INTRODUCTION

Top Abstract Introduction Materials and Methods Results and Discussion References

Plasmodium falciparum, a parasite responsible for the most severe forms of malaria, is a major cause of morbidity and mortality in the world. In areas where it is endemic, most people are frequently infected but the main part of the clinical burden is borne by young children and pregnant women, while few adults are usually clinically affected. Interestingly, among pregnant women, primiparae are more susceptible to malaria than multiparae (16). Our work aims at understanding the rationale for the higher susceptibility of primigravidae to malaria, in order to help in targeting and performing oriented public health programs.

It has long been known that repeated and continuous exposure to malaria parasites during infancy leads to premunition, an immune state defined as the symptomless persistence of a low number of parasites. The development of premunition has been extensively studied, and there is now good evidence that it is related to the acquisition of a repertoire of protective antibodies reactive against polymorphic molecules prominently exposed at the surface of the infected erythrocyte (2, 12). The best characterized of these molecules are collectively referred to as P. falciparum-erythrocyte membrane protein 1 (PfEMP1). Although the P. falciparum genome is able to encode around one hundred PfEMP1 molecules, these molecules are expressed not simultaneously but one at a time. In addition to their antigenic properties, PfEMP1 molecules mediate the cytoadherence of infected erythrocytes to a variety of endothelium cells that express receptors. Receptors potentially bound by P. falciparum-infected erythrocytes include CD36, intercellular adhesion molecule 1 (ICAM-1), thrombospondin, and chondroitin sulfate A (CSA), which are bound in vitro by some PfEMP1 members (1, 19). In addition, E selectin, vascular cell adhesion molecule 1 (VCAM-1), and platelet endothelial cell adhesion molecule 1 (PECAM1) are bound by P. falciparum-infected erythrocytes, but their properties of binding to PfEMP1 remain to be assessed.

In pregnant women, a large amount of mature parasites is sequestered in the maternal space of the placenta (22). It has been suggested that parasites could cytoadhere to the syncytiotrophoblast, the outermost line of the placenta in contact with maternal blood (6). It has been demonstrated that the syncytiotrophoblast expresses a large amount of CSA (13). It has also been shown that erythrocytes parasitized by mature parasites from pregnant women bind to the syncytiotrophoblast by adhering to CSA, while parasites from nonpregnant hosts do not usually bind to CSA (6, 13). Consequently, parasites from pregnant women may express a PfEMP1 variant not expressed in parasites from nonpregnant subjects. We hypothesized that women are not immune to placental parasites before being pregnant and that a first pregnancy allows them to develop an immune response against the CSA-parasite binding receptor that protects them from malaria during subsequent pregnancies.

In a cross-sectional study carried out in Yaoundé, Cameroon, we assessed the ability of sera from pregnant women to agglutinate parasite isolates from pregnant women and a P. falciparum strain that binds only to CSA (and was therefore used to mimic placental parasites). We next described the acquisition of immunity against pregnancy-associated parasites (PAPs) in women longitudinally monitored in Ebolowa, Cameroon, during their first two pregnancies. Finally we present evidence that antibodies directed against PAPs acquired during the first infected pregnancy inhibit the cytoadherence of placental parasites to the human syncytiotrophoblast and may account for the lower frequency of malaria in multigravidae.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results and Discussion References

Samples from Yaoundé. In this study, we enrolled all women delivering babies in the maternity wards of Nkolndongo, Yaoundé, Cameroon, from June 1996 to April 1997, after they gave their oral informed consent. Women delivering during weekends were excluded. After the women had delivered, blood samples were taken by puncture and plasma was frozen. A crush smear was made from an excised piece of placenta. Placental blood thick films were air dried, Giemsa stained, read by microscopy over 50 fields at a ×1,000 magnification, and considered positive when P. falciparum parasites or malarial pigments were observed. Peripheral blood parasites were cryopreserved. Nonpregnant subjects (women and men) were recruited in the dispensaries of Nkolndongo and Messa, in the same town. Plasma samples from all participants were frozen, and parasites, if any were isolated, were cryopreserved.

Serum samples from Ebolowa. To study the evolution of P. falciparum-reactive antibodies during pregnancy, we used plasma samples collected in a longitudinal study conducted from 1991 to 1992 in Ebolowa, Cameroon (5). In this study, 50 women were monitored from their first to second pregnancies. Blood samples were drawn in the 6th month of first pregnancies, at first delivery, six months after delivery, in the 6th month of second pregnancies (if any), and then at second delivery. Cord blood samples were also collected. All sera were kept frozen at 20°C until being tested.

Parasite cultures. The RP5 P. falciparum line (a gift from J. Gysin, Laboratoire de Génétique et d'Immunologie, IMTSSA, Parc du Pharo, Marseille, France) binds to CSA and not to the other known receptors of P. falciparum (8) and consequently binds to the human syncytiotrophoblast (14). In our laboratory, the binding phenotype was maintained by a fortnight flotation on plasmagel (18). Three parasite isolates from pregnant women, four from nonpregnant women, and the RP5 strain were thawed and cultivated in candle jars according to standard procedures (21) at a 5% hematocrit with 10% heat-inactivated human AB serum added to RPMI 1640-HEPES (25 mM). All tests were performed when parasites were in the late stage (from late trophozoite to young schizont). Parasites from isolates were used during the first life cycle.

