Small peptides patterned after the N-terminus domain of SNAP25 inhibit SNARE complex assembly and regulated exocytosis


Address correspondence and reprint requests to A. Ferrer-Montiel, Instituto de Biología Molecular y Celular, Universidad Miguel Hernández, Avenue Del Ferrocarril s/n, 03202 Elche, Alicante, Spain. E-mail:


Synthetic peptides patterned after the C-terminus of synaptosomal associated protein of 25 kDa (SNAP25) efficiently abrogate regulated exocytosis. In contrast, the use of SNAP25 N-terminal-derived peptides to modulate SNAP receptors (SNARE) complex assembly and neurosecretion has not been explored. Here, we show that the N-terminus of SNAP25, specially the segment that encompasses 22Ala-44Ile, is essential for the formation of the SNARE complex. Peptides patterned after this protein domain are potent inhibitors of SNARE complex formation. The inhibitory activity correlated with their propensity to adopt an α-helical secondary structure. These peptides abrogated SNARE complex formation only when added previous to the onset of aggregate assembly. Analysis of the mechanism of action revealed that these peptides disrupted the binary complex formed by SNAP25 and syntaxin. The identified peptides inhibited Ca2+-dependent exocytosis from detergent-permeabilized excitable cells. Noteworthy, these amino acid sequences markedly protected intact hippocampal neurones against hypoglycaemia-induced, glutamate-mediated excitotoxicity with a potency that rivaled that displayed by botulinum neurotoxins. Our findings indicate that peptides patterned after the N-terminus of SNAP25 are potent inhibitors of SNARE complex formation and neuronal exocytosis. Because of their activity in intact neurones, these cell permeable peptides may be hits for antispasmodic and analgesic drug development.

Abbreviations used

basic saline solution


circular dichroism






high-performance liquid chromatography




sodium dodecyl sulphate polyacrylamide gel electrophoresis


synaptosomal associated protein of 25 kDa


SNAP receptors


vesicle-associated membrane protein (synaptobrevin)

Neurotransmitter release is a highly regulated cascade that proceeds through an orchestrated sequence of protein–protein interactions that culminate in the fusion of neurotransmitter-loaded vesicles in response to Ca2+ influx in synaptic junctions (DeBello et al. 1993, 1995; Jahn et al. 2003). This is a complex process mediated by the so-called SNAP receptors (SNARE) proteins, which include the membrane-associated proteins syntaxin and synaptosomal associated protein of 25 kDa (SNAP25), and the vesicle-associated membrane protein (VAMP; Metha et al. 1996; Brunger 2001; Bruns and Jahn 2002). These proteins directly govern vesicle docking and fusion through the formation of an sodium dodecyl sulphate (SDS)-resistant, ternary complex referred to as the SNARE complex (Chen and Scheller 2001). The physiological relevance of SNARE proteins in regulated exocytosis is underscored by the discovery that they are the targets of botulinum and tetanus neurotoxins, a family of naturally occurring neurotoxins that potently block neurosecretion (Schiavo et al. 2000). Furthermore, it has been shown that SNARE proteins represent the minimal fusion machinery, as evidenced by the catalysis of vesicle fusion by SNAREs reconstituted in synthetic liposomes (Weber et al. 1998).

The SNARE complex is formed by a coiled coil structure. High-resolution structural analysis of the SNARE complex core shows a parallel four helix bundle formed by the interaction of two helices from SNAP25, one from syntaxin and one from VAMP (Fasshauer et al. 1997; Sutton et al. 1998; Xiao et al. 2001). The core of the four helix bundle is composed of well-defined layers formed by interacting side-chains from each of the four α-helices (Sutton et al. 1998). This compact structural arrangement is consistent with the tremendous resistance of the SNARE complex to denaturation by SDS (Hayashi et al. 1994). Mutation of the inner interacting layers results in the destabilization of the protein complex (Chen and Scheller 2001; Jahn et al. 2003). Similarly, botulinum neurotoxins that cleave the SNARE proteins and peptides that mimic their amino acid sequence notably interfere with the stability and formation of the SDS-resistant protein complex (DeBello et al. 1993; Gutierrez et al. 1995a, 1995b; Martin et al. 1996; Metha et al. 1996; Ferrer-Montiel et al. 1998a; Apland et al. 1999; Schiavo et al. 2000; Melia et al. 2002).

Several studies have shown that small ≤20 mer peptides patterned after the C-terminus of SNAP25 inhibit complex assembly and neurotransmitter release (Cornille et al. 1995; Gutierrez et al. 1995a, 1997; Apland et al. 1999). The biological activity of these peptides was specific as random sequences of the same amino acid composition were inert modulating exocytosis (Gutierrez et al. 1995a, 1997). The inhibitory activity of these amino acids sequences further corroborated an essential role of the C-terminus of SNAP25 in complex formation. These observations, along with the finding that clostridial neurotoxins BoNTs A and E cleave a portion of the SNAP25 carboxy region, highlighted the need of an unaltered membrane-proximal C-terminus domain of the SNARE complex (Blasi et al. 1993; Lawrence et al. 1997; Schiavo et al. 2000). In marked contrast, the role of the SNAP25 N-terminus remains more elusive, although it has been reported that antibodies that target this domain affect complex assembly and neuronal exocytosis (Xu et al. 1999).

