In 2012, in collaboration with the laboratory of Drs. Kang Zhang (UCSD) and Stephen Friend (Sage), we built a quantitative model of aging using measurements at more than 450,000 methylation (CpG) markers from the whole blood of 656 human individuals, aged 19 to 101. This model was able to very accurately predict age from the state of the methylome, which we showed was impacted by gender and genetic variants. We found our aging model was upheld in other human tissues, and that it revealed an advanced aging rate in tumor tissue [Hannum et al. Molecular Cell 2013]. 

More recently, we sought to advance the field of epigenetic aging in two ways. First, we partnered with Dr. Howard Fox (formerly TSRI, now University of Nebraska) to explore the effects of viral infection on epigenetic aging rate. In our study, HIV+ men on anti-retroviral therapy had significantly advanced epigenetic aging versus controls [Gross et al. Molecular Cell 2016]. Second, we worked jointly with Dr. Peter Adams (Sanford Burnham Prebys) to study the effects of pro-longevity interventions on epigenetic age in mice. Rapamycin treatment decreased the rate of epigenetic aging, as did caloric restriction and the genetic condition of dwarfism [Wang et al. Genome Biology 2017]. 

Most recently, we have used epigenetic profiles to translate age between humans and dogs [Wang et al., Cell Systems (2020) and BioRxiv (2019)]. Mammals progress through similar physiological stages during life, from early development to puberty, aging, and death. Yet, the extent to which this conserved physiology reflects conserved molecular events is unclear. Here, we map common epigenetic changes experienced by mammalian genomes as they age, focusing on evolutionary comparisons of humans to dogs, an emerging model of aging. Using targeted sequencing, we characterize the methylomes of 104 Labrador retrievers spanning a 16 year age range, achieving >150X coverage within mammalian syntenic blocks. Comparison with human methylomes reveals a nonlinear relationship which translates dog to human years, aligns the timing of major physiological milestones between the two species, and extends to mice. Conserved changes center on specific developmental gene networks which are sufficient to capture the effects of anti-aging interventions in multiple mammals. These results establish methylation not only as a diagnostic age readout but as a cross-species translator of physiological aging milestones.

Figure 2: A non-linear transformation from dog to human age.

Looking for datasets and GitHub source codes?

Visit the Data & Models page for links to Aging Models.

Select Press: