Bending “On the Rocks”—ACocktail of Biophysical Modules to...

Bending “On the Rocks”—A Cocktailof Biophysical Modules to Build EndocyticPathways

Ludger Johannes1,2, Christian Wunder1,2, and Patricia Bassereau3,4,5

1Institut Curie—Centre de Recherche, Traffic, Signaling and Delivery Group, 75248 Paris Cedex 05, France2CNRS UMR144, France3Institut Curie—Centre de Recherche, Membrane and Cell Functions Group, 75248 Paris Cedex 05, France4CNRS UMR168, France5Universite Pierre et Marie Curie, F-75252 Paris, France

Correspondence: [email protected]

Numerous biological processes rely on endocytosis. The construction of endocytic pits isachieved by a bewildering complexity of biochemical factors that function in clathrin-de-pendent and -independent pathways. In this review, we argue that this complexity can beconceptualized by a deceptively small number of physical principles that fall into two broadcategories: passive mechanisms, such as asymmetric transbilayer stress, scaffolding, linetension, and crowding, and active mechanisms driven by mechanochemical enzymesand/or cytoskeleton. We illustrate how the functional identity of biochemical modulesdepends on system parameters such as local protein density on membranes, thus explainingsome of the controversy in the field. Different modules frequently operate in parallel in thesame step and often are shared byapparently divergent uptake processes. The emergence of anovel endocytic classification system may thus be envisioned in which functional modulesare the elementary bricks.

Endocytosis is an intracellular traffickingevent in which the plasma membrane is the

donor compartment (Howes et al. 2010; McMa-hon and Boucrot 2011; Gonnord et al. 2012).Virtually any cellular function receives at somepoint endocytic input. Striking examples in-clude nutrient uptake, compartmentalizationand termination of intracellular signalingevents, establishment of morphogen gradients,and cellular entry of pathogens and pathogenicfactors. In this article, we focus our discussion

on endocytic processes at the plasma mem-brane. It is important to point out that theprinciples that we will be discussing (curvaturegeneration, motor activity, etc.) apply to otherintracellular trafficking steps, notably also to en-dosomal sorting.

The endocytic process is also an active fieldof discovery concerning the molecular mecha-nisms of intracellular trafficking, owing to theease at which material can be added at controlledrates and concentrations into the system, and

Editors: Sandra L. Schmid, Alexander Sorkin, and Marino Zerial

Additional Perspectives on Endocytosis available at www.cshperspectives.org

Copyright # 2014 Cold Spring Harbor Laboratory Press; all rights reserved; doi: 10.1101/cshperspect.a016741

Cite this article as Cold Spring Harb Perspect Biol 2014;6:a016741

1

on July 16, 2021 - Published by Cold Spring Harbor Laboratory Press http://cshperspectives.cshlp.org/Downloaded from

http://cshperspectives.cshlp.org/

very early molecular events can be visualizedwith advanced microscopy techniques (Merri-field 2004; Kaksonen et al. 2006; Kumari et al.2010). A confusing complexity of biochemicalfunctions has thereby been identified, which op-erate in the different phases of the endocyticprocess: initiation, cargo recruitment, mem-brane bending, and scission.

Despite an ever-increasing data density, ithas often proven difficult to unambiguously as-cribe a functional identity to given traffickingfactors or complexes, which we here term “bio-chemical modules,” or “modules” for short. Forexample, BAR-domain proteins have been po-sitioned as curvature inducers or sensors (Do-herty and McMahon 2009; Frost et al. 2009).Furthermore, given molecular functions oftenare accomplished by several complementarymodules, such as scission in the clathrin path-way—which appears to be driven by the mech-anochemical enzyme dynamin (Pucadyil andSchmid 2009)—by helix-inserting epsin (Bou-crot et al. 2012), and possibly also by the actincytoskeleton (Liu et al. 2006), at least in yeast.

In this review, we argue that a true under-standing of module identity requires two things.First, an exact description of the module’s func-tional potentialities is required. This typicallycomes from model membrane reconstitutionstudies in which experimental parameters canbe precisely tuned. The second requirement isthe transposition of these potentialities intothe cellularcontext in which the parameter spaceis determined by expression levels and activitystates. As aconsequence, onlya limited spectrumofpossibleactivitystateswillbeusedinthecellularcontext, and it is from these operationally relevantactivity states that endocytic processes are built.

PHYSICAL PRINCIPLES

General Physical Mechanisms to BendMembranes

Membrane trafficking strongly relies on the ca-pacity of lipid membranes to bend. The energythat is required for this depends directly on themembrane’s bending rigidity modulus k that inturn is a characteristic of membrane composi-tion (see Box 1) (Helfrich 1973). Typical k val-ues range between 10 and 60 times the thermalenergy kBT for membranes in a fluid state(Marsh 2006). In general, acyl chain unsatura-tion decreases k, whereas cholesterol and acylchain saturation increase membrane rigidity(Rawicz et al. 2000). As an example, the bendingrigidity of a membrane made from “raft lipids”(basically sphingolipids and cholesterol) andphospholipids is of the order of 60 kBT (Rouxet al. 2005), whereas it is of the order of 20–30 kBT for a lipid mixture made of unsaturatedphospholipids and cholesterol (Pan et al. 2009).Lipid bilayers have a flat geometry as long as theareas of both leaflets are equal. When the leafletsare asymmetric (i.e., the number of lipids is dif-ferent or proteins are inserted in an unbalancedfashion), the membrane bends. The free energyper unit area is given by (Helfrich 1973):

Fbending ¼k

2� ðC � C0Þ2; ð1Þ

where C is the imposed membrane curvature,and C0 is the spontaneous membrane curvature(i.e., the relaxed equilibrium curvature of themembrane in the absence of external mechani-cal action) (Box 2).

BOX 1. BENDING MODULUS

This parameter is a mechanical characteristic of the membrane and represents the energy requiredto bend it. The bending modulus depends on lipid composition. In general, membranes madefrom lipids with unsaturated acyl chains have a lower bending modulus than membranes madefrom lipids with saturated acyl chains, and cholesterol increases the bending modulus. For fluidmembranes, the bending modulus ranges between 10 and 100 kT (k is the Boltzmann constantand T is the absolute temperature, kT = 4 � 10221 N/m at room temperature).

L. Johannes et al.

2 Cite this article as Cold Spring Harb Perspect Biol 2014;6:a016741



Asymmetric Protein Insertion/LipidCompaction

In vivo, different mechanisms can produce anonzero C0. Proteins that insert amphipathicor hydrophobic structures from one side of themembrane into the bilayer can produce mem-brane bending (Fig. 1A) (Antonny 2011). Theresulting effect of wedge insertion depends onprotein density on the membrane, on the depth

of insertion, and also on the protein’s detailedshape, conical versus cylindrical (Campelo et al.2008).

Similar effects are expected for conicaltransmembrane proteins (Aimon et al. 2014)or multimers of membrane proteins forminga structure with a noncylindrical shape (forinstance, reticulon) (Fig. 1A) (Hu et al. 2008).At low protein density, it is expected that the

BOX 2. SPONTANEOUS CURVATURE

Parameter that corresponds to the unstressed curvature of the membrane at equilibrium. Starting froma flat membrane (spontaneous curvature C0 = 0), a nonzero spontaneous curvature can be producedby a number of mechanisms (asymmetric change in membrane area in one of the two leaflets,scaffolding, etc.; see text for details) (Figs. 1 and 2). The spontaneous curvature of a membranedepends on the surface density of curvature-generating proteins. By convention, positive spontane-ous curvature refers to cases in which the membrane invaginates toward the curvature-inducingprotein, and negative curvature to cases in which the membrane invaginates away from the curva-ture-inducing protein.

Helix insertion Conical transmembrane protein

Asymmetric insertion

Asymmetric lipidcompaction

Lipid compactionB

A

Sid

e vi

ewTo

p vi

ew

Stress release

ShigatoxinGb3

Figure 1. Membrane-bending mechanisms based on asymmetric transbilayer stress. (A) Asymmetric helixinsertion into one leaflet, or asymmetric shape of transmembrane proteins. (B) Lipid compaction.

Biophysics of Membrane Bending

Cite this article as Cold Spring Harb Perspect Biol 2014;6:a016741 3



spontaneous membrane curvature C0 is direct-ly proportional to the fractional density ofproteins f bound or inserted in the mem-brane (Leibler 1986; Campelo et al. 2008):C0ðfÞ ¼ C0f, with �C0 being the effective spon-taneous curvature of the protein, a molecularprotein-specific parameter. In the case of trans-membrane proteins, �C0 can be related to aneffective shape of the protein (Aimon et al.2014). The clustering of such proteins wouldlead to the spontaneous formation of a mem-brane bud.

