A key issue for understanding exocytosis is elucidating the various protein interactions and the associated conformational transitions underlying soluble N-ethylmeleimide-sensitive factor attachment protein receptor (SNARE) protein assembly. To monitor dynamic changes in syntaxin 1A (Syx) conformation along exocytosis, we constructed a novel fluorescent Syx-based probe that can be efficiently incorporated within endogenous SNARE complexes, support exocytosis, and report shifts in Syx between ‘closed’ and ‘open’ conformations by fluorescence resonance energy transfer analysis. Using this probe we resolve two distinct Syx conformational transitions during membrane depolarization-induced exocytosis in PC12 cells: a partial ‘opening’ in the absence of Ca2+ entry and an additional ‘opening’ upon Ca2+ entry. The Ca2+-dependent transition is abolished upon neutralization of the basic charges in the juxtamembrane regions of Syx, which also impairs exocytosis. These novel findings provide evidence of two conformational transitions in Syx during exocytosis, which have not been reported before: one transition directly induced by depolarization and an additional transition that involves the juxtamembrane region of Syx. The superior sensitivity of our probe also enabled detection of subtle Syx conformational changes upon interaction with VAMP2, which were absolutely dependent on the basic charges of the juxtamembrane region. Hence, our results further suggest that the Ca2+-dependent transition in Syx involves zippering between the membrane-proximal juxtamembrane regions of Syx and VAMP2 and support the recently implied existence of this zippering in the final phase of SNARE assembly to catalyze exocytosis.
Syntaxin 1A (Syx), a plasma membrane (PM) neuronal Q-SNARE (soluble N-ethylmeleimide-sensitive factor attachment protein receptor), is a major protein component of the machinery involved in the maturation steps through which a vesicle undergoes before it can release a neurotransmitter (Sørensen, 2004), steps such as docking, priming, and fusion (Wojcik and Brose, 2007). During the priming process, sequential formation of the neuronal trimeric SNARE complex occurs (Brunger, 2001; Bruns and Jahn, 2002; Chen and Scheller, 2001; Jahn and Südhof, 1999). Initially, Syx assembles with PM SNARE, SNAP-25, to form the binary t-SNARE complex (Dun et al., 2010), which is followed by assembly of the vesicular SNARE, VAMP2, with the complex, yielding the trimeric SNARE complex, SNAREpin (Fasshauer and Margittai, 2004). The assembly of SNAREpin is a highly regulated multistep process going through pre-fusion partially zippered trans-complexes to the post-fusion fully zippered cis-complex, comprising Ca2+-independent and Ca2+-dependent intermediates (Malsam et al., 2008; Melia, 2007). Importantly, it is known that Syx undergoes one or more conformational changes upon its interaction with its SNARE partners and with regulatory proteins throughout the steps leading to secretion. However, the details of these conformations remain elusive. Indeed, conformational changes in Syx have been the subject of numerous studies. The majority of the approaches involved in vitro interactions of soluble protein motifs, studies of purified proteins reconstituted in liposomes, and use of X-ray crystallography, all of which provided important, yet limited, knowledge about the conformational changes occurring in membrane-bound Syx in neuronal or neuroendocrine cells and their relevance to events occurring during secretion. In particular, examination of the X-ray structure of the neuronal SNARE complex including the transmembrane regions of Syx and VAMP2 led to the hypothesis that the juxtamembrane region of Syx may play an important role in SNARE complex assembly (Stein et al., 2009). In accordance with this hypothesis, by using a reconstituted membrane fusion system, zippering of this region with the corresponding region in VAMP2 has recently been implicated in the SNARE complex assembly required for efficient fusion (Hernandez et al., 2012).
In this study we generated a novel Syx intramolecular fluorescence resonance energy transfer (FRET) reporter probe that is incorporated within endogenous SNARE complexes and reports dynamic conformational changes in Syx, in a neuronal-like cellular environment, during exocytosis. This probe enabled us to resolve two discrete secretion-related conformational changes in Syx in PC12 cells and provided two novel findings. First, Syx undergoes two distinct conformational transitions during exocytosis: a ‘partial opening’ induced by depolarization but in the absence of Ca2+ entry and further ‘opening’ that occurs upon Ca2+ entry. Second, the conserved juxtamembrane region of Syx plays a crucial role in the Ca2+-dependent ‘opening’ of Syx, probably at the final stage of the SNARE complex assembly. Thus, this probe enables one to test and validate Syx conformational transitions associated with specific interactions already documented in cell-free studies and to gain insights about novel interactions in vivo.
Construction of intramolecular Syx-based FRET probes
It is generally accepted that Syx can shift between two conformational states. In its ‘closed’ conformation the Habc domain folds back onto the SNARE motif (H3 domain), which is involved in forming a coiled-coil SNARE complex with the SNARE motifs of SNAP-25 and VAMP2 (Fig. 1A) (Margittai et al., 2003a; Verhage et al., 2000), thus preventing the formation of the SNARE complex that drives vesicle fusion. The ‘closed’ conformation constitutes a key intrinsic property of isolated Syx when not assembled into the SNARE complex (Chen et al., 2008), although a small percentage of Syx may spontaneously open (Margittai et al., 2003b). To enter t-SNARE and trimeric SNARE complexes, Syx must assume the ‘open’ conformation, subsequently exposing the H3 domain (Jahn and Scheller, 2006; Sutton et al., 1998).