Here, we have further addressed this issue by using a step-wise deletion strategy in conjunction with the synthesis and assay of small peptides modelled after N-terminus segment encompassing 1Met-44Ile. We chose this protein domain because it lies N-terminal of the conserved ionic layer in the centre of the SNARE complex (Sutton et al. 1998). Deletion of this SNAP25 domain prevented the assembly of the core complex. Interestingly, removal of the segment 1Met-21Leu did not impede complex formation, whereas deletion of the 22Ala-44Ile region fully deterred the interaction of the SNARE proteins. Similarly, a synthetic peptide mimicking the amino acid sequence of the 22Ala-44Ile domain blocked complex formation and Ca2+-dependent exocytosis more potently than that mirroring the 1Met-21Leu segment. The inhibitory activity of the most potent peptide was only detected if they were added prior to the onset of complex formation, consistent with the proposed zippering model from N- to C-terminus for assembly (Lin and Scheller 1997). Collectively, our observations highlight a critical role for the N-terminus domain of SNAP25, especially of the segment comprising 22Ala-44Ile, in the stability of the SNARE complex. Noteworthy, peptides patterned after this protein domain protected neurones against excitotoxicity, presumably by inhibiting the release of L-glutamate in intact neurones. In support of this tenet, these peptides abrogated catecholamine release from detergent-permeabilized chromaffin cells. Thus, sequences mimicking the 22Ala-44Ile region provide new pharmacological tools that may help to further our understanding of the mechanism of neuronal exocytosis.

Experimental procedures


Recombinant proteins and peptides cDNA plasmids encoding the VAMP cytosolic domain (kindly provided by Dr R. Jahn), syntaxin cytosolic portion (kindly provided by Dr R. Scheller) and full-length SNAP-25 (kindly provided by M.C. Wilson) were cloned into the pGEX-KG vector to obtain a glutathione-S-transferase (GST) fusion constructs containing a thrombin cleavage site after the GST.

Peptides SNAP25_N1 (Ac-MAEDADMRNELEEMQRRADQL-NH2), SNAP25_N2 (Ac-ADESLESTRRMLQLVEESKDAGI-NH2), SNAP25_N3 (Ac-ELEEMQRRADQLA-NH2), SNAP25_N4 (Ac-LESTRRMLQLVEE-NH2), and fluorescein-SNAP25_N4 were purchased from DiverDrugs SL (Barcelona, Spain). Peptide purity was ≥ 95%. Peptide identity was confirmed by matrix-assisted laser desorption ionization time of flight (MALDI–TOF) spectrometry.

Deletion of SNAP25

Deletions were obtained by one-step inverse PCR with the proofreading Pfu turbo DNA polymerase and two restriction digestions (Cabedo et al. 2002). Externally oriented primers with SacII unique restriction sites were used to amplify the whole vector except the region to be deleted. The following primers were used: (a) Δ[1–21] sense 5′-GATTCCGCGGGCTGATGAGTCCCTGGG-3′, antisense 5′-CTAA-CCGCGGGGAGTCTAGAATTCCACCACC-3′; (b) Δ[22–44] sense 5′-GATTCCGCGGAGGACTTTGGTTATGTTGG-ATG-3′, and antisense 5′-CATACCGCGGCAGCTGGTCAGCCCTCCTC-3′; (c) Δ[1–44], sense 5′GATTCCGCGGAGGACTTTGG-TTATGTTGGATG-3′ and antisense 5′-CTAACCGCGGGGAGTC-TAGAATTCCACCACC-3′. PCR products were digested (in the amplification buffer) with 20 U of DpnI overnight at 37°C to remove the methylated plasmid templates, subsequently purified, digested with SacII and ligated. Designed deletions were verified by restriction digestion with SacII and by automatic sequencing.

Expression and purification of recombinant SNARE proteins

Recombinant bacterially expressed SNARE proteins were obtained as described (Blanes-Mira et al. 2001, 2002). Briefly, GST-fusion proteins were expressed in the BL21DE3 strain. Protein expression was induced with 1 mm isopropyl-β-d-thiogalactopyranoside (Sigma, St Louis, MO, USA) for 5 h at 30°C. Bacterial cultures were pelleted, washed with lysis buffer (10 mm phosphate pH 7.4, 136 mm NaCl, 2.7 mm KCl), digested with 0.1 mg/mL lysozime (Sigma) for 10 min at 22°C in lysis buffer, supplemented with 2 mm phenymethylsulphonyl fluoride (Sigma), 5 mm iodoacetamide (Sigma), 5 mm EDTA and sonicated (3 × 45 s) in a Branson 250 sonifier at 4°C. Lysates were solubilized with 1% Triton X-100 for 20 min at 4°C and cleared by centrifugation at 20 000 g for 30 min at 4°C. SNAP25 and VAMP were purified from the sonicated bacteria supernatant by affinity chromatography on glutathione agarose (Pharmacia, Uppsala, Sweden) following manufacturer instructions. Purification of recombinant proteins was carried out in 20 mm HEPES pH 7.4, 100 mm NaCl, 0.05% n-octyl-β-d-glucopyranoside (OG), 5 mm dithiothreitol (DTT) and cleaved with thrombin for 3 h at 23°C and dialysed against 20 mm HEPES pH 7.0, 80 mm KCl, 20 mm NaCl, 0.1% OG. Syntaxin was obtained from the bacterial pellet by washing the precipitated with 50 mm Tris–HCl pH 8.0, 10 mm EDTA, 100 mm NaCl, 1.0% Triton X-100 in a polytron. Protein was recovered from inclusion bodies by incubating the solubilized pellet with 50 mm Tris–HCl pH 8.0, 10 mm EDTA, 100 mm NaCl, 1.0%N-lauroyl-sarcosine, at 4°C overnight. Extracted protein was diluted 1 : 10 in washing buffer (10 mm HEPES pH 7.4, 0.1% OG) and loaded on to glutathione-agarose resin. Bound material was purified in washing buffer and cleaved with thrombin for 3 h at room temperature. Purified proteins were stored at −80°C. Concentration was assayed with the BCA kit (Pierce, Rockford, IL, USA), and purity verified by sodium dodecyl sulphate polyacrylamide gel electrophoresis (SDS–PAGE) analysis.

