Antibody Fc effector functions and IgG3 associate with decreased HIV-1 risk

Abstract

HVTN 505 is a preventative vaccine efficacy trial testing DNA followed by recombinant adenovirus serotype 5 (rAd5) in circumcised, Ad5-seronegative men and transgendered persons who have sex with men in the United States. Identified immune correlates of lower HIV-1 risk and a virus sieve analysis revealed that, despite lacking overall efficacy, vaccine-elicited responses exerted pressure on infecting HIV-1 viruses. To interrogate the mechanism of the antibody correlate of HIV-1 risk, we examined antigen-specific antibody recruitment of Fcγ receptors (FcγRs), antibody-dependent cellular phagocytosis (ADCP), and the role of anti-envelope (anti-Env) IgG3. In a prespecified immune correlates analysis, antibody-dependent monocyte phagocytosis and antibody binding to FcγRIIa correlated with decreased HIV-1 risk. Follow-up analyses revealed that anti-Env IgG3 breadth correlated with reduced HIV-1 risk, anti-Env IgA negatively modified infection risk by Fc effector functions, and that vaccine recipients with a specific FcγRIIa single-nucleotide polymorphism locus had a stronger correlation with decreased HIV-1 risk when ADCP, Env-FcγRIIa, and IgG3 binding were high. Additionally, FcγRIIa engagement correlated with decreased viral load setpoint in vaccine recipients who acquired HIV-1. These data support a role for vaccine-elicited anti-HIV-1 Env IgG3, antibody engagement of FcRs, and phagocytosis as potential mechanisms for HIV-1 prevention.

Keywords: AIDS vaccine; AIDS/HIV; Vaccines.

Figure 1. ADCP, FcγRIIa, and IgG3 significantly correlate with decreased HIV-1 risk.

HIV-1 Env Con S gp140 ADCP magnitude (score) (A), FcγRIIa binding MFI (B), and IgG3 breadth score (C) by infected/uninfected outcome. ADCP (OR = 0.47, P = 0.001), FcγRIIa (OR = 0.48, P < 0.001), and Env IgG3 breadth (OR = 0.326, P < 0.001) were inversely correlated with HIV-1 acquisition risk. Box-and-whisker plots show the median line and interquartile ranges. n = 125 uninfected, 25 infected participants.

Figure 2. Envelope-specific IgA modifies the association of ADCP and FcγRIIa with HIV-1 risk.

HIV-1 envelope Con S gp140 ADCP magnitude (score) (A) and FcγRIIa binding MFI (B) by infected/ uninfected outcome, stratified by Env IgA positivity. The associations between ADCP/FcγRIIa and HIV-1 acquisition risk are stronger among vaccinees without detectable Env IgA than those with detectable Env IgA (ADCP OR = 0.16, P = 0.006; FcγRIIa OR = 0.37, P < 0.001). The interaction is significant for ADCP (ratio of ORs = 4.3, P = 0.039) but not for FcγRIIa (ratio of ORs = 1.5, P = 0.43). ADCP and FcγRIIa did impact risk in IgA-positive participants (ADCP IgA⁺ OR = 0.68, P = 0.011; FcγRIIa IgA⁺ OR = 0.56, P < 0.001). Statistics are summarized in Table 3. P values were calculated using the Wald test. Box-and-whisker plots show the median line and interquartile ranges. n = 125, uninfected, 25 infected participants.

Figure 3. Multiple immune responses among vaccinees contribute to HIV-1 acquisition risk.

