Munc18a Does Not Alter Fusion Rates Mediated by Neuronal SNAREs, Synaptotagmin, and Complexin*

Background: Munc18-1 is an important factor for synaptic transmitter release, but its molecular mechanism remains an enigma. Results: Munc18a does not affect fusion kinetics. It can sequester syntaxin-1A molecules from the syntaxin-1A·SNAP-25 t-SNARE complex. Conclusion: Munc18a has no effect on complete fusion in conjunction with synaptotagmin-1, complexin-1, and neuronal SNAREs. Significance: This work provides new insights into Munc18-1 and its interactions with other synaptic proteins.

Synaptic neurotransmission involves the release of neurotransmitter into the synaptic cleft upon an action potential.
The release of neurotransmitters into the synaptic cleft is mediated by the fusion of neurotransmitter-laden synaptic vesicles with the plasma membrane in the active zone upon an increase of the local Ca 2ϩ concentration following depolarization. There are several synaptic proteins that have key roles in the Ca 2ϩtriggered fusion of synaptic vesicles with the plasma membrane: the neuronal soluble N-ethylmaleimide-sensitive factor (NSF) 7 attachment protein receptors (SNAREs) synaptobrevin-2/VAMP2 (vesicle-associated membrane protein), syntaxin-1A, and SNAP-25 (synaptosome-associated protein 25); synaptotagmin; complexin; and nSec1/Munc-18a (1)(2)(3). Whereas SNARE proteins can provide the minimal energy to drive membrane fusion (4) and synaptotagmin-1 is the Ca 2ϩ sensor for fast synchronous release (5), the molecular roles of complexin and Munc18a are less certain (6 -9). Here we focus on the mechanism of action of Munc18a.
There is intensive debate as to whether Sec1/Munc18 (SM) proteins, which are required components of all membrane trafficking pathways, play a role in facilitating or regulating SNARE complex assembly or if they are involved in the fusion process itself (10,11). Evidence that the neuronal SM protein Munc18a is involved in more than "just" facilitation or regulation comes primarily from Munc18a knock-out studies in mice (12) that resulted in complete loss of neurotransmitter release. However, Munc18a knockout also resulted in a 70% reduction in total syntaxin-1 protein level in brain lysates of the knock-out mice and reduced syntaxin-1 mRNA levels (13). Nonetheless, the residual syntaxin-1 was correctly targeted to the synapse and formed SDS-resistant SNARE complex (14), so the reduction in syntaxin-1 levels in the Munc18a knock-out mice may not explain the complete loss of neurotransmitter release. Moreover, amperometry of neuroendocrine cell exocytosis along with overexpression of Munc18a and selected mutants showed that Munc18a has an effect on the initial release rate (15).
So far, there is little evidence from reconstitution experiments that Munc18a plays a major role in fusion itself, although stimulatory effects by SM proteins on SNARE-dependent lipid mixing were observed using ensemble fluorescence assays (16 -18). A stimulatory effect of Munc18a was also observed in conjunction with neuronal SNAREs using a single vesicle-vesicle lipid-mixing assay (19). However, all of these experiments only used neuronal SNAREs in addition to Munc18a. Subsequent ensemble experiments that reconstituted neuronal SNAREs in liposomes along with neuron-specific proteins synaptotagmin-1 and complexin-1 showed only a relatively small increase of spontaneous (i.e. at zero Ca 2ϩ concentration) lipid mixing and content mixing upon Munc18a inclusion when complexin-1 was present (see Fig. 3 (D and F) in Ref. 20).
Reconstitution experiments with a different subset of synaptic proteins showed that Munc18a acts in conjunction with NSF and ␣SNAP (soluble NSF adaptor protein), locking syntaxin-1A in a closed conformation and setting the stage for ternary trans-neuronal SNARE complex by the action of Munc13 and an approaching vesicle with synaptobrevin-2 (8). In this reconstitution experiment, Munc18a played an essential role to facilitate trans-SNARE complex formation in conjunction with NSF, ␣SNAP, and Munc13. Although an important step toward understanding the mechanism of Munc18, this study did not provide direct evidence that Munc18a stimulates fusion itself, because the amount of ensemble lipid mixing was roughly the same for a minimal system consisting only of neuronal SNAREs, the C2AB fragment of synaptotagmin-1, and 0.5 M Ca 2ϩ as well as for the more complete system consisting of the same proteins plus NSF, ␣SNAP, Munc18a, and Munc13 (see Fig. 4D in Ref. 8). However, it is possible that the ensemble lipid-mixing method used for this particular comparison may have masked an effect of Munc18a by opposing effects on the association of vesicles prior to fusion and fusion itself because ensemble fluorescence assays generally cannot distinguish between these two effects (21).
The function of SM proteins may be similar or different in other systems. As a case in point, Wickner and co-workers (22) showed that in the yeast vacuolar fusion system, efficient fusion with reconstituted proteins requires the HOPS complex, which contains the SM protein Vps33. HOPS, with its SM subunit, appears to be involved primarily in tethering and in a proofreading mechanism, rather than fusion itself (23)(24)(25), because its function can be replaced by synthetically induced tethering of vacuolar vesicles (26). However, the HOPS tethering complex causes a much more significant increase in the rate of fusion that cannot be explained by a simple increase in the number of trans-SNARE complexes (27), suggesting that it makes SNARE complexes more fusion-competent prior to fusion (e.g. by preventing transient antiparallel SNARE associations (28)).
Munc18a interacts with neuronal SNARE proteins throughout the synaptic vesicle cycle in complexes involving syntaxin-1A and the ternary SNARE complex (syntaxin-1A⅐SNAP-25⅐synaptobrevin-2) (29 -31). Moreover, Munc18a co-localizes with syntaxin-1A and SNAP-25 in fixed cultured neurons and forms nanoclusters in the plasma membrane (32); knockdown of both syntaxin isoforms, syntaxin-1A and syntaxin-1B, disrupts the co-localization of Munc18a to  Munc18a tightly binds to syntaxin-1A in a "closed" conformation in which its Habc domain associates in cis with its SNARE domain (29,33), which in turn hinders the SNARE complex assembly (31, 34 -36). Interestingly, in neuroendocrine cells, transfection of syntaxin alone results in its mislocalization, whereas co-transfection of both Munc18a and syntaxin results in trafficking to the plasma membrane (37,38). Thus, one function of Munc18a may be to stabilize or "chaperone" syntaxin, which also agrees with the aforementioned knock-out studies of Munc18a in mice that showed a reduced level of syntaxin-1 in the plasma membrane (13).
Binding of Munc18a to the closed conformation of syntaxin-1A is not always required because rescue experiments in cultured syntaxin-1A-deficient neurons showed that the syntaxin-1A Habc domain is not required for evoked synaptic vesicle fusion, although it was required for spontaneous synaptic vesicle fusion (39). In contrast, the N-terminal residues of syntaxin-1A are required for binding of the ternary SNARE complex to Munc18a and may thereby serve as a recruiting factor for Munc18a (18). Because the Habc domain is needed for the closed conformation of syntaxin-1A and the tight interaction with Munc18a, this study suggests that the N-terminal and Habc domains mediate different functions in synaptic vesicle fusion.
In this paper, we start from the observation that a minimal system consisting of neuronal SNAREs and the soluble C2AB fragment of synaptotagmin-1 exhibited the same amount of lipid mixing as a more complete reconstitution that also included Munc18a, NSF, ␣SNAP, and Munc13 (8). At variance with that work, here we used full-length synaptotagmin-1 and included complexin-1, in a single vesicle-vesicle content-mixing assay. We recently used this assay to decipher the functions of complexin-1 (40) and found that it qualitatively reproduces effects observed in neuronal cultures for both spontaneous and Ca 2ϩ -triggered release for a variety of complexin-1 truncations and mutations, lending credence to this approach to investigate mechanistic questions.
In the context of studying the potential effect of Munc18a in vesicle fusion, our single vesicle content-mixing assay enables us to distinguish between fusion and vesicle association (often called "docking," but we prefer not to use this term because it has a different meaning during the life cycle of a synaptic vesicle (1)). We investigated the effects of Munc18a on fusion and vesicle association between synaptic vesiclemimicking vesicles (with reconstituted synaptobrevin-2 and synaptotagmin-1, referred to as SV vesicles) and plasma membrane-mimicking vesicles (with reconstituted syntaxin-1A and SNAP-25, referred to as PM vesicles) in the presence of complexin-1. We then used a single molecule assay to demonstrate that Munc18a is capable of sequestering syntaxin-1A from preformed t-SNARE complex. We also tested a double mutant that mimics phosphorylation of Munc18a(S306E/S313E) (41-44) as well as the role of the N-terminal residues of syntaxin-1A. To understand the molecular basis of the effect of Munc18a on vesicle association and fusion, we used purified proteins to demonstrate that Munc18a can interact specifically with the neuronal binary (syntaxin-1A⅐SNAP-25) SNARE complex (also referred to as t-SNARE complex) as well as the syntaxin-1A⅐SNAP-25⅐synaptobrevin-2 (ternary) SNARE complex.

