Abstract Body

Background: A subset of aging individuals with chronic, treated, HIV infection develop neurocognitive decline. Epigenetic factors, such as DNA methylation, have been implicated in the aging process and the development of chronic diseases. We hypothesize that long-term treated HIV infection has had a detectable impact on DNA methylation in these individuals, and this epigenetic change is associated with their cognitive impairment.

Methods: Within the Pitt Mens Study, we identified 8 men with cognitive decline (3 seropositive and 5 seronegative), and matched them with 8 seropositive controls (matched on age and length of time seropositive) and 8 seronegative controls (matched on age). For each of these 24 men we obtained a PBMC pellet from a recent clinic visit and an archival PBMC collected on average ten years earlier. The DNA methylation profile was then determined for each time point using the Illumina HumanMethylation450K microarray, and the change in methylation status at each of the 480,000 CpG sites present on the microarray was calculated.

Results: Over the ten-year time period, DNA methylation levels changed little in any of the seropositive and seronegative control samples. In stark contrast, DNA methylation was drastically perturbed in every seropositive and seronegative sample with cognitive decline. The CpG sites affected are different in each individual: no consistent Methylation Variable Positions (MVPs) or Differentially Methylated Regions (DMRs) were found by bioinformatic analysis, and pairwise comparisons showed that the magnitude and direction of the methylation change at each site varied greatly between individuals. The perturbations were similar in both seropositive and seronegative samples, but the average age of onset of cognitive decline was 10.5 years younger in the seropositive samples.

Conclusions: We determined that the development of cognitive decline in our samples is accompanied by a chaotic change in DNA methylation. This may be due to changes in the populations of cells that make up the PBMC samples obtained for each person, but it is more likely that this “methylation entropy” reflects random changes in DNA methylation in the genome of affected individuals. It is not clear what initial events trigger this chaotic change, nor is it clear how this increase in entropy affects cognition, but we suggest that DNA methylation may be a biomarker of risk for cognitive decline, and should also be investigated in other diseases of aging in HIV-positive individuals.