Modeling of Plasmodium falciparumInfected Erythrocyte Cytoadhesion in Microvascular Conditions: Chondroitin-4-Sulfate Binding, A Competitive Phenotype

¹Unité de Parasitologie Expérimentale, Faculté de Médecine, Université de la Méditerranée, and ²Unité de Parasitologie, Institut de Médicine Tropicale du Service de Santé des Armées, Marseille, France; ³Faculté de Médecine, de Pharmacie et d'Odonto-Stomatologie, Université du Mali, Bamako, Mali; ⁴Instituto de Apoio a Pesquisa em Patologias Tropical, Rondonia, Brazil

Received 22 July 2002; revised 17 September 2002; electronically published 30 December 2002.

    Financial support: European Union for Research and Technical Development (contract no. QLK2-CT2000-00109 and IC18-CT98-0362); Ministère de l'Education Nationale, de la Recherche et de la Technologie (Program Paludisme + 2000); Délégation Générale pour l'Armement/Projet d'Etude Amont (no. 980814); Conselho Nacional de Desenvolvimento Cientifico e Tecnologico foundation fellowship from Brazil (to P.A.N.).
     Reprints or correspondence: Dr. Jürg Gysin, Unité de Parasitologie Expérimentale, EA 3282, Faculté de Médecine, Université de la Méditerranée, 27 Boulevard Jean Moulin, 13385 Marseille Cedex 5, France .

     The cytoadhesion of Plasmodium falciparuminfected erythrocytes (IEs) to the endothelial cells of microvessels is considered to be involved in the pathogenesis of malaria [1, 2]. These adhesion capacities result in the sequestration of late-stage IEs in the microvasculature of various organs, which prevents the clearance of IEs by the spleen. Numerous endothelial receptors for IE cytoadhesion have been identified [24], and their role in the pathogenesis of severe malaria has been investigated. The chondroitin-4-sulfate (CSA)binding phenotype has been implicated in malaria during pregnancy [5, 6], but no other clear correlation has been established between IE adhesion phenotype in vitro and disease syndrome in patients.

     In areas where P. falciparum is endemic, immune protection is acquired by age 10 years, which thereby limits malaria-related morbidity and mortality in young children. This protection is based mostly on the development of antibodies directed against variant IE surface antigens [7]. However, the prevalence and severity of malaria increase during the first pregnancies of previously protected women [8, 9]. This heightened susceptibility is associated with the accumulation of IEs in the intervillous spaces of the placenta [10], although parasites may be absent from peripheral blood [11, 12]. The much higher prevalence of the CSA cytoadhesion phenotype among the IEs sequestered in the placenta [5, 6] than in the peripheral blood [1316] suggests that the CSA cytoadhesion phenotype plays a major role for causing malaria during pregnancy. The prevalence of this phenotype is low in nonpregnant patients, and the serum of multigravidae women contains antibodies recognizing IE^CSA and inhibiting their adhesion, whereas such antibodies are detected at only very low levels in the serum of children and men or women before the third trimester of their first pregnancy [17, 18]. These observations led to the hypothesis that CSA-binding infected erythrocytes (IE^CSA) could develop only in pregnant women [19].

     However, this hypothesis is inconsistent with various observations. CSA is expressed at the surface of endothelial cells in the microvasculature of target organs for sequestration, such as the brain and the lung [2023]. All IE^CSA from placenta or peripheral blood isolates cytoadhere to endothelial cells [6, 2426]. The development of this phenotype is not restricted to pregnant women; IE^CSA have been detected in numerous field isolates collected from children, men, and nonpregnant women [1416, 26]. IE^CSA cytoadhere comparably with that of other phenotypes in flow conditions and better resist increasing wall shear stresses [2730]. Finally, in male and nonpregnant female Saimiri sciureus infected with the FCR3 strain, specific sequestration of IE^CSA was observed, despite the absence of a placenta [31].

     Therefore, why should IE^CSA sequestration be limited to the placenta? We investigated this question by modeling IE cytoadhesion in the microvasculature of the brain and lung. The polyphenotypic aspect of field isolates was simulated by mixing IE^CSA with IE^CD36, which are abundant in field isolates, and IE^ICAM-1, which are thought to be involved in cerebral malaria [32]. The possible interactions, synergy, and competition among these 3 phenotypes were investigated at physiological pH, by static, and flow-based cytoadhesion assays to endothelial cells from target organs for sequestration, expressing the 3 receptors studied. Under such conditions, the CSA-binding phenotype bound 10 times more efficiently than did the CD36 and intercellular adhesion molecule 1 (ICAM-1) phenotypes. This higher level of binding was correlated with a higher resistance to wall shear stress in interactions between the IE^CSA and the endothelial cells. All the characteristics of CSA-mediated cytoadhesion were consistent with a possible role for this phenotype in the initiation of obstruction in the microvessels of target organs for sequestration. Finally, comparison of 2 P. falciparum strains, FCR3 and IPL/BRE1, revealed differences in cytoadhesion capacity, which possibly corresponded to variants of the 3 phenotypes studied.

