CROI 2015 Program and Abstracts

Abstract Listing

Poster Abstracts

birth weight percentiles. In mice, progesterone supplementation improved fetal weight deficits induced by cART. In this study we investigated the molecular mechanisms leading to cART-associated P4 level alterations. Methods: Expression levels of key enzymes of P4 synthesis and metabolismwere assessed by qPCR on placenta tissue from HIV+ cART-exposed (Study group, N=33) and HIV-negative women (Control group N=15). Plasma P4 and human prolactin (hPL) levels were quantified at gestational week 33-37 by EIA. Human choriocarcinoma (BeWo) cells were treated with increasing doses of hPL for 24h and 20 α HSD expression and P4 levels were measured by qPCR and EIA respectively. P4 levels in cART-exposed BeWo cells were assessed with or without 20 α HSD inhibition. Results: Similarly to our findings at mid-gestation, P4 levels were significantly lower in the study group compared to the control group (median [IQR]: 131.0 [93.3-158.9] vs. 171.1 [139.6-198.8] ng/mL respectively, p=0.014). Placental expression of most P4 metabolism enzymes was similar between groups. Only the P4-eliminating enzyme 20 α HSD was significantly higher in the study group (Table 1). hPL, the main regulatory hormone for 20 α HSD, was significantly lower in the study group compared to controls (median [IQR]: 0.50 [0.38-0.72] vs. 0.77 [0.48-0.89] ng/mL, respectively, p=0.043). 20 α HSD expression significantly correlated with hPL levels in women’s plasma at GW 33-37 (Spearman’s r=-0.822, p<0.0001). In BeWo cells hPL down-regulated 20 α HSD expression and P4 production in a dose-dependent manner (p <0.0001). cART exposed BeWo cells produced significantly less P4 compared to controls (median [IQR] 2.8 [2.5-3.1] vs 3.6 [3.5-3.7] ng/mL, p=0.028). P4 levels were restored by inhibiting 20 α HSD activity (3.6 [3.4-4.0] ng/mL).

Expression levels of P4 metabolism enzymes Conclusions: Our data suggest that low P4 levels observed in cART-exposed HIV+ pregnant women could be the result of higher levels of 20 α HSD induced by low hPL levels. We describe a newmechanism by which cART maybe influencing maternal hormone balance during pregnancy, and identify potential new therapeutic targets that may improve birth outcomes for HIV+ women on cART.

TUESDAY, FEBRUARY 24, 2015 Session P-T10 Poster Session

Poster Hall

2:30 pm– 4:00 pm Immune Mechanisms in MTCT 903 Generating HIV Neutralization in MilkWith Neutralizing IgG/dIgA Antibodies Infusion Genevieve Fouda 1 ; Josh Eudailey 1 ; Erika Kunz 1 ; Joshua Amos 1 ; Jonathan Himes 1 ; Lisa Colvin 1 ; XinyueWang 2 ; Keith Reimann 2 ; Barton Haynes 1 ; Sallie Permar 1 1 Duke University, Durham, NC, US; 2 University of Massachusetts Medical School, Boston, MA, US

Background: Breastfeeding is responsible for almost half of infant HIV-1 infections in resource-limited areas. While antiretroviral prophylaxis reduces transmission, it is difficult to implement and does not eliminate the risk of postnatal HIV infection. Inhibition of virus in breast milk by passively-infused broadly neutralizing antibodies (bNab) could be a potential strategy to reduce the infectious virus reservoir in milk. To investigate the distribution of passively-infused bNabs to the breast milk compartment, we assessed the kinetics of binding and neutralizing antibody responses in plasma and milk of lactating rhesus monkeys following passive immunization with the IgG and dimeric IgA (dIgA) forms of the bnAb B12. Methods: The first generation bnAb B12 Ig variable regions were engineered in either rhesus IgG1Fcs or rhesus IgA Fcs with rhesus J chain. Rhesus recombinant B12 versions were administered intravenously at a dose of 5mg/kg to three (dIgA) or four (IgG) hormone-induced, lactating female rhesus monkeys. Blood and milk were collected prior to and from 1 hour to seven weeks post-infusions. Levels of the infused antibodies and tier 1 (HIV MW965) virus neutralization in TZM-bl cells were measured at each time-point. Results: B12 IgG and dIgA peaked in plasma 1-6 hours post-infusion, whereas the peak in milk was slightly delayed following IgG infusion (24-72 h) as compared to dIgA infusion (6-24h). The median peak B12 plasma concentration was similar between the two groups of animals (IgG median: 87,503 ng/ml, dIgA median 72,905 ng/ml, p=0.57). In contrast, the peak B12 milk concentration in dIgA infused animals was up to two logs higher than in animals infused with IgG (Median, IgG:59 ng/ml, dIgA: 5,462 ng/ml, p=0.06). Interestingly, both B12 IgG and B12 dIgA were still detectable in plasma and milk seven weeks after infusion. The peak neutralizing activity in plasma occurred 6 to 24h post infusion and was comparable between animals immunized with IgG and dIgA (median ID 50 against MW965: 2,313.5 versus 1,821, p=0.22). Neutralizing activity peaked in breast milk 24-72 h after IgG infusion and 24 h after dIgA infusion and trended one log higher following dIgA than IgG infusion (median ID 50 50.5 versus 526, p=0.06). Conclusions: Our results indicate distinct kinetics of the transfer of systemic IgG and dIgA bnAbs to the breast milk compartment and suggest that maternal passive immunization

Poster Abstracts

with dIgA bNabs may be an effective way to achieve elimination of infectious virus in milk. 904 Specificity of V3-specific Neutralizing Responses in HIV-1 InfectedWomen David R. Martinez ; Genevieve Fouda; NathanVandergrift; Celia LaBranch; David Montefiori; Xiaoying Shen;Thomas Denny; GeorgiaTomaras; Sallie Permar Duke Human Vaccine Institute, Durhamc, NC, US

Background: We recently completed a humoral immune correlate analysis of MTCT risk in the U.S. Women and Infant Transmission study (WITS), a historical cohort followed before availability of antiretroviral therapy. We observed that plasma IgG responses against the Envelope (Env) third variable (V3) loop and tier 1 virus neutralization predicted reduced risk of MTCT. To follow up on this study, we investigated the binding specificity of maternal plasma anti-V3 IgG antibodies and their contribution to the tier 1 virus neutralization response. Methods: Maternal plasma samples (n=248) were tested for IgG binding against a multi-clade panel of Env V3 peptides by a binding antibody multiplex assay to identify amino acid differences associated with IgG binding strength. A peptide competition neutralization assay with SF162 V3, scrambled and no peptide was used to measure the

543

CROI 2015