To better understand Syx conformational changes associated with the exocytotic process in a living cell, we constructed double-labeled fluorescent Syx probes that may report conformational changes in Syx by FRET. We explored the recent crystal structure of Syx in order to rationalize our design of the probes. In the available structure, the N-terminus of Syx and its juxtamembrane region, connecting the transmembrane anchor and the H3 domain, are in proximity in the ‘closed’ rather than in the ‘open’ conformation (Dulubova et al., 1999; Misura et al., 2001). Accordingly, we fused two fluorescent molecules to Syx via flexible linkers: cyan fluorescent protein (CFP) to the N-terminus and yellow fluorescent protein (YFP) to the juxtamembrane region (Fig. 1A). We predicted that the two fluorophores would reside in proximity when Syx is in the ‘closed’ conformation, yielding a high FRET signal. Conversely, the ‘open’ conformation of Syx should robustly cause the fluorophores to separate, leading to a decreased FRET signal (Fig. 1A). Two probes were constructed: (1) CSYS (CFPNT-Syx-YFPdisatl-H3-Syx), with YFP inserted in the middle of the polybasic juxtamembrane region (KARRKK), and (2) CSYS-5RK, with YFP inserted between the H3 domain and the polybasic sequence (Fig. 1B). As control probes, we generated CSYS-Open and CSYS-5RK-Open, each with two point mutations, L165A and E166A, inserted at the linker region between the Habc and H3 domains, previously shown to shift the equilibrium of Syx toward the ‘open’ conformation (Dulubova et al., 1999; Richmond et al., 2001). Although several intramolecular FRET probes, based on SNAP-25, were previously reported (An and Almers, 2004; Takahashi et al., 2010; Wang et al., 2008), providing valuable insights about SNARE complex formation in living cells, to the best of our knowledge, no such Syx-based FRET probes have been reported. We reasoned that Syx-based probes will prove more sensitive in reporting the assembly of SNARE proteins associated with exocytosis in PC12 cells. Unlike the SNAP-25-based probes, prone to dilution by endogenous SNAP-25 [found in large excess over Syx in PC12 cells (>10-fold) (Knowles et al., 2010)], exogenously expressed Syx-based probes are more likely to compete efficiently with the relatively small amounts of endogenous Syx and be incorporated efficiently into native SNARE complexes.
Next, we performed experiments to determine whether CSYS and CSYS-5RK can form binary t-SNARE and trimeric SNARE complexes, as does native Syx (Fig. 1C–E). Since both probes showed similar results, they were collectively termed CSYS in these experiments. Co-immunoprecipitation analysis performed in Xenopus oocytes co-expressing metabolically labeled CSYS and SNAP-25, using either anti-Syx or anti-SNAP-25 antibodies, revealed that CSYS effectively associates with SNAP-25 to form t-SNARE complexes (Fig. 1C). In addition, SDS-resistant SNARE complexes were detected from oocytes co-expressing metabolically labeled CSYS, SNAP-25, and VAMP2, upon immunoprecipitation with either anti-Syx or anti-SNAP-25 antibodies, confirming the ability of CSYS to form trimeric SNARE complexes (Fig. 1D). To demonstrate that our probes are as effective as native Syx in forming trimeric complexes with SNAP-25 and VAMP2 as does native Syx, we assessed the SDS-resistant complexes formed in oocytes by CSYS and compared them to those formed by native Syx (Fig. 1E). Indeed, CSYS readily formed SDS-resistant complexes which contained also SNAP-25 and VAMP2, similarly to those formed by native Syx (Fig. 1E; note the difference in mobility of Syx- and CSYS-containing trimeric complexes).
CSYS probes can report the conformational ‘opening’ of Syx
Next, we analyzed the conformations adopted by CSYS and CSYS-5RK, using the spectral FRET technique in Xenopus oocytes (Etzioni et al., 2011; Zheng et al., 2003). The probes were efficiently targeted to the PM and exhibited a high FRET signal under resting conditions (static FRET; Fig. 2A). As expected, the static FRET signal did not change over a wide range of expression levels because of a ∼1∶1 donor-to-acceptor ratio [supplementary material Fig. S1A–C; (Berlin et al., 2010)]. Surprisingly, the FRET signals of CSYS and CSYS-5RK were significantly different, although the position of the YFP fluorophore was shifted only 4 a.a. within the polybasic juxtamembrane region (Fig. 2A, right panel). This prompted us to investigate the importance of this highly conserved region and to generate an additional probe, CSYS-5RK/A, in which the 5 positively charged residues in CSYS-5RK were neutralized (Fig. 1B). CSYS-5RK/A had a FRET signal similar to that of CSYS, but it was significantly different from that of CSYS-5RK (Fig. 2A, right panel). These results suggest that changes within the polybasic region affect the conformation of Syx.
