DNA encodes several characteristics of an individual, but it turns out that these sequences alone don’t always determine everything about a person. Instead, a second set of instructions in the form of chemical tags that are added to the DNA, called epigenetic modifications, influences whether or not and to what extent a gene is expressed. One of the most common markers is DNA methylation, or the addition of a methyl group to a cytosine directly next to a guanine (CpG).
As the field began connecting these epigenetic changes to aging, researchers explored the potential for using these methylation profiles as windows into this process, launching the lengthy development of epigenetic clocks over the next decade and a half.1-3 The Scientist continues to celebrate its 40th anniversary, take a look back on some of the magazine’s highlights of epigenetic milestones.
2011: Scientists Measure Age from Spit
To take a closer look at the relationship between methylation and age, researchers at the University of California, Los Angeles (UCLA) studied these DNA patterns from saliva in a cohort of twins between the ages of 21 and 55. They identified 88 sites where these chemical marks correlated with age, and these associations held up in a larger group of a general population. From this, they trained a model on two gene profiles with the strongest association to age. When they applied the model to their original group’s data, it predicted the individuals’ ages within five years. The researchers envisioned this type of testing to help with forensic analyses and medical screening.
2013: Building an Epigenetic Clock from DNA Methylation
People may expect that all of their cells age at the same rate, but in 2013, geneticist Steve Horvath at UCLA showed that different tissues can have varied biological ages. Comparing the methylation patterns of 51 different tissues and cell types, Horvath found 353 CpG sites where the methylation patterns changed over time and used these to develop an epigenetic clock. For example, this prediction tool estimated embryonic or stem cell tissue to be nearly zero based on its lack of methylation. The findings received mixed reviews from scientists: some felt that studying these patterns can offer insights into aging, whereas others pointed out that, without a mechanism behind this process, all the field can say is that DNA methylation is associated with aging.
2017: Biological Aging Goes Commercial
As the concept of a biological age influenced by lifestyle and genetic factors became mainstream, companies broadened their product line to include direct-to-consumer kits to determine one’s biological age. Zymo Research released the first of these tests based on the clock developed by Horvath’s group but included additional CpG sites from other researchers. Since the kit did not have US Food and Drug Administration approval, the company could not provide any health-related information about individuals’ results. Yet researchers, including Horvath, questioned the value of the kit for the general population since scientists still don’t fully understand the influences of DNA methylation or how to intervene in this process.
2019: A Cross-Species Epigenetic Clock Predicts Animal Ages
One of the greatest challenges in basic science research is translating findings from animal studies to humans. Epigenetic researchers ran into this issue as well, when a clock trained on one species failed to work in another. When geneticists Bernardo Lemos and Meng Wang, both then at the Harvard University, first studied methylation patterns in ribosomal DNA, they were following up on previous indications that the nucleolus, where this specialized DNA is kept, was related to longevity. The researchers showed that methylation increased with age in mice and built a clock from 72 of these sites. Surprisingly, these DNA locations were conserved in dogs and humans and predicted individuals’ ages more accurately than other tools. Scientists hope that these patterns can also provide insight into the functions of the nucleolus in addition to aging.
2023: Epigenetic Clocks Zeroes in on Super-agers
Understanding aging through epigenetic clocks can help researchers suggest how people can remain healthy as they get older. This often involves studying DNA and its chemical dimmer switches in long-lived people. However, reports have suggested that some age claims are inaccurate, such as the current record holder for the world’s oldest person, Jeanne Calment. In response, Horvath developed an epigenetic clock focused on validating older individuals’ ages where he prioritized collecting samples from people over 100 years old from around the world. After identifying 33,000 CpG sites with strong and weak correlations to age, he trained a machine learning model to predict how old a person was. Although the clock was more accurate in predicting older individuals’ ages compared to earlier clocks, Horvath could not put Calment’s age to the test, as the institution holding her sample said it would be unethical to do so since she cannot consent to the study.
2024: T Cells Tell Time Differently
While Horvath’s group showed that different tissues can age at distinct rates in the body, in 2013, one group showed that some cells can live incredibly long lives—sometimes outliving their host—by measuring time in a unique way. T cells are adaptive immune cells that remain inactive in the body until they encounter a specific antigen, at which point they rapidly divide. A research team, led by University of Minnesota immunologist David Masopust, showed that they were able to maintain the same population of T cells for over 10 years, transferring these to new animals 17 times. Studying these cells’ epigenetic profiles, they showed that T cells methylation patterns correlated to their number of proliferations and not their chronological time. The findings could offer insights into understanding the immune system during aging.
2026: Different Variables Tune the Settings of Epigenetic Clocks
After decades of datasets, epigenetics and the clocks made from them are robust tools for scientists. However, several variables influence how well these predictive models compare against each other. Various clocks use different training data, and the sample source, health patterns studied, and tissues included can all influence how the model measures the sample. Horvath, who was working at the longevity biotechnology company Alto Labs at the time, said that this is particularly an issue with commercial kits, where they use saliva against clocks trained on blood or measure methylation using a different technology. Horvath added that the value of these tests for individuals is also uncertain, and that some tests that are currently offered are still early in how well they are validated. But, he and his team are working on validating similar tools and anti-aging therapies with AI so that they can become clinically useful.
- Fraga MF, Esteller M. Epigenetics and aging: The targets and the marks. Trends Genet. 2007;23(8):413-418.
- Maegawa S, et al. Widespread and tissue specific age-related DNA methylation changes in mice. Gen Res. 2010;20(3):332-340.
- Shumaker DK, et al. Mutant nuclear lamin A leads to progressive alterations of epigenetic control in premature aging. Proc Natl Acad Sci USA. 2006;103(23):8703-8708.
