The Effects of Anti-retroviral Therapy on Epigenetic Age Acceleration Observed in HIV-1-infected Adults

Background: HIV-1 infection is associated with acceleration of age-related methylation patterns in peripheral blood and brain of infected individuals although the relative contributions of HIV-1 infection versus its treatment to the observed accelerations in biological aging have not yet been investigated.

Methods: In this longitudinal study of the effects of antiretroviral therapy (ART) on epigenetic aging patterns, we extracted DNA from peripheral blood mononuclear cells from 15 HIV-1-infected individuals infected at three time points: 6 months-1year pre-ART, 6-12 months post-initiation of ART, and 18-24 months after initiating ART. We compared these trajectories with those of 15 age-matched uninfected control participants at three time points with similar intervals. Methylation studies were performed using the Infinium methylation 450 arrays. We examined four epigenetic clock measurements: Age acceleration residual (AAR), Extrinsic (EEAA), Phenotypic (PEAA), and Grim (GEAA) epigenetic age acceleration. Weighted correlation network (WGCNA) analysis was used to identify clusters of highly co-methylated CpGs.

Results: We found that prior to the initiation of ART all four epigenetic measures were significantly higher in HIV-1-infected individuals compared with uninfected individuals (0.001 for AAR, =0.008 for EEAA, =0.012 for GEAA, <0.001 for PEAA using Wilcoxon rank sum tests between serostatus groups). These effects persisted after the initiation of ART, although the magnitude of these differences diminished. At 18-24 months post-ART initiation (time point 3), PEAA and GEAA were no longer significantly different between HIV-1-infected and uninfected individuals (=0.059 for PEAA, =0.11 for GEAA), while AAR and EEAA remained significantly higher in HIV-1-infected individuals compared with uninfected individuals. We further examined for global patterns of methylation differences between HIV-1-infected and uninfected at each time point, and found 14 groups of co-methylated CpGs that were significantly different between groups at baseline, and remained different after the initiation of ART. Conclusion: We confirm that epigenetic age acceleration associated with HIV-1 infection is most dramatic before ART initiation, and this observation is consistent across four epigenetic clock measurements, as well as in additional groups of co-methylated CpGs identified using WGCNA. Following initiation of ART, there is a partial reduction in age acceleration in all measures, with loss of any significant difference in PEAA and GEAA between serostatus groups. Our findings support the need for future studies examining for a link between epigenetic age acceleration and clinical outcomes in HIV-1-infected individuals.