Fig 1.
SETD2 activity is not required for HIV infection but regulates HIV expression.
(A.) Schematic of experimental design for panels B-F. (B). Jurkat cells were exposed to EPZ-719 at 500nM for 3 days, then total cell protein lysate was western blotted for H3K36me3 and β-tubulin. (C). EPZ-719 treated or control (DMSO) exposed cells (3 days) were infected with HIV-drEGFP, and infection measured at 2dpi by flow cytometry. (D). Bar chart of HIV infection for EPZ-719 or DMSO exposed cells. (E, F). HIV-dreGFP infected cells were enriched for productively infected cells by flow sporting at 3dpi, then cultured in the presence of EPZ-719 500nM or DMSO for an additional 17 days, and the emergence of a latently infected population (GFP-) was measured by flow cytometry over time. (G). CEM5.25 cells were pretreated with EPZ-719 at 500nM or control (DMSO) for three days, then infected with the replication competent HIV strain NL4-3. Infected cells were then cultured in the presence of EPZ-719 or control for an additional four days. At indicated timepoints, the fraction of infected cells (GFP+) was measured by flow cytometry. Each datapoint represents the average of triplicate samples. Error bars represent the standard deviation of the mean. Asterisk represents a comparison where P<0.05 (T Test).
Fig 2.
EPZ-719 inhibition of HIV expression is reversible.
Jurkat cells were pretreated with EPZ-719 at 500nM or DMSO for three days, then infected with HIV-dreGFP. At 2dpi, productively infected cells (GFP+) were enriched by flow sorting, then cultured for an additional two weeks in EPZ-719 or DMSO. At 2wpi the culture was divided and exposure conditions were reversed for a subset of the cells before an additional week of culture. At 3wpi, the level of H3K36me3 and total H3 was measured by western blot (A), and viral gene expression measured by flow cytometry (B, C). Each datapoint represents the average of triplicate samples. Error bars represent the standard deviation of the mean. Asterisk represents a comparison where P<0.05 (T test).
Fig 3.
H3K36me3-modified histones are present across the HIV proviral genome and are depleted by EPZ-719 exposure.
Jurkat cells were infected with HIV-dreGFP then sorted to enrich for actively infected (GFP+) cells. Infected cells were then cultured in the presence of EPZ-719 at 500nM or control vehicle (DMSO) for 8 days, then cellular protein abundance was analyzed by western blot (A) and viral gene expression measured by flow cytometry (B). The cells were then analyzed by Cleavage Under Targets & Release Using Nuclease (CUT&RUN) using antibodies against H3K36me3, H3K4me3 and control IgG. (C) Normalized coverage of sequencing reads across the HIV reference genome is displayed using Integrative Genomics Viewer version 2.17.
Fig 4.
SETD2 activity is required for HIV expression in primary CD4 T cells.
500nM EPZ-719 was added to resting (A) or activated (B) CD4 T cells for the length of time indicated. Total protein extracts were then harvested and western blotted for H3K36me3 and total histone H3. (C) Schematic overview of primary cell model of HIV latency. CD4 T cells were activated for 2d, then exposed to EPZ-719 or DMSO for an additional 2d before being infected with HIV-dreGFP. At 2dpi, GFP+ cells were then flow sorted and cultured for an additional 8 days in the presence of EPZ-719 (500nM), UNC1999 (1μM), EPZ-719+UNC1999 or DMSO. Viral gene expression was then measured by flow cytometry and the percentage of cells with active expression (GFP+) calculated (D, E). Each datapoint represents the average of triplicate samples. Error bars represent the standard deviation of the mean. Asterisk represents a comparison where P<0.05 (T Test).
Fig 5.
SETD2 knockout in primary CD4 T cells reduces HIV expression.
(A). Schematic overview of SETD2 knockout in primary CD4 T cells. CD4 T cells were activated for 2 days, then infected with HIV-dreGFP. At 2dpi, cells were nucleofected with ribonucleoparticles (RNPs) targeting SETD2, Tat, or non-targeting (NT) control. At 8d post targeting, total cell protein lysate was extracted and western blotted for SETD2, H3K36me3, total histone H3 and α-tubulin (B). Uninfected cells and infected but not nucleofected cells (“No Zap”) are also shown. Viral gene expression was measured by flow cytometry (C,D). Each datapoint represents the average of triplicate samples. Error bars represent the standard deviation of the mean. Asterisk represents a comparison where P<0.05 (T test).
Fig 6.
EPZ-719 affects cellular gene expression but not HIV RNA levels.
