Afucosylated Plasmodium falciparum-specific IgG is induced by infection but not by subunit vaccination

Abstract

Plasmodium falciparum erythrocyte membrane protein 1 (PfEMP1) family members mediate receptor- and tissue-specific sequestration of infected erythrocytes (IEs) in malaria. Antibody responses are a central component of naturally acquired malaria immunity. PfEMP1-specific IgG likely protects by inhibiting IE sequestration and through IgG-Fc Receptor (FcγR) mediated phagocytosis and killing of antibody-opsonized IEs. The affinity of afucosylated IgG to FcγRIIIa is up to 40-fold higher than fucosylated IgG, resulting in enhanced antibody-dependent cellular cytotoxicity. Most IgG in plasma is fully fucosylated, but afucosylated IgG is elicited in response to enveloped viruses and to paternal alloantigens during pregnancy. Here we show that naturally acquired PfEMP1-specific IgG is strongly afucosylated in a stable and exposure-dependent manner, and efficiently induces FcγRIIIa-dependent natural killer (NK) cell degranulation. In contrast, immunization with a subunit PfEMP1 (VAR2CSA) vaccine results in fully fucosylated specific IgG. These results have implications for understanding protective natural- and vaccine-induced immunity to malaria.

Trial registration: ClinicalTrials.gov NCT02647489.

Fig. 1. Background and study workflow.

a IgG1 specific for the merozoite antigen GLURP and two members of the PfEMP1 family expressed on the surface of IEs were analyzed in this study. Most PfEMP1 variants facilitate sequestration of IEs to vascular endothelium (exemplified here by VAR6), while VAR2CSA-type PfEMP1 mediate IE sequestration in the placental syncytiotrophoblast and intervillous space. b Plasma samples were split and used to purify total plasma IgG1 and antigen-specific IgG1, using protein G-coupled sepharose and solid-phase absorption with recombinant antigens, respectively. Eluted IgG1 was digested with trypsin and the glycopeptides analyzed by liquid chromatography mass spectrometry (LC-MS). Examples of MS spectra of total IgG1 (left) and antigen-specific (anti-VAR6) IgG1 (right) from one sample is shown. c The fractions of the different glycosylation traits of the Fc glycan depicted were calculated from LC-MS spectra.

Fig. 2. Fc fucosylation of naturally acquired P. falciparum-specific IgG depends on antigen location and exposure.

a Fc fucosylation levels of total plasma IgG1 (gray, n = 127) and IgG1 specific for VAR2CSA (orange, n = 117), VAR6 (green, n = 121), and GLURP (blue, n = 88) in Ghanaian pregnant women (left four panels). Fc fucosylation levels of total plasma IgG1 from unexposed Dutch women (n = 5) were included for comparison (right panel). Medians and densities are shown. Statistically significant pairwise differences between antigen-specific IgG and total IgG (multiple two-sided Wilcoxon signed-rank tests with Bonferroni correction) are indicated (****P < 0.0001). b–e Correlations of b VAR2CSA-, c VAR6-, d GLURP-specific, and e total IgG1-Fc fucosylation levels with parity. P values and correlation coefficients are shown. Statistical significance of correlations (Spearman’s correlations. *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001.

Fig. 3. Fc fucosylation levels of VAR2CSA-specifc IgG is temporally stable.

a Fc fucosylation levels of total plasma IgG1 (gray, n = 72) and IgG1 with specificity for VAR2CSA (orange, n = 50), VAR6 (green, n = 65), and GLURP (blue, n = 43) in non-pregnant Ghanaian women exposed to VAR2CSA during one or more previous pregnancies. Fc fucosylation levels of total plasma IgG1 from unexposed Dutch females (n = 5) are included as controls. Medians and densities are shown. Statistically significant pairwise differences between antigen-specific IgG and total IgG (multiple two-sided Wilcoxon signed-rank tests with Bonferroni correction) are indicated (****P < 0.0001). b Correlation between fucosylation levels of VAR2CSA-specific IgG1 and parity. c Correlation between fucosylation levels of VAR2CSA-specific IgG1 and time since last pregnancy. Statistical significance of correlations are shown (Spearman’s correlations. *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001).

Fig. 4. VAR2CSA-specific IgG induced by subunit vaccination is not Fc afucosylated.

Fc fucosylation levels of total (gray) and VAR2CSA-specific (orange) plasma IgG1 in German vaccinees (n = 32) without (a) and in Beninese vaccinees (n = 18) with (b) natural exposure to P. falciparum. Medians and densities are shown. Statistically significant pairwise differences between antigen-specific IgG and total IgG (two-sided Wilcoxon signed-rank test) are indicated (****P < 0.0001).

Fig. 5. Afucosylated PfEMP1-specific IgG induces NK-cell-mediated ADCC.

a VAR2CSA-specific IgG fucosylation levels of samples with low (open symbols) and high fucose (filled symbols) levels. b Comparison of VAR2CSA-specific IgG levels and c CD107a expression between highly fucosylated (filled symbols) and afucosylated anti-VAR2CSA IgG1 (open symbols) samples, respectively. Mean from three independent experiments and P values from two-sided Mann–Whitney tests are shown. d VAR2CSA-specific, human monoclonal antibody PAM1.4 as either fucosylated or afucosylated IgG1 was titrated in the same assay and VAR2CSA-binding or e degranulation activity (CD107a expression) on NK92-CD16a cells was measured. IgG: non-immune human IgG. Data represent mean values ± SD from three independent experiments.