CBP/p300 lysine acetyltransferases inhibit HIV-1 expression in latently infected T cells

Abstract

HIV-1 latency is regulated by chromatin modifying enzymes, and histone deacetylase inhibitors (HDACi) cause reactivation of provirus expression. Surprisingly, we observed that inhibitors of the CBP/p300 acetyltransferases also cause reversal of latency in T cells. CBP/p300 inhibitors synergize with various latency reversing agents to cause HIV-1 reactivation. In contrast, inhibition of CBP/p300 impaired reversal of latency by the HDACi SAHA, indicating that CBP/p300 must contribute to acetylation on the HIV-1 LTR associated with HDACi-mediated latency reversal. CBP/p300 inhibition caused loss of H3K27ac and H3K4me3 from the LTR, but did not affect association of the inhibitor protein BRD4. Furthermore, inhibition of the additional lysine acetyltransferases PCAF/GCN5 or KAT6A/KAT6B also caused reversal of latency, suggesting that protein acetylation has an inhibitory effect on HIV-1 expression. Collectively, these observations indicate that transcription from the HIV-1 LTR is controlled both positively and negatively by protein acetylation, likely including both histone and non-histone regulatory targets.

Keywords: Biological sciences; Immunology.

Graphical abstract

Figure 1

Inhibitors of CBP/p300 acetyltransferase activity promote HIV-1 transcription (A) Molecular structure of the highly selective CBP/p300 acetyltransferase inhibitors A-485 and iP300w. (B and C) JLat10.6 cells were incubated with the indicated concentration of CBP/p300 inhibitor for 24 h. Subsequently, HIV-1 expression was assessed by flow cytometry and is reported as the GFP delta (Δ) Mean Fluorescence Intensity (MFI) (B) and the proportion of GFP expressing cells (C) (n = 3, mean ± SD, unpaired t test). (D) Cellular viability was determined for JLat10.6 cells treated as in (B) and (C) (n = 3, mean ± SD, unpaired t test). (E) Schematic representation of the Red-Green-HIV-1 full-length dual reporter virus. GFP expression is directed by the 5′ LTR and is a measure of proviral transcription, while an internal constitutive PGK promoter drives mCherry expression allowing for the determination of infection independent of LTR activity. A frameshift mutation in env renders the virus replication incompetent. (F) Representative flow cytometry scatterplots of RGH infected Jurkat E6-1 T cells that 1-day post-infection, were treated with a vehicle control (DMSO), 5 μM A-485, or 5 μM iP300w for 48 h. Latently infected cells are in Q1 (mCherry+), productive HIV-1 infected T cells are in Q2 (GFP+/mCherry+), noise generated by viral recombination is in Q3 (GFP+), and uninfected cells are in Q4. (G and H) Jurkat E6-1 T cells were transduced with RGH at a multiplicity of infection that caused ∼20% of cells to be infected. 24 h post-infection, cells were incubated with a vehicle control (DMSO), or the indicated concentration of A-485 or iP300w. Following 48 h, HIV-1 expression was determined by flow cytometry and is reported as the percent of productively infected cells (Q2/(Q1+Q2)x100) (G) and the GFP Mean Fluorescence Intensity (MFI) of the mCherry+ population (Q1 and Q2) (H) (n = 3, mean ± SD, unpaired t test). Statistical significance is indicated at ∗p < 0.05, ∗∗p < 0.01, or ∗∗∗p < 0.001, with n.s. denoting non-significant p ≥ 0.05.

