Abstract
Presynaptic inhibitory G protein-coupled receptors (GPCRs) can decrease neurotransmission by inducing interaction of Gβγ with the soluble N-ethylmaleimide-sensitive factor attachment protein receptor (SNARE) complex. We have shown that this action of Gβγ requires the carboxyl terminus of the 25-kDa synaptosome-associated protein (SNAP25) and is downstream of the well known inhibition of Ca2+ entry through voltage-gated calcium channels. We propose a mechanism in which Gβγ and synaptotagmin compete for binding to the SNARE complex. Here, we characterized the Gβγ interaction sites on syntaxin1A and SNAP25 and demonstrated an overlap of the Gβγ- and synaptotagmin I -binding regions on each member of the SNARE complex. Synaptotagmin competes in a Ca2+-sensitive manner with binding of Gβγ to SNAP25, syntaxin1A, and the assembled SNARE complex. We predict, based on these findings, that at high intracellular Ca2+ concentrations, Ca2+-synaptotagmin I can displace Gβγ binding and the Gβγ-dependent inhibition of exocytosis can be blocked. We tested this hypothesis in giant synapses of the lamprey spinal cord, where 5-HT works via Gβγ to inhibit neurotransmission (Blackmer et al., 2001). We showed that increased presynaptic Ca2+ suppresses the 5-HT- and Gβγ-dependent inhibition of exocytosis. We suggest that this effect may be due to Ca2+-dependent competition between Gβγ and synaptotagmin I for SNARE binding. This type of dynamic regulation may represent a novel mechanism for modifying transmitter release in a graded manner based on the history of action potentials that increase intracellular Ca2+ concentrations and of inhibitory signals through Gi-coupled GPCRs.
Modification of neurotransmitter release by presynaptic G protein-coupled receptors (GPCRs) is ubiquitous and vital to nervous system function. Gβγ inhibits synchronous exocytosis directly at the SNARE complex, the heart of the fusion machinery (Silinsky, 1984; Man-Son-Hing et al., 1989; Blackmer et al., 2001, 2005; Takahashi et al., 2001; Chen et al., 2005; Gerachshenko et al., 2005) and consequently modifies the quantal size of release by causing transient fusion of vesicles in both lamprey giant synapses and mammalian chromaffin cells (Chen et al., 2005; Photowala et al., 2006). G proteins may also inhibit exocytosis through inhibition of voltage-gated calcium channels, blunting the influx of Ca2+ required for exocytosis (Herlitze et al., 1996; Ikeda, 1996).
Synaptotagmin I is believed to mediate the requirement for Ca2+ in the initiation of exocytosis. Before the fusion event, vesicles are primed by formation of a stable protein structure, the ternary SNARE complex, comprising the plasma membrane t-SNAREs SNAP25 and syntaxin and the vesicular SNARE VAMP2 (Fig. 1A). Synaptotagmin, as the putative Ca2+ sensor, is thought to act as the trigger by which Ca2+ influx effects the fusion of primed vesicles (Bai et al., 2004; Maximov and Sudhof, 2005; Wang et al., 2006).
Given its central role in triggering exocytosis, interactions between synaptotagmins and the SNARE complex have been studied extensively. Upon Ca2+ binding to its C2 domains (Fig. 1A) during regulated exocytosis, synaptotagmin interacts with phospholipids and the SNARE complex (Perin et al., 1990; Schiavo et al., 1997). This interaction occurs through the H3 SNARE domain of syntaxin1A and the C terminus of SNAP25 (Bennett et al., 1992; Chapman et al., 1995; Gerona et al., 2000; Bai et al., 2004). Although the mechanism by which the SNARE complex initiates fusion is under debate (Rizo et al., 2006), interactions with synaptotagmin may cause stabilized ternary SNAREs to zip up, uniting the vesicular and plasma membranes (Margittai et al., 1999; Sorensen et al., 2006). Other evidence suggests that ternary SNARE complex formation is a concerted process with more than one transition state during initial and sustained dilation of a protein fusion pore (Han and Jackson, 2006).
Despite the controversy regarding SNARE complex-mediated membrane fusion, widespread consensus holds that synaptotagmin is the Ca2+ sensor for stimulated exocytosis and may provide the cue for full fusion versus kiss-and-run fusion. Mutagenesis studies, for example, indicate that fusion is inhibited in the presence of synaptotagmin mutants with diminished SNARE interactions (Bai et al., 2004). Furthermore, overexpression of different isoforms of synaptotagmin in PC-12 cells leads to different fusion pore opening durations, suggesting that the interaction of synaptotagmin with the SNARE complex and the membrane affects the nature of fusion pore formation and dilation (Wang et al., 2001). Gβγ signaling is another mediator of fusion pore dynamics (Blackmer et al., 2005; Chen et al., 2005; Gerachshenko et al., 2005; Photowala et al., 2006), but it is unclear how the Ca2+-dependent synaptotagmin/SNARE complex interaction is actively modulated by Gβγ (Searl and Silinsky, 2005).
