Dissecting the mechanism of atlastin-mediated homotypic membrane fusion at the single-molecule level.

Homotypic membrane fusion of the endoplasmic reticulum (ER) is mediated by dynamin-like GTPase atlastin (ATL). This fundamental process relies on GTP-dependent domain rearrangements in the N-terminal region of ATL (ATLcyto), including the GTPase domain and three-helix bundle (3HB). However, its conformational dynamics during the GTPase cycle remain elusive. Here, we combine single-molecule FRET imaging and molecular dynamics simulations to address this conundrum. Different from the prevailing model, ATLcyto can form a loose crossover dimer upon GTP binding, which is tightened by GTP hydrolysis for membrane fusion. Furthermore, the α-helical motif between the 3HB and transmembrane domain, which is embedded in the surface of the lipid bilayer and self-associates in the crossover dimer, is required for ATL function. To recycle the proteins, Pi release, which disassembles the dimer, activates frequent relative movements between the GTPase domain and 3HB, and subsequent GDP dissociation alters the conformational preference of the ATLcyto monomer for entering the next reaction cycle. Finally, we found that two disease-causing mutations affect human ATL1 activity by destabilizing GTP binding-induced loose crossover dimer formation and the membrane-embedded helix, respectively. These results provide insights into ATL-mediated homotypic membrane fusion and the pathological mechanisms of related disease.

Myomerger Induces Membrane Hemifusion and Regulates Fusion Pore Expansion.

Membrane fusion is a crucial mechanism in a wide variety of important events in cell biology from viral infection to exocytosis. However, despite many efforts and much progress, cell-cell fusion has remained elusive to our understanding. Along the life of the fusion pore, large conformational changes take place from the initial lipid bilayer bending, passing through the hemifusion intermediates, and ending with the formation of the first nascent fusion pore. In this sense, computer simulations are an ideal technique for describing such complex lipid remodeling at the molecular level. In this work, we studied the role played by the muscle-specific membrane protein Myomerger during the formation of the fusion pore. We have conducted µs length atomistic and coarse-grained molecular dynamics, together with free-energy calculations using ad hoc collective variables. Our results show that Myomerger favors the hemifusion diaphragm-stalk transition, reduces the nucleation-expansion energy difference, and promotes the formation of nonenlarging fusion pores.

Munc18c accelerates SNARE-dependent membrane fusion in the presence of regulatory proteins α-SNAP and NSF.

Intracellular vesicle fusion is driven by the soluble N-ethylmaleimide-sensitive factor attachment protein receptors (SNAREs) and their cofactors, including Sec1/Munc18 (SM), α-SNAP, and NSF. α-SNAP and NSF play multiple layers of regulatory roles in the SNARE assembly, disassembling the cis-SNARE complex and the prefusion SNARE complex. How SM proteins coupled with NSF and α-SNAP regulate SNARE-dependent membrane fusion remains incompletely understood. Munc18c, an SM protein involved in the exocytosis of the glucose transporter GLUT4, binds and activates target (t-) SNAREs to accelerate the fusion reaction through a SNARE-like peptide (SLP). Here, using an in vitro reconstituted system, we discovered that α-SNAP blocks the GLUT4 SNAREs-mediated membrane fusion. Munc18c interacts with t-SNAREs to displace α-SNAP, which overcomes the fusion inhibition. Furthermore, Munc18c shields the trans-SNARE complex from NSF/α-SNAP-mediated disassembly and accelerates SNARE-dependent fusion kinetics in the presence of NSF and α-SNAP. The SLP in domain 3a is indispensable in Munc18c-assisted resistance to NSF and α-SNAP. Together, our findings demonstrate that Munc18c protects the prefusion SNARE complex from α-SNAP and NSF, promoting SNARE-dependent membrane fusion through its SLP.

The Ykt6-Snap29-Syx13 SNARE complex promotes crinophagy via secretory granule fusion with Lamp1 carrier vesicles.

