• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of pnasPNASInfo for AuthorsSubscriptionsAboutThis Article
Proc Natl Acad Sci U S A. Aug 25, 2009; 106(34): 14303–14308.
Published online Aug 10, 2009. doi:  10.1073/pnas.0902976106
PMCID: PMC2732825

The N-terminal peptide of the syntaxin Tlg2p modulates binding of its closed conformation to Vps45p


The Sec1/Munc18 (SM) protein family regulates intracellular trafficking through interactions with individual SNARE proteins and assembled SNARE complexes. Revealing a common mechanism of this regulation has been challenging, largely because of the multiple modes of interaction observed between SM proteins and their cognate syntaxin-type SNAREs. These modes include binding of the SM to a closed conformation of syntaxin, binding to the N-terminal peptide of syntaxin, binding to assembled SNARE complexes, and/or binding to nonsyntaxin SNAREs. The SM protein Vps45p, which regulates endosomal trafficking in yeast, binds the conserved N-terminal peptide of the syntaxin Tlg2p. We used size exclusion chromatography and a quantitative fluorescent gel mobility shift assay to reveal an additional binding site that does not require the Tlg2p N-peptide. Characterization of Tlg2p mutants and truncations indicate that this binding site corresponds to a closed conformation of Tlg2p. Furthermore, the Tlg2p N-peptide competes with the closed conformation for binding, suggesting a fundamental regulatory mechanism for SM–syntaxin interactions in SNARE assembly and membrane fusion.

Keywords: Sec1/Munc18 protein, SNARE, membrane fusion

Eukaryotic cell growth and survival require membrane-bound vesicles to transport proteins and membrane between the various organelles within the cell and to the plasma membrane for secretion. These conserved vesicle trafficking mechanisms require exquisite regulation to ensure specificity (1, 2). Crucial components are the SNARE proteins, which form a parallel four-helix bundle called the SNARE complex to bridge the vesicle and target membranes for fusion (3, 4). The Sec1/Munc18 (SM) protein superfamily regulates SNARE complex assembly and membrane fusion through direct interactions with their cognate SNAREs (for review, see refs. 5 and 6). The SM family is divided into four subfamilies: Sec1p, Vps45p, Sly1p, and Vps33p. The Sec1p family is exocytic and includes yeast Sec1p and mammalian Munc18 isoforms; Munc18a is neuronal specific, whereas Munc18c functions in multiple cell types. The Vps45p family regulates endosomal trafficking, whereas the Sly1p family regulates trafficking between the ER and the Golgi, and Vps33p functions in trafficking to the vacuole/lysosome. The different SM proteins are structurally similar, consisting of three mixed α-helical and β-sheet domains arched around a central cleft, suggesting a conserved function (710).

Although SM proteins are thought to perform similar roles in the regulation of vesicle fusion, a wealth of conflicting data from SM homologues in different trafficking steps and species has made the mechanism of SM action unclear (for review, see ref. 5). SM proteins appear to have both positive and negative roles in vivo, which may reflect different aspects of their function in the SNARE assembly/disassembly cycle. Consistent with these different roles, SM proteins interact not only with individual SNAREs, but also with assembled SNARE complexes. Most of the SM proteins interact directly with syntaxins at one of two distinct binding sites, which are separated by >30 Å. One site is formed by the central cleft of Munc18a and the closed conformation of syntaxin 1a (Fig. 1A, mode 1; ref. 7). In contrast, other SMs bind individual syntaxins and SNARE complexes, using a hydrophobic pocket in domain 1 of the SM protein to bind the N-terminal peptide of the syntaxin (Fig. 1A, mode 2). These include Sly1p–Sed5p (9, 11), Vps45p–Tlg2p (12, 13), and Munc18c–syntaxin 4 (10). Thus, the N-peptide interaction appears to be the most common mode of binding, and the closed conformation interaction has been hypothesized to be specific to the neuronal Munc18a (6). Additionally, the yeast exocytic Sec1p binds assembled SNARE complexes in the absence of the syntaxin N-peptide (14); this mode may also be used by other SMs (13, 15). Moreover, several SM proteins have been shown to interact with nonsyntaxin SNAREs (13, 1618). The multiple distinct modes of interaction between SM proteins and SNARE partners suggested that the SMs may not share a common mode of action (5), although the high degree of sequence and structural similarity between them argues that SMs should have a common function in vesicular trafficking.

Fig. 1.
The SM protein Vps45p binds the syntaxin Tlg2p. (A) Representation of the SM-syntaxin binding modes. SM proteins bind to the closed syntaxin conformation (mode 1), the N-peptide of syntaxin (mode 2), or both sites of syntaxin (mode 3). Shading of the ...