We have shown that the nine C-terminal residues of SNAP25 that interact with synaptotagmin provide an important binding site for Gβγ. The resulting Gβγ/SNARE interaction inhibits transmitter release in vertebrate central synapses and mammalian neuroendocrine cells (Blackmer et al., 2005; Gerachshenko et al., 2005). We now identify Gβγ binding sites on the SNARE complex. In addition, we identify a role for Ca2+ and demonstrate competition between Gβγ and synaptotagmin for binding to SNARE complexes. We propose that modulation of exocytosis is subject to convergent and complementary pathways directed to alteration of synaptotagmin/SNARE complex interactions in which presynaptic Ca2+ both stimulates fusion and modifies the impact of Gi/o-coupled inhibitory GPCRs on exocytosis.
Materials and Methods
Plasmids. The open reading frames for the SNARE component proteins were subcloned into the pGEX6p1 vector (GE Healthcare, Chalfont St. Giles, Buckinghamshire, UK) for expression in bacteria. Individual truncation mutants were generated via a PCR-based strategy.
SNARE Protein Purification. Recombinant bacterially expressed glutathione transferase (GST) fusion proteins were expressed in Escherichia coli strain BL21(DE3). Protein expression was induced with 0.1 mM isopropyl β-d-thiogalactoside for 16 h at room temperature. Bacterial cultures were pelleted, washed with 1× phosphate-buffered saline, and then resuspend in lysis buffer [50 mM NaCl, 50 mM Tris, pH 8.0, 5 mM EDTA, 0.1% Triton X-100, 2 mM phenylmethylsulfonyl fluoride, and 1 mM dithiothreitol (DTT)]. Cells were lysed with a sonic dismembrator at 4°C. GST-SNAP25 and GST-VAMP2 were purified from cleared lysates by affinity chromatography on glutathione-agarose (GE Healthcare), following the manufacturer's instructions. While the proteins were bound to the column, the buffer was exchanged to 20 mM HEPES, pH 7.4, 100 mM NaCl, 0.05% n-octyl β-d-glucopyranoside (OG), and 5 mM dithiothreitol. In some cases, the proteins were eluted by cleaving from GST with PreScission protease (GE Healthcare) for 4 h at 4°C and then dialyzed extensively against 20 mM HEPES, pH 7.0, 80 mM KCl, 20 mM NaCl, and 0.05% OG. GST-syntaxin1A (lacking the transmembrane domain) was purified from the sonicated bacterial supernatant by affinity chromatography on glutathione-agarose (GE Healthcare) in 10 mM HEPES, pH 7.4, 0.05% OG, and 2 mM DTT. Protein concentrations were determined with a Bradford assay kit (Pierce, Rockford, IL), and purity was verified by SDS/PAGE analysis.
Binary and Ternary SNARE Complex Reassembly. An equimolar ratio (3 μM) of GST-syntaxin1A on glutathione-agarose and SNAP25 were incubated overnight at 4°C in 20 mM HEPES, pH 7.4, 100 mM NaCl, 0.1% OG, and 2.0 mM dithiothreitol. The binary t-SNARE complex was washed three times with phosphate-buffered saline and eluted from the column by removing GST with PreScission protease (GE Healthcare) for 4 h at 4°C. Equimolar protein-protein interaction was confirmed by SDS-PAGE/Coomassie staining analysis. To reassemble ternary SNARE complex, an equimolar ratio (3 μM) of GST-syntaxin1A on glutathione-agarose, VAMP2, and SNAP25 was incubated overnight at 4°C in 20 mM HEPES, pH 7.4, 100 mM NaCl, 0.1% OG, and 2.0 mM DTT. The SNARE proteins were washed and eluted from the column by removing GST. SDS- and heat-stable SNARE complex formation was verified by SDS/PAGE analysis with or without a 15-min incubation at 100°C.
Gβγ Purification and Labeling. Gβ1γ1 was purified from bovine retina as described previously (Mazzoni et al., 1991). Recombinant Gβ1γ2 was expressed in Sf9 cells and purified via a His6 tag on Gγ2 using nickel-nitrilotriacetic acid affinity chromatography (Sigma-Aldrich, St. Louis, MO). Fluorescence labeling of Gβ1γ1 and binding assays were conducted as described previously (Phillips and Cerione, 1991). In brief, purified Gβ1γ1 was dialyzed into labeling buffer (20 mM HEPES, pH 7.4, 5 mM MgCl2, 150 mM NaCl, and 10% glycerol), then mixed with 2-(4′-maleimidylanilino)naphthalene-6-sulfonic acid (MIANS) in a 5-fold molar excess. The reaction proceeded for 3 h at 4°C before quenching with 5 mM 2-mercaptoethanol. The MIANS-Gβ1γ1 complex was separated from unreacted MIANS using a PD-10 desalting column (GE Healthcare). MIANS-Gβ1γ1 was stored in aliquots at –80°C.
