Dissecting the Causal Mechanism of X-Linked Dystonia-Parkinsonism by Integrating Genome and Transcriptome Assembly

Abstract

X-linked Dystonia-Parkinsonism (XDP) is a Mendelian neurodegenerative disease that is endemic to the Philippines and is associated with a founder haplotype. We integrated multiple genome and transcriptome assembly technologies to narrow the causal mutation to the TAF1 locus, which included a SINE-VNTR-Alu (SVA) retrotransposition into intron 32 of the gene. Transcriptome analyses identified decreased expression of the canonical cTAF1 transcript among XDP probands, and de novo assembly across multiple pluripotent stem-cell-derived neuronal lineages discovered aberrant TAF1 transcription that involved alternative splicing and intron retention (IR) in proximity to the SVA that was anti-correlated with overall TAF1 expression. CRISPR/Cas9 excision of the SVA rescued this XDP-specific transcriptional signature and normalized TAF1 expression in probands. These data suggest an SVA-mediated aberrant transcriptional mechanism associated with XDP and may provide a roadmap for layered technologies and integrated assembly-based analyses for other unsolved Mendelian disorders.

Keywords: DYT3; Parkinson’s disease; SVA; TAF1; XDP; dystonia; genome assembly; intron retention; retrotransposon; transcriptome assembly.

Figure 1. XDP associated genomic region and experimental design

(A) Genomic segment associated with XDP on Xq13.1 with seven variants reportedly shared among probands and not observed in controls: five single nucleotide variants, annotated as Disease-specific Single-nucleotide Changes (DSCs)-1,2,3,10,12; a SINE-VNTR-Alu (SVA) retrotransposon inserted antisense to TAF1; and a 48-bp deletion. (B) Experimental workflow showing the number of XDP probands (black), carrier females (mixed), and controls (red), with the number of clones for each cell line.

Figure 2. Haplotypes observed among XDP probands

(A) Allelic diversity of XDP haplotypes reconstructed from de novo assembly and CapSeq. All known DSCs (red) were detected with 47 additional variants shared among probands compared to controls for the predominant haplotype (n=373, 93% of XDP probands). Variations are shown in 5′ to 3′ orientation spanning the region. (B) Five recombinations (denoted by ®) with alleles observed for two recombinant haplotypes that narrowed the XDP causal locus. (C) Recombination between DSC1 and DSC3 in pedigree 27 produced haplotype H7, with all alleles shown. Dotted rectangle represents the narrowed XDP region shared among all haplotypes based on recombinations, with reversion to the reference allele observed at DSCn3 (See also Table S1 and Table S2). See key for all annotations.

Figure 3. Characterization of iPSC-derived NSCs and NGN2-induced cortical neurons

(A) Heatmap of relative expression of pluripotency, neural stem cell, neuronal and glial genes in NSCs and iNs based on RNAseq. (B) Representative images from proband, carrier female, and control iNs showing processes stained with doublecortin (DCX), βIII-tubulin/Tuj, and MAP2. (C) Ca²⁺ mobilization in iNs visualized via Fluo-4AM. Upper and lower panels show Fluo-4AM fluorescence before and after, respectively, KCl treatment in control (left panels), carrier (middle panels) and XDP (right panels) lines. (D) Representative traces show relative change in fluorescence intensity (ΔF/F) induced by KCl (upper panels) and kainate (lower panels) in control (left), carrier (middle) and patient (right) lines. Traces represent individual cells (n = 10-15 cells).

Figure 4. De novo assembly of TAF1 transcript structure and differential expression of splice variants

(A) Transcript structure from de novo assembly depicts TAF1 isoforms previously annotated in Ensembl and additional splice variants detected in this study. For each transcript, boxes denote exons in black (Ensembl-annotated) or pink (this study). Brown triangle indicates genomic position of the SVA. Notation is provided for the cell type in which each transcript was detected. Extension of the transcript assembled from Illumina short reads by the PacBio data are indicated by a dashed orange line with additional exons represented by orange boxes. The genomic coordinate reflects the insertion of SVA (2627 bp). (B) Relative expression abundance of each TAF1 transcript in controls (x-axis) and relative change in TAF1 transcripts in XDP probands compared to controls (y-axis) in NSC (left) and iNs (right). Error bars reflect FDR correction of 95% confidence interval. (C) Relative expression of each exon of cTAF1 in XDP NSCs relative to controls. Black dashed line represents no change.

Figure 5. Aberrant expression of TAF1 intron 32 and transcriptome-wide significance in XDP NSCs

(A) Composite plot demonstrates normalized Illumina sequencing coverage of TAF1 intron 32 in control (blue) and XDP (red) samples across three cell lines. Brown triangle and the vertical brown line indicate the SVA insertion site while shadowed areas represent TAF1 coding regions. Solid horizontal lines intersecting the Y axis show the average sequencing coverage of the TAF1 coding region in control (blue) and XDP (red) samples. X axis represents the genomic coordinates of human X chromosome with the SVA inserted. (B) Transcriptome-wide levels of IR among all 258,852 annotated introns in XDP vs. control NSCs (x-axis) plotted against significance levels (y-axis, log₁₀ transformed). Significant IR changes (FDR < 0.05) are marked in orange. (C) Expression correlations among TAF1 intron 32 expression, overall TAF1 expression and TAF1 transcripts in fibroblasts (left), NSCs (middle) and iNs (right). Colors indicate Spearman correlation coefficients. Rows and columns are clustered based on Euclidean distance.

Figure 6. Excision of the SVA rescues aberrant splicing and expression in intron 32 and expression of TAF1

(A) Sashimi plot depicting coverage and splicing in intron 32 of TAF1 in control, XDP, and SVA-excised (ΔSVA) proband NSCs. (B) Normalized RNA-Seq counts in intron 32 of TAF1 5′ to the SVA insertion (left) and TAF1 (right) in proband NSCs, corresponding ΔSVA clones, and control cells (one clone per individual). (C) Relative expression of intron 32 splice variant in fibroblasts (Fibro), iPSCs, NSCs, iNs, NSC-derived cortical neurons, and GABAergic neurons from XDP, control, and ΔSVA lines. Graphs represent mean (+SEM) from clones generated for each cell type. See methods for total numbers and biological replicates of each genotype. Unpaired two-tailed t-test (fibro) or one-way ANOVA with Tukey’s multiple comparisons test was performed on each cell type. *p<0.05, ***p<0.001, ****p<0.0001, or n.s. = not significant.

Figure 7. Co-expression modules with strongest enrichment for DEG in NSCs and neurons are enriched for cell growth and ER stress response

(A) Modules with significant enrichment for differentially expressed genes (DEGs) in NSCs (left) and iNs (right). The number of DEGs indicate the number of genes included in WGCNA analyses for each cell type. Color represents the significance of enrichment, and the number indicates the number of overlapping genes. Modules with the most significant enrichments for DEGs at FDR levels are outlined and the overlap between modules is represented in the Venn diagram with the corresponding enrichment p-value (center). (B) Significantly enriched gene ontology terms in 110 overlapping genes from Module 2 in NSCs and Module 5 in iNs.