The interaction of tPA with NMDAR1 drives neuroinflammation and neurodegeneration in α-synuclein-mediated neurotoxicity

Abstract

The thrombolytic protease tissue plasminogen activator (tPA) is expressed in the CNS, where it regulates diverse functions including neuronal plasticity, neuroinflammation, and blood-brain-barrier integrity. However, its role in different brain regions such as the substantia nigra (SN) is largely unexplored. In this study, we characterize tPA expression, activity, and localization in the SN using a combination of retrograde tracing and β-galactosidase tPA reporter mice. We further investigate tPA's potential role in SN pathology in an α-synuclein mouse model of Parkinson's disease (PD). To characterize the mechanism of tPA action in α-synuclein-mediated pathology in the SN and to identify possible therapeutic pathways, we performed RNA-seq analysis of the SN and used multiple transgenic mouse models. These included tPA deficient mice and two newly developed transgenic mice, a knock-in model expressing endogenous levels of proteolytically inactive tPA (tPA Ala-KI) and a second model overexpressing proteolytically inactive tPA (tPA Ala-BAC). Our findings show that striatal GABAergic neurons send tPA⁺ projections to dopaminergic (DA)-neurons in the SN and that tPA is released from SN-derived synaptosomes upon stimulation. We also found that tPA levels in the SN increased following α-synuclein overexpression. Importantly, tPA deficiency protects DA-neurons from degeneration, prevents behavioral deficits, and reduces microglia activation and T-cell infiltration induced by α-synuclein overexpression. RNA-seq analysis indicates that tPA in the SN is required for the upregulation of genes involved in the innate and adaptive immune responses induced by α-synuclein overexpression. Overexpression of α-synuclein in tPA Ala-KI mice, expressing only proteolytically inactive tPA, confirms that tPA-mediated neuroinflammation and neurodegeneration is independent of its proteolytic activity. Moreover, overexpression of proteolytically inactive tPA in tPA Ala-BAC mice leads to increased neuroinflammation and neurodegeneration compared to mice expressing normal levels of tPA, suggesting a tPA dose response. Finally, treatment of mice with glunomab, a neutralizing antibody that selectively blocks tPA binding to the N-methyl-D-aspartate receptor-1 (NMDAR1) without affecting NMDAR1 ion channel function, identifies the tPA interaction with NMDAR1 as necessary for tPA-mediated neuroinflammation and neurodegeneration in response to α-synuclein-mediated neurotoxicity. Thus, our data identifies a novel pathway that promotes DA-neuron degeneration and suggests a potential therapeutic intervention for PD targeting the tPA-NMDAR1 interaction.

Keywords: Dopaminergic neurons; Glunomab; Microglia; NMDAR; Neuroinflammation; Parkinson’s disease; Substantia nigra; T-cell; tPA; α-synuclein.

Fig. 1

tPA is localized to presynaptic terminals associated with DA-neurons. (A) Western blot of tPA, TH, and actin in the SN and hippocampus of WT mice (n = 3–4). Actin was used as a loading control. (B) In-gel zymography showing tPA proteolytic activity in the SN and hippocampus of WT mice. Human tPA was used as a positive control (n = 3). (C) Coronal sections of the SN showing tPA (red) and TH (green) staining in WT mice. White squares indicate 60x closeup (right panels) and white dashed lines depict SNpc (n = 3). (D) Western blot of crude synaptosome protein extracts from SN of WT mice. Synaptosome and cytosol fractions were run and developed in the same gel and membrane (n = 3). (E) In-gel zymography showing tPA proteolytic activity, and (F) quantification of band intensity of supernatants from crude synaptosome preparations from the SN after 5 mM or 50 mM KCl treatment (n = 3–4). tPA-KO mice were used as a negative control in (A), (B), and (D). (G) Confocal images of the SN showing tPA (red), TH (white), and VAMP-2 (green). Merged image shows colocalization of tPA and VAMP-2 (yellow and white arrows). Panels next to merged image (xy) are orthogonal view indicating xz (bottom panel) and yz (right panel) colocalization (n = 3; MOC = 0.4 ± 0.04). Data are shown as mean ± SEM, *p < 0.05; two-tailed t-test. Scale bar = 500 μm (C), 20 μm (close up, C), 10 μm (G)