In a biological context, the directionalityof membrane deformation is functionally rele-vant. If a cytosolic protein binds to the cyto-plasmic leaflet and induces membrane bendingtoward the cytosolic side, this curvature is ar-bitrarily called “positive.” In contrast, the cur-

vature induced toward the cytosolic space bya protein bound to the exoplasmic leaflet istermed “negative.” Although many cytosolicproteins drive positive curvature, insertionsinto bilayers can in principle produce any typeof curvature (Campelo et al. 2008), and patho-gens can exploit the induction of negative cur-vature at the plasma membrane to enter into, orbud out of cells (Solon et al. 2005; Romer et al.2007; Welsch et al. 2007; Ewers et al. 2010; Ross-man and Lamb 2011). In the case of toxins thatspecifically bind to glycosphingolipids, it hasbeen suggested that negative curvature mightbe produced by the cone shape geometry ofprotein–lipid complexes, possibly accompa-nied by compaction of the lipid head groups(Fig. 1B) (Romer et al. 2007; Sens et al. 2008).Acyl chain splaying under toxin-bound lipid

P ′

Scaffolding

Concentration

Curvature-sensing BAR domain Curvature-inducing BAR domain

Increased proteindistanceProtein

crowding

CrowdingC

B

A

Membrane-boundprotein

C2

C1

Side viewTop view

Chirality-induced budding

Center of chirality withlipid packing defect

Figure 2. Membrane-bending mechanisms based on (A) chirality, (B) scaffolding, and (C) crowding. See text fordetails. (Panel A from Sarasij et al. 2007.)

L. Johannes et al.




head groups would impose mechanical strainthat is released by bending. The exact moleculardetails still need to be worked out.

Other examples for positive and negativemembrane curvature-inducing proteins thatfunction at the level of endosomes are discussedelsewhere in this collection, including Henneet al. (2013), Burd and Cullen (2014), and Gau-treau et al. (2014).

Chirality

Another often-neglected effect could also ac-count for membrane deformation: the intrinsicasymmetry (chirality) of membrane material,proteins, and/or (glyco)lipids (Fig. 2A). Whensuch chiral molecules are organized in so-calledorientation fields, topological defects emerge,leading to constraints that exist in the flat mem-brane state, and which can be relaxed throughthe formation of buds or tubules (Sarasij et al.2007). For the example of cholera toxin it hasindeed been described that its multivalent bind-ing to the glycosphingolipid GM1 induces theemergence of a new organization of the lipidphase, termed textured phase (Watkins et al.2011). The clustering of glycolipids or glycosyl-phosphatidylinositol (GPI)-anchored proteins,possibly leading to their organization in orien-tation fields, may be favored by their coupling tothe dynamic cortical actin cytoskeleton (Go-swami et al. 2008; Gowrishankar et al. 2012).The mechanisms of this coupling still need tobe worked out, and may involve transmembraneproteins that themselves interact with the actinmachinery via their cytosolic tails.

Scaffolding

Spontaneous membrane curvature can further-more result from the tight binding of proteinswith a rigid and curved backbone onto a mem-brane (Fig. 2B). Local membrane deformationand thus �C0, results then from the adhesion be-tween the membrane and the protein surface, asin a mold. This scaffolding process is relevant,for instance, for the BAR-domain proteinsuperfamily (Frost et al. 2009; Mim and Unger2012) whose backbones consist of dimers with

different intrinsic curvature and orientation,depending on the specific member of the fami-ly. N-BAR proteins contain membrane-insert-ing amphipathic helices, in addition to BARdomains. In this case, both elements can syn-ergistically produce curvature. Furthermore,the curvature induction capacity of a givenBAR-domain protein critically depends on itslocal density (see section on Inducing versusSensing).

Crowding

Recent work showed that even in the absence ofbilayer insertion and scaffolding, membranebudding and tubulation can occur if mem-brane-bound proteins are concentrated in do-mains (Fig. 2C) (Stachowiak et al. 2010, 2012).Here, the driving force is a crowding mecha-nism in which the steric interactions betweenproteins or with the membrane create lateralpressure that results in spontaneous bending.

Coats

Some proteins self-assemble into rigid, curved,shell-like structures forming coats, which,bound to the membrane, impose their curva-ture onto the bilayer. The coat shape depends onthe shape of the individual components and onthe lateral interactions; it can be spherical as inthe case of clathrin, COPI or COPII (McMahonand Mills 2004; Stagg et al. 2007), or tubular-likefor dynamin (Schmid and Frolov 2011; Fergu-son and De Camilli 2012), and F-BAR-domainproteins (e.g., FBP17 or FCHo in human cells)(Masuda and Mochizuki 2010). The typical ra-dii for these structures are of the order of 50–120 nm for the spherical coats. For dynamin,the membrane tube radius is 10 nm, whereasthe radii imposed by F-BAR domains are larg-er ranging from typically 30 nm to .50 nm(Henne et al. 2007; Frost et al. 2008). Of note,the use of the coat term in the context of dyna-min and F-BAR-domain-containing proteinsmay appear unconventional. Yet, these proteinsclearly have the capacity to self-assemble ontomembranes to form a rigid structure, and there-fore conform to the definition of coats.





Local Pulling Force

Ultimately, membrane tubes can be extendedwhen applying a local force, the amplitude ofwhich (in the order of up to tens of piconew-tons) depends on the membrane’s bending ri-gidity and its lateral tension (Derenyi et al.2002). Molecular motors bound to the mem-brane (such as kinesins and myosins) and mov-ing along cytoskeleton filaments (microtubulesin the case of kinesins, and actin in the case ofmyosins) have been shown to be able to producesuch forces (Fig. 3) (Roux et al. 2002; Kosteret al. 2003). The tubule radii are quite narrow,usually well below 200 nm, and sometimes aslittle as 10 nm.

Line Tension

We have listed here a series of bending mecha-nisms involving proteins inserted into or boundonto membranes. Early theoretical predictions(see, for instance, Julicher and Lipowsky 1993)and later in vitro experiments using giant lipo-somes (e.g., Baumgart et al. 2003) have shownthat membranes made of pure lipid mixturescan also bend locally, in the absence of proteinsand transbilayer asymmetry, providing that lip-id domains are present (Box 3). The force thatproduces membrane bending in this case is theline tension g at domain interfaces, which re-sults from misalignments (e.g., curvature orheight mismatches) that cause hydrophobicacyl chain segments to be exposed to bulk watermolecules or hydrophilic head groups fromneighboring lipids (Fig. 4A) (Semrau andSchmidt 2009; Johannes and Mayor 2010). Anexample is long acyl chain glycosphingolipids(GSLs) that are clustered by bacterial toxins to

form a domain that is thicker than the bulkmembrane (Romer et al. 2007). Depending onlipid composition and temperature, line tensioncan vary between 0.1 and 10 pN in in vitro sys-tems (Garcia-Saez et al. 2007; Tian et al. 2007).Line tension produces a constriction in the at-tempt to reduce the perimeter of the domainboundary and, thereby, the energy penalty.This effect is counterbalanced by the bendingenergy of the lipid domain and membrane ten-sion that tends to flatten the membrane. Ifmembrane tension is strongly reduced by de-flating the giant liposome, line tension can besufficient to allow the membrane domain tobud and detach from the limiting membrane(Box 4) (Lipowsky 1993; P Bassereau, unpubl.).

We can estimate the smallest size Rmin of thebud that can be formed by this process. Theenergy required to generate a vesicle can be cal-culated by integrating Equation 1 on a sphere,which yields Fbending = 8pk, independently ofthe size of the bud. The line-tension energy isFline ¼ 2pRming. The lowest size corresponds tothe bud formed at zero tension, which can beestimated as,

Rmin �4k

g� 300 nm,

for typical membrane parameters of k � 40 kBTand g � 2 pN (Sens et al. 2008). Larger buds areexpected to form as soon as membrane tensionincreases. This very simplistic estimate impliesthat line tension alone cannot explain the for-mation of narrow curvature radii that charac-terize transport carriers on cells, and that pro-teins are necessary for biological membranebending.

Cytosol

Plasma m

embrane

Pulling

Motors

Microtubule

Local pulling forces

Figure 3. Motor-driven formation of membrane tubules on microtubule tracks.