Fluorescence Binding and Competition Assay. All fluorescence measurements were performed in a fluorescence spectrophotometer (Cary Eclipse; Varian, Inc., Palo Alto, CA) at room temperature. In general, MIANS-Gβ1γ1was diluted into 0.5 ml of assay buffer (20 mM HEPES, pH 7.5, 5 mM MgCl2, 1 mM DTT, 0.1 M NaCl, and 1 mM EDTA) to a final concentration of 20 nM. The MIANS fluorescence was monitored with excitation at 322 nm and emission at 417 nm. All proteins purified as GST fusion proteins were cleaved from GST with PreScission protease (GE Healthcare) before analysis. The fluorescent changes caused by the addition of SNARE complexes were monitored continuously. Note that the amplitude of the fluorescence increase is not a measure of the affinity of the complex; rather, it reflects the specific site on fluorescently labeled Gβγ of interaction with each protein. There was no nonspecific binding of the free probe to the SNARE proteins and MIANS-Gβγ was resistant to photobleaching under experimental conditions (data not shown). For Gβγ/synaptotagmin competition assays, various concentrations of synaptotagmin with SNARE proteins were added to labeled Gβ1γ1 with the noted Ca2+concentrations, and fluorescence changes were monitored. The EC50 concentrations were determined by sigmoidal curve fitting using Prism software (GraphPad Software, San Diego, CA).
GST Pull-Down Assay. GST-SNAP25, GST-VAMP2, and GST-syntaxin1A (5 μg of each protein) were incubated with various amount of Gβ1γ2 proteins for 1 h at 4°C and washed three times with assay buffer (20 mM HEPES, pH 7.0, 80 mM KCl, 20 mM NaCl, and 0.1% OG). The complex was eluted with 20 μl of SDS sample buffer followed by separation via SDS-PAGE. Precipitated Gβ was detected using Western blotting with a rabbit anti-pan Gβ antibody (Santa Cruz Biotechnology, Santa Cruz, CA).
Electrophysiology and Microinjection. Lamprey spinal cord preparations (Petromyzon marinus) were performed as described previously (Wickelgren, 1977). Whole-cell patch-clamp recordings were achieved with a modified blind technique (Blanton et al., 1989) using pipettes (5–10 MΩ) containing solution of the following composition: 102.5 mM MeSO4, 1 mM NaCl, 1 mM MgCl2, 5 mM HEPES, and 0.05 mM EGTA, brought to a pH of 7.4 with KOH. All paired recordings to measure the effect of protein injection in this study were made at a minimum of 5 min after injection and less than 100 μm from the presynaptic injection site. Recombinant Gβ1γ2 protein was stored at –20°C in a solution containing 100 mM MeSO4, 5 mM HEPES, and 0.05 mM EGTA at pH of 7.4. Concentrations ranged from 3 to 5 mg/ml. For microinjection into the presynaptic terminal, 50% (by volume) of 1 M KMeSO4 was added to the Gβ1γ2-containing solution. Protein was pressure microinjected through presynaptic microelectrodes using Picospritzer II (General Valve, Fairfield, NJ).
Imaging. Fluorescence images were recorded with a modified confocal microscope (MRC600; Bio-Rad Laboratories, Hercules, CA) run with custom software available at http://alford.bios.uic.edu. Reticulospinal axons were labeled retrogradely with a dextran amine-conjugate form of the Ca2+-sensitive dye Fluo-4 (high affinity; Invitrogen, Carlsbad, CA). The dye was applied using a suction pipette fitted to the cut end, immediately after the end of the spinal cord was cut. The tissue was then incubated overnight for the dye to be transported throughout the axons. Images were collected at high speed by scanning a laser over a single line at 500 Hz. Imaging data were analyzed using NIH ImageJ software (http://rsb.info.nih.gov/ij/) on a Macintosh computer. ImageJ was used to calculate the brightness value (range of possible values, 10 bit) for each pixel in a field of view. For each individual axon of interest, the brightness values were measured, and after background subtraction, images were normalized to the baseline level of fluorescence to give (F + 1)/F values.
Calibration of the Ca2+Sensitive Dyes. The dye sensitivity to Ca2+ was determined using the same optical path (confocal microscope) that was used to measure Ca2+ transients in the tissue. Dyes (5 μM) were prepared in blends of two Ca2+ buffer standards (10 mM K2EGTA, 100 mM KCl, and 30 mM MOPS) and (10 mM CaEGTA, 100 mM KCl, and 30 mM MOPS) both at pH 7.2 (Invitrogen) to make an 11-point standard curve. The absolute Ca2+ values of the mix were corrected for the low temperatures used in this study (10°C). The dye buffer mix was placed between cover slips cooled from below with a liquid cooling system to 10°C and imaged from above with the 40× water immersion lens.
Western blots are representative of at least three independent experiments. Data points are presented as mean ± S.E.M. (*, p < 0.01; **, p < 0.001, two-tailed Student's t test).
Results
Gβγ Interacted with Individual SNARE Proteins and Ternary SNARE Complex. To understand the relationship between Gβγ and the SNARE complex in the modulation of exocytosis, we characterized sites of interaction by two complementary in vitro approaches, GST pulldown assays and fluorescent binding assays. In GST pulldowns, both components of the t-SNARE (syntaxin1A and SNAP25) interacted with Gβγ individually. In addition, the vesicular SNARE component, VAMP2, was found to interact directly with Gβγ (Fig. 1B). To demonstrate that the GST pulldown assay provides specific binding, we performed pulldown assays with Gα subunit and heterotrimeric G protein that contains Gαβγ. The Gβγ subunit, but not the Gα subunit or the heterotrimeric G proteins, interacts with SNARE proteins, suggesting that G protein interaction with SNARE occurs upon inhibitory G protein-coupled receptor activation that induces Gα and Gβγ dissociation (Supplementary Figure S1).
