No abstract

Despite the success of combination antiretroviral therapy (cART), HIV persists in long lived latently infected cells in the blood and tissue, and treatment is required lifelong. Recent clinical studies have trialed latency-reversing agents (LRA) as a method to eliminate latently infected cells; however, the effects of LRA on the central nervous system (CNS), a well-known site of virus persistence on cART, are unknown. In this study, we evaluated the toxicity and potency of a panel of commonly used and well-known LRA (panobinostat, romidepsin, vorinostat, chaetocin, disulfiram, hexamethylene bisacetamide [HMBA], and JQ-1) in primary fetal astrocytes (PFA) as well as monocyte-derived macrophages as a cellular model for brain perivascular macrophages. We show that most LRA are non-toxic in these cells at therapeutic concentrations. Additionally, romidepsin, JQ-1, and panobinostat were the most potent at inducing viral transcription, with greater magnitude observed in PFA. In contrast, vorinostat, chaetocin, disulfiram, and HMBA all demonstrated little or no induction of viral transcription. Together, these data suggest that some LRA could potentially activate transcription in latently infected cells in the CNS. We recommend that future trials of LRA also examine the effects of these agents on the CNS via examination of cerebrospinal fluid.

BackgroundDifferent classes of latency reversing agents (LRAs) are being evaluated to measure their effects in reactivating HIV replication from latently infected cells. A limited number of studies have demonstrated additive effects of LRAs with the viral protein Tat in initiating transcription, but less is known about how LRAs interact with Tat, particularly through basic residues that may be post-translationally modified to alter the behaviour of Tat for processive transcription and co-transcriptional RNA processing.ResultsHere we show that various lysine and arginine mutations reduce the capacity of Tat to induce both transcription and mRNA splicing. The lysine 28 and lysine 50 residues of Tat, or the acetylation and methylation modifications of these basic amino acids, were essential for Tat transcriptional control, and also for the proviral expression effects elicited by histone deacetylase inhibitors (HDACi) or the bromodomain inhibitor JQ1. We also found that JQ1 was the only LRA tested that could induce HIV mRNA splicing in the absence of Tat, or rescue splicing for Tat lysine mutants in a BRD4-dependent manner.ConclusionsOur data provide evidence that Tat activities in both co-transcriptional RNA processing together with transcriptional initiation and processivity are crucial during reactivation of latent HIV infection. The HDACi and JQ1 LRAs act with Tat to increase transcription, but JQ1 also enables post-transcriptional mRNA splicing. Tat residues K28 and K50, or their modifications through acetylation or methylation, are critical for LRAs that function in conjunction with Tat.Electronic supplementary materialThe online version of this article (10.1186/s12977-018-0421-6) contains supplementary material, which is available to authorized users.