L. Johannes et al.




Control of the Carrier Size

Most proteins that deform membranes throughthe generation of spontaneous membrane cur-vature or by a crowding mechanism produceelongated membrane tubules and not sphericalbuds. Tubule diameter is in principle relatedto the spontaneous curvature of the protein,but no constraints besides membrane tension

and total protein quantity intrinsically regulatethe length of these tubules. This is also true forcoat-forming proteins polymerizing as cylin-drical structures (i.e., BAR-domain proteins).Yet, there are many cases in which the numberof cargoes per transport carrier has to be strict-ly controlled, such as for synaptic vesicle for-mation, or the internalization of receptors oradhesion molecules. This requirement could

BOX 3. LINE TENSION

Line tension is related to the energy cost at domain interfaces. For instance, if a domain is thicker thanthe surrounding bulk membrane, lipids are misaligned at the domain interface, which leads tounfavorable contacts between hydrophobic acyl chains and hydrophilic head groups from neigh-boring lipids, or water molecules from the solvent (Fig. 4A). Line tension in lipid membranes has beenmeasured with values ranging between 0.1 and 10 pN.

Heightmismatch

B

A

Line-tension-driven clustering

Totalperimeter

Total mechanicalenergy

Total areaof domains

Acyl chainexposureto water Perimeter

reduced

Bending

Line-tension-driven membrane bending

Top

view

Mag

nific

atio

nof

sid

e vi

ew

Figure 4. Line tension. (A) Membrane bending. (B) Membrane-mediated clustering.





in principle be met in two ways: a tight timing ofscission or additional mechanical constraints bywhich spontaneous curvature sets the size of thespherical carrier (its diameter). It is the lattersolution that appears to have been evolved pref-erentially with the emergence of rigid coats thatprovide high fidelity for transport carrier for-mation at the plasma membrane (e.g., synapticvesicles for quantal neurotransmitter release),but also for trafficking between the endoplas-mic reticulum (ER) and Golgi. The ancestraltubular carrier structures remain in use atmany intracellular trafficking interfaces forless tightly normalized functions.

Scission

A lipid membrane can be disrupted upon ap-plication of a high stretching force. Lysis tensiontypically ranges between 1023 and 1022 N/m(Evans and Smith 2011). In biological systems,detachment of the transport carrier from thelimiting membrane occurs in a nonleaky fash-ion, which is achieved, possibly via a hemifis-sion state, by mechanisms leading to the squeez-ing or pinching of the neck of the spherical budor, locally, of a tubule.

Line Tension

The scission of a bud or tubule requires over-coming an energy barrier in the process. In apure lipid mixture, line tension at the edgesof lipid domains can provide a constrictionforce (Fig. 5), as discussed above. It has indeedbeen argued on theoretical and experimentalgrounds that membrane tubules can undergoline-tension-driven scission along the domainedge (Allain et al. 2004; Roux et al. 2005).

Membrane Tension and Pulling Forces

Interestingly, the increase of membrane tensionthat results from a reduced tubule radius con-siderably decreases the time for scission afterphase separation (Allain et al. 2004). In vivo,pulling forces that are applied to the buddingstructure by molecular motors or polymerizingactin filaments are expected to reduce the ener-gy barrier involved in neck or tubule fission (Liuet al. 2006).

Scission Proteins

Dynamin or dynamin-like proteins inducemembrane scission by converting chemical en-ergy of GTP into a squeezing force on the mem-brane neck (Chappie et al. 2011; Campelo andMalhotra 2012; Shnyrova et al. 2013). A recentpaper has shed light on the fission mechanism(Morlot et al. 2012). Twisting of the helical dy-namin structure results in local constriction of

BOX 4. MEMBRANE-MEDIATED CLUSTERING MECHANISM

Proteins can cluster in the absence of direct protein–protein interactions. Such membrane-mediatedclustering can be driven by hydrophobic mismatch in the case of transmembrane proteins (i.e., thethickness of the lipid membrane does not match the height of the hydrophobic part of the transmem-brane protein), and more generally bya reduction of the line tension (Fig. 4B). Membrane fluctuationscan also drive clustering owing to entropic effects.

Line-tension-driven scission

Figure 5. Line-tension-driven scission. The forma-tion of a lipid domain in a membrane tubule leadsto spontaneous constriction at domain interfaces toreduce line tension. Schematically, a light blue do-main (left) is juxtaposed to a dark blue domain(right). (Courtesy of Storm et al. 2004.)

L. Johannes et al.




the membrane under the dynamin coat, whichincreases the local membrane elastic energy atthe edge of the dynamin polymer and signifi-cantly reduces the energy barrier for fission.This situation is in principle similar to line ten-sion at domain interfaces that we have discussedabove. The boundary forces here result from thediscontinuity between the constricted tube coat-ed with dynamin, and the noncoated tube with alarger radius. Scission then takes place at thislocation with a kinetics essentially controlledby the membrane elasticity (bending rigidityand membrane tension). Another set of experi-ments proposes that membrane constriction bydynamin rings is assisted by the tilt of the PHdomain upon GTP hydrolysis, creating a wedgeeffect (Shnyrova et al. 2013). These mechanismsmay work alone or in combination, and mayapply to other proteins or protein assembliesthat use nucleotide hydrolysis to constrict themembrane.

Helix Insertion

Local destabilization of the membrane upon hy-drophobic insertion is an alternative mechanismfor fission. In a recent paper, a combination ofthermodynamic, in vitro and in vivo, approacheswas used to show that increasing the density ofinsertions in the membrane drives membranedeformation from tubulation to fission, inde-pendently of nucleotide hydrolysis (Boucrotet al. 2012). The fact that in N-BAR proteinsthe amphipathic helices are attached to BAR do-mains limits the density of insertions into themembrane and thereby the fission efficiency.

Experimental Systems

As seen above, different mechanisms account formembrane bending and scission. To discrimi-nate stringently between these, in vitro experi-ments with purified proteins and model mem-branes are used. A typical approach consists inincubating small 50- to 100-nm-diameter lipo-somes in the presence of proteins whose poten-tial to bend membranes is under scrutiny. Mem-brane shape changes are monitored by electronmicroscopy (see, for instance, Peter et al. 2004;

Mattila et al. 2007). In the case of organizedcoats, structural data can be deduced (Daninoet al. 2004; Frost et al. 2008). In the presence ofhigh-protein concentrations, the spontaneouscurvature of the ligand can in principle be de-duced from the diameter of the tubules (Takeiet al. 1998).

In some cases, the effect of protein bindingon membrane curvature can be addressed withsmall liposomes and light scattering (Antonny2011) or quantitative high throughput fluores-cence microscopy (Bhatia et al. 2010). However,under these conditions, one only has accessto bulk protein concentrations, which makes itdifficult to compare the data with theoreticalmodels. Giant liposomes (also called GUVs,for “giant unilamellar vesicles”) represent anexperimental system that has a number of ad-vantages over the bulk approaches. GUVs en-able the monitoring of membrane deformationand protein distribution with optical microsco-py, and provide at the same time control overmembrane tension (Sens et al. 2008). Mem-brane nanotubes can be pulled with controlledcurvature radii, and the mechanical action ofproteins on membranes can be deduced fromforce measurements and compared with theo-retical models (Sorre et al. 2012). Protein distri-bution can be tuned as a function of membranecurvature by continuous changing of the tuberadius under concomitant observation by fluo-rescence microscopy (Sorre et al. 2009, 2012;Roux et al. 2010; Singh et al. 2012). Membranefission can be followed in real time and analyzedin detail (Romer et al. 2010; Morlot et al. 2012).Membrane tubes can also be pulled from a pla-nar lipid bilayer using a patch-clamp micro-pipette (Bashkirov et al. 2008), allowing for elec-trophysiology measurements. The constrictiveeffect of proteins can then be monitored viathe conductivity of the nanotube.

Tension-free tubules can be obtained afterprotein binding either onto lipid-coated planarglass substrates, or lipid-coated beads (Rouxet al. 2006; Pucadyil and Schmid 2008). In thesesetups, many tubes can be observed at the sametime.

In this section of the review, we have seenhow various protein modules drive membrane





curvature changes, exploiting a limited set ofphysical principles. In the following section,we discuss some cell biological aspects of endo-cytosis from the perspective of these physicalconsiderations.

