PRRT2 Regulates Synaptic Fusion by Directly Modulating SNARE Complex Assembly

Abstract

Mutations in proline-rich transmembrane protein 2 (PRRT2) are associated with a range of paroxysmal neurological disorders. PRRT2 predominantly localizes to the pre-synaptic terminals and is believed to regulate neurotransmitter release. However, the mechanism of action is unclear. Here, we use reconstituted single vesicle and bulk fusion assays, combined with live cell imaging of single exocytotic events in PC12 cells and biophysical analysis, to delineate the physiological role of PRRT2. We report that PRRT2 selectively blocks the trans SNARE complex assembly and thus negatively regulates synaptic vesicle priming. This inhibition is actualized via weak interactions of the N-terminal proline-rich domain with the synaptic SNARE proteins. Furthermore, we demonstrate that paroxysmal dyskinesia-associated mutations in PRRT2 disrupt this SNARE-modulatory function and with efficiencies corresponding to the severity of the disease phenotype. Our findings provide insights into the molecular mechanisms through which loss-of-function mutations in PRRT2 result in paroxysmal neurological disorders.

Keywords: PRRT2; SNARE proteins; neurotransmitter release; paroxysmal dyskinesia; synaptic fusion.

Graphical abstract

Figure 1

PRRT2 Inhibits Synaptic SNARE-Mediated Fusion under Reconstituted Conditions (A) Purity of the recombinant PRRT2 purified using bacterial expression setup is confirmed using Coomassie-stained SDS-PAGE analysis. (B) Identity of the purified protein is verified using western blot analysis with a PRRT2 antibody. The native protein from the P2 synaptosomal fraction was used as positive control. (C) Purified PRRT2 reconstituted into VAMP2- or t-SNARE-containing liposomes using detergent dilution and dialysis method is analyzed using Coomassie-stained SDS-PAGE analysis. (D) Cryoelectron microscopy analysis of reconstituted t-liposomes shows that PRRT2 incorporation does not alter the size or other physical attributes of the liposomes. Representative micrographs of t-SNARE liposomes with or without PRRT2 (1:1) incorporated are shown. (E) PRRT2 introduced in either the v- or t-liposomes inhibited synaptic SNARE-mediated fusion of liposomes monitored by NBD dequenching assay. Negative control with soluble cytoplasmic domain of VAMP2 (CDV) added in excess to titrate out the t-SNAREs shows that PRRT2 is not inherently fusogenic. Representative fusion curves are shown. (F) PRRT2 displays a typical dose-response curve corresponding to a unique bio-molecular interaction. The dose-response curve was constructed from the maximal fusion levels observed for varying PRRT2 concentrations. Dotted line shows a single exponential decay fit, which estimates the half-maximal response at a 1:1 PRRT2/t-SNARE ratio. Average values and SDs from a minimum of three independent experiments are shown. (G) Schematic of the single-vesicle fusion analysis. Fluorescent-labeled lipid (ATTO647-PE) included in v-SNARE-containing liposomes (v-SUV) enables us to track the association and fusion of single vesicles with the free-standing planar bilayer using confocal microscopy. Typically, the vesicle appears in the field of view when it loosely tethers to the bilayer (left), progresses to firmly dock concomitant with an increase in the fluorescence signal (middle), and then proceeds to fuse (right), evidenced by the radial diffusion of the fluorescent lipids in the bilayer after transfer due to membrane fusion. Our automated software enables the tracking of individual vesicles and classify the different stages, leading to fusion. (H) PRRT2 reduces the fraction of vesicles that proceed to fuse in the single-vesicle analysis. In the absence of PRRT2, ∼60% of all observed vesicles proceed to full fusion. PRRT2 introduced in the v-SUVs significantly reduces (∼50%) the fraction of fused vesicles. Average values and SDs from a minimum of three independent experiments are shown.

Figure 2

PRRT2 Blocks Ca²⁺-Regulated Exocytosis under Both Reconstituted and Physiological Conditions (A) Reconstituted fusion assay with synaptotagmin in the v-SNARE vesicles shows that PRRT2 also blocks synaptotagmin-regulated fusion, and the block is not altered by addition of Ca²⁺. Representative fusion curves are shown. (B) Direct comparison of the PRRT2 fusion block under both synaptotagmin-free and synaptotagmin ± Ca²⁺ conditions shows the extent of inhibition (one copy of PRRT2 per t-SNARE) is comparable under all conditions tested, suggesting that SNARE proteins are the primary target for PRRT2. Average values and SDs from three or four independent experiments are shown. (C) Western blot analysis shows the PC12 cells lack endogenous PRRT2, thus providing a virgin environment to test the functional capabilities of PRRT2 under physiologically relevant conditions. (D) TIRF microscopy setup used to observe single exocytotic events in PC12 cells transfected with VAMP2-pHluorin (±PRRT2), stimulated by local perfusion of KCl supplemented with 2.5 mM Ca²⁺. Live cell imaging analysis using TIRF microscopy shows that PRRT2 inhibits evoked fusion. It blocks both partial and full-fusion events, as characterized by their distinctive fluorescence signatures (Figure S2B). Average values and SDs from three or four independent experiments, with a minimum of 50 cells under each conditions, are shown. Statistical significance was established using a Wilcoxon-Mann-Whitney test to account for the non-normal distribution of the data.