Fig. 2

DARPP-32 GABAergic neurons in the striatum send tPA⁺projections to the SN. (A, B) Retrograde tracing strategy using tPAβGAL reporter mice showing CTB-488 injection site (green) in the SN. (C) Representative coronal section images of the dorsomedial striatum showing β-Gal⁺ cells (black) and CTB-488⁺ neurons (green) in tPAβGAL^+/− mice 1 week after CTB-488 injection in the SN. White arrows indicate double-positive CTB-488 and β-Gal cells. Only the striatum showed CTB-488⁺ staining in neurons (average number of CTB-488⁺ in striatum per mm² = 272.3 ± 41.6; n = 3; green). (D) Coronal sections of the striatum showing DARPP-32 (red), TH (white), CTB-488 (green), and DAPI (blue) staining in tPAβGAL^+/− mice 1 week after CTB-488 injection in SN. White arrows indicate double-positive CTB-488 and DARPP-32 cells in the dorsomedial striatum (n = 4). (E) Confocal images of the SN showing tPA (red), the blood vessel marker podocalyxin (PODO; white) and VGAT (green). PODO staining was used to show that tPA in the SN is mainly associated with GABAergic presynaptic axons and not blood vessel (white open arrows). Merged image (xy) shows colocalization of tPA and VGAT (yellow and white arrows). Panels next to merged image (xy) show orthogonal view indicating xz (bottom panel) and yz (right panel) colocalization (n = 3; MOC = 0.39 ± 0.05). Scale bar= (B) 500 μm (C) 50 μm (D) 100 μm (E) 10 μm

Fig. 3

tPA-KO mice are protected from DA-neuron degeneration and behavioral deficits after rAAV2-hα-SYN injection. (A) Coronal sections of the SN showing tPA staining and (B) quantification of tPA fluorescence intensity in the SN 4 weeks after rAAV2-hα-SYN or rAAV2-control injection in WT mice (n = 8–9). (C) Representative images of the SN showing DA-neuron degeneration in the injected SN (TH, white), and (D) quantification of TH⁺ neurons in coronal sections of the SN 4 weeks after rAAV2-hα-SYN or rAAV2-control injection in WT and tPA-KO mice. (E) Quantification of sensorimotor bias in a corridor task 4 weeks after rAAV2-hα-SYN or rAAV2-control injection in WT and tPA-KO mice. White dashed lines represent the SNpc (n = 14–20). These experiments were conducted using 9- to 11-week-old WT and tPA KO mice. Data are shown as mean ± SEM, N.S = not significant, *p < 0.05, **p < 0.01; ****p < 0.0001. (B) 2-tailed t-test, (D, E) 1-way ANOVA followed by Tukey post hoc test. Scale bar = 500 μm

Fig. 4

Genes and proteins associated with innate immune response are upregulated by overexpression of hα-SYN in the SN of WT mice, but not in tPA-KO mice. (A) Heatmap showing combined differentially expressed genes in the SN between uninjected and injected SN for WT and tPA-KO mice 4 weeks after rAAV2-hα-SYN injection in the SN. Yellow rectangles indicate genes associated with innate and adaptive immune response downregulated in tPA-KO mice compared to WT mice 4 weeks after rAAV2-hα-SYN (n = 4). Heatmap was generated using the raw counts per millions of the selected genes. (B) q-PCR validation of selected genes of interest in the SN 4 weeks after rAAV2-hα-SYN injection in the SN. RLP38 was used as the housekeeping gene (n = 4). (C) Coronal sections of the SN showing MHC-I, CD16/32, and C1q staining and their respective quantifications indicating fluorescence intensity fold change relative to the uninjected site 4 weeks after rAAV2-hα-SYN injection in WT and tPA-KO mice (n = 6–10). Confocal images of the SN showing TH (white), (D) MHC-I, or (E) CD16/32 (green) staining in WT and tPA-KO mice 4 weeks after rAAV2-hα-SYN injection (n = 3–5). Arrows indicate MHC-I or CD16/32 in contact with TH⁺ cell bodies or axons in WT mice after rAAV2-hα-SYN injection. Confocal images are shown as maximum intensity projections. For (A) and (B) experiments were conducted using 12- to 13-week-old WT and tPA KO mice. For (C-E) experiments were conducted using 9- to 11-week-old WT and tPA KO mice. Data are shown as mean ± SEM, N.S = not significant, *p < 0.05 **p < 0.01; ***p < 0.001; ****p < 0.0001. (a-c) 1-way ANOVA followed by Tukey post hoc test. Scale bar= (C) 500 μm (D, E) 10 μm

Fig. 5

Lack of tPA reduced proinflammatory microglia and infiltrated T-cells in the SN after overexpression of hα-SYN. (A) Confocal images of the SN showing TMEM119 (red), TH (white), and MHC-I (green) staining 4 weeks after rAAV2-hα-SYN injection in WT and tPA-KO mice. Arrows indicate MHC-I colocalization with TMEM119 (MOC = 0.8 ± 0.1; n = 3). (B) Representative images of the SN showing CD3 (red) and TH (white) staining with closeup showing CD3⁺ cells in contact with TH cell bodies and axons (arrows). (C) Quantification of CD3⁺ T cells associated with the SNpc 4 weeks after rAAV2-hα-SYN in WT and tPA-KO mice (n = 5). (D) Confocal images of the SN showing CD3 (red), TH (white), and CD4 (green) staining or (E) CD3 (red), TH (white), and CD8 (green) staining in WT and tPA-KO mice 4 weeks after rAAV2-hα-SYN injection (n = 5). Orange arrows show the presence of CD4⁺ or CD8⁺ T cells associated with TH⁺ cell bodies and axons. (E) White arrows show the presence of CD8⁺/CD3⁻ cells. Confocal images are shown as maximum intensity projections. These experiments were conducted using 9- to 11-week-old WT and tPA KO mice. Data are shown as mean ± SEM, N.S = not significant, *p < 0.05; **p < 0.01; 1-way ANOVA followed by Tukey post hoc test. Scale bar= (A) 10 μm (B) 250 μm (B close up) 20 μm (D, E) 20 μm