CELL BIOLOGY OF ENDOCYTICMEMBRANE BENDING

From Apparently Simple to Complex:Construction of Endocytic Pits

In going from simple to more evolved organ-isms, an apparent complexification of the endo-cytic machinery can be observed. In the clathrinpathway of mammalian cells, several moleculeshave the potential to participate in membranebending, including clathrin itself via a rigid coateffect (Dannhauser and Ungewickell 2012),BAR-domain-containing scaffolding proteins(Henne et al. 2010; Cocucci et al. 2012), andhelix-inserting proteins such as epsins (Fordet al. 2002). In comparison, the endocytic path-way in yeast appears to be simpler. Clathrin isnot strictly required, and BAR-domain proteinscome in at late stages of endocytic membraneinvagination for a likely function in scission thatis mostly driven by actin (Kaksonen et al. 2005).For certain bacterial toxins and animal virusesit has even been argued that initial endocyticmembrane bending does not strictly requirecytosolic machinery and is induced by GSL re-organization (Romer et al. 2007; Ewers et al.2010). In this case, the signal that is sent tothe cytosol appears to be high membrane cur-vature, which would then be recognized by cy-tosolic machinery such as actin (Romer et al.2010) and BAR-domain proteins for furtherprocessing into the cell. Evolutionary complex-ification thereby appears to be achieved by anincreasing number of functional modules thatcontribute to the bending process.

Does multimodality represent redundancy?At this stage, no clear answer can be given. Thedepletion of all three epsins from mammaliancells prevents clathrin-dependent transferrinendocytosis (Boucrot et al. 2012). This blockwas not at the level of membrane bending, butrather for scission. Although this has allowed

identifying an unexpected function of epsins isin the latter reaction (see below), it is still likelythat epsins also contribute to initial bending,but in a redundant manner. Such redundancymight also explain why BAR-domain proteinssuch as FCHos (Henne et al. 2010; Cocucci et al.2012) have variable phenotypes in different ex-perimental systems, and why genetic deletion ofwell-studied BAR-domain proteins such as am-phiphysin does not produce a generalized up-take phenotype in the clathrin pathway (DiPaolo et al. 2002). Redundancy likely improvesthe robustness of the endocytic system, whichgains importance as organisms become multi-cellular and need to integrate the capacity ofsending and receiving precisely defined signals.

The increased complexity of endocytic mem-brane-bending mechanisms results in markedchanges concerning the morphology of endo-cytic carriers. Uptake structures that are gener-ated by clathrin-independent endocytosis arenotoriously pleomorphic (Hansen and Nichols2009), and often have tubular or cisternal shapes(Kirkham et al. 2005; Massol et al. 2005; Romeret al. 2007). In contrast, clathrin-coated vesiclesare of defined sizes between 50 nm (synapticvesicles) and 80–120 nm (Fotin et al. 2004).This normalization in size is achieved at theprice of bringing high amounts of clathrin,adaptors, and accessory proteins onto the mem-brane whose intrinsic self-organization capacityimpinges the preferred spherical shape of thevesicle. In all cases of clathrin- and caveolin-independent endocytosis, highly organizedcoatlike structures could not be detected at sitesof membrane invagination (Howes et al. 2010),which likely explains the pleomorphic nature ofresulting transport carriers. As we have seenabove, the bacterial Shiga toxin (Romer et al.2007), cholera toxin, and the polyoma virusSV40 (Ewers et al. 2010) reorganize GSLs inthe exoplasmic leaflet of the plasma membrane,and thereby drive the clathrin-independentformation of narrow tubular endocytic invagi-nations. It is not yet clear whether similar mech-anisms also operate in the case of clathrin-inde-pendent uptake of endogenous proteins.

The simple-to-complex scheme becomesmore convoluted when looking at actin. Recent

L. Johannes et al.




time-resolved electron tomography studiesshow that, in yeast, actin forces appear to be re-quired for the formation of initial curvature(Kukulski et al. 2012, but see also Idrissi et al.2012), and to push the membrane tubule towardthe interior. The specificity of the yeast systemis that the membrane invagination process hasto overcome high membrane tension owingto turgor pressure. Correspondingly, actin inthe mammalian clathrin pathway appears to bestrictly required only in specific conditions ofhigh membrane tension caused experimentallyby osmotic swelling or mechanical stretching(Boulant et al. 2011). In yeast, an additional el-ement emerges from the fact that the organismis surrounded by a rigid cell wall, like in plants.We can thus hypothesize that the actin cytoskel-eton can exert inward force by pushing onto thiscell wall, in contrast to mammalian cells inwhichthe counteracting force would be minimal ow-ing to the flexibility of the plasma membrane.

In clathrin-independent endocytosis inhigher eukaryotes, actin is clearly a key element(Howes et al. 2010), likely by favoring receptorclustering (Goswami et al. 2008) and scission(Romer et al. 2010). Whether actin also drivesmembrane bending directly remains to be ana-lyzed.

Inducing versus Sensing

An important question that has caused someconfusion is whether BAR-domain proteinsare curvature sensors or inducers. Already inthe pioneering paper on the structure of theBAR domain of amphiphysin, both aspectswere brought up (Peter et al. 2004). Subsequentstudies clearly documented the appearanceof tubular structures upon overexpression ofBAR-domain proteins (reviewed in Frost et al.2009). A recent study has allowed conceptualiz-ing this apparent contradiction (Sorre et al.2012). Using amphiphysin as a model, it wasshown that the protein’s mechanical effects ona membrane tubule depend on protein density(Fig. 2B). At very low densities, curvature sens-ing prevails (i.e., strong enrichment of the pro-tein on highly curved membrane segments),and no mechanical effects are observed. In con-

trast, a scaffold is formed around the tubule atelevated amphiphysin densities, imposing veryhigh mechanical strain onto the membrane.These findings illustrate that one needs to inte-grate information on local protein densities onmembranes when reasoning about the questionof curvature sensing or induction in the cellularcontext.

From Apparently Simple to Complex:Membrane Scission

Biological membrane scission is another exam-ple for which the from-simple-to-complex par-adigm appears to apply. The GTPase dynaminhas received the most attention in this context,owing to its preponderant role in the clathrinpathway. Dynamin clearly is a mechanoenzyme(Doherty and McMahon 2009; Mettlen et al.2009; Chappie et al. 2011). Recent studies haveshown that scission occurs spontaneously at theinterface of the dynamin lattice and the baremembrane once dynamin has reduced the ener-gy barrier sufficiently by constricting the mem-brane tubule or bud neck (Morlot et al. 2012).

Yet dynamin is not the whole scission story,even in the clathrin pathway. Under certain ex-perimental conditions of overexpression, epsinsvia their helix insertion capacity can function-ally replace dynamin in the clathrin-dependentuptake of transferrin (Boucrot et al. 2012). Inthis study, an intricate interplay has been re-vealed in which BAR domains limit the scissionprocess that is driven by membrane-insertinghelices. It remains yet to be determined howthis general finding applies to other endocyticprocesses.

Many of the clathrin-independent endocy-tosis processes are dynamin independent, or atleast not strictly reliant on dynamin. The primeexample is yeast. Here, BAR-domain proteinsare recruited at late stages before scission, andit has been argued on theoretical grounds thata scaffold of BAR domains would create a lipiddomain whose interface forces then drive thespontaneous squeezing of the invaginated tu-bule leading to scission (Fig. 5) (Liu et al.2009). As we have seen above, physicists callsuch domain boundary forces line tension,





and theoretical and experimental studies hadalready linked line tension to scission (Allainet al. 2004; Roux et al. 2005; Liu et al. 2006).The most direct evidence for a function of linetension in biological membrane scission comesfrom the Shiga toxin system. In a study on modeland cell membranes it was shown that Shigatoxin-containing endocytic invaginations arepoised such that an appropriate trigger inducesmembrane reorganization leading to line ten-sion and scission (Romer et al. 2010). In cells,this trigger appears to be actin, whose bindingto lipids can indeed change the phase behaviorof a membrane system (Liu and Fletcher 2006).Whether this type of scission mechanism alsocontributes to the clathrin pathway is a possibil-ity that still needs to be addressed directly.

A Rafty Bending Business

It is now clearly established that the plasmamembrane is compartmentalized. Such com-partmentalization appears to be driven by activeprocesses linked to the actin cytoskeleton (Gow-rishankar et al. 2012). Active membrane do-mains could in principle lead to the spontane-ous formation of membrane buds, owing to thetendency of the system to reduce the line-ten-sion energy penalty that exists at domain inter-faces (Fig. 4A) (Baumgart et al. 2003; Bacia et al.2005). However, the resulting curvature radiiare in the micrometer range, and cannot bythemselves explain the formation of narrow in-vaginations that characterize endocytic pits.

