CROI 2017 Abstract e-Book

Abstract eBook

Oral Abstracts

60 IMMUNE CORRELATES OF HIV ACQUISITION Krystelle Nganou Makamdop 1 , Slim Fourati 2 , Jianfei Hu 1 , Amy Ransier 1 , Farida Laboune 1 , Rafick-Pierre Sekaly 2 , Daniel Douek 1 1 NIAID, Bethesda, MD, USA, 2 Case Western Reserve Univ, Cleveland, OH, USA

Background: A predictive correlate of susceptibility to HIV infection derived from the baseline inflammatory status of an individual would contribute towards understanding not only the pathogenesis of HIV transmission but also to interpret results from prophylactic vaccine studies. A retrospective study on samples from the HVTN505 trial was set out to investigate immune correlates for HIV acquisition. Because the study surveyed over 2500 at-risk volunteers for several years during which some individuals contracted HIV infection, HVTN505 provides an unequivocally perfect set of samples for a study on biomarkers predictive for HIV acquisition. Methods: Evaluation of multiple plasma and cellular biomarkers was performed in HVTN505 participants who became HIV infected and participants who remained uninfected. All measurements were performed on several time points before infection. Soluble biomarkers measured by ELISA/luminex in plasma included 32 cytokines or chemokines of innate and adaptive cell function. Cellular responses measured in PBMC included frequencies of various innate and T cell subsets and their overall activation status as determined by flow-cytometry (FCM). All PBMC timepoints were sorted into 7 subsets covering innate (mDC, NK, monocytes) and adaptive (CD4 and CD8) cell subsets, which underwent deep sequencing for transcriptome analysis by RNA-sequencing (RNA-Seq). A multi-omics data analysis including RNA-Seq, FCM and ELISA/luminex assessments was conducted to evaluate the set of markers that best predicts HIV acquisition. Results: By themselves, no plasma cytokines/chemokines were able to predict HIV acquisition. Analysis of the FCM data using unsupervised and supervised approaches revealed that an elevated activation status of immune cells prior to infection was associated to an increased risk of infection. RNA-Seq of sorted immune cells revealed the activation of the canonical interferon signaling pathway in particularly mDCs was the best predictor of HIV acquisition (AUC>0.701). An integrative analysis of the RNA-Seq expression and the FCM data revealed a direct positive correlation between activation of the IFN signaling pathway in mDCs and higher frequency of activated effector T-cells, which are markers of increased risk of HIV acquisition. Conclusion: Our study reveals that higher expression of IFN signaling pathway in mDCs leading to an increase of activated effector immune cells several months before infection is identified as a new and strong correlate of HIV acquisition. 61 MARKED CHANGES IN CELLULAR BUT NOT VIRUS TRANSCRIPTOME IN CD4-LOW HIV-INFECTED CELLS Joseph P. Casazza 1 , Eli Boritz 2 , David R. Ambrozak 2 , Amy R. Henry 2 , Farida Laboune 2 , Sam Darko 2 , Jianfei Hu 2 , Christine Grech 2 , Daniel Douek 2 , Richard A. Koup 2 1 NIH, Bethesda, MD, USA, 2 VRC, NIAID, Bethesda, MD, USA Background: Advances in the identification of HIV-infected CD4 T cells and next-generation sequencing nowmake it possible to compare the viral and cellular transcriptome at both early and late time points in infection. Methods: Primary CD4 T cells from 5 different donors were activated with PHA-P, rested for 24 h in the presence of IL-2 and then infected with HIV BAL . Seventy-two hours after infection cells were bulk sorted based on CD4 down-regulation and surface staining of gp120 by PG9 and VRC07. Viral and cellular transcriptomes were characterized by Illumina RNA-seq methodology. Results: Results- The cellular transcriptome of HIV-infected CD4 T cells, when compared to HIV-uninfected cells, was characterized by a pattern of transcription consistent with activation of the NF-κB pathway, and increased transcription of markers of cell cycle progression. Unlike the transcription signature for NF-κB which show no clear effect of CD4 down regulation, markers of cell cycle progression clearly decreased with CD4 downregulation. Additionally, transcription of the T cell master regulatory genes TBX21, RORC, BLC6 and PRDM1 all increased with CD4 down regulation. Based on the frequency of D1 and D4A7 splices, the pattern of HIV RNA splicing did not seem to change significantly with CD4 downregulation, but total HIV reads increased with CD4 down regulation. The median frequency of HIV reads in CD4 bright Env - cells was 0.09(range 0.05-0.18)%, 4.95(0.93- 6.66)% in the CD4 bright Env + cells, 9.2(3.6-10.8)% in the CD4 dim Env + cells and 16.6(12.2-19.9)% in the CD4 null, Env + cells. Conclusion: These data show a pattern of gene transcription consistent with a sequential change in cellular gene expression with progression of infection. These changes do not appear to be due to a change in the splicing pattern of HIV RNA; rather, a dramatic increase in the frequency of intracellular HIV RNA is observed which may be responsible for the changes in the cellular transcriptome observed. Our ability to use bnAbs to identify HIV-infected cells before CD4 down regulation suggest that bnAbs may have therapeutic utility in viremic and latently infected individuals. 