Interspecies complementation identifies a pathway to assemble SNAREs

Abstract

Unc18 and SNARE proteins form the core of the membrane fusion complex at synapses. To understand the functional interactions within the core machinery, we adopted an "interspecies complementation" approach in Caenorhabditis elegans. Substitutions of individual SNAREs and Unc18 proteins with those from yeast fail to rescue fusion. However, synaptic transmission could be restored in worm-yeast chimeras when two key interfaces were present: an Habc-Unc18 contact site and an Unc18-SNARE motif contact site. A constitutively open form of Unc18 bypasses the requirement for the Habc-Unc18 interface. These data suggest that the Habc domain of syntaxin is required for Unc18 to adopt an open conformation; open Unc18 then templates SNARE complex formation. Finally, we demonstrate that the SNARE and Unc18 machinery in the nematode C. elegans can be replaced by yeast proteins and still carry out synaptic transmission, pointing to the deep evolutionary conservation of these two interfaces.

Keywords: Biological sciences; Cell biology; Functional aspects of cell biology; Molecular biology; Molecular neuroscience; Neuroscience.

Graphical abstract

Figure 1

Syntaxin domains differentially contribute to neurotransmission (A) (Top) Cartoons depicting syntaxin domains and the chimeras generated by swapping in the corresponding domains from yeast Ssop1. All the chimeras and the rescuing wild-type control (‘worm SYX-1’) are GFP-tagged and integrated into the syntaxin null background, syx-1. (Bottom) Representative locomotion trajectories collected for 1 min. Scale bar represent 1mm. (B) Average locomotion rates (speed) of 20 animals are compared for the same four strains. Data are displayed as scatter dot plots with mean and SEM; each point represents an animal. (C) Average paralysis time courses after aldicarb exposure (n = 3 independent experiments on 20 worms per experiment). Error bars represent SEM n.s. > 0.05; ∗∗∗ <0.001 (Student’s two-tailed t-test).

Figure 2

The syntaxin Habc is required for neurotransmission (A) (top) Cartoons depicting syntaxin with the Habc domain swapped in from the placozoan Trichoplax adhaerens, the choanoflagellate Monosiga brevicollis, and yeast Saccharomyces cerevisiae. All the chimeras and the wild-type control (worm SYX-1) are GFP-tagged and expressed in the syx-1 null strain. The GFP tag mildly decreases function compared to the true wild-type (WT) control. (bottom) Representative locomotion trajectories collected for 1 min. Scale bar represent 1mm. (B) Average locomotion speed of 20 animals compared for the same four strains. (C) Average paralysis time courses after aldicarb exposure (n = 3 independent experiments on 20 worms per experiment). (D) Representative traces of endogenous miniature postsynaptic currents (minis) recorded from the body muscle of syntaxin chimeras. (E) Quantification of the mini frequency. Neuronal expression of GFP-tagged worm syntaxin-rescued mini frequency of syntaxin null animals, but not to wild-type levels (WT, 48.5 ± 7.0 minis/second; n = 6 vs. worm SYX-1, 33 ± 2.7 minis/second; n = 11 vs. syx-1 null, 0.03 ± 0.019 minis/second; n = 6). Mini frequency in Trichoplax Habc chimeras (29.9 ± 3.2 minis/second; n = 9) was not different from rescued worm syntaxin. The average rate of fusion measured from choanoflagellate Habc chimeras (12.9 ± 1.9 minis/second; n = 12) and yeast-Habc chimeras (0.1 ± 0.06 minis/second; n = 8) was significantly lower than that measured from the syntaxin-rescued strain. Speed and mini frequency are displayed as scatterplots with mean and SEM; each point represents a single animal. n.s. > 0.05; ∗ <0.05; ∗∗∗ <0.001 (Student’s two-tailed t-test). Error bars in aldicarb curves represent the SEM.