Figure 2

Effect of CBP/p300 inhibitors in combination with LRAs (A) JLat10.6 cells were pre-treated with the indicated concentration of A-485 or iP300w for 1 h, after which 4 nM PMA was added. Following 24 h, HIV-1 expression was determined by flow cytometry and is indicated as the GFP delta (Δ) mean fluorescence intensity (MFI) (n = 3, mean ± SD, unpaired t test). (B) Drug interaction between 4 nM PMA and the indicated concentration of either A-485 or iP300w was determined using Bliss Independence Modeling. Presented is the difference between the predicted and the observed fractional HIV-1 transcription response to the given drug combination whereby values greater than 0 indicate a synergistic interaction. See Materials and Methods for more details (n = 3, mean ± SD, unpaired t test). (C) JLat10.6 cells were treated with a vehicle control (DMSO), 10 μM A-485, or 10 μM iP300w for 1 h prior to the addition of the indicated latency reversing agent. Following 24 h, HIV-1 expression was assessed by flow cytometry and is reported as the GFP delta (Δ) mean fluorescence intensity (MFI). The concentration of LRAs used were 4 nM PMA, 4 nM PEP005, 1 μM SAHA, 10 μM JQ1, 15 μM IACS-9571 (n = 3, mean ± SD, unpaired t test). (D) Representative flow cytometry scatterplots for JLat10.6 cells treated as indicated in (C). Q1 contains latent (GFP-) cells while cells possessing transcriptionally active provirus are shown in Q2 (GFP+). (E) Bliss Independence Modeling was applied to determine combinatorial drug interaction as in (B). Values greater than 0 indicate synergy while values less than 0 signify an antagonistic relationship (n = 3, mean ± SD, unpaired t test). Statistical significance is indicated at ∗p < 0.05, ∗∗p < 0.01, or ∗∗∗p < 0.001, with n.s. denoting non-significant p ≥ 0.05.

Figure 3

PROTAC-mediated degradation of CBP/p300 reactivates latent HIV-1 (A) JLat10.6 cells were incubated with the indicated concentration of dCBP-1. Following 24 h, whole cell lysates were extracted and subjected to immunoblotting with the indicated antibody. (B) JLat10.6 cells were incubated for 24 h with the indicated concentration of dCBP-1. After 24 h, HIV-1 expression was assessed by flow cytometry and is depicted as the change (Δ) in GFP mean fluorescence intensity (MFI) (n = 4, mean ± SD, unpaired t test). (C) JLat10.6 cells were incubated with a vehicle control or 1 μM dCBP-1. Following the indicated amount of time, HIV-1 expression was assessed by flow cytometry and is reported as the GFP delta (Δ) mean fluorescence intensity (MFI) (n = 3, mean ± SD, unpaired t test). (D) JLat10.6 cells were incubated in the presence of 1 μM dCBP-1 for 3 h. Subsequently, the indicated LRA was added, and HIV-1 expression was examined 24 h later by flow cytometry and is depicted as the delta (Δ) GFP mean fluorescence intensity (MFI). The concentrations of LRAs used were 4 nM PMA, 4 nM PEP005, 1 μM SAHA, 10 μM JQ1, 15 μM IACS-9571 (n = 4, mean ± SD, unpaired t test). (E) Representative flow cytometry scatterplots for JLat10.6 cells treated as indicated in (D). Q1 contains latent (GFP−) cells while the transcriptionally active population is present in Q2 (GFP+). (F) Bliss Independence Modeling was applied to determine combinatorial drug interactions between dCBP-1 and the indicated LRA. Values greater than 0 indicate synergy while values less than 0 signify an antagonistic relationship (n = 4, mean ± SD, unpaired t test). Statistical significance is indicated at ∗p < 0.05, ∗∗p < 0.01, or ∗∗∗p < 0.001, with n.s. denoting non-significant p ≥ 0.05.