MATERIAL AND METHODS

     Parasites and cells.     Parasites of the FCR3 [33, 34] and IPL/BRE1 (Bre) [24] strains, as well as the clone D6 (selected from the FCR3^CSA subpopulation by micromanipulation), were studied. IEs were cultured in candle jars, as described elsewhere [35]. The following cell types, which were cultured, as described elsewhere [36], without being immortalized, were used for cytoadhesion assays without cytokine stimulation: Saimiri brain microvascular endothelial cell (SBEC) 1D, C2, 3A, and 17 [36] and human lung endothelial cells (HLECs) adapted to culture from primary explants [25]. Their constitutive expression of CSA, CD36, and ICAM-1 already has been described elsewhere [24, 25, 36].

     Selection by panning.     Subpopulations of the FCR3 and Bre strains were selected by 3 successive pannings of mature-stage IEs on cellular CSA, CD36, or ICAM-1, as described elsewhere [36, 37]. The phenotype of each subpopulation was checked by cytoadhesion inhibition assays on SBEC 1D (expressing CSA, CD36, and ICAM-1), as described elsewhere [37]. When the level of inhibition of a monophenotypic subpopulation by the corresponding specific inhibitor decreased to <90%, we panned the subpopulation once more or thawed a cryostabilate made immediately after the last panning of the subpopulation.

     Parasite labeling.     To enable us to differentiate among the phenotypes of IEs in mixtures that were allowed to cytoadhere simultaneously to SBEC 1D or HLECs, we labeled the parasites with fluorescent probes. Aliquots of 200 L of mature-stage IE pellets were washed with RPMI 1640 at pH 6.8 (cytoadhesion medium) and were incubated for 1 h at 37°C with 300 L of cytoadhesion medium containing the following markers at the indicated final concentrations: 12 g/mL 4,6-diamidino-2-phenylindole (DAPI), 12 mg/mL Texas Reddextran 10,000 (Molecular Probes), and 8 mg/mL fluorescein isothiocyanatedextran 10,000 (Sigma-Aldrich). IEs then were thoroughly washed, and the suspension was enriched by gelatin flotation. The IEs were finally resuspended in cytoadhesion medium at pH 6.8 or pH 7.2, at a hematocrit level of 5 × 10⁶ IEs/mL, for cytoadhesion assays in static or flow conditions.

     Viability test.     We checked that the fluorescent labels were not toxic by assessing the viability of labeled IEs. The labeled IEs were dispensed into 96-well plates (Becton Dickinson; 612 wells for each marker). Parasite growth was assessed after 2 invasions by adding 1 Ci of ³H-hypoxanthine with a specific activity of 14.1 Ci/mmol (NEN Products) to each well 72 h after labeling. The plates were incubated with the ³H-hypoxanthine for 24 h and then were frozen and thawed to lyse the IEs. The contents of the well were collected on standard filter microplates (Unifilter GF/B; Packard) and were washed with a cell harvester (FilterMate Cell Harvester; Packard). Filter microplates were dried, and 25 L of scintillation cocktail (Microscint O; Packard) was added to each well. The incorporation of radioactivity into the IEs was measured by means of a scintillation counter (Top Count; Packard). Parasite growth was assessed by calculating the geometric mean counts per minute of the 612 wells. The growth of unlabeled IEs also was assessed and defined as 100% viability.

     Static cytoadhesion assays.     The subpopulations of fluorescently labeled IEs were mixed in pairs, with each subpopulation comprising 50% of the mixture, or all 3 subpopulations were mixed together, with each subpopulation accounting for 33% of the mixture. Cytoadhesion microassays then were performed on 12-well immunofluorescence assay slides (Bio-Rad), as described elsewhere [37]. The slides were observed () by exhaustive photon reassignment microscopy [38] (exhaustive photon reassignment; CELLscan; Scanalytics). The assays were performed in triplicate.

fig.ommitted

Figure 1.        Identification of the phenotypes of the cytoadherent infected erythrocytes (IEs). Each IE phenotypic subpopulation was fluorescently labeled with 4,6-diamidino-2-phenylindole (DAPI), fluorescein isothiocyanate (FITC)dextran, or Texas Reddextran 10,000, as described in . Mixtures of the 3 phenotypes were allowed to cytoadhere to human lung endothelial cells, and the cytoadherent IEs were observed by exhaustive photon reassignment microscopy. Two examples of transmitted images are presented in panels A and A , with the corresponding fluorescent images obtained with DAPI (B and B ), FITC (C and C ), and Texas Red (D and D ) filter sets. By counting the IEs labeled with each marker, we determined the proportions of the various phenotypes among the cytoadherent IEs. Scale bar, 10 m.

     Flow-based cytoadhesion assays.     HLECs were cultured in microslides (VD/3530-050; Camlab), as described elsewhere [39]. In brief, 45 L of a 1 × 10⁶ HLECs/mL suspension was introduced into the microslides by capilarity, and the cells were allowed to settle for 2 h at 37°C. The microslides were immersed in DME-F12 supplemented with 10% vol/vol of fetal calf serum, 15 g/mL of endothelial cell growth supplement, and 10 g/mL of gentamicin and were connected to a peristaltic pump, which changed the medium in the microslide every 15 min. After 48 h, the cells formed a confluent monolayer, and cytoadhesion assays in flow conditions were performed, as described elsewhere [4042]. Microslides were connected to a precise infusion/withdrawal syringe pump (model 210P; KD Scientific) to control the flow of IE suspension or cell-free medium through the microslide. The other extremity of the microslide was connected to reservoirs containing the IE suspension or cytoadhesion medium (RPMI 1640), through an electronic 3-way valve. The flow rate (Q) required to give the desired wall shear stress () was calculated from the dimensions of the microslide and the viscosity of the medium as follows: Q = ²/6, where  is the medium viscosity, and w and h are the internal width and height, respectively, of the microslide. Cytoadhesion was observed by means of an inverted epifluorescence microscope (Eclipse TE200; Nikon), using a Plan Fluor ELWD 40×/0.60 objective (Nikon).

