A peptide that mimics the C-terminal sequence of SNAP-25 inhibits secretory vesicle docking in chromaffin cells.

Excitation-secretion uncoupling peptides (ESUPs) are inhibitors of Ca2+-dependent exocytosis in neural and endocrine cells. Their mechanism of action, however, remains elusive. We report that ESUP-A, a 20-mer peptide patterned after the C terminus of SNAP-25 (synaptosomal associated protein of 25 kDa) and containing the cleavage sequence for botulinum neurotoxin A (BoNT A), abrogates the slow, ATP-dependent component of the exocytotic pathway, without affecting the fast, ATP-independent, Ca2+-mediated fusion event. Ultrastructural analysis indicates that ESUP-A induces a drastic accumulation of dense-core vesicles near the plasma membrane, mimicking the effect of BoNT A. Together, these findings argue in favor of the notion that ESUP-A inhibits ATP-primed exocytosis by blocking vesicle docking. Identification of blocking peptides which mimic sequences that bind to complementary partner domains on interacting proteins of the exocytotic machinery provides new pharmacological tools to dissect the molecular and mechanistic details of neurosecretion. Our findings may assist in developing ESUPs as substitute drugs to BoNTs for the treatment of spasmodic disorders.

Ca 2ϩ -dependent exocytosis is a highly regulated process in neural and endocrine cells. The molecular components and events involved in the exocytotic cascade are fast-emerging thanks to the convergence of genetic, biochemical, pharmacological, and biophysical approaches (1)(2)(3)(4)(5)(6)(7). The information accrued in recent years has led to the formulation of the SNAP 1 receptors (SNARE) model to describe the final steps of the secretion cascade (1)(2)(3)(4)(5)(6)(7)(8)(9). The SNARE hypothesis distinguishes three distinct stages in the pathway, namely docking, priming, and fusion. Vesicle docking refers to the process by which cargo vesicles are targeted to the plasma membrane at the active zone, although they are not competent for Ca 2ϩ -triggered fusion (4). After docking, vesicles are activated by an ATP-dependent step known as vesicle priming. Primed vesicles are readily releasable in response to a transient Ca 2ϩ elevation that triggers the fusion event.
Biochemically, docking is associated with the formation of a 7 S ternary complex involving the vesicle membrane protein synaptobrevin, also known as vesicle-associated membrane protein (VAMP), which is the vesicle SNARE (v-SNARE), and two plasma membrane proteins SNAP-25 and syntaxin, which comprise the target SNARE (t-SNARE) (2-6, 10 -12). Vesicle priming is initiated by binding of the soluble proteins NSF (for N-ethylmaleimide-sensitive factor, an ATPase) and SNAPs (for soluble NSF attachment proteins, which are not related to  to the SNAP receptors (2)(3)(4)(5)(6). Specifically, VAMP binds to the SNAP-25-syntaxin heterodimer forming a ternary core complex that serves as a receptor for SNAPs and recruits NSF forming a 20 S complex (2-6, 9, 10, 13, 14). ATP hydrolysis by bound NSF energizes the secretory vesicles to the primed state, in which they are ready to fuse with the plasma membrane and release their content in response to the Ca 2ϩ signal (1)(2)(3)(4)(5)(6)(7). The identification of an ATP-dependent, slow component in the final steps of the exocytotic cascade lends support to the SNARE hypothesis.
The discovery that Clostridial neurotoxins target-specific components of the v-SNARE and t-SNARE has contributed to our knowledge of the molecular entities comprising the exocytotic machinery, as well as the molecular events involved in Ca 2ϩ -mediated neuroexocytosis. Botulinum neurotoxins (BoNT) B, D, F, and G, and the structurally related tetanus toxin specifically cleave VAMP at different sites (15)(16)(17), whereas BoNT A and E cleave SNAP-25 at the C terminus (18,19), and BoNT C cuts syntaxin and SNAP-25 (20,21). Cleavage of any of these proteins prevents the formation of the core complex and abrogates Ca 2ϩ -triggered exocytosis (4).