HIV-dreGFP infected Jurkat cells that had been exposed to EPZ-719 (500nM) or DMSO for three days before infection and 8 days after infection, were profiled by bulk RNAseq. (A). Volcano plot of differentially expressed genes (DEGs), with significantly downregulated (pvaladj<0.05) genes in blue and upregulated genes shown in red. The three most significantly upregulated and downregulated genes are labeled. Not sig = not significantly different (grey). (B). The upregulated and downregulated genes were analyzed using Enrichr to find enrichment of known protein/DNA binding sites from the ENCODE database within the gene set. (C). The abundance of unique viral reads as a percentage of total unique reads in the dataset was calculated (left panel), and the abundance of Gag viral RNA (Gag-vRNA) was also examined by quantitative PCR (right panel, arbitrary units). The analysis was performed on biological triplicate samples. Error bars represent the standard deviation of mean. NS = not significant, T test (P>0.05).
Fig 7.
EPZ-719 affects HIV RNA splicing.
HIV RNA splicing was quantified by a method in which primers including a random barcode sequence of nucleotides were used for reverse transcription of RNA from HIV-dreGFP infected Jurkat cells exposed to EPZ-719 or DMSO, followed by PCR for the major 4kb and 1.8kb species of viral RNA, and Illumina sequencing. (A). The location of the primer binding sites and the HIV slice donor (D1-D4) and acceptor (A1-A7) sites are shown on a map of the HIV genome. (B). The fraction of vRNAs that are spliced is shown. (C). The percentage of completely spliced vRNAs (1.8kb) is shown. (D). The fraction of spliced transcripts using different initial acceptors is shown. (E). The fraction of spliced transcripts using different final acceptors is shown. Each bar represents the average of three independent biological replicates. Error bars represent the standard deviation of the mean. Asterisk indicates significant difference (P<0.05, Students T test).
Fig 8.
Analysis of HIV integration sites in the presence of EPZ-719.
Genomic DNA derived from Jurkat cells infected with HIV-dreGFP in the presence of EPZ-719 (500nM) or control (DMSO) was used to identify 2531 and 1421 integration sites respectively. For these integration sites, we compared their chromosomal distribution (A), their association with genes (B), the normalized expression level of cellular genes associated with HIV integration sites that were infected in the presence of EPZ-719 or DMSO control (C), and the Spearman correlation between the number of genes and the number of integration sites across chromosomes (D). (E). Each integration site was annotated as belonging to one of 15 different chromatin/transcriptional environments using ChromHMM.
Fig 9.
LEDGF associates with cellular chromatin in a H3K36me3-independent manner.
2D10 cells were exposed to EPZ-719 at 500nM or control vehicle (DMSO) for three days, then fractionated into total cellular lysate, soluble lysate and chromatin-associated proteins. Protein samples were then analyzed by western blot for the indicated targets.
Fig 10.
H3K36me3 depletion enhances sensitivity of latent HIV to reactivation with an HDAC inhibitor.
(A). Schematic overview of experimental design. Jurkat cells were exposed to EPZ-719 (500nM) or DMSO for three days, then infected with HIV-dreGFP. At 48hpi, productively infected cells were isolated by flow sorting, then cultured in EPZ-719 (500nM) or DMSO for an additional two weeks. Latently infected cells (GFP-) were then enriched by flow sorting and stimulated with vorinostat (VOR) (500nM) or prostratin (500nM) in the presence of EPZ-719 (500nM) or DMSO. 24h after stimulation, the reactivation of the latently infected population was measured by flow cytometry. (B, C). Flow cytometry and bar chart of HIV-dreGFP reactivation in response to vorinostat (VOR) or prostratin. (D). Protein extracts from stimulated and control cells were isolated and histone acetylation (H3K9ac and H3K27ac) as well as H3K36me3 were measured for each population by western blot. Each datapoint represents the average of triplicate samples. Error bars represent the standard deviation of the mean. Asterisk represents a comparison where P<0.05 (T test).
Fig 11.
EPZ-719 enhances latency reversal by a vorinostat/EEDi combination in 2D10 cells.
2D10 cells were incubated with 500nM EPZ-719 for three days prior to stimulation for 24h with the small molecules indicated (EED226 = EED inhibitor, AZD5582 = non canonical NF-κB agonist, iBET151 = BRD4 inhibitor, VOR = vorinostat/HDAC inhibitor). Cells were then analyzed for reactivation of latent HIV by flow cytometry (A, B). (C) The abundance of cellular histone modifications was determined by western blot. (D). HIV Gag RNA was quantified by real-time PCR. Statistically significant differences are highlighted (P<0.05, one-way ANOVA).