Skip to main content
Premium Trial:

Request an Annual Quote

Structural Variants Influence Brain Gene Expression, Disease Risk

NEW YORK — Researchers have uncovered thousands of structural variants that affect how genes are regulated in the human brain, as well as brain disease risk.

Specifically, the Icahn School of Medicine at Mount Sinai team uncovered nearly 171,000 structural variants from four cohorts and examined how these variants influenced gene expression in the brain and other molecular phenotypes. They noted that structural variants have been implicated in brain-related disorders and traits in the past but that their effects have not been well examined in brain tissue to date.

"[M]any genetic studies just identify novel SVs but never take it to the next level in understanding their functional impact," senior author Towfique Raj, an associate professor at Mount Sinai, said in an email. "There's a significant gap in the field in our understanding of the functional impact of structural variants in the genome."

In their new study, which appeared Monday in Nature Neuroscience, the researchers used quantitative trait locus analyses to examine the effect of the structural variants on histone modifications, mRNA expression, mRNA splicing, and protein levels. Through this, they found more than 3,000 SNPs linked to at least one molecular phenotype within human brain samples.

Raj and his colleagues first analyzed sequencing data on 1,881 samples from four different cohorts, the Religious Orders Study, the Memory and Aging Project, the Mayo Clinic, and the Mount Sinai Brain Bank. All of these data were made available through the Accelerating Medicines Partnership in Alzheimer's Disease (AMP-AD) Knowledge Portal. Using seven different tools to capture variations, they homed in on 170,966 high-confidence structural variants.

They then examined associations between the structural variants and not only mRNA but also protein expression and H3K9ac peaks in various brain regions. Most structural variants that affected protein levels also affected gene expression levels, they found, and structural variants associated with mRNA and proteins shared the same direction of effect in more than 87 percent of the pairs, indicating that their effects reach further down the regulatory cascade.

But the researchers also noted that some structural variations affected gene expression but not other parts of the regulatory cascade, suggesting that other factors may mediate their effect.

Meanwhile, they examined whether any of the structural variants tagged variants known to be involved in neurodegenerative disease from genome-wide association studies.

More than 800 common SVs were associated with 534 traits, and of those, 344 SVs were linked to a molecular phenotype. In particular, 47 SVs were found in linkage disequilibrium with variants associated with brain-related conditions including schizophrenia, autism, bipolar disorder, and multiple sclerosis. For instance, a deletion found upstream of SRR, which is involved in glutamatergic neurotransmission and plasticity, is in linkage disequilibrium with variants implicated by GWAS in schizophrenia. The deletion was additionally linked to an H3K9ac peak and reduced SRR RNA and protein expression.

The researchers further uncovered four structural variants in the MAPT region that were highly correlated with progressive supranuclear palsy. These variants tagged haplotypes previously reported to have links to the condition, as well as Parkinson's disease. "We went further to show that many of these SVs also affect other molecular traits. We can begin to better understand the molecular mechanisms," Raj said.

He added that he and his colleagues are now using targeted long-read sequencing and ddPCR to validate the structural variants.

"We acknowledge that more long-read [whole-genome sequencing] data is necessary to accurately identify SVs," he said, adding that "we are also working with collaborators on using model systems iPSC or organoid and CRISPR-Cas9 to further understand the impact of SVs on molecular phenotypes. This is a really understudied area in genomics."