Molecular Mechanisms Underlying Neurotransmitter Release

Abstract

Major recent advances and previous data have led to a plausible model of how key proteins mediate neurotransmitter release. In this model, the soluble N-ethylmaleimide-sensitive factor (NSF) attachment protein (SNAP) receptor (SNARE) proteins syntaxin-1, SNAP-25, and synaptobrevin form tight complexes that bring the membranes together and are crucial for membrane fusion. NSF and SNAPs disassemble SNARE complexes and ensure that fusion occurs through an exquisitely regulated pathway that starts with Munc18-1 bound to a closed conformation of syntaxin-1. Munc18-1 also binds to synaptobrevin, forming a template to assemble the SNARE complex when Munc13-1 opens syntaxin-1 while bridging the vesicle and plasma membranes. Synaptotagmin-1 and complexin bind to partially assembled SNARE complexes, likely stabilizing them and preventing fusion until Ca²⁺ binding to synaptotagmin-1 causes dissociation from the SNARE complex and induces interactions with phospholipids that help trigger release. Although fundamental questions remain about the mechanism of membrane fusion, these advances provide a framework to investigate the mechanisms underlying presynaptic plasticity.

Keywords: Munc13; Munc18; NSF; SNAPs; SNAREs; complexin; membrane fusion; neurotransmitter release; synaptotagmin.

Figure 1.

Working model of the basic steps that lead to neurotransmitter release. (a) Diagram showing the localization of synaptobrevin (red) and synaptotagmin-1 (blue) on a synaptic vesicle, and of αSNAP (brown) bound to a 1:2 complex between SNAP-25 (green) and syntaxin-1 (SNARE motif yellow; H_abc domain orange) on the plasma membrane. Helices formed by the SNAREs are represented by cylinders. αSNAP binding to the syntaxin-1-SNAP-25 complex hinders binding to synaptobrevin and SNARE complex formation. (b) Diagram illustrating that, even if trans-SNARE complexes between synaptobrevin, syntaxin-1 and SNAP-25 are formed, binding of αSNAP to these complexes prevents fusion, ensuring that neurotransmitter release does not occur through non-regulated pathways. (c) Diagram showing syntaxin-1 adopting a closed conformation that binds tightly to Munc18–1 (violet). This binary complex constitutes the starting point of the pathway that leads to neurotransmitter release. Closed syntaxin-1 may be available on the plasma membrane or may form after NSF dissociates an αSNAP-bound syntaxin-1-SNAP-25 complex (from panel a) or an αSNAP-bound trans-SNARE complex (from panel b). (d-e) The conserved C-terminal region of Munc13–1 (cyan) bridges the vesicle and plasma membranes through respective interactions involving the C₂C domain and the C₁-C₂B region, and opens syntaxin-1 by binding to the linker between the syntaxin-1 H_abc domain and SNARE motif. This action likely facilitates binding of Munc18–1 to synaptobrevin, forming a template complex (d) that initiates SNARE complex assembly upon binding of SNAP-25 to syntaxin-1 and synaptobrevin (e). Synaptotagmin-1 likely facilitates assembly by binding to SNAP-25 (not shown in d for simplicity). The synaptotagmin-1 C₂B domain binds to the partially assembled SNARE complex through the primary interface and to the plasma membrane through the polybasic region (e). Munc13–1 facilitates SNARE complex assembly but also limits the number of SNARE complexes that form and hinders C-terminal zippering by bridging the two membranes in an approximately perpendicular orientation that characterizes a loose primed state (LS). This inhibitory action may be aided by formation of Munc13–1 clusters. (f) Further but not complete C-terminal zippering of the SNARE complex, which is favored by binding of complexin (pink), forces the two membranes closer together and Munc13–1 must bridge the membrane in a slanted orientation, forming a tight primed state (TS) that has a much higher probability of release upon Ca²⁺ influx than LS. Complexin and synaptotagmin-1 stabilize this state and prevent disassembly of the trans-SNARE complexes by NSF/αSNAP, but hinder final zippering to prevent premature fusion. (g-h) Ca²⁺ binding to synaptotagmin-1 causes dissociation from the SNARE complex, relieving the inhibition and thus allowing final C-terminal zippering and synaptic vesicle fusion. The dissociated synaptotagmin-1 molecules, or other synaptotagmin-1 molecules that were not bound to the SNAREs (shown in the middle) accelerate fusion through interactions with the lipids, perhaps because they perturb the bilayers, bridge the two membranes and/or induce membrane curvature. The increase in vesicular release probability caused by accumulation of Ca²⁺ and DAG during repetitive stimulation is proposed to arise because Ca²⁺ and DAG bind to the C₂B and C₁ domains of Munc13–1, respectively, and favor the slanted orientation, shifting the equilibrium from LS to TS. The following features are not shown for simplicity: (b, e-h) the linker region between the two SNAP-25 SNARE motifs, which remains anchored on the plasma membrane through plamitoylation; (f-h) the N- and C-terminal regions of complexin, which bind to the membranes; (e-h) Munc18–1, which may remain bound to the N-terminal region of syntaxin-1; (e-f) additional synaptotagmin-1 molecules that may bind to the SNARE complex through the tripartite interface or form a synaptotagmin-1 oligomeric ring that has been proposed to prevent fusion before Ca²⁺ influx and is dissociated upon Ca²⁺ binding.