We then tested the ability of the probes to report structural rearrangements related to the ‘opening’ of Syx. Co-expressed SNAP-25 significantly and dose-dependently reduced the FRET signals of CSYS and CSYS-5RK to levels similar to those obtained by the corresponding Open probes (Fig. 2Ba,b), confirming previous observations regarding the ‘opening’ of Syx by SNAP-25 (Jahn and Scheller, 2006; Sutton et al., 1998). Thus, we concluded that reductions in the FRET signals of CSYS and CSYS-5RK most probably report the ‘opening’ of Syx.
CSYS-5RK/A also reported a SNAP-25-mediated ‘opening’, similarly to CSYS and CSYS-5RK (Fig. 2Bc). Indeed, concomitant co-immunoprecipitation analysis in oocytes co-expressing metabolically labeled CSYS-5RK or CSYS-5RK/A with SNAP-25 and VAMP2 revealed that CSYS-5RK/A is as effective as CSYS-5RK in binding SNAP-25 and VAMP2 (supplementary material Fig. S2). Importantly, these results indicate that, although the neutralization of the juxtamembrane region of Syx affects the initial probe conformation (see above), it does not affect the ability to associate with its SNARE partners and to report structural ‘openings’ in vivo.
High K+- depolarization induces conformational transitions in CSYS probes in PC12 cells
Our next aim was to use our FRET probes to investigate conformational changes in Syx associated with SNARE complex formation in a physiologically relevant setting of secreting PC12 cells. All our results in PC12 cells discussed hereafter (unless otherwise noted) were obtained with both CSYS and CSYS-5RK probes, which yielded similar results; hence, they are collectively termed CSYS. Several preliminary analyses were performed. First, we verified that CSYS targeted properly the PM in PC12 cells. Indeed, 90% of cells transfected with CSYS exhibited a fluorescent signal at the PM region, indicating PM expression (see the membrane expression in Fig. 3B). Second, we evaluated the impact of CSYS on secretion. Briefly, we used an established secretion assay in PC12 cells in which fluorescence decline of mRFP-tagged vesicular neuropeptide Y (NPY-mRFP; dimming of cells as they release the granular marker) is monitored in response to membrane depolarization induced by perfusion of a high [K+] (hK) solution [Fig. 3Aa; (Singer-Lahat et al., 2007)]. More than 70% of the cells displayed a significant amount of secretion that was practically eliminated when intracellular [Ca2+] elevation was blocked in the presence of cadmium (Cd) (Fig. 3Ab; supplementary material Fig. S3Aa). This corroborated the occurrence of the well-documented dependence of secretion on Ca2+ entry via voltage-gated Ca2+-channels under our experimental conditions. Importantly, the expression of CSYS in these cells did not change the depolarization-induced elevation of the cytosolic Ca2+ level (supplementary material Fig. S3Ab) and significantly enhanced secretion (Fig. 3Ab), suggesting that CSYS can associate with endogenous SNARE partners and form functional exocytic complexes. Third, realizing that the SNARE functionality of CSYS is of utmost importance in serving as a reporter of SNARE-conformational changes during exocytosis, we sought to rigorously challenge the ability of CSYS to substitute for native Syx and to support secretion in cells transfected with the light chain of BoNT-C1, which cleaves Syx and inhibits membrane fusion (Schiavo et al., 1995). To this end, we generated a CSYS mutant, CSYS(R), bearing a mutation [K253I; (Lam et al., 2008)] in the Syx sequence that conferred resistance to BoNT-C1 (Fig. 3B; supplementary material Fig. S3B). Indeed, whereas secretion triggered by hK was reduced to 15% in cells expressing CSYS and BoNT-C1, secretion in cells expressing CSYS(R) and BoNT-C1 was rescued to 65% (Fig. 3C; the expression levels of CSYS-5RK and CSYS-5RK(R) were similar; partial SNAP-25 cleavage by BoNT-C1 could contribute to the incomplete rescue). Thus, we concluded that CSYS could substitute for endogenous Syx, could be successfully incorporated into endogenous SNARE complexes, and could support exocytosis. However, the validity of this conclusion is dependent on the ability of BoNT-C1 to cleave endogenous Syx in the presence of the overexpressed cleavage-resistant CSYS(R). We verified this by showing that Syx was equally sensitive to BoNT-C1 in the absence and presence of CSYS or CSYS(R) (supplementary material Fig. S3B). Taken together, the results of the above preliminary analyses validated the suitability of CSYS to serve as a reporter for Syx's conformational changes upon depolarization-induced exocytosis in secreting PC12 cells.
Next, conformational changes associated with SNARE complex formation were monitored by dynamic FRET changes in response to hK depolarization in PC12 cells expressing CSYS. Time series images of PC12 cells were acquired before and during hK depolarization and fluorescent intensities were collected from the PM (Fig. 4A), from which the FYFP/FCFP ratio was calculated (Berlin et al., 2010; Hein et al., 2005). Remarkably, significant reductions in FRET following exposure to hK solution were evident already in single cells (Fig. 4A) and were reproducible in more than 70% of the CSYS-expressing cells. We verified that these FRET changes exhibited an intra-molecular interaction with no contribution from an intermolecular interaction (supplementary material Fig. S1D), thus reporting conformational changes associated with ‘opening’ of CSYS. Fig. 4B shows a significant decrease of ∼5% in the average normalized FRET ratio, reporting a conformational shift of CSYS toward the ‘open’ state upon hK stimulation. To further substantiate this conclusion, we stimulated cells expressing CSYS-Open with an hK solution. As predicted, no significant changes in FRET were observed (Fig. 4C).