Recently, synthetic peptides that mimic the amino acid se-quence of segments from synaptotagmin (22,23), SNAPs (24), synaptobrevin (25) and SNAP-25 (26,27) were shown to be specific inhibitors of neurosecretion. These peptides, for which the term ESUP was coined to highlight their activity (26), are useful pharmacological tools to probe the functional role of distinct protein components in the secretory machinery, to dissect the contribution of specific domains in the proteinprotein interactions that mediate the process and to identify steps in the exocytotic cascade. The use of this new set of reagents, however, remains limited because the mechanism underlying their inhibitory activity is unkonwn. Here, we characterize the molecular steps of the exocytotic process that are sensitive to the blocking activity of a 20-mer peptide that mimics the amino acid sequence of the C-terminal domain of SNAP-25 (SNAP-25-(187-206): SNKTRIDEANQRATKM-LGSG; denoted as ESUP-A). We find that ESUP-A arrests Ca 2ϩ -dependent secretion from permeabilized chromaffin cells by inhibiting vesicle docking.

Reagents-[ 3 H]Noradrenaline was from Du Pont. t-butoxycarbonyl
and Fmoc amino acids, with standard side chain protecting groups, were obtained from Applied Biosystems (Foster City, CA), NovaBiochem (La Jolla, CA), or Peninsula Laboratories (Belmont, CA). Solvents, reagents, and resins for peptide synthesis were obtained from Applied Biosystems (Foster City, CA). All other reagents were of analytical grade from Sigma. BoNT A was a gift of B. R. DasGupta (University of Wisconsin, Madison, WI).
Peptide Synthesis and Purification-ESUP-A (SNKTRIDEANQRAT-KMLGSG) and ESUP-A RDM (TDSSGREMIKANKQLANGTR) were synthesized by t-butoxycarbonyl or Fastmoc® Fmoc chemistries in an Applied Biosystems 431A automated solid-phase peptide synthesizer, and cleaved as described (26). Cleaved peptides were purified by reversephase HPLC with a VydaC C-18 semipreparative column. Samples of crude peptide (10 -20 mg) were dissolved in 0.1% trifluoroacetic acid, applied to the column, and eluted with a linear gradient of 90% acetonitrile in 0.1% trifluoroacetic acid. Eluted peaks were monitored by absorbance measurements at 214 nm, pooled, and lyophilized. Peptide purity was assessed by reverse-phase HPLC in a VydaC C-18 analytical column.
Chromaffin Cell Cultures-Chromaffin cells were prepared from bovine adrenal glands by collagenase digestion and further separated from debris and erythrocytes by centrifugation on Percoll gradients as described (26,28). Cells were maintained in monolayer cultures at a density of 625,000 cells/cm 2 and were used between the 3rd and 6th day after plating. All the experiments were performed at 37°C.
Determination of Catecholamine Release from Detergent-permeabilized Chromaffin Cells-Secreted [ 3 H]noradrenaline was determined in digitonin-permeabilized cells as described (26,29). Briefly, cells were incubated with [ 3 H]noradrenaline (1 Ci/ml) in Dulbecco's modified Eagle's medium supplemented with 0.56 mM ascorbic acid during 4 h. Thereafter, monolayers were washed four times with a Krebs/HEPES basal solution: 15 mM HEPES, pH 7.4, with 134 mM NaCl, 4.7 mM KCl, 1.2 mM KH 2 PO 4 , 1.2 mM MgCl 2 , 2.5 mM CaCl 2 , 0.56 mM ascorbic acid, and 11 mM glucose. Cell permeabilization was accomplished with 20 M digitonin in 20 mM Pipes, pH 6.8 with 140 mM monosodium glutamate, 2 mM MgCl 2 , 2 mM Mg-ATP, and 5 mM EGTA. This incubation was carried out in the absence or presence of 100 M ESUP-A, ESUP-A RDM , or 100 nM dithiothreitol (DTT)-reduced BoNT A as indicated. BoNT A was reduced with 10 mM DTT for 30 min at 37°C. Following permeabilization, media were discarded, and cells were incubated for 10 additional 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 M buffered Ca 2ϩ solution. Media were collected, and released catecholamines as well as the total cell content were determined by liquid scintillation counting. Statistical significance was calculated using Student's t test with data from four or more independent experiments.
Vesicle density distribution was estimated in digitized micrographs using the NIH Image analysis program. For each chromaffin cell, the distance between the dense core granules and the plasma membrane was measured and binned using a bin width of 0.15 m. To obtain a distance distribution histogram (Fig. 5), the number of granules occurring at given point was plotted as a function of the average distance between the plasma membrane, defined at 0 m, and the nuclear membrane, at ϳ3 m. Data are given as mean with number of cells n Ն 5 for each treatment; the number of vesicles (v) counted per cell was 240 Յ v Յ 290, and the experimental error 10% Յ ⑀ Յ 15%.