In vitro reconstitution of the SDS-resistant SNARE complex

Syntaxin and VAMP were incubated at final concentrations of 3.0 µm in the presence/absence of the peptides at the indicated concentrations for 2 h at 4°C. Thereafter, SNAP25 (3.0 or 0.3 µm, as indicated) was added and the reactions (V = 15 µL) proceed in 20 mm HEPES pH 7.4, 100 mm NaCl, 1.0% OG, 2.0 mm DTT at 4°C for the indicated times, and were stopped by the addition of SDS–PAGE sample buffer. Peptide blockade activity was evaluated on SDS–PAGE (12%) by the disappearance of the 75 kDa band corresponding to SDS-resistant ternary core complex. Gels were digitized and quantified as described (Blanes-Mira et al. 2001).

Circular dichroism spectroscopy

Circular dichroism (CD) was carried out in a JASCO J-810 Spectropolarimeter equipped with a computer-controlled temperature cuvette holder. CD data for the Far-UV CD spectra (region 195–250 nm) were recorded with a 1-mm path length cell containing 100 µm peptides in 20 mm Tris–HCl pH 8.0. All spectra were recorded at 25°C and at 50 nm/min (response time of 1 s), averaged (five scans), and corrected for the buffer contribution. CD signals (in mdegrees) were converted to mean ellipticity (θ, mdegrees cm2/dmol) using the relationship θν = 100 × CD signal/(C × N × l), were C denotes the peptide concentration, N the number of residues and l the path length. Secondary structure elements were inferred by fitting the CD spectra as described (Blanes-Mira et al. 2001).

Chromaffin cell cultures

Chromaffin cell cultures were prepared from bovine adrenal glands by collagenase digestion and further separated from debris and erythrocytes by centrifugation on Percoll gradients as described (Gomis et al. 1994). Cells were maintained in monolayer cultures at a density of 625 000 cells/cm2 and were used between the third and sixth day after plating. All the experiments were performed at room temperature.

Determination of catecholamine release from detergent-permeabilized chromaffin cells

Secreted noradrenaline and adrenaline was determined in digitonin-permeabilized cells as described (Gutierrez et al. 1995b). Briefly, monolayers were washed 4 times with a Krebs/HEPES basal solution (containing in mM: 15 HEPES pH 7.4 with 134 NaCl, 4.7 KCl, 1.2 KH2PO4, 1.2 MgCl2, 2.5 CaCl2, 0.56 ascorbic acid and 11 glucose). Cell permeabilization was accomplished with 20 µm digitonin in 20 mm Pipes, pH 6.8 with 140 mm monosodium glutamate, 2 mm MgCl2, 2 mm Mg-ATP and 5 mm EGTA. This incubation was carried out in the absence or presence of peptides. Following permeabilization, media were discarded and cells were incubated for 10 min in digitonin-free medium in presence or absence of peptides. Basal secretion was measured in 5 mm EGTA, whereas stimulated secretion was measured in a medium containing 10 µm buffered Ca2+ solution. Media were collected and released catecholamines as well as the total cell content were determined by high-performance liquid chromatography (HPLC) using an electrochemical detector. Statistical significance was calculated using Student's t-test with data from four or more independent experiments.

Hippocampal cultures

Mixed hippocampal neuronal/glial cultures were prepared as described (Ferrer-Montiel et al. 1998b; Valera et al. 2002). Briefly, hippocampi were dissected from E17–E19 rat embryos, and incubated at 37°C for 15 min in basic saline solution (BSS, containing: 137 mm NaCl, 3.5 mm KCl, 0.4 mm KH2PO4, 0.33 mm Na2HPO4·7 H2O, 5 mm TES pH 7.4 and 10 mm glucose) with 0.025% trypsin (Invitrogen, Carlsbad, CA, USA). Trypsin was diluted by rinsing the tissue three times for 5 min each with BSS. Tissue was then dissociated by several passes through a siliconized Pasteur pipette, first unpolished and then with a fire-polished pipette. Cells were centrifuged for 5 min at 200 g and pellets resuspended in BSS. Cells were plated at 2 × 105 viable cells/cm2 in poly-l-lysine-coated (0.2 mg/mL, Sigma) dishes. The culture medium had the following composition: minimum essential medium (Earle's salts, Invitrogen) supplemented with 10% (v/v) heat-inactivated horse serum (Sigma), 10% (v/v) heat-inactivated fetal bovine serum (Sigma), 1 mm glutamine and 22 mm glucose. Cultures were maintained at 37°C in a 5% CO2 atmosphere and half of the medium was renewed every 2–3 days. At day 5 after plating, glial proliferation was inhibited by the addition of 80 µm 5-fluoro-2′-desoxiuridine.

Glucose deprivation induced cell death

Neurones cultured at 14 days in vitro were used. Culture medium was removed and neurones were rinsed with fresh medium without glucose and added sucrose to maintain osmolarity (Ferrer-Montiel et al. 1998b; Valera et al. 2002). Plates were then returned to the incubator for 3 h. Botulinum neurotoxins E and B (Sigma) were added 24 h before the excitotoxic insult. Peptides (100 µm) were added 2 h before the glucose deprivation and were present during the insult. Glucose deprivation was terminated by changing the cultures to a glucose-containing medium (22 mm glucose) plus 10 µm MK-801 to prevent additional NMDA receptor activation. Cell death was blindly assessed 18–24 h post-insult using the trypan blue (0.4%) exclusion assay.


