Arg206 of SNAP-25 is essential for neuroexocytosis at the Drosophila melanogaster neuromuscular junction

An analysis of SNAP-25 isoform sequences indicates that there is a highly conserved arginine residue (198 in vertebrates, 206 in the genus Drosophila) within the C-terminal region, which is cleaved by botulinum neurotoxin A, with consequent blockade of neuroexocytosis. The possibility that it may play an important role in the function of the neuroexocytosis machinery was tested at neuromuscular junctions of Drosophila melanogaster larvae expressing SNAP-25 in which Arg206 had been replaced by alanine. Electrophysiological recordings of spontaneous and evoked neurotransmitter release under different conditions as well as testing for the assembly of the SNARE complex indicate that this residue, which is at the P1′ position of the botulinum neurotoxin A cleavage site, plays an essential role in neuroexocytosis. Computer graphic modelling suggests that this arginine residue mediates protein–protein contacts within a rosette of SNARE complexes that assembles to mediate the fusion of synaptic vesicles with the presynaptic plasma membrane.


Introduction
Convincing evidence for the key role of the three neuronal SNARE proteins in neurotransmitter release derives from its complete blockade exerted by tetanus and botulinum neurotoxins (Schiavo et al., 1992;Blasi et al., 1993;Schiavo et al., 1993a;Rossetto et al., 2001). Botulinum neurotoxin type A (BoNT/A) cleaves SNAP-25 nine residues from the C-terminus (Schiavo et al., 1993b;Binz et al., 1994) suggesting that this part of the molecule plays a central role in the assembly and/or function of the neuroexocytosis machinery. We have begun a detailed investigation of this SNAP-25 region by site-directed mutagenesis in Drosophila melanogaster coupled to electrophysiological recording of the Drosophila larva neuromuscular junction (NMJ). We started with an analysis of the available sequences of neuronal SNAP-25 and related isoforms reported in databases. As expected, mapping of the conservation profiles into the structure of the SNARE complex (Chen et al., 2002) shows a high conservation for residues involved in intermolecular contacts within each SNARE four-helix bundle Jahn and Scheller, 2006;Sorensen, 2009). We have, as well, considered external residues which are essential for possible protein-protein contacts that are necessary to form a rosette of SNARE complexes. The formation and involvement of such a rosette in neuroexocytosis is suggested by several findings and considerations Hua and Scheller, 2001;Tokumaru et al., 2001;Keller et al., 2004;Montecucco et al., 2005;Rickman et al., 2005;Lu et al., 2008;Sorensen, 2009). There is a single amino acidic residue that is fully conserved among species and isoforms of SNAP-25, not involved in the zippering of the SNARE complex, and this is Arg198 (mouse and human SNAP-25 numbering), which points outside of the four-helix bundle SNARE complex in the C-terminal of SNAP-25. This residue is at Arg206 of SNAP-25 is essential for neuroexocytosis at the Drosophila melanogaster neuromuscular junction the P 1 Ј site of the peptide bond cleaved by BoNT/A (de Paiva et al., 1993).
On this basis, we hypothesize that site-directed substitution of Arg198 with a neutral residue could interfere, in vivo, with the function of the neuroexocytosis apparatus in such a way as to be detectable by electrophysiological recordings at the neuromuscular junction. To test this possibility, we used Drosophila melanogaster, as an animal model. Arg198 of human SNAP-25 corresponds to Arg206 of D. melanogaster SNAP-25. We generated transgenic lines of D. melanogaster carrying SNAP-25 with an alanine at position 206. The transgene was expressed in a wild-type background and resulted in a significant reduction of both evoked and spontaneous neurotransmitter release at the neuromuscular junction in third instar larvae, relative to that in controls, in the absence of overt NMJ morphological defects. This finding is discussed and interpreted in terms of a rosette model for the apparatus that mediates fusion of synaptic vesicles with the presynaptic membrane.