Figure 3

Matching the Habc domain with Unc18 provides only minimal rescue (A) Average locomotion rates (speed) in, from left to right: syntaxin null animals; syx-1 unc-18 double mutants; the yeast-Habc chimera; the yeast-Habc chimera overexpressing worm UNC-18; and the yeast-Habc chimera overexpressing Sec1p (yeast Unc18). (B) Quantification of the mini frequency in the same five strains, all were similarly defective in synaptic transmission: syx-1 null: 0.03 ± 0.019 minis/second, n = 6; syx-1 unc-18 null: 0.08 ± 0.017 minis/second, n = 10; yeast-Habc chimera: 0.10 ± 0.057 minis/second, n = 8; yeast-Habc chimera overexpressing worm UNC-18: 0.02 ± 0.005 minis/second, n = 17; yeast-Habc chimera overexpressing Sec1p: 0.13 ± 0.021 minis/second, n = 22. Note, syx-1 null and yeast-Habc are reproduced from Figure 2. (C) Representative traces of endogenous miniature postsynaptic currents (minis) recorded from the body muscle. (D) Quantification of docked synaptic vesicles at acetylcholine synapses in the syx-1 unc-18 double mutants, the yeast-Habc chimera, the yeast-Habc chimera overexpressing worm UNC-18, and the yeast-Habc chimera overexpressing Sec1p (yeast UNC-18). All had similarly reduced docking compared to wild-type animals (docked SVs/per ACh synapse: the wild type, 3.9 ± 0.48, n = 21; syx-1 unc-18, 0.95 ± 0.15 docked SV/synapse, n = 20; yeast-Habc chimera, 0.35 ± 0.14, n = 31; yeast-Habc chimera overexpressing UNC-18: 1.3 ± 0.21, n = 19; yeast-Habc chimera overexpressing Sec1p: 1.9 ± 0.31, n = 15). (E) Representative electron micrographs of the neuromuscular junctions in the ventral nerve cord in the respective strains. Arrows indicate docked vesicles. All micrographs are displayed at the same magnification. Scale bar represents 100 nm. Grouped data are displayed as scatterplots with mean and SEM. In locomotion assays and physiological assays, each point represents one animal; in EM, each point represents a synapse. n.s. > 0.05; ∗∗∗ <0.001 (Student’s two-tailed t-test for locomotion and physiological assays; Mann-Whitney for EM).

Figure 4

Synaptic transmission is restored with two interaction interfaces: Unc18 – Habc and Unc18 – SNARE domain (A) Average locomotion speed in, from left to right: the chimeric yeast-Habc chimera overexpressing Sec1p (yeast Unc18); the yeast-Habc chimera overexpressing the Sec1p chimera (yeast Unc18) with the SNARE interactions restored; syntaxin mutants overexpressing the full yeast SNARE complex and Sec1p without a matching Habc interaction; syntaxin mutants overexpressing the full yeast SNARE complex and Sec1p with a matching Habc interaction. (B) Quantification of the mini frequency in the same four strains. When the two interaction surfaces are restored, synaptic transmission is rescued: yeast-Habc chimera overexpressing Sec1p: 0.13 ± 0.021 minis/second, n = 22; yeast-Habc chimera overexpressing the Sec1p chimera: 9.62 ± 1.636 minis/second, n = 12; overexpression of yeast SNARE complex with worm Habc + overexpression of Sec1p: 0.09 ± 0.033 minis/second, n = 11; overexpression of yeast Habc-SNARE with yeast Habc + overexpression of Sec1p: 6.70 ± 2.095 minis/second, n = 11). Note, yeast-Habc chimera + overexpression of Sec1p are reproduced from Figure 3. (C) Representative traces of endogenous miniature postsynaptic currents (Minis) recorded from the body wall muscle. (D) Quantification of docked synaptic vesicles in the yeast-Habc chimera overexpressing Sec1p; the yeast-Habc chimera overexpressing Sec1p with the SNARE interactions restored; syntaxin mutants overexpressing the full yeast SNARE complex and yeast Sec1p without a matching Habc interaction; syntaxin mutants overexpressing the full yeast SNARE complex and Sec1p with a matching Habc interaction; and wild-type animals. When the two interaction surfaces are restored, synaptic vesicle docking is restored (docked SVs/per ACh synapse: yeast-Habc chimera + overexpression of Sec1p: 1.9 ± 0.31, n = 15; yeast-Habc chimera + overexpression of the Sec1p chimera: 3.8 ± 0.47, n = 16; overexpression of the yeast SNARE complex with worm Habc + overexpression of Sec1p: 0.95 ± 0.093, n = 19; overexpression of the yeast SNARE complex with yeast Habc + overexpression of Sec1p: 6.9 ± 0.49, n = 15; wild type, 3.9 ± 0.48, n = 21). Note that the wild type, and the yeast-Habc chimera with overexpression of Sec1p, are the same data as Figure 3D. (E) Representative electron micrographs of the neuromuscular junctions in the ventral nerve cord in the respective strains. Arrows indicate docked vesicles. All micrographs are displayed at the same magnification. Scale bar represents 100nm. Grouped data are displayed as scatterplots with mean and SEM. In locomotion assays and physiological assays, each point represents one animal; in EM, each point represents a synapse. n.s. > 0.05; ∗∗ <0.01; ∗∗∗ <0.001 (Student’s two-tailed t-test for locomotion and physiological assays; Mann-Whitney for EM).