The N-terminus of SNAP25 modulates SNARE complex assembly

To identify peptides in the N-terminal half of SNAP25 that modulate the stability of the SNARE complex, we first investigated the role of the N-terminus of SNAP25 in the formation of the SNARE aggregate using a step-wise deletion approach of segment 1Met-44Ile. The rationale was based in the differential propensity of specific regions within this segment to form coiled-coil structures and to adopt an α-helical structure. As shown in Fig. 1(a), the segment encompassing residues 1Met-21Leu displays a 100% prediction for coiled-coil formation and 70% probability to fold into an α-helix, while the region 22Ala-44Ile does not show significant probability to adopt a coiled-coil arrangement nor α-helical secondary structure. Hence, we deleted these two regions, and evaluated the ability of the deleted SNAP25 species in the assembly of the SNARE complex in vitro (Fig. 1b). For this task, the SNARE complex was reconstituted in vitro by incubating equimolar amounts of the three recombinant SNARE proteins, and its formation was analysed by SDS–PAGE. Figure 1(b) shows that incubation of SNAP25, VAMP and syntaxin at 4°C overnight leads to the formation of an SDS-resistant complex of 75 kDa (control lane), that is sensitive to temperatures ≥ 90°C (90°C lane). Replacement of SNAP25 by Δ[1–21] did not affect the formation of the SDS-resistant complex (Figs 1b and c), although it significantly reduced the thermal stability of the SNARE complex (Fig. 1d). In contrast, removal of amino acids 1Met-44Ile or 22Ala-44Ile from SNAP25 fully prevented the assembly of the SDS-resistant complex (Figs 1c and d, Δ[1–44] and Δ[22–44] lanes). Collectively, these results demonstrate that the region 22Ala-44Ile of SNAP25 is important for the assembly of the SNARE complex, while segment comprising 1Met-21Leu affects the stability of the ternary complex, consistent with observations reported by others (Chapman et al. 1994; Yang et al. 2000).

Figure 1.

Deletion of domain 1Met-44Ile in the N-terminus half of SNAP25 notably affects the stability of the SNARE complex. (a) Coiled-coil and α-helix secondary predictions of the N-terminal half of SNAP25. Analysis was performed with the programs Agadir ( and Coils ( (b) Schematic representation of the deleted regions and effect of SNAP25 deletion species Δ[1–21], Δ[1–44] and Δ[22–44] on the in vitro assembly of the SNARE complex. Complex formation was accomplished by incubating equimolar amounts of the SNARE proteins at 4°C overnight, and analyzed by SDS–PAGE. (c) Protein gels were digitized and quantified as described (Blanes-Mira et al. 2001). Complex formation was normalized with respect to that obtained for SNAP25 wild type (control). Data are mean ± SEM with n = 2. (d) SNARE complex formation as a function of the temperature. Ternary complexes formed by SNAP25 or Δ[1–21] were heated to the specified temperatures for 5 min before SDS–PAGE analysis. The extent of complex assembly was normalized with respect to that formed at 25°C. Experimental points were fitted to a logistic equation inline imagewhere CF is the fraction (%) of SNARE complex formation, T0.5 is the denaturing temperature and p the slope of the curve. Denaturating temperatures were 76 ± 0.2°C for SNAP25 wild type, and 72 ± 0.4°C for deletion Δ[1–21] (mean ± SEM, n = 3).

Peptides patterned after the N-terminus domain of SNAP25 inhibit SNARE complex formation

Because truncation of internal protein segments may significantly alter the protein structure, we questioned whether the dramatic effect of the 22Ala-44Ile region on SNARE complex assembly was specific or it was the consequence of the resulting protein structure. To address this issue, we proposed the use of peptides patterned after these SNAP25 regions as specific pharmacological tools (Fig. 2a). Peptides SNAP25_N1 (Ac-MAEDADMRNELEEMQRRADQL-NH2) and SNAP25_N2 (Ac-ADESLESTRRMLQLVEESKDAGI-NH2) were synthesized by solid phase and their effect on the stability of the reconstituted SNARE complex was evaluated. To favour the inhibitory activity of the peptides, the SNAP25 concentration was reduced 10-fold (to 0.3 µm) with respect to that of VAMP and syntaxin. In these experiments, peptides were incubated with syntaxin and VAMP at 4°C for 2 h before the assembly of the SNARE complex was initiated by addition of SNAP25. As depicted in Fig. 2(b), the three SNARE proteins assembled into a ternary complex in both the absence and presence of either peptide. Notice, however, that the amount of SNARE complex formed was lower in the presence of 1 mm of either peptide, as evidenced by the lighter intensity of the ∼75 kDa protein band and the appearance of a ∼27 kDa band corresponding to SNAP25. Quantification of these observations indicate that SNAP25_N1 inhibited a ∼25% of complex formation, while the extent of complex inhibition for SNAP25_N2 increased to ∼50% (Fig. 2c). These results support the finding that the segment 22Ala-44Ile, in the middle of the N-terminal half of the SNAP25, plays a critical role in the assembly of the SNARE complex.

Figure 2.

Peptides SNAP25_N1 and SNAP25_N2 encompassing the domain 1Met-44Ile of SNAP25 N-end inhibit the in vitro assembly of the SNARE complex. (a) Amino acid sequence of peptides SNAP25_N1 and SNAP25_N2 encompassing domains 1Met-21Leu and 22Ala-44Ile, respectively. (b) Effect of SNAP25_N1 and SNAP25_N2 peptides on SNARE complex assembly. Peptides (1 mm) were incubated with 3 µm VAMP and 3 µm syntaxin at 4°C for 2 h before SNARE complex formation by addition of 0.3 µm SNAP25. The reaction proceed at 4°C overnight. Complex assembly was analyzed by SDS–PAGE. Control denotes the formation of the protein aggregate in the absence of peptides. (c) Quantification of inhibitory effect of peptides on the formation of SNARE complex. Gels were digitized and quantified as described (Blanes-Mira et al. 2001). Data are mean ± SEM, n = 2.

Analysis of the content of secondary structure elements by circular dichroism, showed that both sequences exhibited a 25% content in α-helix in aqueous buffer that was significantly increased by addition of TFE, as evidenced by the appearance of CD minima at 207 and 220 nm (Figs 3a and b). As illustrated in Fig. 3(c), the secondary structure of SNAP25_N1 increased to 60% in the presence of ∼50% TFE, while that of SNAP25_N2 was augmented to ∼77%. Thus, the inhibitory activity of both peptides appears to correlate with their propensity to adopt an α-helix secondary structure.

