Excess phosphoserine-129 α-synuclein induces synaptic vesicle trafficking and declustering defects at a vertebrate synapse

Abstract

α-Synuclein is a presynaptic protein that regulates synaptic vesicle (SV) trafficking. In Parkinson's disease (PD) and dementia with Lewy bodies (DLB), α-synuclein aberrantly accumulates throughout neurons, including at synapses. During neuronal activity, α-synuclein is reversibly phosphorylated at serine 129 (pS129). While pS129 comprises ∼4% of total α-synuclein under physiological conditions, it dramatically increases in PD and DLB brains. The impacts of excess pS129 on synaptic function are currently unknown. We show here that compared with wild-type (WT) α-synuclein, pS129 exhibits increased binding and oligomerization on synaptic membranes and enhanced vesicle "microclustering" in vitro. Moreover, when acutely injected into lamprey reticulospinal axons, excess pS129 α-synuclein robustly localized to synapses and disrupted SV trafficking in an activity-dependent manner, as assessed by ultrastructural analysis. Specifically, pS129 caused a declustering and dispersion of SVs away from the synaptic vicinity, leading to a significant loss of total synaptic membrane. Live imaging further revealed altered SV cycling, as well as microclusters of recently endocytosed SVs moving away from synapses. Thus, excess pS129 caused an activity-dependent inhibition of SV trafficking via altered vesicle clustering/reclustering. This work suggests that accumulation of pS129 at synapses in diseases like PD and DLB could have profound effects on SV dynamics.

FIGURE 1:

Characterization of pS129 α-synuclein. (A) NMR structure of human α-synuclein showing the pS129 modification. Ribbon model was generated in UCSF Chimera 1.16 using the UniProt PDB file: 1XQ8 of lipid micelle-bound α-synuclein (Ulmer et al., 2005). (B) Coomassie gel (left; 2 μg/lane) and Western blot (right; 0.2 μg/lane) on recombinant WT and phosphoserine129-modified (pS129) human α-synuclein. Both proteins ran on a 12% SDS–PAGE primarily in monomeric form at ∼17 kDa. Western blot for total α-synuclein was performed using a pan-synuclein antibody that recognizes a common epitope in the N-terminal domain. (C) Western blots using pS129-specific α-synuclein antibodies confirmed the identity and stability of pS129 α-synuclein. Each lane contained either 250 ng (left blot) or 500 ng (middle and right blots) of protein.

FIGURE 2:

WT and pS129 α-synuclein preferentially bind membrane lipids enriched at synapses. (A) Protein-lipid overlay showing that WT α-synuclein bound preferentially to PA, PS, PI, PI(4)P, PI(4,5)P₂, and PI(3,4,5)P₃, as well as mitochondrial CL. WT α-synuclein was detected with a mouse monoclonal antibody (Abcam; ab1903 [clone 4D6]) (B) Quantification showing the lipid binding distribution for WT α-synuclein. Bars indicate mean ± SEM from n = 4 experiments and are colored according to their enrichment in SVs, blue; PM, green; or mitochondria (Mito; magenta). (C and D) pS129 α-synuclein exhibited a similar lipid binding profile to WT α-synuclein. pS129 α-synuclein was detected with a rabbit monoclonal (D1R1R; Cell Signaling). Bars indicate mean ± SEM from n = 7 experiments. TG, triglyceride; DAG, diacylglycerol; PA, phosphatidic acid; PS, phosphatidylserine; PC, phosphatidylethanolamine; PG, phosphatidylglycerol; CL, cardiolipin; PI, phosphatidylinositol; PI(4)P, phosphatidylinositol 4-phosphate; PI(4,5)P₂, phosphatidylinositol 4,5-bisphosphate; PI(3,4,5)P₃, phosphatidylinositol 3,4,5-trisphosphate; Chol, cholesterol; SPH, sphingomyelin; SM4, 3-sulfogalactosylceramide.

FIGURE 3:

Enhanced binding and oligomerization of pS129 α-synuclein on synaptic membranes. (A) Isolated membranes from mouse brain synaptosomes were incubated with 2 μM WT or pS129 α-synuclein. After washing, any bound WT or pS129 was detected by Western blotting using a pan-synuclein antibody (Abcam 53726). While WT α-synuclein was readily recruited to synaptic membranes, pS129 showed enhanced binding and oligomerization into dimers and trimers. N-cadherin was used as a loading control for the membranes. (–) indicates membranes that were incubated with cytosol but not supplemented with WT or pS129. (B) Band intensity analysis revealed a 2.2-fold increase in pS129 recruitment to synaptic membranes, compared with WT α-synuclein. (C and D) Oligomerization of pS129 on synaptic membranes was also greater than that observed for WT α-synuclein. Bars in B-D indicate mean ± SEM from n = 9–10 experiments; ** indicates p < 0.005 and **** indicates p < 0.0001 by Student’s t test.

