell and tissue scales, little is known about how RhoA regulates junction dynamics on a subcellular scale or how junctional RhoA activity changes over short time scales. Here, we describe transient, localized accumulations of active RhoA (“Rho flares”) at junctions in the epithelium of gastrula‐stage Xenopus laevis embryos. Typically we observe more flares near cells undergoing cell shape changes (e.g., dividing cells), indicating that Rho flares may be the result of junction remodeling in response to cell shape changes. In order to investigate


the cause and consequence of Rho flares, we co‐imaged active Rho with fluorescently‐tagged junction proteins. Intriguingly, the tight junction proteins ZO‐1 and Occludin are locally decreased prior to the Rho flare and, along with Claudin‐6, are reinforced following the flare. Using a newly‐developed assay that allows us to sensitively detect local barrier breaches during live imaging, we observe that some Rho flares are preceded by barrier breaches, which are repaired following the flare. Additionally, Rho flares are accompanied by decreases in junction length and local increases in F‐actin, E‐cadherin, and α‐catenin that return to baseline levels soon after the flare. This indicates that RhoA effectors work in concert with the adherens junction to reinforce the tight junction. Taken together, our data suggest that localized activation of RhoA is needed to repair and remodel junctions to maintain the barrier function and integrity of developing epithelia.

E96 Self‐organization of skin organoid with hair primordia formation in vitro. M. Lei1,2, L. Schumacher3, Y. Lai4, W. Juan4, C. Yeh1, P. Wu1, T. Jiang1, R. Baker3, R. Widelitz1, L. Yang2, C. Chuong1,4; 1Department of Pathology, University of Southern California, Los Angeles, CA, 2College of Bioengineering, Chongqing University, Chongqing, China, 3Mathematical Institute, University of Oxford, Oxford, United Kingdom, 4Integrative Stem Cell Center, China Medical University, Taichung, Taiwan The transformation process from dissociated multipotent stem or progenitor cells into an organized tissue is a poorly understood and yet fundamental issue in regenerative medicine. Previously we developed a planar hair forming procedure that forms hairs robustly on the back of the nude mice. However, due to its in vivo nature, we do not know the cellular events during this morphogenetic process. To explore this process, we now developed an in vitro transwell 3‐dimensional mixed culture, in which dissociated newborn mouse epidermal and dermal cells can self‐organize to form planar hairy skin. With more than 500 time‐lapse movie segments, to our surprise, we found dissociated cells go through a complex morphogenetic detour to become a planar skin, from small aggregates – dermal / epidermal cysts– fused cysts –hair‐bearing planar skin. Collective cell behavior analyses showed a series of morphological phase transition. To analyze key molecular events between transitions of different phases, we use transcriptome analyses to profiling molecular expression at different time points. Non‐biased clustering of molecular profiling data categorizes them into different phases. In the early phase, adhesion molecules, and morphogens such as PDGF, insulin‐like growth factors are more involved. In the late phase, Wnts and extracellular matrix molecules are more involved. Functional perturbations show timely physical‐chemical coupling of these molecular events with phase transition are essential for tissue self‐organization. When transplanted to the backs of nude mice, these reconstituted cultured skin grafts robustly formed hairs. While cells from adult mice do not form hairs easily, by comparing transcriptome analyses and cellular events, we made progress toward making adult cells to form new hairs. The results demonstrate the self‐organizing process from dissociated progenitor cells to planar reconstituted skin with newly induced hair follicles. The in vitro model set up the platform to further analyze molecular process in detail, and can help us identify molecules that favor hair formation. Funding Source: This work is supported by NIH grants AR42177, AR60306, China Postdoctoral Science Foundation (2016M590866), Fundamental Research Funds for the Central Universities (106112015CDJRC231206), Special Funding for Postdoctoral Research Projects in Chongqing (Xm2015093), and a fellowship from the China Scholarship Council (2011605042).


E97 Direct visualization of multinucleated giant cell formation. J.J. Faust1,2, W. Christenson3,4, K. Doudrick5, R. Ros3,4, T.P. Ugarova1,2; 1Center for Metabolic and Vascular Biology, Mayo Clinic, Scottsdale, AZ, 2Molecular and Cellular Biology, Arizona State University, Tempe, AZ, 3Department of Physics, Arizona State University, Tempe, AZ, 4Center for Biological Physics, Arizona State University, Tempe, AZ, 5Department of Civil and Environmental Engineering, University of Notre Dame, Notre Dame, IN The formation of multinucleated giant cells (MGCs) is known to occur as a result of macrophage fusion, but surprisingly no published study to date has visualized fusion in context with living specimens. Consequently, the mechanism(s) and kinetics that govern MGC formation in vitro remain ill‐defined. Through the use of a variety of live imaging techniques and a novel micropatterned glass surface that imparts spatial control of macrophage fusion, this work provides the first time‐resolved views of MGC formation. MGC formation followed three distinct patterns: First, mononucleated macrophages fused to establish a founder population of multinucleated cells. Second, founders acquired additional mononucleated macrophages to generate MGCs. Third, MGCs fused with neighboring MGCs to establish syncytia. Founders initiated fusion rapidly 9.5 ± 0.5 hours after the application of IL‐4 and MGCs formed exclusively on micropatterns. Juxtaposed membranes appeared to coalesce within 180 ± 68 seconds and the entire fusion event required 51 ± 26 minutes as defined by first observable intercellular contact to full nuclear integration. Fusion was accompanied by contraction of the cell body and subsequently by rapid extension of lamellipodia. Taken together, this work indicates that an as‐of‐yet unidentified mechanism involving intercellular interaction between margins, but not “cellocytosis” accounts for the initiation of macrophage fusion leading to MGC formation in vitro.

E98 Human Kidney Organoids from Genome‐Modified Pluripotent Stem Cells Establish a High‐Penetrance Model of Polycystic Kidney Disease. N.M. Cruz1, B.S. Freedman1; 1Department of Medicine, Division of Nephrology, Kidney Research Institute, and Institute for Stem Cell and Regenerative Medicine, University of Washington, Seattle, WA Human pluripotent stem cells (hPSCs) can differentiate into nephron‐like kidney organoids containing diverse epithelial cell types, which can be used to model the cell biology of human disease. Polycystic kidney disease (PKD) is an extremely common Mendelian disorder in which kidney tubular epithelial cells give rise to large, fluid‐filled cysts that disrupt kidney function. PKD is caused genetically by loss‐of‐function mutations in PKD1 or PKD2, whose gene products form a receptor‐channel complex in primary cilia. Organoids derived from genome‐modified PKD1 ‐/‐ or PKD2 ‐/‐ hPSCs form cysts from kidney tubules, but the penetrance of this phenotype is low. Using a variety of PKD hPSCs, we investigated the potential of differentiation protocol, culture conditions, and genetic background to modulate PKD organoid cystogenesis. To model PKD, induced hPSCs from four ADPKD and ARPKD patients or genome‐modified PKD1 ‐/‐ and PKD2 ‐/‐ hPSCs were differentiated into kidney organoids using two different published protocols. Organoids were cultured under 2D or 3D conditions with or without forskolin, an agonist of adenylyl cyclase. PKD organoids and cysts were counted, measured, and analyzed for kidney tubule marker immunofluorescence and cilia, and by time‐lapse microscopy, compared to non‐PKD controls. Compared to induced hPSCs from different patients, genome‐modified hPSCs exhibited much more uniform differentiation into kidney organoids, enabling specific disease modeling experiments. Under


novel three‐dimensional growth conditions, PKD1 ‐/‐ and PKD2 ‐/‐ organoids budded off translucent cysts within the first month of differentiation, which further expanded to diameters approaching 1 cm over several months in culture. Penetrance of the cystogenesis phenotype was increased to ~ 80 % in PKD1 ‐/‐ and PKD2 ‐/‐ organoids, compared to < 10 % in isogenic control organoids. Cysts included proximal and distal tubular cells, were hyperproliferative, and increased upon adenylyl cyclase stimulation. Cystic epithelia elaborated primary cilia, and the PKD2 gene product, polycystin‐2, was lost from primary cilia in both PKD1 and PKD2 mutants. In conclusion, PKD1 ‐/‐ and PKD2 ‐/‐ kidney organoids establish a high‐penetrance system for human PKD cystogenesis in vitro. Loss of polycystin‐2 from primary cilia is a common defect in PKD hPSCs and organoids. Cyclic AMP signaling enhances human PKD cystogenesis, similar to mouse models. Our findings indicate that PKD cystogenesis from kidney tubules is a cell‐intrinsic process that can be modeled in a relatively simple cell culture system in vitro. The modularity of kidney organoid cultures makes them a powerful new pre‐clinical model in which to investigate PKD pathophysiology and evaluate candidate therapeutics.

E99 The role of division orientation in embryonic patterning. L.I. Rathbun1, J. Amack1, H. Hehnly1; 1Cell and Developmental Biology, State University of New York Upstate Medical University, Syracuse, NY During the cleavage period of zebrafish embryonic development (2‐64‐cells/embryo), early embryos undergo a number of synchronous planar cell divisions before switching to asynchronous oriented divisions. Planar cell divisions occur when the central spindle orients parallel to the yolk‐sac syncytium, and oriented divisions occur when the spindle orients perpendicular, initiating cellular remodeling necessary for embryonic patterning. We hypothesize that this process is regulated by astral microtubules (MTs), which connect the central spindle to the cell cortex. By utilizing live‐cell video microscopy to track and measure astral MTs (labeled with EMTB‐3xGFP), we noted that early cleavage period (2‐16‐cells) mitotic cells lose astral MT contacts with the cortex during metaphase to anaphase transition. This suggests that astral MTs do not contact the cell cortex during planar cell divisions, possibly because cells do not need to orient their spindles yet. To determine the developmental stage at which astral MTs are able to contact the cortex during metaphase to anaphase transition, we measured cell area and the ratio between cell diameter and spindle length. Strikingly, a 44 + 1% decrease in cell area was measured between each division from 4‐ to 64‐cell stage. In pre‐16 cell embryos, we found that although the spindle decreases in length, 6.3% from 4‐ to 8‐cell embryos and 16.9 % from 8‐ to 16‐cell embryos, it is not proportional to the simultaneous decrease in cell diameter. Thus, we conclude that the spindle occupies 14% more of the cell length during the blastula period (>128‐cells/embryo) compared to a cleavage period embryo (2‐64‐cells). With these findings, we present a testable model where the ratio of spindle length to cell length needs to pass a certain threshold before astral MTs can contact the cortex and contribute to mitotic spindle orientation. To further test this model we used two approaches to investigate astral MT‐cortex contact: 1) transgenic zebrafish live cell imaging and 2) a micro‐surgical method of fixed embryo mounting to allow for high resolution microscopy. Using these tools, we observed that myosin, an actin‐binding motor protein, and aPKC, a polarity protein, could be utilized as molecular markers that label the apical membrane and cell periphery in developing zebrafish. Our fluorescent myosin transgenic line has proven useful in not only characterizing MT cortical contacts, but also in early embryonic patterning where cells can subdivide into 4 main zones post 16‐cell stage. Utilizing these developmental and molecular tools we will investigate intercellular events that position the mitotic spindle and subsequently determine early cellular patterning during zebrafish embryonic development.


Microsymposium 12: Membrane Trafficking and Signaling

E100 Cellular, molecular and structural mechanisms of phosphorylation‐dependent inhibition of IRSp53 by 14‐3‐3. D.J. Kast1, R. Dominguez1; 1Physiology, University of Pennsylvania, Philadelphia, PA Filopodia are dynamic actin‐rich protrusions that play essential roles in numerous cellular processes such as cell migration, phagocytosis and axon guidance. The BAR‐domain protein IRSp53 couples Rho‐GTPase signaling to membrane and cytoskeleton dynamics, acting as a potent inducer of filopodia formation. We previously showed that IRSp53 exists in a closed autoinhibited state stabilized by intramolecular interactions, and that it is combinatorially activated by interaction with either Cdc42, via a unique CRIB‐PR domain, or downstream cytoskeletal effectors, via its SH3 domain. Activation is characterized by a conformational change into a more open state. Using live‐cell imaging, biochemical and structural approaches, we show here that 14‐3‐3 inhibits IRSp53 in a phosphorylation‐dependent manner, stabilizing the closed, inactive conformation. In live cells, 14‐3‐3 inhibits IRSp53‐dependent filopodia dynamics and the ability of cells to chemotax. Using a phospho‐IRSp53 FRET sensor purified from mammalian cells, we show that 14‐3‐3 binding impedes (or reverses) the activating conformational change induced by Cdc42 or the downstream cytoskeletal effector Eps8. Using phosphoproteomics, we identified four phosphorylation sites on IRSp53 between the CRIB‐PR and SH3 domains, which mediate the binding of 14‐3‐3. Individually, these sites binds 14‐3‐3 with micromolar affinity, but when combined into specific pairs, their affinities increase ~200‐fold. Crystal structures of 14‐3‐3 bound to IRSp53 phosphopeptides show that 14‐3‐3 binds asymmetrically to pairs of phospho‐sites. In 14‐3‐3‐expressing cells, the expression of non‐phosphorylatable mutants of IRSp53 restores wild‐type‐like filopodia dynamics and the ability of cells to chemotax. The data support a model whereby IRSp53 is combinatorially inhibited by intramolecular interactions and asymmetric binding of 14‐3‐3 and activated by the binding of Cdc42 and downstream effectors.

E101 The Role of Rab30 at the Interface between the Recycling Endosome & the trans‐Golgi Network. K.L. Zulkefli1, F.J. Houghton1, P. Gosavi1, I.S. Mahmoud1, P.A. Gleeson1; 1Biochemistry Molecular Biology, Bio21 Molecular Science Biotechnology Institute, University of Melbourne, Melbourne, Australia The recycling endosome compartment is a pleiomorphic subpopulation of endosomes, acting as a major sorting node for anterograde transport and for recycling of cargo to the plasma membrane. The trans‐Golgi network also serves as a sorting hub, and functions to regulate the final post‐translational modifications of nascent proteins, transport of cargo to the plasma membrane as well as receiving incoming cargo from retrograde trafficking pathways. There is membrane flow between the recycling endosomes and the trans‐Golgi network and these two compartments share similar properties. However, the interface between the recycling endosome and the trans‐Golgi network is not well defined. Here we show that the previously poorly characterised Rab30 may function at the interface between these compartments. We demonstrate that Rab30 is localised to both the recycling endosome and the trans‐Golgi network. In addition, we explored the role of Rab30 in the trafficking events


between these compartments and identified the putative interacting partners of Rab30. Overexpression of Rab30 resulted in an increase in the rate of retrograde trafficking of furin and TGN38 from endosomal compartments to the trans‐Golgi network. Conversely, Rab30 silencing resulted in the decrease in the rate of retrograde transport of TGN38 to the trans‐Golgi network and a modest retention of transferrin receptor in the recycling endosomes. The silencing of Rab30 also resulted in an altered distribution of the recycling endosomes and a dispersal of the Golgi complex. Silencing of Rab11 lead to the appearance of dynamic Rab30‐positive tubules emanating from the Golgi complex which were also found to be loaded with transferrin receptor. Interactive partners of Rab30 are being identified by immunoprecipitation of Rab30‐GFP expressed stably in HeLa cells and mass spectrometric analysis to date, include components found both in the recycling endosome and the Golgi complex. Taken together, these results indicate a role for Rab30 in regulating the membrane flux and trafficking events between the recycling endosome and the trans‐Golgi network.

E102 Pre‐allocated Subset of Clathrin‐coated Pits Controls Competition and Cooperation of Receptor Signaling. D. Kim1, Y. Kwon1, H. Ahn2, J. Noh1, M. Jeong3, S. Park1, K. Zhou1, N. Lee4, S. Ryu1; 1Life Sciences, Pohang University of Science and Technology, Pohang, Korea, South, 2School of Interdisciplinary Bioscience and Bioengineering, Pohang University of Science and Technology, Pohang, Korea, South, 3Integrative Biosciences Biotechnology, Pohang University of Science and Technology, Pohang, Korea, South, 4Physics, Pohang University of Science and Technology, Pohang, Korea, South Clathrin‐mediated endocytosis plays indispensable roles for not only receptor internalization but signaling. However, it is unknown how clathrin‐coated pits (CCPs) manage to perform their functions for entire receptors while only a small number of CCPs are expressed in a plasma membrane and one CCP can accommodate tens of receptors at most. Here, we developed a single‐molecule diffusion‐based molecular colocalization technique to resolve receptors residing in a single CCP on the plasma membrane of living cells. We introduced a new dimension of super‐resolution microscopy in addition to space and time, the diffusivity; if two molecules are colocalized, the diffusivity of the two must be the same. Using this method, we found that there exist a subset of CCPs pre‐allocated for epidermal growth factor receptor (EGFR) in its basal state, which is utilized for EGFR endocytosis when activated by EGF. The pre‐allocated subset for EGFR was shared with ErbB2 but not transferrin receptor and beta2‐adrenergic receptor, providing a critical insight on the mechanism of CCPs to control competition among receptors by limiting the capacity for one receptor at the early stage of cargo selection. Furthermore, the dimerization between EGFR and ErbB2 was greatly augmented inside a CCP compared to a plasma membrane, implying that a pre‐allocated subset of CCPs serves as a signaling platform to generate a strong cooperation between shared receptors by increasing the local concentration of the receptors in a single CCP.


E103 Specific exosome subtypes are generated in different endosomal compartments and selectively regulated by stress sensor mTORC1. S. Fan1, S.M. Perera1, B. Kroeger1, C. Alves1, I. Stefana1, J.F. Morris1, C. Wilson1, A.L. Harris2, D. Goberahan1; 1Department of Physiology, Anatomy and Genetics, Oxford University, Oxford, United Kingdom, 2The Weatherall Institute of Molecular Medicine, Oxford University, Oxford, United Kingdom Cells can adapt to nutrient starvation and growth‐inhibitory drugs by communicating with their neighbours, through secretion of both soluble factors and exosomes. Exosomes are secreted nanovesicles, which are thought to originate from late endosomal multivesicular bodies. They carry complex bioactive cargos to target cells. In cancer, these multi‐molecular signals can reprogramme recipient cells to support tumour cell growth and migration, and also to prime new pre‐metastatic niches. The overarching aim of this study was to determine how exosome signalling is modulated by cellular stress signals. We have focused on the role of the amino acid‐sensitive mechanistic Target of Rapamycin Complex 1 (mTORC1) in this process. Our approach has involved an in vivo model of exosome biogenesis that we have recently developed. It takes advantage of the exceptionally large prostate‐like secondary cells of the Drosophila male accessory glands. This work was complemented by studies using a range of human cancer cell lines to analyse exosome secretion and function, after genetic, pharmacological and starvation‐induced modulation of mTORC1 activity. We have used a wide range of methods to isolate exosomes from cultured cells and to characterise their physical, biochemical and functional properties. Using in vivo super‐resolution (3D‐SIM) imaging, we show for the first time that exosomes are not only formed in Rab7‐positive late endosomes, but also in Rab11‐positive recycling endosomes. These exosome populations carry different cargos and their Rab signature indicates their endosomal birth place. Indeed Rab11‐positive compartments seem to be the major sites of exosome biogenesis in fly secondary cells. Using different cancer cell lines, including HCT116 colorectal cancer cells, we show that blocking a rapamycin‐resistant form of mTORC1, controlled by extracellular glutamine levels, leads to a change in the cargo of secreted exosomes. Vesicles carrying the well‐established exosome marker, CD63, switch to Rab11‐positive exosomes, which are secreted through a different trafficking pathway and are also enriched with the lipid raft‐associated protein Caveolin‐1. These stress‐induced exosomes induce changes in recipient cells that may alter the cellular microenvironment and promote cancer cell survival. We conclude that there are at least two evolutionarily conserved, spatially separated exosome biogenesis routes involving different endosomal compartments and contributing to exosome heterogeneity. Furthermore, in human cancer cells, these cellular passages are differentially regulated by nutrient stress and mTORC1‐inhibitory drugs, so that cells switch their exosome signals in response to such treatments, allowing them to adapt the tumour microenvironment.


E104 Growth factor signaling regulation of a Rab‐effector switch required for ciliogenesis initiation. V. Walia1, M. Vetter2, C. Insinna1, Q. Lu1, D. Ritt1, S. Specht1, J. Stauffer1, D. Morrison1, E. Lorentzen2, C.J. Westlake1; 1Center for Cancer Research, National Cancer Institute, Frederick, MD, 2Max Plank Institute of Biochemistry, Martinsried, Germany Defects in primary cilium assembly and signaling are associated with a growing list of genetic diseases. Cell cycle arrest via deprivation of serum growth factors stimulates primary cilia assembly in cultured cells. However, the signaling pathways regulating ciliogenesis are not known. We previously showed that Rabin8, a Rab8 GEF, is recruited by Rab11 to centrosomal vesicles within minutes of serum removal to activate Rab8, which is required for cilium elongation. Importantly, Rabin8 pre‐ciliary trafficking on Rab11 endosomes marks the earliest ciliogenesis initiation step reported thus far in the literature. Here we describe the discovery of a growth factor signaling pathway regulating a Rab11‐effector switch that functions in ciliogenesis initiation. A screen of serum growth factors identified lysophosphatidic acid (LPA) and its G‐protein coupled receptor, LPAR1, as the upstream regulators of Rabin8 transport and ciliation. LPA‐LPAR1 signals through PI3K‐Akt to block Rab11‐Rabin8 interaction, Rabin8 pre‐ciliary trafficking and ciliogenesis. We show that Rabin8 is a substrate of Akt, however its phosphorylation by Akt does not directly affect Rab11 interaction. We next hypothesized that Rabin8 interaction with Rab11 is regulated indirectly by Akt via phosphorylation of another Rab11 effector. Using biochemical approaches, we identified candidate Rab11‐effector proteins displaying serum‐sensitive binding to Rab11. RNAi depletion of candidate proteins showed that WDR44/Rabphillin‐11 promotes serum‐independent Rabin8 centrosomal trafficking and initiates early ciliogenesis processes at the mother centriole. An Akt phosphorylation site was mapped to the Rab11 binding domain in WDR44 important for Rab11 interaction. A phospho‐mimetic WDR44 mutant interacts more strongly with Rab11 than the phospho‐inactive protein, and blocks ciliogenesis in RPE‐1 cells and zebrafish embryos. FIP3 is another Rab11 effector that has previously shown to function in ciliary receptor trafficking in complex with Rab11 and Rabin8. Our data suggest that FIP3 can also regulate RPE‐1 ciliogenesis and is required for Rabin8 pre‐ciliary trafficking. Moreover, we show that WDR44 and FIP3 competes for the same binding site on Rab11. Together our findings demonstrate that LPA‐LPAR1 signaling promotes Akt phosphorylation of WDR44 which stabilizes its interaction with Rab11 and prevents ciliogenesis. Serum starvation mediated deactivation of Akt disrupts this effector‐based interaction and promotes a switch to a Rab11‐Rabin8 complex that delivers membranes to the mother centriole needed to initiate early ciliogenesis processes.

E105 Organization of the Leukotriene Synthetic Complex in Neutrophils as an Indicator of Disease State. A.B. Schmider1, M. Godin1, H.D. Elliott2, R.J. Soberman1; 1Nephrology, Massachusetts General Hospital, Charlestown, MA, 2Cell Biology, Harvard Medical School, Boston, MA The recruitment and activation of neutrophils is a prominent component of tissue injury in inflammatory diseases such as Anti‐Neutrophil Cytoplasmic Autoantibody (ANCA)‐Associated Vasculitis (AAV). The chemotactic lipid, leukotriene (LT)B4 plays a role in the initial recruitment of neutrophils from the vasculature and controls the participation of these cells in the key pathological process of swarming. LTs are fatty acid mediators generated from arachidonic acid metabolism that initiate and amplify innate


and adaptive immune responses. The core members of this structure are the enzyme, 5‐lipoxygenase (5‐LO) and the scaffold protein, 5‐lipoxygenase‐activating protein (FLAP). We previously showed in mouse neutrophils that priming with granulocyte macrophage colony‐stimulating factor and subsequent treatment with complement component 5a, causes 5‐LO and FLAP to organize on the nuclear envelope in a multistep fashion resulting in LTB4 synthesis. We paired Stochastic Optical Reconstruction Microscopy (STORM) with cluster analysis and Fluorescence Lifetime Imaging Microscopy (FLIM) to show the reorganization of 5‐LO and FLAP into supramolecular clusters on the nuclear envelope of human neutrophils in response to priming and activation from neutrophils of patients with AAV and healthy controls. Primary antibodies against 5‐LO and FLAP were used for all microscopy experiments. STORM data was obtained using secondary antibodies labeled with Cy3b and AlexaFluor 647. STORM data is generated as a map of intensity signals and presented as pseudo colored images based on defined parameters. To test the hypothesis that LT synthetic complexes were organized into larger structures, we developed unbiased automated cluster analysis algorithms to analyze primary STORM data. Using AlexaFluor 488 and 594 secondary antibodies, FLIM determined the closeness between 5‐LO and FLAP. Our data suggest that in blood peripheral neutrophils from patients with AAV, 5‐LO and FLAP formed clusters with a greater number of molecules, larger areas and higher densities compared to neutrophils from healthy patients. Overall, we demonstrate that assembly of the LT complex into supramolecular structures is a marker for disease state.

E106 Otoferlin is a sensory hair cell specific calcium sensor for membrane fusion and neurotransmitter release during the encoding of sound. C.P. Johnson1, P. Chatterjee1, R. Tanguay1; 1Biochemistry, Oregon State University, Corvallis, OR Congenital hearing loss is a common disorder, with approximately 1 out of every 600 children suffering from profound deafness. A central player in hearing is the transmembrane protein otoferlin. More than 60 pathogenic mutations in otoferlin are known, and up to 8% of all forms of prelingual autosomal recessive hearing loss are due to otoferlin mutations. Currently the only known bodily function for otoferlin is in mediating neurotransmitter release from auditory and vestibular hair cells, making it a sensory hair cell specific protein. While otoferlin is essential for neurotransmitter release, the functions and mechanisms by which otoferlin regulates exocytosis and endocytosis remain unknown. Further, the large size of otoferlin has prevented rescue experiments and prohibited viral‐based gene transfer, representing a major hurdle toward the use of gene therapy as a treatment. To address these issues we have conducted in vitro recombinant protein reconstitution assays, as well as high resolution fluorescent microscopy imaging, RNAseq and rescue experiments on otoferlin depleted zebrafish. In zebrafish, expression of otoferlin occurs early in development, and depletion of otoferlin results in hearing and balance defects. Analysis of fluorescent confocal imaging indicates otoferlin localizes to the apical and basolateral areas of sensory hair cells, with loss of otoferlin resulting in abnormal ribeye and VGlut3 localization. Loss of otoferlin also resulted in abnormal endocytosis of fluorescent dye into hair cells and defective synaptic vesicle cycling as determined from path clamp measurements. Further, analysis of RNAseq data indicates that loss of otoferlin is associated with changes in the expression patterns of several hair cell specific proteins. Measurements on recombinant protein indicate that otoferlin binds both calcium, phosphatidylinositol bisphosphate (PI(4,5)P2) lipids, and SNARE proteins. In addition, results of in vitro reconstituted membrane fusion assays demonstrate that otoferlin can act as a calcium sensitive regulator of SNARE mediated membrane fusion, and provide a molecular level explanation for the loss of otoferlin phenotype in our zebrafish studies. We have also conducted the first otoferlin


rescue experiments, and have found that mouse otoferlin restored both hearing and balance in zebrafish depleted of endogenous otoferlin. Remarkably, truncated forms of otoferlin retaining the C‐terminal region also rescued the otoferlin depletion phenotype, suggesting that truncated forms of otoferlin can restore hearing and balance. In summary, our results suggest otoferlin is a hair cell specific protein that confers calcium sensitivity to SNARE mediated exocytosis, and that truncated forms of otoferlin may be used in gene therapies.

E107 The Lowe Syndrome phosphoinositide phosphatase dOCRL restricts innate immune activation by regulating endosomal traffic. S.J. Del Signore1, S.A. Biber1, K.S. Lehmann1, T.L. Eskin1, A.A. Rodal1; 1 Rosenstiel Basic Medical Sciences Research Center, Department of Biology, Brandeis University, Waltham, MA Lowe Syndrome arises from mutations in OCRL, a phosphatidylinositol‐5‐phosphatase with roles in endocytic and endolysosomal trafficking, cytokinesis, and ciliogenesis (Mehta, 2014). However, it remains unclear how loss of OCRL results in tissue and organ‐level pathologies in patients. Here, we found that mutation of the Drosophila melanogaster homolog of OCRL causes innate immune activation, and that this phenotype is cell‐autonomous to hemocytes. docrl mutant hemocytes exhibit broad structural and functional trafficking defects: clathrin, Rab5, and Vps35 compartments are fragmented and redistributed to a perinuclear aggregate, while Rab7 and Rab11 compartments are significantly expanded. These changes correlate with decreased motility of Rab5 early endosomes, defective scavenger receptor endocytosis, and defective autophagosome‐lysosome traffic. To pinpoint which among these diverse defects drives immune activation, we genetically manipulated individual trafficking steps. We found that hemocyte‐specific inhibition of Rab5 or Rab11, or activation of Rab7 in otherwise wild‐type larvae phenocopied the immune activation in docrl mutants. By contrast, blocking endocytosis by overexpression of temperature sensitive or dominant negative dynamin did not cause immune activation. Further, we found that the converse endosomal manipulations in docrl mutant hemocytes‐ activation of Rab5 or Rab11, or inhibition of Rab7, each suppressed docrl‐induced hemocyte defects. To determine how these trafficking defects promote immune activation, we examined signaling pathways implicated in immunity, and found specifically that Toll signaling was aberrantly activated in hemocytes. docrl induced immune activation correlates with mis‐sorting of the Toll ligand Spåtzle (Spz) to a Rab11 compartment, excessive secretion of Spz into the hemolymph, and aberrant Toll pathway activation. Together, these data suggest that innate immune activation specifically arises due to a shift from Rab11‐dependent recycling route to a Rab7‐dependent late endosomal route. Thus, dOCRL regulation of endosomal traffic is critical for maintaining hemocytes in a poised, but quiescent state. Our results suggest mechanisms by which endosomal misregulation of signaling pathways may contribute to the pleiotropic symptoms in Lowe syndrome.


E108 Clathrin light chain isoform specific effects alter EGFR trafficking, signaling and increase metastatic potential. P. Chen1, N. Bendris1, Y. Hsiao2, C.R. Reis1, M. Mettlen1, H. Chen3, S. Yu2, S.L. Schmid1; 1Cell biology, UTSW medical center, Dallas, TX, 2Department of Clinical Laboratory Sciences and Medical Biotechnology, National Taiwan University College of Medicine, Taipei, Taiwan, 3Institute of Statistical Science, Academia Sinica, Taipei, Taiwan Signaling receptors are internalized and regulated by clathrin‐mediated endocytosis (CME). Vertebrates express two clathrin light chain isoforms, CLCa and CLCb: integral components of the endocytic machinery whose differential functions remain unknown. We reported that CLCb is highly expressed in non‐small cell lung cancer (NSCLC) cells and associated with poor patient prognosis. Engineered single CLCb‐expressing NSCLC cells, as well as ‘switched’ cells that predominantly express CLCb, exhibit increased rates of CME and altered clathrin‐coated pit dynamics due to increased expression of dynamin 1 (Dyn1). Up‐regulation of CLCb enhances EGF‐dependent Akt and GSK3β phosphorylation to activate Dyn1, modulate EGFR trafficking, and enhance cancer cell migration in vitro, and metastasis in vivo. We define a molecular link between signaling, CME and the increased cancer metastasis.

Kaluza Minisymposium

A4 Phase Separation of Multivalent Adaptor Proteins. S. Banjade1,2, P. Li1, H.C. Cheng1, Q. Wu1, B. Chen1, R.V. Pappu3, M.K. Rosen1,4; 1Department of Biophysics, UT Southwestern Medical Center, Dallas, TX, 2Current: Weill Institute for Cell and Molecular Biology, Cornell University, Ithaca, NY, 3Department of Biomedical Engineering, Washington University in St. Louis, St. Louis, MO, 4HHMI, UT Southwestern Medical Center, Dallas, TX Eukaryotic cells organize many biochemical activities in various membrane‐less organelles and membrane‐bound clusters. The physical mechanisms behind the formation of such macroscopic assemblies remain poorly defined. Using the adhesion receptor Nephrin and its cytoplasmic partners Nck and N‐WASP as a model system, we discovered that multivalent interactions between these proteins give rise to phase‐separation into 3D droplets in solution and form micron‐sized clusters on the surface of model membranes. Nephrin has a cytoplasmic tail that, upon phosphorylation at three sites, recruits Nck to the membrane. Nck has three SH3 domains which recruit N‐WASP to the membrane through N‐WASP's proline‐rich motifs (PRMs). The dynamics and thermodynamics of these phase‐separated clusters depend upon the valency of the phosphotyrosine motifs and the SH3 domains, suggesting that phase‐separation is governed by polymerization of the proteins through multivalent interactions. Dependence of phase‐separation on the number of phosphotyrosine motifs suggests how kinases may control such phenomena in biology. In addition to the interactions of the ordered domains, a disordered region in Nck enhances the ability of the complexes to phase separate by interacting with the SH3 domains, demonstrating an important role of intrinsically disordered regions of adaptor proteins during this process. The phase‐separation also leads to a sharp change in activity of N‐WASP toward the Arp2/3 complex, which nucleates actin polymerization, signifying how such phenomenon could lead to switch‐like activation in signaling systems. The widespread presence of multivalent proteins in cellular bodies suggest that our findings with this model system may have broad implications in the understanding of the functions of such assemblies and the mechanisms of their formation.


A5 Extracellular vesicles from Trypanosoma brucei mediate virulence factor transfer and cause host anemia. A. Szempruch1, S. Sykes2, R. Kieft2, L. Dennison2, A. Becker2, A. Gartrell2, W. Martin3, E. Nakayasu4, I. Almeida5, J. Harrington2, S. Hajduk2; 1Division of Biology and Biological Engineering, California Institute of Technology, Pasadena, CA, 2Biochemistry and Molecular Biology, University of Georgia, Athens, GA, 3Animal Health Research Center, University of Georgia, Athens, GA, 4Biological Sciences, Pacific Northwest National Laboratory, Richland, WA, 5Biological Sciences, University of Texas El Paso, El Paso, TX Trypanosoma brucei cycles between an insect vector and a mammalian host, causing human African sleeping sickness and nagana in cattle. Infection of both humans and animals results in anemia through non‐hemolytic clearance of erythrocytes. The human pathogenic T. b. rhodesiense is able to inactivate the primate‐specific, innate immune particle trypanosome lytic factor (TLF) through expression of the serum resistance‐associated protein (SRA), resulting in human infectivity. Here we show that bloodstream trypanosomes produce membrane nanotubes which originate from the flagellar membrane and give rise to extracellular vesicles (EV). Proteomic analysis of purified EV showed an enrichment in flagellar proteins, and analysis of EV from SRA expressing cells showed that SRA is an EV component. Image‐stream analysis showed EV rapidly interacted at the posterior end of T. brucei, resulting in lipid exchange through membrane fusion. Fluorescence microscopy showed that EV proteins localized to the flagellar pocket and entered the endocytic pathway. Purified EV were capable of transferring the SRA protein to T. b. brucei and transfer resulted in resistance to TLF. In co‐culture transwell experiments, T.b. rhodesiense transferred TLF resistance without direct cell contact. We also observed that EV were able to fuse with mammalian erythrocytes, depositing parasite lipid and VSG protein, which altered the physicochemical properties of the erythrocyte plasma membrane. Erythrocytes treated ex vivo with EV were rapidly cleared when introduced to naïve mice and direct injection of EV caused significant loss of the total circulating erythrocyte population. This demonstrates that T. brucei EV are capable of transferring functional proteins and may serve a role as a cellular communication molecule, while affecting host cells.

A6 Angiopellosis as an Alternative Mechanism of Cell Extravasation. T. Allen1; 1Molecular Biomedical Sciences, NC State University, Raleigh, NC Stem cell therapy remains a promising strategy in the field regenerative medicine. Intravenous injection of stem cells has been proven as safe and effective methods for cell delivery to damaged/inflamed areas. However, for stem cells in circulation to exert their therapeutic effect, they must cross the blood vessel wall and enter the surrounding damaged tissue. The mechanism of this phenomenon has yet to be characterized. Using intra‐vital microscopy and a transgenic zebrafish embryo model, we discovered a previously unreported novel mechanism of stem cell extravasation, which we have termed Angio‐pello‐sis, which differs distinctly from the prevailing diapedesis method of extravasation. The transgenic tg(fli1a:egfp) zebrafish strain allows for unique visualization of blood vessels during the development, as they produce green fluorescent protein (gfp) exclusively in the endothelial cells of blood vessels. Multiple cell types were used, including leukocytes and cardiac, and mesenchymal stem cells. These cells were fluorescently labelled and injected intravenously into the blood stream of the embryos at 48 hours post fertilization and then imaged to capture the mechanism of extravasation.


The stem cells, once injected, underwent a distinct method of extravasation that was markedly different from other cells found in the blood stream, such as leukocytes. In this unreported mechanism, the vascular wall undergoes an extensive remodeling to allow the cell to exit the lumen, while the cell itself remains distinctively passive in activity. We term this new cell extravasation process as Angio‐pello‐sis (Angio = related to blood vessels; pello = expel). Both individual cells as well as multicellular clusters were able to utilize this process to extravasate, pointing to a potential means for circulating tumor clusters to extravasate. Inert polymer microspheres that were coated with the membrane of cardiac stem cells also extravasated through angiopellosis, while uncoated microspheres were unable to exit. This suggests a cell membrane specific signal, prompting the extravasation event, which only occurs in specific situations. However, when membrane molecules essential for diapedesis were inhibited, stem cells maintained there ability to extravasate trhough angiopellosis, while leukocytes lost all ability to undergo diapesesis. This suggests that while membrane recognition is required for both extravasation methods, the molecules between angiopellosis and diapedesis are distinct from each other. Angiopellosis represents an alternative mechanism of cell extravasation in contrast to the extravasation method of diapedesis. A better understanding of this mechanism will lead to improved methods for stem cell therapies, and increase cell extravasation knowledge

A7 The developmental polarity and morphogenesis of a single cell. D. Bonazzi1, A. Haupt2, J. Julien3, H. Tanimoto2, R. Seddiki2, M. Romao4, D. Delacour2, D. Salort5, M. Piel4, A. Boudaoud6, N. Minc2; 1Pathogenesis of Vascular Infections, Institut Pasteur, Paris, France, 2Institut Jacques Monod, Paris, France, 3Max Planck Institute for Dynamics and Self‐Organization, Göttingen, Germany, 4Institut Curie, Paris, France, 5UMR 7238 CNRS, Université Pierre et Marie Curie, Paris, France, 6ENS CNRS, Laboratoire Joliot‐Curie, Lyon, France Cell shape influences many basic cellular functions, such as growth, migration or division, and defects in cell morphogenesis can give rise to many diseases. But how does a cell develop and maintain its shape? The rod‐shape fission yeast S. pombe is an established model to dissect mechanisms regulating polarized growth and morphogenesis. In these cells, spatial protein landmarks at the plasma membrane direct actin assembly, secretory apparatus and cell wall synthesis for polarized growth at cell tips. In here, using quantitative microscopy and modeling, we investigate how these cells become rod in the first place, by monitoring how round germinating spores break symmetry to become rod‐shaped vegetative cells. Spores exhibit an initial near‐isotropic growth phase during which a single cortical polarity domain wanders around the surface. Very interestingly, we found that these oscillating Cdc42‐GTP domains scale with local cell‐surface radii of curvature over an 8‐fold range. This scaling depends on formin‐nucleated cortical actin cables and the fusion of secretory vesicles transported along these cables. These results suggest that actin networks may allow local curvature sensing, uncovering a morphogenetic principle for how polarity domains define their size according to cell geometry. Later on during germination, spores break symmetry to establish a single polarity axis, in a process termed outgrowth. Strikingly, we found that this occurs at a precise fold‐change in spore volume, concomitantly with the stabilization of polarity domains and a singular mechanical rupture in the outer spore wall, which encases the spore. This rupture occurs when the mechanical stress on the spore wall exceeds a threshold, and releases the constraints of this wall on cell growth, and stabilizes polarity. In conclusion, these dynamic studies shed light on a ubiquitous developmental process in unicellular organisms, and bring fundamental knowledge on the complex feedbacks between geometry, polarity and mechanics for cell morphogenesis.


BIBLIOGRAPHY: Bonazzi, D., et al., Actin‐Based Transport Adapts Polarity Domain Size to Local Cellular Curvature, Curr Biol, 2015, 25(20): p. 2677‐83. Bonazzi, D., et al., Symmetry breaking in spore germination relies on an interplay between polar cap stability and spore wall mechanics, Dev Cell, 2014, 28(5): p. 543‐46.

A8 Distinct Circuits for the Formation and Retrieval of an Imprinted Olfactory Memory. X. Jin1,2, N. Pokala1,3, C.I. Bargmann1; 1The Rockefeller University/HHMI, New York, NY, 2Society of Fellows, Harvard University, Cambridhe, MA, 3New York Institute of Technology, New York, NY Memories formed early in life are particularly stable and influential, representing privileged experiences that shape enduring behaviors. To understand the sites and molecules that enable this privileged learning, we developed an aversive olfactory imprinting paradigm in the nematode C. elegans. Exposing newly‐hatched larvae to pathogenic bacteria, e.g. Pseudomonas aeruginosa PA14, can generate an aversive memory of bacterial odors which sustained into adulthood (4 days). By contrast, training of adults results in a medium‐term memory that lasts for less than a day (Zhang et al, 2005). This long‐lasting imprinted aversion is specific to the experienced pathogen and has a critical period in the first larval stage. Through chemical‐genetic silencing of candidate neurons, we identified a circuit essential for memory formation but not for memory retrieval (interneurons AIB and RIM), and a complementary circuit essential for memory retrieval but not for memory formation (interneurons AIY and RIA). The RIM memory formation neurons synthesize the neuromodulator tyramine, which is required in the L1 stage for learning. This learning signal is transmitted to the AIY memory retrieval neurons by the tyramine receptor SER‐2, which is required for imprinted aversion but not for adult learned aversion. Tyramine modulation bridges the two subcircuits by linking tyramine production during learning with memory retrieval days later. Functional calcium imaging indicates that the early imprinting experience modifies neuronal activity and output of the memory circuit. Among several neurons examined, changes in RIA best express the context and specificity of the imprinted memory. Combining classical neuroethology, molecular genetics, and functional imaging, we have mapped distinct groups of neurons required for the formation and retrieval of an imprinted memory, defined the neuromodulation that enables this privileged learning (tyramine and SER‐ 2), and identified neuronal activity changes associated with memory specificity. These findings provide insight into neuronal substrates of different forms of memory, and lay a foundation for further understanding of early learning. Reference: Zhang, Y., Lu, H., and Bargmann, C.I. (2005). Pathogenic bacteria induce aversive olfactory learning in Caenorhabditis elegans. Nature 438, 179‐184.


A9 Adaptive evolution to the deletion of essential genes. G. Liu1,2, J. Yong1, M. Yurieva3, S. Kandhadayar Gopalan3, L.Z. Jaron1, J.S. Lim1, M. Poidinger3, G.D. Wright1, F. Zolezzi3, H. Choi4, N. Pavelka3, G.I. Rancati1; 1Institute of medical biology, Agency for Science, Technology and Research, Singapore, Singapore, 2School of biological science, Nanyang Technological University, Singapore, Singapore, 3Singapore Immunology Network, Agency for Science, Technology and Research, Singapore, Singapore, 4Saw Swee Hock School of Public Health, National University of Singapore, Singapore, Singapore Essential and non‐essential genes are classified based on the viability of the corresponding deletion mutant cells. However previous studies in budding yeast challenged this view by showing that mutants deleted of essential or non‐essential genes are capable of accumulating compensatory mutations and resume viability, suggesting that essentiality is not an intrinsic gene property but depends on the cellular capacity to evolve. To identify essential genes the lethality of which could be bypassed by short‐term evolutionary processes, a four‐layered screen was performed on all previously dubbed essential genes (1106). We identified 88 of such genes and renamed them as evolvable genes. While DNA content analysis of mutants deleted of evolvable genes showed accumulation of large‐scale genome changes, short‐term evolutionary experiments indicated that majority of mutants improved growth rate upon passaging, suggesting that by accumulating compensatory mutations cells evolved to the gene deletion. Network feature and conservation analyses showed that evolvable genes tend to have intermediate properties in respect to non‐essential and the other essential genes, suggesting gene essentiality form a gradient. Moreover, evolvable genes were enriched in protein complexes involved in intracellular trafficking of proteins. Interestingly independently generated mutant strains carrying the same gene deletion or deletion of genes involved in the same protein complex carried specific sets of aneuploidies, suggesting that aneuploidy could function as an adaptive mutations and that cells respond to inactivation of different component of the same functional submodule in similar ways. Accordingly, we showed that the presence of either specific aneuploidies or a specific gene on the aneuploid chromosome was required and sufficient to bypass the lethality of genes belonging to the same functional module. Dissection of the molecular mechanism at the basis of the adaptation showed that cells bypassed the crippled function not by fixing the broken machinery but by tinkering with pre‐existing components and using them for novel tasks. Taken together our results suggest that gene essentiality forms a continuum of characteristic at the gene and network‐specific properties. Moreover, cellular systems respond to the depletion of different components of a given functional complex using similar evolutionary strategies that are brought about by whole‐chromosome aneuploidy.

A10 Single‐chromosome aneuploidy commonly functions as a tumor suppressor. J. Sheltzer1; 1Cancer Center, Cold Spring Harbor Laboratory, Cold Spring Harbor, NY Chromosomal instability and whole‐chromosome aneuploidy are hallmarks of human malignancies. The prevalence of chromosome segregation errors in cancer – first noted more than 100 years ago – has led to the widespread belief that aneuploidy plays a crucial role in tumor development. In this work, we describe the first systematic test of this hypothesis. We transduced congenic euploid and trisomic fibroblasts with 14 different oncogenes or oncogene combinations, thereby creating genetically‐matched cancer cell lines that differ only in karyotype. Surprisingly, nearly all aneuploid cell lines divided slowly in vitro, formed few colonies in soft agar, and grew poorly as xenografts, relative to


matched euploid lines. Similar results were obtained when comparing a near‐diploid human colorectal cancer cell line with derivatives of that line that harbored extra chromosomes, and when comparing chromosomally‐stable and chromosomally‐unstable primary cell lines. Only a few aneuploid lines grew at close to wild‐type levels, and no aneuploid line exhibited greater tumorigenic capabilities than its euploid counterpart. These results demonstrate that rather than promoting tumorigenesis, aneuploidy can very often function as a tumor suppressor. Moreover, our results suggest one potential way that cancers can overcome the tumor suppressive effects of aneuploidy: upon prolonged culture in vitro or in vivo, cell lines with simple aneuploidies developed recurrent chromosomal aberrations that were absent from their euploid counterparts and that were associated with enhanced growth. Thus, the genome‐destabilizing effects of single‐chromosome aneuploidy may facilitate the evolution of balanced, high‐complexity karyotypes that are frequently found in advanced malignancies.

A11 TRPA1 channels regulate cochlear amplification through active shape changes of supporting cells in the inner ear. A.C. Velez‐Ortega1, R. Stepanyan1, S.E. Edelmann1, C. Park2, K.Y. Kwan3, G.P. Sinha1, D.P. Corey3, G.I. Frolenkov1; 1Dept. of Physiology, University of Kentucky, Lexington, KY, 2Dept. of Head Neck Surgery, University of California, Los Angeles, Los Angeles, CA, 3Dept. of Neurobiology, Harvard Medical School, Boston, MA TRPA1 channels are sensors for noxious stimuli in a subset of nociceptive neurons. They are also expressed in the mammalian inner ear. Given that Trpa1–/– mice exhibit normal hearing, balance, and sensory mechanotransduction, the function of TRPA1 channels in the inner ear remains unknown. We assessed TRPA1 expression in the organ of Corti using antibodies against the human PLAP reporter expressed in Trpa1–/– mice, and found broad PLAP labeling in the sensory hair cells as well as in several types of non‐sensory supporting cells, including the Hensen’s cells (HeCs), Deiters’ cells (DCs) and pillar cells (PCs). Next, to assess the presence of functional TRPA1 channels, we evaluated (i) the permeation of the FM1‐43 dye through TRPA1 channels in the presence of an agonist, (ii) patch clamp recordings of TRPA1‐mediated inward currents and (iii) intracellular Ca2+ responses after the local application of TRPA1 agonists. Our results indicated the presence of functional TRPA1 channels in several types of supporting cells. Yet HeCs exhibited the most robust and long‐lasting responses after stimulation with 4‐HNE, an endogenous TRPA1 agonist produced by lipid peroxidation. Thus, HeCs seem to be the major sensors of oxidative damage in the cochlea. TRPA1‐initiated Ca2+ responses often propagated from HeCs to other types of supporting cells and were accompanied by prominent shape changes in DCs and PCs. Using flash photolysis of caged Ca2+ compounds, we confirmed that PCs possess Ca2+‐dependent contractile machinery. Noise exposure is known to increase the oxidative stress in the cochlea and to cause the generation of 4‐HNE for several days. Therefore, we hypothesized that 4‐HNE generation after acoustic overstimulation would lead to TRPA1‐initiated changes of the supporting cell shape that would alter the geometry of the organ of Corti and modify cochlear amplification. To test this, we exposed young adult mice to mild noise and evaluated the recovery of hearing thresholds and cochlear amplification over time. Consistent with our hypothesis, we found a longer‐lasting inhibition of cochlear amplification in wild type mice than in Trpa1–/– littermates. Two weeks after the noise exposure, we observed a relative reduction in the auditory nerve responses compared to brain‐stem responses in Trpa1–/– mice, which represents a hallmark of noise‐induced cochlear synaptopathy. Our results indicate that the non‐sensory supporting cells of the hearing organ detect tissue damage via the activation of TRPA1 channels and subsequently modulate cochlear amplification through active cell


shape changes. This novel mechanism of cochlear regulation protects the organ of Corti after acoustic trauma or other types of cochlear injuries. Supported by NIDCD/NIH (R01_DC009434)

A12 Toward quantitative understanding of kinetochore‐microtubule interactions based on the molecular properties of individual components. A.V. Zaytsev1; 1Department of Physiology, Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA Accurate chromosome segregation crucially depends on the dynamic attachments between chromosomal kinetochores and spindle microtubules. Recent years brought tremendous progress in identifying molecular components of the kinetochores, but quantitative studies of how interactions between individual proteins and microtubules enable accurate chromosome segregation are lagging behind. Our work focuses on the Ndc80 complex, a highly conserved kinetochore protein that links kinetochores to spindle microtubules. With single molecule fluorescence microscopy we have shown that multisite phosphorylation at the N‐terminus of the Hec1 subunit of Ndc80 elicits a linear decrease in the NDC80 binding affinity to microtubules. This unusual behavior can be explained by conformational plasticity of the Hec1 terminus, which enables it to function as a phosphorylation‐controlled “rheostat” to control the NDC80 microtubule‐binding affinity. Experiments in cells expressing NDC80 complexes with increasing number of phosphomimetic mutations also show that the number of kinetochore‐bound microtubules decreases gradually. Strikingly, when we used our molecular results to model the changes in kinetochore‐microtubule affinity in response to phosphorylation, modeling results depended strongly on the overall organization of the kinetochore‐microtubule binding interface. With a traditional view of this interface as containing repetitive microtubule‐binding sites, the model predicted a switch‐like response that did not match the gradual dependency in cells. When the interface was assumed to contain a randomly distributed molecular “lawn” of the Ndc80 complexes, an excellent match to our in vivo studies has been obtained. In this conceptually different view of the kinetochore interface the individual NDC80 protein complexes are not restricted to binding only to the microtubule that belongs to this attachment site. Instead, the Ndc80 molecules can alternate between binding several nearby microtubules, forcing microtubules to compete for these molecular links. Thus, the important implication of our study is that the unconstrained interactions between Ndc80 complexes and microtubules are essential for the kinetochores’ ability to monitor and respond accurately to the number of microtubule attachments during mitosis.

ORAL PRESENTATIONS‐ TUESDAY

Oral Presentations‐Tuesday, December 6 Symposium 5: Cellular Communities

S11 Quorum Sensing and its Control. B. Bassler1; 1Molecular Biology, Princeton and HHMI, Princeton, NJ Bacteria communicate with one another via the production and detection of secreted signal molecules called autoinducers. This cell‐to‐cell communication process, called “Quorum Sensing”, allows bacteria to synchronize behavior on a population‐wide scale. Behaviors controlled by quorum sensing are usually ones that are unproductive when undertaken by an individual bacterium acting alone but become effective when undertaken in unison by the group. For example, quorum sensing controls virulence factor production, biofilm formation, and the excretion of public goods such as enzymes that solubilize solid food sources making them accessible for consumption. We developed small molecule quorum‐sensing agonists and antagonists to discover the principles underlying the exquisite selectivity quorum‐sensing receptors have for their cognate ligands. Our results suggest mechanisms bacteria use in the wild to ensure the proper ligand has interacted with its partner receptor prior to eliciting signal transduction. We suggest that, in their native environments, bacteria encounter mixtures of autoinducers produced by other species occupying the same niche. Precise autoinducer discrimination enables a particular species of bacteria to respond exclusively to its own signal even in the face of fierce competition. This ability prevents the leakage of benefits of quorum‐sensing‐controlled public goods to non‐kin. Beyond learning about fundamental principles underlying quorum sensing, another use for our synthetic molecules is to control quorum sensing on demand. Indeed, our most potent quorum sensing modulators protect animals from quorum‐sensing‐mediated killing by pathogenic bacteria and prevent biofilm formation. These results validate the notion that targeting quorum sensing has potential for antimicrobial drug development.

S12 Modeling human brain development and disease in stem cell derived 3D organoid culture. J.A. Knoblich1, M. Lancaster1, M. Renner1; 1IMBA, Institute of Molecular Biotechnology, Vienna, Austria The human brain is highly unique in size and complexity. While many of its characteristics have been successfully studied in model organisms, recent experiments have emphasized unique features that cannot easily be modeled in animals. We have therefore developed a 3D organoid culture system derived from human pluripotent stem cells that recapitulates many aspects of human brain development. These cerebral organoids are capable of generating several brain regions including a well‐organized cerebral cortex. Furthermore, human cerebral organoids display stem cell properties and progenitor zone organization that show characteristics specific to humans. Finally, we use both RNAi and patient specific iPS cells to model microcephaly, a human neurodevelopmental disorder that has been difficult to recapitulate in mice. This approach reveals premature neuronal differentiation with loss of the microcephaly protein CDK5RAP2, a defect that could explain the disease phenotype. Our data


demonstrate an in vitro approach that recapitulates development of even this most complex organ, which can be used to gain insights into disease mechanisms.

S13 Primary functions for "secondary" metabolites in microbial communities. D.K. Newman1; 1BBE and GPS, California Institute of Technology, Pasadena, CA Conventionally, microorganisms are grown in the laboratory under conditions where doubling times are less than an hour. While it is well appreciated that microorganisms typically grow much more slowly in nature, often in multicellular communities called “biofilms”, we know very little about how they sustain their metabolic activity in these contexts. Yet for over a century, microbiologists have noticed that many bacteria produce colorful pigments at the tail end of growth, leading to the classification of these molecules as “secondary” metabolites. In this talk, I will discuss my group’s efforts to test the hypothesis that these pigments play critical roles in bacterial physiology under conditions where oxidants are limited, such as within the anoxic zone of biofilms. I will mainly focus on the opportunistic pathogen, Pseudomonas aeruginosa, and its ability to leverage its production of phenazines—a versatile class of redox‐active pigments—to aid in its survival. I will explain the implications of our phenazine findings for understanding the physiological functions of “secondary” metabolites more broadly, and how this knowledge may be leveraged to control microbial communities.

Symposium 6: Logic of Signaling

S14 Dissecting regulatory circuits: from cells to tissues. A. Regev1; 1MIT and Broad Institute of MIT and Harvard/HHMI, Cambridge, MA No abstract available.

S15 Anastasis, recovery from the brink of apoptotic cell death, in normal development and disease. G. Sun1, X.A. Ding1, Y. Argaw1, J.O. Wong1, D. Cai1, J. Mondo1, J.P. Campanale1, a. Mishra1, C. Conner1, V. Balasanyan1, Y. Sallak1, D.J. Montell1; 1MCDB, University of California, Santa Barbara, CA Apoptosis is a programmed form of cell suicide that is conserved in all multicellular organisms and functions to remove excessive or damaged cells during development and stress. Excessive apoptosis contributes to degenerative diseases, whereas inappropriate inhibition of apoptosis can cause cancer or autoimmune disease. Activation of executioner caspases is a critical step and until recently, has been considered the point of no return. New work demonstrates that under some circumstances cells can survive caspase activation in a process named anastasis. I will describe the physiological significance of this survival pathway and insights into the underlying molecular mechanisms.


Minisymposium 13: Cell Death and Genome Instability

M131 RIPK1 mediates axonal degeneration by promoting inflammation and necroptosis in ALS. Y. Ito1, D. Ofengeim1, J. Yuan1; 1Cell Biology, Harvard Medical School, Boston, MA Mutations in the optineurin (OPTN) gene have been implicated in both familial and sporadic amyotrophic lateral sclerosis (ALS). However, the role of this protein in the central nervous system (CNS) and how it may contribute to ALS pathology are unclear. Our recent study found that optineurin actively suppressed receptor‐interacting kinase 1 (RIPK1)–dependent signaling by regulating its turnover. Loss of OPTN led to progressive dysmyelination and axonal degeneration through engagement of necroptotic machinery in the CNS, including RIPK1, RIPK3, and mixed lineage kinase domain–like protein (MLKL). Furthermore, RIPK1‐ and RIPK3‐mediated axonal pathology was commonly observed in SOD1G93A transgenic mice and pathological samples from human ALS patients. Thus, RIPK1 and RIPK3 play a critical role in mediating progressive axonal degeneration. Furthermore, inhibiting RIPK1 kinase may provide an axonal protective strategy for the treatment of ALS and other human degenerative diseases characterized by axonal degeneration.