Figure 3.

Peptides SNAP25_N1 (▴) and SNAP25_N2 (•) exhibit significant propensity to adopt an α-helical conformation. (a and b) Far-UV CD spectra of peptides SNAP25_N1 and SNAP25_N2 at increasing percentages of TFE (0, 5, 10, 20, 30 and 50%). Peptide concentration was 100 µm in 20 mm Tris pH 8.0. CD spectra represent the average of five scans, and were corrected for the buffer contribution. (c) Quantification of the α-helical content as a function of the percentage of TFE. α-Helical values were inferred as described (Blanes-Mira et al. 2001).

Identification of core sequences that inhibit SNARE complex assembly

We next investigated whether shorter sequences that modulate the stability of the SNARE complex could be identified. For this purpose, peptides SNAP25_N3 (Ac-ELEEMQRRADQLA-NH2) corresponding to the α-helical region detected in the N-terminus (Fig. 1), and SNAP25_N4 (Ac-LESTRRMLQLVEE-NH2) that encompasses the core of the 22Ala-44Ile region were synthesized and assayed (Fig. 4a). Because short peptides exhibit lower potency destabilizing coiled coil complexes, the in vitro reconstitution of SNARE complexes was modified to accomplish peptide : SNAP25 ratios greater than 5000 : 1. Specifically, we used in vitro translated [35S]SNAP25 instead of the bacterially produced protein. Incubation of [35S]SNAP25 with recombinant VAMP and syntaxin gave rise to the assembly of the SNARE complex, which was readily detected as a radioactive band of 75 kDa that disappeared upon heating the samples to 90°C (Fig. 4b, control and 90°C lanes). Pre-incubation of syntaxin and VAMP with 3 mm SNAP25_N3 or SNAP25_N4 significantly inhibited the formation of the SNARE complex (Fig. 4b). Consistent with the stronger potency of peptide SNAP25_N2, peptide SNAP25_N4 exhibited higher inhibitory activity than peptide SNAP25_N3 (Fig. 4c). The inhibitory activity of both peptides was dose-dependent. As for larger peptides, CD analysis shows that both peptides exhibit a conspicuous propensity to adopt a α-helical secondary structure, as evidenced by the strong TFE-inducted α-helical content (Fig. 4d). Taken together, these results substantiate the tenet that peptides mimicking the 22Ala-44Ile segment (Fig. 1a) of the N-terminal half of SNAP25 are potent inhibitors of SNARE complex formation. The tendency of the peptides to acquire an α-helical structure seems important for their inhibitory activity, although the influence of additional properties that contribute to define their blockade efficacy cannot be ruled out. For instance, it did not escape to our attention that the first 20 amino acids of SNAP25 are poorly conserved between different species while segment 22Ala-44Ile is virtually invariant throughout evolution (Risinger et al. 1993).

Figure 4.

Peptides SNAP25_N3 (•) and SNAP25_N4 (▴) patterned after the core segment potently inhibit the formation of the SNARE complex. (a) Amino acid sequence of peptides SNAP25_N3 and SNAP25_N4. (b)  In vitro reconstituted SNARE complex using [35S]SNAP25 in the absence (control) and presence of 3 mm SNAP25_N3 or SNAP25_N4. Control samples heated to 90°C for 5 min are also displayed (90°C). VAMP and syntaxin were used at 3 µm. Complex formation was carried out at 4°C overnight and analysed by SDS–PAGE. (c) Extent of SNARE complex formation obtained from SDS–PAGE gels by fluorography. Values denote mean ± SEM, with n = 3. (d) Quantification of α-helical content as a function of the percentage of TFE. The secondary structure element was calculated from Far-UV CD spectra of peptides as described (Blanes-Mira et al. 2001). Peptide concentration was 100 µm in 20 mm Tris pH 8.0.

Peptides prevent the formation of the SNARE complex

To gain insights on the inhibitory activity of these peptides, we compared the kinetics and extent of complex formation in two conditions: (a) inhibitory peptide SNAP25_N2 was added prior complex assembly and (b) peptide SNAP25_N2 was supplied 15 s after the onset of complex formation. The SNAP25_N2 peptide was selected because of its stronger efficacy inhibiting complex formation, as evidenced by dose–response curves that reveal an IC50 of 0.62 ± 0.15 mm for SNAP25_N2 and 1.5 ± 0.3 mm for SNAP25_N4 (Fig. 5a).

Figure 5.

Peptide SNAP25_N2 inhibits the assembly of the SNARE aggregate if added previous to the onset of complex formation. (a) Dose–response relationship for the inhibitory activity of the SNAP25_N2 (▪) and SNAP25_N4 (•) peptides. Experimental data were fitted to a Hill relation inline image where IC50 is the concentration of peptide that prevents the formation of complex to half maximum and nH is an index of the steepness of the slope. The values obtained were IC50 0.62 ± 0.15 mm and nH = 1.15 ± 0.20 for SNAP25_N2, and IC50 1.5 ± 0.3 and nH = 1.30 ± 0.10 for SNAP25_N4 (mean ± SEM, n = 2). (b) Kinetics of SNARE complex formation at 4°C in the absence of SNAP25_N2 peptide. (c) SNAP25_N2 (2 mm) was incubated with VAMP and syntaxin for 10 min at 4°C. Complex formation was initiated by addition of SNAP25. (d) SNAP25_N2 peptide was added 15 s after initiation of complex assembly. The extent of aggregate formation as a function of reaction time was monitored by SDS–PAGE.