To quantify the relative contributions to Gβγ binding by the individual SNARE components, we employed a sensitive and quantitative fluorescence assay for measuring protein-protein interactions (Phillips and Cerione, 1991). Gβ1γ1 was labeled with the environmentally sensitive fluorescent probe MIANS. This probe serves to report interaction with a binding partner because it undergoes enhanced fluorescence when the local environment becomes more hydrophobic. In these assays, labeled Gβγ (final concentration, 20 nM) was used, whereas interactions with different concentrations of individual SNARE proteins or complexes were assessed through fluorescence intensity changes. Therefore, apparent affinities were estimated by determining concentrations of binding proteins at which MIANS fluorescence reached 50% of maximum (EC50).
This assay was used to quantify the interaction of Gβγ with the ternary SNARE complex and its constituent proteins (Fig. 1, C and D). The ternary SNARE complex indeed caused a dose-dependent increase in the fluorescence of MIANS-Gβγ, consistent with our previous findings that Gβγ and the SNARE complex interact in vitro (Blackmer et al., 2001, 2005) (Fig. 1C and EC50 = 0.27 μM). Likewise, we tested the apparent Gβγ-binding affinities of the individual SNARE components (Fig. 1D). Each of the three components caused a fluorescence increase, confirming direct interactions. The relative affinities followed the order syntaxin1A (EC50 = 0.33 μM), VAMP2 (0.94 μM), and SNAP25 (1.07 μM).
Gβγ Interacted with the Syntaxin1a H3 Motif and the C Terminus of SNAP25. Syntaxin1A has two major regions, the H3 SNARE domain and the modulatory Habc region (Fig. 2A). Truncation mutants representing the two major syntaxin1A regions were generated (Fig. 2A), and EC50 values for binding to 20 nM MIANS-labeled Gβγ were determined. The syntaxin1A H3 domain showed a right-shifted binding curve (EC50 = 0.89 μM; Fig. 2B) relative to the entire molecule (EC50 = 0.33 μM). Syntaxin1A (1–190) did not induce a significant Gβγ fluorescence change. These results, which were confirmed by GST pulldown assays (data not shown), indicate that Gβγ binds to the syntaxin1A H3 SNARE domain. However, because Gβγ bound to the H3 domain with an affinity lower than that for intact syntaxin1A, we cannot rule out a masked minor contribution of the Habc domain or the possibility that the isolated domain does not fold optimally (Jarvis et al., 2002).
Recent evidence suggests that nine amino acids from the C terminus of SNAP25 are crucial for interaction with Gβγ as well (Blackmer et al., 2005; Gerachshenko et al., 2005). Botulinum toxin A (BoNT/A) treatment, which removes nine C-terminal residues of SNAP25 (SNAP25Δ9) (Fig. 2C), diminishes Gβγ-mediated inhibition of exocytosis (Blackmer et al., 2005; Gerachshenko et al., 2005). In addition, a ternary SNARE complex comprising VAMP2, syntaxin1A, and SNAP25Δ9 shows reduced Gβγ binding compared with the wild-type SNARE complex (Gerachshenko et al., 2005). Here, we quantified the direct interaction of Gβγ with the C terminus of SNAP25. A truncation mutant of SNAP25 that mimics the BoNT/A cleavage product (SNAP25Δ9) was analyzed for its binding to Gβγ. Relative to full-length SNAP25, SNAP25Δ9 showed significantly decreased affinity for Gβγ (Fig. 2D).
In presynaptic terminals of neurons in situ and in cultured PC-12 and chromaffin cells, BoNT/E, which cleaves 26 amino acids from the C terminus of SNAP25 (Fig. 2C), inhibits Ca2+-dependent exocytosis (Ferrer-Montiel et al., 1998; Chen et al., 2001), consistent with an observed loss of interaction between synaptotagmin and SNAP25 (Chen et al., 1999; Gerona et al., 2000). We investigated whether a further truncated SNAP25 (SNAP25Δ26) also weakens binding to Gβγ. Fluorescent binding assays demonstrated that deletion of 26 amino acids caused a great loss of Gβγ binding (Fig. 2D). These data indicate that 26 C-terminal amino acids of SNAP25 are also critical for Gβγ binding.