Protein Purification
Synaptotagmin-1-Full-length rat synaptotagmin-1 (Syt1) was expressed in Escherichia coli from plasmid pJ414 (DNA 2.0, Menlo Park, CA) with a C-terminal decahistidine tag proceeded by a PreScission protease cleavage site to remove the tag and all cysteine residues changed to alanine except for the cysteine residue at position 277. Syt1 was expressed in E. coli BL21 (DE3) by growing the cells to A 600 of 0.6 -0.8 at 37°C and then induced at 20°C for 12-16 h with 0.5 mM isopropyl ␤-D-1thiogalactopyranoside (IPTG). The cells from 6 liters of induced culture were harvested and suspended in 200 ml of synaptotagmin buffer A (50 mM sodium phosphate, pH 7.4, 600 mM NaCl, 2 mM DTT, and 10% glycerol (w/v)) supplemented with phenylmethylsulfonyl fluoride (PMSF) to 1 mM and two EDTA-free Complete Protease Inhibitor Mixture tablets (Roche Applied Science). Cells were lysed by three passes through the Emulsiflex C5 homogenizer (Avestin, Ottawa, Canada) at 15,000 p.s.i. Cell debris and inclusion bodies were removed by centrifugation at 8,000 rpm in a Beckman JA-20 rotor (Beckman Coulter, Brea, CA) for 10 min. The supernatant was then centrifuged at 8,000 rpm for 10 min in the same rotor. Membranes were collected by centrifugation at 40,000 rpm in a Beckman Ti-45 rotor (Beckman Coulter) for 1 h. Membranes were washed by homogenization in 100 ml of synaptogamin buffer A (50 mM sodium phosphate, pH 7.4, 600 mM NaCl, 2 mM DTT, 1 mM PMSF, 10% glycerol (w/v)) supplemented with two EDTA-free Complete Protease Inhibitor Mixture tablets (Roche Applied Science). Membranes were harvested again by centrifugation in a Ti-45 rotor (Beckman Coulter) at 40,000 rpm for 1 h. The pellet was suspended in 100 ml of synaptotagmin buffer B supplemented with 2 EDTA-free Complete Protease Inhibitor Mixture tablets and snap frozen in two 50-ml aliquots in liquid nitrogen. One aliquot of membranes was solubilized in the presence of 1.5% dodecylmaltoside (Anatrace, Maumee, OH) for 1 h at 4°C. The extract was clarified by centrifugation in a Ti 45 rotor (Beckman Coulter) at 40,000 rpm for 35 min. The supernatant was bound to a 4-ml bed volume of Ni-NTA beads (Qiagen, Hilden, Germany) equilibrated in synaptotagmin buffer A containing 1.5% dodecylmaltoside (Anatrace) by stirring at 4°C for 1 h. Beads were harvested by centrifugation and poured into a column, attached to an AKTA Prime FPLC system (GE Healthcare), and washed with 50 ml of synaptotagmin buffer A supplemented with 110 mM n-octyl-␤-D-glucopyranoside (OG) (Anatrace) and 10 mM imidazole and then eluted with synaptotagmin buffer A supplemented with 110 mM OG and 500 mM imidazole. Protein-containing fractions were combined and injected on a Superdex 200 (16/60) column (GE Healthcare) equilibrated in 20 mM sodium phos-phate, pH 7.4, 300 mM NaCl, 2 mM DTT, 110 mM OG, 10% glycerol (w/v) (synaptotagmin buffer B). Peak fractions were then combined, and 100 g of prescission protease (GE Healthcare) was added to cleave the tag. Prescission protease-cleaved Syt1 was diluted with 20 mM sodium phosphate, pH 7.4, 2 mM DTT, 110 mM OG, 20 M EGTA, 10% glycerol (w/v) (synaptotagmin buffer C) to bring the NaCl to 100 mM and then injected on a MonoS 5/50 column (GE Healthcare), washed with synaptotagmin buffer C containing 100 mM NaCl, and eluted with synaptotagmin buffer C containing 600 mM NaCl. Protein-containing fractions were dialyzed against synaptotagmin buffer C to lower the salt concentration and used for reconstitution into proteoliposomes.
Complexin-1-Full-length wild type complexin-1 (Cpx) was expressed in E. coli using BL21 (DE3) cells with a thrombincleavable N-terminal hexahistidine tag from plasmid pET28a (Novagen). Cells were grown in Luria Broth to an A 600 between 0.6 and 0.8, and protein expression was induced by the addition of IPTG to 0.5 mM for 12-16 h at 20°C. Cells were harvested by centrifugation and resuspended in 20 mM HEPES, pH 7.5, 500 mM NaCl, 2 mM DTT, 10 mM imidazole (complexin buffer A) supplemented with 1 mM PMSF and EDTA-free Complete Protease Inhibitor Mixture tablets (Roche Applied Science). Cells were lysed by passing them through the Emulsiflex C5 homogenizer (Avestin) at 15,000 p.s.i. three times. Lysate was clarified by centrifugation in a Ti-45 rotor (Beckman Coulter) for 35 min at 40,000 rpm. Supernatant was bound to a 4-ml bed volume of Ni-NTA beads (Qiagen) in batch, stirring for 1 h at 4°C. The beads were harvested by centrifugation and poured into a column, attached to an AKTA Prime system (GE Healthcare), and washed with 60 ml of complexin buffer A containing 25 mM imidazole and then eluted with complexin buffer A containing 450 mM imidazole. Protein-containing fractions were combined, and 100 units of thrombin were added. This was then dialyzed against 500 ml of 20 mM HEPES, pH 7.5, 50 mM NaCl, 4 mM DTT (complexin buffer B) overnight at 4°C. PMSF was added to 1 mM to quench the thrombin activity, and the mixture was injected onto a MonoQ column 5/50 (GE Healthcare) equilibrated with complexin buffer B. The column was washed with 20 column volumes of complexin buffer B, and the protein eluted with a linear NaCl gradient in complexin buffer B from 50 to 500 mM NaCl over 30 column volumes. The protein-containing fractions were combined and dialyzed against 20 mM HEPES, pH 7.5, 100 mM NaCl, 4 mM DTT overnight at 4°C. The protein concentration was measured by absorbance at 280 nm, and aliquots were frozen in liquid nitrogen.
Wild Type Munc18a and Munc18a(S306E/S313E)-Wild type Munc18-1a (referred to here as Munc18a) and Munc18a (S306E/S313E) were expressed in E. coli with an N-terminal, TEV protease-cleavable hexahistidine tag from plasmid pPROExHTa. Proteins were expressed by the addition of IPTG to late log phase cultures (A 600 ϳ0.6 -0.8) to 0.5 mM at 20°C for 16 -18 h. Cell pellets from 6 liters of culture were resuspended in 200 ml of Munc18a lysis buffer (20 mM sodium phosphate, pH 8, 300 mM NaCl, 10 mM imidazole), supplemented with 4 EDTA-free Complete Protease Inhibitor Mixture tablets (Roche Applied Science) and 1 mM PMSF and lysed by passing them three times through an Emulsiflex C5 homogenizer (Avestin) at 15,000 p.s.i. three times.
The lysate was clarified in the Ti45 rotor (Beckman Coulter) at 40,000 rpm for 35 min. The supernatant was applied to a 5-ml Ni-NTA-agarose column (Qiagen), washed on an ATKA Prime FPLC system (GE Healthcare) with 100 ml of Munc18a lysis buffer supplemented with 50 mM imidazole, and then eluted with Munc18a lysis buffer supplemented with 300 mM imidazole. Protein-containing fractions were supplemented with 1 mM EDTA and 5 mM DTT, 100 g of TEV protease was added to the combined protein-containing fractions to remove the hexahistidine tag, and the mixture was dialyzed overnight against 1 liter of 20 mM HEPES, pH 7.5, 50 mM NaCl, 1 mM EDTA, 5 mM DTT (monoQ buffer A). The TEV protease-cleaved Munc18a was then applied to a MonoQ 5/50 column (GE Healthcare) equilibrated in monoQ buffer A, and the column was eluted in a linear gradient from 50 to 500 mM NaCl. Protein-containing fractions were dialyzed against 20 mM HEPES, pH 7.5, 180 mM NaCl, 20 M EGTA, 0.1% ␤-mercaptoehanol for 3 h and used immediately afterward in fusion and docking assays.
Full-length Syntaxin-1A and Synaptobrevin-2-Full-length rat syntaxin-1A (Stx1a) and synaptobrevin-2 were prepared as described previously (45). These proteins were expressed separately in E. coli with an N-terminal, TEV protease-cleavable, hexahistidine tag from plasmid pTEV5 (46). Proteins were expressed overnight at 25°C in autoinducing medium (47) in E. coli strain C43 (48). Cell pellets from 8 liters of culture were suspended in 400 ml of 50 mM sodium phosphate, pH 8, 1 M NaCl, 5 mM EDTA, and 1 mM PMSF supplemented with Complete Protease Inhibitor Mixture tablets (Roche Applied Science) and broken by three passes through a M-110-EH microfluidizer (Mircrofluidics Corp., Newton, MA) at 15,000 p.s.i. Inclusion bodies were removed by two consecutive 10-min spins at 10,000 rpm in a JA-14 (Beckman Coulter) rotor, and the membrane fraction was collected by centrifugation at 40,000 rpm for 2 h in a Ti-45 (Beckman Coulter) rotor. Membranes containing syntaxin-1A were further washed with a buffer containing 10 mM Tris-H 2 SO 4 , pH 7.5, 10 mM EDTA, 10% glycerol (w/v), centrifuged at 40,000 rpm for 1 h in a Ti-45 rotor. Membrane pellets were resuspended in 20 mM HEPES, pH 7.5, 500 mM NaCl, 1 mM tris(2-carboxyethyl)phosphine (TCEP), and 10% glycerol (w/v) and centrifuged for an additional 1 h in the same rotor. Membranes were suspended to a concentration of 5 mg/ml in 20 mM HEPES, pH 7.5, 500 mM NaCl, 1 mM TCEP, 10 mM imidazole, and 10% glycerol (w/v), 1 mM PMSF, and EDTAfree Complete Protease Inhibitor Mixture. Dodecylmaltoside (Anatrace) was added to 2% (w/v), and after incubation at 4°C for 1 h, the sample was centrifuged for 35 min at 40,000 rpm in a Ti-45 (Beckman Coulter) rotor, and the supernatant was loaded onto a 1-ml column of Ni-NTA-agarose (Qiagen). The column was washed with 20 mM HEPES, pH 7.5, 300 mM NaCl, 1 mM TCEP, 20 mM imidazole, 110 mM OG, and 10% glycerol (w/v), and the proteins were eluted in the same buffer containing 450 mM imidazole and 1 M NaCl. 1 mM EDTA was immediately added to the pooled fraction and loaded onto a Superdex 200 HR 10/300 GL column (GE Healthcare) that was equilibrated with 20 mM HEPES, pH 7.5, 300 mM NaCl, 1 mM TCEP, 110 mM OG, and 10% glycerol (w/v). Protein fractions were pooled and digested with 100 g TEV protease for 30 min at ambient temperature, after which the TEV protease had pre-cipitated. TEV was removed by centrifugation at 5,000 rpm for 10 min in an Eppendorf model 5804 R tabletop centrifuge (Eppendorf North America, Hauppauge, NY). The sample in the supernatant was again digested with 100 g of TEV protease for 30 min at ambient temperature, after which the reaction was complete and the TEV protease had precipitated. TEV was removed by centrifugation at 5,000 rpm for 10 min in the same rotor.
The N-terminally truncated construct of syntaxin-1A (Stx1a (10 -288)) was expressed with an N-terminal thrombin-cleavable glutathione S-tranferase tag from plasmid pGEX-KG (49). After induction with 0.1 mM IPTG at A 600 of 0.8, the proteins were expressed overnight at 16°C at 100 rpm in Luria broth in BL21 (DE3) E. coli cells (Novagen, EMD Chemicals, Gibbstown, NJ). Cell pellets from 1 liter of culture were suspended in 40 ml of sodium phosphate, pH 8, 1 mM dithiothreitol, 0.2% Triton X-100, 1% dodecylmaltoside, and 0.2 mM PMSF supplemented with Complete Protease Inhibitor Mixture tablets (Roche Applied Science), and the cells were disrupted by sonication at 5/15-s on/off pulse for 5 min at 50% power using a Sonicator Ultrasonic Processor XL-2020 (Misonix, Farmingdale, NY). Inclusion bodies were removed by a 10-min 15,000-rpm spin in a JA-20 (Beckman Coulter) rotor and further cleared by centrifugation at 40,000 rpm for 35 min in a Ti-45 (Beckman Coulter) rotor. The supernatant was incubated for 2 h with a 1.5-ml slurry of glutathione-Sepharose 4 Fast Flow beads (GE Healthcare) that was pre-equilibrated in the same buffer. The beads were then collected in a column and washed with 80 column volumes of buffer containing 50 mM sodium phosphate, pH 8, 1 mM dithiothreitol, and 0.2% Triton X-100 and then with 30 column volumes of buffer containing 40 mM Tris-HCl, pH 8, 150 mM NaCl, 1 mM dithiothreitol, and 0.8% OG. The beads were then incubated for 90 min at ambient temperature with 100 units of bovine ␣-thrombin (Hematologic Technologies, Essex Junction, VT) in buffer containing 40 mM Tris-HCl, pH 8, 150 mM NaCl, 1 mM dithiothreitol, and 3% OG. The supernatant containing cleaved ⌬N syntaxin(10 -288) was flowed off the column, after which the thrombin was quenched with 1 mM PMSF. Although the GST-syntaxin-1A construct was designed to include the N terminus of syntaxin-1A, the application of thrombin proteolysis also removed nine additional residues at the N terminus as determined by Edman sequencing. Removal of these nine N-terminal residues has no significant effect on fusion efficiency and kinetics for spontaneous and Ca 2ϩ -triggered fusion with neuronal SNAREs, synaptotagmin-1, and complexin-1 (Fig. 1, G and H).