Recent studies are beginning to resolve this quandary. Although the closed conformation seemed to be the predominant mode of the Munc18a–syntaxin 1a interaction, the syntaxin 1a N-peptide appears to also be important for the regulation of SNARE complex assembly, the interaction between Munc18a and assembled neuronal SNARE complexes, and stimulation of membrane fusion (15, 1922). The N-peptide alone is not sufficient for binding of syntaxin 1a to Munc18a, but these functional data suggested that Munc18a may bind the syntaxin 1a N-peptide similarly to other SM proteins. Moreover, when the original Munc18a-syntaxin 1a structure was re-examined, density corresponding to several residues of syntaxin 1a's N-peptide bound to Munc18a was observed, confirming that syntaxin 1a can bind Munc18a through dual modes (Fig. 1A, mode 3; ref. 22).

Here, we focus on yeast Vps45p, which interacts with the N-peptide of Tlg2p (Fig. 1A, mode 2). Mutational analyses had previously suggested that only this N-peptide binding site was important for the Vps45p–Tlg2p interaction (12, 13). Additionally, NMR studies of truncated Tlg2p had suggested that, unlike syntaxin 1a and Sso1p (23, 24), Tlg2p does not form a stable closed conformation (12). The above data, along with homology to other SM–syntaxin pairings, led to the hypothesis that the N-peptide of Tlg2p is solely responsible for the interaction between Vps45p and Tlg2p. However, the absence of a phenotype for mutants that disrupt binding between Vps45p and Tlg2p (13), the evidence for dual binding modes for Munc18a-syntaxin 1a, and recent data indicating that residues outside of the N-peptide of syntaxin 16 stabilize the binding of syntaxin 16 to mammalian Vps45 (22) led us to re-examine the possibility of additional interaction modes between Vps45p and Tlg2p. We discovered a second binding site for Vps45p on Tlg2p, which corresponds to a Tlg2p closed conformation (Fig. 1A, mode 1). We demonstrate that either of these Vps45p binding sites is sufficient for Tlg2p function in vivo, but abrogation of both results in a phenotype similar to tlg2Δ cells. Furthermore, we show that the N-peptide of Tlg2p modulates the affinity of the closed conformation binding site, indicating a role for the N-peptide in controlling accessibility of Tlg2p for SNARE complex assembly. These data suggest that a common mechanism for SM–syntaxin interactions is a dual mode, whereby the SM interacts with two distinct sites on the syntaxin, the N-peptide and the closed conformation.


Vps45p Interacts with Tlg2p in the Absence of the Tlg2p N-Peptide.

Various truncations of Tlg2p were created to examine the role that individual domains play in the interaction with Vps45p and were designed using secondary structure predictions, sequence alignments with other syntaxin homologues, and domains delineated by NMR experiments (12). The full-length cytosolic Tlg2p(1-318) protein contains: the N-peptide region; the Habc domain, which acts as an autoinhibitory domain in many syntaxin-type SNAREs; a short linker region; and the SNARE motif region, which is used for binding to the other SNARE proteins (Fig. 1B). We also purified wild-type Vps45p and Vps45p–L117R, a mutant with abrogated binding to the Tlg2p N-peptide (13). Each protein is monomeric and predominantly α-helical, as determined by size exclusion chromatography (SEC) and circular dichroism (CD) (Fig. S1 A and B). The exception is Tlg2p(221-318), which is predominantly unfolded, as expected for an isolated SNARE motif region (25, 26).

The interaction between the recombinant Vps45p and Tlg2p(1-318) proteins was investigated. Vps45p and Tlg2p(1-318) were incubated together, and SEC was used to separate the complex from the free proteins (Fig. S1C). The individual elution peaks of both Vps45p and Tlg2p(1-318) shift to a larger apparent molecular weight when these two proteins are mixed, indicating that Vps45p interacts with the cytosolic region of Tlg2p, corroborating previous studies (12, 13). We next demonstrated that a construct lacking the N-peptide, Tlg2p(37-318), also interacts with Vps45p (Fig. 2A). This result was surprising, because previous studies had indicated that Tlg2p lacking the N-peptide could not bind Vps45p (12, 13). Our results clearly demonstrate that Vps45p can interact with another binding site in the C-terminal region of Tlg2p.

Fig. 2.
The N-peptide of Tlg2p is not required for the Vps45p–Tlg2p interaction. (A) The interaction between Vps45p and Tlg2p(37-318) was analyzed by SEC as described in SI Text. Vps45p and Tlg2p(37-318) were incubated together and applied to a Superdex ...