Gβγ and Synaptotagmin Competed in Binary and Ternary SNARE Complexes in a Ca2+-Dependent Manner. It is noteworthy that the Gβγ-binding domains of the SNARE complex, the syntaxin1A H3 domain and the C terminus of SNAP25, are also known to constitute functionally important Ca2+-dependent synaptotagmin binding sites (Chapman et al., 1995; Gerona et al., 2000). The overlapping of binding sites suggests that there might be competition between Gβγ and synaptotagmin on the complexed SNAREs. To test whether Gβγ competes with synaptotagmin for binding to SNARE proteins in complex, we conducted competition assays in vitro. We used a mutant of synaptotagmin I (K326A, K327A) that exhibits wild-type SNARE binding affinity but holds a reduced propensity to oligomerize and to form insoluble aggregates in response to high concentrations of Ca2+ (Bai et al., 2004). In fluorescent binding assays, synaptotagmin does not interact significantly with MIANS-Gβγ (data not shown). We therefore tested competition between Gβγ and synaptotagmin for binding to the binary t-SNARE complex, comprising syntaxin1A and SNAP25. The interaction between Gβγ and binary t-SNARE complex in the absence of synaptotagmin was normalized to 100%. In 50 μM Ca2+, the interaction between MIANS-Gβγ and 1 μM binary t-SNARE complex was almost completely inhibited by 250 nM synaptotagmin I (Fig. 3A). Increasing synaptotagmin concentrations competed more effectively with Gβγ, resulting in reduced fluorescence changes (Fig. 3B), further indicating that Gβγ and synaptotagmin compete for SNARE binding.
Synchronous vesicle fusion is initiated by Ca2+, but additional Ca2+influx may modify large dense core vesicle fusion properties and possibly alter the quantal size of synaptic neurotransmitter release (Elhamdani et al., 2001, 2006; Harata et al., 2006). Because this modulation may reflect the Ca2+ sensitivity of synaptotagmin (Morimoto et al., 1995), we tested whether competition with Gβγ is also Ca2+ sensitive. We found that the competition between synaptotagmin and Gβγ for binary t-SNARE complex binding is dependent on intracellular Ca2+concentrations. In buffer lacking Ca2+, synaptotagmin caused a slight but statistically significant inhibition of the ability of t-SNAREs to increase the fluorescence of MIANS-Gβγ (Fig. 3C), possibly because of a limited Ca2+-independent interaction between synaptotagmin and t-SNAREs (Bennett et al., 1992; Sollner et al., 1993). However, upon addition of physiologically relevant concentrations of Ca2+, we observed an increased inhibition of Gβγ binding by synaptotagmin, with maximum effectiveness at 50 μM Ca2+ (Fig. 3C).
Interactions of both Gβγ and synaptotagmin with the ternary SNARE complex modify the dilation of the vesicle fusion pore (Bai et al., 2004; Photowala et al., 2006). Thus, we tested the hypothesis that Gβγ competes with synaptotagmin for binding to the ternary SNARE complex. We first measured the Gβγ/SNARE interaction by determining the change in fluorescence of 20 nM MIANS-Gβγ upon addition of the ternary SNARE complex. Addition of 1 μM SNARE complex induced an increase of fluorescence intensity (Fig. 3D). This increase was reduced by inclusion of 250 nM synaptotagmin, indicating that Gβγ and synaptotagmin compete for binding to the ternary SNARE complex. In addition, competition by synaptotagmin and the resulting reduced fluorescence changes were dependent on the concentration of synaptotagmin (Fig. 3E) and the concentration of Ca2+ (Fig. 3F). Together, these data suggest that indeed Gβγ and synaptotagmin compete for binding to the ternary SNARE complex.
Gβγ and Synaptotagmin Competed for Binding to the Individual t-SNAREs (Syntaxin 1A and SNAP25) but Not the Vesicular SNARE (VAMP2). The H3 domain of syntaxin1A and the C terminus of SNAP25 provide interaction regions for Gβγ and synaptotagmin. We saw significantly more efficient competition between Gβγ and synaptotagmin in the context of binary t-SNAREs than ternary SNARE complexes. Therefore, we measured the individual contributions of SNARE components to competition between Gβγ and synaptotagmin.
First, to determine whether Gβγ and synaptotagmin compete for binding to syntaxin1A or SNAP25, the interaction of 20 nM MIANS-Gβγ with syntaxin1A and SNAP25 was monitored with or without preincubation of the SNARE proteins with 20 nM synaptotagmin. The interactions between Gβγ and SNARE proteins were unaffected by Ca2+ in the absence of synaptotagmin (data not shown). In the absence of Ca2+, we observed a slight, statistically insignificant decrease in Gβγ binding to both syntaxin1A (Fig. 4A) and SNAP25 (Fig. 4B) upon preincubation of the SNARE proteins with 20 nM synaptotagmin. However, in 100 μM Ca2+, the ability of synaptotagmin to compete for binding to SNAREs was greatly enhanced. Syntaxin1A binding to Gβγ was reduced by 48% (Fig. 4A), and binding of SNAP25 to Gβγ was reduced by 78% (Fig. 4B). Thus, Gβγ competes with Ca2+-synaptotagmin for binding to both syntaxin1A and SNAP25.