Membrane domains that are generated byactive mechanisms could favor the formationof narrow tubules if the domain material itselfcontributed to bending. There are several wayshow that could occur, including the clusteringof proteins or lipids whose intrinsic shapes (re-viewed in McMahon and Gallop 2005) or ori-entations (Sarasij et al. 2007) tune membranecurvature (Figs. 1 and 2). However, one needs toconsider whether sufficiently stable molecularassemblies can be achieved by purely actin-driv-en clustering. The example of GSLs may be cho-sen to illustrate this point. The reorganization ofcellular GSLs by bacterial Shiga (Romer et al.2007) and cholera (Ewers et al. 2010) toxins

leads to membrane bending and the formationof narrow tubular endocytic membrane invagi-nations. In principle, one might envision thatactin dynamics on the cytosolic leaflet ofthe membrane might cause a similar (glyco)-sphingolipid reorganization to drive membranebending in raft lipid-dependent pathways likethe CLIC/GEEC formation process (reviewedin Chadda et al. 2007). In such a case, onemust assume the existence of tight transbilayercoupling mechanisms that yet need to be de-scribed in further detail. Another possibility isthe spatiotemporal coupling between active ac-tin-driven and passive membrane-mediatedclustering mechanisms (Box 4). In this scenario,an interplay between active contractility, re-modeling of the cytoskeleton, and transbilayercoupling results in transient focusing of passivemolecules (Chaudhuri et al. 2011), such as GSL-toxin complexes. These GSL-toxin complexescould then cluster further by membrane-medi-ated mechanisms favored by line-tension forcesresulting from packaging defaults such as heightand curvature mismatches (Fig. 4A) (Johannesand Mayor 2010). The efficacy of toxin-drivenmembrane bending would thereby be increased,notably at low bulk toxin concentrations, as onewould expect to find in the pathological situa-tion. Whether this active–passive couplingmechanism actually applies to the toxins, and,if so, whether it has some general bearing onclathrin-independent endocytosis, remains tobe addressed directly.

The concerted interplay between active andpassive clustering mechanisms of raft-type ma-terials might also favor the creation of mem-brane environments that are poised to carryout specific biological functions. For example,the sorting of lipids has been found, on theo-retical and experimental grounds, to be mostefficient in membrane systems that are close tophase separation (Sorre et al. 2009; Tian andBaumgart 2009). This is explained by lipid–lip-id interactions that become dominant overmixing entropy as membranes approach lipid-demixing conditions. Of interest, GSLs can becosorted under such conditions (Safouane et al.2010), which may contribute to the generationof specific compositional environments that are

L. Johannes et al.




programmed to carry out downstream func-tions. A striking example is the above-men-tioned actin-driven scission process on Shigatoxin-induced membrane invaginations (Ro-mer et al. 2010). Here, tubular endocytic invag-inations at physiological temperature are veryclose to demixing. An appropriate trigger leadsto domain formation, line-tension-driven tu-bule squeezing, and spontaneous scission. Inother words, by inducing tubular plasma mem-brane invaginations with yet-to-be defined spe-cific compositions, Shiga toxin preprogramsthe next step of its entry into cells, which istubule scission. At first sight, it might appearsurprising to propose that biological membranecompositions with thousands of proteins andlipids could be tuned to position the systemin such states close to demixing. Exactly suchevents have recently been described however(Heimburg and Jackson 2005; Lingwood et al.2008; Polozov et al. 2008). Compositional tun-ing opens unprecedented opportunities to bio-chemically impose, in many physiological con-texts, the concept of induced domain formationas a driver for cell biological functions.

STAGING

Clathrin-mediated endocytosis appears to in-volve many of the functional modules that wehave described so far to achieve budding in avery robust manner. The time sequence of thisprocess is highly orchestrated (Taylor et al. 2011;see also Merrifield and Kaksonen 2014). Nophysical model exists so far to globally describehow these modules could be functionally cou-pled. A simple biochemical view would be thatthe binding sequence is driven by the sequentialrecruitment of protein modules. However, weargue that the coupling between curvature andprotein density on the membranes can producethe time sequence for the recruitment of down-stream module partners. The shaping of themembrane induced by the first deforming mod-ules could be the signal that triggers recruitmentof the next modules. Such a model requires thatthe modules involved in fission should appearlast for correct budding, thus involving proteinseither with a low density on the membrane

(controlled by a reduced affinity for the mem-brane, or a low bulk concentration) or alterna-tively that are poor curvature sensors.

In vitro experiments on amphiphysin anddynamin have pointed toward curvature as avery efficient signal at moderate protein densi-ties (Roux et al. 2010; Sorre et al. 2012). More-over, feedback between different modules canexist as it has been described for dynamin, actin,and N-BAR proteins that can cooperatively workto efficiently catalyze membrane scission (Tayloret al. 2012). Such feedback is even more difficultto include in a physical modeling. Nevertheless,it is urgent to better understand how/why thefirst shaping module is recruited to bend mem-brane. Recent live-cell total internal reflectionfluorescence microscopy (TIRF) imaging exper-iments with single-molecule sensitivity on cla-thrin-coated pits of mammalian cells have re-vealed that recruitment of two AP2 adaptors tothe PIP2-containing membrane, stabilized byone clathrin triskelion with oligomerization ca-pacity, can provide the first brick for further coatrecruitment and subsequent bending (Cocucciet al. 2012). In yeast, a different sequence occursas shown byexperiments correlating 3D electronmicroscopy and fluorescence microscopy (Ku-kulski et al. 2011, 2012). Clathrin binding doesnot induce bending, which instead is producedby the actin network. However, immunogoldlabeling with high-resolution transmission elec-tron microscopy (EM) have challenged this viewand suggested that the branched actin network isinvolved in the elongation of an existing invag-ination and not in the initial bending (Idrissiet al. 2012). This difference may have arisenfrom different sensitivities of the methods thatwere used in both studies to detect initial mem-brane deformation.

CONCLUDING REMARKS AND FUTUREPERSPECTIVES

Classification of endocytic pathways has been achallenge ever since the discovery of clathrin-independent uptake processes. In this review,we have argued that a limited set of physicalprinciples can be used to group biochemical ma-chinery into functional modules. We anticipate





that individual modules will be shared betweendifferent endocytic events, and that specificitywithin a given uptake process arises from de-fined cocktails of modules. Based on a true un-derstanding of how these modules function, itshould become possible to assign them to spe-cific uptake processes, and more precisely to spe-cific steps within the uptake process (initiation,bending, invagination . . .). Such classificationof endocytic events would thus consist of a chainof defined module identity, rather than a reduc-tionist limitation to one functional element.

The key aspect here is on mechanistic under-standing in terms of physical/mechanical prin-ciples. We believe that this must come from acombination of approaches that range frommodel membrane reconstitution with precisecontrol over experimental conditions and accessto parameters such as forces that are difficultto measure in complex systems, to systems biol-ogyexplorations of integrated circuits. The strivefor mechanistic understanding calls for inter-disciplinary expertise from physics for conceptsand workstations for quantitative measure-ments, from chemistry for tailor-made tools,and from biology for relevant working hypoth-eses and their exploration into the physiologicaland pathological context of cells and organisms.

The back-assignment of mechanistic con-clusions from in vitro explorations to the celland organism situation is of critical importance.In vitro experiments should be strictly quantita-tive, consistent between experimental condi-tions, and referring to the invivo concentrations.To illustrate this point, one may recall that BAR-domain proteins like amphiphysin sense or in-duce curvature, depending on their concentra-tion on membranes (Sorre et al. 2012). For theassignment of their module identity within agiven endocytic program, it is thus of critical im-portance to have access to concentration valueswithin the context of the cell. The advent ofquantitative imaging approaches holds out thepromise of accessing this type of information inboth a dynamic and highly localized manner.

The assignment of module identity withinendocytic pathways is rendered complex by ad-ditional considerations. A functional interplaybetween modules is one of these. As we have

seen, actin (via line tension) (Stechmann et al.2010) and dynamin (as a mechanoenzyme)(Chappie et al. 2011) contribute to scission intheir own right. They possibly also interactfunctionally if it is confirmed in vivo that dyna-min activity is enhanced on tensed membranes(Roux et al. 2006; Ferguson et al. 2009). Anothersource for complexity comes from the mechan-ical state of the membrane substrate itself. Forexample, membrane tension is buffered by cav-eolae (Sinha et al. 2011), and influences endo-cytic processes (Gauthier et al. 2012). As cellbiologists dig their way deeper into these com-plexities, they increasingly meet and team upwith physicists and chemists who have becomeinterested in biological materials as fertile sub-strates for their own scientific imagination.