We next tested the role of Ca2+ in the conformational shift of CSYS. In the presence of Cd, which completely blocked intracellular [Ca2+] elevation (supplementary material Fig. S3A) and secretion (Fig. 3Ab) in response to hK stimulation, CSYS only partially ‘opened’ upon hK stimulation (Fig. 4D; in one of these experiments, BAPTA-AM, a membrane-permeant Ca2+ chelator, was also included to further rule out any local [Ca2+] rise). This Cd-resistant partial ‘opening’ demonstrates that Syx undergoes, in the absence of intracellular [Ca2+] elevation, a depolarization-dependent, yet Ca2+-independent, conformational transition that does not support by itself exocytosis. Notably, no similar conformational change upon hK stimulation was detected in CSYS-Open (Fig. 4C), strongly suggesting that the partial Ca2+-independent ‘opening’ of CSYS represents a physiologically related transition and not a stimulation-induced non-specific conformational change in the probe. To better understand the nature of the partial Ca2+-independent ‘opening’ of CSYS, we investigated whether it represents an intermediate step that can be transformed into a ‘full opening’ in the presence of Ca2+. To address this issue, we performed a two-step hK stimulation, first, stimulation in a Ca2+-free solution, followed by a second stimulation in a Ca2+-containing solution (Fig. 4E). In the absence of Ca2+, a ‘partial opening’ of CSYS occurred, the extent of which was similar to that observed in the presence of Cd (compare Fig. 4D,E). Upon addition of Ca2+, an additional ‘opening’ occurred, to a level similar to that of the one-step stimulation in the presence of Ca2+ (compare Fig. 4D,E). Importantly, no such additional ‘opening’ occurred upon prolonged incubation of the cells in hK solution with no Ca2+ added (Fig. 4F); namely, the addition of Ca2+ is responsible for the further decrease of the FRET ratio. These results suggest that a two-component conformational transition takes place during hK stimulation. The Ca2+-independent, but depolarization-dependent, ‘partial opening’ of CSYS may possibly be an intermediate structure along a sequential pathway leading to a ‘full opening’. Notably, we failed to detect conformational changes induced by membrane depolarization, using a dynamic FRET assay in oocytes expressing CSYS, either clamped to different depolarized voltages (supplementary material Fig. S4A) or subjected to hK solution (supplementary material Fig. S4B). This suggests that the ‘partial’, as well as the ‘full openings’, apparent in PC12 cells, require a secreting cell environment. This also rules out the possibility of an hK solution-related artifact.
Taken together, our results indicate the ability of CSYS to resolve conformational transitions in the process of being incorporated into endogenous SNARE complexes in secreting cells. We suggest, for the first time, that the ‘full opening’ of Syx during depolarization-induced exocytosis is mediated by two separate mechanisms related to Ca2+-independent and Ca2+- dependent steps.
The polybasic juxtamembrane region of Syx is important for Ca2+-dependent conformational transitions in Syx during exocytosis
The results of the static FRET analysis in oocytes suggested that changes in the juxtamembrane region of Syx affect the conformation of Syx (Fig. 2A). Recent X-ray structure implicated this region of Syx in the assembly of the neuronal SNARE complex (Stein et al., 2009). Taken together, this led us to hypothesize that the juxtamembrane region of Syx may be an important component of Syx's depolarization-induced conformational transitions (Fig. 4). We tested our hypothesis by using CSYS-5RK/A (in which all the basic residues of the juxtamembrane region of CSYS-5RK are neutralized; Fig. 1B). Importantly, this probe retains the ability to associate with its SNARE partners (supplementary material Fig. S2) and to report structural ‘openings’ in vivo (Fig. 2Bc). Using the dynamic FRET assay in hK stimulated PC12 cells (as done in Fig. 4), we compared the conformational transitions in CSYS-5RK/A with those in CSYS-5RK. Cells expressing CSYS-5RK/A exhibited a smaller, but statistically significant (P<0.05), reduction in the FRET ratio compared with cells expressing CSYS-5RK (Fig. 5A). This conformational transition of CSYS-5RK/A was unaffected by Cd (Fig. 5B), suggesting that it may reflect the Ca2+-independent transition of CSYS-5RK. Indeed, it was similar to that of CSYS-5RK in the presence of Cd, which was tested in the same experiment (Fig. 5C). These results indicate that neutralization of the basic residues abolished the Ca2+-dependent hK-induced transition of Syx ‘opening’. The neutralization also impaired exocytosis as hK-induced secretion from cells expressing CSYS-5RK/A was smaller than that from cells expressing CSYS-5RK (Fig. 5D; the expression levels of CSYS-5RK and CSYS-5RK/A were similar). Hence, using our reporters, we found that the full ‘opening’ of Syx during the Ca2+-dependent transition absolutely depends on the positive charges in the juxtamembrane region. Our findings strongly suggest a crucial role for this region in Ca2+-dependent structural transition of Syx that underlie efficient exocytosis.
