CELL BIOLOGY Lipid transport by TMEM24 at plasma …RESEARCH ARTICLE CELL BIOLOGY Lipid transport by...

RESEARCH ARTICLE SUMMARY◥

CELL BIOLOGY

Lipid transport by TMEM24 atER–plasma membrane contactsregulates pulsatile insulin secretionJoshua A. Lees,* Mirko Messa,* Elizabeth Wen Sun, Heather Wheeler, Federico Torta,Markus R. Wenk, Pietro De Camilli,† Karin M. Reinisch†

INTRODUCTION: Insulin is secreted by pan-creatic b cells in response to glucose stimu-lation. Its release is controlled by the interplayof calcium and phosphoinositide signalingpathways. A rapid release phase, in whichinsulin containing granules that are alreadydocked and primed at the plasma membrane(PM) undergo exocytosis, is followed by slowrelease. In this second phase, granules aredocked and primed and then released in aseries of bursts, each triggered by a spike incytosolic Ca2+.

RATIONALE: To better understand the mo-lecular basis underlying insulin secretion, wecharacterized TMEM24, a protein enriched inneuroendocrine cells previously suggested tobe required for a normal secretory response.

RESULTS:We found that TMEM24 is an en-doplasmic reticulum (ER) protein that concen-trates at ER-PM contact sites, where it tethersthe two bilayers. TMEM24 binding “in trans” tothe PM is negatively regulated by phosphoryla-tion in response to elevation of cytosolic Ca2+,so that TMEM24 transiently dissociates from thePM as Ca2+ concentration spikes and then re-associates with this membrane upon dephos-phorylation. Additionally, TMEM24 containsa lipid transport module of the synaptotagmin-like, mitochondrial and lipid-binding protein(SMP) family, which we structurally character-ized and showed to bind glycerolipids with apreference for phosphatidylinositol (PI). Thus,TMEM24 helps deliver PI, which is synthesizedin the ER, to the PM, where it is converted tophosphatidylinositol-4,5-bisphosphate [PI(4,5)P2]

to replenish pools of this lipid hydrolyzed dur-ing glucose-stimulated signaling. Supportinga key role of TMEM24 in the coordination ofCa2+ and phosphoinositide signaling, the lipidtransport function of TMEM24 is essential for

sustaining the intracellularCa2+ oscillations that triggerbursts of insulin granulerelease and hence insulinsecretion. PI(4,5)P2 is re-quired for Ca2+-dependentexocytosis. It also controls

the activity of PM ion channels that regulatecytosolic Ca2+ levels and is the precursor of IP3,which also helps to modulate cytosolic Ca2+

by triggering Ca2+ release from the ER. Thus,in insulin-secreting cells, TMEM24 participatesin coordinating Ca2+ and phosphoinositide sig-naling pathways to cause pulsatile insulin sec-retion (see the figure).

CONCLUSION: Our findings implicate ER-PMcontact sites and an ER resident lipid-transferprotein in the direct regulation of PM phospho-inositide pools, offering fresh insights into themechanisms of cellular phosphoinositide dynam-ics. More specifically, they elaborate the mech-anisms underlying insulin secretion, which isimpaired in patients with type II diabetes, andmay ultimately have therapeutic ramifications.▪

RESEARCH

Lees et al., Science 355, 709 (2017) 17 February 2017 1 of 1

The list of author affiliations is available in the full article online.*These authors contributed equally to this work and are listed inalphabetical order.†Corresponding author. Email: [email protected] (P.D.C.);[email protected] (K.M.R.)Cite this article as J. A. Lees et al., Science 355, eaah6171(2017). DOI: 10.1126/science.aah6171

TMEM24 activity cycle at ER-PM contacts.Glucose stimulation of insulin-secreting cells triggers Ca2+ influx, phospholipase C–dependent PI(4,5)P2cleavage, and granule exocytosis. Ca2+-stimulated phosphorylation causes TMEM24 dissociation from the PM and interruption of SMP-mediated PItransfer that allows PI(4,5)P2 resynthesis. Lowered PI(4,5)P2 levels attenuate the excitatory response. Dephosphorylation allows TMEM24 to return tothe PM to replenish PI(4,5)P2, permitting a new cycle of Ca

2+ elevation and secretion.

ON OUR WEBSITE◥

Read the full articleat http://dx.doi.org/10.1126/science.aah6171..................................................

on June 4, 2021

http://science.sciencemag.org/

Dow

nloaded from


RESEARCH ARTICLE◥

CELL BIOLOGY

Lipid transport by TMEM24 atER–plasma membrane contactsregulates pulsatile insulin secretionJoshua A. Lees,1* Mirko Messa,1,2,3,4* Elizabeth Wen Sun,1,2,3,4 Heather Wheeler,1,2,3,4

Federico Torta,6 Markus R. Wenk,6 Pietro De Camilli,1,2,3,4,5† Karin M. Reinisch1†

Insulin is released by b cells in pulses regulated by calcium and phosphoinositide signaling.Here, we describe how transmembrane protein 24 (TMEM24) helps coordinate thesesignaling events. We showed that TMEM24 is an endoplasmic reticulum (ER)–anchoredmembrane protein whose reversible localization to ER-plasma membrane (PM) contacts isgoverned by phosphorylation and dephosphorylation in response to oscillations incytosolic calcium. A lipid-binding module in TMEM24 transports the phosphatidylinositol4,5-bisphosphate [PI(4,5)P2] precursor phosphatidylinositol between bilayers, allowingreplenishment of PI(4,5)P2 hydrolyzed during signaling. In the absence of TMEM24,calcium oscillations are abolished, leading to a defect in triggered insulin release. Ourfindings implicate direct lipid transport between the ER and the PM in the control of insulinsecretion, a process impaired in patients with type II diabetes.

Eukaryotic cells harbor close appositions (10 to30 nm) between the plasmamembrane (PM)and the endoplasmic reticulum (ER), theorganelle where the synthesis of most mem-brane lipids begins. These ER-PM contact

sites are important in cellular calcium dynamicsand also play critical roles in lipid homeostasisand signaling (1–4). Identification and functionalcharacterization of the protein machinery at ER-PM appositions, including lipid-transfer proteinsthat carry lipids between membranes, have pro-vided key insights into the molecular mecha-nisms underlying processes that occur at thesesites (5–10).Here, we investigated the properties of TMEM24,

one of two similar proteins, TMEM24/C2CD2L andC2CD2, which bioinformatics studies predict tocontain a module typically found in proteins thatact at membrane contact sites (11, 12). TMEM24,which is highly expressed in professional neuro-secretory cells, is a regulator of insulin secretion(13). Insulin is produced by pancreatic b cells,where it is generated in the Golgi complex via

the cleavage of proinsulin, then packaged intogranules and stored (14). A rise in blood glucoseactivates tightly coupled calcium and phospho-inositide signaling pathways in these cells andleads to granule exocytosis, which occurs in twophases (15, 16). Immediately after glucose stim-ulation, a rise in intracellular calcium levels trig-gers secretion of the readily releasable granulepool (~5%), consisting of granules already dockedand primed at the PM (17). Next, a reserve gran-ule pool undergoes docking and priming stepsand is therefore released more gradually in a se-ries of small bursts dependent on pulsatile risesin calcium levels (15, 18). Insulin secretion isimpaired in patients with type II diabetes, high-lighting the importance of understanding themechanisms that underlie this process (14).Here, we probed the molecular mechanism by

which TMEM24modulates insulin exocytosis andidentified TMEM24 as a lipid transport proteinresident at ER-PM contact sites implicated in thecross-talk between calcium and PM phospho-inositide dynamics that regulates insulin release.We found that TMEM24 plays a role in transport-ing glycerolipids with a preference for phosphati-dylinositol (PI) from its site of synthesis in the ERto the PM, where it is a precursor for phospha-tidylinositol (4,5)-bisphosphate [PI(4,5)P2]. At thePM, PI(4,5)P2 has multiple roles in the cell, includ-ing in signal transduction (19, 20) and in the directcontrol of insulin granule exocytosis (14, 21–23)and PM ion channel function (14, 22, 23). Where-as TMEM24 localization to ER-PM contact sitesand its activity in transporting PI to the PM areregulated by calcium dynamics, the lipid trans-fer function of TMEM24 is required for calciumpulsatility.

Results and discussionTMEM24 is an ER-anchored proteinthat localizes at ER-PM contactsBioinformatics studies predict that the trans-membrane protein TMEM24 comprises a lipid-binding module of the tubular lipid-binding(TULIP) superfamily (12) (Fig. 1A), which weconfirmed by crystallography (see below). Allpreviously characterized members of the intracel-lular subfamily of TULIP proteins containing thismodule—knownas synaptotagmin-like,mitochond-rial, and lipid-binding protein (SMP) domain—areconcentrated atmembrane contact sites (10, 11, 24),suggesting that TMEM24 might also localize tosites of membrane apposition. To address thispossibility, we assessed localization of versionsof the human protein [TMEM24-enhanced greenfluorescent protein (EGFP)] in INS-1 cells, a ratinsulinoma cell line that, like pancreatic islets,exhibits high levels of endogenous TMEM24 (13)(fig. S1, A and B), or in HeLa cells (TMEM24-mCherry) that are not specialized for triggeredsecretion and have undetectable endogenous levelsof TMEM24 (13) (fig. S1, A and B). Confocalmicroscopy of the transfected cells showed thatmoderate levels of fluorescent TMEM24 coloca-lizedwith cotransfected ERmarkers,mRFP-Sec61bor ER-oxGFP, indicating a localization through-out the ER (Fig. 1, B and C, left insets in C). Inaddition, hot spots of fluorescencewere observedat the cell periphery, in a pattern consistent witha concentration at ER-PM contacts (24) (Fig. 1, Band C, right insets in C).To rule out a possible subcellular mistarget-

ing of exogenous TMEM24 due to the fluorescenttags, we assessed the localization of the overex-pressed untagged protein and of the endogenousprotein using newly generated antibodies directedagainst TMEM24 (Fig. 1D and fig. S1D), fol-lowed by confocal microscopy. An equatorial focalacquisition plane showed that overexpresseduntagged TMEM24 immunofluorescence accu-mulated at the cell periphery in spots with thecharacteristic pattern for ER-PM contact sites(fig. S1D, upper panel) and partially colocalizedwith the ER marker calreticulin (Fig. 1D andfig. S1D). This peripheral localization was furtherconfirmed by image acquisition of the ventralplane of cells, which highlighted the portions ofthe ER in close proximity with the plasma mem-brane (fig. S1D, lower panel). The same localizationwas observed for endogenous TMEM24 (Fig. 1D),which also partially colocalized with the ERmark-er protein disulfide isomerase (PDI). Absence ofTMEM24 immunoreactivity in INS-1 clusteredregularly interspaced short palindromic repeats(CRISPR)/Cas9–edited cells lacking TMEM24 val-idated the specificity of the signal (Figs. 1D and6B and fig. S1C).The localization of TMEM24 in the ER with a

concentration in the cortical ER differs from thereported localization of endogenous TMEM24 im-munoreactivity on insulin granules based on acommercially available TMEM24 antibody [Aviva#ARP47080_P050 (13)]. We confirmed this findingusing the same commercial antibodies. However,we found that such immunoreactivity was still

RESEARCH

Lees et al., Science 355, eaah6171 (2017) 17 February 2017 1 of 14

1Department of Cell Biology, Yale University School ofMedicine, New Haven, CT 06520, USA. 2Department ofNeuroscience, Yale University School of Medicine, NewHaven, CT 06510, USA. 3Howard Hughes Medical Institute,Yale University School of Medicine, New Haven, CT 06510,USA. 4Program in Cellular Neuroscience, Neurodegeneration,and Repair, Yale University School of Medicine, New Haven,CT 06510, USA. 5Kavli Institute for Neuroscience, YaleUniversity School of Medicine, New Haven, CT 06510, USA.6Department of Biochemistry, Yong Loo Lin School ofMedicine, National University of Singapore, 117599Singapore.*These authors contributed equally to this work and are listed inalphabetical order. †Corresponding author. Email: [email protected] (P.D.C.); [email protected] (K.M.R.)

on June 4, 2021


Dow

nloaded from


present in TMEM24 knockout (KO) INS-1 cells,suggesting that colocalization with insulin gran-ules reflected nonspecific binding (fig. S1E). Incontrast, no staining of insulin granules was pro-duced by our new antibodies (Fig. 1E).Several other proteins resident at ER-PM con-

tact sites, such as the extended synaptotagmins(E-Syts) or oxysterol-binding protein-related pro-

tein (ORP) family members, function as tethersthat promote the expansion of contacts when over-expressed (6, 10, 24). To assess whether TMEM24also functions as a tether, we overexpressedTMEM24-EGFP in HeLa cells, together with horse-radish peroxidase (HRP) tagged with an ER reten-tion motif [ssHRP-KDEL (25)] and imaged thecells by electron microscopy (EM) cytochemistry.

The presence of electron-dense HRP reaction prod-uct in the lumen of the ER facilitates identifi-cation of this organelle in EM images. Consistentwith the tethering function of TMEM24, over-expression of TMEM24 resulted in a robust increasein PM profiles engaged in direct close appositionwith the ER relative to control cells expressingssHRP-KDEL alone (Fig. 1, E and F).


Fig. 1. TMEM24 is a tether at endoplasmicreticulum (ER)–plasmamembrane (PM) contactsites. (A) Schematic representation of TMEM24.(B) Overexpression of human TMEM24-EGFP inINS-1 insulinomacellsmonitored byconfocalmicros-copy. TMEM24-EGFP partially colocalizes with ERmarker mRFP-Sec61b throughout the ER and is ad-ditionally concentrated at hot spots at the cellperiphery (arrows), as expected for ER-PM contactsites. (C) Overexpression of TMEM24-mCherry inHeLa cells monitored by confocal microscopy.TMEM24-mCherry colocalizes with ER marker ER-oxGFP throughout the reticular and cortical ER butis greatly enriched in the cortical ER (right insets).(D) Immunofluorescence for TMEM24 and ERmarker PDI in mixed TMEM24 KO (asterisks) andWT INS-1 insulinoma cells.TMEM24mainly localizesto theperipheryof the cells. (E) Immunofluorescencefor TMEM24 and insulin inWTand TMEM24KO INS-1 cells.TMEM24 immunoreactivity is onlyobserved inWT cells, is concentrated at the cell periphery, anddoes not colocalize with insulin granules. (F) Repre-sentative EM images of HeLa cells transfected withssHRP-KDEL alone orwith full-length TMEM24-EGFPorDCTermTMEM24-EGFP. Arrowheads indicate ER-PMcontacts.Note the increase in contact site numberafter overexpression of TMEM24-EGFP. (G) Quantita-tion of experiments described in Fig. 1F. Number(left panel; mean ± SEM) and average length (rightpanel; mean ± SEM) of ER-PM contacts increasesrelative to control only when full-length TMEM24 isoverexpressed. Scale bars are 5 mm for (B) and (D),10 mm for (C), and 200 nm for (F). P values are


Fig. 2. TMEM24 N-terminal TM domain and C-terminal unstructuredregion are required for localization to ER-PM contacts. (A to F) Com-parative analysis of localization of full-length, DTM, and DCTerm TMEM24-EGFPwhen overexpressed in COS-7 [(A) to (C)] and INS-1 [(D) to (F)] cells monitoredby confocal microscopy. mRFP-Sec61b was used as ER marker. Given thedifferent shape of the two cell types (top drawings), the plasma membranewas observed “en face” in the thin lamellipodium of the COS-7 cells but incross section in the INS-1 cell. Full-length TMEM24 [(A) and (D)] has a diffuselocalization in the ER but clusters at hot spots that represent ER-PM contactsites. The reticular shape of the ER is optimally appreciated in the “en face”view of the COS-7 cell.The construct lacking the transmembrane region (DTM)[(B) and (E)] localizes primarily at the PM, where it has a diffuse distribution

(as clear by the different views of the COS-7 cell and the INS-1 cell). The con-struct lacking the C-terminal region (DCTerm) [(C) and (F)] precisely colocalizeswith the ER marker Sec61b. Scale bar is 2 mm. (G) Purified human TMEM24lacking the transmembrane domain (residues 36 to 706) was incubated withliposomes containing 10% PE and either 20% PS or 5% other test lipid, withthe remainder PC. Liposomes were sedimented and analyzed for associatedprotein. DTM-TMEM24 directly binds acidic membranes without specific lipidpreference. PA and PI are less charged than the other phosphoinositides in thepanel, and their charge, closer to the hydrophobic layer, may be more shieldedby the head groups of neighboring phospholipids. (H) Purified SMP-C2 frag-ment of TMEM24 (residues 76 to 414) does not sediment with membranes inthe presence of EDTA or calcium. Liposome compositions are as in (G).

RESEARCH | RESEARCH ARTICLEon June 4, 2021


Dow

nloaded from


The transmembrane segment mediatesTMEM24 ER association and theC-terminal region mediates PM associationTMEM24must bind the PM to function as an ER-PM tether. The protein comprises a very shortlumenal N-terminal domain, a transmembrane(TM) segment, the SMP domain, a C2 domain, anda predicted unstructured, polybasic C-terminalregion (residues 415 to 706) (12, 13) (Fig. 1A). Todetermine which portions of TMEM24 are re-sponsible for the tethering of the ER to the PM,we monitored the localization of EGFP fusionsof TMEM24 deletion constructs and of the ERmarker mRFP-Sec61b coexpressed as a control(Fig. 2). Both INS-1 (professional secretory) andCOS-7 (fibroblast-like) cells were used for thisanalysis.TheERdistributionof full-lengthTMEM24was particularly striking in “en face” views of thevery thin edges of COS-7 cells, where intense spots

representedER-PMcontactsites(Fig.2A).ATMEM24construct lacking the TM segment was partiallydiffuse throughout the cytosol but concentratedselectively at the PM, as clearly seen in equatorialsections of INS-1 cells (Fig. 2, B and E). However, incontrast to full-length TMEM24 (Fig. 2, A andD),it decorated the PM profile continuously, ratherthan at hot spots, as would be the case if it wererestricted to the cortical ER (Fig. 2, B and E).Thus, this construct, which lacks the TM anchor,contains thePMbindingdeterminant(s). Conversely,a construct lacking the C-terminal region followingtheC2domain localized exclusively to the reticularand not the cortical ER (Fig. 2, C and F), indi-cating that this region of the protein mediates as-sociation with the PM. Accordingly, overexpressionof this construct does not promote the formationof contact sites, as confirmed by EM (Fig. 1, Fand G). The SMP or C2 domain fragments had

a diffuse cytosolic localization, and deletion of theseregions from TMEM24 resulted in constructs withthe same localization as full-length TMEM24 (fig.S2A). These two domains, therefore, are not re-quired for the localization of the protein toER-PMcontact sites.To assess whether TMEM24 might interact di-

rectly with the bilayer of the PM, rather than viaa protein factor, we expressed the cytosolic por-tion of TMEM24 (residues 36 to 706) recombi-nantly and used the purified protein in bindingassays with a panel of liposomes of varying compo-sitions (Fig. 2G). TMEM24 bound acidic liposomescontaining either phosphatidylserine (PS) [see also(26)] or phosphoinositides, without preference fora specific phosphoinositide, including PI(4,5)P2,a key determinant of PM identity. Supporting thishypothesis, depletion of this phosphoinositidein the PM by light-induced recruitment of the


Fig. 3. TMEM24 localization at ER-PM contact sites is regulated by aphosphorylation-dephosphorylation mechanism. (A) TMEM24 localiza-tion at ER-PM contacts, as monitored by TIRF microscopy, is not lost uponPM PI(4,5)P2 depletion by blue light–induced CRY2-OCRL 5-phosphatasedomain recruitment to the PM in COS-7 cells. (B) HeLa cells transfected withTMEM24-EGFP were analyzed by TIRFmicroscopy. In basal conditions,TMEM24forms PM puncta corresponding to ER-PM contact sites (left).Treatment withthapsigargin induces partial dissociation of TMEM24 from the PM, as revealedby loss of fluorescence in the TIRF plane (middle). Subsequently, TMEM24rapidly reassociated with the PM, overshooting basal fluorescence intensity(right). (C) Quantitation of (B). (D) HeLa cells overexpressing TMEM24-EGFPwere preincubated with the PKC inhibitor bisindolylmaleimide II before stim-

ulation with thapsigargin. PKC inhibition prevented TMEM24 dissociationfrom the PM but did not block excess TMEM24 PM accumulation duringrecovery. (E) In vitro phosphorylation of a cytosolic TMEM24 construct bypurified PKC. 32P incorporation onto TMEM24 by PKC (autoradiogram ininset) (30 min, 37°C) followed Michaelis-Menten kinetics with a Km of 0.81 ±0.085 mM. (F) Calcineurin/PP2B inhibition by pretreating HeLa cells over-expressing TMEM24-EGFP with cyclosporine before thapsigargin treatmentinhibited PM reassociation of TMEM24. Thapsigargin-mediated dissociationfrom the PM was unaffected by cyclosporine, which did not impair PKC ac-tivity. (G) CaMK inhibition by KN-93 did not affect TMEM24 dissociation fromthe PM. Recovery of TMEM24 is partially attenuated under KN-93 treatmentdue to loss of downstream calcineurin/PP2B activation by CaMK.



Dow

nloaded from


5-phosphatase domain of Lowe oculocerebrorenalsyndrome protein (OCRL) fused to cryptochrome2 (CRY2) (27, 28) did not affect the localizationof TMEM24 to contact sites in COS-7 cells (Fig. 3A).Thus, TMEM24 binds the PM selectively becausethe cytoplasmic leaflet of this membrane is high-ly acidic owing to an enrichment of PS and phos-phoinositides, which function redundantly.Consistent with the data obtained in living cells,

we found that lipid bilayer binding in vitro wasdependent on the C-terminal portion of TMEM24

because a fragment comprising only the SMPand C2 domains (residues 76 to 414) did not bindacidic liposomes. Nor did the same construct bindliposomes in the presence of calcium (Fig. 2H),which induces membrane interactions for a num-ber of previously characterized C2 domains (29).Sequences in the TMEM24 C terminus are highlyconserved (fig. S2B), indicative of functional im-portance, and are enriched in basic residues, whichare well suited to mediate interactions with acidicmembranes.

Regulation of TMEM24 localization bycalcium-dependent phosphorylationand dephosphorylationThe critical role of calcium dynamics in regulatedsecretion and the previously observed require-ment of TMEM24 for efficient insulin release (13)prompted us to investigate whether calcium regu-lates TMEM24 localization. To this end, we treatedHeLa cells transfected with TMEM24-EGFP withthapsigargin, which elevates cytosolic calciumlevels by blocking the ER calcium pump (SERCA)


Fig. 4. TMEM24 harbors a phosphatidylinositol (PI)–binding SMP mod-ule. (A) TMEM24 (76 to 260) adopts an SMP domain fold. TMEM24 SMPdomain is modeled as a dimer by analogy to the E-Syt2 SMP domains (atright). Disordered segments are represented as dashed lines. TMEM24 helixa2 and strand b13 move closer compared with the corresponding helix andstrand in the liganded E-Syt2 structure. Four lipid molecules bound by E-Syt2dimer are indicated. (B) Native electrospray ionization mass spectrometry(ESI-MS) analysis of TMEM24 SMP domain purified from Expi293 cells indi-cates dimerization, with masses consistent with one phospholipid moleculebound per monomer.Whereas the monomeric fraction exists in both apo and

holo form, the dimeric fraction is mostly complexed with lipid. (C) SEC ofpurified TMEM24 SMP domain preincubated with liver PI indicates that PIbinds and comigrates with TMEM24 SMP. Although TMEM24 SMP migratesas a dimer in solution, a small shift in apparent molecular weight upon bindingto PI occurs, consistent with a change in conformation but not oligomerizationstate. (D) Analysis of individual fractions from 14.0 to 16.0 ml by SDS-PAGEconfirms redistribution of the liganded SMP domain. (E) Fractions between14.0 and 16.0 ml retention volume were pooled; lipid was extracted andresolved by thin-layer chromatography. The PI band did not appear in thesefractions when TMEM24 SMP or liver PI was subjected separately to SEC.



Dow

nloaded from



Fig. 5. TMEM24 functions as a lipid-transfer pro-teinwith a preference for PI. (A) In vitro lipid-transferassays with TMEM24. (Top) Heavy donor (blue) andlight acceptor (yellow) liposomes were mixed in thepresence or absence of a recombinant N-terminallyHis12-tagged construct comprising the cytosolic por-tion of TMEM24 (residues 36 to 706). After lipidtransfer (15 or 30 min),TMEM24 was degraded withprotease and liposomes were separated by centrif-ugation. Donor and acceptor liposomes were separately analyzed for lipid content by HPLC or scintillation counting. (Bottom left) TMEM24 was tethered todonor liposomes containing PS via the basic C terminus and to acceptor liposomes containing 2 mole % (mol %) DOGS-NTA-Ni2+ via its N-terminal His12 tag.Transfer to acceptor liposomes of PS or PI, originally present at 20 mol % only in the donor liposomes, was measured by HPLC analysis after 30-min transfer.(Bottom right) TMEM24 was tethered to donor liposomes containing DOGS-NTA-Ni2+ via its N-terminal His12 tag and to acceptor liposomes containing PSvia the basic C terminus. Transfer to acceptor liposomes of 14C-PE or 3H-PI, originally present only in the donor liposomes, was measured by scintillationcounting after 15-min transfer. (B) Quantitation of the assays described in (A). TMEM24 preferentially transfers PI over PS (top panel) and PE (bottompanel). Mean ± SD indicated. (C) Schematic representation of light-induced TMEM24 PM recruitment strategy. TMEM24 was rapidly recruited to the PMby light-induced dimerization of CRY2 and CIBN modules. A version of TMEM24 with its C terminus replaced by a CRY2-mCherry module was expressedin COS-7 cells. The CIBN module is localized to the PM via a C-terminal CAAX box. (D) Dynamics of TMEM24-CRY2-mCherry (black) and DSMP-TMEM24-CRY2-mCherry (red) fluorescence before, during, and after blue-light stimulation (20 pulses, 200 ms each), as monitored by TIRF microscopy. (E) TIRFmicroscopy of PI(4)P marker iRFP-P4M (left) and the PI(4,5)P2 marker iRFP-PH-PLCd (right) before, during, and after blue-light stimulation. Note theincrease in both phosphoinositides due to transfer of PI to the PM by TMEM24 after blue-light stimulation (black). PM recruitment of DSMP-TMEM24-CRY2-mCherry does not affect PM levels of PI(4)P or PI(4,5)P2 (red). P values are P = 0.0049 and P < 0.0001 for (B) upper and lower panels, respectively, and P <0.0001 for both (E) panels.



Dow

nloaded from



Fig. 6. Oscillatory localization of TMEM24 at the PM affects insulinsecretion in insulinoma cells. (A) (Top) High-glucose stimulation of WTINS1 cells triggers oscillations of TMEM24 (TMEM24-mCherry, top) and cal-cium (GcaMP6s, bottom) opposite in phase as monitored by TIRF micros-copy. (Inset) Depletion of TMEM24 (peak inverted) reaches maximum afterthe peak of calcium influx. Dotted lines indicate corresponding peaks. (Bottom)Kymograph shows that TMEM24 and calcium oscillations are opposite in phaseafter glucose stimulation. (B) WT and TMEM24 KO INS1 cell lysates wereresolved by SDS-PAGE and immunoblotted with antibodies against TMEM24and insulin, showing no difference in insulin biogenesis in the KO cells. Topband in anti-insulin panel is proinsulin, whereas bottom band is insulin.GAPDH was used as a loading control. (C) Calcium oscillations (monitored by GcaMP6s) are abrogated after glucose stimulation in TMEM24 KO INS1 cells.(D) Exogenous overexpression of TMEM24-mCherry in KO cells rescues glucose-induced calcium oscillations. Dotted lines indicate corresponding peaks.(E) Quantitation of Ca2+ oscillations per minute in WT, TMEM24 KO, and TMEM24 KO INS1 cells rescued with either full-length or DSMP TMEM24 uponglucose stimulation. (F) Dynamics of PI(4,5)P2 recovery in WT (black) and TMEM24 KO (red) INS1 cells overexpressing PI(4,5)P2 marker iRFP-PHPLCd aftertreatment with carbachol to deplete PI(4,5)P2 at the PM.TMEM24 KO cells show delayed recovery of basal PI(4,5)P2 level at the PM.This delay is rescued byexogenous overexpression of TMEM24-EGFP (dashed line). Black dots indicate time points for which significance is reported (n.s., no significant difference;*P < 0.05; **P < 0.01; ****P < 0.0001). Significances of difference for WT versus KO and KO versus KO+TMEM24 are indicated above and below traces,respectively. (G) Amount of secreted insulin from WTor TMEM24 KO INS1 cells after glucose stimulation was assayed by ELISA. Stimulated insulin secretion wasnormalized to secretion under basal conditions (mean ± SEM). (H) TMEM24 KO INS1 cells were transfected with full-length TMEM24-EGFP or DSMP-TMEM24, andinsulin secretion was assayed by ELISA. Only TMEM24 containing the lipid-transfer module was able to rescue insulin secretion. (I) Anti-insulin immunofluorescenceof WT (top) and TMEM24 KO (bottom) INS1 cells shows that the typical punctate distribution of insulin granules is not perturbed by loss of TMEM24 expression (asquantified at right). Scale bar is 5 mm. P values are 0.0003 for WTversus TMEM24 KO, 0.0005 for WTversus KO+DSMP, 0.0019 for KO versus KO+FLTMEM24, and0.0029 for KO+FL TMEM24 versus KO+DSMP in (E), 0.0408 in (G), and 0.0428 for control versus FL TMEM24 and 0.0031 for FL TMEM24 versus DSMP in (H).



Dow

nloaded from


and inducing indirectly (as a result of the depletionof calcium within the ER) the opening of store-operated calcium channels in the PM (30, 31).Upon thapsigargin treatment, TMEM24-EGFPrapidly dissociated from the PM, presumably re-distributing from the cortical to the reticular ER,then reassociated with this membrane to overshootbasal levels (Fig. 3, B and C).Inspection of polybasic sequences in the

TMEM24 C-terminal region, which mediate PMassociation, revealed an enrichment in serine andthreonine residues, including within sites that fita consensus (32, 33) for the Ca2+-dependent pro-tein kinase C (PKC) (fig. S2B). Phosphorylation byPKC and downstream kinases, which results in theacidification of the C-terminal segment, could ac-count for TMEM24’s dissociation from the acidicPM. Conversely, dephosphorylation could enableTMEM24 reassociation with the PM and couldresult in overshooting if TMEM24 is basally phos-phorylated at resting conditions. Protein serine/

threonine phosphatase 2B (PP2B/calcineurin) de-phosphorylates substrates, including PKC sub-strates, in a calcium-dependent manner and couldbe a player in this mechanism.Consistent with such a scenario, TMEM24 lo-

calization to or away from ER-PM contact sitesafter thapsigargin treatment was altered by drugsthat block either PKC or PP2B. Thus, when PKCwas inhibited by the bisindolylmaleimide II (34),thapsigargin treatment no longer induced disso-ciation of TMEM24 from the PM (Fig. 3D). Con-versely, TMEM24 reassociation with the PM afterthapsigargin treatment was impaired by the PP2Binhibitor cyclosporin A (35) (Fig. 3F). Additionally,in vitro phosphorylation of purified TMEM24 (Fig.3E) indicated that TMEM24 is a substrate for PKC.Inhibition of another major family of calcium-

dependent serine/threonine kinases, the Ca2+/calmodulin-dependent protein kinases (CaMKs),via the drug KN-93 (36), did not affect the disso-ciation of TMEM24 from PM upon thapsigargin

treatment, arguing against the involvement ofthis class of Ca2+-dependent protein kinases inthis effect (Fig. 3G). The loss of dephosphorylationovershoot in the presence of this inhibitor mayreflect indirect effects of CaMKs on PP2B activity(37, 38).

Structural features of the SMP domain

We expressed the putative SMP domain (resi-dues 76 to 260) of TMEM24 recombinantly inEscherichia coli, purified it by affinity and size-exclusion chromatography (SEC), crystallized it,and used the single-wavelength anomalous dis-persion method to determine its structure at 3.0-Åresolution (table S1). As predicted (39), the domainadopts a TULIP fold, with two alpha helices anda highly curved antiparallel beta sheet forming acornucopia-like structure (Fig. 4A). The fold isthe same as for the SMP module present in theE-Syts (40) and predicted for components of theER-mitochondrion encounter structure (ERMES)


Fig. 7. Model for TMEM24 modulation of pulsatile insulin secretionthrough effects on phosphoinositide and calcium dynamics. (Top left)In insulin-secreting cells at resting state, ER-anchored TMEM24 localizes toER-PM contact sites via interaction between its basic C-terminal region andthe PM. In this state,TMEM24 transfers PI from the ER to the PM, maintainingthe pool of PM phosphoinositides, including PI(4)P and PI(4,5)P2. PI(4,5)P2interacts with voltage-gated calcium channels, increasing their responsive-ness to depolarization caused by glucose stimulation. PI(4,5)P2 also par-ticipates in priming of insulin secretory granules in preparation for exocytosis.(Right) Upon glucose stimulation or membrane depolarization, calcium influxthrough the calcium channels stimulates insulin release while activating thePKC pathway, causing phosphorylation of the TMEM24 C terminus. ThisC-terminal modification is responsible for the dispersion of TMEM24 from

ER-PM contacts into the reticular ER.Transfer of PI to the PM by TMEM24 isabolished. Calcium also activates PLCd, which is responsible for cleavage ofPI(4,5)P2 to form DAG and IP3, partially depleting PI(4,5)P2 from the PM andfurther elevating levels of cytosolic calcium via activation of IP3 receptorsin the ER membrane. The loss of PI(4,5)P2 eventually promotes the closedstate of PM calcium channels and terminates IP3 production. (Bottom left)Calcium-induced activation of the S/T-phosphatase calcineurin/PP2B resultsin dephosphorylation of the TMEM24 C terminus, allowing TMEM24 torepopulate ER-PM contact sites. PI transfer to the PM resumes to replenishPI(4)P and PI(4,5)P2 pools, reenabling calcium channel opening and priminga new population of insulin granules. The oscillatory interplay between cal-cium and TMEM24 continues in a cyclical fashion for the duration of glucosestimulation.



Dow

nloaded from


complex (39) (root-mean-square deviation be-tween the TMEM24 and E-Syt2 SMP domainsis 12.4 Å over 128 Ca positions).In the crystal, the TMEM24 SMP domains are

arranged as trimers. This oligomerization statelikely is a crystal packing artifact, because bothSEC and mass spectrometry analyses indicate thatthe domain is dimeric in solution (Fig. 4, B and C).In both E-Syt2 and the ERMES complex subunitMmm1, the SMP domains are arranged as head-to-head dimers (40, 41), and most likely theTMEM24 TULIP modules associate similarly,with the cornucopia tips pointed away from theinterface (Fig. 4A). Dimerized in this way, TMEM24contains a cavity lined with hydrophobic residueswell suited as a lipid binding pocket. The bindingsite is similar to that in E-Syt2, except that in E-Syt2the cavity connects to solvent via a seam alongmostof its length. The structural elements borderingthe seam in E-Syt2 are closer together in TMEM24(TMEM24 helix a2 and strand b13 in Fig. 4A),eliminating the seam and narrowing the cavity ascompared with E-Syt2. As a result, the particularsof lipid binding in TMEM24may be different fromthose in E-Syt2. Indeed, a mass spectrometry anal-ysis of a TMEM24-lipid complex indicates thatonly one glycerolipid is bound per SMP monomer(Fig. 4B) rather than two, as in E-Syt2. We haveso far been unable to obtain a crystal structure oflipid-bound TMEM24, and we suspect that crystalcontacts induce distortions in the TMEM24 SMPdomain to make it incompatible with lipid binding.Thus, the details of how TMEM24 interacts withlipids remain unclear.

