Early development of broadly neutralizing antibodies in HIV-1-infected infants

Abstract

Eliciting protective neutralizing antibodies (NAbs) against HIV-1 is daunting because of the extensive genetic and antigenic diversity of HIV-1. Moreover, broad and potent responses are uncommon even during persistent infection, with only a subset of adults developing broadly neutralizing antibodies (bNAbs) that recognize viral variants from different HIV-1 clades. It is not known whether bNAbs can also arise in HIV-1-infected infants, who typically progress to disease faster than adults, presumably in part due to an immature immune system. Here, we show that bNAbs develop at least as commonly in infants as in adults. Cross-clade NAb responses were detected in 20/28 infected infants, in some cases within 1 year of infection. Among infants with breadth of responses within the top quartile, neutralization of tier 2 or 3 variants from multiple clades was detected at 20 months after infection. These findings suggest that, even in early life, there is sufficient B cell functionality to mount bNAbs against HIV-1. Additionally, the relatively early appearance of bNAbs in infants may provide a unique setting for understanding the pathways of B cell maturation leading to bNAbs.

Figure 1

Neutralization of panel viruses. (a) Initial screen of 28 infant samples against 8 viruses. The first column shows infant ID, followed by the last HIV-1 negative timepoint, where ‘−1.0’ indicates testing of cord blood sample, and ‘0.0’ indicates delivery timepoint. The plasma timepoint tested is followed by time post-HIV-1 infection (PI), calculated as time from first HIV-1 detection. The column labeled ‘Clade’ indicates the infecting virus clade. BH285 was infected with a clade C recombinant virus but portions of the envelope gene could not be readily assigned to any known subtype. Panel virus clades and tier designations^,, are shown above virus names. In the second to last column, ‘+’ and ‘−’ indicate the presence or absence of cross-clade NAbs against a Tier 2 virus, respectively. The last column shows the number of Tier 2 virus clades neutralized. IC₅₀ values, shown as reciprocal plasma dilutions from ≥ 2 independent experiments, are color coded as shown in the key. Gray shading indicates that 50% neutralization was not achieved at the lowest plasma dilution tested (1:100). The 7 samples that neutralized a Tier 2 virus from all 4 clades are shown above the horizontal line. (b) Neutralization profile of 7 samples with bNAbs against 23 viruses. Infant IDs are shown in the first column, followed by the months post-infection at which samples were tested. IC₅₀ values are from ≥ 2 independent experiments and are color coded as in (a). The last column shows percentage of viruses neutralized. SIV was included as a negative control.

Figure 2

Kinetics of NAb breadth for 7 infants. Graphs show IC₅₀ values against 8 viruses, 2 from each subtype A–D (shown in the key in upper right corner) over time. Viruses are color-coded by clade, as shown in the key. Black arrows denote when HIV-1 was first detected. Error bars indicate standard error of the mean based on 2 independent experiments.

Figure 3

Association between passive and de novo NAbs. (a) Correlation between passive and de novo infant average log₂(IC₅₀) values. Samples were available for 22/28 infants. (b) Correlation between passive and de novo log₂(IC₅₀) for infants with NAbs against Tier-2 viruses from at least 2 different clades identified in Figure 1a and with plasma sample available within the first week of life (n = 11). (c) Correlation between passive and de novo log₂(IC₅₀) for infants without NAbs against Tier-2 viruses from at least 2 different clades identified in Figure 1a and with plasma sample available within the first week of life (n = 11). For all panels, infant de novo IC₅₀ values were obtained from the last available timepoint (the plasma sample used in Fig. 1), while passive IC₅₀ values were obtained within the first week of life. For panels (a) and (b), red dot represents data point for BG505. Pearson’s r, 95% confidence intervals, and P-values shown in black and gray indicate analysis excluding and including BG505, respectively.

Figure 4

Comparison of set-point viral load and Env-specific IgG for infants with and without bNAbs. Unpaired t-test with Welch’s correction comparing (a) set-point viral load of infants with (n=6, excluding BB391, who had no SVL measurement) and without (n = 21) bNAbs, and (b) Env-specific (Q461.d1 gp120) log-transformed end-point titers (EPT) of plasma from infants with (n = 7) and without (n = 21) bNAbs identified in Figure 1. Horizontal lines in (a) and (b) represent mean and 95% confidence intervals.