Microglia affect α-synuclein cell-to-cell transfer in a mouse model of Parkinson's disease

Abstract

Background: Cell-to-cell propagation of α-synuclein (α-syn) aggregates is thought to contribute to the pathogenesis of Parkinson's disease (PD) and underlie the spread of α-syn neuropathology. Increased pro-inflammatory cytokine levels and activated microglia are present in PD and activated microglia can promote α-syn aggregation. However, it is unclear how microglia influence α-syn cell-to-cell transfer.

Methods: We developed a clinically relevant mouse model to monitor α-syn prion-like propagation between cells; we transplanted wild-type mouse embryonic midbrain neurons into a mouse striatum overexpressing human α-syn (huα-syn) following adeno-associated viral injection into the substantia nigra. In this system, we depleted or activated microglial cells and determined the effects on the transfer of huα-syn from host nigrostriatal neurons into the implanted dopaminergic neurons, using the presence of huα-syn within the grafted cells as a readout.

Results: First, we compared α-syn cell-to-cell transfer between host mice with a normal number of microglia to mice in which we had pharmacologically ablated 80% of the microglia from the grafted striatum. With fewer host microglia, we observed increased accumulation of huα-syn in grafted dopaminergic neurons. Second, we assessed the transfer of α-syn into grafted neurons in the context of microglia activated by one of two stimuli, lipopolysaccharide (LPS) or interleukin-4 (IL-4). LPS exposure led to a strong activation of microglial cells (as determined by microglia morphology, cytokine production and an upregulation in genes involved in the inflammatory response in the LPS-injected mice by RNA sequencing analysis). LPS-injected mice had significantly higher amounts of huα-syn in grafted neurons. In contrast, injection of IL-4 did not change the proportion of grafted dopamine neurons that contained huα-syn relative to controls. As expected, RNA sequencing analysis on striatal tissue revealed differential gene expression between LPS and IL-4-injected mice; with the genes upregulated in tissue from mice injected with LPS including several of those involved in an inflammatory response.

Conclusions: The absence or the hyperstimulation of microglia affected α-syn transfer in the brain. Our results suggest that under resting, non-inflammatory conditions, microglia modulate the transfer of α-syn. Pharmacological regulation of neuroinflammation could represent a future avenue for limiting the spread of PD neuropathology.

Keywords: Alpha-synuclein; Interleukin-4; Lipopolysaccharide; Microglia; Parkinson’s disease; Prion-like.

Fig. 1

Elimination of microglia using a CSF1R inhibitor in a model of α-syn cell-to-cell transfer. a Experimental time line for the α-syn cell-to-cell transfer model. b Huα-syn expression in the SN and striatum by immunohistochemistry following unilateral stereotaxic injection in the SN. Surviving grafted neurons expressing TH within the striatum. The asterisk is in the center of the graft; SN; substantia nigra, Str; striatum, scale bars: 500 μm

Fig. 2

PLX treatment significantly decreases microglia numbers in the mouse striatum. a Microglia marker (Iba-1) protein as determined by western blotting shows significant depletion in striatal brain homogenates; n = 6 control, n = 8 PLX. b The density of microglia was reduced in striatal brain tissue stained for Iba-1. Quantification is shown above the images; n = 8 control, n = 8 PLX. ** = p < 0.01, *** = p < 0.001, **** = p < 0.0001. The error bars represent S.E.M. c Representative images of microglia through the striatum that contained the neural graft used for quantifying microglia density. The insert images are higher magnification taken from the same brain section; high magnification image scale bars = 500 μm; low magnification image scale bars = 50 μm. d Representative images of microglia morphology from the contralateral striatum and ipsilateral striatum. Quantification of microglia morphology (area/perimeter) is shown below the images; n = 10 control, n = 10 PLX. ** = p < 0.01, *** = p < 0.001, **** = p < 0.0001. The error bars represent S.E.M.

Fig. 3

Fewer microglia following PLX treatment increased α-syn cell-to-cell transfer within the graft. a Confocal orthogonal reconstructions of grafted mouse TH-positive neurons (red) displaying intracellular puncta of transferred huα-syn (white); DAPI stain (blue) indicates nuclei. White box outlines cell of interest containing huα-syn puncta. The percentage of TH-positive neurons containing huα-syn. Control (circles) n = 18, PLX (squares) n = 10. * = p < 0.05. The error bars represent S.E.M. b Confocal 3D reconstructions of Iba-1 positive microglia (green), grafted mouse TH-positive neurons (red) displaying intracellular puncta of transferred huα-syn (white); DAPI stain (blue) indicates nuclei. White box outlines cell of interest containing huα-syn puncta. White arrows indicate huα-syn punctae. The percentage of Iba-1-positive microglia containing huα-syn. Control (circles) n = 18, PLX (squares) n = 10. The error bars represent S.E.M.