ESUP-A Inhibits the Slow Component of the Ca 2ϩ -evoked NE
Release from Permeabilized Chromaffin Cells-Ca 2ϩ -evoked exocytosis from permeabilized chromaffin cells can be kinetically dissected in two distinct components (31, 32) ( Fig. 1): a fast component that may be detected few seconds after Ca 2ϩ application, lasts ϳ1 min, and releases ϳ40% of the total [ 3 H]NE, followed by a slow component that proceeds for ϳ15 min. This biphasic behavior of catecholamine release has been associated with the presence of at least two distinct pools of secretory granules at different stages of the exocytotic cascade (3)(4)(5)(6)33). The fast component appears to represent the release of neurotransmitter from a population of primed, readily releasable vesicles, whereas the slow component corresponds to a pool of docked vesicles that must undergo priming prior to fusion and other releasable pools of nondocked granules (3-6, 31, 33). We investigated if ESUP-A targeted a specific step of the secretory pathway. As shown in Fig. 1, incubation of permeabilized chromaffin cells with 100 M ESUP-A inhibited ϳ60% of catecholamine release by primarily altering the slow phase of secretion. These data suggest that the pool of primed vesicles is insensitive to the action of ESUP-A.
To gain further insights on the specific secretory step inhibited by ESUP-A, we attempted to dissociate the two distinct components of the exocytotic process by performing a double Ca 2ϩ pulse secretion assay. The rationale considered that a short (2 min) Ca 2ϩ pulse would primarily deplete the pool of primed vesicles, with minor effects on docked vesicles. A subsequent longer (10 min from chromaffin cells that is highly insensitive to 100 M ESUP-A (ϳ10% inhibition). In contrast, ESUP-A blocked ϳ80% of the Ca 2ϩ -dependent exocytosis elicited by the second, longer Ca 2ϩ pulse. A synthetic peptide with the same amino acid composition, yet random sequence (ESUP-A RDM ) blocked only ϳ10% of the Ca 2ϩ -evoked neurosecretion, confirming that the ESUP-A inhibitory activity is sequence-specific. Taken together, these data suggest that ESUP-A inhibits Ca 2ϩ -evoked catecholamine secretion by preventing vesicle docking or priming at the active zone, yet do not argue in favor of one or the other.
ESUP-A Blocks the Energy-dependent Steps of Ca 2ϩ -mediated Exocytosis in Chromaffin Cells-Catecholamine release from chromaffin cells is an energy-dependent process requiring ATP hydrolysis to induce and stabilize the primed state of secretory vesicles before the Ca 2ϩ -induced fusion event (1-7, 31, 32). Vesicle priming is composed primarily of an ATP-dependent component followed by a temperature-sensitive step that leads to the Ca 2ϩ -induced exocytosis (32)(33)(34). As shown in Fig. 3, incubation of permeabilized chromaffin cells with 100 M of ESUP-A attenuated ϳ60% of the Ca 2ϩ -induced catecholamine secretion when 2 mM ATP was present in the incubation and stimulation media (Fig. 3). The inhibitory effect was sequence-specific as evidenced by the inertness of ESUP-A RDM . Removal of ATP from the incubation and stimulation media diminished ϳ40% of the Ca 2ϩ -evoked release of [ 3 H]NE (32,34). Note that the ATP-independent secretion was not affected by ESUP-A. As expected, ESUP-A did not affect the temperature-sensitive step that follows the ATP-dependent transition (data not shown). These findings suggest that ESUP-A interferes with the steps that precede the ATP-dependent activation of secretory vesicles.
ESUP-A Induces Accumulation of Secretory Granules Near the Plasma Membrane of Chromaffin Cells-The morphology of secretory granules in chromaffin cells treated with peptides was examined by electron microscopy (Fig. 4). A main goal was to detect changes in secretory granule distribution induced by ESUP-A. For comparison purposes, the effect of BoNT A, known to prevent vesicle docking (35,36), was investigated. Electron micrographs of permeabilized, unstimulated chromaffin cells display a random distribution of dense core secretory vesicles (Fig. 4A). Ca 2ϩ stimulation induces a significant depletion of granules, as evidenced by the reduced number of vesicles in the cytosol (Fig. 4B), in accord with other reports (33). Notably, the presence of ESUP-A produced an accumulation of vesicles near the plasma membrane (Fig. 4C), similar to that produced by BoNT A (Fig. 4D).

