NEW YORK (GenomeWeb) – New research suggests that some genetic variants contributing to Alzheimer's disease act by influencing alternative gene splicing events.
Using RNA sequencing, researchers from the Icahn School of Medicine at Mount Sinai, Columbia University, and elsewhere analyzed post-mortem brain tissue samples from 450 individuals from prospective aging cohorts, searching for pre-messenger RNA splicing events and related genetic variants linked to Alzheimer's disease. The findings appeared online today in Nature Genetics, and highlighted splicing quantitative trait loci (sQTL) affecting the expression of more than 3,000 genes.
"Overall, we deepen our understanding of genetic regulation in the transcriptome of the aging brain," the authors wrote, "and provide a foundation for the formulation of mechanistic hypotheses for Alzheimer's disease and other neurodegenerative diseases."
In particular, the team noted that Alzheimer's-associated alleles in and around the PICALM, CLU, and PTK2B genes seemed to contribute to disease risk by altering gene splicing. Such findings were complemented by the group's transcriptome-wide association studies, which led to 21 new and known risk genes, pointing to an over-representation of genes from an autophagy-lysosomal pathway.
"Our transcriptome-wide reference map of RNA splicing in the aging cortex is a new resource that provides insights for many different neurologic and psychiatric diseases," senior author Philip De Jager, director of Columbia University's Center for Translational and Computational Neuroimmunology, said in a statement.
For the study, the researchers used Affymetrix arrays to genotype 450 individuals enrolled through the Religious Order Study or Memory and Aging Project. They also did paired-end RNA sequencing on 542 post-mortem brain samples with the Illumina HiSeq instrument, focusing on the dorsolateral prefrontal cortex (DLPFC), and were left with 450 samples after the quality control and genotype-matching steps.
"Given the high abundance of alternative splicing in the brain, we hypothesized that other genetic variants associated with Alzheimer's disease could also affect pre-mRNA splicing, possibly by disrupting efficient binding of splicing factors," the authors explained.
From thousands of alternative splicing events detected in more than 16,500 genes, the team searched for differential splicing events and gene expression changes associated with Alzheimer's disease. The analysis highlighted 84 differentially spliced genes, including 52 altered slicing events that were significantly associated with Alzheimer's disease in a follow-up set of 301 brain samples from the Alzheimer's disease-free and –affected individuals from the Mount Sinai Brain Bank. And 11 of the differentially expressed genes corresponded with genes showing altered splicing in an induced pluripotent stem cell model for the disease.
The researchers brought in SNP data for their sQTL analysis, which unearthed more than 9,000 sQTLs, including many overlapping with SNPs previously linked to epigenetic features. Variants found in past genome-wide association studies of Alzheimer's disease were also over-represented in the set of sQTLs identified in the post-mortem DLPFC samples, they reported.
To build on those findings, the team did two TWAS to look for genes with alternative splicing or expression profiles impacted by genetic variants, leading to genes at known Alzheimer's disease-related loci and at eight loci not linked to the disease in the past.
"[T]hese new insights into genetic mechanisms in the aging brain will help offer new strategies and directions for RNA-targeted biomarkers and therapeutic intervention in Alzheimer's disease," co-first author Towfique Raj, a neuroscience, genetics, and genomics researcher at the Icahn School of Medicine, said in a statement.
Raj noted that antisense oligonucleotide compounds that target RNA splicing have already been proposed for treating brain-related conditions ranging from spinal muscular atrophy or amyotrophic lateral sclerosis to Huntington disease.