TMEM24 binds and transportsglycerolipids with a preference for PI

It has been reported that other TULIP-domainproteins bind and transport lipids (5, 40, 42), sug-gesting that TMEM24 might as well. To identifypotential lipid ligands for TMEM24, we over-expressed its SMP domain in mammalian cells(Expi293) and used mass spectrometry to identifylipids that copurify with it. A similar approachwas successful in identifying the ligands of otherlipid transport proteins (6, 40, 43). Similar to theE-Syts and SMP-domain proteins in the ERMEScomplex (40, 41), the TMEM24 SMPmodule boundglycerophospholipids (table S2). The proportionof the lipids bound [37.9% phosphatidylcholine(PC), 15.9% phosphatidylethanolamine (PE),35% phosphatidylinositol (PI), 10.0% phospha-tidylserine (PS), rest other glycerophospholipids](table S2) reflects their relative abundance incells, except for PI, which constitutes only 5 to10% of total cellular glycerophospholipid and wasenriched in the TMEM24 bound fraction (44)(table S2). These data, particularly when con-trasted with the lipid analysis of the SMP do-main of E-Syt2 (40), indicated PI as a potentialpreferred TMEM24 ligand. Accordingly, PI comi-grated with a TMEM24 SMP dimer, as assessedby SEC and thin-layer chromatography (TLC), con-firming binding (Fig. 4, C to E).We carried out in vitro transfer assays to de-

termine whether TMEM24 can transfer PI be-tween membranes (Fig. 5A). In these experiments,

a cytosolic N-terminally dodecahistidine (His12)–tagged TMEM24 construct was tethered between“heavy” (~400 nm in diameter, filled with 25%sucrose) and “light” (~100 nm in diameter, nosucrose) liposomes to permit the transfer of lipidsbetween them. One set of liposomes also con-tained DOGS-NTA-Ni lipid, allowing capture ofTMEM24 via the His12 tag, whereas the secondset contained PS for binding the TMEM24 Cterminus (Fig. 5A, top panels). After incubation,we removed the protein by proteolysis and sep-arated the heavy and light liposomes by cen-trifugation, then monitored transport of selectedlipids using high-performance liquid chromatog-raphy (HPLC) (to compare PI and PS transport) orscintillation counting (for PI versus PE) (Fig. 5A,lower panel). TMEM24 transferred both PE andPS less efficiently than PI, consistent with a pref-erence for PI transport (Fig. 5B).Next, we explored whether TMEM24may trans-

fer glycerolipids, and more specifically PI, to thePM in living cells. Enrichment of PI in the PM islowest among the glycerolipids (5). Furthermore,this pool is rapidly metabolized by kinases andphospholipases and thus needs to be continu-ously replenished to maintain steady-state levelsof its downstream products at rest and to com-pensate for acute reductions in response to phy-siological or pharmacological manipulations.We examined whether TMEM24 might partic-ipate in PI replenishment at the PM by deter-mining whether its acute recruitment affectedlevels of its downstream metabolites phospha-tidylinositol 4-phosphate (PI4P) and PI(4,5)P2,two lipids whose levels can be assessed in vivo bygenetically encoded fluorescence probes. To thisend, we modified the optogenetics approachdescribed above (27): The N-terminal domainof the transcription factor CIB1 (CIBN) was fusedto a C-terminal CAAX prenylation motif to induceits constitutive targeting to the PM, and the C-terminal polybasic sequence of TMEM24 was re-placed by an in-frame CRY2-mCherry cassetteto make its recruitment to the PM, via interactionwith CIBN, inducible by blue light (Fig. 5C).In COS-7 cells expressing these constructs, blue

light induced a reversible interaction of TMEM24-CRY2-mCherry with PM-anchored CIBN. This in-teraction recruited ER-anchored TMEM24 to thePM contact sites (Fig. 5D). The kinetics of the re-cruitment were slower than previously observedfor a cytosolic protein (27) because TMEM24 istethered to the ER and could interact with thePM only by diffusion through the ER membrane.Levels of PI4P (P4M probe) (Fig. 5E, left panel)and PI(4,5)P2 (PH-PLCd probe) (Fig. 5E, rightpanel) underwent rapid increases coincident withthe recruitment to the PM of the TMEM-CRY2-mCherry fusion protein (Fig. 5E, black lines). Incontrast, levels of these phosphoinositides didnot change using a TMEM24-CRY2-mCherry con-struct that lacked the SMP module (Fig. 5E, redlines). The only modest, ~10% increase in PMPI4P and PI(4,5)P2 after TMEM24 PM recruit-ment likely reflects the occurrence of homeostaticmechanisms that prevent the excess accumula-tions of these lipids in the PM.

Taken together, these data support the hy-pothesis that TMEM24 binds and transfers PIfrom the ER to the PM, where it can be convertedto PI4P and PI(4,5)P2 by enzymes resident there.

Pulsatile PM localization of TMEM24after glucose stimulation ininsulin-secreting cells and link tocalcium oscillations

Intracellular calcium levels in insulin-secretingcells oscillate during the slow-release phase ofsecretion after glucose stimulation (19, 20). Toelucidate whether the properties of TMEM24as described above are relevant in a physiolog-ical context, we investigated its dynamics andeffect on PI(4,5)P2 signaling upon glucose stim-ulation in INS-1 cells. We analyzed the localiza-tion of TMEM24 (TMEM24-mCherry) at the PMand of calcium levels (GCaMP6s) in response toglucose. Pulses of TMEM24 dissociation from thePM occurred opposite in phase with oscillationsin cytosolic calcium, suggesting a clear relationbetween the two events (Fig. 6A and fig. S3A).Stretching the time axis, however, revealed thatpeaks of calcium oscillations slightly precededpeaks of TMEM24 dissociation (Fig. 6A, inset),consistent with the hypothesis that calcium-dependent phosphorylation is responsible for thedissociation of TMEM24.It was shown that calcium oscillations are

abolished upon reduction of PI(4,5)P2 below acritical level in INS-1 cells. TMEM24 could be partof the mechanism that sustains the oscillations.Indeed, calcium oscillations were abrogated inTMEM24 KO cells (Fig. 6, B to E, and fig. S3B)and were rescued in KO cells transfected with afull-length TMEM24 construct (Fig. 6, D and E,and fig. S3C). In contrast, a DSMP construct didnot rescue calcium pulsatility (Fig. 6E), indicat-ing a requirement for the lipid transfer functionof TMEM24.To assess more directly the contribution of

TMEM24 to PI(4,5)P2 resynthesis, wild-type (WT)and TMEM24 KO cells were stimulated withcarbachol, which triggers phospholipase C (PLC)–dependent PI(4,5)P2 cleavage at the PM and in-creases intracellular calcium concentrations (19).In WT cells, PM PI(4,5)P2 depletion, as assessedby the iRFP-PH-PLCd signal in total internal re-flection fluorescent (TIRF) microscopy, recoveredand then overshot the basal levels. A similar over-all dynamic was observed in TMEM24 KO cells.However, the recovery was delayed (Fig. 6F), andsuch delay was rescued by expression of TMEM24(Fig. 6F), consistent with the hypothesis thatTMEM24plays a role in replenishing PMPI(4,5)P2.Because PMPI(4,5)P2 levels did eventually recovereven in the absence of TMEM24, there are clearlyadditional TMEM24-independent mechanisms tosupply phosphoinositides to the PM, as expected gi-ven theoccurrenceofotherPI transportproteins (45).

The TMEM24 lipid transfer moduleis required for stimulatedinsulin exocytosis

A previous report implicated TMEM24 in theregulation of insulin secretion (13) but did not




Dow

nloaded from


elucidate the molecular basis for its action. Weconfirmed the effect of TMEM24 on insulin re-lease by comparing glucose-stimulated insulinsecretion in WT and TMEM24 KO INS-1 cells,finding that secretion is strongly reduced in thelatter (Fig. 6G). Furthermore, the lipid transferfunction of TMEM24 was required for secretionbecause transfection of TMEM24 KO cells witha full-length TMEM24 construct rescued stimu-lated secretion, whereas a construct lacking theSMP lipid-transfer module did not (Fig. 6H). Sim-ilarly, overexpression of TMEM24 in WT INS-1cells enhanced evoked insulin secretion, whereasoverexpression of the SMP-deletion construct didnot (fig. S3D).We ruled out that the effect of TMEM24 on

evoked secretion was because of an effect on in-sulin granule biogenesis, because immunoblottingof lysates of WT and TMEM24 KO INS-1 cells(Fig. 6B) for both proinsulin and insulin revealedsimilar levels in the two cell populations. Further-more, immunofluorescence on WT or TMEM24KO INS-1 cells using an antibody against insulindid not show obvious differences in insulin gran-ule number or their overall distribution in thecell (Fig. 6I). These data, which are in agreementwith observations of Pottekat et al. (13), are con-sistent with a secretory defect owing to impairedgranule exocytosis.We suggest that one mechanism by which ab-

lation of this protein attenuates triggered exocy-tosis is through an impaired delivery of PI to thePM, a process that becomes critical for the regen-eration of PI(4,5)P2 during the recovery from itsPLC-dependent consumption. Lowered levels ofPI(4,5)P2 produced by the absence of TMEM24will in turn result in ablation of calcium oscilla-tions and hence a defect of Ca2+-dependent exocy-tosis. Direct effects of lowered levels of PI(4,5)P2or its metabolite diacylglycerol (DAG) on theexocytic machinery are also possible, becausePI(4,5)P2 is required for calcium-triggered exo-cytosis (21, 46) and DAG is required for the pri-ming steps leading to granule fusion (47, 48).Experiments on broken neuroendocrine cells haveshown that efficient secretory granule exocytosisrequires transport of PI to the PM and PI(4,5)P2synthesis in this membrane, most likely to re-plenish PI(4,5)P2 pools degraded upon cell lysis(21, 46, 49). A cytosolic PI transport protein (PITP)plays a role in this process (46). Our resultssuggest that the SMP-domain protein TMEM24may have a synergistic role with PITP in insulin-secreting cells, although TMEM24-mediated lipidtransfer is additionally coordinated with calciumdynamics.

Concluding remarks

In summary, our data demonstrate that TMEM24is an ER-PM contact site tether. After glucosestimulation in insulin-secreting cells and in re-sponse to a rise in intracellular calcium, theTMEM24 C terminus is phosphorylated andhence acidified, leading to TMEM24 disasso-ciation from the acidic PM and ER-PM contacts.Subsequent dephosphorylation within the cell al-lows TMEM24 to return to these sites (Fig. 7).

