The SATB1-MIR22-GBA axis mediates glucocerebroside accumulation inducing a cellular senescence-like phenotype in dopaminergic neurons

Abstract

Idiopathic Parkinson's disease (PD) is characterized by the loss of dopaminergic neurons in the substantia nigra pars compacta, which is associated with neuroinflammation and reactive gliosis. The underlying cause of PD and the concurrent neuroinflammation are not well understood. In this study, we utilize human and murine neuronal lines, stem cell-derived dopaminergic neurons, and mice to demonstrate that three previously identified genetic risk factors for PD, namely SATB1, MIR22HG, and GBA, are components of a single gene regulatory pathway. Our findings indicate that dysregulation of this pathway leads to the upregulation of glucocerebrosides (GluCer), which triggers a cellular senescence-like phenotype in dopaminergic neurons. Specifically, we discovered that downregulation of the transcriptional repressor SATB1 results in the derepression of the microRNA miR-22-3p, leading to decreased GBA expression and subsequent accumulation of GluCer. Furthermore, our results demonstrate that an increase in GluCer alone is sufficient to impair lysosomal and mitochondrial function, thereby inducing cellular senescence. Dysregulation of the SATB1-MIR22-GBA pathway, observed in both PD patients and normal aging, leads to lysosomal and mitochondrial dysfunction due to the GluCer accumulation, ultimately resulting in a cellular senescence-like phenotype in dopaminergic neurons. Therefore, our study highlights a novel pathway involving three genetic risk factors for PD and provides a potential mechanism for the senescence-induced neuroinflammation and reactive gliosis observed in both PD and normal aging.

Keywords: MicroRNAs; Parkinson's disease; gene expression regulation; glucosylceramides; lysosomes; mitochondria; neuroinflammation; senescence.

FIGURE 1

Knockout (KO) of SATB1 leads to MIR22HG de‐repression and subsequent downregulation of GCase in dopaminergic (DA) neurons. (a) Western blot analysis and quantification reveal a significant decrease in GCase protein levels in SATB1‐KO compared to wild‐type (WT) DA neurons (n = 6). (b) RNA expression profiling of MIR22HG, GBA, and GBAP1 in WT and SATB1‐KO DA neurons shows differential expression (Day 50, n = 3). (c) Genomic data display the MIR22HG gene with ATAC‐seq enrichment tracks from WT and SATB1‐KO DA neurons, SATB1 ChIP‐seq from WT DA neurons, and H3K9ac ChIP‐seq data from human substantia nigra as a regulatory region marker. (d) Quantification of GCase protein expression in SATB1‐KO SK‐N‐MC cells treated with a miR22‐3p inhibitor or scrambled control (N = 4, n = 12). (e) Quantification of GCase protein expression in WT SK‐N‐MC cells treated with a miR22‐3p mimic or scrambled control (N = 3, n = 9). Data are presented as mean ± S.E.M. Two‐way ANOVA was performed for b. Student's t‐test was performed for a, d, and e. * p < 0.05; ** p < 0.01; *** p < 0.001; ns = not significant.

