CROI 2019 Abstract eBook

Abstract eBook

Poster Abstracts

331 A NEW PUBLIC DATABASE FOR NEAR FULL-LENGTH HIV SINGLE-GENOME SEQUENCES Wei Shao 1 , Jigui Shan 1 , Wei-Shau Hu 2 , Elias K. Halvas 3 , Brian Luke 1 , John W. Mellors 3 , John M. Coffin 4 , Mary F. Kearney 2 1 Leidos Biomedical Research, Inc, Frederick, MD, USA, 2 National Cancer Institute, Frederick, MD, USA, 3 University of Pittsburgh, Pittsburgh, PA, USA, 4 Tufts University, Boston, MA, USA Background: Despite the success of ART, HIV persists in reservoirs and viremia rebounds if treatment is interrupted. To facilitate understanding of the genetic structure and dynamics of the HIV reservoir, we developed a public database for the storage and meta-analyses of near full-length (NFL) HIV genomic RNA and proviral sequences that persist in donors on ART or that rebound after ART is interrupted. Methods: The database was constructed with MySql, an open source relational database management system. Built-in sequence analysis and annotation tools were programmed in multiple programming languages. We collected HIV-1 NFL proviral sequences from published papers. Patient demographics information is also available where provided. Results: The HIV NFL sequence database is a regularly-updated MySql- based relational database. HIV NFL DNA and RNA sequences obtained by single-genome amplification and sequencing from donors are available for meta-analyses. Sequences can be queried by host characteristics (e.g. gender, duration of infection prior to ART, duration of ART suppression) if provided by the original paper, sample type (e.g. PBMC, lymph node, or CSF), sequence type (DNA or RNA from persistent viremia), or sequence characteristics (e.g. intact, hypermutant, or large internal deletion). Tools for characterizing NFL are available including annotating the sequences, determining intactness, identifying indels, pre-mature stop codons, drug resistance mutations and more. Host integration sites of NFL proviruses are included when available. Users can query the database and analyze the data dynamically through a robust website. Conclusion: Conclusions: This public HIV NFL sequence database and accompanying tools provide easy querying and analyses of HIV DNA and RNA genomes that persist on ART. Meta-analyses of HIV NFL genetics will contribute to our understanding of HIV persistence and may reveal targets for potential future curative interventions. The database will be extended to include other retroviruses as sequences are available. 332 TRANSCRIPTOMIC CORRELATES OF HIV RESERVOIR SIZE DURING ANTIRETROVIRAL THERAPY Wim Trypsteen 1 , Mohamed Abdel-Mohsen 2 , Xutao Deng 3 , Christopher D. Pilcher 4 , Steven G. Deeks 4 , Linos Vandekerckhove 1 , Satish K. Pillai 3 1 Ghent University, Ghent, Belgium, 2 Wistar Institute, Philadelphia, PA, USA, 3 Blood Systems Research Institute, San Francisco, CA, USA, 4 University of California San Francisco, San Francisco, CA, USA Background: HIV is an obligate intracellular pathogen that depends on the cellular machinery of the host to complete its life cycle. Therefore, examining the relationships between the transcriptomes of CD4+ T cells from ART- suppressed HIV-infected individuals and latent reservoir markers can reveal valuable host signatures involved in HIV persistence. Methods: Global transcriptome profiling of blood-derived CD4+ T cells from 69 HIV-infected individuals on suppressive ART for 1-2 years was performed by RNAseq. Genes with counts above 10 Reads Per Kilobase Million (RPKM) across samples were analyzed. Levels of HIV markers were assessed via qPCR assays for cell-associated HIV RNA, total (pol) HIV DNA and 2LTR circles. CD4+ T cell transcriptomes were correlated with reservoir markers and ART initiation timing in a guilt-by-association manner and involved the construction of correlation matrices (Pearson) with subsequent gene set enrichment analysis if a large enough signal was present (>10 genes). Significance was determined using a False Discovery Rate adjusted p-value (q-value) cutoff of 0.05. Gene sets were retrieved from the molecular signature database and include BIOCARTA, KEGG and REACTOME pathways. Results: HIV DNA, RNA and 2LTR markers were significantly correlated with expression levels of 5, 141 and 15 genes, respectively. The top correlated genes with total HIV DNA levels were ST3GAL5 (q<0.0022), and PDCD1 (q<0.0031), both showing a positive correlation. A larger signature was identified for HIV RNA levels, and associated pathways included: NFAT signaling, P53, NOTCH signaling, RIG-I and EDG1-induced T cell impairment. HIV 2LTR levels were correlated with MAD2L2 expression (q<0.0041). ART initiation timing was

correlated with 469 genes and pathways included: mTOR, p53, IL2-induced STAT5 and MYC signaling. Conclusion: Our study revealed biologically intuitive associations between host genetic pathways and HIV persistence. The strong positive correlation between HIV DNA and expression of the PDCD1 gene encoding PD-1 affirms T cell exhaustion as a mechanism underlying HIV persistence. The ST3GAL5 gene product catalyzes the production of GM3 which is a key regulator of cellular proliferation, trafficking, and survival; this finding is consistent with clonal proliferation as a key factor determining reservoir size. These data also demonstrate the profound and sustained effect of early versus late ART on CD4+ T cell function, and identifies several potential targets for immunotherapy.

Poster Abstracts

333 EX VIVO CD4+ T-CELL DIFFERENTIATION IMPROVES HIV RESERVOIR QUANTIFICATION Elizabeth R. Wonderlich 1 , Krupa Subramanian 1 , Ann Wiegand 2 , Carol Lackman-Smith 1 , Michael J. Bale 2 , Mars Stone 3 , Peter Bacchetti 4 , Frank Maldarelli 2 , Michael P. Busch 3 , Roger Ptak 1 , Mary F. Kearney 2 , Deanna Kulpa 5 1 Southern Research Institute, Frederick, MD, USA, 2 National Cancer Institute, Frederick, MD, USA, 3 Blood Systems Research Institute, San Francisco, CA, USA, 4 University of California San Francisco, San Francisco, CA, USA, 5 Emory University, Atlanta, GA, USA Background: Quantifying the number of cells harboring latent, replication- competent HIV provirus is critical in evaluating the efficacy of interventions aimed at reducing the viral reservoir. However, the low frequency of these cells makes this measurement extremely challenging. The quantitative viral outgrowth assay (QVOA) is based on ex vivo activation of resting CD4+ T cells to measure HIV persistence during anti-retroviral therapy (ART). Recent studies have shown that QVOA does not detect all latently infected cells assayed, potentially due to sub-optimal virus reactivation under standard culture and activation conditions. Here, we applied our observation that differentiation into effector CD4+ T cells more effectively promotes HIV latency reversal to improve proviral reactivation in the QVOA, termed differentiation QVOA (dQVOA). Methods: Peripheral blood samples from virally suppressed donors (n=12) were enriched for resting CD4+ T cells and plated in replicate limiting dilution. Cells were then either activated according to the standard QVOA procedure, or induced to differentiate into effector lineages prior to activation. CD4+ T cell phenotypes were assessed by flow cytometry and culture supernatants were evaluated for p24 via ELISA. The frequency of infected cells was calculated using the maximum likelihood method. Single-genome sequencing (SGS) was performed on supernatants to characterize expressed and replicating viral sequences and to compare them to the integrated proviral population. Results: Coupling CD4+ T cell differentiation with activation in dQVOA induced a 14-fold average increase (95% CI 4- to 24-fold) in the estimated frequency of cells with replication competent HIV compared to standard QVOA, indicating that promoting effector lineage differentiation significantly increases

CROI 2019 121