We observed a direct interaction between Gβγ and VAMP2 (Fig. 1, B and D). An earlier study suggested that synaptotagmin interacts with VAMP2 (Fukuda, 2002), leading to the possibility that this protein forms an additional site of competition between Gβγ and synaptotagmin. However, the existence of a direct synaptotagmin/VAMP2 interaction remains controversial (Chapman et al., 1995). Consequently, we monitored the Gβγ interaction with VAMP2 with or without preincubation of VAMP2 with synaptotagmin in the presence or absence of Ca2+. No alteration in Gβγ binding to VAMP2 by synaptotagmin was observed, suggesting that Gβγ and synaptotagmin do not share a binding region on VAMP2 (Fig. 4C). This result may explain why competition at the ternary SNARE complex seems to favor Gβγ binding compared with competition at the binary t-SNARE: in the ternary SNARE complex, Gβγ binding is enhanced by an additional binding site within VAMP2 that is unaffected by synaptotagmin I. Thus, its ability to compete with Gβγ in the ternary SNARE complex is decreased.
The C Terminus of SNAP25 in the Ternary SNARE Complex Was Important for Gβγ-Synaptotagmin Competition. We have shown that BoNT/A, which specifically removes the C-terminal nine amino acids from SNAP25, greatly reduced the inhibitory effect of Gβγ on regulated exocytosis (Blackmer et al., 2001; Gerachshenko et al., 2005). We propose that this reduction can be explained if removal of these residues diminishes Gβγ competition with synaptotagmin for binding to the t-SNAREs and the ternary SNARE complex in vivo. Therefore, we examined the contribution of the C terminus of SNAP25 in the competition between Gβγ and synaptotagmin within the ternary SNARE complex. It is noteworthy that whereas BoNT/A cleavage of SNAP25 reduced binding of synaptotagmin and Gβγ individually, significant interactions were still observed (Fig. 2D). Thus, it was important to quantify the interactions using the fluorescent binding assay to determine whether this cleavage shifted the Gβγ/synaptotagmin competition toward synaptotagmin binding.
To test the relative contributions of the C terminus of SNAP25 in binding Gβγ and synaptotagmin, we quantified binding to ternary SNARE complexes comprising syntaxin1A, VAMP2, and SNAP25Δ9. Competition by synaptotagmin for Gβγ binding to SNARE complex that contains SNAP25Δ9 was increased compared with the wild-type SNARE complex at 50 μM Ca2+ (Fig. 5A). This finding is consistent with a model in which the nine C-terminal amino acids of SNAP25 represent an important determinant mediating exclusion of synaptotagmin binding by Gβγ.
We then determined the concentration dependence of the synaptotagmin competition. The strength of synaptotagmin competition in SNARE complexes formed with SNAP25Δ9 increased dramatically compared with the wild type SNARE complex (Fig. 5B). These results suggest that the C-terminal nine amino acids of SNAP25 are more critical for Gβγ than for synaptotagmin binding. These data support the dramatic loss of inhibitory effect of Gβγ in the BoNT/A-treated synapse (Gerachshenko et al., 2005).
The 5-HT Mediated Regulation of Exocytosis Was Diminished with High [Ca2+] at the Lamprey Giant Synapse. What role does Gβγ-synaptotagmin I competition for SNARE complexes play at synapses? At lamprey giant synapses, we have shown that 5-HT inhibits transmitter release through direct Gβγ/SNARE complex interactions. 5-HT mediated Gβγ does not inhibit voltage-gated Ca2+ channels in this preparation but acts directly on the ternary SNARE complex to inhibit exocytosis (Blackmer et al., 2001; Takahashi et al., 2001; Gerachshenko et al., 2005). From our in vitro study, we hypothesized that Gβγ-mediated inhibition requires competition between Gβγ and Ca2+-dependent synaptotagmin binding on the SNARE complex. These findings suggested that the effect of GPCR-mediated activation of Gβγ would be mitigated by raising the synaptic terminal Ca2+ concentration.
We tested this possibility using paired recordings between giant axons and postsynaptic neurons in lamprey synapse. To increase action potential-evoked intracellular Ca2+ concentrations, we raised extracellular Ca2+ concentrations (Fig. 6, A–C). We first quantified the consequent effect on evoked presynaptic Ca2+ entry. Axons were filled with a Ca2+ sensitive dye (Fluo-4 dextran high affinity) and imaged with confocal microscopy. Action potential-evoked presynaptic fluorescence transients were imaged under increasing extracellular Ca2+ (2.6, 10, and 50 mM; Fig. 6, B and C, and data not shown). The relationship between extracellular Ca2+ concentrations and the fluorescence transient was alinear (Fig. 6E). Therefore, to calculate the relationship between extracellular Ca2+ and the intracellular Ca2+ transient amplitude, we determined the peak intracellular Ca2+ during the transients. The affinity of the dye (1.2 μM) was obtained from fluorescence curves that showed a 120-fold increase in fluorescence between 0 μM Ca2+ and saturation (41 μM) at a physiological temperature. Using these data and the resting [Ca2+] in lamprey axons (approximately 100 nM), we estimated the peak transient intracellular Ca2+ (Fig. 6E). Extracellular Ca2+ (50 mM) slightly increased basal presynaptic intracellular Ca2+ (Fig. 6D). However, the evoked fluorescence transient, and therefore its peak Ca2+ concentration, was markedly increased (Fig. 6E). Note that absolute Ca2+ concentrations calculated in this way underestimate the concentration at the vesicle fusion site because we cannot image with sufficient resolution to resolve the close spatial relationship between Ca2+ channels and the release machinery (Adler et al., 1991). Consequently, we used these data to calculate the relative amplitude of the Ca2+ transient at extracellular concentrations of 10 and 50 mM (Fig. 6F) normalized to control (2.6 mM).