     IE suspensions at the desired pH were allowed to flow over HLECs for 10 min at 0.04 Pa. Cytoadhesion medium (RPMI 1640) at the desired pH then was allowed to flow through the microslide for 10 min at 0.04 Pa to remove nonadherent IEs. The level of cytoadhesion then was determined by counting the number of adherent IEs on 5 randomly distributed fields (0.081 mm²). The proportion of each subpopulation within the total population of adherent IEs was determined by counting the fluorescent IEs for each excitation wavelength (200 IEs for each experiment).

     Adhesion of IEs from flow.     We determined the ability of each subpopulation to cytoadhere to HLECs at different wall shear stresses. The IE suspensions were allowed to flow over the HLECs for 10 min at 0.04 Pa, 5 min at 0.08 Pa, 2.5 min at 0.16 Pa, 75 s at 0.32 Pa, 38 s at 0.64 Pa, and 19 s at 1.28 Pa. The cells were washed for the same period of time with cytoadhesion medium, and the cytoadherant IEs were counted on 5 fields, as described above. This procedure allowed the same number of IEs to flow over the HLECs at each wall shear stress.

     Detachment of cytoadherent IEs by increasing flow.     To determine the resistance of each phenotype to wall shear stress, cytoadherent IEs were subjected to successive washes (10 min) at increasing flow rates corresponding to stresses from 0.08 to 2.56 Pa. The number of IEs that continued to adhere to the HLECs and the proportions of each phenotype among the remaining adherent IEs were determined for each level of stress. We then calculated the stress required to detach 50% of the adherent IEs (SD₅₀). All experiments were performed at a mean room temperature of 24°C.

     Statistical analysis.     The results of the cytoadhesion assays are expressed as mean ± SD of at least 3 different experiments conducted on different days, with a different fluorescent marker used for each subpopulation every time. The Mann-Whitney U test was used to evaluate the statistical significance of the data obtained from cytoadhesion assays.

RESULTS

     Viability tests and toxicity of the fluorescent labeling.     The possible toxicity of the fluorescent labels used for IEs were investigated by ³H-hypoxanthine incorporation. Only DAPI labeling (12 g/mL) was found to be lethal (). This was expected, and we investigated whether the interaction between DAPI and the parasite DNA was likely to affect the cytoadhesion characteristics of the IEs during the 2 h after labeling. Three aliquots of the CSA-binding subpopulation of the FCR3 strain (FCR3^CSA) were labeled, each aliquot being labeled with a different probe, and a fourth aliquot was incubated in cytoadhesion medium without a marker to serve as a control. We assessed the adhesion of these IE populations, enriched by gelatin flotation, to SBEC 1D, in static conditions and found no statistically significant difference between the 4 aliquots ().

fig.ommitted

Table 1.          Cytotoxicity of the fluorescent labelings.

     Cytoadhesion of each IE subpopulation assayed independently.     These controls were performed to verify the ability of both endothelial cell types to support the cytoadhesion of all the IE phenotypic subpopulations independently, in all the experimental conditions used. Since IE^CD36 and IE^ICAM-1 are not monophenotypic subpopulations, each being able to interact with CD36, ICAM-1, and possibly other nonidentified receptors [37], these independent cytoadhesion trials provided a better idea of the ability of the endothelial cells to allow for the cytoadhesion of each IE subpopulation than a quantitative determination of the expression of the receptors at the endothelial cell surface. For each experimental condition, we determined the ratio of the IE^CSA adhesion level divided by the IE^CD36 or IE^ICAM-1 adhesion level (). Twenty-one of the 24 ratios obtained ranged from 1 to 3.

fig.ommitted

Table 2.          Cytoadhesion level of each infected erythrocyte (IE) subpopulation assayed independently.

     Static competition assays.     We performed static cytoadhesion assays on SBEC 1D and HLECs, which express CSA, ICAM-1, and CD36. The final hematocrit level of all the parasite suspensions was 5 × 10⁶ IEs/mL, corresponding to half a monolayer, which thereby prevents effects on cytoadhesion interactions due to steric hindrance. For both strains, regardless of whether 2 or 3 phenotypes were competing together, the proportion of each phenotype in the cytoadherent IE population differed considerably from that in the initial mixture (). The assays of competition between Bre subpopulations on SBEC 1D (data not shown) provided similar results to those obtained with HLECs used as target cells ( and ). In competition assays that involved Bre subpopulations at both pH levels, as well as FCR3 subpopulations at pH 7.2, the CSA phenotype was the most competitive, resulting in an increase in the proportion of IE^CSA from 33% or 50% in the initial mixture (3 or 2 phenotypes competing, respectively) to 74%96% in the cytoadherent populations (.0001  P < .0439). Assays of competition between the CSA- and ICAM-1binding phenotypes (data not shown) gave results similar to those obtained from assays of competition between the CSA- and the CD36-binding phenotypes. At pH 6.8, competition between the FCR3 subpopulations resulted in a weak but significant predominance of the CSA phenotype (), except for FCR3^CSA versus FCR3^CD36 on HLECs (P = .6746; ).