Next, we tested the possibility that the Ca2+-dependent structural transition may relate to zippering between the juxtamembrane regions of Syx and VAMP2 during SNARE complex assembly along the fusion pathway, as suggested in (Hernandez et al., 2012; Stein et al., 2009). To this end, we tested the prediction that neutralization of the basic residues in the juxtamembrane region of CSYS should impair interaction between CSYS and VAMP2. Interaction between Syx and VAMP2 (in the absence of SNAP-25) was previously assayed in vitro and was found to be very weak (Fasshauer et al., 1997; Hazzard et al., 1999) or undetectable (Gao et al., 2012). Here, we first tested for this interaction in vivo, using our FRET probes. We found that in oocytes co-expressing VAMP2 with CSYS or with CSYS-5RK (in the absence of co-expressed SNAP-25) the FRET of the probes was significantly decreased (by ∼20%), establishing that VAMP2 interacts with both Syx probes and mediates their ‘opening’ (Fig. 6A). As expected from a weak interaction between VAMP2 and Syx, this reduction in FRET was significantly smaller than that obtained in oocytes co-expressing SNAP-25 with CSYS (Fig. 6B). In agreement with our prediction, the FRET of CSYS-5RK/A remained unaffected (Fig. 6A). Namely, neutralization of the basic residues in the juxtamembrane region abolished the interaction with VAMP2, strengthening our hypothesis of the role of the juxtamembrane region of Syx when interacting with VAMP2. Interestingly, concomitant co-immunoprecipitation analysis, using Syx antibody, performed on the same oocytes, revealed that VAMP2 associates with both CSYS-5RK/A and CSYS-5RK (albeit the association with CSYS-5RK/A was weaker by ∼20%; both associations were very weak and probably reflect interaction mainly mediated by the corresponding SNARE motifs; supplementary material Fig. S5). Taken together, this data demonstrates the superior sensitivity and specificity of our Syx-based FRET reporters, even exceeding those of biochemical analyses, in capturing conformational changes of Syx arising from subtle changes in its interaction(s) with protein partners.
In this study we generated novel Syx intramolecular FRET probes, termed CSYS, which are able to report dynamic changes in Syx structure in the process of being incorporated into endogenous SNARE complexes formed along the pathway to exocytosis. Using these probes we resolved, for the first time in vivo, secretion-related conformational transitions in Syx. We showed that full ‘opening’ of Syx is mediated by two separate mechanisms occurring during Ca2+-independent and Ca2+-dependent steps of depolarization-induced efficient exocytosis. Notably, although Ca2+-independent structural changes in Syx have previously been implicated in vesicle exocytosis (see discussion below), this is the first case showing that such events can be induced by membrane potential depolarization. Furthermore, we found that the Ca2+-dependent ‘opening’ of Syx is contingent on positive charges in its juxtamembrane region, neutralization of this region impairs exocytosis. This finding, obtained in living cells, strongly highlights the crucial role of this region in structural transition(s) of Syx that underlie efficient exocytosis and validates predictions made by previous structural (Stein et al., 2009) and membrane fusion-related studies in cell-free reconstituted systems (Hernandez et al., 2012).
Our results open the question regarding the precise physiological correlates and molecular events underlying the two conformational transitions assumed by Syx. The Ca2+-independent partial ‘opening’ of Syx, induced by membrane potential depolarization, may report conformational changes occurring during vesicle docking, priming, and entry into partially assembled SNAREpins. Such conformational changes, controlled by SNARE-associated regulatory factors, are pre-Ca2+ intermediates documented in numerous reconstituted assays in relation to vesicle docking and priming reactions (Malsam et al., 2008; Melia, 2007; Parisotto et al., 2012; Südhof and Rothman, 2009). It is quite likely that Syx itself cannot sense membrane electric field and trigger these molecular events. We have examined this possibility and failed to detect any similar conformational changes in CSYS-expressing oocytes that were voltage-clamped to different depolarized voltages (supplementary material Fig. S4). In addition, we demonstrated (Fig. 5) that eliminating all charged residues located in the vicinity of the membrane, which could potentially serve as a voltage sensing module, does not eliminate the voltage sensitivity of our probe. A more plausible explanation is that the voltage sensitivity arises from a separate voltage-sensitive protein, serving as a voltage sensor, which is not present in the oocyte. Such a mechanism would be similar, for instance, to the well-characterized voltage sensing mechanism of Ca2+ release from sarcoplasmic reticulum (SR) in skeletal muscle. There, the L-type Ca2+ channel, located in the plasma membrane, senses depolarization and signals, via direct contact, to the Ca2+ channel located in the SR [for a review, see (Iino, 1999)]. There are several candidate proteins that could similarly signal to Syx, via direct physical contact, by serving as a voltage sensor: (1) the neuronal voltage-dependent Ca2+ channels of N- or L-type (Bergsman and Tsien, 2000; Bezprozvanny et al., 1995; Wiser et al., 1996) [for a review, see (Catterall, 2000)], (2) the voltage-dependent K+ channel Kv2.1, which is the predominant channel in neuroendocrine cells and is thought to interact directly with Syx to enhance vesicle priming by promoting assembly and/or stabilization of the t-SNARE complex (Feinshreiber et al., 2010; Michaelevski et al., 2003; Singer-Lahat et al., 2007), or (3) G protein-coupled receptors (Linial et al., 1997). All these proteins undergo conformational changes upon depolarization (Ben-Chaim et al., 2006; Hille, 2001).