SNAP-25 and SNAP-25(K76C) Constructs for Vesicle Reconstitution and Single Molecule Experiments-Cysteine-free
SNAP-25 (C84S, C85S, C90S, and C92S) or single cysteinecontaining SNAP-25(K76C) (C84S, C85S, C90S, C92S, and K76C) was expressed from plasmid pTEV5 (46) with an N-terminal TEV protease-cleavable hexahistidine tag. The proteins were expressed overnight in autoinducing medium (47) in E. coli strain BL21(DE3) at 30°C. Cells from 4 liters of culture were resuspended in 200 ml of 50 mM sodium phosphate, pH 8.0, 300 mM NaCl, 20 mM imidazole, and 10% glycerol (w/v) supplemented with 1 mM PMSF and four EDTA-free protease inhibitor mixture tablets. Cells were lysed by three passes through an Emulsiflex C5 homogenizer (Avestin) at 15,000 p.s.i. The lysate was clarified by centrifugation in the Ti-45 rotor for 1.5 h at 40,000 rpm. The supernatant was bound to a 5-ml Ni-NTA column by flowing the lysate onto the column at 1 ml/min using an AKTA Prime FPLC system (GE Healthcare). The column was washed with 150 ml of SNAP-25 buffer containing 50 mM sodium phosphate, pH 8.0, 300 mM NaCl, 10% glycerol (w/v), supplemented with 50 mM imidazole, and eluted with buffer containing 50 mM sodium phosphate, pH 8.0, 300 mM NaCl, 10% glycerol (w/v), and 350 mM imidazole. Proteincontaining fractions were combined, DTT was added to 5 mM, EDTA was added to 1 mM, and 150 g of TEV protease was added to remove the hexahistidine tag. This mixture was dialyzed against buffer containing 20 mM HEPES, pH 7.5, 100 mM NaCl, 4 mM DTT, and 10% glycerol (w/v) overnight at 4°C. The TEV-cleaved SNAP-25 was concentrated in a 15-ml Amicon ultracentrifugal concentrator with a 10,000 molecular weight cut-off dialysis cassette (Millipore, Billerica, MA) to 5 ml and injected onto the Superdex 200 (16/60) column (GE Healthcare) equilibrated with buffer containing 20 mM HEPES, pH 7.5, 100 mM NaCl, 4 mM DTT, 10% glycerol (w/v). Protein-containing fractions were combined and then concentrated using a 3,000 molecular weight cut-off dialysis cassette (Millipore). The concentration of SNAP-25 was measured by absorbance at 280 nm, and aliquots were frozen in liquid nitrogen.
Purified SNAP-25(K76C) was incubated with a 10-fold molar excess of TCEP for 10 min followed by incubation with a 10-fold molar excess of Alexa647 c2 maleimide at room temperature for 2 h. The reaction mix was then transferred to 4°C overnight. A Sephadex G-50 column (Sigma-Aldrich) was used for crude separation of free Alexa647 from proteins. An additional dialysis step was done using a 10,000 molecular weight cut-off dialysis cassette (Millipore) at 4°C against buffer containing 20 mM HEPES, pH 7.5, 90 mM NaCl to remove remaining free dye. The concentrations of SNAP-25(K76C) and Alexa647 were measured by absorbance at 280 and 651 nm, and aliquots were frozen in liquid nitrogen.