The kinetics of the complex formation was monitored at 4°C (Fig. 5b). As shown, the 75 kDa band was detectable 15 s after incubation of the three SNARE proteins, and complex formation reached steady-state in 30–60 s. Incubation of syntaxin and VAMP with 2 mm SNAP25_N2 peptide at 4°C for 10 min before supplying SNAP25 resulted in full inhibition of SNARE complex formation, at least within the 60 min explored (Fig. 5c). In contrast, the SNAP25_N2 did not inhibit complex assembly when it was added 15 s after the mixing of the three SNARE proteins (Fig. 5d). Thus, peptide SNAP25_N2 prevents complex formation but does not disrupt pre-assembled complexes.

Mechanistically, peptides patterned after the N-terminus of SNAP25 may disrupt the binary complex assembled by SNAP25 and syntaxin, or may impede the interaction of VAMP with the pre-organized binary aggregate. In an attempt to address this question, we next studied the effect of peptide SNAP25_N2 on the arrangement of the binary complex formed by syntaxin and SNAP25. Assembly of SNAP25–syntaxin binary complexes was readily monitored by non-denaturating PAGE (Fig. 6a). Incubation of syntaxin with 2 mm SNAP25_N2 peptide fully inhibited the constitution of the binary complex with SNAP25 (Fig. 6a, lane 3). The peptide also disrupted pre-assembled SNAP25-syntaxin binary complexes suggesting the peptide competes with SNAP25 for interacting with syntaxin (Fig. 6a, lane 5). Abrogation of the binary complex by the SNAP25_N2 peptide notably reduced the extent of SNARE complex assembly, although it did not completely prevent its formation (Fig. 6a, lanes 4 and 6). In contrast, the interaction of the three SNARE proteins was not altered when the peptide was co-incubated with VAMP or added to the pre-assembled ternary aggregate, which is consistent with the high affinity and stability of the ternary core complex (Fig. 6a, lanes 7 and 8). Similar results were obtained when the SDS-resistant SNARE complex was arranged and analyzed under the same conditions (Fig. 6b). These results demonstrate that peptides patterned after the N-terminus of SNAP25 prevent the formation or disrupt the binary complex assembled by syntaxin and SNAP25, but are unable to affect the constituted SNARE complex. Thus, the inhibitory activity of peptides patterned after the N-terminus of SNAP25 is due primarily to interference with SNARE complex assembly. Furthermore, they appear to be competitive SNAP25 antagonists as a 10-fold increment of the SNAP25 concentration of SNAP25 virtually abolished the blockade activity of SNAP25_N2 (data not shown).

Figure 6.

Peptide SNAP25_N2 abrogates the assembly of the binary complex formed by SNAP25 and syntaxin. (a) Effect of SNAP25_N2 peptide on the formation of the binary and ternary protein complexes. Protein interactions were analysed by non-denaturating PAGE. Lane 1 displays SNAP25-syntaxin binary complex; lane 2 shows the ternary aggregate; lanes 3 and 4, SNAP25_N2 peptide was pre-incubated for 60 min with syntaxin; lanes 5 and 6, peptide SNAP25_N2 was incubated with the pre-formed binary complex; lane 7, peptide was co-administered with VAMP; lane 8, peptide was added to the assembled ternary aggregate. Peptide concentration was 2 mm. Binary complex formation proceed for 2 h at 23°C, while ternary complex assembly proceed at 4°C overnight. (b) Similar samples as in (a) analysed by SDS–PAGE. Lane 1, molecular weight standards; lane 2, SNARE complex; lane 3, SNARE complex heated to 90°C for 5 min; lanes 4, 5, 6 and 7 correspond to lanes 4, 6, 7 and 8 displayed in (a).

Identified peptides potently inhibit Ca2+-evoked catecholamine release

The inhibitory activity on SNARE complex arrangement implies that these peptides may modulate the release of neurotransmitter in excitable cells. To study this issue, we evaluated the effect of peptides SNAP25_N2 and SNAP25_N4 on Ca2+-evoked catecholamine release from detergent-permeabilized chromaffin cells. As illustrated in Fig. 7, digitonin-treated primary cultures of chromaffin cells readily release both adrenaline and noradrenaline in response to a 10-min pulse of 10 µm Ca2+. The extent of the catecholamine secretion was severely impaired by pre-incubation of permeabilized chromaffin cells with SNAP25_N2 and SNAP25_N4 peptides (Fig. 7). Both peptides block Ca2+-stimulated secretion in a concentration-dependent manner. The potency blocking catecholamine release paralleled their activity inhibiting SNARE complex formation. As depicted in Fig. 7, SNAP25_N2 peptide was more active than its shorter counterpart SNAP25_N4. Thus, these observations indicate that peptides patterned after the N-terminus of SNAP25 block-regulated exocytosis.

Figure 7.

Peptides SNAP25_N2 and SNAP25_N4 inhibit Ca2+-evoked catecholamine release from digitonin-permeabilized chromaffin cells. Concentration-dependent inhibition of Ca2+-stimulated release by peptides SNAP25_N2 and SNAP25_N4. Digitonin permeabilization lasted 10 min. Net adrenaline and noradrenaline secretion is the amount evoked by 10 µm Ca2+ minus that stimulated with 5 mm EGTA for 10 min Net release was normalized with respect to that obtained in the absence of peptides. Values are mean ± SEM with n = 4. Experiments were performed at room temperature. NA, nor adrenaline; A, adrenalin.