In the Drosophila larval salivary gland, developmentally programmed fusions between lysosomes and secretory granules (SGs) and their subsequent acidification promote the maturation of SGs that are secreted shortly before puparium formation. Subsequently, ongoing fusions between non-secreted SGs and lysosomes give rise to degradative crinosomes, where the superfluous secretory material is degraded. Lysosomal fusions control both the quality and quantity of SGs, however, its molecular mechanism is incompletely characterized. Here we identify the R-SNARE Ykt6 as a novel regulator of crinosome formation, but not the acidification of maturing SGs. We show that Ykt6 localizes to Lamp1+ carrier vesicles, and forms a SNARE complex with Syntaxin 13 and Snap29 to mediate fusion with SGs. These Lamp1 carriers represent a distinct vesicle population that are functionally different from canonical Arl8+, Cathepsin L+ lysosomes, which also fuse with maturing SGs but are controlled by another SNARE complex composed of Syntaxin 13, Snap29 and Vamp7. Ykt6- and Vamp7-mediated vesicle fusions also determine the fate of SGs, as loss of either of these SNAREs prevents crinosomes from acquiring endosomal PI3P. Our results highlight that fusion events between SGs and different lysosome-related vesicle populations are critical for fine regulation of the maturation and crinophagic degradation of SGs.

The STX17-SNAP47-VAMP7/VAMP8 complex is the default SNARE complex mediating autophagosome-lysosome fusion.

Autophagosome-lysosome fusion mediated by SNARE complexes is an essential step in autophagy. Two SNAP29-containing SNARE complexes have been extensively studied in starvation-induced bulk autophagy, while the relevant SNARE complexes in other types of autophagy occurring under non-starvation conditions have been overlooked. Here, we found that autophagosome-lysosome fusion in selective autophagy under non-starvation conditions does not require SNAP29-containing SNARE complexes, but requires the STX17-SNAP47-VAMP7/VAMP8 SNARE complex. Further, the STX17-SNAP47-VAMP7/VAMP8 SNARE complex also functions in starvation-induced autophagy. SNAP47 is recruited to autophagosomes following concurrent detection of ATG8s and PI(4,5)P2 via its Pleckstrin homology domain. By contrast, SNAP29-containing SNAREs are excluded from selective autophagy due to inactivation by O-GlcNAcylation under non-starvation conditions. These findings depict a previously unknown, default SNARE complex responsible for autophagosome-lysosome fusion in both selective and bulk autophagy, which could guide research and therapeutic development in autophagy-related diseases.

A Mechanistic Understanding of the Modes of Ca2+ Ion Binding to the SARS-CoV-1 Fusion Peptide and Their Role in the Dynamics of Host Membrane Penetration.

The SARS-CoV-1 spike glycoprotein contains a fusion peptide (FP) segment that mediates the fusion of the viral and host cell membranes. Calcium ions are thought to position the FP optimally for membrane insertion by interacting with negatively charged residues in this segment (E801, D802, D812, E821, D825, and D830); however, which residues bind to calcium and in what combinations supportive of membrane insertion are unknown. Using biological assays and molecular dynamics studies, we have determined the functional configurations of FP-Ca2+ binding that likely promote membrane insertion. We first individually mutated the negatively charged residues in the SARS CoV-1 FP to assay their roles in cell entry and syncytia formation, finding that charge loss in the D802A or D830A mutants greatly reduced syncytia formation and pseudoparticle transduction of VeroE6 cells. Interestingly, one mutation (D812A) led to a modest increase in cell transduction, further indicating that FP function likely depends on calcium binding at specific residues and in specific combinations. To interpret these results mechanistically and identify specific modes of FP-Ca2+ binding that modulate membrane insertion, we performed molecular dynamics simulations of the SARS-CoV-1 FP and Ca2+ions. The preferred residue pairs for Ca2+ binding we identified (E801/D802, E801/D830, and D812/E821) include the two residues found to be essential for S function in our biological studies (D802 and D830). The three preferred Ca2+ binding pairs were also predicted to promote FP membrane insertion. We also identified a Ca2+ binding pair (E821/D825) predicted to inhibit FP membrane insertion. We then carried out simulations in the presence of membranes and found that binding of Ca2+ to SARS-CoV-1 FP residue pairs E801/D802 and D812/E821 facilitates membrane insertion by enabling the peptide to adopt conformations that shield the negative charges of the FP to reduce repulsion by the membrane phospholipid headgroups. This calcium binding mode also optimally positions the hydrophobic LLF region of the FP for membrane penetration. Conversely, Ca2+ binding to the FP E801/D802 and D821/D825 pairs eliminates the negative charge screening and instead creates a repulsive negative charge that hinders membrane penetration of the LLF motif. These computational results, taken together with our biological studies, provide an improved and nuanced mechanistic understanding of the dymanics of SARS-CoV-1 calcium binding and their potential effects on host cell entry.