FIGURE 2

SATB1 knockout (KO) leads to reduced GCase levels and increased vulnerability in vitro and in vivo. (a), Representative images from a SA‐β‐Gal assay comparing wild‐type (WT) and Satb1‐KO N2A cells. Blue cells indicate senescent cells. (b), Quantification of the senescence assay in (a) (WT, N = 9 and n = 1277; KO, N = 10 and n = 1061). (c) Western blot analysis of GCase protein levels in N2A Satb1‐KO cells confirms the reduction of GCase protein (WT, n = 3; KO, n = 3). (d) Measurement of β‐glucosidase enzymatic activity in N2A Satb1‐KO cells compared to controls (n = 23). Cell viability assays demonstrate the increased dose‐dependent vulnerability of N2A Satb1‐KO to treatment with 6‐OHDA (n ≥ 21) (e) and H₂O₂ (f) compared to control cells (n ≥ 10). (g) Cell viability assays in WT and KO N2A cells with and without GBA overexpression. GBA overexpression rescues Satb1‐KO vulnerability to 6‐OHDA treatment (n = 8/condition). (h) Representative tyrosine hydroxylase (TH) immunofluorescent staining and quantification in mice receiving a stereotaxic injection with a shRNA‐Satb1 virus and a control vector, as well as a contralateral injection with a shRNA‐Satb1 virus and a GBA‐overexpressing virus. Injected substantia nigra pars compacta is shown and was quantified using unbiased stereological cell counting of TH⁺ cells (n = 7) (scale bar: 500 μm). Data are presented as mean ± S.E.M. Two‐way ANOVA was performed for e, f, and g. Student's t‐test was performed for b, c and d. Paired Student's t‐test was performed for h. * p < 0.05; ** p < 0.01; *** p < 0.001, ns = not significant.

FIGURE 3

Knockout (KO) of SATB1 disrupts lysosomal function leading to α‐SYN accumulation which can be rescued with GBA overexpression. (a) Fluorescence microscopy images and quantification of lysosomal content in wild‐type (WT) and Satb1‐KO cells with and without overexpression of GBA or SATB1. Lysosomal content (a) (n ≥ 89/condition) and cathepsin‐D activity (b) (n = 4/condition) are increased in Satb1‐KO cells but can be normalized by extrinsic expression of either SATB1 or GBA. c, Triton X‐100 insoluble α‐SYN levels in N2A^WT (n = 4) and N2A^{Satb1‐ KO} (n = 4) cells. Representative Western blot and quantification of α‐SYN monomers and oligomers are shown. (d) Co‐transfection of GBA or SATB1 along with α‐SYN (A53T) reduced triton X‐100 insoluble α‐SYN when compared to A53T transfection alone. Representative Western blot and quantification are shown (n = 3/condition). (e) Treatment of human DA neurons with recombinant GCase leads to a significant reduction in LAMP1 levels and normalizes the elevation of α‐SYN (n = 3/condition). Data are presented as mean ± S.E.M. Two‐way ANOVA was performed for a, b and d. Student's t‐test was performed for c. * p < 0.05; *** p < 0.001, ns = not significant.

FIGURE 4

SATB1 knockout (KO) disrupts mitochondrial function and turnover which can be rescued by GBA overexpression. (a) Transmission electron microscopy–based ultrastructural analysis of mitochondria in wild‐type (WT) and Satb1‐KO N2A cells. N2A^Satb1‐KO cells exhibit severe accumulation of abnormal mitochondria. (b) The ratio of mitochondrion diameter to cristae length was significantly lower in Satb1‐KO cells than in WT cells (WT, n = 113; KO, n = 106). (c) Measurement of basal respiration and ATP production in N2A^WT and N2A^Satb1‐KO. (d) Qualitative analysis of mitochondrial activity using a MitoTracker Red CMX probe in WT and N2A Satb1‐KO neurons. (e), Oxygen consumption rate measurement in N2A^WT, N2A^Satb1‐KO, and N2A^Gba‐KOcells. Satb1 and Gba‐KO cells show significant impairment in basal respiration and ATP production. Overexpression of SATB1 or GBA rescues ATP production and respiration, while SATB1 does not improve the phenotype in Gba‐KO cells, indicating that Satb1 acts upstream of Gba (n ≥ 7/condition). (f) Downregulation of α‐SYN is insufficient to restore oxygen respiration and ATP production in Satb1‐KO neurons (n = 8). (g) Increased vulnerability of N2A^{Satb1‐ KO} cells to the glycolysis inhibitor 2‐deoxyglucose compared to controls (n = 6). Data are presented as mean ± S.E.M. Two‐way ANOVA was performed for c, e, f and g. Student's t‐test was performed for b. ** p < 0.01; *** p < 0.001, ns = not significant.