Figure 2.

The SNARE complex and membrane fusion. (a) The stalk model of membrane fusion, which postulates that the two membranes have to be brought into proximity, the bilayers bend and the proximal leaflets fuse to yield the so-called stalk intermediate, then the distal leaflets fuse to form the fusion pore (31). (b) Ribbon diagrams of the NMR structure of the syntaxin-1 H_abc domain (orange) (50) and the crystal structure of the SNARE complex formed by the SNARE motifs of syntaxin-1 (yellow), SNAP-25 (green) and synaptobrevin (red) (150) (PDB IDs 1BR0 and 1SFC, respectively). N and C indicate N- and C-termini, respectively. (c) Model postulating that formation of the SNARE complex brings the vesicle and plasma membranes together, inducing membrane fusion. C indicates the C-termini of synaptobrevin and syntaxin-1. (d) Diagram illustrating the formation of extended membrane-membrane interfaces induced by SNARE complex formation, which have been observed by cryo-EM (61). In (c,d), the H_abc domain is not shown.

Figure 3.

The 20S complex. (a) Cryo-EM structure of the 20S complex formed by NSF, αSNAP, synaptobrevin (red), syntaxin-1 (yellow) and SNAP-25 (green) (190) (PDB ID 3J96). The four molecules of αSNAP that bind around the SNARE four-helix bundle are colored in orange and brown in an alternative fashion. Similarly, the six molecules of NSF are colored in blue and cyan in an alternative fashion. The positions of the D1, D2 and N-domains of NSF, as well as the N-terminal hydrophobic loop of αSNAP, are indicated. (b) Model proposing that the core machinery that mediates yeast vacuolar fusion is a 20S complex formed by Sec18 (blue), Sec17 (orange) and the four vacuolar SNAREs (yellow, green and red) (143). Note that the small ellipse at the membrane-proximal end of each αSNAP molecule represents its N-terminal hydrophobic loop and insertion of the loop into the bilayers is postulated to make a key contribution to membrane fusion.

Figure 4.