62LB SINGLE CELL TRANSCRIPTIONAL PROFILING REVEALS A NOVEL POPULATION OF MUCOSAL TFH CELLS Andrew R. Rahmberg 1 , Premeela A. Rajakumar 1 , James M. Billingsley 2 , Surinder P. Gill 1 , R. Paul Johnson 1 1 Emory Univ, Atlanta, GA, USA, 2 Yerkes Natl Primate Rsr Cntr, Decatur, GA, USA Background: Despite extensive study, our understanding of the molecular events that determine the susceptibility of CD4+ T cells to SIV/HIV infection and the cellular events that follow lentiviral infection have been limited by our ability to track events that occur in single cells and analyze gene expression, including viral gene expression, on a single cell basis. New techniques that permit high-throughput analysis of gene expression in single cells can be used to deconvolute cell types in apparently homogenous populations of bulk cells. Methods: We utilized acutely simian immunodeficiency virus (SIV)-infected rhesus macaques to isolate individual infected and uninfected CD4+ T cells from the intestinal mucosa, the primary site of viral replication in acute infection. Using high-throughput microfluidic quantitative real-time PCR of single cells, we measured expression of five viral transcripts used to define SIV-infected cells along with 91 cellular genes chosen for potential relevance in the viral replication cycle. Results: Single cell analysis of over 300 single jejunal CD4+ T cells obtained 10 days after intravenous SIV infection revealed that approximately 20% of these cells were SIV- infected. Comparison of gene expression using multiple statistical methods identified PD-1 and CXCR5 as being the most significantly differentially expressed genes between infected and uninfected cells. The coexpression of PD-1 and CXCR5 on CD4+ T cells defines T follicular helper (Tfh) cells. However, Tfh have been classically associated with secondary lymphoid tissue. Flow cytometric analysis of jejunal samples from uninfected macaques identified a distinct population of PD-1+ CXCR5+ CD4+ T cells, with multiple phenotypic characteristics of classical Tfh cells, including expression of BCL-6 and IL-21. Transcriptional profiling of a panel of 70 Tfh-associated genes verified the similarity of this novel population to classical Tfh. PD-1+ CXCR5+ cells from jejunum contained an average of 3.4 SIV gag DNA copies/cell at the peak of acute infection. This level of infection over 10-times higher than that of bulk memory CD4+ T cells was observed despite low levels of cell surface expression of the SIV coreceptor CCR5. Conclusion: This study is the first single cell gene expression analysis of primate lentivirus-infected cells, and identified a novel and highly susceptible target cell population in vivo during acute infection. 63 HIGH HIV PERMISSIVITY OF T FOLLICULAR REGULATORY CELLS IS RELATED TO Ki67 EXPRESSION Shannon Miller 1 , Brodie Miles 1 , Joy Folkvord 2 , Amie Meditz 3 , Martin McCarter 1 , Mario Santiago 1 , David N. Levy 4 , Samantha MaWhinney 1 , Elizabeth Connick 2 1 Univ of Colorado Denver, Aurora, CO, USA, 2 Univ of Arizona, Tucson, AZ, USA, 3 Boulder Community Hosp, Boulder, CO, USA, 4 New York Univ Coll of Dentistry, New York, NY, USA Background: HIV replication is concentrated within follicular T cells in B cell follicles (F) during asymptomatic disease. Whether T follicular regulatory cells (TFR) are more permissive than other subsets and what factors drive permissivity are not understood. Methods: Disaggregated HIV- human tonsil cells were spinoculated with R5-tropic GFP reporter virus (NLYUV3-GFP), 3 R5-tropic transmitted/founder viruses (T/F) (CH58, CH470, CH40) or mock spinoculated at day 0 and cultured for 2 days. Some cells were cultured for 2 days alone or with imatinib, a T cell proliferation inhibitor, prior to spinoculation. GFP or p24 expression were determined by flow cytometry. Due to CD4 downregulation by HIV, we gated on CD3+CD8- cells, and defined subsets as: extrafollicular (EF; CXCR5-CD25- CD127+/-), EF Treg (CXCR5-CD25+CD127-), TFH (CXC5+CD25-CD127+/-) and TFR (CXCR5+CD25+CD127-). CCR5, HLA-DR/CD38, CD69 and Ki67 expression were determined prior to

Oral Abstracts

24

CROI 2017