ACKNOWLEDGMENTS

The work in the authors’ laboratories on sub-jects related to the current review is supportedby Agence Nationale pour la Recherche (ANR-09-BLAN-283, ANR-11 BSV2 018 03, ANR-11BSV2 014 03) and Institut National du Cancer(PLBIO11-022-IDF-JOHANNES). The Basser-eau and Johannes teams are members of Lab-ex CelTisPhyBio. A Marie Curie InternationalReintegration Grant within the Seventh Euro-pean Union Framework Programme (FP7-RG-277078) supports C.W.

REFERENCES�Reference is also in this collection.

Aimon S, Callan-Jones A, Berthaud A, Pinot M, ToombesGE, Bassereau P. 2014. Membrane shape modulates trans-membrane protein distribution. Dev Cell (to be pub-lished).

Allain JM, Storm C, Roux A, Ben Amar M, Joanny JF. 2004.Fission of a multiphase membrane tube. Phys Rev Lett93: 158104.

Antonny B. 2011. Mechanisms of membrane curvaturesensing. Annu Rev Biochem 80: 101–123.

Bacia K, Schwille P, Kurzchalia TV. 2005. Sterol structuredetermines the separation of phases and the curvatureof the liquid-ordered phase in model membranes. ProcNatl Acad Sci 102: 3272–3277.

Bashkirov PV, Akimov SA, Evseev AI, Schmid SL, Zimmer-berg J, Frolov VA. 2008. GTPase cycle of dynamin iscoupled to membrane squeeze and release, leading tospontaneous fission. Cell 135: 1276–1286.

L. Johannes et al.




Baumgart T, Hess ST, Webb WW. 2003. Imaging coexistingfluid domains in biomembrane models coupling curva-ture and line tension. Nature 425: 821–824.

Bhatia VK, Hatzakis NS, Stamou D. 2010. A unifyingmechanism accounts for sensing of membrane curva-ture by BAR domains, amphipathic helices and mem-brane-anchored proteins. Sem Cell Dev Biol 21: 381–390.

Boucrot E, Pick A, Camdere G, Liska N, Evergren E, McMa-hon HT, Kozlov MM. 2012. Membrane fission is promot-ed by insertion of amphipathic helices and is restricted bycrescent BAR domains. Cell 149: 124–136.

Boulant S, Kural C, Zeeh JC, Ubelmann F, Kirchhausen T.2011. Actin dynamics counteract membrane tension dur-ing clathrin-mediated endocytosis. Nat Cell Biol 13:1124–1131.

� Burd C, Cullen PJ. 2014. Retromer: A master conductor ofendosome sorting. Cold Spring Harb Perspect Biol doi:10.1101/cshperspect.a016774.

Campelo F, Malhotra V. 2012. Membrane fission: The bio-genesis of transport carriers. Ann Rev Biochem 81: 407–427.

Campelo F, McMahon HT, Kozlov MM. 2008. The hydro-phobic insertion mechanism of membrane curvaturegeneration by proteins. Biophys J 95: 2325–2339.

Chadda R, Howes MT, Plowman SJ, Hancock JF, Parton RG,Mayor S. 2007. Cholesterol-sensitive Cdc42 activationregulates actin polymerization for endocytosis via theGEEC pathway. Traffic 8: 702–717.

Chappie JS, Mears JA, Fang S, Leonard M, Schmid SL, Milli-gan RA, Hinshaw JE, Dyda F. 2011. A pseudoatomicmodel of the dynamin polymer identifies a hydrolysis-dependent powerstroke. Cell 147: 209–222.

Chaudhuri A, Bhattacharya B, Gowrishankar K, Mayor S,Rao M. 2011. Spatiotemporal regulation of chemical re-actions by active cytoskeletal remodeling. Proc Natl AcadSci 108: 14825–14830.

Cocucci E, Aguet F, Boulant S, Kirchhausen T. 2012. The firstfive seconds in the life of a clathrin-coated pit. Cell 150:495–507.

Danino D, Moon KH, Hinshaw JE. 2004. Rapid constrictionof lipid bilayers by the mechanochemical enzyme dyna-min. J Struct Biol 147: 259–267.

Dannhauser PN, Ungewickell EJ. 2012. Reconstitution ofclathrin-coated bud and vesicle formation with minimalcomponents. Nat Cell Biol 14: 634–639.

Derenyi I, Julicher F, Prost J. 2002. Formation and interac-tion of membrane tubes. Phys Rev Lett 88: 238101.

Di Paolo G, Sankaranarayanan S, Wenk MR, Daniell L, Pe-rucco E, Caldarone BJ, Flavell R, Picciotto MR, Ryan TA,Cremona O, et al. 2002. Decreased synaptic vesicle recy-cling efficiency and cognitive deficits in amphiphysin 1knockout mice. Neuron 33: 789–804.

Doherty GJ, McMahon HT. 2009. Mechanisms of endocy-tosis. Annu Rev Biochem 78: 857–902.

Evans E, Smith BA. 2011. Kinetics of hole nucleation inbiomembrane rupture. New J Phys 13: 095010.

Ewers H, Romer W, Smith AE, Bacia K, Dmitrieff S, Chai W,Mancini R, Kartenbeck J, Chambon V, Berland L, et al.2010. GM1 structure determines SV40-induced mem-

brane invagination and infection. Nat Cell Biol 12: 11–18.

Ferguson SM, De Camilli P. 2012. Dynamin, a membrane-remodelling GTPase. Nat Rev Mol Cell Biol 13: 75–88.

Ferguson S, Raimondi A, Paradise S, Shen H, Mesaki K,Ferguson A, Destaing O, Ko G, Takasaki J, Cremona O,et al. 2009. Coordinated actions of actin and BAR pro-teins upstream of dynamin at endocytic clathrin-coatedpits. Dev Cell 17: 811–822.

Ford MG, Mills IG, Peter BJ, Vallis Y, Praefcke GJ, Evans PR,McMahon HT. 2002. Curvature of clathrin-coated pitsdriven by epsin. Nature 419: 361–366.

Fotin A, Cheng Y, Sliz P, Grigorieff N, Harrison SC, Kirch-hausen T, Walz T. 2004. Molecular model for a completeclathrin lattice from electron cryomicroscopy. Nature432: 573–579.

Frost A, Perera R, Roux A, Spasov K, Destaing O, EgelmanEH, De Camilli P, Unger VM. 2008. Structural basis ofmembrane invagination by F-BAR domains. Cell 132:807–817.

Frost A, Unger VM, De Camilli P. 2009. The BAR domainsuperfamily: Membrane-molding macromolecules. Cell137: 191–196.

Garcia-Saez AJ, Chiantia S, Schwille P. 2007. Effect of linetension on the lateral organization of lipid membranes. JBiol Chem 282: 33537–33544.

Gauthier NC, Masters TA, Sheetz MP. 2012. Mechanicalfeedback between membrane tension and dynamics.Trends Cell Biol 22: 527–535.

� Gautreau A, Oguievetskaia K, Ungermann C. 2014. Func-tion and regulation of the endosomal fusion and fissionmachineries. Cold Spring Harb Perspect Biol doi: 10.1101/cshperspect.a016832.

Gonnord P, Blouin CM, Lamaze C. 2012. Membrane traf-ficking and signaling: Two sides of the same coin. SeminCell Dev Biol 23: 154–164.

Goswami D, Gowrishankar K, Bilgrami S, Ghosh S, Raghu-pathy R, Chadda R, Vishwakarma R, Rao M, Mayor S.2008. Nanoclusters of GPI-anchored proteins are formedby cortical actin-driven activity. Cell 135: 1085–1097.

Gowrishankar K, Ghosh S, Saha SCR, Mayor S, Rao M.2012. Active remodeling of cortical actin regulates spa-tiotemporal organization of cell surface molecules. Cell149: 1353–1367.

Hansen CG, Nichols BJ. 2009. Molecular mechanisms of cla-thrin-independent endocytosis. J Cell Sci 122: 1713–1721.

Heimburg T, Jackson AD. 2005. On soliton propagation inbiomembranes and nerves. Proc Natl Acad Sci 102: 9790–9795.

Helfrich W. 1973. Elastic properties of lipid bilayers: Theoryand possible experiments. Zur Naturforschung 28c: 693–703.

