Maternal Neutralization-Resistant Virus Variants Do Not Predict Infant HIV Infection Risk

Abstract

Mother-to-child transmission (MTCT) of HIV provides a setting for studying immune correlates of protection. Neutralizing antibodies (NAbs) are suggested to contribute to a viral bottleneck during MTCT, but their role in blocking transmission is unclear, as studies comparing the NAb sensitivities of maternal viruses have yielded disparate results. We sought to determine whether transmitting mothers differ from nontransmitting mothers in the ability to neutralize individual autologous virus variants present at transmission. Ten transmitting and 10 nontransmitting HIV-infected mothers at high risk of MTCT were included in this study. Full-length HIV envelope genes (n = 100) were cloned from peripheral blood mononuclear cells obtained near transmission from transmitting mothers and at similar time points from nontransmitting mothers. Envelope clones were tested as pseudoviruses against contemporaneous, autologous maternal plasma in neutralization assays. The association between transmission and the log2 50% inhibitory concentration (IC50) for multiple virus variants per mother was estimated by using logistic regression with clustered standard errors. t tests were used to compare proportions of neutralization-resistant viruses. Overall, transmitting mothers had a median IC50 of 317 (interquartile range [IQR], 202 to 521), and nontransmitting mothers had a median IC50 of 243 (IQR, 95 to 594). Transmission risk was not significantly associated with autologous NAb activity (odds ratio, 1.25; P = 0.3). Compared to nontransmitting mothers, transmitting mothers had similar numbers of or fewer neutralization-resistant virus variants, depending on the IC50 neutralization resistance cutoff. In conclusion, HIV-infected mothers harbor mostly neutralization-sensitive viruses, although resistant variants were detected in both transmitting and nontransmitting mothers. These results suggest that MTCT during the breastfeeding period is not driven solely by the presence of maternal neutralization escape variants.

Importance: There are limited data demonstrating whether NAbs can prevent HIV transmission and infection in humans, and for this reason, NAbs have been studied in MTCT, where maternal antibodies are present at the time of transmission. Results of these studies have varied, perhaps because of differences in methods. Importantly, studies often used cultured viruses and samples from time points outside the window of transmission, which could confound findings. Here, we considered the role of maternal NAbs against individual maternal virus variants near the time of transmission. We found no evidence that NAbs are associated with protection from infection. In fact, depending on the cutoff used to define neutralization resistance, we found evidence that nontransmitting mothers have more neutralization-resistant virus variants. These results suggest that lack of virus transmission in the early breastfeeding period is not simply due to an absence of maternal neutralization escape variants and likely includes multiple factors.

FIG 1

Neighbor-joining trees of maternal HIV envelope genes. Envelope clones from nontransmitting (A) and transmitting (B) mothers. Each envelope sequence is identified by a maternal ID number followed by the timing of isolation and envelope ID. P number, pregnancy week; W number, week after delivery. Samples are color coded on the basis of average IC₅₀s, as shown at the right, when tested against contemporaneous maternal plasma. Viruses resistant to neutralization at a 1:50 dilution of maternal plasma (IC₅₀, 25) are in bold. Reference sequences are not colored.

FIG 2

Autologous NAb responses of nontransmitting and transmitting mothers. (A) IC₅₀s of nontransmitting and transmitting mothers are color coded by mother. Each square or dot represents the IC₅₀ of an individual envelope variant tested against contemporaneous autologous plasma. Data are shown as raw IC₅₀s on a log₂ scale; black bars represent medians and IQRs, and the dotted line represents an IC₅₀ of 25 (resistant to neutralization at a 1:50 dilution). The P value is based on logistic regression with clustered standard errors using log₂-transformed data. (B) Percentages of virus variants with IC₅₀s below the detection limit, which we defined as the midpoint between 0 and the lowest dilution tested (IC₅₀, 25); IC₅₀s of <100; and IC₅₀s of <200 for nontransmitting (black) and transmitting (gray) mothers. The P values were determined by t tests comparing the proportions of neutralization-resistant viruses of transmitting and nontransmitting mothers at the IC₅₀s shown.