M132 Helicobacter pylori infection promotes genomic instability through a caspase‐dependent pathway without promoting cell death. A. Zamperone1, D. Cohen1, A. Muesch1; 1Developmental and Molecular Biology, Albert Einstein College of Medicine, Bronx, NY Infection with Helicobacter pylori (H.p.), a human adapted bacterium that colonizes the gastric mucosa, is considered the largest risk factor for the development of gastric cancer. H.p. strains that express cytotoxin‐associated gene A (CagA) present a much higher cancer risk than CagA –negative strains, and CagA acts as an oncogene in mice. But how this non‐obligate cytotoxin contributes to the cancer risk remains incompletely understood. A significant factor for the transformation risk in chronically infected patients appears to be the ability of H.p. to inflict DNA damage, including Double‐Stand Breaks (DSBs), in its host. The latter are particularly poised to give rise to mutations and DNA translocations because they are repaired with a high error rate. The mechanism by H. p. induces DNA DSBs and whether CagA is involved, is not yet clear. To address these questions, we adapted a human gastric epithelial primary culture system that allows for the continuous passage of mucosal gastric cells with little genetic drift (Schlaermann et al., 2016). 3D spheroids of mucus‐secreting cells are transferred onto 2D substrates prior to infection with H.p. Our rationale was that primary epithelial cells are more likely to mimic in vivo signaling processes than commonly employed immortalized gastric cancer cell lines. We report here, for the first time, that infection with Helicobacter pylori triggers Caspase‐Activated DNAse (CAD) to generate DSBs in gastric epithelial cells without actually driving the cells into apoptosis. We characterized the underlying signaling cascade that leads from H.p. 's CagA to the activation of the c‐Jun activated kinase (Jnk) that promotes and regulates histone H2A.x‐phosphorylation and CAD activity in infected cells. Recently, the serine/threonine kinase Par1/MARK2 was identified as a CagA‐target in the context of the loss of apical‐basolateral polarity observed in gastric ulcers after H.p. infection. In this study we determined that Par1b‐inhibition by CagA is also essential for the DNA‐damage pathway by causing Jnk activation via Rho‐Kinase, a known downstream effector of Par1b signaling. This pathway


could represent one of the molecular mechanisms of H.p. ‐related early events that can lead to carcinoma initiation in infected patients.

M133 Cause and mechanism of Xenopus hybrid unviability. R. Gibeaux1, R. Heald1; 1Department of Molecular Cell Biology, University of California, Berkeley, CA The two frog species Xenopus laevis and Xenopus tropicalis diverged from a common ancestor ~65 million years ago and now differ at multiple levels. X. tropicalis has 20 chromosomes and is smaller (~4 cm), whereas X. laevis has 36 chromosomes and is larger (~10 cm). Scaling at the organismal and genome levels is accompanied by differences in the size of the egg as well as nuclei and spindles formed in egg extracts. Despite these differences, hybrid embryos can be produced by cross‐fertilization. Strikingly, while the hybrid produced when X. laevis eggs are fertilized by X. tropicalis sperm (le × ts) is viable and gives rise to adult frogs of intermediate size, the reverse hybrid (te × ls) dies as late blastulae. Interestingly, this difference in survival suggests that while X. laevis and X. tropicalis proteins and other components are compatible and able to function together in a living organism, the maternal contribution is crucial, and unlike the case for X. laevis egg, either the molecular contents or the size of the X. tropicalis egg cannot support a hybrid genome. Our results show that te × ls hybrid death is cell autonomous and due to the X. laevis DNA, which leads to chromosome mis‐segregation during mitosis. We have found that this mis‐segregation happens from early development and is not due to the increased size of the hybrid genome. Whole genome sequencing revealed that the chromosome loss is selective for the X. laevis chromosomes 3L and 4L and additional results suggest that this loss is due to the incompatibility of these two chromosomes with the X. tropicalis cytoplasm. By analysis of transcriptomic, proteomic, metabolomic and lipidomic data, we are investigating how unbalanced gene expression at the mid blastula transition leads to a metabolic crisis, which results in unprogrammed te × ls hybrid cell and embryo death.

M134 Catastrophic Chromosome Shattering in Micronuclei is Coupled to Inaccurate Reassembly by Canonical End Joining. P. Ly1,2, L.S. Teitz3,4, D.H. Kim1,2, O. Shoshani1,2, H. Skaletsky3,5, D. Fachinetti6, D.C. Page3,4,5, D.W. Cleveland1,2; 1Ludwig Institute for Cancer Research, La Jolla, CA, 2Department of Cellular and Molecular Medicine, University of California San Diego, La Jolla, CA, 3Whitehead Institute for Biomedical Research, Cambridge, MA, 4Department of Biology, Massachusetts Institute of Technology, Cambridge, MA, 5Howard Hughes Medical Institute, Cambridge, MA, 6CNRS UMR144, Institut Curie, Paris, France Chromosome missegregation into a micronucleus can cause complex and localized genomic rearrangements known as chromothripsis, but the underlying mechanisms remain poorly understood. To directly examine the fate of missegregated chromosomes, we developed an inducible CENP‐A chimera replacement strategy that selectively and functionally inactivates the Y centromere in otherwise diploid human cells. Using this approach, we identified a temporal cascade of events that are initiated upon centromere inactivation involving chromosome missegregation, fragmentation, and religation that span three consecutive cell cycles. Following centromere inactivation, a micronucleus harboring the Y chromosome is formed in the first cell cycle. Extensive chromosome shattering, producing up to 53 dispersed fragments from a single chromosome, is triggered by premature


micronuclear condensation during mitotic entry of the second cycle. Lastly, haphazard re‐joining of DNA fragments by non‐homologous end joining (NHEJ) facilitates intra‐chromosomal rearrangements in the third cycle. Thus, in addition to aneuploidy, errors in cell division can also provoke further genomic instability through fragmentation of micronuclear DNAs in mitosis coupled to aberrant NHEJ‐mediated reassembly in the subsequent interphase.

M135 Determining the missegregation rates of individual human chromosomes. J.T. Worrall1, S.E. McClelland1; 1Centre for Molecular Oncology, Barts Cancer Institute , London, United Kingdom Recurrent aneuploidies are associated with cancer, and ageing in humans, however mechanisms underlying these are unclear. Human chromosomes vary greatly in size, gene density, heterochromatin content and structure among other properties, and therefore may behave differently to one another under cellular stresses affecting chromosome replication and segregation. However, it is not fully understood whether some chromosomes are more prone to missegregation, and whether this varies depending on the specific mechanism causing aneuploidy. To address these important questions we are performing the first comprehensive examination of chromosome missegregation rates of individual human chromosomes under different cellular stresses. To do this we have employed fluorescence In‐Situ hybridization (FISH) imaging of centromeres, coupled to high throughput analysis using the ImageStream®X cytometer, allowing us to score missegregation events in tens of thousands of cells and generate highly accurate rates of chromosome gain or loss for each chromosome. Findings are then validated using traditional microscopy and centromeric FISH assays. Using this approach we have identified striking differences in rates of chromosome gain and loss between individual chromosomes. For example, under conditions where improper kinetochore‐microtubule attachments are promoted (nocodazole washout treatment), we observe high rates of chromosome instability specifically for chromosomes 1, 3, 4, 10, 14 and 21. In contrast, following chromosome missegregation induced by defects in DNA replication in the preceding cell cycle (aphidicolin treatment), chromosomes 7, 13, 14, 18, 18 and X are particularly unstable. Lastly, chromosomes X and 18 were frequently observed to be unstable across multiple conditions and cell lines. This suggests a general instability of these chromosomes, and possibly explains the frequent loss of chromosome 18q in colorectal cancer, and the loss of sex chromosomes associated with ageing in humans. Taken together this new approach to comprehensively analyze aneuploidy patterns following different cellular insults may have the power to predict mechanistic causes of aneuploidy in cancer and aging.

M136 Small Molecule Antagonists of the Deubiquitinase USP7 Interfere With Ubiquitin Binding. I.E. Wertz1; 1Departments of Molecular Oncology and Early Discovery Biochemistry, Genentech, Inc., South San Francisco, CA Deubiquitinase (DUB) enzymes are rapidly emerging as attractive drug targets. In particular, ubiquitin specific protease‐7 (USP7) is a DUB that promotes tumorigenesis by numerous mechanisms; thus, developing chemical inhibitors of USP7 activity may be of therapeutic benefit. However, identifying bona fide selective inhibitors of DUB catalytic activity has been challenging. For example, compound screening assays are often based on DUB substrates conjugated to fluorescent reporters that may be


subject to biochemical artifacts. We therefore embarked on an NMR‐based screening campaign to select small molecule fragments that bind the USP7 catalytic domain in a discrete fashion. In this way we identified a fragment series that specifically binds in the USP7 catalytic site. Further Rapid Overlay of Chemical Structure (ROCS) searches to find compounds with similar three‐dimensional properties, combined with re‐mining of lead fragments, identified two additional series of compounds. Surprisingly, X‐Ray crystallographic and NMR structural studies revealed these compounds bind the USP7 palm region that is adjacent to, but distinct from, the catalytic site. Chemical optimization yielded compounds having enhanced potency in biochemical and cellular assays that retain USP7 selectivity. Further structural analysis determined that these compounds achieve inhibition by interfering with ubiquitin binding to the USP7 catalytic cleft. In sum, our studies identify discrete, functionally significant ligand binding sites within the USP7 catalytic domain. The optimization of compounds that inhibit USP7 by blocking substrate binding provides new opportunities for the development of USP7 inhibitors, and may be a strategy more broadly applicable to inhibition of additional DUBs.

M137 Covalent irreversible inhibitors of USP7 as small molecule cancer immunotherapy agents. J. Wang1, J. Wu1, L. Wang2, I. Sokirniy1, H. Wang1, P. Nguyen1, J. Weinstock1, M. Mattern1, W.W. Hancock2, S. Kumar1; 1Progenra, Malvern, PA, 2Children’s Hospital of Philadelphia, Philadelphia, PA Tumors employ various checkpoint strategies to suppress the immune system’s ability to recognize and destroy cancer cells. Checkpoint inhibitors such as anti‐PD1 antibodies unleash anti‐tumor immune response against a variety of tumors. Regulatory T cells (Tregs) that are immunosuppressive in the tumor microenvironment are also a critical component in tumor immune evasion. Therefore depletion of Tregs or impairment of Treg function is considered an attractive therapeutic approach. USP7, a deubiquitylase (DUB) and a master regulator of several cancer signaling pathways has recently emerged as an essential factor in maintaining Treg functions. Conditional deletion of USP7 in Tregs results in impairment of their function leading to immune activation. Progenra has previously reported several USP7 inhibitors that exhibited anti‐tumor activities in variety of tumor models. Here we describe characterization of a series of covalent irreversible inhibitors that selectively target USP7 catalytic site. Pharmacokinetic studies revealed sustained USP7 inhibition after short term inhibitor treatment and subsequent changes in the level and ubiquitylation of various pharmacodynamic markers including the Treg lineage specific transcription factor Foxp3. USP7 inhibitors also impaired Treg functions and exhibited robust anti‐tumor activity against syngeneic lung tumor models in immunocompetent mice. In addition, Progenra’s USP7 inhibitors are shown to increase the effects of anti‐PD1 antibody and cancer vaccines. As such, USP7 inhibitors possess unique dual activity capable of directly suppressing tumor growth as well as promoting anti‐tumor immunity via inhibiting Treg functions. These studies strongly suggest that USP7 inhibitors can be combined with currently approved cancer immunotherapy agents to improve their therapeutic efficacy.


M138 Targeting endoplasmic reticulum‐resident proteins for the treatment of B cell cancer. C.A. Tang1, J. Del Valle2, C.A. Hu1; 1The Wistar Institute, Philadelphia, PA, 2Department of Chemistry, The University of South Florida, Tampa, FL The IRE‐1/XBP‐1 pathway is the most conserved endoplasmic reticulum (ER) stress response. We discovered that genetic deletion of XBP‐1 significantly decelerates the progression of chronic lymphocytic leukemia (CLL) in Emu‐TCL1 mice. We set up a high throughput in vitro screening platform to look for effective inhibitors that can block the IRE‐1/XBP‐1 pathway. We discovered a specific inhibitor, B‐I09, which blocks the IRE‐1/XBP‐1 pathway with high potency and efficacy. B‐I09 clearly suppresses activation of the IRE‐1/XBP‐1 pathway as evidenced by the decreased mRNA and protein levels of XBP‐1 in intact cells. B‐I09 specifically targets mouse CLL cells in vivo by inducing apoptosis. We also discovered that genetic deficiency of XBP‐1 compromises the BCR signaling which is crucial for the survival of CLL cells. To test whether pharmacological inhibition of XBP‐1 could enhance the effect of inhibitors to the BCR signaling, we combined B‐I09 with Bruton’s Tyrosine Kinase (BTK) inhibitor, ibrutinib, to treat human CLL cells. A true and strong synergistic effect in pharmacology was observed using the Chou‐Talalay combination index method. Such results suggested that B‐I09 can decelerate the growth of CLL either as a single agent or in combination with ibrutinib. Our finding in synergism may be important because B‐I09 may help ibrutinib to achieve higher cytotoxicity in CLL cells at a lower dose, addressing ibrutinib’s toxicity issue. Our studies also led us to discover that IRE‐1 interacts with another ER‐resident protein called STING. STING is critical for cytoplasmic DNA sensing and interferon production. The agonists of STING are recently shown to activate anti‐tumor functions of T and NK cells, resulting in efficient control of cancer in mouse models. We showed that the IRE‐1/XBP‐1 pathway is required for the interferon‐producing function of STING, and that the agonists of STING selectively trigger mitochondria‐mediated apoptosis in normal and malignant B cells. Upon stimulation, STING was degraded less efficiently in B cells, implying that prolonged activation of STING can lead to apoptosis. Transient activation of the IRE‐1/XBP‐1 pathway partially protected agonist‐stimulated malignant B cells from undergoing apoptosis. In CLL‐bearing mice, injection of the STING agonist, 3'3'‐cGAMP, induced apoptosis and regression of leukemia. Similarly efficacious effects were elicited by 3'3'‐cGAMP injection in syngeneic or immunodeficient NSG mice grafted with malignant B cells. These data suggested that, in addition to boosting anti‐tumor immune responses, STING agonists can directly eradicate CLL as well as other types of B cell malignancies in vivo. IRE‐1 and STING are thus useful ER‐resident protein targets for the treatment of B cell cancer.

M139 Endosomal interactions with mitochondria during mitochondrial outer membrane permeabilization. N.R. Brady1, A. Hamacher‐Brady1; 1Molecular Microbiology Immunology, Johns Hopkins University Bloomberg School of Public Health, Baltimore, MD Mitochondrial outer membrane permeabilization (MOMP) by Bax commits mitochondria to the activation of intrinsic apoptosis. Bax transitions from an inactive cytosolic monomer to a membrane‐inserted oligomer which mediates MOMP, a process involving protein and lipid interactions. We have recently reported that in response to a variety of cell death triggers, XIAP E3 ligase activity leads to mitochondrial ubiquitination, associated with XIAP‐mediated recruitment of early


and late endosomal machinery to and into mitochondria. Importantly, inter‐organellar communication is emerging as a key component of organelle function and regulation. Therefore, here we tested the hypothesis that during apoptosis, XIAP‐mediated endosomal interactions with mitochondria regulate Bax activation. By high‐resolution live cell imaging we show that endosomes interact with the outer mitochondrial membrane during and following the onset of MOMP, in a XIAP‐dependent manner. We further demonstrate that endocytic trafficking targets cholesterol, a known inhibitor of Bax pore formation activities, to regions of Bax clustering at the mitochondrial outer membrane. We will discuss a new model suggesting that endocytic trafficking can control mitochondrial apoptotic signaling.

M140 Genetically encoded fluorogenic protease reporters visualize spatiotemporal dynamics of apoptosis in vivo. X. Shu1; 1Cardiovascular Research Institute, Department of Pharmaceutical Chemistry, University of California San Francisco, San Francisco, CA Proteases play fundamental roles in animal development, homeostasis and disease. Genetically encoded fluorogenic protease reporters, of which the fluorescence is turned on by protease activity, are ideal for imaging spatiotemporal dynamics of protease activity in intact animals. First, based on a bacterial phytochrome‐derived monomeric infrared fluorescent protein (mIFP) that incorporates endogenous biliverdin as the chromophore, we have redesigned mIFP so that its chromophore incorporation is regulated by protease activity. Upon protease activation, the infrared fluorogenic protease reporter (named iProtease) becomes fluorescent with no requirement of exogenous cofactor. To demonstrate biological applications, we have designed an infrared fluorogenic executioner‐caspase reporter (named iCasper), which reveals spatiotemporal coordination between cell apoptosis and embryonic morphogenesis, as well as dynamics of apoptosis during brain tumorigenesis in Drosophila. Second, we have developed a green fluorogenic protease reporter by genetically modifying the self‐assembling split GFP so that the two parts cannot self‐assemble until the modifications are “unmasked” by proteases. The reporter, dubbed “ZipGFP”, achieves ~10‐fold fluorescence increase upon proteolytic cleavage. We have demonstrated that ZipGFP‐based executioner caspase reporter is able to visualize spatiotemporal dynamics of caspase activity or apoptosis in the embryonic brain of zebrafish, which is an important model vertebrate for understanding embryonic development. iCasper and ZipGFP caspase reporter will thus be important tools in imaging apoptosis in vivo, which plays essential roles in animal development and disease. They will also be useful to dissect signaling networks that lead to apoptosis by hypothesis‐driven studies as well as unbiased high throughput screening in live animals such as zebrafish and Drosophila. ZipGFP and iProtease may be further used to design reporters of many other proteases, which play essential roles for many biological processes and are tightly regulated by cell signaling networks.


Minisymposium 14: Cell Mechanics

M141 Excitable RhoA dynamics drive pulsed contractions in the early C. elegans embryo. J.B. Michaux1, F.B. Robin2, W. McFadden3, E.M. Munro1; 1Molecular Genetics and Cell Biology, University of Chicago, Chicago, IL, 2INSERM, Biologie du Développement Paris Seine, Paris, France, 3Biophysics Graduate Program, University of Chicago, Chicago, IL Pulsed contractility is a widespread mode of actomyosin contractility that plays diverse roles in tissue morphogenesis. However, the mechanisms that govern pulsed contractility remain poorly understood in any system. Here, we identify a core mechanism for pulse generation in the early C. elegans embryo. Using single molecule imaging and particle tracking analysis, we provide definitive evidence that pulsed actomyosin accumulation is governed almost entirely by local dynamic control over actomyosin assembly and disassembly. Using fast multicolor near TIRF microscopy, we show that pulsed activation and inactivation of RhoA precedes, respectively, the accumulation and disappearance of its downstream targets F‐actin, Myosin II and Anillin. RhoA activity peaks before the onset of contraction and locally pulsed activation of RhoA persists in the complete absence of Myosin II, ruling out requirements for Myosin II activity or contractility in pulse initiation. We present evidence that RhoA feeds back rapidly and positively to promote its own activation and drive the upswing of each pulse, while delayed F‐actin‐dependent accumulation of the RhoGAPs RGA‐3/4 provides negative feedback to terminate each pulse. Simple models constrained by these experimental data confirm that this combination of fast positive and delayed negative feedback is sufficient to generate locally excitable RhoA dynamics and reproduce the temporal dynamics of local RhoA and RGA‐3/4 accumulation. We propose that mechanochemical coupling of excitable RhoA dynamics to an actomyosin cortex is a key underlying feature of pulsed contractility that is differently tuned in different contexts to produce a diversity of morphogenetic outcomes.

M142 A mechanosensitive cdk1 threshold shapes cortical contractions in oocytes. J. Bischof1, C.A. Brand2, K. Somogyi1, U.S. Schwarz2, P. Lénárt1; 1Cell Biology and Biophysics, EMBL, Heidelberg, Germany, 2Institute for Theoretical Physics, University of Heidelberg, Heidelberg, Germany Cells are divided by contraction of the actin cortex patterned in space and time in a tightly controlled manner. In cytokinesis of somatic cells, spindle microtubules provide the dominant cues that have been extensively studied and mechanisms are beginning to be understood. By contrast, in large dividing cells, in particular meiotic oocytes and eggs, striking shape changes termed surface contraction waves (SCWs) are observed that appear to be patterned independent of the spindle. That SCWs are linked to the cell cycle and that they may play important roles in the spatial organization of the egg cytoplasm has long been established, but the molecular mechanisms underlying contractility and patterning SCWs in space and time are not at all understood. Here, we investigated SCWs during meiotic divisions of starfish oocytes. We show that the SCWs manifest as a band of cortical contractility traveling across the oocyte just preceding anaphase and polar body formation. This contractility is mediated by recruitment of non‐muscle myosin II (Myo II) to the cortex regulated by the conserved RhoA‐Rok‐Myo II cascade, the same molecular module that drives the contraction of the cytokinetic ring in somatic cells. However, unlike somatic cells this cascade is


activated independent of microtubules and is directly controlled by the key cell cycle kinase, cdk1‐cyclin B. We further show that cdk1‐cyclin B forms a concentration gradient across the cell which is initially established by import of cdk1‐cyclinB into the off‐centre located nucleus, and is maintained up until anaphase. The SCW is directly patterned by this cdk1‐cyclin B gradient with the SCW wave front following a fixed threshold of cdk1‐cyclin B across the cell as cyclin B is degraded at anaphase. Intriguingly, we find that the threshold of cdk1‐cyclin B at which the SCWs occur depends on the mechanical properties of the cell, with SCWs triggered in softer cells at higher thresholds and thus exhibiting stronger SCWs starting earlier, while in harder cells SCWs are triggered at a lower threshold and are thus weaker and shorter. Together, we show that SCWs are driven by the highly conserved RhoA‐Rok‐Myo II pathway controlled directly by a spatiotemporal gradient of cdk1‐cyclin B. Integrating these data in a computational model quantitatively recapitulates this complex cellular behaviour patterned by a biochemical activity gradient triggered at mechanosensitive thresholds.

M143 Propagating cortical actomyosin‐generated tension to intercellular adhesion complex by myosin‐1c. V. Tang1; 1Cell and Developmental Biology, University of Illinois, Urbana‐Champaign, Urbana, IL Myosin 1c has been shown to play a role in E‐cadherin stabilization at epithelial junction through the regulation of E‐cadherin recycling. Using deconvolution and structured illumination microscopy, we found that myosin 1c does not colocalize with the E‐cadherin complex at the cell junction or on the lateral intercellular interface. Thus, myosin 1c cannot be directly participating in E‐cadherin recycling and must stabilize the cell junction via another mechanism. We found that application of intercellular tension induces myosin 1c accumulation at the lateral plasma membrane between epithelial cells, i.e. cell‐cell interface. We hypothesized that the accumulation of myosin 1c at cell‐cell interface allows coupling of actomyosin contractility to the generation of intercellular tension, thus is essential for tension‐induced junction maturation. In agreement with our hypothesis, a point mutation of myosin‐1c at the ATP‐binding pocket that abrogates its actin motility activity acts as a dominant negative mutant and prevented junction recruitment of a‐actinin‐4. In addition, knockdown of myosin 1c KD prevented junction recruitment of both the upstream regulator of a‐actinin‐4, synaptopodin, as well as a‐actinin‐4. However, myosin‐1c localization or tension‐induced myosin‐1c targeting to cell‐cell interface were unaffected by knockdown of a‐actinin‐4 or synaptopodin, suggesting that myosin‐1c is upstream of both synaptopodin and a‐actinin‐4. Interestingly, application of cyclic intercellular tension rescued tension‐dependent a‐actinin‐4 accumulation, indicating that myosin‐1c KD did not disrupt the overall integrity of the junctional complex. On the other hand, pulsatile tension failed to rescue myosin‐1c KD and a‐actinin‐4 recruitment. Thus, myosin‐1c functions not only to propagate but also modulate cortical tension generated by cellular contractility. These results indicate that the primary role of myosin 1c in E‐cadherin regulation is through a mechanotransduction mechanism rather than structural or biochemical recycling mechanisms.


M144 Dynamic mechanosensory response of Spectrin restricts cell adhesion molecules at the fusogenic synapse during cell‐cell fusion. R. Duan1, J. KIM1, K. Shilagardi1, E. Schiffhauer2, S. Son3, T. Luo4, D.A. Fletcher3, D.N. Robinson2, E.H. Chen1,5; 1Molecular Biology and Genetics, Johns Hopkins University School of Medicine, Baltimore, MD, 2Cell Biology, Johns Hopkins University School of Medicine, Baltimore, MD, 3Bioengineering, University of California, Berkeley, Berkeley, CA, 4Modern Mechanics, University of Science and Technology of China, Hefei, China, 5Molecular Biology, UT Southwestern Medical Center, Dallas, TX Cell‐cell fusion is critical for the conception, development and physiology of multicellular organisms. Our previous studies have shown that cell‐cell fusion is an asymmetric and invasive process involving an attacking cell and a receiving cell. While the attacking fusion partner extends actin‐propelled membrane protrusions into the receiving cell, the latter mounts a Myosin II‐mediated mechanosensory response to resist the pushing force from the attacking cell. The pair of pushing and resisting forces put the fusogenic synapse under high mechanical tension to promote fusogen engagement and cell membrane merger. In a genetic screen for additional actin cytoskeletal regulators in cell‐cell fusion, we have uncovered an essential function for Spectrin in Drosophila myoblast fusion. Spectrin is a membrane skeleton protein critical for maintaining cell membrane integrity and cell shape. Myoblast fusion is severely disrupted in alpha and betaheavy spectrin double mutant embryos, and both alpha and betaheavy Spectrin are enriched at the fusogenic synapse on the side of the receiving cell as Myosin II. Strikingly, unlikely the well‐characterized stable scaffold formed by Spectrin in erythrocytes and axons, Spectrin exhibits dynamic accumulation to the fusogenic synapse in response to the invasive protrusions from the attacking cell. Micropipette aspiration (MPA) and Atomic Force Microscopy (AFM) experiments on cultured cells that are induced to fuse reveal rapid mechanosensory accumulation of Spectrin. However, distinct from Myosin II, which exhibits mechanosensory response to cellular dilation, the mechanosensory response of Spectrin is specific to shear forces. Coarse‐grained molecular dynamic simulation show overlapping domains of dilation and shear in the receiving cell along the invasive protrusions of ~250 nm radius, providing an explanation for the overlapping accumulation of Spectrin and Myosin II at the fusogenic synapse. We further demonstrate that the function of the mechanosensory accumulation of Spectrin at the fusogenic synapse is to restrict cell adhesion molecules (CAMs), since the CAMs required for cell‐cell fusion are largely diffused in alpha and betaheavy spectrin double mutant embryos. We suggest that Spectrin restricts CAMs via biochemical interactions with adaptor proteins. Taken together, this study has revealed a surprising dynamic response of Spectrin to mechanical forces and has general implications for understanding the mechanosensory function of Spectrin in various physiologically relevant cellular processes.

M145 Direct coupling of the actin cytoskeleton to a viral fusogen drives cell‐cell fusion. K. Chan1, M.D. Vahey2, S. Son2, E.M. Schmid2, D.A. Fletcher1,3,4; 1UC Berkeley/UC San Francisco Graduate Group in Bioengineering, Berkeley, CA, 2Department of Bioengineering, University of California, Berkeley, CA, 3Department of Bioengineering Biophysics Program, University of California, Berkeley, CA, 4Physical Biosciences Division, Lawrence Berkeley National Laboratory, Berkeley, CA Cell‐cell fusion occurs during normal development and tissue repair, as well as during infection by some viruses and bacteria, likely to promote pathogen spread. Although membrane fusion of intracellular


vesicles and during viral entry have been well studied, the mechanism by which the plasma membranes of two cells fuse is poorly understood. Fusogenic membrane proteins and the actin cytoskeleton have been suggested to work together to mediate the merger of the two plasma membranes, but how actin assembly and fusogens are coordinated is not known. We find that the reovirus fusion‐associated transmembrane (FAST) protein p14 drives cell‐cell fusion by directly coupling to the actin cytoskeleton. Using a fluorescent cell‐cell fusion assay and biochemical perturbations, we show that branched actin network assembly plays a critical role in p14‐mediated cell‐cell fusion. Specifically, p14 binds to an adaptor protein, Grb2, through an SH2 domain binding motif in its cytoplasmic tail. Grb2 binding to N‐WASP then directly couples branched actin network assembly to p14. When Grb2 binding is disrupted through point mutations in the SH2 domain binding motif, cell‐cell fusion is abrogated. Fusion can be restored by coupling the mutant p14 to an actin nucleation promoting factor in an inducible manner. Unlike other types of cell‐cell fusion that may require distinct molecular pathways to trigger actin polymerization necessary for fusion, p14 drives cell‐cell fusion by directly coupling branched actin network assembly to its cytoplasmic tail. This work demonstrates a direct role for the actin cytoskeleton in virus‐mediated cell‐cell fusion.

M146 Spatial Regulation of RhoA Reveals Zyxin‐mediated Elasticity of Stress Fibers. P.W. Oakes1,2, E. Wagner3, C.A. Brand4, D. Probst4, M. Linke4, U.S. Schwarz4, M. Glotzer3, M.L. Gardel2; 1Department of Physics, University of Rochester, Rochester, NY, 2Institute for Biophysical Dynamics, James Franck Institute and Department of Physics, University of Chicago, Chicago, IL, 3Molecular Genetics and Cell Biology, University of Chicago, Chicago, IL, 4Institute for Theoretical Physics and BioQuant, Heidelberg University, Heidelberg, Germany The mechanics of the actin cytoskeleton regulates cell morphogenesis during essential physiological processes. While cellular contractility is known to be largely RhoA‐dependent, how localized biochemical signals translate into cell‐level responses is not well understood. Here, we use an optogenetic approach to spatiotemporally increase RhoA activity to probe the mechanics of the actin cytoskeleton in adherent fibroblasts. Local activation of RhoA results in local increases in F‐actin and myosin, but not de novo formation of stress fibers or focal adhesions. It also stimulates local contraction, leading to increased traction forces that rapidly propagate across the cell via stress fibers and drive actin flow towards the region of heightened RhoA. Surprisingly, the flow reverses direction when local RhoA activation stops. These data are explained with a physical model, which demonstrates that stress fibers are elastic‐like, even at time scales exceeding turnover of constituent proteins. We identify zyxin as a regulator of stress fiber mechanics, as they are fluid‐like in its absence. Such molecular control of mechanics likely plays critical roles in regulation of morphogenic events.

M147 Single molecular force sensitivity and threshold for the activation of B cell receptors. W. Liu1, Z. Wan1; 1School of Life Sciences, Tsinghua University, Beijing, China B lymphocytes show remarkable mechanosensing capability by utilizing the surface expressed B cell receptors (BCRs) to sense antigens. However, the sensitivity and threshold for the activation of BCRs resulting from the stimulation by mechanical forces are unknown. Here we systemically addressed this question using a ds‐DNA based tension gauge tether (TGT) system serving as a predefined single molecular force gauge ranging from 12 to 56 pN to restrict BCR and antigen tension. Utilizing total


internal reflection fluorescence microscopy, we observed that IgM‐BCR activation is sensitive to single molecular forces. We observed an obvious multi‐threshold effect with a low‐end 12‐16 pN force which triggered a weak activation, whereas a 23‐43 pN force initiated a medium‐level activation, and a high‐end 50‐56 pN force produced a strong activation. Such patterned dependence on mechanical forces with multi‐threshold effects did not strictly rely on myosin IIA, dynein and integrin. However, the break‐through of the high‐end but not the medium‐level and low‐end mechanical force threshold in IgM‐BCR activation required myosin IIA and the inside‐out activation of integrin. Surprisingly, we found that the activation of the isotype‐switched IgG‐BCR or IgE‐BCR only requires an extremely low threshold of less than 12 pN, providing an explanation for their rapid activation in response to antigen stimulation. Mechanistically, we found that the cytoplasmic tail of the IgG‐BCR heavy chain is both required and sufficient to account for the low mechanical force threshold to activate the IgG‐BCR. These results defined the single molecular force sensitivity and threshold that are required to activate different isotyped BCRs.

M148 Specialized actin architecture mechanically tunes T cell synaptic contact lifetime. S. Kumari1, Y. Poh1, M. Melo1, M. Mak2, E. Vasile1, R. Kamm1, R. Geha3, D.J. Irvine1; 1Massachusetts Institute of Technology, Cambridge, MA, 2Dept. of Biomedical Engineering, Yale university, New haven, CT, 3Harvard Medical School, Cambridge, MA T cells form strong yet transient contacts with cognate antigen‐presenting cells (APC) called 'immunological synapses' (IS). The stability and duration of the IS are critical determinants of subsequent T cell responses. Once synapse is established, the T cell‐intrinsic mechanisms that drive its disengagement, thus controlling its lifetime, remain unknown. Actin cytoskeleton is conceived to influence IS lifetime, however the precise mechanism by which actin regulates lifetime of mature synapse remains unclear, as the actin cytoskeleton would also be required for preceding step of synapse formation as well as T cell migration following synapse disengagement. We interrogated the role of actin cytoskeleton in regulating synapse lifetime by probing the cytoskeletal makeup of the T cell synaptic interface, using a combination of synthetic and natural antigen presenting surfaces, super‐resolution microscopy and genetic dissection tools, during the synapse disengagement phase. We identified a specialized actin organization, TCR‐Associated Force‐transducing F‐actin Foci (TAFF), plays an important mechanical role in stabilizing synapses by generating traction forces and modulating T cell receptor (TCR)‐associated mechanotransduction. The TAFF are nucleated continuously by Wiskott‐Aldrich Syndrome Protein (WASP) at the TCR clusters in the synapse and represent high‐tension zones. During T cell‐APC disengagement, when the mature stable synapse transitions into a motile synapse (‘kinapse’), cellular WASP levels and consequently TAFF are downregulated, facilitating synapse disengagement. Consistently, WASP‐deficient mouse T cells as well as the Wiskott‐Aldrich Syndrome (WAS) patient T cells, both engage synapses initially, but their synapses lack TAFF, exhibit poor traction and impaired mechanotransduction, and prematurely transition into kinapses. Reconstitution of a non‐degradable form of WASP could rescue the mechanical tension as well as the synapse stability. Together, our results outline a novel and major mechanosensory and tension consolidating cytoskeletal mechanism at the T cell synapse that tunes synapse lifetime.


M149 Ena/Vasp proteins selectively regulate activated T cell trafficking in vivo by mediating transendothelial migration. M.L. Estin1,2, J. Jacobelli1,2; 1Dept. Biomedical Research, National Jewish Health, Denver, CO, 2Dept. Immunology and Microbiology, University of Colorado Denver, Denver, CO The cytoskeletal effector proteins Vasodilator‐stimulated phosphoprotein (VASP) and Ena‐VASP‐like (EVL) regulate cell morphology, adhesion, and migration in various cell types. However, the role of these cytoskeletal effectors in T cell motility, adhesion, and trafficking has not been characterized. The goal of this work was to determine the role of Ena/VASP cytoskeletal effectors in T cell migration in vivo. Our results indicate that EVL and VASP selectively regulate activated T cell trafficking but not naïve T cell trafficking to the lymph nodes and spleen. We also investigated activated T cell trafficking during inflammation using a model of Multiple Sclerosis, in this setting EVL and VASP deficiency severely impaired activated T cell trafficking to the inflamed central nervous system (CNS) of mice with experimental autoimmune encephalomyelitis (EAE). EVL/VASP deletion resulted in impaired alpha‐4 integrin expression and function, but unexpectedly did not cause alterations in shear‐resistant adhesion and motility of T cells on primary brain microvascular endothelial cells in vitro. Instead, deletion of EVL and VASP caused a significant impairment in the ability of T cells to initiate and complete the diapedesis step of transendothelial migration. Furthermore, upon blockade of alpha‐4 integrin, T cell diapedesis became equivalent between control and EVL/VASP‐deficient T cells. Together, these data suggest that that EVL and VASP selectively regulate activated T cell trafficking by specifically mediating the diapedesis step of transendothelial migration in an alpha‐4 integrin‐dependent manner.

M150 Coiled‐coils as molecular motors: entropic polymer engines. M. Jahnel1,2, D.H. Murray2, M. Zerial2, S.W. Grill1,2; 1BIOTEC, TU Dresden, Dresden, Germany, 2Max Planck Institute of Molecular Cell Biology and Genetics, Dresden, Germany Coiled‐coils are ubiquitous protein motifs present in over 5% of all eukaryotic proteins. Interestingly, very long coiled‐coils (L > 30 nm) are often found in proteins that can generate and sense forces. Thus extended coiled‐coil proteins take part in some of the most intricate cellular processes: from vesicle tethering to the transport of cargo, from the maintenance of chromosome structure to the attachment of the kinetochore, from the motion of the cytoskeleton to cytokinesis. But how can slender molecules reliably generate mechanical forces in an environment that is constantly agitated by Brownian motion? Here, we discuss a polymer model that consist of a structural cycle of lengthy coiled‐coil proteins. Tethered on one side, a selective binding event on the opposite side leads to a reduction in the polymer's persistence length by disrupting the regular packing of the coiled‐coil. This sudden increase in the flexibility of an extended protein structure generates an entropic collapse force that pulls its ends together. The model will quantitatively discuss our recent discovery [1] that the unusually long membrane tethering protein early endosome antigen 1 (EEA1) – a roughly 200 nm long coiled‐coil protein involved in membrane trafficking – is rather rigid and extended when only tethered at the base, yet becomes much more flexible upon binding of Rab5:GTP to its free end. Importantly, using a reconstituted tethering system and single‐molecule optical tweezer experiments, we demonstrate that this increased flexibility indeed results in an entropic force that pulls two objects together. Moreover, we show that


GTP‐hydrolysis acts as a timer for this interaction and that strategic mutations to the coiled‐coil can interfere with this effect. Taken together, our experimental results and theoretical predictions describe an unusual thermodynamic cycle in which the persistence length of a coiled‐coil biopolymer is modulated by binding and NTP hydrolysis. Compared to traditional molecular motors, switching a slender molecule from rigid to flexible and back by transient binding events could be another effective mechanism for generating forces on the jiggling molecular level. References: [1] Murray, D.H. and Jahnel, M. et al., An entropic collapse force mediates vesicle tethering in early endosomes, Nature, accepted, 2016

Minisymposium 15: Cell Polarity and Morphogenesis

M151 Unraveling PAR polarity protein interactions with single‐cell biochemistry. D.J. Dickinson1,2, B. Goldstein1,2; 1Biology, University of North Carolina, Chapel Hill, NC, 2Lineberger Comprehensive Cancer Center, University of North Carolina, Chapel Hill, NC The conserved PAR protein system mediates polarization in a wide variety of animal cell types including epithelia, neurons and stem cells. PAR proteins localize asymmetrically within a cell and direct the polarization of downstream cellular processes. Although some in vitro binding interactions between PAR proteins are known, a fundamental unanswered question is how PAR proteins are organized biochemically into a signaling network that can mediate cell polarization. To address this question, we developed a method to quantify protein‐protein interactions at the single‐molecule level, and with temporal resolution, in single cells. We used this method to measure interactions between PAR proteins in staged C. elegans zygotes, which are single cells that follow a defined and stereotyped polarization program. We have identified PAR protein complexes that are specifically up‐ or down‐regulated as the zygote polarizes. In particular, our data identify PAR‐3 oligomerization as a key regulatory node for controlling cell polarization: PAR‐3 forms large oligomers during polarity establishment, but these oligomers are absent both before and after polarization. Quantitative live imaging experiments suggest that PAR‐3 oligomerization contributes to polarity establishment by facilitating redistribution of the PAR‐3/PAR‐6/aPKC complex to the anterior half of the zygote. Once polarity establishment is complete, disassembly of PAR‐3 oligomers occurs together with a loss of PAR‐3 binding to the membrane and to PAR‐6/aPKC, suggesting that control of cell polarity shifts to a different polarity complex during polarity maintenance. The timing of this shift is linked to the cell cycle, because disassembly of PAR‐3 oligomers in stably polarized cells requires the activity of the cell cycle kinase PLK‐1. Ongoing experiments that combine single‐cell biochemistry, quantitative live cell imaging, CRISPR/Cas9‐induced targeted mutations and computational modeling in this simple system will reveal how regulated interactions between PAR proteins contribute to the establishment and maintenance of cell polarity.


M152 Patterned Toll receptor expression organizes epithelial cell intercalation. A. Paré1, A. Vichas1, C. Fincher1, Z. Mirman1, D. Farrell1, A. Mainieri1, J. Zallen1; 1Sloan‐Kettering Institute, New York, NY Tissue elongation through convergent extension is a common mechanism throughout animal development that requires the coordinated polarization and movement of large groups of cells. Elongation of the head‐to‐tail body axis in Drosophila is a striking example of convergent extension. This process involves the spatially restricted recruitment of contractile and adhesive proteins to complementary cellular domains, driving globally organized cell intercalation. However, the mechanisms used by cells to sense their spatial orientation relative to the global body axes have long been mysterious. We recently showed that three cell‐surface receptors in the Toll receptor family, Toll‐2, Toll‐6, and Toll‐8, are expressed in a repeating pattern of stripes, creating a spatial code that cells use as a compass to determine directionality within the tissue. We propose a model in which heterophilic interactions between different Toll receptors expressed on neighboring cells enhance contraction and destabilize adhesion at specific cell‐cell interfaces, coordinating cell rearrangements across the tissue. Toll family receptors have been well studied in the context of innate immunity, where they serve as crucial receptors for the detection of pathogen‐derived molecules and endogenous damage‐associated proteins. Our findings suggest a novel role for this protein family in mediating cell‐cell interactions during development. Therefore, it appears that Toll receptors are used in a variety of cellular contexts, not only to distinguish self from non‐self by the immune system, but also to distinguish cells from their neighbors. These studies point to a relatively unexplored and potentially conserved link between Toll receptor signaling and activation of the actomyosin network to induce changes in cell shape and behavior. We are now using genetic, cell biological, and biochemical approaches to identify the signaling mechanisms by which Toll family receptors regulate cell polarity and tissue structure.

M153 Biomechanical analysis of cell behaviors during neural plate convergent extension. D.S. Vijayraghavan1, J.H. Shawky1, L.A. Davidson1,2,3; 1Bioengineering, University of Pittsburgh, Pittsburgh, PA, 2Developmental Biology, University of Pittsburgh, Pittsburgh, PA, 3Computational and Systems Biology, University of Pittsburgh, Pittsburgh, PA At the start of neurulation, the dorsal ectodermal epithelium of the embryo, comprised of the neural plate and non‐neural tissue, undergoes convergent extension (CE). The lateral edges of the neural plate converge towards the dorsal midline while the tissue extends in the perpendicular anterior posterior (AP) direction. To achieve this anisotropic tissue deformation, dorsal ectoderm cells must coordinate and orient their behaviors. Increasing evidence from epithelial morphogenetic events such as Xenopus epiboly and Drosophila wing outgrowth suggest tissue tension may direct cell behaviors. We seek to elucidate whether the tissue mechanical environment regulates the cell behaviors that facilitate neural plate CE in Xenopus laevis embryos. From high‐resolution confocal time‐lapse sequences we performed a kinematic analysis of the apical dorsal surface of Xenopus embryonic ectoderm as the neural plate is shaped. Quantifying and correlating cell‐level deformations with local tissue deformation revealed distinct topological and cell shaping patterns between the neural and non‐neural ectoderm during tissue CE. Within the neural plate, cells undergo directed rosette‐based and T1 transitions, rearranging into narrow AP elongated multicellular arrays. Cells at the lateral border of the neural plate alter their apical shape becoming AP elongated. To understand if these behavioral programs are cell autonomous, we performed a series of


grafting experiments to introduce cells to different external tissue environments. When the neural plate is grafted into isotropically spreading non‐neural explants neural cells continue to form rosettes and undergo apical constriction. Alternatively, non‐neural cells grafted into the center of the converging and extending neural plate stretch in the AP direction and mirror the tissue level deformation. Grafted cells had similar phenotypes to neural plate border cells. We hypothesized that border cells may passively deform due to anisotropic tissue tensions generated during CE. To test this, we used dorsal tissue explants that maintain their ability to converge and extend in isolation. Physically constrained explants whose AP extension was blocked exhibited altered the tissue strains. Rosettes persist in the neural plate in both constrained and freely elongating control explants indicating neural cells maintain their rearrangement program. Border cell elongation and orientation in constrained explants significantly differ from control free explants. These results give insights into the multiscale mechanics of neural plate convergent extension and suggest processes that link cell behaviors to tissue level deformations. Further exploration of this relationship may reveal how aberrant mechanics cause defective morphogenesis.

M154 Organ sculpting by patterned extracellular matrix elasticity. J. Crest1, A. Diz‐Muñoz2, D. Chen1, D.A. Fletcher2, D. Bilder1; 1Molecular and Cell Biology, University of California‐Berkeley, Berkeley, CA, 2Bioengineering Biophysics, University of California‐Berkeley, Berkeley, United States All organs have precise forms that are critical for their functions. Studies of organogenesis focus primarily on cells, but how the surrounding extracellular matrix (ECM) works in tandem to shape an organ remains poorly understood. Here we analyze a simple morphogenetic event –elongation of the Drosophila egg —using direct biophysical measurements of both cells and ECM. These data reveal robust mechanical anisotropy in the ECM‐based basement membrane (BM), but not in cells, during elongation. Atomic force microscopy on in vivo BM documents a stiffness gradient that develops symmetrically along the anterior‐posterior axis, while elongation‐defective mutants retain a uniformly soft BM. Genetic manipulation shows that the instructive determinant is relative rather than absolute BM stiffness, creating differential resistance to isotropic tissue expansion that drives elongation. The stiffness gradient requires morphogen‐like signaling to regulate A‐P BM incorporation, as well as planar‐polarized organization to homogenize circumferential stiffness. These results demonstrate how development can sculpt an organ by patterning mechanical anisotropy within its ECM, rather than its cells.

M155 The molecular signature of loser cells reveals key signalling pathways involved in cell competition. E. Piddini1; 1The Gurdon Institute, University of Cambridge, Cambridge, United Kingdom Cell competition is a process similar to natural selection at the cellular level. During cell competition less fit cells (losers) are eliminated by a population of fitter cells (winners) and this maximises tissue and organism fitness. Even though it is over four decades ago since the discovery of cell competition, the molecular properties that earmark cells as losers have not been identified and it is largely unknown what pre‐existing conditions and differences between winners and losers are responsible for initiating the process.


By comparing the transcriptome of functionally unrelated Drosophila loser mutants (harbouring mutations in ribosomal genes or in the cullin gene mahjong), we have recently identified a molecular signature composed of hundreds of genes differentially expressed in loser cells in their naïve state, i.e. prior to cell competition. This led us to identify 3 key stress signalling pathways as major players in cell competition: the JNK pathway, the JAK/STAT pathway and the oxidative stress response. We find that JNK signaling, chronically activated in loser cells, restricts their intrinsic growth rate. Conversely the JAK/STAT pathway activated in winner cells during cell competition fuels their proliferation, boosting cell competition. Strikingly, we found that loser cells chronically activate the oxidative stress response and that activation of this pathway, even in the absence of actual oxidative damage, is sufficient to turn cells into losers and prime cells for their elimination by wild‐type cells. This is the first known signal, shared among multiple loser mutations, whose activation precedes and causes cell competition. Altogether these findings provide important new mechanistic insights on how cell competition occurs.

M156 Directional tissue migration regulation in early mouse embryo. T.A. Omelchenko1,2, A. Hall2, K.V. Anderson1; 1Developmental Biology Program, Memorial Sloan Kettering Cancer Center, New York, NY, 2Cell Biology Program, Memorial Sloan Kettering Cancer Center, New York, NY The directional collective cell migration is a driving force for embryonic development. Cell movements and cell‐cell adhesion mediated by spatially localized activation of Rho GTPase signaling is required for the formation of tissues and organs. However the cellular and biochemical basis of how cells self‐organize and coordinate their behaviors in mammalian embryos is poorly understood. We previously showed that beta‐Pix, a small Rho GTPase Rac/Cdc42 exchange factor, and Rac1 are required for polarized protrusive activity and collective epithelial migration of the extraembryonic endoderm, which is required for establishment of the anterior‐posterior body axis. Here, using live imaging of early mouse embryos expressing fluorescent reporters, we identify essential roles for beta‐Pix ‐ Cdc42 signaling in collective mesoderm migration. Deletion of beta‐Pix or Cdc42 in the epiblast (the embryo proper) blocks embryonic morphogenesis leading to early lethality at E8.5. We characterized patterns of mesoderm migration and cell protrusive activity and found that beta‐Pix and Cdc42 promote directionality and persistence of migration through formation of specialized protrusions. Intriguingly, beta‐Pix null mesodermal cells show increased motility but not directional migration, leading to widening and thickening of the mutant embryo. Our results indicate that during mammalian embryogenesis protrusive activity provides predominant mechanism regulating directionality of collective cell migration. These studies provide a framework for investigating protrusion‐dependent mechanism of collective cell behaviors that are disrupted in developmental diseases and in cancer.

M157 Genetic and infectious causes of microcephaly caused by NDE1 mutations and Zika virus. D.J. Doobin1, A. Rosenfeld2, A. Carabalona1, V. Racaniello2, R.B. Vallee1; 1Pathology and Cell Biology, Columbia University Medical Center, New York, NY, 2Microbiology and Immunology, Columbia University Medical Center, New York, NY Microcephaly is a brain developmental malformation characterized by an abnormally small neocortex and head size. Recent work from our lab has elucidated the cellular mechanisms responsible for a severe form of microcephaly caused by human mutations in NDE1, a protein involved in cytoplasmic dynein


regulation. Using in utero electroporation of NDE1 short‐hairpin RNA in embryonic rat brain, we have observed three independent points of severe cell cycle arrest in proliferating neural progenitors. These occur during regulation of primary cilia at the G1‐to‐S transition, during G2‐dependent apical interkinetic nuclear migration (INM), and at the G2‐to‐M transition (Doobin et al., Nature Commun., 2016, in press). Overexpression of NDEL1, a NDE1 paralogue, could functionally compensate for NDE1 loss, except at the G2‐to‐M transition, suggesting a novel role for NDE1 in mitotic entry in this system. To gain insight into the mechanisms controlling NDE1 behavior during the neural progenitor cell cycle we have begun to examine the effects of NDE1 phosphorylation by Cdk1. We find that expression of a NDE1 double phosphomutant is sufficient to block apical INM, whereas a single site phosphomutant causes the G2‐M arrest. These results reveal that the severity of NDE1‐associated microcephaly results not from defects in mitosis, but rather the inability of neural progenitors to ever reach this stage. By applying this understanding of microcephaly with a similar methodology to organotypic mouse brain slices, we have begun to investigate the mechanism of Zika virus‐induced microcephaly. Using diverse isolates of Zika virus we observe differences in viral pathology in mouse embryonic brain slice preparations, though nearly all strains can replicate in this system. We observe a severe decrease in mitotic events in the mouse brain progenitor cells, and evidence for subsequent defects in postmitotic neuronal migration and viability as well. By studying both genetic and infectious causes of microcephaly, we hope to obtain a more complete understanding of the cell biology of this disorder, and that of other malformations of cortical development. Supp. by HD40182 and NINDS F30NS095577.

M158 Nanoscale architecture of cadherin‐based cell adhesions. C. Bertocchi1, Y. Wang1, A. Ravasio1, Y. Hara1, Y. Wu1, T. Sailov2, M.A. Baird3, M.W. Davidson4,5, R. Zaidel‐Bar1,6, Y. Toyama1,7,8, B. Ladoux1,9, R. Mege9, P. Kanchanawong1,6; 1Mechanobiology Institute , National University of Singapore, Singapore, Singapore, 2Singapore Centre on Environmental Life Sciences Engineering, Nanyang Technological University, Singapore, Singapore, 3National Heart Lung and Blood Institute, National Institutes of Health, Bethesda, MD, 4National High Magnetic Field Laboratory, The Florida State University, Tallahassee, FL, 5Department of Biological Science, The Florida State University, Tallahassee, FL, 6Department of Biomedical Engineering, National University of Singapore, Singapore, Singapore, 7Department of Biological Sciences, National University of Singapore, Singapore, Singapore, 8Temasek Life Sciences Laboratory, National University of Singapore, Singapore, Singapore, 9Institut Jacques Monod, Université Paris Diderot and CNRS , Paris, France Cadherin‐mediated adhesions are highly dynamic supramolecular complexes formed at sites of cell‐cell contacts which play essential roles in multiple aspects of tissue morphogenesis and multicellularity, including cell recognition and sorting, boundary formation, coordinated cell movements, and the maintenance of tissue polarity. Although the mechanosensitivity of cadherin‐mediated adhesions are well documented, their ultrastructural basis remains still unknown, thus obscuring insights into the underlying molecular mechanisms. To elucidate the molecular architecture of cadherin‐based complexes, we used superresolution microscopy to map the nanoscale organization of key cell‐cell junction proteins, in a biomimetic system whereby cells form cadherin‐mediated complexes on biomimetic cadherin‐coated substrate recapitulating key attributes of early cell‐cell contacts. Such a system offers better optical accessibility, while overcoming spatial resolution limitations due to the imaging depth and the highly variable 3D geometry of the native cell‐cell junctions, and, thus is highly amenable to superresolution imaging and allows high precision (~10‐nm) protein position measurements. Here we report a surprisingly well‐organized molecular architecture stratified along the vertical (z) axis, with the cadherin‐catenin layer and the actin compartment separated by ~30 nm, interposed by a vinculin‐containing interface zone. We observed that vinculin can undergo a


conformational activation, spanning between the cadherin‐catenin layer and the actin compartment. Nanoscale positioning of vinculin is determined by alpha‐catenin, while vinculin conformational state is controlled by contractility and/or Abl kinase phosphorylation on the residue Y822 of vinculin. Vinculin activation, in turn, modulates the positioning of VASP and zyxin, inducing VASP‐mediated actin polymerization, thus resulting in a positive feedback loop that leads to junction strengthening. In conclusion, our measurements reveal a modular nanoscale architecture of cadherin‐based adhesions, suggesting a control principle whereby vinculin serves as a molecular clutch that integrates mechanical and biochemical signals to differentially engage the cadherin‐catenin complexes to the actomyosin contraction machinery under different contexts such as developmental processes or diseases states.

M159 The Mechano‐Responsive E‐cadherin/LGN Complex Instructs Epithelial Cell Division Orientation. M. Gloerich1, K.C. Hart1, J.M. Bianchini1, K.A. Siemers1, D.J. Cohen1, J. Tan2, J. Sim3, B.L. Pruitt3,4,5, W.J. Nelson1,4; 1Biology, Stanford University, Stanford, CA, 2Biophysics Program, Stanford University, Stanford, CA, 3Mechanical Engineering, Stanford University, Stanford, CA, 4Molecular and Cellular Physiology, Stanford University, Stanford, CA, 5Bioengineering, Stanford University, Stanford, CA During epithelial cell proliferation, the planar orientation of division maintains a single‐layered tissue architecture, while the division angle within this plane controls the direction of tissue expansion. Spatial cues from the extracellular environment instruct division orientation, but how cells sense and translate those signals to the mitotic spindle is not well understood. Here we demonstrate that the cell‐cell adhesion protein E‐cadherin is a mechano‐regulated landmark for mitotic spindle orientation. E‐cadherin binds directly to LGN, a core component of the spindle orientation machinery. This complex directs the mitotic recruitment of NuMA and stabilizes cortical associations of astral microtubules at cell‐cell contacts to establish planar cell divisions. Tensile forces on E‐cadherin, through contractility of the actomyosin cytoskeleton or externally applied mechanical strain, regulate junctional LGN recruitment. As a consequence, E‐cadherin and LGN align planar division with anisotropic tissue tension. Thus, epithelial division orientation is coupled to cell‐cell adhesion and tissue tension by the mechano‐responsive E‐cadherin/LGN complex. These results define a mechanism by which cells can translate adhesive and mechanical cues to orient cell divisions, which has broad implications for our understanding of the organization of metazoan tissues.

M160 A Sterile 20 family kinase regulates oogenesis by tuning contractile ring proteins on germline intercellular bridges. K.N. Rehain1, A.C. Love2, M. Werner2, I. Macleod3, J. Yates iii3, A.S. Maddox2; 1Curriculum in Genetics and Molecular Biology, University of North Carolina, Chapel Hill, NC, 2Biology, University of North Carolina, Chapel Hill, NC, 3Chemical Physiology, The Scripps Research Institute, La Jolla, CA Intercellular bridges are actin‐based rings that form stable cytoplasmic connections between cells, promoting communication and coordinated action. These bridges connect germ cells in cysts/syncytia, and are highly conserved features of germlines throughout metazoa. Moreover, these connections are required for fertility. The leading hypothesis in the field proposes that intercellular bridges arise from the stabilization of the cytokinetic ring during incomplete cytokinesis. Regulatory and structural components that localize to and promote cytokinetic ring closure are also present at and regulate stably


open intercellular bridges. Given this paradox, the mechanism of intercellular bridge stabilization is unclear. We examined intercellular bridge dynamics throughout the C. elegans oogenic germline, and found that changes in bridge size dynamics (ex. growing to shrinking) correlate with changes in meiotic stage, suggesting that bridge stability is in part regulated by cell cycle progression. To control for this difference, we limited the remainder of our studies to a single meiotic stage in which intercellular bridges are continuously open, but are highly variable in size. The multi‐domain scaffold protein anillin is enriched on and regulates both cytokinetic rings and intercellular bridges. In C. elegans, canonical anillin ANI‐1 and an endogenous truncation homolog ANI‐2 differentially regulate intercellular bridge size and stability. We predicted that the amount of ANI‐2, which aids bridge stability, would positively correlate with intercellular bridge size, while ANI‐1, which limits bridge stability, would have a negative relationship. As we predicted, ANI‐2 levels positively correlate with intercellular bridge size. However, ANI‐1 levels also positively correlated with bridge size, indicating that negative regulation of intercellular bridges by ANI‐1 is not dependent on the amount of ANI‐1 present. To better understand how ANI‐1 is regulated, we immunoprecipitated ANI‐1 from worm lysates and used mass spectrometry to identify ANI‐1 interacting proteins. We identified GCK‐1, a conserved serine/threonine kinase, as a novel ANI‐1 interactor. We found that together with CCM‐3, a known GCK‐1 binding partner, GCK‐1 inhibits excess localization of ANI‐1 and non‐muscle myosin II (NMY‐2) to intercellular bridges and promotes intercellular bridge stability. The shorter anillin ANI‐2 also limits NMY‐2 localization to bridges and may promote bridge stability via this mechanism. This idea is supported by others’ work showing that inhibition of NMY‐2 stabilizes intercellular bridges in the Drosophila egg chamber. Together with our findings, this suggests an ancient and conserved mechanism for the stabilization of intercellular bridges via myosin inhibition.

