Vaccine-Induced, High-Magnitude HIV Env-Specific Antibodies with Fc-Mediated Effector Functions Are Insufficient to Protect Infant Rhesus Macaques against Oral SHIV Infection

Abstract

Improved access to antiretroviral therapy (ART) and antenatal care has significantly reduced in utero and peripartum mother-to-child human immunodeficiency virus (HIV) transmission. However, as breast milk transmission of HIV still occurs at an unacceptable rate, there remains a need to develop an effective vaccine for the pediatric population. Previously, we compared different HIV vaccine strategies, intervals, and adjuvants in infant rhesus macaques to optimize the induction of HIV envelope (Env)-specific antibodies with Fc-mediated effector function. In this study, we tested the efficacy of an optimized vaccine regimen against oral simian-human immunodeficiency virus (SHIV) acquisition in infant macaques. Twelve animals were immunized with 1086.c gp120 protein adjuvanted with 3M-052 in stable emulsion and modified vaccinia Ankara (MVA) virus expressing 1086.c HIV Env. Twelve control animals were immunized with empty MVA. The vaccine prime was given within 10 days of birth, with booster doses being administered at weeks 6 and 12. The vaccine regimen induced Env-specific plasma IgG antibodies capable of antibody-dependent cellular cytotoxicity (ADCC) and phagocytosis (ADCP). Beginning at week 15, infants were exposed orally to escalating doses of heterologous SHIV-1157(QNE)Y173H once a week until infected. Despite the induction of strong Fc-mediated antibody responses, the vaccine regimen did not reduce the risk of infection or time to acquisition compared to controls. However, among vaccinated animals, ADCC postvaccination and postinfection was associated with reduced peak viremia. Thus, nonneutralizing Env-specific antibodies with Fc effector function elicited by this vaccine regimen were insufficient for protection against heterologous oral SHIV infection shortly after the final immunization but may have contributed to control of viremia. IMPORTANCE Women of childbearing age are three times more likely to contract HIV infection than their male counterparts. Poor HIV testing rates coupled with low adherence to antiretroviral therapy (ART) result in a high risk of mother-to-infant HIV transmission, especially during the breastfeeding period. A preventative vaccine could curb pediatric HIV infections, reduce potential health sequalae, and prevent the need for lifelong ART in this population. The results of the current study imply that the HIV Env-specific IgG antibodies elicited by this candidate vaccine regimen, despite a high magnitude of Fc-mediated effector function but a lack of neutralizing antibodies and polyfunctional T cell responses, were insufficient to protect infant rhesus macaques against oral virus acquisition.

Keywords: ADCC; Fc-mediated antibody function; pediatric HIV vaccine; rhesus macaque.

FIG 1

Experimental design. In the vaccine group, 12 neonatal rhesus macaques (Table 1) were immunized with 2 × 10⁸ PFU of MVA-HIV Env, HIV Env protein (15 μg) mixed with 3M-052-SE, and 5 × 10¹⁰ ChAdOx1.tSIVconsv239 viral particles at week 0. Booster immunizations of 2 × 10⁸ PFU each of MVA-HIV Env, HIV Env protein in 3M-052-SE, and MVA.tSIVconsv239 were provided at weeks 6 and 12. A second cohort of 12 age-matched RM received control MVA immunizations at weeks 0, 6, and 12. Beginning at week 15, animals were challenged weekly with SHIV-1157(QNE)Y173H viral stock diluted 1:1,000 in RPMI 1640 medium until infected. After 13 exposures, uninfected infants (n = 11) were exposed to a 1:100 SHIV dose for 7 weeks, a dose that was increased to 1:10 for seven more exposures in animals not infected by the 1:100 dose (n = 4). Two infants remained negative and became infected after challenge with 1:2 dilution of virus stock (RM19) or undiluted (1:1) virus (RM10) (Table 1). SHIV exposures are indicated by arrows with distinct shades of red based on virus dilution.

FIG 2

C.1086 Env-specific antibody responses. Plasma concentrations of 1086.c gp120-specific IgG (A) and IgA (B) were measured by ELISA and BAMA, respectively. Salivary IgG and IgA levels, measured by BAMA, are reported as specific activity in nanograms of 1086.c gp120 IgG or IgA per μg of total IgG (C) or IgA (D). Dashed lines represent the cutoff for positivity, defined as mean antibody levels in control animals plus 3 standard deviations (SD). Panels E and F illustrate the Spearman correlation between plasma and saliva vaccine-induced IgG and IgA levels, respectively. Each symbol represents an individual animal of the 12 vaccinated animals.