Protein Reconstitution in Liposomes
We used the same membrane compositions and protein densities as in our previous studies (45,50). Likewise, the reconstitution protocol was similar (50) with several changes; the concentration of 2-mercaptoethanol in the dialysis and vesicle-free buffers was changed from 1 to 0.1%, an extra dialysis step was added to ensure removal of Ca 2ϩ and detergent, all dialysis buffers contained 0.8 g/liter Chelex 100 resin (Bio-Rad) and 2.5 g/liter Bio-beads SM2 (Bio-Rad), and the Sepharose CL-4B column (GE Healthcare) was packed under constant pressure.
Briefly, SV vesicles were reconstituted with both synaptotagmin-1 and synaptobrevin-2 (except when noted otherwise), whereas PM vesicles were reconstituted with syntaxin-1A and SNAP-25A, using the lipid compositions described previously. The protein/lipid ratios used were 1:200 for synaptobrevin and syntaxin and 1:800 for synaptotagmin. A 3-5-fold excess of SNAP-25 (with respect to syntaxin-1A) and 3.5 mol % PIP 2 (unless noted otherwise) were added to the protein/lipid mixture for PM vesicles only; the excess ratio of SNAP-25 to syntaxin-1A essentially eliminates changes of formation of nonproductive 2:1 syntaxin-1A⅐SNAP-25 complexes (51). Dried lipid films were dissolved in 110 mM OG buffer containing purified proteins. Detergent-free buffer (20 mM HEPES, pH 7.4, 90 mM NaCl, 0.1% 2-mercaptoethanol) was then added to the protein-lipid mixture until the detergent concentration reached the critical micelle concentration of 24.4 mM. The vesicles were then subjected to size exclusion chromatography using a Sepharose CL-4B column, packed under near constant pressure by gravity with a peristaltic pump (GE Healthcare) in a 5-ml column with a 2-ml bed volume that was equilibrated with buffer V (20 mM HEPES, pH 7.4, 90 mM NaCl, 20 M EGTA, 0.1% 2-mercaptoethanol) followed by dialysis into 2 liters of detergent-free buffer V supplemented with 5 g of Bio-beads SM2 and 0.8 g/liter Chelex 100 resin. After 4 h, the buffer was changed with 2 liters of fresh buffer V containing Bio-beads and Chelex, and dialysis continued for 12 h. During the preparation of SV vesicles, 50 mM sulforhodamine B (Invitrogen) was present in all solutions prior to the size exclusion chromatography step. As described previously (51), the presence and purity of reconstituted proteins were confirmed by SDS-PAGE of the vesicle preparations, and the directionality of the membrane proteins (facing outward) was assessed by chymotrypsin digestion followed by SDS-PAGE. The size distributions of the SV and PM vesicles were analyzed by cryo-EM, as described previously (45).

Single Vesicle-Vesicle Content-mixing Assay
After a surface passivation with polyethylene glycol (PEG) molecules to eliminate nonspecific binding of vesicles, a #1.5 glass coverslip (VWR International) was assembled into a flow chamber and coated with neutravidin (0.1 mg/ml) (50). PM vesicles at a 100ϫ dilution were immobilized on the PEG-coated surface for a 30-min incubation period in buffer V in the presence or absence of 1 M Munc18a (or mutant Munc18a(S306E/S313E)) at room temperature (ϳ25°C). This incubation period was followed by three rounds of 150-l buffer washes with buffer V along with Munc18a if it was present in the previous stage. SV vesicles were loaded into the flow chamber in buffer V along with 2 M complexin-1 and Munc18a if it was present in the previous stage. For this loading step, the SV vesicle concentrations were adjusted to account for reduction of PM vesicle association when Munc18a was present during the incubation period; the PM vesicle stock solution was diluted 1,000 times in the absence of Munc18/Munc18a(S306E/S313E) or diluted 100 times in the presence of Munc18a or diluted 300 times in the presence of Munc18a(S306E/S313E). The system was incubated for 30 min, followed by three rounds of washes with buffer V supplemented with 2 M complexin-1, and 1 M Munc18a if it was present in the previous stage. Buffer V, supplemented with 500 M Ca 2ϩ , 2 M complexin, and 1 M Munc18a if it was present in the previous stage, was injected into the flow chamber at a speed of 2 ml/min by a motorized syringe pump. The TTL signal feeding to the syringe pump was delayed by a fixed time with respect to the start of the movie recording by a 9520 digital delay pulse generator (Quantum Composers, Bozeman, MT). Prior to Ca 2ϩ injection, we monitored spontaneous fusion events for a period of 1 min in the same channel that was also used for Ca 2ϩ injection. Real-time traces with at least one florescent jump after the triggering were analyzed. The details of automated data analysis were reported in previous work (52). Histograms of Ca 2ϩ -triggered fusion events were calculated with a 1-s time bin and fitted to a two-exponential decay function. The faster time constant of the fitted two-exponential decay function is defined as the synchronization time constant of Ca 2ϩ -triggered fusion.