Liposome fusion assays that recapitulate the dependence of synaptic vesicle fusion on Munc18–1, Munc13–1 and synaptotagmin-1. (a,b) Assays that monitored lipid mixing (a) and content mixing (b) of VSyt1-liposomes containing synaptobrevin (P:L ratio 1:500) and synaptotagmin-1 (P:L ratio 1:1,000) with T-liposomes containing syntaxin-1 (P:L ratio 1:800) bound to SNAP-25, in the presence of NSF and αSNAP. Under these conditions, there is no fusion without other additions (T + VSyt1, black curves). Inclusion of Munc18–1 and a Munc13–1 C-terminal fragment spanning the C₁, C₂B, MUN and C₂C domains (C₁C₂BMUNC₂C) leads to highly efficient Ca²⁺-dependent fusion (+M13 +M18, red curves), but there is no fusion when only one of these proteins is included (+M18, orange curves; +M13, blue curves) (85). Note that in the complete reaction with all the reagents there is some lipid mixing before Ca²⁺ addition, indicating that a few SNARE complexes are formed, but there is no content mixing, showing that there is not fusion. (c) Content mixing assays performed under analogous conditions but with V-liposomes containing only synaptobrevin (P:L ratio 1:500) in the presence of NSF, αSNAP, Munc13–1 C₁C₂BMUNC₂C and WT Munc18–1 (T + V, black curve), Munc18–1 D326K (red curve) or Munc18–1 L348R (blue curve). The D326K mutation that unfurls the Munc18–1 loop leads to efficient fusion before Ca²⁺ addition, whereas the L348R mutation that impairs synaptobrevin binding to Munc18–1 strongly hinders fusion (140). Note that the Ca²⁺-dependent fusion observed with WT Munc18–1 is highly efficient even though synaptotagmin-1 was absent [compare black curve in panel (c) with red curve in panel (b)]. (d) Content mixing assays performed under analogous conditions but with VSyt1-liposomes containing synaptobrevin at 1:10,000 P:L ratio and synaptotagmin-1 at 1:1,000 ratio (T + VSyt1, red curve) or V-liposomes containing only synaptobrevin at 1:10,000 P:L ratio (T + V, black curve), in the presence of NSF, αSNAP, Munc13–1 C₁C₂BMUNC₂C and WT Munc18–1. Note that inclusion of synaptotagmin-1 led to much more efficient fusion at this low synaptobrevin densities, and that experiments with VSyt1-lipsomes bearing synaptotagmin-1 with mutations in the Ca²⁺-binding sites of both C₂ domains (T + VSyt1 C₂A*B*, blue curve) led to similar fusion to that observed without synaptotagmin-1, showing that stimulation of fusion depends on Ca²⁺ binding to synaptagmin-1 (145).

Figure 5.

Structures of the SNARE complex assembly machinery. (a) Ribbon diagram of the crystal structure of Munc18–1 (violet) bound to closed syntaxin-1 (SNARE motif in yellow, N-terminal region in orange) (102) (PDB ID 3C98). The positions of the Munc18–1 domains (D1, D2, D3a and D3b), of the furled loop that covers the synaptobrevin binding site and of the syntaxin-1 N-peptide (N-pep) are indicated. (b) Superposition of crystal structures of the Munc18–1 homologue Vps33 (violet) bound to the SNARE motif of the syntaxin-1 homologue Vam3 (yellow) or the SNARE motif of the synaptobrevin homologue Nyv1 (red) (7) (PDB IDs 5BV0 and 5BUZ, respectively). (c) Domain diagram of Munc13–1. CaMb = calmodulin-binding region. (d) Model of the structure of a fragment spanning the C₁ (salmon), C₂B (blue) and MUN (cyan) domains of Munc13–1 built from the crystal structure of this fragment (177) and completing the Ca²⁺-binding region of the C₂B domain with the crystal structure of this domain bound to Ca²⁺ (137) (PDB IDs 5UE8 and 3KWU, respectively). Zinc ions are shown as yellow spheres and Ca²⁺ ions as orange spheres. N indicates the N-terminus of the fragment, where the Munc13–1 N-terminal region should emerge. A homology model of the Munc13–1 C₂C domain (blue) (116) is shown at its expected position at the C-terminus of the MUN domain. The DAG-binding site of the C₁ domain, the Ca²⁺/PIP₂-binding site of the C₂B domain and the membrane-binding site of the C₂C domain, predicted to bind to synaptic vesicles, are indicated. A peptide corresponding to the juxtamembrane region of synaptobrevin is shown in red in the position observed in the crystal structure of this peptide bound to the Munc13–1 MUN domain (163) (PDB ID 6A30). Note that the position of residue 92, which is close to the TM region, is far from the expected membrane-binding region of the C₂C domain.