FIG. 2. ESUP-A blocks catecholamine secretion in permeabilized chromaffin cells evoked by the second of two consecutive
Ca 2؉ pulses. Secretion was elicited by 10 M of free Ca 2ϩ or in basal media (5 mM EGTA) for 2 min (1 st pulse). Media were collected, and cells were incubated for 5 min in Ca 2ϩ -free buffer. Thereafter, a 10 min Ca 2ϩ stimulation was applied (2 nd pulse). Three conditions were assayed: Control, cells incubated with 100 M ESUP-A (SNKTRIDEANQRATK-MLGSG); and cells exposed to 100 M ESUP-A RDM , a 20-mer peptide with the same amino acid composition of ESUP-A but random sequence (TDSSGREMIKANKQLANGTR). The counts/min released from control cells under basal conditions were ϳ4,000 and increased to ϳ13,000 after stimulation with 10 M Ca 2ϩ . The total number of counts obtained from detergent-permeabilized cells was ϳ130,000. Thus, the normalized basal and the Ca 2ϩ -evoked release represent the 3.5% and ϳ10% of the total secretion, respectively. Data are given as mean Ϯ S.E. with n (number of experiments performed in triplicate) ϭ 4. Other conditions are the same as described in the legend to Fig. 1.

FIG. 4. ESUP-A induces changes in chromaffin granule distribution after Ca 2؉ stimulation of digitonin-permeabilized cells.
Cells were permeabilized as described in presence or absence of peptide. Secretion was triggered with 10 M Ca 2ϩ for 10 min. Thereafter, cells were collected by low speed centrifugation (1,000 ϫ g, 2 min), fixed, and processed for electron microscopy. Frequency distribution histograms show that unstimulated (control) chromaffin cells are characterized by the occurrence of two populations of cargo vesicles (Fig. 5A, left): a predominant population with a mean distance to the plasma membrane of ϳ0.5 m and a second most frequent with a mean distance of ϳ2.3 m. The relative frequency of occurrence of these two vesicle populations was not significantly altered by treating the cells with peptides or BoNT A (Fig. 5, B-D, left). Ca 2ϩ -stimulated exocytosis in control cells (Fig. 5A, right) or cells incubated with 100 M ESUP-A RDM (Fig. 5B, right) decreased the total number of vesicles, as evidenced by the reduction in the area of both granule populations (Figs. 4B and 5A, right). By contrast, treatment with 100 M ESUP-A or 100 nM BoNT A notably skewed the distribution of vesicles toward the plasma membrane, resulting in the virtual disappearance of the population centered at ϳ2.3 m and the ensuing increase of the granule population centered at ϳ0.5 m (Fig. 5, C and D,  right). The majority of vesicles gathered at the active zone, however, are not in tight contact with the membrane, lending support to the notion that ESUP-A prevents vesicle docking.
A Molecular Mechanism Underlying the Inhibitory Activity of ESUP-A-The main result of our study is that a 20-mer peptide that mimics the C terminus of SNAP-25, effectively inhibits Ca 2ϩ -dependent exocytosis in endocrine cells by preventing the docking of cargo vesicles at the plasma membrane. Catecholamine release in permeabilized chromaffin cells exhibits a biphasic time course characterized by an early ATP-independent component, and a slow, ATP-dependent phase (31)(32)(33)(34)(35). The ATP-independent phase presumably represents the Ca 2ϩ -mediated steps that ultimately lead to fusion, while the ATP-dependent phase may be associated with vesicle priming events (1)(2)(3)(4)(5)(6)(7)33). The final stages of neurosecretion in endocrine cells, therefore, may be described according to the following model (32,34), where A denotes the state that leads to the docked state B, and C and CЈ represent the energy-dependent intermediates in the process of vesicle activation that lead to the primed state D; Ca 2ϩ triggers the process of fusion. Several lines of evidence support this mechanism: (i) the identification of an ATP-dependent step ([ATP]) before the actual Ca 2ϩ -triggered fusion event (BfC) (32); (ii) the involvement of a temperature-sensitive process (⌬T) between the ATP-dependent step and the Ca 2ϩ -mediated fusion event, which appears to be the overall rate-limiting factor during priming (CfCЈ) (34); (iii) the abrogation of vesicle docking by Clostridial neurotoxins (AfB), arresting the ATP-dependent steps of the cascade (3)(4)(5)(6)(35)(36)(37). Although this model is an oversimplification of the complex exocytotic cascade, it provides a basic framework to understand the molecular mechanism underlying the inhibitory activity of excitation-secretion uncouplers such as ESUP-A. Given the mechanism depicted in Scheme 1, our results with ESUP-A are compatible with the concept that the peptide pre-  (Table I) (35)(36)(37).