Henne WM, Kent HM, Ford MGJ, Hegde BG, Daumke O,Butler PJG, Mittal R, Langen R, Evans PR, McMahon HT.2007. Structure and analysis of FCHo2 F-BAR domain: Adimerizing and membrane recruitment module that af-fects membrane curvature. Structure 15: 839–852.

Henne WM, Boucrot E, Meinecke M, Evergren E, Vallis Y,Mittal R, McMahon HT. 2010. FCHo proteins are nucle-ators of clathrin-mediated endocytosis. Science 328:1281–1284.





� Henne WM, Stenmark H, Emr SD. 2013. Molecular mech-anisms of the membrane sculpting ESCRT pathway. ColdSpring Harb Perspect Biol 5: a016766.

Howes MT, Mayor S, Parton RG. 2010. Molecules, mecha-nisms, and cellular roles of clathrin-independent endo-cytosis. Curr Opin Cell Biol 22: 519–527.

Hu J, Shibata Y, Voss C, Shemesh T, Li Z, Coughlin M,Kozlov MM, Rapoport TA, Prinz WA. 2008. Membraneproteins of the endoplasmic reticulum induce high-cur-vature tubules. Science 319: 1247–1250.

Idrissi FZ, Blasco A, Espinal A, Geli MI. 2012. Ultrastruc-tural dynamics of proteins involved in endocytic bud-ding. Proc Natl Acad Sci 109: E2587–E2594.

Johannes L, Mayor S. 2010. Induced domain formation inendocytic invagination, lipid sorting, and scission. Cell142: 507–510.

Julicher F, Lipowsky R. 1993. Domain induced budding ofvesicles. Phys Rev Lett 70: 2964–2967.

Kaksonen M, Toret CP, Drubin DG. 2005. A modular designfor the clathrin- and actin-mediated endocytosis ma-chinery. Cell 123: 305–320.

Kaksonen M, Toret CP, Drubin DG. 2006. Harnessing actindynamics for clathrin-mediated endocytosis. Nat RevMol Cell Biol 7: 404–414.

Kirkham M, Fujita A, Chadda R, Nixon SJ, Kurzchalia TV,Sharma DK, Pagano RE, Hancock JF, Mayor S, PartonRG. 2005. Ultrastructural identification of uncoated cav-eolin-independent early endocytic vehicles. J Cell Biol168: 465–476.

Koster G, VanDuijn M, Hofs B, Dogterom M. 2003. Mem-brane tube formation from giant vesicles by dynamicassociation of motor proteins. Proc Natl Acad Sci 100:15583–15588.

Kukulski W, Schorb M, Welsch S, Picco A, Kaksonen M,Briggs JAG. 2011. Correlated fluorescence and 3D elec-tron microscopy with high sensitivity and spatial preci-sion. J Cell Biol 192: 111–119.

Kukulski W, Schorb M, Kaksonen M, Briggs JA. 2012. Plasmamembrane reshaping during endocytosis is revealed bytime-resolved electron tomography. Cell 150: 508–520.

Kumari S, Mg S, Mayor S. 2010. Endocytosis unplugged:Multiple ways to enter the cell. Cell Res 20: 256–275.

Leibler S. 1986. Curvature instability in membranes. J Phys I47: 507–516.

Lingwood D, Ries J, Schwille P, Simons K. 2008. Plasmamembranes are poised for activation of raft phase coales-cence at physiological temperature. Proc Natl Acad Sci105: 10005–10010.

Lipowsky R. 1993. Domain-induced budding of fluid mem-branes. Biophys J 64: 1133–1138.

Liu AP, Fletcher DA. 2006. Actin polymerization serves as amembrane domain switch in model lipid bilayers. Bio-phys J 91: 4064–4070.

Liu J, Kaksonen M, Drubin DG, Oster G. 2006. Endocyticvesicle scission by lipid phase boundary forces. Proc NatlAcad Sci 103: 10277–10282.

Liu J, Sun Y, Drubin DG, Oster GF. 2009. The mechano-chemistry of endocytosis. PLoS Biol 7: e1000204.

Marsh D. 2006. Elastic curvature constants of lipid mono-layers and bilayers. Chem Phys Lipids 144: 146–159.

Massol RH, Larsen JE, Kirchhausen T. 2005. Possible role ofdeep tubular invaginations of the plasma membrane inMHC-I trafficking. Exp Cell Res 306: 142–149.

Masuda M, Mochizuki N. 2010. Structural characteristics ofBAR domain superfamily to sculpt the membrane. SemCell Dev Biol 21: 391–398.

Mattila PK, Pykalainen A, Saarikangas J, Paavilainen VO,Vihinen H, Jokitalo E, Lappalainen P. 2007. Missing-in-metastasis and IRSp53 deform PI(4,5)P2-rich mem-branes by an inverse BAR domain-like mechanism. JCell Biol 176: 953–964.

McMahon HT, Boucrot E. 2011. Molecular mechanism andphysiological functions of clathrin-mediated endocyto-sis. Nat Rev Mol Cell Biol 12: 517–533.

McMahon HT, Gallop JL. 2005. Membrane curvature andmechanisms of dynamic cell membrane remodelling.Nature 438: 590–596.

McMahon HT, Mills IG. 2004. COP and clathrin-coatedvesicle budding: Different pathways, common approach-es. Curr Opin Cell Biol 16: 379–391.

Merrifield CJ. 2004. Seeing is believing: Imaging actin dy-namics at single sites of endocytosis. Trends Cell Biol 14:352–358.

� Merrifield C, Kaksonen M. 2014. Endocytic accessory fac-tors and regulation of clathrin-mediated endocytosis.Cold Spring Harb Perspect Biol doi: 10.1101/cshper-spect.a016733.

Mettlen M, Pucadyil T, Ramachandran R, Schmid SL. 2009.Dissecting dynamin’s role in clathrin-mediated endocy-tosis. Biochem Soc Trans 37: 1022–1026.

Mim C, Unger VM. 2012. Membrane curvature and its gen-eration by BAR proteins. Trends Biochem Sci 37: 526–533.

Morlot S, Galli V, Klein M, Chiaruttini N, Manzi J, HumbertF, Dinis L, Lenz M, Cappello G, Roux A. 2012. Membraneshape at the edge of the dynamin helix sets location andduration of the fission reaction. Cell 151: 619–629.

Pan J, Tristram-Nagle S, Nagle JF. 2009. Effect of cholesterolon structural and mechanical properties of membranesdepends on lipid chain saturation. Phys Rev E 80: 021931.

Peter BJ, Kent HM, Mills IG, Vallis Y, Butler PJ, Evans PR,McMahon HT. 2004. BAR domains as sensors of mem-brane curvature: The amphiphysin BAR structure. Sci-ence 303: 495–499.

Polozov IV, Bezrukov L, Gawrisch K, Zimmerberg J. 2008.Progressive ordering with decreasing temperature of thephospholipids of influenza virus. Nat Chem Biol 4: 248–255.

Pucadyil TJ, Schmid SL. 2008. Real-time visualization ofdynamin-catalyzed membrane fission and vesicle release.Cell 135: 1263–1275.

Pucadyil TJ, Schmid SL. 2009. Conserved functions ofmembrane active GTPases in coated vesicle formation.Science 325: 1217–1220.

Rawicz W, Olbrich KC, McIntosh T, Needham D, Evans E.2000. Effect of chain length and unsaturation on elasticityof lipid bilayers. Biophys J 79: 328–339.

Romer W, Berland L, Chambon V, Gaus K, Windschiegl B,Tenza D, Aly MR, Fraisier V, Florent J-C, Perrais D, et al.2007. Shiga toxin induces tubular membrane invagina-tions for its uptake into cells. Nature 450: 670–675.

L. Johannes et al.




Romer W, Pontani LL, Sorre B, Rentero C, Berland L, Cham-bon V, Lamaze C, Bassereau P, Sykes C, Gaus K, et al.2010. Actin dynamics drive membrane reorganizationand scission in clathrin-independent endocytosis. Cell140: 540–553.

Rossman JS, Lamb RA. 2011. Influenza virus assembly andbudding. Virology 411: 229–236.

Roux A, Cappello G, Cartaud J, Prost J, Goud B, Bassereau P.2002. A minimal system allowing tubulation with molec-ular motors pulling on giant liposomes. Proc Natl AcadSci 99: 5394–5399.

