Membrane Fusion pp. 171-172 (2011)
Chap. 7. Chasing the Trails of SNAREs and Lipids Along the Membrane Fusion Pathway
Young Yoon, Dae-Hyuk Kweon and Yeon-Kyun Shin

IV. SMALL MOLECULES AS INTERROGATORS FOR SNARE-MEDIATED MEMBRANE FUSION

It is widely accepted that the nucleation of the SNARE complex at the membrane distal N-terminal region corresponds to vesicle priming, whereas complex formation at the membrane-proximal C-terminal end drives membrane fusion. C-terminal complex formation may be assisted by auxiliary proteins such as synaptotagmins and complexins. Recently, Yang et al. (2010) asked if such sequential SNARE zippering could be blocked by small molecules. Finding small molecule SNARE blockers appears to be relatively tangible because coiled coils are known to be excellent targets for small molecules. For example, a few molecules were identified to inhibit formation of the six-helical bundle of HIV gp41 that is considered a key structural intermediate for membrane fusion required for vial entry (Balogh, Wu, Zhou, & Gochin, 2009; Cai & Gochin, 2007; Frey et al., 2006). A recombinant leucine zipper, in which a core-forming residue was changed to alanine, was shown to bind ring-structured small compounds such as benzene, cyclohexane, and toluene (Doerr, Case, Pelczer, & McLendon, 2004; Gonzalez, Plecs, & Alber, 1996). Furthermore, it was also shown that a designed four-helix bundle protein accepts anesthetics in a core hydrophobic layer (Zhang & Johansson, 2003, 2005). The existence of natural compounds that block SNARE complex formation has been recently verified (Jung et al., 2009), although the chemical properties of the candidate compounds were unknown.

Among natural compounds, polyphenols deserve special attention because they are abundant in natural products, have ring structures, and are known to be versatile in protein binding. Some polyphenols in diets can cross the bloodbrain barrier (Ehrnhoeferet al., 2006; Kocisko et al., 2003), suggesting that they can target SNARE proteins in central nervous system once identified as SNARE blockers. Yang et al. (2010) showed that certain small hydrophobic molecules (SHM) enabled a layer-by-layer control of SNARE zippering by intercalating into various points of the SNARE complex. As an initial step, 39 polyphenolic compounds representing 12 subgroups were screened for an inhibitory activity against SNARE-driven proteoliposome fusion. Ten polyphenol compounds showed significant inhibitory effect on membrane fusion, and the degree of SNARE complex formation, assessed by Western blotting, was reduced for all 10 compounds. By virtue of the simplicity of the in vitro fusion assay, this observation suggests that inhibition of membrane fusion by SHM was likely a direct consequence of inhibition of SNARE zippering.

Yang et al. performed further experiments using the three most potent compounds: delphinidin, cyanidin, and myricetin, and found out that these three SHMs intercalate into the inner layer of the SNARE core complex during helical bundle formation. Delphinidin and cyanidin bind to the N-terminal part of SNARE motif, inhibiting N-terminal nucleation of SNARE complex formation. On the other hand, myricetin has two binding sites: one in the middle of the SNARE motif and the other in the N-terminal part. This second binding site has an important functional consequence: It is observed that myricetin stops SNARE-mediated membrane fusion at the hemifusion state. Neurotransmission and SNARE complex formation in neuronal PC 12 cells also support the results from the in vitro study.

On the basis of these observations, it is proposed that the dynamic SNARE zippering process could be stopped at the point of interest by wedging an appropriate SHM into the zipper (Fig. 3). Delphinidin and cyanidin wedges into the SNARE zipper at the far N-terminus (P20~V42 of VAMP2), thereby preventing N-terminal nucleation of SNARE complex formation. Meanwhile, myricetin stops subsequent zippering toward the membrane-proximal part by wedging into the middle region of the SNARE zipper. The detailed Ala-scanning study by Yang et al. shows that the four-stranded coiled coil up to residue M46 with the frayed C-terminal halves may represent the trans SNARE complex blocked by myricetin.

FIGURE 3 Small molecule SNARE inhibitors. Several small molecules stop SNARE zippering by wedging the complex. After stopping membrane fusion by this wedge-like action of small molecules membrane structure was analyzed to correlate the SNARE conformation to membrane fusion stage. (A) Predocking state of vesicle. (B) The N-terminal partial complex is formed as observed by 4°C preincubation. This partial complex docks secretary vesicles near the plasma membrane. (C) In one of the previously popular SNARE-driven membrane fusion models, the assembled four-helical bundle with TMDs still at separate membranes waits for a calcium trigger. In this structure, hemifusion is assumed to be achieved by full zippering of the SNARE motifs, and inner leaflet mixing is coupled to TMD interaction. (D) calcium promotes full SNARE zippering extending to the TMDs, which opens fusion pore. (E) Delphinidin/cyanidin wedges into the SNARE zipper at the far N-terminus (P20~V42 of VAMP2), thereby preventing N-terminal nucleation of SNARE complex formation and even outer leaflet mixing. (F) Myricetin wedges into the SNARE zipper midway (N49~L63 of VAMP2), hampering residual C-terminal zippering and inner leaflet mixing. Hemifusion assay raises the possibility that this halfway zippering is sufficient to induce hemifusion. Results also raise the possibility that the N-terminal partial complex shown in (B) may be present only at 4°C but this partial complex seems to be enough for hemifusion at physiological temperatures. (G) In the trans-complex, VAMP2 coiling extends up to M46 and the residual C-terminal region is still frayed. (See Color Insert.)