How does ESUP-A prevent vesicle docking? During docking, a ternary complex involving VAMP, SNAP-25, and syntaxin is formed, the so-called SDS-resistant complex (3)(4)(5)(6)38). A sequence of protein-protein interactions starts with binding of SNAP proteins to the ternary aggregate and recruitment of NSF. The next step involves ATP hydrolysis that energizes and rearranges the core complex, making it competent for Ca 2ϩinduced fusion (3)(4)(5)(6). Since the C-terminal domain of SNAP-25 binds tightly to VAMP during docking (5,38), it is conceivable that VAMP is a complementary binding partner for the peptide. How does ESUP-A binding to VAMP block exocytosis? A plausible mechanism considers that ESUP-A competes with SNAP-25 for binding to VAMP and, thereby, prevents the formation of the critical SDS-resistant complex, comprising SNAP-25-VAMP-syntaxin (3-6, 38, 39). Absence of this ternary complex would hinder the association of SNAP and NSF proteins and the subsequent ATP-hydrolysis, therefore, preventing vesicle docking, priming, and fusion. Two findings suggest that the peptide binds directly to VAMP and interrupts the subsequent chain of protein-protein interaction events that lead to vesicle fusion. First, ESUP-A promotes the accumulation of secretory vesicles near the active zone (Figs. 4 and 5). Second, the peptide arrests the ATP-dependent maturation of the secretory granules (Fig. 3). Direct binding measurements of ESUP-A to VAMP in vitro have produced inconclusive results, presumably arising from low affinity (micromolar) of the interaction. Indeed, given a micromolar affinity constant, and as-suming a moderate dissociation rate for the peptide-VAMP complexes, it is likely that the majority of the bound peptide would dissociate in a 10-s wash (40). Development of ESUP peptides with higher affinity should provide decisive evidence. Table I, BoNT A and ESUP-A produce comparable effects on the Ca 2ϩ -regulated fusion events that lead to transmitter release. Both BoNT A and ESUP-A inhibit the slow component of the Ca 2ϩ -dependent exocytosis without affecting the fast component. Both attenuate to a similar extent the ATP-primed exocytosis and not the ATP-independent exocytosis, and both increase the number of vesicles accumulated near the active zone. It appears, therefore, that BoNT A and ESUP-A block exocytosis in chromaffin cells by inhibiting the process of vesicle docking.

ESUP-A Mimics the Action of BoNT A on Exocytosis-As indicated in Scheme 2 and summarized in
The finding that ESUPs are blockers of neurotransmitter release suggests that BoTxs disable the fusion process with such efficacy by a synergistic action of cleaving the substrate molecules that interact through noncovalent interactions to assemble into a fusion complex and of releasing peptide products that block by inhibiting vesicle docking. This implies that saturation with peptides designed to mimic putative sequences that bind to complementary partner sequences on interacting proteins, and combinations thereof, may abrogate vesicle fusion and transmitter release. The notion embodied in the ESUP activity suggests alternative pathways to regulate synaptic vesicle exocytosis and provides novel pharmacological tools to unravel the molecular components and details of the secretory cascade. The fact that these synthetic peptides mimic the action of Clostridial neurotoxins provides clues to develop peptide-based agents that may have practical medical application as potential therapy in disorders associated with involuntary muscle spasms.

TABLE I Inhibitory activities of BoNT A and ESUP-A on exocytosis
Fast and slow components denote the net catecholamine release in a double Ca 2ϩ -pulse secretion assay (Fig. 2). ATP-dependent and independent exocytosis refer to Ca 2ϩ -evoked release in presence or absence of 2 mM ATP at 37°C, respectively (Fig. 3)