Skin cells from women aged 53–68 “aged” 8 times faster by DNA markers after six months in a dish without dividing than cells in living people: metformin slowed these “clocks,” while rapamycin had little effect

Skin cells from women aged 53–68 “aged” 8 times faster by DNA markers after six months in a dish without dividing than cells in living people: metformin slowed these “clocks,” while rapamycin had little effect

Jerome Mertens’s group at the University of California, San Diego, cultured fibroblasts, the connective tissue cells of the skin, from four women for six months. The paper was published on November 1 in Aging Cell.

Cellular age is estimated using “clocks,” formulas that calculate age in years from DNA methylation (chemical marks that change with age) or patterns of gene activity. Cells collected from donors of different ages provide only a snapshot: this comparison cannot track the same cells over time or test drugs on them as they age.

Fibroblasts divide infrequently in the skin but spend decades exposed to oxidative stress and inflammation. In a 2022 study in Nature Aging, methylation clocks advanced with time in culture even in dividing cells that had not entered senescence (a stable arrest of cell division). The authors of the new paper prevented division: they grew the cells into a continuous layer, where crowding restricted division, and left them for six months without passaging.

The cells did not enter senescence: the marker proteins p16 and p21 were absent, and the cells resumed dividing when passaged. Methylation clocks advanced 8,2 times faster than calendar time, compared with the matching pace in living donors, accumulating about five years in six months. Gene activity clocks advanced 22,5 and 48,3 times faster than calendar time. The authors suggest that these clocks outpace methylation clocks because gene activity is more flexible and responds more sharply to the loss of constraints that maintain balance in tissue, whereas methylation age accumulates more gradually.

The researchers compared gene activity in the cultured cells with two datasets from human fibroblasts collected at different ages. Of 66 gene groups whose activity decreased in culture, 60 were associated with age in humans; of 31 groups whose activity increased, 20 were age-associated. For 24 groups identified both in culture and in both datasets, the changes followed the same direction as those associated with age, but proceeded faster in culture, several times faster for most groups. In cells from two donors, variation between cells increased over five months, which is considered a sign of aging. The proportion of cells falling into the “oldest” quarter based on age-related gene activity rose from 9–15% to 28–41%.

In a 2015 study, Mertens and colleagues showed that neurons converted directly from skin cells retain the donor’s age-related patterns of gene activity, while neurons generated through reprogrammed stem cells lose them. The researchers therefore converted cells from three donors into neurons after six months in culture, and the accumulated age was retained. These neurons were 4,2 years older by methylation clocks and 6,6–21,5 years older by gene activity clocks than neurons generated from the same cells before the prolonged culture period. This approach produces neurons of two ages from a single skin sample. The authors suggest that these neurons could be used to study the transition from mild memory impairment to Alzheimer’s disease.

The authors tested metformin and rapamycin on the donors’ cells for six months, using one concentration of each drug. They state that both drugs have extended lifespan in animals. Metformin reduced the rate of methylation clock progression from 8,2× to −2,8×: on average, the clocks ran backward. One gene activity clock slowed from 48,3× to 11×, while the other showed no change. Metformin altered the age-related course of activity in 151 genes. Rapamycin produced only a trend toward slower clock progression that did not reach statistical significance and affected substantially fewer genes. The authors see these findings as consistent with animal experiments: metformin slowed aging clocks in male cynomolgus macaques in a 40-month experiment, while rapamycin extended mouse lifespan but improved only a few signs of aging.