To determine the apparent affinity between Vps45p and this second binding site, we developed a native EMSA using Tlg2p(37-318) protein fluorescently labeled with Alexa Fluor 488 dye [Tlg2p(37-318)*]. This well-established method has been widely used in the field of protein–nucleic acid interactions (27). First, increasing concentrations of Vps45p were equilibrated with a trace amount of Tlg2p(37-318)*. The reactions were electrophoresed in a horizontal 6% nondenaturing polyacrylamide gel to separate the bound from unbound Tlg2p(37-318)*, visualized with a Fuji FLA-5000 Fluor Imager, and analyzed as described in SI Text (Fig. 2B). Because dissociation of the complex during loading, electrophoresis and migration through the gel can occur (although this is somewhat mitigated by the caging effect of the polyacrylamide gel), our calculations are for an apparent affinity of each interaction. The apparent affinity (Kd,app) of Vps45p for Tlg2p(37-318)* is 150 ± 31 nM, with an apparent Hill coefficient (n) of 1.5 ± 0.2. The slightly elevated Hill coefficient could be caused by weak positive cooperativity in binding; however, this interpretation is inconsistent with our SEC data, and minor deviations from unity are common in binding experiments with limiting labeled material (27).

Additional evidence that the C-terminal binding site on Tlg2p does not require the N-peptide is demonstrated by the interaction between Tlg2p(37-318)* and the mutant Vps45p–L117R protein. We hypothesized that although the L117R mutation disrupts the ability of the N-peptide to interact with Vps45p, it should retain the Tlg2p(37-318) binding site. When measured by EMSA, the Kd,app = 200 ± 13 nM (n = 1.1 ± 0.1), indicating that the L117R mutation does not disrupt binding of Vps45p to the C-terminal binding site on Tlg2p. Collectively, these results describe a previously unidentified mode of interaction between Vps45p and Tlg2p, one that does not require the N-peptide of Tlg2p, nor the hydrophobic pocket on domain 1 of Vps45p.

Tlg2p N-Peptide Modulates Affinity of the C-Terminal Binding Site.

We used competition EMSA to analyze the ability of various Tlg2p truncations, and mutants thereof, to compete with Tlg2p(37-318)* for binding to Vps45p. In these experiments, the Vps45–Tlg2p(37-318)* mobility shift is used to probe the binding competency of unlabeled Tlg2p variants. In this assay, increasing concentrations of competitor protein are incubated with Tlg2p(37-318)* and Vps45p. If the competitor competes for binding to Vps45p, the fraction of bound Tlg2p(37-318)* decreases as the concentration of competitor increases. The apparent affinity (Kc,app) of the Vps45p–competitor interaction is calculated from a fit of the data to the Lin and Riggs equation (28).

To validate this competition assay, we tested the ability of unlabeled Tlg2p(37-318) to compete with Tlg2p(37-318)* for binding to Vps45p. The Kc,app of this interaction is 280 ± 4 nM (Table 1), which is less than a 2-fold difference from the Kd,app measured by direct titration. Tlg2p(1-318), which contains the N-peptide, competes for binding to Vps45p with a similar Kc,app (190 ± 10 nM; Fig. 3A and Table 1) to Tlg2p(37-318). This similar affinity was surprising, because data from Munc18a and mammalian Vps45 studies (22) led us to expect that the combination of both binding sites would lead to a considerably tighter apparent affinity for Tlg2(1-318).

Table 1.
Competition of Tlg2p truncations and mutants for Vps45p binding
Fig. 3.
The N-peptide of Tlg2p negatively affects the binding of Tlg2p(37-318). (A) The full-length cytosolic Tlg2p(1-318) protein competes with the Tlg2p(37-318)*–Vps45p interaction. Increasing concentrations of Tlg2p(1-318) were added to Vps45p–Tlg2p(37-318)*. ...

To elucidate the relationship between the two Tlg2p binding sites, we wanted to directly test the binding of the N-peptide to Vps45p; however, poor solubility of the labeled peptide precluded further study. Instead, we examined the unlabeled Tlg2p(1-33) peptide by competition EMSA. We did not expect the N-peptide to compete with Tlg2p(37-318)* for Vps45p binding, because the two binding sites on the SM proteins are not adjacent to each other, and because Munc18a appears to simultaneously bind both sites on syntaxin 1a (22). Unexpectedly, Tlg2p(1-33) strongly competed with Tlg2p(37-318) for binding to Vps45p (Kc,app = 35 ± 3 nM; Fig. 3B and Table 1). Several potential explanations could account for this tight Kc,app. The Tlg2p N-peptide could bind to the hydrophobic pocket on domain 1 of Vps45p and modulate the conformation of Vps45p such that it can no longer bind the C-terminal binding site on Tlg2p. Alternatively, the N-peptide could bind directly to Tlg2p(37-318)* and compete for Vps45p binding, or the N-peptide could bind to a different site on Vps45p that precludes binding of Tlg2p(37-318)*. To discriminate between these possibilities, we analyzed the ability of the N-peptide to compete for binding to the Vps45p–L117R mutant. We discovered that, even at concentrations as high as 1 μM, Tlg2p(1-33) was unable to compete with Tlg2p(37-318)* for binding to Vps45p–L117R. Thus, the N-peptide cannot compete for binding to Vps45p when its hydrophobic pocket binding site has been disrupted. These results suggest that the two binding sites are not independent; in fact, they appear to be mutually exclusive. This apparent negative allostery between the two binding sites likely explains the weaker apparent affinity observed for the cytoplasmic region of Tlg2p compared with the N-peptide alone, although steric hindrance and/or interactions between the N-peptide and the 37-318 region cannot currently be ruled out.