FIGURE 5

SATB1‐KO‐mediated GBA reduction causes lipid accumulation in DA neurons. (a), Transmission electron microscopy (TEM) image showing lipid accumulation in SATB1‐KO DA neurons (* indicates lipid inclusion). (b) SATB1‐KO DA neurons show a significant increase in the number of lipid vesicles per cell (cells analyzed for TEM: WT, n = 88; KO, n = 110). (c) Immunofluorescence images using anti‐GluCer antibodies and ActinGreen demonstrating increased GluCer levels in SATB1‐KO DA neurons. (d) Dot blot analysis and quantification of GluCer levels in SATB1‐KO DA neurons relative to controls (normalized to protein amount, n = 4). (e) Treatment of human DA neurons with recombinant GCase leads to the presence of intracellular recombinant GCase (n = 4). (f) Significant reduction of GluCer accumulation in human DA neurons following treatment with recombinant GCase (n = 4). Data are presented as mean ± S.E.M. Two‐way ANOVA was performed for f. Student's t‐test was performed for b and d. * p < 0.05; ** p < 0.01, ns = not significant.

FIGURE 6

GluCer directly impairs mitochondria and lysosome function leading to p21 and p16 upregulation and a senescence‐like phenotype in DA neurons. (a), LysoTracker staining of wild‐type (WT) N2A cells treated with DMSO control, 2.5 μM GluCer, or 40 μM GluCer for 6 days, demonstrating a significant dose‐dependent increase in lysosomal content (integrated intensity) upon lipid treatment (N = 3, cells analyzed for lysotracker: DMSO n = 456, 2.5 μM GluCer n = 461, and 40 μM GluCer n = 523). (b) Seahorse analysis of lipid‐treated mesencephalic DA (mDA) neurons showing dose‐dependent alterations in oxygen consumption rate (2 ng/mL is 2.8 μM). (c) Treatment with GENZ‐112638, an inhibitor of GluCer synthesis, mitigates the effects of lipid treatment on mitochondrial function. (d, e) SA‐β‐Gal‐based senescence assay performed to evaluate senescence induction after 7 days of GluCer treatment in N2A cells (d, N = 6, DMSO n = 1909 and 2.5 μM GluCer n = 3253) and mature WT human DA neurons (e, N = 3, DMSO n = 528 and 40 μM GluCer n = 345), demonstrating that GluCer treatment is sufficient to induce senescence in both cell types. (f) RNA‐seq analysis of WT human DA neurons treated with DMSO (n = 3) or 40 μM GluCer (n = 3) revealing significantly increased p16 (Cdkn2a) and p21 (Cdkn1a) expression as shown by TPM values and a positive tendency for p19 (Cdkn2d). Data are presented as mean ± S.E.M. Two‐way ANOVA was performed for a, b, c, and f. Student's t‐test was performed for d and e. * p < 0.05; *** p < 0.001, ns = not significant. (g), Overview of the SATB1‐MIR22HG‐GBA‐GluCer pathway. The senescence regulator SATB1 acts as a negative regulator of MIR22HG expression. Decreased SATB1 levels lead to increased MIR22HG expression. Following nuclear processing, miR‐22‐3p targets GBA mRNA, reducing its transcription. This results in reduced GCase activity and the consequent accumulation of its substrate GluCer, a cerebroside that impairs mitochondrial and lysosomal function. GluCer accumulation leads to elevated ROS production and increased p21, p16 and S100A9 expression, ultimately driving cellular senescence. Importantly, the genes SATB1, MIR22HG, and GBA (red stars) are associated with PD. During aging, SATB1 and GBA levels decrease (red hourglasses), while MIR22HG, GluCer, p21/p16, and ROS levels increase (green hourglasses), rendering DA neurons more vulnerable. Both S100A9 and miR22 induce cellular senescence (Shi et al., ; Xu et al., 2011) (figure generated in biorender, agreement number: SR261VU25Q).