A defined mechanistic correlate of protection against Plasmodium falciparum malaria in non-human primates

Abstract

Malaria vaccine design and prioritization has been hindered by the lack of a mechanistic correlate of protection. We previously demonstrated a strong association between protection and merozoite-neutralizing antibody responses following vaccination of non-human primates against Plasmodium falciparum reticulocyte binding protein homolog 5 (PfRH5). Here, we test the mechanism of protection. Using mutant human IgG1 Fc regions engineered not to engage complement or FcR-dependent effector mechanisms, we produce merozoite-neutralizing and non-neutralizing anti-PfRH5 chimeric monoclonal antibodies (mAbs) and perform a passive transfer-P. falciparum challenge study in Aotus nancymaae monkeys. At the highest dose tested, 6/6 animals given the neutralizing PfRH5-binding mAb c2AC7 survive the challenge without treatment, compared to 0/6 animals given non-neutralizing PfRH5-binding mAb c4BA7 and 0/6 animals given an isotype control mAb. Our results address the controversy regarding whether merozoite-neutralizing antibody can cause protection against P. falciparum blood-stage infections, and highlight the quantitative challenge of achieving such protection.

Conflict of interest statement

A.D.D., D.G.W.A., G.J.W., and S.J.D. are named inventors on patents relating to use of PfRH5 vaccines and anti-PfRH5 antibodies for prevention or treatment of malaria. The remaining authors declare no competing interests.

Figures

Fig. 1

Chimeric mAbs have desired properties for a challenge study. Panel a shows GIA with murine (dashed lines) and chimeric (solid lines) versions of 2AC7 (red lines) and 4BA7 (black lines) against FVO-strain parasites. Error bars indicate range of two independent experiments (with each experiment’s datum being the mean of three replicate wells); lines connect the mean of the experiments. Results of −20 to 20% GIA are regarded as negative. Panel b shows absence of ELISA-detectable interaction of A. nancymaae C1q with mAbs bearing the hIgG1Δnab Fc. Horizontal dotted line indicates background (mean plus two standard deviations of control wells coated with mAbs but not receiving plasma); points and error bars indicate mean and range of two replicate wells per condition. Panels c and d show SPR data demonstrating that the hIgG1Δnab Fc abrogates binding to A. nancymaae Fcγ receptors. Panel c shows reference-subtracted (Fc2-1) binding-response curves for the injection of various antibody analytes (indicated by line characteristics, as shown in the legend to the right) over chips with FcγRIa (left), FcγRIIa (middle), and FcγRIIIa (right) ligands captured on Fc2; in each case, analyte is injected from 0 to 60 s. Panel d summarizes peak (60 s) binding of each antibody to each receptor, expressed as percentage of the response obtained on the same receptor with the wild-type human IgG1 mAb control (EBL040-hIgG1). Panel e shows results of preliminary pharmacokinetic study: anti-PfRH5 ELISA-measured plasma mAb concentrations from days 3 to 22 after administration of a single dose of 30 mg/kg c4BA7 to two A. nancymaae. Points are mean of two animals’ results; error bars show the range, although are not visible where replicates were in close agreement. Line shown is the fitted one-phase exponential decay curve

Fig. 2

A GIA-inducing, but not a non-GIA-inducing, anti-PfRH5 mAb protects against blood-stage P. falciparum. Panels a–c show timecourse of parasitemia in animals receiving EBL040-hIgG1Δnab (panel a), c4BA7 (panel b), or c2AC7 (panel c), each at 100 mg/kg body weight (n = 6 per group). Upper horizontal dashed line indicates the 200,000 parasites/µL (p/µL) threshold for initiation of antimalarial treatment (Rx) because of hyperparasitemia. “P” indicates treatment of an animal due to hyperparasitemia; “+” indicates a single animal found dead on day 19; occasional unexpected deaths have previously been recorded among Aotus both before and during P. falciparum challenge. Panel d shows cumulative parasitemia up to day 10 (on which the first animal in the study reached a treatment endpoint) as per the prespecified primary analysis. Individual datapoints and median are shown. ***P < 0.0001 by one-tailed t test with Welch’s correction. Panel e shows cumulative microscopically-detected parasitemia up to day 35 (the end of the study) vs. qPCR-measured parasitemia on day 35 in the c2AC7 group. Two animals had no detectable parasitemia by either method. Panels f–h show timecourse of parasitemia in the lower-dose study, annotated as for panels a–c; “A” indicates treatment due to anemia. Panel f shows animals receiving no mAb (n = 4); panel g shows animals receiving c4BA7 (n = 3) at 33 mg/kg; panel h shows animals receiving c2AC7 at a range of doses from 33 mg/kg down to 4 mg/kg, as indicated by the legend (n = 2 per dose level)

Fig. 3

Plasma mAb concentrations, GIA and outcome. Panels a and b show ELISA-measured plasma mAb concentrations in the high-dose and lower-dosing studies, respectively, up to the point of reaching a treatment endpoint. Points indicate data from individual animals; lines link group medians. Arrowheads indicate mAb top-up dose administration on days 7 and 14. Panel c shows the relationship between time-weighted mean (TWM) c2AC7 mAb concentration and peak parasitemia in individual animals, combining data from the high-dose and lower-dosing studies (Spearman correlation coefficient rS = −0.8, P = 0.001). Panel d shows timecourses of parasitemia (black circles, left axis) against ELISA-measured c2AC7 concentrations (pink crosses, right axis). Each subplot represents an individual animal from the 100 mg/kg c2AC7 group. Parasitemia data is as shown in panel 2c and ELISA data is as shown in panel 3a. Panel e shows the relationship between concentration of total protein G-purified IgG and GIA (against FVO parasites). Due to small sample volumes, samples from pairs of Aotus (collected on day 10 after challenge) were pooled prior to IgG purification. Results shown for the 100 mg/kg c2AC7 group (n = 6 animals) are the median (points) and range (error bars) of the three such pairs, with each datum being the median of three independent assays. Results for the 33 and 16 mg/kg c2AC7 groups (n = 2 animals each) are each from a single pair of animals, showing the median (points) and range (error bars) of three independent assays. All assays had triplicate wells at each tested IgG concentration. Results are not shown for animals from other groups, in which GIA did not reach 50% at the maximum tested concentration: GIA was <20% at 0.75 mg/mL of IgG from animals in the EBL040, c4BA7 and no mAb control groups, and <30% at 0.75 mg/mL for the 8 and 4 mg/kg c2AC7 groups