NEW YORK (GenomeWeb) – Results from a large Parkinson's disease study are pointing to new genetic pathways that may contribute to the disease.
Researchers from Genentech, the National Institute on Aging, Data Tecnica International, and 23andMe brought together data for hundreds of thousands of individuals with or without Parkinson's disease for a multi-stage genome-wide association analysis and meta-analysis seeking new genetic contributors to the neurodegenerative condition.
Along with the dozens of common variants implicated in Parkinson's disease in the past, the team uncovered 17 previously unreported risk loci — a collection that was used to start assessing candidate genes in the context of expression data and expression quantitative trait loci (eQTL). The findings appeared online today in Nature Genetics.
"The identification of these candidate genes allows for the prioritization of functional studies to determine causal genes for [Parkinson's disease] and possible therapeutic targets," co-senior authors Robert Graham and Tushar Bhangale, investigators from Genentech, and their co-authors wrote.
For the first stage of the study, the researchers used custom Illumina arrays to genotype 6,476 individuals of European ancestry with Parkinson's disease from a 23andMe research cohort, as well as 302,042 unaffected controls.
Variants at a dozen loci were over-represented in the Parkinson's disease cases, they reported, including one new site and multiple regions enriched for active regulatory marks. Although several known loci turned up in the analysis, they did not see significant ties for variants in or around the CHMP2B or TMEM229B genes that were reported in the past.
The team folded in data for another 13,708 cases and 95,282 controls for its meta-analysis, uncovering 35 loci with significant Parkinson's disease associations. Of those, 29 were subsequently tested in a replication cohort made up of 5,851 individuals with the condition and 5,866 without, all genotyped with the Illumina NeuroX array.
In the process, the researchers identified 17 new Parkinson's disease-related loci, though one could not be assessed using genotype data available for the validation cohort. With the new risk loci — and two-dozen risk loci described in the past — they looked for candidate genes with a so-called neurocentric strategy that took into account eQTL data, expression profiles, and clues about genes with possible neurological effects.
"The candidate gene nomination strategy confirmed several known [Parkinson's disease] risk genes, including GBA, LRRK2, SNCA, and MAPT," the authors wrote. "Among the 41 [Parkinson's disease] risk loci, a total of 29 loci had either a protein-altering or a cis-eQTL variant linked to the index SNP."
In particular, the team noted, the candidate gene set was rife with representatives from pathways involved in lysosomal activity or autophagy. New candidate genes also turned up in mitochondrial pathways, as did risk loci overlapping with transcription factors regulating neuronal survival pathways.