We then tested the effect of increased Ca2+ on 5-HT-mediated presynaptic inhibition. At a control extracellular Ca2+ concentration (2.6 mM), 5-HT (600 nM) reduced the EPSC to 55.9 ± 7.3% of control (n = 5; P < 0.05). The EPSC fully recovered upon 5-HT removal (Fig. 7A, top). Extracellular Ca2+ (50 mM) increased the EPSC amplitude to 129.7 ± 4.7% of control (n = 5; P < 0.01), but 600 nM 5-HT caused no inhibition (Fig. 7A, bottom; EPSC was 97.1 ± 8.1% of the EPSC modified by 50 mM Ca2+; n = 5 pairs). At 50 mM extracellular Ca2+, even a saturating dose of 5-HT (10 μM) resulted in only a slight reduction in the EPSC [response was 71.3 ± 0.9% of the control response modified by high Ca2+ (n = 3, P > 0.05), compared with a reduction to 20% of control in normal Ca2+ (Takahashi et al., 2001)] (Fig. 7, B and C). We also confirmed that 5-HT does not change Ca2+ influx in lamprey synapses (Blackmer et al., 2001; Takahashi et al., 2001) treated with 50 mM extracellular Ca2+ concentration (Supplementary Figure S2). Thus, increased intracellular Ca2+ transients attenuated GPCR mediated presynaptic inhibition.
High Extracellular Ca2+Attenuated the Presynaptic Inhibitory Effect of Gβγ. We have shown that the reduction of the EPSC upon activation of 5-HT receptors is caused by liberated Gβγ (Blackmer et al., 2001; Gerachshenko et al., 2005). Here, we examined the effects of manipulating intracellular Ca2+ on EPSC inhibition by free Gβγ introduced through direct injection of purified Gβγ protein into presynaptic axons. We compared Gβγ-mediated inhibition of exocytosis at a control (2.6 mM) and a raised Ca2+ concentration (10 mM); the latter increased the EPSC amplitude (Fig. 8A) to 122 ± 2% of control (n = 3). In four more paired recordings, we probed the effect of raising Ca2+ concentrations on inhibition by presynaptic Gβγ. In control Ca2+ (2.6 mM), Gβγ injection significantly reduced the EPSC, as described previously (Blackmer et al., 2001), to 37 ± 12% (Fig. 8, B and C). In the same pairs, the Ca2+ concentration was raised to 10 mM and Gβγ inhibition was reversed (10 mM; Fig. 8, B and C; EPSC was 100 ± 9% of control). Thus raised Ca2+ transient amplitudes significantly attenuated the inhibitory effect of presynaptic Gβγ. The effects of both 5-HT and presynaptic Gβγ injection are attenuated by enhancing the amplitude of the action potential evoked presynaptic Ca2+ transient.
Discussion
In primed vesicles, Ca2+ effects exocytosis in part by increasing synaptotagmin's affinity for the primed SNARE complex (Perin et al., 1990; Schiavo et al., 1997; Yoshihara and Littleton, 2002; Bai et al., 2004; Wang et al., 2006). Gβγ liberated by inhibitory Gi/o-coupled GPCRs inhibits neurotransmitter release by at least two mechanisms: inhibition of presynaptic Ca2+ entry at many synapses by modulating voltage-gated calcium channels (Herlitze et al., 1996; Ikeda, 1996), as well as interaction with the ternary SNARE complex late in membrane fusion (Blackmer et al., 2005; Gerachshenko et al., 2005). Here we show that Gβγ interacts with SNARE proteins and that this binding is not directly modified by calcium. Gβγ interacts with VAMP2 at a site different from synaptotagmin's interaction. Perhaps the Gβγ/VAMP2 interaction strengthens Gβγ binding to the SNARE complex, whereas occlusion of synaptotagmin I binding occurs through the t-SNARE components (syntaxin1A and SNAP25) in the ternary SNARE complex. In addition, Gβγ binds directly to the t-SNAREs SNAP25 and syntaxin1A at sites similar to those used by synaptotagmin to interact with these proteins in a Ca2+-dependent manner. This sets up a potent competition with the Ca2+ sensor to inhibit regulated exocytosis. This differential Ca2+ requirement can lead to a Ca2+-sensitivity of the Gβγ inhibition of SNARE/synaptotagmin complex formation: at low and moderate Ca2+concentrations, the presence of inhibitory signals generates Gβγ that competes with synaptotagmin for binding to SNARE, leading to inhibition of exocytosis. At high Ca2+concentrations, the affinity of synaptotagmin for SNARE increases, whereas the Gβγ/SNARE affinity remains steady, and thus high synaptic activity overrides the inhibition by Gβγ.
