Login / Signup

Brain proteome-wide and transcriptome-wide association studies, Bayesian colocalization and Mendelian randomization analyses revealed causal genes of Parkinson's disease.

Siquan ZhouYe TianXuejiao SongJingyuan XiongGuo Cheng
Published in: The journals of gerontology. Series A, Biological sciences and medical sciences (2022)
How genome-wide associated loci confer risk for Parkinson's disease is unclear. To reveal causal risk genes through their effects on brain proteins to provide new pathogenesis insights for Parkinson's disease. Proteome-wide and transcriptome-wide associations were determined by functional summary-based imputation leveraging data from genome-wide association summary (56,306 Europeans, 1.4 million controls), brain proteomes (528 cases from two separate datasets) and transcriptome (452 cases), followed by Mendelian randomization, Bayesian colocalization, cell-type specific and brain regional expression, and drug-gene interaction analyses. As a result, genetically regulated protein abundances of 11 genes were associated with Parkinson's disease. Five genes (CD38, GPNMB, TMEM175, RAB7L1 and HIP1R) were colocalized. Four genes (GPNMB, SEC23IP, CD38 and DGKQ) demonstrated Mendelian randomized correlations (P < 8.10×10 -5). Higher GPNMB level (1.47, 1.28-1.68) and lower CD38 level (0.319, 0.24-0.43) were causally associated with higher risk of Parkinson's disease, consistent with transcriptomic evaluations. CD38 and GPNMB were preferentially enriched in astrocytes and oligodendrocyte precursor cells, respectively. And CD38 and GPNMB were suggested to be the targets of many oncological drugs from Drug-Gene Interaction database. In conclusion, utilizing multidimensional data, GPNMB and CD38 were prioritized as the causal genes of Parkinson's disease, crucial for mechanistic and therapeutic investigations.
Keyphrases