CROI 2019 Abstract eBook

Abstract eBook

Poster Abstracts

372 HIV-1 PROVIRAL CLONE-SPECIFIC DIFFERENCES IN RESPONSE TO LATENCY REVERSING AGENTS

lower after bryostatin, RMD, GD-9620, PMA/ionomycin, and anti-CD3/anti-CD28 beads treatment, respectively. These differences persisted at 48 hours. Conclusion: Latency reversal agents reactivate virus transcription in HIV-2 infection. Whereas levels of HIV-1 and -2 caRNA were similar during latency reversal, when normalized to HIV DNA copy number, statistically significantly less supernatant virus RNA was produced during HIV-2 latency reversal. This suggests that a post-transcriptional block may affect the ability of HIV-2 to produce virions during reactivation from latency. 374 VPU CONTROLS HIV-1 LATENCY THROUGH MODULATION OF THE NF- κ B PATHWAY Gloria Martrus 1 , Rebecca Cornelis 1 , Johannes Brandi 1 , Smitha Srinivasachar Badarinarayan 2 , Gwladys W. Haslé 1 , Annika Niehrs 1 , Jonathan Chow 3 , Leonard U. Hess 1 , Julia K. Bialek 1 , Angelique Hölzemer 1 , Wilfredo García Beltran 4 , Frank Kirchhoff 2 , Daniel Sauter 2 , Marcus Altfeld 1 1 Heinrich Pette Institute, Hamburg, Germany, 2 Ulm University Medical Center, Ulm, Germany, 3 SQZ Biotechnologies, Watertown, MA, USA, 4 Ragon Institute of MGH, MIT and Harvard, Cambridge, MA, USA Background: The long-lived and persistent latent viral reservoir in memory T cells represents one of the main obstacles for an HIV-1 cure. Novel therapies that aim at purging the HIV-1 reservoir using latency reversal agents (LRAs), targeting cellular host proteins, have limited effects in vivo and can induce severe side effects, emphasizing the need for alternative approaches to reverse HIV-1 latency. Accessory HIV-1 proteins play an important role in optimizing viral replication, enabling HIV-1 to evade host restriction and immunity, but their role in regulating HIV-1 latency remains largely unknown. Methods: CRISPR/Cas9 knockout was used to disrupt the accessory HIV-1 genes vpu, nef and vif in the latently HIV-1-infected J89 T cell line. The proportion of HIV-1-GFP+ (i.e. HIV-1 reactivated) cells was traced using flow cytometry and immunofluorescence microscopy. We re-introduced 89.6 Vpu protein and vpu/vpuR45K DNA in J89Δvpu cells using Cell Squeeze® and Amaxa® Cell Line Nucleofector® Kit V technologies, respectively. NK-κB pathway activity was quantified using a dual luciferase reporter assay and ImageStreamX Mark II Imaging Flow Cytometer technique. Results: Our data showed that J89Δvpu cells completely lost control over viral latency, while knockout of nef and vif had no impact. Re-introduction of Vpu protein alone restored HIV-1 latency in a median of 55% of J89Δvpu cells. The proportion of latently HIV-1-infected (HIV-1-GFP-) cells significantly inversely correlated with the proportion of Vpu-FLAG+ J89Δvpu cells. Furthermore, Vpu- FLAG+ J89Δvpu cells showed reduced tetherin surface levels, demonstrating functionality of the Vpu-FLAG protein. As Vpu has been reported to suppress the NF-κB pathway, we measured NF-κB p65 nuclear translocation and observed that J89Δvpu cells showed higher NF-κB activation levels than parental J89 cells. Introduction of vpuR45K, encoding a Vpu mutant that selectively fails to inhibit NF-κB activation, did not affect HIV-1 gene expression in J89Δvpu cells, while re-introduction of wild-type vpu restored HIV-1 latency, indicating a critical role of the NF-κB pathway. Conclusion: These data identified Vpu as a viral protein involved in the maintenance of HIV-1 latency through a modulation of the NF-κB pathway, and provides strong rationale to screen for novel Vpu inhibitors as potential HIV-1 LRAs. 375 FORMATION OF THE REPLICATION-COMPETENT HIV-1 RESERVOIR COINCIDES WITH ART INITIATION Melissa-Rose Abrahams 1 , Sarah B. Joseph 2 , Nigel Garrett 3 , Lynn Tyers 1 , Matthew Moeser 2 , Nancie Archin 2 , Olivia Council 2 , David Matten 1 , Shuntai Zhou 2 , Colin Anthony 1 , Salim Abdool Karim 3 , David M. Margolis 2 , Sergei L. Kosakovsky Pond 4 , Carolyn Williamson 1 , Ronald Swanstrom 2 1 University of Cape Town, Cape Town, South Africa, 2 University of North Carolina at Chapel Hill, Chapel Hill, NC, USA, 3 CAPRISA, Durban, South Africa, 4 Temple University, Philadelphia, PA, USA Background: Although antiretroviral therapy (ART) is highly effective at suppressing HIV-1 replication, the virus persists in a latent reservoir during therapy. The HIV-1 reservoir is present in all HIV-infected people, even when ART is initiated soon after infection, suggesting that it forms early and persists despite long-term viral suppression. Methods: We investigated the temporal origins of the long-lived reservoir in 10 women from the CAPRISA 002 acute infection cohort who initiated treatment in chronic infection. After a median of 4.5 yrs of untreated infection, nine of

