The histone deacetylase inhibitor vorinostat (SAHA) increases the susceptibility of uninfected CD4+ T cells to HIV by increasing the kinetics and efficiency of postentry viral events

Abstract

Latently infected cells remain a primary barrier to eradication of HIV-1. Over the past decade, a better understanding of the molecular mechanisms by which latency is established and maintained has led to the discovery of a number of compounds that selectively reactivate latent proviruses without inducing polyclonal T cell activation. Recently, the histone deacetylase (HDAC) inhibitor vorinostat has been demonstrated to induce HIV transcription from latently infected cells when administered to patients. While vorinostat will be given in the context of antiretroviral therapy (ART), infection of new cells by induced virus remains a clinical concern. Here, we demonstrate that vorinostat significantly increases the susceptibility of CD4(+) T cells to infection by HIV in a dose- and time-dependent manner that is independent of receptor and coreceptor usage. Vorinostat does not enhance viral fusion with cells but rather enhances the kinetics and efficiency of postentry viral events, including reverse transcription, nuclear import, and integration, and enhances viral production in a spreading-infection assay. Selective inhibition of the cytoplasmic class IIb HDAC6 with tubacin recapitulated the effect of vorinostat. These findings reveal a previously unknown cytoplasmic effect of HDAC inhibitors promoting productive infection of CD4(+) T cells that is distinct from their well-characterized effects on nuclear histone acetylation and long-terminal-repeat (LTR) transcription. Our results indicate that careful monitoring of patients and ART intensification are warranted during vorinostat treatment and indicate that HDAC inhibitors that selectively target nuclear class I HDACs could reactivate latent HIV without increasing the susceptibility of uninfected cells to HIV.

Importance: HDAC inhibitors, particularly vorinostat, are currently being investigated clinically as part of a "shock-and-kill" strategy to purge latent reservoirs of HIV. We demonstrate here that vorinostat increases the susceptibility of uninfected CD4(+) T cells to infection with HIV, raising clinical concerns that vorinostat may reseed the viral reservoirs it is meant to purge, particularly under conditions of suboptimal drug exposure. We demonstrate that vorinostat acts following viral fusion and enhances the kinetics and efficiency of reverse transcription, nuclear import, and integration. The effect of vorinostat was recapitulated using the cytoplasmic histone deacetylase 6 (HDAC6) inhibitor tubacin, revealing a novel and previously unknown cytoplasmic mechanism of HDAC inhibitors on HIV replication that is distinct from their well-characterized effects of long-terminal-repeat (LTR)-driven gene expression. Moreover, our results suggest that treatment of patients with class I-specific HDAC inhibitors could induce latent viruses without increasing the susceptibility of uninfected cells to HIV.

FIG 1

(A) Combination reporter viruses are produced by cotransfection of 293T cells with plasmids bearing an egfp-containing HIV core lacking an envelope gene (pNL4-3-ΔE-EGFP), bla-Vpr, and an HIV env with known coreceptor tropism. Virions package the bla-Vpr protein and the egfp-HIV core. (B) Fusion between virions and CD4⁺ T cells is detectable by flow cytometry, as bla-Vpr-mediated CCF2 cleavage alters cellular fluorescence. Cells undergoing reverse transcription, uncoating, nuclear import, integration, and LTR-dependent gene expression are identified by Nef-mediated CD4 downregulation and EGFP accumulation. Numbers represent percentages of positive fusion and productive infection events (top and bottom, respectively) within the total number of events recorded. (C) In combination with the memory markers CCR7 and CD45RO, T_N, T_CM, T_EM, and T_TE subsets can be identified. Cells undergoing fusion or LTR-driven EGFP expression are shown in blue.

