Lysosomal stress drives the release of pathogenic α-synuclein from macrophage lineage cells via the LRRK2-Rab10 pathway

Abstract

α-Synuclein and LRRK2 are associated with both familial and sporadic Parkinson's disease (PD), although the mechanistic link between these two proteins has remained elusive. Treating cells with lysosomotropic drugs causes the recruitment of LRRK2 and its substrate Rab10 onto overloaded lysosomes and induces extracellular release of lysosomal contents. Here we show that lysosomal overload elicits the release of insoluble α-synuclein from macrophages and microglia loaded with α-synuclein fibrils. This release occurred specifically in macrophage lineage cells, was dependent on the LRRK2-Rab10 pathway and involved exosomes. Also, the uptake of α-synuclein fibrils enhanced the LRRK2 phosphorylation of Rab10, which was accompanied by an increased recruitment of LRRK2 and Rab10 onto lysosomal surface. Our data collectively suggest that α-synuclein fibrils taken up in lysosomes activate the LRRK2-Rab10 pathway, which in turn upregulates the extracellular release of α-synuclein aggregates, leading to a vicious cycle that could enhance α-synuclein propagation in PD pathology.

Keywords: Cell biology; Molecular biology.

Graphical abstract

Figure 1

Lysosomal overload induces extracellular release of insoluble α-synuclein from macrophages and microglial cells (A) A flowchart of the experiment to detect insoluble α-synuclein and other soluble proteins in media and lysates from cells loaded with α-synuclein PFFs. sup: supernatant, ppt: precipitate. (B–G) The release of insoluble α-synuclein upon exposure to chloroquine (CQ), chlorpromazine (CP) or lidocaine (LC) in RAW264.7 cells (B-D) or MG6 cells (E-G) that were loaded with α-synuclein PFFs. α-Synuclein in ppt fractions, as well as cathepsin B (cat B) and α-tubulin in sup fractions, from media and lysates are shown. Bar graphs show densitometric analysis of α-synuclein in the ppt fraction of media (C, F) and mature cathepsin B in the supernatant fraction of media (D, G). Relative levels compared to PFF-only sample are shown. Data represent mean ± SEM, n = 4, one-way ANOVA with Tukey’s test. (H–J) The release of α-synuclein and mature cathepsin B from mouse primary microglia (H), and the lack of release from SH-SY5Y cells (I) or mouse primary neurons (J) that were loaded with PFFs and treated for 5 h with lysosomotropic agents or PBS (control). In I and J, the media from cells incubated with lysis buffer were loaded in the rightmost lane as a positive control. See also Figure S1.

Figure 2

The released α-synuclein has undergone prior intracellular uptake and delivery to lysosomes (A) Colocalization of internalized α-synuclein (red) and lysosomes (LAMP1, green) in RAW264.7 cells 12 h after the addition of α-synuclein PFFs. Nuclei were stained with DRAQ5 (blue). Arrowheads indicate α-synuclein-positive lysosomes. Bar = 10 μm. (B) Time-dependent increase in the colocalization of α-synuclein and LAMP1, as shown by percentage of α-synuclein-positive area in LAMP1-positive area. Mean ± SEM, n = 3 independent experiments. (C, D) Inhibitory effect of Dynasore (Dyna) and cytochalasin D (CytoD) on CQ-induced release of α-synuclein from RAW264.7 cells loaded with α-synuclein PFFs. Unaltered levels of released cathepsin B in media and α-tubulin in cell lysates are also shown. Quantification of relative levels of α-synuclein in ppt fractions of media is shown in D. Data represent mean ± SEM, n = 4, one-way ANOVA with Tukey’s test. (E, F) Inhibitory effect of Dynasore (Dyna) and cytochalasin D (CytoD) on the uptake of α-synuclein PFFs in RAW264.7 cells. Quantification of relative levels of α-synuclein in ppt fractions of cell lysates is shown in F. Data represent mean ± SEM, n = 4, one-way ANOVA with Tukey’s test. See also Figure S2.

Figure 3

Exosomes are involved in the release of insoluble and seed-competent α-synuclein (A) Confirmation of insolubility of the α-synuclein released from RAW264.7 cells loaded with α-synuclein PFFs and exposed to the indicated lysosomotropic agents. The collected media were treated with Triton X-100 to solubilize exosomal structures and then ultracentrifuged. α-Synuclein, but not an exosomal marker Alix, remained detectable in the precipitate fraction after Triton X-100 treatment. (B) Left: schematic diagram of ultracentrifugation-free exosome purification method using magnetic beads. Upper right: detection of α-synuclein and Alix in exosomal and flow-through fractions of media from CQ-treated RAW264.7 cells. Triton X-100 treatment prior to exosome purification resulted in the shift of α-synuclein from the exosome to flow-through fractions. Lower right: quantification of α-synuclein in the exosome fraction. Mean ± SEM, n = 3, one-way ANOVA with Tukey’s test. (C) Inhibitory effect of an exosome inhibitor GW4869 on the release of insoluble α-synuclein, Alix and cathepsin B from RAW264.7 cells upon CQ treatment. (D, E) Quantification of the release of insoluble α-synuclein and mature cathepsin B, as shown in C. Mean ± SEM, n = 4, one-way ANOVA with Tukey’s test. (F) FRET fluorescence of α-synuclein biosensor cells treated with the media from RAW264.7 cells that had been treated with PFF or PFF+CQ. Bar = 10 μm. (G) Quantification of the fluorescent dots in biosensor cells treated with the indicated media, as shown in F. Mean ± SEM, n = 4, one-way ANOVA with Tukey’s test. See also Figure S3.