Identified peptides are cell-permeable and protect hippocampal neurones against hypoglycaemia-induced death

We further investigated the cellular activity of these peptides in primary neurones from the rat hippocampus. In these experiments we pursued the topic of whether the peptides may modulate exocytosis in intact neurones, thus concomitantly evaluating the membrane permeability of the peptides. We used a functional assay that measures the neuronal death in primary hippocampal cultures induced by glucose-deprivation. Hypoglycaemia-evoked neuronal death is due to excessive exocytosis of l-glutamate (Monyer et al. 1992). As illustrated in Fig. 8(a), exposure of hippocampal neurones to glucose-free medium for 3 h evoked neuronal death that was fully prevented by the presence of the uncompetitive NMDA receptor antagonist MK-801. This result substantiates the notion that hypoglycaemia-induced neuronal death is caused primarily by exposure of neurones to l-glutamate released into the extracellular medium and emphasizes that blockade of glutamate exocytosis should prevent neuronal death (Monyer et al. 1992). As anticipated, pre-treatment of hippocampal cell cultures with 100 pm BoNT E or BoNT B resulted in approximately 50% reduction of hypoglycaemia-induced neuronal death (Monyer et al. 1992). A dose–response curve for BoNT E neuroprotective activity showed an IC50 of 3.0 ± 1.0 pm, and a maximum neuroprotection of 46 ± 5% (Fig. 8b). These results imply that under hypoglycaemic conditions, neurones secrete a fraction of the glutamate by an exocytotic mechanism that involves the SNARE proteins, and indicate that inhibition of fusion of synaptic vesicles prevents subsequent neuronal damage.

Figure 8.

Peptides patterned after the N-terminus of SNAP25 protect hippocampal neurones from glucose deprivation-evoked death. (a) Neuroprotection elicited by peptides emulating the N-terminus of SNAP25, BoNTs E and B and MK801 against 3 h glucose deprivation. Neuronal viability was assessed blindly 18–24 h postinsult using the trypan blue exclusion assay. Neurones were incubated with neurotoxins 24 h before insult, and with peptides 3 h before glucose deprivation. MK-801 (10 µm) and peptides (100 µm) were present during the 3 h glucose deprivation. Values are given as mean ± SD with n≥ 3000 neurones, and n≥ 4. (b) Dose–response curves of the BoNT E and SNAP25_N2 peptide neuroprotective activity against glucose deprivation. Solid lines depict the best fit to the equationinline image where D denotes normalized cell death, IC50 the concentration of compound that reduces cell death to half-maximum and nH is the hill coefficient. The IC50 for BoNT E was 3.0 ± 1 pm, and for SNAP25_N2 was 1.8 ± 0.3 µm. The maximal neuroprotection was 46% ± 5 for BoNT E and 40% ± 4 for SNAP25_N2. Values are mean ± SEM with n = 4, and n = 2000 neurones. (c) Illustrative photograph (left) showing that fluorescently labelled SNAP25_N4 accumulates into the cytosol of PC12 cells. Fluorescein-conjugated peptide (1 µm) was incubated for 2 h, cells were extensively washed with phosphate-buffered saline, and cellular uptake analysed by confocal microscopy. (Right) Light-transmitted picture of cells shown in the left panel. Scale bar denotes 10 µm.

This finding provides a sensitive assay to investigate the in vivo inhibitory activity of molecules that affect the stability or formation of the SNARE complex. As displayed in Fig. 8(a), incubation of hippocamapal cultures with 100 µm of peptides SNAP25_N1, SNAP25_N2 and SNAP25_N4 attenuated the neural death triggered by the hypoglycaemic insult by 15 ± 2%, 40 ± 3.2% and 28 ± 2.8% (mean ± SEM), respectively. Peptide SNAP25_N3 was not neuroprotectant at this concentration. Note that the observed neuroprotective activity nicely paralleled the potency of these peptides inhibiting SNARE complex formation. A dose–response curve for SNAP25_N2 reveals an IC50 of 1.8 ± 0.3 µm, with a maximum neuroprotective of activity of 40 ± 4% (Fig. 8b). Because these sequences abrogate SNARE complex assembly and inhibit Ca2+-dependent exocytosis in permeabilized cells, their neuroprotective activity in intact neurones imply that they can translocate through the plasma membrane. To demonstrate this conclusion, we tagged the N-end of SNAP25_N4 peptide with fluorescein to monitor its cell permeability. PC12 cells were incubated with 1 µm fluorescein-SNAP25_N4 for 2 h, washed and analysed by confocal microscopy. As shown in Fig. 8(c), the fluorescently labelled SNAP25_N4 derivative readily accumulated into the cytosol of the cells. Taken together, these data demonstrate that peptides patterned after the N-terminus of SNAP25 are cell permeable peptides that inhibit neuronal exocytosis by interfering with vesicle fusion.


The central aim of this study was to investigate if peptides derived from the N-terminus of SNAP25 modulate the stability and function of the SNARE core complex. We addressed this question because, at variance with the intense research performed on the role of the SNAP25 C-end domain, the implication of the N-terminal domain has remained poorly investigated. As a result, BoNT A and E and peptides patterned after the C-terminus of SNAP25 have been used to understand the dynamics of complex formation, as well as of the neurosecretory cascade (Gutierrez et al. 1997; Ferrer-Montiel et al. 1998b; Apland et al. 1999). Recently, however, an hexapeptide derived from the N-terminal domain of SNAP25 was reported to modulate the stability of the SNARE core complex and to inhibit Ca2+-evoked neurosecretion in chromaffin cells (Blanes-Mira et al. 2002). Notably, topical formulations containing this peptide displayed antiwrinkle activity in humans, similar to that exhibited by BoNT A (Benedetto 1999; Blanes-Mira et al. 2002). In addition, deletion of N-terminus domains or the use of specific antibodies suggest a critical contribution of this protein segment to SNARE complex assembly and regulated exocytosis (Chapman et al. 1994; Xu et al. 1999; Yang et al. 2000).