Fig. 6

tPA-mediated DA-neuron degeneration is independent of its proteolytic activity in the SN. (A) Diagram depicting point mutation in exon 14 in tPA Ala-KI mice (top) and tPA Ala-BAC construct used to generate tPA Ala-BAC transgenic mice (bottom). (B) In-gel zymography (top; blue and white) and western blot (bottom; black and white) of whole-brain protein extract from WT, tPA Ala-KI, tPA Ala-BAC, and tPA-KO mice showing lack of tPA activity in tPA Ala-KI and tPA Ala-BAC mice but the presence of tPA protein in these mice via western blot (black arrow indicates endogenous tPA and tPA Ala; red arrow indicate the size shift in tPA Ala-Cerulean in the tPA Ala-BAC mice). (C) Luminex assay showing tPA levels in whole brain protein extracts in WT, tPA Ala-KI, tPA Ala-BAC, and tPA-KO mice (n = 3). tPA-KO mouse brain protein extracts were used as a negative control for in-gel zymography, western blot, and Luminex assay (b-c). (D) Representative images of the SN showing DA-neuron degeneration in the injected SN (TH; white), and (E) quantification of TH⁺ neurons in coronal sections of the SN 4 weeks after rAAV2-hα-SYN in WT, tPA Ala-KI, and tPA Ala-BAC mice. These experiments were conducted using 15-week-old WT mice, 13- to 14-week-old tPA Ala-KI mice, and 12- to 18-week-old tPA Ala-BAC mice. White dashed lines represent the SNpc (n = 11–12). Data are shown as mean ± SEM, N.S = not significant, **p < 0.01. 1-way ANOVA followed by Tukey post hoc test. Scale bar = 500 μm

Fig. 7

tPA-NMDAR1 interaction is necessary to induce DA-neuron degeneration in the SN. Representative images of the SN staining with (A) TH and (C) MHC-I, and the respective quantification of (B) DA-neuron survival and (D) MHC-I fluorescence intensity fold change in the SN 4 weeks after rAAV2-hα-SYN injection in tPA Ala-BAC mice treated with glunomab or isotype control (6 ug/day) (n = 5). (E) Confocal image of the SN showing CD3⁺ cells (red) in contact with TH⁺ cell bodies and axons. (F) Quantification of CD3⁺ T cells associated with the SNpc 4 weeks after rAAV2-hα-SYN in tPA Ala-BAC mice treated with glunomab or isotype control (n = 5). These experiments were conducted using 14- to 16-week-old tPA Ala-BAC mice. Data are shown as mean ± SEM, N.S = not significant, *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001. (B) 2-tailed t-test; (D, F)1-way ANOVA followed by Tukey post hoc test. Scale bar= (A, C) 250 μm, (E) 20 μm

Fig. 8

Intravenous delivery of glunomab reduces neuroinflammation and DA-neuron degeneration and rescues behavioral deficits after overexpression of hα-SYN. Representative images of the SN stained for (A) TH and (C) MHC-I, and the respective quantification of (B) DA-neuron survival and (D) MHC-I fluorescence intensity fold change in the SN 4 weeks after rAAV2-hα-SYN injection in WT mice treated with glunomab (10 mg/kg/week) or vehicle (n = 6–10). (E) Confocal image of the SN showing CD3⁺ cells (red) in contact with TH⁺ cell bodies and axons. (F) Quantification of CD3⁺ T cells associated with the SNpc 4 weeks after rAAV2-hα-SYN in WT mice treated with glunomab or vehicle (n = 7–10). (G) Quantification of sensorimotor bias in a corridor task 4 weeks after rAAV2-hα-SYN in WT mice treated with glunomab (10 mg/kg/week) or vehicle (n = 9–10). Glunomab treatment was performed on 12- and 16-week-old WT mice. For vehicle treatment, age-matched 12- and 16-week-old WT mice were used. Data are shown as mean ± SEM, N.S = not significant, *p < 0.05; ****p < 0.0001. (B, F, G) 2-tailed t-test; (D) 1-way ANOVA followed by Tukey post hoc test. Scale bar= (A, C) 250 μm (E) 20 μm