An Epigenetic Clock Using Retrotransposon DNA Methylation
An aging clock can be produced via machine learning from any sufficiently large set of biological data that changes with age. Researchers are creating a great many different aging clocks these days, even as there are efforts to standardize to a few generally accepted clocks. The push for standardization is conducted in order to better direct resources in the scientific community to passing the remaining hurdles that prevent clocks from being used to quantify the effectiveness of potential rejuvenation and age-slowing therapies. In principle the assembly of enough mouse and human data coupled with calibration studies for specific interventions could achieve this result for a large fraction of present research and development - but efforts applied to one clock are largely irrelevant to all of the others, so having a panoply of clocks and no consensus on the best is a hindrance.
Reactivation of retroelements in the human genome has been linked to aging. However, whether the epigenetic state of specific retroelements can predict chronological age remains unknown. We provide evidence that locus-specific retroelement DNA methylation can be used to create retroelement-based epigenetic clocks that accurately measure chronological age in the immune system, across human tissues, and pan-mammalian species.
We also developed a highly accurate retroelement epigenetic clock compatible with EPICv.2.0 data that was constructed from CpGs that did not overlap with existing first- and second-generation epigenetic clocks, suggesting a unique signal for epigenetic clocks not previously captured. We found retroelement-based epigenetic clocks were reversed during transient epigenetic reprogramming, accelerated in people living with HIV-1, and responsive to antiretroviral therapy. Our findings highlight the utility of retroelement-based biomarkers of aging and support a renewed emphasis on the role of retroelements in geroscience.