Characterization of VAR2CSA-deficient Plasmodium falciparum-infected erythrocytes selected for adhesion to the BeWo placental cell line

Abstract

Background: Malaria in pregnancy is characterized by accumulation of infected erythrocytes (IE) in the placenta. The key ligand identified as mediating this process is a Plasmodium falciparum erythrocyte membrane protein 1 family member, termed VAR2CSA. VAR2CSA appears to be the main ligand responsible for adhesion to chondroitin sulphate A (CSA). Whether other PfEMP1 molecules can also mediate placental adhesion, independent of CSA binding, is unclear.

Methods: The parasite line CS2 carrying a disrupted var2csa gene (CS2KO) was selected for adhesion to the BeWo choriocarcinoma cell line, which has been proposed as a model for placental malaria. The selected and control IE were tested for adhesion to placental sections and flow cytometry was used to measure recognition of IE by three serum sets from malaria-exposed men and women.

Results: Wild-type CS2 adhere to BeWo and placental tissue via CSA. CS2KO IE were successfully selected for adhesion to BeWo, and adhered by a CSA-independent mechanism. They bound to immobilized ICAM-1 and CD36. BeWo-selected CS2KO bound at moderate levels to placental sections, but most binding was to placental villi rather than to the syncytiotrophoblast to which IE adherence occurs in vivo. This binding was inhibited by a blocking antibody to CD36 but not to ICAM-1. As expected, sera from malaria-exposed adults recognized CS2 IE in a gender and parity dependent manner. In one serum set, there was a similar but less pronounced pattern of antibody binding to selected CS2KO IE, but this was not seen in two others. One var gene, It4var19, was particularly abundant in the selected line and was detected as full length transcripts in BeWo-selected IE, but not unselected CS2KO.

Conclusion: This study suggests that IE with characteristics similar to the CS2KO have a limited role in the pathogenesis of placental malaria. VAR2CSA appear to be the major ligand for placental adhesion, and could be the basis for a vaccine against pregnancy malaria.

Figure 1

Adhesion of CS2KO IE to BeWo cells following repeated cycles of selection. Adhesion is expressed as IE bound per 100 BeWo nuclei, in the presence (black bars) or absence (white bars) of CSA 10 μg/ml. Data shown are the mean + SEM of at least two independent experiments, performed in triplicate.

Figure 2

Adhesion of (a) CS2 IE and (b) CS2KO9 IE to immobilized receptors, expressed as IE bound/mm². Data shown are the mean + SEM of two or more independent experiments, performed in triplicate. CSA, chondroitin sulphate A; CSA+10 μg/ml, Adhesion to CSA in presence of free CSA at 10 μg/ml; HCSPG, human chondroitin sulphate proteoglycan; ICAM-1, intercellular adhesion molecule 1. HA, hyaluronic acid; FG, fibrinogen.

Figure 3

Adhesion to human placental tissue cryosections. Adhesion levels of IE in the intervillous space (IVS), to syncytiotrophoblast (SCT) and over the villus (V) is shown compared to untreated (control) sections present on each slide. (a) Adhesion of CS2 (white bars) and CS2KO9 (black bars) IE in the presence of CSA (100 μg/ml) or following chondroitinase ABC pretreatment (cABC) (0.5 units/mL). (b) Adhesion of CS2 (white) and CS2KO9 (black bars) IE in the presence of CSA or following anti-CD36 pre-treatment (0.5 μg/ml). (c) Adhesion of E8B (white) and CS2KO9 (black) IE in presence of CSA or following pretreatment with anti-ICAM-1 (10 μg/ml). Adhesion was counted at 40× magnification and expressed as average IE bound per field (+SEM) for at least 3 independent experiments.

Figure 4

IgG recognition by sera from malaria-exposed individuals. (a) CS2 (b) CS2KO9 and (c) E8B IE were incubated with sera from Malawian men, primigravid and multigravid women (cohort 1). (d) CS2 and (e) CS2KO9 IE were incubated with sera from men, primigravid and multigravid women from Madang, Papua New Guinea (cohort 2). (f) CS2 and (g) CS2KO IE were incubated with sera from men, primigravid, and multigravid women from Malawi (cohort 3). Data presented are mean fluorescence intensities (MFI) for individual samples. Bars indicate median MFI for the population. Significant differences in responses were found as follows: (a) cohort 1, CS2: men and primigravidae (PG; p < 0.0001), men and multigravidae (MG; p < 0.0001) and PG and MG (p = 0.001); (b) cohort 1, CS2KO9: men and PG (p = 0.0011), men and MG (p < 0.0001), and PG and MG (p = 0.047); (c) cohort 1, E8B: men and MG (p = 0.047) (f) cohort 3, CS2, men and PG (p = 0.001), men and MG (p = 0.0013).

Figure 5

(a) Northern blot analysis of CS2KO9 and CS2KO consecutively probed with IT4var19 and var exon two sequences (exon2). (b) Quantities of different var gene transcripts in BeWo selected CS2KO9 and unselected CS2KO parasites. Absolute quantification of var gene transcripts was performed by quantitative RT-PCR using standard curves of plasmids containing the var gene target sequences. Presented data is normalized against 18S rRNA levels so that molecules of var cDNA from equivalent numbers of cells for each parasite line are compared. Data are presented as mean + standard deviation.