Our conclusions are supported by several critical point mutations. First, we examined an N-peptide mutant, Tlg2p(1-318)–F9A/L10A. The dual alanine mutations abolish the interaction with the hydrophobic pocket in Vps45p (12) and therefore should not affect binding at the C-terminal binding site. Accordingly, this mutant competes for binding to Vps45p with a Kc,app of 280 ± 23 nM, similar to the unlabeled Tlg2p(37-318) (Table 1). Next, we took advantage of the observation that residue I233 in syntaxin 1a directly contacts Munc18a in the crystal structure (7); mutation of this residue to Ala, or the equivalent I236 in the Drosophila syntaxin, abrogates mode 1 binding to their respective SM proteins (22, 29). We introduced the analogous mutation (I285A) into Tlg2p and found that I285A completely disrupts the ability of the truncated Tlg2p(37-318)–I285A to compete for Vps45p binding (Fig. 3C and Table 1). In contrast, the full-length Tlg2p(1-318)–I285A mutant competes for binding to Vps45p with a Kc,app of 250 ± 12 nM (Fig. 3C and Table 1). Thus, even though the I285A mutation blocks C-terminal binding, the presence of the N-peptide is sufficient to compete with the Tlg2p(37-318)*–Vps45p interaction. These results support our observation that the N-peptide can negatively affect binding at the C-terminal site.

Mutation of Both Binding Sites Abrogates the Function of Tlg2p in Vivo.

Previous studies indicated that the N-peptide binding mode is not required for the function of Sly1p or Vps45p in vivo (13, 16), although recent evidence indicates a role for the syntaxin N-peptide in Caenorhabditis elegans (30, 31). To determine the functional importance of the Tlg2p C-terminal binding site, we assessed the ability of Tlg2p containing the I285A mutation to complement the CPY trafficking defect of cells lacking endogenous Tlg2p. To measure the effect of the loss of one or both Vps45p binding sites on Tlg2p function, full-length TLG2 (encoding residues 1–397) and mutant tlg2 constructs were expressed in tlg2Δ cells producing a CPY-invertase fusion protein that is used to quantitatively measure the amount of CPY secreted (Fig. 4A) (32). Fig. 4B demonstrates that wild-type cells secrete ≈5% of the CPY-invertase fusion protein, reflecting the fact that the CPY fusion protein is properly sorted to the vacuole in these cells (32). In contrast, tlg2Δ cells secrete ≈25% of this fusion protein (Fig. 4B) (33). Production of either wild-type Tlg2p or the Tlg2p–F9A/L10A mutant complements the CPY-invertase secretion phenotype of tlg2Δ cells (Fig. 4B), consistent with our previous finding that abolition of the N-peptide mode of binding by the Vps45p–L117R mutant did not affect CPY sorting (13). Interestingly, mutation of Tlg2p at I285A, which disrupts the C-terminal binding mode in vitro, also rescued the CPY-invertase secretion phenotype (Fig. 4B), indicating that neither binding site individually is essential for Tlg2p function. However, expression of the triple mutant, Tlg2p–F9A/L10A/I285A, in which both binding sites are disrupted, does not rescue the CPY secretion defect (Fig. 4B). Although it is possible that the overexpression observed for the Tlg2p–F9A/L10A mutant could result in sufficient levels of functional Tlg2p, this is clearly not the case for Tlg2p-F9A/L10A/I285A. Thus, Tlg2p function in CPY secretion requires the presence of at least one of the two Vps45p binding sites.

Fig. 4.
Abrogation of the two individual binding modes is required to disrupt Tlg2p function. (A) The expression levels of wild-type Tlg2p or Tlg2p mutants were monitored in tlg2Δ cells. Tlg2p constructs were expressed from a plasmid and their presence ...

Vps45p Interacts with the Closed Conformation of Tlg2p.

