CROI 2020 Abstract eBook

Abstract eBook

Poster Abstracts

virus and then were treated with Texas red-labeled control or anti-HIV-C1C2 fusion proteins decorated EVs. After 24h, cells were analyzed by flow cytometry. TZM-bl cells were in vitro infected with HIVNL4-3P2A-Nef and two days later were treated with labeled EVs, decorated with C1C2 (control) or anti-HIV-C1C2 proteins. Internalization of decorated EVs was assessed after 18h incubation using fluorescence microscopy. Results: Tracking data revealed that HTEVs clustered and moved in tandem with HIV virions, in contrast to negative controls which did not form clusters and tracked independently of virions. HTEVs significantly inhibited HIV infection in Jurkat E6.1 cells (n=3) and CD4+ T cells (n=5 donors) with respect to negative controls (p<0.05, paired t-test). HTEVs efficiently directed EVs into latently- infected, reactivated T lymphoid cells and in vitro HIV-1-infected TZM-bl cells. Conclusion: HTEVs suppress HIV infection and selectively target infected cells ex vivo following latency reversal. HTEVs may facilitate the clearance of the latent HIV reservoir by delivering cytotoxic cargo specifically to infected cells. 358 IN VITRO MODEL FOR STUDYING T-CELL PROLIFERATION/SURVIVAL DRIVEN BY HIV INTEGRATION Machika Kaku 1 , John K. Yoon 1 , John M. Coffin 1 1 Tufts University, Boston, MA, USA Background: The latent HIV reservoir is maintained through clonal expansion of T cells containing latent proviruses. Several clusters of HIV integration sites have been identified in a fraction of highly expanded T cell clones from patients on long-term ART. These clusters are characterized by proviruses in few introns in the same orientation as the host target gene. Similar patterns are hallmarks of insertional mutagenesis by oncogenic retroviruses; in these cases the integrated provirus causes aberrant gene expression, leading to cancer if the target gene is a proto-oncogene. The HIV provirus clusters in expanded cell clones likely reflect the alteration of host gene expression to promote growth/ survival of the host cell, contributing to the persistence of the latent reservoir during ART, and possibly promoting tumor development in infected cells. Methods: Primary human CD4+ T cells were infected with an HIV vector and reactivated every other week for 3 months to allow the T cells to proliferate and rest repeatedly. Each donor sample was divided into three independent replicates. Integration site analysis on randomly fragmented genomic DNA after each reactivation event was used to monitor provirus dynamics: clonal expansion of an infected cell was determined by the observation of multiple DNA fragments with different breakpoints and identical HIV-host junction sites from the Illumina sequencing library. Results: We observed expansion of clones containing specific HIV provirus insertions, supporting our model of T cell expansion during HIV infection. We did not observe expansion of cell with proviruses in the known gene, but we identified large provirus clusters in one small intron of the STAT3 gene associated with extensive clonal expansion. Interestingly, there was a recently reported case of an AIDS-related B cell lymphoma with an HIV integration in the same region and orientation of STAT3. Conclusion: Although in our pilot experiment we did not observe clonal expansion of cells with proviruses in the genes identified in vivo, we did observe significant expansion of cells that contained a provirus in a small region in the STAT3 gene in each of six replicates from two donors. This in vitro systemwill be an important tool for identifying new genes that HIV may disrupt to promote proliferation of the host cell to play a direct role in HIV-related cancer. 359LB AMINOBISPHOSPHONATES REVERSE LATENCY IN HIV-SEROPOSITIVE INDIVIDUALS Sara Selitsky 1 , Matthew Clohosey 1 , Marie Anne Iannone 1 , Susana Garcia-Recio 1 , Jennifer Kirchherr 1 , Carolina Garrido 1 , Yinyan Xu 1 , Shahryar Samir 1 , Ann Marie Weideman 1 , Michael Hudgens 1 , Nilu Goonetilleke 1 , Joel Parker 1 , Natalia Soriano- Sarabia 1 1 University of North Carolina at Chapel Hill, Chapel Hill, NC, USA Background: We hypothesized that aminobisphosphonates (N-BPs), such as pamidronate (PAM), zoledronate (Zol) and alendronate (ALN), that inhibit the formation of farnesyl pyrophosphate groups, used for protein prenylation disrupting cell signaling, may induce reversal of HIV latency. Therefore, here we explore their potential as novel latency reversing agents. Methods: Latency reactivation capacity of N-BPs was analyzed ex vivo in nine HIV-seropositive individuals on suppressive antiretroviral therapy (ART). Resting CD4 T cells were isolated and left untreated or treated with PHA, PAM or Zol, and then cell-associated HIV RNA (caRNA) levels and replication-competent HIV were