Expression of the SNAP-25 R206A mutant in the nervous system of D. melanogaster
The expression of SNAP-25 R206A using the nervous system driver ElavGAL4, is illustrated in Fig. 1A, which shows a representative semiquantitative PCR performed on third instar larval brain. In Fig. 1B The average miniature end-plate potential (MEPP) amplitudes (in current-clamp conditions) were: 0.97±0.06 mV and 0.94±0.02    Fig. 3C).
Spontaneous fusions occur in a random manner, each event being independent from the previous and next one. Accordingly, the frequency distribution of MEPCs inter-event intervals is well described by a Poisson distribution in both controls and SNAP-25 R206A mutants (Fig. 3D), indicating that the random nature of spontaneous fusion was not altered by the presence of SNAP-25 R206A .
The presence of SNAP-25 R206A at the D. melanogaster neuromuscular junction does not alter the Ca 2+ dependency of neuroexocytosis or the time course of formation of the SNARE complex SNAP-25 is involved in the control of Ca 2+ dynamics (close to) at the active zone (Pozzi et al., 2008) and interacts with the voltagegated Ca 2+ channel via its C-terminus (Catterall, 1999;Zhong et al., 1999). Indeed, it has been well documented that high external Ca 2+ concentrations can partially rescue the neuroexocytosisblocking effect of botulinum neurotoxin type A which cleaves SNAP-25 at the Gln198-Arg199 peptide bond (Schiavo et al., 1993b;Binz et al., 1994). To test for the possibility that the Arg-Ala replacement affects the Ca 2+ sensitivity of neuroexocytosis, the Ca 2+ dependence of evoked neurotransmitter release was measured. We observed a reduced Ca 2+ sensitivity in SNAP-25 R206A compared with SNAP-25 WT ( Fig. 5A) with no change in the slope of the linear fit to the data of voltage change recorded at different  extracellular Ca 2+ concentrations (Fig. 5B). This result suggests that the Ca 2+ cooperativity of the process is unchanged between SNAP-25 R206A and SNAP-25 WT transgenic lines.
The probability of fusion of a synaptic vesicle is strictly dependent on the correct formation of SNARE complexes (Sudhof and Rothman, 2009;Sorensen, 2009) and the results obtained here could reflect defective SNARE complex assembly. We therefore assayed the extent and time-course of the formation of SNARE complexes using GST-SNAP-25 WT or the mutant GST-SNAP-25 R198A and the cytoplasmic domains of VAMP and syntaxin using established methods Fasshauer et al., 2002;Hu et al., 2002). Supplementary material Fig. S1A shows that no difference in the formation of the SNARE complex in vitro was detected between the wild-type and mutated proteins. Measuring the time-course of the SNARE complex formation is a more sensitive and relevant assay. Supplementary material Fig. S1B shows that both wild-type and mutated SNAP-25 very rapidly formed a SNARE complex with VAMP and syntaxin as found previously (Fasshauer et al., 2002) with no difference in the time course. This is a necessary control, the result of which was almost expected as the mutated residue points outward in the crystallographic structure of the SNARE complex (Sutton et al., 1998;Brunger et al., 2009) and C-terminal mutations of SNAP-25 3279 SNARE supercomplex at the Drosophila NMJ have been shown not to affect SNARE assembly (Sorensen et al., 2006).

The presence of SNAP-25 R206A does not alter morphology and morphometry of the D. melanogaster neuromuscular junction
As other Drosophila synaptic mutants with alteration in synaptic function were reported to show a morphological change at the level of the NMJ (Budnik and Ruiz-Canada, 2006), we analysed the morphology and morphometry of the NMJ after staining with the neuronal-membrane-specific anti-HRP antibody. Fig. 6A shows representative images of the NMJ of muscles 6 and 7 in the abdominal segment A3 of both SNAP-25 WT and SNAP-25 R206A larvae. Morphometrical analysis indicates that neither the total number of boutons per NMJ (Fig. 6B) nor the diameter of type 1b and 1s boutons were changed in SNAP-25 R206A compared with SNAP-25 WT (Fig. 6C). We did not extend this analysis to the electron microscopic level because the removal from SNAP-25 of  the C-terminal segment beginning with Arg198 by BoNT/A has been previously shown not to change the appearance of the synaptic terminal at this level (Neale et al., 1999).

