CROI 2017 Abstract e-Book

Abstract eBook

Poster and Themed Discussion Abstracts

de novo to reconstruct the HIV genome. Integration site analysis was performed in two steps. First, reads were mapped to a HIV reference to profile the LTR. Any reads containing an LTR end motif then had the LTR sequence removed, the remaining human sequence was mapped to the hg19 genome using Novoalign to determine the integration locus. Results: Near full-length sequence was obtained frommixes of uninfected and infected cell lines and then from 6 HIV infected patients with a range of total HIV DNA as measured by qPCR (1669 – 13186 copies per million CD4 T-cells). The depth of sequencing coverage achieved for each sample was positively correlated with HIV-DNA (p=0.013). The mean sequencing coverage for patient samples was 6x. Consistent with current literature, all primary cell integration sites mapped to intronic regions of the human genome. Conclusion: We present the first Next Gen-based enrichment protocol which allows near full-length proviral HIV DNA sequence analysis from a broad range of HIV total DNA with integration site identification. The preliminary study of six patients supports the current understanding that a high frequency of deletions are present in the reservoir and that HIV preferentially integrates within intronic regions of highly expressed genes. A qualitative sequence-based correction of qPCR (“q²PCR”) is likely to provide a more accurate reflection of the true reservoir size. Additionally, a comprehensive map of annotated integration sites could also help identify cells most likely to constitute the reservoir. 302 HIV-1 INTEGRATION SITES ANALYSIS REVEALS DIFFERENT COMPARTMENTS OF THE HIV RESERVOIR Eva Malatinkova 1 , Wei Wang 2 , Stephanie Laufs 2 , Ward De Spiegelaere 1 , Ning Wu 2 , Saira Afzal 2 , Raffaele Fronza 2 , Sabine Kinloch-de Loes 3 , Manfred Schmidt 2 , Linos Vandekerckhove 1 1 HIV Cure Rsr Cntr, Ghent Univ, Ghent, Belgium, 2 German Cancer Rsr Cntr and Natl Cntr for Tumor Diseases, Heidelberg, Germany, 3 Royal Free Hosp, London, UK Background: Silently persisting latent HIV-1 reservoir remains a major obstacle to an HIV-1 cure. Comparison of the viral reservoir composition by identifying the HIV-1 integration sites (IS) and clonality between blood and tissue and between different patient cohorts will lead to unreaveling the mechanisms of reservoir persistence and guide therapeutic approaches towards a functional cure. Methods: Viral IS were identified in 15 patients of a cross-sectional study, enrolled in two clinical centers (Ghent, BE and London, UK) including three patient cohorts: early treated patients with antiretroviral therapy (ART) started during seroconversion (Early ART, n=5), patients with late ART initiation during chronic phase of HIV-1 infection (Late ART, n=5) and long-term non-progressors (LTNP, n=5). Patients within Early ART and Late ART cohorts were treated uninterrupted for a median of 10 years with undetectable viral load for at least 4 years. Viral IS were determined by linear amplification mediated PCR (LAM-PCR) both in peripheral blood mononuclear cells (PBMCs) and rectal biopsies. A semi- quantitative estimation of clonal size was done by determining the number of retrieved sequences (retrieval frequency) of individual vector-genome junctions (cell clones). The relative sequence count of all detected IS was calculated in relation to all sequences which could be mapped to a definite position in the genome. Results: A total of 1271 IS were obtained from PBMCs or rectal biopsies from the 15 patients, these represented 1268 different integration events. Only 4 shared IS were observed between PBMCs and rectal biopsies. 44 IS (3.5%) were associated with more than 1000 sequence reads, revealing that only a small fraction of infected cells may be derived from expanded clones. Interestingly, no integrations into the BACH2 or MKL2 gene regions, which were reported to be associated with clonal expansion in previous studies, were found in the whole IS pool. No significant hotspots of virus integrations were found. Conclusion: The HIV-1 proviral repertoire present in PBMCs and rectal biopsies revealed integrations that were not clustered in specific genomic regions without substantial signs of clonal expansion. In line with this, we did not find any integration in the previously described gene regions that were associated with clonal cell expansion. Furthermore, almost no overlap in clones was observed between the PBMCs and rectal biopsies, indicating the compartmentalized HIV-1 reservoir. 