CBL0100 was kindly provided by Cleveland Labs. Recombinant TNFα was (Cat. #BDB554618) purchased from Fisher Scientific. Prostratin (Cat. #P0077) was purchased from Sigma-Aldrich. Dimethyl sulfoxide (DMSO) (BP231-100) was purchased from Fischer Scientific.

Mouse anti-RNA Pol II (Cat. #sc47701), anti-SUPT16H (Cat. #sc165987), and IgG (Cat. #sc2025) were obtained from Santa Cruz biotechnology. Anti-CD3 (Cat. #16003785) and anti-CD28 (Cat. #16029805) antibodies were purchased from Ebioscience.

HIV-1 IIIB and HIV-1 NL4-3 strains were required from the NIH AIDS reagent program. VSV-G pseudotyped HIV-1 NL4-3 luciferase virus (HIV-1 Luc) was created by co-transfecting HEK293T cells with pCG-VSV-G and pNL4-3-Luc (ΔEnv). Transfection was performed using TransIT®-293 transfection reagent (Cat. #MIR 2700, Mirus), according to manufacturer's instruction. Envelope deleted DHIV-nef wild type (wt) and DHIV-nef NF-κB mutant (ΔκB) plasmids were kindly provided by Vicente Planelles (Bosque and Planelles, 2009). VSV-G-pseudotyped DHIV viruses were prepared as described above. Retroviruses expressing HIV-1 Tat (pQCXIP-Tat) were created by co-transfecting HEK293T cells with pQCXIP-Tat, pCG-VSV-G, and pCG-Gag/Pol as described previously (Huang et al., 2015; Power et al., 2015). pQCXIP empty vector was used as a negative control.

Jurkat T cell, JLTRG, J-LAT A2, J-LAT-A1, U1/HIV-1 cells were required from the NIH AIDS reagent program and cultured in RPMI-1640, 10% Fetal Bovine Serine (FBS), 1x penicillin-streptomycin media. Primary CD4⁺ T cells (Sanguine Biosciences, Lonza) and PBMCs were cultured in RPMI-1640, 20 mmol/L HEPES, 1x non-essential amino acids, and 1 × sodium butyrate. Primary CD4⁺ T cells and PBMCs were also supplied with IL2 (30 IU/ml) every 2 days. HEK293 were maintained in Dulbecco's Modified Eagle Medium with 10% FBS.

HIV-1 positive, cART-treated, aviremic patients were recruited through the AIDS clinic at the Strong Memorial Hospital of University of Rochester Medical Center (Rochester, New York) to donate whole blood via leukapheresis. Study subjects were treated with ART for >3 years, had a suppressed viral HIV-RNA level (<50 copies/ml) for at least 6 months, and normal CD4⁺ T cell count (>300 cells/mm³). This study was approved by the University of Rochester Research Subjects Review Board (#RSRB00053667). PBMCs from three de-identified, HIV-1 positive, cART-naïve, viremic patients were provided by James Kobie at URMC. PBMCs from the whole blood of patients were isolated via the Ficoll-Hypaque gradient method. In brief, whole blood was diluted in 1x PBS (without Ca²⁺ or Mg²⁺) at a 1:1 ratio. Then 26 ml of the cell suspension was overlaid onto 14 ml of Ficoll-Paque (GE Healthcare) in a 50 ml conical tube. Samples were then centrifuged for 20 min at 966 × g (without braking and at room temperature). Next, the buffy coat band from the interface was collected and transferred to a new sterile 50 ml conical tube. PBMCs were then washed three times using 1x PBS with 5 mM EDTA. Cells rested in the complete RPMI media in the presence of Nevirapine (600 nM) for 3 days and CD8⁺ T cells were depleted by negative selection using the CD8 MicroBeads (Miltenyi Biotec) following the manufacturer's instruction.

Viability was assessed by measuring ATP levels following CBL0100 treatment using Promega Cell-Titer-Glo Luminescence cell viability kit (G7570), following the manufacturer's instructions. The luciferase signal was measured using the Cytation 5 imaging reader (BioTek).

Supernatants from HIV-1 infected primary CD4⁺ T cells or Jurkat T cells were collected. The protocol of p24 assay provided by the manufacturer (ABL Science) was used to determine viral p24 production (pg/ml) with a minor modification in wash procedure: five washes of 280 μl 1x wash buffer whenever indicated.

J-LAT clones were seeded at 4 × 10⁶ cells/ml. Cells were treated with DMSO or stimulated with 10 ng/ml TNFα, or 1 μM Prostratin. Cells were also co-treated with DMSO or CBL0100 (0.1 μM). Cells were kept in culture for 24 h. Viable cells were then sorted using an Accuri C6 flow cytometer (BD Biosciences). Percentage of GFP+ cells were analyzed by using Flow Jo. The same procedure was used for THP89GFP cells.