Figure 4

LRRK2 and Rab10 mediate the release of insoluble α-synuclein under lysosomal stress (A–F) Suppression of CQ-induced release of α-synuclein and mature cathepsin B by knockdown of LRRK2 or Rab10. The media and lysates of RAW264.7 cells (A-C) or MG6 cells (D-F) that were pretreated with the indicated siRNAs and then treated with PFF and CQ were analyzed by immunoblotting. Quantification of α-synuclein in insoluble fractions (B, E) and mature cathepsin B in soluble fractions (C, F) from media of RAW264.7 cells (B, C) or MG6 cells (E, F) are shown. Data represent relative levels compared to CQ (−) sample. Mean ± SEM, n = 3, one-way ANOVA with Tukey’s test. (G–L) Immunoblot analysis of media and lysates of PFF-treated RAW264.7 cells (G-I) or MG6 cells (J-L) that were treated with LRRK2 kinase inhibitors (MLi-2, GSK) and CQ. Quantification of insoluble α-synuclein (H, K) and mature cathepsin B (I, L) from media of RAW264.7 cells (H, I) or MG6 cells (K, L) are shown. Data represent relative levels compared to CQ (−) sample. Mean ± SEM, n = 3, one-way ANOVA with Tukey’s test. See also Figure S4.

Figure 5

Internalization of α-synuclein PFFs induces the LRRK2 phosphorylation of Rab10 (A–F) Immunoblot analysis of Rab10 phosphorylation in RAW264.7 cells (A-C) or MG6 cells (D-F) treated with α-synuclein PFFs, monomers, PBS (negative control) or CQ (positive control). Densitometric analysis of phospho-Rab10 divided by total Rab10 in RAW264.7 (B) or MG6 (E) cell lysates, as well as mature cathepsin B in media from RAW264.7 (C) or MG6 (F) cells, are shown. Data represent relative values compared to PBS-treated sample. Mean ± SEM, n = 4, one-way ANOVA with Tukey’s test. (G–I) Time-dependent increase of Rab10 phosphorylation in RAW264.7 cells (G) or MG6 cells (H, I) after treatment with α-synuclein PFFs. Quantification of relative phospho-Rab10 levels in MG6 cells is shown in I. Mean ± SEM, n = 6, one-way ANOVA with Tukey’s test. (J–L) The increase of phospho-Rab10 in mouse embryonic fibroblasts (MEF) (J) or SH-SY5Y cells overexpressing LRRK2 (J, K) upon exposure to α-synuclein PFFs or CQ. Relative levels of phospho-Rab10 divided by total Rab10 in SH-SY5Y cells were shown in L. Mean ± SEM, n = 5, one-way ANOVA with Tukey’s test. (M, N) Inhibition of α-synuclein PFF-induced Rab10 phosphorylation in RAW264.7 cells by the co-treatment of PFF and Dynasore or cytochalasin D. Relative levels of phospho-Rab10 divided by total Rab10 were shown in N. Mean ± SEM, n = 6, one-way ANOVA with Tukey’s test. See also Figure S5.

Figure 6

Internalization of PFFs causes accumulation of LRRK2 and Rab8/Rab10 on lysosomal surface with close proximity (A–C) Confocal microscopic detection of proximity ligation (PL) signals (red) between LAMP1 (cytosolic tail) and LRRK2 (A), LRRK2 and Rab10 (B) or LRRK2 and Rab8 (C) in RAW264.7 cells exposed to α-synuclein PFFs for 12 h, or CQ for 3 h. Nuclei were stained with DRAQ5 (blue). Scale bars = 20 μm. (D–F) Counting of the particles of PL signals emitted between LAMP1 and LRRK2 (D), LRRK2 and Rab10 (E) or LRRK2 and Rab8 (F), as shown in A-C. The total number of particles in the field were divided by that of nuclei to calculate particle number per cell. Data represent mean ± SEM, n = 6, one-way ANOVA with Tukey’s test. See also Figures S6–S8.

Figure 7

A schematic model of reciprocal regulation between LRRK2 activation and insoluble α-synuclein release, with its dependence on cell types A diagram showing the lysosomal stress-mediated formation of a positive feedback loop between α-synuclein and LRRK2 that facilitates α-synuclein release. Internalization of α-synuclein fibrils causes lysosomal overload stress that enhances the LRRK2 phosphorylation of Rab10. The activated LRRK2-Rab10 pathway in turn mediates extracellular release of insoluble α-synuclein wrapped in exosomes. The α-synuclein-LRRK2 positive feedback loop may be induced in phagocytes but not in neuronal cells. In microglia, the internalized insoluble α-synuclein is rapidly re-released upon lysosomal stress, whereas in neurons, such insoluble α-synuclein remains in cells and may later form inclusions incorporating endogenous α-synuclein.