303 ENRICHMENT OF DEFECTIVE PROVIRUSES LACKING TAT-REV DURING LONG-TERM ART Francesco R. Simonetti 1 , Paolo Cattaneo 1 , Alessia Lai 1 , Elizabeth M. Anderson 2 , Stefano Rusconi 1 , Melissa Tosiano 3 , John Mellors 3 , Frank Maldarelli 2 , Massimo Galli 4 , Claudia Balotta 1 1 Univ of Milan, Milan, Italy, 2 NCI, Frederick, MD, USA, 3 Univ of Pittsburgh, Pittsburgh, PA, USA, 4 Luigi Sacco Univ Hosp, Milan, Italy Background: Within a few years of ART initiation, HIV DNA in peripheral blood reaches a stable plateau and shows a proviral landscape in which 98% of genomes are not replication competent. However, the composition of these proviruses is heterogeneous, and little is known about their relative dynamics and transcription during ART. We characterized cell associated HIV DNA and RNA measuring LTR, gag and tat-rev from the blood of 75 patients. Methods: We conducted a cross-sectional study on patients with chronic HIV infection sampled at diagnosis (N=14) and after short (N=14) or long-term ART (N=30). We also enrolled patients with acute infection sampled at diagnosis (N=8) and after early ART (N=8). Duplex assays were designed to quantify, by ddPCR, single and double positive (DP) targets on the same DNA fragment: LTR and gag, tat-rev exon-1 and -2, LTR and tat-rev exon-1. We measured plasma HIV RNA by iSCA and cell-associated RNA (us-gag, ms-tat-rev and polyA) in chronic patients on ART. Results: Patients on ART had lower HIV DNA levels and showed higher ratios of LTR to internal regions compared to viremic patients (e.g. mean LTR/gag: 3.4 vs 1.9, p=0.0002), reflecting the loss of productively infected cells with intact HIV DNA. Copies of double positive DNA were significantly lower on ART, with DP tat-rev being the lowest (p=0.035), consistent with prior reports that a greater fraction of proviruses have deletions in these genes. In comparing patients on short and long term ART (median: 1.2 vs 11.4y), there was no significant difference in HIV DNA levels, but LTR to tat-rev ratio was significantly higher in patients with long-term treatment (4.6 vs. 2.7, p<0.0001), and correlated with years on ART (r=0.62, p<0.0001), suggesting selection of proviruses lacking tat-rev over time. In acutely treated patients, DP LTR-gag and tat-rev DNA showed a greater fold reduction (7 and 16 folds, respectively) compared to chronic patients, likely due to early ART preventing further accumulation of defective species. Of the measurements of transcriptional activity, only us-gag and polyA RNA correlated with single and double positive DNA assays, with polyA RNA and DP LTR-gag DNA showing the strongest correlation (r=0.6, p=0.0006). Conclusion: Our data show rapid selection of deleted proviruses upon ART initiation and continued selection on long-term ART of proviruses with deletion in tat-rev sequences. The absence of genes required for proviral transcription and translation may favor the survival of infected cells. 304 HIV INTEGRATION SITE/ORIENTATION CONFER SELECTIVE ADVANTAGE FOR PERSISTENCE ON ART Hadega Aamer 1 , Malika Aid 2 , Sherry McLaughlin 1 , Thomas R. Sibley 3 , Evan J. Silberman 3 , Wenjie Deng 3 , Paul Edlefsen 4 , Rafick-Pierre Sekaly 5 , Lisa Frenkel 3 , James Mullins 3 1 Seattle Children’s Rsr Inst, Seattle, WA, USA, 2 Harvard Univ, Boston, USA, 3 Univ of Washington, Seattle, WA, USA, 4 Fred Hutchinson Cancer Rsr Cntr, Seattle, WA, USA, 5 Case Western Reserve Univ, Cleveland, OH, USA Background: Recent studies have shown that HIV provirus integration may play a role in infected cell persistence and proviral latency. We hypothesized that the HIV integration site (IS) and proviral orientation relative to the human gene confers a proliferative advantage to the infected cell. Methods: We compiled published and our unpublished IS data from in vitro infections and in vivo IS data from individuals on suppressive antiretroviral therapy (ART). Provirus orientation and biological pathway associations found in cells persisting during ART compared to in vitro infected cells were evaluated by a parametric bootstrap likelihood ratio test and MsigDB using Fisher’s exact test. Results: A total of 55,365 IS were identified within genes (51,554 in vitro and 3,811 in vivo), 92% of which were in introns. Proliferation, detected by identical IS in different cells, was higher in vivo (11.5%) than in vitro (<1%). No proviral orientation bias was found in vitro, but integration in the reverse orientation with respect to the gene was detected in vivo (59%, p=0.005), across exons and introns. Pathway analysis revealed significant enrichment in pathways associated with cancer, Treg-modulation, IL10 targets, mitotic spindle, axonal guidance signaling, MHC-I, and the G2M checkpoint. These pathways were significantly more enriched in vivo than in vitro. A bias for integration in the reverse orientation was found among these pathway genes. However, examination of the cancer gene pathway, which was the most enriched pathway in vitro and in vivo, revealed a significant forward orientation bias (p<0.0001) among proviruses found in proliferating cells and a reverse bias among proviruses in non-proliferating cells.

Poster and Themed Discussion Abstracts

CROI 2017 119