Minisymposium 16: Dark Matters in Signaling and Differentiation

M161 Ciliary PI(4,5)P2 dictates fall of primary cilia and rise of cell cycle. S.C. Phua1, T. Inoue1; 1Cell Biology, Johns Hopkins University, Baltimore, MD The primary cilia, which appears during the G0/G1 state of the cell cycle, serves as an external sensor for signals such as Hedgehog. We report here an unexpected role for actin polymerization in the process of cilia disassembly after growth stimulation. We performed live‐cell and super‐resolution fluorescence imaging to demonstrate that quiescent cells relocates phosphoinositide 5‐phosphatase, Inpp5e, when exiting the G0 state and alters ciliary distribution of phosphoinositide (4,5)‐bisphosphate (PI(4,5)P2). Distal accumulation of PI(4,5)P2 in cilia alleviates basal inhibition of actin polymerization, releasing ciliary vesicles in a myosin II‐dependent manner. Proteomic analyses on extracellular ciliary vesicles further revealed an enrichment of IFT‐B components and Hedgehog effectors. Cilia decapitation occurs prior to G1 entry, and G1 duration was prolonged by experimental perturbation of cilia decapitation using a cilia‐targeted actin inhibitor. This newly defined program of cilia disassembly may find significance in the regulation of quiescence exit during normal development and cancer.


M162 A bacterially‐produced aphrodisiac regulates choanoflagellate mating. A. Woznica1, J.P. Gerdt2, J. Clardy2, N. King1; 1Molecular and Cell Biology, University of California, Berkeley, Berkeley, CA, 2BCMP, Harvard Medical School, Boston, MA The earliest animals evolved in oceans teeming with bacteria, and bacterial symbionts continue to profoundly shape the biology of their animal hosts. Choanoflagellates, the closest living relatives of animals, provide a simple and evolutionarily relevant model to help uncover fundamental mechanisms by which bacteria influence animal biology. We recently discovered that the bacterium, Vibrio fischeri, induces mating in the choanoflagellate, S. rosetta. Within minutes of exposure to Vibrio bacteria, S. rosetta haploid cells form mating‐swarms through cell aggregation. Pairs of swarming cells then mate, undergoing cell and nuclear fusion, followed by meiotic recombination. We have determined that the mating induction factor produced by Vibrio is a secreted chondroitin sulfate (CS) lyase that is sufficient to induce choanoflagellate mating at nanomolar concentrations. Interestingly, similar extracellular glycosaminoglycan lyases play essential roles in sperm‐egg binding and fertilization in diverse animals, from sea stars to humans. This discovery that a bacterial cue regulates choanoflagellate mating is the first example of bacteria driving mating in eukaryotes, and raises the possibility that environmental bacteria or bacterial symbionts regulate mating in animals as well.

M163 Autism susceptibility gene TAOK2 mediates dendritic spine maturation through Septin7 phosphorylation. S. Yadav1,2, J.A. Oses‐Prieto3, C. Peters1,2, A. Burlingame3, L.Y. Jan1,2, Y.N. Jan1,2; 1Physiology, University of California San Francisco, San Francisco, CA, 2Howard Hughes Medical Institute, San Francico, CA, 3Pharmaceutical Chemistry, University of California San Francisco, San Francisco, CA Abnormalities in dendritic spines are manifestations of several major neurological and psychiatric diseases. TAOK2 is one of the genes in the 16p11.2 locus, deletions and duplications of which are associated with autism and schizophrenia. Here, we show that the serine/threonine kinase encoded by TAOK2 localizes to dendritic spines and that its kinase activity is required for spine maturation. TAOK2 depletion leads to unstable dendritic protrusions, ultimately resulting in mislocalized synapses with aberrant accumulation of glutamate receptors in the dendritic shaft. Using chemical‐genetics and mass spectrometry, we identified several TAOK2 phosphorylation targets in the brain, including the cytoskeletal GTPase Septin7. We show in vitro and in vivo that TAOK2 directly phosphorylates Septin7 at an evolutionary conserved site on its C‐terminal tail, a modification required for its interaction with the synapse scaffolding protein PSD95 and for stable dendritic spine formation. Our study provides a mechanistic cell biological basis by which TAOK2 signaling orchestrates postsynaptic stability and compartmentalization, with implications toward understanding the potential role of TAOK2 in neurological diseases associated with the 16p11.2 region.


M164 Actomyosin contractility modulates Wnt signaling through adherens junction stability. E.T. Hall1, E. Hoesing1, E.M. Verheyen1; 1Molecular Biology and Biochemistry, Simon Fraser University, Burnaby, BC Mechanical force and cytoskeletal tension are increasingly being found as key factors regulating signal transduction pathways during development and tissue homeostasis. Previous studies have shown that mechanical force can perturb Wnt signaling in developing tissues, yet a molecular mechanism remains elusive. In an in vivo RNAi screen designed to identify novel phosphatase and kinase regulators of the Wnt/Wg pathway in Drosophila, we found that reduction of myosin phosphatase components, protein phosphatase type 1β (PP1β)[encoded by the flapwing (flw) gene] and myosin phosphatase targeting protein (Mypt)‐75D , disrupted Wnt signaling. Mypt‐75D and PP1β are known to function together to dephosphorylate and inactivate the actomyosin contractile motor protein, non‐muscle myosin II (NM II). The Wnt pathway regulates the stabilization and localization of its key effector protein β‐catenin (β‐cat)[ Drosophila Armadillo (Arm)]. β‐cat is also an integral component of the adherens junctions (AJ), mediating cell‐cell adhesion. Upon Wnt pathway activation, a multi‐protein destruction complex is disrupted via recruitment to the cell membrane allowing stabilization and translocation of β‐cat to the nucleus to promote target gene expression. Knockdown of flw and Mypt‐75D in epithelia reduced Wnt target gene expression, but did not affect Wnt ligand expression or stability. The recruitment of the destruction complex to the cell membrane under active Wnt signaling was not affected by knockdown of flw or Mypt‐75D , yet β‐cat failed to accumulate in the cytoplasm and nucleus, and was preferentially enriched at AJs. These cells also exhibited elevated NM II activation and corresponding cellular contraction. Gain or loss of function mutations of NM II resulted in similar Wnt pathway disruptions. Subsequent epistasis experiments confirmed that Wnt defects found upon reduction of myosin phosphatase activity were mediated through increased activation of NM II. Genetic interaction experiments confirmed actomyosin contractility directly regulated Wnt pathway activity. Expression of activated NM II or constitutively active formin proteins, which causes elevated filamentous (F)‐actin and accumulation of AJ proteins, including Arm/β‐cat, also led to reduced Wnt target genes. The inhibitory effect on Wnt could be rescued by reducing levels of F‐actin. Collectively from these results we have identified a mechanism in which actomyosin contractility modulates Wnt pathway activation, by affecting sequestration of Arm/β‐cat proteins to the AJ. This compensatory mechanism may be a physiological response to preferentially maintain tissue integrity over growth and patterning during development and morphogenesis.

M165 Tissue stiffness and hypoxia regulate breast cancer stem cells through ILK. M. Pang1,2, M.J. Siedlik1, S. Han2, M. Stallings‐Mann3, D.C. Radisky3, C.M. Nelson3; 1Department of Chemical and Biological Engineering, Princeton University, Princeton, NJ, 2Department of Molecular Biology, Princeton University, Princeton, NJ, 3Department of Cancer Biology, Mayo Clinic Cancer Center, Jacksonville, FL The microenvironment is a key determinant of cell phenotype during tumor progression, and breast tumors are both more stiff and hypoxic than nonmalignant breast tissue. Using an engineered culture model, we found that stiff and hypoxic microenvironments promote the development of breast cancer stem cells (CSCs) through modulation of integrin‐linked kinase (ILK). Depleting ILK blocks stiffness and hypoxia‐dependent acquisition of CSC marker expression and behavior, whereas ectopic expression of ILK stimulates CSC development under softer or normoxic conditions. Stiff microenvironments also


promote tumor formation and metastasis in ovo, and depleting ILK significantly abrogates the tumorigenic and metastatic potential of invasive breast cancer cells. We further find that the ILK‐mediated phenotypes induced by stiff and hypoxic microenvironments are regulated by PI3K/Akt. Analysis of human breast cancer patient samples reveals an association between substratum stiffness, ILK, and CSC markers, as ILK and CD44 were colocalized in cancer cells found in tumor regions predicted to be stiff. Our data suggest a role for ILK as a critical mechanotransducer that modulates breast CSC development in response to tissue mechanics and oxygen tension.

M166 A Novel Role of Sonic Hedgehog Signaling in Differentiated Human Airway Epithelia. S. Mao1,2,3, A. Shah4, L.R. Reznikov5, T. Moninger2,3, L.S. Ostedgaard2,3, M.J. Welsh1,2,3; 1Molecular physiology and biophysics, University of Iowa, Iowa City, IA, 2Internal Medicine, University of Iowa Hospitals and Clinics, Iowa City, IA, 3Howard Hughes Medical Institute , Iowa City, IA, 4University of Chicago, Chicago, IA, 5Department of Physiological Sciences, University of Florida, Iowa City, IA In vertebrates, Sonic Hedgehog (Shh) signaling is transduced in primary cilia, which are single sensory microtubule organelles present on most eukaryotic cells. Shh signaling is essential for regulating lung development and has been reported to be aberrantly activated in some lung diseases. In contrast to many cell types that have a single primary cilium, airway epithelia can contain many motile cilia. It is unknown whether Shh signaling is also transduced in motile cilia. Moreover, there is much less information about Shh in human tissue compared to mice. Therefore, we studied human differentiated airway epithelia obtained from adults to explore the expression and function of Shh signaling in adulthood. We found that ciliated airway epithelia expressed the core components of the Shh signaling pathway. The ligand Shh was present in both basal media and apical airway surface liquid. Its receptors Patched 1, co‐receptor Smoothened, regulators Gli2, Gli3 and SuFu were differentially located on motile cilia. Surprisingly, Shh in differentiated airway epithelia did not elicit the canonical Shh signaling pathway that regulates transcription during development. Instead, we identified a non‐canonical Shh signaling pathway that reduced intracellular cAMP levels. As a result, Shh signaling decreased cAMP‐dependent ciliary beating, inhibited cAMP‐activated CFTR current activity and reduced airway surface liquid pH. Because these changes have been observed in the airways of people with chronic obstructive pulmonary disease (COPD), we studied differentiated cultures of human airway epithelia obtained from people with COPD. We observed that Shh ligand expression was significantly increased in COPD airway epithelia. Our data identify Shh signaling components expressed in differentiated ciliated airway epithelia, and uncover a novel function of Shh signaling in regulating cAMP‐dependent physiology. They also suggested a potential role of Shh signaling in the pathogenesis of airway disease.

M167 The sperm‐derived TRP family channel TRP‐3 induces a calcium rise in the fertilized oocyte in C. elegans. J. Takayama1, H. Okada1, S. Onami1; 1RIKEN QBiC, Kobe, Japan Calcium waves during fertilization trigger animal embryogenesis. How the fertilizing sperm induces the calcium waves has been explained either by the action of a soluble sperm‐derived factor introduced upon fusion or the surface interaction between the sperm and egg before fusion. The involvement of a sperm plasma membrane calcium channel, which is assumed by the sperm conduit model, has not been demonstrated because of the lack of molecular evidences. Here we show that the sperm calcium‐


permeable channel TRP‐3 induces a calcium rise in the oocyte upon fusion in C. elegans. To visualize the calcium response during in vivo fertilization, we established a high‐speed confocal imaging system and image processing methods, and found that the calcium response consists of a rapid local rise that occurs immediately after sperm entry near the sperm entry point followed by a traveling wave that propagates throughout the oocyte. Mutant analysis demonstrated that the local calcium rise is absent and the onset of the traveling wave is delayed in the trp‐3 mutant. Transgenic rescue experiments for the trp‐3 mutant using several sperm‐specific drivers demonstrated that the more TRP‐3 channel in the sperm, the larger the local calcium rise and the earlier the traveling wave is generated. This relationship between the local calcium rise and the onset of the traveling wave was reproduced by a numerical simulation, which assumes that the oocyte has the calcium‐induced calcium release (CICR) machinery. High‐speed calcium imaging and simulation suggest that TRP‐3 acts as a calcium conduit that allows calcium entry after sperm‐oocyte fusion. Moreover, visualization of the localization dynamics of TRP‐3::TagRFP‐T during fertilization demonstrated that fertilization in C. elegans takes place as a direct plasma membrane fusion and that TRP‐3 is transferred from the sperm plasma membrane to the oocyte plasma membrane upon fusion. These results suggest that the sperm‐derived TRP‐3 channel induces the local calcium rise probably by allowing calcium influx after sperm‐oocyte fusion, which then triggers the traveling wave by the CICR machinery, consistent with the sperm conduit model. Currently we are investigating which calcium channel is responsible for the propagation of the traveling wave. Two independent germline‐specific knockdown of the sole IP3 receptor ortholog itr‐1 did not show any detectable changes in the calcium response patterns. Therefore, we are conducting an RNAi screen targeted to all cation channels expressed in the germline with improved calcium indicator and image processing methods.

M168 Uncovering novel substrates and functions for the calcineurin phosphatase in human cells. C.P. Wigington1, N.P. Damle1, J. Roy1, S.E. Cho1, N.E. Davey2, Y. Ivarsson3, M.S. Cyert1; 1Biology, Stanford University, Stanford, CA, 2Conway Institute of Biomolecular and Biomedical Sciences, University College, Dublin, United Kingdom, 3Chemistry‐BMC, Uppsala University, Uppsala, Sweden Protein phosphatases play essential roles in every signaling pathway; however, systems‐level understanding of phosphatase signaling networks is lacking due to the inherent challenges associated with proteome‐wide identification of their substrates. Calcineurin (CN) is the conserved Ca2+/calmodulin‐activated protein phosphatase and target of the widely prescribed immunosuppressant drugs, FK506 and Cyclosporin A. CN is ubiquitously expressed and plays critical roles in the immune, nervous, skeletal and cardiovascular systems, as well as during development. However, only 50 substrates are currently attributed to this phosphatase. CN utilizes conserved docking surfaces to interact with substrates via Short Linear Motifs (SLiMs) termed PxIxIT and LxVP, which occur preferentially in intrinsically disordered domains and are challenging to identify due to sequence degeneracy and low affinity for calcineurin. We are applying novel experimental and computational approaches to systematically identify CN‐interacting SLiMs within the human proteome with the goal of ultimately establishing the human CN signaling network. We directly identified novel CN‐binding sequences by performing unbiased peptide phage display selections with human CN using a library containing all predicted disordered regions in the human proteome. These SLiM sequences directly identified novel candidate CN substrates, including the nucleoporin, NUP153, indicating a previously uncharacterized role for CN in regulating nuclear pore structure and/or function. We have shown that NUP153 phosphorylated by ERK is directly dephosphorylated by CN in vitro. Additionally, NUP153 contains a conserved PxIxIT sequence that is required for its interaction with CN in vivo. Furthermore, a NUP153 mutant lacking this SLiM is dephosphorylated less efficiently by CN in vitro and shows altered dephosphorylation in vivo. Finally, all novel CN‐binding SLiMs were used to computationally predict CN‐


interacting proteins in the human proteome using both structure‐ and sequence‐based methods. These analyses identified several additional nuclear pore components as possible substrates of CN, and we are currently expanding our investigations of CN‐dependent regulation of these structures. In addition to nuclear pore proteins, combining experimental and computational identification of CN‐binding SLiMs has uncovered a host of new candidate substrates for CN including ion channels, kinases, transcription factors and receptors. Together, these studies suggest new points of cross‐talk between CN and other signaling pathways in human cells and will ultimately allow us to establish the first comprehensive signaling network for this critical Ca2+‐dependent regulator of human health.

M169 SPV‐1, a RhoGAP with a novel F‐BAR domain, regulates calcium signaling in the C. elegans spermatheca during ovulation events. J. Bouffard1, A.R. Asthagiri1,2, R. Zaidel‐Bar3,4, E.J. Cram5; 1Bioengineering, Northeastern University, Boston, MA, 2Chemical Engineering, Northeastern University, Boston, MA, 3Biomedical Engineering, National University of Singapore, Singapore, Singapore, 4Mechanobiology Institute, National University of Singapore, Singapore, Singapore, 5Biology, Northeastern University, Boston, MA Biological tubes are found in organisms across the animal kingdom, and proper regulation of contractility in these tubes is vital for many life processes. The spermatheca, a tissue in the reproductive system of the nematode C. elegans, provides an ideal model for detailed studies of such tubes. The spermatheca is composed of 24 smooth muscle‐like cells that form a contractile tube and 4 cells that form a syncytial valve. During ovulation events the spermatheca is stretched significantly by oocyte entry, remains distended for a regulated period of time while the eggshell forms, and finally initiates coordinated contraction of the cells, opening the valve and expelling the egg. Having highly conserved genes and parallels with many smooth muscle systems, mechanisms at work in the spermatheca potentially represent fundamental biological responses to cell and tissue level deformations in biological tubes. SPV‐1, a RhoGAP with a novel F‐BAR domain, is a key regulator of contractility in the spermatheca. When spv‐1 is lost, the spermatheca contracts immediately upon oocyte entry, leading to faster egg transit and perturbed egg shapes. Calcium signaling drives contraction in the spermatheca, so we used the genetically encoded calcium sensor GCaMP to determine if the loss of spv‐1 alters calcium signaling. We monitored calcium activity in the spermatheca during ovulation by acquiring movies of live, intact animals on a widefield fluorescence microscope. To analyze our imaging data we developed semi‐automated image processing and signal analysis pipelines in Fiji and Matlab. These pipelines extract metrics from the GCaMP datasets, enabling quantitative observations and comparisons across multiple ovulation events in different worms. We find that the loss of spv‐1 results in a more rapid onset of calcium signaling upon oocyte entry, and elevated calcium levels throughout ovulation events. Although the maximum calcium signal does not vary significantly between wildtype and spv‐1 mutants, mutants lack proper spatiotemporal regulation of calcium signaling. The loss of the worm Filamin fln‐1 is known to alter calcium signaling, so we examined if the additional loss of spv‐1 further alters calcium signaling. Our analyses show that the fln‐1 spv‐1 double mutants also exhibit a more rapid onset of calcium signaling, elevated calcium levels throughout ovulation events, and spatiotemporal misregulation of calcium signaling compared to fln‐1 single mutants. These results show that FLN‐1 and SPV‐1 act in parallel to modulate calcium signaling in the spermatheca. Ongoing work aims to identify effectors that SPV‐1 uses to regulate calcium signaling, and to understand how the individual spermathecal cells are communicating to coordinate tissue level responses.


M170 Coordination of directional outgrowth and patterning by Wnt5a and Fgf signaling interaction. Y. Yang1, B. Gao2; 1Developmental Biology, Harvard School of Dental Medicine, Boston, MA, 2Biochemistry, University of Hong Kong, Hong Kong, Hong Kong Formation of tissue and organs with complex morphology that exhibits three‐dimensional asymmetry requires proper regulation of both patterning and directional growth. However, it is unclear how these two events are precisely coordinated. The Wnt/Planar Cell Polarity (PCP) signaling is emerging as a critical evolutionarily conserved mechanism whereby directional information is conveyed. The uniform directionality of PCP is thought to be established by global cues that break the original symmetrical localization of core PCP proteins such as Vangl2. However, the molecule nature and functional mechanisms of the global cues remain largely unknown, particularly in vertebrates. Here taking the mammalian developing limbs as a model system, we found that the well‐known AER‐derived limb patterning signaling molecules Fgfs are required to regulate PCP in a Wnt5a‐dependent manner. In addition to its permissive role to allow Fgf signaling to orient PCP, we further demonstrate with definitive genetic evidence that Wnt5a gradient acts as a global cue that is instructive in establishing PCP. By genetically altering the direction of Wnt5a gradient, we showed that the direction of Wnt5a gradient determines the direction of Vangl2 asymmetrical localization, a molecular readout of PCP. Therefore, our results indicate that directional growth and patterning of developing limbs are coordinately regulated by Wnt5a and Fgf signaling, which maybe a fundamental mechanism that ensures proper morphogenesis of the limb and other organs.

Minisymposium 17: Genome Replication, and Gene Regulation

M171 Spatial organization of topologically associated domains in individual chromosomes. S. Wang1,2, J. Su1,3, B.J. Beliveau4, B. Bintu1,5, J.R. Moffitt1,2, C. Wu4, X. Zhuang1,2,5; 1Howard Hughes Medical Institute, Cambridge, MA, 2Department of Chemistry and Chemical Biology, Harvard University, Cambridge, MA, 3Department of Molecular and Cellular Biology, Harvard University, Cambridge, MA, 4Department of Genetics, Harvard Medical School, Cambridge, MA, 5Department of Physics, Harvard University, Cambridge, MA The spatial organization of chromatin critically impacts genome function. Recent high‐throughput chromosome‐conformation‐capture studies have revealed topologically associated domains (TADs) as a conserved feature of chromatin organization. How TADs are spatially organized in individual chromosomes remains, however, largely unknown. Here we developed an imaging method to directly map the spatial positions of numerous genomic regions along individual chromosomes inside the nucleus and used this method to trace the positions of TADs in four human interphase chromosomes. We observed that chromosome folding deviated from the ideal fractal‐globule model at large length scales. Our single‐chromosome images revealed that TADs were spatially organized into two physical compartments that were positioned in a highly polarized fashion in individual chromosomes. Moreover, we observed that active and inactive X chromosomes differed from each other and from autosomes in terms of folding and compartmentalization, suggesting that the spatial organization of TADs can be changed in response to regulation.


M172 ORC is not essential for DNA replication in human cell lines, but important for repressing Rb‐E2F and Polycomb group regulated genes. E. Shibata1, M. Kiran1, Y. Shibata1, S. Singh1, S. Kiran1, A. Dutta1; 1Biochemistry and Molecular Genetics, University of Virginia, Charlottesville, VA Genomic stability requires the orderly replication of our chromosomes once in each cell‐cycle. Chromosomal replication initiates from plastic origins of replication spaced about once every hundred kilobases. The six‐subunit Origin Recognition Complex (ORC) is believed to be an essential initiator protein that binds to origins of replication and initiates DNA replication in eukaryotes, similar to DnaA in bacteria. In a model of replication initiation believed to be universal in eukaryotes, ORC helps recruit CDC6 and CDT1 proteins, which in their turn recruit MCM2‐7 proteins to license the origins for replication initiation. We have discovered that human cell lines in culture proliferate with intact chromosomal origins of replication after CRISPR‐Cas9 mediated genetic disruption of both alleles of ORC2 or ORC1. Although the six‐subunit ORC is believed to associate with the origins as a complex, chromosomal association of ORC6 is relatively unaffected by loss of the critical subunits of ORC. Loading of CDC6 and CDT1 on chromatin is unaffected, while the loading MCM2‐7 is significantly decreased, but not absent. Surprisingly, cell proliferation rates, duration of S phase, spacing of origins of replication and fork elongation rates are relatively unaffected despite the knockout of ORC2 or ORC1. There is no evidence of excessive DNA damage or checkpoint activation as would be expected if DNA replication was compromised. In contrast, Retinoblastoma/E2F‐ or polycomb‐regulated genes are specifically de‐repressed, suggesting that ORC is critical for gene repression by these key critical transcription regulators. An shRNA screen of cellular genes that become essential in the absence of ORC identifies CDC6 as becoming more important for proliferation of the ORC knockout cells. The ORC‐depleted cells become critically dependent on CDC6 for survival and DNA replication, suggesting that factors downstream of ORC are sufficient to initiate DNA replication in the absence of the six‐subunit ORC. Thus, although known universally as an essential replication initiation factor critical for genomic stability, ORC is dispensable for replication initiation and genomic stability in established human cell lines while being critical for transcription regulation. The mechanism by which ORC helps repress E2F/Rb‐ and polycomb‐regulated genes is being investigated.

M173 The mechanism of DNA damage and chromothripsis from chromosome segregation errors. A. Spektor1,2, C. Zhang3,4, N. Umbreit 2,5, M. Feldman2,5, A. Ahsan2,5, Y. Zhang6, D.S. Pellman2,5; 1Radiation Oncology, Dana‐Farber Cancer Institute, Boston, MA, 2Cell Biology, Harvard Medical School , Boston, MA, 3Medical Oncology, Dana‐Farber Cancer Institute, Boston, MA, 4Biomedical Informatics, Harvard Medical School, Boston, MA, 5Pediatric Oncology, Dana‐Farber Cancer Institute, Boston, MA, 6Single Cell Sequencing Program, Dana‐Farber Cancer Institute, Boston, MA Tumorigenesis and resistance to therapy are generally believed to arise through a classical Darwinian process, with gradual, multigenerational accrual of mutations. However, recent cancer genome sequencing suggests that many cancers may accumulate a large number of mutations rapidly, perhaps during the course of a single cell cycle. The most dramatic example of such “punctuated equilibrium‐type” rapid genome evolution is chromothripsis, a novel mutational process characterized by massive chromosome rearrangements and a unique pattern of DNA copy number variations, all curiously restricted to one or a few chromosomes. The mechanism(s) leading to chromothripsis have been


unclear, but our group previously proposed that it could result through physical isolation of chromosomes in abnormal nuclear structures called micronuclei. Micronuclei are deficient in key nuclear functions such as nuclear envelope stability, DNA replication, and transcription, and because of these abnormalities develop DNA damage upon entry into S‐phase. By combining live‐cell imaging with single‐cell whole genome sequencing, we recently showed that micronucleation can lead to a large number of rearrangements on the missegregated chromosome within one cell cycle, in some cases recapitulating all the hallmarks of chromothripsis. Here we describe our latest efforts to elucidate the mechanism of chromothripsis in detail, by examining the timing and cell cycle dependence of events that lead to DNA breakage and chromosome reassembly. Our preliminary results indicate that chromosome rearrangements and chromothripsis can occur even in the absence of nuclear envelope rupture during interphase, but may require reincorporation of the micronucleated chromosome into one of the daughter cell nuclei. We are also using inhibitor compounds and gene depletions/knockouts to identify the genetic requirements for chromothripsis. Together, these experiments advance our understanding of the ways by which chromosome segregation errors shape cancer genomes.

M174 Non‐coding Heterochromatic RNAs Promote Genomic Instability and Tumor Formation in vivo. Q. Zhu1, N. Hoogn1, A.S. Aslanian2,3, T. Hara1, K.H. Miga4, C. Benner5, S. Heinz5, J. Yates iii3, T. Hunter2, I. Verma1; 1Laboratory of Genetics, The Salk Institute for Biological Studies, La Jolla, CA, 2Molecular Biology Laboratory, The Salk Institute for Biological Studies, La Jolla, CA, 3Department of Chemical Physiology, The Scripps Research Institute, La Jolla, CA, 4Department of Bioengineering, University of California, Santa Cruz, Santa Cruz, CA, 5Department of Cellular and Molecular Medicine, University of California, San Diego, La Jolla, CA Non‐coding RNAs are emerging as an important player in a variety of regulatory processes in human health. Heterochromatic satellite RNAs are a heterogeneous population of non‐coding RNA (20 to 1,500 nucleotides in length) believed to be expressed from large swaths of repetitive sequences at the centromere and telomere of a variety of eukaryotic chromosomes. It has been shown that these satellite RNAs are massively transcribed into non‐coding RNAs in a variety of human cancers, including BRCA1‐mutant breast cancer. In certain cases, this class of non‐coding satellite RNAs represent up to ~50% of the total cellular RNAs. However, the consequences of the aberrant overexpression have not been studied. In fission yeast, the expression of satellite RNAs is vital for the establishment and maintenance of the pericentromeric chromatin. In mammalians, these satellite RNAs are rarely transcribed in adult organisms under normal homeostatic conditions. We have found that the overexpression of satellite RNA in cultured mammalian cells causes DNA damage response, cell cycle checkpoint activation, and chromosome segregation defects. Here, we provide the first evidence that overexpression of satellite RNAs have oncogenic potential in vivo. We have employed the CRISPR technology to activate the endogenous satellite RNAs and find that overexpression of endogenous satellite RNAs causes DNA damage response and chromosome breakages and redials, which is reminiscence of the signature chromosomal abnormality in BRCA1‐deficient Fanconi Anemia tumor cells. Using a lentivirus gene delivery system we show that satellite RNA overexpression induces tumors in mice. Moreover, we generated transgenic mice that ubiquitously overexpress satellite RNA. These mice exhibit developmental defects, including high percentage of runt pups and hydrocephaly, as well as high incidence of tumor formation. Strikingly, we observe accelerated tumorigenesis upon inactivation of tumor suppressor p53. These tumors again show severe chromosome breaks and radial formation. In conclusion, our data from cultured cells and engineered mouse models demonstrate that overexpression of non‐coding satellite RNAs found in large number of human tumors adversely affects


the genomic stability and promotes tumorigenesis. These studies lay the foundation for developing novel therapeutic strategies that block the effects of non‐coding satellite RNAs in cancer cells.

M175 Expression of a non‐canonical mRNA facilitates meiotic kinetochore remodeling. J. Chen1, A. Tresenrider1, M. Chia2, F. van Werven2, E. Ünal1; 1Molecular and Cell Biology, University of California, Berkeley, Berkeley, CA, 2The Francis Crick Institute, London, United Kingdom A hallmark of sexual reproduction is meiosis, a dynamic developmental program that generates haploid gametes from diploid progenitors. To achieve ploidy reduction, cells have evolved meiotic‐specific modifications to gene expression, chromosome structures and dynamics, as well as chromosome segregation machinery such as kinetochores. Timely coordination of all these remodeling events is crucial to the success of meiosis. In budding yeast, non‐canonical, dynamic transcriptional and translational controls are widespread in meiosis, and have been proposed as a key way to impose temporal regulation on meiotic‐specific remodeling events. Here we provide a molecular mechanism for how expression of a non‐canonical mRNA facilitates remodeling of the kinetochores. Previously, we have found that a single kinetochore subunit, Ndc80, is downregulated in meiotic prophase I to ensure meiosis I‐specific chromosome segregation. We now demonstrate that this downregulation occurs in part due to regulated expression of two NDC80 transcript isoforms in meiosis. We find that vegetative cells express a short isoform of NDC80 (NDC80 ORF), but meiotic cells express a long uORF translating transcript isoform (NDC80 luti) after meiotic entry. Interestingly, NDC80 luti is coding‐incompetent due to translation of the uORFs in its 5’ leader sequence, and is both necessary and sufficient to downregulate Ndc80 protein. It also inhibits expression of NDC80 ORF through transcriptional interference and chromatin modifications. Collectively, expression of NDC80 luti represses Ndc80 protein synthesis in meiotic prophase I. As cells exit from meiotic prophase I, the coding‐competent isoform NDC80 ORF is upregulated, and Ndc80 is rapidly resynthesized for timely kinetochore reassembly and chromosome segregation. Lastly, expression of NDC80 luti and NDC80 ORF requires the master transcription factors Ime1 and Ndt80, respectively, which likely control expression of other luti‐RNAs that are uncovered from a genome‐wide study. Altogether, our study provides mechanistic insight into how meiotic‐specific gene expression is temporally coupled to alterations on kinetochores composition, and potentially to other cellular remodeling events.

M176 Brr6 Restricts ncRNA‐Mediated Transcriptional Repression at the Nuclear Rim. A. de Bruyn Kops1, C. Guthrie1; 1Biochemistry, UCSF, San Francisco, CA Correlation between transcriptional regulation and positioning of genes at the nuclear envelope is well established in eukaryotes and a large body of information links nuclear envelope components with transcriptional activation and repression. However, the mechanisms by which the envelope contributes to gene regulation are not fully understood. It is now appreciated that eukaryotic transcription is pervasive, generating numerous ncRNAs that can impact transcription of coding genes by design or as a result of deficient surveillance. We show that Brr6, an essential envelope transmembrane protein required for mRNA export and envelope lipid homeostasis in budding yeast, plays a role in ncRNA surveillance at the nuclear envelope. We find that BRR6 is required for rim positioning and efficient transcription of PAB1 and GAL7 loci and


that the brr6‐1 mRNA export defect likely stems from decreased PAB1 expression. Interestingly, the GAL7 transcription block is specific for the 3’ end of the ORF and we detect an aberrant ncRNA that runs into the GAL7 3’ UTR in brr6‐1. Similarly, a ncRNA upstream of PAB1 extends into the adjacent CHD1 3’ UTR and CHD1 3’ ORF transcript levels are also reduced in brr6‐1. Importantly, conditions that restrict this ncRNA restore CHD1 as well as PAB1 expression, suggesting that the brr6‐1 transcription defects stem from failed termination of ncRNAs. In keeping with this, brr6‐1 is synthetically lethal with sen1‐1, a component of the NNS ncRNA termination complex. Consistent with established connections between ncRNAs and chromatin deacetylation in yeast, histone acetylation in the GAL7 ORF decreases in brr6‐1 and deleting the HDA1 deacetylase reverses this effect and partially restores transcription. These results are consistent with a model in which Brr6 functions in transcriptional regulation by restricting ncRNA‐mediated repression of coding genes at the nuclear envelope.

M177 Evolution of promoter directionality. S. Churchman1; 1Department of Genetics, Harvard Medical School, Boston, MA Many eukaryotic promoters drive transcription in two directions. At the promoters of coding genes, transcription leads to stable mRNA in the sense direction and to unstable non‐coding RNA in the antisense direction. It remains unknown whether the unstable transcript produced by divergent transcription serves a function or simply constitutes transcriptional noise. In this study, we investigated whether promoters are shaped by evolutionary pressures to produce transcription in both directions or instead inherently yield bidirectional transcription. Analysis of Saccharomyces cerevisiae nascent transcripts revealed that promoters of young genes tend to promote bidirectional transcription. Furthermore, promoters placed in foreign trans environments lost directionality, indicating that cis‐elements and trans‐acting factors co‐evolve to promote directional transcription. Unexpectedly, we found that promoters emerge frequently in DNA placed in a foreign environment, and these fortuitous new promoters produced equal transcription in both directions. Thus, without evolutionary pressure, promoters are intrinsically bidirectional. Finally, we found that similar promoters arose naturally in evolution and were either purged through mutation or retained to foster gene re‐organization. Our data reveal that promoters are born bidirectional and are subsequently shaped by evolution to bias transcription of coding transcripts while suppressing non‐coding transcriptional noise.

M178 Cryo‐EM visualization of promoter binding by the human general transcription factor TFIID. R.K. Louder1, A. Patel1, Y. He2,3, J. Fang4, E. Nogales2,4,5,6; 1Biophysics Graduate Group, University of California, Berkeley, Berkeley, CA, 2Molecular Biophysics and Integrative Bioimaging Division, Lawrence Berkeley National Laboratory, Berkeley, CA, 3Northwestern University, Department of Molecular Biosciences, Evanston, IL, 4Howard Hughes Medical Institute, University of California, Berkeley, Berkeley, CA, 5QB3 Institute, University of California, Berkeley, Berkeley, CA, 6Department of Molecular and Cell Biology, University of California, Berkeley, Berkeley, CA In eukaryotes, the general transcription factor IID (TFIID) plays a central role in the initiation of RNA polymerase II (Pol II)‐dependent transcription by nucleating pre‐initiation complex (PIC) assembly at the core promoter. TFIID is a large multi‐subunit complex comprising the TATA‐binding protein (TBP) and 13 TBP‐associated factors (TAFs), which specifically interact with a variety of core promoter DNA sequences, general transcription factors, transcriptional activators, and epigenetic marks. Despite its


importance in coordinating transcription initiation, a detailed mechanistic understanding of TFIID’s activities remains elusive. The high degree of conformational plasticity within the TFIID architecture has largely limited structural analysis to low‐resolution electron microscopy (EM) studies. However, by improving the biochemical preparation of TFIID complexes and by exploiting the latest technological advances within the field of 3D cryo‐EM, we have recently been able to visualize promoter binding by human TFIID in close‐to‐atomic‐level detail. Our high‐resolution structures have revealed the full architecture of the promoter‐binding core of TFIID and show which components of TFIID are involved in direct recognition of conserved core promoter sequences. By solving cryo‐EM structures of TFIID in multiple conformations in both promoter‐bound and unbound states, we are able to explain with unprecedented detail the molecular mechanisms by which TFIID fulfills its primary function of properly positioning TBP on the core promoter, which ultimately determines the placement of Pol II relative to the transcription start site. By superimposing the structure of promoter‐bound TFIID with the structures of other transcriptional assemblies, we have gained further insights into the role of TFIID in PIC assembly, such as how TFIID provides protein‐protein contacts for the recruitment of downstream PIC components. We also propose a major role for TFIID in providing additional levels of control during transcription initiation. For instance, the large‐scale flexibility of TFIID is likely to be an important property that allows sequential conformational states that provide regulatory checkpoints throughout the process of PIC assembly and transcription initiation. Ultimately, our structures provide a framework for a molecular understanding of the complex mechanism underlying TFIID function, shedding new light into the overlapping roles of TFIID as both a coactivator and a general platform for PIC assembly in the coordination of transcription initiation.

M179 A polyglutamine domain enables transcriptional reprogramming in response to a transient pH change. L.J. Holt1, I. Gutiérrez2; 1Institute for Systems Genetics, New York University Langone Medical Center, New York, NY, 2Molecular and Cell Biology, University of California, Berkeley, Berkeley, CA Polyglutamines (polyQs) are known to form aggregates in pathogenic contexts, such as Huntington disease. However, little is known about their normal biological role. Saccharomyces cerevisiae has around 100 polyQ containing proteins, many of which are DNA or RNA binding proteins. The SWI/SNF chromatin‐remodeling complex is highly enriched in polyQs, including one of the largest polyQ domains in the yeast genome within the SNF5 subunit. We found that the SNF5‐polyQ is required for efficient transcription of some glucose‐repressed genes including ADH2. It has been reported that carbon starvation leads a decrease in cytosolic pH from 7.4 to ~6 and that histidines, which contain an imidazole ring with a pKa of around 6.5, are enriched within Q‐rich sequences. Therefore, we hypothesized that pH changes and these histidine residues would contribute to the regulation of glucose‐repressed genes. Consistent with this idea, prevention of cytosolic acidification or mutation of SNF5 histidines to alanine prevented expression of ADH2. To better relate pH changes to gene expression, we developed single‐cell reporters for cytosolic pH and ADH2 expression. Initially, all cells acidify to pH ~6.5. Subsequently, the population bifurcates: Around 60% of cells further acidify to pH ~6, while a second subpopulation recovers to pH ~7. Only the latter subpopulation induces ADH2. Therefore, a transient pH decrease to below 6.5 followed by recovery to pH 7 is required for expression of ADH2. PolyQ domains are known to assemble into higher order structures. Therefore, we hypothesized that a pH‐dependent oligomerization could be important for ADH2 expression. In support of this idea, we found that a synthetic spidroin domain from spider silk, which is soluble at pH 7 but oligomerizes at pH 6, partially complemented the


function of the polyQ domain. During glucose starvation, many SWI/SNF‐dependent genes are inactivated and a new set of genes is induced. We speculate that the SNF5 polyQ domain acts as a “reset” switch to allow this redistribution of the SWI/SNF complex upon starvation. These results provide further evidence of an important biological role for polyQ domains beyond the disease context.

M180 Regulation of estrogen‐responsive genes in single human cells. J. Rodriguez1, C.C. Chow2, D.R. Larson1; 1Laboratory of Receptor Biology Gene Expression, National Cancer Institute, Bethesda, MD, 2Laboratory of Biological Modeling, National Institute of Diabetes and Digestive and Kidney Diseases, Bethesda, MD One of the distinguishing features of the human genome is the existence of long‐range regulation by distal regulatory elements such as enhancers. This regulation is facilitated by the non‐random conformation of chromosomes on multiple length scales. However, these chromosome contacts may only exist in a fraction of cells, suggesting they are also dynamically regulated. In addition to this spatial compartmentalization, transcription also displays temporal compartmentalization: genes are transcribed at irregular intervals in episodic ‘bursts’ of RNA synthesis. The relationship between these two phenomena and the regulation of genes in response to upstream signals is largely unknown. Here, we describe the use of single‐molecule live‐cell imaging to dissect the regulation of a human estrogen‐responsive gene under endogenous regulation. The estrogen response is a robust and rapid transcriptional paradigm, inducing hundreds of target genes within an hour of estradiol treatment. The estrogen receptor and its co‐factors are often bound both proximally and distally within enhancers. One of the first identified and highly induced estrogen responsive genes is TFF1, which also exhibits complex dynamic regulation such as oscillatory behavior. Using CRISPR/Cas9 we successfully integrated 24 MS2 stem loops into multiple TFF1 alleles of the same cell at the endogenous location, allowing us to observe the synthesis of nascent RNA in real time. Using this direct observation of transcription dynamics, we find behavior which is inconsistent with the prevailing stochastic models of transcription. Although this gene is highly induced, we observe very short active periods (~ minutes) and periods of inactivity which range from minutes to many hours. Despite these intrinsic dynamics, alleles are not independent but rather show correlated bursting in the same nucleus. To determine the impact of the enhancer on transcriptional dynamics we deleted a portion of the enhancer and observed effects on both the duration of the active states and the frequency of observing these active states. Finally, we derive a mathematical model of transcription which yields mechanistic insight into how cells are able to ‘sense’ changes in estrogen.

Minisymposium 18: Quality Control and Organelle Trafficking

M181 Oxidized sterols mediate Vms1 stress‐responsive translocation to mitochondria. J. Rutter1, J. Nielson1; 1Department of Biochemistry, University of Utah/HHMI, Salt Lake City, UT Mitochondrial dysfunction contributes to a variety of human diseases. Cells have developed a mutli‐layered system of mitochondrial quality control. One important component in mitochondrial protein degradation is VCP/Cdc48‐associated Mitochondrial Stress‐responsive (Vms1), which together with the AAA‐ATPase Cdc48, translocates to damaged mitochondria under stress and aids in the proteasome‐dependent degradation of mitochondrial proteins. Mitochondrial targeting of Vms1 is mediated by a


conserved Mitochondrial Targeting Domain (MTD), which is inhibited by an intermolecular interaction with an N‐terminal Leucine Rich Sequence (LRS) under unstressed conditions. We have determined the 2.7Å crystal structure of Vms11‐417, which includes the MTD/LRS interaction. We have also identified a mitochondrial, oxidized sterol that is necessary and sufficient for Vms1 localization to mitochondria. Furthermore, we provide evidence that the oxidized sterol receptor and LRS compete for interaction with the MTD. These data support a model in which mitochondrial stress induces the production of the oxidized sterol receptor, thereby promoting Vms1 recruitment to mitochondria.

M182 Analysis of a novel phosophodegron‐dependent selectivity mechanism in stationary phase mitophagy. H. Abeliovich1, P. Kolitsida1, M. Rackiewicz2, J. Dengjel2; 1Biochemistry, Food Science and Nutrition, Hebrew University of Jerusalem, Rehovot, Israel, 2Biology, University of Fribourg, Fribourg, Switzerland Mitophagy, or the autophagic degradation of mitochondria, is an important housekeeping function of eukaryotic cells that prevents the accumulation of defective mitochondria due to oxidative damage and spontaneous mutations. The culling of defective mitochondria has been suggested to delay the onset of aging symptoms, and defects in mitophagy have been linked to late onset hereditary disorders such as Parkinson’s disease and type II diabetes. We previously showed that different mitochondrial matrix proteins undergo mitophagy at different rates. An attractive model, supported by existing data, is that fission and fusion of mitochondria are linked to the segregation of defective mitochondrial compartments from undamaged ones, in a ‘distillation’ mechanism that leads to the selective turnover of defective compartments. We now demonstrate that dynamic mitochondrial matrix protein phosphorylation and dephosphorylation generate a segregation principle that would couple with mitochondrial fission and fusion to generate such a distillation process. Our data support a model wherein differences in protein‐protein interactions between differentially phosphorylated proteins of the same species can drive a microscopic phase separation which, coupled with fusion‐fission dynamics could account for the observed selectivity.

M183 Energy Need‐Dependent PINK1 Phosphorylates Mitofilin to Maintain Mitochondrial Crista Junctions in a Pathway Linked to Parkinson’s Disease. P. Tsai1, C. Lin2, C. Schoor3, J. COUTHOUIS4, R. Wu2, O. Ross5, A. Gitler4, D. Winter3, X. Wang1; 1Neurosurgery, Stanford University School of Medicine, Palo alto, CA, 2Neurology, National Taiwan University Hospital, Taipei, Taiwan, 3Biochemistry and Molecular Biology, University of Bonn, Bonn, Germany, 4Genetics, Stanford University School of Medicine, Palo alto, CA, 5Neuroscience, Mayo Clinic, Jacksonville, FL Mitochondrial crista structure is dynamically remodeled to support the changes of respiratory rate. Heightened cellular energy requirements trigger the increase of crista junctions and membrane areas leading to mitochondrial crista condensation. However, the molecular mechanism governing energy need‐dependent crista condensation remains elusive. Here we show that in Drosophila upon elevation of cellular energy demands, the Parkinson’s disease (PD)‐linked Ser/Thr kinase PINK1 phosphorylates the inner mitochondrial membrane protein Mitofilin (Mic60), to maintain the growing number of crista junctions and consequently to enhance ATP production. Upregulation of Mitofilin in PINK1 null flies restores their crista structure and ATP levels, and remarkably rescues their behavioral defects and


dopaminergic neurodegeneration. Furthermore, we discover that mitofilin (IMMT) mutations increase risk for PD in humans, and expression of these disease‐linked human mitofilin mutations in Drosophila impairs crista junction formation and causes locomotion deficits. These results illuminate a novel PD pathway linked to maintenance and plasticity of crista junctions in vivo.

M184 Selective destruction of mitochondrial membrane proteins through the Mitochondrial‐Derived Compartment Pathway. A.M. Litwiller1, C.E. Hughes1, T.J. Campbell1, A.L. Hughes1; 1Department of Biochemistry, University of Utah School of Medicine, Salt Lake City, UT Mitochondrial dysfunction is a hallmark of aging, and underlies the development of many diseases. Cells maintain mitochondrial health through a number of quality control systems that can detect dysfunctional mitochondria and repair or eliminate problematic organelles. Using yeast as a model system, we recently discovered a new mitochondrial quality control system, the mitochondrial‐derived compartment (MDC) pathway, which eliminates proteins from dysfunctional mitochondria by autophagy. Unlike many common mitochondrial autophagy pathways, this system selectively removes a subset of membrane proteins from the mitochondrial inner and outer membranes, while leaving the remainder of the organelle intact. Selective removal of preexisting proteins is achieved by sorting of proteins into a mitochondrial‐derived compartment, or MDC. MDCs form at sites of ER‐mitochondrial contact, and after formation, are subsequently released from mitochondria by the canonical mitochondrial fission machinery and eliminated by autophagy. We have screened the mitochondrial proteome to catalog the substrates of this pathway, and found that the primary targets of this system are proteins that rely on the mitochondrial surface receptor Tom70 for import into mitochondria. The largest class of these substrates is the mitochondrial nutrient carrier proteins, which promote nutrient exchange across the mitochondrial inner membrane. Based on this observation, combined with our preliminary results that MDC formation is triggered by metabolic stress, we propose that the MDC pathway acts as a mechanism to rapidly modulate mitochondrial nutrient transporters in response to fluctuations in intracellular nutrients.

M185 Reticulons Play Key Roles in the Endoplasmic Reticulum (ER) Stress Surveillance Pathway in a Sphingolipid‐dependent Manner. M. Niwa1, F. Pina‐Nunez1, J.T. Chao1, F. Yagisawa1; 1Division of Biological Sciences, UCSD, La Jolla, CA The Endoplasmic reticulum (ER) is a gateway for production of secretory and membrane proteins and is a major site for both lipid synthesis. The ER cannot be synthesized de novo but arises only from existing ER; therefore, each daughter cell must inherit the ER from the mother cell at every division. While a large body of research has uncovered various cell cycle checkpoints and regulatory mechanisms that ensure proper segregation of nuclear genomic material, much less is known about whether similar types of regulatory mechanisms or checkpoints operate for segregation of cytoplasmic components, including organelles like the ER. Previously, we reported an ER stress surveillance mechanism (ERSU) in S. cerevisiae that ensures the inheritance of a functional ER by the daughter cell. When the demand on ER function is high (known as ‘ER stress’), transmission of the ER into the daughter cell is blocked, and concomitantly, the septin complex at the bud neck is re‐localized, preventing the cell from proceeding to cytokinesis. When the


functional capacity of the ER is recovered, the cells re‐enter the cell cycle. These events are independent of the well‐known Unfolded Protein Response (UPR), but rather by the MAP kinase Slt2. Cells deficient in slt2 (slt2Δ) fail to delay inheritance of functionally stressed ER and to halt cytokinesis, leading to cell death, illustrating the functional significance of the ERSU pathway. To further our understanding of the molecular mechanisms of the ERSU pathway, we used live‐cell imaging and microfluidics to investigate the ER inheritance process in cells exposed to ER stress. We found that establishment of the ER inheritance block requires functional asymmetry between the cortical ER (cER) and perinuclear ER (pnER) and that this functional asymmetry, which depends on Reticulons 1 and 2 and Yop1, helps change the spatial orientation of the tubular ER, diminishing its ability to enter the daughter cell. We also found that ER stress causes a transient increase in sphingolipid synthesis. Cells defective in sphingolipid synthesis or cells treated with inhibitors of specific step(s) in the sphingolipid biosynthetic pathway fail to block both inheritance of a stressed ER and cytokinesis. However, these cells were able to induce the UPR normally, demonstrating that the failure to initiate the ERSU pathway was not due to an inability to perceive and respond to ER stress. Finally, addition of sphingolipids leads to block of ER into the daughter cell in a manner dependent on Rtn1, Rtn2, and Yop1 and septin relocalization. The molecular mechanisms by which changes in sphingolipids induce Rtn1‐, Rtn2‐, and Yop1‐dependent ER functional asymmetry and by which such ER functional asymmetry causes the ER inheritance block will be discussed.

M186 BiP AMPylation by dFic is required for vision in Drosophila. A.K. Casey1, A.T. Moehlman2, J. Zhang1, H. Kramer2,3, K. Orth1,4; 1Department of Molecular Biology, University of Texas Southwestern Medical Center, Dallas, TX, 2Department of Neuroscience, University of Texas Southwestern Medical Center, Dallas, TX, 3Department of Cell Biology, University of Texas Southwestern Medical Center, Dallas, TX, 4Howard Hughes Medical Institute, Dallas, TX BiP is an ER chaperone protein with a critical and central role in the regulation of the unfolded protein response (UPR) and Endoplasmic Reticulum (ER) homeostasis. Previously, we discovered that the protein dFic is responsible for the reversible AMPylation of BiP in Drosophila melanogaster and that this post‐tanslational modification is regulated by induction of ER stress. We also found that dfic null flies have vision defects using electroretinograms (ERGs) and phototaxis assays. We hypothesized that the vision defects of dfic null flies are due to changes in BiP AMPylation. To test this hypothesis, we generated flies that express either wild type BiP or mutant forms of BiP that cannot be AMPylated by dFic. Using ERGs, the vision of flies expressing wild type BiP appears normal whereas our BiP mutant flies display ERG defects comparable to that of dfic null flies. This supports our hypothesis that post‐translational regulation of BiP by AMPylation is important for vision in flies. Furthermore, dfic and BiP mutant flies vision defects in are increased under certain stresses, suggesting a connection to the UPR. Using a mixture of genetics, biochemistry, and mass spectrometry analysis, our current goal is to further understand how AMPylation of BiP is regulated and how AMPylation impacts the function of BiP in the ER.


M187 A conformational RNA zipper promotes non‐conventional XBP1 mRNA splicing. J. Peschek1, D. Acosta‐Alvear1, A.S. Mendez1, P. Walter1; 1Department of Biochemistry and Biophysics, UC San Francisco/HHMI, San Francisco, CA The kinase/endonuclease IRE1 is the most conserved signal transducer of the unfolded protein response (UPR), an intracellular signaling network that monitors and regulates the protein folding capacity of the endoplasmic reticulum (ER). Upon sensing protein folding perturbations in the ER, IRE1 initiates the unconventional splicing of XBP1 mRNA cumulating in the production of the transcription factor XBP1s, which expands the ER folding capacity. The splicing mechanism requires two coordinated cleavage events and the subsequent ligation of the correct exon ends. How this reaction is orchestrated with fidelity remains an outstanding question. We reconstituted the splicing reaction in vitro using its core components IRE1, XBP1 mRNA and tRNA ligase. We showed that a conserved, RNA‐intrinsic conformational change acts to keep the severed exons together thus ensuring splicing efficiency and fidelity. This conformational change occurs within a conserved RNA structure of XBP1 mRNA. It allows zippering of the exons by base‐pairing and concomitant ejection of the intron after IRE1‐mediated cleavage. We confirmed this mechanism in cells using XBP1‐derived splicing reporters. Our results support a model in which the XBP1 mRNA serves as an active component of the splicing mechanism in the IRE1‐mediated UPR pathway.

M188 Tetraspanins as novel regulators of secretory granule biogenesis. C.J. Ma1,2, J.A. Brill1,2,3; 1Cell Biology Program, The Hospital for Sick Children, Toronto, ON, 2Institute of Medical Science, University of Toronto, Toronto, ON, 3Department of Molecular Genetics, University of Toronto, Toronto, ON Regulated secretion is a cellular process of fundamental importance in which biologically active molecules are released in response to external stimuli. Cargoes of regulated secretion are stored in long‐lasting secretory granules (SGs), but the underlying mechanisms behind the biogenesis and maturation of these SGs are still poorly understood. The Drosophila melanogaster larval salivary gland is an excellent model to address these mechanisms because it is enriched with hormonally regulated SGs known as glue granules. To identify key players involved in biogenesis and maturation of SGs, we employed a candidate RNAi screen using publicly available transgenic lines. The effects of gene knockdown were scored by examining SG size. From this screen, we identified a surprisingly large number of endocytic genes required for proper SG maturation, indicating that glue granules might be a type of lysosome related organelle (LRO). As the tetraspanin CD63 is known to localize to LROs and is crucial for LRO biogenesis, we decided to characterize CD63 in our model system. When expressed in larval salivary glands, mammalian CD63‐GFP strongly localized to the limiting membrane of SGs. In addition, the expression level of CD63‐GFP was positively correlated with the size of SGs. Knocking down Drosophila CD63 (Tsp29Fa) resulted in smaller SGs, whereas expressing mammalian CD63‐GFP in a Tsp29Fa RNAi background restored SGs to normal size. These results show that tetraspanins not only localize to SGs, but also play a functional role in SG biogenesis. To further investigate how CD63 affects SG maturation, we are conducting a secondary screen to determine whether CD63 suppresses other positive hits from our initial RNAi screen. We discovered that CD63 expression suppresses the small granule phenotype caused by knockdown of AP‐2 or ESCRT‐II, but not by knockdown of Rab5 or Syx16. Glue granules were previously thought to be derived solely from post‐Golgi secretory trafficking.


However, we have discovered that biogenesis of these granules also requires CD63 and endocytic trafficking. Our results implicate glue granules as a new type of LRO and identify tetraspanins as conserved regulators of LRO/SG biogenesis.

M189 Monoubiquitination of syntaxin 5 regulates Golgi membrane dynamics during the cell cycle. S. Huang1, D. Tang1, Y. Wang1; 1Molecular, Cellular, and Developmental Biology, University of Michigan, Ann Arbor, MI Partitioning of the Golgi membrane into daughter cells during mammalian cell division occurs through a unique disassembly and reassembly process. Several converging lines of evidence have suggested that monoubiquitination plays an essential role in the regulation of post‐mitotic Golgi membrane fusion. Monoubiquitination, as a regulatory signal, occurs during mitotic Golgi disassembly and is required for subsequent Golgi reassembly by p97. Previously, we have identified a Golgi‐localized ubiquitin E3 ligase HACE1 that is involved in mitotic Golgi disassembly; reduced HACE1 activity is human cancer such as Wilms’ tumor results in fragmented Golgi. We have also discovered the p97/p47 binding protein VCIP135 as a deubiquitinating enzyme whose activity is required for post‐mitotic Golgi reassembly. VCIP135 is inactivated by phosphorylation in metaphase and activated upon dephosphorylation in telophase. Here we report the identification of the Golgi t‐SNARE syntaxin 5 (Syn5) as the ubiquitinated substrate. Syn5 is monoubiquitinated by the ubiquitin ligase HACE1 in early mitosis and deubiquitinated by the deubiquitinase VCIP135 in late mitosis. Syn5 ubiquitination on Lysine 270 (K270) in the SNARE domain impairs the interaction between Syn5 and the cognate v‐SNARE Bet1, but increases its binding to p47, the adaptor protein of p97. Expression of the Syn5 K270R mutant in cells impairs post‐mitotic Golgi reassembly. Therefore, monoubiquitination of Syn5 in early mitosis disrupts SNARE complex formation. Subsequently, ubiquitinated Syn5 recruits p97/p47 to the mitotic Golgi fragments and promotes post‐mitotic Golgi reassembly upon ubiquitin removal by VCIP135. Overall, this study reveals both the substrate and the mechanism of ubiquitin‐mediated regulation of Golgi membrane dynamics during the cell cycle.

M190 A STRIPAK complex mediates axonal transport of autophagosomes and dense core vesicles through PP2A regulation. A.L. Neisch1, T.P. Neufeld1, T.S. Hays1; 1Genetics, Cell biology and Development, University of Minnesota, Minneapolis, MN An accumulation of cellular debris is a hallmark of several neurodegenerative diseases. In neurons, autophagy has a crucial function in the clearance of misfolded proteins, protein aggregates and damaged organelles to maintain cellular homeostasis. Autophagosomes, the organelles that encapsulate this cellular debris, form at distal sites in the axon and are transported along microtubules to proximal sites of lysosomal fusion, resulting in the degradation of the encapsulated cellular material. The microtubule motor protein, dynein, is known to drive the transport of autophagosomes to sites of lysosomal fusion in mammalian cultured neurons (Maday et al., 2012). To identify proteins that regulate the attachment and transport of autophagosomes, we carried out an RNAi based screen in Drosophila motoneurons. We visually screened larvae for the accumulation of autophagosomes within synaptic terminals, a phenotype that reflects the disruption of transport and which we previously observed upon depletion of the dynein heavy chain. Intriguingly, we identified the Striatin family protein, Connector of kinase to AP‐1 (CKA), and show that CKA is required for autophagosome transport in neurons. CKA


functions within a larger multi‐protein complex, Striatin‐interacting phosphatase and kinase (STRIPAK), to regulate autophagosome transport. In addition, CKA co‐immunoprecipitates with dynein and can directly bind to the autophagosomal membrane protein Atg8 in pull down assays in vitro. We show using in vivo imaging that the STRIPAK complex also regulates the transport of dense core vesicles, but not mitochondrial transport. These results suggest that the STRIPAK complex functions locally to regulate organelle transport, rather than non‐specifically disrupting all organelle transport through a global impact on the microtubule cytoskeleton. Within the STRIPAK complex, CKA functions as a regulatory subunit of the phosphatase PP2A. PP2A is required within the STRIPAK complex to mediate axonal transport, and we show that CKA‐deficient, adult Drosophila exhibit PP2A‐dependent locomotor defects. Together, our data provide evidence that CKA can act to link dynein to autophagosomes and that the STRIPAK complex provides phospho‐regulatory activity that is important for regulating axonal transport of a subset of organelles. Our data suggest that CKA function within the STRIPAK complex is crucial to prevent axonal transport defects that contribute to neurodegeneration. Maday, S., K.E. Wallace, and E.L. Holzbaur. 2012. Autophagosomes initiate distally and mature during transport toward the cell soma in primary neurons. J Cell Biol. 196:407‐17.

