Detection of naturally acquired, strain-transcending antibodies against rosetting Plasmodium falciparum strains in humans

Abstract

Strain-transcending antibodies against virulence-associated subsets of P. falciparum-infected erythrocyte surface antigens could protect children from severe malaria. However, the evidence supporting the existence of such antibodies is incomplete and inconsistent. One subset of surface antigens associated with severe malaria, rosette-mediating Plasmodium falciparum Erythrocyte Membrane Protein one (PfEMP1) variants, cause infected erythrocytes to bind to uninfected erythrocytes to form clusters of cells (rosettes) that contribute to microvascular obstruction and pathology. Here, we tested plasma from 80 individuals living in malaria-endemic regions for IgG recognition of the surface of four P. falciparum rosetting strains using flow cytometry. Broadly reactive plasma samples were then used in antibody elution experiments in which intact IgG was eluted from the surface of infected erythrocytes and transferred to heterologous rosetting strains to look for strain-transcending antibodies. We found that seroprevalence (percentage of positive plasma samples) against allopatric rosetting strains was high in adults (63%-93%) but lower in children (13%-48%). Strain-transcending antibodies were present in nine out of eleven eluted antibody experiments, with six of these recognizing multiple heterologous rosetting parasite strains. One eluate had rosette-disrupting activity against heterologous strains, suggesting PfEMP1 as the likely target of the strain-transcending antibodies. Naturally acquired strain-transcending antibodies to rosetting P. falciparum strains in humans have not been directly demonstrated previously. Their existence suggests that such antibodies could play a role in clinical protection and raises the possibility that conserved epitopes recognized by strain-transcending antibodies could be targeted therapeutically by monoclonal antibodies or vaccines.

Keywords: PfEMP1; antibodies; epitopes; immunology; rosette formation; severe malaria; virulence.

Fig 1

Recognition of infected erythrocytes expressing rosetting or DC8 PfEMP1 variants by human IgG. (A) Adults’ plasma samples. (B) Childrens’ plasma samples. Four rosetting parasite strains (expressing the PfEMP1 variants 11019VAR1, HB3VAR6, TM284VAR1, and ITVAR60) and two DC8-expressing human brain endothelial cell-binding parasite strains (expressing 3D7_PFD0020c and ITVAR19) were incubated with human plasma at 1/10 dilution. The infected erythrocytes expressing the specific variant of interest were detected with rabbit anti-PfEMP1 IgG as described in Fig. S1, and the human IgG bound to those cells was detected with an Alexa Fluor 488-conjugated anti-human IgG (gamma chain) antibody at 1/1,000 dilution. The corrected median fluorescence intensity (cMFI) was determined by subtracting non-specific background staining as described in the methods and Fig. S2. Each plasma sample was tested in duplicate in each experiment, and the mean of the corrected MFI from two independent experiments for each parasite strain is shown. The number of positive plasma samples (cMFI >100) for each parasite strain is shown at the base of each column. The age in years of each individual at the time of plasma donation is shown in square brackets if known. Samples from children with severe malaria are underlined. *Samples used in eluted antibody experiments.

Fig 2

Method of detection of strain-transcending IgG in eluates. (A) Schematic diagram of the acid elution method. (B) Successful detection of IgG in eluates. Left, dot-plots showing the gating strategy used to identify mature infected erythrocytes in whole cultures. Forward and side scatter were used to exclude debris and set a gate on all erythrocytes, then mature pigmented-trophozoite and schizont-infected erythrocytes were identified using Vybrant Dyecycle Violet (DNA stain, 1/2,500 dilution) and Ethidium Bromide (DNA/RNA stain, 20 µg/mL) as the DNA and RNA high population (40). Right, fluorescence intensity histogram of ITvar60R+ mature infected erythrocytes stained with neat eluate of plasma M2 from ITvar60R+ (blue), compared to a concentration-matched human IgG control (gray). Bound human IgG was detected with an Alexa Fluor 647-conjugated anti-human IgG (gamma chain) antibody at 1/1,000 dilution. A bimodal staining pattern is expected if the eluted antibodies recognize a VSA such as PfEMP1, which is expressed by some but not all of the infected cells in the culture.

Fig 3

Summary of recognition of homologous and heterologous rosetting infected erythrocytes by eluted antibodies using flow cytometry. Eight rosetting parasite strains were stained with antibodies eluted from ITvar60R+ (eight individual donors; four adults M2, M3, M4, M6, and 4 children G7, G9, G12, and G14) or TM284R+ (three individual donors; two adults M2 and M4 and 1 child G9). Mature infected erythrocytes were identified using Vybrant Dyecycle Violet at 1/2,500 dilution (DNA stain) and ethidium bromide at 20 µg/mL (DNA/RNA stain) (40). Eluates were used neat, and human IgG bound to mature infected erythrocytes was detected with an Alexa Fluor 647-conjugated anti-human IgG (gamma chain) antibody at 1/1,000 dilution. Corrected median fluorescence intensities (cMFI) were adjusted for background staining as described in the methods. Due to the limited availability of eluate, the experiments were performed once.

Fig 4

Rosette-disrupting activity of eluates. Rosette frequency of parasite strains TM284R+, 9197UNR+, and HB3R+ (indicated in the key) incubated in the presence of a negative control eluate that does not recognize mature infected erythrocytes, a positive control (antisera to the relevant PfEMP1 DBLα domain or heparin), or selected eluates which showed surface recognition against the tested parasite strain. The plasma and eluting strain for each eluate is indicated on the X-axis. Results from three independent experiments are shown as individual data points. Bar heights show the mean of the three experiments, and error bars indicate the SD. Significant differences between the negative control and tested eluate by student’s t-test or one-way ANOVA are shown, ***P < 0.001, *P < 0.05. All differences between the negative and positive controls were statistically significant (P < 0.0001).