fig.ommitted

Figure 2.        Static cytoadhesion competition assays. FCR3 (A, B, D, and E) and IPL/BRE1 (C and F) phenotypic subpopulations selected on the basis of their cytoadhesion to cellular chondroitin-4-sulfate (CSA), CD36, or intercellular adhesion molecule 1 (ICAM-1) were labeled and mixed 2 by 2 (50/50; A, B and C) or all 3 were mixed together (33/33/33; D, E, and F) and were allowed to cytoadhere under static conditions to Saimiri brain microvascular endothelial cell (SBEC) 1D (A and D) or human lung endothelial cells (HLECs; B, C, E, and F), at pH 6.8 or 7.2. Cells were washed, and proportion of each phenotype within cytoadherent infected erythrocytes (IEs) was determined by fluorescence. Data are mean ± SD of experiments carried out at least in triplicate.

     Cytoadhesion assays in flow conditions.     Because of the difficulty of growing SBEC 1D in microslides, cytoadhesion assays in flow conditions were carried out only with HLECs as the target cells. We investigated the ability of the various subpopulations to cytoadhere to HLECs, at physiological pH, under increasing wall shear stresses (). With the exception of Bre^ICAM-1, all the subpopulations behaved identically, with a sharp decrease in cytoadhesion with increasing wall shear stress. The decrease was slower for Bre^ICAM-1, but these cells had initial cytoadhesion levels only about one-fourth of those of the other subpopulations. At wall shear stresses >0.4 Pa, we observed no significant cytoadhesion.

fig.ommitted

Figure 3.        Effect of wall shear stress on the cytoadhesion of the FCR3 and IPL/BRE1 phenotypic subpopulations to human lung endothelial cells (HLECs). Chondroitin-4-sulfate (CSA), CD36- or intercellular adhesion molecule 1 (ICAM-1)binding subpopulations of the FCR3 (A) or IPL/BRE1 (B) strains were allowed to cytoadhere at pH 7.2 onto HLECs under the following conditions: for 10 min at 0.04 Pa, 5 min at 0.08 Pa, 2.5 min at 0.16 Pa, 75 s at 0.32 Pa, or 37 s at 0.64 Pa. Cells then were washed with cytoadhesion medium (RPMI 1640), at the corresponding wall shear stress and for the same time. No. of cytoadherent infected erythrocytes (IEs) per mm² was then determined. Data are mean ± SD of 5 measures.

     We then investigated the resistance to wall shear stress of the interactions between the various IE subpopulations and HLECs, at both pH levels, by testing various flow constraints of the medium flushed through the microcapillary (). At pH 6.8, the FCR3^CSA and Bre^CSA that adhered at 0.04 Pa resisted wall shear stresses >0.64 and >1.28 Pa, respectively. The SD₅₀ of FCR3^CSA was 2.11 Pa, whereas, for Bre^CSA, only 44% of the initially adherent IEs were detached by the highest wall shear stress used. Significant numbers of adherent FCR3^CD36 were detached at 0.08 Pa (FCR3^CD36 SD₅₀, 0.43 Pa), whereas adherent Bre^CD36 resisted wall shear stresses >0.64 Pa, with no more than 30% of the initially adherent IEs detached. Finally, FCR3^ICAM-1 were continually detached, resulting in a total loss of 95% of the initially cytoadherent IEs (SD₅₀, 0.25 Pa), whereas cytoadherent Bre^ICAM-1 completely resisted wall shear stresses of up to 0.64 Pa (SD₅₀, 1.68 Pa).

fig.ommitted

Figure 4.        Effect of increasing wall shear stresses on the cytoadherent FCR3 and IPL/BRE1 phenotypic subpopulations. Chondroitin-4-sulfate (CSA), CD36- or intercellular adhesion molecule 1 (ICAM-1)binding subpopulations of the FCR3 (A and C) or IPL/BRE1 (B and D) strains were allowed to cytoadhere at pH 6.8 (A and B) or pH 7.2 (C and D) onto human lung endothelial cells for 10 min, at a wall shear stress of 0.04 Pa. Cells then were washed with cytoadhesion medium (RPMI 1640) at the corresponding pH and were allowed to flow over the cells at 0.04 Pa for 10 min. No. of cytoadherent infected erythrocytes (IEs) per mm² then was determined. Cells then were washed with medium flushed through the slide at increasing wall shear stress (successively 0.08, 0.16, 0.32, 0.64, 1.28, and 2.56 Pa), with each wall shear stress maintained for 10 min, and no. of IEs that remained cytoadherent was determined. Data are mean ± SD of 5 measurements.

     All interactions were weaker at pH 7.2 than at pH 6.8. For the FCR3 strain, almost all the adherent IEs resistant to wall shear stresses higher than those found in the placenta were of the CSA phenotype (SD₅₀ = 1.16 Pa), and the cytoadherent IEs of the CSA-binding clone D6 resisted all wall shear stresses tested (data not shown). In contrast, 76% of FCR3^ICAM-1 were detached at 0.08 Pa and <35% of the cytoadherent FCR3^CD36 resisted the lowest wall shear stress found in microvessels (SD₅₀ = 0.06 and 0.07, respectively). Bre^CSA behaved similarly at both pH values, with 34% of cytoadherent IEs detached at 2.56 Pa (pH 7.2). For Bre^ICAM-1 and Bre^CD36, significant decreases in cytoadhesion were observed at 0.08 and 0.16 Pa, respectively, with SD₅₀ at 0.13 and 0.47 Pa, respectively.

