Our DNA is made up of genes that vary drastically in size. In humans, genes can be as short as a few hundred molecules known as bases or as long as two million bases. These genes carry instructions for constructing proteins and other information crucial to keeping the body running. Now a new study suggests that longer genes become less active than shorter genes as we grow older. And understanding this phenomenon could reveal new ways of countering the aging process.

Luís Amaral, a professor of chemical and biological engineering at Northwestern University, says he and his colleagues did not initially set out to examine gene length. Some of Amaral’s collaborators at Northwestern had been trying to pinpoint alterations in gene expression—the process through which the information in a piece of DNA is used to form a functional product, such as a protein or piece of genetic material called RNA—as mice aged. But they were struggling to identify consistent changes. “It seemed like almost everything was random,” Amaral says.

Then, at the suggestion of Thomas Stoeger, a postdoctoral scholar In Amaral’s lab, the team decided to consider shifts in gene length. Prior studies had hinted that there might be such a large-scale change in gene activity with age—showing, for example, that the amount of RNA declines over time and that disruptions to transcription (the process through which RNA copies, or transcripts, are formed from DNA templates) can have a greater impact on longer genes than shorter ones.

Stoeger, Amaral and their team used a machine-learning algorithm to identify features that best explained changes in RNA from 17 different tissues, including heart, brain and kidney, in male mice that were four, nine, 12, 18 and 24 months old. (The strain of mouse used in this study is considered “very old” at 24 months.) This analysis revealed a clear and consistent pattern across tissues: longer transcripts became less abundant than shorter transcripts in older animals. This imbalance in long- and short-gene expression provided a possible explanation for why they couldn’t find a specific set of genes whose expression was changing. While the particular genes being expressed varied from experiment to experiment, overall, shorter genes appeared to become more active than longer genes as animals aged, according to Amaral. “You will always find hundreds of genes that seem to change, but once you see it in terms of this linear trend, everything makes sense,” he says. (Amaral notes, however, that while changes in transcription are the likeliest explanation for his and his colleagues’ findings, other processes, such as the degradation of RNA, may also be at play.)

The team repeated this experiment using data collected from various types of postmortem human tissue, as well as tissues extracted at specific ages in other animals. They found this age-associated imbalance in gene-length-related expression was consistent across organisms. The human findings were particularly exciting, because unlike the mice, which were genetically identical and raised in the same laboratory conditions, the humans lived different lives and died of different causes at different times, Amaral says. “The fact that you find the same pattern despite this diversity really says that this is something robust,” he says. “That result dramatically increases my confidence in this being a true and important pattern.”

When Amaral and his colleagues looked at the longest and shortest transcripts, they found that the top 5 percent of genes with the shortest transcripts included many linked to shorter life spans, such as those involved in maintaining the length of telomeres (DNA sequences at the ends of chromosomes that become shorter with age) and immune function. And they found that the top 5 percent of genes with the longest transcripts included ones linked with longevity, such as neuronal activity and transcriptional regulation. They also examined the effects of 12 antiaging interventions on the balance of short- and long-gene activity by reassessing data from previously published animal experiments. Seven of these interventions—which included rapamycin and resveratrol, two antiaging drugs—led to a relative increase in long-gene transcripts, suggesting that this aging-associated imbalance may be reversible. The findings were published in December in Nature Aging.

This study fits with previous work, according to Maria Ermolaeva, a group leader at the Leibniz Institute of Aging in Germany, who was not involved in the study. For example, researchers have shown that the accumulation of DNA damage during aging has a stronger effect on longer genes; the longer the gene is, the more likely it is to develop a problem that cannot be repaired, she says. Such unrepaired DNA lesions stall the process of transcription, leading to a reduction in the transcripts produced from longer genes. “The authors of the new study might have observed the global consequences of this previously described molecular phenomenon,” Ermolaeva says.

The transcriptome imbalance the authors observe with age “is an interesting association,” but whether this process drives aging remains to be seen, says João Pedro de Magalhães, a professor of molecular biogerontology at the University of Birmingham in England, who also was not involved in this study. “I wouldn’t discard it as a possibility, but I think you will need some pretty strong evidence that we don’t have yet,” he says. It could be that length-associated transcriptome changes are simply a reflection of other aging-related processes, such as an uptick in immune system activity. Small genes are often associated with immune function—and immune processes such as inflammation tend to become more active as we get older, de Magalhães adds. “So it makes some sense that you would see patterns in terms of gene length, because it reflects the processes that are being altered with age.”

Amaral speculates that the imbalance in transcription could be caused by the accumulation of harmful exposures—viral infections, for example—over the life span that gradually alters the cellular machinery required to successfully transcribe longer genes. “Maybe aging is a measure of this imbalance—the greater the imbalance, the more aged you are, the more aged your tissue is,” he adds. In future experiments, Amaral hopes to examine how injuries influence the transcriptome imbalance in younger organisms—and to potentially see if antiaging interventions could help restore the imbalance that occurs after potentially damaging exposures.

There are plenty of open questions to address, such as how, exactly, the transcriptional machinery gets altered with age, Amaral says. “We hope that this study will get people excited to do experiments that could help us unravel what’s going on in greater depth.”