We showed previously that the histone lysine methyltransferase (HKMT) H3K27me3 (EZH2) is the catalytic subunit of Polycomb repressive complex 2 (PRC2) and is required for the maintenance of HIV-1 latency in Jurkat T cells. Here we show, by using chromatin immunoprecipitation experiments, that both PRC2 and euchromatic histone-lysine N-methyltransferase 2 (EHMT2), the G9a H3K9me2-3 methyltransferase, are highly enriched at the proviral 5′ long terminal repeat (LTR) and rapidly displaced upon proviral reactivation. Clustered regularly interspaced short palindromic repeat(s) (CRISPR)-mediated knockout of EZH2 caused depletion of both EZH2 and EHMT2, but CRISPR-mediated knockout of EHMT2 was selective for EHMT2, consistent with the failure of EHMT2 knockouts to induce latent proviruses in this system. Either (i) knockout of methyltransferase by short hairpin RNA in Jurkat T cells prior to HIV-1 infection or (ii) inhibition of the enzymes with drugs significantly reduced the levels of the resulting silenced viruses, demonstrating that both enzymes are required to establish latency. To our surprise, inhibition of EZH2 (by GSK-343 or EPZ-6438) or inhibition of EHMT2 (by UNC-0638) in the Th17 primary cell model of HIV latency or resting memory T cells isolated from HIV-1-infected patients receiving highly active antiretroviral therapy, was sufficient to induce the reactivation of latent proviruses. The methyltransferase inhibitors showed synergy with interleukin-15 and suberanilohydroxamic acid. We conclude that both PRC2 and EHMT2 are required for the establishment and maintenance of HIV-1 proviral silencing in primary cells. Furthermore, EZH2 inhibitors such as GSK-343 and EPZ-6438 and the EHMT2 inhibitor UNC-0638 are strong candidates for use as latency-reversing agents in clinical studies.

IMPORTANCE Highly active antiretroviral therapy (HAART) reduces the circulating virus to undetectable levels. Although patients adhering to the HAART regimen have minimal viremia, HIV persists because of the existence of latent but replication-competent proviruses in a very small population of resting memory CD4⁺ T cells (~1 in 10⁶ cells). Latency remains the major obstacle to a functional cure for HIV infection, since the persistent reservoir almost invariably rebounds within 2 to 8 weeks when HAART is interrupted. In latently infected cells, the HIV genome is stably integrated into the host chromosome but transcriptionally repressed because of epigenetic silencing mechanisms. We demonstrate here that multiple histone lysine methyltransferases play a critical role in both the establishment and maintenance of proviral silencing in cells obtained from well-suppressed patients. Drugs that inhibit these enzymes are available from oncology applications and may find a use in reversing latency as part of a reservoir reduction strategy.