Single Vesicle-Vesicle Association Experiments
This assay is similar to the single vesicle-vesicle content-mixing assay described above with some modifications in order to enable higher throughput. We followed a protocol for studying the association of single free SV vesicles with immobilized PM vesicles described previously (53,54). A saturated layer of unlabeled PM vesicles was immobilized on an imaging surface via biotin/neutravidin interactions. Specifically, the PM vesicle solution was diluted 10 times with buffer V. 100 l of the diluted PM vesicle solution was injected into the sample chamber and incubated for 30 min, followed by buffer exchange (1 ϫ 200 l of buffer V) for 6 s. Next, 100 l of solution containing wild type or Munc18a mutant (1 M) was injected into the sample chamber and incubated for 30 min, followed by buffer exchange (1 ϫ 200 l of buffer V) for 6 s. Subsequently, 100 l of 500 -1,000ϫ diluted SV vesicle solution was injected into the flow chamber in the presence of 2 M complexin-1. After an incubation period of 30 min, unbound SV vesicles were removed by buffer exchange (2 ϫ 200 l of buffer V for ϳ20 s). 20 M EGTA was included to ensure the absence of free Ca 2ϩ ions. For the control with soluble synaptobrevin-2(1-96), an additional 30-min incubation period with 10 M synaptobrevin-2(1-96) was included between the Munc18a incubation and SV vesicle incubation stages.
For the preassociated vesicle experiment, the Munc18a and the SV vesicle incubations were reversed with the same conditions: SV vesicles preassociated with PM vesicles for 30 min and then incubated in the flow chamber with or without Munc18a for 30 min. Associated SV vesicles were counted after one round of buffer wash (6 s) to remove any SV vesicles that may have become unbound by the action of Munc18a.
Sample slides with five channels were monitored in a wide field total internal reflection fluorescence microscope (Nikon) using an electron multiplying charge-coupled device (CCD) camera (iXonϩ DV 897E, Andor Technology). A program (smCamera) written in Cϩϩ was used for data acquisition and analysis (available from Taekjip Ha, University of Illinois). 10 images were taken at random locations within each channel on the quartz slide. Details regarding software, slide assembly, and imaging protocols are as described previously (55).
As described in our previous work (53,54), the PM vesiclecovered surfaces were saturated and produced a homogeneous distribution for each surface preparation with red laser excitation (633 nm) of the DiD-labeled immobilized vesicles. Our preparation of a reproducible, homogeneous, and saturated surface of immobilized PM vesicles ensures that the number of associated SV vesicles is directly related to the association probability. To minimize variation among different slides, for each set of comparisons between different conditions and/or mutants, the same protein preparations and surface preparations (quartz slide with immobilized vesicles) and incubation times were used, and the experiments were carried out in separate channels on the same slide. We found that the relative differences and ratios were statistically similar for different protein and surface preparations.

Single Molecule SNAP-25 Experiments
In order to probe release of SNAP-25 from t-SNARE complex by Munc18a, PM vesicles were reconstituted with syntaxin-1A and a 100:1 mixture of unlabeled and labeled SNAP-25. For labeled SNAP-25, the protein was mutated at position 76 (K76C) and labeled with Alexa647. The reconstitution was performed using the same protocol as used for all other single vesicle experiments in this work. Likewise, the PM vesicles with the mixture of labeled and unlabeled SNAP-25 molecules were immobilized on an imaging surface via biotin/neutravidin interactions with the same protocol as used for all other single vesicle experiments in this work. Fluorescence intensity from single vesicles was monitored before and after the Munc18a addition; specifically, 1 M Munc18a was injected into the chamber and incubated for 120 min, followed by buffer exchange (3 ϫ 200 l of vesicle buffer for ϳ20 s). A slide with multiple channels was monitored in a wide field total internal reflection fluorescence microscope, and data were acquired with a CCD camera and analyzed with the smCamera program, similar to the single vesicle-vesicle association experiments described above. 10 images were taken at random locations within each channel on the quartz slide. The number of vesicles decorated with Alexa647-labeled SNAP-25 was determined by counting the number of Alexa647 fluorescent spots from emission of Alexa Fluor upon excitation at 633 nm before and after the Munc18a addition.

Pull-down Assay
Purified Stx1a(1-267) and Stx1a(25-267) were incubated with GST or GST-SNAP-25 in phosphate-buffered saline (PBS) solution (20 mM Na 2 HPO 4 /NaH 2 PO 4 , pH 7.4, 150 mM NaCl, 1 mM DTT) overnight at 4°C and then incubated with Munc18a for 1.5 h and subsequently combined with equilibrated glutathione-agarose beads (final protein concentration ϳ7 M) and incubated for an additional 1.5 h at 4°C. Samples were spun at 850 relative centrifugal force for 5 min at 4°C, followed by separation of supernatant and bead fractions. The latter were washed twice with 0.5 ml of PBS (spun at 850 relative centrifugal force for 5 min at room temperature), combined with SDS, boiled, and analyzed by SDS-PAGE.

Isothermal Titration Calorimetry
ITC experiments were conducted using a VP-ITC instrument (Microcal) at 25°C. Samples were dialyzed against 1 liter of PBS overnight at 4°C, filtered, and degassed prior to each run. Munc18a (cell) concentrations ranged from 9 to 11 M, t-SNARE and SNARE complex (syringe) concentrations ranged from 100 to 180 M, and titration injection volumes varied from 7 to 10 l, 35-45 injections total, with 8 -24 s/injection and 4 min between injections. Microcal Origin version 7.0 software was used to integrate and analyze the measured heat released on binding. Fitting the data to a single site binding model gave the equilibrium association constant K a , the enthalpy of binding ⌬H, and the stoichiometry n.

Fluorescent Labeling of the Soluble Synaptobrevin-2 Fragment
Purified single cysteine soluble fragment of synaptobrevin-2(1-96) (S79C) was dialyzed against labeling buffer (20 mM HEPES, pH 7.0, 150 mM NaCl) overnight at 4°C and subsequently incubated with excess Alexa488-maleimide (Life Technologies) for a molar ratio of 1:2 (synaptobrevin-2(1-96)/Alexa488) in 10 mM TCEP for 24 h at 4°C. Labeled protein (synaptobrevin-2(1-96)-A488) was subsequently dialyzed against 20 mM MES, pH 6.0, 75 mM NaCl, 1 mM EDTA, and 1 mM DTT for 2 h at 4°C and loaded on a MonoS column (GE Healthcare). Fractions containing labeled protein were pooled and concentrated. Pre-formed Munc18a⅐t-SNARE complexes were purified on an S200 gel filtration column. All proteins were in excess of synaptobrevin-2(1-96)-A488 (50 nM). BSA (0.1-0.6 mg/ml) was used to minimize nonspecific binding of synaptobrevin-2(1-96)-A488 to the plastic well walls. Proteins were incubated separately and/or together in wells for 0.5-2 h at room temperature, followed by a 10-min "warm-up" run in the plate reader at 25°C. The final component (synaptobrevin-2(1-96)-A488 or t-SNARE(s)) was added to start the reaction. Fluorescence anisotropy measurements were recorded every 20 -30 s for 40 -50 min. Anisotropy (r) was calculated using the equation, r ϭ (I VV Ϫ GI VH )/(I VV ϩ 2GI VH ), where G ϭ I HV /I HH , I is fluorescence intensity, the first subscript letter represents the direction of excitation light, and the second subscript letter represents the direction of emission light for horizontally (H) and vertically (V) polarized light. Baselines were subtracted from all spectra, and data were analyzed using Prism version 5.0 (GraphPad Software). This anisotropy assembly assay was validated in Ref. 31, where it was shown that tagged synaptobrevin-2 assembles into an SDS-resistant complex, which is the gold standard for an assembled, stable neuronal SNARE complex.