Figure 6.

Model of how the Munc13–1 C-terminal region controls vesicular release probability and presynaptic plasticity. The model postulates that the Munc13–1 C-terminal region (cyan) can bridge the vesicle and plasma membranes in two orientations: i) an approximately perpendicular orientation that is favored in the absence of Ca²⁺ and allows initiation of SNARE complex assembly, but hinders C-terminal zippering and membrane fusion (a); and ii) a slanted orientation that allows full zippering of the SNARE complex and membrane fusion, and that is favored by Ca²⁺-binding to the C₂B domain and DAG binding to the C₁ domain (b) (116, 177). These two orientations have been proposed to underlie the formation of two primed states (LS and TS), with TS having a much higher probability for release upon Ca²⁺ influx (104). The equilibrium can be shifted toward TS before Ca²⁺ influx by factors such as complexins (Figure 1e,f), and by accumulation of DAG and Ca²⁺ during repetitive stimulation, leading to enhanced release probability.

Figure 7.

Coupling of synaptotagmin-1, SNARE and complexin function. (a) Domain diagram of complexin-1 with selected residue numbers indicated above and ribbon diagram of the crystal structure of the SNARE complex bound to a complexin-1 fragment (30) (PDB ID 1KIL). Synaptobrevin is in red, syntaxin-1 in yellow, SNAP-25 in green and the complexin-1 fragment in orange (accessory helix) and pink (central helix). (b) Ribbon diagrams of the NMR structures of the Ca²⁺-bound C₂A and C₂B domains of synaptotagmin-1 (49, 134) (PDB IDs 1BYN and 1K5W, respectively). Ca²⁺ ions are shown as orange spheres. In the diagram of the C₂B domain, selected residues that were implicated in binding to the SNARE complex and/or PIP₂ are represented by spheres and color coded (R322 and K325 blue; K324 and K326, cyan; E295 and Y338, magenta; R398 and R399, purple). (c-e) Ribbon diagrams of a representative conformer of the ensemble of NMR structures of the synaptotagmin-1 C₂B domain bound to the SNARE complex via the polybasic region (16) (c), of the crystal structure of synaptotagmin-1 C₂AB bound to the SNARE complex through the primary interface (193) (d) and of the crystal structure of synaptotagmin-1 C₂AB bound to a complexin-1-SNARE complex through the tripartite interface (194) (e) (PDB IDs 2N1T, 5KJ7 and 5W5C, respectively). The C₂A domain is not shown. (f) Model showing how the synaptotagmin-1 C₂B domain can bind simultaneously to the plasma membrane through the polybasic region and to the SNARE complex through the primary interface while complexin-1 can bind to the opposite side of the SNARE complex. The model was constructed by superimposing the structures of panels (a) and (d). Models of the plasma membrane and a synaptic vesicle are represented by gray spheres, with a PIP₂ molecule in the plasma membrane shown in red. The dashed lines indicate that the C-termini of the synaptobrevin and syntaxin-1 SNARE motifs should be close to the vesicle and plasma membranes, respectively; hence, the SNARE complex cannot be fully zippered as in the structure shown. Note also that the complexin-1 accessory helix points straight toward the vesicle in this orientation and hence should hinder C-terminal zippering of the SNARE complex. (g) Model of how the Ca²⁺-bound synaptotagmin-1 C₂B domain is expected to bind to the plasma membrane in an approximately perpendicular orientation that allows insertion of both Ca²⁺-binding loops into the bilayer. This orientation is incompatible with binding to the SNARE complex in the three modes of panels (c-e). Note that R322 and K325 can readily bind to PIP₂ whereas K324 and K326 point away from the membrane. The color codes of panels (a, b) are also used in panels (c-g). N and C indicated the N- and C-termini of the SNARE four-helix bundle, respectively.