Figure 4

CBP/p300 inhibition induces HIV-1 expression in CD4⁺ cells ex vivo (A) Experimental design for RGH infection of primary CD4⁺ T cells. CD4⁺ T cells isolated from an uninfected participant were infected with the RGH dual reporter virus at a low MOI. 3 days, post-infection, cells were incubated with a vehicle control (DMSO), 10 μM A-485, or 10 μM iP300w for 24 h after which proviral expression was assessed by flow cytometry. (B) Representative flow cytometry scatterplots of primary CD4⁺ T cells treated as in (A). Latent infections are in Q1 (mCherry+), productively infected T cells are in Q2 (GFP+/mCherry+), noise generated by viral recombination is in Q3 (GFP+), and uninfected cells are in Q4. (C and D) Following treatment of primary CD4⁺ T cells as in (A), proviral transcription was determined by flow cytometry and is expressed as the percentage of productive infections (C) and the GFP mean fluorescence intensity (MFI) of infected cells (D) (n = 3, mean ± SD, one-way ANOVA with Dunnett’s multiple comparisons test). (E) Viability was determined for participant derived CD4⁺ T cells treated as in (A) (n = 3, mean ± SD, one-way ANOVA with Dunnett’s multiple comparisons test). (F and G) Primary CD4⁺ PBMCs isolated from people living with HIV-1 who are receiving ART were treated with a vehicle (DMSO), 10 μM A-485, or 10 μM iP300w and were either left unstimulated or were treated with anti-CD3/anti-CD28. Following 24 h, intracellular RNA was extracted, and RT-PCR was preformed using oligos specific for multiply spliced Tat-Rev HIV-1 mRNA transcripts. HIV-1 mRNA expression is normalized to GAPDH (n = 3, mean ± SD, unpaired t test). (H) Primary CD4⁺ PBMCs treated as in (F) and (G) were assessed for cellular viability (n = 2, mean, unpaired t test). Statistical significance is indicated at ∗p < 0.05, ∗∗p < 0.01, or ∗∗∗p < 0.001, with n.s. denoting non-significant p ≥ 0.05.

Figure 5

Effect of CBP/p300 inhibition on T cell activation (A) Heatmap depiction of selected upregulated genes following PMA/ionomycin treatment of Jurkat T cells as identified by RNA-seq analysis. (B–G) Jurkat T cells were incubated 24 h in the presence of DMSO (Vehicle), 10 μM A-485, 10 μM iP300w, or 4 nM PMA/1 μM ionomycin. Subsequently, intracellular RNA was extracted and analyzed by RT-PCR using oligos specific for the indicated mRNA transcript. RNA expression is normalized to GAPDH transcript (n = 3, mean ± SD, unpaired t test). (H) Representative flow cytometry scatterplots following surface staining of CD69 or CD25 receptors. Prior to membrane receptor staining, Jurkat T cells were incubated for 24 h with a vehicle control (DMSO), 10 μM A-485, 10 μM iP300w, or 4 nM PMA/1 μM ionomycin. (I and J) Following treatment of Jurkat T cells as in (H), surface staining was performed using PE-Cy7 conjugated anti-CD69. CD69 expression was determined by flow cytometric analysis and is reported as the Mean Fluorescence Intensity (MFI) of PE-Cy7 (I) and the percentage of cells presenting surface CD69 (J) (n = 3, mean ± SD, unpaired t-test). (K.and L) As in (I, J) but surface staining was performed using PE-Cy7 conjugated anti-CD25 (n = 3, mean ± SD, unpaired t-test). Statistical significance is indicated at ∗p < 0.05, ∗∗p < 0.01, or ∗∗∗p < 0.001, with n.s. denoting non-significant p ≥ 0.05.

Figure 6

Tat dependency for CBP/p300 regulation of HIV-1 (A) JLatA72 cells possessing an integrated HIV-1 LTR-GFP reporter that does not express Tat, were incubated 1 h with a vehicle control (DMSO), 10 μM A-485, or 10 μM iP300w prior to the addition of 4 nM PMA, 4 nM PEP005, 1 μM SAHA, 10 μM JQ1, or 15 μM IACS-9571. Following 24 h, LTR transcriptional activity was assessed by flow cytometry and is reported as the GFP delta (Δ) mean fluorescence intensity (MFI) (n = 3, mean ± SD, unpaired t test). (B) Representative flow cytometry scatterplots of JLatA72 cells treated as in (A). Cells harboring transcriptionally silent provirus (GFP−) are located in Q1 while cells with LTR expression (GFP+) are in Q2. (C) Determination of synergy between A-485 and iP300w with the indicated latency reversing agent using Bliss Independence Modeling. Drug concentrations are the same as those used in (A). Data are presented as the difference between the predicted and the observed fractional HIV-1 expression response to the given drug combination. See Materials and Methods for more details (n = 3, mean ± SD, unpaired t test). Statistical significance is indicated at ∗p < 0.05, ∗∗p < 0.01, or ∗∗∗p < 0.001, with n.s. denoting non-significant p ≥ 0.05.

