Shared genetic risk between corticobasal degeneration, progressive supranuclear palsy, and frontotemporal dementia

Abstract

Corticobasal degeneration (CBD), progressive supranuclear palsy (PSP) and a subset of frontotemporal dementia (FTD) are neurodegenerative disorders characterized by tau inclusions in neurons and glia (tauopathies). Although clinical, pathological and genetic evidence suggests overlapping pathobiology between CBD, PSP, and FTD, the relationship between these disorders is still not well understood. Using summary statistics (odds ratios and p values) from large genome-wide association studies (total n = 14,286 cases and controls) and recently established genetic methods, we investigated the genetic overlap between CBD and PSP and CBD and FTD. We found up to 800-fold enrichment of genetic risk in CBD across different levels of significance for PSP or FTD. In addition to NSF (tagging the MAPT H1 haplotype), we observed that SNPs in or near MOBP, CXCR4, EGFR, and GLDC showed significant genetic overlap between CBD and PSP, whereas only SNPs tagging the MAPT haplotype overlapped between CBD and FTD. The risk alleles of the shared SNPs were associated with expression changes in cis-genes. Evaluating transcriptome levels across adult human brains, we found a unique neuroanatomic gene expression signature for each of the five overlapping gene loci (omnibus ANOVA p < 2.0 × 10^-16). Functionally, we found that these shared risk genes were associated with protein interaction and gene co-expression networks and showed enrichment for several neurodevelopmental pathways. Our findings suggest: (1) novel genetic overlap between CBD and PSP beyond the MAPT locus; (2) strong ties between CBD and FTD through the MAPT clade, and (3) unique combinations of overlapping genes that may, in part, influence selective regional or neuronal vulnerability observed in specific tauopathies.

Figure 1

Fold enrichment plots of enrichment versus nominal −log₁₀ p-values (corrected for inflation) in corticobasal degeneration (CBD) below the standard GWAS threshold of p < 5×10⁻⁸ as a function of significance of association with progressive supranuclear palsy (PSP, left panel) and frontotemporal dementia (FTD, right panel) and at the level of −log₁₀(p) ≥ 0, −log₁₀(p) ≥ 1, −log₁₀(p) ≥ 2 corresponding to p ≤ 1, p ≤ 0.1, p ≤ 0.01, respectively. Blue line indicates all SNPs.

Figure 2

‘Conjunction’ Manhattan plot of conjunction and conditional −log₁₀ (FDR) values for corticobasal degeneration (CBD) (black) given progressive supranuclear palsy (PSP; CBD|PSP, red) and frontotemporal dementia (CBD; CBD|FTD, orange). SNPs with conditional and conjunction −log₁₀ FDR > 1.3 (i.e. FDR < 0.05) are shown with large points. A black line around the large points indicates the most significant SNP in each LD block and this SNP was annotated with the closest gene, which is listed above the symbols in each locus.

Figure 3

Average regional gene expression of MAPT and MOBP from 6 postmortem subjects from the Allen Brain Science Institute mapped into a three dimensional reconstruction (‘inflated’ view) of the gray/white matter boundary of the cerebral cortex (fsaverage subject from FreeSurfer). Top panels illustrate the lateral view and bottom panels illustrate the medial view of the left cerebral hemisphere.

Figure 4

Color-coded bar plots illustrating mean regional transcript levels for MAPT, MOBP, CXCR4, EGFR and GLDC using data from the Allen Brain Science Institute. Values represent z-scores averaged across the individual probes and across the 6 postmortem subjects.

Figure 5

Voxel based morphometry results illustrating atrophy patterns for FTD (red), CBD (blue), PSP (yellow), and overlap of all three tauopathies. Results are shown as T-map overlays (p<0.001) of volume loss in patients versus controls, and heat map reflects overlap between tauopathies. Atrophy patterns were derived from a cohort of individuals clinically diagnosed and pathologically confirmed to have CBD (n=6), FTD Pick’s disease (n=5), or PSP (n=17), which was compared to a cohort of cognitively normal healthy older controls (n=180, mean±SD 66.1±8.5 years). For detailed methods and results, see Supplemental Material.

Figure 6

Network interaction graph illustrating physical interactions (pink), co-expression (purple), predicted (orange), pathway (aqua), co-localization (blue), gene interactions (green) and shared protein domains (khaki) for CXCR4, MOBP, EGFR, GLDC, and MAPT.