Biomolecules, Vol. 12, Pages 1779: SNARE Modulators and SNARE Mimetic Peptides

2.1. SNARE ComplexUpon cleavage by clostridial neurotoxins, VAMP2 and syntaxin 1, larger fragments containing most of the SNARE motif are released from the membrane, whereas a large fragment of SNAP-25 remains membrane-bound. SNARE cleavage products do not support membrane fusion and exert a dominant negative effect by forming non-productive SNARE complexes with intact SNARE proteins. Injection or viral delivery of SNAP-25 fragments into Aplysia cholinergic neurons, where such fragments resembled cleavage products generated by botulinum neurotoxins BoNT A and E (residues 1–180 and 1–197), reduced postsynaptic currents by 50% and inhibited glutamate release in cultured hippocampal neurons by 40–60% [43]. Well over 20 such functional peptides have been reported to date (Figure 2). Their sequences are further detailed in the Supplementary Table S1, and a few notable examples are detailed below. When transfected into insulinoma cells, a SNAP-25 fragment (residues 1–197) inhibited insulin secretion evoked by depolarization or high glucose to levels comparable to BoNT A application [44]. Notably, in permeabilized PC12 cells treated with BoNT E, calcium-dependent norepinephrine release is partially rescued by adding a 26-mer peptide corresponding to the cleaved C-terminal coil of SNAP-25 [45]. Rescue can be enhanced by the introduction of a D186A mutation into the peptide sequence. This result indicates that the continuity of the C-terminal helix is not essential for SNAP-25 function. However, a similar approach applied to VAMP2 cleaved by BoNT D failed to rescue exocytosis. A cytoplasmic fragment of VAMP2 released by the tetanus toxin (residues 1–93) inhibited neurotransmitter release at an Aplysia cholinergic synapse [46]. A similar fragment of VAMP2 (residues 1–96) inhibited fusion-pore expansion and blocked the release of neurotransmitters and peptides from transfected astrocytes [47]. VAMP2 (residues 1–94) and syntaxin 1 fragments (residues 1–265) have been used extensively to block lipid and content mixing in liposome-based reconstitution assays [48]. Notably, the treatment of mouse N2A and human SH-SY5Y neuroblastoma cell lines but not of neurons with BoNT C or BoNT C/D, targeting syntaxin 1 and VAMP2, resulted in significant cytotoxicity [49].Cytotoxicity was also observed with syntaxin 1 (201–245) and VAMP2 (25–52) peptides, which can form aberrant SNARE complexes. A similar approach was applied to SNARE proteins that are not natural substrates for clostridial neurotoxins. A SNAP-23 fragment lacking 8 C-terminal amino acids (modeled after SNAP-25 cleaved by BoNT E) inhibited constitutive exocytosis of glutamate transporter EAAC1 in transfected C6 glioma cells [50]. SNAP-25 cleavage by BoNT A, C and E releases short C-terminal peptides containing the C-terminal part of the second SNARE motif. Multiple studies indicated that these peptides inhibit membrane fusion mediated by SNAP-25. A 20-mer peptide encompassing the C-terminal region of SNAP-25 (residues 187–206) blocked calcium-evoked catecholamine release from permeabilized chromaffin cells with IC50 ~20 μM [51]. A 26-mer peptide modeled after SNAP-25 cleavage by BONT E also inhibited exocytosis in permeabilized chromaffin cells [52]. Finally, SNAP-25 C-terminal peptides (residues 187–206, 170–189 and 181–206) inhibited neurotransmission in an Aplysia cholinergic synapse by ~40% after injection into the presynaptic neuron [53].Several peptides derived from the SNARE motif sequence of various SNARE proteins have been demonstrated to inhibit SNARE-dependent membrane