Discussion
The present finding that the replacement of the conserved Arg198 residue of SNAP-25 with alanine affects the probability of neurotransmitter release is very relevant because it identifies a key residue in this process of paramount importance for the function of the nervous system. However, as the process is a complex one, the end result is open to different interpretations. We provided evidence against the possibilities that: (1) this change might have affected the assembly of the SNARE complex; (2) the calcium concentration dependence of the phenomenon could have been altered; (3) the morphology and/or morphometry of the NMJ was changed. This residue is not involved in the protein-protein interactions between the three SNARE proteins, and points outward, suggesting that it may be involved in the interaction with other proteins of the neuroexocytosis machinery. There are several scattered pieces of evidence that a number of SNARE complexes are involved in the exocytosis of vesicles and granules. Multimers containing a variable number of SNARE complexes have been observed under various circumstances in vitro Antonin et al., 2000) and were isolated from detergent-treated squid synaptosomes (Tokumaru et al., 2001). The Ca 2+ -cooperativity of neurotransmitter release was found to be linked to the number of SNARE proteins (Stewart et al., 2000). Remarkably, star-shaped oligomers, comprising three to four SNARE complexes, were isolated from detergent-treated homogenized bovine brain with a monoclonal antibody directed toward the acetylated N-terminus of SNAP-25 (Rickman et al., 2005). In vitro experiments performed with different biophysical approaches indicate the formation of oligomers of three to ten SNARE complexes (Karatekin et al., 2010;Yersin et al., 2003;Lu et al., 2008).
Using a peptide that inhibits granule fusion in a chromaffin cell line, Hua and Scheller (Hua and Scheller, 2001) estimated that a minimum of three SNARE complexes is sufficient to support exocytosis. In the same PC12 cells, mutagenesis of the transmembrane domain of the SNARE presynaptic membrane protein syntaxin led to an estimate of five to eight syntaxin molecules being involved in catecholamine release (Han et al., 2004). However, careful experiments performed with mouse spinal cord motoneurons and with the frog NMJ intoxicated with botulinum neurotoxins suggested a higher figure (Raciborska et al., 1998;Keller and Neale, 2001;Keller et al., 2004;Montecucco et al., 2005). The involvement of multi SNARE complexes in neuroexocytosis also accounts for the strikingly long duration of BoNT/A poisoning (>3 months in human skeletal muscles) (Eleopra et al., 1998;Meunier et al., 2003) and for the fact that transfection of BoNT/A-truncated SNAP-25 inhibits exocytosis (Huang et al., 1998;Wei et al., 2000;Fang et al., 2008) [for a discussion see Montecucco et al. (Montecucco et al., 2005)]. On the basis of this rich set of data indicating the role of an oligomeric SNARE supercomplex in exocytosis, we have modelled here a ten-SNAREcomplex rosette and have run molecular dynamics simulations (Fig. 7). Such modelling is justified by the fact that the three SNARE proteins are sufficient to promote membrane fusion in vitro and can, therefore provide relevant information on proteinprotein contacts among SNARE complexes in membranes (Sudhof and Rothman, 2009;Karatekin et al., 2010). Given the necessity of forming a ring of 'petals' to define a central area where 3280 Journal of Cell Science 123 (19) membrane fusion may take place, such an approach shows that few protein-protein contacts between SNARE complexes are involved, and Arg198 is at the centre of the contact areas between the petals of the rosette. Changing the number of petals of the rosette between eight and 13 results in little alteration in the amount of proteinprotein contacts between each SNARE complex, while Arg198 remains in a central position within the area of contact. The simple replacement of this charged residue with a helix-promoting, but uncharged, alanine residue does not appear to change the secondary structure, and indeed we did not find any effect of the alanine replacement in the rate and extent of assembly of the SNARE complex in vitro (supplementary material Fig. S1). However, this single mutated SNAP-25 is sufficient to decrease the number of miniature end plate potential events as well as evoked end plate potentials. It is even more remarkable that it does so in a wild-type SNAP-25 background at the Drosophila larva NMJ. Although further analysis with expression of the mutant SNAP-25 in a null background is required to substantiate the present results, the consistency of the data reported here suggests that the presence of a single mutant SNARE complex in the very critical region of protein-protein contact between the petals of the rosette is sufficient to block the activity of the rosette. This does explain the dominantnegative nature of the mutation introduced here with respect to neuroexocytosis, and fits well with the specific action of BoNT/A and BoNT/C which cleave SNAP-25 just before and after Arg198, respectively, causing a dominant-negative effect with the characteristic long duration of action of these two neurotoxins (Eleopra et al., 1997;Eleopra et al., 1998;Meunier et al., 2003). The present model also explains the remarkable findings that the replacements of Lys201 and Leu203 with a negatively charged glutamate residue do not affect SNARE assembly but inhibits exocytosis in chromaffin cells (Criado et al., 1999;Gil et al., 2002) as these changes are disturbing the protein-protein contacts between adjacent SNARE complexes.