The Ca2+-dependent full ‘opening’ of Syx reported by CSYS may reflect one or more structural changes in Syx involved in the final phase of SNARE assembly when cis-complexes are formed. The CSYS-5RK/A probe with the neutralized juxtamembrane region demonstrates that the polybasic juxtamembrane region is crucial for these Ca2+-dependent structural changes reported by CSYS (Fig. 5A). According to the X-ray structure analysis of the neuronal SNARE complex (Stein et al., 2009), the juxtamembrane regions of Syx and VAMP2 form helices continuous with their SNARE motifs and transmembrane regions. Furthermore, a recent study of membrane fusion intermediates in a cell-free system, shows that efficient fusion up to an extended form of hemifusion requires zippering beyond the core SNARE complex to the juxtamembrane regions of Syx and VAMP2 (Hernandez et al., 2012). Taken together, the Ca2+-dependent full ‘opening’ reported by CSYS most likely reflects zippering between the juxtamembrane regions of Syx and VAMP2. According to this interpretation, it is predicted that neutralization of basic residues in the juxtamembrane region of CSYS impairs the interaction between CSYS and VAMP2. In agreement with our prediction, FRET analysis detected an interaction between VAMP2 and CSYS, but could not detect similar interaction between VAMP2 and CSYS-5RK/A (Fig. 6). Interestingly, CSYS-5RK/A compromised, but did not entirely eliminate hK-induced secretion (Fig. 5D), as had been demonstrated before for native Syx with a neutralized juxtamembrane region (Lam et al., 2008). Indeed, as previously discussed (Stein et al., 2009), structural perturbation of the juxtamembrane interaction between VAMP2 and Syx (as was done in earlier studies by manipulating the juxtamembrane regions' length and the amino acid composition) should be less disruptive for fusion than is perturbation of the interaction within the four-helix bundle of the SNARE motifs. Overall, our results suggest a role for the polybasic juxtamembrane region of Syx in facilitating exocytosis and validate, in a neuronal-like environment, the notion that this final phase of SNARE assembly is directly coupled to efficient membrane fusion.
It should be noted that the polybasic juxtamembrane region of Syx has been recognized as a lipid binding domain. Consequently, it was proposed that Syx–acidic phospholipids interactions are critical in determining the energetics of the SNARE-mediated fusion event by sequestering fusogenic lipids to sites of fusion (Lam et al., 2008). Notably, electrostatic interactions of the polybasic residues with PIP2 (James et al., 2010; Murray and Tamm, 2009) have recently been shown to underlie clustering and segregation of Syx into distinct microdomains where synaptic vesicles undergo exocytosis (van den Bogaart et al., 2011). Thus, it is quite possible that the conformational transitions linked to the juxtamembrane region of Syx, detected in our study, also reflect interaction of this region with PIP2, hence, complementing the implication of this region in the efficient assembly of the complete fusion machinery in spots of exocytosis.
Materials and Methods
Double-labeled Syx (CSYS) cDNA was generated first by fusing enhanced cyan fluorescent protein (CFP), flanked by EcoRI and XhoI sites to the N-terminus of the rat Syx cDNA in pGEMHJ vector, followed by insertion of NcoI restriction sites after the Arg-263 residue, which lies between the transmembrane anchor and the H3 domain of Syx. Next, the enhanced yellow fluorescent protein (YFP) was fused into the aforementioned restriction site via Gly-Gly-Gly-Ser linkers. CSYS-5RK and CSYS-5RK/A were generated based on CSYS cDNA, using point mutation site-directed mutagenesis. The open forms of the constructs were generated by introducing two point mutations, L165A and E166A, at the linker region between the Habc and H3 domain of CSYS (Dulubova et al., 1999), using point mutation site-directed mutagenesis (Edelheit et al., 2009). The BoNT-C1-resistant construct, CSYS(R), was generated by introducing one point mutation, K253I, at the BoNT-C1 recognizing sequence (Lam et al., 2008). For PC12 transfection, the constructs were cloned into pcDNA3 vector using EcoRI and XbaI restriction sites.
