CROI 2020 Abstract eBook

Abstract eBook

Poster Abstracts

Conclusion: High-throughput, single-copy HIV env sequencing can reveal genetic changes that occur within large virus populations after bNAb infusion. The incorporation of UMIs in full-length env sequencing greatly improves accuracy, providing a robust method to study dynamics of Ab-resistant HIV. 188 ANALYSIS OF COMPARTMENTALIZATION OF HIV-1 IN BONE MARROW Thuy T. Nguyen 1 , Jenna Honeycutt 1 , Christopher Nixon 1 , Oksana Zakharova 1 , Christopher Evans 1 , Joann D. Kuruc 1 , Ben Murrell 2 , Cynthia L. Gay 1 , Douglas D. Richman 3 , J. V. Garcia 1 1 University of North Carolina at Chapel Hill, Chapel Hill, NC, USA, 2 Karolinska Institutet, Stockholm, Sweden, 3 University of California San Diego, La Jolla, CA, USA Background: HIV infection results in hematological disorders frequently observed in the late stages of disease. Little is known about the virus in the bone marrow (BM). We evaluated the compartmentalization of HIV in BM from participants not on antiretroviral therapy (ART). Methods: Full HIV env was sequenced from BM and peripheral blood (PB) plasma using PacBio technologies. High-quality consensus sequences (FLEA software) were used to build phylogenetic trees (FastTree v.2.1.11) and assessed for compartmentalization by distance-based tests a) genetic diversity with the average pairwise distance (APD), b) divergence using a test for panmixia, c) Wright’s measure of population subdivision (Fst), d) nearest-neighbor statistics and tree-based tests a) Simmonds Association Index, b) Slatkin- Maddison, c) correlation coefficients (MEGA7, HYPHY v.2.3.14). Viral tropism was determined at 2.5% of false-positive rate (https://coreceptor.geno2pheno. org/). Compartmentalization was established when results from phylogeny and compartmentalization analyses were concordant. Results: Paired PB and BM samples were collected from 3 participants. Participant 1 was ART-naïve. Participants 2 and 3 were ART-experienced but stopped treatment 12 and 2 months before sample collection, respectively. Participant 3 self-discontinued his ART multiple times. The medians [IQR] log HIV-RNA (copies/mL) in BM (4.85 [4.06-4.97]) and in PB (4.84 [3.69-4.97]) were not statistically different (p = 0.25). The HIV populations had a low diversity (median APD of 1.29% in BM vs 0.94% in PB). The APD was statistically different between BM and PB from participant 3 (1.78% in BM vs 2.65% in PB, p = 0.0005). Participant 3 had two viral populations with genetic distance of 22.44%, compartmentalized virus in BM (fig.1), and CXCR4-tropic virus present at 3.04% in BM and 7.27% in PB. CXCR4- tropic virus was also found in PB from participant 1 (6.59%). Neither participant 1 nor 2 had compartmentalized virus in BM. Conclusion: We demonstrate viral compartmentalization and the presence of CXCR4-tropic virus in the BM. HIV-1 compartmentalization has been previously shown in the central nervous system and correlated with neurocognitive impairment suggesting that compartmentalization in other tissues might have pathogenic consequences.