We hypothesized that the C-terminal binding site is comprised of the Habc domain, linker, and SNARE motif regions folded into a closed conformation akin to that observed for other syntaxins (Fig. 1A, mode 1). To test this, Tlg2p(37-192) containing the Habc domain and Tlg2p(221-318) containing most of the linker and the SNARE motif (Fig. 1B) were incubated separately with Vps45p and their binding was analyzed by SEC and EMSA. Neither region is able to directly bind to Vps45p, nor compete with Tlg2p(37-318)* for binding to Vps45p (Fig. S2 and Table 1). Moreover, one prediction of the closed conformation model is that residues critical for the Munc18a-closed syntaxin interaction would also be important for Vps45p–Tlg2p(37-318). Results from the I285A mutants support this idea. The I285 residue is present in the SNARE motif region of Tlg2p, and, by analogy to the syntaxin 1a-Munc18a structure, should contact Vps45p when Tlg2p is closed. We found that the I285A mutation abrogated binding to the C-terminal binding site of Tlg2p (Fig. 3C), indicating that Tlg2p is bound to Vps45p in a conformation structurally similar to closed syntaxin 1a.

To directly probe the closed conformation of Tlg2p, we made mutations in the Habc domain designed to destabilize the closed conformation (Fig. 5A). The closest structural homologues to Tlg2p were used to create a homology model for the Habc domain (see SI Text). Using this model, and our previous “open” Sso1p mutants as guides (23, 34), we designed and created putative open Tlg2p mutants. The triple alanine mutant Tlg2p (37-318)–K134A/K137A/K163A shows several characteristics consistent with an open conformation, which would be caused by the SNARE motif residues no longer packing against the Habc domain. The Tlg2p(37-318) mutant's α-helical secondary structure is decreased ≈20% and its apparent molecular weight, as monitored by SEC, is increased when compared with wild-type Tlg2p(37-318) (Fig. 5 B and C). Supporting our hypothesis that Vps45p interacts with the Tlg2p closed conformation, this mutant is unable to compete for binding to Vps45p (Fig. 5D).

Fig. 5.
The Tlg2p(37-318) K134A/K137A/K163A open mutant. (A) Homology model of Tlg2p Habc domain (amino acids 71–193) with the K134A/K137A/K163A mutations indicated. Molecular graphics were generated with PyMOL (http://pymol.sourceforge.net). (B) The ...


Here, we have used a fluorescent EMSA to reveal a previously uncharacterized binding site for the endosomal Sec1/Munc18 protein, Vps45p, on the syntaxin Tlg2p. We show that a construct lacking the N-peptide of Tlg2p interacts tightly with Vps45p (Fig. 2), demonstrating the presence of a C-terminal binding site. The Vps45p–L117R mutant, which does not bind the N-peptide of Tlg2p, binds to Tlg2p(37-318) with a similar affinity as wild type. Moreover, mutation of Tlg2p residues F9 and L10 to alanine, which disrupts binding of the N-peptide to Vps45p, competes for binding to Vps45p similarly to the N-peptide deletion (Table 1). This Tlg2p–Vps45p interaction explains the lack of a trafficking defect when the Vps45p–L117R mutant is expressed in yeast as the sole copy of Vps45p (13). Our results indicate that abrogation of both the N-peptide and closed conformation binding sites is needed to disrupt Tlg2p function (Fig. 4). Although previous qualitative studies suggested that the Tlg2p N-peptide was both necessary and sufficient for Vps45p interaction (12, 13), some of these studies did not test the appropriate truncated constructs, whereas others were performed at low concentrations with tagged proteins, in which binding may not have been readily detectable. We propose that similar quantitative binding studies may reveal a C-terminal binding site for other SM homologues that have previously been demonstrated to bind their cognate syntaxin's N-peptide, such as Sly1p. The presence of a second binding site would explain the lack of trafficking defects observed in mutants that abolish the Sly1p–Sed5p N-peptide interaction (16).

Several results strongly support our finding that this C-terminal binding site for Vps45p is a closed conformation of Tlg2p (Fig. 1A, mode 1). First, we previously showed that deletion of the first 230 residues of Tlg2p in vps45Δ cells leads to endosomal SNARE complex formation (35), suggesting that in the absence of Vps45p, the closed conformation of Tlg2p inhibits SNARE complex assembly. Second, we have recently demonstrated that either removal of the Tlg2p Habc domain or the addition of Vps45p increases the rate of endosomal SNARE complex assembly in vitro (36). Here, we show that binding of Vps45p to the C-terminal binding site requires the Habc, linker and SNARE motif regions of Tlg2p. In addition, the Tlg2p(37-318)–I285A mutant is unable to compete for binding to Vps45p; I285 is a residue predicted from analogous mutations in other SMs (7, 22, 29) to interact directly with Vps45p only when Tlg2p is in a closed conformation. Furthermore, the Tlg2p(37-318)–K134A/K137A/K163A mutant, which appears to form an open conformation as analyzed by CD and SEC, no longer competes for binding to Vps45p. Finally, evidence from mammalian Vps45 suggests that residues in the C-terminal region of syntaxin16 contribute to the overall binding affinity for Vps45, although no binding was detected in the absence of the syntaxin16 N-peptide (22).