Analysis of the N-terminal primary structure of SNAP25 revealed the presence of two domains with distinct forecast propensities to form coiled coils and to adopt an α-helical secondary structure. Notably, the segment that exhibited the lowest predicted propensity to coiled coil and α-helix formation was the most critical for SNARE complex assembly. Deletion of the region encompassing residues 22Ala-44Ile from SNAP25 completely prevented the in vitro formation of the core SNARE complex. In marked contrast, removal of the segment 1Met-21Leu yielded SNARE complexes that dissociated at a lower temperature. Our results suggest that the protein domain comprising residues 22Ala-44Ile is critical for the interaction of SNAP25 with the SNARE proteins, while the 1Met-21Leu segment contributes to the thermal stability of the complex. These observations are in agreement with results using antibodies targeting these two protein domains (Xu et al. 1999). Taken together, our findings suggest that peptides patterned after these two protein segments, especially those emulating the segment 22Ala-44Ile, may efficiently modulate the stability and formation of the SNARE core complex. To substantiate this notion, we investigated the inhibitory activity of peptides mimicking domains 1Met-21Leu and 22Ala-44Ile of SNAP25. Both sequences reduced the in vitro formation of the SNARE core complex, being the peptide mimicking the 22Ala-44Ile segment the most potent. This conclusion was further underscored by the inhibitory activity of the 13-mer peptide SNAP25_N4 comprising the core sequence of the SNAP25 segment 22Ala-44Ile. Noteworthy, the in vitro inhibitory activity of SNARE complex formation of these peptides was paralleled by their efficacy abolishing Ca2+-evoked neurosecretion in chromaffin cells as well as in neurones. Most significant was the discovery that these peptides were able to translocate through the plasma membrane of intact cells, as evidenced by their neuroprotective activity of hippocampal neurones against excitotoxicity. Blockade of regulated exocytosis was observed at a concentration as low as 1 µm, which is 500-fold lower than that required to target the SNARE complex in vitro. This finding is consistent with the notion that the in vivo exocytosis-competent SNARE complex is the trans configuration, while in vitro predominates the cis form. It has been reported that the trans-SNARE complex is energetically less stable that the cis form and, thus more amenable to modulation by small molecules (Weber et al. 1998; Brunger 2001; Chen and Scheller 2001; Bruns and Jahn 2002; Melia et al. 2002).

Mechanistically, peptides mimicking the N-end of SNAP25 disrupt the interaction of the parental protein with syntaxin, as clearly evidenced by the full abrogation of the binary complex formed in vitro by both SNARE protein. The ternary complex was markedly destabilized when syntaxin was pre-incubated with the SNAP25_N2 peptide. In contrast, pre-assembled SNARE complexes were not affected by the peptide, as demonstrated by the lack of efficacy of the most potent amino acid sequence. Furthermore, an increment in the concentration of SNAP25 fully prevented the inhibitory activity of the peptides. These findings indicate that these peptides act as competitive antagonists of SNAP25 for the assembly of the SNARE complex.

This conclusion along with their high propensity to adopt an α-helical secondary structure, may be used to gain insights of their binding site by using the three-dimensional structure of the SNARE complex as a guide. For this task, the N-terminus of SNAP25 on the structure was trimmed to leave the sequences of the SNAP25_N2 and SNAP25_N4 (Fig. 9). As illustrated, a model of the peptide SNAP25_N4 docked on syntaxin and the C-terminal domain of SNAP25 shows that this peptide interacts with syntaxin segment 199His-217Met and SNAP25 domain 142Arg-160Leu, respectively. Notably, the SNAP25_N4 peptide can establish up to seven interactions with syntaxin and the SNAP25 C-end. As shown, residue 30Arg on the peptide forms a salt bridge with 145Glu on the C-terminal of SNAP25, while 31Arg and 28Ser interact with 206Glu on syntaxin. Furthermore, 36Glu on SNAP25_N4 pairs with 213His on syntaxin. In addition of these interacting pairs, several hydrophobic interactions notably contribute to peptide binding (Fig. 9, inset). Similar results are obtained for peptide SNAP25_N2. In contrast, peptides SNAP25_N1 and SNAP25_N3 are stabilized with fewer interactions, namely 19Asp on SNAP25_N3 with 142Arg on SNAP25 C-end, and 17Arg on SNAP25_N3 with 196Glu on syntaxin (not shown). Therefore, the higher number of interacting surface area exhibited by SNAP25_N2 and SNAP25_N4 provides a structural explanation of their higher efficacy and potency precluding the SNARE complex assembly and abrogating neuronal exocytosis.

Figure 9.

Structural model illustrating the putative binding site of peptides SNAP25_N2 on the SNARE complex. (Top) Putative interaction of SNAP25_N2 peptide on the SNARE complex preventing the interaction of the N-terminus domain of SNAP25. (Bottom) Enlargement showing the interactions of peptide SNAP25_N4 with syntaxin (red), the C-terminus of SNAP25 (yellow) and VAMP (blue). Peptide is shown in magenta. Interactions are highlighted.

In conclusion, the most salient contribution of this study is the discovery of cell-permeable peptides patterned after the SNAP25 N-terminus that efficiently prevent the formation of the SNARE complex and potently inhibit regulated exocytosis from excitable cells. These peptides significantly protected primary neurones against glucose-deprivation induced neurodegeneration. Therefore, our findings imply that peptides that target the binary complex assembled by SNAP25 and syntaxin may be novel therapeutics to efficiently attenuate dysfunctional exocytosis such as that characteristic of spasmodic disorders, excitotoxicity and pain transduction. This notion is further substantiated by the development of argireline, a hexapeptide encompassing the 12Glu-17Arg amino acid sequence, as an active antiwrinkle agent for human use (Blanes-Mira et al. 2002). Thus, the peptides SNAP25_N2 and SNAP25_N4 should be considered hit compounds for novel antispasmodic and analgesic drugs that complement BoNT-based therapies.


This work was supported by grants from La Fundación La Caixa (01/085–00 to AF-M), the Spanish Ministry of Science and Technology (MCYT) (SAF2000-0142 to AF-M and BMC2002-00845 to LMG), the Instituto de la Salud Carlos III (FIS-01/1162 to AF-M), The Generalitat Valenciana (GV01-01 to AF-M), the EU Biotechnology BIO4-CT97-2086 and the Fundación Ramón Areces to EPP.