Background
Identifying non-invasive biomarkers of HIV persistence is critical to the development and evaluation of cure strategies. Cell-free DNA (cfDNA), largely released into circulation by dying cells, has proven to be an invaluable biomarker in the oncology space. We recently demonstrated that plasma-derived cfDNA levels and epigenetic features predict HIV rebound kinetics following antiretroviral therapy (ART) cessation. Here, we investigated the relationship between plasma cfDNA features and intact HIV reservoir decay rates during ART.
Methods
Longitudinal blood samples were collected retrospectively from people with HIV (PWH) on ART (n=65) enrolled in the UCSF SCOPE cohort who maintained viral suppression for >2 years. The intact proviral DNA assay (IPDA) was used to measure the HIV reservoir in CD4+ T cells, and intact reservoir decay rates were calculated using random-slope linear regression. cfDNA was isolated from 1 mL of plasma and abundance and fragment sizes were measured using capillary electrophoresis. Baseline cfDNA samples were subject to methylome profiling using an enzyme-based method, followed by whole-genome sequencing on the Illumina Novaseq. Bismark and methylKit software packages were used for alignment to the human genome reference and methylation analyses. ~170,000 CpG dinucleotide sites were considered for correlation analyses, principal component analysis (PCA), cfDNA tissue-of-origin prediction using MethAtlas, and plasma virome profiling using an in-house metagenomics pipeline.
Results
Methylation levels at 5 cfDNA CpG sites were significantly associated with intact HIV reservoir decay rates (p<10-6, Spearman’s rho). The most significant association was observed at chromosome 18, position 62858396 (PHLPP1 gene; Fig. 1A). PCA of the entire cfDNA methylome revealed that principal component 2 was strongly associated with reservoir decay rates (Spearman’s rho=-0.79, p=1.94e-07; Fig. 1B). HIV DNA was detectable in ~47% of cfDNA samples. No significant associations were observed between reservoir decay and plasma cfDNA levels, fragment sizes, tissue of origin, or virome composition.
Conclusions
Measuring methylation levels at specific cell-free DNA CpG sites may enable prediction of latent HIV reservoir clearance rates, facilitating clinical monitoring of cure strategies. Predictive methylation signatures may reveal insights into HIV persistence mechanisms. Methylation of PHLPP1, a regulator of T cell Akt signaling, may impact the reservoir by modulating T cell survival.
Table or Figure