FIG 2

Vorinostat enhances HIV infection of CD4⁺ T cells in a dose- and time-dependent manner. (A) Four-hour pretreatment of unstimulated primary CD4⁺ T cells with vorinostat enhanced LTR-driven EGFP expression by combination reporter viruses pseudotyped with a patient-derived CXCR4-tropic Env in a dose-dependent manner. (B) Four-hour pretreatment of unstimulated CD4⁺ T cells with 200 nM panobinostat or romidepsin also enhanced EGFP expression. (C) Infection of activated, polarized CD4⁺ T cells and Jurkat cells with reporter viruses bearing CXCR4-tropic Envs is also enhanced by 4-h pretreatment with vorinostat. (D) Pretreatment of cells with vorinostat increases their susceptibility to reporter viruses pseudotyped with VSV-G or CCR5-tropic HIV Env, suggesting enhanced infection is independent of receptor or coreceptor usage. (E) HIV infection of unstimulated primary CD4⁺ T cells was enhanced by 4-h pretreatment with vorinostat compared to addition 24 or 48 h after infection. The data represent means and standard errors of the mean. *, P < 0.05; **, P < 0.01. Exact P values are provided in the text.

FIG 3

Vorinostat does not increase viral fusion with CD4⁺ T cells. (A) Unstimulated primary CD4⁺ T cells were infected with X4- or R5-tropic combination reporter viruses and spinoculated at 1,200 × g for 2 h, and fusion levels were determined by bla-Vpr-mediated CCF2 cleavage. Four-hour vorinostat pretreatment did not affect fusion levels of either X4- or R5-tropic HIV. (B) Vorinostat pretreatment did not increase fusion levels in the absence of spinoculation for either R5- or X4-tropic HIV. Ten- to 15-fold-higher concentrations of HIV were used to compensate for the reduction in fusion in the absence of spinoculation. The data represent means and standard errors of the mean. (C) Pretreatment with vorinostat did not affect the percentages of CD4⁺ T cells in T_N, T_CM, T_EM, or T_TE subsets fusing with HIV but enhanced the likelihood of infection following fusion. Representative data from one of five patients are shown. The numbers represent the percentages of cells undergoing fusion or LTR-driven EGFP expression in each CD4⁺ T cell subset.

FIG 4

Vorinostat enhances the kinetics of postentry viral events, including reverse transcription and integration. Uninfected CD4⁺ T cells were infected with X4-tropic reporter viruses with or without 4-h pretreatment with vorinostat. Prior to infection or at various time points after infection, the coreceptor antagonist AMD3100 (A), fusion inhibitor T20 (B), reverse transcriptase inhibitor efavirenz (C), or integrase inhibitor raltegravir (D) was added. Infection values were normalized to uninfected controls. The graphs represent average levels among three replicates. Error bars are omitted to prevent figure congestion.

FIG 5

Efficiency of postentry viral events, including reverse transcription and nuclear import, is improved by vorinostat. Primary unstimulated (A) and polarized, activated (B) CD4⁺ T cells were infected with X4-tropic reporter viruses and spinoculated. At various time points following infection, the cells were harvested and DNA was extracted for qPCR of 2-LTR circles, followed by deep sequencing to confirm specificity. Product abundances were determined using Genomics Workbench 5. The data were normalized by input DNA and sequencing-chip size.

FIG 6

Vorinostat enhances productive infection and replication of HIV in a spreading infection. Primary CD4⁺ T cells were stimulated with 100 IU/ml IL-2, infected with X4-tropic replication-competent virus, and spinoculated. Cells and supernatants were harvested at days 1, 3, 5, 7, 9, 11, and 13 postinfection. (A) Pretreatment with vorinostat increased the percentage of CD4⁺ T cells infected by replication-competent HIV, as measured by EGFP accumulation using flow cytometry. (B) p24 produced from primary CD4⁺ T cells, as measured by ELISA. Pretreatment with vorinostat dramatically increased the levels of p24 throughout multiple rounds of infection, as evidenced by a steady increase by 13 days postinfection. The data were normalized to ng p24 per 1 × 10⁶ viable cells. The data represent means and standard errors of the mean.

FIG 7

Specific inhibition of HDAC6 by tubacin recapitulates the effect of vorinostat on uninfected CD4⁺ T cells. (A) Four-hour pretreatment of CD4⁺ T cells with 2 μM tubacin did not affect fusion of R5- or X4-tropic combination reporter viruses in the presence or absence of spinoculation. Ten- to 15-fold-higher viral concentrations were used in the absence of spinoculation to compensate for lower fusion levels. (B) Pretreatment of CD4⁺ T cells with 2 μM tubacin increased the percentage of CD4⁺ T cells infected by HIV for R5- and X4-tropic viruses. The data represent means and standard errors of the mean.