Epigenetic silencing of HIV-1 by the histone H3 lysine 27 methyltransferase enhancer of Zeste 2

Abstract

Latent HIV proviruses are silenced as the result of deacetylation and methylation of histones located at the viral long terminal repeat (LTR). Inhibition of histone deacetylases (HDACs) leads to the reemergence of HIV-1 from latency, but the contribution of histone lysine methyltransferases (HKMTs) to maintaining HIV latency remains uncertain. Chromatin immunoprecipitation experiments using latently infected Jurkat T-cell lines demonstrated that the HKMT enhancer of Zeste 2 (EZH2) was present at high levels at the LTR of silenced HIV proviruses and was rapidly displaced following proviral reactivation. Knockdown of EZH2, a key component of the Polycomb repressive complex 2 (PRC2) silencing machinery, and the enzyme which is required for trimethyl histone lysine 27 (H3K27me3) synthesis induced up to 40% of the latent HIV proviruses. In contrast, there was less than 5% induction of latent proviruses following knockdown of SUV39H1, which is required for H3K9me3 synthesis. Knockdown of EZH2 also sensitized latent proviruses to external stimuli, such as T-cell receptor stimulation, and slowed the reversion of reactivated proviruses to latency. Similarly, cell populations that responded poorly to external stimuli carried HIV proviruses that were enriched in H3K27me3 and relatively depleted in H3K9me3. Treating latently infected cells with the HKMT inhibitor 3-deazaneplanocin A, which targets EZH2, led to the reactivation of silenced proviruses, whereas chaetocin and BIX01294 showed only minimal reactivation activities. These findings suggest that PRC2-mediated silencing is an important feature of HIV latency and that inhibitors of histone methylation may play a useful role in induction strategies designed to eradicate latent HIV pools.

Fig. 1.

The E4 clone model for HIV-1 latency and structure of the HIV-1 LTR. (A) Proviral insertion site in the latently infected Jurkat T-cell clone (E4). The provirus, derived from HIV-1_NL4-3, expresses the wild-type Tat protein and carries the fluorescent d2EGFP protein in place of the Nef gene as a marker of transcriptional activation. The provirus is integrated into the fourth exon of the centromere protein P gene (CENPP) on chromosome 9. (B) Induction of E4 cells by T-cell receptor stimulation or TNF-α activation. Cells were stimulated for 18 to 24 h with 0.125 μg/ml anti-CD3 MAb (left panel), 0.125 μg/ml anti-CD3 MAb plus 1 μg/ml anti-CD28 MAb (middle panel), or 10 ng/ml TNF-α (right panel). Flow cytometry was used to measure expression of d2EGFP, which provided an indirect measurement of HIV transcription.

Fig. 2.

Latent HIV-1 proviruses carry EZH2 and trimethylated histones at the viral LTR. ChIP assays were performed on unstimulated latently infected E4 cells and E4 cells stimulated for 30 min with 10 ng/ml TNF-α. Each sample was analyzed using a series of primers spanning the HIV-1 provirus (Fig. 1). The GAPDH promoter was used as a control. (A) RNAP II); (B) total histone H3; (C) total acetylated histone H3; (D) EZH2; (E) H3K27me3; (F) H3K9me3. Note that EZH2 is located in the Nuc-1 region of latent proviruses. After proviral induction by TNF-α, EZH2 is displaced and there is a concomitant loss of H3K27me3. Error bars represent the standard deviations of 4 separate real-time PCR measurements of the individual ChIP samples.

Fig. 3.

Latent HIV-1 proviruses that respond poorly to cellular activation signals carry high levels of H3K27me3 at their promoters. (A) Experimental scheme. E4 cells were activated for 24 h with a suboptimal level of TNF-α (2 ng/ml) and then sorted into d2EGFP⁺ and d2EGFP⁻ populations. The sorted cells were then expanded over the next 12 days, and d2EGFP expression levels were monitored by flow cytometry. By 12 days both cell populations had reverted to a silenced state. (B) ChIP assays measuring the distribution of RNAP II and chromatin-associated proteins at the HIV promoter (−116 to +4; left panel) and Nuc-1 region (+4 to +134; right panel) in the cells grown from the sorted d2EGFP⁻ and d2EGFP⁺ populations. Error bars represent the standard deviations of 4 separate PCR measurements.

Fig. 4.

