Integrated genetic analysis and single cell-RNA sequencing for brain image-derived phenotypes and Parkinson's disease

Abstract

Background

Previous studies have reported Parkinson's disease (PD) patients usually have changes in brain image-derived phenotypes (IDPs). However, the role of genetic factors in their association and biological mechanism remains unclear. We aimed to unveil genetic and biological links between brain IDPs and PD.

Methods

Using genome-wide association study (GWAS) summary statistics and single-cell RNA sequencing (scRNA-seq) data, we performed a comprehensive analysis between 624 brain IDPs and PD. The genetic correlations and causality were examined by linkage disequilibrium score regression (LDSC), two-sample bidirectional Mendelian randomization (MR) and meta-analysis. Potential shared genes were identified using MAGMA and PLACO. Finally, pathway enrichment using FUMA and Metascape, and scRNA-seq analysis were performed to determine biological mechanisms and gene expression atlas across various cell types in brain tissue.

Results

LDSC revealed that 50 brain IDPs were genetically correlated with PD (P < 0.05), in which 5 IDPs, exhibited putative causality on PD through MR (P < 0.05). For instance, we identified that the increased volume of the right thalamus (IVW: OR = 2.08, 95 % CI: 1.33 to 3.25, PFDR = 0.03) was positively correlated with the risk of PD, which was also supported by replicated MR (IVW: OR = 1.63, 95 % CI: 1.17–2.26, PFDR = 0.02) in FinnGen and meta-analysis (OR = 1.78, 95 % CI: 1.36–2.31, PFDR = 5.00 × 10⁻⁴). Additionally, we identified 56 unique pleiotropic genes, such as FAM13A, with notable enrichment in neuronal cells. Biological mechanism analysis revealed these genes were enriched in brain tissues and a variety of pathways such as negative regulation of neuron apoptotic processes.

Conclusion

We indicated the shared genetic architecture and biological mechanisms between brain IDPs and PD. These findings might provide insights on the therapeutic intervention and early prediction of PD at the brain imaging level.