FIGURE 4:

pS129 enhances α-synuclein-induced SV “microclustering” in vitro. (A) Synapsin I (6 μM) robustly clustered SVs in vitro, as shown by the rapid, biphasic increase in turbidity at OD 405 nm (black reference curve). Addition of 2 μM WT or pS129 α-synuclein, maintaining the physiological 3:1 M ratio of synapsin 1: α-synuclein, had little impact on synapsin-SV condensation under these conditions. Data points represent mean ± SEM from n = 4 experiments. (B and C) Even when synapsin 1 and α-synuclein were in equimolar ratios (8 μM for each), α-synuclein had little impact on the rate of synapsin-SV condensation. (D) In the absence of synapsin, both 8 μM (n = 2) and 24 μM (n = 4) WT and pS129 α-synuclein induced moderate SV clustering in vitro, however pS129 produced greater, dose-dependent effects. Control curves show results with WT and pS129 α-synuclein protein alone, in the absence of SVs, demonstrating that the increased turbidity is due to SV microclustering. (E) Normalized data from panel D. Arrow indicates the 24-h time point from which samples were imaged (see [G–L]). (F) Working model illustrating how WT and pS129 α-synuclein could impact SV microclustering and synapsin-SV clustering (see also Hoffmann et al., 2021). (G–L) EM images (37,000x) showing SV microclustering induced by α-synuclein. At the beginning of the incubation period, purified SVs were generally monodispersed and appeared as single vesicles (H). After 24 h of incubation with either 8 or 24 μM WT α-synuclein or pS129 α-synuclein, small SV microclusters were visible (arrows in I–L). In both cases, pS129 enhanced SV microclustering. Scale in (G) applies to (H–L).

FIGURE 5:

pS129 and WT α-synuclein localize to SV clusters in lamprey axons. (A) Diagram showing the microinjection strategy for α-synuclein into lamprey giant RS axons. (A1) Inset shows a small region of a whole mounted lamprey spinal cord immunostained for the injected α-synuclein (pS129) and endogenous SV2, a marker of SV clusters. Dotted lines denote the borders of a single giant RS axon. (B) Characterization of the WT and pS129-specific α-synuclein antibodies used for IF experiments. Western blotting confirms the antibody is specific for human α-synuclein with little or no cross-reactivity against endogenous synuclein in rat brain or lamprey CNS. Antibodies used are as indicated. (C) Giant RS axons were injected with recombinant WT or pS129 α-synuclein and subsequently immunostained for α-synuclein (green) and SV2 (magenta). High resolution confocal images were obtained using a Zeiss LSM980 Axio Examiner with Airyscan2 (63X, 1.4 NA objective). Dotted lines indicate border of injected axon. White box in leftmost panel indicates the ROIs for the subsequent images. Arrows indicate SV clusters colabeled with α-synuclein and SV2. Solid line in the rightmost merged image indicates the synaptic position from which the intensity analysis was performed. (D) Histogram analysis of both raw (left) and normalized (right) fluorescence intensities showed a strong colocalization between injected WT α-synuclein and the SV2-positive SV clusters. (E) Peak analysis confirmed the strong colocalization between WT α-synuclein and SV2. (F and G) pS129 also colocalized to SV clusters. Data were obtained from n = 29-30 synapses, four to five axons per condition.

FIGURE 6:

Excess pS129 α-synuclein does not affect the SV clusters at resting lamprey synapses. (A) Electron micrograph of an unstimulated control lamprey synapse showing a large, densely-packed cluster of SVs. (B and C) After introducing low (5–8 μM) or high (10–20 μM) concentrations of pS129, there was no obvious effect on the SV clusters. Dotted lines indicate PM evaginations. Asterisk indicates postsynaptic density. Scale bar in A applies to B and C. (D–H) Quantification revealed no significant difference in the number of SVs, number and size of Cists, nor CCP/Vs (Clathrin Coats), at synapses treated with pS129. Only the PM evaginations were larger. (I) There was no change in the total synaptic membrane after treatment with pS129. Bars in D–I indicate mean ± SEM from n = 22–39 synapses, two axons/animals. “ns” indicates “not significant” and * indicates p < 0.05 by one-way ANOVA.

FIGURE 7:

Excess pS129 α-synuclein causes a severe depletion of SVs at stimulated lamprey synapses. (A–C) Electron micrographs showing a progressive loss of SVs after introducing pS129 at low (5–8 μM) or high (10–20 μM) concentrations to stimulated lamprey RS synapses (20 Hz, 5 min). pS129 also caused a moderate expansion of the PM (dotted lines) and build-up of atypical vesicular “Cist” (C), which are likely endosomes. Circles indicate CCP/Vs. Asterisks indicate postsynaptic density. Scale bar in A applies to B and C. (D–F) three-dimensional reconstructions of the synapses from panels A–C. Blue spheres = SVs; magenta ribbons = Cist; yellow and white spheres = CCPs and CCVs, respectively. PM is indicated by green ribbons, and AZ is shown as a red slab. The loss of SVs and build-up of Cist are especially prominent. Insets show the CCP/V distribution at these synapses. Red arrows show several SV microclusters that appear to be dispersed from the main SV cluster. (G–I) Morphometric analysis revealed that pS129 caused a depletion of SVs at synapses and that the remaining SVs had a slightly larger diameter. (J–L) The loss of SVs was partially compensated by an expansion of the PM, as well as an increase in the number and size (perimeter) of Cist. (M–N) pS129 did not alter the number or distribution of CCP/Vs, unlike our previous studies on WT α-synuclein (Busch et al., 2014; Medeiros et al., 2017; Banks et al., 2020). (O) A total membrane analysis revealed a significant loss of synaptic membrane, primarily due to SV depletion. Bars indicate mean ± SEM from n = 55–67 synapses, four axons/animals. * indicates p < 0.05, and “ns” indicates “not significant” by one-way ANOVA.