     Competition assays in flow conditions.     The competitive cytoadhesion assays performed under flow conditions provided similar results for both strains (), as well as for the CD36- and ICAM-1binding phenotypes (data not shown). At pH 6.8, the proportions of the phenotypes among cytoadherent IEs were similar to those in the initial mixture. At pH 7.2 and for both strains, the IE^CSA strongly dominated in competition with another phenotype (86%90%) or if all 3FCR3 subpopulations were allowed to compete with each other simultaneously (80%). When the 3 Bre subpopulations were allowed to compete, the CSA phenotype predominated among the cytoadherent IEs (56%; ), but considerable differences were observed between experiments (). However, the marked predominance of the CSA phenotype was restored as wall shear stress was increased ().

fig.ommitted

Figure 5.        Cytoadhesion competition in flow conditions at pH 6.8 and 7.2. FCR3 (A and C) and IPL/BRE1 (B and D) phenotypic subpopulations selected for cytoadhesion to cellular chondroitin-4-sulfate (CSA), CD36, or intercellular adhesion molecule 1 (ICAM-1) were labeled and mixed 2 by 2 (50/50; A and B) or all 3 together (33/33/33; C and D). Infected erythrocytes (IE) were allowed to cytoadhere under flow conditions to human lung endothelial cells for 10 min (0.04 Pa). Cells then were washed for 10 min (0.04 Pa), and proportion of cytoadherent IE corresponding to each phenotype was determined by fluorescence. Data are mean ± SD of experiments carried out at least in triplicate.

fig.ommitted

Figure 6.        Cytoadhesion competition of the IPL/BRE1 subpopulations in flow conditions and resistance to increasing wall shear stress at pH 7.2. A, Proportions of each phenotype within the cytoadherent infected erythrocytes (IEs) of the 4 experiments, which correspond to the mean result presented in , are presented individually. B, Phenotypic subpopulations were mixed together in equal proportions and were allowed to cytoadhere for 10 min on human lung endothelial cells under a wall shear stress of 0.04 Pa. Cells were washed for 10 min with cytoadhesion medium (RPMI 1640) at a wall shear stress of 0.04 Pa, and proportion of each phenotype within the cytoadherent IEs was determined by fluorescence. Cells then were washed with medium flushed through the slide to give increasing wall shear stress (successively 0.08, 0.16 and 0.32 Pa), with each wall shear stress maintained for 10 min, and proportion of each phenotype among the IEs that continued to be cytoadherent was determined. Data are from 1 representative experiment.

DISCUSSION

     The recently described cryptic cycle of IE^CSA [43] may account for the low prevalence of this phenotype in field isolates [5, 13, 15, 26, 44 ], as well as for its absence from the peripheral blood of a large number of pregnant women (34 of 7 patients reported by Fried and Duffy [5] and 68 of 18 patients reported by Beeson et al. [15]). However, the restriction of antiIE^CSA antibodies to pregnant women seems to be a strong argument for the limitation of this phenotype to pregnancy. Among the hypotheses that may explain this supposed limitation, we investigated the possible interactions among IEs of different phenotypes and between IEs and endothelial cells that could interfere with IE^CSA cytoadhesion.

     We developed an interactive model of cytoadhesion under physiological pH and flow conditions. The wall shear stresses that were used corresponded to placental and postcapillary venule conditions [45]. The endothelial cells expressed CSA (with 45% of 4-sulfated disaccharides [46], as required for optimal interaction [47]), CD36, and ICAM-1. The polyphenotypic composition of field isolates (e.g., [1416, 26, 48]) was mimicked by mixing IE^CSA, IE^CD36, and IE^ICAM-1. IE^CSA is a monophenotypic subpopulation and can be considered to be IEs interacting exclusively with CSA, whereas it must be understood that IE^CD36 and IE^ICAM-1 were selected for their ability to interact with CD36 or ICAM-1 and are able to use CD36, ICAM-1, and possibly other receptors to cytoadhere. However, IE^ICAM-1 interacted exclusively by rolling on endothelial cells, as described elsewhere [27].

     The patterns of adhesion obtained under increasing wall shear stresses were similar to those already described elsewhere [2729, 49]. In all cases, the ability to establish the interaction between IEs and endothelial cells was limited to wall shear stresses consistent with a sequestration in postcapillary venules, as described elsewhere [5052]. Once the cytoadhesion was established at 0.04 Pa, important variations of the resistance of the attachment to increasing wall shear stresses were observed among the 3 phenotypes, as described elsewhere [27, 28], and for each phenotype among the 2 strains. At physiological pH, only IE^CSA cytoadhere sufficiently strongly to resist the wall shear stresses observed in the microvasculature (0.11 Pa [53]). Thus, IE^CSA is the best adapted of the 3 phenotypes studied to sequester in the microvasculature of target organs. Given its cytoadhesion characteristics, the presence of the CSA phenotype may be essential to initiate the obstruction in small microvessels and to maintain this obstruction against the resulting increases in wall shear stress and pressure.

