CROI 2017 Abstract e-Book

Abstract eBook

Poster and Themed Discussion Abstracts

An upregulated high mobility group A1 (HMGA1) gene encodes a protein that repressed reporter transcription from HIV LTR in cell lines (Eilebrecht et al., 2013). In the present study, we investigated 1) whether the degree of HMGA1 upregulation by SAHA correlated with levels of HIV reactivation, and 2) whether knocking down HMGA1 would improve the ability of SAHA to reactivate latent HIV. Methods: An in vitro model was used to generate latent HIV infection in resting primary CD4 T cells from 24 different blood donors. Cells were treated for 24h with 1mM SAHA or its solvent dimethyl sulfoxide (control). Expression of HIV and HMGA1 RNA was assessed by droplet digital PCR. Pearson correlation analysis between HMGA1 upregulation and HIV RNA induction was performed using 21 paired SAHA-treated and control samples. To induce HMGA1 knockdown, cells were treated for 3 days with GapmeR reagents from Exiqon, Inc. before SAHA treatment. GapmeRs contain locked nucleic acid base pairs, enter cells by gymnosis without a need for transfection, and induce stable knockdowns via target RNA degradation by RNase H. Samples from 3 blood donors were used to test whether HIV expression following SAHA treatment was greater in the samples with HMGA1 knockdown than untreated ones. Expression in SAHA-treated samples was normalized to expression in DMSO-treated samples, and a t-test was used to determine whether the difference was significant. Results: Fold induction of HIV RNA expression upon SAHA treatment negatively correlated with the induction in HMGA1 expression (R=-0.5, p=0.02). HIV non-responders to SAHA consistently had the greatest upregulation of HMGA1 (Fig 1). GapmeRs induced, on average, 33% knockdown (p=0.03). In the presence of GapmeRs, expression of HIV following SAHA treatment was higher than in the cells incubated in parallel in the absence of GapmeRs (fold change 2.1, p=0.09). Conclusion: To our knowledge, our data are first to demonstrate a possible role of the host factor HMGA1 in HIV reactivation from latency in primary CD4 T cells. The data suggest that inhibiting HMGA1 may improve the ability of SAHA to reactivate latent HIV.

Poster and Themed Discussion Abstracts

308 INVESTIGATING THE MECHANISMS OF HIV-1 LATENCY BY HIGH-RESOLUTION MICROSCOPY Luca Sardo 1 , Angel Lin 1 , Svetlana Khakhina 1 , Lucas Beckman 1 , Weam Elbezanti 1 , Tara Jaison 1 , Harshad D. Vishwasrao 2 , Hari Shroff 2 , Christopher J. Janetopoulos 1 , Zachary A. Klase 1 1 Univ of the Scis, Philadelphia, PA, USA, 3 NIH, Bethesda, MD, USA Background: Biochemical evidence demonstrated that the state of chromatin condensation and the recruitment of regulators of transcription on the HIV-1 LTR can control virus latency during antiretroviral therapy. However, our understanding of these mechanisms in a living cell are limited. Optical microscopy can visualize events in real-time at a single cell level, but the cell body and organelles can hamper the staining and visualization of the nucleus. In facts, high resolution microscopy techniques require fixation of the sample or expression of exogenous fluorescently tagged proteins, which are respectively limited by loss of time resolution and potential artifacts associated with over-expression. Therefore, high resolution microscopy tools to study endogenous nuclear events in real-time are in need. Methods: Transcriptionally competent nuclei were isolated from cell line models of HIV-1 latency and immunostained without fixation for endogenous chromatin markers and chromatin modifying enzymes. Latency reversing agents (LRAs) were supplemented to isolated nuclei to induce chromatin remodeling. State of the art confocal and structural illumination multicolor live microscopy coupled with computational image analysis were used to visualize and quantify chromatin changes in real-time. Results: The spatial organization and accumulation of novel HIV-1 LTR transcription regulators RUNX1 and STAT5 proteins was observed in response to LRAs treatment at unprecedented resolution in unfixed specimens. Microscopy observations were validated by flow cytometry and biochemical assays. In similar experiments we employed CRISPR/ dCas9-SunTag to visualize the HIV-1 LTR and to study its association with chromatin markers in response to virus reactivation. Conclusion: We have developed a high resolution, live imaging approach to investigate the in-vivo molecular mechanisms involved in HIV-1 transcriptional latency. 309 CURRENT LATENCY REVERSING AGENTS REACTIVATE A SMALL FRACTION OF LATENT PROVIRUSES Anthony R. Cillo 1 , Joshua C. Cyktor 1 , Michele Sobolewski 1 , Taylor Buckley 1 , Joseph Hesselgesser 2 , Tomas Cihlar 2 , Romas Geleziunas 2 , Jeffrey Murry 2 , John W. Mellors 1 1 Univ of Pittsburgh, Pittsburgh, PA, USA, 2 Gilead Scis, Foster City, CA, USA Background: Small molecule latency reversing agents (LRAs) are currently under intense investigation for their ability to reactivate latent proviruses, but the fraction of all inducible proviruses that can be reactivated by LRAs is unknown. Methods: Resting CD4+ T (rCD4) cells were isolated by negative selection from participants on long-term ART. The frequency of infected rCD4 cells was evaluated by qPCR targeting HIV-1 pol. rCD4 cells were pulse-treated with LRAs of interest (Table) at pharmacologically achievable concentrations, and then serially diluted and cultured for 7 days. Separate serially diluted rCD4 cells were treated with medium control or anti-CD3/CD28 as negative and positive controls, respectively. Virion production (HIV-1 RNA in supernatants) was measured after 7 days by qRT-PCR (LOD 40 cps/ml). Maximum likelihood estimates were used to quantify the fraction of proviruses that were reactivated to produce virions (Cillo, et al. PNAS 2014). Results: rCD4 cells were isolated from leukapheresis products from 8 donors on suppressive ART for an average of 9 years (range: 3-17), and pulsed with 40 nM RMD for 4 hours, 17.5 nM PAN for 30 minutes, 25 uM JQ1 for 30 minutes, or 10 nM BRY for 1 hour. Pulses with 2 LRAs were at the same concentration and duration as single LRAs. Median fold- increase in virion production over medium control was 5-fold for RMD, 1-fold for PAN, 3-fold for JQ1, 1.7-fold for BRY, 19-fold for BRY+RMD, and 22-fold for RMD+JQ1 compared with 592-fold for anti-CD3/CD28. Despite increases in virion production, only a small fraction of proviruses (between 0.012% and 0.12%) were reactivated with any LRA or combination of LRAs, compared to 3.7% of proviruses reactivated with anti-CD3/CD28 (Table). When the fraction of proviruses that were reactivated is expressed as a percent of the anti-CD3/CD28 positive control (set at 100%), only 1.4% of inducible proviruses were reactivated with RMD, 0.48% for PAN, 0.26% for JQ1, 0.31% for BRY, 3.3% for RMD+JQ1, and 1.8% for RMD+BRY. Conclusion: Although latency reversing agents targeting HDAC, PKC, and bromodomain and extraterminal domain (Brd4) can stimulate virion production from resting CD4 cells, only a very small fraction of proviruses (<<1%) are reactivated to produce virions. Combinations of latency reversing agents increased the fraction of proviruses reactivated, but only to a few percent of that achieved with T cell activation. The “shock and kill” strategy will require latency reversing agents with much greater activity.

CROI 2017 121