Skip to main content
Premium Trial:

Request an Annual Quote

Personal Methylome, Transcriptome Profiling Over Time Reveals Illness-Linked Changes

NEW YORK (GenomeWeb) – Alterations to a person's methylome or transcriptome over time could point to the onset of illness, according to a new analysis.

In the study, reported today in Nature Medicine, Stanford University's Michael Snyder and colleagues followed changes to both his methylome and transcriptome of over the course of three years, finding that that changes to his methylome corresponded with spikes in his blood glucose levels — he has type 2 diabetes — while transcriptomic changes were typically linked to viral infections.

"Methylation is well known to be associated with a change in gene activity — the presence of methyl groups usually shuts off gene expression," Snyder said in a statement. "We're thinking that the regulatory sequence of these metabolic genes are being perturbed, and that's causing dysregulation in my glucose, contributing to my diabetes."

Snyder has previously undergone multi-omic profiling, which uncovered genetic variants that predisposed him to developing type 2 diabetes, which was then diagnosed.

The researchers collected blood samples from Snyder 57 times during a 36-month period and extracted both DNA and RNA from peripheral blood mononuclear cells for analysis. Within this timespan, Snyder developed six viral infections and experienced two periods with elevated fasting glucose levels and glycated hemoglobin A1c levels that reached diabetic levels.

He and his colleagues performed RNA-seq on samples from all the time points and MethylC-seq on samples from 28 time points. In general, the researchers noted a high correlation among DNA methylation patterns from the various time points, which they noted was expected as all the samples came from the same person. Still, it suggested the methylome was was largely stable over time.

They found, though, that his methylome and transcriptome had distinct dynamics that related to different physiological conditions.

The researchers unearthed genomic regions that were differentially methylated between adjacent time points. Most of these methylation differences occurred at time points corresponding to when Snyder experienced changes to his blood glucose levels. When they focused in on those events, they noted that the highest number of differentially methylated regions arose between 80 days and 90 days prior to when the actual change in glucose occurred.

"So because it's occurring a little bit before, it's a sign that the methyl changes might actually be somewhat responsible for, or contribute to the dysregulation of glucose," Snyder said.

The researchers also examined differences in methylation at gene promoters at adjacent time points and also linked these differences to blood glucose changes, noting that the genes whose promoters experienced the most change were enriched for involvement in glucose and diabetes-related pathways.

When they examined differences in transcription between time points, however, the researchers found that genes whose expression changed were enriched for involvement in immunological processes. They noted some differences linked to glucose-related genes, though to a lesser extent.

The researchers found that these transcription changes were instead linked to the viral infections Snyder experienced. For instance, as compared to healthy time points, 116 genes were differentially expressed during his two adenovirus infections, including 56 genes whose expression changed at specific time points during viral infection.

This difference in pattern suggested to the researchers that short-term, acute changes might be reflected in transcriptional changes, while chronic ones might be captured by the methylome.

These findings also indicated to them that personal omic profiling could help monitor disease. Further, as methylation changes cropped up in Snyder before any clinical symptoms did, he and his colleagues posited that such profiling could lead to early disease detection or even prevention.

To gauge the generalizability of their findings — as this study relied on a single individual — Snyder and his colleagues are now analyzing data collected from 100 people in a similar study.