All SNAP25 molecules in the vesicle-plasma membrane contact zone change conformation during vesicle priming.

In neuronal cell types, vesicular exocytosis is governed by the SNARE (soluble NSF attachment receptor) complex consisting of synaptobrevin2, SNAP25, and syntaxin1. These proteins are required for vesicle priming and fusion. We generated an improved SNAP25-based SNARE COmplex Reporter (SCORE2) incorporating mCeruelan3 and Venus and overexpressed it in SNAP25 knockout embryonic mouse chromaffin cells. This construct rescues vesicle fusion with properties indistinguishable from fusion in wild-type cells. Combining electrochemical imaging of individual release events using electrochemical detector arrays with total internal reflection fluorescence resonance energy transfer (TIR-FRET) imaging reveals a rapid FRET increase preceding individual fusion events by 65 ms. The experiments are performed under conditions of a steady-state cycle of docking, priming, and fusion, and the delay suggests that the FRET change reflects tight docking and priming of the vesicle, followed by fusion after ~65 ms. Given the absence of wt SNAP25, SCORE2 allows determination of the number of molecules at fusion sites and the number that changes conformation. The number of SNAP25 molecules changing conformation in the priming step increases with vesicle size and SNAP25 density in the plasma membrane and equals the number of copies present in the vesicle-plasma membrane contact zone. We estimate that in wt cells, 6 to 7 copies of SNAP25 change conformation during the priming step.

Synaptobrevin2 monomers and dimers differentially engage to regulate the functional trans-SNARE assembly.

The precise cell-to-cell communication relies on SNARE-catalyzed membrane fusion. Among ∼70 copies of synaptobrevin2 (syb2) in synaptic vesicles, only ∼3 copies are sufficient to facilitate the fusion process at the presynaptic terminal. It is unclear what dictates the number of SNARE complexes that constitute the fusion pore assembly. The structure-function relation of these dynamic pores is also unknown. Here, we demonstrate that syb2 monomers and dimers differentially engage in regulating the trans-SNARE assembly during membrane fusion. The differential recruitment of two syb2 structures at the membrane fusion site has consequences in regulating individual nascent fusion pore properties. We have identified a few syb2 transmembrane domain residues that control monomer/dimer conversion. Overall, our study indicates that syb2 monomers and dimers are differentially recruited at the release sites for regulating membrane fusion events.

The hookup model of the HOPS complex in autophagosome-lysosome fusion.

Macroautophagy/autophagy is a highly conserved process that involves the degradation of proteins, damaged organelles, and other cytoplasmic macromolecules. Autophagosome-lysosome fusion is critical for successful substrate degradation and is mediated by SNARE proteins. The fusion process requires additional vesicle docking and tethering-regulating factors. Our recent work has uncovered a functional model of autophagosome-lysosome fusion. We demonstrated that the six-subunit homotypic fusion and vacuole protein sorting (HOPS) complex can be assembled by two subcomplexes, the VPS39-VPS11 subcomplex (HOPS-2) and the VPS41-VPS16-VPS18-VPS33A subcomplex (HOPS-4). VPS39 binds with RAB2 on the autophagosome and VPS41 binds with RAB39A on the lysosome, which then promotes membrane tethering and autophagic SNARE-mediated membrane fusion. Moreover, we have revealed that ALS- and FTD-related C9orf72 is a guanine exchange factor (GEF) for RAB39A. In this punctum, we discuss how the C9orf72-RAB39A-HOPS axis function regulates autophagosome-lysosome fusion.

RUNDC1 negatively mediates the fusion of autophagosomes with lysosomes via regulating SNARE complex assembly.