fusion. Peptides derived from the N-terminal region of SNAP-25 inhibited SNARE-complex formation and exocytosis from permeabilized chromaffin cells. Highly potent peptide containing residues 22–44 also protected primary hippocampal neurons against glutamate-induced neurotoxicity and disrupted the SNAP-25/syntaxin 1 binary complex [54]. The hexapeptide Ac-EEMQRR-NH2 (called argireline) patterned after the N-terminal part of the first SNAP-25 SNARE motif (residues 12–17) interfered with the assembly of the SNARE complex, and inhibited calcium-dependent catecholamine release from permeabilized chromaffin cells [55]. Argireline penetrated through the skin epidermal layer and displayed anti-wrinkle activity in human subjects. Syntaxin 1 SNARE motif peptides (residues 229–251 and 197–219) inhibited calcium-induced insulin secretion from permeabilized pancreatic beta-cells, increased basal insulin release and had no effect on insulin release induced by non-hydrolysable GTP analogs [56]. Similar myristoylated peptides (residues 233–245 and 200–212) inhibited glucose-induced insulin secretion from intact beta-cells [57]. The introduction of a longer syntaxin 1 peptide containing the entire SNARE motif (residues 162–265) by a patch electrode inhibited neurotransmitter release in primary hippocampal neurons [58].Most peptides cannot diffuse efficiently through biological membranes. To improve intracellular delivery, SNARE-derived peptides were fused with protein-transduction sequences (e.g., TAT sequence). TAT-fused syntaxin 1 peptide (residues 202–265) was effectively taken up by intact cells and inhibited regulated exocytosis in pancreatic beta-cells [59] and PC12 cells [60]. Furthermore, 17-mer peptides derived from the N-terminal region of SNARE motif in VAMP2 and VAMP8 inhibited SNARE-complex formation and reduced mast-cell degranulation [61]. When fused with protein-transduction sequences, these peptides improved the symptoms of atopic dermatitis in mouse models.SNARE motif peptides were also used to test the parallel zippering model of SNARE-protein function. This model predicts that the zippering is initiated at the N-terminus of the SNARE motif and proceeds towards the C-terminus. Therefore, N-terminal peptides are predicted to inhibit membrane fusion more effectively than C-terminal peptides, by arresting the nucleation of the SNARE complex. In liposome-fusion assays with reconstituted neuronal SNARE proteins, the N-terminal VAMP2 peptide (residues 29–56) potently blocked membrane fusion [62]. In contrast, the C-terminal peptide (residues 5792) accelerated membrane fusion, likely by structuring the membrane-proximal coiled coil region in the t-SNARE.A search for peptide inhibitors of SNARE complex formation and membrane fusion was conducted using the combinatorial 17-mer α-helix-constrained peptide library [63]. A potent hit was identified, with amino acid sequence: acetyl-SAAEAFAKLYAEAFAKG-NH2, unrelated to SNARE-protein sequences. The peptide blocked calcium-evoked catecholamine secretion from permeabilized chromaffin cells and glutamate release from primary hippocampal neurons.

In summary, functional peptides derived from the SNARE motif sequence potently inhibit membrane fusion by forming non-productive SNARE complexes with endogenous SNAREs. These peptides are active in a wide range of functional assays and systems from in vitro reconstituted liposome-fusion assays to permeabilized and intact cell-based assays. When fused with protein-transduction sequences, the peptides can efficiently penetrate through the biological membranes, opening a path to their use in basic and applied science.

