CROI 2019 Abstract eBook

Abstract eBook

Poster Abstracts

in both HIV DNA (n=6, p=0.03) and IUPM (n=10, p=0.002), with no viral outgrowth observed in QVOAs following three of these HIVE assays. Lastly, combinations of LRAs and ABT-199 did not significantly decrease levels of HIV DNA or IUPM (n=10, p=0.17, p=0.18, respectively). Conclusion: This study provides further evidence that ex vivo, latently infected CD4+ T cells exhibit a resistance to CD8+ T cell killing that is not seen in primary cell models of latency. ABT-199 is a clinical stage Bcl-2 inhibitor that, in combination with CD8+ T-cells and LRAs, enabled substantial reductions in HIV reservoirs. However, appreciable levels of ABT-199 induced bystander toxicity emphasize the need for further studies into the mechanisms underlying these observations to develop more targeted approaches. 362 BCL-2/XL ANTAGONISTS REDUCE HIV RESERVOIRS FROM IN VITRO MODELS NOT EX VIVO CD4 CELLS Yanqin Ren 1 , Amanda Macedo 2 , Szuhan Huang 1 , Dora Chan 2 , Ronald Truong 2 , Thomas Rohwetter 2 , Talia Mota 1 , Colin Kovacs 3 , Erika Benko 3 , Alberto Bosque 2 , R. Brad Jones 1 1 Weill Cornell Medicine, New York, NY, USA, 2 George Washington University, Washington, DC, USA, 3 Maple Leaf Medical Clinic, Toronto, ON, Canada Background: HIV cure is obstructed by long-lived viral reservoirs. We previously reported that HIV reservoirs in ex vivo patient CD4+ T-cells are resistant to CTL-mediated elimination. CD4+ TCM cells from HIV-infected individuals express elevated levels of the pro-survival protein Bcl-2, which may contribute to this resistance. We tested the abilities of Bcl-2 family antagonists (also known as senolytics) in combination with the LRA bryostatin to eliminate infected cells from a primary cell model of latency and from ex vivo patient CD4+ T-cells. Methods: Primary CD4+ cell latency models were generated with cells from HIV- donors (‘cultured TCMmodel’, Bosque lab). Ex vivo patient CD4+ T-cells were isolated from leukapheresis samples. HIV Eradication assays (HIVE) were conducted by treating cells with senolytics (ABT-199, A-1155463 or A-1331852) alone, or combined with Bryostatin. Cell viability and phenotypes were assessed by flow cytometry. HIV DNA was measured by digital droplet PCR (ddPCR). Inducible replication competent HIV reservoirs were measured by QVOA and expressed as infectious units per million cells (IUPM). Results: Treatment of cells with the Bcl-2 inhibitor ABT-199 resulted in high levels of non-specific cell death at 1µM, and moderate levels at 100nM. The selective Bcl-xL inhibitors A-1155463 and A-1331852, showed only modest effects on cell viability. Combinations with bryostatin substantially reduced toxicity. Care was taken to standardize IUPM calculations for input of viable cells in each condition. The latency model showed substantial decreases in both HIV DNA and IUPM for both concentrations of ABT-199, and for A-1331852 as single agents (HIV DNA p<0.0001; IUPM p≤0.03), and in combinations with bryostatin (HIV DNA p<0.0001; IUPM p<0.01). In analogous experiments using ex vivo CD4+ T-cells from 6 ARV-treated participants, we did not observe significant decreases in IUPM in any individual, nor across the sample set (p>0.15), with similar results for HIV DNA (p>0.3). Conclusion: Our results are consistent with previous reports in demonstrating the elimination of infected cells from latency models with ABT-199, and extend this to Bcl-xL inhibitors. Unexpectedly, combination with an LRA was dispensable for these effects. These latency-model results were not recapitulated in ex vivo patient CD4+ T-cells, where elimination of reservoir- harboring cells was not detected. It is of interest to determine if the addition of CTL to these combinations may result in reservoir reductions. 363 IN VIVO ANTIVIRAL EFFECT OF DASATINIB IN HUMANIZED MICE INFECTED WITH HIV-1 Maria Salgado 1 , Javier Martinez-Picado 1 , Cristina Gálvez 1 , Sara Rodriguez-Mora 2 , Victor Urrea 1 , Elena Mateos 2 , José Alcamí Pertejo 2 , Mayte Coiras 2 1 IrsiCaixa Institute for AIDS Research, Badalona, Spain, 2 Institute of Health Carlos III, Madrid, Spain Background: Dasatinib is a tyrosine kinase inhibitor currently used for treating chronic myeloid leukemia (CML). It has also been shown to interfere with HIV-1 infection in vitro and ex vivo, mostly by the following mechanisms: 1) preserving the antiviral effect of the innate factor SAMHD1; 2) hindering HIV-1 proviral integration and NF-κB-mediated viral transcription in CD4 T cells isolated from patients with CML on treatment; 3) blocking proliferation of CD4 T cells induced by TCR-mediated stimulus or homeostatic cytokines such as IL-2 and IL-7. In this