Macroautophagy/autophagy is an essential pro-survival mechanism activated in response to nutrient deficiency. The proper fusion between autophagosomes and lysosomes is a critical step for autophagic degradation. We recently reported that RUNDC1 (RUN domain containing 1) inhibits autolysosome formation via clasping the ATG14-STX17-SNAP29 complex to hinder VAMP8 binding. We showed that RUNDC1 colocalizes with LC3 and associates with mature autophagosomes in cell lines and the zebrafish model. We utilized liposome fusion and in vitro autophagosome-lysosome fusion assays to demonstrate that RUNDC1 inhibits autolysosome formation. Moreover, we found that RUNDC1 clasps the ATG14-STX17-SNAP29 complex via stimulating ATG14 homo-oligomerization to inhibit ATG14 dissociation, which in turn prevents VAMP8 from binding to STX17-SNAP29. Our results demonstrate that RUNDC1 is a negative regulator of autophagy that restricts autophagosome fusion with lysosomes and is crucial for zebrafish survival in nutrient-deficient conditions. Here, we summarize our findings and discuss their implications for our understanding of autophagy regulation.

SARS-CoV-2 hijacks a cell damage response, which induces transcription of a more efficient Spike S-acyltransferase.

SARS-CoV-2 infection requires Spike protein-mediated fusion between the viral and cellular membranes. The fusogenic activity of Spike depends on its post-translational lipid modification by host S-acyltransferases, predominantly ZDHHC20. Previous observations indicate that SARS-CoV-2 infection augments the S-acylation of Spike when compared to mere Spike transfection. Here, we find that SARS-CoV-2 infection triggers a change in the transcriptional start site of the zdhhc20 gene, both in cells and in an in vivo infection model, resulting in a 67-amino-acid-long N-terminally extended protein with approx. 40 times higher Spike acylating activity, resulting in enhanced fusion of viruses with host cells. Furthermore, we observed the same induced transcriptional change in response to other challenges, such as chemically induced colitis and pore-forming toxins, indicating that SARS-CoV-2 hijacks an existing cell damage response pathway to optimize it fusion glycoprotein.

Inconspicuous Yet Indispensable: The Coronavirus Spike Transmembrane Domain.

Membrane-spanning portions of proteins' polypeptide chains are commonly known as their transmembrane domains (TMDs). The structural organisation and dynamic behaviour of TMDs from proteins of various families, be that receptors, ion channels, enzymes etc., have been under scrutiny on the part of the scientific community for the last few decades. The reason for such attention is that, apart from their obvious role as an "anchor" in ensuring the correct orientation of the protein's extra-membrane domains (in most cases functionally important), TMDs often actively and directly contribute to the operation of "the protein machine". They are capable of transmitting signals across the membrane, interacting with adjacent TMDs and membrane-proximal domains, as well as with various ligands, etc. Structural data on TMD arrangement are still fragmentary at best due to their complex molecular organisation as, most commonly, dynamic oligomers, as well as due to the challenges related to experimental studies thereof. Inter alia, this is especially true for viral fusion proteins, which have been the focus of numerous studies for quite some time, but have provoked unprecedented interest in view of the SARS-CoV-2 pandemic. However, despite numerous structure-centred studies of the spike (S) protein effectuating target cell entry in coronaviruses, structural data on the TMD as part of the entire spike protein are still incomplete, whereas this segment is known to be crucial to the spike's fusogenic activity. Therefore, in attempting to bring together currently available data on the structure and dynamics of spike proteins' TMDs, the present review aims to tackle a highly pertinent task and contribute to a better understanding of the molecular mechanisms underlying virus-mediated fusion, also offering a rationale for the design of novel efficacious methods for the treatment of infectious diseases caused by SARS-CoV-2 and related viruses.

Synaptophysin chaperones the assembly of 12 SNAREpins under each ready-release vesicle.

The synaptic vesicle protein Synaptophysin (Syp) has long been known to form a complex with the Vesicle associated soluble N-ethylmaleimide sensitive fusion protein attachment receptor (v-SNARE) Vesicle associated membrane protein (VAMP), but a more specific molecular function or mechanism of action in exocytosis has been lacking because gene knockouts have minimal effects. Utilizing fully defined reconstitution and single-molecule measurements, we now report that Syp functions as a chaperone that determines the number of SNAREpins assembling between a ready-release vesicle and its target membrane bilayer. Specifically, Syp directs the assembly of 12 ± 1 SNAREpins under each docked vesicle, even in the face of an excess of SNARE proteins. The SNAREpins assemble in successive waves of 6 ± 1 and 5 ± 2 SNAREpins, respectively, tightly linked to oligomerization of and binding to the vesicle Ca++ sensor Synaptotagmin. Templating of 12 SNAREpins by Syp is likely the direct result of its hexamer structure and its binding of VAMP2 dimers, both of which we demonstrate in detergent extracts and lipid bilayers.