In contrast, NMR evidence suggested that Tlg2p does not adopt a stable closed conformation (12). However, the Tlg2p construct used in that study (amino acids 60–283) lacked additional N-terminal residues and a key section of the SNARE motif, which may have destabilized the closed conformation. Analogous SNARE motif residues in Sso1p, although unstructured, are ordered in the crystal structure of the closed conformation and pack against the Habc domain, possibly contributing to the stability of the closed conformation (23). Similarly, SNARE motif residues of Tlg2p(37-318) that are C-terminal to residue 283 may stabilize the closed conformation. These would be predicted to pack in close proximity to the location of residues K134, K137, and K163, which we found to be important for stabilizing the closed Tlg2p conformation (Fig. 5). Consistent with its open conformation and lack of additional SNARE motif residues, the Tlg2p(60-283) truncation protein does not compete for binding to Vps45p (Table 1). It remains possible that Tlg2p is not tightly closed and is in equilibrium between closed and open states; in this case, the closed conformation would be stabilized by binding to Vps45p. Further biophysical and/or structural characterization will be necessary to test these models. Nonetheless, our results indicate that, by binding to the closed conformation of Tlg2p, Vps45p joins Munc18a as proteins that possess both modes 1 and 2 of binding to their cognate syntaxin proteins.

Our results also demonstrate that the Tlg2p N-peptide modulates the affinity of the Tlg2p closed conformation for Vps45p, indicating that these two sites do not bind simultaneously, e.g., using either mode 1 binding or mode 2, but excluding mode 3. Our competition assays reveal that the N-peptide cannot compete with the Vps45p–L117R mutant, confirming that binding of the Tlg2p N-peptide to the hydrophobic pocket on domain 1 of Vps45p is required for this competition. This result was surprising, because the two distinct binding sites in both Munc18a and mammalian Vps45 appear to positively influence each other (mode 3; ref. 22.) These discrepancies likely reflect real differences between the SM proteins in diverse organisms and suggest that the N-peptide may be an attractive target for regulating the accessibility of the syntaxin SNARE motif for SNARE complex assembly (switching from mode 1 to 2). Alternatively, these results may suggest that, rather than the Munc18a cleft, Vps45p may use a different surface to bind to closed Tlg2p, one that overlaps with the N-peptide binding pocket. The fact that the Vps45p–L117R mutant binds the closed Tlg2p with similar affinity to the wild type argues strongly against this possibility.

In conclusion, our use of quantitative in vitro binding analyses led to the discovery of a closed conformation binding site on Tlg2p for the yeast SM protein Vps45p. Together with previous studies, these results now reveal that Vps45p uses all of the binding modes observed for other SM proteins and their cognate syntaxins. These diverse modes of binding to SNAREs and SNARE complexes allow SM proteins to function at multiple key stages during the SNARE assembly and membrane fusion processes. The plethora of different binding modes provides ample opportunities for exquisite regulation of SNARE-mediated membrane fusion in diverse trafficking steps. Further quantitative in vitro studies, combined with in vivo analyses of specific SM and SNARE mutants, will tease apart the detailed regulatory mechanisms. A comprehensive molecular understanding necessitates the expansion of these studies to include the function of SMs in conjunction with tethering complexes, Rab GTPases, and specialized regulators, such as Munc13, synaptotagmins, and complexins.

Materials and Methods

Protein Purification and Analyses.

Recombinant Tlg2p and Vps45p constructs with N-terminal His6 tags were expressed in Escherichia coli and purified by affinity, ion-exchange, and SEC. The Tlg2p(1-33) peptide containing a cysteine at the C terminus was synthesized (Quality Controlled Biochemicals). For SEC experiments, protein samples were run on a Superdex 200 10/30 column (GE) and monitored using the absorbance at 280 nm. Complex formation was verified by running elution fractions on SDS/PAGE and staining with Coomassie blue. Further purification and SEC details are given in SI Text.


