CROI 2020 Abstract eBook

Abstract eBook

Oral Abstracts

116 EFFICACIOUS RHCMV/SIV VECTORS ELICIT BROADLY CROSS-REACTIVE SIV-SPECIFIC CD8+ T CELLS Benjamin N. Bimber 1 , Shaheed Abdulhaqq 1 , Abigail Ventura 1 , Eric McDonald 1 , Daniel Douek 2 , Scott Hansen 1 , Jonah Sacha 1 , Louis J. Picker 1 1 Oregon Health and Sciences University, Portland, OR, USA, 2 NIH, Bethesda, MD, USA Background: RhCMV68-1/SIV vaccines demonstrate a profound ability to protect against SIV challenge, with half of all vaccinated rhesus macaques clearing viremia shortly after infection. A hallmark of RhCMV68-1 vaccines is the induction of CD8 T cells that are non-classically restricted, either by MHC-II or MHC-E molecules. MHC-E restricted cells are necessary for RhCMV68-1 mediated protection, and characterizing these unconventional cells is essential to understand this unique immune response and to improve vaccine efficacy. Methods: We developed novel single-cell methods to isolate and characterize MHC-E restricted CD8 T cells. CD8+ T cells from RhCMV68-1/SIV vaccinated rhesus macaques were stimulated with antigen in vitro (epitopic peptides or autologous SIV-infected CD4+ T cells) and responding cells were isolated on the basis of surface trapped TNF-a and CD69 expression. Next, we performed single cell RNA-seq (scRNA-seq) using the 10x Genomics platform, which enables simultaneous capture of transcriptome data and TCR clonotype from individual cells. As validation, full length TCR alpha/beta pairs were synthesized and used to transduce CD8 T-cells from SIV-naïve macaque, which were used in similar recognition assays. Results: We characterized MHC-E restricted TCR clonotypic hierarchies from four RhCMV68-1/SIV vaccinated rhesus macaques over more than 2 years. In each animal, a small number of broadly cross-reactive TCRs represents the entire MHC-E restricted response to SIV-infected cells, with a single clone recognizing up to 7 distinct epitopes. TCR alpha/beta transductants replicated the in vivo pattern of antigen recognition. While these TCRs are specific, we further demonstrate the peptide/MHC avidity of these MHC-E restricted clones is significantly lower than conventional MHC-Ia clones. Conclusion: These data indicate that the broad, MHC-E-restricted epitope recognition is accomplished by a small number of T cell clones using highly cross-reactive TCRs with low functional avidity relative to classical responses. These results provide insight into the mechanisms underlying RhCMV/SIV vector efficacy and demonstrate a novel set of methods that could be used to study any T cell population. 117 PD-1 BLOCKADE AT TIME OF ART WITHDRAWAL FACILITATES EARLY POST-PEAK VIRAL CONTROL Afam Okoye 1 , Derick M. Duell 1 , Benjamin Varco-Merth 1 , Morgan Chaunzwa 1 , Matthew Lidell 1 , Hannah Behrens 1 , Jeremy Smedley 1 , Michael K.Axthelm 1 , Scott Hansen 1 , Steven G. Deeks 2 , Nicolas Chomont 3 , Jeffrey D. Lifson 4 , Sharon R. Lewin 5 , Louis J. Picker 1 1 Oregon Health and Sciences University, Portland, OR, USA, 2 University of California San Francisco, San Francisco, CA, USA, 3 Centre de Recherche du CHUM, Montreal, QC, Canada, 4 Frederick National Laboratory for Cancer Research, Frederick, MD, USA, 5 Doherty Institute for Infection and Immunity, Melbourne, VIC, Australia Background: Previous studies evaluating the ability of PD-1 blockade to reduce viral reservoirs in SIV+monkeys on ART have failed to demonstrate significant therapeutic benefit. Here, we evaluated whether PD-1 blockade, around the time of ART discontinuation, could facilitate induction of long-term post-ART control of SIV replication through the functional enhancement of SIV-specific T cells around the time of viral rebound. To address this, we initiated PD-1 blockade in SIV-infected rhesus macaques (RM) receiving ART, starting prior to ART release, with low antigen exposure, or at time of ART release with high antigen exposure. Methods: 30 RM were IV inoculated with 200 ffu of SIVmac239 and after 12 days began receiving ART (tenofovir/emtricitabine/dolutegravir). After sustained virus suppression (<15 RNA copies/ml), RM were randomized into 3 groups (n=10 each) that received; A) 9 biweekly doses of a rhesusized anti-PD1 mAb at 3mg/Kg starting 45 days prior to ART release (low antigen exposure); B) 6 biweekly doses of anti-PD1 starting 3 days prior to ART release (high antigen exposure); or C) an isotype control mAb at the same dosing frequency as Group A. Plasma viral loads (pvl; RNA copies/ml) were quantified by qRT-PCR and T cell dynamics assessed by flow cytometry. Results: Anti-PD1 induced increases in CD4+ and CD8+memory T cell proliferation but had no effect on the frequency of viral blips in the low antigen exposure group vs. controls prior to ART release, suggesting PD-1 blockade did not induce SIV production. Following ART release, all RM rebounded within