Minisymposium 19: Recent Developments in Autophagy and ESCRT Biology

M191 Non‐Canonical ESCRT Structures and Functions. M. Gu*1, D. LaJoie*2, O.S. Chen*1, M.S. Ladinsky3, M.J. Redd4, L. Nikolova5, P.J. Bjorkman3, W.I. Sundquist1, K.S. Ullman2, A. Frost1,6; 1Department of Biochemistry, University of Utah, Salt Lake City, UT, 2Department of Oncological Sciences, Huntsman Cancer Institute, University of Utah, Salt Lake City, UT, 3Cal‐Tech Division of Biology and Biological Engineering, Caltech, Pasadena, CA, 4HSC Imaging Core Facility, University of Utah, Salt Lake City, UT, 5HSC Electron Microscopy Core Facility, University of Utah, Salt Lake City, UT, 6Department of Biochemistry and Biophysics, University of California, San Francisco, San Francisco, CA *These authors contributed equally to this work. The molecular mechanism for sealing newly‐formed nuclear envelopes was unclear until the recent discovery that ESCRT‐III proteins mediate this process. CHMP7, in particular, was identified as an early‐acting factor that recruits other ESCRT‐III proteins to the nuclear envelope. A fundamental aspect of the varied roles of ESCRT factors is their recruitment by site‐specific receptors, yet the central question of how the ESCRT machinery is targeted to nuclear membranes has remained outstanding. We have identified an inner nuclear membrane protein, LEM2, as a key and conserved factor that directly recruits CHMP7 and downstream ESCRT‐III proteins to breaches in the nuclear envelope.

M192 Buckling of the cell membrane by ESCRT‐III. A. Roux1,2, N. Chiaruttini*1, J. Moser von Filseck1, S. Scheuring3, L. Redondo‐Morata3, M. Girardin1, M. Lenz4; 1Biochemistry, University of Geneva, Geneva, Switzerland, 2Chemical Biology, National Centre for Competitive Reseach (NCCR), Geneva, Switzerland, 3U1006, INSERM/Université Aix‐Marseille, Marseille, France, 4LPTMS, CNRS/Université Paris‐Sud, Orsay, France Cells and organelles are delimited by lipid bilayers. Since these membranes are impermeable to most solutes, in order to exchange material with their environment, organelles and cells have developed a


large protein family involved in budding membranes to form membrane carriers. These carriers transport material between organelles. Proteins involved in intracellular membrane traffic can remodel the membrane by several ways. Clathrin, for example, polymerizes into a spherical cage onto the membrane, forcing it to curve. Here we describe a recently discovered protein complex called ESCRT‐III, which has the property of forming spirals at the surface of the lipid bilayer. This unique structural feature did not suggest any known mechanism by which it could deform the membrane. It was theoretically proposed that, while growing into a spiral, it accumulates stress energy which can be released by buckling of the central part of the spiral. By using high‐speed AFM and biophysical tools to measure membrane elasticity we show how the elastic and polymerization properties of the ESCRT‐III filament are compatible with such model. We further investigate the dynamics of the complex.

M193 Antiviral factors fashioned from ESCRT functions. D.M. Downhour1, L. Rheinemann2, G. Marcenne2, J. McCullough2, A.N. McKeown1, W.I. Sundquist2, N.C. Elde1; 1Human Genetics, University of Utah, Salt Lake City, UT, 2Biochemistry, University of Utah, Salt Lake City, UT Many enveloped viruses require interactions with components of ESCRT pathway to escape from infected cells. Among the ESCRT proteins, viruses co‐opt members of the family of charged multi‐vesicular body proteins (CHMPs) to promote the release of budding virus particles from cellular membranes. Unlike genes encoding many immune defense functions, CHMP genes are highly conserved, reflecting the complex coordination of numerous membrane trafficking functions. These tight genetic constraints raise the possibility that ESCRT functions are always vulnerable to viral exploitation. However, our evolutionary analysis of CHMP genes across diverse primates identified numerous duplicated CHMP gene copies undergoing rapid evolution, a genetic signature of evasion from viral exploitation. These genes arose as processed pseudogenes, a byproduct of retrotransposition, and based on previous studies of synthetic constructs are novel candidates for antiviral restriction of viruses, including HIV. We observed striking inhibition of HIV replication with the expression of a CHMP3 retrogene from squirrel monkeys. Importantly, retroCHMP3 exhibited reduced cytotoxicity compared with synthetic CHMP3 genes. Combining evolutionary reconstruction with confocal microscopy allowed us to narrow the genetic basis of reduced cytotoxicity to a handful of amino acid substitutions in retroCHMP3 compared to the parental gene. These findings reveal a natural means of blocking HIV infections through modification of ESCRT functions and suggest a mechanism for host‐pathogen evolution where retrogenes derived from highly conserved host genes adapt to provide previously unrecognized activities against pathogens. Our work also supports the emerging idea that a substantial share of genomic and cellular complexity reflects highly entangled histories of conflict with pathogens stretching back before the origins of eukaryotic cells.

M194 SMCR8 functions as negative autophagy regulator by inhibiting ULK1 kinase activity and gene expression. C. Behrends1; 1Goethe University, Institute of Biochemistry II, Medical School, Frankfurt Am Main, Germany Autophagy is an intracellular degradation and recycling pathway during which cargoes such as damaged organelles, misfolded proteins or pathogens are engulfed in the cytoplasm by double membrane


structures termed phagophores. Mature autophagosomes are formed by elongation and closure of phagophores and autophagosomal content is degraded in autolysosomes after fusion of autophagosomes with lysosomes. Rab GTPases are small G‐proteins of the Ras superfamily and recruit effector proteins during their GTP‐bound state. Together with their regulators Rab GAPs (GTPase activating protein) and Rab GEFs (guanine nucleotide exchange factor), Rab GTPases are cellular coding hubs for intracellular trafficking and regulate vesicle budding, transport, targeting and fusion. Our knowledge of the role of Rab GTPases and its regulators in autophagy is currently limited. Therefore, we performed an image‐based focused RNAi screen simultaneously monitoring early to late autophagosomes at endogenous level and identified 34 Rab machinery components passing our stringent selection criteria. SMCR8 interaction proteomics revealed association with the ULK1 kinase and C9ORF72 GEF complexes with overlapping binding regions. Accordingly, SMCR8 modulated formation and maturation of autophagosomes. Furthermore, we revealed increased ULK1 kinase activity in SMCR8 depleted cells and in addition, increased ULK1 gene expression. The latter phenotype involved chromatin association of SMCR8 with the ULK1 gene locus. Collectively, we established SMCR8 as dual negative regulator of ULK1.

M195 A Membrane‐Associated Autophagy Signaling Pathway for Controlling Peroxisome Fate. V. Denic1; 1Molecular and Cellular Biology, Harvard University, Cambridge, MA Eukaryotic cells maintain the quality of their organelles by controlling organelle turnover via autophagy. Recent work has shown that autophagy receptor proteins enable selective organelle destruction by triggering local autophagosome initiation. With the exception of mitochondrial autophagy in mammalian cells, which is driven by cytosolic receptor recruitment to depolarized mitochondria, molecular mechanisms that commit damaged organelles for destruction by selective autophagy are not well understood. To discover novel mechanisms that regulate selective autophagy, we studied Atg36, the sole autophagy receptor that drives peroxisome degradation in the budding yeast S. cerevisiae. Atg36 is anchored to the cytosolic face of peroxisomes by the integral membrane protein Pex3 but remains latent until phosphorylated by the cytosolic kinase Hrr25. We affinity‐purified Pex3 from detergent‐solubilized membranes to reveal the existence of a novel multi‐subunit complex comprising the Pex3‐Atg36 subcomplex in association with the Pex1/6/15 AAA‐ATPase membrane subcomplex. Pex1/6 had previously been shown to use ATP hydrolysis energy to extract ubiquitinated Pex5 from the peroxisome membrane, but we found that Pex5 is not involved in Atg36 regulation. Nonetheless, an active site mutation that disrupts the ATPase activity of Pex6 without disrupting membrane complex formation induced constitutive Atg36 phosphorylation. We could demonstrate that Pex1/6/15 AAA‐ATPase subcomplex has a novel function as a direct repressor of Atg36 phosphoactivation by protein re‐targeting to the mitochondrial outer membrane and biochemical reconstitution. Our work uncovers a novel mechanism by which autophagy receptors are kept inactive on the cytosolic surface of functional organelles via interactions with membrane‐associated factors. As many receptors reside constitutively on target organelles, but trigger organelle degradation only upon phosphoactivation, we propose that local receptor inhibition may be a general mechanism for controlling organelle turnover.


M196 Role of the human Vps15 kinase in PI3K complex I mediated autophagy regulation. G. Stjepanovic1,2, M.G. Lin1, S. Baskaran1, L. Carlson1, J.H. Hurley1,2; 1MCB, UC Berkeley, Berkeley, CA, 2LBNL, Berkeley, CA Autophagy is basic catabolic pathway for the controlled degradation of cellular components. The class III phosphatidylinositol 3‐kinase complex (PI3KC3) phosphorylates phosphatidylinositol to generate phosphatidylinositol 3‐phosphate, which is essential to initiate autophagy. We have previously reported the structural architecture of the complex and suggested a mechanism for the regulation of the lipid kinase activity by the regulatory signals. However, a long standing question about role of serine/threonine protein kinase Vps15 in the PI3KC3 complex remained unanswered. In this study we demonstrate that catalytic Vps15 mutations prevent delivery of cargo proteins to the vacuole and lead to inhibition of autophagy in yeast. Our hybrid structural biology approach describes the molecular mechanism of Vps15 mediated assembly of the PI3KC3 complex. We show the full range of PI3KC3 domain movements that occur in solution and determinate the critical role of conformation flexibility in PI3K complex activity. Our findings reveal the major mechanism of regulation of PI3K lipid kinase activity required in early stages of autophagosome formation.

M197 Membrane recognition by the Atg4 family of proteases. K. Kauffman1, J. Jin1, B. Mugo1, A. Lystad2, T.J. Melia1; 1Cell Biology, Yale School of Medicine, New Haven, CT, 2Department of Molecular Medicine, University of Oslo, Oslo, Norway Autophagy depends upon the transient formation of a protein‐lipid adduct between LC3 proteins and phosphatidylethanolamine lipids (forming LC3‐PE on the autophagosome membrane). LC3‐PE drives membrane growth, captures cargo and is necessary to close the autophagosome. Nonetheless, LC3‐PE eventually becomes an impediment to the completion of autophagy, and must be removed from the outside of mature autophagosomes to allow fusion of autophagosomes with lysosomes. The Atg4 proteases drive LC3‐PE removal through proteolysis at the peptide‐lipid bond. However, the principle Atg4 protease, Atg4B, is abundant and constitutively active in the cytoplasm and yet LC3‐PE proteolysis is limited spatially and temporally to mature autophagosomes. How proteolysis is restricted on autophagosomes is unknown. We recently demonstrated that the pathogen, Legionella pneumophila, specifically aborts autophagy in infected host cells by accelerating the cleavage of LC3‐PE, leaving immature autophagosome membranes with no path to complete closure or engage the lysosome. To target membrane‐bound LC3‐PE, Legionella expresses a unique bacterial protease which becomes activated only at the surface of autophagosomes and thus is a specific and highly efficient LC3‐PE degrader. Here, we apply similar in vitro and in vivo tools to establish how Atg4 proteins determine temporal and spatial specificity. The ability to recognize and cleave LC3‐PE by each of the four Atg4 isoforms is not at all predicted by measures of Atg4 activity in solution. Instead membrane‐proximal cleavage is driven by interfacial activation or inhibition of individual Atg4 proteins. For example, Atg4B which has been exhaustively studied in solution is 1000‐fold less active when substrates like LC3‐PE are anchored in intact membranes. In contrast, Atg4C and Atg4D are almost exclusively membrane‐active. Further, mutation or removal of auto‐inhibitory domains in Atg4A, Atg4C or Atg4D proteases selectively enhances LC3‐PE family turnover rates without activating these enzymes towards unprimed soluble LC3 family substrates.


We propose that this interfacial activation mechanism allows the cell to tune the lifetime and homolog composition of the LC3‐PE protein families on the maturing autophagosome.

M198 Dynamic regulation of autophagy in neurons during aging. A.K. Stavoe1, E.L. Holzbaur1; 1Physiology, University of Pennsylvania, Philadelphia, PA Autophagy defects have been implicated in neurodegenerative disorders, including Parkinson’s and Alzheimer’s Diseases. While the role of autophagy in cellular homeostasis has been studied in yeast and mammalian cell culture, much less in known about autophagy in neurons, especially during aging. Autophagosome biogenesis is known to occur preferentially at the distal tips of axons, and autophagosomes mature as they transit to the soma. We used dual‐color, live‐cell imaging of dorsal root ganglion (DRG) neurons to determine the temporal dynamics and regulation of autophagy in neurons from young (P90‐120) and aging (P480‐540) mice. We identified a significant decrease in both autophagosome biogenesis and flux during aging. In addition, we show that autophagosome assembly factors display disrupted kinetics in aged neurons compared to young adult neurons. Stepwise assembly of autophagosomes results in productive biogenesis events in neurons from young adult mice. Conversely, in 70% of neurons from aged mice, early autophagosome assembly factors coalesce into stalled isolation membranes and do not form complete autophagosomes. Surprisingly, we found a significant increase in distal autophagosome maturation in aged neurons compared to young adult neurons, as determined by autophagosome fusion with lysosomes and acidification. Our studies will help us elucidate how aging modulates autophagy, which will further our understanding of how misregulation of autophagy contributes to neurodegenerative diseases.

M199 Visualization of the autophagosomal maturation: The ATG‐conjugation systems are important for degradation of the inner autophagosomal membrane. I. Koyama‐Honda1, K. Tsuboyama1, Y. Sakamaki2, M. Koike3, N. Mizushima1; 1Department of Biochemistry and Molecular Biology, The University of Tokyo, Graduate School and Faculty of Medicine, Tokyo, Japan, 2Research Center for Medical and Dental Sciences, Tokyo Medical and Dental University, Tokyo, Japan, 3Departments of Cell Biology and Neuroscience, Juntendo University Graduate School of Medicine, Tokyo, Japan Macroautophagy is an evolutionarily conserved degradation pathway. Cytoplasmic proteins and organelles are sequestered into the double‐membrane structure, the autophagosome, which then fuses with the lysosome to become the autolysosome. We characterized how the autophagosome maturates into the autolysosome in mammalian cells using the autophagosomal SNARE syntaxin17 as a novel autophagosome marker. Immediately after syntaxin17 recruitment, several small lysosomes fuse with an autophagosome. Acidification initially occurs between the outer and inner autophagosomal membranes, followed by whole‐autophagosome acidification due to degradation of the inner membrane. Syntaxin17 dissociates from the membrane thereafter. Next, we found that syntaxin17‐positive autophagosome‐like structures could be generated in cells lacking the ATG conjugation systems, whose counterparts are essential for autophagosome formation in yeast. Using syntaxin17 together with other markers, we show that the ATG conjugation systems are important for two critical steps in autophagosome formation and maturation: (1) efficient transition from the isolation membrane (unclosed autophagosome) to the autophagosome and (2) efficient degradation of the inner


autophagosomal membrane. Accordingly, autophagic activity in ATG conjugation‐deficient cells is strongly suppressed. These results suggest that the ATG conjugation systems, which are likely required for the closure/fission of the autophagosomal edge, are not absolutely essential for autolysosome formation but important for efficient degradation of the inner autophagosomal membrane.

M200 Digitizing Ubiquitin Signaling for Mitophagy. A. Ordureau1, J.A. Paulo1, S.P. Gygi1, J.W. Harper1; 1Cell Biology, Harvard Medical School, Boston, MA Mitochondrial quality control is essential to the health of the cell. As such numerous mechanisms have evolved to control both the abundance of damaged mitochondria in the cell as well as the state of protein folding within mitochondria itself. Our work has focused on the role of ubiquitin in marking damaged mitochondria via a signaling pathway involving the PINK1 kinase and the PARKIN ubiquitin ligase, two genes mutated in Parkinson’s Disease. Using quantitative proteomics, we have found that PARKIN builds K6, K11, K48, and K63 chains on dozens of mitochondrial outer membrane (MOM) proteins. Subsequent to ubiquitin chain synthesis, a program of autophagy adaptor recruitment is initiated and is driven via the TBK1 protein kinase, which phosphorylates autophagy adaptor OPTN to promote mitophagy by increasing affinity of these receptors for ubiquitin chains. These studies suggest a self‐reinforcing positive feedback mechanism that coordinates TBK1‐dependent autophagy adaptor phosphorylation with the assembly of ubiquitin chains on mitochondria to facilitate efficient mitophagy, and mechanistically links genes mutated in Parkinson’s disease and amyotrophic lateral sclerosis (TBK1 and OPTN) in a common selective autophagy pathway. In order to further understand this system, we are developing new proteomic tools and procedures that allow us to quantify dozens of primary ubiquitylation events on numerous PARKIN substrates on the MOM. This system allows the kinetics of modification of individual lysine residues as well as a determination of abundance on individual substrates. We demonstrate that the relative abundance of PARKIN substrates on mitochondria differs by 2 orders of magnitude, and that the stoichiometry of ubiquitylation on individual sites various greatly across the substrates. This approach provides a means by which to digitize ubiquitylation events, thereby allowing the development of a quantitative framework for understanding complex signal‐dependent ubiquitylation cascades.

Microsymposium 13: Nuclear Structure, Function, and Movement

E109 Control of gene expression by nucleus centration in mouse oocytes. M. Almonacid1, S. El‐Hayek1, A. Othmani2, W.W. Ahmed3,4,5, P. Davidson3,4,5, C. Campillo6, C. Sykes3,4,5, A. Genovesio2, M. Verlhac1; 1Center for Interdisciplinary Research in Biology (CIRB), CNRS, INSERM, PSL Research University, Paris, France., College de France, Paris, France, 2Scientific Center for Computational Biology, CNRS‐INSERM‐ENS, PSL Research University., Institut de Biologie de l Ecole Normale Supérieure, Paris, France, 3CNRS‐UMR168, Paris, France, 4UPMC, Paris, France, 5Laboratoire Physico‐Chimie, Institut Curie, Centre de Recherche, Paris, France, 6LAMBE, Université Evry Val d Essonne, Evry, France Nucleus position can act as a morphogenetic cue in many cells conveying both spatial and temporal information. In a variety of species, the location of the oocyte nucleus defines the future axis of the embryo. In worms, sea urchin, frog, some fishes and amphibians, the nucleus marks the animal pole and


in Drosophila it defines the future antero‐posterior axis of the embryo. However, in mammals, the nucleus is centrally located and a failure to do so correlates with a poor outcome of oocyte development (Brunet & Maro 2007). Using a multidisciplinary approach, optical trapping and modelling, we have recently discovered how the nucleus robustly localizes in a very original manner at the cell centre in the mouse female gamete, the oocyte (Almonacid 2015). Reintroduction of the actin nucleator Formin 2 induces a directional movement of the nucleus towards the centre in Formin 2 ‐/‐ oocytes, that harbour initially off‐centred nuclei. Using this robust assay for nucleus centration, we are now addressing the biological significance of nucleus positioning in the mouse oocyte. We have observed that nuclear architecture is disturbed in Formin 2 ‐/‐ oocytes. A bioinformatic approach on 3D reconstructed nuclei allowed to identify about 100 discriminating features between control and Formin 2 ‐/‐ nuclei, notably less chromatin condensation and more chromatin/Lamin A contacts in Formin 2 ‐/‐. Consistent with perturbed nuclear architecture, RNA sequencing of control versus ‐/‐ oocytes allowed the detection of about 2300 genes showing significant (p< 0.05) deregulation. We investigate in greater details this non‐canonical function of a microfilament nucleator in the regulation of oocyte transcription. In particular we address the importance of mechanotransduction via Formin 2‐nucleated microfilaments in the control of oocyte transcription, a step critical for gamete quality. Formin 2 ‐/‐ oocytes are rescued with full length Formin 2, the actin nucleating domain of the protein, or its N‐terminal domain, not involved in actin nucleation, and compared for nuclear architecture by Principal Component Analysis (PCA) on their 100 most discriminating features, and transcriptomic profile. Our work proposes a novel role for an actin nucleator in chromatin organization and gene expression with potential impact on the very first steps of development. References : Brunet S & Maro B. (2007). Germinal vesicle position and meiotic maturation in mouse oocyte. Reproduction 133 : 1069–1072. Almonacid M, Ahmed WW, Bussonnier M, Mailly P, Betz T, Voituriez R, Gov NS, Verlhac MH. (2015). Active diffusion positions the nucleus in mouse oocytes. Nat Cell Biol 17:470‐9.

E110 Nucleosome–nucleosome interactions via histone tails and linker DNA regulate nuclear rigidity. Y. Shimamoto1,2,3, S. Tamura4, K. Maeshima2,4; 1Center for Frontier Research, National Institute of Genetics, Shizuoka, Japan, 2Department of Genetics, SOKENDAI University, Shizuoka, Japan, 3AMED‐PRIME, Japan Agency for Medical Research and Development, Tokyo, Japan, 4Structural Biology Center, National Institute of Genetics, Shizuoka, Japan Cells in the human body generate and respond to significant mechanical stresses during migration, contraction, and adhesion. The nucleus, a micron‐sized cellular organelle that packages genomic DNA and regulates gene expression, is subjected to such stresses while maintaining its structural and functional stability. Growing evidences suggest that chromatin, a macromolecular assembly of DNA and associated proteins, plays a major role in regulating the mechanical properties of nuclei. However, the related physical and molecular mechanisms remain largely unknown, due primarily to the lack of an assay that dissects nuclear micromechanics with controlled chromatin perturbations. Here, we combined the microneedle‐based force measurement setup we developed previously (Shimamoto et al., Cell, 2011) with controlled biochemical assays on isolated HeLa‐cell nuclei (Maeshima et al., EMBO J., 2016), to examine their rigidity and elasticity and how these mechanics change with different biochemical perturbations of chromatin. Our major findings were as follows: 1) The nuclei were significantly rigid and elastic when the buffer’s Mg2+ level was sufficiently high (e.g., 5 mM), at which nuclear chromatin formed a highly compacted state. Lowering the buffer’s Mg2+ level (<1 mM) resulted in less rigid nuclei, associated with chromatin decompaction. 2) The digestion of linker DNA by a restriction endonuclease led to a significant loss of nuclear rigidity. These nuclei also lost their sensitivity


to Mg2+ levels on overall rigidity. 3) The acetylation of histone H4 tails, which mediate the interaction between nucleosomes, was induced by trichostatin A, a histone deacetylase inhibitor, and significantly decreased nuclear rigidity. Notably, the composition of other chromatin assembly factors, cohesin, CTCF, and condensin II, and the major nuclear component, lamin, remained unperturbed in all experimental conditions. To conclude, we propose a simple mechanical model for chromatin‐based nuclear‐mechanics regulation and suggest that chromatin compaction possesses the advantage of protecting genome integrity against mechanical perturbations.

E111 Nuclei migrate through constricted spaces using microtubule motors and actin networks in C. elegans hypodermal cells. C. Bone1, Y. Chang1, N. Cain1, D.A. Starr1, S. Murphy1; 1Molecular and Cellular Biology, University of California, Davis, Davis, CA Cellular migrations through constricted spaces are a critical aspect of many developmental and disease processes including hematopoiesis, inflammation, and metastasis. A limiting factor in these events is nuclear deformation. Here, we establish an in vivo model where nuclei can be visualized while moving through constrictions and use it to elucidate mechanisms for nuclear migration. C. elegans hypodermal P‐cell larval nuclei traverse a narrow space of around 200 nm, about 5% their width. This constriction is blocked by fibrous organelles, structures connecting the muscles to cuticle through P cells. Fibrous organelles are removed just prior to nuclear migration, when nuclei and lamins undergo extreme morphological changes to squeeze through the space. Both actin and microtubule networks are organized to mediate nuclear migration. The LINC complex, consisting of the SUN protein UNC‐84 and the KASH protein UNC‐83, recruits dynein and kinesin‐1 to the nuclear surface. Both motors function in P‐cell nuclear migration, but dynein, functioning through UNC‐83, plays a more central role as nuclei migrate toward minus ends of polarized microtubule networks. Thus, the nucleoskeleton and cytoskeleton are coordinated to move nuclei through constricted spaces.

E112 Stiffness‐dependent maintenance of Mkl1/SRF‐signaling requires Emerin and Lamin A. M.K. Willer1, C.W. Carroll1; 1Dept. of Cell Biology, Yale University, New Haven, CT The nuclear lamina (NL) is a protein‐rich meshwork composed of filamentous lamins and their associated inner nuclear membrane proteins. While the NL lines the inside of the nuclear envelope, its physical connection to the cytoskeleton enables force transmission from the cellular environment to the nuclear interior. Recently, Lamin A and Emerin, an inner nuclear membrane protein, were shown to regulate the activity of the transcription factor/co‐activator complex SRF/Mkl1, which controls expression of genes involved in cell migration and contraction. A critical step in the activation of SRF/Mkl1 is accumulation of Mkl1 in the nucleus in response to chemical and mechanical stimuli. Here, we show that Emerin promotes Mkl1 nuclear localization in a mechanosensitive manner. The extent of nuclear Mkl1 after serum stimulation increased in normal fibroblasts as cells were cultured on increasingly stiff substrates. In contrast, Mkl1 nuclear localization did not change in Emerin null (EmnKO) fibroblasts as substrate stiffness increased. As a result, the levels of nuclear Mkl1 after serum stimulation were similar between EmnKO and control cells on soft substrates whereas the loss of Emerin significantly reduced Mkl1 nuclear localization on stiff substrates. Unexpectedly, reduced nuclear Mkl1 in serum stimulated EmnKO cells did not lead to a defect in SRF/Mkl1‐dependent transcription over short


time scales. Instead, Emerin was required for sustained SRF/Mkl1 activity. By analyzing the levels of several SRF/Mkl1 target genes in different culture conditions, we determined that the inability of EmnKO fibroblasts to properly regulate Mkl1 localization does not result from a pre‐existing defect in SRF/Mkl1 activity, suggesting that Emerin is essential for the transduction of mechanical cues into a sustained transcriptional response. Consistent with this possibility, we found that EmnKO fibroblasts also have defects in focal adhesion assembly. Similar to Emerin loss, depletion of Lamin A specifically impaired Mkl1 nuclear localization in cells cultured on stiff substrates. Moreover, Lamin A depletion in EmnKO fibroblasts further reduced Mkl1 nuclear localization, suggesting that Lamin A has additional Emerin‐independent roles in SRF/Mkl1 activation. These results have implications for our understanding of how mutations in the broadly expressed factors SRF, Mkl1, Lamin A, and Emerin lead to tissue‐specific diseases and identify the NL as a potential novel regulator of pathologic fibrosis in response to tissue stiffening after wound healing.

E113 Genome variation in an osteosarcoma cell line after pore migration. J. Irianto1, Y. Xia1, C.R. Pfeifer1, J. Ji1, C.M. Alvey1, M. Tewari1, R.A. Greenberg2, D.E. Discher1; 1School of Engineering and Applied Sciences, University of Pennsylvania, Philadelphia, PA, 2Cancer Biology, Abramson Family Cancer Research Institute, Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA Migration through micron‐size constrictions has been reported with various cell types to rupture the nucleus, release nuclear‐localized GFP, and cause localized accumulations of ectopic 53BP1 – a DNA repair protein – suggestive of DNA damage. Here, constricted migration of a U2OS osteosarcoma line causes DNA breaks based on both electrophoretic ‘comet’ measurements and foci of phosphorylated‐histone (γH2AX), but without affecting endogenous 53BP1. Migration also causes cytoplasmic mis‐localization of DNA repair proteins, several of which are known from mouse knockouts or mutants to change chromosome copy numbers. Partial knockdown of these factors in U2OS cells causes DNA breaks and aberrant levels of DNA without affecting migration. Migration‐induced nuclear damage is reversible for wild‐type U2OS and sub‐cloned lines, but DNA array and RNA‐seq analyses reveal lasting genomic differences and heterogeneity, with hundreds of megabases in gains and losses in various chromosomes. Phenotypic differences that arise from constricted migration are further illustrated by a clone with highly elongated morphology that depends on microtubules.

E114 Mutant lamin A causes elevated ribosome biogenesis and protein translation in Hutchinson‐Gilford Progeria Syndrome. A.L. Buchwalter1, M.W. Hetzer1; 1Molecular and Cell Biology Laboratory, The Salk Institute for Biological Studies, La Jolla, CA The nuclear lamina is an intermediate filament meshwork that scaffolds the nucleus. In Hutchinson‐Gilford progeria syndrome (HGPS), a mutant form of lamin A accumulates, causing distortion and thickening of the lamina and sequestration of nuclear proteins1. We used metabolic labeling to investigate how protein homeostasis is altered in HGPS and were surprised to find more rapid protein turnover proteome‐wide. We show that global protein synthesis is increased through disrupted nucleolar organization and enhanced ribosome biogenesis. Expression of mutant lamin A disrupts the nucleoplasmic lamin meshwork 2 and depletes heterochromatin marks3, allowing increased transcription of ribosomal RNA and nucleolar expansion. Synthesis of ribosomal protein subunits is also coordinately


upregulated, and more ribosomes are actively engaged in translation in HGPS‐derived cells. Increased flux through the energetically costly processes of ribosome biogenesis and translation is consistent with several previously observed features of HGPS, including a high metabolic rate4, depletion of cellular ATP5, and accumulation of ROS6. Since one of the key roles of the mTOR signaling network is to tune protein synthesis rates, these findings help explain why the mTOR inhibitor rapamycin has shown promise as a treatment for HGPS7 and point towards direct modulation of ribosome biogenesis as another possible therapeutic target. Diminishing protein translation increases lifespan in several systems8. We find a surprising link between the nuclear lamina and regulation of ribosome biogenesis, and show for the first time that over‐production of protein correlates with shortened lifespan in premature aging. References: 1. Goldman, R.D., et al. PNAS 2004. 2. Vidak, S., et al. Genes Dev 2015. 3. Shumaker, D.K., & Goldman, R.D. PNAS 2006. 4. Merideth, M., et al. NEJM 2008. 5. Rivera‐Torres, J., et al. J Proteomics 2013. 6. Kubben, N., et al. Cell 2016. 7. Cao, K., et al. Sci Transl Med 2011. 8. Steffen, K.K., & Dillin, A. Cell Metabolism 2016.

E115 Suppressing ERK1/2 phosphorylation of TAN Line Component FHOD1 rescues defective nuclear movement in fibroblasts expressing muscular dystrophy lamin A mutants. S. Antoku1, W. Wu2, C. Östlund2, H.J. Worman2, G.G. Gundersen1; 1Department of Pathology and Cell Biology, Columbia University, New York, NY, 2Department of Medicine, Columbia University, New York, NY Nuclei in skeletal muscle occupy stereotypic positions that are disrupted in muscular dystrophy. Mice expressing the human disease lamin A variant, lamin A H222P, develop muscular dystrophy and cardiomyopathy and exhibit upregulated ERK1/2 activity (Muchir, A., et al., JCI, 2007). Reducing ERK1/2 activity with MEK1/2 inhibitors ameliorates symptoms in these mice (Muchir, A. et al., HMG, 2009). In NIH3T3 fibroblasts, expression of muscular dystrophy causing lamin A variants disrupts rearward nuclear movement during centrosome reorientation (Folker, E. PNAS, 2011), providing an opportunity to test whether ERK1/2 signaling contributes to nuclear movement defects associated with muscular dystrophy. Treatment of fibroblasts expressing lamin A H222P or several other muscular dystrophy disease variants of lamin A, with MEK1/2 (selumetinib and PD98059) or ERK1/2 (SCH772984) inhibitors rescued the defect in nuclear movement. Consistent with these results, hyperactivation of ERK1/2 activity in fibroblasts by expressing constitutively active MEK1, MEK1 S218D/S222D, inhibited nuclear movement. Inhibiting the peak of ERK1/2 activation with selumetinib during growth factor stimulation of nuclear movement, resulted in earlier initiation and completion of nuclear movement than in untreated controls. Dorsal actin cables and TAN lines that provide the force and coupling for nuclear movement, formed earlier and in greater numbers in ERK1/2 inhibited cells without affecting the rate of actin retrograde. Conversely, hyperactivation of ERK1/2 resulted in the inhibition of TAN line formation without affecting the formation of actin cables. Among TAN line components, we found that FHOD1 and nesprin2G were phosphorylated by ERK1/2. Furthermore, we mapped the phosphorylation sites. Expression of a phosphomimetic mutant of the single conserved ERK1/2 phopshorylation site of FHOD1 inhibited nuclear movement in fibroblasts, whereas expression of the unphosphorylatable FHOD1 mutant rescued the nuclear movement defect in fibroblasts overexpressing lamin A H222P. Importantly, our results establish ERK1/2 phosphorylation of FHOD1 as a key negative regulatory event in nuclear movement and suggest that its upregulation by disease causing lamin A mutants may contribute to muscular dystrophy and cardiomyopathy.


E116 Nucleus‐nucleus interactions are regulated by two distinct genes linked to Emery‐Dreifuss Muscular Dystrophy and Centronuclear Myopathy. M.A. Collins1, T.R. Mandigo1, J.M. Camuglia1, E.S. Folker1; 1Department of Biology, Boston College, Chestnut Hill, MA Syncitia, cells with multiple nuclei, are less prevalent than their mononucleated counterparts. However, there are many examples of syncytia in biology. The developing Drosophila blastoderm and skeletal muscle are commonly studied examples as are less frequently studied examples such as placenta and osteoclasts. Little is known about how the presence of many nuclei in each of these cells impacts their specific biology. Fundamental to understanding the biology of syncytia is understanding how and when each nucleus interacts with the other nuclei. In muscle, the rapid movement of nuclei to the cell center to join already incorporated nuclei suggests attractive interactions. Conversely, there are repulsive interactions between nuclei later in muscle development which are utilized to keep nuclei spaced distant from one another. Because muscle exhibits both attractive and repulsive interactions that are separated in developmental time, it is an ideal syncytia to identify the mechanisms and functions of nucleus‐nucleus interactions. In Drosophila, movement of nuclei in muscle occurs in three distinct stages. After the completion of fusion, nuclei separate into two clusters, one positioned dorsally and the other positioned ventrally. Each cluster then moves directionally toward its respective pole with fewer than 2% of nuclei change directions. During directional movement, the nuclei remain in tightly associated clusters. Finally, after nuclei reach the muscle end they then move back into the cell center and maximize the distance between adjacent nuclei. Disruption of two different genes, bocksbeutel (dEmerin) and klarsicht (dNesprin), blocked initial separation of nuclei into two distinct clusters. Live‐embryo time‐lapse microscopy demonstrated occasionally a nucleus could escape the cluster. These nuclei moved directionally as in controls and moved more quickly than controls indicating that the force generating machinery and the spatial cues were functional. Thus, bocksbeutel and klarsicht were necessary to regulate the necessary repulsive interactions between nuclei. Conversely, disruption of Amphiphysin, inhibited the attractive interactions between nuclei. This was evident by the regular dissociation of nuclei from clusters and the presence of these nuclei in the center of the muscle. Together these data indicate that nuclei do indeed have attractive and repulsive interactions in skeletal muscle and that these interactions are regulated by distinct proteins at specific developmental times.

E117 LINC complexes support epidermal keratinocyte adhesion and differentiation. R.M. Stewart1, D. Reyes Aguilar2,3, V.J. Horsley3,4, M.C. King1,3; 1Cell Biology, Yale School of Medicine, New Haven, CT, 2University of Maryland, Baltimore County, Baltimore, MD, 3Sackler Institute for Biological, Physical and Engineering Sciences, New Haven , CT, 4Molecular, Cell and Developmental Biology, Yale University, New Haven, CT The mechanical integrity of tissues is supported by the cytoskeleton, which is organized at the multicellular scale, and is coordinated by intercellular adhesions. Nuclear envelope‐spanning LINC (Linker of Nucleoskeleton and Cytoskeleton) complexes integrate the nucleus into this mechanical network, while also providing a mechanism to allow for force transmission from adhesions and the cytoskeleton to the nuclear interior. Epidermal defects are frequently observed in diseases of the nuclear envelope (“nuclear envelopathies” and “laminopathies”), but their origin remains elusive.


We (and others) have shown that nuclear position responds to formation of cell‐cell adhesions in epithelial cells, suggesting a link between intercellular junctions and the nucleus. Further, we have shown that mice deficient for the LINC complex component SUN2 exhibit defects in hair follicle structure and integrity, in part due to defective cell‐cell adhesion. Epithelial sheets generated from SUN2‐null cells are also mechanically weak, but the mechanisms underlying this defect are poorly defined. Here, we have further investigated the mechanisms underlying these in vivo and in vitro defects using a model for the complete ablation of LINC complexes (achieved by use of a SUN1/SUN2 double knock‐out (dKO) mouse model). We observe two types of epidermal defects in this model system. First, we find that SUN dKO cells fail to form intact mechanical units, leading to defective epithelial sheet formation in vitro and defective tissue organization in vivo. We link this in vitro defect to a failure to reorganize the actomyosin network to close “wounds” in the sheet, which likely underlies the mechanical defects reported previously. Second, ablation of LINC complexes leads to precocious keratinocyte differentiation as revealed by transcriptome analysis of cultured keratinocytes in vitro and marker staining in vivo. This supports a model in which the association of basal keratinocytes with the basal lamina drives a LINC‐complex dependent signal to repress differentiation until cells exit the basal layer, where they take on their suprabasal fate. Taken together, our results suggest that the LINC complex acts to both dynamically organize the cytoskeleton and to mechanically communicate extracellular signals to the nuclear interior to support epidermal differentiation and homeostasis.

Microsymposium 14: Trafficking Dynamics and Imaging

E118 Cargo sorting during protein secretion. J. Von Blume1, B. Blank1; 1Molecular Medicine, Max Planck Institute of Biochemistry, Martinsried, Germany Sorting of proteins for transport to the cells surface or for secretion from cells is essential for the extracellular activities of proteins like insulin, collagens, antibodies, surface receptors, secreted proteases, and GPCRs to name a few. A major regulator of intracellular protein distribution is the trans‐Golgi network (TGN) that sorts secretory proteins into specific carriers to transport them to their final destination. The sorting of lysosomal hydrolases at the TGN by the mannose‐6‐phosphate receptor is well established. However, the sorting mechanism for secreted proteins remains poorly understood. In recent years we have identified major components required for secretory cargo sorting at the TGN. These components include F‐actin, cofilin, the Ca2+ ATPase SPCA1 at the cytosol and Ca2+ and the Golgi resident protein Cab45 in the lumen of the TGN. Although Ca2+ and Cab45 are known to play a role in sorting the mechanism remained unclear. Here, to gain a further mechanistic inside into this process we used biochemistry to reconstitute, using purified components, the sorting of cargo molecules. We show that Cab45 changes its conformation and oligomerizes in a Ca2+ dependent manner. These oligomers specifically recruit cargo molecules such as LysozymeC and cartilage oligomeric protein (COMP) in vitro. The CRISPR/Cas9 technology allowed us to confirm that Cab45 acts as a Ca2+‐dependent cargo sorter in living cells. Furthermore, super‐resolution microscopy revealed that the organization of Cab45, SPCA1 and cargo occurs in specific domains of the TGN. We believe that Cab45 is a sorter of secretory protein and may represent a unique way to export cargo independent of a bona fide cargo receptor at the TGN.


E119 Assembly and activation of dynein–dynactin by the cargo adaptor protein Hook3. C.M. Schroeder1, R.D. Vale2; 1Division of Basic Sciences, Fred Hutchinson Cancer Research Center, Seattle, WA, 2Dept. of Cellular and Molecular Pharmacology, University of California, San Francisco, San Francisco, CA Metazoan cytoplasmic dynein moves processively along microtubules with the aid of dynactin and an adaptor protein that joins dynein and dynactin into a stable ternary complex. We examined how Hook3, a cargo adaptor involved in Golgi and endosome transport, forms a motile dynein–dynactin complex. We show that the conserved Hook domain interacts directly with the dynein light intermediate chain 1 (LIC1) subunit. By solving the crystal structure of the Hook domain and using structure‐based mutagenesis, we identify two conserved surface residues that are each critical for LIC1 binding. Hook proteins with mutations in these residues fail to form a stable dynein–dynactin complex, revealing a crucial role for LIC1 in this interaction. We also identify a region of Hook3 specifically required for an allosteric activation of processive motility. Our work reveals the structural details of Hook3’s interaction with dynein and offers insight into how cargo adaptors form processive dynein–dynactin motor complexes.

E120 SNARE proteins entropically expand membrane fusion pores. S. Thiyagarajan1, Z. Wu2,3, O. Bello3, S. Auclair3, W. Vennekate3, S.S. Krishnakumar3, E. Karatekin2,3,4,5, B. O'Shaughnessy6; 1Physics, Columbia University, New York, NY, 2Cellular and Molecular Physiology, Yale University, New Haven, CT, 3Nanobiology Institute, Yale University, West Haven, CT, 4Molecular Biophysics and Biochemistry, Yale University, New Haven, CT, 5Fondamentales et Biomédicales, Centre National de la Recherche Scientifique , Paris, France, 6Chemical Engineering, Columbia University, New York, NY Secretion of neurotransmitters and hormones involves creation of a nanoscale membrane fusion pore through which vesicle contents are released. To establish the fusion pore, cells use a fusion machinery whose core consists of the SNARE proteins. Cognate SNAREs (v‐ and t‐SNAREs) in apposing membranes bind in a zippering action to form SNARE complexes, drawing membranes together, but beyond this the mechanism of SNARE‐mediated fusion is unknown. Neither is the mechanism known that controls the fate of the fusion pore once formed: pores fluctuate and can dilate into fully developed pores (“full fusion”) or reseal irreversibly after releasing a fraction of the contents (“kiss and run”). Here we combine mathematical modeling and single pore experimental measurements to characterize fusion pores and the role of SNAREs. The model is motivated by experiments in which we measured the conductance of single fusion pores connecting nanolipoprotein particles (NLPs), reconstituted with complementary v‐SNAREs (vNLPs), to voltage‐clamped membrane patches of cells ectopically expressing flipped neuronal/exocytotic t‐SNAREs. From conductance traces we obtained the distribution of fusion pore sizes versus number of v‐SNAREs reconstituted into the vNLPs. Even one SNARE was sufficient to form a fusion pore, but as the number of SNAREs per NLP face increased from 0 to 15 the mean pore size increased and the distribution of pore sizes became dramatically broader. To understand these results we mathematically modeled the pore as a simple toroid, connecting the NLP to a planar membrane. The SNAREs are fully or partially zippered: zippered SNAREs occupy a ring at the fusion pore waist, while unzippered SNAREs are free to roam. We calculated the distribution of pore sizes versus number of SNAREs, assuming sufficient time for equilibration. Model predictions for pore size distributions were fitted to experimental distributions for different numbers of SNAREs, using the


TMD zippering energy and membrane tension as fitting parameters. Best fit values were ~8 kT and ~0.5 pN nm‐1 respectively. The model showed that introduction of SNAREs caused fusion pores to dilate dramatically. The driving force was the TMD and linker domain zippering energy, which encouraged more SNAREs to zipper at the fusion pore waist. Entropic (crowding) effects of the zippered TMDs at the waist then expanded the pore, since a bigger pore gave the zippered SNAREs more entropy. Our results suggest that SNARE proteins, in addition to creating fusion pores, may critically regulate them post‐fusion: one or two SNAREs suffice to create a pore, but several SNAREs are required to fully dilate the pore for efficient contents release during exocytosis.

E121 Two distinct modes of exocytosis revealed by automated computer‐vision image analysis. F.L. Urbina1, S. Gomez2, S.L. Gupton1; 1Cell Biology and Physiology, University of North Carolina: Chapel Hill, Chapel Hill, NC, 2Biomedical Engineering, University of North Carolina: Chapel Hill, Chapel Hill, NC During development, neurons acquire a unique, highly polarized morphology that requires rapid expansion of the plasma membrane. This increase in membrane surface area occurs primarily through secretion, in which exocytic vesicles dock with the plasma membrane and SNARE complex formation between v‐SNAREs in the vesicle membrane and t‐SNAREs in the plasma membrane lead to lipid fusion. Two modes of exocytic vesicle fusion have been suggested. Full‐vesicle fusion (FVF), the classic mode of exocytosis, is characterized by the rapid expansion of the fusion pore, which leads to secretion of vesicular cargo into the extracellular space and insertion of vesicular membrane into the plasma membrane. In the kiss‐and‐run mode of fusion (KNR), a transient fusion pore opens leading to expulsion of vesicular cargo, but then the pore closes without fully dilating, allowing the vesicle to retain its identity. Presumably FVF would add to the expanding plasma membrane of a developing neuron, whereas KNR would not. Differentiation between these modes of exocytosis is complicated due to the size and speed of the fusion events. To visualize fusion events in developing neurons, we imaged the v‐SNARE VAMP2 luminally tagged with a pH‐sensitive variant of GFP (pHluorin) by total internal reflection fluorescence microscopy. VAMP2‐pHluorin exocytic events are characterized by a rapid appearance of fluorescence upon initial opening of the fusion pore, followed by a slower decay of fluorescence. Here we report a high‐throughput, unbiased detection and analysis of these exocytic events using a newly developed MATLAB and R based software. All detected events were sensitive to tetanus toxin, which cleaves VAMP2, preventing SNARE complex formation and fusion. The peak fluorescence of vesicle fusion events was unimodal, indicating vesicles contained similar VAMP2‐pHluorin compositions. However, the half‐life of fluorescence decay of events was unexpectedly bimodal, consistent with two modes of exocytosis. By quantitatively assessing the fate of VAMP2‐pHluorin after the initial fusion, we were able to separate the exocytic events into two populations corresponding to the bimodal peaks in the distribution of event half‐life. The longer lived events were characterized by VAMP2‐pHluorin diffusion within the plasma membrane, consistent with FVF, whereas the shorter lived events did not exhibit diffusion, consistent with KNR. Using this software we demonstrate that the axon guidance cue netrin‐1 increases the frequency of FVF exocytosis in developing neurons, consistent with the effect of netrin‐1 on increasing plasma membrane area during axon branching. Further, we identify an E3 ubiquitin ligase that biases the mode of exocytosis toward FVF and is required for axon branching.


E122 Running against time: a kinetic study of endophilin. K.R. Poudel1, A. Su1, T. Ho1, J. Bai1; 1Basic Sciences, Fred Hutchinson Cancer Research Center, Seattle, WA Time is a limiting factor for many cellular processes. Neurons are particularly sensitive to time as they are specialized for rapid communication. For instance, synaptic vesicle endocytosis occurs between tens of milliseconds and a couple of seconds to recycle vesicle membranes from the plasma membrane. By contrast, endocytosis in non‐neuronal cells occurs slowly, often taking minutes to complete. While a similar set of protein machinery is involved in both slow and rapid endocytosis, it is unknown whether the difference in time duration impacts the action of these proteins. To address this issue, we have investigated the dynamic action of endophilin, because it is a critical protein for both rapid and slow endocytosis. The ability of endophilin to shape membranes is well‐documented. Biochemical studies have shown that endophilin converts flat membranes into bilayer tubules and vesicles in vitro. However, how it acts within physiological time windows is not known. In order to address this issue, we developed kinetic assays to monitor how endophilin binds membranes, induces local membrane deformation, and produces large‐scale changes of membrane morphology. We find that endophilin undergoes a sensing‐to‐bending transition, a previously unknown feature of membrane‐bending proteins. This transition takes about 30 milliseconds. As a result, endophilin does not immediately deform membranes upon its arrival. The bending function is activated only after two amphipathic helices become ready for membrane insertion. Surprisingly, we find that endophilin is fairly slow in producing large‐scale membrane changes. While local membrane deformation occurs rapidly, it takes minutes for endophilin to induce the extensive morphological changes for membrane tubulation and vesiculation. Therefore, we propose that endophilin’s action depends on how much time it has on endocytic membranes. In other words, it is the kinetic property, rather than the thermodynamic equilibrium, that dominates endophilin action during rapid endocytosis. Our step‐by‐step analyses have shown for the first time the order of sensing, wedging, and bending actions taken by endophilin to sculpt membranes. These findings initiate a new direction to address the temporal limit of membrane‐remodeling proteins for various biological activities.

E123 Functional recruitment of dynamin requires multimeric interactions with SH3 domain containing proteins for efficient endocytosis. D. Perrais1, M. Rosendale1, T. Van1, D. Grillo‐Bosch1, I. Gauthereau1, D. Choquet1, M. Sainlos1; 1Interdisciplinary Institute for Neuroscience, CNRS and University of Bordeaux, Bordeaux, France Clathrin mediated endocytosis is controlled by a number of cytosolic parameters as well as extensive protein‐protein interactions. An essential interaction occurs between the proline/arginine rich domain of dynamin and the Src homology domain 3 (SH3) of various proteins, including amphiphysin, endophilin, intersectin and syndapin. Indeed, disruption of these interactions by isolated SH3 domains or a 15‐mer peptide named D15, bearing a PxRPxR motif, blocks endocytosis in living cells. D15 has become a popular tool to rapidly disrupt dynamin meditated endocytosis when injected in cells. It is however used at high concentrations (1‐2 mM) and its specificity for various SH3 domain containing proteins has not been assessed. Moreover, dynamin and associated proteins are probably concentrated in ring‐like structures around the neck of nascent vesicles, bringing multiple interaction sites in close


proximity, a mechanism still poorly understood in living cells. To gain insight into the parameters of such interactions, we took a multidisciplinary approach using peptide design and characterisation of binding with pull‐down and surface plasmon resonance assays, quantitative live cell imaging assays of endocytosis (Merrifield et al. Cell 2005) and rescue of endocytosis with dynamin mutants in dynamin triple knock‐out (TKO) cells (Park et al. J Cell Science 2013). We found that elongated and divalent D15‐derived peptides bind more strongly to multimers of SH3 domains and are more effective at blocking endocytosis than D15. Moreover dynamin mutated in the D15 motif could partially rescue dynamin activity, and an additional mutation in flanking PxxPxR made dynamin completely inactive in TKO cells. From this data we conclude that dynamin interacts with multiple SH3 domain containing proteins, most likely amphiphysin, to be recruited efficiently to the neck of nascent vesicles and mediate vesicle scission in cells.

E124 The dynamics of chemokine receptor CXCR4 endocytosis. L.K. Rosselli‐Murai1, M. DeNies2, M. White3, A.P. Liu1,2,3; 1Department of Mechanical Engineering, University of Michigan, Ann Arbor, MI, 2Department of Molecular and Cellular Biology, University of Michigan, Ann Arbor, MI, 3Department of Biomedical Engineering, University of Michigan, Ann Arbor, MI Chemokine receptor CXCR4 signaling has been reported to st‐‐‐‐‐imulate metastasis of several tumors, with carcinoma‐associated fibroblasts being an important source of the chemokine CXCL12 (CXCR4 agonist). Upon CXCR4 activation, a number of signaling pathways are activated ultimately leading to enhanced cell migration. CXCR4 regulation by clathrin‐mediated endocytosis (CME) is an integral part of CXCR4 signaling, however very little is known about how clathrin‐coated pits (CCPs) mediate CXCR4 internalization and integrate with downstream signaling cascade. In order to understand the relationship between CCP dynamics and CXCR4 signaling and function, we used live cell dual color total internal reflection fluorescence (TIRF) microscopy. We have established an approach where CXCR4 is site‐specifically biotinylated by an endoplasmic reticulum‐expressed E. coli biotin ligase BirA. Using a bicistronic tetracycline‐regulatable expression construct, we controlled the levels of biotinylated CXCR4 which are subsequently labeled by fluorescent streptavidin. To validate the functionality of streptavidin‐labelled CXCR4, we performed an in vitro wound healing assay and observed that addition of CXCL12 accelerated cell migration while the presence of CXCR4 antagonist AMD3100 suppressed CXCL12‐induced cell migration. Combining TIRF microscopy with unbiased computational analysis, we quantitatively detected, tracked, and analyzed the dynamics of CXCR4 endocytosis via CCPs in live cells expressing fluorescently tagged CXCR4 and EGFP‐clathrin light chain. Our results show that CXCR4 is constitutively internalized through CCPs both in the presence or absence of CXCL12. However, the CCP lifetime analysis revealed a remarkably higher accumulation of CXCR4 in short‐lived CCPs upon CXCL12 stimulation. It has been well stablished that short‐lived CCPs, also known to be functionally abortive, are a distinct subpopulation of CCPs that failed to internalize to form a vesicle. Although CXCL12 is thought to be important to the recruitment of CXCR4 into CCPs leading to its internalization, it is not clear whether the clathrin structures required for signaling are the same as CCPs that function in endocytosis, given the heterogeneous nature of CCPs in size, lifetime, and composition at the cell membrane. Interestingly, our data also showed that while CXCL12 addition reduced CCPs initiation density, it significantly increased the frequency of short‐lived CCPs, further suggesting a potential role of short‐lived CCPs and signaling activation. The combination of our site‐specific receptor labeling and live cell imaging of receptor and CCPs during agonist addition represents a powerful approach to reveal mechanistic details of CXCR4 endocytic signaling network through CCPs.


E125 High‐throughput superresolution imaging of clathrin‐ and actin‐mediated endocytosis in yeast. M. Mund1, A. Picco2, M. Kaksonen2, J. Ries1; 1Cell Biology and Biophysics, European Molecular Biology Laboratory, Heidelberg, Germany, 2Department of Biochemistry, University of Geneva, Geneva, Switzerland Clathrin‐mediated endocytosis is a key process to take up cargo from the cell surface. More than 50 proteins form the endocytic machinery, which invaginates the plasma membrane and pinches off a vesicle. Due to its complexity and small size below the diffraction limit, the structural organization of endocytic proteins in situ is largely unknown. The endocytic machinery has to provide the force necessary to invaginate the membrane. In yeast, polymerizing actin provides this force with a high efficiency and remarkable temporal and spatial regularity. In mammalian cells, actin polymerization appears to be most required under conditions of increased membrane tension. The spatio‐temporal control of actin polymerization is closely linked to the structural organization of the endocytic proteins, which control actin filament nucleation, cross‐linking, capping and disassembly. Here, we use high‐throughput single‐molecule localization microscopy (PALM/STORM) to quantitatively analyze how 17 endocytic proteins are organized at endocytic sites in Saccharomyces cerevisiae. Achieving the highest possible resolution with this method requires static structures, which is why cells are fixed and time resolution is lost. This leads to heterogeneity among the imaged endocytic sites, which complicates their interpretation. To determine the endocytic time point, we image a second reference structure and integrate our images with data from live‐cell fluorescence imaging. This allows us to directly visualize the structural relation between polymerized actin and actin interacting proteins to ultimately understand how actin polymerization is regulated in situ. We discovered that several actin nucleation promoting factors and their regulators show a striking ring‐shaped organization around the clathrin coat. The ring organization is established prior to actin polymerization. This lateral pre‐patterning of the endocytic site provides an elegant explanation for an efficient force generation by the actin machinery and force transfer to the membrane.

E126 Using quantitative super‐resolution microscopy to optimize multivalent Her2‐targeting ligands. O. Golfetto1, R. Jorand1, S. Biswas1, D.L. Wakefield1, S.J. Tobin1, K.N. Avery1, K. Meyer1, J.C. Williams1, T. Jovanovic‐Talisman1; 1Molecular Medicine, Beckman Research Institute‐City of Hope, Duarte, CA Monoclonal antibodies (mAbs) are critical components of many current and next generation cancer therapeutics. In a number of clinical studies, combination of mAbs targeting the same antigen but different epitopes (e.g., trastuzumab and pertuzumab) extends progression‐free survival in breast cancer patients1. According to our hypothesis, the mechanism underlying such improved clinical outcomes involves antibody‐mediated clustering of the surface‐expressed antigens: this reorganization ultimately leads to more comprehensive signaling inhibition, down‐regulation of the targeted receptor, and/or improved antibody‐dependent cell‐mediated cytotoxicity. Herein, we examined the ability of therapeutic mAbs to affect clustering of human epidermal growth factor receptor 2 (Her2). Her2 clustering was quantitatively interrogated with pointillistic super‐resolution microscopy techniques2 – which can measure the distribution of proteins with nanometer scale precision and single molecule sensitivity. We established the organization of Her2 tagged with photoactivatable GFP (paGFP)


overexpressed in breast cancer MDA‐MB‐468 cells and determined that Her2‐paGFP largely behaves like endogenous protein. Importantly, treating these cells with therapeutic mAbs profoundly affected Her2 organization in a valency dependent manner: while Her2 clustering was affected minimally by the monovalent Fab domain of trastuzumab, divalent trastuzumab or pertuzumab had a distinct and pronounced clustering effect. Critically, a combination of divalent trastuzumab and divalent pertuzumab produced the highest level of clustering. To demonstrate that multivalent ligands can effectively replace a second antibody, we leveraged recently discovered meditope technology3. Meditopes are cyclic peptides that can bind tightly to a broad array of re‐engineered, meditope‐enabled mAbs (memAbs) without affecting antigen binding; thus they are a rapid and site‐specific means to manipulate and/or add functionality to mAbs. Compared to the clustering induced by the combination of trastuzumab and pertuzumab, a tetravalent meditope Fc analog combined with memAb trastuzumab not only produced greater clustering but was also more effective at inhibiting EGF‐dependent activation of Her2, EGFR and Akt. Thus, the multivalent meditope ligands and memAbs efficiently change geography and activation of Her2 receptor; and they may represent a highly effective approach for cancer therapy. 1. Swain, S. M., Baselga, J. et al. The New England journal of medicine 372, 724, (2015). 2. Sengupta, P., Jovanovic‐Talisman, T. et al. Nat. Methods 8, 969, (2011). 3. Donaldson, J. M., Zer, C. et al. Proc Natl Acad Sci U S A 110, 17456, (2013).

