Epigenomic and clinical analyses of striatal DAT binding in healthy individuals reveal well-known loci of Parkinson's disease

Abstract

Background: Striatal dopamine transporter (DAT) binding is a sensitive and specific endophenotype for detecting dopaminergic deficits across Parkinson's disease (PD) spectrum. Molecular and clinical signatures of PD in asymptomatic phases help understand the earliest pathophysiological mechanisms underlying the disease. We aimed to investigate whether blood epigenetic markers are associated with inter-individual variation of striatal DAT binding among healthy elderly individuals. We also investigated whether this potential inter-individual variation can manifest as dysfunction of particular cognitive domains. Omics studies conducted on endophenotypes of PD among healthy asymptomatic individuals can provide invaluable insights into early detection, disease mechanisms, and potential therapeutic targets for PD.

Method: We conducted a blood epigenome-wide association study of striatal DAT binding on 96 healthy individuals using the Illumina EPIC array. For functional annotation of our top results, we employed the enhancer-gene mapping strategy using a midbrain single-nucleus multimodal dataset. Finally, we conducted several investigative regression analyses on several neuropsychological tests across five cognitive domains to assess their association with striatal DAT binding among 250 healthy subjects.

Results: We identified seven suggestive (P-value<10^-5) CpG probes. Specifically, three probes were colocalized with three risk loci previously identified in PD's largest Genome-Wide Association Study (GWAS). UCN5A and APOE loci were identified as suggestive DMRs associated with striatal DAT binding. Functional analyses prioritized the FDFT1 gene as the potential target gene in the previously reported CTSB GWAS locus. We also showed that delayed recall memory impairment was correlated with reduced striatal DAT binding, irrespective of age.

Conclusion: Our study suggested epigenetic and cognitive signatures of striatal DAT binding among healthy individuals, providing valuable insights for future experimental and clinical studies of early PD.

Keywords: Cognitive domains; DAT scan; Epigenome-wide association study; Healthy individuals; Parkinson's disease.

Graphical abstract

Fig. 1

The figures depict the Manhattan and Q-Q plots of mean, right, and left striatal DAT binding EWAS among 96 healthy individuals. Each figure has two panels: a) This panel shows the Manhattan plot with the positional gene mapping of the top DMPs. The blue line shows the suggestive threshold (P-value<10⁻⁵) of our EWAS. Loci previously reported as top PD GWAS hits are highlighted in green. b) This panel illustrates the Q-Q plot of the P-values of our EWAS results with the bias- and inflation-corrected lambda of our analysis computed by the BACON package.

Fig. 2

The figure shows two top significant methylation probes of our EWAS on striatal DAT binding and their positional position and vicinity to two previously reported GWAS hits of the largest PD GWAS to date: a) This panel zooms into one of the top GWAS hits of the Nalls et al. study in the 7q11.21 locus and its linkage disequilibrium (LD) structure. The top SNP rs76949143 in this locus lies between the TPST1 and KCTD7 genes. It also shows the location of our top EWAS result (cg08118333 methylation probe) between the KCTD7 and RABGEF1 genes. The lower part shows the direct linear correlation of cg08118333 methylation with right striatal DAT binding among 96 healthy individuals obtained from our EWAS analysis. b) This panel zooms into one of the other top GWAS hits of the Nalls et al. study in the 12q24.33 locus and its LD structure. The top SNP rs11610045 in this locus is located between the GALNT9 and FBRSL1 genes. It also shows the location of our top EWAS result (cg18316254 methylation probe) between the GALNT9 and FBRSL1 genes. The lower part shows the inverse linear correlation of cg18316254 methylation with mean striatal DAT binding among 96 healthy individuals obtained from our EWAS analysis.

Fig. 3

The upper part of the figure shows the genomic position of one of the suggestive DMPs associated with right striatal DAT binding and its vicinity to the CTSB locus previously reported in the largest PD GWAS. The lower parts illustrate the overlap of this DMP with the chromatin-accessible region of the midbrain of nine healthy elderly individuals. It also shows our enhancer-gene mapping strategy using a midbrain single-nucleus multimodal dataset, which prioritized the FDFT1 gene as the potential candidate gene in this PD GWAS locus. a) This panel zooms into the CTSB GWAS locus, previously reported in the largest PD GWAS conducted by Nalls et al., and its LD structure and P-value. The top SNP rs1293298 in this locus is located in the CTSB gene. It also shows the location of our top EWAS DMP (cg14458991 methylation probe) in the FDFT1 gene next to the CTSB gene and a previously reported DMP in PD EWAS (cg17155499) in the NEIL2 gene. The lower part shows the overlap of the cg14458991 with the ATAC peaks obtained from the single-nucleus multimodal dataset of midbrain of 9 post-mortem healthy elderly individuals. b) This panel shows the direct linear correlation of the cg14458991 methylation probe with right striatal DAT binding among 96 healthy individuals obtained from our EWAS analysis. c) The lower part of the panel shows the heatmap result of our enhancer-gene mapping analysis. The rows are four potential genes in this locus. The columns show nine multimodal samples obtained from the midbrain of post-mortem healthy elderly individuals. Each cell's colors depict the degree to which ATAC peak regions correlate with gene expression. The upper part shows the scatter plot of the correlation between the expression of the top gene, FDFT1, and the chromatin accessibility of that region.