FIG 3

Prechallenge antibody function of vaccinated infant macaques. (A) Avidity score, determined by SPR, of week 15 plasma IgG specific for 1086.c gp120 or V1V2 or for the consensus clade C V3 (gp70). Each symbol represents a single animal. Note that only 11 of 12 animals were included in the testing for 1086.c V1V2 avidity due to limited plasma volumes. (B) Tier 1b clade C I6644.v2.c33 neutralization titers of vaccinated infants at week 14 and week 15. (C and D) Longitudinal data for ADCC endpoint titers and maximum granzyme B activity, with each line representing an individual animal. Dashed lines indicate the limit of detection. (E) The percentage of monocyte-independent, NK cell-mediated ADCC activity at week 15. (F) ADCP scores for vaccinated animals prior to vaccination at week 0 and week 14. (G) Plasma IgG binding to cells infected with HIV 1086.c is shown over time for individual vaccinated animals. Each time point shows data for all 12 of the vaccinated animals if not indicated otherwise.

FIG 4

Vaccine-induced 1157ipd3N4 and SHIV1157(QNE)Y175H Env-specific antibody responses. (A) Plasma concentration of 1157ipd3N4 gp120-specific IgG over time in the 12 vaccinated infant rhesus macaques. (B) Avidity scores of plasma IgG specific for 1157ipd3N4 gp120 (n = 12) or gp70-V1V2 SHIV1157(QNE)Y375H (n = 10). Each symbol represents an individual animal; horizontal lines represent the medians. Note that only 10 animals could be tested for the avidity of antibodies to gp70-V1V2 SHIV1157(QNE)Y375H due to the limited plasma volumes available from infant rhesus macaques. (C) ADCC endpoint titers for plasma antibodies specific to 1157ipd3N4 gp120 in the 12 vaccinated animals.

FIG 5

SIV Gag-specific T cell responses in PBMCs (n = 12) and peripheral lymph nodes (n = 12) at week 14. Each bar in panels A and B represents the sum of single cytokine responses of SIV Gag-specific CD4⁺ (left graphs) or CD8⁺ T cells (right graphs) for each vaccinated animal at week 14 in PBMCs (A) or lymph nodes (B). Cytokines measured include gamma interferon (IFN-γ), IL-2, IL-17, and TNF-α.

FIG 6

Challenge outcome. (A) Longitudinal plasma viral load measurements as assessed by RT-PCR from control (n = 12) and vaccinated (n = 12) cohorts of infant RM are displayed in copies per milliliter plasma. Shaded areas represent the challenge doses: light gray, 1:000, weeks 0 to 13; medium gray, 1:100, weeks 14 to 21; dark gray, 1:10, weeks 22 to 28; darkest gray, 1:2 or undiluted. (B) The number of challenges required for infection is plotted for control (n = 12) and vaccinated (n = 12) animals. Horizontal lines represent the medians. (C) Kaplan-Meier survival curves for any dose of viral stock dilutions are shown for control and vaccinated infants. (D and E) Peak viremia (D) and area-under-the curve (AUC) viremia from weeks 0 to 10 postinfection (PI) (E) in control (n = 12) and vaccinated (n = 12) animals. Control and vaccinated animals are indicated by orange or blue lines/symbols, respectively, with each symbol representing an individual animal; horizontal lines indicate the medians.

FIG 7

CD4⁺ T cell activation. PBMCs from week 15 after vaccination were gated on CD3⁺ CD4⁺ T cells and assessed for surface expression of CD195 (CCR5), CD69, and CD279 (PD1) and intracellular expression of Ki-67 and TNF-α. TNF-α-positive T cell frequencies between control (n = 12) and vaccinated (n = 12) animals were compared by Mann-Whitney test.

FIG 8

Correlation between 1086.c Env-specific antibody responses and challenge outcome. (A) Kaplan-Meier plot to demonstrate the relationship between ADCC endpoint titers and number of challenges required for infection when vaccinated animals are categorized as having a low (n = 6) or high (n = 6) ADCC titer based on the median ADCC endpoint titer of 10⁵ in comparison to control animals (n = 12). Mantel-Cox log rank test was applied to determine differences in the risk of infection between groups. (B) Graph of the Spearman rank correlation between ADCC endpoint titers and number of challenges required for infection of each vaccinated animal (n = 12). Animals with ADCC titers below or above the median ADCC endpoint titer are indicated by empty or filled blue circles, respectively. Panels C to F illustrate correlation between Env-specific antibody responses postinfection and challenge outcome. Note that limited plasma volumes did not allow us to assess postinfection antibody responses in all 12 vaccinated animals. Panels C and D show the negative correlation of endpoint 1086.c ADCC titers at week 1 postinfection with the number of challenges required for infection (C) and peak viremia (D) for 9 vaccinated animals. Panel E illustrates that higher ADCC activity at week 1 postinfection was associated with reduced peak viremia (n = 9). (F) In animals that required fewer challenge exposures to infection, the percentage of NK cell-mediated ADCC activity at week 4 postinfection was higher than in animals that required a greater number of exposures to infection (n = 8).