In Vitro Experiments to Better Understand Epigenetic Age
Epigenetic clocks to assess age are constructed via the application of machine learning to epigenetic data at various ages, examining which CpG sites are methylated and which are not, and identifying weighted combinations of specific sites that correlate well with chronological age. Researchers have shown that an epigenetic age greater than chronological age, known as epigenetic age acceleration, correlates with greater mortality and burden of age-related disease. It is therefore thought that epigenetic clocks may allow for the rapid assessment of potential rejuvenation therapies, an important goal for the research and development community.
At present it is slow and expensive to quantify the degree to which any given approach to the treatment of aging is actually useful in mice, and very much more challenging in human patients, given the need for long-term studies. The ability to run a quick, low-cost assay immediately before and immediately after treatment, to obtain an accurate assessment of biological age, would greatly speed up development, allowing the research community to focus more rapidly on approaches that work, versus those that are marginal.
Unfortunately, there is as yet little understanding of what exactly is being measured by epigenetic clocks. How does the methylation of specific CpG sites relate to the underlying mechanisms of aging, or specific consequences of those mechanisms? That is a black box, and clocks likely reflect only some of the mechanisms and changes of aging. Without knowing which aspects of aging determine epigenetic age for any given epigenetic clock, the only way to trust that the clock will usefully measure the effects of a potential rejuvenation therapy is to calibrate it against that therapy in long-term studies. Which rather defeats the point, as then we are right back to the slow and untenable present situation.
Today's interesting open access paper is an example of the way in which researchers are starting to make inroads into understanding how specific aspects of aging relate to epigenetic age in various clocks. The authors here picked one clock and performed a variety of in vitro studies on cells in an attempt to illuminate the relationships between age-related damage and change in function and the epigenetic age as assessed by the chosen clock. The data and conclusions are interesting, but it is worth bearing in mind that this is just a first step on a likely lengthy road.
The relationship between epigenetic age and the hallmarks of ageing in human cells
The excitement following the development of epigenetic clocks has been tinged with uncertainty as to the meaning of their measurements (i.e., epigenetic age, EpiAge). This uncertainty is compounded by the fact that different epigenetic clocks appear to measure different features of aging. Our investigations using the Skin&blood clock uncovered many features of epigenetic aging, of which two are particularly important. First, epigenetic aging initiates at very early point of life when pluripotency ceases. This process evidently continues through development, postnatal growth, maturity, and adulthood until death, as epigenetic clocks are applicable to the entire lifespan. Therefore, epigenetic aging is not an auxiliary phenomenon but an integral part of the deterministic process of life. Despite this fact, epigenetic aging is not refractive to the influence of external factors that can alter its rate. Indeed, experiments demonstrate the malleability of the rate of epigenetic aging.
At a higher level of consideration, the innate nature and inevitability of epigenetic aging contrasts with the stochasticity of wear and tear, which is presumed to exert a measurable aging effect only later in life when damage outstrips repair. This, however, does not argue against the relevance of wear and tear and cellular senescence. Instead, these distinct stochastic processes are likely to synergize with epigenetic aging in manifesting the overall phenotypical features of aging. If a successful strategy against aging is to be found, then these distinct and parallel aging mechanisms must be addressed; for example, by the removal of senescent cells, together with the retardation of epigenetic aging.
Another pivotal point concerns the ticking of the clock. It is intuitive to assume that this ticking is owed to dynamic changes of methylation on age-related CpGs in all cells in a tissue. Our observations with cell clones suggest that the ticking of the epigenetic clock is, at the very least, a measure of change in cell composition with age. This change can perceivably occur through expansion or reduction of a subpopulation of cells with different ages within the tissue. It was previously shown that mouse muscle stem cells are considerably younger. Therefore, such changes can conceivably result from alterations in the relative amounts of stem cells and non-stem cells, although the impact of stem cells from many more different tissues to aging requires further empirical investigations.
It was particularly important to address the question of the relationship between epigenetic aging and cellular senescence, as previous reports were equivocal in their conclusions. Here, using primary cells from many individual donors, the results are clear that cellular senescence, although undoubtedly a major contributor to the aging phenotype, is not associated with epigenetic aging, as measured by the Skin&blood clock. Similarly, DNA damage, and genomic instability have been hypothesized and proffered as means by which cells undergo epigenetic aging. Here, using different primary cell types derived from multiple donors, irradiated in different ways (acute or continuous) at different doses and dose rates, we did not observe any measurable impact on the rate of epigenetic aging.
Collectively, the results described here with primary cells from a large number of donors and multiple cell types, as well as in vivo mouse experiments previously reported, indicate that nutrient sensing, mitochondrial function, stem cell exhaustion, and altered cell-cell communication affect epigenetic aging as measured by Skin&blood clock, but cellular senescence, telomere attrition, and genomic instability do not. The connection of epigenetic aging to four of the hallmarks of aging implies that these hallmarks are also mutually connected at a deeper level. If so, epigenetic clocks will be instrumental in identifying the underlying unifying mechanisms. The absence of a connection between the other aging hallmarks and epigenetic aging suggests that aging is a consequence of multiparallel mechanisms, crudely divided into deterministic pathways: those associated with epigenetic aging and stochastic ones, which are independent of epigenetic aging and may result instead from wear and tear.