Tlg2p(37-318) was labeled [Tlg2p(37-318)*] at Cys-316 with maleimide-conjugated Alexa Fluor 488 dye. The apparent affinity (Kd,app) between Vps45p and Tlg2p(37-318)* was determined by EMSA using a protocol adapted from ref. 37. Additional details are given in SI Text. The data were fit to a sigmoidal dose–response function to determine the half-maximal saturation point (Kd,app) and the apparent Hill coefficient (n), by using Eq. 1 in SI Text. Competition experiments were performed similarly to the direct titrations, except that increasing concentrations of unlabeled Tlg2p constructs (“competitor”) were incubated with a subsaturated concentration of Vps45p–Tlg2p(37-318)*. The apparent equilibrium dissociation of the competitor (Kc,app) and the shape factor were determined as described in SI Text.

CPY Secretion Assays.

The ability of various Tlg2p constructs to complement the CPY trafficking defects of tlg2Δ cells was monitored by using a CPY-invertase fusion protein as described (32).

Supplementary Material

Supporting Information:


We thank Jen Songer (1979–2008), A. Lachapelle, B. Farley, and J. Pagano for technical assistance; J. Kahana (Harvard University) for pJK213; L.N. Carpp (Instituto Oswaldo Cruz/FIOCRUZ) for pCOG066; and C. Carr, B. Kobertz, and members of M.M.'s laboratory for critical reading of this manuscript. N.J.B. is a Prize Fellow of the Lister Institute of Preventive Medicine. This work was supported by Biotechnology and Biological Sciences Research Council Project Grant 17/C19548 (to N.J.B) and National Institutes of Health Grants GM081422 (to S.P.R.) and GM068803 (to M.M.).


The authors declare no conflict of interest.

This article is a PNAS Direct Submission.

This article contains supporting information online at www.pnas.org/cgi/content/full/0902976106/DCSupplemental.


