CROI 2020 Abstract eBook

Abstract eBook

Poster Abstracts

the landscape of proviral variation and evolution within, and between large numbers of hosts. 376 COMBINED ASSAYS SHED NEW LIGHT ON HIV-1 PROVIRAL SEQUENCE AND LINKED INTEGRATION SITE Basiel Cole 1 , Laurens Lambrechts 1 , Maria Artesi 2 , Vincent Hahaut 3 , Wojciech Witkowski 1 , Ytse Noppe 1 , Annemieke Dhondt 4 , Keith Durkin 5 , Anne Van Den Broeke 6 , Linos Vandekerckhove 1 1 HIV Cure Research Center, Ghent University, Ghent, Belgium, 2 GIGA German Institute of Global and Area Studies, Liege, Belgium, 3 Université Libre de Bruxelles, Brussels, Belgium, 4 Ghent University Hospital, Ghent, Belgium, 5 University of Liège, Liege, Belgium, 6 Institut Jules Bordet, Brussels, Belgium Background: HIV-1 infection remains incurable due to the establishment of a persistent viral reservoir, capable of rebounding upon treatment interruption. Evidence has shown that only a small proportion of this reservoir contains intact proviruses, and these are maintained, at least in part, by clonal expansion. We have generated integration site (IS) data down to single cell level on patients on cART using different approaches, comparing strengths/weaknesses and complementarity of the methods. Methods: Two patients (PT1, PT2) underwent leukapheresis and CD4+ T-cells were isolated. DNA was extracted and IS were sequenced using Integration Site Loop Amplification (ISLA). DNA from the same extract was analyzed using Pooled CRISPR Inverse PCR sequencing (PCIP-seq), a new long-read NGS method, to generate both IS and adjacent proviral genomes. CD4+ T-cells were stimulated, stained for two epitopes of p24, and double positive cells were single-cell sorted. After DNA amplification, near full-length (NFL) proviral genomes and corresponding IS were sequenced. Subsequently, all data were subjected to an in-depth comparison. Results: Using ISLA, we recovered 144 IS for PT1, and 201 IS for PT2. The former displays a limited degree of clonality (7%, 4 clones) while the latter is highly clonal (75%, 13 clones). PCIP yielded 80 IS for PT1 and 161 IS for PT2. Comparison showed that most clonal IS were detected by both ISLA and PCIP, validating the results of PCIP. Moreover, NFL genomes from 4 clones were identified by PCIP in PT2. One of them contained a 115 bp deletion, disrupting the packaging signal. The second one is located in the MLLT3 gene, which protein product has been shown to interact with HIV-1 Tat. Importantly, both of these clones were detected using the p24 stimulation assay while the other two, integrated within centromeric regions, were not detected with the assay. Conclusion: Comparing PCIP to ISLA, we show that PCIP is a potent method to retrieve both IS and linked proviral genome. Next to that, we show that while the stimulation assay biases towards proviruses that are translationally competent, it does not bias towards replication competent ones. The fact that the stimulation assay does not reveal intact proviruses in centromeric regions hints to deep latency and the inability of the assay to reactivate these. We conclude that the PCIP yields the most comprehensive overview of proviruses and their associated IS, while the stimulation assay adds functional data on translational competency. 377 ULTRADEEP ANALYSIS OF PRETHERAPY HIV PREDICTS GENETICALLY COMPLEX RESERVOIRS Kristi Huik 1 , Junko Hattori 1 , Valerie F. Boltz 1 , Jason W. Rausch 1 , Wei Shao 2 , Mary F. Kearney 1 , John M. Coffin 3 , Frank Maldarelli 1 1 National Cancer Institute, Frederick, MD, USA, 2 Leidos Biomedical Research, Inc, Frederick, MD, USA, 3 Tufts University, Boston, MA, USA Background: Measuring the genetic characteristics of HIV populations is essential to understanding the formation of HIV reservoirs that persist during antiretroviral therapy (ART). Analysis of plasma HIV using new next generation sequencing (NGS) approaches using primer ID [ultrasensitive single genome sequencing (uSGS)] and advanced bioinformatic analyses (Boltz et al 2016), yields large HIV sequence datasets with the same, low PCR error and recombination rate as standard SGS. We used uSGS to determine population parameters (replicating population size, in vivo recombination rate). We also extended the uSGS approach to characterize cell associated (CA) HIV RNA and DNA derived from peripheral blood lymphocytes (PBLs). Methods: Plasma samples were obtained from chronically infected ART naïve individuals (N=6) enrolled in HIV studies at the NIH in 2000-2002. uSGS of HIV RT [HXB2nt 2704-2943 and 3046-3253] using primer-ID in the Illumina NGS yielded 400-nt sequences. Replicating population sizes were estimated as previously described (Maldarelli et al 2013) and linkage disequilibrium