Microsymposium 15: Cell Shape and Signaling

E127 Dissecting the role of RhoA, ‐B and ‐C during host‐pathogen interaction. J. Halfen1, J. Kollasser1, L. Gröbe2, P. Hagendorff3, R. Geffers3, A. Steffen1, C.H. Brakebusch4, K. Rottner5, T.E. Stradal1; 1Cell Biology, Helmholtz Centre for Infection Research, Braunschweig, Germany, 2Flow Cytometry and Cell Sorting, Helmholtz Centre for Infection Research, Braunschweig, Germany, 3Genome Analytics, Helmholtz Centre for Infection Research, Braunschweig, Germany, 4Biotech Research and Innovation Centre, University of Copenhagen, Copenhagen, Denmark, 5Molecular Cell Biology, Technical University Braunschweig, Braunschweig, Germany Some enteropathogenic bacteria (e.g. Salmonella enterica Typhimurium and Shigella flexneri) ensure their protection from the immune system by invasion of host cells. This uptake is achieved through manipulation of mammalian Rho GTPases by specific virulence factors (e.g. SopB, SopE and SopE2 in S. enterica) which often activate or even directly mimic guanine nucleotide exchange factors (GEFs). Rho GTPases organize the actin cytoskeleton of mammalian cells and are thus prominently involved in processes such as cell migration or division. The GTPase Rac1, for instance, is essential for membrane ruffling and hence implicated in triggered bacterial invasion of e.g. S. enterica or S. flexneri. In contrast, RhoA signals to myosin II mediated contraction and stress fiber formation. Interestingly, for S. enterica a new invasion pathway involving RhoA has recently been identified. However, the details of this pathway are still not fully elucidated and thus explored in future studies. Furthermore, we are investigating whether S. flexneri also employs RhoA for cell invasion as it harbors a bacterial GEF for this GTPase, namely IpgB2. Finally, RhoA has two paralogs in mammals, RhoB and ‐C, the role of which during bacterial infections is still not clearly defined. We here set out to determine the role of RhoA, ‐B and ‐C for cell invasion by S. enterica and S. flexneri. To do so, the murine fibroblast cell line NIH‐3T3 was employed for generating isogenic lines disrupted in the genes encoding all three Rho proteins, RhoA, ‐B and‐C, either individually or in various combinations, using the CRISPR/Cas9 technology. All cell lines were probed for the absence of respective protein by


Western Blotting and genomically sequenced in the targeted locus. Validated cell lines were subjected to GeneChip analyses for evaluation of expression profiles. Finally, invasion rates of S. enterica and S. flexneri were determined using gentamycin protection assays. Double but not single knockout (KO) cell lines displayed a strongly altered morphology as compared with wildtype cells, including long extensions, multiple nuclei and increased cell size. In addition, invasion rates were reduced in all KO clones compared with wildtype cells, with the greatest reduction observed for S. flexneri in the RhoA KO cells by approx. 80%. In future work, the differential contributions of RhoA, ‐B, and ‐C to establishment and maintenance of cell morphology, signaling, gene expression and invasion by specific pathogens will be determined.

E128 Regulatory mechanism of JMY in actin‐based autophagosome movement. X. Hu1, R. Mullins1; 1Cellular and Molecular Pharmacologyy, University of California, San Francisco, San Francisco, CA Junction‐Mediating and regulatorY protein (JMY) was initially discovered as a co‐activator of p53‐dependent apoptosis, which forms a ternary complex with other p53 regulators: p300 and strap. Our previous work revealed that JMY is also a unique regulator of the cytoskeleton that promotes assembly of actin filaments in two ways: (1) by stimulating activity of Arp2/3 complex and (2) by directly nucleating filaments via three actin‐binding, WH2 domains. We now address two basic questions: (1) How is JMY’s actin nucleation activity regulated? (2) What is the role of JMY in the cytoplasm? In live cells we find that JMY forms puncta that migrate on growing actin comet‐tails. Similar to whamm, JMY‐positive puncta also contain autophagy markers LC3 and DFCP. Remarkably, although LC3 does not affect the nucleation activity of the C‐terminal WWWCA domain of JMY, it enhances the actin nucleation activity of the full‐length protein. This enhancement is produced by interaction with an N‐terminal region of JMY, which contains a cryptic actin nucleation activity. In addition, we discovered that some JMY puncta are immobile and that these non‐moving puncta co‐localize with strap but not LC3. Biochemical studies confirmed that strap binds both WWWCA domain and N‐terminus of JMY and inhibits their actin nucleation activity. These results reveal previously unknown positive (LC3) and negative (strap) regulators of actin nucleation by JMY. LC3 accelerates JMY‐positive autophagosome propulsion whereas strap prevents JMY puncta to nucleate actin and to further form actin‐based comet tail.

E129 Role of cytoskeleton in morphological changes of blood platelets. A. Mathur1, S. Dmitrieff1, S. Correia1, R. Gibeaux1,2, I. Kalinina1, T. Quidwai1, J. Ries1, F. Nedelec1; 1Cell Biology and Biophysics, European Molecular Biology Laboratory, Heidelberg, Germany, 2Department of Molecular Cell Biology, University of California, Berkeley, CA Platelets are 2‐3μm sized discoid cell fragments that help maintain haemostasis. Activation of platelets, caused by chemical factors released upon injury, results in drastic morphological changes brought about by a repertoire of cytoskeletal proteins. During the first step of platelet activation, a microtubule marginal band that runs along platelet periphery undergoes coiling while the platelet changes from a disc to a sphere shape. Both actin and microtubules are implicated in this process but their respective roles remain elusive. We use a combination of experimental and quantitative analysis techniques to assess mechanical properties of this system by direct measurement of the marginal band morphology in mice and humans.


The structure of the band was analyzed using electron tomography, which provided information about individual microtubules. The dynamics of coiling process was observed with live cell fluorescence microscopy in conjunction with a microfluidic system. This enabled visualization of marginal band shape change in response to platelet agonists or cytoskeletal perturbations. A large population of platelets was analyzed to infer the intrinsic variations in properties of the marginal band. Additionally, we developed a mathematical model for describing marginal band mechanics and simulated the coiling process with the software Cytosim. Our approach enabled us to get novel insight into mechanics of resting and coiling marginal band. First, we show that resting platelet size scales with its polymerized tubulin content, in agreement with a simple viscoelastic model of the marginal band. In addition, our measurements of the marginal band coiling, as triggered by platelet agonist adenosine diphosphate, indicate a short, elastic, and a long, viscous, timescale of response. We also found marginal band coiling to be strongly dependent on presence of polymerized actin. Furthermore, our data suggests that platelets with longer marginal bands have a higher propensity to coil, consistent with our theoretical model described by antagonistic forces of cortical actin tension and marginal band rigidity. We thus established an experimental and theoretical framework to analyze platelet size and their behavior during activation. This is of particular interest as platelet size is a parameter often altered in pathological condition.

E130 Keratins regulate β‐cell mitochondrial morphology, motility and homeostasis. J.S. Silvander1, S.M. Kvarnström1, A. Kumari‐Ilieva1, C.M. Alam1, D.M. Toivola1; 1Cell Biology, Åbo Akademi University, Turku, Finland Type 1 diabetes (T1D) is an autoimmune disease that in most cases develops at young age, and is manifested as insufficient insulin production from β‐cells in the endocrine pancreatic. The molecular mechanisms in the β‐cells that underlie T1D development are still fairly unknown. We have previously shown that keratin (K) intermediate filaments K8 and K18 are the primary keratins in β‐cells and that absence of K8 leads to mislocalization of the glucose transporter, GLUT2, reducing glucose signaling in these cells. Moreover, the total insulin amount is reduced and insulin vesicles are irregular in the K8 knockout (K8‐/‐) mouse pancreas, suggesting a role for K8 in β‐cell glucose metabolism. Since mitochondria are essential in the process leading from glucose stimulation to production of insulin in β‐cell, the aim was to investigate the involvement of keratins in beta‐cell mitochondrial morphology and dynamics. Transmission electron microscopy analysis showed that the mitochondria in β‐cells in vivo from K8‐/‐ mice are rounder in shape with diffuse and less defined cristae, compared to more elongated, electron dense mitochondria in wild type K8+/+ β‐cells. Morphometry‐analysis shows that the number of mitochondria is also increased in K8‐/‐ β‐cells. Moreover, 3D‐reconstruction of K8‐/‐ β‐cells, using TOM20 as a mitochondrial marker, revealed a more fragmented mitochondrial network compared to K8+/+ β‐cells. Interestingly, in the absence of K8/K18 the levels of the mitochondrial fusion protein mitofusin 2 (MFN2) is decreased, and the cytolinker trichoplein (TCHP), known to bind both K8/K18 and MFN2, is decreased half‐way in β‐cells. Furthermore, the total protein amount of cytochrome c and COX IV is decreased in the K8‐/‐ β‐cells, which could be a result of the less defined mitochondrial cristae and a disturbance in mitochondrial homeostasis. Mitochondrial motility studies of isolated K8‐/‐ β‐cells lead to an increase in mitochondrial motility, compared to K8+/+ mitochondria. Taken together, β‐cell K8/K18 maintain mitochondrial morphology, motility and network homeostasis, possibly through TCHP, and may this way modulate beta‐cell insulin secretions.


E131 SHANK3 Structure Reveals a RAS‐associated Domain Regulating Integrin Activation. J. Lilja1, T. Zacharchenko 2, M. Georgiadou1, H. Kreienkamp3, I. Barsukov2, J. Ivaska1; 1Turku Centre for Biotechnology, University of Turku, Turku, Finland, 2University of Liverpool , Liverpool, United Kingdom, 3University Medical Center Hamburg‐Eppendorf, Institute for Human Genetics , Hamburg, Germany SHANK3, a synaptic scaffold protein, is widely expressed outside of the central nervous system with predominantly unknown function. In humans, SHANK3 missense mutations are associated with autism spectrum disorders. However, how these mutations influence SHANK3 function remains unclear. Here, we identify SHANK1 and SHANK3 as inhibitors of integrin activation. By solving the structure of the N‐terminal region of SHANK3, we find that the SPN‐domain is an unexpected Ras‐association domain with high affinity for GTP‐bound Ras and Rap G‐proteins. In cells, SHANK3 prevents G‐protein‐mediated integrin activation by sequestering active Rap1 and R‐Ras. Autism‐related mutations within the SHANK3 SPN‐domain (R12C and L68P) disrupt G‐protein interaction and fail to counteract Rap1 signaling and integrin activation. Consistently, SHANK3 silencing triggers increased Rap1 activity on the plasma membrane, cell spreading, migration and invasion. Altogether, we establish SHANKs as critical regulators of integrin‐dependent processes and uncover the relevance of SHANK3 autism‐related mutations at the molecular level.

E132 β2 integrins mediate the formation of a focal adhesion‐like cytoskeleton and signaling platform during phagocytosis. V. Jaumouillé1, T. Liu2, E. Betzig2, C.M. Waterman1; 1National Heart, Lung, and Blood Institute, National Institutes of Health, Bethesda, MD, 2Janelia Research Campus, Howard Hughes Medical Institute, Ashburn, VA Professional phagocytes are immune cells specialized in the uptake and clearance of large particulate material including infiltrated microbes, apoptotic cells and debris. αMβ2 and αXβ2 integrins (also called complement receptor 3 and 4) are highly expressed in macrophages, neutrophils and dendritic cells, and are thought to be the main phagocytic receptors for many pathogens, and to mediate the clearance of the billions of cells that are turned over every day. In other contexts, such as cell adhesion and migration, integrin functions are intimately linked to the dynamics of the actin cytoskeleton. For instance, at the focal adhesion, integrins are connected to the actin cytoskeleton through a molecular clutch, and the retrograde flow of actin is thought to be critical to switch integrins to a high affinity conformation and to generate the traction force exerted on the substrate. Although it is established that phagocytosis requires actin, the dynamics of actin during β2 integrin‐dependent phagocytosis remains largely unknown. We sought to characterize the molecular requirements for actin dynamics during β2‐dependent phagocytosis and to test the hypothesis that actin flow along the target particle is critical for integrin binding, outside‐in signaling, and particle engulfment. Using live cell imaging, we observed that engagement of complement‐coated particles lead to a dramatic reorganization of the actin cytoskeleton, characterized by the formation of large ruffles that wrap around the particle, whereas completion of uptake is accompanied by a burst of actin polymerization and contraction of the cortex. Bursts of actin polymerization and ruffle formation coincide with the recruitment of the Arp2/3 complex, which is required for particle internalization. In addition, phagosome formation lead to the assembly of focal adhesion‐like signaling platforms at the phagocytic cup, characterized by the recruitment of Vinculin, α‐Actinin, Zyxin, Arp2/3 and Myosin II, as well as tyrosine phosphorylation of


Paxillin, FAK and Syk. Contrary to previous reports, we found that Src family kinases, Syk and FAK/Pyk2 activities are important for particle engulfment. Tyrosine kinases are required for the recruitment of Vinculin, which reinforce the molecular clutch that couples integrins with the actin cytoskeleton, independently of Myosin II. Our observations indicate that professional phagocytes co‐opt the focal adhesion‐lamellipodium machinery to engulf large particulate material via broad spectrum β2 integrins.

E133 Identifying signaling connections in cancer cell motility using partial correlation analysis of simultaneously imaged Rho GTPase and RhoGEF activities. D.J. Marston1, M. Vilela2, G. Glekas1, G. Danuser2, J. Sondek1, K.M. Hahn1; 1Dept of Pharmacology, UNC Chapel Hill, Chapel Hill, NC, 2Department of Cell Biology, University of Texas Southwestern Medical Center, Dallas, TX RhoGEFs directly activate Rho family GTPases leading to the regulation of cell shape through alterations in cytoskeletal dynamics. Consequently, cellular morphodynamics requires the exquisite spatiotemporal activation of diverse RhoGEFs acting in concert. Here, we describe a general strategy to produce RhoGEF biosensors. A pair of fluorescent proteins was inserted in a region of the RhoGEF where relief of autoinhibition altered FRET between the fluorescent pair. The design was optimized using high‐content screening to produce Vav2 and Tiam1 biosensors where activation causes changes in FRET emission ratios of 210% and 75%, respectively. The optimized Vav2 biosensor is responsive to Src kinase and epidermal growth factor receptor (EGFR) while the Tiam1 biosensor reports activation by Src, Rap1 and EGFR. The design strategy works for diverse RhoGEFs, including: p115RhoGEF, B‐Pix, Tim and Asef. We used the RhoGEF biosensors together with updated versions of our Rho family GTPase biosensors to examine the relative dynamics of GTPases and RhoGEFs during constitutive migration of MDA‐MB‐231 breast cancer cells. The RhoGEF Vav2 and the GTPases Rac1, Cdc42 and RhoA were all active at the edge of protruding/retracting regions of the cell. Vav2, Rac1 and Cdc42 had activation patterns that positively correlated with edge velocity, while RhoA showed a negative correlation. We modified our GTPase biosensors to enable simultaneous imaging of RhoGEF and GTPase activity. This revealed that the activity of Vav2, Rac1 and Cdc42 are closely correlated at the cell edge. Partial correlation analysis of simultaneous imaging data was used to dissect the role of Vav2 in regulating Rac1 and Cdc42 activity. Rac1 activation relevant to protrusion initiation was specifically dependent on Vav2. This was confirmed using shRNA knockdown of Vav2. This study demonstrates that partial correlation analysis of multiple biosensors can be used to reveal how upstream molecules regulate different target proteins without disruption of normal biological processes.

E134 Single Molecule Force Measurements in Living Cells Reveal a Minimally Tensioned Integrin State. S.J. Tan1, A.C. Chang1, A.H. Mekhdjian1, A.K. Denisin2, B.L. Pruitt3, A.R. Dunn1; 1Chemical Engineering, Stanford University, Stanford, CA, 2Bioengineering, Stanford University, Stanford, CA, 3Mechanical Engineering, Stanford University, Stanford, CA Integrins mediate cell adhesion to the extracellular matrix and enable the construction of complex, multicellular organisms; yet fundamental aspects of integrin‐based adhesion remain poorly understood. Notably, the magnitude of the mechanical load experienced by individual integrins within living cells is unclear, due principally to limitations inherent to existing techniques. We used Förster resonance energy transfer (FRET)‐based molecular tension sensors (MTSs) to measure the distribution of


mechanical forces exerted by individual integrin heterodimers in living cells. By taking advantage of the sensors’ modular nature, we engineered MTS variants that display a minimal RGD sequence derived from fibronectin (MTSRGD) or the 9th and 10th type III domains of fibronectin (MTSFN), and that respond to low (0‐7 pN) or intermediate (8‐11 pN) loads. Using these sensors, we found that a large fraction of integrins exert modest forces of <11 pN. For integrins specifically within large adhesion complexes, termed focal adhesions, we found that the measured force per integrin fell into distinct subpopulations, suggesting the presence of specific adhesion structures or modes of attachment. The presence of the fibronectin synergy site, a secondary binding site specifically for α5β1 integrin present in MTSFN but not MTSRGD, led to increased recruitment of α5β1 integrin to adhesions but not to an increase in overall cellular traction generation. The presence of the synergy site did, however, increase cells’ resistance to detachment via externally applied load. Based on these data, we suggest that a large pool of engaged, but minimally tensioned integrins may provide a synergy site‐dependent adhesive reserve that imparts cells and tissues with mechanical integrity in the presence of widely varying mechanical loads. In ongoing work, we take advantage of the unique capabilities of MTSs to determine how subpopulations of load‐bearing integrins are altered in response to external mechanical perturbations, and how (and whether) distinct integrin subtypes bear differing levels of mechanical load.

E135 A role for fibrosis in promoting pro‐tumor immune response in breast cancer. O. Maller1, L. Cassereau1, A. Drain 1, B. Ruffell2, I. Acerbi1, M. Broz3, J. Munson4, M. Swartz5, M.F. Krummel3, L.M. Coussens6, V.M. Weaver1; 1Department of Surgery , University of California San Francisco , San Francisco, CA, 2Moffitt Cancer Center , Tampa, FL, 3Department of Pathology, University of California San Francisco , San Francisco, CA, 4Department of Biomedical Engineering, University of Virginia, Charlottesville, VA, 5Institute for Molecular Engineering, University of Chicago, Chicago, IL, 6Department of Cell Developmental Biology, Oregon Health Sciences University, Portland, OR We established a positive correlation between a fibrotic phenotype in human breast tumors — especially HER2 and basal‐like breast cancer subtypes — and CD45 and CD68 positive immune cell infiltration. We were interested in elucidating how this fibrotic phenotype may influence the immune response. To address this question, we examined if matrix stiffness alters the function of STAT3, a central regulator of tumor inflammation. We hypothesize that tissue fibrosis promotes STAT3 signaling in mammary tumor cells and alters the cytokine milieu to induce a pro‐tumor immune response . We found that ECM stiffness directly enhanced STAT3 phosphorylation in tumor cells both in vitro and in vivo . Our data suggest the fibrotic phenotype promotes STAT3 activity, enhancement of which may drive a pro‐tumor immune response. Indeed, we observed several alterations in cytokines and immune cell populations upon STAT3 ablation consistent with anti‐tumor immune response. Interestingly, our data also suggest STAT3 knockout in tumor cells doesn’t necessary influence immune cell infiltration, but rather their differentiation in mammary tumors. Finally, we investigated if matrix stiffness has potentiated macrophage differentiation when cultured with specific immunosuppressive cytokines. Overall, our work reveals a novel mechanistic insight into how a pro‐tumor immune response stems from the interplay between fibrosis and STAT3 signaling in tumor cells. As such, our findings may stimulate an interest in exploring combinational treatment options with anti‐fibrotic agents and immunotherapy.


Microsymposium 16: Development and Invasion

E136 Phosphorylation of the lipid droplet‐associated protein Jabba by casein kinase 2 is critical for Drosophila oogenesis. E.A. McMillan1, S.M. Longo1, M.D. Smith1, S.A. Broskin1, N. Singh1, T.I. Strochlic1; 1Department of Biochemistry and Molecular Biology, Drexel University College of Medicine, Philadelphia, PA In most female animals, reproductive potential is highly dependent on nutritional and metabolic status. Numerous studies indicate that lipid metabolism, in particular, plays a critical role in regulating oogenesis and early embryogenesis. During oogenesis, developing oocytes accumulate large amounts of sterols and triglycerides that can be utilized for energy production or for the synthesis of other essential lipid species. Consequently, perturbations in lipid metabolism are often associated with infertility and several human reproductive disorders. Yet relatively little is understood regarding the molecular mechanisms that regulate lipid metabolism in the ovary. As key mediators of intracellular signaling pathways, protein kinases are also crucial for oogenesis and embryonic development. Specific kinases transduce developmental cues for the formation of a mature oocyte. However, the complete complement of kinases involved in metabolic signaling within the ovary is unknown. To address this, we performed a screen in which we reduced the expression of a subset of the Drosophila kinome by shRNA in the adult female germline and assessed effects on fertility. From this screen we identified and subsequently validated that the catalytic activity of casein kinase 2 (CK2) is critical for oogenesis in Drosophila. To further investigate the role of CK2 in oogenesis, we designed a novel biochemical strategy to identify specific substrates of this kinase in the Drosophila ovary. Using this approach, we have determined that CK2 phosphorylates a specific isoform (isoform F) of the lipid droplet‐associated protein Jabba. Knockdown of this Jabba isoform by shRNA also results in defects in oogenesis, suggesting that the role of CK2 in regulating female fertility in the fly may be, in part, through phosphorylation of Jabba. Supporting this notion, we further demonstrate that CK2‐dependent phosphorylation of Jabba is required for its localization to lipid droplets and for proper lipid metabolism in the Drosophila ovary. Collectively, these findings define a novel function for CK2 in oogenesis and establish a mechanistic link between kinase‐mediated metabolic signaling and female reproduction.

E137 The distal appendage protein CEP164 is critical for embryonic development and multiciliated cell differentiation. S.S. Siller1,2,3, H. Sharma1,2,3, F. Li1, M.J. Holtzman4, H. Colognato1,2,3, B.C. Holdener5, K. Takemaru1,2,3; 1Department of Pharmacological Sciences, Stony Brook University, Stony Brook, NY, 2Medical Scientist Training Program, Stony Brook University, Stony Brook, NY, 3Graduate Program in Molecular and Cellular Pharmacology, Stony Brook University, Stony Brook, NY, 4Department of Medicine, Washington University School of Medicine, St. Louis, MO, 5Department of Biochemistry and Cell Biology, Stony Brook University, Stony Brook, NY Cilia are ancient microtubule organelles often categorized as either immotile/primary or motile based on microtubule arrangement (9+0 for immotile versus 9+2 for motile cilia). Multiciliated cells are a specialized cell type with hundreds of motile cilia that serve to provide mucociliary transport in the respiratory tract, maintain cerebrospinal fluid flow in the brain ventricles, and move the ovum through the fallopian tubes. Ciliary dysfunction is linked with a wide range of clinical phenotypes, which manifest


in a spectrum of disorders termed ciliopathies. In past years, intense effort has uncovered multiple novel mutations in ciliopathy patients for genes that encode ciliary proteins that localize to the distal appendage/transition fiber of centrioles/basal bodies, a structure necessary for proper ciliary vesicle recruitment and formation, basal body docking, and intraflagellar transport particle recruitment. One such protein found to be mutated in the ciliopathy nephronophthisis is centrosomal protein of 164 kilodaltons (CEP164). Several recent studies indicate the importance of CEP164 in primary cilia formation; however, no study to date has examined the role of CEP164 in mammalian model organisms or in multiciliated cell differentiation. Here, we report the first characterization of both a whole body and multiciliated cell‐specific CEP164 knockout (KO) mouse model. Whole‐body CEP164 KO mice exhibit embryonic lethality around E10.5 with abnormalities in forebrain, heart looping, and posterior trunk. To gain further insight into CEP164 function in ciliogenesis in a physiological context, we generated multiciliated cell‐specific CEP164 KO mice. These mice show hydrocephalus, loss of airway cilia, and dramatic deficits in spermatid development. In primary cultures of multiciliated cells, dramatic reductions in the number of cilia are noticeable in the absence of CEP164. Surprisingly, though, a small fraction of CEP164 KO ciliated cells maintain the capacity to extend cilia, albeit often short and stubby. We additionally extend previous data to demonstrate that, in multiciliated cells, CEP164 is critical for basal body docking as well as for recruitment of the transition fiber protein Chibby. Of particular note, we also find that CEP164 is not required for proper ultrastructural organization of either the transition fiber or the transition zone. Taken together, our data suggests that CEP164 is necessary for both normal mammalian development as well as multiciliated cell differentiation. In the future, this mouse model should contribute to the interrogation of tissue‐specific functions of CEP164.

E138 A new platform for dynamically and reversibly modulating matrix stiffness reveals YAP inhibition of β‐catenin drives mechanosensitive neural stem cell differentiation during a critical time window. S. Rammensee1,2, M. Kang1,2, K. Georgiou1,2, D. Schaffer1,2, S. Kumar1,2; 1Bioengineering, University of California, Berkeley, Berkeley, CA, 2Bioengineering, University of California, San Francisco, San Francisco, CA It is now clear that biophysical inputs from the stem cell niche, including extracellular matrix (ECM) stiffness, can critically regulate lineage commitment. However, it remains poorly understood when stem cells are most strongly sensitive to mechanical inputs, or how such inputs interface with more canonical lineage‐commitment signals. The objective of this study is to address these questions in adult hippocampal neural stem cells (NSCs) by determining the critical time window of ECM stiffness‐dependent neuronal commitment, and mechanisms through which mechanotransductive signals interact with classical neurogenic signals. Previously we had shown that when NSCs are exposed to signals that induce differentiation into a mix of neurons and astrocytes, soft ECMs strongly favor neurogenic commitment. To determine the time window when this fate commitment is most sensitive to ECM stiffness cues, we conducted “stiffness pulse” experiments using an oligonucleotide‐crosslinked polyacrylamide gel we created that allows reversible stiffening and softening through the delivery of complementary crosslinker and release oligos. These identified 12‐36h as the window during which ECM stiffness maximally (~2‐fold soft vs. stiff) impacts neurogenic commitment. In considering signaling pathways that may mediate the effects of ECM stiffness on cellular fate decisions during this window, we focused on YAP, a Hippo pathway effector whose nuclear localization and transcriptional co‐activation have been implicated in mechanotransduction. Surprisingly, we did not observe strong nuclear localization with increased stiffness but instead saw increased overall YAP levels. Pulsed


overexpression of YAP during the 12‐36 h window suppressed neurogenesis on soft ECMs, and shRNA silencing of YAP rescued neurogenesis on stiff ECMs. Notably, overexpression of a YAP mutant with impaired TEAD‐binding still suppressed neurogenesis, implying that YAP transcriptional co‐activation is dispensable for mechanosensitive NSC lineage commitment. To investigate alternative mechanisms by which YAP could inhibit neurogenesis, we hypothesized that YAP may be suppressing β‐catenin, which strongly promotes NSC neurogenesis. Indeed, co‐immunoprecipitation confirmed a YAP/β‐catenin interaction during the 12‐36 h window, and overexpression of YAP reduced β‐catenin‐dependent transcription. Moreover, a YAP mutant with impaired β‐catenin‐binding did not suppress neurogenesis on soft ECMs. Together, our findings point to a new role for YAP in mechanosensitive stem cells, independent of its transcriptional functions and based on antagonism of differentiation signals. We also introduce a powerful new ECM platform that allows reversible modulation of stiffness for probing temporally restricted cell behaviors.

E139 Protein Kinase C signalling controls the de novo cell polarity establishment in the mammalian early embryogenesis. M. Zhu1, M. Zernicka‐Goetz1; 1Department of Physiology, Development and Neuroscience, University of Cambridge, Cambridge, United Kingdom Cell polarity is a universal feature of all cells and it is crucial for tissue morphogenesis and differentiation. In the early mammalian embryo, the embryonic cells ‐ blastomeres do not polarise until a certain stage, despite the presence of cell‐cell contact. In the mouse embryo, cell polarity is established de novo at the 8‐cell stage when the blastomeres compact, and Par complex becomes enriched in the apical and E‐cadherin in the basolateral domain. Such cellular symmetry breaking event also governs the first cell fate segregation in the mouse embryo development: the presence of apical domain commits the cells to the trophectoderm cell fate, whilst the lack of apical domain allows the cells to form pluripotent inner cell mass. One fundamental question is what induces the de novo cell polarity in the 8‐cell stage mouse embryo. Here by using a combination of pharmacological, embryological and newly developed bioimaging approaches, we have identified the PKC family members as essential activators for cell polarity establishment. Multiple methods to inhibit the activity of PKC or its upstream activator PLC consistently abolished apicobasal polarity, while the spatially and temporally controlled activation of PKC enhanced recruitment of the Par complex members to the membrane. We further found that PKC activates polarity through a Rho GTPase – myosin mediated pathway. PKC translocates to the plasma membrane at the 8‐cell stage where it restricts Rho GTPases activity to the apical domain. Rho GTPases then activates myosin to build the apical domain. These findings shed light on the mechanism of de novo cell polarity formation, as well as help us to understand the timing of the first lineage commitment event.

E140 Interleukin 6 directly activates pancreatic tumor cell invasion via Cdc42 signaling. G.L. Razidlo1, M.A. McNiven1; 1GI Basic Research, Biochemistry Molecular Biology, Mayo Clinic, Rochester, MN Inflammation is a major driver of tumor progression and metastasis, though the mechanisms of how specific inflammatory cytokines drive metastatic invasion are not known. Particularly in pancreatic cancer (PDAC), one of the most lethal forms of cancer, the complex tumor microenvironment is


comprised of inflammatory cells and signals that act to prime the tumor cells and surrounding area for metastasis. Invasive cell migration is regulated by the Rho family of small GTPases, including Rac and Cdc42, which activate downstream effectors to induce actin cytoskeletal remodeling. These pathways are potently activated in invasive tumor cells by multiple mechanisms to contribute to metastasis. Interleukin‐6 (IL‐6) is a potent inflammatory cytokine that is elevated in PDAC patients, and is required for PDAC formation in mice. Further, IL‐6 also increases tumor cell invasion in vitro. We investigated whether IL‐6 could directly activate invasive migration through Rho GTPase signaling in pancreatic cancer cells. Indeed, IL‐6 stimulation led to rapid and robust activation of Cdc42 in PANC1 cells within 5 minutes, which was sustained for over one hour. Intriguingly, IL‐6 had no effect on Rac1 activity, suggesting specificity for Cdc42. Canonical IL‐6 signaling activates the kinase JAK, which phosphorylates and activates the STAT transcription factors. We determined that IL‐6 signals through JAK to activate Cdc42, as treatment with a JAK inhibitor prevented IL‐6‐mediated Cdc42 activation. Importantly, Cdc42 activation was required for IL‐6‐induced invasion, as knockdown of Cdc42 rendered cells insensitive to the pro‐invasive effects of IL‐6. Strikingly, loss of Cdc42 prevented IL‐6‐dependent invasion even in the presence of activated STAT3. These data suggest that IL‐6 promotes metastatic invasion through Cdc42. Thus, in combination with its pleiotropic effects on tumor growth and progression, IL‐6 is also a direct activator of pro‐invasive GTPase signaling. Supported by P50 CA102701 and R01 CA104125.

E141 The regulation of nuclear ErbB3 by the androgen receptor in prostate cancer cells. M.K. Jathal1,2, T.M. Steele1,2, S. Siddiqui2, P.M. Ghosh1,2,3; 1Urology, University of California Davis, DAvis, CA, 2Research, Veterans Affairs Northern California Healthcare System, Mather, CA, 3Biochemistry and Molecular Medicine, University of California Davis, Davis, CA Introduction: We previously demonstrated that protein levels of the receptor tyrosine kinase (RTK) ErbB3, a member of the epidermal growth factor receptor (EGFR) family, are regulated by the androgen receptor (AR) in prostate cancer cells. ErbB3 levels increased in more aggressive, castration‐resistant cell lines compared to less aggressive, androgen‐sensitive ones. RTKs are typically localized in the membrane where they undergo ligand‐activated signaling but investigators have also reported the presence of nuclear ErbB3. The goal of these studies was to determine the localization of upregulated ErbB3 and whether it was activated when required for signal transduction to downstream targets. Materials and Methods: ErbB3 localization was investigated in vitro (human prostate cancer cell lines) and in vivo (human xenografts in athymic male mice) by manipulating the AR surgically (castration) or pharmacologically (serum‐depletion, ligand treatment with dihydrotestosterone (DHT) or enzalutamide (Enz)). ErbB3 and AR expression and activity were analyzed using immunoblot/fractionation/ immunohistochemistry/reporter gene assays. Results: High‐definition analysis of ErbB3 in human prostate tumor tissues demonstrated its expression on the nuclear membrane. Castration caused opposing effects on ErbB3 localization in androgen‐dependent vs castration‐resistant tumors. In vitro, ErbB3 localization could be altered using DHT or Enz. Androgen‐sensitive LNCaP cells express nuclear and cytoplasmic ErbB3 protein. Cytoplasmic ErbB3 increased upon androgen deprivation but nuclear ErbB3 decreased when an AR‐specific ligand, DHT was added. Cell viability was greatly increased in LNCaP cells co‐cultured with an ErbB3‐specific ligand, HRG‐1β, and DHT but not in response to HRG‐1β alone. In contrast, castration‐resistant Rv1 cells do not display this additive effect. Surprisingly, transient wild‐type ErbB3 overexpression decreased cell viability and kinase‐deficient mutant ErbB3 (mErbB3) increased viability in castration‐resistant tumor cells but not androgen‐sensitive ones treated with HRG‐1β. In a second castration‐resistant line (C4‐2), mErbB3


was cytoplasmic, indicating that the mutated kinase sites did not alter ErbB3 localization. This suggests that the kinase domain of ErbB3 may no longer be required for the survival of castration‐resistant cells. Conclusions: AR inhibition increases cytoplasmic ErbB3 in androgen‐sensitive tumor cells, indicating a bypass mechanism for survival. Tumor cells may then sequester overexpressed ErbB3 in the nucleus for later release back into the cytoplasm and eventually to the plasma membrane from where it can initiate pro‐survival signaling. Accordingly castration‐resistant tumor cells display increased baseline levels of cytoplasmic ErbB3,

E142 Expansions of mitotic kinase families in Stentor coeruleus, the model for single cell regeneration. S.B. Reiff1, W.F. Marshall1; 1Biochemistry Biophysics, University of California San Francisco, San Francisco, CA The giant unicellular ciliate Stentor coeruleus has the ability to fully regenerate after being cut in half, in a way that perfectly preserves the original cell polarity and structure. This ability has made it a classical model system for studying single‐cell regeneration, which remains one of the long‐standing mysteries of cell biology, as the molecular details have remained largely unstudied. To begin to develop molecular tools, our laboratory had developed a system for RNAi knockdown of Stentor genes. We have also recently assembled a Stentor coeruleus draft genome, in which we find that not only do Stentor’s introns appear to possess the smallest average intron length of any organism described to date at 15 bp, but Stentor also seems to use the standard genetic code, unlike other ciliates. Ultimately, however, we wish to understand how the regeneration process is coordinated at the molecular level. To identify candidates for RNAi knockdown we analyzed the kinome of Stentor by looking for protein kinase domains among the predicted protein coding genes. Stentor was found to encode more than 2000 kinases, making up a colossal 6% of the total protein coding genes, and a few of these kinases were found to exhibit novel domain architectures. The high number of kinases in Stentor is partially due to expansions of kinase families absent in animals and yeast; over 12% of the kinome consists of the calcium‐dependent CDPK family, originally identified in plants. However, there are also large expansions in many mitotic kinase families such as CDKs, PLKs, NDRs, and NEKs. Interestingly, some of the events that occur during regeneration parallel the events of cell division, and thus we are highly interested in exploring whether the cell co‐opts certain cell division signaling pathways for regeneration. Initial RNAi experiments on Stentor CDKs have resulted in cells with defective cell division, and experiments to test their regeneration ability are in progress. RNAi screening of the other mitotic kinase genes is ongoing, and will ultimately reveal which of these kinases help to coordinate the many different precisely timed cellular events required for successful regeneration. In the future, a better understanding of the mechanisms behind single cell regeneration will have important implications for basic biology as a whole, and will reveal how these single cells can establish and maintain their polarity and cortical organization with such a high degree of precision.


E143 Ganoderma lucidum extract significantly decreases stemness properties in Inflammatory Breast Cancer cells via STAT3 regulation. T.J. Rios1, Y. Loperena1, M.Y. Lacourt1, P. Lopez2, Y. Yamamura2, L.A. Cubano3, M.M. Martinez‐Montemayor1; 1Department of Biochemistry, Universidad Central del Caribe‐School of Medicine, Bayamon, PR, 2Department of Basic Sciences (Microbiology), Ponce Health Sciences University, Ponce, PR, 3Department of Anatomy and Cell Biology, Universidad Central del Caribe‐School of Medicine, Bayamon, PR Inflammatory Breast Cancer (IBC) is a rare, aggressive and lethal type of breast cancer (BC). IBC is associated with a negative outcome and an elevated risk of recurrence and metastasis compared with non‐inflammatory locally advanced BC. A population of cancer stem cells (CSCs) are thought to be responsible for cancer initiation, progression, metastasis, recurrence and drug resistance; making it vital to understand the molecular mechanisms that regulate stem cell properties. CSCs display a CD44+/CD24− surface marker phenotype that is associated with the Janus Kinase 2 (JAK2)/signal transducer and activator of transcription 3 (STAT3) pathway to promote oncogenesis and maintenance of the stem cell phenotype. A study reported that CD44+ tumor‐initiating BC cells had preferential activation of STAT3, suggesting it may be a potential therapeutic target. Hence, we developed a constitutively active SUM‐149 IBC cell STAT3 mutant (cSTAT) to determine the role of STAT3 in IBC and stemness. Evidence suggests that enriched aldehyde dehydrogenase (ALDH) activity is a hallmark of CSCs. The ALDH+ and ALDH+/CD44+/CD24− subpopula ons of BC cells express higher levels of pSTAT3. Moreover, our preliminary data indicates that Ganoderma lucidum extract (GLE) decreases the activation of STAT3 in IBC in vitro and in vivo. We hypothesize that GLE decreases stem cell properties via STAT3 regulation in IBC cells. Our cytometry results show that cSTAT cells have a significantly larger population of CD44+/CD24‐ cells compared to wild type (wt) SUM‐149. More importantly, our results demonstrate that GLE significantly decreases the population of CD44+/CD24‐ in cSTAT cells. To further analyze if GLE is decreasing stem cell properties, we performed western blot analysis of stem cell‐related transcription factors that have a strong correlation with the CSCs phenotype (Nanog, Oct4, and Sox2). Our results show that GLE significantly decreases the expression of these transcription factors in wt SUM‐149 cells. In addition, our GLE significantly decreases ALDH expression in wt SUM‐149 cells compared to control. We conclude that GLE decreases stem cell properties via STAT3 regulation in both of our IBC cell models. Finally, we highlight the effectiveness of GLE in decreasing stem cell properties and identified STAT3 as a potential IBC therapy target. This work was supported by NIGMS #GM111171 (MMM), NCRR #RR003035, NIMHD #MD007583 (MMM), NCRR #RR016470, GM103475 (UPR‐pilot MMM), NIMHD #MD008149 (MMM), NIMHD #MD007587 (MMM), NIGMS #GM110513 (LAC, TRF), Title‐V‐PPOHA #P031M105050 and Title‐V‐Cooperative #P031S130068 U.S. DOE (LAC), NIMHD #MD007579. The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH or the U.S. DOE.


E144 YAP Promotes Myogenic Differentiation via the MEK5‐ERK5 pathway. T.H. CHEN1,2, C.H. CHANG2,3, C.P. CHU2; 1Department of Life Science, National Tsing Hua University, Hsin‐Chu, Taiwan, 2Institute of Cellular and System Medicine, National Health Research Institutes, Zhunan, Taiwan, 3Department of Internal Medicine, Changhua Christian Hospital, Changhua, Taiwan Myogenesis is known to be regulated by many signaling pathways which involve the permanent stop of DNA synthesis, exit of cell cycle, activation of muscle‐specific gene expression and the fusion of single myotube into multinucleated muscle fiber. The MEK5/ERK5 MAP kinase cascade is essential for the early step of myogenesis. Although many substrates of ERK5 have been discovered, its upstream activator during muscle differentiation is unidentified. Yes‐associated protein (also know as YAP) is a transcriptional coactivator in the Hippo pathway which regulates cell proliferation, differentiation and apoptosis. In this study, we generated C2C12 stable cell lines expressing YAP (C2C12‐YAP cells) and found that ERK5 and MEK5 were activated as evidenced by the increased phosphorylation of threonine 218 and tyrosine 220 on ERK5 and of Serine 311 and threonine 315 on MEK5 in C2C12‐YAP cells as compared to those control C2C12 (C2C12‐vector) cells. C2C12‐YAP stable cells also differentiated into myotubes better than C2C12‐vector cells, and expressed elevated levels of myogenin, a transcription factor regulating myogenesis, and elevated levels of MHC, a skeletal muscle marker, as compared to C2C12‐vector cells. Conversely, treatment of inhibitors of c‐Abl, Src or MEK5 inhibited myogenesis of C2C12 myoblasts. Specific interactions between YAP and proteins in ERK5 pathway such as MEKK3 and ERK5 were illustrated by co‐immunoprecipitation experiments. Site‐directed mutagenesis experiments further revealed that YAP with the PPGY motif (amino acids 178‐181) on MEKK3, and that transfection of MEKK3 Y181F inhibited MEK5/ERK5 activation and myogenic differentiation. These data suggest that YAP promotes muscle differentiation by activating the Abl/Src/MEKK3/MEK5/ERK5 kinase cascade. TAZ, a YAP structure‐related protein, has been shown to positively regulate myoblast differentiation. Over‐expression of TAZ in C2C12 myoblasts activated ERK5 and promoted myogenesis. Co‐immunoprecipitation revealed that TAZ associated with c‐Abl, but not MEKK3 and ERK5. These results indicate that like YAP, TAZ promotes myogenesis by activating the ERK5 signaling pathway, although it interacts with different proteins upstream of ERK5.

Microsymposium 17: Cell Division in Development and Disease

E145 Quantification of centromeric protein dynamics during early embryogenesis of C. elegans. L. Smith1, C.A. Barnhardt2, P.S. Maddox2; 1Curriculum in Genetics and Molecular Biology, University of North Carolina, Chapel Hill, NC, 2Department of Biology, University of North Carolina, Chapel Hill, NC Centromeres are a fundamental component of all cell divisions; they are required for kinetochore establishment, microtubule attachment, and chromosome segregation. The Histone‐3 variant CENtromeric‐Protein‐A (CENP‐A) replaces H3 in centromeric nucleosomes and is accepted as the defining mark of centromeres. CENP‐A is a fairly unique mitotic protein, as its subcellular localization does not change throughout the cell cycle, even during the gap phases. Here we report the dynamic behavior of CENP‐A during embryogenesis of the holocentric organism C. elegans. Despite morphological differences, holocentric chromosomes (like those found in C. elegans) are subject to the same molecular rules as those found in other organisms, such as humans, that have regionally localized


centromeres. In order to understand CENP‐A dynamics during the early cell cycles of the C. elegans embryo, we developed an analysis pipeline to temporally align 4D images of GFP::CENP‐A (HCP‐3 in C. elegans, the only copy in the animal is tagged) from early embryonic development. This temporal alignment allowed us to create profiles of CENP‐A dynamics throughout the developmental path of an entire cell lineage. Using our custom analysis pipeline, we have found that in early C. elegans embryos, CENP‐A levels within each nucleus decrease after every cell cycle. We also find that CENP‐A levels double at the end of the cell cycle, representing the loading of new CENP‐A nucleosomes, after the doubling observed for canonical histones during S phase of each cell cycle. This pattern is distinct from that seen in human tissue culture cells, however it is similar to that reported in plants and fission yeast. In order to quantitatively compare changes in CENP‐A loading rates across a cell lineage, we calculated the rate constant of new CENP‐A assembly within each cell cycle. We also measured protein turnover rate and stability using fluorescence recovery after photobleaching (FRAP) experiments in early embryos. From these data, we were able to calculate how the levels of CENP‐A are modulated in each cell cycle through development of a specific cell lineage. Overall, our data represent the first developmental map of the centromere epigenetic mark and highlight epigenetic plasticity over developmental time.

E146 A Novel Mutual Protection Mechanism of Centromere Regulation and an Application for Predicting Cancer Patient Response to Adjuvant Radio‐ and Chemotherapy based on Centromere Misregulation. W. Zhang1,2, J. Mao1, W. Zhu3, A.K. Jain4,5, K. Liu6,7, J.B. Brown6,7,8, A. Williamson2, J. Garbe1, M.A. LaBarge1, M. Stampfer1, M. Rape2, G.H. Karpen1,2; 1Biological Systems and Engineering Division, Lawrence Berkeley National Laboratory, Berkeley, CA, 2Department of Molecule and Cell Biology, University of California, Berkeley, Berkeley, CA, 3Department of Translational Bioinformatics, Cellular Biomedicine Group, Inc., Shanghai, China, 4Department of Therapeutic Radiology, Yale School of Medicine, New Haven, CT, 5Ashland Bellefonte Cancer Center, Ashland, KY, 6Environmental Genomics and Systems Biology Division, Lawrence Berkeley National Laboratory, Berkeley, CA, 7Department of Statistics, University of California, Berkeley, Berkeley, CA, 8Environmental Bioinformatics, University of Birmingham, Birmingham, United Kingdom Chromosomal instability (CIN) is a cancer hallmark that contributes to tumor heterogeneity and other malignant properties. Aberrant centromere function causes CIN through chromosome missegregation in experimental systems. CENP‐A is a histone H3 variant found at centromeres. Centromere is regulated epigenetically by replenishing CENP‐A chromatin during cell divisions. This process requires a dedicated CENP‐A chaperone called HJURP and other factors. However, we are lack of a comprehensive understanding about whether these proteins are involved in clinical cancers, and how centromere misregulation may contribute to cancer development. Here, basing on our previous findings in Drosophila melanogaster, we identified a conserved mutual protection mechanism between CENP‐A and HJURP between flies and human cells, where their physical interaction protects each other from being targeted for degradation. Moreover, we showed that HJURP stability is regulated both in vitro and in vivo by the Anaphase Promoting Complex/Cyclosome (APC/C) ubiquitin E3 ligase. We further demonstrated that CENP‐A and APC/C compete for the same binding site on HJURP. Using isogenic breast cancer cell culture progression series, we found that many centromere proteins are progressively overexpressed at the protein level during tumorigenesis. These results led us to hypothesize that overexpression of centromere proteins may cause centromere misregulation and CIN, and contribute to human cancers. To test the idea, we developed a CES (Centromere and kinetochore gene Expression


Score) signature that quantitates the mRNA levels of 14 key genes required for centromere structure in cancers. High tumor CES values strongly correlate with increased copy number alterations and mutation frequencies across many cancer types, and prognosticate poor patient survival for many cancers. High CES values also signify high levels of genomic instability that sensitize cancer cells to additional genotoxicity. Indeed, the CES signature forecasts patient response to adjuvant chemo‐ or radiotherapy for lung and breast cancers. In conclusion, we demonstrate regulation of HJURP by APC/C ubiquitin E3 ligase, mutual protection between the key centromere proteins CENP‐A and HJURP, and a mode of misregulation of centromere protein genes in human cancers. We show the efficacy of the CES signature as a potential biomarker to identify patients who likely responds to specific treatments, thus spare the non‐responding patients from less effective treatments. These findings are expected to help address the over‐treatment problem prevalent in cancer treatments. Our approach validates the critical importance of incorporating basic knowledge of chromosome segregation pathways into cancer research and clinical applications.

E147 H3.3 Serine 31 phosphorylation at pericentromeric heterochromatin regulates chromosome segregation. C.A. Day1, A.K. Langfald1, S.R. Fadness1, K.T. Vaughan2, E.H. Hinchcliffe1; 1Hormel Institute, University of Minnesota, Austin , MN, 2Department of Biological Sciences, Universtiy of Notre Dame, Notre Dame, IN Histone H3.3 differs from canonical H3.1 by five amino acid substitutions – one, Ser31, lies at the base of the unstructured N‐terminal tail region. H3.3 Ser31 phosphorylation during mitosis is restricted to the pericentromeric heterochromatin (PCH), which surrounds the region of the chromosome that directs kinetochore assembly. We recently reported that H3.3 Ser31 mitotic phosphorylation contributes to p53 activation in response to chromosome missegregation, which acts as an aneuploidy “checkpoint” or “fail‐safe” (Hinchcliffe, Day, et al., 2016, Nat Cell Biol). Long‐term imaging of cells that missegregate a chromosome reveals that both daughters arrest for >60 hrs, directly demonstrating the activation of the “fail‐safe”. Masking Ser31 phosphorylation by antibody microinjection short‐circuits this “fail‐safe” by preventing p53 activation in response to chromosome missegregation. Here we show that H3.3 Ser31 phosphorylation is also required for proper chromosome segregation during anaphase. We expressed the H3.3 S31A non‐phosphorylatable mutant in normal, chromosomally stable cultured cells. Live imaging revealed that the mutant acts as a dominant negative, inducing significant increases in lagging chromosomes during anaphase (26% S31A vs 9% WT: N >30 cells each). Therefore, decreasing the amount of H3.3 in the PCH that can be phosphorylated at Ser31 alters the ability of chromosomes to properly align and segregate during mitosis. Our work supports the emerging hypothesis that the PCH is critical for sensing inter‐chromatid tension during mitosis (Nerusheva et al., 2014, Genes Dev). Interestingly, recurring single nucleotide driver mutations in H3.3 (but not H3.1) are found in multiple cancers: these cause K27M or G34R amino acid substitutions: both are residues flanking Ser31. When we expressed these mutations in normal cultured cells, they mimicked the S31A dominant negative phenotype, inducing chromosome segregation errors (25% K27M vs 31% G34R: N >30 each). This suggests that driver mutations found in human cancers can generate aneuploidy by disrupting proper PCH architecture and altering the generation of tension that is essential for accurate chromosome segregation. Finally, long‐term time‐lapse of S31A expressing cells revealed that unlike WT, only one of the daughters arrested for >60 hrs, while the other cell continued to divide. This reveals that suppression of Ser31 phosphorylation associated with mutant allele expression can override the fail‐safe, allowing a low level of aneuploid cell proliferation. Thus, H3.3 K27M or G34R driver mutations may


act during tumorigenesis by inducing chromosome missegregation during mitosis and abrogating the fail‐safe normally suppressing the proliferation of the resulting aneuploid daughter cells.

E148 Centrosome amplification drives spontaneous tumor development in mammals. M.S. Levine1, B. Bakker2, B. Boeckx3, D. Lambrechts3, F. Foijer2, A.J. Holland1; 1Molecular Biology and Genetics, Johns Hopkins University, School of Medicine, Baltimore, MD, 2European Research Institute for the Biology of Ageing, University of Groningen, UMCG, Groningen, Netherlands, 3Oncology, Vesalius Research Center, VIB, Leuven, Belgium Centrosome number is normally tightly controlled in cycling cells, so that upon entering mitosis a cell contains two centrosomes to organize the bipolar spindle and ensure equal chromosome segregation. However, extra centrosomes are frequently observed in both solid and non‐solid cancers and are correlated with high tumor grade and poor prognosis. Despite the strong correlation between supernumerary centrosomes and tumor development, it remains unclear whether and how extra centrosomes contribute to tumor development. To address this question, we developed a mouse model in which the levels of Polo‐like kinase 4 (Plk4), the master regulator of centrosome copy number, can be inducibly elevated to promote the creation of extra centrosomes in the absence of additional genetic defects. We show that a modest increase in Plk4 levels drives robust and sustained centrosome amplification across multiple tissues in vivo. Centrosome amplification increased the frequency of chromosome misseggregation and aneuploidy. To address the role of extra centrosomes in tumorigenesis, we drove centrosome amplification in the APCmin intestinal neoplasia model. Supernumerary centrosomes promoted an increase in tumor initiation events in the intestine, but did not influence tumor progression. Additionally, we tested the effect of centrosome amplification in the absence of additional genetic insults. Importantly, we provide the first evidence to show that centrosome amplification is sufficient to promote spontaneous tumorigenesis. Tumors in Plk4 overexpressing animals exhibited robust centrosome amplification and exhibited striking aneuploidy and ongoing chromosome segregation errors. This recapitulates the karyotypic alterations observed in genetically unstable human tumors. Together, our findings demonstrate that extra centrosomes can be a driver of tumorigenesis in mammals.

E149 Zika infection disrupts centriole biogenesis. A.T. Kodani1,2, J.F. Reiter1, K. Knopp1, J. DeRisi1; 1Biochemistry and Biophysics, UCSF, San Francisco, CA, 2Genetics and Genomics, Boston Children's Hospital, Boston, MA Zika virus (ZIKV) is an arbovirus transmitted to humans by mosquito bites and sexual transmission. Outbreaks of the virus in Brazil and Central America have been linked to a rise in congenital neurodevelopmental conditions, including microcephaly and Guillain‐Barre syndrome. Recent studies have revealed that ZIKV can infect human neural stem cells growing as brain organoids and abrogate neurogenesis in mice; however, the cellular mechanism by which brain development is disrupted remains unknown. Here we demonstrate that the Brazilian and Costa Rican ZIKV strains disrupt centriole biogenesis, a phenotype commonly associated with the autosomal recessive disorder, primary microcephaly (MCPH). Cells infected with ZIKV produce supernumerary Centrin foci that over‐accumulate the MCPH protein CEP63 and CCDC14, critical regulators of centriole duplication. We


demonstrate that CEP63 and CCDC14 interact with the Zika protease/helicase, NS3. Moreover, expression of the active form of NS3 leads to the formation of overabundant Centrin foci that accumulate CEP63 and CCDC14. These results suggest that the neurodevelopmental defects in ZIKV infected patients arise as a result of abrogated centriole biogenesis.

E150 Chromatin spatially regulates cytokinesis. D. Beaudet1, T. Akhshi2, J. Phillipp1, A.J. Piekny1; 1Biology, Concordia University, Montreal, QC, 2Biochemistry, University of Toronto, Toronto, ON The spatial regulation of cytokinesis requires several inputs. During cytokinesis, an actomyosin ring assembles and ingresses at a precise cortical location to form two daughter cells. Coupling ring positioning with the segregation of chromosomes is crucial to avoid aneuploidy, which is a precursor to genomic instability and cancer. The central dogma is that the mitotic spindle spatially controls the contractile ring, but we found that chromatin also restricts cortical polarity for ring positioning. This mechanism may be crucial in cells where chromatin is asymmetrically positioned to help govern polarity maintenance during division. We hypothesize that a Ran‐GTP gradient, which persists around chromosomes during mitosis and is crucial for mitotic spindle assembly, also regulates cortical contractility during cytokinesis. In support of this hypothesis, we found that chromatin position correlates with the polarization of cultured mammalian cells lacking polymerized microtubules and forced to exit mitosis. Ran‐GTP is the chromatin‐associated cue, since altering Ran‐GTP levels, or ectopically localizing Ran‐GTP to the membrane influences the localization of contractile proteins. Furthermore, Ran‐GTP functions together with the mitotic spindle, because the loss of active Ran increases cytokinesis failure in cells with disrupted spindles. One of the molecular targets of Ran‐GTP is the highly conserved protein anillin, a major regulator of cytokinesis that binds to RhoA, an upstream regulator of actomyosin contractility, actin, myosin and phospholipids. Anillin contains a nuclear localization signal (NLS) within its C‐terminus that is required for the polarization of cells lacking microtubules, and for cytokinesis in cells with intact spindles. We found that this site, which is autoinhibited, helps to change anillin’s affinity for cytokinesis regulators. This data supports a model in which a chromatin pathway influences anillin function to regulate cortical polarity during cytokinesis.

E151 Combinatorial Drug Treatments Identify Lamin A as a Novel Aurora B Substrate Required for Nuclear Lamina Assembly During Telophase. K.T. Vaughan1, E. Hinchcliffe2, K.L. Huegel1, S. Guo1, R. Hill1, A. Zeleniak1, M. Joyce1, W. Boggess1, Z. Schafer1, P. Vaughan1, R. Taylor1; 1Biological Sciences, University of Notre Dame, Notre Dame, IN, 2Hormel Institute, University of Minnesota, Austin, MN To test the ability of Microtubule Stabilizing Agents (MTSAs) to synergize with mitotic kinase inhibitors in killing tumor cells, we analyzed Epothilone B (EpoB) in combination with inhibitors of Aurora B (AurB), Polo‐like Kinase 1 (Plk1) and Monopolar Spindle 1 (Mps1). Uniquely among these treatments, the combination of EpoB and Aurora B Kinase Inhibitors (ABKIs) induced a significant accumulation of cells with highly‐lobulated nuclei that correlated with cell death after 48 hours. Time‐lapse analysis of EpoB/ABKI–treated cells revealed a prometaphase delay followed by premature DNA decondensation without chromosome segregation. Because the lobulated nuclei resembled the nuclear lamina defects in progeria disease, we analyzed the impact of EpoB/ABKI treatment on nuclear lamins. EpoB/ABKI


treatment disrupted assembly of LMNA during telophase, with asymmetric accumulation of nuclear lamins around the decondensing chromatin. In vitro kinase assays and MS/MS analysis revealed phosphorylation of LMNA by AurB at a single site (Ser‐628). Cells expressing a S628A mutant mimicked EpoB/ABKI treatment; cells expressing a S628E mutant were resistant to ABKI treatment. Furthermore, blot overlay assays and siRNA‐based depletion revealed a phosphorylation‐dependent interaction between LMNA and BAF (Barrier‐to‐Autointegration Factor) that was disrupted by EpoB/ABKI treatment. The benefits of this new combinatorial treatment were tested in cell line and animal models of pancreatic cancer. Human epithelial and stromal pancreatic cancer lines responded to EpoB/ABKI treatment and a novel EpoB/ABKI/gemcitibine combination was significantly more effective than gemcitabine alone. Median survival rates of a mouse orthotopic xenograft model of pancreatic cancer were significantly higher in mice treated with the EpoB/ABKI/gemcitabine combination versus gemcitabine alone. These studies identify LMNA as an unknown mitotic substrate for AurB that produces a novel defect in nuclear lamina assembly during combinatorial MTSA/ABKI treatment. This novel treatment induces enhanced cell death for cultured human pancreatic cell lines and improved median survival for an orthotopic xenograft model of pancreatic cancer.