As TMEM24 is alternately phosphorylated andthen dephosphorylated, it localizes to ER-PMcontacts in a pulsatile manner opposite inphase with the variations in calcium. We showfurther that TMEM24 uses an SMP-family lipidtransfer module to transport PI, the precursorof PI(4,5)P2, from its site of synthesis in the ERto the PM (Fig. 7) and thus helps to replenishPI(4,5)P2 pools, which are degraded during PLCsignaling. Our data do not exclude that TMEM24may also be involved in the transfer of otherglycerolipids, a possibility that we have not ex-tensively explored.Moreover, we found that the lipid transfer

function of TMEM24 plays a critical role in cal-cium pulsatility. Because PI(4,5)P2 itself affectsthe activity of PM ion channels (22), including inbeta cells (23), and inositol 1,4,5-trisphosphate(IP3) generated from PI(4,5)P2 modulates calciumoscillations through a cross-coupling betweenPLC and IP3 receptor activity (50, 51), TMEM24most likely participates in calcium oscillationsthrough its role in replenishing PM PI(4,5)P2pools: Its disassociation from contact sites athigh cytosolic calcium allows PM PI(4,5)P2 todecrease through PLC activity. This ensures thatthe IP3 receptors (regulated by IP3) and the PMion channels that control calcium influx andare regulated by PI(4,5)P2 are inactivated, low-ering cytosolic calcium levels. In response tothis change, TMEM24 reassociates with thePM and helps replenish PI(4,5)P2 pools (Fig. 7and legend). Without calcium oscillations, in-sulin exocytosis would be impaired, as observedin both TMEM24 knockdown (13) and KO cells(Fig. 6G).Thus, in insulin secreting cells, TMEM24 co-

ordinates phosphoinositide and calcium dynam-ics, and it likely serves a similar function in otherpeptide-secreting neurosecretory cells. Phospho-inositide and calcium signaling pathways areconserved in all eukaryotes, suggesting TMEM24as a paradigm for proteins that couple these pro-cesses in other cell types. The TMEM24 paralogC2CD2 is one candidate that could have such afunction, as could other lipid-transfer proteinfamily members.Our results clearly point to a role for ER-PM

contact sites and their resident protein TMEM24in phosphoinositide metabolism and signaling.That membrane contact sites are key participantsin these processes is also supported by otherrecent work. As examples, Nir2 has recently beenimplicated in recycling PI(4,5)P2–derived phos-phatidic acid (PA) from the PM to the ER forconversion to PI after signaling events (52, 53),and we recently proposed that the E-Syts maysimilarly be involved in the recycling of anotherPI(4,5)P2 metabolite, DAG (5). Additionally, ORPfamily members, which have been shown tocounter-exchange PI4P to transport other lip-ids (PS and sterol) against a gradient (6, 7, 9),function in regulating PI4P levels at membranecontact sites (54). Our findings, while helpingto illuminate the complex workings underlyinginsulin secretion and pulsatility, also implicateER-PM contact sites and a resident lipid-transfer

protein in the direct regulation of PM phos-phoinositide pools, offering fresh insights intothe mechanisms of cellular phosphoinositidedynamics.

Materials and methodsAntibodies and reagents

Antibodies were obtained from the followingcommercial sources: rabbit anti-TMEM24 (AvivaBiotechnology), mouse anti-Calreticulin (EnzoLife Science), rabbit anti-Proinsulin/Insulin, guineapig anti-Insulin (AbCam), mouse anti-mRFP (Rock-land) and mouse anti-Actin (MP Biomedical).Mouse anti-GAPDH (glyceraldehyde-3-phosphatedehydrogenase ) was generated in our lab, whileRabbit anti-TMEM24 was generated by YenZymAntibodies LLC using a C-terminal peptide fromTMEM24 with the following sequence: CLRSGT-KLIFRRRPRQKE. Alexa488, Alexa594, andAlexa647conjugated secondary antibodies and Thapsigarginwere from Life Technologies; BisindolylmaleimideII, KN-93, Cyclosporine and Carbamoylcholinechloride (Carbachol) were from Tocris. Lipids werefrom Avanti Polar Lipids and American Radio-labeled Chemicals.

Plasmid construction

TMEM24 coding sequences (full-length, SMPdomain, C2 domain and C-terminal region) wereamplified byPCR fromhumancDNA. ThePCRprod-uct was cloned into pEGFP-N1 (BD Clontech) usingXhoI or BglII and EcoRI sites. Truncated forms(DTM, DSMP, DC2, DCterm) of TMEM24 were ob-tained by site-directed mutagenesis (QuikChangeII XL, Agilent Technology) of full-length TMEM24-EGFP. TMEM24 fragments amplified by PCRwerecloned into pCDF-C6PusingEcoRI or BamHI andNotI sites for bacterial expression with an N-terminal GST tag. pCMV6-AN-His (Origene) wasmodified to include an N-terminal His12 tag.TMEM24 DTM (36-706) was cloned using XhoIand NotI sites into pCMV6-AN-12His for Expi293cell expression. TMEM24SMP (76-260)was clonedinto pcDNA3.1-3xFLAG using BamHI and NotIsites for Expi293 cell expression as anN-terminal3xFLAG-tagged protein. The human TMEM24-mCherry vector resulted from PCR amplifica-tion of the coding sequence of TMEM24 (fromTMEM24-EGFP) followedby cloning intopmCherry-N1 (BD Clontech) using XhoI and EcoRI sites.To assemble the TMEM24-CRY2-mCherry vectorTMEM24, CRY2 and mCherry fragments werePCRamplified fromTMEM24-EGFP, 5-phosphatasedomain OCRL-CRY2 and pmCherry-C1, respec-tively. The amplification products were ligatedby using the In-Fusion cloning Kit (BD Clontech).TheresultingTMEM24-CRY2-mCherryplasmidwasused to generate DSMP-TMEM24-CRY2-mCherryvector by site directedmutagenesis (QuikChangeII XL, Agilent Technology). pGP-CMV-GCaMP6s[Addgene plasmid # 40753 (55)], the untaggedversion of TMEM24was fromOrigine (SC310158)andpSpCas9n(BB)-2A-Puro (PX462) V2.0 [Addgeneplasmid # 62987 (56)]. ER-oxGFP, mRFP-Sec61b,ssHRP-KDEL, PH-PLCd-iRFP, P4M-iRFP, CAAX-GFP-CIBN and OCRL-CRY2 were previouslydescribed (24).




Dow

nloaded from


Cell cultures and transfectionHeLa, COS-7 and HEK cells were cultured at37°C and 5% CO2 in DMEMmedium (Life Tech-nology) containing 10% fetal calf serum, 1 mMsodium pyruvate, 100 U/ml penicillin, 100 mg/mlstreptomycin and 2 mM glutamax (Life Technol-ogy). N2A and PC12 cells were cultured in RPMIcontaining 10% fetal calf serum, 1 mM sodiumpyruvate, 100 U/ml penicillin, 100 mg/ml strep-tomycin and 2 mM glutamax (Life Technology).INS1 cells (passage 30 to 50)were plated in RPMI-1640 with 11.1 mM D-Glucose and supplementedwith 10% fetal bovine serum, 100 U/ml penicillin,100 mg/ml streptomycin, 10 mM HEPES-Na pH7.4, 2 mM L-glutamine, 1 mM sodium pyruvate(Life Technologies), and 50 mM b-mercaptoethanol(Sigma Aldrich). 18 hours before experiments thestandard medium with 11.1 mM glucose was re-placedwith freshmediumcontaining 5mMglucose.In all our experiments, plasmids were trans-

fected by using Lipofectamine 2000 (Life Technol-ogy) and cells were cultured for 24 hours beforeanalysis.

Immunofluorescence

Cells were plated and grown on 5 mg/ml humanfibronectin (Millipore)-coated glass coverslipsand fixedwith 4%paraformaldehyde/4% sucrosein 0.1M sodiumphosphate buffer (pH 7.2) at roomtemperature. Coverslips werewashedwith 50mMNH4Cl pH 7.2, then blocked and permeabilized in0.1% Triton X-100 and 3% bovine serum albuminin PBS. Primary and secondary antibody incuba-tions were performed using the same buffer. Cov-erslips were rinsed in 1X PBS and mounted inProlongGold+DAPI (Life Technology). Immuno-fluorescence data acquisition was performed usingspinning disk confocal microscope (see below).

Live imaging

For live imaging, cells were incubated at 37°C inthe following buffer: 136 mM NaCl, 2.5 mM KCl,2 mM CaCl2, 1.3 mM MgCl2, 3 mM D-glucoseand 10 mM HEPES-Na pH 7.4. Spinning-diskconfocal microscopy was performed using theImprovision UltraVIEW VoX system (Perkin-Elmer) built on a Nikon Ti-E inverted micro-scope, equipped with PlanApo objectives (60×1.45-NA) and controlled by Volocity (Improvi-sion) software. TIRF microscopy was carried outon a Nikon TiE microscope equipped with 60×1.49-NA and 100× 1.49-NA objectives. Excitationlights were provided by 488 nm (for GFP) and561 nm (mCherry) diode-pumped solid-statelasers. An optical fiber coupled the lasers to theTIRF illuminator and an acousto-optic tunablefilter controlled the output from the lasers. Fluo-rescent signals were detected with an EM-CCDcamera (DU-887; Andor) and acquisition was con-trolled by Andor iQ software. Images typically weresampled at 0.2 Hz with exposure times in the 4to 6 s range.

Electron microscopy

Cells were fixed in 1.3% glutaraldehyde-0.1Msodium cacodylate. To activate the HRP reac-tion, cells were incubated in 0.5 mg/mL 3′3’-

Diaminobenzidine (DAB) plus 0.005% H2O2in 0.1M ammonium phosphate. They were post-fixed with 1% OsO4 in 1.5% K4Fe(CN)6 and 0.1Msodium cacodylate, en bloc stained with 2% ura-nyl acetate in 50 mM sodium maleate buffer pH5.2, dehydrated and embedded in Embed 812.Randomly selected cells were imaged at low mag-nification and cell perimeter was measured usingiTEM by Olympus. Higher magnification imagesof selected cells were taken and ER-PM contactsites were counted and measured in length. AllEM reagents were purchased from Electron Mi-croscopy Sciences (Hatfield, PA).

Neuronal culture preparation

All experiments involving mice were performedin accordance with the Yale University Institu-tional Animal Care and Use Committee. Briefly,primary cultures of hippocampal neurons weredissected from P0-P2 brains. Hippocampal neu-rons were dissociated by papain treatment andtrituration. Cells were counted in the presence ofTrypan Blue and seeded at 8.3 × 104 cells/cm2 onpoly-D-lysine-coated, glass-bottomedPetri dishes(MatTek). After two hours, the serum-based me-dia was carefully removed and replaced with neu-ronal growthmedia [Neurobasal A supplementedwith B27 andGlutamax (All reagents fromGibco)]andmaintained for at least 21 days in vitro (DIV)at 37°C and 5% CO2.All cultures were fed every 4 days with 10%

cortical enriched media and 15% glial enrichedmedia in freshly made neuronal growth media.

In vitro phosphorylation

In vitro TMEM24 phosphorylation by PKC wasperformed by incubating various concentrations(0-5 mM) of purified His12-TMEM24(36-706) with2 units of commercial PKC (Promega) in PKCreaction buffer (200mMHEPES pH 7.4, 16.7 mMCaCl2, 10 mM DTT, 100 mM MgCl2 and 0.6 mg/ml PS) in presence of 50 mM[g-32P]-ATP, (1.5 mCi/sample). Samples were incubated at 37°C for30min and the reaction stopped by immediatelyaddingLaemmli Stop buffer (57). Phosphorylationof TMEM24 was assessed by autoradiographyof the SDS–polyacrylamide gel electrophoresis(SDS-PAGE) gel and densitometric analysis ofthe bands.