Figure 3

PRRT2 Impedes SNARE-Dependent Docking/Priming of Vesicles (A) Tethering/docking of the vesicles surveyed by neutralizing the pH in all vesicles with 50 mM ammonium chloride under TIRF conditions shows that PRRT2 expression does not substantially alter the number of vesicles at or near the PM. Average values and SDs from three independent experiments with a minimum of 25 cells per condition are shown. (B) EM analysis of serial sections of PC12 cells with (left bottom) and without (left top) PRRT2 expression confirms that PRRT2 does not alter the density and distribution of the dense core vesicles. The proportional distribution of the dense core vesicles (dark spots) from the PM is shown. The average distribution from three independent experiments (∼25 cells in total) is shown. (C) TIRF microscopy-based analysis reveals that PRRT2 inhibits the docking of individual v-SNARE liposomes to a t-SNARE-containing planar supported bilayer. To get an accurate estimate of the docked vesicles, VAMP2 protein with mutations in the C-terminal half (L70D, A74R, A81D, and L84D, termed VAMP2-4X) that eliminates fusion activity was used, and the number of firmly docked vesicles was estimated after a 10 min incubation followed by an extensive buffer wash. The number of docked vesicles normalized to the PRRT2-free condition from three to five independent experiments is shown. The error bars indicate the SEM. (D and E) PRRT2 has a moderate effect on the membrane fusion process. (D) The single-vesicle fusion analysis showing that PRRT2 does not change the efficiency of the fusion process as the percentage of firmly docked vesicles that ultimately fuse is unaltered by the inclusion of PRRT2. However, PRRT2 introduces a slight delay in the fusion of the docked vesicles. The percentage survival curve of docked vesicles (E) reveals that the vesicles containing PRRT2 on an average take longer to fuse compared with control vesicles. Average and SEM from five independent single-vesicle fusion analyses are shown.

Figure 4

The N-Terminal Proline-Rich Domain of PRRT2 Blocks SNARE Complex Assembly (A) Assembly of the SNARE complex followed using FRET between Oregon green-labeled t-SNARE and Texas red-labeled VAMP2 introduced in the N terminus (SNAP25 residue 20 and VAMP2 residue 28) shows that PRRT2 blocks the initial engagement of the SNARE complex and does so in a concentration-dependent manner. PRRT2/t-SNARE ratios of 1:1 (red), 2:1 (green), and 3:1 (magenta) are shown. Excess soluble VAMP2 (black) was used as a negative control. (B) Consistent with the polarized assembly of the SNARE proteins, a similar level of inhibition was observed for different PRRT2/t-SNARE ratios (same color scheme as in A) for the labels introduced in the C terminus (SNAP25 residue 193 and VAMP2 residue 75). Averages and SDs from four independent experiments are shown. (C and D) PRRT2 contains a large, intracellular N-terminal proline-rich domain (orange), followed by a membrane associated region (light purple) and a trans-membrane domain (dark purple), connected by a short, flexible loop (green) (C). In situ removal of the PRRT2 N-terminal domain using an engineered TEV protease site (denoted by the arrow) (PRRT2^TEV) as confirmed by SDS-PAGE analysis results in complete loss of PRRT2 function in the lipid-mixing fusion assay (D), indicating that the proline-rich region is the effector domain. The fusion curves for PRRT2 and PRRT2^TEV with (open) or without (filled) TEV protease treatment are shown. Representative fusion curves and average values and SDs from three or four independent experiments are shown.

Figure 5

PKD-Associated Mutations in PRRT2 Disrupt Its SNARE-Modulatory Function (A) List of PKD-associated mutations in PRRT2 identified by Gardiner et al. (2015). The key mutations (highlighted in red) that did not affect the protein levels and exhibit no non-sense-mediated mRNA decay were tested in this study. (B) Lipid-mixing assay showing that the PKD mutations G305W (magenta) and X341L (green) disrupt the SNARE-inhibitory function of PRRT2 (red) to different extents. In comparison, the non-pathogenic P216H (yellow) mutation has no effect on PRRT2 function. Negative control with soluble cytoplasmic domain of VAMP2 (CDV) added in excess to titrate out the t-SNAREs is also shown. Representative fusion curves are shown. (C) Dose analysis shows that the loss of function of G305W mutation (red) in the reconstituted fusion assay is complete and not reversed at a higher dosage of PRRT2. In contrast, the wild-type (blue) PRRT2 shows a typical dose curve. Averages and SDs for three independent trials are shown. (D) Consistently, the single-vesicle exocytosis assay in PC12 cells shows that the G305W mutation abrogates the SNARE-modulatory function of PRRT2 under physiological conditions. Average values and SDs from a minimum of three independent experiments are shown. (E) Single-vesicle docking analysis showing that the PKD-associated G305W mutation introduced in the same molar ratio as the wild-type in the v-SNARE liposomes (Figure S5A) does not impede the SNARE-dependent docking/priming of vesicles to the supported bilayers. Average values and SEMs of three to five independent experiments are shown. (F) FRET-based SNARE assembly assay reveals that the loss of function phenotype results from the inability of G305W mutant (red) to inhibit the SNARE engagement, in contrast to the wild-type (blue) PRRT2. Average values and SDs from three or four independent experiments are shown.