Fifty nanograms of p24 of NL4-3 (X4) was used to infect 1 × 10⁶ cells in the presence of DMSO or CBL0100 at a series of concentrations. Cells were washed once with 1 × PBS, re-suspendended in fresh media, and re-treated with CBL0100. At 48 h post-infection, supernatant was collected and the viral load was quantified by p24 ELISA assay.

The Planelles model of HIV-1 latency was used to generate latently infected CD4⁺ T cells with slight modifications (Bosque and Planelles, 2009; Hayashi et al., 2017; Huang et al., 2017). Briefly, naïve CD4⁺ T cells were isolated from two healthy PBMC donors and were stimulated with anti-CD3/CD28 antibodies that were precoated on a Nunc-Immuno Maxi Sorp plate (Thermo Scientific). Cells were incubated with complete medium supplemented with 10 ng/ml of TGFβ, 2 μg/ml of anti-humanIL-12, and 1 μg/ml of anti-human IL-4 (R&D Systems) for 3 days. At day 7, cells were infected with HIV-1 Luc and allowed to revert back to a quiescent/latent state. At day 11, cells were stimulated with anti-CD3/CD28 (+) or mock stimulated with DMSO (−) and co-treated with CBL0100 (0.1 μM) or DMSO for another 3 days. Luciferase signal was measured using One Glo luciferase assay kit (Promega) following the manufacturer's instructions.

ChIP assays followed this previously described protocol (Laspia et al., 1989) with minor changes. Briefly, 1 × 10⁶ U1/HIV-1 cells were activated with 10 ng/ml TNFα, or mock activated with DMSO, and co-treated with either DMSO or CBL0100 (0.1 uM). At 24 h, the cells were cross-linked using 0.5% formaldehyde. After quenching the cross-linking reaction with glycine (125 mM), the cells were washed with cold 1 × PBS and lysed in 1 × CE buffer (10 mM HEPES-KOH, pH 7.9, 60 mM KCl, 1 mM EDTA, 0.5% Nonidet P-40, 1 mM DTT, and protease inhibitor mixture). Cell lysate was centrifuged at 700 × g for 10 min at 4°C to pellet the nuclei. The nuclei pellet was resuspended in 1 × SDS lysis buffer (1% SDS, 10 mM EDTA, 50 nM Tris-HCl, pH 8.1, and protease inhibitor mixture) to a final concentration of 3 × 10⁷ cells/ml, and the nuclear lysate was sonicated for 2 min using a Fisher Scientific™ model 505 sonic dismembrator (1-s on and 1-s off cycles at 50% impulse) to fragment DNA to an average size of ~500 bp. Cellular debris was spun down, and the supernatant was diluted 10-fold with 1 × ChIP dilution buffer (0.01% SDS, 1% Triton X-100, 1.2 mM EDTA, 16.7 mM Tris-HCl, pH 8.1, 150 nM NaCl, and protease inhibitor mixture) and incubated with 5 μg of antibodies against SUPT16H, RNA Pol II (CTD4H8) or control antiserum for overnight with rotation at 4°C. Fifty microliters of protein A/G beads were pre-equilibrated with 0.5 mg/ml BSA (Fisher Scientific) and 0.125 mg/ml calf thymus DNA (Trevigen) for 2 h at 4°C and then added to each sample, which were incubated for another 2 h. Beads were collected and then washed once with each of the following: low salt buffer (0.1% SDS, 1% Triton X-100, 2 mM EDTA, 20 mM Tris-HCl, pH 8.1, 150 mM NaCl), high salt buffer (0.1% SDS, 1% Triton X-100, 2 mM EDTA, 20 mM Tris-HCl, pH 8.1, 500 mM NaCl), and LiCl buffer (0.25 M LiCl, 1% Nonidet P-40, 1% sodium deoxycholace, 1 mM EDTA, 10 mM Tris-HCl, pH 8.1). Beads were then washed twice with 1 × TE buffer (10 mM Tris-HCl, pH 8.1, 0.1 mM EDTA). Immunoprecipitated protein-DNA complexes were eluted twice with fresh elution buffer (1% SDS, 0.1 M NaHCO₃) for 1.25 h at room temperature. Eluates and nuclear lysates (“input”) were heated at 65°C for overnight, to reverse cross-links, in the presence of 0.2 M NaCl. Samples were then treated with 1 μl of 20 mg/ml proteinase K (Life Technologies), 10 μl of 2 M Tris-HCl (pH 6.5), and 10 μl of 0.5 M EDTA for 2 h at 50°C. Released DNA was extracted with phenol/chloroform, precipitated by ethanol, and then resuspended in 100 μl of water. One to three microliters of each sample was used for qPCR using primer sets for amplifying the nuc-1 region of LTR. The qPCR was performed on the CFX Connect™ real time PCR detection system (Bio-Rad), in a 20-μl volume using the following program: 95°C for 1 min and 50 cycles of 95°C for 15 s and 60°C for 30 s.