Optogenetic depletion of PI(4,5)P2 at theplasma membrane and light-inducedrecruitment of TMEM24-CRY2-mCherry

5′-dephosphorylation of plasma membranePI(4,5)P2 by blue-light global illumination wasperformed as previously described (28, 58). Briefly,COS-7 cellswere transfectedwithCRY2-5-ptaseOCRL,EGFP-CIBN-CAAX, TMEM24-mCherry and thePI(4,5)P2 reporter PH-PLCd-iRFP. In transfectedcells, 20-30 pulses of blue light (488 nm, 200 ms,4 s apart) induce immediate binding of CRY2tomembrane-associated CIBN, thus bringing the5-ptase to the plasma membrane. A similar ap-proach was used to recruit TMEM24 to the plas-ma membrane to assess lipid transfer. Cells weretransfected with TMEM24-CRY2-mCherry (orits nonfunctional variant DSMP-TMEM24-CRY2-

mCHerry), EGFP-CIBN-CAAX, and the PI4P [P4M-iRFP (59)] or PI(4,5)P2 [iRFP- PH-PLCd (60)]probe. Blue light illumination induced TMEM24recruitment to the plasma membrane. TIRF mi-croscopy was performed as described above, di-merization was induced by 50x100 ms 488 nmlight pulses delivered through the evanescentfield. Because of spectral overlap of EGFPwith theabsorption spectrum of CRY2 (61), EGFP-taggedCIBN expression and localization could only beimaged when dimerization was promoted.

Generation of TMEM24 KO INS1 cellsby CRISPR/Cas9

TMEM24 KO INS1 cells were generated accord-ing to (56). Briefly, sgRNAs matching the firstexon of rat TMEM24 were designed using an on-line CRISPR Design Tool (http://tools.genome-engineering.org). Based on sequence analysisand quality score the oligos GE1 (CACCGCGG-GGGTTCATTGCGCGAGC) and GE2 (CACCGTG-CTGCGTTTGCGGGCGACC) were selected for theKO generation. After annealing, the oligonucleo-tides were cloned into the pSpCas9n(BB)-2A-Puroplasmid. Following sequence quality control theresulting vector was transfected into parental WTINS1 cells. After 24-hour transfection, cells wereselected by using 1-3 mg/ml puromycin (Sigma)and incubated for a total of 72 hours. Resistantclones were selected and growth in standardmedium. These clones were analyzed by PCR-genotyping and immunoblotting using an anti-TMEM24 antibody.

Insulin secretion assay

INS1 cell lines (WT or TMEM24 KO) were platedonto 24-well plates at a density of 800000 cells/well, transfected as described above and grownto 100% confluency before the secretion assay.Standard medium was replaced with fresh me-dium containing 5 mM glucose 18 hours beforeexperiments. Stimulation of Insulin secretionwasperformed as described in (62). Briefly, cells werewashed in 37°C warm HBSS (114 mM NaCl,4.7 mMKCl, 1.2 mMKH2PO4, 1.16 mMMgSO4,20 mMHEPES pH 7.2, 2.5 mM CaCl2, 25.5 mMNaHCO3, and 0.2% bovine serum albumin) con-taining 3 mM glucose. After 1 hour at 37°C, cellswere incubated in either basal HBSS or HBSSwith 25 mM glucose (0.8 ml volume) for 1 hourat 37°C. Supernatants were collected and cen-trifuged at 2000 rpm for 10-15 min at 4°C to re-move any cellular debris or floating cells. Cellswere independently processed for the analysis ofinsulin content. Enzyme-linked immunosorbentassay (ELISA) was performed by using a rat in-sulin ELISA kit (Mecodia) according to manu-facturer’s instructions.

Real Time Q-PCR

Total RNAwas extracted fromHeLa, COS-7, INS1cells and cortical neuronal by using RNeasyMiniKit (QIAGEN). cDNA was obtained from2 mg of RNA with the iScriptTM cDNA synthesiskit (Bio-Rad). The resulting DNA was subjected toreal-time qPCR with SYBR green fast kit (Bio-Rad)and prevalidated primers (Bio-Rad) according to




Dow

nloaded from

http://tools.genome-engineering.orghttp://tools.genome-engineering.orghttp://science.sciencemag.org/

manufacturer’s instructions. Quantification re-sults were expressed in terms of the cycle thresh-old (Ct). All real-time qPCR reactions were runin triplicate, and the Ct values were. Data werenormalized to the internal controls: the refer-ence gene S26 and HPRT. Differences betweenthe mean Ct values of each gene and those ofthe reference gene were calculated as DCt =Ctgene Ctreference and represented as 2 DCt as shownin fig. S1B.

Protein expression and purification

TMEM24 fragment bacterial expression constructswere transformed into BL21(DE3) RIL CodonPlus (Agilent) or C43(DE3) (Lucigen) E. coli cells.LB precultures were diluted into large-scale ex-pression cultures and grown at 37°C to an A600 of0.6-0.8, then protein expression was induced with0.5 mM isopropyl-b-D-thiogalactoside (IPTG) at26°C for 12-18hourswith shaking. Selenomethionine-substituted TMEM24(76-260) was produced aspreviously described (63). Cells recovered by cen-trifugation were resuspended in TNT500 buffer(50 mM Tris-Cl, 500 mM NaCl, 0.5 mM TCEP, 1xcomplete EDTA-free protease inhibitor cocktail(Roche), pH 7.5), then lysed in an Emulsiflex-C5cell disruptor (Avestin) at a pressure of 10,000-15,000 psi. Lysates were clarified by centrifuga-tion at 27,000xg for 45 min. Soluble TMEM24protein was bound to Glutathione Sepharose 4Fast Flow (GEHealthcare) resin and washed with3 × 10 bed volumes of TNT500 buffer. Protein wascleaved from the resin for 10-15 hours in TNT500buffer with PreScission protease. Cleaved proteinwas eluted, concentrated in a 3 kDa NMWCOAmicon centrifugal filtration device, and loadedon a Superdex 200 16/60 column (GEHealthcare).Peak fractions were recovered and concentrated.His12-TMEM24(36-706)was expressed inExpi293

cells (Invitrogen) according to manufacturer in-structions for 72 hours. Cells recovered by cen-trifugation were resuspended in TNT500 bufferand lysed by four freeze-thaw cycles in liquidnitrogen and at 37°C, respectively. Insolublemate-rial was pelleted at 15,000xg for 40min. TMEM24was recovered from the soluble fraction by in-cubation with Ni-NTA Agarose (Qiagen), fol-lowed by extensive washing. Protein was elutedwith 200 mM imidazole, concentrated, and frac-tionated in TNT500 buffer on a Superdex 200 10/30 GL column (GE Healthcare).

Protein crystallization, structuredetermination, and refinement

Selenomethionine-derivatized TMEM24(76-260)at 3.5 mg/ml was crystallized by vapor diffusionusing the sitting drop method in 100-200 mMMgCl2, 100 mM bis-tris propane, pH 6.5-8.0, 45%PEG 200, mixing the protein and well solutionin equal volumes (1 ml each). Crystals were re-covered in cryo-loops and flash frozen. All datawere collected at the Advanced Photon Source(NE-CAT beamline 24-ID-E) at 0.9792 Å to max-imize the selenium anomalous signal and datawere processed in XDS (64). We used the single-wavelength anomalous dispersion method to ob-tain de novo phases. Selenium positions were

identified using SHELX (65), and phases werecalculated and further improved by solvent flat-tening and noncrystallographic 6-fold averag-ing, as implemented in Phenix (66). COOT (67)was used for model building. Phases calculatedfrom partial models were iteratively combinedwith experimental phases to improve the mapsin unmodeled regions. The model was refined to2.96 Å using Phenix (66). Refinement was carriedout using least squares, real space, TLS andindividual B-factor, and torsion angle annealingoptions andwas alternated withmanual rebuildingin COOT. Figures were prepared using PyMol(v 1.7.2.0) software.

Lipid analysis by LC/MS

3xFLAG-TMEM24(76-260)was expressed in Expi293cells according to the manufacturer’s instruc-tions for 72 hours. Protein was immunoprecipi-tated, along with bound lipids, using anti-FLAGM2 affinity gel (Sigma), followed by elution with3xFLAG peptide (Sigma) and size exclusion chro-matography. Purified sample was sent to AvantiPolar Lipids, Inc. for analysis, where it was di-luted inHPLC-grade water and extracted into 2:1(v/v) chloroform:methanol. The chloroform layerwas dried and reconstituted with an internalstandard containing molecular species of PC,PE, PI, PS, PA, PG, and SM for quantitative mea-surement. The sample was injected onto a 100 ×2.1 mm i.d., 3.5 mmHILIC columnunder gradientconditions to resolve the phospholipids accord-ing to class and polarity. The molecular speciesfor each class were detected by an AB Sciex 4000QTrap using multiple reaction monitoring for~250 compounds. These compounds were quan-tified above a signal to noise of 10:1 and summedby lipid class.

Native TMEM24-lipid complex analysisby LC/MS

3xFLAG-TMEM24(76-260)was expressed in Expi293cells and purified as above. Purified protein sam-ple was exchanged into 20 mM ammonium ace-tate pH 6.8 using a Vivaspin 0.5 mL filter. Nativemass spectrometry analyses were performed inpositive ion mode. The sample was directly in-jected at a flowof 1 mL/min into aquadrupole time-of-flight (QTOF) 6550 Agilent mass spectrometerequipped with an Agilent 1260 HPLC systemthrough a SilicaTip nanospray emitter (New Ob-jective, USA). The nanoelectrospray voltage wasset to 1600 V (Vcap), temperature 280°C, dryinggas flow 11 L/min and fragmentor 175 V. The sam-ples were injected in 20 mM ammonium acetatepH 6.8 and spectra acquired at a scan rate of1 spectrum/sec. Data were collected over a massrange of 700-8000 m/z.

Liposome preparation

Lipid solutions in chloroform were mixed inglass test tubes, evaporated to dryness underN2 gas, then dried 30 min under vacuum. Forliposome sedimentation assays, lipid films wererehydrated under 20 mM Tris pH 7.5, 100 mMNaCl, 0.5 mM TCEP at 85°C for 1 hour, thenvortexed to generate crude liposomes. Crude lipo-

somes were frozen in liquid nitrogen and thawedat the rehydration temperature 7 times. Lipo-someswere used immediately or stored at –80°C.For lipid transfer assays, lipid films were rehy-drated under 20 mM Tris pH 7.5, 100 mM NaCl(for light liposomes) or 20mMTris pH7.5, 100mMNaCl, 0.75 M sucrose (for heavy liposomes) at 37°Cfor 1 hour. Rehydrated lipids were vortexed togenerate crude liposomes. Crude liposomes werefrozen in liquid nitrogen and thawed at 37°C 5times, then extruded through 100 nm (light lip.)or 400 nm (heavy lip.) pore size cellulose acetatefilters in a mini-extruder (Avestin). Liposomeswere clarified by centrifugation at 16,100xg for15 min. Light liposome supernatants were trans-ferred to a clean tube, while heavy liposome pel-lets were washed twice, then resuspended, all insucrose-free buffer. Liposomes were stored at 4°Cand used within 48 hours.

Liposome sedimentation assays

100 mg liposomeswere incubatedwith 1 mg proteinfor 15 min at room temperature, then liposomeswere pelleted at 16,100xg for 30 min at 4°C. Sup-ernatants were saved and pellets resuspended inbuffer. Equal proportions of supernatants andpellets were analyzed by SDS-PAGE and stainedwithCoomassie Brilliant Blue. Destained gelswereimaged electronically using aGE ImageQuant LAS4000 and the appropriate bands were quantitatedusing ImageJ (ImageJ, U. S. National Institutes ofHealth, Bethesda, Maryland, USA, https://imagej.nih.gov/ij, 1997-2016) to determine the proportionof liposome-bound protein.

In vitro TMEM24 SMP binding to PI

Purified TMEM24 (76-260)was incubated 48 hoursat 4°C with liposomes composed of 100% liver PIat a 10-fold molar excess over protein in 200 mltotal reaction volume. Control reactions were pre-pared by replacing either protein or lipids withbuffer. Protein and liposome mixtures were re-solved on a Superdex 200 10/300 Increase column(GE Healthcare). Peak fractions were sampled forSDS-PAGE, thenpooled. Lipidswere acid-extractedas previously described (68). The organic phasewas evaporated to dryness under nitrogen gas.Dried lipids were resuspended in 20 ml chloro-form:methanol (1:1), spotted on a silica gel 60TLC plate (EMD) pretreatedwith 50%methanol/1% potassium oxalate and resolved with freshlypreparedmobile phase [water:acetic acid:methanol:acetone:chloroform (7:16:12:15:32)] against a liverPI standard. Plateswere stainedwith iodine vaporand imaged.

