Cerebellar Heterogeneity and Selective vulnerability in Spinocerebellar Ataxia Type 1 (SCA1)

Abstract

Heterogeneity is one of the key features of the healthy brain and selective vulnerability characterizes many, if not all, neurodegenerative diseases. While cerebellum contains majority of brain cells, neither its heterogeneity nor selective vulnerability in disease are well understood. Here we describe molecular, cellular and functional heterogeneity in the context of healthy cerebellum as well as in cerebellar disease Spinocerebellar Ataxia Type 1 (SCA1). We first compared disease pathology in cerebellar vermis and hemispheres across anterior to posterior axis in a knock-in SCA1 mouse model. Using immunohistochemistry, we demonstrated earlier and more severe pathology of PCs and glia in the posterior cerebellar vermis of SCA1 mice. We also demonstrate heterogeneity of Bergmann glia in the unaffected, wild-type mice. Then, using RNA sequencing, we found both shared, as well as, posterior cerebellum-specific molecular mechanisms of pathogenesis that include exacerbated gene dysregulation, increased number of altered signaling pathways, and decreased pathway activity scores in the posterior cerebellum of SCA1 mice. We demonstrated unexpectedly large differences in the gene expression between posterior and anterior cerebellar vermis of wild-type mice, indicative of robust intraregional heterogeneity of gene expression in the healthy cerebellum. Additionally, we found that SCA1 disease profoundly reduces intracerebellar heterogeneity of gene expression. Further, using fiber photometry, we found that population level PC calcium activity was altered in the posterior lobules in SCA1 mice during walking. We also identified regional differences in the population level activity of Purkinje cells (PCs) in unrestrained wild-type mice that were diminished in SCA1 mice.

Keywords: Cerebellum; Heterogeneity; Neurodegeneration; SCA1; Selective vulnerability.

Figure 1.. Assessment of PCs dendritic atrophy and synaptic loss in across cerebellar regions in 18-week-old Atxn1^154Q/2Q mice.

A. Top. Schematic of sagittal section of mouse cerebellar vermis (Left) and hemispheres (Right) from Allan Brain Atlas. Vermal lobules II, V/VI, VII, and X and hemispheres lobules Crus 1, crus 2, paramedian, and copula pyramidis were measured and are labeled with red numbers. Bottom. Calbindin, vglut2, and merged images for WT (top) and Atxn1^154Q/2Q mice (bottom). B. Quantification of molecular layer (ML) thickness. C. Quantification of calbindin intensity (relative intensity normalized to wild-type sex matched littermates). D. Quantification of excitatory synapse loss in molecular layer by ratio of VGLUT2/molecular layer width measurements. Black bars represent WT, pink bars represent Atxn1^154Q/2Q mice. Student’s t-test, error bars represent SEM, ** = p < 0.005, **** = p < 0.0001. N=7-8 for WT, N=5-8 for Atxn1^154Q/2Q mice.

Figure 2.. Assessment of gliosis across cerebellar vermis and hemispheres 18-week-old Atxn1^154Q/2Q mice.

A. Images of GFAP staining in the cerebellar vermis lobules II, primary fissure, VII, and X, and hemisphere lobules crus1, crus1, paramedian, and copula pyramidis. WT (top), Atxn1^154Q/2Q mice (bottom). B. Quantification of intensity of GFAP signal for all eight regions of the cerebellum, relative to wild-type littermate and sex matched controls. Black bars represent WT, pink bars represent Atxn1^154Q/2Q mice. Student’s t-test, error bars represent SEM, * = p < 0.01. N=6-68for WT, N=5-8 for Atxn1^154Q/2Q mice. C. Images of Iba1 staining in the cerebellar vermis lobules II, primary fissure, VII, and X, and hemisphere lobules crus1, crus1, paramedian, and copula pyramidis. WT (top), Atxn1^154Q/2Q mice (bottom). D. Quantification of density of Iba1 positive cells (microglia) in the molecular layer for all eight regions of the cerebellum. Black bars represent WT, pink bars represent Atxn1^154Q/2Q mice. Student’s t-test, error bars represent SEM, * p < 0.05, ** =p < 0.01. N=6-8 for WT, N=5-8 for Atxn1^154Q/2Q mice.

Figure 3.. Assessment of PCs CF synapses, GFAP expression and microglia density across the cerebellar vermis hemispheres 18-week-old wild-type mice.

A. Quantification of excitatory climbing fiber synapses in the molecular layer by ratio of VGLUT2 / molecular layer width. B. Quantification of GFAP intensity across the four cerebellar vermis regions in wild-type mice normalized to Lobule II, and to Crus 1 in hemispheres. C. Quantification of density of Iba1 positive cells (microglia) in the molecular layer for all eight regions of the cerebellum. One-way ANOVA with Tukey post-hoc test, error bars represent SEM, * p < 0.05, *** p < 0.005, ****p < 0.001, N = 6-8 for WT, N=5-8 for Atxn1^154Q/2Q mice.

Figure 4.. SCA1 intracerebellar transcriptomics identifies shared and unique disease pathways in Atxn1^154Q/2Q mice.