Munc18a Has No Effect on Fusion
Kinetics-We used our single vesicle-vesicle content-mixing assay to study both spontaneous and Ca 2ϩ -triggered fusion (50, 51) with modifications to protein preparation, reconstitution, and single vesicle imaging as described under "Experimental Procedures." Syntaxin-1A and SNAP-25 were reconstituted in PM vesicles decorated with biotin-PE lipids for surface immobilization. Synaptobrevin-2 and synaptotagmin-1 were reconstituted in "SV" vesicles that contained encapsulated sulforhodamine B as a content indicator. PM vesicles were then immobilized on the surface of a flow chamber and incubated with or without Munc18a for 30 min. SV vesicle solution was then injected into the flow chamber in the presence of complexin-1; Munc18a was included if it had been present in the previous incubation stage. Ca 2ϩ solution was then injected into the sample chamber using the same buffer, which included complexin-1 and/or Munc18a if it had been present in the previous stage. Prior to Ca 2ϩ injection, spontaneous fusion events were monitored for a 1-min period. Upon Ca 2ϩ injection, fusion events were monitored, and time delays of fusion events with respect to the time point of Ca 2ϩ injection were plotted as binned histograms (Fig. 1, A-F).
Munc18a had no effect on the normalized Ca 2ϩ -triggered fusion histograms (Fig. 1, A-F), the synchronization time constant (Fig. 1G), or the probability of spontaneous fusion (Fig.  1H), regardless of whether full-length Stx1a or the N-terminally truncated Stx1a(10 -288) fragment were used. Moreover, there was no effect if Munc18a was incubated concurrently when adding free SV vesicles to the immobilized PM vesicles (Fig. 2). We note that the histograms were normalized with respect to the total number of fusion events, so they measure the effect of Munc18a on the intrinsic spontaneous and Ca 2ϩ -triggered fusion kinetics of associated vesicle-vesicle pairs. The ability to measure intrinsic fusion kinetics is an advantage compared with ensemble fusion experiments, where effects on docking and fusion cannot be separated (21).
Munc18a Binds to the Binary t-SNARE Complex-Our vesicle fusion assay uses PM vesicles that bear assembled Stx1a⅐SNAP-25 binary (t-SNARE) complexes, so we asked whether Munc18a could bind to the t-SNARE complex. We used ITC to measure the binding thermodynamics of Munc18a to the soluble t-SNARE complex. For direct comparison, we carried out ITC measurements of Munc18a-binding to the soluble ternary neuronal SNARE complex, as described previously (31). Control experiments in which t-SNAREs or SNARE complex were injected into buffer confirm that both complexes remain intact over the course of an ITC run (i.e. no dissociation occurred upon dilution). Remarkably, the binding energetics of Munc18a to the t-SNARE complex and to the ternary SNARE complex are very similar (Fig.  3 and Table 1). However, the affinity for both complexes is much weaker than for syntaxin-1A alone (K d of 1.4 nM (31, 56)).
Because binding of Munc18a to the ternary complex depends upon the N-terminal residues of Stx1a, we tested whether this is also true for the binary complex. Stx1a  or the N-terminally truncated mutant Stx1a(25-267) was incubated with GST-SNAP-25 overnight, and the resulting t-SNARE complex was then mixed with Munc18a and glutathione-agarose. GST-SNAP-25 bound Munc18a in the presence of Stx1a(1-267) but not Stx1a(25-267) (Fig. 4), indicating that binding of Munc18a to the t-SNARE complex is dependent on the conserved N-terminal residues of syntaxin-1A.
In all three cases, we found no significant difference in the amount of neuronal SNARE complex formed or the rate of complex formation (Fig. 5). In the case of sequentially added Munc18a, the protein concentrations correspond to Munc18a⅐t-SNARE binding occupancies of 20 -80%; we were not able to achieve saturated binding due to the insolubility of Munc18a at higher concentrations. However, the results were no different starting from fully saturated complexes with a 1:1:1 molar ratio of Munc18a⅐Stx1a(1-267)⅐SNAP-25. In summary, Munc18a has no detectable effect on the rate of cytoplasmic, soluble SNARE complex formation starting from the soluble, non-membraneanchored t-SNARE complex, and Munc18a binds to the t-SNARE complex with similar affinity as to the ternary t-SNARE complex. APRIL 17, 2015 • VOLUME 290 • NUMBER 16

JOURNAL OF BIOLOGICAL CHEMISTRY 10525
Munc18a Reduces SV-PM Vesicle Association-Although we did not observe an effect of Munc18a on the Ca 2ϩ -triggered fusion histograms and spontaneous fusion probabilities (Fig. 1), we found that the number of associated SV vesicles was reduced when Munc18a had been preincubated with the immobilized PM vesicles; in fact, we compensated for this reduction in associated SV vesicles by increasing the SV vesicle concentration accordingly compared with control without Munc18a (see "Experimental Procedures"). We studied this effect quantitatively by using a single vesicle-vesicle association assay that enabled higher throughput (53,54). Similar to the fusion experiments discussed above, immobilized PM vesicles were preincubated with Munc18a before SV vesicles along with complexin-1 (see "Experimental Procedures" for details). The number of associated SV vesicles per field of view was counted and used as a measure of association efficiency between pairs of free SV and immobilized PM vesicles. For a series of conditions, the same surface preparation and immobilized PM vesicles were used, and the free SV vesicle concentrations and incuba-tion times were kept the same as well. Because experiments with different protein and surface preparations cannot be compared directly, we plotted the results as fractional probabilities relative to the respective controls without Munc18a for each series of experiments that used the same surface and PM vesicle preparation (Fig. 7).
Munc18a reduced the number of associated SV-PM vesicles compared with the case without Munc18a (Fig. 7A). As a control, when the cytoplasmic domain of synaptobrevin-2 (synaptobrevin-2(1-96)) was added to the immobilized PM vesicles at 10 M concentration, SV-PM vesicle association was essentially abolished regardless of the presence of Munc18a (Fig. 7A). This experiment confirms that SV-PM vesicle association depends on trans-SNARE complex formation because the soluble synaptobrevin-2(1-96) fragment blocks trans-SNARE complex formation by sequestering t-SNARE (syntaxin-1A⅐SNAP-25) complexes.
A larger reduction by Munc18a on vesicle association was observed when SV vesicles were reconstituted with only synap- tobrevin-2 (i.e. without synaptotagmin-1) (Fig. 7E). Thus, the presence of synaptotagmin-1 masks part of the effect of Munc18a on SV-PM vesicle association. One possible explanation could be that the interaction between reconstituted synaptotagmin-1 in the SV vesicles and the membrane of the immobilized PM could be responsible. Synaptotagmin-1 interacts with membranes containing PIP 2 even in the absence of Ca 2ϩ (57,58), so we performed an experiment without PIP 2 in the PM vesicle preparation. However, Munc18a reduced vesicle association to a similar degree (Fig. 7B) as that observed in the presence of PIP 2 , implying that the effect by Munc18 on vesicle association is not caused by interference of the synaptotagmin-PIP 2 interaction. Finally, Munc18a had no effect on SV-PM vesicle association when it was added after vesicles had been associated by trans-SNARE complex formation (Fig. 7C).
Munc18a Sequesters Syntaxin-1A from Membrane-reconstituted t-SNARE Complex-The reduction of SV-PM vesicle association by Munc18a was similar when the N-terminally truncated Stx1a(10 -288) construct was used (Fig. 7, D and F). Because the observed reduction of SV-PM vesicle association does not depend on the N terminus of syntaxin-1A, yet the interaction between Munc18a and the t-SNARE complex depends on it (Fig. 4), the effect of Munc18a cannot be due to its interaction with the t-SNARE complex. Thus, we asked whether the observed reduction in SV-PM vesicle association could instead be caused by a sequestration reaction where Munc18a would capture closed syntaxin-1A and, thus, release SNAP-25.
To test this hypothesis, we labeled SNAP-25 at residue 76 and reconstituted a mixture of labeled and unlabeled SNAP-25 molecules into PM vesicles together with unlabeled syntaxin-1A. The same reconstitution and immobilization protocols were performed as for all other single vesicle experiments in this work. We adjusted the ratio of labeled to unlabeled SNAP-25 molecules in order to obtain at most one labeled SNAP-25 molecule per PM vesicle. Note that we used a soluble SNAP-25 construct (i.e. without palmitoylation), so once a SNAP-25 molecule was dislodged from a t-SNARE complex it was expected to diffuse away from the surface. We monitored fluorescence from single SNAP-25 molecules before and 120 min after the Munc18a addition (Fig. 8). Consistent with our hypothesis of syntaxin-1A sequestration, we observed that incubation of Munc18a results in reduction of about 50% of surface-localized SNAP-25 molecules.
Effects of Reducing the Munc18a-Stx1a Interaction by Phosphorylation-Phosphorylation of Munc18a greatly reduces interaction with syntaxin-1A (41), and a phosphorylation-mimicking mutant Unc-18(S322E) of Caenorhabditis elegans Unc-18 (S322E) reduces binding to syntaxin-1A. However, the role of phosphorylation in neurons was found to be relatively subtle; the two phosphorylation sites in Munc18a appear to be important for post-tetanic potentiation (59). In our system, similar to wild type Munc18a, the phosphorylation mimic mutant of Munc18a(S306E/S313E) had no effect on the normalized Ca 2ϩtriggered fusion histograms (Fig. 1, A-F), the synchronization time constant (Fig. 1G), or the probability of spontaneous fusion (Fig.  1H), regardless of whether full-length Stx1a or the N-terminally truncated Stx1a(10 -288) fragment were used. The phosphorylation mimic Munc18a(S306E/S313E) also reduced vesicle association but to a lesser degree (Fig. 7A). The lessened effect on SV-PM vesicle association by the phosphorylation mimic is correlated with its reduced affinity to syntaxin-1A compared with wild type Munc18a, as determined by bio-layer interferometry experiments (Fig. 9). Thus, these experiments further corroborate that the sequestration of syntaxin-1A by Munc18a is related to the interaction between Munc18a and syntaxin-1A.
The reduction of SV-PM vesicle association by Munc18a(S306E/S313E) was similar when the N-terminally truncated syntaxin-1A construct Stx1a(10 -288) was used (Fig.  7, D and F). Moreover, the affinity of the phosphorylation mimic Munc18a(S306E/S313E) to the N-terminally truncated syntaxin-1A construct was reduced relative to wild-type syntaxin-1A (Fig. 9). This suggests that the syntaxin-1A sequestration reaction by Munc18a(S306E/S313E) is independent of the presence of the N-terminal residues of syntaxin-1A, as concluded above for wild type Munc18a.