Quantitative PCR was used to measure the levels of different HIV-1 transcripts as described previously (Zhu et al., 2012). Briefly, 5 × 10⁵ U1/HIV-1 cells were activated with 10 ng/ml TNFα, or mock activated with DMSO, and co-treated with either DMSO or CBL0100 (0.1 μM). Cells were then collected and subjected to mRNA extraction (RNeasy® Mini Kit, Qiagen) and reverse transcription (iScript™ cDNA Synthesis Kit, Bio-Rad). qPCR was conducted using iTaq™ Universal SYBR® Green Supermix (Bio-Rad) on a CFX Connect™ Real-Time PCR System (Bio-Rad), in a 20 μl volume using the following program: 95°C for 1 min and 40 cycles of 95°C for 15 s and 60°C for 30 s. The following qPCR primers were used: initiation, elongation-1, elongation-2, and gag mRNA (see Table S1). GAPDH was used as a reference gene. The same qPCR method was used to quantify HIV-1 gag mRNA levels in DHIV-nef viruses infected Jurkat cells.

Nested qPCR assay to quantify HIV-1 viral load was done as described previously (Mousseau et al., 2015). Specifically, the supernatant of cells subjected to HIV-1 infection or HIV-1 reactivation was collected, and the viral RNA was extracted using a QIAmp® Viral RNA kit (Qiagen) following the manufacturer's instruction. The RNA samples were treated with DNase I (Invitrogen) for 10 min at 25°C, and then inactivated by EDTA at 65°C for 10 min. Reverse transcription coupled qPCR assay was carried out using the Superscript® III One-Step RT-PCR System with Platinum® Taq High Fidelity (Life Technologies) in a total volume of 50 μl. The HIV-1 gag-specific primers P1 and P2 were used for PCR with the following thermal cycles: 50°C for 30 min, 94°C for 2 min, 16 cycles 94°C for 15 s, 62°C for 30 s, 68°C for 60 s, and a final elongation step at 65°C for 5 min. The pre-amplified RT-PCR products were purified via QIAquick® PCR Purification Kit (Qiagen) and a second nested qPCR assay was conducted using TaqMan® Universal PCR Master Mix (Applied Biosystems) in a 25 μl reaction. P3, P4 primers and probe were used for the second PCR. Cycling conditions included UNG incubation (50°C for 2 min), followed by initial denaturation (95°C, 10 min), 45 cycles of amplification (95°C, 15 s; 50°C, 20 s; 60°C, 1 min). All PCR assays were conducted on a CFX connect™ real time PCR detection system (Bio-Rad). A serial dilution of HIV-1 IIIB virus of known concentration was used to create a standard curve for the absolute quantification.

Statistical analysis was performed using Student's t-test or one-way analysis of variance (ANOVA) followed by Tukey's multiple-comparison test (GraphPad Prism 5.0 software). A P-value of < 0.05 was considered statistically significant. Dose response curve was created by nonlinear regression model and the 50% inhibitory concentrations (IC₅₀) was calculated using GraphPad Prism 5.0 software.