Edmond Atindaana 1 , Erin T. Larragoite 2 , Sarah Emery 1 , Kalyani Pyaram 1 , Cleo Burnett 1 , Cheong-Hee Chang 1 , Vicente Planelles 2 , Jeffrey Kidd 1 , Alice Telesnitsky 1 1 University of Michigan, Ann Arbor, MI, USA, 2 University of Utah, Salt Lake City, UT, USA Background: Latent HIV-1 infections are a major obstacle to an HIV-1 cure, and efforts are ongoing to understand and develop strategies to eliminate this reservoir. One of these strategies is the “shock and kill” approach, where virus expression is induced by latency reversing agents (LRAs), allowing infected cells to either expose themselves for clearance. Some studies have shown inconsistency in the potency of LRAs using different models of HIV-1 latency. Since HIV-1 integrates quasi-randomly into active genes, it seems likely that proviruses might be subject to locus-specific gene regulation. Hence, different classes of LRAs may have differential effects on proviral-clones’ response. Methods: Here, we infected Jurkat cells with an Env-Vpr-PuroR virus harboring gfp in the nef ORF to generate a polyclonal proviral pool, with proviruses marked with “zipcodes”-sequence tags within viral sequences that identify clonal progeny of individual integration events. The GFP negative subpopulation was FACS-sorted and treated with different classes of LRAs. Latency reactivation was quantified by the frequency of GFP positive cells, and the total amount of virus released. We also determined the extent of reactivation per proviral clone by quantifying the zipcodes in released virion RNA by high-throughput sequencing for each LRA treatment. Results: Our results suggest that only a fraction of the proviral clones were reactivated by any tested LRA, and clones responded to LRAs to differing extents. Some clones were unique to specific LRAs, with similar classes of LRAs reactivating similar proviral clones. Clonal analysis of the class I-specific histone deacetylase inhibitor (HDACi), entinostat and pan HDACi, SAHA treatments revealed proviral clones that were only reactivatable by SAHA but not entinostat, suggesting HDACs other than class I may play a role in HIV latency. Characterizing individual cell clones revealed differences from the total population’s behavior. For example, while one LRA combination showed additive reactivation when monitored by total virion release in this pool, clonal analysis revealed that a few proviral clones were reactivated to greater extents than they were by one LRA, while other clones’ expression appeared to be reduced by dual LRA Conclusion: The total levels of reactivation can misrepresent the clonal behavior of a latent pool in response to different LRA classes, and mechanisms that act to reactivate one clone may act to silence another. 1 Harvard University, Boston, MA, USA, 2 Brigham and Women’s Hospital, Boston, MA, USA Background: HIV-2 infection is associated with lower plasma virus loads and slower disease progression when compared to HIV-1 infection. Prior work has suggested that, in viremic participants, levels of total viral DNA are similar between HIV-1 and HIV-2 infection but HIV-2 cell-associated RNA (caRNA) levels may be lower. We hypothesized that this difference may extend to virus latency during treated infection and investigated the effects of latency reversal agents (LRAs) on HIV-2 reactivation. Methods: We recruited participants with HIV-1 or HIV-2 infection and isolated PBMCs fromwhole blood. Blood draw volumes precluded the isolation of pure resting CD4+ T cell populations. Reversal of HIV latency was measured ex vivo following 24- or 48-hour exposures to a panel of LRAs with different mechanisms of action: bryostatin 10nM, romidepsin 20nM (RMD), the TLR7 agonist GS-9620 100nM, PMA/ionomycin, and anti-CD3/anti-CD28 beads (1:1 ratio). Total HIV DNA, cell-associated RNA (caRNA), and supernatant viremia levels were determined by validated real-time quantitative PCR (qPCR) assays. Results were normalized for input HIV DNA copy number. Results: While HIV caRNA/DNA ratios were consistently lower for HIV-2 reactivation (3.2 – 11.6, HIV-2; 5.3 – 54.1, HIV-1) for the LRA conditions we tested after 24 and 48 hours of drug exposure, no statistically significant differences were observed in HIV caRNA levels between HIV-1 and HIV-2. Interestingly, levels of supernatant viremia were significantly lower during HIV-2 reactivation, when compared to HIV-1. HIV-2 supernatant viremia levels at 24 hours, corrected for HIV DNA input, were 39-, 77-, 10-, 152-, and 117-fold

Poster Abstracts

373 HIV-2 DYNAMICS DURING LATENCY REVERSAL Fernanda Ferreira 1 , Athe Tsibris 2 , Daniel R. Kuritzkes 2

CROI 2019 134