You’re Only As Old As Your Transcriptional Profile.

Getting older is a dangerous hobby. With increasing age comes greater susceptibility to chronic and acute conditions like heart disease, strokes and cancer. But while these trends are well-studied and clear, less is known about the specific mechanisms of aging that contribute to greater disease risk over time. Most large-scale studies to date have focused strictly on genetic factors like gene mutations, which provide only a partial picture of the entire cellular landscape.

A recent paper in Nature Communications describes a different approach. Here, the authors examined the “transcriptional landscape of age” by assaying blood from thousands of people to discover which genes were expressed, and particularly which of those vary in abundance with age. They also compared the levels of DNA methylation — a chromosomal modification that inactivates expression of nearby genes — in samples from individuals of varying age.

In the primary sample set of whole blood from nearly 15,000 individuals of European descent, the research team found about 900 genes expressed significantly higher in samples from younger individuals, and another 600 more abundant in the blood of older individuals. When expression of these genes was verified against other populations — Native Americans, Hispanic Americans and African American cohorts — between 30 and 75 percent of the identified genes showed similar patterns related to age. While these sample sets were smaller (between 400–500 individuals each), the results suggest there may be a core set of genes across populations that are expressed only in younger, or in older, individuals.

The authors explored the possible functions of these genes by performing clustering analyses of genes that tended to be abundant in the same samples, or co-expressed, and determining the known cellular pathways associated with those genes. Among genes expressed preferentially in younger individuals, they found components of DNA repair pathways, genes associated with mitochondria and metabolism and clusters linked to naïve immune system response. The genes up-regulated in older individuals were instead involved in regulation of cellular structure, degradation of specialized polysaccharides called glycosaminoglycans as well as the innate and adaptive immunity. While some of these functions have been linked to aging in other studies, others are implicated here for the first time.

Interestingly, the age-associated differences among the immune-related gene clusters held true even when the authors tested individual cell types from the blood samples. This means that the differences are not just due to the relative abundance of certain cell types — like naïve T cells in younger individuals, or memory T cells in the older population — but that differences in expression of at least some genes persists across cell types. The authors found an additional hint of this phenomenon by testing a set of brain tissue samples against their initial gene sets. In both cerebellum and frontal cortex cells, around 20–25% of the genes identified in blood samples had similar expression patterns associated with age.

The results from the authors’ DNA methylation study were somewhat surprising. Earlier work has identified a number of DNA methylation sites correlated with chronological age — but these were not significantly enriched near the genes showing the highest differences in expression. Likewise, the promoters of these age-associated genes were not preferentially methylated. Instead, nearby regulatory regions — gene enhancers and insulators — showed much better correlation with the expression of these specific genes, suggesting that these genes undergo active regulation, which may be fine-tuned or altered over time.

Perhaps the most striking result from this study was the predictive success of the genes identified. Based on the expression of “age-correlated” and “age-anti-correlated” genes, the researchers could make predictions about the age of the individual donating each sample. Those predictions were generally fairly accurate, with an average difference from chronological age of less than eight years. But when the “transcriptomic age prediction” was higher than the actual age, there were consistent health factors — high blood pressure, high cholesterol, high BMI — associated with those individuals. This suggests not only that gene expression may be useful for estimating age and understanding the mechanisms underlying the aging process, but it could also help to identify individuals with higher health risks, whose transcriptional profiles mark them “older” than their calendar years.

Charlie Hatton | Meta Staff Writer