Presumably, an ideal LPA should also block acute HIV-1 replication, which may still exist in HIV-positive, cART-treated, aviremic patients, and contributes to residual viremia and/or reservoir replenishment (Darcis et al., 2017). We thus first determined the inhibitory effect of CBL0100 on acute HIV-1 replication in vitro (Figure 1A). The p24 ELISA assay of cell supernatant showed that CBL0100 efficiently inhibits HIV-NL4-3 replication in Jurkat cells in a dose-dependent manner (IC₅₀ = 0.055 μM) (Figure 1B). Cell viability was also measured for CBL0100-treated Jurkat cells, and showed that there was either no or mild cytotoxicity at tested concentrations (0.05, 0.1, 0.2 μM) (Figure S1A). CBL0100 was used at the concentration of 0.1 μM for all subsequent experiments unless otherwise indicated. We next determined the inhibitory effect of CBL0100 on HIV-1 IIIB replication in primary CD4⁺ T cells isolated from three healthy donors. At 3 days post-of-infection (dpi), CBL0100 alone moderately decreased the p24 level in the supernatant of HIV-infected CD4⁺ T cells across the three donors tested, without any cytotoxicity (Figure 1C, Figure S1B). However, the inhibitory effect of CBL0100 almost disappeared at 5 dpi (data not shown). cART potently inhibited viral replication at both time points and was not enhanced by the addition of CBL0100 (Figure 1C). Furthermore, we determined the inhibitory effect of CBL0100 using PBMCs isolated from three HIV-positive, cART-naïve, viremic donors. Overall, CBL0100 alone significantly reduced the viral RNA level in the supernatant of cultured PBMCs across all donors (Figure 1D). Specifically, it had a greater inhibitory effect on donors 1 and 2, while its effect on donor 3 was mild (Figure 1D). The addition of CBL0100 further enhanced cART to block HIV-1 replication for donors 1 and 3 but not donor 2, as cART alone was already potent for donor 2. Collectively, these data suggest that in certain individuals when used in combination with cART, CBL0100 can generate an intensified effect and quickly control viral output during acute infection.

We next decided to determine whether CBL0100 inhibits the reactivation of latently infected HIV-1 proviruses, given that viral reservoirs are a major concern for HIV/AIDS patients (Archin et al., 2014). Two J-LAT cell lines, A1 and A2, were stimulated with TNFα to reactivate latent HIV-1, which was determined by measuring GFP expression. CBL0100 treatment (0.1 μM) led to a decrease in both the GFP+ cell population and GFP mean fluorescence intensity (MFI) in both J-LAT A1 and A2 cells without any cytotoxicity; however, its effect on A2 cells was more striking (Figures 2A,B, Figure S2C). No significant effect of CBL0100 was observed in un-stimulated cells.

In addition to memory CD4⁺ T cells, cells from the monocyte/macrophage lineage may be another source of HIV-1 latent reservoirs (Kumar et al., 2014). We further determined the inhibitory effect of CBL0100 in two monocytic cell line models of HIV-1 latency, THP89GFP and U1/HIV-1 cells. For THP89GFP cells, CBL0100 treatment at the higher but non-toxic concentration (0.2 μM) potently blocked TNFα-induced HIV-1 reactivation, which was observed by measuring GFP+ cell population and MFI (Figure 2C, Figure S2D). For U1/HIV-1 cells, CBL0100 treatment (0.1 μM) resulted in a remarkable decrease of intracellular HIV-1 gag mRNA following TNFα stimulation without any cytotoxicity (Figure 2D, Figure S2E). Furthermore, we determined whether CBL0100 blocks HIV-1 reactivation through the use of different stimulators. We found that CBL0100 inhibited prostratin-induced HIV-1 reactivation in both J-LAT A2 and THP89GFP cells. Additionally, the suppressive effect of CBL0100 on prostratin-induced cells was similar to that of TNFα-induced cells (Figures 2B,C, Figure S2A,B). Together, these results suggest that CBL0100 is a potent LPA, which inhibits HIV-1 reactivation triggered by multiple stimulators in multiple types of reservoir cells.

To further verify the potency of CBL0100 with respect to inhibiting HIV-1 reactivation, we moved our studies from cell lines to primary cells. First, we tested CBL0100 by using a primary CD4⁺ T cell model of HIV-1 latency previously described by Planelles' group (Bosque and Planelles, 2009). In this model, HIV-1 latency is established in non-polarized, memory CD4⁺ T cells infected with VSV-G pseudo-typed HIV-luciferase virus (Figure 3A). HIV-1 reactivation was induced by treating these cells with anti-CD3/CD28 antibodies (+). CBL0100 treatment (0.1 μM) potently inhibited HIV-1 reactivation in anti-CD3/CD28 stimulated CD4⁺ T cells isolated from two healthy donors, which was determined by measuring the intracellular luciferase activity (Figure 3B). CBL0100 treatment also showed similar effects in un-stimulated cells, indicating that it can also block the basal transcription of HIV-1 proviruses. These results, using primary cells, support our hypothesis that CBL0100 can be used as an LPA to block the reactivation of HIV in latent reservoir cells.

To further validate these observations, we examined the effects of CBL0100 on HIV-1 reactivation in latent reservoir cells using PBMCs isolated from HIV-positive, cART-treated, aviremic patients. Specifically, we depleted CD8⁺ T cells from the PBMCs of four HIV-positive, cART-treated, aviremic donors and stimulated them with C+ for a period of 3 days (Figure 3C). CBL0100 (0.1 μM) treatment strikingly suppressed HIV-1 reactivation, reaching ~95% reduction of viral RNA output as compared to the DMSO control (+) for all donors, while presenting minimal cytotoxicity (Figure 3D, Figure S4). Individual suppression ranged from 83 to 99.9% for these donors (Figure S3). On the other hand, un-stimulated cells did not produce a measurable viral RNA output, which is likely due to the resting state of these cells.