Plasmid generation and germline transformation in Drosophila
Two isoforms of SNAP-25 [SNAP-25 R206A and one wild-type SNAP-25 WT ] were produced by PCR using primers carrying point mutations as necessary. The ORFs coding for SNAP-25 was amplified from the full-length cDNA derived Fig. 7. Model of the SNARE complex rosette apparatus of neuroexocytosis. The number of SNARE complexes used here (ten) derives from an analysis of different experiments performed with botulinum neurotoxins type A and E on the frog neuromuscular junction (Raciborska et al., 1998) and on mouse spinal cord motoneurons (Keller and Neale, 2004). Clearly, the number ten is an approximate value, but it should be noted that using 8 to 13 complexes does not significantly change the position of the SNAP-25 Arg with respect to the neighbouring partner. from w 1118 flies using the following oligonucleotides: sense SNAP-25 WT , 5Ј-CGGAATTC ATGCCAGCGGATCCATCTGAAG-3Ј; antisense SNAP-25 WT , 5Ј-CTCTCGAGT TACTTTAATAGTTGATGTGCC-3Ј; antisense SNAP-25 R206A , 5Ј-CTCTCGAG TTACTTTAATAGTTGATGTGCCgcTTGATTAGC-3Ј. Sense and antisense primers have at their 5Ј ends the restriction sites for EcoRI and XhoI, respectively (underlined nucleotides). Mutated sites are in lower case letters.
After the initial denaturation step at 94°C for 3 minutes, amplification was achieved through 35 cycles at 94°C for 30 seconds, 55°C for 30 seconds, and 72°C for 2 minutes. A final extension reaction was carried out for 7 minutes at 72°C. The obtained 655 bp PCR products were ligated into the pCR2.1-TOPO vector (Invitrogen, USA) which was used to transform One Shot TOP10 E. coli cells. Positive clones were detected by -galactosidase screening and sequencing (BRM Genomics, Italy). Using conventional restriction enzyme digestion techniques, sequences were extracted with EcoRI and XhoI and ligated into the pUAST Drosophila transformation vector.
The resulting plasmids, pUAS-SNAP-25 WT and pUAS-SNAP-25 R206A were again sequenced to ensure that the mutation occurred only at the position of interest and to verify the proper orientation inside the vector of each ORF (open reading frame). These plasmids were then used for embryo transformation. P-element-mediated germline transformation was done using a Drosophila embryo injection service (Best Gene Inc, CA, USA). Briefly, the constructs with the w + marker were injected into w 1118 embryos. Nine independent transformant lines were established for each construct. FM7 (X chromosome), CyO (chromosome 2) and TM3(sb) (chromosome 3) balancers were used to balance these lines.
Lines carrying the UAS-SNAP-25 WT transgene and the UAS-SNAP-25 R206A transgene were chosen on the basis of their homozygous viability. These new alleles of SNAP-25 were expressed in flies with a SNAP-25 wild-type background using the GAL4/UAS system. ElavGAL4 virgin females were crossed to homozygous UAS-SNAP-25 WT and UAS-SNAP-25 R206A males.
Flies were raised on a standard yeast-glucose-agar medium and were maintained at 23°C, 70% relative humidity, in a 12 hour light:12 hour dark cycle. Experiments were carried out on F1 third instar larvae.
Total RNA was extracted from 12 brains (four for each line) of third instar larvae with TRIzol (Invitrogen, USA) following the manufacturer's instructions. RNA was treated with RQ1 RNAse-free DNAse (Promega, Italy) to avoid DNA contamination, and then resuspended in RNase-free water.
For each sample, 1 mg of RNA was used for the first-strand cDNA synthesis, employing oligonucleotides dT 20 and SuperScript II (Invitrogen, USA) according to the manufacturer's instructions.
The size of the PCR products were 210 bp in the case of the SNAP-25 isoforms and 164 bp in the case of rp49. The samples were analysed by electrophoresis (1% agarose gel) and the resultant digital image of the gel was captured using a UVtransilluminator. The intensities of the bands were quantified using Image Pro Plus 6.0 (Media Cybernetics, USA). WT and R206A SNAP-25 isoform expression was calculated relative to the median values of the rp49 housekeeping mRNA.