(A) Principal components analyses with a scatterplot of the first (PC1) and second (PC2) principal components are shown. In A, summary variables are assessed as described in Supplemental Table 2, with major contributors described in the legend of this figure. Each measurement from an infected vaccinee is represented by a red dot and each measurement from an uninfected vaccinee is represented by a black circle. Ellipses representing 95% confidence regions are shown. **Principal components were significantly inversely correlated with HIV-1 risk (PC1: OR = 0.366, P < 0.001 and PC2: OR = 0.409, P < 0.001). (B) Prediction accuracy of SuperLearner models for different marker sets (all models adjust for the baseline variables age, BMI, and behavioral risk score). CV-AUC is the cross-validated area under the receiver operating characteristic curve. The 95% CI to the right of 0.5 indicates the ability to predict HIV-1 acquisition. Fx Ab = functional antibody variables ADCP, FcγRIIa, FcγRIIIa; no markers = model with baseline variables only. P values were calculated using the Wald test. n = 125 uninfected and 25 infected participants.

Figure 4. Antibody Fc effector function, IgG3, and CD8⁺ T cells stratify vaccinees into vaccine efficacy greater than zero and at (or near) zero.

The ADCP immune measurement stratifies vaccinees into vaccine efficacy >0 (VE [1]) (black lines) and vaccine efficacy near 0 (VE [0]) (red lines), similar to CD8⁺ T cell Env polyfunctionality score. Ignorance intervals (solid lines) and 95% estimated uncertainty intervals (EUIs) (dashed lines) for VE by subgroup of vaccinees were defined by cutting above and below the median response magnitude of each of the immune response measurements of ADCP, FcγRIIa, and CD8 Env method of Gilbert et al. (25). Each panel assumes no (A), medium (B), or high (C) potential selection bias. The amount of potential selection bias is specified by a model that varies defined sensitivity parameters over the range [0, 0], [–0.5, 0.5], or [–1, 1], respectively. n = 25 infected and 125 uninfected participants.

Figure 5. Host genetics (FcγRIIa SNPs) significantly modified the correlation of ADCP with HIV-1 infection risk.

**Interaction P = 0.0006 and Q value = 0.026, *within genotype group P < 0.05 after adjusting for all covariates and 2 cellular variables. P values were calculated based on the Wald test. Con S gp140 phagocytosis (A), FcγRIIa binding (B), and Env IgG3 (C) associations with FCGR2A-intron13-645-G/A. Q value was adjusted for 7 × 2 statistical tests with respect to 7 FcγRIIa SNPs and 2 primary Fc effector functions. In the first analysis, we adjusted for age, race, behavior risk, and BMI. In the second analysis, in addition to adjusting these covariates, we also adjusted for 2 cellular responses, Env CD4⁺ T cell polyfunctionality score (PFS) and Env CD8⁺ T cell PFS. Box-and-whisker plots show the median line and interquartile ranges. There are 9 infected and 60 uninfected vaccinees with the GG genotype and 16 infected and 62 uninfected in the GA/AA genotype group.

Figure 6. Host setpoint viral load negatively correlates with FcγRIIa binding.

FcγRIIa and representative imputed viral load setpoint (A and B) and observed viral load (measured as RNA copies/mL) (C) are displayed. Each standard deviation increase in FcγRIIa associated with a 0.53 log₁₀ viral load decrease in viral load setpoint (P = 0.027, 95% CI = 0.06 to 1.00, linear regression shown in A). Dichotomized cases and viral load are displayed (B), with time since infection and median in C. For FcγRIIa low-binding participants, n = 13 (at baseline), 14 (at week 2), 5 (at week 4), 10 (at week 6), 11 (at week 8), 9 (at week 10), 8 (at week 12), 6 (at week 14), 7 (at week 16), 5 (at week 20), 2 (at week 24), 2 (at week 28), 2 (at week 32), and 2 (at week 36). For FcγRIIa high-binding participants, n = 12 (at baseline), 13 (at week 2), 9 (at week 4), 9 (at week 6), 8 (at week 8), 9 (at week 10), 6 (at week 12), 7 (at week 14), 8 (at week 16), 8 (at week 20), 9 (at week 24), 7 (at week 28), 6 (at week 32), 5 (at week 36), 4 (at week 40), 4 (at week 44), 4 (at week 48), 2 (at week 64), 3 (at week 56), and 2 (at week 72). Box-and-whisker plots show the median line and interquartile ranges. P values were calculated using the Wald test.