Super-resolution imaging reveals α-synuclein seeded aggregation in SH-SY5Y cells

Abstract

Aggregation of α-synuclein (α-syn) is closely linked to Parkinson's disease (PD) and the related synucleinopathies. Aggregates spread through the brain during the progression of PD, but the mechanism by which this occurs is still not known. One possibility is a self-propagating, templated-seeding mechanism, but this cannot be established without quantitative information about the efficiencies and rates of the key steps in the cellular process. To address this issue, we imaged the uptake and seeding of unlabeled exogenous α-syn fibrils by SH-SY5Y cells and the resulting secreted aggregates, using super-resolution microscopy. Externally-applied fibrils very inefficiently induced self-assembly of endogenous α-syn in a process accelerated by the proteasome. Seeding resulted in the increased secretion of nanoscopic aggregates (mean 35 nm diameter), of both α-syn and Aβ. Our results suggest that cells respond to seed-induced disruption of protein homeostasis predominantly by secreting nanoscopic aggregates; this mechanism may therefore be an important protective response by cells to protein aggregation.

Fig. 1. Super-resolving the replication of intracellular α-syn aggregates inside cells.

a Schematic of antibody-DNA-PAINT imaging. Cells at defined time points were immunostained with MJFR-14-6-4-2 conjugated with the docking strand (DS). The complementary imaging strand (IS), linked to a fluorescent Cy3b dye, interacts with the docking strand stochastically at a nanomolar concentration and hence can “paint” the boundary of the given object. The two microscopic images show that super-resolved (SR) images can distinguish two aggregated particles compared with conventional diffraction-limited (DL, thioflavin T-stained) images. b Growth of intracellular α-syn aggregates over time. SH-SY5Y cells are incubated with a mixture of PFF seeds (final 2.5 µM monomer equivalent of sonicated α-syn aggregates) and Bioporter for 4 h, then rinsed with PBS and cultured with fresh DMEM/10% FBS (time = 0 h after rinsing). Nuclei are stained blue with Hoechst dye; super-resolved α-syn aggregates are shown in red. c, d Intracellular aggregates increase in number and length after seed transduction. Statistical difference was determined by one-way ANOVA with Tukey’s multiple comparison test. N (total number of cells imaged) = 30, 29, 29, 29, 30 for 0, 4, 24, 48, 72 h, respectively. The cells were separately pooled for each biological replicate (n = 3) and the median calculated for each pool; those three medians were then used to calculate the mean (horizontal bars), standard deviation (error bars), and p values. *p < 0.05, **p < 0.01, ***p < 0.001, n.s. not significant; the difference of the means is not significant at the 0.05 level. e Kinetic analysis of the formation of intracellular aggregates; the median values from c are fitted to a minimal model of replication (see “Methods” for details). The best fit for the replication rate within cells is 0.22 h⁻¹, with 95% credible interval CI between 0.14 and 0.41. f The percentage of cells containing >50 aggregates grows over time, while the percentage of cells containing <10 aggregates falls.

Fig. 2. The protein clearance machinery shapes α-syn aggregation in SH-SY5Y cells.

After seeding with PFFs, SH-SY5Y cells were incubated DMEM/10% FBS supplemented with proteasomal and/or lysosomal modulators for 24 h, then α-syn aggregates detected with antibody-DNA-PAINT. BTZ: Bortezomib (5 µM); CFZ: Carfilzomib (1 µM); MG132 (1 µM); 3-MA: 3-Methyladenine (2.5 mM); BafA1: Bafilomycin A1 (0.4 µM); rapamycin (60 nM). a Proteasomal inhibition reduces α-syn aggregation in SH-SY5Y cells. b Autophagy-lysosomal pathway does not remove α-syn efficiently in SH-SY5Y cells. Nuclei are stained blue with Hoechst dye; super-resolved α-syn aggregates are shown in red. Inhibiting the proteasome decreased the number of aggregates per cell (c) but not the aggregate length (d). In contrast, inhibiting autophagy did not affect the number of α-syn aggregates (e) but stimulating autophagy with rapamycin restored protein clearance; neither treatment affected aggregate length (f). Statistical difference was determined by one-way ANOVA with Tukey’s multiple comparison test. The cells were separately pooled for each biological replicate (n = 3) and the median calculated for each pool; those three medians were then used to calculate the mean (horizontal bars), standard deviation (error bars), and p values. *p < 0.05, **p < 0.01, ***p < 0.001, n.s. not significant; the difference of the means is not significant at the 0.05 level. For proteasomal inhibition, N = 30, 29, 29, 30, 28, 27 for Ctrl 24 h, Seeded 24 h, +BTZ, +MG132, +CFZ, +MG132/BafA1, respectively, from three biological replicates. For autophagy inhibition/induction, the total number of cells imaged N = 21, 24, 24, 22, 23 for Ctrl 24 h, Seeded 24 h, +3-MA, +BafA1, +Rapamycin, respectively, from three biological replicates. Scale bars in a and b (from top to bottom): 10, 10, and 1 µm.

Fig. 3. Seeded cells secrete small aggregates.

Culture medium was collected at defined time points after seeding and imaged with AD-PAINT. a Super-resolution images of extracellular aggregates. Secreted aggregates accumulate over time following seeding (b), while their length (c) and eccentricity (d) increase slightly. Statistical difference was determined by two-way ANOVA with Tukey’s multiple comparison test. *p < 0.05, **p < 0.01, ***p < 0.001. Mean ± SD from three biological replicates. Scale bars in a: 500 nm (main panels) and 25 nm (insets).

Fig. 4. Seeded cells secrete high levels of Aβ and α-syn aggregates.

Immunodepletion of Aβ and α-syn from culture medium was carried out using 6E10 or Syn211 antibody. The immunodepleted media were then imaged with AD-PAINT. a Number of aggregates per µm² after immunodepletion. b Percentage of aggregates >75 nm in the detected population. c Numbers of aggregates removed by depletion with antibodies against Aβ and α-syn at defined time points. The numbers were calculated using the mean values of medium only controls and immunodepleted media as shown in a. Secreted aggregate of both Aβ and α-syn accumulate over time.

Fig. 5. Schematic summary of α-syn seeded aggregation in SH-SY5Y cells.

The schematic summarizes the cellular responses after α-syn seeds are internalized measured in this study. In a, the cells can only be seeded by the fibrillar form of α-syn aggregates after sonication, not by monomeric form, in an inefficient way (shown as low seeding probability). The α-syn seeds that escape from autophagy-dependent degradation effectively replicate in cells by recruiting endogenous α-syn monomers. Proteasomes, which are responsible for protein degradation and quality control in cells, facilitate the replication process by accelerating fibril fragmentation. The resulting α-syn aggregates are secreted by the cells and accumulate in the cytosol. b In contrast under basal conditions, aggregates that form in the cytosol are either degraded or secreted by the cell; hence there is no accumulation of aggregates.