Isolation of tritiated PI

COS-7 cells were incubated in inositol-freeDMEMsupplemented with dialyzed FBS and 20 mCi/ml3H-myo-inositol (American Radiolabeled Chem-icals) for 72 hours at 37°C. Cells were scraped afterincubation of plate on ice in 0.6 ml 4.5% per-chloric acid for 15 min., then pelleted and washedthree timeswith ice-cold 0.1MEDTA. Lipids wereacid-extracted as described above, then separatedby TLC, as above. Tritiated phosphoinositides onthe plate were detected by autoradiography. Bands




Dow

nloaded from

https://imagej.nih.gov/ij/https://imagej.nih.gov/ij/http://science.sciencemag.org/

containing 3H-phosphatidylinositol were scrapedfrom the TLC plate into a microcentrifuge tubeand PI was acid-extracted as before.

Lipid transfer assays

100 ml of light and heavy liposomes (each at1 mM final concentration) were incubated withrotation in the presence of 10 ml of 0.1 mg/mlprotein or 10 ml of buffer for 15 (scint. assay) or30 (HPLC assay) minutes, then quenched by theaddition of 100 ml stop cocktail (600 mM imid-azole, 10 mM EDTA, 0.1 mg/ml elastase, pH 8.0)and incubation for 60 min at room temperature.Heavy and light liposomes were separated bycentrifugation at 16,100xg for 15 min at roomtemperature. Supernatants were reserved, whileliposome pellets were washed once, then resus-pended in sucrose-free buffer. For radioactivelipids, supernatant and pellet fractions weremixed with EcoScint scintillation fluid and mea-sured by liquid scintillation counting. For un-labeled lipids, equal proportions of supernatantand pellet fractions were acid-extracted as de-scribed above. The organic phase was transferredto a clean tube and evaporated to dryness undernitrogen gas. Lipid content of each fraction wasanalyzed as described below.

Lipid head group analysis andHPLC quantitation

Extracted lipids were resuspended and deacy-lated in 40%methylamine:water:n-butanol:meth-anol (36:8:9:47) for 1 hour at 50°C and evaporatedto dryness under vacuum. Head groups were dis-solved and extracted into ultrapurewater againstn-butanol:petroleum ether:ethyl formate (20:40:1).The aqueous phase, containing the lipid headgroups, was removed and evaporated to drynessunder vacuum. Lipid head groups were then dis-solved in ultrapure water and filtered through0.22 mm cellulose acetate Spin-X filters (Corn-ing). Head groups were resolved in a disconti-nuous KOH gradient on an IonPac AS11-HC (2 ×250 mm) analytical column using a Dionex ICS-3000 ion chromatography system. Each lipid classwas separately quantitated in the Chromeleonsoftware suite by integration of the correspond-ing chromatogram peak areas.

Quantification and statistical analysis

Fluorescent signals were quantified with FijiImage J (http://fiji.sc/wiki/index.php/Fiji) or Volocity3D Image Analysis software (Improvision). Im-munoblots were analyzed by Fiji ImageJ or ImageStudio (Odyssey). Graphs weremade using GraphPad Prism (Graph Pad Software).Statistical analyses were performed by Graph

Pad Prism using Student’s t test for independentsamples, one-way (grouped samples) or two-way(two independent variables) ANOVA, followedby Tukey’s multiple comparison post hoc test.

REFERENCES AND NOTES

1. S. Tavassoli et al., Plasma membrane: Endoplasmic reticulumcontact sites regulate phosphatidylcholine synthesis. EMBORep. 14, 434–440 (2013). doi: 10.1038/embor.2013.36;pmid: 23519169

2. J. C. M. Holthuis, T. P. Levine, Lipid traffic: Floppy drives anda superhighway. Nat. Rev. Mol. Cell Biol. 6, 209–220 (2005).doi: 10.1038/nrm1591; pmid: 15738987

3. C. J. Stefan et al., Osh Proteins regulate phosphoinositidemetabolism at ER-plasma membrane contact sites. Cell 144,389–401 (2011). doi: 10.1016/j.cell.2010.12.034

4. C. J. Stefan, A. G. Manford, S. D. Emr, ER-PM connections:Sites of information transfer and inter-organellecommunication. Curr. Opin. Cell Biol. 25, 434–442 (2013).doi: 10.1016/j.ceb.2013.02.020; pmid: 23522446

5. Y. Saheki et al., Control of plasma membrane lipid homeostasisby the extended synaptotagmins. Nat. Cell Biol. 18, 504–515(2016). doi: 10.1038/ncb3339; pmid: 27065097

6. J. Chung et al., PI4P/phosphatidylserine countertransport atORP5- and ORP8-mediated ER-plasma membrane contacts.Science 349, 428–432 (2015). doi: 10.1126/science.aab1370;pmid: 26206935

7. B. Mesmin et al., A four-step cycle driven by PI(4)P hydrolysisdirects sterol/PI(4)P exchange by the ER-Golgi tether OSBP.Cell 155, 830–843 (2013). doi: 10.1016/j.cell.2013.09.056;pmid: 24209621

8. A. T. Gatta et al., A new family of StART domain proteins atmembrane contact sites has a role in ER-PM sterol transport.eLife 4, (2015). doi: 10.7554/eLife.07253; pmid: 26001273

9. J. Moser von Filseck et al., Phosphatidylserine transport byORP/Osh proteins is driven by phosphatidylinositol4-phosphate. Science 349, 432–436 (2015). doi: 10.1126/science.aab1346; pmid: 26206936

10. A. G. Manford, C. J. Stefan, H. L. Yuan, J. A. Macgurn,S. D. Emr, ER-to-plasma membrane tetheringproteins regulate cell signaling and ER morphology.Dev. Cell 23, 1129–1140 (2012). doi: 10.1016/j.devcel.2012.11.004; pmid: 23237950

11. A. Toulmay, W. A. Prinz, A conserved membrane-bindingdomain targets proteins to organelle contact sites. J. Cell Sci.125, 49–58 (2012). doi: 10.1242/jcs.085118; pmid: 22250200

12. V. Alva, A. N. Lupas, The TULIP superfamily of eukaryoticlipid-binding proteins as a mediator of lipid sensing andtransport. Biochim. Biophys. Acta 1861 (8 Pt. B), 913–923(2016). pmid: 26825693

13. A. Pottekat et al., Insulin biosynthetic interaction networkcomponent, TMEM24, facilitates insulin reserve poolrelease. Cell Rep. 4, 921–930 (2013). doi: 10.1016/j.celrep.2013.07.050; pmid: 24012759

14. J. C. Hou, L. Min, J. E. Pessin, Insulin granule biogenesis,trafficking and exocytosis. Vitam. Horm. 80, 473–506 (2009).doi: 10.1016/S0083-6729(08)00616-X; pmid: 19251047

15. P. Rorsman, E. Renström, Insulin granule dynamics inpancreatic beta cells. Diabetologia 46, 1029–1045 (2003).doi: 10.1007/s00125-003-1153-1; pmid: 12879249

16. B. A. Spurlin, D. C. Thurmond, Syntaxin 4 facilitatesbiphasic glucose-stimulated insulin secretion from pancreaticbeta-cells. Mol. Endocrinol. 20, 183–193 (2006).doi: 10.1210/me.2005-0157; pmid: 16099818

17. S. Daniel, M. Noda, S. G. Straub, G. W. Sharp, Identification ofthe docked granule pool responsible for the first phase ofglucose-stimulated insulin secretion. Diabetes 48, 1686–1690(1999). doi: 10.2337/diabetes.48.9.1686; pmid: 10480595

18. T. K. Bratanova-Tochkova et al., Triggering and augmentationmechanisms, granule pools, and biphasic insulin secretion.Diabetes 51 (suppl. 1), S83–S90 (2002). doi: 10.2337/diabetes.51.2007.S83; pmid: 11815463

19. A. Wuttke, J. Sågetorp, A. Tengholm, Distinct plasma-membrane PtdIns(4)P and PtdIns(4,5)P2 dynamics insecretagogue-stimulated beta-cells. J. Cell Sci. 123, 1492–1502(2010). doi: 10.1242/jcs.060525; pmid: 20375060

20. S. Thore, A. Wuttke, A. Tengholm, Rapid turnover ofphosphatidylinositol-4,5-bisphosphate in insulin-secreting cellsmediated by Ca2+ and the ATP-to-ADP ratio. Diabetes 56,818–826 (2007). doi: 10.2337/db06-0843; pmid: 17327453

21. T. F. Martin, PI(4,5)P2-binding effector proteins for vesicleexocytosis. Biochim. Biophys. Acta 1851, 785–793 (2015).doi: 10.1016/j.bbalip.2014.09.017; pmid: 25280637

22. B. Hille, E. J. Dickson, M. Kruse, O. Vivas, B.-C. C. Suh,Phosphoinositides regulate ion channels. Biochim. Biophys.Acta 1851, 844–856 (2015). doi: 10.1016/j.bbalip.2014.09.010;pmid: 25241941

23. B. Xie et al., Plasma membrane phosphatidylinositol4,5-bisphosphate regulates Ca2+-influx and insulin secretionfrom pancreatic b cells. Cell Chem. Biol. 23, 816–826 (2016).doi: 10.1016/j.chembiol.2016.06.009

24. F. Giordano et al., PI(4,5)P(2)-dependent and Ca2+-regulatedER-PM interactions mediated by the extended synaptotagmins.

Cell 153, 1494–1509 (2013). doi: 10.1016/j.cell.2013.05.026;pmid: 23791178

25. T. Schikorski, S. M. Young Jr., Y. Hu, Horseradish peroxidasecDNA as a marker for electron microscopy in neurons.J. Neurosci. Methods 165, 210–215 (2007). doi: 10.1016/j.jneumeth.2007.06.004; pmid: 17631969

26. N. B. Caberoy, Y. Zhou, G. Alvarado, X. Fan, W. Li, Efficientidentification of phosphatidylserine-binding proteins by ORFphage display. Biochem. Biophys. Res. Commun. 386, 197–201(2009). doi: 10.1016/j.bbrc.2009.06.010; pmid: 19520055

27. O. Idevall-Hagren, E. J. Dickson, B. Hille, D. K. Toomre,P. De Camilli, Optogenetic control of phosphoinositidemetabolism. Proc. Natl. Acad. Sci. U.S.A. 109, E2316–E2323(2012). doi: 10.1073/pnas.1211305109; pmid: 22847441

28. M. J. Kennedy et al., Rapid blue-light-mediated induction ofprotein interactions in living cells. Nat. Methods 7, 973–975(2010). doi: 10.1038/nmeth.1524; pmid: 21037589

29. J. Rizo, T. C. Südhof, C2-domains, structure and functionof a universal Ca2+-binding domain. J. Biol. Chem. 273,15879–15882 (1998). doi: 10.1074/jbc.273.26.15879;pmid: 9632630

30. J. Liou et al., STIM is a Ca2+ sensor essential for Ca2+-store-depletion-triggered Ca2+ influx. Curr. Biol. 15, 1235–1241(2005). doi: 10.1016/j.cub.2005.05.055; pmid: 16005298

31. P. G. Hogan, R. S. Lewis, A. Rao, Molecular basis of calciumsignaling in lymphocytes: STIM and ORAI. Annu. Rev. Immunol.28, 491–533 (2010). doi: 10.1146/annurev.immunol.021908.132550; pmid: 20307213

32. K. Nishikawa, A. Toker, F. J. Johannes, Z. Songyang,L. C. Cantley, Determination of the specific substrate sequencemotifs of protein kinase C isozymes. J. Biol. Chem. 272,952–960 (1997). doi: 10.1074/jbc.272.2.952; pmid: 8995387

