NEW YORK – A team led by investigators at the National Institutes of Health has demonstrated that a gene-focused analytical approach can uncover previously unappreciated genetic contributors to Alzheimer's disease (AD), pointing to the possibility of applying a similar strategy to other polygenic conditions.
"[O]ur gene-constrained approach to gene identification utilized a bioinformatic approach distinct from GWAS," William Simonds, a researcher with the National Institute of Diabetes and Digestive and Kidney Diseases, noted in an email, adding that such approaches "may complement the power of GWAS for the identification of AD risk genes and may be more broadly applicable for other polygenic diseases."
For a paper appearing in the American Journal of Human Genetics on Tuesday, Simonds and colleagues from the NIH, the University of Pennsylvania, and Partek relied on a gene-constrained analytical approach, using a "gene function impacting variant rate" (GFIVR) bioinformatics strategy focused on moderate- and high-risk variants found with the Ensembl Variant Effect Predictor that fell in protein-coding parts of the genome.
When the team employed this GFIVR-centered strategy to assess publicly available data for almost 181,400 AD cases or controls, it tracked down 660 genes suspected of contributing to increased AD risk, particularly in individuals with African or African American ancestry.
"Because of the known disparity in prevalence of AD among subjects of African ancestry, we intentionally biased our gene discovery efforts toward this population," Simonds explained, noting that past studies suggest individuals of African ancestry are roughly twice as likely to develop AD as those of European ancestry.
After bringing in RNA sequence data from nearly two dozen brain regions derived from 2,728 individuals with or without AD, and performing further analyses with OmicsNet network visualization and Ingenuity Pathway Analysis tools, the team was left with nine candidate genes that were not part of the Alzheimer’s Disease Sequencing Project (ADSP) Gene Verification Committee's list of verified AD-related genes or loci previously identified through genome-wide association studies.
While Simonds cautioned that the African ancestry-focused analyses "may have overlooked other critical genes more important in those of European ancestry," he said that "five of the nine genes we did identify were already implicated in other neurologic or neurodegenerative diseases, suggesting we are on the right track with our approach."
For follow-up experiments, the researchers looked more closely at one of the genes, a heterotrimeric G protein beta family-encoding gene called GNB5, which was previously linked to normal neuronal development, Simonds explained.
In a transgenic mouse model carrying APP and PSEN1 gene mutations, for example, they found that alterations affecting one copy of the GNB5 gene coincided with increased formation of AD-related brain deposits such as amyloid-beta plaque and neurofibrillary tangles.
"The only known function of GNB5 is as a stabilizer of regulator G-protein signaling proteins," Simonds wrote. "This strongly suggests that regulation of G-protein signaling is somehow important for the pathogenesis of AD," he noted, adding that "more research is necessary to identify which receptors and pathways specifically regulated by GNB5 may influence the formation of amyloid plaques and neurofibrillary tangles in AD."