Fig. 4

Altering the microglia activation state using a pro-inflammatory versus anti-inflammatory environment. a Experimental time line for the α-syn cell-to-cell transfer model. b Grafted neurons positive for TH survived all three experimental conditions (left column of images; scale bars, 100 μm). The contrast in microglial morphology using immunohistochemistry for Iba-1 between the contralateral and ipsilateral striatum (right two column of images). The inset images are of the ipsilateral striatum indicating microgliosis around the graft and where the larger images are taken from; scale bars, 10 μm and 500 μm c Ipsilateral microglial morphology as assessed by the area:perimeter index (hydraulic radius). The horizontal line represents the mean area:perimeter index score of microglia from the contralateral striatum. Microglia in the LPS group had significantly higher area:perimeter index than the control. Area:perimeter index was significantly different between control and IL-4 (control, n = 20; LPS, n = 20; IL-4, n = 23; p < 0.05). d There was no significant difference in the density of microglia from the ipsilateral striatum between control, LPS-, or IL-4-injected mice. e Inflammatory cytokines measured using the MesoScale pro-inflammatory panel 1 assay in graft tissue and f in striatal tissue; control, n = 10; LPS, n = 8–12; IL-4, n = 12. * = p < 0.05, ** = p < 0.01, *** = p < 0.001. The error bars represent S.E.M.

Fig. 5

RNA seq analysis of pro-vs. anti-inflammatory environment in the model of αsyn cell-to-cell transfer. a MDS plot maintains distance between samples based on normalized CPM expression data for all mapped genes. b Volcano plot of RNA-Seq data where –log10(p-value) is plotted against the log2 fold change expression difference between LPS and IL-4 treatments for every mapped gene. The horizontal dotted red line corresponds to a FDR of 0.05 and the vertical dotted lines correspond to expression changes of 2-fold. Color corresponds to the median gene expression (CPM) in the PBS treatment group. c The top 50 signature genes were given for each of 10 cell types in Gokce et al. 2016. The ratio of genes for each cell type that are also significantly different (IL-4 vs. LPS: FDR < 0.05) are plotted. d Hierarchical clustering (eucledian distance, average linkage) of the RNA-seq results FPKM for the top 50 signature microglial genes with ratio of median LPS/median IL-4 expression. e Hypothesis driven gene ontology enrichment. A subset of biological process gene categories was selected. The enrichment significance measuring the overlap between gene members in these categories and significantly DE expressed genes is plotted. CPM; count per million, FDR; false discovery rate, FPKM; Fragments Per Kilobase Million, MSD; Multidimensional scaling

Fig. 6

Increased huα-syn in the striatum of LPS-injected mice. a Western blots of striatal samples probed for huα-syn, pS129 α-syn, LC3B-1, LAMP1 and p62 in control, LPS-injected, and IL-4-injected mice. b Quantification of huα-syn, LC3B-1, and LAMP1 normalized to β-actin levels within the contralateral and ipsilateral Str of control, LPS-, and IL-4-injected mice. Pon S stain for loading reference. * = p < 0.05; error bars represent S.E.M. Pon S; Ponceau S, Str; striatum

Fig. 7

Cell-to-cell transfer of hu-αsyn into Iba-1-positive and TH-positive cells following microglia activation. a Orthogonal confocal reconstructions of microglial cells (Iba-1, green) containing intracellular puncta of hu-αsyn (red); nuclei are stained blue. White box outlines cell of interest containing huα-syn puncta. The percentage of Iba-1-positive microglia containing hu-αsyn (control, n = 20; LPS, n = 20; IL-4, n = 23). * = p < 0.05; error bars represent S.E.M. b Orthogonal confocal 3D reconstructions of TH-positive dopaminergic neurons (green) containing intracellular puncta of hu-αsyn (red); nuclei are stained blue. White box outlines cell of interest containing huα-syn puncta. White arrows indicate huα-syn punctae. The percentage of TH-positive neurons containing hu-αsyn (control, n = 20; LPS, n = 20; IL-4, n = 23). * = p < 0.05; error bars represent S.E.M.