A. Venn diagram of DEGs in each cerebellar cortical region. 1655 DEGs in CCAV (green), 1755 DEGs in CCPV (purple), and 1444 DEGs in CCH (blue). B. Number of up and down regulated DEGs in CCAV, CCPV, and CCH. C. Expression levels of Magenta genes across cerebellar cortical regions of wild-type and Atxn1^154Q/2Q mice. Heatmaps of gene expression levels of Magenta genes for each genotype in each cerebellar cortical region. CCAV (green), CCPV (blue), and CCH (purple). Wild-type (pink), Atxn1^154Q/2Q (teal). Each column represents an individual sample. Heatmaps were generated by plotting the counts per million (cpm) of each gene in individual samples. Expression is normalized per gene in the heat map representation. D. Heatmaps of top up and down regulated DEGs in between WT and Atxn1^154Q/2Q mice in each cerebellar cortical region: CCAV (C), CCPV (D), and CCH (E). Each column represents an individual sample, WT in pink, Atxn1^154Q/2Q in teal, females with hashed boxes, males with solid boxes. Heatmaps were generated by ranking the top 15 up and down regulated DEGs based on logFC value and plotting the counts per million (cpm) of each gene in individual samples. Expression is normalized per gene in the heat map representation. E. KEGG pathway analysis of enriched pathways of top 500 DEGs between WT and Atxn1^154Q/2Q mice in each cerebellar cortical region. Dot size represents number of genes in the enriched pathway, and color of the dot represents p-value. F. Average pathway activity in each cerebellar cortical region of WT and Atxn1^154Q/2Q mice.

Figure 5.. Transcriptomic analysis in wild-type mice identifies differently regulated genes across the cerebellar cortex.

A. Volcano plots of all DEGs between the anterior and posterior vermis (left), the anterior vermis and the cerebellar hemispheres (middle), and the posterior vermis and cerebellar hemispheres (right). Red dots represent significant DEGs based on both p-value and logFC, gray dots represent DEGs significant by one metric or that did not reach significance. B. Heatmaps of top up and down regulated DEGs in between the anterior and posterior vermis (left), the anterior vermis and the cerebellar hemispheres (middle), and the posterior vermis and cerebellar hemispheres (right). Heatmaps were generated by ranking the top 15 up and down regulated DEGs based on logFC value and plotting the counts per million (CPM) of each gene in individual samples. Expression is normalized per gene in the heatmap representation. C. KEGG pathway analysis of enriched pathways of top 500 DEGs between the anterior and posterior vermis, the anterior vermis and the cerebellar hemispheres, and the posterior vermis and cerebellar hemispheres. Dot size represents gene ratio (number of genes in the enriched pathway divided by the total number of DEGs), and color of the dot represents p-value. D. Cck mRNA expression across cerebellar cortical regions in WT mice by RTqPCR. n = 4 per genotype. Error bars = SEM, One-way ANOVA with Tukey’s test. * p < 0.05.

Figure 6.. Reduced heterogeneity in SCA1 cerebellar regions.

A.-B. Transcriptomic analysis in wild-type and in SCA1 mice identifies differently regulated genes across the cerebellar cortex and genes uniquely expressed in either of the regions in pairwise comparison. C. Volcano plots of all DEGs between the anterior and posterior vermis (left), the anterior vermis and the cerebellar hemispheres (middle), and the posterior vermis and cerebellar hemispheres (right). Red dots represent significant DEGs based on both p-value and logFC, gray dots represent DEGs significant by only one metric or that did not reach significance.

Figure 7.. ATXN1 affects the calcium activity in PCs from lobule IV and lobule VI.

A. Example of optic fiber implant in lobules IV/V and lobule VI (right) and schematics of experiment and example traces of calcium activity (green) and movement (blue) (left). B. Total frequency, amplitude and area under the curve (AUC) of PCs calcium events over ten minutes exploration in the arena in lobules IV (LIV) and VI (LVI) of wild-type mice (grey) and SCA1 (red). C. Frequency, amplitude and area under the curve (AUC) of PCs calcium events over ten minutes of exploration in the arena in lobules IV (LIV) and VI (LVI) during ambulation in wild-type mice (grey) and SCA1 (red). D. Frequency, amplitude and area under the curve (AUC) of PCs calcium events over ten minutes of exploration in the arena in lobules IV (LIV) and VI (LVI) while wild-type mice (grey) and SCA1 (red) were stationary (non-walking). Data is average ± SEM with individual events presented as dots. n = 344-527 events N= 5 wild-type and N = 6 Atxn1^154Q/2Q mice 16-18 weeks of age. One-way ANOVA with Tukey’s post hoc * p< 0.05, ** p< 0.001, *** p < 0.0001.

Figure 8.. Regional pattern of PC activity change is absent in ATXN1 mice during walking.

A. Frequency, amplitude and area under the curve (AUC) of PCs calcium events over ten minutes exploration in the arena in lobules IV (LIV) and VI (LVI) of wild-type mice. B. Frequency, amplitude and area under the curve (AUC) of PCs calcium events during periods of walking and non-walking in lobules IV (LIV) and VI (LVI) of wild-type mice. Data is average ± SEM with individual events presented as dots. n = 218-512 events from N = 5 wild-type mice 16-18 weeks of age. One-way ANOVA with Tukey’s post hoc * p< 0.05, ** p< 0.001, *** p < 0.0001. C. Frequency, amplitude and area under the curve (AUC) of PCs calcium events over ten minutes of exploration in the arena in lobules IV (LIV) and VI (LVI) while walking and not walking of SCA1 mice. D. Frequency, amplitude and area under the curve (AUC) of PCs calcium events over ten minutes of exploration in the arena in lobules IV (LIV) and VI (LVI) while walking and not walking of SCA1 mice. Data is average ± SEM with individual events presented as dots. n = 300-527 events from N = 6 Atxn1^154Q/2Q mice 16-18 weeks of age. One-way ANOVA with Tukey’s post hoc * p< 0.05, ** p< 0.001, *** p < 0.0001.