E152 Cell‐cell cooperativity modulates proliferation and mitotic defects of phenotypically heterogeneous invading cancer cells. E. Summerbell1,2, J. Konen1,2, A.I. Marcus2; 1Graduate Program in Cancer Biology, Emory University, Atlanta, GA, 2Hematology and Medical Oncology, Emory University, Atlanta, GA Collective cell invasion, where adherent cells migrate as a single unit, is a major mode of cancer invasion and metastasis. Within collective invasion packs, pioneering leader cells direct the invasive front while follower cells adhere to leader cells and travel behind them. To further study these distinct cell types and how they cooperate during collective invasion, we developed a method termed Spatiotemporal Genomic and Cellular Analysis (SaGA) wherein we select, isolate, and amplify single living cells within collective invasion packs using image‐guided techniques. We have selected and maintained rare leader cell lines and follower cell lines from H1299 lung cancer cells. We hypothesize that specialized leader and follower cells promote mutually beneficial cell‐cell cooperativity to provide a unique invasive advantage. Isolated leader cells remain highly invasive whereas follower cells have limited invasive abilities. Gene expression analysis identified numerous differentially regulated pathways between leader and follower cells, including many cell cycle‐related genes that are more highly expressed in follower cells than leader cells. In proliferation assays and colony formation assays, follower cells are highly proliferative whereas leader cells proliferate more slowly. The addition of follower conditioned media (CM) increases the proliferation of leader cells, whereas the addition of leader CM inhibits follower cell growth. Cell cycle analysis by flow cytometry shows that leader cells undergo a significant G1 arrest 20 hours post‐serum starvation but not follower cells. Leader cells also commonly display mitotic defects that are rare in follower cells, such as cytokinetic instability, division into 3+ daughter cells, fusion of daughter cells, and uneven cell division. Overall 69.4% of leader cells display mitotic defects whereas only 6.1% of follower cells exhibit such defects. Co‐culture of leader and follower cells abrogates leader cell mitotic failure (inability to create two normal daughter cells) from 33.1% to 2.7%; adding follower CM to leader cells achieves a similar effect. Interestingly, leader cells or leader CM negatively impact follower cell cytokinesis, mitotic success, and overall mitotic rate, suggesting that leader cells may hinder follower cell proliferation. Our data suggest a mechanism wherein follower cells


promote the proliferation and mitotic success of leader cells but where leader cells negatively impact follower cell proliferation, perhaps to increase follower cell invasive potential. Understanding how leader‐follower cooperativity regulates collective cell invasion will provide crucial knowledge of mechanisms driving cancer metastasis, ultimately allowing us to improve patient prognosis and survival.

E153 Apoptotic effects of green tea in leukemic cells from acute promyelocytic leukemia mice. C.O. Torello1, R.N. Shiraishi1, F.I. Della Via1, F. Martins1, T.C. L. Castro1, M.C. Alvarez De Prax1, M.L. Queiroz1, S.T. Saad1; 1Hemocentro/Unicamp, University of Campinas , Campinas, Brazil Beneficial effects of green tea consumption have been described, including the ability to reduce cancer development. Polyphenols are the main chemical constituents of green tea extract and have been identified as the most effective substances that can inhibit tumorigenesis. Acute myeloid leukemia is an aggressive hematologic malignancy and its development and progression involve deregulation in the apoptotic process. In this concern, the aim of this study was to investigate green tea effects in acute promyelocytic leukemia (APL) mice. A total of 1 × 106 leukemic cells obtained from hCG‐PML‐RARa transgenic mice were intravenously injected in the tail vein of 12‐ to 16‐week‐old NOD.CB17‐Prkdcscid/J mice, after 4‐6 h of sublethal cobalt irradiation with 2 Gy. The hematologic counts were monitored weekly, and the following criteria were used for the diagnosis of leukemia: presence of at least 1% of blast in peripheral blood associated with leukocytosis above 30 000 cells/L, hemoglobin levels below 10 g/dL, and thrombocytopenia below 500 × 103 cells/L (He et al, 1997). Twelve days after transplantation, mice were then submitted to daily oral treatment (gavage) with 250 mg/kg/day green tea or vehicle only (water) for 5 consecutive days and were sacrificed; bone marrow (BM) and spleens were collected to the assays. Percentage of apoptotic cells in the BM and spleen of leukemic mice were evaluated using Annexin V–FITC/PI for cell staining. Treatment with green tea significantly increased the mean number of apoptotic cells in the BM (29.4 ± 5.2 vs untreated 21.0 ± 2.1 %, P < 0.05) and spleen (13.9 ± 3.1 vs untreated 9.2 ± 1.9 % P < 0.05) of mice. We next studied changes in caspase activity in response to green tea treatment by flow‐cytometry analysis. We found that green tea treatment induced an increase in the median fluorescence intensity (MFI) of cleaved caspase‐3 in the BM (83.9 ± 3.6 vs untreated 72.6 ± 4.7, P < 0.05) and in the spleen (75.5 ± 28.2 vs untreated 55.8 ± 7.3, P < 0.01); cleaved caspase‐8 in the BM (117.3 ± 9.9 vs untreated 89.1 ± 12.3, P < 0.005) and in the spleen (118.0 ± 31.5 vs untreated 81.5 ± 14.8, P < 0.001); and cleaved caspase‐9 in the BM (138.2 ± 52.4 vs untreated 85.8 ± 12.9, P < 0.001) and in the spleen (121.7 ± 49.2 vs untreated 76.5 ± 21.9, P < 0.001) of leukemic mice. Green tea treatment in APL mice induces apoptosis of cells in the bone marrow and spleen, confirmed by activation of caspase‐3, ‐8 and ‐9.

Microsymposium 18: Signaling and Bioengineering

E154 Engineering phagocytosis to understand conserved mechanisms of membrane‐proximal signal transduction. A.P. Williamson1, M.A. Morrissey1, R.D. Vale1; 1Cellular and Molecular Pharmacology, UCSF/HHMI, San Francisco, CA Recognition and engulfment of dying cells by phagocytes is important to maintain tissue homeostasis and prevent further tissue damage. Phagocytosis requires coordination of dynamic cellular behaviors in


response to specific receptor‐ligand interactions. Elegant genetic work has identified many essential molecules but how they collaborate in space and time to promote phagocytosis remains poorly understood. Therefore, we developed a cellular reconstitution platform amenable to direct, real‐time imaging to study phagocytosis. Our engineered system converts inept phagocytes into efficient ones, enabling us to appreciate the signaling behaviors associated with phagocytic proficiency. We report that 1) phagocytic receptors form mobile microclusters in response to ligation, 2) microclusters recruit cytosolic signaling molecules to the plasma membrane and spatially organize actin dynamics at the phagocytic synapse, and 3) the natural ligand‐receptor pair can be replaced by an engineered module, indicating that phagocytic signaling can be reprogrammed towards specific targets. From these data, we conclude that the conserved phagocytic signaling system bears unexpected mechanistic similarity to the more well‐studied pathways regulating T cell triggering in response to non‐self antigens. The observed mechanistic parallels are important because phagocytic signaling is found in organisms thought to lack adaptive immune systems. We are also in a unique position to adapt tools developed to understand other signaling systems to our studies of phagocytic signaling. For instance, our ability to replace the endogenous receptor with a heterologous receptor‐ligand pair indicates that phagocytosis can be directed towards a specific target in a manner conceptually similar to the Chimeric Antigen Receptor used to program cytotoxic T cells to kill tumor cells. Our ability to engineer phagocytic signaling has allowed us to further the understanding of a fundamental cellular process and in the future may allow us to re‐program phagocytosis towards targets of clinical import.

E155 The Hi‐HOST Phenome Project: Interpreting human genetic diversity through high‐throughput cell biology. L. Wang1, M.I. Alvarez1, L.C. Glover1, K.J. Pittman1, K.D. Gibbs1, D.C. Ko1,2; 1Department of Molecular Genetics Microbiology, Duke University, Durham, NC, 2Department of Medicine, Duke University, Durham, NC Background: Genome‐wide association studies (GWAS) have successfully identified hundreds of genetic variants associated with human diseases and traits. However, elucidating the cell biological mechanisms whereby genetic differences impact genes and pathways to affect disease remains a formidable challenge. We contend that GWAS of cellular traits provides a means to decipher how human genetic variation regulates cellular pathways, while also providing a system for experimental validation and mechanistic studies. Pathogen‐induced traits are particularly compelling, because pathogens can serve to probe fundamental pathways in cell biology that continue to impact susceptibility and severity of disease. Methods and Results: We developed a novel discovery platform for GWAS of cellular traits named Hi‐HOST (high throughput human in vitro susceptibility testing) that utilizes nearly a thousand genotyped lymphoblastoid cell lines from different human populations. We previously validated this system using a single trait, demonstrating that human genetic variation in the methionine salvage pathway associated with Salmonella‐induced cell death (pyroptosis) was also associated with survival in sepsis patients. Here, we quantified pathogen invasion and replication, cytokine levels, and host cell death using 9 different pathogens, and consequently carried out family‐based GWAS analysis on 148 cellular traits. Nearly all traits (95%) had repeatability of measurement between 50‐95% and most (85%) had significant heritability more than 10%. A total of 69 common genetic variants exceeded the commonly used threshold for genome‐wide significance (p<5 x 10‐8) for traits involving Yersinia pestis, Chlamydia trachomatis, Toxoplasma gondii and Salmonella species. Follow‐up experiments using luciferase reporters, immunofluorescence, RNAi, and CRISPR editing have validated functional significance and


revealed mechanisms connecting individual SNPs to nearby genes and cell biological processes. We have also connected cellular GWAS to GWAS of infectious disease (for typhoid fever, Salmonella bacteremia, and trachoma) through both single SNP and enrichment analyses, revealing cellular responses and pathways that play a significant role in risk of specific infectious diseases. Conclusions: We have 1) generated an atlas of genetic variants associated with cellular traits of 9 pathogens using cellular GWAS; 2) succeeded in identifying and experimentally validating known and new genes that regulate host‐pathogen interactions; and 3) identified cellular pathways that play an important role in risk of infectious diseases. The Hi‐HOST Phenome Project is a novel hypothesis‐generating engine for interpreting human genetics through the lens of cell biology.

E156 Optogenetics‐based dissection of chemotactic signaling and adaptation in neutrophils. B.R. Graziano1, D. Gong1, K.E. Anderson2, A.R. Goldberg1, O.D. Weiner1; 1Cardiovascular Research Institute, UCSF, San Francisco, CA, 2The Babraham Institute, Cambridge, United Kingdom Neutrophils are cells of the innate immune system that employ chemotaxis to track down and engulf pathogens. To sense small differences in chemoattractant concentration over a wide dynamic range, components of the neutrophil chemotaxis signaling network adapt to the current level of stimulus: A sustained step input of chemoattractant will generate only transient downstream signaling responses. Following adaptation, cells are again sensitive to further changes in chemoattractant concentration. A variety of sensory systems rely on the process of adaptation to measure changes in input rather than its absolute level, allowing them to dramatically increase the dynamic range of stimuli they are capable of responding to. For many such systems, adaptation occurs through stimulus‐dependent negative feedback, typically by post‐translational modification of the receptor. Such mechanisms are dispensable in neutrophils however, and the origins of adaptation within these cells remains unknown. To probe the source(s) of adaptation in neutrophil chemotactic signaling, we employed light‐responsive phytochrome proteins to modulate the temporal dynamics of signaling intermediates downstream of the receptor (here phosphoinositide 3‐kinase (PI3K) activity) in human neutrophil‐like HL‐60 cells. We find that persistent optogenetically‐driven PIP3 production results in only transient activation of Rac, indicating that the portion of the signaling cascade lying downstream of PIP3 is capable of generating adaptation. We further identify the guanine nucleotide exchange factor P‐Rex1 as the primary Rac activator in this adaptive circuit; and, in combination with drug treatments, that actin dynamics are essential for turning off Rac in response to sustained PI3K activity. When cells are instead stimulated with chemoattractant, neither P‐Rex1 nor actin dynamics are required for proper Rac activation or deactivation, respectively: Their roles in this process only become ‘unmasked’ when sustained input is applied directly at the level of PIP3 production. Together, these observations highlight the value of our optogenetics‐based approach in uncovering protein functions in pathways with significant feedback and redundancy.


E157 Systemic insulin sensitivity is regulated by GPS2 through inhibition of AKT ubiquitination and activation in adipose tissue. C.T. Cederquist1, M. Lee1, C. Martinez‐Calejman2, V. Hayashi1, J. Orofino1, C. Lentucci1, D. Guertin2, S. Fried3, M.D. Cardamone1, V. Perissi1; 1Biochemistry, Boston University School of Medicine, Boston, MA, 2University of Massachusetts Medical School , Worcester, MA, 3Mount Sinai School of Medicine, New York, NY Adipose tissue (AT) is a flexible endocrine organ that plays a critical role in the regulation of total body metabolism. Within the AT, insulin signaling plays an important role in regulating the proper balance between lipid storage and utilization. Regulation of insulin action is based on a tightly coordinated signaling network that relies on the activation of the AKT kinase, a downstream mediator of insulin receptor. Recent evidences indicate that AKT recruitment to the membrane, and thus activation, upon stimulation with EGF and IGF, is regulated through K63 ubiquitination. However, it is currently unknown whether K63 ubiquitination of AKT similarly contributes to the regulation of downstream insulin signaling events. Because impairment of insulin action in AT is associated with metabolic syndrome and Type 2 Diabetes, a complete understanding of the regulatory mechanisms governing AKT in the insulin signaling cascade is critical. Recent studies indicate that G protein pathway suppressor 2 (GPS2) plays an important role in regulating metabolism and inflammation. Our lab has specifically shown that nuclear GPS2 is required for the activation of a specific PPARg target gene program. Cytosolic GPS2 plays an important anti‐inflammatory role by controlling the activation of TNFa signaling. Our work reveals GPS2 functions in these different cellular compartments rely on a conserved regulatory strategy based on the inhibition of ubiquitin conjugating complexes responsible for the formation of K63 ubiquitin chains. Recent data indicate that GPS2 directly inhibits Ubc13 enzymatic activity, suggesting that GPS2‐mediated regulation may extend to other signaling pathways. Here we report that non‐proteolytic K63 ubiquitination of AKT is required for its phosphorylation and activation upon insulin stimulation. As loss of GPS2‐mediated inhibition of Ubc13 activity promotes AKT ubiquitination and activation and enhances insulin signaling, our findings demonstrate that GPS2 is a previously unrecognized component of the insulin signaling cascade. In support of this role, adipo‐specific ablation of GPS2 (AKO) in mice promotes sustained insulin signaling in AT, even under fasting conditions. AKO mice are associated with increased body weight, larger adipose tissue depots, and adipocyte hypertrophy compared to WT mice under chow conditions. AKO mice display impaired lipolysis under fasting conditions and increased lipogenesis. Intriguingly however, AKO mice have elevated levels of adiponectin and overall improved insulin sensitivity despite their significantly increased fat mass. Together, these results reveal a novel layer of regulation in the insulin pathway and describe how GPS2 modulates systemic metabolism by regulating insulin signaling in AT.

E158 TGFβ Regulates the Stability of Sox9. G. Coricor1, R. Serra1; 1Cellular, Developmental and Integrative Biology, University of Alabama at Birmingham , Birmingham , AL One out of two people in the U.S. will experience some form of Osteoarthritis (OA) in their lifetime. Members of the Transforming Growth Factor‐β (TGFβ) superfamily are important factors that stimulate chondrocyte matrix biosynthesis. Mice with a dominant‐negative mutation of the TGFβIIR develop a degenerative joint disease resembling OA. Another key chondrogenic factor, SOX9, is important for


maintaining chondrocyte function. Patients with OA show decreased levels of SOX9 and TGFβ receptors. Our overall hypothesis is that TGFβ regulates SOX9 to maintain articular cartilage. We utilized two cell culture models for these experiments, ATDC5 cells, and primary bovine articular chondrocytes. We showed that TGFβ stabilizes SOX9 protein and increases phosphorylation of SOX9 in both ATDC5 and bovine articular chondrocytes. SOX9 can be phosphorylated on three Serine residues S64, S181, and S211. We mutated these serine sites to alanine then over expressed the mutants in ATDC5 cells to determine which phosphorylation site is important for TGFβ‐mediated stability of Sox9. We found that mutation of Serine 211 is required to maintain normal SOX9 protein levels and that TGFβ was not able to increase the stability of SOX9‐S211A mutant. The results suggest that TGFβ may phosphorylate SOX9 on S211 to stabilize it. The S211 site is in a motif that is targeted by p38 kinase. TGFβ can signal through a canonical/Smad‐mediated pathway or non‐conical pathways, including p38. We tested the hypothesis that Smad2/3 and/or p38 regulate the phosphorylation and stability of SOX9. First, we determined that knock down of Smad2/3 using siRNA, blocked TGFβ‐mediated stability of Sox9. Then we showed that stabilization of SOX9 by TGFβ was blocked in cells treated with p38 inhibitors suggesting that TGFβ requires both Smad2/3 and p38 activity to regulate SOX9 protein levels. Previously, we showed that Papss2 is a downstream target of TGFβ. We utilized siRNA for SOX9 to show that TGFβ requires SOX9 to regulate Papss2. We also showed that p38 is required for TGFβ‐mediated regulation of PAPSS2 suggesting a new mechanism for TGFβ‐mediated gene regulation in chondrocytes. Understanding how TGFβ stabilizes SOX9 may lead to identification of preventative targets of both age‐related and post‐traumatic OA.

E159 TULP3 and Tubby function as general adapters for trafficking of structurally diverse integral membrane proteins to the primary cilium. H.B. Badgandi1, S. Hwang1, I. Shimada1, S. Mukhopadhyay1; 1Cell Biology, University of Texas Southwestern Medical Center, Dallas, TX The primary cilium is a paradigmatic organelle for studying compartmentalized signaling. The ciliary membrane is enriched in multiple integral membrane proteins including certain rhodopsin family GPCRs, and proteins linked to polycystic kidney disease, such as the single‐pass transmembrane protein fibrocystin and the TRP channel family proteins polycystin‐1/2 (PC1/2). While soluble proteins gain access to cilia in a size‐dependent manner, pathways regulating integral membrane protein trafficking to cilia are poorly understood. Here, we determine the tubby family protein, TULP3 functions as a general adapter for ciliary trafficking of structurally diverse integral membrane cargo. We determine that the ciliary localization of at least 15 out of 23 reported and two novel rhodopsin family cilia‐targeted GPCRs, the orphan GPR19 and the purinergic P2RY1 is dependent on TULP3. The founding tubby family member, TUB also localizes to cilia, similar to TULP3, and preferentially determines trafficking of a subset of these GPCRs to neuronal cilia. Notably, other integral membrane proteins, such as the polycystins PC1/2, which traffic to cilia interdependently in a complex, also require Tulp3 for ciliary trafficking. Using minimal ciliary localization sequences (CLSs) from GPCRs and fibrocystin, we show these motifs to be sufficient and TULP3‐dependent for ciliary trafficking. Furthermore, using two independent and orthogonal assays (proximity biotinylation and chemical cross‐linking) we detect direct binding between the diverse chimeric CLSs and TULP3/TUB mediated by the C‐terminal tubby domain. The tubby domain specifically binds to phosphoinositide 4,5‐bisphosphate (PI(4,5)P2), and we demonstrate that the interactions between diverse CLSs and TULP3 are membrane PI(4,5)P2‐dependent, suggestive of a coincidence detection model. Cilia are PI(4,5)P2‐deficient compartments, due to from localization and emrichment of 5' phosphatases such as Inpp5e. Upon Inpp5e depletion we detect


accumulation of Tulp3‐associated cargo such as the orphan GPCR, Gpr161 and PC2 in cilia suggesting that dislodgement of Tulp3 from ciliary membrane and subsequent release of associated cargo is coupled to turnover in cilia. We propose a three‐step model for TULP3/TUB‐mediated ciliary trafficking, including capture of diverse membrane cargo by tubby domain in a PI(4,5)P2‐dependent manner, ciliary delivery by intraflagellar transport complex‐A binding to TULP3/TUB N‐terminus, and subsequent release into PI(4,5)P2‐deficient ciliary membrane. By defining Tulp3/Tub as a key factor in trafficking diverse types of integral membrane cargoes, our study has broader implications in understanding the molecular basis of cilia‐generated signaling in ciliopathies, including polycystic kidney disease.

E160 Axonal growth and guidance responses regulated by NADPH oxidase‐derived reactive oxygen species. H.S. Roeder1, D.M. Suter1; 1Department of Biological Sciences, Purdue University, West Lafayette, IN NADPH oxidases (NOX) are a class of membrane‐bound multimeric complex proteins whose sole purpose is to produce endogenous reactive oxygen species (ROS), specifically superoxide. At high concentrations, ROS are known to have detrimental effects on proteins, DNA, and lipids; a process collectively known as oxidative stress. At physiological concentrations, however, ROS are known to regulate many cellular processes including cell proliferation, differentiation, migration, and apoptosis. Our lab has previously shown that ROS play a physiological role in regulating F‐actin structure and dynamics in growth cones as well as neurite outgrowth of Aplysia bag cell neurons. Furthermore, our recent in vivo studies in zebrafish (Danio rerio) suggest a role of NOX‐derived ROS in optic nerve development; however, the underlying mechanisms are currently unknown. In order to investigate whether NOX‐derived ROS play a role in axonal growth and guidance responses to specific guidance cues, we cultured zebrafish retinal ganglion cells (RGC’s) in vitro. Cells treated with either netrin‐1 or brain‐derived neurotrophic factor (BDNF) showed axonal growth rates similar to control cells; however, when NOX was pharmacologically inhibited, growth rates of RGC’s exposed to netrin‐1 but not to BDNF were significantly reduced. Contrarily, growth rates of RGC neurons exposed to slit2, a soluble cue known to cause growth cone collapse and neurite retraction, increased to that of control rates following NOX inhibition. Furthermore, positive turning of RGC growth cones towards a gradient of BDNF as well as negative turning away from a slit2 or netrin‐1 gradient was completely abolished, when NOX was inhibited. These findings suggest that specific guidance cue receptors can signal to NADPH oxidases to regulate growth and guidance responses of zebrafish retinal axons.

E161 Modeling Glioma‐Associated Genomic Rearrangements Using Somatic Genome Editing. P.J. Cook1, A. Ventura1, R. Benezra1; 1Cancer Biology and Genetics, Memorial Sloan Kettering Cancer Center, New York, NY Recent advances in next‐generation sequencing technology have led to the identification of a large number of novel recurrent gene fusions in human cancer patients. These gene fusions are primarily the result of tumor‐specific genomic rearrangements and are largely uncharacterized in regards to their potential to act as oncogenic drivers. Previously, we demonstrated that CRISPR‐Cas9 genome editing technology could be used to model oncogenic fusions more accurately by directly inducing the underlying genomic rearrangement, generating a lung cancer model that more faithfully recapitulated the human disease. Here, we explored the hypothesis that an analogous approach can be employed to


rapidly interrogate the oncogenic potential of genomic rearrangements recently identified in human glioma, a type of cancer where efficacious treatment options are not readily available. Specifically, we have selected a set of genomic deletions, inversions, and translocations that are predicted to generate oncogenic fusion kinases and are recurrently found in human gliomas. We demonstrate that these gene fusions can be efficiently modeled in vitro using our CRISPR‐based approach within primary cultures of adult mouse neural stem cells. Importantly, we also show that one of these rearrangements, a chromosomal deletion resulting in a previously uncharacterized tyrosine kinase receptor fusion, leads to the generation of invasive high‐grade glioma with high penetrance in an orthotopic transplantation model. Analysis of single cell derived clones indicated that tumorigenesis was specifically dependent on the presence of the deletion allele. These tumors displayed constitutive PI3K/AKT signaling in culture, and were highly sensitive to genetic or pharmacological inhibition of the fusion kinase both in‐vitro and in‐vivo. Our work demonstrates that the oncogenic potential of previously uncharacterized genomic rearrangements can be rapidly and efficiently tested in vivo, and provides a means to generate accurate preclinical models of highly lethal human cancers.

E162 The Tetraspanin CD82 Regulates Hematopoietic Stem Cell Interactions with the Bone Marrow Microenvironment. C.A. Saito Reis1, K.D. Marjon1, E.M. Pascetti1, K.L. Karlen1, R.J. Dodd1, C.M. Termini1, J.M. Gillette1; 1Pathology, University of New Mexico Health Science Center, Albuquerque, NM Hematopoietic stem and progenitor cells (HSPCs) are responsible for the continued production of new blood and immune cells. Cell signaling and adhesion within the bone marrow microenvironment tightly regulate the self‐renewal and differentiation of HSPCs. The tetraspanin CD82 is a membrane scaffold protein highly expressed on HSPCs. Previous work in our laboratory showed that human CD34+ HSPCs pre‐treated with a CD82 blocking antibody had impaired bone marrow homing. Therefore, we hypothesize that the CD82 scaffold regulates HSPC adhesion and signaling critical for establishing and/or maintaining interactions within the bone marrow. To test this hypothesis, we are utilizing a global CD82 knock out (CD82KO) mouse to evaluate changes in HSPC fitness and function. To assess HSPC fitness, we performed a competitive repopulation assay, measuring the ability of HSPCs to properly migrate and engraft within a lethally irradiate mouse. We find a significant increase in the amount of WT HSPCs that repopulate the blood of recipient mice when compared to CD82KO HSPCs. These data suggesting that CD82 expression regulates stem cell fitness in a competitive environment. Next, we performed a competitive homing assay, where we find a larger proportion of CD82KO cells localized within the blood rather than the bone marrow. Therefore, these data suggest that the loss of CD82 leads to a disruption or delay in HSPC homing or perhaps that CD82KO cells are unable to be maintained in the bone marrow. Additionally, HSPC adhesion within the bone marrow was assessed in vivo by treating mice with AMD3100, a drug used clinically to mobilize HSPCs into the blood. From these experiments we find a significant increase of CD82KO HSPCs mobilized to the blood when compared to WT, suggesting a reduction in the CXCR4‐mediated bone marrow interactions within the CD82KO mice. More recently, HSPC migration was assessed in vitro by live‐cell confocal microscopy. Isolated CD82KO HSPCs plated on fibronectin show a significant decrease in their migration velocity and track length when compared to WT HSPCs, suggesting a specific role for CD82 in HSPC migration. In addition, in vitro cell spreading assays were performed where cytoskeletal organization was measured by confocal imaging of phalloidin stained HSPCs. In these data, we find a significant increase in cell spreading of the CD82KO HSPCs. We speculate that this increase in cell spreading could be due to changes in Rho GTPase activity within the CD82KO HSPCs. Together, these data demonstrate that the CD82 scaffold plays a critical role regulating


HSPC migration, adhesion and repopulation potential. Future studies will analyze how CD82 regulates specific signaling within the bone marrow niche.

E.B. Wilson Medal Presentation and Address

A13 From Laminin to Lamin and p53: What determines the differentiated state? M.J. Bissell1; 1Biological Systems Engineering, Lawrence Berkeley National Laboratory, Berkeley, CA “If there is one generalization that can be made from all tissue and cell culture studies with regards to the differentiated state, it is this: Since most, if not all, functions are changed in culture qualitatively and/or quantitatively, there is no constitutive gene expression in higher organisms; i.e. the differentiated state is unstable and the (micro)environment regulates gene expression.” (Mina Bissell, International Review of Cytology, 1981) Then I spent the next 4 decades to examine if the above statement is correct and searched for mechanisms. I will discuss the significance of interactions between the extracellular matrix and the cytoskeleton with nucleus and chromatin, and why it needs to be dynamic and reciprocal. I will also discuss the mechanism of how the normal cells maintain homeostasis and what may go wrong when they forget and become cancerous.

ORAL PRESENTATIONS‐ WEDNESDAY

Oral Presentations‐Wednesday, December 7

Symposium 7: Nuclear Organization

S16 The role of heterochromatin in a complex organism: stabilizing the genome. S.M. Gasser1,2; 1Director, Friedrich Miescher Institute for Biomedical Research, Basel, Switzerland, 2University of Basel, Basel, Switzerland Epigenetic marks are known to influence gene expression, in part through their ability to sequester DNA away from the transcriptional machinery. Histone H3K9 methylation is a conserved modification that correlates broadly with gene repression in organisms from fission yeast to man. In C. elegans, di‐ and tri‐K9 methylation is abundant on repetitive elements (RE), including both transposons and simple repeats, and coats both pseudogenes and silent tissue‐specific genes. Using a double mutant that eliminates the two C. elegans H3K9 histone methyltransferases, SET‐25 and MET‐2, we find that H3K9me is dispensable for development1,2 , although worms become ts sterile. This correlates with extensive DNA damage‐driven apoptosis in the germline, but there is no elevation in either mitotic or meiotic chromosome mis‐segregation. Instead, we find that the loss of H3K9methylation leads to the promiscuous and widespread expression of all classes of repetitive elements (DNA and RNA transposons, and simple repeats) in both germline and somatic tissues. The loss of transcriptional silencing correlates with an accumulation of insertions and deletions at repetitive sequences, and renders worms sensitive to replication fork stalling, but not ionizing radiation. RNA‐DNA hybrids accumulate in the absence of H3K9me even without exogenous stress. We conclude that a key function of H3K9me is to ensure the stability of a repeat‐rich genome, most likely by suppressing the transcription of simple repeats. This is distinct from the role of H3K9me in sequestering chromatin at the nuclear envelope, which contributes to the stability of cell fate decisions,2 . Zeller, P., Padeken, J., van Schendel, R., Kalck, V., Tijsterman, M., Gasser, S.M. (2016) Histone H3K9me‐mediated repression is dispensable for development, but suppresses RNA:DNA hybrid‐associated repeat instability. Nature Genetics , in press. Gonzalez‐Sandoval, A., Towbin, B.D., Kalck, V., Cabianca, D.S., Gaidatzis, D.,Hauer, M.H., Geng, L., Wang, L., Yang, T., Wang,. X., Zhao, K., Gasser, S.M. (2015) Perinuclear Anchoring of H3K9‐methylated Chromatin Stabilizes Induced Cell Fate in C. elegans embryos . Cell, 163 , 1333‐47

S17 Following Single mRNAs from Birth to Death in Living Cells. R.H. Singer1,2; 1Anatomy and Structural Biology, Albert Einstein College of Medicine, Bronx, NY, 2Janelia Research Campus of the HHMI, Ashburn, VA Single molecule microscopy has been an essential tool to analyze the dynamic properties of messenger RNA. Newly developed technologies in optical microscopy and novel methods for RNA tagging have allowed us to detect and track individual mRNA molecules in single living cells, yielding insights that could not have been obtained through any ensemble measurements. These approaches employ a plethora of imaging techniques, ranging from multiphoton microscopy, high‐speed real‐time widefield microscopy, single molecule tracking, super‐registration microscopy and fluorescence fluctuation


analysis. The kinetics of the key elements of RNA metabolism: initiation, elongation, termination, export, cytoplasmic trafficking, localization, translation and decay have become tractable at the single molecule level in real time in living cells. For instance it was possible to measure the initiation frequency and elongation rates of single mRNA transcription or translation, or the probability and the time to be transported through a nuclear pore. We have been dedicated most recently to developing and implementing these technologies to further the understanding of dynamics of mRNA regulation in living tissues, particularly neural tissues of a transgenic mouse where the mRNAs synthesized from targeted genes contain stem‐loops from a phage bound by a fluorescent capsid protein. By observing RNA in its native environment, we may find new regulatory mechanisms not envisioned in the study of cultured cells. Supported by funding from NIH. Further information can be found at www.singerlab.org.

Minisymposium 20: Cell Division‐ Chromosome and Cytoskeletal Dynamics

M201 Single‐molecule localization microscopy reveals a dynamic architecture within the synaptonemal complex in C. elegans meiosis. S. Köhler1, M. Wojcik2, K. Xu2, A.F. Dernburg1; 1Molecular Cell Biology, University of California, Berkeley, Berkeley, CA, 2Chemistry, University of California, Berkeley, Berkeley, CA Meiosis is the specialized cell division process that separates homologous chromosomes to produce haploid gametes. Proper segregation of homologous chromosomes depends on meiotic recombination, which results in the formation of chiasmata, physical links between homologs that enable them to bi‐orient and separate during the first meiotic division. Meiotic recombination is regulated by a mysterious mechanism known as “crossover interference”, which restricts the number of crossovers resulting in widely‐spaced chiasmata. The nematode C. elegans has particularly strong interference, such that only a single crossover occurs per pair of chromosomes in each meiotic cell. This implies that once a crossover is designated, an inhibitory signal can act along the length of paired chromosomes to inhibit subsequent crossover formation. Although it remains controversial, there is strong evidence that this interference signal is mediated by the synaptonemal complex (SC) between homologous chromosomes. The SC is a proteinaceous structure that bridges the axes of paired homologous chromosomes. Because this structure is only ~100 nm wide, it has been difficult to interrogate its internal organization using conventional light microscopy. Using a combination of STORM and PALM super‐resolution microscopy, we have analyzed the three‐dimensional structure of the synaptonemal complex in intact meiotic nuclei from C. elegans. We took advantage of novel genome editing methods to tag SC proteins at distinct sites using CRISPR/Cas9. By localizing these epitope tags with respect to an internal reference, we are building a three‐dimensional model of SC organization in late pachytene. This approach has enabled us to investigate the relationship between SC structure and function. Surprisingly, we found that the SC structure is reorganized upon crossover formation, and that mutations that prevent crossover formation affect this reorganization. Moreover, we identified a novel mutation in one of the four known SC proteins that can prevent these structural changes. This mutation results in a dramatic increase in crossovers, based on both cytological analysis and genetic recombination mapping by whole genome sequencing. The effects of this mutation reinforce the hypothesis that conformational changes in the SC underlie crossover interference.


M202 BAF forms a rigid shell around anaphase chromosomes to prevent micronucleation. M. Samwer1, P.S. Schmalhorst2, M. Schneider1, R. Höfler1, D.W. Gerlich1; 1Institute of Molecular Biotechnology of the Austrian Academy of Sciences (IMBA), Vienna, Austria, 2Institute of Science and Technology Austria (IST Austria), Vienna, Austria Most eukaryotic cells package all of their chromosomes into a single nucleus, which is important for the maintenance of genome integrity. During mitosis, higher eukaryotes disassemble the nuclear envelope to disperse mitotic chromosomes into the cytoplasm, so that they individually move on the mitotic spindle. Following the segregation to the spindle poles, each set of chromosomes is enwrapped by membranes to form a single cell nucleus in each of the nascent daughter cells. Nuclear reassembly is mediated by the interaction of inner nuclear membrane proteins with chromatin receptors. Why chromosomes are not enwrapped individually by these chromatin‐membrane interactions, however, is not known. Using image‐based RNAi screening, we identified barrier‐to‐autointegration factor (BAF) as a critical factor for packaging each set of anaphase chromosomes into a single nucleus. BAF was previously implicated in nuclear reassembly, attributed to its high‐affinity binding to DNA and inner nuclear membrane proteins containing a LEM (LAP2, emerin, MAN1) domain. Unexpectedly, RNAi phenotype rescue experiments with mutant BAF versions showed that LEM‐domain binding was not required to shape a single nucleus. Instead, BAF’s DNA‐DNA crosslinking activity was essential to prevent nuclear fragmentation into micronuclei. Using purified chromatin and recombinant BAF proteins, we found by atomic force microscopy that BAF forms a rigid shell at the chromatin surface, depending on its DNA‐DNA crosslinking activity. Thus, BAF regulates mechanical properties of chromatin to shape a single nucleus during mitotic exit.

M203 Microtubules push chromosomes apart in anaphase. C. Yu1, S. Redemann2, H. Wu3, T. Yoo1, T. Müller‐Reichert2, D.J. Needleman1,4; 1School of Engineering and Applied Sciences, Harvard University, Cambridge, MA, 2Medical Faculty Carl Gustav Carus, Technische Universität, Dresden, Germany, 3Physics, Harvard University, Cambridge, MA, 4Molecular and Cellular Biology and Center for Systems Biology, Harvard University, Cambridge, MA The processes which drive chromosome motion in anaphase are poorly understood and different mechanisms have been proposed to operate in meiosis and mitosis and in different organisms, including depolymerization of kinetochore microtubules, pulling apart spindle poles by cortical pulling forces, and pushing apart spindle poles by the elongation of microtubules which extend from the pole to the center of the spindle. We have investigated anaphase in the first mitotic division of C. elegans and found that, contrary to previous speculations, spindle pole separation and chromosome separation are two mechanistically distinct processes. We argue that while spindle pole separation is driven by cortical pulling forces, chromosome separation results from a new population of microtubules which appear between chromosomes at the onset of anaphase and push chromosomes apart. We characterized the behaviors of this newly discovered population of microtubules using a combination of laser ablation, nonlinear microscopy, fluorescence recovery after bleaching, larger scale electron tomography reconstructions, and mathematical modeling. Our results suggest that these microtubules continually nucleate in a small region between chromosomes throughout anaphase, and push chromosomes apart by a combination of polymerization and sliding. Additional preliminary results indicate that the same processes could be the


primary driver of chromosome motion in anaphase in C. elegans meiosis and in mitotic human tissue culture cells. Thus, a common underling mechanism may drive chromosome segregation in anaphase in diverse systems: newly generated microtubules between chromosomes push them apart.

M204 Meiotic drive depends on spindle asymmetry induced by cortical proximity. T. Akera1, L. Chmatal1, K. Yang1, C. Janke2, R.M. Schultz1, M.A. Lampson1; 1Biology, University of Pennsylvania, Philadelphia, PA, 2Signaling, Neurobiology and Cancer, Institut Curie, Orsay, France Meiotic drive violates Mendel’s First Law of Genetics through biased segregation of selfish chromosomes to the egg. Although it has significant consequences for chromosome evolution, the underlying cell biological mechanisms remain elusive. Biased segregation should in theory depend on differences between the half spindle destined for the egg and the other half destined for the polar body. Experimental evidence for such asymmetry is scarce, however, and it is unknown how asymmetry would be established with a consistent orientation relative to the oocyte cortex. Here, we show that biased orientation of selfish chromosomes towards the egg depends on spindle asymmetry induced by cortical proximity. Using mouse oocytes as a model system, we find that spindles become asymmetric only after migration to the cortex, with the cortical side of the spindle enriched for more stable microtubules and for tyrosinated alpha‐tubulin. By manipulating spindle position, we demonstrate that cortical proximity induces spindle asymmetry. We also show that asymmetry is lost when the gradient of Ran GTPase activity is disrupted, suggesting that Ran‐mediated cortical polarization is the underlying mechanism controlling spindle asymmetry. Furthermore, we find that active Cdc42 GTPase, which is enriched at the polarized cortex is required for spindle asymmetry. Importantly, loss of spindle asymmetry abolishes the bias in chromosome orientation. These findings provide the first direct evidence that meiotic drive depends on spindle asymmetry and the first mechanistic insight, indicating that signals from the oocyte cortex regulate microtubules to induce spindle asymmetry.

M205 SFI1 recruits USP9X to stabilize the microcephaly protein STIL. A.T. Kodani1,2, J.F. Reiter1, C.A. Walsh2; 1Biochemistry and Biophysics, UCSF, San Francisco, CA, 2Genetics and Genomics, Boston Children's Hospital, Boston, MA Centrioles function as core components of neuronal stem cell proliferation and cell fate during neurogenesis. In mammalian cells, effective centriole duplication is essential for cell division, ciliogenesis and developmental signaling. Here, we identify a role for the peri‐centrosomal protein SFI1 as a critical regulator of centriole duplication and biogenesis. Similar to its role in yeast and Tetrahymena, loss of SFI1 or a loss of function human mutation in SFI1 that disrupts brain morphology abrogates centriole duplication and centrosome organization. We demonstrate that SFI1 interacts with and controls the function of the deubiquitinating enzyme, USP9X, a gene mutated in female‐restricted X‐linked syndromic mental retardation‐99 (MRXS99F), an X‐linked neurodevelopmental disorder characterized by cognitive delays and brain malformations. The deubiquitinating activity of USP9X directly removes the proteasome targeting Lys48‐ubiquitin linkages on the primary microcephaly protein STIL, a critical regulator of centriole duplication mutated in primary microcephaly 7 (MCPH7). The USP9X‐mediated removal of ubiquitin stabilizes STIL, potentiating its recruitment of other critical regulators of centriole duplication to the centrosome. In addition, de novo USP9X mutations that cause MRXS99F disrupt STIL


stability and centriole biogenesis. Together, these results reveal that SFI1 works through its interactor USP9X to stabilize STIL and promote centriole duplication, a fundamental cell biological events critical for human brain development.

M206 Human microcephaly protein RTTN interacts with STIL and is required for assembly of full‐length centrioles. H. Chen1,2, C. Wu2,3, Y. Lin2, W. Wang3, T.K. Tang2; 1Graduate Institution of Life Sciences, National Defense Medical Center, Taipei, Taiwan, 2Institute of Biomedical Sciences, Academia Sinica, Taipei, Taiwan, 3National Yang‐Ming University, Taipei, Taiwan Centrioles are the key fundamental elements of centrosomes and cilia. Mutations in many centriolar protein‐encoding genes have been reported to be associated with a neurodevelopmental disorder, primary microcephaly (MCPH). Here, we characterize the role of a newly identified human microcephaly protein, RTTN, and found that it acts as a key player for assembly of full‐length centrioles. Using three‐dimentional structured illumination microscopy (3D‐SIM) approach, we found that RTTN is recruited to the proximal end of the procentriole at early S phase and colocalized with STIL. Further studies demonstrated that RTTN directly interacts with STIL and acts downstream of STIL‐mediated centriole assembly. Crispr/Cas9‐mediated RTTN knockout in p53‐deficient RPE1 cells induced amplification of procentriole‐like structures that lack POC5 and POC1B, two known distal half centriolar proteins. Further analysis showed that depletion of RTTN did not affect the initial step of procentriole assembly but inhibited the recruitment of POC5 and POC1B to the distal half centrioles. These data suggests that STIL‐RTTN interaction is required for assembly of full‐length centrioles. To assess the function of disease‐associated RTTN mutations, we replaced endogenous RTTN with wild‐type or missense mutants of RTTN‐GFP transgenes. Interestingly, a mutation of RTTN (A578P) that causes MCPH in humans shows a low affinity for STIL binding and inhibits centriole duplication. Together, our results reveal the role of RTTN in centriole assembly and provide an underlining mechanism for human microcephaly patients carrying the RTTN mutation (A578P).

M207 Epsilon‐tubulin deletion in human cells results in loss of centrioles and a futile procentriole formation/disintegration cycle. J.T. Wang1, C. Hoerner1,2, D. Kong3, J. Loncarek3, T. Stearns1,4; 1Department of Biology, Stanford University, Stanford, CA, 2Department of Medicine ‐ Division of Oncology, Stanford School of Medicine, Stanford, CA, 3Laboratory of Protein Dynamics and Signaling, Center for Cancer Research ‐ Frederick, National Cancer Institute, National Institutes of Health, Frederick, MD, 4Department of Genetics, Stanford School of Medicine, Stanford, CA The tubulin superfamily has 6 members; alpha‐, beta‐ and gamma‐tubulin are found in all eukaryotes, whereas the other tubulin family members, zeta‐, epsilon‐, and delta‐tubulin, form an evolutionarily co‐conserved module (ZED) that has been lost in some, but retained in most organisms, including mammals1. Where retained, epsilon‐tubulin is always present, along with one or both of delta‐ and zeta‐tubulin. Zeta‐tubulin was recently shown to be required for orientation of basal bodies in multi‐ciliated cells, and to be a component of the basal foot1, but the molecular functions of the other ZED tubulins are not well‐understood. Epsilon‐tubulin has been reported to be required for centriole duplication in Xenopus egg extract2, and in several unicellular eukaryotes it is required for centriole ultrastructure and triplet microtubule formation3,4,5,6. However, the function of epsilon‐tubulin in the well‐characterized somatic centriole


cycle of mammalian cells has not been examined. We deleted the TUBE1 epsilon‐tubulin gene using CRISPR/Cas9 in human hTERT RPE‐1 cells. Recent work has shown that loss of centrioles in mammalian cells results in a p53‐dependent arrest, and we found that homozygous mutation of TUBE1 could only be made in TP53‐/‐ cells. Remarkably, TUBE1‐/‐; TP53‐/‐ RPE1 cells lack normal centrioles, consistent with sequential loss following centriole duplication failure, but about half (~50%) of the asynchronous cells do have multiple centriole‐like foci, which are positive for labeling with centrin and SASS6. Our hypothesis was that these correspond to immature centrioles made each cycle by the de novo centriole synthesis pathway, which are then lost during passage through mitosis. Consistent with this, these foci are present in S phase TUBE1‐/‐; TP53‐/‐ cells, but not in G1 cells, and do not mature into microtubule‐nucleating centrosomes. Correlative light and electron microscopy showed that these foci have some aspects of centriole structure, but lack triplet microtubules, are shorter than control centrioles, and have other unusual structural features. We conclude that in human cells, similar to unicellular eukaryotes, epsilon‐tubulin is required for centriole structure and triplet microtubule formation. In mammalian cells, this results in loss of normal centrioles, accompanied by a futile cycle of de novo formation of aberrant procentrioles followed by disintegration each cell cycle. 1. Turk et al., 2015, Cur Biol 25, 2177 2. Chang et al., 2003, Nat Cell Biol 5, 71 3. Goodenough and St. Clair, 1975, J Cell Biol 66, 480 4. Dutcher et al., 2002, Mol Biol Cell 13, 3859 5. Dupuis‐Williams et al., 2002, J Cell Biol 158, 1183 6. Ross et al., 2013, J Cell Sci 126, 3441

M208 FLIRTing with cell division during development to study the spatiotemporal regulation of protein function in vivo. S. Sundaramoorthy1, T. Davies1, Y. Zhuravlev1, S. Hirsch1, M. Shirasu‐Hiza2, J. Dumont3, J.C. Canman1; 1Pathology and Cell biology, Columbia University Medical Center, New York, NY, 2Genetics and Development, Columbia University Medical Center, New York, NY, 3Cell division and Reproduction, Institut Jacques Monod, Paris, France We developed an infrared laser‐based optogenetic technology called FLIRT (Fast Local InfraRed Thermal‐control) to manipulate fast‐acting temperature‐sensitive (ts) protein function at precise times and locations during intricate cellular processes. Molecular control of complex cellular activity, such as cell motility or cell division, involves a high degree of spatiotemporal regulation, which is difficult to study using traditional genetic approaches. Fast‐acting ts genetics permit studies of the temporal regulation of protein function by simply shifting to the restrictive temperature during a complex cellular process. To harness the power of fast‐acting ts mutants for spatiotemporal studies, FLIRT works by focusing an infrared laser at distinct sub‐cellular structure(s) or on specific cells to locally heat and inactivate ts proteins at precisely defined moments during a complex cellular process while simultaneously monitoring the kinetic effects on that process in vivo . Here we use FLIRT to probe the spatiotemporal regulation of the core cytokinesis machinery; however, FLIRT is applicable to any cellular or developmental process accessible by light microscopy. Cytokinesis, the physical division of one cell into two, is essential for the development of all multicellular organisms. We use FLIRT in the roundworm C. elegans to study the spatiotemporal regulation of cytokinesis, both in single cell zygotes and in specific cells of multicellular embryos, using our growing collection of fast‐acting (~20 sec) and reversible ts mutants that affect cytoskeletal proteins and regulators. Cytokinesis is driven by constriction of a


contractile ring composed of formin‐nucleated f‐actin and the motor myosin‐II. We found that FLIRT irradiation of one cell in a 2‐to‐4 cell or 4‐to‐8 cell stage embryo of formin(ts) or myosin‐II(ts) mutants causes cytokinesis failure in the irradiated cell and not in the neighboring non‐irradiated cells . Also, FLIRT irradiation of half of the contractile ring in formin(ts) or myosin‐II(ts) mutant embryos halts ring constriction on the irradiated side of the cell, but not on the non‐irradiated side. This suggests the contractile ring is made up of individual contractile units that can function independent of a full contractile ring. Importantly, FLIRT irradiation does not inhibit cytokinesis in control embryos lacking ts mutations, or in formin(ts) and myosin‐II(ts) embryos irradiated outside of the division plane (in the polar cell cortex). These results highlight the high degree of spatial control that can be achieved with FLIRT. We are currently expanding our analysis to all of our fast‐acting ts mutants in worms; in the future, this technology can be applied to study other model systems including yeast, flies, and human cells.

M209 Membrane composition is important in contractile ring anchoring in Schizosaccharomyces pombe. C.E. Snider1, A.H. Willet1, J. Chen1, K.L. Gould1; 1Department of Cell and Developmental Biology, Vanderbilt University, Nashville, TN Cytokinesis is the final step in cell division in which two daughter cells are physically separated. Accurate spatial and temporal coordination of cytokinesis with mitosis is necessary for faithful segregation of the genome. In order for most eukaryotic cells to divide they assemble an actin‐ and myosin‐based contractile ring (CR) that is physically linked to the plasma membrane. Precise lipid composition of the plasma membrane is known to impact cytokinesis. Phosphoinositides in particular are important for positioning the cleavage furrow and for maintaining physical linkage of the CR to the plasma membrane in animal cell cytokinesis, although the mechanism is not completely understood. Schizosaccharomyces pombe also utilize an actomyosin CR to complete cytokinesis much like animal cells. However, it is unclear what role lipid composition plays in S. pombe cytokinesis and whether it is conserved. Here, we investigated the effect of changing plasma membrane composition via deletion of the PI4‐kinase scaffold Efr3. We observed that S. pombe cells lacking Efr3 often divided with off‐centered septa. Interestingly, this phenotype results from sliding of the CR away from the cell center during anaphase rather than a failure to correctly position the assembling CR. These results suggest that plasma membrane lipid composition in S. pombe stabilizes the initial central positioning of the CR and promotes accurate completion of cell division. We reason that the affinity of several membrane‐binding proteins for the plasma membrane is reduced when the lipid composition is altered in the absence of Efr3, and that one or more of these proteins is important for stabilizing the CR. We have identified two proteins, the RhoGEF Rgf1 and the dual PH domain containing protein Opy1, that have reduced localization to the division site in efr3∆ cells. Previously, we determined the essential cytokinetic scaffold Cdc15 oligomerizes and binds membranes through its F‐BAR domain at the division site. Cdc15 mutants defective in oligomerization display a similar phenotype to efr3∆, with CR sliding from the cell middle during anaphase. Interestingly, the frequency and distance of CR sliding is exacerbated when both alleles are combined, suggesting multiple mechanisms exist for anchoring the CR to the plasma membrane in its medial position.


M210 Integrated model of cytokinetic ring constriction and septation in fission yeast reproduces experimental values of ring tension. S. Wang1, S. Thiyagarajan1, B. O'Shaughnessy2; 1Physics, Columbia University, New York, NY, 2Chemical Engineering, Columbia University, New York, NY Regardless of the organism, a major obstacle to establishing the mechanisms of cytokinesis is that contractile ring constriction is coupled to other processes such as contractility and flow of the actomyosin cortex, membrane addition, cytoplasmic flow and, in cell‐walled organisms such as fission yeast, septation, the inward growth of new cell wall in the wake of the constricting ring that completes the daughter cell walls. The ultimate goal is a quantitative model of the coupled system of ring constriction with these other processes. Here we developed such a model for fission yeast, whose cytokinetic ring is particularly well characterized. We integrated a 3D molecularly explicit simulation of the constricting ring, highly constrained by known amounts and biochemistry of many identified components, with a simulation describing the synthesis of the septum at its leading edge by beta glucan synthases (Bgs) in the plasma membrane. The Bgs‐mediated septum growth was assumed to be mechanosensitive and coupled to contractile ring tension. To our knowledge, this is the first quantitative account of the total constriction process during cytokinesis. The septum provided an inward‐growing edge to which the ring is attached. As constriction progressed, the length of the edge decreased and its curvature increased; both effects influenced the ring organization and tension. Influence occurred also in the reverse direction: at each location along the ring, the ring tension affected the mechanosensitive septum growth rate so as to reduce irregularities and maintain near‐circularity of the septum. Growth in small valleys in the septum edge was accelerated by the centripetal force from tension, tending to fill in the valleys, while growth at bumps was decelerated, tending to flatten the bumps. We quantified the non‐circularity of simulated septum edges by their roughness; edge roughness statistics from our simulations closely reproduced the statistics from measurements of septum edges in live cells (Thiyagarajan et al., J. Cell Sci., 2016). The ring tension strongly suppressed roughness and enforced proper septum closure, furnishing the two daughter cells with sealed cell walls. We used the model to compute ring tensions, and we compared these with ring tensions measured experimentally on live protoplast fission yeast cells using a novel method. The experimentally measured tensions varied from ~400 to ~800 pN as the ring constricted. The model reproduced these values, and showed that tension is generated by myosin‐II clusters pulling on actin filaments whose barbed ends are anchored to the plasma membrane, allowing tension to build up. The model also reproduced measurements in myo2‐E1 mutant cells which showed ~35% of the normal tension.

Minisymposium 21: Cell Migration and Invasion

M211 Moving beyond animal cell migration: deep evolutionary conservation of “alpha‐motility" and “blebbing‐motility”. L. Fritz‐Laylin1, S. Lord1, R.D. Mullins1; 1Molecular and Cellular Pharmacology, University of California San Francisco, San Francisco, CA Various cells scattered throughout the eukaryotic tree crawl across surfaces or through three‐dimensional environments. Evidence now indicates that cell crawling is not a single behavior, but rather


a collection of processes, driven by different molecular mechanisms. Understanding these mechanisms and their evolutionary relationships first requires narrowly defining mechanical modes of locomotion, and then identifying phenotypic and molecular markers of each. The best studied mode of cell migration is the adhesion‐based migration fibroblasts and epithelial cells, a mode limited to cells of the animal lineage. In contrast, we find two modes of crawling widely dispersed among eukaryotic phyla. The first, known as “blebbing” motility, involves delamination of the cell membrane resulting in bubble‐like protrusions. The second is characterized by dynamic, actin‐filled pseudopods and weak, nonspecific adhesion to external substrates, a mode we refer to as “alpha‐motility.” We are currently defining gene complements required for each mode, and using the resulting data to predict crawling motility in new species. For example, we use this approach to predict alpha‐motility in Chytrid fungi, a prediction we have verified using microscopy and small molecule inhibitors of actin cytoskeletal components. By developing mechanistic definition of distinct modes of crawling motility, and expanding our phylogenetic analysis to many eukaryotes, we are identifying genetic markers of alpha‐ and blebbing‐motility that can be used to understand the evolution of these key eukaryotic behaviors.

M212 Curvotaxis directs cell migration through cell‐scale topographical landscapes. L. Pieuchot1, M. Vassaux2, T. Cloatre1, I. Brigaud1, T. Petithory1, J. Milan2, M. Bigerelle3, K. Anselme1; 1IS2M, CNRS UMR 736, Mulhouse, France, 2ISM, CNRS UMR 7287, Marseille, France, 3LAMIH, CNRS UMR 8201, Valenciennes, France Cells evolve in vivo in complex 3D environments from which they integrate signals of different nature. For example, cells can detect the presence of nano‐ and micro‐scale topographic patterns in the extra‐cellular matrix, and these physical features can trigger stem cell differentiation to a specific lineage. Despite a large body of study on nano/micro topography, very little is known about how cell‐scale curvature ‐ which is a ubiquitous trait of natural biotopes ‐ can affect the physiology of the cell. To probe specifically this kind of topography, we develop a series of 3D sinusoidal surfaces presenting a continuous array of cell‐scale bumps and valleys with very low nanometric roughness, and monitored cell behavior on these biomimetic landscapes. We found that cells avoid convex regions during their migration and position their nuclei in concave valleys. This new sensing ability ‐ which we term “curvotaxis”‐ depends on acto‐myosin activity and relies on a dynamic interplay between the nucleus and the cytoskeleton. Computational modeling suggests that the nucleus acts as a mechano‐sensing organelle that drives cell migration towards valleys. Further analysis show that substratum concavity increases cellular and nuclear sphericity, lowers stress fiber tension and down‐regulates a subset of genes involved in stem cells differentiation. All together, this work identifies curvotaxis as a new guiding mechanism that enables adherent cells to read cell‐scale curvature gradients and suggest that cell‐scale curvature might be a true component of the stem cell niche.

M213 Intracellular mechanics and fluid dynamics in rapidly‐moving amoeboid cell motility. C.K. Chan1, T.Y. Tsai2, E.F. Koslover3, R.M. Garner1, A. Hadjitheodorou1, J.A. Theriot1; 1Biochemistry, Stanford University School of Medicine, Stanford, CA, 2Systems Biology, Harvard Medical School, Boston, MA, 3Physics, University of California, San Diego, San Diego, CA Amoeboid cell motility in neutrophils is characterized by persistence in travel, rapid directional changes, and dynamic morphology. To understand the mechanics of such versatile motility, we use both the neutrophil‐like HL60 cell line and primary human neutrophils to investigate the coordination of force‐


generating components of the cytoskeleton (particularly front‐rear coupling of actin and myosin), the interaction between these components and other cellular components such as the nucleus and the plasma membrane, the physical properties of the intracellular environment during amoeboid motility (including intracellular fluid flows and cytoplasmic viscoelasticity), and the morphological space that these motile cells explore, both with respect to overall cell shape and intracellular protein distributions. Our research on the coordination between actin and myosin using wide‐field fluorescence and TIRF microscopy suggests that rapid directional change is achieved through tight front‐rear coupling, with time‐delayed coupling between actin polymerization that drives the cell front followed by an increase in myosin II activity at the rear after 9‐15 seconds. This temporal coordination contributes to the timely re‐orientation of the uropod and rapidly aligns the cell in the new direction of travel. We also observed that the frequent changes in cell morphology during motility are accompanied by dynamic cytoplasmic flow of untethered organelles, such as granules and lysosomes, and re‐positioning of larger organelles, such as the nucleus. Through high‐frame rate fluorescent imaging and particle‐tracking analysis (BNEW), we find that untethered particles and organelles within the cytoplasm are shuttled along cytoplasmic flow that behaves as a viscous fluid. Re‐positioning of larger organelles, such as the single‐lobed nuclei in HL60 cells and multiple‐lobed nuclei in primary human neutrophils, also appears to be coupled to overall changes in cell shape and movement behavior. Finally, we mapped the neutrophil morphological space that contributes to motility behavior by cell shape mode and global protein distribution analysis. We find that amoeboid motility can be decomposed into several major shape modes, and that the shape space explored by a large population of cells is similar to the shape space explored by a single cell over time. Given this understanding, we aim to extend our analysis to make predictions on amoeboid motility behavior through the tracking of cell shapes and correlation with changes in intracellular protein distributions.