33. E. L. Huttlin et al., A tissue-specific atlas of mouse proteinphosphorylation and expression. Cell 143, 1174–1189 (2010).doi: 10.1016/j.cell.2010.12.001; pmid: 21183079

34. D. Toullec et al., The bisindolylmaleimide GF 109203X is apotent and selective inhibitor of protein kinase C. J. Biol. Chem.266, 15771–15781 (1991). pmid: 1874734

35. J. Liu et al., Calcineurin is a common target of cyclophilin-cyclosporin A and FKBP-FK506 complexes. Cell 66, 807–815(1991). doi: 10.1016/0092-8674(91)90124-H; pmid: 1715244

36. M. Sumi et al., The newly synthesized selective Ca2+/calmodulin dependent protein kinase II inhibitor KN-93reduces dopamine contents in PC12h cells. Biochem. Biophys.Res. Commun. 181, 968–975 (1991). doi: 10.1016/0006-291X(91)92031-E; pmid: 1662507

37. R. M. Mulkey, S. Endo, S. Shenolikar, R. C. Malenka,Involvement of a calcineurin/inhibitor-1 phosphatase cascadein hippocampal long-term depression. Nature 369, 486–488(1994). doi: 10.1038/369486a0; pmid: 7515479

38. I. Kameshita, A. Ishida, H. Fujisawa, Phosphorylation andactivation of Ca2+/calmodulin-dependent protein kinasephosphatase by Ca2+/calmodulin-dependent protein kinase II.FEBS Lett. 456, 249–252 (1999). doi: 10.1016/S0014-5793(99)00958-8; pmid: 10456318

39. K. O. Kopec, V. Alva, A. N. Lupas, Homology of SMP domains tothe TULIP superfamily of lipid-binding proteins provides astructural basis for lipid exchange between ER andmitochondria. Bioinformatics 26, 1927–1931 (2010).doi: 10.1093/bioinformatics/btq326; pmid: 20554689

40. C. M. Schauder et al., Structure of a lipid-bound extendedsynaptotagmin indicates a role in lipid transfer. Nature 510,552–555 (2014). doi: 10.1038/nature13269; pmid: 24847877

41. A. P. AhYoung et al., Conserved SMP domains of the ERMEScomplex bind phospholipids and mediate tether assembly.Proc. Natl. Acad. Sci. U.S.A. 112, E3179–E3188 (2015).doi: 10.1073/pnas.1422363112; pmid: 26056272

42. H. Yu et al., Extended synaptotagmins are Ca2+-dependent lipidtransfer proteins at membrane contact sites. Proc. Natl. Acad.Sci. U.S.A. 113, 4362–4367 (2016). doi: 10.1073/pnas.1517259113; pmid: 27044075

43. K. Maeda et al., Interactome map uncovers phosphatidylserinetransport by oxysterol-binding proteins. Nature 501, 257–261(2013). doi: 10.1038/nature12430; pmid: 23934110

44. J. L. Sampaio et al., Membrane lipidome of an epithelial cellline. Proc. Natl. Acad. Sci. U.S.A. 108, 1903–1907 (2011).doi: 10.1073/pnas.1019267108; pmid: 21245337

45. Y. J. Kim, M.-L. G. Hernandez, T. Balla, Inositol lipid regulationof lipid transfer in specialized membrane domains. Trends CellBiol. 23, 270–278 (2013). doi: 10.1016/j.tcb.2013.01.009;pmid: 23489878

46. J. C. Hay, T. F. Martin, Phosphatidylinositol transfer proteinrequired for ATP-dependent priming of Ca2+-activated




Dow

nloaded from

http://fiji.sc/wiki/index.php/Fijihttp://dx.doi.org/10.1038/embor.2013.36http://www.ncbi.nlm.nih.gov/pubmed/23519169http://dx.doi.org/10.1038/nrm1591http://www.ncbi.nlm.nih.gov/pubmed/15738987http://dx.doi.org/10.1016/j.cell.2010.12.034http://dx.doi.org/10.1016/j.ceb.2013.02.020http://www.ncbi.nlm.nih.gov/pubmed/23522446http://dx.doi.org/10.1038/ncb3339http://www.ncbi.nlm.nih.gov/pubmed/27065097http://dx.doi.org/10.1126/science.aab1370http://www.ncbi.nlm.nih.gov/pubmed/26206935http://dx.doi.org/10.1016/j.cell.2013.09.056http://www.ncbi.nlm.nih.gov/pubmed/24209621http://dx.doi.org/10.7554/eLife.07253http://www.ncbi.nlm.nih.gov/pubmed/26001273http://dx.doi.org/10.1126/science.aab1346http://dx.doi.org/10.1126/science.aab1346http://www.ncbi.nlm.nih.gov/pubmed/26206936http://dx.doi.org/10.1016/j.devcel.2012.11.004http://dx.doi.org/10.1016/j.devcel.2012.11.004http://www.ncbi.nlm.nih.gov/pubmed/23237950http://dx.doi.org/10.1242/jcs.085118http://www.ncbi.nlm.nih.gov/pubmed/22250200http://www.ncbi.nlm.nih.gov/pubmed/26825693http://dx.doi.org/10.1016/j.celrep.2013.07.050http://dx.doi.org/10.1016/j.celrep.2013.07.050http://www.ncbi.nlm.nih.gov/pubmed/24012759http://dx.doi.org/10.1016/S0083-6729(08)00616-Xhttp://www.ncbi.nlm.nih.gov/pubmed/19251047http://dx.doi.org/10.1007/s00125-003-1153-1http://www.ncbi.nlm.nih.gov/pubmed/12879249http://dx.doi.org/10.1210/me.2005-0157http://www.ncbi.nlm.nih.gov/pubmed/16099818http://dx.doi.org/10.2337/diabetes.48.9.1686http://www.ncbi.nlm.nih.gov/pubmed/10480595http://dx.doi.org/10.2337/diabetes.51.2007.S83http://dx.doi.org/10.2337/diabetes.51.2007.S83http://www.ncbi.nlm.nih.gov/pubmed/11815463http://dx.doi.org/10.1242/jcs.060525http://www.ncbi.nlm.nih.gov/pubmed/20375060http://dx.doi.org/10.2337/db06-0843http://www.ncbi.nlm.nih.gov/pubmed/17327453http://dx.doi.org/10.1016/j.bbalip.2014.09.017http://www.ncbi.nlm.nih.gov/pubmed/25280637http://dx.doi.org/10.1016/j.bbalip.2014.09.010http://www.ncbi.nlm.nih.gov/pubmed/25241941http://dx.doi.org/10.1016/j.chembiol.2016.06.009http://dx.doi.org/10.1016/j.cell.2013.05.026http://www.ncbi.nlm.nih.gov/pubmed/23791178http://dx.doi.org/10.1016/j.jneumeth.2007.06.004http://dx.doi.org/10.1016/j.jneumeth.2007.06.004http://www.ncbi.nlm.nih.gov/pubmed/17631969http://dx.doi.org/10.1016/j.bbrc.2009.06.010http://www.ncbi.nlm.nih.gov/pubmed/19520055http://dx.doi.org/10.1073/pnas.1211305109http://www.ncbi.nlm.nih.gov/pubmed/22847441http://dx.doi.org/10.1038/nmeth.1524http://www.ncbi.nlm.nih.gov/pubmed/21037589http://dx.doi.org/10.1074/jbc.273.26.15879http://www.ncbi.nlm.nih.gov/pubmed/9632630http://dx.doi.org/10.1016/j.cub.2005.05.055http://www.ncbi.nlm.nih.gov/pubmed/16005298http://dx.doi.org/10.1146/annurev.immunol.021908.132550http://dx.doi.org/10.1146/annurev.immunol.021908.132550http://www.ncbi.nlm.nih.gov/pubmed/20307213http://dx.doi.org/10.1074/jbc.272.2.952http://www.ncbi.nlm.nih.gov/pubmed/8995387http://dx.doi.org/10.1016/j.cell.2010.12.001http://www.ncbi.nlm.nih.gov/pubmed/21183079http://www.ncbi.nlm.nih.gov/pubmed/1874734http://dx.doi.org/10.1016/0092-8674(91)90124-Hhttp://www.ncbi.nlm.nih.gov/pubmed/1715244http://dx.doi.org/10.1016/0006-291X(91)92031-Ehttp://dx.doi.org/10.1016/0006-291X(91)92031-Ehttp://www.ncbi.nlm.nih.gov/pubmed/1662507http://dx.doi.org/10.1038/369486a0http://www.ncbi.nlm.nih.gov/pubmed/7515479http://dx.doi.org/10.1016/S0014-5793(99)00958-8http://dx.doi.org/10.1016/S0014-5793(99)00958-8http://www.ncbi.nlm.nih.gov/pubmed/10456318http://dx.doi.org/10.1093/bioinformatics/btq326http://www.ncbi.nlm.nih.gov/pubmed/20554689http://dx.doi.org/10.1038/nature13269http://www.ncbi.nlm.nih.gov/pubmed/24847877http://dx.doi.org/10.1073/pnas.1422363112http://www.ncbi.nlm.nih.gov/pubmed/26056272http://dx.doi.org/10.1073/pnas.1517259113http://dx.doi.org/10.1073/pnas.1517259113http://www.ncbi.nlm.nih.gov/pubmed/27044075http://dx.doi.org/10.1038/nature12430http://www.ncbi.nlm.nih.gov/pubmed/23934110http://dx.doi.org/10.1073/pnas.1019267108http://www.ncbi.nlm.nih.gov/pubmed/21245337http://dx.doi.org/10.1016/j.tcb.2013.01.009http://www.ncbi.nlm.nih.gov/pubmed/23489878http://science.sciencemag.org/

secretion. Nature 366, 572–575 (1993). doi: 10.1038/366572a0; pmid: 8255295

47. D. J. James, T. F. J. Martin, CAPS and Munc13: CATCHRs thatSNARE Vesicles. Front. Endocrinol. (Lausanne) 4, 187 (2013).doi: 10.3389/fendo.2013.00187; pmid: 24363652

48. L. Kang et al., Munc13-1 is required for the sustained release ofinsulin from pancreatic beta cells. Cell Metab. 3, 463–468(2006). doi: 10.1016/j.cmet.2006.04.012; pmid: 16697276

49. D. A. Eberhard, C. L. Cooper, M. G. Low, R. W. Holz, Evidencethat the inositol phospholipids are necessary for exocytosis.Loss of inositol phospholipids and inhibition of secretion inpermeabilized cells caused by a bacterial phospholipase C andremoval of ATP. Biochem. J. 268, 15–25 (1990). doi: 10.1042/bj2680015; pmid: 2160809

50. L. D. Gaspers et al., Hormone-induced calcium oscillationsdepend on cross-coupling with inositol 1,4,5-trisphosphateoscillations. Cell Reports 9, 1209–1218 (2014). doi: 10.1016/j.celrep.2014.10.033; pmid: 25456123

51. A. T. Harootunian, J. P. Kao, S. Paranjape, R. Y. Tsien,Generation of calcium oscillations in fibroblasts by positivefeedback between calcium and IP3. Science 251, 75–78 (1991).doi: 10.1126/science.1986413; pmid: 1986413

52. Y. J. Kim, M.-L. L. Guzman-Hernandez, E. Wisniewski, T. Balla,Phosphatidylinositol-phosphatidic acid exchange by Nir2at ER-PM contact sites maintains phosphoinositide signalingcompetence. Dev. Cell 33, 549–561 (2015). doi: 10.1016/j.devcel.2015.04.028; pmid: 26028218

53. C.-L. L. Chang, J. Liou, Phosphatidylin

CELL BIOLOGY Lipid transport by TMEM24 at plasma …RESEARCH ARTICLE CELL BIOLOGY Lipid transport by...

Documents

Transcript of CELL BIOLOGY Lipid transport by TMEM24 at plasma …RESEARCH ARTICLE CELL BIOLOGY Lipid transport by...