189 INTRAHOST EVOLUTION OF HIV-2 p24 CORRELATES WITH PROGRESSION TO AIDS Michael T. Boswell 1 , Angelica Palm 2 , Sara Karlson 2 , Fredrik Månsson 2 , Hans Norrgren 2 , Antonio Biague 3 , Zacarias J. Da Silva 3 , Marianne Jansson 2 , Patrik Medstrand 2 , Sarah Rowland-Jones 1 , Joakim Esbjörnsson 2 1 University of Oxford, Oxford, UK, 2 Lund University, Lund, Sweden, 3 Laboratoire National de Santé Publique, Bissau, Guinea-Bissau Background: HIV-2 infection will progress to AIDS in most patients without treatment, albeit at half the rate of HIV-1 infection. Prolines in p24 have been associated with low HIV-2 viral loads and enhanced processing of T cell epitopes. HIV-2 long-term non-progression has been linked to maintenance of Gag-specific CD8+ T cell responses. Lower evolutionary rates and positive selection on conserved residues in HIV-2 env have been associated with slower progression to AIDS. Here, we hypothesise that the intrahost evolution of HIV-2 p24 impacts on disease progression rates. Specific aims: 1) To determine intrahost HIV-2 p24 sequence variation and evolution. 2) To determine site-specific selection pressure in HIV-2 p24 within hosts. 3) To determine how HIV-2 p24 evolution associates with disease progression Methods: Twelve treatment-naïve patients from the Guinea-Bissau Police cohort with longitudinal CD4+ T cell data and clinical follow-up were included in the analysis. Gag amplicons of 735 nucleotides were amplified, cloned and sequenced from 25 blood plasma samples by nested PCR, TOPO-TA cloning, and Sanger sequencing. The sequences were analysed by Bayesian phylogenetics. Results: In total, 371 heterochronous HIV-2 p24 sequences from 12 male patients with a median age of 29.5 years at enrolment were analysed. CD4+ T cell data was used to stratify patients into faster and slower disease progressor groups. Faster progressors had lower CD4% levels at the midpoint of follow-up and faster CD4% decline rates. The time to AIDS was approximately twice as fast among faster progressors than slower progressors (9.4 vs. 20.6 year, P<0.001). Slower progressors’ p24 sequences were more likely to have the G6A, I12V and A119P variants (OR=4.2, P<0.001). P24 evolved under negative selection in both groups (dN/dS=0.13). Synonymous and nonsynonymous substitution rates were higher in faster than slower progressors (4.5x10-3 vs. 1.6x10-3 and 5.9x10-4 vs. 2.7x10-4 substitution/site/year, P<0.001). Viral evolutionary rates had a strong negative correlation with CD4% level (ρ=-0.78, P=0.02), but not decline rates. Conclusion: Faster evolution in HIV-2 p24 was associated with lower CD4% level and faster time to AIDS 190 PLATELETS CARRY INFECTIOUS HIV IN cART-SUPPRESSED PATIENTS IN IMMUNOLOGICAL FAILURE Morgane Bomsel 1 , Fernando Real 1 , Claude Capron 2 , Alexis Sennepin 1 , Aiwei Zhu 1 , Jacques Izopet 3 , Eliseo Eugenin 4 , Elisabeth Rouveix 2 , Elisabeth Cramer- Bordé 2 1 INSERM, Paris, France, 2 Assistance Publique – Hôpitaux de Paris, Paris, France, 3 INSERM, Toulouse, France, 4 University of Texas at Galveston, Galveston, TX, USA Background: Beside hemostasis, human platelets exert several immune functions and interact with infectious pathogens including HIV. We investigated whether platelets from cART-treated patients contain infectious HIV in vivo and addressed the significance and clinical implications of HIV sheltering by platelets in AIDS. Methods: Infectious HIV content in platelets was quantified by qPCR, FISH- Flow, microscopy, and reporter cell assays using platelet-rich-plasma (PRP) from 78 HIV-infected cART-treated adult patients. The capacity of platelet containing HIV to propagate infection was evaluated by culturing human primary macrophages with PRP with or without the platelet activation-blocker Abciximab (anti-integrin alpha(IIb)/beta(3) Fab). The presence of HIV in platelets was correlated with patient clinical status and parameters over >3 years. Results: We demonstrate that platelets from HIV-infected patients shelter infectious HIV in vivo, despite successful viral suppression by the combined antiretroviral therapy (cART) and in strong correlation with low blood CD4+T-cell counts (<350cells/microL). Patient platelets carrying HIV can propagate infection to macrophages in vitro in a process prevented by blocking platelet-macrophage interaction with Abciximab. Comparative phylogenetic analyses of virus found in peripheral blood and platelet samples prior to and > 1 year after cART initiation indicate that viruses contained in platelets do not originate from a latent reservoir established prior to therapy. Moreover, 88% of virally suppressed patients sheltering HIV in platelets are immunological nonresponders and fail to restore a proper immune status over > 1year of cART

Poster Abstracts

63

CROI 2020