UC San Diego

Our group recently developed a quantitative model for human aging based on genome-wide DNA methylation patterns using measurements at 470,000 CpG markers from whole blood1. It is highly accurate at predicting age, and can also discriminate relevant factors in aging, including gender and genetic variants. With this model, there is evidence that metabolic traits such as body mass index and diabetes impact biological aging. In addition, these methylation patterns are strongly correlated with cellular senescence and aging. The top age-predictive CpG marker in our model, and 3 out of the top 12 markers, reside on ELOVL2, a gene involved in elongation of fatty acids. Because of the links between lipid metabolism and aging2–4, we decided to study the relationship between this gene and aging. We found that in WI-38 and IMR-90 human fibroblasts, two common model cell lines of aging, ELOVL2 methylation increases and expression decreases substantially with passage number, and that knocking down ELOVL2 negatively influences cellular aging phenotypes. In mice, we found that mutant ELOVL2 results in an eye aging phenotype which may relate to age-related macular degeneration (AMD). In a screen of aging-related molecules, we identified Aza-2-deoxycytidine (5-Aza) and L-ascorbic acid 2-phosphate (VcP) as modulators of aging phenotypes. We observed amelioration of aging phenotypes, and increased expression and decreased methylation of ELOVL2 upon administration of VcP and 5-Aza in aging cell lines. Furthermore, we observed a reduction of AMD phenotypes upon intraocular injection with 5-Aza. We hypothesize that VcP and 5-Aza can modulate aging phenotypes by influencing DNA methylation, and that ELOVL2 plays a role in aging.