Epigenetic clock and other biomarkers of aging

What is a biomarker of aging and why do we need it?

We all probably know someone who “looks young for their age” and someone who “hasn’t aged well”:

A biomarker of aging allows us to turn this subjective description into objective medical reality. It enables us to conclusively say: yes, you are 50 years old, but you have the health of a 35-year-old. But you, young man, should take better care of yourself — your biological age is 10 years higher than chronological age, and this is fraught with a 48% increase in the risk of death.

What does aging have to do with the probability of death? Everything. In humans, as in practically all mammals, aging is accompanied by an exponential increase in mortality risk:

As can be seen from the above graph, at age 30, the annual risk of dying equals to about 1 in a thousand, but by age 80 it rises 100-fold. In fact, it is this age-related increase in the probability of death that gerontologists define as aging. Intriguingly, not every organism ‘ages’. There are species that actually ‘get younger’ with age — their likelihood of death falls and fertility grows:

Thus, the ultimate goal of life extension therapies is to learn how to roll back everyone’s biological age to the level of today’s healthy 25-year-old, and freeze it there. In turn, the task of a good aging biomarker is to separate biological age from chronological (passport) age. That is, to reliably show where exactly on this curve a given person falls:

Therefore, a good aging biomarker should be highly correlated with mortality risk, so that it can, in turn, be used to determine biological age. For example, if the biomarker shows that your current annual probability of death is 1/1000, then you are biologically 30, and if it is 1/100, then you are 60. Regardless of what your passport says. Because people die according to their health, not their passport.

A good aging biomarker should also demonstrate bidirectional dynamics: if we ‘rejuvenate’ an organism with a proven anti-aging therapy, we should see a decrease in biological age.

What is Epigenetics?

Epigenetics is a mechanism of controlling genes. In fact, there are several such mechanisms: methylation of the genes themselves, acetylation or methylation of histones on which these genes are “wound up,” and some other things that fall under the definition of epigenetic control.

Why do genes need to be controlled at all? First, because the DNA is identical in all cells of an organism, but very different genes must be active in a brain cell than in a skin cell. Also, because different genes are responsible for different stages of organismal development — a caterpillar and a butterfly have the same set of genes but very different profiles of their activity. The same is true for humans: one set of our genes is active while we are in the womb, another in our childhood, and yet others as we sexually mature and get old.

As it turns out, with age, the on/off profile of different genes changes in very similar ways between people. What is even more interesting, it changes in a similar way in mice and other animals. That is, the epigenetic aging of a mouse resembles the epigenetic aging of a person, only accelerated by a factor of 40:

We observed tissue-specific age-associated changes in DNA methylation [in mice] that was directionally consistent with those observed in humans. These findings lend further support to the notion that changes in DNA methylation are associated with chronological age and suggest that these processes are often conserved across tissues and between mammalian species.

Moreover, we even share some of the ‘gears’ in our epigenetic aging clocks — the genes that make up those clocks:

The differentially methylated regions in mice have high sequence conservation in humans and the pattern of methylation is also largely conserved between the two species.

So What Is an Epigenetic Clock?

At a simple level, the “epigenetic clock” is just a set of on/off gene switches that best correlates with age. With what age, one might ask, chronological or biological?

Initially, with both — after all, for wild animals both chronological and biological ages are quite close, save for some stochastic differences. They do not drink or smoke, and do frequent McDonald’s. Therefore, initially, their methylation (epigenetic) clocks are set (calibrated) according to the chronological age of each species, and then various ways of accelerating or slowing aging are tested on these animals in order to check whether the effects that prolong their life also slow down these clocks, and, conversely, those effects that shorten life, speed those clocks up.

This is what we observe in humans, too. In smokers, diabetics, AIDS patients, or people with Down syndrome (who age much faster), the biological age turned out to be higher than their chronological age. Conversely, in mice that received various anti-aging therapies, the ticking of their epigenetic clock was observably slower.

Also, a high degree of correlation between the methylation clock and age was observed in species other than mice or humans — in rotifers, chimpanzees, and even whales:

So What is So Special About These Clocks?

What is special about methylation clocks is the fact that they are highly correlated with mortality. For example, in this large-scale study by Horvath et al. on thirteen thousand people, it was found that each year of difference between methylation age and chronological age (i.e. if you are 45, and the methylation clock shows 46) increases one’s mortality risk by 2% to 4%.

This observation worked in both directions and had a cumulative effect: those whose methylation clock was 10 years ahead of their age had a 48% increase in their risk of death (1.04¹⁰ = 1.48), and those whose clocks were 5 years “younger” than their age were 18% less likely to die.

In another study, a high correlation between methylation clocks and the risk of lung cancer in smokers was shown. Moreover, both the risk and correlation coefficient increased with age:

We also showed that the ability for IEAA to predict lung cancer was strongest among individuals ages 70 and older. A one unit difference in IEAA was associated with a 2.5-fold increase in lung cancer among individuals ages 70+, compared to an only 50% increase when considering the entire 50+ year old sample.

Moreover, this study identified just 10 important methylation sites (for comparison, the “Horvath clock” mentioned above has 353 such sites) in which abnormal methylation of at least 6 out of those 10 was correlated with several-fold increase in the risk of death — both from all causes and from cancer or cardiovascular disease:

Finally, this study has demonstrated something very encouraging for rejuvenation prospects even of the elderly. The Japanese researchers have shown that defects in mitochondrial respiration are reversible and are thus not caused by accumulation of damage, but rather by epigenetic (programmed) changes. When skin cells of 100-year old people were subjected to epigenetic rejuvenation by using Yamanaka (OSKM) factors, all mitochondrial respiration defects disappeared:

We reprogrammed human fibroblast lines by generating iPSCs, and showed that the reprogramming of fibroblasts derived from elderly subjects restored age-associated respiration defects. Therefore, these age-associated phenotypes found in elderly fibroblasts are regulated reversibly and are similar to differentiation phenotypes in that both are controlled by epigenetic regulation, not by mutations in either nuclear or mtDNA. Given that human aging can be seen as a consequence of a programmed phenomenon, it is possible that epigenetic regulation also controls human aging.

Epigenetic Clocks and Anti-Aging Interventions

The main point of aging biomarkers is to use them to help us find the most effective ways to fight aging itself. Therefore, immediately after the methylation clock has established itself as such a potential biomarker, scientists rushed to investigate it to see if it reflects the effectiveness of various established anti-aging therapies in mice.

One study in mice to assess the effects of rapamycin, caloric restriction, and life-prolonging genetic mutations on the methylation clock has confirmed that all these interventions have a positive effect on the methylation clock:

We have formulated an epigenetic-aging model in mice and used it to find evidence that lifespan-extending conditions slow an epigenetic clock in mice livers.

Interestingly, the same researchers also published a parallel article, where they tried to identify which areas of the genome are subject to age-related changes in methylation. After analyzing 42 million (!) methylation sites, the authors came to the conclusion that the main objects of age-related changes are promoters and enhancers of highly expressed genes.

These observations could be interpreted as providing evidence for the programmed aging hypothesis — it seems that the key age-related change is the change in the expression profile of just a few key genetic regulators at the top of the hierarchy of homeostatic control, and this change cascades down to other age-related changes.

Very similar findings, indicating several key genes, are also described by Vadim Gladyshev and colleagues in their latest work:

The sites contributed unequally to the mDNAm clock and formed several distinct clusters (~30) associated with genes Hsf4, Kcns2, Map10, Tns2, Wnt3a, and Zscan2. We found that 17 out of 18 CpG sites common to subset 1 and 2 clocks were also present among 90 CpG sites of the mDNAm clock. Most of these 17 CpG sites were located within introns of Ciita, Cd200r4, Rasgef1c, Wnt3a, and Zscan2, and several were clustered. CpG sites in introns of particular genes often play a role of their secondary enhancers, and we note that several identified genes are involved in development, differentiation, and tissue morphogenesis, consistent with a program-like behavior.

The slowing down of methylation clock due to various anti-aging interventions was also shown in this paper:

We found that the mouse methylation clock is affected by biological interventions and as such we suggest that the prediction of the clock reflects not only chronological age but also biological age.

The same paper had a nice graphical representation of age-related changes in methylation levels of 329 methylation sites in different tissues:

The above graph shows how some genes are activated with age, others are silenced, and yet others remain unchanged.

By the way, a human twins study confirmed the correlation of methylation clock with mortality. The higher the biological age of one of the twins, the higher the probability that he/she will die first:

This hypothesis was supported by a classical survival analysis showing a 35% (4–77%) increased mortality risk for each 5‐year increase in the DNAm age vs. chronological age. Furthermore, the intrapair twin analysis revealed a more‐than‐double mortality risk for the DNAm oldest twin compared to the co‐twin and a ‘dose–response pattern’ with the odds of dying first increasing 3.2 (1.05–10.1) times per 5‐year DNAm age difference within twin pairs, thus showing a stronger association of DNAm age with mortality in the oldest-old when controlling for familial factors. In conclusion, our results support that DNAm age qualifies as a biomarker of aging.

In summary, there now seems to be plenty of biological evidence that epigenetic clock not only reflects but also controls aging. Therefore, the next logical step is to test the hypothesis that stopping or reversing epigenetic clock (for example, with the help of OSKM factors) would also stop or reverse the aging process. Thankfully, there seem to be very encouraging preliminary data in support of that hypothesis.