Structural basis of synaptic vesicle assembly promoted by α-synuclein

Abstract

α-synuclein (αS) is an intrinsically disordered protein whose fibrillar aggregates are the major constituents of Lewy bodies in Parkinson's disease. Although the specific function of αS is still unclear, a general consensus is forming that it has a key role in regulating the process of neurotransmitter release, which is associated with the mediation of synaptic vesicle interactions and assembly. Here we report the analysis of wild-type αS and two mutational variants linked to familial Parkinson's disease to describe the structural basis of a molecular mechanism enabling αS to induce the clustering of synaptic vesicles. We provide support for this 'double-anchor' mechanism by rationally designing and experimentally testing a further mutational variant of αS engineered to promote stronger interactions between synaptic vesicles. Our results characterize the nature of the active conformations of αS that mediate the clustering of synaptic vesicles, and indicate their relevance in both functional and pathological contexts.

Figure 1. CEST experiments probing the membrane interactions of αS_A30P and αS_E46K.

CEST experiments were recorded at a ¹H frequency of 700 MHz (see Methods section), using protein concentrations of 300 μM and 0.06% (0.6 mg ml⁻¹) of DOPE:DOPS:DOPC lipids in a ratio of 5:3:2 and assembled into SUVs. ¹H–¹⁵N HSQC spectra were recorded by using continuous wave saturation (170 Hz or 350 Hz) in the ¹⁵N channel at offsets ranging between −28 kHz and +28 kHz; an additional spectrum, saturated at −100 kHz, was recorded as a reference. Data recorded using a saturation bandwidth of 350 Hz are shown here (the data measured using a saturation bandwidth of 170 Hz are shown in Supplementary Fig. 2). For comparison, the plots in panels b and e are drawn using αS_WT data from our previous investigation. (a–c) CEST surfaces for αS_A30P (a) αS_WT (b) and αS_E46K (c). (d–f) CEST saturation along the sequences of αS_A30P (d), αS_WT (e) and αS_E46K (f). The green lines refer to the averaged CEST profiles measured using offsets at +/− 1.5 kHz, and the profiles for +/− 3 kHz and +/− 5 kHz are shown in black and red, respectively. (g–i) Schematic illustration (see Materials and Methods) of the equilibrium between surface attached/detached local conformations in the membrane-bound states αS_A30P (g) αS_WT (h) and αS_E46K (i). The major differences in the data of αS_A30P, αS_WT and αS_E46K are located in the anchor region. Overall, these three variants of αS maintain the same topological properties at the surfaces of synaptic-like SUVs.

Figure 2. MAS ssNMR spectra of αS_A30P and αS_E46K bound to SUVs.

(a) ¹³C–¹³C DARR correlation spectra (aliphatic regions) recorded at −19 °C using a 50 ms mixing time at a MAS rate of 10 kHz. We used a 1:65 protein:lipid ratio in both cases, and spectra of αS_A30P and αS_E46K are shown in the left and right panels, respectively. Residues are indicated using the single letter convention. The highest signal intensities in the spectra of the samples studied here were obtained by performing the measurements at −19 °C. Under these conditions the lipid mixtures used here are in the gel phase, enabling ¹³C–¹³C DARR spectra to be measured with significantly increased signal-to-noise ratios but without affecting the pattern of chemical shifts; the latter are consistent with those measured at 4 °C (ref. 28). No variations in the number of observed resonances or in the chemical shifts were observed using protein:lipid ratios ranging from 1:30 to 1:200 (ref. 28). (b) ¹H–¹³C correlation via INEPT transfer recorded at 4 °C at a MAS rate of 10 kHz. The experiments were performed at a ¹H frequency of 700 MHz using a 3.2 mm E^Free probe. Atom names ca, cb, cg, cd and ce are used for C^α, C^β, C^γ, C^δ and C^ɛ atoms, respectively.

Figure 3. Vesicle assembly induced by αS.

(a) Molecular details of the double-anchor mechanism described in this work. SUVs of 50 nm in diameter were modelled to mimic as closely as possible the experimental conditions in this study (see Methods section). αS was modelled with the N-terminal anchor in an amphipathic α-helical conformation (red) and bound to the lower vesicle. The region 65–97 (cyan) of αS was modelled in an amphipathic α-helical conformation bound to the upper vesicle. The C-terminal fragment (residues 98–140) and the linker region 26–59 are shown in pink and grey colours, respectively. With this topology the modelling reveals that a single αS molecule could simultaneously bind two vesicles that are up to 150 Å apart. (b,c) Cryo-EM (b) and STED (c) images acquired on SUVs at a concentration of 0.5 mg ml⁻¹. (d,e) Cryo-EM (d) and STED (e) images measured on SUVs following a 12 h incubation with 200 μM αS_WT. (f,g) Cryo-EM (f) and STED (g) images acquired on SUVs following a 12 h incubation with 200 μM αS_Sw.

Figure 4. Stepwise representation of SUV interactions and fusion promoted by αS.

The scheme shows the stepwise mechanism of vesicles assembly as probed from images obtained in vitro by cryo-EM, which are also shown. Disordered cytoplasmatic αS (red) binds dynamically to the surface of SUVs (green), as described in this study. SUVs coated with αS assemble with fast kinetics as a consequence of the double-anchor mechanism promoted by the αS molecules decorating their surfaces. The tethered SUVs, which are initially assembled together in dimeric, trimeric, tetrameric and higher order states, eventually fuse to form larger vesicles. With the increasing size of the fused vesicle, we observed preferential fusion events at the termini of the aggregated vesicles. This observation can be explained by the higher affinity of αS for significantly curved membrane surfaces, which increases the concentration of bound αS at the termini of the elongated vesicles thereby promoting a stronger double-anchor mechanism in these loci.

Figure 5. Clustering of synaptic vesicles promoted by αS.

SVs purified from rat brain were incubated for 48 h at 37 °C. The concentrations during the incubation were 0.5 mg ml⁻¹ and 85 μM for the SVs and the αS variants, respectively. (a–c) dSTORM imaging of SVs alone (a) and SVs incubated with αS_WT (b) and with αS_Sw (c). The images were collected using a previously described protocol. Scale bars, 1 μm. To generate fluorescent SVs, we used a primary antibody that is specific for synaptotagmin 1 and a secondary antibody that is covalently linked to an ATTO 647 N dye. 10,000 fluorescence frames with an exposure time of 10 ms were recorded. The field of view imaged covered 1,997 × 1,997 camera pixels, corresponding to an area on the sample of ∼20 × 20 μm². (d) To assess the level of clustering of the SVs, we adapted an approach that has previously been successfully employed to analyse protein self-assembly. For each dSTORM image, clusters of SVs were identified on the basis of the distances between the centres of mass of the SVs. In particular two or more vesicles were associated with a specific cluster if their distances apart are less than 60 nm. The distribution of SVs in clusters of different sizes is reported using orange, green and blue histograms for SVs, SVs in the presence of αS_WT and SVs in the presence of αS_Sw, respectively. Cryo-EM images (scale bar, 50 nm) show representative clusters of different size.