Earlier studies of CBL0100 have indicated that it interferes with NF-κB activity through the inhibition of FACT complex (Gasparian et al., 2011). We investigated whether CBL0100 elicits its antiretroviral effect on HIV-1 replication and reactivation through a similar mechanism. We first examined the effects of CBL0100 on HIV-1 5′ LTR promoter activity with or without HIV-1 Tat transactivator, in TZM-bl cells which harbor an integrated LTR-luciferase construct. At the basal level (–Tat), there was a modest reduction of luciferase expression between DMSO and CBL0100 treated cells (Figure 4A). However, in the presence of Tat (+Tat) CBL0100 treatment led to a ~97% reduction of luciferase expression as compared to DMSO (Figure 4A). These results confirm that CB0L100 indeed inhibits the HIV-1 Tat-LTR mediated viral transcription. We next tested whether the inhibitory effect of CBL0100 on HIV-1 transcription is dependent on the NF-κB binding site of the LTR. We determined the drug effect of CBL0100 on the viral expression of VSV-G pseudo-typed HIV-1 DHIV-nef viruses that contain either a wild-type (wt) or a mutated NF-κB binding site (ΔκB) in Jurkat cells (Bosque and Planelles, 2009) (Figure 4B). Interestingly, CBL0100 showed a similar inhibitory effect on the wt and ΔκB viruses, suggesting that CB0100 does not require an intact NF-κB site at the 5′ LTR to suppress HIV-1 transcription; which is different from its anti-tumor activity (Gasparian et al., 2011). For this assay, we also included raltegravir as a negative control, as it does not inhibit post-integrated steps of HIV-1 life cycle.

To further understand how CBL0100 blocks HIV-1 reactivation through inhibition of viral transcription, we turned to the U1/HIV-1 latency model. We measured the different HIV-1 transcripts (initiated or elongated) using specific primer sets in TNFα-stimulated, CBL0100-treated U1/HIV-1 cells. Surprisingly, CBL0100 treatment (0.1 μM) slightly increased the initiated transcript in TNFα-stimulated cells but has no obvious effect in un-stimulated cells (Figure 4C). However, CBL0100 reduced the elongated transcripts in both TNFα-stimulated and un-stimulated cells, with more significant effects observed in TNFα stimulated cells—particularly when comparing the longer elongated transcripts (Figure 4C). These data suggest that CBL0100 may specifically target HIV-1 transcriptional elongation vs. initiation. HIV-1 transcriptional elongation is mainly mediated by RNA polymerase II (Pol II), which overcomes pausing at the nucleosome-1 (nuc-1) region of the HIV-1 5′ LTR promoter (Laspia et al., 1989; Kim et al., 2002). Therefore, we sought to determine the effect of CBL0100 on the occupancy of Pol II at nuc-1 in TNFα-stimulated U1/HIV-1 cells, through the use of a chromatin immunoprecipitation (ChIP) assay. Elongated HIV-1 transcripts are less efficiently produced at basal level and in the absence of Tat (Laspia et al., 1989), while TNFα stimulation enhances their production. Our results consistently showed that TNFα stimulation greatly increased the Pol II occupancy at nuc-1, but CBL0100 treatment (0.1 μM) dramatically blocked such increase (Figure 4D). At the basal condition, without TNFα stimulation, Pol II occupancy at nuc-1 also seemed to be reduced following CBL0100 treatment. Furthermore, Gaspian et al. has previously shown that both Pol II and FACT proteins are depleted at the promoter region of NF-κB dependent genes in TNFα-stimulated, CBL0100-treated tumor cells, and that their transcription is also severely down-regulated (Gasparian et al., 2011).

We next determined the effect of CBL0100 on the occupancy of SUPT16H, a FACT protein, at nuc-1 in TNFα-stimulated U1/HIV-1 cells using another ChIP assay. Consistently, we identified that the occupancy of SUPT16H at nuc-1 substantially decreased with CBL0100 treatment (0.1 μM) at both the basal level and following TNFα stimulation (Figure 4E). Overall, our mechanistic studies suggest that CBL0100 inhibits HIV-1 transcriptional elongation by reducing the occupancy of both Pol II and FACT complex at the 5′ LTR promoter region of HIV-1.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.