     We investigated the cytoadhesion behavior of the phenotypic subpopulations when mixed in equal proportions, to allow for synergistic or competitive interactions. We observed a striking predominance of the CSA phenotype among the IEs that cytoadhered at pH 7.2 (the ratios between the adhesion levels of IE^CSA and that of the other subpopulations ranged from 6 to 25). This finding was unexpected in view of the adhesion level of each subpopulation assayed individually (ratios ranging from 1 to 6). It may result from the ability of some phenotype(s) to regulate the cytoadhesion of other(s) or from differences in the competitiveness of the phenotypes for cytoadhesion. For instance, IE^CSA established a stable interaction with the endothelial cells faster than did the other phenotypes (data not shown), which allowed interference with the cytoadhesion of IE^CD36 and IE^ICAM-1 possibly by modification of the endothelial cell surface. Because of the differences in the cytoadhesion competitiveness observed among the 3 phenotypes studied, the importance of a phenotype in sequestration cannot be estimated in adhesion assays to purified receptors or cells expressing only 1 type of receptor. Our data also demonstrate the importance of the experimental conditions used. Performing prevalence assays at pH 6.8 might bias the results toward an underestimation of the role of highly competitive minor phenotypes and overestimation of the role of poorly competitive major phenotypes.

     In light of our data, it appears that the notion of cytoadhesion phenotype might not be homogeneous enough to observe correlation between the phenotypes detectable in the peripheral blood of patients and the severity of the clinical case. Because of the difference in their cytoadhesion resistance to wall shear stress, FCR3^CD36 and Bre^CD36 may not result in the same complications in vivo. Identical considerations apply when comparing the homogeneous strong-binder Bre^CSA subpopulation, or D6 clone, with the heterogeneous FCR3^CSA subpopulation. The pathogenicity of IE may rely less on the endothelial receptor with which they interact than on the strength of the interaction or on the length of the CSA chain with which they can interact [6, 37]. These variations also may account for organ tropism. A weak adhesion may limit the sequestration in organs were the wall shear stress low, like the placenta, whereas a strong interaction would allow for sequestration in the microvasculature.

     It is paradoxical that the most adapted phenotype for microvessel sequestration should be limited to placental cytoadhesion. Besides the observation that some serum samples from children and men inhibit CSA-mediated cytoadhesion (authors' unpublished data), the demonstration of the existence of variants of the CSA phenotype raises the possibility that the immune response may not be quantitatively equal in all patients and may depend on the variants with which they are infected. In addition, since IE^CSA cytoadhesion is optimal for the total clogging of microvessels, contacts between the antigen-presenting cells (APCs) of the immune system and IE^CSA in the deep microvasculature may be limited. In contrast, immune competent cells in placenta have direct access to IE^CSA, which facilitates the build-up of an antibody-associated immune response.

     If confirmed with field isolates, our results raise important questions concerning the role of the CSA phenotype in Plasmodium species biology. Although their is no doubt that they can cytoadhere in microvascular conditions, IE^CD36 and IE^ICAM-1 nevertheless seem to be inefficient at causing microvessel occlusion. In the absence of a more competitive phenotype, their sequestration may be limited to "large" microvessels (always accessible to APCs), like 20-m diameter postcapillary venules, where partial clogging would not result in a significant increase of the wall shear stress. For a more complete occlusion of the microvessels or for sequestration in small microvessels, the IEs of a more competitive phenotype may be necessary. With a cryptic cycle and cytoadhesion characteristics conferring a special immune status, IE^CSA may constitute a reservoir of parasites, possibly regulated by phenotype switching [6, 54] to prevent the permanent obstruction of microvessels. Further analysis of the various hypothesis discussed here will require a direct determination of the phenotypes of the IEs sequestered in organs such as the brain and lung, by flushing IEs from organ necropsy samples or by using specific antibodies able to recognize IE^CSA in organ cryosections [55].

References

1. 

Berendt AR, Turner GDH, Newbold CI. Cerebral malaria: the sequestration hypothesis. Parasitol Today 1994; 10:4124.

2. 

Wahlgren M, Treutiger CJ, Gysin J. Cytoadherence and rosetting in the pathogenesis of severe malaria. In: Wahlgren M, Perlmann P, eds. Malaria: molecular and clinical aspects. Amsterdam: Harwood Academic Publishers, 1999:289327.

3. 

Siano JP, Grady KK, Millet P, Wick TM. Plasmodium falciparum: cytoadherence to (v)3 on human microvascular endothelial cells. Am J Trop Med Hyg 1998; 59:779. 

4. 

Flick K, Scholander C, Chen Q, et al. Role of nonimmune IgG bound to PfEMP1 in placental malaria. Science 2001; 293:2098100.

5. 

Fried M, Duffy PE. Adherence of Plasmodium falciparum to chondroitin sulfate A in the human placenta. Science 1996; 272:15024.

6. 

Gysin J, Pouvelle B, Fievet N, Scherf A, Lepolard C. Ex vivo desequestration of Plasmodium falciparuminfected erythrocytes from human placenta by chondroitin sulfate A. Infect Immun 1999; 67:6596602.

7. 

Bull PC, Lowe BS, Kortok M, Molyneux CS, Newbold CI, Marsh K. Parasite antigens on the infected red cell surface are targets for naturally acquired immunity to malaria. Nat Med 1998; 4:35860.

8. 

Brabin BJ. An analysis of malaria in pregnancy in Africa. Bull World Health Organ 1983; 61:100516. 

9. 