FIGURE 8:

Excess pS129 α-synuclein induces SV declustering and dispersion at lamprey synapses. (A–C) EM micrographs showing a control, stimulated synapse (20 Hz, 5 min), compared with those treated with either WT or pS129 α-synuclein (“high” concentration: 10–20 μM). While WT α-synuclein induced moderate SV declustering, this was dramatically enhanced in the presence of excess pS129 α-synuclein, resulting in much smaller SV clusters. (D) three-dimensional reconstruction and (Di–iii) serial micrographs of a pS129 α-synuclein-treated synapse showing the SV dispersion. Red arrows indicate small, SV microclusters or single SVs dispersed away from the synapse. (E) SV distribution analysis. pS129 α-synuclein caused a reduction in SVs at all distances measured, which was greater than that caused by WT α-synuclein at distances >150 nm from the AZ. (F–K) By all measures, the nearest neighbor analyses revealed that pS129 α-synuclein caused greater dispersion of SVs, compared with WT α-synuclein. Note that the average distances between SVs significantly increased with pS129, compared with WT and controls (F–G). Furthermore, the nearest neighbor SV distributions fit a two summed Gaussian curve, and compared with controls, WT and pS129 α-synuclein caused a pronounced tail (H). A log–log density plot of data points from (F) also appears more disperse, leading to an increase in the area and eccentricity of the elliptical fits (H–K). Data in (F–K) are from n = 3,016–12,143 SVs (55–67 synapses, two to four axons/condition). * indicates p ≤ 0.05, compared with control, by one-way ANOVA.

FIGURE 9:

Live imaging confirms that pS129 α-synuclein disrupts SV trafficking and clustering in vivo. (A) Lamprey giant axons were imaged using lattice light sheet microscopy (LLSM) during physiological recordings and microinjections. (B) Action potentials were evoked through the recording microelectrode. (C) Giant axon filled with Alexa dye (red). Synapses were labeled with FM 1-43 dye (green) via activity-dependent uptake. (D–F) Axon prelabeled with Alexa 642 hydrazide and FM 1-43, followed by pS129 α-synuclein (“high” concentration: 10–20 μM) coinjected with Alexa 568 hydrazide. Images show a single LLSM plane through the axon. (G) Another axon with FM 1-43 labeled synapses and injected with pS129. Image shows a 5 µm thick z-stack. (H) Action potentials were applied at 20 Hz (4000 stimuli), which caused highly variable FM destaining in the presence of pS129. Each trace represents FM destaining measured from a single SV cluster. (I–K) During stimulation FM 1-43 labeled SV microclusters (resolution limited points [Potcoava et al., 2021]) were mobile throughout the interior of the axon. Panel I show all mobile puncta in one time point z-stack, which was determined by expressing the frame as a ratio of the mean of 10 frames immediately prior. Panels J and K show an example of one of these SV microclusters leaving a larger synapse. (L) Average destaining kinetics reveal that WT and pS129 α-synuclein reduced the mean rate of destaining, as compared with control axons. Red bar indicates the stimulus (20 Hz, 4000 stimuli). Shaded areas show SEM. (M) pS129 caused a significant increase in the number of mobile FM1-43 labeled puncta within a 50-µm length of axon over a period of 200 s. ** indicates p < 0.01 by one-way ANOVA. Data in (L and M) are from n = 22-33 synapses, three to four axons per experimental condition.

FIGURE 10:

Working model illustrating how pS129 α-synuclein impacts synapses. (A) Diagram showing normal SV trafficking under physiological conditions. α-Synuclein participates in SV exocytosis, endocytosis, and SV clustering. Upon activity, some α-synuclein is pS129, which regulates synaptic transmission and plasticity (Ramalingam et al., 2023). Lamprey synapses are known to exhibit clathrin-mediated and bulk endocytosis for recycling SVs, though other nonclathrin mechanisms may also participate, which are not shown for simplicity. (B) When introduced acutely to synapses, excess pS129 α-synuclein causes a severe loss of SVs, which is likely due to aberrant clustering/declustering and/or altered postendocytic reclustering. SVs appear to break into small microclusters of SVs, which may be facilitated by the enhanced membrane binding and oligomerization of pS129. Such alterations of the SV clustering dynamics would allow SVs to float away from the synapse exacerbating synaptic transmission defects.