Immunoprecipitation (IP) and immunoblotting (IB)
In IP experiments de-folliculated Xenopus oocytes were metabolically labeled 4 hrs after mRNA injection by incubation in NDE solution containing 0.3 mCi/ml of [35S]methionine/cysteine for 2 days until homogenization was achieved. Six to eight oocytes were homogenized in 1 ml of medium composed of 20 mM Tris (pH 7.4), 5 mM EDTA, 5 mM EGTA, and 100 mM NaCl [containing protease inhibitor cocktail (Roche)]. Debris was removed by centrifugation for 10 min at 4°C. After the addition of Triton X-100 to a final concentration of 1%, followed by microcentrifugation for 15 min at 4°C in a desktop centrifuge, antiserum [Syx 1A antibody (Alomone), SNAP-25 antibody (BD Transduction lab) or VAMP2 antibody (Abcam)] was added to the supernatant and the homogenate was incubated for 16 hrs. Then, the antibody–antigen complex was incubated for 1 h at 4°C with Protein A–Sepharose and then pelleted by centrifugation for 1 min at 8000×g. Immunoprecipitates were washed four times with immunowash buffer (150 mM NaCl, 5 mM EDTA, 50 mM Tris (pH 7.5), and 0.1% Triton X-100); the final wash contained no Triton X-100. Samples were boiled in SDS gel loading buffer and electrophoresed on SDS- 8% or 12% polyacrylamide gel together with standard molecular mass markers. Digitized scans were derived by PhosphorImager (Amersham Biosciences), and relative intensities were quantified by ImageQuant.
In IB experiments proteins expressed in Xenopus oocytes or in PC12 cells were co-immunoprecipitated by Syx 1A antibody (Alomone), subjected to western blot analysis and detected by Syx antibody (Sigma) with the ECL detection system (Pierce protein research products, Thermo Scientific).
Fluorescence resonance energy transfer (FRET) in Xenopus oocytes
Xenopus oocyte preparation
Xenopus oocytes were prepared and injected with mRNA as previously described (Dascal and Lotan, 1992; Peleg et al., 2002). Each probe concentration was determined in a calibration experiment in order to obtain the same expression levels among all constructs in every experiment. Oocytes were injected with the following mRNA concentrations (ng/oocytes) of the various probes: 5 CSYS, 5 CSYS-5RK, 15 CSYS-5RK/A, and 15 of various probes in the ‘open’ form. Oocytes were injected with various CSYS probes alone or together with increasing concentrations of SNAP-25 and/or VAMP2 (ranging from 5 to 25 ng/oocytes), for a dose-dependent response and in order to detect the concentration that produces the highest effect.
Static FRET assay in oocytes
Oocytes were imaged in ND96 solution in a 0.7-mm glass-bottom dish. Fluorescence emissions from CFP and/or YFP-tagged proteins were collected from the animal hemisphere of the oocyte with a Zeiss inverted confocal microscope (Zeiss Axiovert LSM 510META), using a 20× 0.75 NA air objective and laser excitations of 405 nm and 514 nm, respectively. We used a spectrum-based method to remove contamination caused by donor emission and for direct excitation of the acceptor. The FRET assay was performed as described in (Zheng and Zagotta, 2004). Briefly, two emission spectra were collected from each oocyte, one with 405 nm excitation and the other with 514 nm excitation. A scaled CFP spectrum, collected from control oocytes expressing CFP-tagged proteins only, were used to normalize the CFP emissions from the spectrum taken from oocytes expressing both fluorophores at 405 nm excitation. This procedure allows one to dissect the YFP emission spectrum, termed F405. F405 has two components: one is due to direct excitation of YFP () and the other is due to FRET (). F405 is normalized to the total YFP emission with 514 nm excitation at the same oocyte, F514. The resulting ratio, termed RatioA, can be expressed as . The direct excitation component in the calculated RatioA, termed RatioA0, is experimentally determined from a large population of oocytes expressing only YFP-tagged proteins. This allows precise calculations of the bleed-through of direct excitation of YFP by the 405 nm laser. The difference between RatioA and RatioA0, (RatioA−RatioA0) is directly proportional to FRET efficiency. The apparent FRET efficiency from an individual cell, Eapp, can be calculated as , where εD and εA are molar extinction coefficients for the donor and acceptor, respectively, at the donor excitation wavelength (Gao et al., 2007; Takanishi et al., 2006).
Dynamic FRET assay in oocytes
Oocytes were voltage clamped and their fluorescence measurements were taken simultaneously as described in (Berlin et al., 2010). Oocytes were kept at a holding potential of −100 mV for 20 sec and were gradually depolarized from −80 to +80 with step increments of 20 mV. Fluorescent signals were collected with a Zeiss 510META confocal microscope using its “channel mode”. Oocytes were excited with a 405-nm laser band only and the emission was filtered through the main beam splitter HFT405/514/633 nm and further separated by a secondary beam splitter, NFT515 nm. CFP and YFP fluorescences were collected by a 470–500 nm and 505–550 nm band pass filter, respectively, and directed into two separate photomultipliers. Under these settings, the leak of YFP into the CFP recording window is purely optical, very low (<1%) (Okamoto et al., 2004), and remained constant regardless of changes in FRET, thus not requiring any corrections. A region in the oocyte membrane area was sequentially imaged every second for 120 sec. For each oocyte, CFP and YFP intensities were normalized to initial CFP and YFP intensities and their FRET ratio (FYFP/FCFP) was calculated. Changes in FRET are reflected in changes in the FRET ratio.