was calculated using DNASP; recombination rate was calculated directly by measuring the rate at which linked alleles become unlinked. To obtain uSGS sequences from PBL, DNA was sheared (avg 10 kb), and subjected to a linear PCR step to add primer-IDs before the uSGS procedure. Results: Longitudinal plasma samples were obtained from chronically infected ART naïve individuals (median CD4=498 c/ul, viral RNA=4.3 log 10 cps/ml). uSGS from plasma derived HIV resulted in total of 17,172 (median 1,252/patient, range 54-3165) sequences from 6 subjects from 2 time points. Maximum replicating population sizes exceeded 10 7 /person. Viral populations were highly polymorphic, but nearly all polymorphisms were in linkage equilibrium. With a single exception, all linked loci (3-12/patient) became unlinked over short periods (30-413 generations). The measured recombination rate (range 0.004-0.07) is similar to previous estimates (Batorsky et al 2011) indicating that virtually all sequences were the product of recent recombination events. Analysis of CA HIV from PBL of one patient revealed HIV was readily recovered with 742 DNA sequences, and 946 RNA sequences. Conclusion: Prior to ART, HIV populations are large (>10 6 -10 7 /person) and compose of variants that undergo frequent recombination. uSGS predicts that viruses rebounding from reservoirs are diverse and likely to have evidence of prior recombination events. 378 HIV CONTROLLERS HAVE LOW FREQUENCIES OF INTACT PROVIRAL DNA Abena Kwaa 1 , Caroline Garliss 1 , Kristen D. Ritter 2 , Gregory Laird 2 , Joel Blankson 1 Johns Hopkins University School of Medicine, Baltimore, MD, USA, 2 Accelevir Diagnostics, Baltimore, MD, USA Background: Elite controllers or suppressors (ES) are subjects who control viral replication without antiretroviral therapy. Studies using standard DNA PCR assays or the quantitative viral outgrowth assay (QVOA) have shown that these subjects have smaller viral reservoirs than chronic progressors (CP) on antiretroviral therapy (ART). However, standard DNA PCR assays measure both defective and replication-competent virus and the QVOA measures only a fraction of the replication-competent reservoir. The objective of this study was to better approximate the size of the latent reservoir in ES by measuring the frequency of CD4+ T cells that contained intact proviral DNA. Methods: Total and intact proviral DNA was measured in unfractionated CD4+ T cells from 9 CPs, 8 treatment naïve ES and 2 viremic controllers (VCs, VL < 1000 copies pre-treatment) on ART with the recently described intact proviral DNA assay (IPDA). CD4+ T cells from 5 ES and the 2 VCs on ART were also cultured in the standard QVOA. Results: The median frequency of total provirus was 24.7 per million CD4+ T cells in the ES, 220.1 per million CD4+ T cells in the 2 VCs on ART and 574.6 per million CD4+ T cells in the CPs on ART. The median frequency of intact provirus was 1.2 per million CD4+ T cells in the ES, 2.83 per million CD4+ T cells in the 2 VCs on ART and 36.2 per million CD4+ T cells in the CPs on ART. While the absolute frequencies of total and intact provirus per million CD4+ T cells were significantly lower in ES than in CP, there was no significant difference in the fraction of total proviral DNA that was found to be intact between these 2 subject groups. There was a positive correlation between the frequency of intact proviral DNA and the frequency of latently infected cells as measured by QVOA in the ES and VCs on ART. Conclusion: We show that ES have a median frequency of both intact proviral DNA and total proviral DNA that are more than 1 log lower than the frequencies seen in CPs. These findings suggest that while the absolute frequency of persistent HIV is lower in ES as compared to CP, the relative composition of that pool of persistent proviruses may not differ significantly. Furthermore, this data has implications for HIV cure strategies as it demonstrates that while this small reservoir size may contribute to the control, it is not an absolute requirement as one ES had a higher frequency of intact proviral DNA than all of the CPs in our study. 379 EARLY THERAPY OF YOUTH WITH ACUTE/RECENT HIV: EFFECT ON HIV DNA & ANTIBODY (ATN 147) Ruth Cortado 1 , Tara Kerin 1 , Justine Ceballos 1 , Eduardo Saad 1 , Kate Mitchell 1 , Jasmine Fournier 2 , Brenda Andrews 2 , Manuel Ocasio 2 , Sue Ellen Abdalian 2 , Robert Bolan 3 , Risa Flynn 3 , Brittany Juhlmann 3 , Mary Jane Rotheram-Borus 1 , Karin Nielsen-Saines 1 , Yvonne Bryson 1 1 University of California Los Angeles, Los Angeles, CA, USA, 2 Tulane University, Metairie, LA, USA, 3 Los Angeles LGBT Center, Los Angeles, CA, USA

Poster Abstracts

CROI 2020 130