Oral Abstracts

115LB HUMAN NK CELLS DEVELOP ANTIGEN-SPECIFIC IMMUNOLOGICAL MEMORY OF HIV Stephanie Jost 1 , Olivier Lucar 1 , Haley Dugan 2 , Marcus Altfeld 3 , Paul Goepfert 4 , R. Keith Reeves 1 1 Beth Israel Deaconess Medical Center, Boston, MA, USA, 2 Ragon Institute of MGH, MIT and Harvard, Cambridge, MA, USA, 3 Heinrich Pette Institute, Hamburg, Germany, 4 University of Alabama at Birmingham, Birmingham, AL, USA Background: Beyond their ability to eliminate infected cells without the need for prior sensitization, murine, non-human primate and human natural killer (NK) cells have been shown to mediate virus-specific recall responses. However, whether NK cells can mediate HIV antigen-specific immunological memory in humans remains to be demonstrated. Methods: Using calcein acetoxymethyl ester- or flow cytometry-based cytotoxicity assays, we tested the ability of bulk and clonally expanded individual NK cells to lyse MHC-devoid K562 cell lines or autologous B-LCLs left unpulsed or pulsed with pools of self-antigen-, CMV-, CMV/EBV/influenza (CEF)-, HIV Gag- or HIV Env-derived overlapping peptides. Study participants included HIV-negative healthy donors (HD; n=6) and people living with HIV (PLWH) virally suppressed on cART (ART; n=14), untreated viremic (UT; n=14) or elite controllers (EC; n=4). Phenotypic analysis was performed using up to 28-color flow cytometry on a BD FACSymphony instrument. Results: Significant anti-HIV Gag activity (range: 8%-50% killing) by bulk NK cells was exclusively detected in half of all PLWH, while killing of BCLs pulsed with the CEF peptide pool, or killing of MHC-devoid K562 cell lines, was comparable between PLWH and HD. NK cells from half of EC had detectable HIV Gag-specific cytotoxic activity and displayed the most robust responses. Strikingly, 35% of all tested NKCL (n=165) generated from 22 PLWH (59% NKCL from 8 ART, 18% NKCL from 14 UT) showed positive responses to HIV (at least twice above killing of unstimulated and above killing of self-peptide-pulsed B-LCLs). Reactive NKCL displayed anti-HIV Gag cytotoxic activity up to 43% specific lysis and anti-HIV Env cytotoxic activity up to 87% specific lysis, within the range of robust cytotoxicity normally found against tumor cells. Phenotypic analysis indicated antigen-specific memory was associated with increased NKG2C and CD57 expression. Accordingly, NKG2C receptor blockade and pulsing with single HIV-derived peptides that bind HLA-E indicated memory NK cell responses likely depend on an HLA-E-dependent recognition mechanism. Conclusion: Collectively, our work presents the first mechanistic evidence for HIV-specific memory NK cells induced by HIV infection in humans. These data suggest that HIV-specific responses mediated by NK cells may have the potential to be harnessed for curative or other therapeutic interventions.

38

CROI 2020