M214 FMNL formins are required for lamellipodial force generation. F. Kage1,2, M. Winterhoff3, V. Haas1,2, J.M. Müller4, T. Thalheim5, A. Freise2, S. Bruehmann3, J. Kollasser1, J. Block1, G. Dimchev1,2, M. Geyer6, C.H. Brakebusch7, T.E. Stradal1, M. Carlier8, M. Sixt4, J. Käs5, J. Faix3, K. Rottner1,2; 1Molecular Cell Biology, Helmholtz Centre for Infection Research, Braunschweig, Germany, 2Molecular Cell Biology, Technical University Braunschweig, Braunschweig, Germany, 3Institute for Biophysical Chemistry, Hannover Medical School, Hannover, Germany, 4Institute of Science and Technology , Klosterneuburg, Austria, 5Institut für experimentelle Physik , Leipzig University, Leipzig, Germany, 6Institute of Innate Immunity, University of Bonn, Bonn, Germany, 7BRIC, University of Copenhagen, Copenhagen, Germany, 8Cytoskeleton Dynamics and Motility, National Centre for Scientific Research, Paris, France Actin polymerization drives lamellipodia protrusion. However, protrusion efficiency and the determinants of force generation remain poorly defined. Lamellipodia are thought to be generated by Arp2/3 complex‐mediated filament branching, but potential contributions of formins are controversially discussed. We dissected the precise functions in protrusion of FMNL2 and FMNL3, two Diaphanous‐related formins regulated by the small Rho‐GTPase Cdc42. I n vitro , FMNL3 nucleated actin filaments more potently than FMNL2, but displayed lower profilin‐dependent elongation activity. Individual and combined RNAi‐based knockdown of both FMNLs in B16‐F1 cells as well as their CRISPR/Cas9‐mediated gene disruption consistently established that both formins not only contribute to the efficiency of protrusion, but also to lamellipodial width, actin filament density and bundling. Importantly, the ~50% reduction of lamellipodial actin filament mass was uncoupled from Arp2/3 complex activity, as incorporation of the latter into lamellipodial networks was unaffected by FMNL2/3 removal. Instead of


reduced velocities of lamellipodial actin network polymerization, as assumed previously, defects in lamellipodia protrusion appeared to be caused by reduced actin filament densities and modified angular distributions of filaments abutting the lamellipodial tip membrane. These structural changes coincided well with the substantial reduction of forces exerted by FMNL2/3‐deficient lamellipodia when compared to controls, as assessed by atomic force microscopy. Finally, major phenotypes of FMNL2/3 loss of function were recapitulated in CRISPR/Cas9‐treated NIH3T3 fibroblasts, confirming the general relevance of our findings. In conclusion, we propose that generation of physiologically relevant pushing forces by lamellipodia does not solely depend on Arp2/3 complex‐mediated branching, but also on the mechanical stability provided by FMNL formin‐generated filaments.

M215 FMN2 is a melanoma metastasis‐promoter that mediates formation of a perinuclear actin/adhesion system to protect nuclei and DNA from damage during confined cell migration. C.T. Skau1, R.S. Fischer1, P.S. Gurel1, H.R. Thiam1,2, A. Tubbs3, M.A. Baird1,4, M.W. Davidson4, M. Piel2, G.M. Alushin1, A. Nussenzweig3, P.S. Steeg5, C.M. Waterman1; 1Cell Biology and Physiology Center, National Heart, Lung, and Blood Institute, National Institutes of Health, Bethesda, MD, 2UMR 144 Institut Curie/CNRS, Institut Curie, Paris, France, 3Laboratory of Genome Integrity, National Cancer Institute, National Institutes of Health, Bethesda, MD, 4Magnet Lab, Florida State University, Tallahassee, FL, 5Center for Cancer Research, National Cancer Institute, National Institutes of Health, Bethesda, MD Cell migration in confined 3D tissue microenvironments is critical for development, immune function, and the dissemination of tumor cells in metastasis. How cells squeeze their large and stiff nuclei through such environments is unclear. We discovered a cytoskeletal structure that prevents damage to the nucleus and genetic material during migration in confined microenvironments. The formin‐family actin filament nucleator FMN2 associates with and generates a perinuclear actin/focal adhesion (FA) system that is compositionally and functionally distinct from previously characterized actin/FA structures. This system controls nuclear shape and positioning in cells migrating on 2D surfaces. In cells migrating in confined 3D microenvironments, FMN2 promotes cell survival by limiting nuclear envelope damage and DNA double‐strand breaks. We found that FMN2 is highly upregulated in a range of human melanomas, and show that disruption of FMN2 in mouse melanoma cells inhibits their extravasation and metastasis to the lung. Collectively our results indicate a critical role for FMN2 in generating a perinuclear actin/FA system that protects the nucleus and DNA from damage to promote cell survival during confined migration, and thus promote cancer metastasis.

M216 Adenomatous polyposis coli (APC)‐mediated actin nucleation is required for directed cell migration. M. Angeles Juanes‐Ortiz1, R. Jaiswal1, A. Badache2, B.L. Goode1; 1Biology, Brandeis University, Waltham, MA, 2Centre de Recherche en Cancerologie de Marseille , Institut Paoli‐Calmettes , MARSEILLE, France Adenomatous polyposis coli (APC) is a tumor suppressor protein with important roles in Wnt signaling as well as cytoskeletal rearrangements that govern cell migration, cell protrusion, and cell adhesion. Because APC is a large multi‐domain protein with a wide spectrum of activities and cellular binding partners, it has been difficult to pinpoint which of these activities and interactions are important for its different in vivo functions. To address this, we generated a separation‐of‐function allele in APC that


specifically disrupts its actin nucleation activity, a function that until now has only been studied in vitro. Our previous studies showed that a C‐terminal 78 kDa fragment of APC (2130‐2843), which includes the ‘basic domain’, interacts with actin monomers, nucleates actin assembly, and directly collaborates with formins via a ‘rocket launcher’ mechanism to promote actin assembly. Here, we mapped the nucleation activity to specific residues in the basic domain, and found that mutations at this site disrupt the ability of APC to stimulate actin nucleation in vitro, both alone and in collaboration with formins. Importantly, this mutation had no effect on APC’s ability to bind or bundle microtubules. In U2OS cells, silencing of endogenous APC led to a striking loss of microtubule and mitochondria staining at the cell cortex, and a pronounced defect in directed cell migration. Expression of an RNAi‐refractive wild‐type full‐length APC rescued all of these defects. However, full‐length mutant APC only rescued the microtubule and mitochondria defects, and not the cell migration defects. Thus, our results indicate that APC’s actin nucleation activity is required for its functions in directed cell migration, but not MT capture/stability or mitochondria transport. These findings also raise the intriguing possibility that APC‐mediated actin nucleation facilitates the migration of colonic crypt epithelial cells, and thus contributes to its role as a tumor suppressor.

M217 Notch signaling plays a crucial role in the signaling network that coordinates cell migration, organ size control and cell shape changes during morphogenesis. A. Kozlowskaja‐Gumbriene1, R. Alexander1, M. McClain1, T. Piotrowski1; 1Stowers Institute for Medical Research, Kansas City, MO Organ morphogenesis is driven by cell shape changes and organ size control. Even though the intracellular mechanisms leading to cell shape changes are fairly well understood, we are only beginning to understand how signaling pathways coordinate cell shape changes of select cells with the development of a whole tissue or embryo. The developing zebrafish lateral line is a powerful model to address this question, as the lateral line sensory organs self‐organize within a collectively migrating group of cells called the lateral line primordium. As the primordium migrates to the tail tip it periodically deposits sensory organs, which thereby form a line. The leading region of the primordium consists of mesenchymal cells, whereas in the trailing region cells apically‐basally polarize and apically constrict to form garlic bulb‐shaped sensory organs. Cell shape changes are driven by cortical tension‐generating actin‐myosin contractions that are transmitted to neighboring cells via apically localized cadherins. Our results show that Notch signaling ensures proper morphogenesis by regulating actin‐myosin driven apical constrictions, as well as the upregulation of cell adhesion and tight junction molecules. We also show that that Notch signaling is an integral part of the feedback loop between Wnt and Fgf signaling that coordinates cell migration with organ size control and the self‐organization of rosette‐shaped sensory organs in the zebrafish lateral line.

M218 HEMODYNAMIC PROFILES TUNE THE ARREST AND EXTRAVASATION OF CIRCULATING TUMOR CELLS. G. Follain1, N. Osmani1, G. Allio1, N. Fekonja1, S. Harlepp2, J.G. Goetz1; 1MN3T, iNSERM U1109, STRASBOURG, France, 2DON, IPCMS, STRASBOURG, France Cancer progression is a complex multi‐step process in which a growing and invasive primary tumor spreads to distant organs to form life‐threatening metastases. While multiple evidences suggest that metastatic seeding is driven by tumor cell‐intrinsic as well as local environmental cues, contribution of


mechanical cues such as vascular architecture or hemodynamic factors has been poorly studied. In our laboratory, we develop pioneer imaging tools allowing us to track single metastatic cells at high‐resolution in both mouse and zebrafish (ZF) models1. Using the latter, we aim to elucidate the impact of blood flow on the arrest, the adhesion and the extravasation of circulating tumor cells in vivo by combining intravital and high‐speed imaging as well as biophysical analysis of blood flow parameters. Because arrest of circulating tumor cells (CTCs) in vivo occurs mostly in hemodynamically favorable vessels, we questioned whether tuning blood flow would alter both their arrest and the extravasation efficacies. For doing so, we exploit the advantages of 2‐days post‐fertilization ZF embryos where hemodynamic profiles can be tuned using pharmacological approaches. We have successfully used several drugs that either increase or decrease flow velocities in the ZF embryo, without impact key traits of tumor cells in vitro (adhesion, migration). Interestingly, blood flow tuning affects multiple steps of tumor cell metastasis in the living ZF embryo. First, using live imaging of CTCs upon injection, we observed that low flow forces favored the initiation of stable tumor cell arrest while high flow forces tend to favor circulation of the CTCs, impeding its arrest. Using optical tweezers applied both in vivo and in vitro, we identified a threshold of flow and adhesion forces favoring successful arrest of CTCs, independently of the mechanical constraints imposed by vessel architecture. We then developed a quantitative mapping of the arrest of CTCs at 3h post‐injection (hpi) over several embryos and observed that tuning flow forces significantly altered the favored location of arrest of CTCs. Finally, using quantitative mapping of extravasation at 24 hpi, we observed that while low flow forces drastically impaired tumor cell extravasation, high flow forces stimulated the extravasation efficacies suggesting that flow forces can stimulate the transmigration of CTCs. Using our intravital CLEM approach, we are currently testing a potential contribution of the endothelium in this process. Altogether, these results demonstrate that hemodynamic profiles at metastatic sites can regulate important steps of the metastasis cascade. 1.Karreman, M. A. et al. Fast and precise targeting of single tumor cells in vivo by multimodal correlative microscopy. J. Cell Sci.129, 444–456, 201

M219 “Mechanobiology of epithelia on native basement membranes and relevance for cancer cell invasion”. M. Plodinec1,2, P. Oertle2, D. Assgeirson2, W. Halfter3, S. Eppenberger Castori1, E.C. Obermann1, A. Glentis4, D. Matic Vignjevic4, R.Y. Lim2; 1Institute of Pathology, University Hospital of Basel, Basel, Switzerland, 2Biozentrum and the Swiss Nanoscience Institute, University of Basel, Basel, Switzerland, 3Eye Hospital, University Hospital Basel, Basel, Switzerland, 4CNRS, Institut Curie, PSL Research University, Paris, France The onset of metastasis occurs when cancer cells invade and breach the basement membrane (BM) that provides mechanical support to epithelial tissues. Yet, it remains unclear what triggers cancer cells to breach the BM, and how ‘triggered’ cells breach the BM. Here, we have established an in vitro assay using native BM interface for culturing epithelial cells. Using atomic force microscopy (AFM) with other high‐resolution microscopies and TER (trans‐epithelial resistance), we have correlated the mechano‐cellular attributes of the BM/epithelium interface to its biochemical and structural properties. We demonstrated that the internal limiting membrane (ILM) isolated from human retinas acts as a native substrate for culturing epithelial cells in terms of BM composition, architecture and stiffness. These are required to act jointly in order to achieve apico‐basal polarity, tissue barrier formation and stiffness properties of the epithelial layer similarly to secretory epithelia in vivo. The native BM serves several advantageous over reconstituted Matrigel, an ECM extracts that originates from mouse tumor ascites. Besides variations in thickness and biochemical composition, we find that Matrigel is mechanically 100‐


fold more compliant (i.e., softer) than native BMs. We show that stiffness and architecture of the native alpha – 5 laminin chain has a key role in activating ß1 integrin and for the establishment of a physiological mechanophenotype. During cancer progression in vivo, cancer cells can perforate BM using proteolysis and the cancer cell invasion is associated with decrease in cellular stiffness and correlated to changes in cell and BM morphology. The role of stromal cells in this process has not yet been resolved.Therefore, we examined if carcinoma‐associated fibroblasts (CAFs) isolated from cancer patients promote cancer cell invasion through a BM. In the presence of CAFs, moderately invasive cancer cells invade in a matrix metalloproteinase‐independent manner. Using live imaging and atomic force microscopy, we could show that CAFs actively pull, stretch and soften the BM, forming gaps through which cancer cells can migrate. By exerting contractile forces, CAFs alter the organization and physical properties of native BM, making it permissive for cancer cell invasion. Finally, we propose that, in addition to proteolysis, mechanical interactions between CAFs and BM represent an alternative mechanism of BM breaching. Given the mechano‐biological relevance, native BMs allow us to further understand how mechanosignaling occurs between the epithelia and the surrounding stromal layers at the BM interface both in physiological and pathological states.

M220 Decreased lamin A in invasive breast cancer: a key node controlling mechanical and biochemical aspects of metastasis. E.S. Bell1,2, P. Shah2, P.M. Davidson1,2, A.L. McGregor1,2, G.R. Fedorchak1,2, P. Isermann1,2, J. Lammerding1,2; 1Meinig School of Biomedical Engineering, Cornell University, Ithaca, NY, 2Weill Institute for Cell and Molecular Biology, Cornell University, Ithaca, NY Changes in nuclear size and shape are used extensively in cancer diagnosis, with greater nuclear abnormalities correlating with worse outcome. The morphological and biophysical properties of the nucleus can be determined by the composition of the nuclear envelope. This includes the nuclear lamina, an intermediate filament network composed of A‐ and B‐type lamins underlying the nuclear membrane. In addition to providing strength and stability to the nucleus, A‐type lamins (A and C) are implicated in diverse cellular processes, such as DNA repair, transcription factor regulation, and chromatin organization, and have been shown to impact signaling networks important for tumorigenesis. Despite a growing understanding of the complex cellular functions of A‐type lamins, the role of lamins in regulating cellular mechanics and signaling in the context of cancer progression remains poorly understood. Using human and mouse model systems, we determined that lamin A/C expression is variable across breast cancer subtypes. The deformability of the nucleus also varies widely across breast cancers, and exhibits a strong inverse correlation with lamin A levels. Notably, cancer cells with greater metastatic capacity show decreases in both lamin A expression and nuclear stiffness. Since deformation of the nucleus can be a barrier to cell transit through tight spaces, we tested the impact of A‐type lamin levels on cell motility through confined microenvironments. We find that cells with naturally occurring lower levels of lamin A/C transit faster through micron‐scale constrictions and collagen matrices, and that confined cell migration can be modulated by adjusting expression of A‐type lamins. This suggests that decreasing lamin A/C levels provides a selective advantage for the transit of metastatic breast cancer cells through tissue environments. Furthermore, increasing lamin A expression results in decreased proliferation of aggressive breast cancer cells. Together, these studies indicate that downregulation of A‐type lamin levels could coordinate both tumor cell invasion and outgrowth, thus providing a key point of control over the development of metastases.


Minisymposium 22: Cell‐Fate Determination in Signaling and Differentiation

M221 Regulation of Cell Turnover During Epithelial Tissue Homeostasis. C. Brock1, S.T. Wallin1, E.A. Sumner2, G.T. Eisenhoffer1,2; 1Genetics, The University of Texas MD Anderson Cancer Center, Houston, TX, 2Graduate School of Biomedical Sciences, The University of Texas Health Science Center at Houston, Houston, TX Epithelial tissues covering our body and internal organs function as guardians against environmental insults. Because epithelial tissues provide the first line of defense, damaged or defective cells within epithelia must be continually eliminated and replaced to sustain a functional barrier. Changes in the balance between cell loss and division have been implicated in numerous human diseases, yet how these two processes influence each other to regulate overall cell numbers within epithelia remain poorly understood. Here we report that apoptotic basal cells produce WNT to stimulate proliferation of neighboring stem cells to maintain population numbers and drive cellular turnover. To investigate cellular turnover in a living epithelium, we identified a set of GAL4 enhancer trap lines that are expressed in discrete epithelial cell types in the developing zebrafish. When combined with UAS effector lines, our epithelial GAL4 lines provide the opportunity to visualize specific cells for imaging, overexpress genes of interest, and target cells for ablation. Using these tools, we created an epithelial wounding assay that allows induction of death specifically in a subset of the basal cells. Live imaging and electron microscopy experiments indicate that apoptotic cells are rapidly cleared by engulfment of the surrounding basal cells. Importantly, the remaining stem cells proliferate to compensate for the cell loss and maintain tissue integrity and function. Inhibition of cell death or WNT signaling eliminated the apoptosis‐induced division and led to failed regeneration, while genetic overexpression of WNT signaling in the context of a damage response led to an increase in overall cell numbers. Transcriptional profiling at defined times during the regeneration process uncovered distinct molecular pathways associated with the observed cellular behaviors. We are now investigating the changes that occur to cellular turnover when damaged cells are not properly eliminated. Together, this study provides a high‐resolution in vivo characterization of epithelial cell turnover and creates a system to rapidly identify new mechanisms controlling tissue homeostasis and the specific alterations that lead to pathologies.

M222 Hippo regulates Hedgehog signaling in digital morphogenesis. C. Tang1,2, X. Wu1,2; 1 Zhejiang University, Institution of Stem Cell and Regenerative Medicine, Hangzhou, China, 2School of Medicine, Zhejiang University, Department of Pharmacology, Hangzhou, China Hedgehog (HH) proteins comprise a family of secreted signaling molecules essential for many fundamental processes in embryonic development within metazoan. Here, we show that TAZ, a biologically potent transcriptional co‐activator in Hippo signaling, serves as an element of the Hedgehog signaling. This function of TAZ is independent from its well‐established role as mediator of Hippo signaling. TAZ restricts HH signaling by complex formation with Protein kinase A(PKA) in the cytoplasm, which further phosphorylates GLI3 in its carboxy‐terminal and sequentially initiates the proteasome‐mediated cleavage of full length GLI3 (GLI3‐F) into its transcriptional repressor form (GLI3‐R). As a result, GLI3‐R enters the nuclei and suppresses HH downstream target gene expression. Conversely, abrogation of TAZ levels by a variety of approaches effectively dampens the interaction between TAZ and PKA, leading to the failure in GLI3‐R formation in cytoplasm. Thus, GLI3‐F moves into the nuclei and induces


HH target gene expression. Moreover, mice lacking SHH gene develop severe defects in digital morphogenesis, while absence of TAZ obviously rescues the phenotype caused by SHH deletion and attenuates the target gene expression in digits as well. These results uncover a novel function of TAZ in regulating HH signaling and highlight the role of the Hippo pathway in coordinating morphogenetic signaling with digital morphogenesis.

M223 Centrosome Amplification Disrupts Renal Development and Causes Cystic Kidney Disease. K. Shim1, L. Dionne1, M. Hoshi1, V. Marthiens2, A. Knoten1, R. Basto2, S. Jain1, M.R. Mahjoub1; 1Dept of Medicine (Nephrology), Washington University, St Louis, MO, 2Institue Curie, Paris, France Cystic kidney diseases are characterized by hyperproliferation of normally quiescent renal epithelial cells, which profoundly alter the organ architecture and impair renal function over time. It is well established that defects in two essential microtubule‐based organelles, the centrosome and cilium, contribute to the cystic transformation of renal epithelial cells. The centrosome‐cilium complex acts as a cellular signaling center to organize and regulate the activity of various developmental signaling pathways. Mutations in proteins that localize and function through the centrosome‐cilium complex cause human disease conditions termed ciliopathies, which include polycystic kidney disease and nephronophthisis. Recent studies have noted the presence of ectopic centrosomal structures (meaning too many centrosomes per cell) in renal epithelial cells isolated from patients and animal models of polycystic kidney disease. Surprisingly, this phenotype has been mostly ignored, and considered a potential secondary effect of cystic cell transformation and proliferation. However, we hypothesized that abnormal centrosome biogenesis may play an important causal role in the pathogenesis of the disease. In this study, we make use of novel genetic mouse models with which we can alter centrosome biogenesis in vivo. We induced the formation of ectopic centrosomes in progenitor cells of the metanephric mesenchyme and the ureteric bud epithelium during the early stages of embryonic kidney development. We demonstrate, for the first time, that the formation of ectopic centrosomes disrupts embryonic kidney development and results in rapid cystogenesis. We also find that ectopic centrosome formation following renal development sensitize kidneys in adult mice, causing cystogenesis after renal injury. These results indicate that ectopic centrosome biogenesis alone is sufficient to trigger cyst formation and growth, even in the absence of mutations in cyst‐inducing genes. These studies further our understanding of the fundamental cellular events that trigger cystogenesis, and characterize a potentially new therapeutic target for treatment of cystic kidney disease.

M224 Protein phosphatases generate specificity in the fission yeast ARK signaling network. L. Deng1, M. Lee1, J.B. Moseley1; 1Biochemistry, Dartmouth, Hanover, NH Protein phosphatases generate specificity in the fission yeast ARK signaling network The AMPK‐related kinase (ARK) subfamily of protein kinases is essential for coordination of cell growth, proliferation, and migration with environmental status. Here, we demonstrate that all fission yeast ARKs are activated by phosphorylation at an identical amino acid sequence in their activation loop. The upstream kinase Ssp1 phosphorylates this shared sequence on each ARK protein. Despite sharing an identical sequence for phosphorylation by Ssp1, different ARKs are phosphorylated, activated, and required for growth under different conditions. We studied the source of specificity in this system, and show that that each ARK is inactivated by a distinct phosphatase to generate signal specificity. For


example, we find that Ssp1 phosphorylates AMPK/Ssp2 according to intracellular energy levels, and this activating phosphorylation is removed by the protein phosphatase PP6. In contrast, energy levels and PP6 activity do not affect Ssp1 phospho‐activation of a second ARK protein Cdr2, which functions to promote cell cycle progression. Rather, Cdr2 phosphorylation by Ssp1 is controlled by cell size, and this activating phosphorylation is reversed by PP2A. A third ARK protein, the cell polarity kinase MARK/Kin1, is rapidly phosphorylated and activated by Ssp1 under heat stress. Genetic and pharmacological data reveal dephosphorylation and inactivation of MARK/Kin1 by a phosphatase distinct from PP6 and PP2A. Our work identifies protein phosphatases as an unexpected source of signal specificity downstream of the master kinase Ssp1. We propose that phosphatases may play a key role in specifying signal output for other systems including the human ARK network.

M225 A Galpha‐s/PKA tumor suppressor pathway restraining YAP1 and GLI in epithelial stem cells. R. Iglesias‐Bartolome1; 1Laboratory of Cellular and Molecular Biology ( LCMB), National Cancer Institute (NCI), Bethesda, MD G‐protein‐coupled receptors (GPCRs) are the largest family of cell‐surface molecules involved in signal transduction, playing central roles in numerous physiological processes and pathological conditions along with their associated G proteins. Importantly, our understanding of the functions of GPCRs and their linked heterotrimeric G proteins in stem cell biology in vivo is still largely incomplete. Here, by using a combination of knockout animals and conditional deletion of genes, we have demonstrate that among Gα proteins, Gαs plays a central role in coordinating self‐renewal and differentiation in epithelial stem cells. In mice, conditional epidermal deletion of the gene coding for the Galpha‐s subunit (Gnas) was alone sufficient to cause an aberrant expansion of the stem cell compartment, resulting in the rapid formation of basal cell carcinoma‐like lesions. Remarkably, this could be phencopied by inactivation of the Gαs signaling partner protein kinase A (PKA). In contrast, expression of an active form of Galpha‐s in the skin caused hair follicle stem cell exhaustion and the consequent hair loss. Mechanistically, we found that Galpha‐s/PKA disruption promotes the concomitant cell autonomous stimulation of Sonic Hedgehog (SHH) pathway and inhibition of Hippo signaling, resulting in the non‐canonical activation of GLI and YAP1. These findings support a central role of Galpha‐s and PKA in stem cell fate decisions in mammals, and reveal a tumor suppressive mechanism by which the Galpha‐s/PKA signaling axis limits the aberrant proliferation of epithelial stem cells, maintaining hair follicle and skin homeostasis.

M226 Comparative forward genetic screens in haploid human cells reveal new regulatory mechanisms in WNT signaling. A.M. Lebensohn1, R. Dubey1, L.R. Neitzel2, O. Tacchelly‐Benites3, E. Yang3, C.D. Marceau4, E.M. Davis5, B.B. Patel1, Z.B. Nejad1, Y. Ahmed3, E. Lee2, J.E. Carette4, R. Rohatgi1; 1Departments of Biochemistry and Medicine, Stanford University School of Medicine, Stanford, CA, 2Department of Cell and Developmental Biology, Vanderbilt University Medical Center, Nashville, TN, 3Department of Genetics and the Norris Cotton Cancer Center, Geisel School of Medicine at Dartmouth College, Hanover, NH, 4Department of Microbiology and Immunology, Stanford University School of Medicine, Stanford, CA, 5Department of Molecular, Cellular, and Developmental Biology, University of Colorado at Boulder, Boulder, CO The comprehensive understanding of cellular signaling pathways remains a challenge due to multiple layers of regulation that may become evident only when the pathway is probed at different levels or


critical nodes are eliminated. The WNT pathway is a fundamental signaling system that plays central roles in embryonic development, regeneration and cancer. The principal control point in WNT signaling is the regulated degradation of the transcriptional co‐activator beta‐catenin by the destruction complex (DC), a multi‐protein assembly composed of scaffolds, kinases and an E3 ubiquitin ligase. In the absence of WNT signals, beta‐catenin is targeted for proteasomal degradation by the DC. Binding of WNT to its receptors on the cell surface causes inactivation of the DC, resulting in beta‐catenin accumulation and activation of target genes. While the overall anatomy of the pathway has been established, its complex circuitry may mask unknown regulatory mechanisms overlaid on the core beta‐catenin degradation module, making it an ideal system for an in‐depth, methodical genetic dissection. In order to discover new regulatory mechanisms, we conducted a systematic forward genetic analysis through reporter‐based screens in haploid human cells—the presence of a single copy of each gene enables true loss‐of‐function screens through insertional mutagenesis. A comparative analysis of genome‐wide screens for negative, sensitizing and positive regulators of WNT signaling, as well as mediators of R‐spondin‐dependent signaling (R‐spondins are secreted ligands that potentiate WNT signaling), revealed new regulatory features at all levels of the pathway, from signal reception to transcriptional activation. We discovered a requirement for the transcription factor TFAP4 downstream of beta‐catenin, a role for the DAX domain of the DC scaffold AXIN2 in coupling beta‐catenin protein abundance and transcriptional activity, and a requirement for glycophosphatidylinositol (GPI) anchor biosynthesis in R‐spondin reception. Screens for suppressors of constitutive signaling induced by loss of the tumor suppressor APC or casein kinase 1alpha revealed two different mechanisms that regulate WNT signaling when distinct components of the DC are lost: regulation of beta‐catenin protein abundance by the RNA‐binding protein SERBP1, and a DC‐independent role for casein kinase 1alpha in control of signaling strength, mediated by the E3 ubiquitin ligase HUWE1. Our findings, supported by experiments in model organisms, highlight new ways in which WNT signaling can be regulated beyond control of beta‐catenin degradation. More generally, we demonstrate the power of forward genetics in human cultured cells to discover components and regulatory principles in signaling pathways.

M227 Asymmetric activity without asymmetric localization: the KA1 domain regulates PAR‐1 activity during cell polarization. A.W. Folkmann1, G. Seydoux1; 1Department of Molecular Biology and Genetics, Johns Hopkins University School of Medicine/HHMI, Baltimore, MD Asymmetric localization of polarity regulators is a hallmark of polarized cells. For example, in C. elegans, the polarity kinase PAR‐1 localizes to the posterior cortex of the zygote under the influence of the anterior kinase PKC‐3. Asymmetric localization of PAR‐1 is thought to drive other cellular asymmetries, including the posterior localization of P granules. We have obtained evidence that PAR‐1 protein asymmetry is not essential, and that PAR‐1 activity is also regulated spatially. Using genome editing, we mutated the PKC‐3 phosphorylation site in PAR‐1 (T983A). As expected, PAR‐1(T983A) was symmetrically distributed throughout the zygote. Surprisingly, however, the P granules still localized to the posterior. The KA1 (kinase‐associated) domain of PAR‐1 is required for PAR‐1 localization to the cortex. We found that deletion of the KA1 domain completely eliminated the ability of PAR‐1(T983A) to polarize P granules and caused PAR‐1 to be active throughout the cytoplasm. Consistent with these results, we found that in vitro, the KA1 domain acts as an intra‐molecular inhibitor of PAR‐1 kinase activity, as has been shown for related kinase in S. cerevisiae KIN2 and human MELK1‐2.


Deletion of the KA1 domain did not affect the polarization function of wild‐type PAR‐1, suggesting that auto‐inhibition is not essential when PAR‐1 is asymmetrically localized. Together these data suggest a model whereby PAR‐1 asymmetry is regulated redundantly by two parallel mechanisms: phosphorylation by PKC‐3, which localizes PAR‐1 protein, and auto‐inhibition by KA1, which localizes PAR‐1 activity. How the latter is regulated spatially is the focus of our current investigations. References: [1] Beullens M, Vancauwenbergh S, Morrice N, Derua R, Ceulemans H, Waelkens E, Bollen M. (2005) Substrate specificity and activity regulation of protein kinase MELK. J Biol Chem. 280(48):40003‐11. [2] Elbert M, Rossi G, Brennwald P. (2005) The yeast par‐1 homologs kin1 and kin2 show genetic and physical interactions with components of the exocytic machinery. Mol Biol Cell. 16(2):532‐49.

M228 Hippo Pathway target Rae1 regulates mitosis, organ size, and feeds back to regulate Hippo Pathway Homeostasis. M. Jahanshahi1, K. Hsiao1, A. Jenny2, C.M. Pfleger1; 1Oncological Sciences, Icahn School of Medicine at Mount Sinai, New York, NY, 2Developmental and Molecular Biology, Albert Einstein College of Medicine, New York, NY The Hippo Pathway is a master regulatory signaling pathway that restricts growth, proliferation, and organ size, promotes apoptosis and informs cell fate decisions. By coordinating these processes the pathway plays key roles throughout development and in mature tissues including proper regulation of the cell cycle, reaching and maintaining a pre‐set organ, and tumor suppression. We identified WD repeat protein Rae1 as a highly conserved target inhibited by the Hippo Pathway downstream of the Warts/Lats kinase that coordinates proliferation and organ size in parallel to Yki/YAP. Activation of the Hippo Pathway promotes Rae1 degradation in Drosophila and mammalian systems. Unlike the Yki/YAP oncogene which is important for cell growth and can promote cell cycle reentry, our studies in Drosophila tissues suggest Rae1 cannot promote cell cycle reentry of differentiated tissues, so is not an “on/off” switch like Yki/YAP. However, in proliferating cells, Rae1 acts like a “rheostat” or “dial” to change the rate of proliferation and to adjust organ size. Decreasing Rae1 levels in actively proliferating cells decreases proliferation and organ size, in part by downregulating cyclin B. In contrast, increasing Rae1 levels in proliferating cells increases proliferation and also increases overall organ size. Genetic interaction studies indicate that the downregulation of Rae1 is important for Hippo Pathway downregulation of cyclin B and organ size and that accumulation of Rae1 is important for the increased cyclin B and organ overgrowth phenotypes upon loss of Hippo Pathway tumor suppressor components such as Hippo and Warts. Surprisingly, Rae1 is required not only for overgrowth but also for survival of such overgrowing tissue. Parallel studies with Yki/YAP revealed a highly conserved role for Rae1 to inhibit Yki/YAP activity and downregulate Yki/YAP levels. Experiments in culture cells and in vivo exploring this unexpected result revealed that Rae1 acts in a feedback circuit to promote activation of the core kinases Hippo/MST and Warts/Lats in part by promoting accumulation of Merlin/NF2, Hippo/MST, and Warts/Lats protein levels and also by promoting membrane recruitment of Warts/Lats. This feedback is post‐transcriptional and likely reinforces the transcriptional feedback mediated by Yki/YAP. Thus, Rae1 is downregulated by the Hippo Pathway to coordinate proliferation and organ size and also to feedback to regulate pathway homeostasis.


M229 Rap1 and Hippo pathway collaborate to polarize directional protrusions in Drosophila border cell migration. Y. Chang1,2, Y. Hsieh2, T. Huang2, D.J. Montell3, C.A. Jang2; 1Institute of Molecular Biology, Academia Sinica, Taipei, Taiwan, 2Institute of Biotechnology, National Cheng Kung University, Tainan, Taiwan, 3Molecular, Cellular and Developmental Biology, University of California, Santa Barbara, Santa Barbara, United States Actin polymerization provides force to project membrane forward and maintain lamellipodia persistence during cell movement, which involves dynamic regulation by actin nucleators and elongators. In collective migration, directional protrusions orient motile cells in response to external cues including graded chemokines or the rigidity of substratum. Such directionality is achieved by coordinated polarity among the migrating cohort, failure of which causes migration delay or even group scattering. However, the mechanism by which polarized protrusions are coordinated remains less understood. Border cells are derived from the follicle epithelium of egg chambers in Drosophila and migrate collectively along guidance gradients. The cluster invades through the germline cells, nurse cells, by dynamic association of E‐cadherin, which offers a well‐defined model to study how group polarity is organized and how mechanical cues affect cell motility in vivo . With genetic analyses, live imaging, and biochemical studies, we explore that a small GTPase, Rap1, is required to polarize border cells to extend protrusions by collaboration with the Hippo signaling pathway. Depletion of Rap1 impeded border cells from reaching the destination, oocytes. Further investigations showed that constitutive activation of Rap1 by Rap1V12 disrupted the front/back polarity in border cell clusters, in terms of evenly‐distributed actin network and disoriented protrusions, and scattered border cells were observed more frequently. By contrast, suppression of Rap1 activity by over‐expressing the dominantly‐negative form, Rap1N17 , made the clusters spin around with fewer protrusions, leading to impaired movement. Consistently, random induction of Rap1V12 resulted in preferential occupancy of border cells in the front of migrating clusters, but conversely, cells expressing Rap1N17 lagged behind the others. To uncover genes involved in Rap1‐dependent polarity, we conducted a forward genetic screen and identified two core components of Hippo pathway, hippo (hpo) and mob as tumor suppressor (mats). Down‐regulation of either gene not only enhanced Rap1V12‐induced migration defect but also increased the protrusion number both in forward and backward directions. In Rap1V12 over‐expression, lack of one hpo allele also further magnified the expression of Enable/Vasodilator‐stimulated phosphoprotein (Ena/VASP) that promotes actin filament elongation. More importantly, pull‐down analysis revealed the interaction of Rap1V12 and Hpo protein. Taken together, our results suggest that at the leading edge of migrating border cells, activated Rap1 binds to Hpo to unleash its downstream effect on inhibiting Ena‐F‐actin dependent protrusions, which thereby advances the migratory group.

M230 Hippo pathway in cell growth, organ size, and tumorigenesis. T. Moroishi1, K. . Guan1; 1Pharmacology, UCSD, La Jolla, CA The Hippo pathway is crucial in organ size control and its dysregulation contributes to tumorigenesis. Core components of the Hippo pathway include the protein kinases of MST1/2, MAP4Ks, and LATS1/2 as well as the transcription co‐activators of YAP/TAZ. Previous studies have established the biochemical mechanism of regulation of the Hippo pathway effector YAP oncoprotein. Lats phosphorylates YAP to promote cytoplasmic localization and degradation, thereby inactivating YAP. We have discovered that


the Hippo pathway is regulated by a wide range of signals, including cell density, G‐protein couple receptor signaling, cytoskeletal structure, and cellular energy levels. Our studies have established molecular mechanisms linking the Hippo‐YAP pathway to extracellular and intracellular signals. Uncontrolled YAP activation is frequently found in a large numbers of human cancers, particularly in uveal melanoma and malignant mesothelioma. A critical role of YAP activation in human cancers is emerging and suggests potential therapeutic value of inhibiting YAP for cancer treatment.

Minisymposium 23: Membrane Traffic Control By Cytoskeletal and Molecular Machines

M231 A three‐dimensional extracellular matrix enhances cell viability by increasing negative membrane curvature to stimulate Arf6/Rac/Pak activity. F. Kai1, G. Ou1, J. Friedland1, C. Frantz1, R. Tourdot2, W. Guo3, C.S. Chen4, R. Radhakrishnan2,5, A. Long6, S. Dumont6, V.M. Weaver1,7,8,9,10; 1Surgery, University of California, San Francisco, San Francisco, CA, 2Chemical and Biomolecular Engineering, University of Pennsylvania, Philadelphia, Philadelphia, PA, 3Biology, University of Pennsylvania, Philadelphia, Philadelphia, PA, 4Biomedical Engineering, Boston University, Boston, MA, 5Bioengineering, University of Pennsylvania, Philadelphia, Philadelphia, PA, 6Department of Cellular and Molecular Pharmacology, University of California, San Francisco, San Francisco, CA, 7Anatomy, University of California, San Francisco, San Francisco, CA, 8Bioengineering and Therapeutic Sciences, University of California, San Francisco, San Francisco, CA, 9Eli and Edythe Broad Center of Regeneration Medicine and Stem Cell Research, University of California, San Francisco, San Francisco, CA, 10Helen Diller Comprehensive Cancer Center, University of California, San Francisco, San Francisco, CA Tumor metastasis is limited by the ability of cells to survive varying microenvironmental stresses. Metastatic cancer cells encounter a heterogeneous two dimensional (2D) or three dimensional (3D) extracellular matrix (ECM) landscapes with variable biophysical properties. Here, we set out to explore whether ECM dimensionality contributes to cell survival, and if so, how? We used micropatterned adhesive surfaces and compliant polyacrylamide gels conjugated with defined ECMs to recapitulate ECM landscapes and applied computational modeling and biophysical and biochemical approaches to examine the role of ECM dimensionality in cell survival. Consistent with prior studies, non‐spread mammary epithelial cells (MECs) plated on either a soft 2D polyacrylamide gels or on rigid micropatterned adhesive islands died whereas their spread counterparts survived. However, the non‐spread cells were able to survive when they were overlaid with either laminin or reconstituted basement membrane (rBM). Computational modeling predicted that MECs encountering a 3D ECM should exhibit reduced cortical tension and increased negative plasma membrane curvature. Indeed, we verified the prediction via AFM nanoindentation and laser ablation and through high resolution imaging of actin dynamics and membrane protrusions. The model also predicted that negative membrane curvature should preferentially recruit pro‐survival proteins that sense or induce negative membrane curvature, such as Arf6 GEFs. Indeed, MECs interacting with a 3D laminin ECM had significantly higher Arf6 activity and increased its downstream pro‐survival Rac/PAK signaling. Loss of function and gain of function studies also revealed that Rac and Pak signaling was required for MEC survival in a 3D ECM. Moreover, knockdown of key Arf6 GEFs or reducing levels of negative curvature‐inducing protein Exo70 compromised MEC survival in a 3D ECM, while loss of key Arf GAPs enhanced the survival of non‐spread MECs in 2D. Importantly, reducing cortical tension of non‐spread MECs on a 2D ECM through myosin inhibition promoted negative curvature in MECs and cell survival. Our results provide the first evidence


demonstrating that ECM dimensionality altered biophysical properties of cells, which modulates Arf6/Rac/Pak signaling axis to dictate cell viability. Our findings also offer a unique perspective for why Arf6 and Rac GTPases and several of their GEFs have been implicated in cancer aggression and metastasis and suggest that targeting the tissue ECM may provide a novel therapeutic approach to ameliorate tumor dissemination. Supp: CIHR, F.B.K., NSF GRFP, G.O., and a DOD Scholar expansion award BC122990, NIH R01 grants CA138818‐01A1, CA085492‐11A1, CA192914‐01 and GM R01 GM059907‐13 and U01 grant CA202241, V.M.W.

M232 Single‐molecule dynamics of IFT particles and motors at the tip of C. elegans cilia. J. Mijalkovic1,2, J. Van Krugten1,2, F. Oswald1,2, S. Acar1,2, E.J. Peterman1,2; 1Physics and Astronomy, VU University, Amsterdam, Netherlands, 2LaserLab, Amsterdam, Netherlands Bidirectional transport driven by motor proteins is essential for the proper distribution of cargo, and therefore vital for many cellular processes. Cilia are polar, microtubule‐based cellular sensing hubs that rely on a process called intraflagellar transport (IFT) for their development, maintenance and function in signal‐transduction (Scholey, J.M., 2003). IFT trains, consisting of cargo and the IFT‐A and IFT‐B protein complexes, are assembled at the ciliary base and driven co‐operatively by kinesin‐II and OSM‐3 motors to the ciliary tip (Wei, et al., 2012, Prevo, et al., 2015). The trains reverse direction at the tip and are transported back to the base by IFT dynein. The mechanism of IFT turnaround at the ciliary tip remains unknown. Here, we employ single‐molecule fluorescence microscopy and single‐particle tracking in the phasmid cilia of living C. elegans to probe IFT tip turnaround. Single‐molecule trajectories reveal direct, pausing and diffusive turns in a sub‐micrometer long turnaround region at the ciliary tip. Strikingly, while most IFT dyneins, OSM‐3's and IFT‐A particles turn almost instantaneously (within 600ms), IFT‐B particles pause on average for 3 s before returning, with pauses lasting as long as 15 s. Further analysis reveals that IFT dynein and OSM‐3 also exhibit diffusive behavior at the ciliary tip, whereas IFT‐A and IFT‐B are more spatially constrained. Our findings suggest that IFT trains dissociate at the tip and re‐associate in retrograde moving trains. IFT‐B, different than the other components, requires substantial remodeling while remaining spatially constrained at the tip, before it docks to an IFT‐dynein driven retrograde train. Stochastic simulations of several tip turnaround scenarios support this model. Our data provides the first in vivo single‐molecule quantification of IFT tip turnarounds, providing new insights into this vital part of bidirectional transport in sensory cilia.

M233 Spatio‐temporal dynamics of small GTPases and IRSp53 drive CLIC/GEEC endocytic vesicle formation via ARP2/3 based actin polymerization. M. Sathe1, G. Muthukrishnan1, M. Thattai1,2, S. Mayor1,3; 1National Center for Biological Sciences Science (TIFR), Bellary Road, Bangalore, India, 2Simons Centre for the Study of Living Machines, (NCBS‐TIFR), Bangalore, India, 3Institute for Stem Cell Biology and Regenerative Medicine, Bangalore, India Numerous endocytic pathways operate simultaneously on the eukaryotic cell surface. Clathrin‐dependent endocytic pathways have been studied extensively in the context of receptor mediated uptake. Clathrin‐ independent endocytic pathways are emerging as critical players in many cellular processes 1,2. The Clathrin Independent Carrier/GPI‐anchored protein Enriched Endosomal Compartments (CLIC/GEEC; CG) endocytic pathway is responsible for the rapid turnover of the plasma


membrane in fibroblasts 3, and internalizes GPI‐anchored proteins as well as a major fraction of the fluid‐phase in many cell types. It is very sensitive to perturbation of actin polymerization and cholesterol levels at the plasma membrane 4, and does not require dynamin function . Although the CG pathway is regulated by small GTPases such as CDC42 5 and Arf1 6, in the absence of any discernible coats on the CLICs or known scission factors, the structural components that assist in CG vesicle generation are yet to be understood. Here, we focus on the spatiotemporal dynamics of molecules regulating the CG pathway. We developed a real‐time, TIRF‐based assay using a pH‐pulsing protocol along with pH‐sensitive pHlourin‐GPI to identify nascent CG endocytic sites. The temporal profile of the recruitment of known modulators of this pathway to the endocytic site shows that the GTPase/GEF pair ‐ ARF1/GBF1 and CDC42 are recruited nearly 40s and 10s prior to scission, respectively. Pharmacological perturbations of ARP2/3 but not formins inhibit CG endocytosis, suggesting a role for ARP2/3 actin nucleators. Correspondingly, the ARP2/3 complex is recruited to the endocytic sites, contemporaneous with ARF1, and F‐Actin build up occurs 15s later. In a screen for the role for BAR domain we identified an I‐BAR domain protein, necessary for CG endocytosis and show that it is also recruited to the forming CG endocytic vesicle. These studies providing a mechanistic understanding of the molecular machinery linking the small GTPases, ARF1 and CDC42 with ARP2/3 and BAR domain proteins towards building a CG endocytic vesicle. 1. Mayor, S., Parton, R. G. & Donaldson, J. G. Clathrin‐independent pathways of endocytosis. Cold

Spring Harb. Perspect. Biol. 6, (2014). 2. Johannes, L., Parton, R. G., Bassereau, P. & Mayor, S. Building endocytic pits without clathrin. Nat.

Rev. Mol. Cell Biol. 42, 1–11 (2015). 3. Howes, M. T. et al. Clathrin‐independent carriers form a high capacity endocytic sorting system at

the leading edge of migrating cells. J. Cell Biol. 190, 675–691 (2010). 4. Chadda, R. et al. Cholesterol‐sensitive Cdc42 activation regulates actin polymerization for

endocytosis via the GEEC pathway. Traffic 8, 702–17 (2007). 5. Sabharanjak, S., Sharma, P., Parton, R. G. & Mayor, S. GPI‐anchored proteins are del

M234 Actin‐dependent ectocytosis and BBSome‐mediated retrieval drive the removal of activated GPCRs from cilia. A.R. Nager1, F. Ye1, V. Herranz‐Pérez2, D. Portran1, J.S. Goldstein1, J. Manuel Garcia‐Verdugo2, M.V. Nachury1; 1Department of Molecular and Cellular Physiology, Stanford University, Stanford , CA, 2Instituto Cavanilles, Laboratorio de Neurobiología Comparada, Universitat de València, CIBERNED, Spain The primary cilium is a microtubule‐based organelle that functions in Hedgehog and numerous G‐Protein Coupled Receptor (GPCR) signaling pathways in mammalian cells. A hallmark of ciliary signaling pathways is the dynamic delivery and removal of ciliary proteins. Yet ciliary trafficking remains incompletely characterized. We find that activated ciliary GPCRs are retrieved back into the cell through β‐Arrestin2‐ and BBSome‐mediated transport. Surprisingly, when retrieval is compromised, activated GPCRs become selectively concentrated into membranous buds at the tip of cilia, and actin‐mediated scission releases ectosomes packaged with activated GPCRs and harboring low levels of ciliary membrane proteins. Actin‐mediated ectosome severing requires the microtubule‐actin crosslinker Drebrin and Myosin VI. Myosin VI and Drebrin localize to the site of ectosome release at the ciliary tip under ectosome‐producing conditions. Signal‐dependent ectocytosis also applies to physiological contexts where it removes from cilia specific GPCRs such as NPY2R that lack retrieval determinants. Moreover, protein removal from cilia affects signaling as simultaneous inhibition of both the retrieval


and ectocytosis pathways blocks a key event of Hedgehog signaling. Our data show that signal‐dependent ectocytosis is an evolutionarily conserved process that enables the tuning and dissemination of ciliary signaling.

M235 Activation cycle of β‐arrestin allowing independent trafficking and signaling functions. K. Eichel1, B. Barsi‐Rhyne1, M. von Zastrow1; 1Department of Psychiatry, University of California, San Francisco, San Francisco, CA How membrane trafficking and cellular signaling are interconnected is a fundamental question in cell biology. We recently discovered a clear example of such reciprocal control between signaling and endocytosis through the study of β‐arrestin, a critical player in the signaling and trafficking of G protein‐coupled receptors (GPCRs), the largest signaling receptor family. Using live‐cell quantitative fluorescence imaging, we found that β‐arrestin, previously thought to exclusively function in an obligate physical complex with its activating GPCR, can carry out trafficking and signaling functions independently after dissociating from its activating GPCR (Nature Cell Biology, March 2016). We discovered this surprising property of β‐arrestin function through study of a particular GPCR, the β1 adrenergic receptor. We now show that this is a general property of β‐arrestin, elicited by numerous GPCRs in which β‐arrestin dissociates from its activating GPCR to capture clathrin‐coated pits (CCPs), stall CCP maturation, and repurpose CCPs into signaling stations on the plasma membrane. We delineate the biochemical cycle of this β‐arrestin signaling and explain how it is energetically driven by the clathrin‐mediated endocytic process. This cycle involves discrete stages of β‐arrestin conformational activation, which enables great diversity in endocytic behaviors of GPCRs. We also show how this novel β‐arrestin cycle can be differentially manipulated by clinically relevant drugs acting on the same GPCR. These results define a new framework for understanding cellular β‐arrestin function and the signaling consequences of clathrin‐mediated endocytosis.

M236 The binding energy of the atlastin crossover conformation drives membrane fusion. J. Winsor1, T.H. Lee1; 1Biological Sciences, Carnegie Mellon University, Pittsburgh, PA The homotypic fusion of endoplasmic reticulum (ER) membranes by the membrane bound GTPase atlastin is required for maintaining the highly branched structure of the ER. The fusion mechanism is not yet well understood, but involves a conformational shift termed crossover, which is catalyzed by GTP hydrolysis. Crossover formation moves atlastin from a “pre‐fusion” extended dimer, with the GTPase heads contacting one another and the three helix bundles (3HBs) ‐ attaching each head to its transmembrane domain ‐ extending in opposite directions, into a compact dimer in which the 3HBs have crossed over one another to form extensive new contacts with one another as well as with the opposing heads. In the compact crossover dimer state, the 3HBs appear to extend from the same membrane, resembling the “post‐fusion” state of previously studied fusion catalysts such as viral fusion proteins and SNAREs. Based on previously studied fusion mechanisms, it may be that the extensive energy released by the substantial buried surface area of the crossover dimer provides the driving force for membrane fusion. However, several observations have complicated this explanation. First, GTP hydrolysis, which catalyzes both dimerization and crossover, has been linked to membrane tethering, a step distinct from and upstream of the fusion step. Second, fusion by atlastin depends critically on membrane insertion of an amphipathic helix in the tail. Consequently, more recent models have proposed that atlastin


crossover acts primarily to tether and hold membranes in close proximity, while insertion of the tail into the bilayer triggers fusion. Here, we constructed a series of mutations in an important linker, connecting the GTPase head to the 3HB, to target crossover without altering other portions of atlastin involved in the fusion reaction. These mutants were analyzed in a series of fluorescence based kinetic measurements. Mutations causing defects in crossover inhibited fusion, but did not inhibit tethering. Furthermore, defects in crossover were not associated with defects in initial head to head dimerization. Most importantly, defects in crossover caused losses in the stability of the atlastin dimer in the crossover state, a property directly related to the Gibbs free energy of the “post‐fusion” state, and these losses were proportional to the level of fusion defect. Taken together the data indicate that atlastin crossover formation underlies fusion rather than tethering, and suggest that formation of the crossover state provides a major source of energy for initiating lipid bilayer fusion.

M237 Membrane fission by protein crowding. W.T. Snead1, C.C. Hayden1, A.K. Gadok1, P. Rangamani2, J.C. Stachowiak1; 1Biomedical Engineering, The University of Texas at Austin, Austin, TX, 2Mechanical and Aerospace Engineering, University of California, San Diego, La Jolla, CA Compartmentalization of biological processes into discrete, membrane‐bound volumes is essential for cellular life. The process of membrane fission drives the formation of new compartments from initially continuous membrane surfaces. Proteins with specific structural features and assembly properties, including helical scaffolds and hydrophobic insertions, are traditionally viewed as the primary mediators of fission in the cell. In contrast, here we report a previously unknown mechanism of membrane fission, which is independent of protein shape and assembly – protein crowding. Specifically, when proteins strongly adhere to the membrane surface at high coverage, in‐plane collisions among diffusing proteins generate substantial surface pressure. If this pressure is not balanced by an equivalent pressure on the opposite membrane surface, the resulting gradient in surface pressure dramatically increases membrane curvature, leading to spontaneous membrane fission. By correlating FRET‐based measurements of the membrane coverage by adhered proteins with multiple, independent measurements of membrane fission including membrane shedding, fluorescence‐based quantification of vesicle size, and electron microscopy, we show that both fission efficiency and the size of fission products are correlated with the coverage of protein on the membrane. Specifically, using Epsin1, a clathrin adaptor previously thought to promote fission through insertion of a wedge‐like hydrophobic helix into the membrane, we show that membrane coverage is the main determinant of fission efficiency and vesicle size, regardless of the hydrophobicity of the inserting helix. Importantly, we show that increasing the size of the adhered protein leads to increased fission efficiency, in line with a crowding mechanism driving fission. Our data contrast starkly with the established view that hydrophobic insertions drive membrane fission directly. Instead, we show that hydrophobic insertions are more important for strongly anchoring proteins to the membrane surface, to create highly crowded coverages of protein that favor fission. Further, our results provide a potential explanation for how fission may have been achieved in early cells, before the evolution of complex fission machines like dynamin. In particular, our findings demonstrate that any protein, regardless of its structure or assembly properties, can in principle drive fission, suggesting that the basic energetic criteria for membrane fission may be met by a much greater variety of proteins than previously thought.


M238 Reconstitution of endocytic actin networks on supported lipid bilayers. E.H. Stoops1, M. Wojcik2, S. Low‐Nam2, J.T. Groves2, K. Xu2, D.G. Drubin1; 1Molecular and Cell Biology, University of California, Berkeley, Berkeley, CA, 2Chemistry, University of California, Berkeley, Berkeley, CA Clathrin‐mediated endocytosis (CME) is a conserved process responsible for the selective uptake of nutrients and plasma membrane components. CME in yeast involves the ordered plasma membrane recruitment, activity, and disassembly of ~60 endocytic proteins. Over half of these proteins are involved in actin assembly. A burst of actin assembly in the late stages of CME generates the force required to overcome the turgor pressure of the yeast cell wall and to drive membrane remodeling events during CME. Previously, our group reconstituted actin tail assembly on microbeads and used this assay to identify novel regulators of endocytic sites. Despite the crucial role of the actin cytoskeleton in endocytosis, key ultrastructural information is lacking. To further investigate actin assembly during endocytosis and to gain insight into the organization of actin filaments around the endocytic site, we have developed an in vitro reconstitution of yeast CME actin assembly on synthetic lipids. Supported lipid bilayers coated with purified yeast WASP (Las17) and incubated in cytoplasmic yeast extracts recruit downstream endocytic proteins. These include the yeast homologues of the actin‐binding proteins fimbrin (Sac6) and amphiphysin (Rvs161/167). Fimbrin recruitment to supported bilayers is abolished by treatment with LatrunculinA, confirming that fimbrin recruitment represents actin assembly. Additionally, electron micrographs of these supported bilayers reveal WASP‐dependent membrane invaginations on the size scale of biological endocytic events (~25‐100nm). Current efforts are focused on using superresolution imaging of WASP‐coated bilayers to localize with high precision the four Arp2/3 activators found at endocytic sites and to visualize surrounding actin filaments. This work promises to reveal fundamental principles underlying CME.

M239 Sec3 Promotes the Binary t‐SNARE Complex Assembly and Membrane Fusion. P. Yue1, Y. Zhang2, K. Mei1, S. Wang1, J. Lesigang2, Y. Zhu1, G. Dong2, W. Guo1; 1Biology, University of Pennsylvania, Philadelphia, PA, 2Max F. Perutz Laboratories, Medical University of Vienna, Vienna, Austria Membrane fusion in the secretory pathway is mediated by the interaction of soluble N‐ethylmaleimide‐sensitive factor‐attachment protein receptors (SNAREs) on secretory vesicles (v‐SNAREs) and target membranes (t‐SNAREs), respectively. While the SNAREs constitute the core machinery for fusion, how the assembly of the SNARE complex is initiated is unknown. Here we report that Sec3, an exocyst component that mediates vesicle tethering during exocytosis, directly interacts with the t‐SNARE protein Sso2 on the plasma membrane. This interaction promotes the formation of a Sso2‐Sec9 “binary” t‐SNARE complex, the first step in SNARE complex assembly, and stimulates the early rate‐limiting step of membrane fusion. To further understand the effect of Sec3 in SNARE assembly, we obtained the crystal structure of the Sec3‐Sso2 complex, which shows that Sec3 induces conformational changes of Sso2 that are crucial for the relief of its auto‐inhibition. Interestingly, specific disruption of the Sec3‐Sso2 interaction in cells blocks exocytosis without affecting vesicle tethering. Our study reveals an activation mechanism for early step of SNARE complex assembly, and uncovers a role of the exocyst in promoting membrane fusion in addition to vesicle tethering.


M240 The Sec61 translocon controls IRE1 activity during the unfolded protein response. A. Sundaram1, R. Plumb1, S. Appathurai1, M. Mariappan1; 1Department of Cell Biology, Nanobiology Institute , Yale University, West Haven, CT Misfolded protein in the endoplasmic reticulum (ER) has been implicated in various disease states such as neurodegenerative, diabetes and cancer. To tackle the increase in misfolded proteins, cells mount an unfolded protein response through activating three membrane proteins including IRE1, PERK, and ATF6 to bring back ER into its functional homeostasis. IRE1 is conserved from yeast to mammalian cells and is essential during ER stress and embryogenesis, respectively. The current mode of IRE1 activation is dependent on BiP (HSP70 molecular chaperone) and/or ligand binding [1]. During the increase in unfolded proteins, BIP is released from the IRE1 luminal domain, which activates IRE1 through self‐oligomerization and transautophosphorylation. On the other hand, unfolded protein is also proposed to directly bind to the luminal groove of IRE1 and promotes its oligomerization. We have previously reported that IRE1 interacts with the Sec61 translocon to efficiently find and cleave its substrate mRNAs during ER stress [2]. Here we show that the Sec61 translocon plays a unique role in regulating IRE1 activity. We find that IRE1 interaction with Sec61 is required for proper modulation of IRE1 activity including its oligomerization and phosphorylation. Hence, IRE1 is promptly inactivated upon ER stress alleviation. In contrast, the loss of IRE1 interaction with the Sec61 translocon leads to improper activation and inactivation of Ire1 signaling. As a consequence of this unregulated IRE1 activation, RIDD (regulated IRE1 dependent mRNA decay) is also augmented in cells, thus compromising the protein‐folding capacity of the ER. Our in vitro biochemical studies further support that the Sec61 translocon controls IRE1 oligomerization status, thereby regulating its RNase activity. Thus, our data suggest that the Sec61 translocon‐mediated tight regulation of IRE1 activation is required to restore ER homeostasis particularly in insulin‐secreting cells during ER stress. Gardner BM, Pincus D, Gotthardt K, Gallagher CM, Walter P. Endoplasmic reticulum stress sensing in the unfolded protein response. Cold Spring Harb Perspect Biol. 2013 Plumb R, Zhang ZR, Appathurai S, Mariappan M: A functional link between the co‐translational protein translocation pathway and the UPR. Elife. 2015

Minisymposium 24: Membrane‐less Organelles

M241 ER compartmentalization facilitates lineage‐specific confinement of protein aggregation during yeast aging. J. Saarikangas1, F. Caudron1, R. Prasad1, D.F. Moreno2, M. Aldea2, Y. Barral1; 1Institute of Biochemistry, ETH Zurich, Zurich, Switzerland, 2Institut de Biologia Molecular de Barcelona, CSIC, Barcelona, Spain Budding yeast segregates age asymmetrically at division by confining aging factors, including protein aggregates, to the mother cell. Tethering to organelles and cytoskeletal elements, as well as geometrical constraints help retaining various protein aggregate species. However, whether and how oligomeric aggregation precursors, which are thought to be more cytotoxic than aggregates themselves, are confined to the aging lineage is unknown. Here, we investigated whether the endoplasmic reticulum (ER)‐membrane, and its compartmentalization helps confine the aggregation process to yeast mother cells. Conventional and super‐resolution microscopy demonstrated that the Hsp104‐labelled aggregates, which formed in aged cells, localized to the ER, being primarily tethered to the nuclear envelope (NE).