measured. RNA-seq was used to explore the N-BPs’ mechanism of action, and flow cytometry was used to analyze the ex vivo effect of N-BPs on immune cell activation and proliferation. Longitudinal PBMC (baseline and weeks 2, 24, 48 post intervention) were obtained from the ACTG A5163. This trial examined the effects of weekly dosing of ALN or placebo (PLB) on low bone density associated with HIV and ART. We measured caRNA levels and total HIV DNA levels. A Wilcoxon matched-pairs signed-rank test was used to analyze patient-specific replicate data across treatment types, and a Mann-Kendall test was used to test for time trend (in either direction) in HIV DNA levels. Results: N-BPs induced reactivation of latent HIV ex vivo (Figure 1A) without causing non-specific activation or other significant alterations on peripheral immune cell populations. RNA-seq analysis showed a correlation between pathways altered by N-BP treatment and those altered following HIV infection (R=0.44, p<0.001). In vivo administration of ALN induced perturbations of the latent reservoir in 8 of 9 participants analyzed who took ALN, that were not detected in participants who took PLB (N=5). Most importantly, treatment with N-BP ALN resulted in a 2.9 to 49.1-fold decrease in total HIV DNA levels in three of eight participants (Figure 1B), while no changes were detected in the PLB arm. Conclusion: We present the first demonstration that N-BPs can reactivate latent HIV in primary CD4 T cells and show that, in vivo, N-BPs can function as latency reversing agents and significantly reduce total DNA levels. These findings, together with the known safety profile of N-BPs, support the need for further clinical testing of N-BPs to reduce persistent HIV infection in vivo.

Poster Abstracts

360 PARALLEL HIV RNA, INTEGRATION SITE, AND PROVIRAL SEQUENCING IN SINGLE RESERVOIR CELLS Kevin Einkauf 1 , Matthew Osborn 2 , Ce Gao2, Jane Blackmer 2 , Chenyang Jiang 1 , Eric Rosenberg 3 , Xu G. Yu 2 , Mathias Lichterfeld 1 1 Brigham and Women's Hospital, Boston, MA, USA, 2 Ragon Institute of MGH, MIT and Harvard, Cambridge, MA, USA, 3 Massachusetts General Hospital, Boston, MA, USA Background: Highly durable latent reservoirs constitute the major barrier to HIV cure. The ability of viral reservoir cells to persist long-termmay depend on the proviral sequence, corresponding integration site and HIV gene expression, but technical limitations have hindered efforts to obtain all three features from single reservoir cells. Here, we describe a novel technology that accomplishes this goal. Methods: PBMC from 2 HIV-infected patients, collected during pre-ART viremia and during suppressive ART, were subjected to a novel assay termed Parallel RNA, Integration Site and Proviral Sequencing (PRIP-Seq). Briefly, PBMC were diluted to single viral reservoir cells, subjected to parallel extraction of cellular DNA and RNA, and exposed to whole-genome (WGA) and whole-transcriptome Results: Paired HIV RNA expression profiles and proviral sequences were determined for 219 total proviruses. HIV transcription was observed in 35% and 31% of cells containing genome-intact and defective proviruses, respectively. Integration sites were simultaneously obtained for 99 of these sequences. Among proviruses with defined integration sites as well as detectable and intact viral promoter regions (n=34), transcriptionally-silent proviruses were 2.8-fold more frequently located in non-genic/pseudogenic regions and were positioned 2.9-fold further away from proximal host transcriptional start sites amplification (WTA). Subsequently, near-full-length proviral sequences, integration sites and the expression of immature and mature HIV RNA transcripts were determined using WGA and WTA products.

CROI 2020 124