Site-directed mutagenesis
pGEX2T-SNAP-25 WT was used as template in a PCR amplification performed using the QuickChange site-directed mutagenesis kit (Stratagene). For the R198A mutation the forward primer was 5Ј-ATTGATGAAGCCA ACCAAGCCGCA ACAAA -GATGCTGGGAAGTGGT-3Ј. The mutated codon is underlined. PCR was performed under the following conditions: initial denaturation for 60 seconds at 95°C, amplification for 18 cycles of 50 seconds at 95°C, 50 seconds at 60°C and 11 minutes 30 seconds at 68°C, final extension for 7 minutes at 68°C. After digestion of the parental DNA for 1 hour at 37°C with DpnI, the amplified plasmids were transformed into E. coli XL-1 Blue competent cells. The presence of the mutation was confirmed by DNA sequencing. The mutated construct was transformed into E. coli BL21 DE3 and the recombinant protein expressed.

Expression and purification of GST-SNAP-25 and GST-SNAP-25 mutants
SNAP-25 WT and its mutant R198A were expressed in E. coli BL21 DE3 as GST fusion proteins and isolated as previously described (Tonello et al., 1998). Briefly, expression of GST-SNAP-25 WT and GST-SNAP-25 R198A was induced for 3 hours at 30°C with 1 mM IPTG (isopropyl--D-thiogalactoside) and proteins were isolated on a glutathione-Sepharose 4B affinity column (Amersham Biosciences) according to the manufacturer's instructions. Resin-bound GST-SNAP-25 WT and GST-SNAP-25 R198A were eluted in a buffer containing 20 mM reduced glutathione and dialysed overnight in 10 mM Hepes buffer pH 7.4 with 150 mM NaCl. Protein purity was checked by SDS-PAGE on a 4-12% gel (Invitrogen).