FRET and secretion procedures in PC12 cells
PC12 cells preparation and transfection
PC12 cells were maintained at 37°C/5% CO2 in Dulbecco's modified Eagle's medium (DMEM) with high glucose (Sigma) supplemented with 10% bovine serum, 5% L-glutamine, 100 U/ml penicillin, and 0.1 mg/ml streptomycin. For imaging, cells were replated to a ∼50% confluence onto poly-L-lysine (Sigma)-coated 35 mm glass-bottom culture dishes and allowed to adhere overnight. Cells were transfected with 0.5 µg vesicular neuropeptide Y attached to red fluorescence protein (NPY-mRFP) cDNA alone or together with 0.3 µg of various CSYS probes cDNA, using Lipofectamine 2000 (Invitrogen). In the BoNT-C1 experiments, cells were transfected with 0.5 µg NPY-mRFP, 1.2 µg of BoNT-C1 and with 0.5 µg CSYS or 0.7 µg CSYS(R) cDNAs. The concentration of each probe was determined before transfection in a calibration experiment in order to obtain the same expression levels among all constructs in every experiment. Imaging experiments were conducted at room temperature, 24 hours after transfection. During the experiment the transfected cells were superfused through a 2-ml bath, with control and 105 mM hK stimulation solutions as described in (An and Almers, 2004). Since Ca+2 currents in PC12 cells are usually small (An and Almers, 2004; Taraska et al., 2003), in both solutions elevated [Ca2+] was used in order to obtain more reliable exocytosis in our PC12 cells.
Dynamic FRET assay in PC12
PC12 cells were imaged using a C-Apochromat 40×/1.2 NA water objective and excited with a 405-nm laser every 5 sec for a total of 400 sec. During the sequential imaging the cells were imaged in control solution for 100 sec before and after 200 sec using 105 mM hK stimulation, for a total imaging time of 400 sec. CFP and YFP fluorescences were collected with the same configuration described in the dynamic FRET assay in oocytes. YFP and CFP intensities at a region of interest (ROI) on the cell's plasma membrane were calculated and background fluorescence was quantified from an ROI in each image defined, in an area containing no fluorescent cells. Background-subtracted fluorescence intensity at each exposure time point was normalized to the average of the initial measurements in each cell (before hK solution was added). The FRET ratio of normalized intensities was denoted as FYFP/FCFP. Changes in FRET are reflected in changes in the FRET ratio.
Secretion assay in PC12
Cells were imaged every 5 sec for 400 sec using a C-Apochromat 40×/1.2 NA water objective. Cells were excited with a 543 nm excitation laser and the fluorescence signal was collected from an ROI region containing the whole cell, using a 560–615 nm bandpass filter. One hundred sec after beginning the measurement, the control solution was replaced by stimulation of 105 mM hK solution for the remaining experiments. Exocytosis was measured as a rapid reduction in fluorescence due to release of the NPY peptide in response to hK stimulation. In the control experiments no stimulation solution was added during the entire 400 sec of measurement or Ca2+ elevation was diminished by using 2 mM Ca2+ and by adding 200 µM cadmium (Ca2+ channel blocker) to the control and stimulation solutions.
Cytosolic Ca2+ imaging
PC12 cells transfected with CSYS were grown to ∼50% confluence in eight-well LabTec (Nunc) clusters with a #1.5 coverglass bottom. The cells were incubated for 1 hour in Hank's solution with 2 µM of fura-2 acetoxy-methyl ester (AM), at 37°C in a 6% CO2 atmosphere. Next, the cells were washed with Hank's solution and monitored for fura-2 fluorescence with a Nikon TMD microscope equipped with a high NA (1.3) ×40 oil immersion fluorescence objective (Nikon). Successive frame pairs (at 340 and 380 nm excitation) were acquired at 0.2-seconds intervals every 0.5–5 s. The Sutter DG-4 light source was used. Emission was measured at 510 nm. The emission signal was recorded with an Isis Photonic Science intensified camera using Universal Imaging Metafluor software. The cells were stimulated by a 105 mM hK solution, which was applied directly into the wells. The results are presented as the emission ratio, obtained by dividing the 340 nm image by the 380 nm image.
All statistical analyses were performed using SigmaStat Software. Results are presented as means±s.e.m. Multiple groups were compared by one-way analysis of variance (ANOVA), followed by a post hoc Bonferroni test or Dunnett's multiple comparison test. Two groups were compared using an unpaired two-tailed t-test. In the dynamic FRET and secretion experiments, the significance was tested ‘within groups’ by comparing each average time point measurement with the average measurements before stimulation, using a two-tailed t-test and ‘between groups’ by comparing each average time point measurement between two or more groups, using a t-test or one-way ANOVA, respectively. Asterisks indicate statistically significant differences as follows: *P<0.05; **P<0.001.
We thank Prof. Uri Ashery for critically reading the manuscript.
S.B., D.G.-A. and I.L. conceived the project; D.G.-A, N.B.-B, S.B., D.C. and I.L. designed the experiments; D.G.-A, N.B.-B. and D.C. performed the experiments; I.L., D.G.-A, N.B.-B, S.B., Y.O. and D.C. wrote the article.
This work was supported by the Israel Academy of Sciences [grant number 99/10 to I.L.]; the United States-Israel Binational Foundation [grant number 2009049 to I.L.]; and the Joan and Jaime Constantiner Institute of Molecular Genetics [grant to D.G.-A., N.B.-B.].