At the synapse, after C-terminal truncation of SNAP-25 with BoNT/A, exocytosis recovers in high Ca2+ in many preparations, including the lamprey (Gerachshenko et al., 2005), presumably because high Ca2+ enhances the residual synaptotagmin/SNARE interaction that remains after BoNT/A treatment. However, the affinity of Gβγ for the modified SNARE complex is much lower, and Gβγ therefore competes less well with synaptotagmin for the same site, particularly at raised intracellular Ca2+. These findings led to a testable hypothesis: at the synapse, larger evoked Ca2+ transients should prevent Gβγ-mediated presynaptic inhibition induced by 5-HT. Accordingly, in the lamprey giant synapse, although 5-HT powerfully inhibits exocytosis through Gβγ interaction with the SNARE complex (Silinsky, 1984; Man-Son-Hing et al., 1989; Blackmer et al., 2001, 2005; Takahashi et al., 2001; Chen et al., 2005; Gerachshenko et al., 2005), we observed that enhancing the Ca2+ transient reversed this inhibition (Figs. 6 and 7). These data are consistent with the biochemical results showing that inhibition of synaptotagmin/SNARE complex interaction by Gβγ is sensitive to Ca2+. Although Ca2+ could have other effects in the synapse, the fact that synaptotagmin is thought to be the major Ca2+ receptor for exocytosis (Bai et al., 2004; Maximov and Sudhof, 2005; Wang et al., 2006) suggests that Gβγ may work physiologically by competing with Ca2+-synaptotagmin for SNARE, leading to inhibition of exocytosis. Gβγ also has many other targets, and enhancing intracellular Ca2+ could modify the effects of these interactions on exocytosis. The most obvious other Gβγ effector is the voltage-gated Ca2+ channel, which has been shown by many investigators to be inhibited by Gβγ. However, Gβγ has no effect on Ca2+ entry even at 50 mM extracellular Ca2+, because 5-HT does not modify evoked Ca2+ entry (Supplementary Figure S2).
This Ca2+-dependent regulation may complete a physiological feedback loop, because Gβγ may control both Ca2+ influx through voltage-gated Ca2+ channels (Herlitze et al., 1996; Ikeda, 1996) and access of the Ca2+ effector synaptotagmin to its site on the SNARE. Our data suggest that this competition is elegantly controlled by overlap of the binding sites of Gβγ and synaptotagmin at the SNARE proteins SNAP25 and syntaxin. In a further level of regulation, during repetitive stimulation, buildup of intracellular Ca2+ concentration may reverse Gβγ inhibition. Thus, depending on presynaptic Gi/o-coupled GPCR activation and Ca2+ influx in response to synaptic activity, a range of efficiencies of exocytosis potentially exists. Thus, we predict that a complex and dynamic regulation of vesicle fusion properties will occur. Finally, at low levels of inhibitory GPCR signaling, Gβγ causes kiss-and-run exocytosis in lamprey synapses (Photowala et al., 2006) and adrenal chromaffin cells (data not shown) (Chen et al., 2005). In hippocampal neurons, sustained stimulus trains forcing high presynaptic Ca2+ concentrations change fusion from a state dominated by kiss-and-run to one favoring full fusion (Harata et al., 2006). However, the exact Ca2+ concentrations required to reverse Gβγ-mediated presynaptic inhibition are likely to exceed those obtained by residual Ca2+ during short bursts of stimuli, in which saturating doses of 5-HT remains effective (Takahashi et al., 2001). Thus, we speculate that the dynamic relationship between inhibitory GPCRs and presynaptic Ca2+ constantly modifies the nature of the vesicle fusion event itself.
Acknowledgments
We thank Dr. T. F. Martin (University of Wisconsin, Madison, Wisconsin) for the synaptotagmin C2AB cDNA, and Drs. K. P. M. Currie, A. Preininger, C. Wells (Vanderbilt University) and T. Blackmer (Oregon Health Sciences University) for helpful discussions.
Footnotes
-
This work was supported by National Institutes of Health grants EY10291 (to H.E.H.) and MH64763 (to S.A.) and by an American Heart Associate predoctoral fellowship (to E.J.Y.).
-
Article, publication date, and citation information can be found at http://molpharm.aspetjournals.org.
-
doi:10.1124/mol.107.039446.
-
ABBREVIATIONS: GPCR, G protein-coupled receptor; SNARE, soluble N-ethylmaleimide-sensitive factor attachment protein receptor; t-SNARE, target membrane-associated snare; SNAP25, 25-kDa synaptosome-associated protein; DTT, dithiothreitol; OG, n-octyl β-d-glucopyranoside; PAGE, polyacrylamide gel electrophoresis; GST, glutathione transferase; MIANS, 2-(4′-maleimidylanilino) naphthalene-6-sulfonic acid; MOPS, 3-(N-morpholino)propanesulfonic acid; BoNT/A, botulinum toxin A; EPSC, excitatory postsynaptic current; 5-HT, 5-hydroxytryptamine.
-
↵ The online version of this article (available at http://molpharm.aspetjournals.org) contains supplemental material.
- Received June 27, 2007.
- Accepted August 22, 2007.
- The American Society for Pharmacology and Experimental Therapeutics