McGregor IA. Epidemiology, malaria and pregnancy. Am J Trop Med Hyg 1984; 33:51725. 

10. 

Walter PR, Garin Y, Blot P. Placental pathologic changes in malaria: a histologic and ultrastructural study. Am J Pathol 1982; 109:33042.

11. 

Watkinson M, Rushton DI. Plasmodial pigmentation of placenta and outcome of pregnancy in West African mothers. Br Med J (Clin Res Ed) 1983; 287:2514. 

12. 

Matteelli A, Donato F, Shein A, et al. Malaria and anaemia in pregnant women in urban Zanzibar, Tanzania. Ann Trop Med Parasitol 1994; 88:47583.

13. 

Rogerson SJ, Chaiyaroj SC, Ng K, Reeder JC, Brown GV. Chondroitin sulfate A is a cell surface receptor for Plasmodium falciparuminfected erythrocytes. J Exp Med 1995; 182:1520.

14. 

Chaiyaroj SC, Angkasekwinai P, Buranakiti A, Looareesuwan S, Rogerson SJ, Brown GV. Cytoadherence characteristics of Plasmodium falciparum isolates from Thailand: evidence for chondroitin sulfate a as a cytoadherence receptor. Am J Trop Med Hyg 1996; 55:7680.

15. 

Beeson JG, Brown GV, Molyneux ME, Mhango C, Dzinjalamala F, Rogerson SJ. Plasmodium falciparum isolates from infected pregnant women and children are associated with distinct adhesive and antigenic properties. J Infect Dis 1999; 180:46472.

16. 

Rogerson SJ, Tembenu R, Dobano C, Plitt S, Taylor TE, Molyneux ME. Cytoadherence characteristics of Plasmodium falciparuminfected erythrocytes from Malawian children with severe and uncomplicated malaria. Am J Trop Med Hyg 1999; 61:46772.

17. 

Fried M, Nosten F, Brockman A, Brabin BJ, Duffy PE. Maternal antibodies block malaria. Nature 1998; 395:8512.

18. 

Ricke CH, Staalsoe T, Koram K, et al. Plasma antibodies from malaria-exposed pregnant women recognize variant surface antigens on Plasmodium falciparuminfected erythrocytes in a parity-dependent manner and block parasite adhesion to chondroitin sulfate A. J Immunol 2000; 165:330916. 

19. 

Beeson JG, Reeder JC, Rogerson SJ, Brown GV. Parasite adhesion and immune evasion in placental malaria. Trends Parasitol 2001; 17:3317.

20. 

Wong VL, Hofman FM, Ishii H, Fisher M. Regional distribution of thrombomodulin in human brain. Brain Res 1991; 556:15.

21. 

Boffa MC, Jackman RW, Peyri N, Boffa JF, George B. Thrombomodulin in the central nervous system. Nouv Rev Fr Hematol 1991; 33:4239.

22. 

Gysin J, Pouvelle B, Le Tonqueze M, Edelman L, Boffa MC. Chondroitin sulfate of thrombomodulin is an adhesion receptor for Plasmodium falciparuminfected erythrocytes. Mol Biochem Parasitol 1997; 88:26771.

23. 

Bajaj MS, Kuppuswamy MN, Manepalli AN, Bajaj SP. Transcriptional expression of tissue factor pathway inhibitor, thrombomodulin and von Willebrand factor in normal human tissues. Thromb Haemost 1999; 82:104752. 

24. 

Robert C, Pouvelle B, Meyer P, et al. Chondroitin-4-sulphate (proteoglycan), a receptor for Plasmodium falciparuminfected erythrocyte adherence on brain microvascular endothelial cells. Res Immunol 1995; 146:38393. 

25. 

Muanza K, Gay F, Behr C, Scherf A. Primary culture of human lung microvessel endothelial cells: a useful in vitro model for studying Plasmodium falciparuminfected erythrocyte cytoadherence. Res Immunol 1996; 147:14963. 

26. 

Traore B, Muanza K, Looareesuwan S, et al. Cytoadherence characteristics of Plasmodium falciparum isolates in Thailand using an in vitro human lung endothelial cells model. Am J Trop Med Hyg 2000; 62:3844

27. 

Cooke BM, Berendt AR, Craig AG, MacGregor J, Newbold CI, Nash GB. Rolling and stationary cytoadhesion of red blood cells parasitized by Plasmodium falciparum: separate roles for ICAM-1, CD36 and thrombospondin. Br J Haematol 1994; 87:16270.

28. 

Cooke BM, Rogerson SJ, Brown GV, Coppel RL. Adhesion of malaria-infected red blood cells to chondroitin sulfate A under flow conditions. Blood 1996; 88:40404.

29. 

Udomsangpetch R, Reinhardt PH, Schollaardt T, Elliott JF, Kubes P, Ho M. Promiscuity of clinical Plasmodium falciparum isolates for multiple adhesion molecules under flow conditions. J Immunol 1997; 158:435864.

30. 

Yipp BG, Anand S, Schollaardt T, Patel KD, Looareesuwan S, Ho M. Synergism of multiple adhesion molecules in mediating cytoadherence of Plasmodium falciparuminfected erythrocytes to microvascular endothelial cells under flow. Blood 2000; 96:22928.

31. 

Pouvelle B, Meyer P, Robert C, Bardel L, Gysin J. Chondroitin-4-sulfate impairs in vitro and in vivo cytoadherence of Plasmodium falciparuminfected erythrocytes. Mol Med 1997; 3:50818.

