In old mouse hearts, exercise turned down DNA switches that aging had turned on

Every cell in the body carries the same DNA, but each cell type uses only part of it. Which genes are on is decided partly by enhancers: short stretches of DNA, often far from the gene itself, that act like dimmer switches. An enhancer works when proteins called transcription factors sit on it and when the DNA around it is unpacked enough to be reached.

With age, the pattern of switches in use shifts. One family of transcription factors, called AP-1, responds to stress and inflammation signals and is known to be more active in many aging cell types. A team working at Memorial Sloan Kettering Cancer Center in New York and the Victor Chang Cardiac Research Institute in Sydney asked two questions in the mouse heart: which switches does aging turn on, and can exercise turn them back down? The work is a preprint, not yet peer-reviewed.

What they did

The main experiment used four groups of mice, with males and females balanced: young and older animals, each either sedentary or trained. The trained mice ran on a treadmill for 45 minutes a day, five days a week, for six months, at a moderate pace. Training started at 5 months (young) or 13 months (older), so the mice were 11 or 19 months old at the end.

The team then read out, one nucleus at a time, which genes were active and which stretches of DNA were open in more than 100,000 heart-cell nuclei, covering the main cell types of the heart. They also mapped where the AP-1 protein c-FOS actually sat on the DNA.

Follow-up experiments tested the mechanism:

• Shorter training in very old mice. 22-month-old mice ran for 10 to 11 weeks; the team measured scar-like collagen in the heart and how well the heart muscle contracted.
• Testing the switches outside the body. About 1,900 short DNA pieces from these enhancers were placed in front of a reporter gene in heart fibroblasts, the cells that make connective tissue, to see how strongly each could drive a gene.
• Blocking the signals. In mice about two years old, the team used a virus to make heart muscle cells produce a broken version of AP-1 that blocks the real one. In another group of old mice, they injected an antibody that neutralizes IL-1β, an inflammatory messenger, every three days for five weeks.

What they found

• Exercise mostly reversed aging's gene changes. Of 176 genes that changed clearly with age, 167 (about 95%) moved the other way with exercise. Inflammation and tissue-scarring programs, which rose with age, were the most prominent.
• A younger "gene age". A gene-activity clock for the mouse heart judged exercised hearts younger than sedentary ones, by about 2 to 3 months.
• More than 100,000 switches. Across heart cell types, aging increased AP-1 binding at more than 100,000 sites, and exercise opposed most of these gains.
• Sleeping potential. The DNA switches gained with age were only weakly open in the heart, yet in the reporter test they drove genes about as strongly as switches that are always on. Aging seems to wake up regulatory DNA that was there all along.
• Inflammation keeps them on. In fibroblasts, IL-1β preferentially pulled AP-1 onto the age-gained switches. Neutralizing IL-1β in old mice for five weeks dimmed those same switches most, and shifted gene activity in the same direction as exercise. Blocking AP-1 in heart muscle cells also pushed age-related gene changes back.
• Function. Short-term training reduced collagen build-up and improved how the heart muscle contracted. The five-week IL-1β block, by contrast, did not clearly improve heart function.

What to keep in mind

• Mice, not people. All the experiments are in mice or mouse cells. The link to humans is indirect: genes near the exercise-responsive switches were over-represented among genes tied to coronary artery disease, atrial fibrillation and high blood pressure in human genetic studies.
• Not yet peer-reviewed. The results may change after review.
• Small groups. Several experiments used few animals; for example, the gene-activity readout after blocking AP-1 compared three treated mice with two controls.
• Not a full reset. Exercise dimmed the AP-1 switches, but a broader age-related pattern in which stretches of DNA are open stayed in place in trained old hearts.
• Molecules first. Most results are molecular; the authors report no clear gain in heart function from neutralizing IL-1β over five weeks, and treated mice gained weight.

Why it matters

The study suggests that part of how a heart's genes age is not locked in: the switches aging turns on seem to need an ongoing inflammatory signal to stay on, and exercise, even started in middle age, damped most of them. In the authors' words, this leaves "opportunities to modify gene regulation late in life". Whether the same switches behave this way in human hearts, and whether that is part of why exercise is good for older hearts, has not been tested.