CROI 2020 Abstract eBook

Abstract eBook

Poster Abstracts

to ascertain the cholesterol pathway involvement in natural resistance to HIV-1 infection. These results, nevertheless, suggest a possible basis in novel preventive and therapeutic approaches against HIV-1 infection

Tegucigalpa, Honduras, 8 Ministry of Health, Belmopan, Belize, 9 Hospital Nacional Rosales, San Salvador, El Salvador Background Polymorphismwithin the human leukocyte antigen (HLA) class I loci represents the strongest genetic modifier of HIV disease progression. In a cross-sectional Mexico/Central America (MEX/CAM) cohort, we have described both canonical and novel associations between specific HLA and HIV disease progression, from which the B*35 and B*39 subtypes featured several risk associations, including the Amerindian B*35:12/14 and B*39:01/05/06 alleles. As more than 6% of the MEX/CAM cohort expressed two alleles of the B*35 and/or B*39 subgroups, we investigated HLA additive effects and in the context of B*35-PX/PY grouping with HIV disease progression. Methods: HLA sequence-based typing was performed on 3213 chronically HIV-1 clade B-infected, ART-naïve individuals fromMexico (n=1679), Guatemala (n=418), Nicaragua (n=254), Honduras (n=402), Panama (n=316), Belize (n=102), and El Salvador (n=42). Univariate and multivariate analyses were performed using the Generalized Linear Model (GLM) to evaluate additive effects between B*35/39 subtypes or between B*35-PX/PY groups. Associations were adjusted by age, gender and location of recruitment, included as possible confounders in multivariate analyses. For B*35-PX/PY analyses, only HLA-B heterozygous individuals were compared in order to exclude confounding effects resulting from HLA homozygosity. Results: Both in univariate and multivariate analyses, expressing one or two copies of any B*35 or B*39 subtype (B*39:02 being the exception) was associated to significantly higher plasma viral load (pVL) and lower CD4 counts (in all cases p<0.05). pVL and CD4 linear regression coefficients were one-fold larger in individuals that co-expressed 2-copies of any B*35/39 in comparison with subjects that expressed 1-copy of any B*35/39, suggesting an additive detrimental effect. We confirmed the B*35-PX group association with poor HIV outcome (both with pVL and CD4), but also observed that B*35-PY alleles were associated to significant lower CD4 counts. Given its similarity with other PX members (B*35:02/03), the Amerindian B*35:12 allele represents a putative newmember of the established B*35-PX HIV risk group. Conclusion: Our results suggest an additive detrimental effect between B*35/39 subtypes, highly frequent in the Mesoamerican Mestizo population. Our results also challenge the B*35-PX/PY hypothesis, indicating that PY alleles can be disease-susceptible and also that differences exist in disease associations within PX/PY grouping. 275 NATURAL RESISTANCE TO HIV-1 CORRELATES WITH IFNA-CONTROLLED STEROL METABOLISM Mara Biasin 1 , Irma Saulle 1 , Salomè Valentina Ibba 2 , Claudio Fenizia 1 , Francesca Vichi 3 , Sergio Lo Caputo 4 , Daria Trabattoni 1 , Mario Clerici 1 1 University of Milan, Milan, Italy, 2 Louisiana State University, New Orleans, LA, USA, 3 Santa Maria Annunziata Hospital, Milan, Italy, 4 University of Bari, Bari, Italy Background: The ER-associated enzyme cholesterol-25-hydroxylase (CH25H) is an Interferon stimulated gene (ISG) which is able to interfere with viral replication through the modulation of cholesterol metabolism. We therefore verified if natural resistance to HIV-1 infection in HIV-exposed seronegative (HESN) subjects is at least partially dependent on a peculiar regulation of sterol biosynthesis pathway mediated by IFN-induced CH25H expression. Methods: Peripheral blood mononuclear cells (PBMCs) and monocyte-derived macrophages (MDMs) isolated from 15 sexually-exposed HESN, their HIV+ partners and 15 healthy controls were analyzed for: 1) percentage of IFNa- producing plasmacytoid Dendritic Cells (pDCs); 2) RNA expression of factors involved in lipoprotein signaling and cholesterol metabolism by Real Time PCR; 3) susceptibility to HIV-1 infection by p24 viral antigen quantification. Results: The increase in IFNa-producing pDCs in both unstimulated and in vitro HIV-infected PBMCs from HESN was coupled with an augmented expression of cholesterol-25-hydroxylase (CH25H) (HESN vs HC: p<0.001 in both cases). The expression of several genes involved in cholesterol metabolism (LXR, ABCA1, SCARB, HMGCS1, PPARg) was modulated as well (>3 fold) in unstimulated as well as in in vitro HIV-1-infected PBMCs and MDMs from HESN. Notably, this resulted in a significantly reduced susceptibility to in vitro HIV-1-infection in PBMCs and MDMs of HESNs (p<0.01). Conclusion: The observation that CH25H, an oxysterols-producing enzyme, is up-regulated in HIV-exposed cells from HESN, is particularly intriguing. This could be related to the activation of the IFNa pathway, resulting in a reduced susceptibility to in vitro HIV-1 infection. Further analyses are needed