In vitro assembly of the SNARE complex
The reaction for assembly of the SNARE complex was performed as described previously Fasshauer et al., 2002;Hu et al., 2002). Recombinant VAMP2 and syntaxin 1A were incubated with either GST-SNAP-25 WT or the R198A mutant (30 pmol each) for different time periods at room temperature in 30 l of 20 mM Hepes buffer pH 7.2, 150 mM NaCl, 1 mM Na 2 EDTA, 1 mM dithiothreitol, 0.8% b-octyl glucoside. The reactions were stopped by the addition of sample loading buffer. The solutions were then incubated at room temperature or boiled at 100°C for 3 minutes (boiling disrupts the SNARE complex). Samples were then subjected to SDS-PAGE using NuPAGE 4-12% Precast gels (Invitrogen), followed by Coomassie Blue staining; images were collected with ChemiDoc TM XRS (BioRad) and densitometric measurements were performed with WCIF ImageJ v1.35.

Electrophysiology
Experiments were performed at 20-22°C on third instar larval body wall preparations dissected in Ca 2+ -free HL3 saline (Stewart et al., 1994) and pinned on the siliconecoated surface (Sylgard 184; Dow Corning, USA) of a 35 mm Petri dish (Bolatto et al., 2003). After dissection, Ca 2+ -free HL3 saline was replaced with 1 mM Ca 2+ HL3. Before starting electrophysiological recordings, third instar larval body walls were left to incubate in fresh 1 mM Ca 2+ HL3 solution for at least 15 minutes.
Electrophysiological recordings were done on fibres 6 or 7 of abdominal segment 3 or 4 using intracellular glass microelectrodes (1.2 mm o.d.; 0.69 mm i.d.; 10-12 MW resistance; Science Products, Germany) filled with a 1:2 solution of 3 M KCl and 3 M CH 3 CO 2 K. Fibres with a membrane resting potential lower than -60 mV were discarded. In each fibre both spontaneous and evoked neurotransmitter release were recorded. No more than one fibre for each larval body wall was utilized for electrophysiological recording.
For quantifying Ca 2+ dependence of both spontaneous and evoked neurotransmitter release, after dissection, body walls were first incubated in 1 mM Ca 2+ for at least 15 minutes. After the spontaneous or evoked release was recorded, the bathing solution was replaced with one containing 3 mM Ca 2+ , taking particular care to maintain in place the intracellular microelectrode and that resting membrane potential remained substantially unchanged (apart from the physiological shift due to the different extracellular Ca 2+ concentrations). After an additional 15-minute incubation, spontaneous or evoked release was recorded again, and the bathing solution was changed to one containing 5 mM Ca 2+ ; following a further 15-minute incubation, a new recording of spontaneous or evoked release from the same fibre was performed. In all experiments, particular care was taken to avoid the microelectrode exiting from the fibre or fibre membrane damage; data from experiments in which this happened were discarded. Because evoked release can influence spontaneous neurotransmitter release, each activity was recorded in a separate set of experiments.
We did not observe any clear shift of resting membrane potential nor visible fibre alterations up to the end of the experiment (usually 60 minutes).

Intracellular recording of spontaneous neurotransmitter release
Spontaneous neurotransmitter release was analysed by intracellularly recording miniature end-plate potentials (MEPPs) under current clamp conditions. In a separate set of experiments miniature end plate currents (MEPCs) under voltage-clamp conditions were also recorded. This technique was used as it provides higher accuracy in the determination of spontaneous neurotransmitter release, and the possibility of using only one microelectrode for voltage-clamping because of the biophysical properties of the larval muscle fibre (Wu and Haugland, 1985). This particular advantage is lost when larger currents (for instance evoked excitatory junctional currents) must be clamped, making it necessary to use the more complex twoelectrode voltage clamp (Wu and Haugland, 1985).
Signals were amplified by a patch-clamp amplifier (Axon 200A; Molecular Probe, USA) in current or voltage-clamp mode. When in voltage clamp mode, microelectrode series resistances were corrected to 95%. Resting membrane potential was clamped at -70 mV. Spontaneous release events were recorded for 240 seconds.