1. Bonifacino JS, Glick BS. The mechanisms of vesicle budding and fusion. Cell. 2004;116:153–166. [PubMed]
2. Wickner W, Schekman R. Membrane fusion. Nat Struct Mol Biol. 2008;15:658–664. [PMC free article] [PubMed]
3. Sutton RB, Fasshauer D, Jahn R, Brunger AT. Crystal structure of a SNARE complex involved in synaptic exocytosis at 2.4-Å resolution. Nature. 1998;395:347–353. [PubMed]
4. Weber T, et al. SNAREpins: Minimal machinery for membrane fusion. Cell. 1998;92:759–772. [PubMed]
5. Toonen RF, Verhage M. Vesicle trafficking: Pleasure and pain from SM genes. Trends Cell Biol. 2003;13:177–186. [PubMed]
6. Toonen RF, Verhage M. Munc18–1 in secretion: Lonely Munc joins SNARE team and takes control. Trends Neurosci. 2007;30:564–572. [PubMed]
7. Misura KM, Scheller RH, Weis WI. Three-dimensional structure of the neuronal-Sec1-syntaxin 1a complex. Nature. 2000;404:355–362. [PubMed]
8. Bracher A, Perrakis A, Dresbach T, Betz H, Weissenhorn W. The X-ray crystal structure of neuronal Sec1 from squid sheds new light on the role of this protein in exocytosis. Structure (London) 2000;8:685–694. [PubMed]
9. Bracher A, Weissenhorn W. Structural basis for the Golgi membrane recruitment of Sly1p by Sed5p. EMBO J. 2002;21:6114–6124. [PMC free article] [PubMed]
10. Hu SH, Latham CF, Gee CL, James DE, Martin JL. Structure of the Munc18c/Syntaxin4 N-peptide complex defines universal features of the N-peptide binding mode of Sec1/Munc18 proteins. Proc Natl Acad Sci USA. 2007;104:8773–8778. [PMC free article] [PubMed]
11. Yamaguchi T, et al. Sly1 binds to Golgi and ER syntaxins via a conserved N-terminal peptide motif. Dev Cell. 2002;2:295–305. [PubMed]
12. Dulubova I, et al. How Tlg2p/syntaxin 16 “snares” Vps45. EMBO J. 2002;21:3620–3631. [PMC free article] [PubMed]
13. Carpp LN, Ciufo LF, Shanks SG, Boyd A, Bryant NJ. The Sec1p/Munc18 protein Vps45p binds its cognate SNARE proteins via two distinct modes. J Cell Biol. 2006;173:927–936. [PMC free article] [PubMed]
14. Togneri J, Cheng YS, Munson M, Hughson FM, Carr CM. Specific SNARE complex binding mode of the Sec1/Munc-18 protein, Sec1p. Proc Natl Acad Sci USA. 2006;103:17730–17735. [PMC free article] [PubMed]
15. Shen J, Tareste DC, Paumet F, Rothman JE, Melia TJ. Selective activation of cognate SNAREpins by Sec1/Munc18 proteins. Cell. 2007;128:183–195. [PubMed]
16. Peng R, Gallwitz D. Multiple SNARE interactions of an SM protein: Sed5p/Sly1p binding is dispensable for transport. EMBO J. 2004;23:3939–3949. [PMC free article] [PubMed]
17. Rodkey TL, Liu S, Barry M, McNew JA. Munc18a scaffolds SNARE assembly to promote membrane fusion. Mol Biol Cell. 2008:5422–5434. [PMC free article] [PubMed]
18. Brandie FM, et al. Negative regulation of syntaxin4/SNAP-23/VAMP2-mediated membrane fusion by Munc18c in vitro. PLoS ONE. 2008;3:e4074. [PMC free article] [PubMed]
19. Dulubova I, et al. Munc18–1 binds directly to the neuronal SNARE complex. Proc Natl Acad Sci USA. 2007;104:2697–2702. [PMC free article] [PubMed]
20. Khvotchev M, et al. Dual modes of Munc18–1/SNARE interactions are coupled by functionally critical binding to syntaxin-1 N terminus. J Neurosci. 2007;27:12147–12155. [PubMed]
21. Rickman C, Medine CN, Bergmann A, Duncan RR. Functionally and spatially distinct modes of munc18-syntaxin 1 interaction. J Biol Chem. 2007;282:12097–12103. [PMC free article] [PubMed]
22. Burkhardt P, Hattendorf DA, Weis WI, Fasshauer D. Munc18a controls SNARE assembly through its interaction with the syntaxin N-peptide. EMBO J. 2008;27:923–933. [PMC free article] [PubMed]
23. Munson M, Chen X, Cocina AE, Schultz SM, Hughson FM. Interactions within the yeast t-SNARE Sso1p that control SNARE complex assembly. Nat Struct Biol. 2000;7:894–902. [PubMed]
24. Chen X, Lu J, Dulubova I, Rizo J. NMR analysis of the closed conformation of syntaxin-1. J Biomol NMR. 2008;41:43–54. [PMC free article] [PubMed]
25. Fasshauer D, Bruns D, Shen B, Jahn R, Brunger AT. A structural change occurs upon binding of syntaxin to SNAP-25. J Biol Chem. 1997;272:4582–4590. [PubMed]
26. Nicholson KL, et al. Regulation of SNARE complex assembly by an N-terminal domain of the t-SNARE Sso1p. Nat Struct Biol. 1998;5:793–802. [PubMed]
27. Ryder SP, Recht MI, Williamson JR. Quantitative analysis of protein-RNA interactions by gel mobility shift. Methods Mol Biol. 2008;488:99–115. [PMC free article] [PubMed]
28. Lin SY, Riggs AD. Lac repressor binding to nonoperator DNA: Detailed studies and a comparison of equilibrium and rate competition methods. J Mol Biol. 1972;72:671–690. [PubMed]
29. Wu MN, et al. Syntaxin 1A interacts with multiple exocytic proteins to regulate neurotransmitter release in vivo. Neuron. 1999;23:593–605. [PubMed]
30. Johnson JR, et al. Binding of UNC-18 to the N-terminus of syntaxin is essential for neurotransmission in Caenorhabditis elegans. Biochem J. 2009;418:73–80. [PubMed]
31. McEwen JM, Kaplan JM. UNC-18 promotes both the anterograde trafficking and synaptic function of syntaxin. Mol Biol Cell. 2008;19:3836–3846. [PMC free article] [PubMed]
32. Darsow T, Odorizzi G, Emr SD. Invertase fusion proteins for analysis of protein trafficking in yeast. Methods Enzymol. 2000;327:95–106. [PubMed]
33. Abeliovich H, Grote E, Novick P, Ferro-Novick S. Tlg2p, a yeast syntaxin homolog that resides on the Golgi and endocytic structures. J Biol Chem. 1998;273:11719–11727. [PubMed]
34. Munson M, Hughson FM. Conformational regulation of SNARE assembly and disassembly in vivo. J Biol Chem. 2002;277:9375–9381. [PubMed]
35. Bryant NJ, James DE. Vps45p stabilizes the syntaxin homologue Tlg2p and positively regulates SNARE complex formation. EMBO J. 2001;20:3380–3388. [PMC free article] [PubMed]
36. Struthers MS, et al. Functional homology of mammalian syntaxin16 and yeast Tlg2p reveals a conserved regulatory mechanism. J Cell Sci. 2009;122:2292–2299. [PMC free article] [PubMed]
37. Pagano JM, Farley BM, McCoig LM, Ryder SP. Molecular basis of RNA recognition by the embryonic polarity determinant MEX-5. J Biol Chem. 2007;282:8883–8894. [PubMed]

Articles from Proceedings of the National Academy of Sciences of the United States of America are provided here courtesy of National Academy of Sciences
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...