276 BASELINE INDUCIBLE HIV p24 INFORMS VIRAL CONTROL DURING INTERFERON-Α MONOTHERAPY Livio Azzoni 1 , Emmanouil Papasavvas 1 , Pablo Tebas 2 , KaramMounzer 3 , Jay Kostman 3 , Ian Frank 2 , Bonnie J. Howell 4 , Daria Hazuda 4 , Daniel Holder 4 , Nicolas Chomont 5 , Costin Tomescu 1 , Michael R. Betts 2 , Leticia Kuri Cervantes 2 , Luis J. Montaner 1 1 Wistar Institute, Philadelphia, PA, USA, 2 University of Pennsylvania, Philadelphia, PA, USA, 3 Philadelphia FIGHT, Philadelphia, PA, USA, 4 Merck & Co, Inc, West Point, PA, USA, 5 Université de Montréal, Montreal, QC, Canada Background: Pegylated (peg) IFNα monotherapy after ART interruption results in increased HIV control in association with NK cell activation. The relationship between inducible or other HIV proviral reservoir measurements with subsequent time to rebound during ART interruption and Peg IFN-α monotherapy are unknown. Methods: 13 individuals randomized to arm 1 of the BEAT-HIV study (NCT02227277: HIV VL < 50 copies/ml on ART, CD4 count > 450/μl) receiving 1 μg/kg of peg-IFNα-2b (Pegintron, Merck) for 20 weeks, interrupting ART at week 4 and resuming it upon viremia (VL > 50 copies/μl, bi-weekly evaluations) or at week 20. P24 SIMOA (iP24) was measured in CD4+ T cells cultured for 16-hour with medium or PMA/Ionomycin using single molecule array (SIMOA). Intact, 5’ defective, 3’ defective and total proviral DNA were measured by Accelevir, Inc. on CD4+ T cells; Integrated HIV proviral DNA was assessed using Alu-gag RT-PCR on CD4+ T cells. Time to viremia was first VL > 50 copies/ml after stopping ART. HIV-specific responses in PBMC: a) T cell - 6-hour cultures of PBMC with of 15-mer gag peptides. B) NK ADCC: 4-hour co-cultures with anti-HIV sera with gp120-coated CEM NKres targets. Multicolor flow cytometry was used to assess HIV-specific degranulation and cytokine production. Associations were tested with Pearson or Spearman tests, and linear regression models. Results: 12 of 13 participants became viremic during ART interruption, one remained suppressed and was imputed to week 20. ip24 was positively associated (p<0.05) with time to viremia (effect estimate 0.362; p= 0.029; Adj R2 = 0.305), first detected VL (Fig 1), and Fc receptor-dependent expression of intracellular MIP1β in CD56dim/CD57neg NK cells, but not with T-cell responses to Gag peptides. Proviral measures were correlated to each other, as expected, but not with time to viremia or level of first VL measured. Conclusion: In vitro-inducible HIV proteins (i.e.: p24 SIMOA), but not total proviral HIV DNA measures, are associated with level of first viral load and time to viremia during peg-IFNα-2b mono therapy. In contrast to expectation that higher latent reservoirs would lead to shorter time to viremia, the immune correlates measured are consistent with NK ADCC response and chemokine responses contributing to viral control off ART

Poster Abstracts

94

CROI 2020