Figure 7

Effect of CBP/p300 inhibition on LTR epigenetic context (A) Schematic representation of the HIV-1 LTR and the primer pairs used for ChIP-qPCR analysis. (B) JLat10.6 cells were incubated with a vehicle control (DMSO), 10 μM A-485, or 10 μM iP300w for 24 h. Subsequently, ChIP was performed using non-specific rabbit IgG and the associated chromatin was analyzed by qPCR (n = 3, mean ± SD, two-way ANOVA with Dunnett’s multiple comparisons test). (C–G) Following treatment of JLat10.6 cells as stated in (B), ChIP was performed using antibodies specific to H3K27ac (C), H3K4me3 (D), H3K9ac (E), H4K5ac (F), or BRD4 (G). Subsequently, enrichment of chromatin was determined by qPCR using oligos directed at the indicated region (n = 3, mean ± SD, two-way ANOVA with Dunnett’s multiple comparisons test). Statistical significance is indicated at ∗p < 0.05, ∗∗p < 0.01, or ∗∗∗p < 0.001, with n.s. denoting non-significant p ≥ 0.05.

Figure 8

CBP/p300 inhibition primes the LTR for RNAPII recruitment (A and B) JLat10.6 cells were treated with a vehicle control (DMSO), 10 μM A-485, or 10 μM iP300w or were pre-treated 1 h with a vehicle control (DMSO), 10 μM A-485, or 10 μM iP300w prior to the addition of 4 nM PEP005. Following 24 h, ChIP was performed using antibodies specific to RNAPII (A) or S2P modified RNAPII (B), and chromatin was assessed by qPCR (n = 3, mean ± SD, two-way ANOVA with Dunnett’s multiple comparisons test). (C and D) JLat10.6 cells were left untreated (Vehicle, DMSO) or were pre-treated 1 h with a vehicle control (DMSO) 10 μM A-485, or 10 μM iP300w prior to the addition of 4 nM PEP005 (C) or 1 μM SAHA (D). Following 24 h, ChIP was performed using anti-H3K27ac and analyzed by qPCR (n = 3, mean ± SD, two-way ANOVA with Dunnett’s multiple comparisons test). Statistical significance is indicated at ∗p < 0.05, ∗∗p < 0.01, or ∗∗∗p < 0.001, with n.s. denoting non-significant p ≥ 0.05.

Figure 9

Inhibitors of various acetyltransferases induce proviral transcription (A) Chemical structure of GSK4027, a highly selective PCAF/GCN5 bromodomain inhibitor, and WM-1119, a highly selective KAT6A/KAT6B acetyltransferase inhibitor. (B and C) JLat10.6 cells were incubated with the indicated concentration of GSK4027 or WM-1119. Following 24 h, HIV-1 transcription was assessed by flow cytometry and is stated as the delta (Δ) GFP mean fluorescence intensity (MFI) (B) and the percentage of GFP positive cells (C) (n = 3, mean ± SD, unpaired t test). (D) JLat10.6 cells were pre-treated with a vehicle control (DMSO), 10 μM GSK4027, or 10 μM WM-1119 for 1 h prior to the addition of the indicated latency reversing agent. After 24 h, HIV-1 expression was assessed by flow cytometry and is reported as the GFP delta (Δ) mean fluorescence intensity (MFI). The concentration of LRAs used were 4 nM PMA, 4 nM PEP005, 1 μM SAHA, 10 μM JQ1, 15 μM IACS-9571 (n = 3, mean ± SD, unpaired t test). (E) Representative flow cytometry scatterplots for JLat10.6 cells treated as indicated in (D). The latent population is in Q1 (GFP−) while the transcriptionally active population is shown in Q2 (GFP+). (F) Bliss Independence Modeling was applied to determine combinatorial drug interaction. Values greater than 0 indicate synergy while values less than 0 signify an antagonistic relationship (n = 3, mean ± SD, unpaired t test). Statistical significance is indicated at ∗p < 0.05, ∗∗p < 0.01, or ∗∗∗p < 0.001, with n.s. denoting non-significant p ≥ 0.05.