2.2. SNAREs/SM ProteinsThe important role of SNARE regulators and their role in the process of intracellular membrane fusion, mediated by SNARE proteins, has been recognized long ago, e.g., [64]. In vivo, SNARE proteins are critically assisted by SM proteins, an evolutionarily conserved protein family essential for intracellular membrane fusion. SM proteins have been shown to interact with individual SNAREs (v-SNAREs, syntaxins) and SNARE complexes by structurally diverse mechanisms [65,66]. SM protein Munc18-1 binds the N-terminal domain of syntaxin 1 in the closed conformation (the domain is folded back onto the SNARE motif, preventing SNARE complex assembly) and, independently, to the N-terminal sequence of syntaxin 1 (~20 N-terminal residues). Several lines of evidence suggest that SM proteins guide and chaperone SNARE proteins to facilitate SNARE-complex assembly, but conflicting data exist on the functional significance of diverse SM/SNARE binding modes. For example, the binding of SM protein Munc-18 to syntaxin 1 N-terminal peptide was shown to be essential for neurotransmission in C. elegans [67]. Other studies indicated that this binding mode is not required but plays a more subtle regulatory role in C. elegans, PC12 cells [68] and mammalian neurons [69]. It remains challenging to use peptides for the analysis of SM/closed syntaxin complexes, because the binding interface is large and convoluted, whereas the binding affinity is low nanomolar or higher. In contrast, the interaction via the short N-terminal sequence of syntaxins is ideally suited for this approach. N-terminal syntaxin 1A peptide (residues 2–16) interfered with Munc18-1/neuronal SNARE-complex assembly and inhibited neurotransmission at the calyx of Held synapse [70]. Peptide containing the D3R mutation, which disrupts the interaction, had no effect.Notably, cell-based secretion assays indicated that syntaxin fragments may inhibit membrane fusion by acting as a sink for endogenous SM proteins and not by forming aberrant SNARE complexes [70,71]. SM proteins Munc18-1 and 3 contain a SNARE-like peptide (residues 327–351) that structurally and functionally resembles the C-terminal peptide of VAMP2 (residues 60–84) [72]. Both peptides activated membrane fusion by SNAREpins in reconstituted liposome-based assays by a similar mechanism. 2.3. Other Functional Interfaces in SNARE Proteins Involved in Membrane FusionSNARE proteins are thought to initiate membrane fusion by bringing two membranes in close opposition. Several studies have indicated that juxtamembrane and transmembrane regions in SNAREs are required to disrupt fusing membranes and to catalyze lipid mixing. Juxtamembrane peptide from VAMP2 (residues 79–94) was required for membrane fusion in reconstituted liposome-based assays but could be functionally substituted with an unrelated membrane-destabilizing peptide [73]. Mutations in conserved hydrophobic or positively charged amino acid residues or extending the length of the juxtamembrane linker region by at least two residues inactivated SNARE function. Vicinal tryptophan residues in the juxtamembrane region are highly conserved in most VAMP2 proteins. In vitro studies with peptides derived from the juxtamembrane and transmembrane regions of Drosophila VAMP2 (residues 75–121) showed that peptides containing two, one or no tryptophans were positioned differently in the micelle, indicating that these residues are important for VAMP2 orientation in the vesicle membrane [74]. The VAMP2 peptide containing both regions (residues 83–116) enhanced stalk and pore formation in highly curved vesicles in the presence of phosphatidylserine [75].In vivo, SNARE complexes are disassembled by AAA ATPase NSF and accessory proteins. Inhibiting NSF function should increase non-productive cis-SNARE complexes and block membrane fusion. TAT sequence-fused 22-mer peptides were derived from NSF-inhibited exocytosis of the von Willebrand factor in endothelial cells [76]. Peptides targeted different domains of NSF responsible for various functions. The most potent peptide was obtained from the sequence involved in NSF homooligomerization. Multiple proteins regulate SNARE function, including the abundant presynaptic protein α-synuclein, linked to Parkinson’s disease pathology. α-synuclein was shown to bind VAMP2 and to promote SNARE-complex assembly in vitro and in vivo [77]. α-synuclein forms oligomers that also bind VAMP2, cluster small vesicles and inhibit membrane fusion [78]. These effects were reversed by 30-mer peptides derived from the α-synuclein C-terminal region involved in VAMP2 interaction, with the most potent peptide containing residues 96–125.Calcium-triggered membrane fusion is of critical importance for nervous, endocrine, and immune systems. SNARE proteins are not intrinsically sensitive to calcium; therefore, a calcium sensor protein is incorporated into the SNARE machinery [79]. The best studied calcium sensors are a family of double C2 domain membrane proteins called synaptotagmins. Synaptotagmin 1 C2B domain