A growing number of researchers are developing and testing various implementations of a DNA methylation biomarker of aging. There is even a US company offering a low-cost test for those who want to give it a try. The quality of resulting data and degree of testing and validation accomplished for these various approaches is quite varied. Some provide only loose correlations with mortality and life expectancy, while others produce estimates of age with a five year margin of error. This depends as much on the intentions of the research team as on the details of construction of the biomarker. Not every team has the funding or time to prove their case very rigorously in large data sets, versus creating an initial proof of concept to show that their approach to the biomarker is worthy of that funding and time.
How do these DNA methylation biomarkers differ form one another? DNA methylation is a form of epigenetic marker, a molecular decoration on DNA that can occur at any CpG site in the genome. This mark determines the pace at which proteins are produced from the related genetic blueprint, which genes are active and which are silent, and is one part of the many regulatory mechanisms that drive changes in cell behavior. All of the switches and dials inside the machinery of a cell can be traced back through chains of cause and consequence to a matter of how much of a particular protein is being produced. A cell's epigenetic configuration is a reaction to the circumstances that cell finds itself in. Some portion of that set of circumstances is due to the age of the tissue within which the cell is situated. Since we all age for the same underlying reasons, we all accumulate the same molecular damage, some of the epigenetic changes that occur with aging are shared, and can in principle indicate the level of damage present in an individual's tissues - a measure of biological age. But which epigenetic marks? That is the question. The choice of CpG sites to evaluate, the weight given in the final score to any one site, and the way in which that score is calibrated against test data: all these are ways in which DNA methylation biomarkers can differ from one another.
The development of at least one reliable, accurate biomarker of aging is an important step in the infrastructure needed for rapid progress in rejuvenation biotechnologies. For approaches based on the SENS vision of damage repair, it is straightforward enough to determine how effective a therapy is within its own paradigm. For example, given the ability to clear senescent cells from aged tissues, researchers can immediately follow such a treatment by measuring how great a percentage of senescent cells have been cleared. A senescent cell clearance therapy that clears half of all such cells is better than one that only clears a quarter of them. That doesn't tell us how great an extension of healthy life span will result from the treatment, however. At the present time the only way to assess that outcome is to wait and see. Waiting to see is, unfortunately, expensive and slow: it is an investment of years and millions of dollars in any earnest study, even in mice. That slows down the pace of progress. An independent biomarker such as DNA methylation might be able to short-cut that waiting game by providing a rapid measure of the degree of rejuvenation achieved immediately following the application of an intervention to treat the causes of aging.
What does the methylation status in the DNA reveal about a person's health, his or her susceptibility to disease or, in short, an individual's mortality risk? Researchers investigated the cases of 1,900 participants of two epidemiological studies called ESTHER and KORA. They used DNA from blood cells as the basis of their investigation. All study subjects were older adults and had provided blood samples when they entered the study. This was up to 14 years ago and many of them had died since then. Methyl groups are only attached to a certain combination of DNA building blocks called CpGs. For almost 500,000 of these positions, the researchers analyzed whether their methylation levels revealed a statistical link to survival. After rigorous statistical review, it finally boiled down to 58 CpGs that showed a strong correlation between methylation status and mortality.
These 58 CpGs were all located in genomic regions for which an association with various diseases is well documented. Interestingly, 22 of the 58 CpGs were identical with methylation positions that the researchers had recently found in a study on the epigenetic impacts of smoking. Of all health risk factors, smoking hence appears to leave the strongest tracks in the genome. "The good news is that the level of DNA methylation is not written in stone. Unlike mutations in the DNA building units, it is reversible. That means, for example, that an unfavorable methylation status may change after smoking cessation and the mortality risk may drop again significantly."
Of the 58 CpGs, the scientists selected those ten with the strongest correlation with mortality. This epigenetic risk profile alone enabled them to predict the so-called all-cause mortality (cancer, cardiovascular diseases, and others). Study participants whose genome exhibited an "unfavorable" methylation status at five or more of these sites had a risk of death within the 14-year observation period that was seven times that of study participants whose methylation at these positions showed no abnormalities. "We were surprised that the methylation status of only ten positions of our genome correlates so strongly with all-cause mortality. We found even stronger links to mortality from cardiovascular diseases. Now it is important to find out which prevention measures are most effective to achieve a beneficial impact on the methylation profile and mortality."
DNA methylation (DNAm) has been revealed to play a role in various diseases. Here we performed epigenome-wide screening and validation to identify mortality-related DNAm signatures in a general population-based cohort with up to 14 years follow-up. In the discovery panel in a case-cohort approach, 11,063 CpGs reach genome-wide significance. 58 CpGs, mapping to 38 well-known disease-related genes and 14 intergenic regions, are confirmed in a validation panel. A mortality risk score based on ten selected CpGs exhibits strong association with all-cause mortality, showing hazard ratios of 2.16 (1.10-4.24), 3.42 (1.81-6.46) and 7.36 (3.69-14.68), respectively, for participants with scores of 1, 2-5 and 5+ compared with a score of 0. These associations are confirmed in an independent cohort and are independent from the 'epigenetic clock'. In conclusion, DNAm of multiple disease-related genes are strongly linked to mortality outcomes.
The recently established epigenetic clock (DNAm age) has received growing attention as an increasing number of studies have uncovered it to be a proxy of biological ageing and thus potentially providing a measure for assessing health and mortality. Intriguingly, we targeted mortality-related DNAm changes and did not find any overlap with previously established CpGs that are used to determine the DNAm age. Our findings are in line with evidence, suggesting that DNAm involved in ageing or health-related outcomes are mostly regulated by DNAm regions other than the established age-related DNAm. The difference could also plausibly result from the fact that DNAm age was originally trained as precisely as possible to track chronological age and might thus be more indicative of natural ageing beyond the effect of disease, as exemplified by the much stronger association of DNAm age with mortality in oldest population (mean age 86.1 years) to whom common chronic diseases, such as CVD and cancer, might not continue to pose predominant risks.