Passive immunization of macaques with polyclonal anti-SHIV IgG against a heterologous tier 2 SHIV: outcome depends on IgG dose

Abstract

Background: A key goal for HIV-1 envelope immunogen design is the induction of cross-reactive neutralizing antibodies (nAbs). As AIDS vaccine recipients will not be exposed to strains exactly matching any immunogens due to multiple HIV-1 quasispecies circulating in the human population worldwide, heterologous SHIV challenges are essential for realistic vaccine efficacy testing in primates. We assessed whether polyclonal IgG, isolated from rhesus monkeys (RMs) with high-titer nAbs (termed SHIVIG), could protect RMs against the R5-tropic tier-2 SHIV-2873Nip, which was heterologous to the viruses or HIV-1 envelopes that had elicited SHIVIG.

Results: SHIVIG demonstrated binding to HIV Gag, Tat, and Env of different clades and competed with the broadly neutralizing antibodies b12, VRC01, 4E10, and 17b. SHIVIG neutralized tier 1 and tier 2 viruses, including SHIV-2873Nip. NK-cell depletion decreased the neutralizing activity of SHIVIG 20-fold in PBMC assays. Although SHIVIG neutralized SHIV-2873Nip in vitro, this polyclonal IgG preparation failed to prevent acquisition after repeated intrarectal low-dose virus challenges, but at a dose of 400 mg/kg, it significantly lowered peak viremia (P = 0.001). Unexpectedly, single-genome analysis revealed a higher number of transmitted variants at the low dose of 25 mg/kg, implying increased acquisition at low SHIVIG levels. In vitro, SHIVIG demonstrated complement-mediated Ab-dependent enhancement of infection (C'-ADE) at concentrations similar to those observed in plasmas of RMs treated with 25 mg/kg of SHIVIG.

Conclusion: Our primate model data suggest a dual role for polyclonal anti-HIV-1 Abs depending on plasma levels upon virus encounter.

Figure 1

SHIVIG characterization. A. SHIVIG binding to soluble HIV and SIV proteins was tested by ELISA. Proteins were captured on the plates and probed with serial dilutions of SHIVIG. Binding was detected as described in Methods. Each data point represents the mean ± SEM (n = 3). Env proteins were derived from the following HIV or SHIV strains: clade A, UG37; B, BaL; C, CN54 and SHIV-1157ip; D, UG21. B. Competitive ELISA with the CD4-binding site-specific bnmAbs b12 and VRC01; gp41 MPER-specific bnmAb 4E10; and CD4-inducible site-specific bnmAb 17b. Plates were coated with HIV-1_CN54 gp120 for b12 and VRC01 analysis or SHIV-1157i gp160 for 4E10 and further incubated with 0.5 μg/ml of the bnmAb indicated along with different concentrations of SHIVIG in duplicates or triplicates. For 17b mAb, HIV-1_CN54 gp120 was captured by polyclonal sheep anti-gp120 Abs coated on the plate and the assay was continued as described in Methods. The y-axis indicates OD percentage of maximal binding, which is determined as the reading without SHIVIG (100% binding is marked by dashed line). The irrelevant mAb Fm-6 (anti-SARS virus) and naïve RM IgG were used as negative controls (not depicted). C. ELISA of SHIVIG with consensus clade C Env peptides. The x-axis designates pools of peptides assigned to HIV-1 Env regions. Dotted line represents background. IDR, immunodominant region; UDR, undefined region in gp41; CH, C-terminal heptad region; TMR, transmembrane region. Each data point represents the mean ± SEM (n = 3). All experiments were repeated at least twice.

Figure 2

In vitro neutralization and effector functions of SHIVIG against the challenge virus, SHIV-2873Nip. A. Neutralizing activity of SHIVIG in PBMC assays with or without NK cells. Serially diluted SHIVIG was assayed with unfractionated human PBMC (red) and PBMC depleted of NK cells (blue) in triplicates. VRC01 was used as a positive control and was analyzed with unfractionated human PBMC (solid grey line) and PBMC depleted of NK cells (dashed grey line) as described in Methods. B. ADCVI activity of SHIVIG and HIVIG (IgG from a pool of HIV-positive donors; positive control; squares) was assessed against SHIV-2873Nip as described in Methods. The graph shows the percentage of virus inhibition (y axis) by increasing concentrations of SHIVIG (red) or HIVIG (grey) normalized by values obtained for negative controls at the same concentrations. C. ADCC activity of serially diluted SHIVIG (red) and SHIVIG-derived Fab (grey) were tested in triplicates with CEM-NKr target cells coated with gp120 of HIV_96ZM651 (clade C) and human PBMC as effector cells. The graph shows killing (in percentage) of target cells in the presence of increasing concentrations of SHIVIG or SHIVIG-Fab. Experiments were repeated in triplicate, and representative mean results are shown.

