Analysis of Fc-dependent effector functions of anti-malaria circumsporozoite protein antibodies

Abstract

Antibodies targeting the malaria circumsporozoite protein (CSP) can prophylactically protect against malaria by targeting Plasmodium parasites before they establish symptomatic blood-stage disease. Engineering the antibody Fc region to more effectively engage immune effector functions has produced therapeutic antibodies with enhanced potency against viral and oncological targets. However, whether Fc-dependent immune effector functions can contribute to the protection of malaria CSP mAbs or be further enhanced via engineering has been limitedly tested. Here, we report that Fc-dependent effector functions are required for achieving maximal protection via prophylactic treatment with the CSP mAb 317. We further report that Fc engineering modulated the activity of multiple CSP mAbs in multiple in vitro assays of effector function. Our studies revealed that the mAbs L9 and CIS43 were more potent drivers of antibody-dependent phagocytosis, NK activation and killing, and complement deposition. In contrast, 317, but not L9 and CIS43, drove enhanced activation of CSP-responsive T-cells after DC acquisition of mAb-complexed antigens. Collectively, our data suggest that effector function represents an important mechanism of anti-CSP antibodies with the potential to enhance activity through Fc engineering.IMPORTANCEMalaria disease imposes a major burden on global health, causing over half a million annual deaths. Recent clinical trials in humans have shown that therapeutic antibodies can provide prophylactic protection against malaria in target populations. However, the cost of goods for therapeutic antibodies is high, and the malaria disease burden is concentrated in resource-challenged regions. Engineering the antibody Fc domain to more efficiently engage the immune system is an appealing strategy to increase the potency of therapeutic antibodies but has been minimally tested for malaria. Here, we present evidence that the Fc domain of malaria therapeutic antibodies can confer protection in animal models and can be engineered for more potent stimulation of diverse parasite-targeting immune responses.

Keywords: antibody functions; antimalarial agents; effector functions; malaria; monoclonal antibodies.

Fig 1

Fc domain of CSP-binding mAb 317, but not L9, was required for maximal protection. (A) Schematic of the challenge model. (B–C) Parasite burden was measured by IVIS optical imaging at d2pi (B) and d5pi (C). Control mice were treated with either PBS (closed markers) or an LS-modified IgG1 isotype control antibody (open markers). Radiance was measured in an ROI enclosing the liver (B) or whole mouse (C) to quantify parasite burden. (D) Concentration of CSP mAbs in serum collected at d5pi. (E) Percentage of purified salivary gland sporozoites staining as nonviable after co-incubation with indicated mAbs. (B–E) Pooled data from two independent experiments are shown, with values normalized to the value of the mAb-untreated IV challenge control group from each experiment. Matched marker shapes (circles and squares) indicate data originating from the same experiment. Results of one-way ANOVA with Kruskal-Wallis post-test are indicated for B, C, and one-way ANOVA with Brown-Forsythe and Welch post-test are indicated for E; all calculated P-values are shown, with statistically significant comparisons highlighted with green text.

Fig 2

Binding analysis of Fc-engineered CSP mAbs. (A–B) Binding affinity of Fc-engineered CIS43 variants to a panel of FcgRs was measured by SPR and to C1q by Octet BLI. (A) shows binding affinity, and (B) shows fold-change in affinity, relative to CIS43-LS mAb, with red shading highlighting increased affinity and blue decreased affinity. Data shown are average values calculated from two technical replicates, except for two receptors (*), FcgR2aH and FcgR3aV, calculated from a single technical replicate. KD values could not be quantitated for the N297G variant due to undetectable binding to all receptors. (C–E) Binding of mAbs to live parasites upon co-incubation with 5 ng/mL mAb was assessed by flow cytometry. (C) Percent of parasites harvested from either mosquito salivary glands (green markers) or midguts (blue markers) opsonized by indicated mAbs is shown with pooling data from two independent experiments, each with three technical replicates. (D, E) gMFI corresponding to mAb binding intensity is shown for opsonized parasites in two independent experiments. All statistically significant (P < 0.05) comparisons were assessed (C–E) using a two-way ANOVA using Sidak’s multiple comparison test to assess differences between Fc variants compared to the LS control Fc, regardless of whether parasites originated from the midgut or salivary glands. Fc mAb variants not tested due to limited parasite numbers, and are indicated by nd (not done).

Fig 3

Antibody-dependent phagocytosis was influenced by the Fc region of CSP mAbs. (A–H) Phagocytosis of rCSP-coated beads by THP-1 monocytes (A–D) or primary human neutrophils (E–H) was measured via flow cytometry, following co-incubation of mAbs at indicated dilution series with rCSP-coated beads. Data show average values from two technical replicates, with error bars indicating SD. Two separate donors were used to source neutrophil technical replicates. P values in B–D and F–H indicate statistically significant differences between the area under the curve of the respective curve of the Fc variant compared to the LS control Fc, using one-way ANOVA with Dunnett’s multiple comparisons test (I) Phagocytosis of live Pb-PfCSP-Luc-GFP parasites by primary human neutrophils, following co-incubation with 5 ng/mL mAbs. Plots show data from 2 to 3 independent experiments for each mAb, each with three technical replicates. Matching marker shapes (circles, squares, and triangles) indicate data originating from the same experiment. P-values in black indicate calculated results of Sidak’s multiple comparisons tests following a two-way ANOVA comparing differences between mAb-LS Fc variants to the LS control, regardless of whether parasites originated from the midgut or salivary gland.

Fig 4

Antibody-dependent NK cell activation and killing were influenced by the Fc region of CSP mAbs. (A–L) CSP antigen-induced expression of NK cell activation markers IFN-γ (A–D), CD107a (E–H), and MIP1-β (I-L) was assessed in NK cells following 5 h of co-incubation of NK cells on rCSP-coated plates pre-incubated with indicated mAbs. (M–P) Specific killing of rCSP-pulsed target cells by NK cells after 4 h co-incubation. All data shows average values of two technical replicates, and error bars indicate SD. NK cells were sourced from unique leukopak donors for each technical replicate. Statistical comparisons of CIS43 or L9 vs 317 with wild-type Fc used two-way ANOVA with Sidak’s (A, E, I, and M), and of LS-Fc variants vs the LS-Fc of each mAb used Dunnett’s correction (B–D, F–H, J–L, and N–P) for multiple comparisons. Level of significance: *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001.

Fig 5

Antibody-dependent complement deposition was influenced by the Fc region of CSP mAbs. (A–D) Deposition of C3 complement protein on rCSP-coated beads was detected by flow cytometry, following 50 min incubation with indicated mAbs. Data plots show gMFI for C3 detection antibody from two technical replicates; error bars indicate SD. Statistical comparisons of CIS43 vs L9 vs 317 with wild-type Fc used two-way ANOVA with Sidak’s (A) and of LS-Fc variants vs the LS-Fc of each mAb used Dunnett’s correction (B, C, and D) for multiple comparisons. Level of significance: *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001.

Fig 6

Activation of CSP-specific reporter Jurkat line by CSP antibody immune complexes. (A–C) A CSP-responsive reporter Jurkat T-cell line was used to assess the ability of immune complexes formed of rCSP antigen and CSP mAbs to stimulate T-cell activation. Assay set-up is schematized in (A), and representative flow plots shown in (C). (B) Percent-activated CSP reporter T-cells in two independent experiments. Each experiment included three biological replicates of PBMCs sourced from unique donors (donors indicated by marker shapes). P-values indicate the results of Sidak’s multiple comparison test with two-way ANOVA, comparing LS and LS-N297G variants of each mAb.