Knockdown of EZH2 and SUV39H1 by shRNA. (A) Western blot of EZH2 levels following superinfection of 2D10 cells (left panels) or E4 cells (right panels) with lentiviruses expressing shRNA. Cells were infected with vectors carrying a scrambled control shRNA or shRNA against EZH2 or SUV39H1 or dually infected with a combination of EZH2 and SUV39H1. Whole-cell extracts from the E4 clone infected with the lentiviral vectors expressing the indicated shRNAs were used for Western blot assays. For each cell line, 18 μg of protein in the RIPA cell extract was probed with antibodies against EZH2. GAPDH was used as a loading control. The relative expression levels of EZH2 as estimated by densitometry of the gels are shown below each band. Protein levels were normalized to the GAPDH loading control. (B) Impact of EZH2 knockdown on H3K27me3 expression in E4 cells. Cell extracts were prepared as described above, and Western blots were probed with antibodies to EZH2, SUV39H1, H3K27me3, and GAPDH.

Fig. 5.

Knockdown of EZH2 by shRNA induces HIV-1 transcription in latently infected cells. The cells were superinfected with lentiviral vectors carrying shRNA against SUV39H1, EZH2, or a combination of both vectors. An empty vector was used as a control. Superinfected cells were selected for 4 days using puromycin (for SUV39H1), blasticidine (for EZH2), or a combination of puromycin and blasticidine for the double infections. The FACs profiles were obtained 6 days postinfection. (A) The latently infected Jurkat T-cell clone E4; (B) G4 clone; (C) 2C5 clone, which carries a Nef⁺ provirus; (D) a mixed population of latently infected Jurkat cells.

Fig. 6.

Knockdown of EZH2 by shRNA potentiates HIV-1 emergence from latency. (A) Flow cytometric analysis of the latent E4 clone. The cells were stimulated for 16 h with 0.125 μg/ml anti-CD3 MAb, 0.125 μg/ml anti-CD3 MAb plus 1 μg/ml anti-CD28 MAb, or 10 ng/ml TNF-α. (B) E4 cells were superinfected with a lentiviral vector carrying SUV39H1 shRNA and then stimulated as described for panel A. (C) E4 cells were superinfected with a lentiviral vector carrying EZH2 shRNA and then stimulated as described for panel A. (D) Histogram summarizing flow cytometric analysis results for E4 cells superinfected by control vectors or vectors carrying SUV39H1 shRNA and/or EZH2 shRNA. Error bars represent the standard deviations of 4 separate experiments performed over a period of several months. The cells were either untreated (Minus) or stimulated for 16 h with 0.125 μg/ml anti-CD3 MAb, 0.125 μg/ml anti-CD3 MAb plus 1 μg/ml anti-CD28 MAb, or 10 ng/ml TNF-α, 5 mM HMBA, or 0.5 μM TSA. Data are expressed as the percentage of EGFP⁺ cells (mean fluorescent intensity of >2 × 10¹).

Fig. 7.

Depletion of EZH2 by shRNA blocks the reversion of activated proviruses to a latent state. (A) Representative FACS profiles of the E4 cell populations during proviral shutdown. E4 cells were stimulated with 10 ng/ml TNF-α for 16 h, washed, and then monitored by flow cytometry at 0, 3, 6, 9, 12, 24, 48, and 72 h after removal of TNF-α. (B) Shutdown of E4 cells carrying EZH2 shRNA. Note the slower progression to latency compared to the control cells shown in panel A. (C) Silencing kinetics of E4 cells. The graph shows the mean fluorescent intensity of E4 cell populations treated as described for panel A. A biphasic curve was fitted to the samples. Note that proviral silencing was slowed by the knockdown of EZH2. (D) Silencing kinetics of G4 cells. The experimental protocol was identical to that described for panel C.

Fig. 8.

The EZH2 inhibitor DZNep activates latent HIV-1 proviruses. (A) E4 cells stimulated with 5-Aza-CdR, a DNA methylase inhibitor, for 72 h. (B) E4 cells stimulated with BIX01294 (G9a inhibitor; 16 h). (C) E4 cells stimulated with chaetocin (SUV39H1 inhibitor; 16 h). (D) E4 cells stimulated with DZNep (EZH2 inhibitor; 72 h). Note that DZNep was the most potent compound for inducing latent proviruses.

Fig. 9.

Proviral reactivation by SAHA is enhanced by DZNep or treatment of cells with shRNA to EZH2. (A) E4 cells were treated with 0, 1, or 5 μM SAHA, and d2EGFP expression was measured by flow cytometry. (B) E4 cells were treated for 2 days with 5 μM DZNep and then treated overnight with 0, 1, or 5 μM SAHA. Note that at each SAHA concentration tested, DZNep was able to further stimulate proviral reactivation. (C) Activation of E4 cells exposed to 0 or 500 nM SAHA. There was less than 6% activation of the cells when using this low compound concentration. (D) Activation of E4 cells superinfected with a lentiviral vector expressing EZH2 shRNA and then exposed to 0 or 500 nM SAHA.