DISCUSSION
Our single vesicle-vesicle content-mixing assay enabled us to discriminate between effects caused by vesicle association and intrinsic fusion rates, thus overcoming a key deficiency of previous ensemble fluorescence methods, where both effects are convoluted (21). Using our single vesicle-vesicle content-mixing assay, we found that Munc18a has no effect on both spontaneous and Ca 2ϩ -triggered fusion rates with neuronal SNAREs, complexin-1, and synaptotagmin-1 (Fig. 1). Role of the Syntaxin-1A N Terminus-The N terminus of syntaxin-1A plays only a small role in the binding affinity between Munc18a and closed syntaxin-1A (30,31). In contrast, the N terminus is essential for the interaction between Munc18a and ternary neuronal SNARE complex (31); for the interaction between Munc18a and t-SNARE complex (Fig. 4); and for colocalization of Munc18a, SNAP-25, and syntaxin-1A (32).
Although the N terminus is required for recruitment of Munc18, its function could be physically separated in neuromuscular junctions of C. elegans by using "split" syntaxin transgenes (18). Moreover, point mutations of Munc18a that interfere with binding to the syntaxin-1A N terminus and disrupt binding to the assembled SNARE complexes all support normal docking, priming, and fusion of synaptic vesicles as well as normal synaptic plasticity in rescue experiments with neuronal cultures using autaptic munc18a null mutant neurons (60). Consistent with these previous results, in our experiments, the syntaxin-1A N terminus had no effect on fusion rates (Fig. 1) and vesicle association (Fig. 7).
Munc18a Sequesters Syntaxin-1A from Reconstituted t-SNARE Complex-When we added Munc18a to immobilized PM vesicles with reconstituted syntaxin-1A and SNAP-25, we observed reduction of SV-PM vesicle association on a ϳ30-min time scale (Fig. 7). This result suggested that Munc18a reduces vesicle association by sequestration of syntaxin-1A (i.e. releasing SNAP-25 from t-SNARE complex). We independently con-  Table 1.

TABLE 1 Thermodynamic parameters of Munc18a interaction with t-SNARE and ternary SNARE complexes
Our results differ slightly from previously published ITC data for Munc18a-binding SNARE complex (see supplemental Table S1 in Ref. 31), which probably reflects differences in the protein constructs and preparations.