1 University of Arizona, Tucson, AZ, USA, 2 National Institute of Allergy and Infectious Diseases, Bethesda, MD, USA, 3 Brigham Young University, Provo, UT, USA Background: Follicular dendritic cells (FDCs) reside in lymphoid follicles and preserve infectious HIV in the form of immune complexes (ICs) on their cell surface. Given the importance of FDCs in HIV-1 pathogenesis, we questioned whether the novel functional cure strategy of an HIV-specific chimeric antigen receptor (CAR) could be applied to deplete FDCs bearing HIV-ICs. In this study we evaluated the capacity of T cells expressing an anti-HIV bispecific CAR (CD4-MBL, containing extracellular human CD4 D1D2 linked to the carbohydrate binding domain of human mannose binding lectin) to recognize and lyse cells bearing HIV-ICs in vitro. Methods: HIV-ICs were prepared using a non-neutralizing antibody to gp160 (Chessie13-39.1) and HIV IIIB . HIV-ICs were used to coat FDCs and CD45 + cells isolated from tonsils of HIV-negative patients. Virus infectivity was verified by co-culture with H9 cells for 3 days followed by QPCR of HIV from culture supernatants. CD4-MBL CAR-T cells were mixed with the following CFSE-labeled cells: FDCs ± HIV-IC, the B cell lines BJAB(env-) and TF228.1.16(env+) ± HIV-IC, and control or HIV-infected autologous CD4+T cells at varying effector:target (E:T) ratios. Target cell lysis was measured using a 4-hour CFSE release assay. CAR-T cell activation was quantified using intracellular flow cytometry to detect CD107a. Results: HIV production by H9 was 14-fold higher when co-cultured with HIV-IC/FDC (15x10 7 virus particles/well) compared to HIV-IC/CD45 + cells (1.1x10 7 particles/well). However no HIV-specific lysis of FDC/HIV-IC was observed in the CFSE release assay at a 3:1 E:T ratio: FDC (6.8±0.8%), FDC/HIV-IC (7.5±0.2%), while prominent specific lysis of TF228.1.16 (env+, 20.2±0.3%) occurred. Similarly in the activation assay, CD4-MBL CAR-T cells failed to respond to bead/HIV-IC (1.7% CD107a + ) compared to bead (1.9% CD107a + ), despite prominent response to bead/anti-CD4 mAb (49.1% CD107a + ). We also found that CAR-T cells were activated by autologous HIV-infected CD4+ T cells (43.9% CD107a + ) but blocked in the presence of mAb to ICAM-1 (11.8% CD107a + ). Conclusion: CD4-MBL CAR-T cells were unable to lyse FDCs bearing infectious HIV-ICs. HIV-ICs failed to induce CAR-T activation. For efficient activation of CAR-T cells, ICAM-1 appeared to be required. Stabilization of cell:cell contact may be necessary for efficient lysis of target cells. 361 BCL-2 INHIBITOR ABT-199 REDUCES RESISTANCE OF HIV RESERVOIRS TO CD8+ T-CELL KILLING Szuhan Huang 1 , Shabnum Patel 2 , Yanqin Ren 2 , Ronald Truong 2 , Dean Magat 2 , Chase McCann 2 , Shuntai Zhou 3 , Sabrina Clark 3 , Elizabeth A. Zale 2 , Amanda Macedo 2 , Talia Mota 2 , Ronald Swanstrom 3 , Alberto Bosque 2 , Catherine M. Bollard 4 , R. Brad Jones 2 1 Weill Cornell Medicine, New York, NY, USA, 2 George Washington University, Washington, DC, USA, 3 University of North Carolina at Chapel Hill, Chapel Hill, NC, USA, 4 Children’s Research Institute, Children’s National Health System, Washington, DC, USA Background: We have recently reported that latently infected CD4+ T cells may resist CD8+ T cell killing through a presently unknown mechanism. Studies by Cummins et al have reported that CD4+ TCM cells from HIV-infected individuals may resist apoptosis due to increased expression levels of Bcl-2. In cancer settings, Bcl-2 antagonists can sensitize tumors overexpressing Bcl-2 to CD8+ T cell killing. Here, we investigate the ability of the Bcl-2 inhibitor ‘ABT- 199’ to reduce the resistance of latently infected cells to CD8+ T cell killing. Methods: Reservoir reduction was assessed by ‘HIV eradication’ (HIVE) assays, where resting CD4+ T cells from ART-suppressed individuals were treated with various combinations of HIV-specific CD8+ T cell effectors, LRAs (bryostatin-1, PMA/Ionomycin, or CD3/CD28 antibodies), and ABT-199. Following treatment, CD4+ T cells were assessed for remaining levels of HIV DNA by ddPCR, and infectious virus (IUPM) by quantitative viral outgrowth assay (QVOA). In “spiked” HIVE assays, a cultured TCM primary cell model of latency was generated with HIV-1 strain NL4-3 and mixed with autologous CD4+ T cells, then treated as above. In all HIVE assays, cell count and viability were monitored by flow cytometry to control for Bcl-2 related toxicity. Results: In “spiked” HIVE assays lower proportions of QVOA wells contained NL4-3 vs patient-virus in conditions treated with HIV-specific CD8+ T cell effectors, suggesting preferential killing of latency model cells. In HIVE assays from 6 participants, combinations of LRAs and CD8+ T cell effectors led to significant decreases in HIV DNA, but not in IUPM (n=7, p=0.02, p=0.3, respectively). The addition of ABT-199 led to consistent, significant decreases

Poster Abstracts

CROI 2019 131