Regulation of Syntaxin3B-Mediated Membrane Fusion by T14, Munc18, and Complexin.

Retinal neurons that form ribbon-style synapses operate over a wide dynamic range, continuously relaying visual information to their downstream targets. The remarkable signaling abilities of these neurons are supported by specialized presynaptic machinery, one component of which is syntaxin3B. Syntaxin3B is an essential t-SNARE protein of photoreceptors and bipolar cells that is required for neurotransmitter release. It has a light-regulated phosphorylation site in its N-terminal domain at T14 that has been proposed to modulate membrane fusion. However, a direct test of the latter has been lacking. Using a well-controlled in vitro fusion assay, we found that a phosphomimetic T14 syntaxin3B mutation leads to a small but significant enhancement of SNARE-mediated membrane fusion following the formation of the t-SNARE complex. While the addition of Munc18a had only a minimal effect on membrane fusion mediated by SNARE complexes containing wild-type syntaxin3B, a more significant enhancement was observed in the presence of Munc18a when the SNARE complexes contained a syntaxin3B T14 phosphomimetic mutant. Finally, we showed that the retinal-specific complexins (Cpx III and Cpx IV) inhibited membrane fusion mediated by syntaxin3B-containing SNARE complexes in a dose-dependent manner. Collectively, our results establish that membrane fusion mediated by syntaxin3B-containing SNARE complexes is regulated by the T14 residue of syntaxin3B, Munc18a, and Cpxs III and IV.

How micron-sized exocrine vesicles release content: A comparison with sub-micron endocrine vesicles.

Exocytosis releases vesicular contents to mediate physiological functions. In this issue, Biton et al. (https://doi.org/10.1083/jcb.202302112) found four modes of releasing micron-sized exocrine vesicles and the underlying mechanisms involving actomyosin and BAR domain proteins. We highlight their discovery, compare it with much smaller/faster neuroendocrine vesicle fusion, and draw distinct and conserved principles regarding their membrane transformations, pore dynamics, and underlying mechanisms.

C9orf72-catalyzed GTP loading of Rab39A enables HOPS-mediated membrane tethering and fusion in mammalian autophagy.

The multi-subunit homotypic fusion and vacuole protein sorting (HOPS) membrane-tethering complex is required for autophagosome-lysosome fusion in mammals, yet reconstituting the mammalian HOPS complex remains a challenge. Here we propose a "hook-up" model for mammalian HOPS complex assembly, which requires two HOPS sub-complexes docking on membranes via membrane-associated Rabs. We identify Rab39A as a key small GTPase that recruits HOPS onto autophagic vesicles. Proper pairing with Rab2 and Rab39A enables HOPS complex assembly between proteoliposomes for its tethering function, facilitating efficient membrane fusion. GTP loading of Rab39A is important for the recruitment of HOPS to autophagic membranes. Activation of Rab39A is catalyzed by C9orf72, a guanine exchange factor associated with amyotrophic lateral sclerosis and familial frontotemporal dementia. Constitutive activation of Rab39A can rescue autophagy defects caused by C9orf72 depletion. These results therefore reveal a crucial role for the C9orf72-Rab39A-HOPS axis in autophagosome-lysosome fusion.

MARCKS Effector Domain, a reversible lipid ligand, illuminates late stages of membrane fusion.

Yeast vacuolar HOPS tethers membranes, catalyzes trans-SNARE assembly between R- and Q-SNAREs, and shepherds SNAREs past early inhibition by Sec17. After partial SNARE zippering, fusion is driven slowly by either completion of SNARE zippering or by Sec17/Sec18, but rapid fusion needs zippering and Sec17/Sec18. Using reconstituted-vacuolar fusion, we find that MARCKS Effector Domain (MED) peptide, a lipid ligand, blocks fusion reversibly at a late reaction stage. The MED fusion blockade is overcome by either salt extraction, inactivation with the MED ligand calmodulin, or addition of Sec17/Sec18. During incubation with MED, SNAREs assemble stable complexes in trans and fusion becomes resistant to antibody to the Qa SNARE. When Q-SNAREs are preassembled, a synthetic tether can replace HOPS for fusion. With a synthetic tether, fusion needs both complete SNARE zippering and Sec17/Sec18 to overcome a MED block. In contrast, when SNARE domains are only two-third zippered, only HOPS will support Sec17/Sec18 driven fusion without needing complete zippering. HOPS thus remains engaged with SNAREs during zippering. MED facilitates the study of distinct fusion stages: tethering, initial trans-SNARE assembly and its sensitivity to Sec17, SNARE zippering, Sec17/Sec18 engagement, and lipid and lumenal mixing.

