Skip to main content
Premium Trial:

Request an Annual Quote

Genetic Variant Effects on Gene Expression Mapped in Elderly Human Brains

Brain DNA

NEW YORK – With the help of single-cell transcriptomics and genetic analyses, an international team has documented the effects of genetic variants, including disease-related ones, on expression in specific cell types in the brains of old individuals.

"[W]e systematically mapped the effect of genetic variation across a wide variety of cellular contexts in the aging brain," co-senior and corresponding author Philip De Jager, a researcher at Columbia University's Center for Translational and Computational Neuroimmunology, and his colleagues wrote in a paper published in Nature Genetics on Thursday. They further noted that "[i]ntegration of these results with genome-wide association studies highlighted the targeted cell type and probable causal gene within Alzheimer's disease, schizophrenia, educational attainment, and Parkinson's disease loci."

For their analyses, the researchers used single-nucleus RNA sequencing to assess more than 1.5 million individual cells in the mood- and cognition-related dorsolateral prefrontal cortex region of the brain. The profiled cells were isolated from frozen post-mortem brain samples from 424 elderly individuals enrolled in the longitudinal Religious Orders Study or the Memory and Aging Project.

At the time of their death, 40 percent of participants had documented dementia, 26 percent showed mild cognitive impairment, and 34 percent were classified as cognitively non-impaired. More than 60 percent fulfilled the criteria for a pathological diagnosis of Alzheimer's disease, the team noted.

Together with participants' genotyping profiles and published epigenomic features for cell types in the brain region, the investigators analyzed snRNA-seq profiles on a median of 3,824 nuclei per person, as well as "pseudo-bulk" RNA-seq data representing transcriptomic data for each cell type.

In the process, they described brain cell clusters defining distinct cell types and subtypes in the dorsolateral prefrontal cortex, while identifying expression quantitative trait loci (eQTL) and target genes within these cells.

Across seven of the eight major cell type clusters found, they flagged more than 10,000 eQTL-impacted genes, dubbed eGenes. By focusing on 64 of the 95 cell subtypes that had sufficient eQTL mapping data, they tracked down nearly 8,100 eGenes with activity at the cell subtype level, including genes influenced by specific eQTLs exclusively within a given cell subtype.

The team validated eGene results for two of the main cell types using induced pluripotent stem cell-derived neurons and astrocytes from a few dozen of the study participants.

The authors noted that the varying proportions of different cell types likely contributed to their ability to find eQTLs in certain cell types, since far more single-cell transcriptomic data was available for abundant cell types such as neurons compared to less common ones. For example, they saw more than 7,300 apparent eGenes in excitatory neurons, while 899 eGenes turned up in the less common microglia.

Even so, the researchers found a new genetic variant influencing the expression of the APOE gene in microglia cells. That variant, in turn, appeared to coincide with cerebral amyloid angiopathy burden but did not show clear ties to Alzheimer's disease pathology once investigators accounted for participants' APOE-4 genotypes.

By focusing on genetic variants previously implicated in Alzheimer's and other conditions, on the other hand, the investigators demonstrated that they could pick out eGenes and cell types expected to be impacted by the risk variants.

In particular, they focused in on 21 eGenes impacted by 20 loci linked to Alzheimer's in the past. An analysis of 11 Parkinson's disease-related loci led to 13 eGenes, meanwhile, and 57 schizophrenia-related loci coincided with 75 eGenes, many of them centered on excitatory neurons.

"While we characterized many disease loci, it is clear that many more eGenes remain to be discovered," the authors concluded. "We also highlighted deeper sequencing to better resolve cell subtypes as an important aspect of the path forward."