Interestingly, aggregates from different cells merged together upon cell‐cell fusion, and this required the fusion of ER and NE of the parental cells. Strikingly, lineage analyses revealed that while disrupting ER compartmentalization did not affect the retention of the pre‐existing large aggregate in the mother cell, it caused premature formation of an aggregate in her daughters. Dissociating the Hsp40 protein Ydj1 from the ER membrane led to a similar phenotype. Furthermore, Ydj1 and the ER diffusion barrier worked in the same pathway in confining aggregation to the mother cell. Taken together, we provide evidence that tethering of aggregate precursors to the ER‐membrane via Ydj1 subjects them to confinement by the ER diffusion barriers and ensures that their coalescence into larger inclusions is restricted to the aging‐lineage. Thus, selective ER‐membrane anchorage together with membrane compartmentalization by diffusion barriers emerge as a general theme for the asymmetric segregation of diverse fate determinants during cell division.

M242 Prion Biology at the Interface of Protein Misfolding and its Cellular Environment. T.R. Serio1; 1Department of Molecular and Cellular Biology, The University of Arizona, Tucson, AZ The protein‐only or prion hypothesis posits that changes in cellular physiology arise from the refolding of host‐encoded proteins (prions) to alternative conformations that alter their normal functions. A central event in this process is the assembly of these alternative conformers into ordered amyloid aggregates, which associate with, incorporate, and thereby template the refolding other conformers of the same protein. However, the accumulation of amyloid is a poor predictor of changes in cellular phenotype, and our recent studies reveal that prion, and more generally amyloid, biology likely arises from unique homeostatic niches created by an intersection of multiple, previously unanticipated constraints of the system. Our long‐term goal is to bridge this gap by developing a system‐based understanding of prion biology that integrates protein quality control pathways, prion protein sequence and conformation, and cell biology. Toward this end, we are exploiting enigmatic aspects of prion biology including sequence variants, dominant negative inhibition, cytoskeletal interactions, strain dominance and prion interactions to reveal the contributors and mechanisms that define balance in the system, create distinct homeostatic niches compatible with each conformational state, and induce transitions along these continuums.

M243 Single‐molecule microscopy reveals constrained diffusion by a polar matrix microdomain. A. von Diezmann1, K. Lasker2, T.H. Mann2, L. Shapiro2, W.E. Moerner1; 1Chemistry, Stanford University, Stanford, CA, 2Developmental Biology, Stanford University School of Medicine, Stanford, CA Robust positioning of a suite of signaling proteins at the Caulobacter cell poles mediates the spatial and temporal control of gene expression. The localization of many of these proteins depends on PopZ, a 177‐residue protein that forms a porous polymeric network at the pole known to exclude ribosomes. However, the effect of the PopZ network on smaller molecules, including the many signaling proteins involved in cellular differentiation, is not well‐understood. This stems partly from the challenge of resolving protein distributions and dynamics within a region smaller than the diffraction limit of light. Here, we applied a recently developed scheme to track the motion of single proteins in three dimensions with 30‐60 nm spatial and 20‐100 ms temporal resolution relative to the super‐resolution image of the ≤ 200 nm microdomain of the PopZ network at the poles of live cells.We found that both


membrane‐bound and cytoplasmic proteins implicated in the developmental circuitry of Caulobacter diffuse many times more slowly when on the membrane adjacent to or inside the volume delimited by PopZ, respectively. By enlarging the PopZ microdomain by overexpression, we show how its spatial extent overlaps with the constrained diffusion of these proteins. To investigate the specificity of this effect, we also tracked the motion of heterologous proteins that lack known binding partners in Caulobacter and find radically different behavior near the pole. Finally, in photobleaching experiments, we found that exchange of signaling proteins between the poles and rest of the cell is minimal over a timescale of minutes. Together, our results imply that the PopZ network modulates the localization of signaling proteins at the poles by the restriction of motion within its nanoscale pores. The distinct subcellular environments of Caulobacter bacterial cells thus appear to stem from a mechanism different from those of membrane‐encapsulated organelles or where superstructure is created by scaffold/adaptor proteins possessing multiple interaction domains.

M244 Developmentally regulated histone/lipid droplet interactions control nuclear histone levels in Drosophila embryos. M.R. Johnson1, M.A. Welte1; 1Biology, University of Rochester, Rochester, NY Lipid droplets (LDs) are central organelles for lipid metabolism. Emerging evidence in multiple systems indicates that LDs also regulate nuclear functions, via the exchange of lipids, transcription factors, and chromatin components. But, how such exchange is regulated remains poorly understood. We address this issue using Drosophila embryos where LDs are associated with specific histones. Previously, it had been proposed that this association buffers the histone supply: when histones cannot bind to LDs, the histone variant H2Av over‐accumulates in nuclei and embryos exhibit hallmarks of DNA damage and reduced viability. However, direct evidence for this buffering model has been lacking. Using Fluorescence Recovery After Photobleaching (FRAP) and photoactivation, we discovered that during early embryogenesis H2Av spreads rapidly through the embryo, at rates similar to diffusion. This spread is not due to LD movement; rather, H2Av exchanges between droplets via the cytoplasm. We find that this dynamic behavior of H2Av is developmentally regulated. While embryos younger than 2 hrs all display high rates of H2Av exchange between LDs, H2Av is stably attached to LDs during cellularization stages. This dramatic transition occurs over a span of less than 15 min and is accompanied by decreased nuclear uptake of H2Av. Together, these observations suggest two states of histone dynamics: Initially H2Av is dynamically attached to LDs, rapidly exchanging between droplets via the cytoplasm; this exchange presumably buffers cytoplasmic H2Av levels and provides the source of histones for import into nuclei. After the transition, H2Av is statically attached to LDs; this abolishes buffering and restricts nuclear delivery of H2Av from LDs. We speculate that this static state protects cellularization‐specific developmental processes from histone overload. Indeed, we find that when histone sequestration to LDs is abolished, nuclei prematurely accumulate heterochromatin protein 1, suggesting that this system may regulate the establishment of heterochromatin. To ultimately determine the physiological role of this switch in dynamics, we are now analyzing its mechanistic basis. We have already shown that the switch is unrelated to zygotic genome activation or the massive intracellular redistribution of LDs, both of which occur at the same time as the transition. Studies on modifications of the histone anchor protein Jabba and of the histones themselves are in progress.


It is becoming increasingly clear that LDs cannot only sequester nuclear proteins, but also proteins from many other cellular compartments. Our work may thus provide a paradigm for how lipid droplets regulate and buffer protein availability in general.

M245 Filaments formed by the translation initiation factor eIF2B: high‐resolution insights into a survival strategy. G. Marini1, E. Nüske1, S. Alberti1, G. Pigino1; 1Max Planck Institute of Molecular Cell Biology and Genetics, Dresden, Germany Self‐assembly on the mesoscale level is emerging as a clever cell mechanism to store proteins, save energy and survive stressful conditions. In baker's yeast, many essential enzymes form enzyme‐specific ordered fibers and mesoscale structures, such as bundles of filaments, in energy‐depleted cells. The stress‐induced polymerization of these enzymes has two features: 1) it silences the enzymatic activity, thus slowing down the metabolic rate of cells; 2) it is readily reversible, with rapid depolymerization upon replenishment of energy, thus allowing yeast cells to quickly re‐enter into the cell cycle. This suggests that filament‐formation may be an energy‐saving strategy to endure extreme conditions of starvation. At present, not much is known about the structural basis of the stress‐induced enzyme self‐assembly. Many questions need to be answered, ranging from what ultrastructural arrangement and cellular distribution is adopted by the self‐assembling proteins, to how polymerization leads to enzymatic inactivation. By using high‐pressure freezing, correlative light and electron microscopy (CLEM), electron tomography, and cryo‐electron microscopy (cryo‐EM), coupled to high‐resolution 3D‐reconstruction techniques, we investigate the architecture and cellular distribution of the filament‐forming eukaryotic translation initiation factor eIF2B. Tomograms of energy‐depleted yeast cells show bundles of eIF2B filaments. Subtomogram averaging reveals an ordered stacking of the heteropentameric eIF2B dimers in filaments, with a 13 nm periodicity. The 3D‐model presents also frequent lateral connections between neighbouring filaments in the bundle, which might contribute in keeping their regular pattern and distance. From a cellular point of view, our structural comparison of stressed and non‐stressed cells reveals a massive rearrangement of cytoplasmic structures, in addition to the enzyme compartmentalization. We speculate that increased crowding and other global changes in the cytoplasm drive the assembly of eIF2B, as well as other enzymes, into filaments. Our analysis of the yeast cytoplasm and our detailed structural analysis of filament‐forming enzymes provides insights into a survival strategy that is used by yeast and many other organisms to cope with extreme environmental conditions and stress.

M246 Metabolite regulated partitioning of metabolic enzymes into RNA granules. K. Begovich1, J.E. Wilhelm1; 1Cell and Developmental Biology, University of California, San Diego, La Jolla, CA There has been an increasing appreciation for the role of phase separation in generating novel intracellular compartments to facilitate biochemical reactions. While most of the focus in this field has centered on the role of these compartments in regulating RNA stability and translation, relatively little is known about how they might be utilized to organize and control metabolic pathways. We have recently completed a visual screen of all 440 metabolic enzymes in the S. cerevisiae GFP strain collection to identify metabolic enzymes that assemble into novel structures. This screen identified 59 metabolic enzymes that form either foci or filaments. Further analysis of this set of metabolic enzymes found that


17 localized to RNA granules suggesting that localization to the RNA granule might provide a novel mechanism for spatial control of metabolic pathway activity. This localization does not appear to be used for substrate channeling since the enzymes that are recruited to the RNA granules are not in consecutive steps of a biosynthetic pathway. Instead, one potential use of this localization appears to be to facilitate balancing between competing biochemical pathways. To test this hypothesis, we focused on two enzymes from cysteine/methionine biosynthesis that are both recruited to RNA granules, Cys4p and Sam1p. We focused on these two enzymes because they are competing pathways and the product of Sam1p, S‐adenosylmethionine (AdoMet), is thought to regulate the mammalian ortholog of Cys4p. Using a structure‐function approach, we have identified mutations in Cys4p that disrupt partitioning of Cys4p into the RNA granule without affecting enzyme activity. These mutations also disrupt the binding of AdoMet to Cys4p arguing that Cys4p localization to RNA granules is regulated by a metabolite in a competing pathway. We have now extended this result using mutations in various metabolic pathways to identify additional metabolites that regulate the partitioning of this class of metabolic enzymes. This represents a novel role for intermediate metabolites in controlling the spatial organization of metabolic pathways in vivo. Furthermore, our pathway analysis also uncovered a metabolic enzyme that is required for the assembly of a specific class of RNA granule arguing that RNA regulation and metabolic pathway control might be more tightly integrated than previously believed.

M247 Three complementary mechanisms of quality control ensure RNP granule functionality and dynamics. D. Mateju1, E. Boczek1, J. Wang1, A. Kopach1, S. Maharana1, A. Patel1, H.O. Lee1, M. Jahnel1, S.W. Grill1, A.A. Hyman1, S. Alberti1; 1Max Planck Institute of Molecular Cell Biology and Genetics, Dresden, Germany Liquid‐liquid phase separation can create functionally distinct reaction compartments consisting of proteins and RNAs, which have major effects on cellular dynamics and physiology. RNA‐binding proteins (RBPs) with domains of low sequence complexity (LC) are the key players that mediate this process of phase separation in cells. Recent data suggests that aberrant phase transitions of these proteins may be closely tied or even causative to the pathogenesis associated with diseases such as ALS. Thus, understanding how physiological phase transitions give rise to dysfunctional RNPs and eventually pathological RBP‐containing aggregates will be key to understanding a range of neurodegenerative diseases. Here, we study several human LC domain‐containing RBPs that are associated with age‐related neurodegenerative diseases such as ALS. We show that in vivo these proteins phase separate to form liquid, membrane‐less compartments. Using an in vitro “aging” assay, we demonstrate that reconstituted RNP compartments assembled from purified RBPs have a strong drive to convert with time from a liquid to an aggregated, pathological state. We identify three mechanisms that regulate the conversion of these RNPs into pathological aggregates: First, we find that heterogeneity in RNP granule composition prevents liquid‐to‐aggregate transitions, due to mutual interactions of RBPs, which frustrate nucleation events that are necessary for aggregation. Second, we show that a dedicated chaperone surveillance machinery prevents intra‐granule interactions between RBPs and misfolded proteins, which have a tendency to accumulate in RNP granules because of their hydrophobic properties and affect RNP granule dynamics and material properties. And third, we uncover that aberrant RNP granules are selectively recognized by protein quality control machinery to be targeted to the aggresome and subsequently degraded by autophagy.


These findings suggest that nature has come up with three interdependent mechanisms of regulating aggregation‐prone RBPs in RNP granules: 1) RNP granule‐intrinsic mechanisms that control the conformational properties of RBPs through RNP granule composition, 2) a specialized chaperone surveillance machinery that keeps RNP granules free of aberrant protein conformations, and 3) a disposal machinery that selectively eliminates irretrievable, aberrant RNP granules in toto. We hypothesize that all three pathways weaken with increasing age and that genetic mutations in pathway components accelerate the manifestation of a disease state.

M248 The role of RNA sequence in controlling polyQ‐protein phase transitions. E.M. Langdon1, H. Zhang2, P. Occhipinti1, A.S. Gladfelter1; 1Biology, University of North Carolina, Chapel Hill , NC, 2Biology, Dartmouth College, Hanover, NH Compartmentalization is central to the spatial and temporal control of cellular biochemistry. In the multinucleate cells of the fungus, Ashbya gossypii, nuclei divide asynchronously and multiple sites of polarity coexist in a continuous cytoplasm. Both of these phenomena require functional compartmentalization of cytoplasm. One mechanism of compartmentalization involves the polyQ‐tract RNA‐binding protein, Whi3, undergoing liquid de‐mixing with its target mRNAs to form phase‐separated droplets. Formation of these assemblies allows the positioning of cyclin transcripts in the vicinity of nuclei as well as formin transcripts at incipient and existing sites of polarized growth. What has remained unclear is how the Whi3 protein functions in two separate cellular processes and how co‐existing Whi3 droplets remain physically distinct. Previous work has shown that cyclin and formin mRNAs can drive Whi3 phase separation in a sequence dependent manner to create droplets with different morphologies and biophysical properties in vitro. This suggests that different mRNAs may promote physically and potentially functionally distinct droplets. Consistent with this, we find that pre‐formed droplets containing recombinant Whi3 with the formin mRNA are immiscible with Whi3 droplets containing the cyclin mRNA, suggesting that the nucleotide sequence or other features of RNAs might be responsible for the spatial segregation of different Whi3 assemblies within the cell. Supporting this, single molecule mRNA FISH shows that cyclin and formin transcripts do not co‐localize in the cell. We next investigated features of the RNA sequences that might contribute to the variation in droplet properties and mRNA localization. Notably, synthetic RNAs that contain variable numbers of Whi3‐binding sites and provide similar valency to the native interactions are insufficient to drive physically distinct phase transitions. This indicates additional features of the RNA contribute to the material properties of Whi3 droplets. We are currently examining which features of mRNA sequences can promote different subcellular localizations and the distinct functions of Whi3 droplets. We hypothesize that the mRNAs themselves are encoding not simply genetic information but also the biophysical properties of the Whi3 assemblies to promote spatially variable activity in a common cytoplasm.

M249 Spatial patterning of RNA granules by RNA‐induced phase separation of an intrinsically‐disordered granule scaffold. J. Smith1, D. Calidas1, H. Schmidt1, T. Lu1, G. Seydoux1; 1Molecular Biology and Genetics, Johns Hopkins University School of Medicine, Baltimore, MD RNA granules are non‐membrane bound cellular compartments that exhibit properties of phase‐separated liquids. The molecular mechanisms that regulate the distribution of RNA granules in cells are not known. The P granules of C. elegans form preferentially in the posterior cytoplasm of the 1‐cell


zygote. Here we present evidence that P granule asymmetry depends on RNA‐induced phase‐separation of the granule scaffold MEG‐3. MEG‐3 is an intrinsically‐disordered protein that phase separates with RNA in vitro. In vivo, MEG‐3 forms a posterior‐rich concentration gradient that is anti‐correlated with a gradient in the RNA‐binding protein MEX‐5. MEX‐5 is necessary and sufficient to suppress MEG‐3 granule formation in vivo, and suppresses RNA‐induced MEG‐3 phase separation in vitro. Our findings support a model whereby sequestration of RNA by MEX‐5 locally suppresses MEG‐3 phase separation and drives granule asymmetry

M250 Single‐particle dynamics underlying the segregation of GFP::PIE‐1 during asymmetric division of the C. elegans zygote. Y. Wu1, B. Han1, E. Griffin1; 1Biology, Dartmouth College, Hanover, NH During the asymmetric division of the C. elegans zygote, the PAR proteins orchestrate the partitioning of cytoplasmic RNA‐binding proteins MEX‐5/6 and PIE‐1 along the anterior/posterior axis. The posterior kinase PAR‐1 increases the mobility of its substrate MEX‐5 in the posterior, leading to the preferential retention of MEX‐5 in the anterior cytoplasm (Tenlen, 2008; Daniels, 2010; Griffin, 2011). In turn, MEX‐5 and MEX‐6 act to increase the mobility of PIE‐1 in the anterior cytoplasm, resulting in the preferential retention of PIE‐1 in the posterior cytoplasm (Wu, 2015). The mechanisms underlying the transduction of spatial information from MEX‐5/6 to PIE‐1 are not understood. In order to determine the mechanisms by which MEX‐5/6 control the mobility of PIE‐1, we have used Near‐TIRF imaging to characterize the dynamics of individual GFP::PIE‐1 particles in the polarized zygote. We find that GFP::PIE‐1 is present in two classes of particles: a rapidly diffusing population that is uniformly distributed and a slow‐diffusing population that is enriched in the posterior cytoplasm. Slow‐diffusing GFP::PIE‐1 particles appear and subsequently disappear from the same position in the embryo, suggesting fast and slow‐diffusing particles frequently interconvert. Slow‐diffusing GFP::PIE‐1 particles appear four times more frequently and persist significantly longer in the posterior cytoplasm, leading to their enrichment in the posterior. We find that MEX‐5/6 controls both the appearance rate and persistence of slow‐diffusing GFP::PIE‐1 particles. Interestingly, GFP::MEX‐5 is present in anteriorly‐enriched slow‐diffusing particles whose dynamics are similar to slow‐diffusing GFP::PIE‐1 particles. Furthermore, the ability of MEX‐5/6 to increase PIE‐1 mobility is sensitive to mutations that disrupt MEX‐5 RNA‐binding and to reduction of MEX‐5 concentration. Based on these results, we propose that MEX‐5/6 compete with PIE‐1 for association with a relatively static cytoplasmic structure such that PIE‐1 binding is biased to the posterior, where MEX‐5/6 concentration are low.


Minisymposium 25: Organ Development, Homeostasis, and Disease

M251 Mammary epithelial cells spatiotemporally coordinate molecular activities and mechanical forces to drive radial intercalation during ductal elongation. N.M. Neumann1,2,3, M.C. Perrone4, J.H. Veldhuis4, G.W. Brodland4, A.J. Ewald1,2,3; 1Center for Cell Dynamics, Johns Hopkins University School of Medicine, Baltimore, MD, 2Department of Cell Biology, Johns Hopkins University School of Medicine, Baltimore, MD, 3Department of Oncology, Johns Hopkins University School of Medicine, Baltimore, MD, 4Department of Civil and Environmental Engineering, University of Waterloo, Waterloo, Canada During pubertal development, the mammary epithelium extensively migrates to form an interconnected ductal network. A low‐polarity, stratified, and highly motile cell population elongates the tissue and transitions to a polarized, simple duct. Receptor tyrosine kinases (RTKs) have been shown to be key regulators of collective migration. However, the spatiotemporal coordination of single‐cell behaviors, molecular activities, and mechanical forces during elongation remain incompletely understood. Branching morphogenesis occurs deep within the optically inaccessible mammary fat pad, which limits the ability to study cellular contributions to morphogenetic processes. To overcome this challenge, our laboratory developed three‐dimensional (3D) ex vivo organotypic culture assays and real‐time confocal imaging techniques. These approaches allow us to visualize cell behaviors and molecular activities at high resolution within tissues. We tracked single cells within the stratified layer to observe their behaviors. We noted that cells are highly protrusive in the direction of elongation and observed ductal elongation driven by radial intercalation. Then, using fluorescent molecular biosensors, we visualized the downstream signaling events in the RTK pathway. We observed individual cells within the stratified layer both migrate and radially intercalate while enriching Ras activity, PIP3, and F‐actin in protrusions. In the protrusions, PIP3 enriched at the site of, but prior to, F‐actin enrichment. Inhibiting actin dynamics prevented protrusion formation and morphogenesis. Next, we used the Cellular Force Inference Toolkit in 3D to analyze force balance equations at each triple, or higher‐order, cellular junction within an organoid. We found that cell migration through a tissue environment requires specific ratios of protrusion tension and posterior interfacial tension gradients. For radial intercalation, the duct requires high basal tension and a time‐varying, increasing posterior interfacial tension gradient as protrusions are initiated. Finally, using finite element forward modeling, we were able to generate in silico organoids with cells undergoing migration and radial intercalation to elongate a duct, solely based on mechanical alterations. We conclude that ductal elongation results from an epithelial motility program that uses directed enrichments of Ras activity, PIP3, and F‐actin in protrusions to drive migration and radial intercalation. These dynamics create a spatial gradient of tensions across cells and time‐dependent changes within the tissues that together accomplish morphogenesis. We now seek to understand how cancer cells re‐access these developmental epithelial motility programs during metastasis.


M252 Tissue‐scale coordination of cellular homeostatic and repair behaviors in live mice. V. Greco1; 1Genetics and Yale Stem Cell Center, Yale, New Haven, CT Tissue repair is fundamental to our survival as tissues are challenged by recurrent damage. During mammalian skin repair, cells respond by migrating and proliferating to close the wound. However, the coordination of cellular repair behaviors and their effects on homeostatic functions in a live mammal remains unclear. Here we capture the spatiotemporal dynamics of individual epithelial behaviors by imaging wound re‐epithelialization in live mice. Differentiated cells migrate while the rate of differentiation changes depending on local rate of migration and tissue architecture. Cells depart from a highly proliferative zone by directionally dividing towards the wound while collectively migrating. This regional co‐existence of proliferation and migration leads to local expansion and elongation of the repairing epithelium. Finally, proliferation functions to pattern and restrict the recruitment of undamaged cells. This study elucidates the interplay of cellular repair behaviors and consequent changes in homeostatic behaviors that support tissue‐scale organization of wound re‐epithelialization.

M253 Patterning the Mammalian Epidermis by the Planar Cell Polarity Pathway. D. Devenport1, M. Cetera1, B. Joyce1, L. Leybova1; 1Department of Molecular Biology, Princeton University, Princeton, NJ The ectodermal structures that decorate the vertebrate skin surface, such as feathers, scales, and body hairs, are remarkable examples of planar cell polarity, where spatially asymmetric patterns display local order and global alignment. The core planar cell polarity (PCP) pathway, which controls the global alignment of mammalian hair follicles, is best understood in the Drosophila wing blade where a cell‐to‐cell propagation mechanism aligns polarity between individual cells. Despite the conservation of PCP components in mammalian hair patterning, how PCP directs the asymmetric morphogenesis of these complex and dynamic multicellular structures is unclear. Previously we demonstrated that hair follicles acquire morphological asymmetry and polarized cell fates during the earliest stages of their development, and that the core PCP pathway is required for both of these events. Here, by combining long‐term, time‐lapse imaging and biophysical manipulations of embryonic hair follicle morphogenesis, we define the PCP dependent and independent events that drive local follicle tilting, cell fate asymmetry, and global follicle alignment. We present a simple model where unidirectional tension across the epithelium generates an imbalance of forces on the two sides of the emerging hair placode, driving follicle bending and anterior‐directed growth. This, in turn, displaces a crucial signal center, the dermal condensate, towards one side of the follicle, whose signals are required for asymmetric cell fates and continued anterior growth. These findings demonstrate how a simple unidirectional, PCP‐dependent cell behavior generates a collectively polarized pattern of complex, multicellular structures.


M254 Feedback inhibition of stem cell divisions equalizes cell production and loss during intestinal epithelial turnover. J. Liang1, L.E. O'Brien1; 1Molecular and Cellular Physiology, Stanford University, Stanford, CA Continual cell turnover is crucial for the long‐term maintenance of healthy organs, but to avoid pathological consequences, new cells must be produced at the precise rate that old cells are lost. Here we identify the signaling basis of feedback inhibition that equalizes cell production and loss during stem cell‐driven turnover of the Drosophila intestinal epithelium. E‐cadherin at the cortex of mature enterocytes inhibits EGF‐dependent stem cell divisions by repressing the EGF maturation factor rhomboid. Single, apoptotic enterocytes disrupt this inhibitory relay by degrading E‐cadherin, which de‐represses rhomboid to activate EGFR and promote divisions of nearby stem cells. Ectopic loss of E‐cadherin in non‐apoptotic cells upregulates rhomboid, causing EGFR hyperactivation and excessive divisions. In contrast, blocking apoptosis precludes EGFR activation and slows down divisions to keep total cells constant. Thus, stem cell divisions require enterocyte apoptosis to relieve proliferative inhibition, ensuring equal rates of cell production and loss.

M255 Epithelial self‐healing is recapitulated by a 3D biomimetic E‐cadherin junction. D.J. Cohen1, M. Gloerich1, W.J. Nelson1; 1Biology, Stanford University, Stanford, CA Epithelial monolayers undergo self‐healing when wounded. During the healing process, cells collectively migrate into the wound site, and the converging edges of the wound collide and form a stable interface. For self‐healing to successfully complete, migrating tissues must form cell‐cell adhesions, and reorganize from the front‐rear polarity characteristic of cell migration to the apical‐basal polarity of an epithelium. However, identifying the underlying ‘stop signal’ that induces colliding tissues to cease migrating and heal remains an open question. Epithelial cells form two types of adhesions in orthogonal planes, integrin‐based adhesions to the basal extracellular‐matrix (ECM) and E‐cadherin‐mediated cell‐cell adhesions on the vertical surfaces between cells. Current biological tools have been unable to probe this multicellular 3D interface to determine the ‘stop signal’. We addressed this problem by developing a unique biointerface that mimicked the 3D organization of epithelial cell adhesions. This ‘minimal tissue mimic’ (MTM) comprised a basal ECM substrate and a vertical surface coated with purified E‐cadherin that was designed for collision by the healing edge of an epithelial monolayer. 3D confocal imaging showed that adhesions formed between cells and the Ecad:Fc MTM resembled the morphology and dynamics of native epithelial cell‐cell junctions, and induced the same polarity transition that occurs during epithelial self‐healing. These results indicate that E‐cadherin presented in the proper 3D context constitutes a minimum essential ‘stop signal’ to induce self‐healing. That the Ecad:Fc MTM stably integrated into an epithelial tissue and reduced migration at the interface suggests that this biointerface is a new, complimentary approach to existing tissue‐material interfaces.


M256 The Matriptase‐Par2‐EGFr signaling pathway regulates cell extrusion, proliferation and survival in the zebrafish embryonic epidermis. A. Schepis1, S. Coughlin1; 1Cardio‐Vascular Research Institute, University California San Francisco, San Francisco, CA Matriptase is a membrane‐bound serine protease involved in epithelial development and homeostasis. Matriptase is synthetized as a zymogen and its activation depends on its inhibitor Hai1. After activation, Matriptase activates the GPCR Par2. In spite of extensive studies in vitro, the roles of the Matriptase/par2 pathway in vivo are still poorly understood. The scope of this work is to unveil new functions for this pathway in vivo. In order to achieve our aim, we choose the zebrafish embryonic epidermis as a model. We use a combination of forward genetic, live imaging and chemical treatments to interfere with the Matriptase/Par2 pathway. The embryonic epidermis comprises two layers: the most external layer, the periderm and the basal layer that lie between the periderm and the internal organs. Zebrafish Hai1a mutants display skin defects in the form of massive cell shedding. Co‐depletion of Hai1a and Matripase‐1 (St14a) rescues the skin defects suggesting that it was caused by Matriptase activation and mis‐regulation. By live imaging, we found out that the cell shedding occurs via cell extrusion. Cell extrusion is a process that enable epithelial layer to eliminate apoptotic cells or live cell avoiding excessive crowding. We were able to rescue the cell extrusion by co‐depletion of Hai1a and Par2b indicating a new pathway that triggers cell extrusion in which Matriptase is upstream Par2. Moreover, we observed that co‐depletion of Hai1a and Par2a caused peridermal keratinocyte over‐proliferation while depletion of Ha1a, Par2b and St14a alone did not. In contrast, the basal layer behaved in completely opposite manner, Hai1a depletion caused basal keratinocyte over‐proliferation. This phenotype was rescued in the background of the Par2b and Matripase depletion. Taken together our data suggest that Matriptase and Par2 take part in a signaling pathway that regulated cell extrusion and proliferation but in an opposite manner in the periderm and the basal layer. Finally, we looked for downstream effector to Par2 by transcription profiling of the mutant keratinocytes. We found out that Hai1a depletion caused enrichment of mmp9 and mmp13 mRNA as well as components of the EGF signaling. Interfering with mmp9, mmp13 and EGFr activity inhibit basal layer proliferation. Surprisingly inhibition of EGFr causes increase of cell extrusion and rescue the increased proliferation in the Par2/Hai1a depletion background. In conclusion, we found the par2 is required for cell extrusion and regulates cell proliferation in an opposite manner in the zebrafish epidermis layers. Our results highlight new roles for the Matriptase/Par2 pathway and the factors activated downstream.

M257 Macrophage‐dependent cytoplasmic transfer drives melanoma metastasis in vivo. M. Roh‐Johnson1, A.N. Shah1, R.E. Hernandez2, C.B. Moens1; 1Basic Sciences, Fred Hutchinson Cancer Research Center, Seattle, WA, 2Pediatric Infectious Disease, Seattle Childrens Hospital, Seattle, WA During cancer progression, interactions between tumor cells and immune cells play critical roles in the initiation of metastasis. In malignant melanoma, it is unclear how a melanocyte transitions from a premalignant nevus to an invasive tumor cell, and which components in the tumor microenvironment regulate this switch. A major limitation to understanding this switch in vivo is the lack of genetically tractable model systems that are amenable to high‐resolution imaging. We have overcome this obstacle by directly visualizing tumor cells and their interactions with stromal cells in zebrafish. We found that injecting either human or zebrafish melanoma cells results in dissemination from the site of injection.


Live imaging revealed that when in contact with tumor cells, macrophages form long, thin, and dynamic protrusions. We hypothesized that these protrusions may be a mechanism for cell‐cell communication between tumor cells and tumor‐associated macrophages that may influence metastatic potential. Consistent with this hypothesis, we varied macrophage recruitment to the primary tumor by altering gene function of myd88, a key regulator in the inflammatory response pathway, and found that depleting zebrafish larvae of myd88 reduced macrophage recruitment to the primary tumor and subsequent tumor cell contact, whereas specific expression of constitutively active myd88 in macrophages increased macrophage recruitment and tumor cell contact. These genetic backgrounds resulted in decreased and increased melanoma dissemination, respectively. These results suggest that first, expression of Myd88 in macrophages is sufficient to drive melanoma cell metastasis; and second, that contact between macrophages and tumor cells promotes tumorigenesis in vivo. Recently it has been shown that tumor progression is accelerated by exchange of active molecules between tumor cells (Zomer et al., 2015). Thus, we tested whether macrophages were capable of transferring cytoplasm to tumor cells to promote metastasis. We found that, indeed, macrophages transfer cytoplasm to tumor cells. Importantly, we discovered that macrophage cytoplasmic transfer was correlated with melanoma cell dissemination. These results suggest that recruitment of macrophages to the primary tumor and cytoplasmic transfer of molecules potentially through cell contact‐mediated protrusions may be a mechanism by which macrophages drive melanoma cell dissemination.

M258 Targeting the active extracellular matrix as anti‐metastatic therapy in a breast cancer model. V. Plaks1, J. Chou1, C. Maynard1, N. Kong1, R. van den Bijgaart1, D. Talmi‐Frank2, I. Solomonov2, C. Koopman1, E. Hadler‐Olsen1, M. Headley3, C. Lin1, M. Owyong1, I. Sagi2, Z. Werb1; 1Anatomy, UCSF, San Francisco, CA, 2Biological Regulation, Weizmann Institute, Rehovot, Israel, 3Pathology, UCSF, San Francisco, CA Currently there is no cure for a metastatic disease and it is therefore critical to target the early events that foster metastasis. It is now also recognized that a favorable microenvironment in the metastatic site, primed by the tumor, is crucial for metastasis. Our study is geared towards deciphering cellular and molecular mechanisms governing the metastatic niche that may lead to novel targeted anti‐metastatic therapeutics. We utilize the multi‐stage MMTV‐PyMT breast cancer mouse model, which shares significant similarities with human breast cancer. By injecting a reporter metastatic cell line into hyperplasia‐bearing mice, we were able to probe the susceptibility of the lung microenvironment to metastatic seeding. We demonstrate that early during mammary tumorigenesis, before metastasis has occurred, a metastatic niche is formed in the lung microenvironment. This niche is initiated in part by tumor‐induced systemic pro‐inflammatory factors and local extracellular matrix remodelers, as matrix metalloproteinases (MMPs). We show that the metastatic niche is associated with MMP9‐expressing CD11b+Gr1+ and other lung stromal cells. MMP9, which is also expressed by tumor cells, appears as a pivotal player in the process and is therefore considered as a desirable therapeutic target. To examine the role of MMP9 activity in lung metastatic colonization, we utilized novel endogenous‐like, function‐specific antibodies (SDS3) that block the transiently‐activated enzyme conformation of MMP9, which presumably contributes to disease progression. The therapeutic potential of SDS3 has been demonstrated in models of inflammatory bowel disease. We show that metastatic seeding within the lung microenvironment can be inhibited by SDS3, not only in experimental metastasis models but also when the lung microenvironment is primed by hyperplastic mammary tumors. Primary tumor burden was not changed with SDS3, suggesting that blocking active MMP9 is effective in preventing early metastasis rather than established tumors.


To study the biodistribution and pharmacokinetics of SDS3, we utilized whole‐body bioluminescence, intravital and ex‐vivo live microscopy as well as flow cytometry. We show that SDS3 is retained in myeloid cells within the microenvironment of mammary tumors and lung metastatic foci. In situ zymography shows high MMP activity in premetastaic MMTV‐PyMT lungs, which is reduced after SDS3. SDS3 also inhibits colony formation of cultured metastatic cells. Our results suggest that a metastatic niche is present in the lungs of hyperplastic mammary tumor‐bearing mice and that it can be targeted by blocking MMP9 activity. Our study offers new insights into effectively blocking the in vivo activity of dysregulated MMPs as early anti‐metastatic therapy of various cancers

M259 MenaINV expression, initiated by macrophage‐tumor cell contact, regulates invadopodium‐dependent tumor cell dissemination during metastasis. C.R. Surve1, M. Weidmann1, J. Pignatelli1, J. Bravo‐Cordero2, G.S. Karagiannis1, M.H. Oktay1,3, J.S. Condeelis1; 1Anatomy and Structural Biology, Albert Einstein college of Medicine, Bronx, NY, 2Medicine, Mount Sinai School of Medicine, New York, NY, 3Pathology, Albert Einstein college of Medicine, Bronx, NY Tumor cell intravasation is an essential step in the metastatic cascade, but its exact mechanism is not completely understood. We have previously shown that the direct physical association of a tumor cell over‐expressing invasive Mena isoforms, a perivascular Tie2hi/Vegfhi macrophage and an endothelial cell, forming a cell triad termed “tumor microenvironment of metastasis” (TMEM), increases vascular permeability, facilitating intravasation of tumor cells. We also identified an invasion gene signature in metastasizing cancer cells having differential expression of Mena isoforms: MenaClassicHi, MenaINVHi and Mena11aLo, scored as MenaCalc. Interestingly, TMEM density and MenaCalc score are functionally interrelated and independent prognostic indicators of distant metastasis in breast cancer patients. MenaINV has been shown to be the key Mena isoform in tumor cells during migration toward blood vessels, intravasation, and metastasis. The precise molecular mechanisms relating TMEM formation and function, and MenaINV expression, had not been elucidated. Here we describe the molecular mechanism. We show here that tumor cell‐macrophage contact uniquely induces MenaINV expression at the mRNA and protein levels in tumor cells, in a Notch1‐dependent manner. Macrophage contact also increases the assembly and ECM degradation activity of invadopodia that are required for tumor cell intravasation in a MenaINV‐dependent step. These results are consistent with our previous findings that MenaINV promotes invadopodium maturation by localizing at invadopodium precursors and increasing cortactin Y421 phosphorylation by inhibiting the phosphatase PTP1b there. The increased cortactin Y421 phosphorylation activates cofilin‐dependent actin polymerization from the invadopodium core causing invadopodium protrusion and ECM degradation. In addition, we show here that the specific knockdown of the Mena gene in mouse mammary tumors (MMTV‐PyMT) abolishes TMEM assembly and function, including TMEM‐dependent vascular permeability, circulating tumor cells and lung metastases. We conclude that macrophage contact upregulates tumor cell MenaINV expression to promote invadopodium maturation, and TMEM assembly and function, resulting in tumor cell intravasation. This is the first description of the molecular mechanism behind the predictive power of two clinically used prognostic markers (TMEM and MenaCalc) and represents a major step in defining new targets for the treatment of metastatic tumors.


M260 Formin‐dependent Adhesions are required to Initiate Invasion by Intact Epithelia. T. Fessenden1,2,3,4, M. Perez‐Neut4,5, G.R. Ramirez‐SanJuan1,2,3,6, Y.M. Beckham1,2,3, M.L. Gardel2,3,4; 1Institute for Biophysical Dynamics, University of Chicago, Chicago, IL, 2Department of Physics, University of Chicago, Chicago, IL, 3James Franck Institute, University of Chicago, Chicago, IL, 4Committee on Cancer Biology, University of Chicago, Chicago, ID, 5Ben May Department for Cancer Research, University of Chicago, Chicago, IL, 6Graduate Program on Biophysical Sciences, University of Chicago, Chicago, IL The onset of local invasion defines the transition from benign to malignant tumors, typified by egress of tumor cells from the primary lesion. Efforts to characterize invasive motility have yet to identify conserved biophysical mechanisms by which cell collectives initiate invasion. To determine how actin cytoskeletal organelles control invasion, we investigated actin nucleators Arp2/3 and formins during collective invasion by mouse tumor organoids and MDCK acini undergoing branching morphogenesis. We found that inhibition of formins, but not Arp2/3, prevented the formation of invasive fronts in both cell types. Surprisingly, MDCK cells depleted of the formin protein Dia1 were not impaired in their ability to form polarized acini. Moreover, we observed no difference in motility within the acinus or in cell scattering assays on coverslips. Dia1‐depleted acini were, however, significantly hindered in their ability to generate and stabilize protrusions into the collagen matrix. To examine interactions between acini and collagen, we performed live cell imaging of actin, myosin and collagen using light sheet and confocal microscopy. We found that prior to invasion, control acini deformed the surrounding collagen matrix both across the acinus as well as in discrete points on individual fibrils. Sites of fibril deformation coincided with sub‐cellular puncta of actin and myosin, which became immobilized on the collagen fibril as cells moved within the acinus. Dia1‐depleted acini exhibited global collagen deformations, but lacked the localized adhesion and contractile force necessary for stable interactions with collagen fibrils. Finally, control but not Dia1‐depleted acini formed plaques enriched in phospho‐tyrosine, a hallmark of mature focal adhesions, which colocalized at sites of fibril deformation. This work identifies Dia1 as an essential regulator of invasion and highlights the biophysical basis by which tissues can regulate noninvasive versus invasive motility.

Minisymposium 26: Use Synthetic Biology to Measure and Manipulate Cell Biology

M261 A Synthetic Cell‐like System to Investigate the Cell Size and Shape Dependence of GTPase Signaling. J.G. Bermudez1, P.G. Torre2, M.C. Good1,2; 1Bioengineering, University of Pennsylvania, Philadelphia, PA, 2Cell and Developmental Biology, University of Pennsylvania, Philadelphia, PA To properly grow and divide eukaryotic cells must be able to measure their dimensions. Computational studies suggest that signal flux through kinase and GTPase pathways depends on cell size and shape. For example, a plasma membrane‐localized GEF and a cytoplasm‐localized GAP are predicted to create a gradient of GTPase activation, such that the fraction of activated GTPase will inversely correlate with cell size. However, it has proven challenging to test these predictions using higher eukaryote cell culture systems. Therefore, we recently developed a synthetic cell platform that enables broad control of cellular dimensions. It is comprised of cell‐like compartments filled with cytoplasmic extracts capable of carrying out complex cell cycle processes in vitro. The cell boundary is tunable, enabling control of


compartment volume, shape and lipid composition. Additionally, we are developing light‐inducible dimerization tools to regulate protein localization to the cell boundary. By combining our novel synthetic cell platform with optogenetics we now are able to directly interrogate the link between cell geometry and intracellular signaling. Our goals are to measure the rates of Cdc42 GTPase activation as a function of cell size, and determine how the spatial patterning of GEFs and GAPs dictates the sensitivity of GTPase signaling to cell size. Here I will describe our results: 1) implementing reversible and irreversible blue‐light control of protein localization inside our synthetic cell‐like system, and 2) the application of these tools to generate tunable patterns of Cdc42 and RhoA activation. These studies provide new insights on the mechanisms by which intracellular signaling pathways sense cell size, and paradigms for understanding the altered signaling responses of enlarged cells observed in cancer and senescence.

M262 Conversion of Extracellular Signals to Programmable Genome Manipulation via CRISPRouter. P.P. Dingal1,2,3, N.H. Kipniss1, Y. Gao4, L.S. Qi1,2,3; 1Bioengineering, Stanford University, Stanford, CA, 2Chemical and Systems Biology, Stanford University, Stanford, CA, 3Stanford ChEM‐H, Stanford University, Stanford, CA, 4Cancer Biology Graduate Program, Stanford University, Stanford, CA We have developed a class of human NOTCH1‐Cas9 chimeric receptors (CRISPRouters) that, upon binding to an extracellular Delta ligand, releases membrane‐tethered Cas9 to manipulate the mammalian genome. The CRISPRouter is modular and exhibits robust ligand‐induced activation or knockout of reporter genes and endogenous genes, resulting in programmable modulation of cellular behavior. This chimeric receptor‐Cas9 technology enables extracellular stimulus‐dependent genome engineering, which is useful for dissecting complex signaling pathways and generating novel microenvironment‐gated cellular functions.

M263 The impact of chromatin dynamics on Cas9‐mediated genome editing in human cells. R.M. Daer1,2, J.P. Cutts1, D.A. Brafman1, K.A. Haynes1; 1School of Biological and Health Systems Engineering, Arizona State University, Tempe, AZ, 2Biological Design Graduate Program, Arizona State University, Tempe, AZ The prokaryotic CRISPR/Cas9 system is being widely pursued as a powerful tool for biomedical applications such as gene therapy, tissue regeneration, and cell line and animal model development. Applying CRISPR/Cas9 tools to early embryos of model organisms and stem cells promises to make generation of single‐ and multi‐gene knockout libraries faster and more cost‐effective. Facultative closed chromatin, a state that renders DNA less accessible to nucleases and transcription factors, plays a critical role in maintaining the plasticity of cell fates during embryonic development and stem cell maintenance and differentiation. Some applications, such as high‐throughput, whole‐genome modifications, require efficient access to target sites within chromatin, the ubiquitous DNA‐protein complexes that organize eukaryotic genomes and regulate gene expression. To achieve this, we must first understand how Cas9 interacts with chromatin. Several studies report reduced Cas9 binding at off‐target sites in genomic regions with silenced chromatin in vivo and at reconstituted nucleosomes in vitro. Our study is the first to demonstrate inhibition of Cas9‐editing by silenced chromatin compared to the unsilenced state at a single locus in human cells. We used a transgenic cell line that allows us to use doxycycline to induce accumulation of silencing chromatin proteins at a target gene. We measured the efficiency of Cas9‐editing via non‐homologous end joining (NHEJ) at sites along a luciferase transgene in three chromatin


states, unsilenced, partially silenced, and fully silenced. We found that partially and fully silenced chromatin reduced Cas9‐editing efficiency when compared to the unsilenced state. Using ChIP‐qPCR, we found that Cas9 binding was not blocked by silenced chromatin, implying that the mechanism for inhibition is likely the NHEJ repair step. Artificial reversal of chromatin silencing with a chromatin‐disrupting siRNA increased luciferase expression and recovered Cas9‐editing efficiency to levels comparable to the unsilenced state. However, large increases in luciferase expression using a targeted transcriptional‐activator reduced Cas9 editing to levels comparable to the fully silenced state. Our work is the first to demonstrate that at a single locus, chromatin dynamics, from fully silenced to highly active states, influence Cas9 editing efficiency. Further, we provide a solution for improving editing at sites located in silenced chromatin to enable application of Cas9 to high‐throughput genome modification.

M264 CRISPR‐based genome‐wide screen identifies protein homeostasis factors that control Tau aggregation and propagation in human cells. J.J. Chen1, D. Nathaniel1, S. Mok1, J. Gestwicki1, M. Kampmann1; 1Institute for Neurodegenerative Diseases, UCSF, San Francisco, CA We co‐developed a CRISPR‐based genetic screening platform that enables the genome‐wide interrogation of cellular pathways that are important in human diseases. Here we use our technology to systematically understand the role of the protein homeostasis network in neurodegeneration. The pathologic aggregation of Tau in the human brain is a hallmark of Alzheimer’s diseases as well as other neurodegenerative diseases that are collectively known as Tauopathies. Strikingly, Tau aggregates can spread across cell‐cell boundaries in a prion‐like manner; this mechanism has been proposed to be an important process in disease progression. Previous evidence suggests that Tau‐related pathogenesis in Alzheimer’s diseases may be exacerbated by imbalances in cellular protein homeostasis pathways. While some factors that influence Tau aggregation have been characterized in vitro, the molecular mechanisms that dictate Tau aggregation remain poorly understood in vivo. Here, we present our CRISPR‐based genome‐wide screen that identified several cellular factors that control Tau aggregation. Our top hits reveal several cellular factors that play important roles in protein folding, highlighting the importance of how protein homeostasis determines the fate of pathologic Tau. We are currently characterizing the molecular mechanisms by which the identified factors control tau proteostasis using biochemistry and cell biology. The identified cellular factors may serve as potential therapeutic targets for treating a variety of Tauopathies.

M265 Single‐cell CRISPRi screens interrogating the unfolded protein response. T. Norman1,2, B. Adamson1,2, M. Jost1,2, J. Nunez1,2, J.S. Weissman1,2; 1Cellular and Molecular Pharmacology, University of California, San Francisco, San Francisco, CA, 2Howard Hughes Medical Institute, University of California, San Francisco, San Francisco, CA Functional genomics efforts typically face a tradeoff between complexity of perturbations (e.g., number of genes queried) and complexity of measured phenotype. RNA‐seq, for example, provides transcriptome‐wide measurements, but reports only average behavior for a handful of samples, while pooled screens often investigate thousands of perturbations, but monitor simple outputs like average growth rate or induction of reporter genes. Recent advances in droplet‐based microfluidics technologies for profiling gene expression of individual cells have the potential to bridge this gap, providing rich phenotypic data at the scale of hundreds of thousands of cells. To build a highly parallel platform for


single‐cell functional genomics, we have paired this technology with a simple method for encoding the identity of a perturbation in an expressed transcript that can be captured during single‐cell RNA‐seq library construction. This strategy, called Perturb‐seq, can for example be used to encode the identity of a CRISPR sgRNA mediating the repression, induction, or deletion of a target gene. It is thus now possible to obtain rich phenotypic profiles of hundreds of perturbations in pooled format with single‐cell resolution. To show the utility of the Perturb‐seq approach, we have conducted a detailed study of the mammalian unfolded protein response (UPR), a set of integrated signaling pathways that responds to stresses sensed at the membrane of the endoplasmic reticulum (ER). Using a traditional genome‐wide screen, we found 66 genetic stressors that induced the UPR when depleted using CRISPRi, including both well‐characterized genes and many unknowns. To assign rich phenotypes, we then applied the Perturb‐seq approach to each of these hits, profiling ~100,000 cells in total. Intriguingly, many of these hits show heterogeneous induction of the UPR that would be missed in traditional population approaches. We separately constructed and profiled all combinatorial single, double, and triple knockdowns of the three canonical UPR induction pathways. Using this information, we can then determine how hit genes activate the UPR, through separate or combined induction of the three canonical pathways, and better understand the programs with which cells monitor and respond to stress.

M266 Mimicking transient activation of protein kinases in living cells. J. Klomp1, A. Ray1, V. Huyot1, K.B. Collins1, A.V. Karginov1; 1Pharmacology, University of Illinois ‐ Chicago, Chicago, IL Under physiological conditions, kinases are activated for a finite period of time and duration of this activation is critical for specific biological outcomes. Mimicking this transient biological activity of kinases is very challenging due to the limitation of existing methods. We report a new strategy enabling transient kinase activation in living cells. The design combines two protein engineering methods that provide independent control of kinase activation and inactivation. Insertion of an allosteric switch, the iFKBP domain, into the catalytic domain of a kinase enables specific and efficient activation of the engineered kinase by rapamycin. Shokat and colleagues demonstrated that specific inactivation of a kinase can be achieved by introducing a functionally silent mutation making the kinase sensitive to allele‐specific (AS) inhibitor, 1NA‐PP1. Using tyrosine kinase Src and Ser/Thr kinase p38 as examples, we demonstrate that a rapamycin‐regulated kinase bearing inhibitor‐sensitizing mutation (RapR‐Src‐as2 and RapR‐p38‐as2) can be efficiently activated by rapamycin and inactivated by 1NA‐PP1. We also show that by using this method we can transiently stimulates Src‐induced phosphorylation of endogenous substrates, transiently regulate Src‐mediated morphological changes, and control their duration. Interestingly, transient activation of Src is followed by secondary Src‐independent morphological processes: slow cell spreading and elevated protrusive activity. We found that the extent of these secondary morphodynamic changes depends on the duration of Src activation. This approach allowed us to dissect the different roles of PI3K and Rac1 signaling in regulation of morphological changes during Src activation and following Src inactivation. Inhibition of PI3K or Rac1 at the time of Src activation impedes Src‐mediated cell spreading, while only PI3K inhibition alters Src induced protrusive activity. However, once Src is activated for 15 min neither PI3K nor Rac1 inhibition is sufficient to affect Src mediated cell spreading or protrusive activity. Following Src inactivation, secondary increase in cell area is PI3K dependent, while secondary stimulation of protrusive activity is Rac1 and PI3K dependent. Thus, this strategy enabled us to determine the role of Src‐PI3K and Src‐Rac1 sequential signaling. In conclusion, we present a novel broadly applicable strategy that provides new opportunities for dissecting kinase‐mediated cellular pathways.


M267 Phenotypic variability and plasticity in influenza A virus measured using multi‐spectral viral strains. M.D. Vahey1, D.A. Fletcher1,2; 1Department of Bioengineering, University of California, Berkeley, Berkeley, CA, 2Physical Biosciences Division, Lawrence Berkeley National Laboratory, Berkeley, CA Influenza A virus (IAV) remains a substantial threat to public health, causing hundreds of thousands to millions of deaths worldwide each year, despite the innate defenses provided by the respiratory epithelium and the availability of antiviral drugs. One reason for the virus’s persistence is its ability to rapidly adapt to changes in the host environment. Diversity in size, shape, and composition of virions produced from even a genetically clonal population may play an important role in its ability to adapt quickly, but studying this pleiomorphism and its connections to infectivity has not been possible due to the lack of appropriate molecular tools. To address the challenge of characterizing individual virions and monitoring how they change in response to cellular and chemical defense mechanisms, we developed replication‐competent influenza A virus strains that harbor small peptide tags on five of the virus’s major structural proteins (HA, NA, M2, M1, and NP). These tags can be non‐disruptively and quantitatively labeled with small‐molecule fluorophores in both live cells and released virions, allowing us to catalog phenotypic diversity at the single‐virus level across thousands of virions using fluorescence microscopy. With this system, we find that the protein composition of individual virions is coupled to morphology, and that proteins with competing activities (the receptor‐binding HA and the receptor‐destroying NA) vary considerably more across the population than proteins that package the genome (i.e. NP). The measured distributions in virion size and composition exhibit plasticity, shifting in response to changes in growth conditions, including temperatures representative of the upper and lower airways; the glycosylation state of the host cells; and treatment with the active form of the antiviral drug oseltamivir. These shifts are observable within a single round of infection, where mutations do not have a chance to accumulate through selection. We propose that compositional and morphological plasticity provides one mechanism by which a pleiomorphic virus may be able to rapidly adapt to changes in its host environment while maintaining virulence.

M268 Temporal design of cancer combinatorial therapy guided by single‐cell dynamics. S. Chen1,2, G. Lahav1; 1Systems Biology, Harvard Medical School, Cambridge, MA, 2Institute of Molecular Biology, Academia Sinica, Taipei, Taiwan Tumorgenesis is an evolutionary process that give rise to high degree of heterogeneity among cancer cells. This heterogeneity usually limits the efficacy of a single anticancer drug to fully eliminate all cells in a tumor. One approach for overcoming therapy resistance is by combining multiple drugs. Nevertheless, the current drug combination designs remain mostly relying on the trial and error approach. Here, we use dynamics of signaling molecules in single cells to help guide the design of drug combinations. Specifically, we showed that inhibition of the oncogene MDMX led to temporal changes in cellular state through biphasic dynamics of p53: an initial post‐mitotic pulse followed by low‐amplitude oscillations. When combined with DNA damage, the effects of MDMX was sharply different in these two phases; in the first phase, MDMX depletion was synergistic with DNA damage in causing cell death, whereas in the second phase depletion of MDMX inhibited cell death. Therefore a quantitative understanding of signal


dynamics and cellular state after drug treatment can be important for designing an optimal schedule of drug combinations.

M269 Systematic genetic analysis of subcellular morphology using yeast phenomics. B.J. Andrews1, E. Styles1, M. Mattiazzi Usaj1, A. Verster2, O. Krause3, H. Friesen1, C. Boone1; 1The Donnelly Centre, University of Toronto, Toronto, ON, 2Genome Sciences, University of Washington, Seattle, WA, 3Electrical and Computer Engineering, University of Toronto, Toronto, ON We have developed experimental and computational pipelines which combine array‐based yeast genetics and automated microscopy for systematic and quantitative cell biological screens or phenomics. In one project, we use the Synthetic Genetic Array (SGA) method to introduce fluorescent markers of key cellular compartments or cell cycle progression, along with sensitizing mutations, into yeast mutant collections. We then perform live cell imaging on the mutant arrays using HTP confocal microscopy to quantitatively assess the abundance and localization of our fluorescent reporters, providing cell biological readouts of specific pathways and cellular structures in response to thousands of genetic perturbations. For automated image analysis, we developed a hybrid computational pipeline that combines outlier detection and classical SVM‐driven phenotype labeling, as well as a neural network‐based approach. We have applied the hybrid analysis strategy to score screens of nineteen subcellular compartments, identifying more than 40 distinct mutant phenotypes. Our analysis has revealed that while many genes determine the phenotype of a single subcellular compartment, there are also a number of highly pleiotropic genes that impact the structure of numerous subcellular compartments. Also, our automated cell biological analysis clearly identified several phenotypes for specific compartments that appear to reflect distinct biology. We complemented our effort to map the genetic determinants of subcellular morphology experimentally with a literature curation endeavor aimed at creating a comprehensive database of subcellular morphological phenotypes associated with genetic perturbations. The literature curation project provided a gold standard for analysis of our experimental data, and also revealed significant gaps in our understanding of less experimentally accessible subcellular compartments. The curated information is assembled in a searchable database, using a systematic rhetoric to describe conditions and observable phenotypes, providing a resource with which to assess the results of my genome‐wide screens, and other cell biological data.

M270 Quantitative characterization of genetic parts and circuits using protoplasts for plant synthetic biology. K.A. Schuamberg1, M.S. Antunes2, T.K. Kassaw2, W. Xu3, C. Zalewski2, J.I. Medford2, A. Prasad1,3; 1School of Biomedical Engineering, Colorado State University, Fort Collins, CO, 2Biology, Colorado State University, Fort Collins, CO, 3Chemical and Biological Engineering, Colorado State University, Fort Collins, CO Plant synthetic biology is of great interest for development of sustainable green technologies to meet basic needs. A key step in the rational design of synthetic gene networks is the quantitative characterization of genetic components to enable predictive modeling of synthetic network properties. However this poses special difficulties for plants since stably transforming plants is time consuming and can take months. An alternative strategy is the use of transient protoplast assays for quantitative testing of components. We developed an experimental test bed for characterizing genetic circuits consisting of externally inducible repressors using protoplasts, but found that transient protoplast assays show


significant experimental variability that makes direct quantitative comparisons yield incorrect results. We designed an experiment to understand the sources of variability, and discovered that protoplast experiments show significant batch effects, i.e. variation between protoplast batches is the greatest source of noise. The data also suggested that the noise was multiplicative in nature, and based on its analysis we developed a mathematical model for the batch effects and successfully tested it using simulated data. We then used the model to normalize the data and make quantitative comparisons between different inducible repressors. We assembled over one hundred genetic components in two plant species, Arabidopsis and Sorghum, and quantitatively studied their characteristics, and a small number of choice candidates were identified for further assembly into a plant toggle switch. Statistical analysis of the data did show that our method effectively removed the batch‐to‐batch variability. Our model was validated by stably transforming plants using a small subset of the characterized repressors, and comparing the quantitative properties between transient expression and stable transformation. Despite the significant uncertainties in the comparison, we showed that protoplast assays could roughly predict quantitative properties of synthetic circuits in stably transformed plants. We also carried out a statistical analysis of the data to uncover design principles for building orthogonal synthetic inducible repressors in plants (Nature Methods, v13, pp94–100, (2016)).

2016 ASCB Meeting-Oral Abstracts

Documents

Transcript of 2016 ASCB Meeting-Oral Abstracts