Figure 5

Syntaxin Habc domain opens Unc18 (A) Expression of P334A UNC-18 mutation “Open UNC-18” in the yeast-Habc chimera background increased the locomotion speed 6-fold (n = 20). (B) Locked open UNC-18 makes yeast-Habc chimeras more sensitive to aldicarb than WT UNC-18— indicating a restoration of ACh release in the open UNC-18 background. (C) Open UNC-18 in yeast-Habc chimera increased the frequency of the endogenous miniature postsynaptic currents compared to yeast-Habc chimeras expressing WT UNC-18 (yeast-Habc chimera + overexpression of UNC-18: 0.02 ± 0.005 minis/second, n = 17; yeast-Habc chimera + overexpression of open-UNC-18: 1.80 ± 0.704 minis/second, n = 12). (D) Representative traces of the endogenous miniature postsynaptic currents from indicated genotypes. (E) Open UNC-18 restores docking to the yeast-Habc chimeras. Representative electron micrographs of the neuromuscular junctions in the ventral nerve cord in the animals expressing the yeast-Habc chimera with worm UNC-18 (left) or “open” worm UNC-18 (right). All micrographs are displayed at the same magnification. Scale bar represents 100 nm. (F) Quantification of docking in the same two strains (yeast-Habc chimera + overexpression of UNC-18: 1.3 ± 0.21 docked SV/synapse, n = 19; yeast-Habc chimera + overexpression of open UNC-18: 4.6 ± 0.36 docked SV/synapse, n = 21). Note: yeast-Habc chimera + overexpression of UNC-18 data and sample micrograph are the same as Figure 3D. Speed, mPSC frequency, and docking are displayed as scatterplots with mean and SEM. In locomotion assays and physiological assays, each point represents one animal; in EM, each point represents a synapse. n.s. > 0.05; ∗∗ <0.01; ∗∗∗ <0.001 (Student’s two-tailed t-test for locomotion and physiological assays; Mann-Whitney for EM). Error bars in aldicarb sensitivity curves represent the SEM.

Figure 6

Model of syntaxin Habc domain function In step 1, Unc18 binds closed syntaxin during trafficking to axons. In step 2, the active zone protein Unc13 converts syntaxin to the open configuration. In step 3, the Habc domain then converts Unc18 to an open conformation. In step 4, open Unc18 binds the SNARE domains of syntaxin and synaptobrevin, to align and nucleate SNARE complex assembly. In step 5, SNAP-25, complexin, and synaptotagmin are recruited by unknown mechanisms to form an SNARE complex fully “primed” for fusion. Our results do not explicitly exclude an alternative sequence of steps; for example, the “opening” of UNC-18 could precede the “opening” of syntaxin.