Intracellular recording of evoked neurotransmitter release
Evoked junction potentials (EJPs) were intracellularly recorded under current-clamp conditions. Resting membrane potential was clamped at -70 mV. The single segmental nerve innervating abdominal segment A3 or A4 was stimulated at 0.5 Hz (stimulus duration 0.1 milliseconds, 1.5 threshold voltage) using a suction microelectrode, filled with extracellular bathing solution and connected to a stimulator (S88, Grass, USA) via a stimulus isolation unit (SIU5, Grass, USA) in a capacitative coupling mode. Signals were amplified in current-clamp mode by a patch-clamp amplifier (Axopatch 200, Axon, USA).

Data analysis
Amplified signals were digitized using a digital A/C interface (National Instruments, USA) and then fed to a PC for both on-line visualisation and off-line analysis using an appropriate software (WinEDR, Starthclyde University; pClamp, Axon, USA). Stored data, were analysed off-line using appropriate software (pClamp, Axon, USA). Statistical analysis and graph construction were carried out using Prism software (GraphPad, USA).

Immunohistochemistry
NMJ synaptic boutons of A3-A4 fibres 6 and 7 were stained in body-wall preparations as previously described (Bolatto et al., 2003) using the presynaptic membrane marker anti-HRP antibody (1:300, Cappel MP Biomedicals, USA). Primary antibodies were detected with FITC-conjugated rabbit anti-goat IgG (1:300; Sigma-Aldrich, USA). Images were acquired with a video-confocal microscope (VCM; Mangoni Biomedica, Pisa, Italy) (Benedetti et al., 1995). Briefly, in VC microscopy an arc lamp is used as a multi-point excitation source and a CCD camera as an image detector. Confocal performance is achieved using an original processing method (Benedetti et al., 1995) (US patent 6.016.367), which allows high spatial resolution and spectral flexibility.
Optical sections of 0.5 m were acquired using a ϫ20 0.75 NA dry objective. Acquired images were then collapsed into a single image with a maximum intensity projection algorithm. Bouton diameter was measured with NIS AR software (Nikon Instruments, Japan). Bouton area was measured on digital images, using ImageJ software. Statistical analysis of bouton areas and diameters was performed with Prism software.

Computational methods
Sequence conservation analysis was performed using the Bayesian method as implemented in the Consurf server (Landau et al., 2005). Molecular dynamic simulations were performed using, as a starting model, the SNARE structure (Chen et al., 2002), from which chain E, corresponding to the protein Complexin, was removed. Simulations were performed for the wild-type and Arg198Ala mutant using the generalized Born model approach for implicit solvation as implemented in AMBER software (Hawkins et al., 1996;Tsui and Case, 2000). A time step of 2 femtoseconds was adopted and the SHAKE algorithm (Ryckaert et al., 1977) was used to constrain chemical bonds. A cut off of 1.6 nm was adopted. Both wild-type and mutated systems were stabilized by 0.2 nanoseconds, during which the temperature was raised from 0 to 300 K coupling the systems to a Langevin thermostat with a collision frequency of 2/picosecond (Loncharich et al., 1992). Subsequently, 3 nanoseconds of production runs were performed for both systems.
We are grateful to B. Davletov for the kind gift of the sample of the recombinant cytoplasmic domain of VAMP/synaptobrevin and syntaxin. The present work was supported by Fondazione Cariparo Progetto "Physiopathology of the Synapse: Neurotransmitters, Neurotoxins and Novel Therapies" and Progetto Strategico of University of Padova: "An in vivo Approach to the Physiopathology of Signal Transduction". M.S. was supported by a University of Padova grant (Assegno di Ricerca grant no. CPDR078721/07 and no. CPDR095880/09). D.Z. was supported by a Fondazione Cariparo grant. S.P. was supported by ANII (Agencia Nacional de Investigación e Innovación, Programa de Apoyo Sectorial a la Estrategia Nacional de Innovación) INNOVA URUGUAY (Agreement no. 8 DCI ALA/2007/19.040 between Uruguay and the European Commission).