Agglutination test. Serum antibodies to infected erythrocytes (IEs) were detected by a modification of the antibody-mediated agglutination assay (11). Serum (2.5 µl) was deposited in a 96-well microtitration plate (U bottom). A parasite culture at the mature stage was washed and resuspended in phosphate-buffered saline, pH 7.4, at an 11% hematocrit, and 22.5 µl of this suspension containing 0.01% acridine orange was added into each well (final hematocrit, 10%; final serum concentration, 10%). After a 90-min rotation at room temperature on a Coulter mixer (a 45° inclination on a 22-round-per-minute rotating dish), 50 µl of phosphate-buffered saline was added and 20 µl of the suspension was examined between an examination slide and a 22- by 22-mm cover slide. Agglutinates were examined under UV and bright-field illumination. The assay result was considered positive when at least five agglutinates of at least three IEs were counted, and the result was quantified by the geometric mean of the five biggest agglutinates.

Inhibition of the cytoadherence to human trophoblast by immune-phase sera. The effect of sera on the cytoadherence of the RP5 strain was assessed by using a modification of the cytoadherence assay previously described (14). Briefly, cytotrophoblasts were purified from a human placenta by negative immunoselection for CD9 (24). Cells were seeded in microwells on plastic dishes coated with parafilm (15) and cultivated for 7 to 15 days before the cytoadherence assay was performed. The RP5 strain was used when presenting with a majority of mature stages. Parasite culture at a 10% hematocrit was incubated with RP5-agglutinating sera and RP5-nonagglutinating sera (randomly chosen from the Yaoundé study) at a 1/10 dilution for 1 h with one shake at the midpoint. The parasite culture was then incubated with the trophoblast culture for 40 min before nonadherent IEs were removed by gentle washing. Preparations were glutaraldehyde fixed and observed microscopically. The number of IEs adhering to 1 mm² of trophoblast culture, divided by the parasite density used in the assay, was counted (15).

Statistical analyses. Proportions of agglutination were compared by contingency table analysis (chi-square test or Fisher's exact test, when appropriate). Differences of means of agglutinating indices between groups were analyzed by nonparametric methods (Mann-Whitney U test). The evolution of paired agglutination indices in the longitudinal study was analyzed with Wilcoxon's test for paired data. The significance limit (P) was 0.05, with a two-tailed risk. Statistics were computerized with StatView 4.5 (Abacus Concepts, Berkeley, Calif.).

	RESULTS AND DISCUSSION

Top Abstract Introduction Materials and Methods Results and Discussion References

Cross-sectional study in Yaoundé. A total of 664 women giving birth to live, single-born, term-delivered children were enrolled. Mothers' ages and parities and birth weights of infants are reported in Table 1. A mean of 23.5% of parturients had parasitized placentas, but primigravidae were more often infected, and susceptibility to malaria decreased with parity (Table 1).

Longitudinal study in Ebolowa. The development during pregnancy of specific antibody to PAPs was confirmed in a longitudinal study in Ebolowa (5), which demonstrated that pregnant women were more susceptible to malaria during first pregnancies than second pregnancies. During first pregnancies, P. falciparum-infected erythrocytes were present in 46.7% of women at the 6th month and in 65.2% at first delivery. Conversely, during second pregnancies, infection rates at the 6th month and at delivery dramatically dropped to 7.7% (1 of 13) and 0 of 5, respectively. The infection rates at 6 months of first and second pregnancies differed significantly (Fisher's exact test [P = 0.01]; Table 4). Compared to women from Yaoundé, pregnant women from Ebolowa were more often infected by P. falciparum and presented a stronger protective effect of parity, likely because Plasmodium transmission and exposure are higher (62 infective bites/year/person versus 15) (9, 10).

Mother to fetus passage of antibodies agglutinating RP5. As agglutination of IEs is mainly mediated by immunoglobulin G and parasite-reactive antibodies should cross the maternofetal barrier, we assessed the agglutinating capacity of 42 pairs of mother and cord blood serum samples against both RP5 and an isolate from a nonpregnant woman. Antibodies against RP5 were detected in 90.3% (28 of 31) of cord blood samples which paired maternal blood and agglutinated RP5 and in none of the 11 cord blood samples which paired maternal blood and did not agglutinate RP5. Among paired agglutinating sera, agglutination indices were correlated in mother and offspring (P < 0.01).

Inhibition of cytoadherence of RP5 to syncytiotrophoblast by RP5-agglutinating serum samples. Figure 1 demonstrates that RP5 binds to the trophoblast by using only CSA. Indeed, exogenous CSA almost totally inhibited this binding, while anti-ICAM-1 monoclonal antibody did not affect cytoadherence. The cytoadherence of RP5 to the trophoblast culture was inhibited by RP5-agglutinating sera and not by nonagglutinating sera (81% ± 4% versus 20% ± 3%; Mann-Whitney U test [P = 0.001]). Cytoadherence inhibition by agglutinating sera was only partial, as some agglutinates of RP5-infected erythrocytes bound to the trophoblast, but the mean binding intensity was consistently lower in the presence of agglutinating sera than with nonagglutinating sera. Among RP5-agglutinating sera, the agglutination index and cytoadherence inhibition rate were not related.

View larger version (10K): [in this window] [in a new window]   FIG. 1.   Effect of sera on the cytoadhesion of RP5 to human syncytiotrophoblast. The ability of RP5 to bind to the human syncytiotrophoblast was tested after parasites were supplemented with CSA (1 mg/ml), after the trophoblast was incubated with 84H10 (a monoclonal antibody against ICAM-1), or after parasites were incubated with RP5-agglutinating sera and RP5-nonagglutinating sera (for details see Materials and Methods). Each point represents one measure.