Figure 3

Study design and SHIVIG pharmacokinetics. A. Animal study design and timeline. Four groups of female RMs were enrolled. Group 1 (n = 6) received 400 mg/kg of SHIVIG, Group 2 (n = 2), 675 mg/kg, and Group 3 (n = 6), 25 mg/kg, respectively. Group 4 (n = 14) RMs served as virus-only controls. Ten animals were enrolled and four additional animals used for virus titration given the identical virus dose served as additional controls. The challenge virus, SHIV-2873Nip, had been titrated i.r. to yield systemic infection (>10⁴ viral RNA copies) in untreated monkeys after a maximum of 5 weekly challenges (small red arrows); the challenge dose was 5,000 50% tissue culture infectious doses (TCID₅₀) measured by TZM-bl assay. SHIVIG infusions (large blue arrows) as well as viral challenges were stopped after the monkeys became viremic (>10⁴ viral RNA copies/ml). B. The number of SHIVIG administrations that RMs from different groups received while their vRNA loads were <10⁴ copies/ml. C-E. SHIVIG pharmacokinetics in RM groups. Large blue arrows indicate biweekly SHIVIG administrations and small red arrows indicate SHIV-2873Nip challenges; C. Group 1 (400 mg/kg of SHIVIG, n = 6). D. Group 2 (675 mg/kg of SHIVIG, n = 2); E. Group 3 (25 mg/kg of SHIVIG, n = 6). The insert represents the same graph showing the lower Y axis range. Serial dilutions of RM plasmas were incubated on ELISA plates to which HIV-1 BaL gp120 had been added and preimmobilized. Binding was assessed as described in Methods. All samples were assayed in triplicate.

Figure 4

Plasma viral RNA (vRNA) loads after SHIV-2873Nip challenges. A-D. Graphs represent vRNA loads after viral challenges for Group 1–4, respectively. Small red arrows indicate 5 low-dose SHIV-2873Nip challenges. Animals in Group 2 B became infected after the second viral challenge and further challenges were omitted. D. Virus-only controls. Three animals from this group could not be challenged at week 4 (inclement weather forced closure of the primate center) and one RM (*RSc-12) became viremic only after the high-dose challenge (30,000 TCID₅₀) at week 7 (large red arrow). During the 2-week interval between the 4^th and 5^th low-dose challenge, this RM had developed SIV Gag-specific proliferative responses (data not shown). E. Statistical analysis of the peak of vRNA loads. The dashed line represents the limit of detection (50 copies/ml). For control Group 4, peak vRNA load was calculated including the animals with protocol breach. Peak viremia levels were compared using negative binomial regression, in order to adequately model the non-normality of the observed peak levels. Raw p-values from pair-wise comparisons with the control group were adjusted using Dunn’s method.

Figure 5

Single-genome analysis (SGA) and C’-ADE. A. Number of SHIV-2873Nip quasispecies at peak viremia. The number of quasispecies was analyzed by SGA of the V1/V2 loop region of SHIV-2873Nip env. For every RM, at least 10 individual clones were obtained by limiting dilution PCR followed by sequencing. Sequence readout was performed using both strands of DNA. Statistical significance was assessed by Mann–Whitney test with Holm correction (P = 0.032). B-D. C’-ADE activity of SHIVIG in SupT1.R5 cells tested against different viruses built from the HIV_NL4-3 backbone and containing the envelope of various SHIV or HIV-C strains as well as a Renilla luciferase (LucR) reporter gene. B. NL-LucR.2873Ni; C. NL-LucR.2873Nipd; and D. NL-LucR.1157ipEL. SHIVIG (red solid and green dashed lines) or nRM IgG (grey solid and dashed lines) were assayed in presence of 10% fresh (solid lines) or heat-inactivated (dashed lines) normal human serum as a source of complement. Each data point represents the mean ± SEM (n = 3). All experiments were repeated at least twice.