32. 

Newbold C, Craig A, Kyes S, Rowe A, Fernandez-Reyes D, Fagan T. Cytoadherence, pathogenesis and the infected red cell surface in Plasmodium falciparum. Int J Parasitol 1999; 29:92737.

33. 

Gysin J, Hommel M, da Silva LP. Experimental infection of the squirrel monkey (Saimiri sciureus) with Plasmodium falciparum. J Parasitol 1980; 66:10039.

34. 

Gysin J, Fandeur T. Saimiri sciureus (karyotype 14-7): an alternative experimental model of Plasmodium falciparum infection. Am J Trop Med Hyg 1983; 32:4617.

35. 

Trager W, Jensen JB. Human malaria parasites in continuous culture. Science 1976; 193:6735.

36. 

Gay F, Robert C, Pouvelle B, Peyrol S, Scherf A, Gysin J. Isolation and characterization of brain microvascular endothelial cells from Saimiri monkeys: an in vitro model for sequestration of Plasmodium falciparuminfected erythrocytes. J Immunol Methods 1995; 184:1528.

37. 

Pouvelle B, Fusai T, Lepolard C, Gysin J. Biological and biochemical characteristics of cytoadhesion of Plasmodium falciparuminfected erythrocytes to chondroitin-4-sulfate. Infect Immun 1998; 66:49506.

38. 

Carter KC, Bowman D, Carrington W, et al. A three-dimensional view of precursor messenger RNA metabolism within the mammalian nucleus. Science 1993; 259:13305.

39. 

Cooke BM, Coppel RL, Nash GB. Preparation of adhesive targets for flow-based cytoadhesion assays. Methods Mol Med 2002; 72:5719.

40. 

Cooke BM, Usami S, Perry I, Nash GB. A simplified method for culture of endothelial cells and analysis of adhesion of blood cells under conditions of flow. Microvasc Res 1993; 45:3345.

41. 

Cooke BM, Nicoll CL, Baruch DI, Coppel RL. A recombinant peptide based on PfEMP-1 blocks and reverses adhesion of malaria-infected red blood cells to CD36 under flow. Mol Microbiol 1998; 30:8390.

42. 

Cooke BM, Glenister FK, Mohandas N, Coppel RL. Assignment of functional roles to parasite proteins in malaria-infected red blood cells by competitive flow-based adhesion assay. Br J Haematol 2002; 117:20311.

43. 

Pouvelle B, Buffet PA, Lepolard C, Scherf A, Gysin J. Cytoadhesion of Plasmodium falciparum ring-stageinfected erythrocytes. Nat Med 2000; 6:12648.

44. 

Ho M, Hickey MJ, Murray AG, Andonegui G, Kubes P. Visualization of Plasmodium falciparumendothelium interactions in human microvasculature: mimicry of leukocyte recruitment. J Exp Med 2000; 192:120511.

45. 

Lawrence MB, Springer TA. Leukocytes roll on a selectin at physiologic flow rates: distinction from and prerequisite for adhesion through integrins. Cell 1991; 65:85973.

46. 

Fusai T, Parzy D, Spillmann D, et al. Characterisation of the chondroitin sulphate of Saimiri brain microvascular endothelial cells involved in Plasmodium falciparum cytoadhesion. Mol Biochem Parasitol 2000; 108:2537.

47. 

Alkhalil A, Achur RN, Valiyaveettil M, Ockenhouse CF, Gowda DC. Structural requirements for the adherence of Plasmodium falciparum infected erythrocytes to chondroitin sulfate proteoglycans of human placenta. J Biol Chem 2000; 275:4035764.

48. 

Newbold C, Warn P, Black G, et al. Receptor-specific adhesion and clinical disease in Plasmodium falciparum. Am J Trop Med Hyg 1997; 57:38998.

49. 

Wick TM, Louis V. Cytoadherence of Plasmodium falciparuminfected erythrocytes to human umbilical vein and human dermal microvascular endothelial cells under shear conditions. Am J Trop Med Hyg 1991; 45:57886.

50. 

Miller LH. Distribution of mature trophozoites and schizonts of Plasmodium falciparum in the organs of Aotus trivirgatus, the night monkey. Am J Trop Med Hyg 1969; 18:8605.

51. 

Luse SA, Miller LH. Plasmodium falciparum malaria: ultrastructure of parasitized erythrocytes in cardiac vessels. Am J Trop Med Hyg 1971; 20:65560.

52. 

MacPherson GG, Warrell MJ, White NJ, Looareesuwan S, Warrell DA. Human cerebral malaria: a quantitative ultrastructural analysis of parasitized erythrocyte sequestration. Am J Pathol 1985; 119:385401.

53. 

Chien S. Blood rheology and its relation to flow resistance and transcapillary exchange, with special reference to shock. Adv Microcirc 1969; 2:89103.

54. 

Roberts DJ, Craig AG, Berendt AR, et al. Rapid switching to multiple antigenic and adhesive phenotypes in malaria. Nature 1992; 357:68992.

55. 

lekana Douki JB, Traore B, Costa FT, et al. Sequestration of Plasmodium falciparuminfected erythrocytes to chondroitin sulfate A, a receptor for maternal malaria: monoclonal antibodies against the native parasite ligand reveal panreactive epitopes in placental isolates. Blood 2002; 100:147883.