forms a primary interface with the partially assembled SNARE complex and enables calcium regulation [80].Recent studies identified potent functional peptides that disrupt the primary interface and block calcium-triggered membrane fusion [81,82]. Because the primary interface is linearly extended, internally crosslinked (stapled) peptides were designed and tested in reconstituted liposome-based assays. The most effective peptide was derived from SNAP-25 (residues 37–53) and contained four nonnatural amino acids for stapling. Remarkably, this peptide potently blocked the stimulated secretion of mucin in human or mouse airway epithelium cells. Calcium-dependent secretion of mucin is mediated by a different set of proteins (SNAREs: VAMP8, syntaxin 3 and SNAP-23; calcium sensor: synaptotagmin 2); however, the primary interface is sufficiently conserved in both neuronal and airway cells. To allow efficient intracellular delivery, the stapled peptide was conjugated with a protein-transduction sequence (PEN or TAT). When applied in aerosol form, the peptide markedly reduced the accumulation of mucin in mouse airways, which has therapeutical implications for the treatment of many lung diseases. 2.4. SNAREs/ion Channels and TransportersThe interaction of syntaxin 1 with N-type Ca2+-channels was initially demonstrated by co-immunoprecipitation studies [83,84] leading to a simple model of how calcium entry and fast neurotransmission can be coupled at the synapse. The interaction was mapped to the cytoplasmic loop II-III in N-type but not Q- or L-type Ca2+-channels [85,86]. The 87-mer peptide derived from the loop sequence, termed the synprint, was shown to block the interaction of the channel with the C-terminal region of syntaxin 1 (residues 181–288). The synprint peptide also interacted with syntaxin 1 and SNAP-25 only in the presence of calcium with EC50 ~ 20 μM. Synprint interaction with syntaxin 1 and SNAP-25 displayed distinct calcium dependence for various P/Q-type Ca2+-channel isoforms [87]. The introduction of synprint peptides into presynaptic cervical ganglion neurons modified neurotransmission by inhibiting fast synaptic responses and augmenting slow asynchronous synaptic responses and paired-pulse facilitation but had no effect on calcium currents [88]. Peptides derived from L-type Ca2+-channels had no effect. The introduction of synprint peptides into Xenopus embryos decreased the calcium sensitivity of neurotransmitter release in spinal neurons [89]. In another study, mutation in the P/Q-type Ca2+-channel (localized in loop I-II) prevented the modulation of channel properties by syntaxin 1 and SNAP-25 in an HEK293 cell-based overexpression model [90]. In addition, voltage-gated K+-channels Kv2.1 have been shown to interact functionally and structurally with neuronal SNARE proteins. Syntaxin 1 inhibited currents of the Kv2.1 channel overexpressed in HEK293 cells by interacting with the C-terminus of the channel [91]. The C-terminal peptide blocked the interaction and relieved channel inhibition, whereas the N-terminal peptide had no effect. Peptides derived from the C-terminal region of Kv2.1 channels disrupted interaction with syntaxin 1 and inhibited regulated exocytosis in cracked PC12 cells [92]. The injection of Kv2.1 C-terminal peptides into oocytes overexpressing Kv2.1 channels and syntaxin 1 reversed the effect of syntaxin 1 on the activation/inactivation of the Kv2.1 channel [93]. Surprisingly, SNAP-25 also inhibited the Kv2.1 current in pancreatic beta-cells by interacting with the N-terminus of the channel [94]. This effect can be reversed by the introduction of the N-terminal Kv2.1 peptide, which disrupts the interaction with SNAP-25, but not by the C-terminal peptide. A fragment of SNAP-25 (residues 1–180) inhibited the Kv2.1 current and enhanced glucose-dependent insulin secretion from pancreatic beta-cells [95]. The SNARE motif of syntaxin 1 inhibited the sodium current of ENaC in Xenopus oocytes and interacted with the C-terminus of ENaC[96]. The list of membrane proteins modulated by SNAREs includes several dozens of ion channels and neurotransmitter transporters, raising significant concerns about the general validity of these observations. SNARE proteins and particularly syntaxin 1 have been shown to interact promiscuously with multiple proteins in pulldown and co-immunoprecipitation assays [97]. Functional studies have often relied on the heterologous overexpression of membrane proteins (ion channels, transporters, SNAREs) in mammalian cell lines and oocytes. The overexpression of these proteins, in particular neuronal SNAREs, has profound effects on intracellular trafficking in many cases [70,71]. No specific sequence motif has emerged as a SNARE binding site, with multiple SNAREs binding to different regions of the same channel protein. The initial hypothesis of the direct coupling of SNARE fusion machinery to calcium channels at the presynapse was refuted by recent studies of active zone proteins involved in channel clustering [98].

留言 (0)

沒有登入
gif