Ca2+ and cAMP open differentially dilating synaptic fusion pores.

Neuronal dense-core vesicles (DCVs) contain neuropeptides and much larger proteins that affect synaptic growth and plasticity. Rather than using full collapse exocytosis that commonly mediates peptide hormone release by endocrine cells, DCVs at the Drosophila neuromuscular junction release their contents via fusion pores formed by kiss-and-run exocytosis. Here, we used fluorogen-activating protein (FAP) imaging to reveal the permeability range of synaptic DCV fusion pores and then show that this constraint is circumvented by cAMP-induced extra fusions with dilating pores that result in DCV emptying. These Ca2+-independent full fusions require PKA-R2, a PKA phosphorylation site on Complexin and the acute presynaptic function of Rugose, the homolog of mammalian neurobeachin, a PKA-R2 anchor implicated in learning and autism. Therefore, localized Ca2+-independent cAMP signaling opens dilating fusion pores to release large cargoes that cannot pass through the narrower fusion pores that mediate spontaneous and activity-dependent neuropeptide release. These results imply that the fusion pore is a variable filter that differentially sets the composition of proteins released at the synapse by independent exocytosis triggers responsible for routine peptidergic transmission (Ca2+) and synaptic development (cAMP).

Rab GTPase regulation of phagosome-lysosome fusion is bypassed in the presence of micromolar Ca2.

Several ATP- and cytosol-dependent fusion processes between membranes of the endocytic and exocytic pathways have been biochemically reconstituted. Here, we present a phagosome-lysosome fusion reaction that is driven by micromolar concentrations of Ca2+ in the absence of ATP and cytosol. Investigating classical fusion and Ca2+-driven fusion (CaFu) side-by-side in vitro, using the same membrane preparations, we show that CaFu is faster than standard fusion (StaFu), leads to larger fusion products and is not blocked by established inhibitors of StaFu. A Ca2+ concentration of ∼120 µM supports maximal membrane attachment, and 15 µM Ca2+ supports maximal membrane fusion, indicating that Ca2+ has both a membrane-binding activity and a fusion-promoting activity. StaFu and CaFu are inhibited by a mutant form of α-SNAP (NAPA) that does not support soluble N-ethylmaleimide-sensitive factor attachment protein receptor (SNARE) activation, and both are inhibited by a mixture of the cytosolic domains of three cognate Q-SNARE proteins, demonstrating a role of SNAREs in Ca2+-driven membrane merger. CaFu is independent of the Ca2+-regulated proteins synaptotagmin-7, calmodulin, and annexins A2 and A7. We propose that CaFu corresponds to the last step of phagosome-lysosome fusion, when a raised Ca2+ concentration from the compartment lumen activates SNAREs for fusion.

Requirement of Cholesterol for Calcium-Dependent Vesicle Fusion by Strengthening Synaptotagmin-1-Induced Membrane Bending.

Cholesterol is essential for neuronal activity and function. Cholesterol depletion in the plasma membrane impairs synaptic transmission. However, the molecular mechanisms by which cholesterol deficiency leads to defects in vesicle fusion remain poorly understood. Here, it is shown that cholesterol is required for Ca2+ -dependent native vesicle fusion using the in vitro reconstitution of fusion and amperometry to monitor exocytosis in chromaffin cells. Purified native vesicles are crucial for the reconstitution of physiological Ca2+ -dependent fusion, because vesicle-mimicking liposomes fail to reproduce the cholesterol effect. Intriguingly, cholesterol has no effect on the membrane binding of synaptotagmin-1, a Ca2+ sensor for ultrafast fusion. Cholesterol strengthens local membrane deformation and bending induced by synaptotagmin-1, thereby lowering the energy barrier for Ca2+ -dependent fusion to occur. The data provide evidence that cholesterol depletion abolishes Ca2+ -dependent vesicle fusion by disrupting synaptotagmin-1-induced membrane bending, and suggests that cholesterol is an essential lipid regulator for Ca2+ -dependent fusion.