Abstract Body

Background:

DNA methylation is an epigenetic modification that mediates retroviral epigenetic silencing. However, the role of DNA methylation in HIV-1 proviruses remains controversial. Challenges in assessing DNA methylation in HIV proviruses have included large input DNA requirements due to harsh bisulfite conversion methods, low sampling of proviruses and viral regions due to bisulfite PCR sequencing, and problems of designing and optimizing bisulfite PCR primers accounting for HIV sequence diversity. Here, we address these challenges and develop a novel targeted enzymatic methylation sequencing approach termed HIV TEM-Seq to overcome limitations of prior assays in assessing HIV provirus methylation.

Methods:

HIV TEM-Seq combines enzymatic methylation sequencing (EM-Seq) and target enrichment of full-length autologous proviral sequences. EM-Seq uses TET2 and T4-BGT to convert 5mc and 5hmc into products protected from deamination from APOBEC3A. Unmodified cytosines are converted to uracil enabling discrimination and quantification of DNA methylation at single nucleotide resolution. Autologous HIV sequences are utilized for custom capture probe design enriching for unmethylated and methylated proviral sequences spanning the entire provirus. Next generation DNA sequencing provides >30X coverage of all proviral methylation sites.

Results:

HIV TEM-Seq captured DNA methylation states across the full-length of HIV proviruses in replication competent ACH-2 and U1 cell lines and confirmed increased CpG methylation at the LTR region of single copy integrated provirus in J-Lat 8.4 and 5A8 latent cell lines. Activation of J-Lat cells with anti-CD3/CD28 and PMA/ionomycin significantly decreased CpG methylation at the LTR region. In proviral DNA from blood and CD4 T cells of people living with HIV, we observed increased non-CpG methylation compared to CpG methylation in HIV proviruses. Confirming prior reports, HIV TEM-Seq did not detect CpG nor non-CpG methylation at LTR regions of in vivo proviruses, suggesting the lack of DNA methylation in the LTR likely promotes the presence of a high transcriptionally active reservoir and permissive latency.

Conclusions:

HIV TEM-Seq will have future utility in assaying block-and-lock strategies that aim to deposit DNA methylation using epigenetic editing approaches and evaluating whether DNA methylation increases in proviruses of older people living with HIV on long term ART for decades.