Roux A, Cuvelier D, Nassoy P, Prost J, Bassereau P, Goud B.2005. Role of curvature and phase transition in lipidsorting and fission of membrane tubules. EMBO J 24:1537–1545.

Roux A, Uyhazi K, Frost A, De Camilli P. 2006. GTP-depen-dent twisting of dynamin implicates constriction andtension in membrane fission. Nature 441: 528–531.

Roux A, Koster G, Lenz M, Sorre B, Manneville J-B, NassoyP, Bassereau P. 2010. Membrane curvature controls dyna-min polymerization. Proc Natl Acad Sci 107: 4141–4146.

Safouane M, Berland L, Callan-Jones A, Sorre B, Romer W,Johannes L, Toombes GE, Bassereau P. 2010. Lipid co-sorting mediated by Shiga toxin induced tubulation.Traffic 11: 1519–1529.

Sarasij RC, Mayor S, Rao M. 2007. Chirality induced bud-ding: A raft-mediated mechanism for endocytosis andmorphology of caveolae? Biophys J 92: 3140–3158.

Schmid SL, Frolov VA. 2011. Dynamin: Functional design ofa membrane fission catalyst. Ann Rev Cell Dev Biol 27:79–105.

Semrau S, Schmidt T. 2009. Membrane heterogeneity—From lipid domains to curvature effects. Soft Matter 5:3174–3186.

Sens P, Johannes L, Bassereau P. 2008. Biophysical view onprotein-induced membrane deformations in endocyto-sis. Curr Opin Cell Biol 20: 476–482.

Shnyrova AV, Bashkirov PV, Akimov SA, Pucadyil TJ, Zim-merberg J, Schmid SL, Frolov VA. 2013. Geometric catal-ysis of membrane fission driven by flexible dynaminrings. Science 339: 1433–1436.

Singh P, Mahata P, Baumgart T, Das SL. 2012. Curvaturesorting of proteins on a cylindrical lipid membrane tetherconnected to a reservoir. Phys Rev E 85: 051906.

Sinha B, Koster D, Ruez R, Gonnord P, Bastiani M, AbankwaD, Stan RV, Butler-Browne G, Vedie B, Johannes L, et al.2011. Cells respond to mechanical stress by rapid disas-sembly of caveolae. Cell 144: 402–413.

Solon J, Gareil O, Bassereau P, Gaudin Y. 2005. Membranedeformations induced by the matrix protein of vesicularstomatitis virus in a minimal system. J Gen Virol 86:3357–3363.

Sorre B, Callan-Jones A, Manneville JB, Nassoy P, Joanny JF,Prost J, Goud B, Bassereau P. 2009. Curvature-driven lipidsorting needs proximity to a demixing point and is aidedby proteins. Proc Natl Acad Sci 106: 5622–5626.

Sorre B, Callan-Jones A, Manzi J, Goud B, Prost J, BassereauP, Roux A. 2012. Nature of curvature coupling of amphi-physin with membranes depends on its bound density.Proc Natl Acad Sci 109: 173–178.

Stachowiak JC, Hayden CC, Sasaki DY. 2010. Steric confine-ment of proteins on lipid membranes can drive curvatureand tubulation. Proc Natl Acad Sci 107: 7781–7786.

Stachowiak JC, Schmid EM, Ryan CJ, Ann HS, Sasaki DY,Sherman MB, Geissler PL, Fletcher DA, Hayden CC.2012. Membrane bending by protein–protein crowding.Nat Cell Biol 14: 944–949.

Stagg SM, LaPointe P, Balch WE. 2007. Structural design ofcage and coat scaffolds that direct membrane traffic. CurrOpin Struct Biol 17: 221–228.

Stechmann B, Bai SK, Gobbo E, Lopez R, Merer G, PinchardS, Panigai L, Tenza D, Raposo G, Beaumelle B, et al. 2010.Inhibition of retrograde transport protects mice fromlethal ricin challenges. Cell 141: 231–242.

Takei K, Haucke V, Slepnev V, Farsad K, Salazar M, Chen H,De Camilli P. 1998. Generation of coated intermediates ofclathrin-mediated endocytosis on protein-free lipo-somes. Cell 94: 131–141.

Taylor MJ, Perrais D, Merrifield CJ. 2011. A high precisionsurvey of the molecular dynamics of mammalian clathrinmediated endocytosis. PloS Biol 9: e1000604.

Taylor MJ, Lampe M, Merrifield CJ. 2012. A feedback loopbetween dynamin and actin recruitment during clathrin-mediated endocytosis. PloS Biol 10: e1001302.

Tian A, Baumgart T. 2009. Sorting of lipids and proteins inmembrane curvature gradients. Biophys J 96: 2676–2688.

Tian A, Johnson C, Wang W, Baumgart T. 2007. Line tensionat fluid membrane domain boundaries measured by mi-cropipette aspiration. Phys Rev Lett 98: 208102–208104.

Watkins EB, Miller CE, Majewski J, Kuhl TL. 2011. Mem-brane texture induced by specific protein binding andreceptor clustering: Active roles for lipids in cellular func-tion. Proc Natl Acad Sci 108: 6975–6980.

Welsch S, Muller B, Krausslich H-G. 2007. More than onedoor—Budding of enveloped viruses through cellularmembranes. FEBS Lett 581: 2089–2097.





2014; doi: 10.1101/cshperspect.a016741Cold Spring Harb Perspect Biol Ludger Johannes, Christian Wunder and Patricia Bassereau Endocytic Pathways

A Cocktail of Biophysical Modules to Build−−Bending ''On the Rocks''

Subject Collection Endocytosis

Endocytosis: Past, Present, and Future

ZerialSandra L. Schmid, Alexander Sorkin and Marino Clathrin-Mediated Endocytosis

Imaging and Modeling the Dynamics of

Marcel Mettlen and Gaudenz Danuser

Endosomal SystemRab Proteins and the Compartmentalization of the

Angela Wandinger-Ness and Marino ZerialClathrin-Mediated EndocytosisEndocytic Accessory Factors and Regulation of

Christien J. Merrifield and Marko Kaksonen

Regulator of Cell Polarity and Tissue DynamicsCargo Sorting in the Endocytic Pathway: A Key

Suzanne Eaton and Fernando Martin-BelmonteSystemThe Complex Ultrastructure of the Endolysosomal

Judith Klumperman and Graça Raposo

Links to Human DiseaseCytoskeleton, Cell Cycle, Nucleus, and Beyond:and Other Endocytic Regulators in the Unconventional Functions for Clathrin, ESCRTs,

et al.Frances M. Brodsky, R. Thomas Sosa, Joel A. Ybe,

Lysosome-Related OrganellesThe Biogenesis of Lysosomes and

Dieckmann, et al.J. Paul Luzio, Yvonne Hackmann, Nele M.G.

Endocytosis of Viruses and BacteriaPascale Cossart and Ari Helenius

Endocytosis, Signaling, and BeyondPier Paolo Di Fiore and Mark von Zastrow

Responds to External CuesLysosomal Adaptation: How the Lysosome

Carmine Settembre and Andrea Ballabio

Clathrin-Independent Pathways of Endocytosis

DonaldsonSatyajit Mayor, Robert G. Parton and Julie G.

MetabolismReciprocal Regulation of Endocytosis and

Amira KlipCostin N. Antonescu, Timothy E. McGraw and

SignalingThe Role of Endocytosis during Morphogenetic

Marcos Gonzalez-Gaitan and Frank Jülicher

Cooperation?Endocytosis and Autophagy: Exploitation or

Sharon A. Tooze, Adi Abada and Zvulun ElazarDiseaseRole of Endosomes and Lysosomes in Human

Frederick R. Maxfield

http://cshperspectives.cshlp.org/cgi/collection/ For additional articles in this collection, see

Copyright © 2014 Cold Spring Harbor Laboratory Press; all rights reserved


http://cshperspectives.cshlp.org/cgi/collection/


http://cshperspectives.cshlp.org/cgi/adclick/?ad=55917&adclick=true&url=http%3A%2F%2Fwww.genelink.com


http://cshperspectives.cshlp.org/cgi/collection/ For additional articles in this collection, see

Copyright © 2014 Cold Spring Harbor Laboratory Press; all rights reserved



http://cshperspectives.cshlp.org/cgi/adclick/?ad=55917&adclick=true&url=http%3A%2F%2Fwww.genelink.com


Bending “On the Rocks”—ACocktail of Biophysical Modules to...

Documents

Transcript of Bending “On the Rocks”—ACocktail of Biophysical Modules to...