Identifying candidate Parkinson's disease genes by integrating genomic data
Original author: Demis A. Kia et al. JAMA Neurol. (2021) (DOI: 10.1001/jamaneurol.2020.5257)
Summary
Sr. scientist – Omics
October 19, 2022
Introduction
Parkinson’s disease (PD) is a neurological disorder observed mostly in elderly individuals. The symptoms include tremors, muscle stiffness, slow motion, and loss of balance. It is caused when neurons of the basal ganglia region of the brain fail to produce the neurotransmitter dopamine which regulates movement. The genomic analysis has identified more than 40 loci associated with PD risk. The causal genes corresponding to each locus and their roles in PD are not clear.
Methodology
To identify the candidate genes researchers have used Genome-Wide Association Studies (GWASs) along with Quantitative Trait Loci (QTL). It provides associations of an individual's genotype with gene expression (eQTLs), splicing, or methylation. The data was obtained from several sources like International Parkinson’s Disease Genomics Consortium, the UK Brain Expression Consortium Braineac dataset, and the Genotype-Tissue Expression (GTEx) Consortium dataset. The GWAS data contained 8055803 genotyped and imputed variants in 26035 PD patients and 403190 controls. The genetic colocalization analysis was done using the Coloc R package which probes whether two phenotypes share common genetic causal variant(s) in each region. For integrating GWAS and QTL data, the Transcriptome-wide association study (TWAS) was done. It probes associations between gene expression and complex diseases or traits.
Results
Researchers found out of 515 genes within 1 Mb of a significant PD risk variant, 470 were expressed at a detectable level across different datasets. They proceeded with these genes. Coloc analysis showed out of 470 genes, 9 in Braineac and 27 in GTEx have strong evidence for colocalization in at least one brain region. TWAS analysis showed 61 genes were found to be significantly associated with PD risk. Five genes (WDR6, CD38, GPNMB, RAB29, and TMEM163) were replicated among Coloc and TWAS analysis.
Braineac dataset was used for exon-level eQTL data. Colocalization analysis found 25 genes having strong evidence for colocalization in at least one exon in at least one region of the brain. For 15 genes the evidence suggested that the association is due to an exon-level splicing event. A combination of colocalization and TWAS analysis identifies six genes (ZRANB, PCGF3, NEK1, NUPL2, GALC, and CTSB) having a putative splicing effect.
It provided 11 candidate genes. Analysis of methylation data showed 3 genes (GPNMB, TMEM163, and CTSB) overlapped with expression and splicing analysis. Cell-type–specific expression of candidate genes was studied. Although no single cell type dominated, glial cell types dominate over neurons. WGCNA analysis provided NUPL2, TMEM163, and ZRANB3 were the most relevant genes.
Researchers further constructed a network of 11 candidate genes. Candidate genes are related to proteins that are also relevant for Mendelian forms of PD and Parkinsonism. It implies disease-specific interaction between candidate genes with known risk genes. Pathway enrichment analysis was done with a core composed of candidate genes and Mendelian genes. It showed enrichment of proteins involved in or regulating the ERBB receptor tyrosine-protein kinase signaling pathways.
Discussion
Researchers have performed a comprehensive analysis by integrating QTL and GWAS data and found 11 candidate genes. It is the key strength of the study. The biological roles of those genes were also explored. CD38 regulates neuroinflammation and it showed enrichment in the astrocyte region. It shows the existence of a group of genes and proteins – associated with the ERBB signaling pathways. It increases the risk of developing sporadic and familial PD.
The current study has some limitations also. For example, it considers only cis-QTL and excludes trans-QTL which corresponds to distant genes of different chromosomes.
Impact of the research
It applied a new methodology in the context of PD. It identified hitherto unknown PD risk genes and validated the biological functions of the genes.
Conclusion
In conclusion, it combines GWAS with QTL data to discover 11 candidate genes whose changes in expression, splicing, or methylation are associated with the risk of PD. Protein-protein interaction network analysis highlights the functional pathways and cell types where these genes have important roles.