Munc18a Does Not Alter Fusion Rates
firmed the sequestration of syntaxin-1A by Munc18a by a single molecule experiment (Fig. 8). Moreover, the phosphorylation mimic mutant of Munc18a with reduced affinity to syntaxin-1A resulted in less reduction of vesicle association (Fig. 7A), corroborating the notion of syntaxin-1A sequestration. The interaction between Munc18a and syntaxin-1A lacking the N-peptide is strong (K d ϭ 10 nM) albeit reduced compared with full-length syntaxin-1A (K d ϭ 1.4 nM) (31). Moreover, the crystal structure and solution scattering analysis of the complex between syntaxin-1A and Munc18a is similar in the presence or absence of the syntaxin N-terminal residues (56). In both cases, Munc18a is expected to sequester syntaxin-1A and interact with it in a closed conformation, consistent with our results.
Sequestration of the cytoplasmic, soluble fragment of syntaxin-1A from t-SNARE complex by Munc18a was previously observed by nuclear magnetic resonance spectroscopy on a 11-h time scale (Fig. 1 in Ref. 8) and by a FRET-based assay on a ϳ30-min time scale using separate SNAP-25 N-and C-terminal SNARE helices (Fig. S1B in Ref. 8). Thus, there is considerable variation of the time scale of syntaxin-1A sequestration depending on the constructs and experimental conditions; this probably explains why there is no apparent sequestration of syntaxin-1A (which would produce much larger enthalpy changes and a smaller K d ) (56) on the ϳ1-h time scale of our ITC experiment (Fig. 3).
A Possible Molecular Mechanism for the Sequestration of Syntaxin-1A by Munc18a-Previously, the conformations of 1:1 binary (syntaxin-1A⅐SNAP-25) complex were studied by single molecule FRET experiments using binary complexes that were reconstituted into supported bilayers (61). These experiments revealed three configurations of the 1:1 syntaxin-1A⅐SNAP-25 binary complex; one configuration corresponds to a parallel three-helix bundle, whereas the other two are flexible configurations with one of the SNAP-25 SNARE domains dissociated. Occasional transitions between these configurations were observed. Adding synaptobrevin-2 entirely suppressed the flexible configurations. The addition of complexin-1, the MUN domain of Munc13, Munc18a, or synaptotagmin-1, also reduced the flexible configurations to varying degrees. The sin-  gle molecule experiments with labeled SNAP-25 presented here (Fig. 8) suggest that Munc18a is capable of sequestering a fraction of syntaxin-1A molecules because we observed a loss of surface-localized SNAP-25 upon incubation with Munc18a. Despite the prolonged incubation with Munc18a, this sequestration of syntaxin-1A molecules was only partial. Combined with the results presented in Ref. 61, we suggest that sequestration of syntaxin-1A by Munc18a occurs only for the fraction of 1:1 syntaxin-1A⅐SNAP-25 complexes that are in the flexible configurations during the incubation period; presumably, the kinetic barrier to sequester syntaxin-1A by Munc18a is lower for the flexible configurations.
Is There an Interaction between Munc18a and Synaptobrevin-2?-Previously, it had been suggested that Munc18a interacts with synaptobrevin-2 based on ensemble lipid-mixing assays (17) and line broadening in solution NMR experiments (62). The lack of a significant difference in affinity between binary and ternary neuronal SNARE complex (Fig. 3) or an effect on fusion rates (Fig. 1) is consistent with the very weak (K D ϳ15 M) affinity reported for the Munc18a-synaptobrevin interaction (62). Because our experiments were performed well below the reported K D of 15 M for this interaction, if the intracellular Munc18a concentration were substantially higher, it is possible that contacts between Munc18a and synaptobrevin-2 could have some role in fusion.
Effect of Munc18 on Vesicle Association Is Independent of PIP 2 -PIP 2 is a charged lipid that is enriched in the plasma membrane and may promote synaptic vesicle recruitment to the active zone (63). When PIP 2 was excluded from PM vesicles in our vesicle association assay, the overall association efficiency was greatly reduced, such that a much higher SV vesicle concentration had to be used in order to obtain a significant number of associated SV-PM vesicle pairs. However, the ratio of vesicle association with and without Munc18a was unchanged (Fig. 7, compare A and B), illustrating that the effect of Munc18a on vesicle association is PIP 2independent. Synaptotagmin-1 interacts with PIP 2 -containing membranes even in the absence of Ca 2ϩ (57) and is thus expected to increase association between SV and PM vesicles, as we observed. When synaptotagmin-1 was eliminated from SV vesicles, the same fractional vesicle associations were observed for control, Munc18a, and Munc18a(S306E/S313E) (Fig. 7, compare A and E), consistent with the notion that the effect of Munc18a in these experiments is to sequester syntaxin-1A and release SNAP-25. However, in the absence of synaptotagmin-1, the reduction of SV-PM vesicle association by Munc18a was more pronounced, suggesting that synaptotagmin-1 interactions with t-SNARE (syntaxin-1A⅐SNAP-25) complexes (64) can compensate the partial reduction by Munc18a; as mentioned above, this interaction also stabilizes t-SNARE complexes (65). Lou et al. (71) shows that Munc18a stimulates single vesicle-vesicle lipid mixing with neuronal SNAREs in the absence of synaptotagmin-1 but that there is no stimulatory effect on single vesicle lipid mixing by Munc18a in the presence of neuronal SNAREs and synaptotagmin-1. Moreover, their experiments with SNAREs that were reconstituted in nanodiscs offer an explanation for the enhancement of lipid mixing: Munc18a increases a state of the trans-SNARE complex where the C-terminal ends of the SNARE domains are close to each other in the absence of synaptotagmin-1, whereas in the presence of synaptotagmin-1, Munc18 has no effect on the states of the trans-SNARE complex (see Fig. 1, C and D, in Ref. 71). Their results suggest that the observed increase in lipid mixing is a consequence of increased vesicle association that occurs when Munc18a increases the C-terminal zippering probability of the neuronal trans-SNARE complex. Although we did not see an effect on soluble ternary SNARE complex formation by Munc18 (Figs. 5 and 6), we can reconcile these observations by noting that assembly of the neuronal SNARE complex is paused under loaded conditions (i.e. when the repul-sive forces between solvated bilayers oppose trans-SNARE complex formation) (66,67). Thus, the C-terminal zippering probability is affected by the presence of the reconstituted transmembrane domains, and in the absence of the transmembrane domains, one would not expect to see the intermediate state observed by Lou et al. (71), consistent with our results.

Munc18a Has No Effect on Spontaneous and Ca 2ϩ -triggered Fusion-Recent work by
Parisotto et al. (20) showed that spontaneous ensemble lipid mixing (i.e. at zero Ca 2ϩ ) is increased in the absence of complexin-1, whereas this effect is smaller in the presence of complexin-1 (see Fig. 3, A and B, in Ref. 20). However, the subsequent further increase in ensemble lipid mixing upon Ca 2ϩ injection is similar, regardless of the presence of Munc18a. Moreover, Fig. 3C in Ref. 20 shows that the subsequent increase in ensemble content mixing upon Ca 2ϩ injection is also unaffected by the presence or absence of Munc18a. Thus, the lack of an effect on Ca 2ϩ triggered ensemble content mixing is entirely in agreement with our single vesicle-vesicle Ca 2ϩ -triggered fusion results (Fig. 1, A-C  We emphasize that the data of Parisotto et al. (20) showing an increase in ensemble lipid mixing by Munc18a was not normalized by the number of associated/docked vesicles, so it can be explained by an increase in vesicle association (i.e. more  vesicles are associated, resulting in a larger signal) rather than an effect on hemifusion or fusion per se. This explanation also applies to other previously published results (16,68). Moreover, we speculate that the apparent difference in the degree of ensemble lipid mixing can be explained by the competition of the two processes, the enhancement of neuronal SNARE zippering in the absence of synaptotagmin-1 versus the sequestration of syntaxin-1A by Munc18a; different experimental conditions may shift the balance between these two processes. For example, the requirement of preincubation of vesicles at 4°C for the ensemble lipid mixing experiments by Shen et al. (17) could arise from shifting the balance between Munc18a-enhanced SNARE zippering versus Munc18a-induced sequestration of syntaxin-1A.
NSF and ␣SNAP will disassemble the t-SNARE complex (8); other experiments suggest that Munc18a then prevents further action of NSF/␣SNAP by sequestering syntaxin-1A and locking it into the closed confirmation and preventing trans-SNARE formation until synaptobrevin-2 and Munc13 arrive (8). We note that the amount of ensemble lipid mixing was roughly the same for a minimal system consisting only of neuronal SNAREs, the C2AB fragment of synaptotagmin-1, and 0.5 M Ca 2ϩ and for the more complete system consisting of the same proteins plus NSF, ␣SNAP, Munc18a, and Munc13 (see Fig. 4D in Ref. 8). One of the reasons for our study was to investigate whether the ensemble lipid-mixing method may have masked an effect of Munc18a on content mixing. Our results now conclusively show that Munc18a has no effect on intrinsic spontaneous and Ca 2ϩ -triggered fusion rates in conjunction with neuronal SNAREs, synaptotagmin-1, and complexin-1. Our study was enabled by our single vesicle-vesicle fusion system, where effects on vesicle association can be separated from effects on spontaneous and Ca 2ϩ -triggered fusion (40,45,51). Südhof (69) noted in his 2013 Nobel Lecture, "The fact that SM proteins are required continuously during SNARE-complex assembly argues for a role either in organizing proper SNARE-complex assembly and in preventing dead-end inappropriate SNARE complexes, or in catalyzing lipid mixing during fusion. At present, no conclusive data argue one way or the other, and this question will clearly keep many of us busy for years to come." Because our results show that Munc18a has no effect on fusion in conjunction with neuronal SNAREs, synaptotagmin-1, and complexin-1, the following consensus view emerges. Munc18a primarily plays a role in proofreading of trans-SNARE complexes rather than directly in fusion itself. This notion has also been suggested for the SM protein-containing HOPS complex in the context of vacuolar fusion (27). However, future reconstitution experiments with all factors involved in synaptic vesicle fusion will have to be carried out in order to study the proposed proofreading role of Munc18a.