NEW YORK – Researchers have identified hundreds of expression quantitative trait loci that have short-lived differences during cellular differentiation. These eQTLs could have later ramifications on phenotype or disease, according to a new analysis.
Researchers from Johns Hopkins University and the University of Chicago analyzed eQTLs using RNA sequencing data obtained from induced pluripotent stem cells as they differentiated into heart muscle cells. As they reported yesterday in Science, they found that some eQTLs had a greater influence early in differentiation, others were more influential later on, and yet others had changeable influences.
"Those associations are like shooting stars," senior author Yoav Gilad, chief of genetic medicine at Chicago, said in a statement. "They appear at one point and never again during development, and they might actually be important to the phenotype of the mature tissue and maybe even disease. But unless you study those particular cell types at that particular time, you'll never see them."
For their analysis, the researchers relied on a panel of induced pluripotent stem cell lines generated from 19 individuals. They then coaxed these iPSCs to develop into cardiomyoctes and collected RNA from them every 24 hours for 16 days, encompassing the differentiation period. Most previous eQTL studies, they noted, have relied on data collected from a single time point.
In all, the researchers sequenced 297 RNA samples, and within this dataset uncovered two groups of cells that tended to cluster. The first group, which contained more of the cells, was marked by greater expression of troponin later in development and by the increasing expression of myogenesis-linked genes as development progressed. The second group, meanwhile, expressed genes linked to KRAS, self-renewal of undifferentiated iPSCs, and decreased neuronal differentiation.
Using the software tool WASP, the researchers identified cis-eQTLs within their dataset at different time points. They reported uncovering a median 111 genes with at least one eQTL at each time point.
They further uncovered 550 genes with a significant dynamic eQTL, including ones whose effect was greatest early on in development and then declined, ones whose effect increased as development progressed, and those whose effect size changed during development.
These dynamic eQTLs, the researchers reported, were enriched for genes with roles in myogenesis and with links to dilated cardiomyopathy. In particular, they noted that two eQTLs — rs7633988 and rs6599234 — are also genome-wide association study variants for QRS duration and QT interval.
The researchers extended their search to include a wider range of regulatory patterns to find 693 genes with nonlinear dynamic eQTLs, a portion of which have a stronger influence during the middle of differentiation.
For instance, they found that rs8107849 is associated with the expression of ZNF606 and that its effect was greatest during days 4 through 11. Similarly, they noted an association between rs28818910 and C15orf39. Rs28818910, they said, has also been linked to body mass index and, more weakly, to red blood cell count. It too only has regulatory effects during intermediate stages of differentiation. This suggested to the researchers that a temporary regulatory effect could have phenotypic consequences.
"To fully understand how genetics impacts disease risk, we'll ultimately have to consider all the different cell types, developmental time points, and environmental conditions that could be relevant to different diseases," co-senior author Alexis Battle, an associate professor of biomedical engineering at Johns Hopkins, said in a statement. "This study is one step in that direction."