Plasmodium vivax Adherence to Placental Glycosaminoglycans

Background Plasmodium vivax infections seldom kill directly but do cause indirect mortality by reducing birth weight and causing abortion. Cytoadherence and sequestration in the microvasculature are central to the pathogenesis of severe Plasmodium falciparum malaria, but the contribution of cytoadherence to pathology in other human malarias is less clear. Methodology The adherence properties of P. vivax infected red blood cells (PvIRBC) were evaluated under static and flow conditions. Principal Findings P. vivax isolates from 33 patients were studied. None adhered to immobilized CD36, ICAM-1, or thrombospondin, putative ligands for P. falciparum vascular cytoadherence, or umbilical vein endothelial cells, but all adhered to immobilized chondroitin sulphate A (CSA) and hyaluronic acid (HA), the receptors for adhesion of P. falciparum in the placenta. PvIRBC also adhered to fresh placental cells (N = 5). Pre-incubation with chondroitinase prevented PvIRBC adherence to CSA, and reduced binding to HA, whereas preincubation with hyaluronidase prevented adherence to HA, but did not reduce binding to CSA significantly. Pre-incubation of PvIRBC with soluble CSA and HA reduced binding to the immobilized receptors and prevented placental binding. PvIRBC adhesion was prevented by pre-incubation with trypsin, inhibited by heparin, and reduced by EGTA. Under laminar flow conditions the mean (SD) shear stress reducing maximum attachment by 50% was 0.06 (0.02) Pa but, having adhered, the PvIRBC could then resist detachment by stresses up to 5 Pa. At 37°C adherence began approximately 16 hours after red cell invasion with maximal adherence at 30 hours. At 39°C adherence began earlier and peaked at 24 hours. Significance Adherence of P. vivax-infected erythrocytes to glycosaminoglycans may contribute to the pathogenesis of vivax malaria and lead to intrauterine growth retardation.


Introduction
Vital organ dysfunction and death from P falciparum malaria results from microvascular obstruction caused by the sequestration of red cells containing mature forms of the parasite. P.falciparum infections in pregnancy are associated with a consistent reduction in birth weight, particularly in primigravidae. This is a major risk factor for neonatal death. Intrauterine growth retardation is associated with the accumulation of P.falciparum infected red cells in the placenta. This accumulation results from the adherence of a specific and relatively conserved red cell surface expressed domain of the antigenically variant membrane protein (PfEMP1) to the placental glycosaminoglycan chondroitin sulphate A, and to secondary receptors such as hyaluronic acid and immunoglobulins [1][2][3][4][5]. Plasmodium vivax is generally regarded as a more benign parasite than P.falciparum, and is associated with a lower mortality, although severe forms may occur occasionally [6]. P.vivax accounts for approximately half of all malaria outside Africa [7]. Like P.falciparum, P.vivax has also exerted a considerable selective pressure on human evolution although pathological processes are less well understood. Plasmodium vivax causes rosetting (adherence to uninfected erythrocytes) [8] but until recently it has not been considered to cytoadhere [9]. P. vivax infections in pregnancy also cause abortions [10] and reduce birthweight which increases the risk of neonatal death, although the mechanism underlying early fetal loss or the intrauterine growth retardation is unclear [11]. We have investigated the adherence of P. vivax (N = 33) to the placental glycosaminoglycans chondroitin sulphate A (CSA) and hyaluronic acid (HA), the putative receptors for P. falciparum placental adherence [10,11].

Ethics statement
This study was a part of clinical studies which have been approved by the Ethics Committee, Faculty of Tropical Medicine, Mahidol University. All participants gave fully informed consent to providing a 5 mL blood sample. Written informed consent was provided by study participants.

Parasites
Synchronous fresh isolates (.80% ring stage) of Plasmodium vivax were obtained from non-pregnant adult patients with acute vivax malaria admitted to the Hospital for Tropical diseases, Bangkok and entered into clinical studies. Malaria parasite species were confirmed by PCR [12]. All blood samples were recorded using a code identifier, and the subsequent experiments were conducted and the results were interpreted blinded to the patient data. Blood samples were taken into heparinized tubes and cultured as described previously [13]. Briefly, blood samples were centrifuged at 500 g at 4uC then plasma was discarded. White blood cells were removed by a CF-11 column or PlasmodipurH filter. After 24 hours of cultivation, trophozoite-infected red cells were used for further experiments. The parasite density of clinical isolates with less than 0.5% parasitaemia was augmented by concentration using a 66% PercollH gradient [14] or a magnetic separation column [15]. The synchronous trophozoite-IRBCs were enriched to 80-90% parasitaemia and the concentrate then adjusted to 1% parasitaemia at 1% haematocrit in PV-MCM media for the static adherence assay [16,17], and to 2% Haematocrit in 1% albumax for the laminar sheer flow adherence assay. Highly synchronized ring stage parasites (.0.5% parasitaemia) were used for assessing the relationship between adherence and stage of parasite development. The P. falciparum A4 clone, selected for adhesion to CSA (kindly provided by Dr David Roberts) was used as the control. Rosette formation (the adherence of two or more uninfected red cells to the infected cells) was counted for 100 IRBCs as described previously [18].

Static adherence assay
Adherence of parasitized erythrocytes to umbilical vein endothelial cells was assessed as described previously [19]. The reagents evaluated as potential receptors in the adherence assay were: purified CD36 and ICAM-1 (kindly provided by Arnab Pain), Thrombospondin (TSP; provided by Rachanee Udomsangpetch), CSA (from bovine trachea Sigma@ cat no C8529, or CSA covalently linked to phosphatidylethanolamine, kindly provided by Stephen Rogerson), CSC (chondroitin sulphate C), de-6-Osulphated CSA [20], dextran sulphate with molecular weight of 500 kD (Sigma), and HA (from bovine vitreous humor, Sigma @ cat no H7630). Assessment of adherence to these receptors was performed as described previously [20]. Briefly, receptors (at 100 ug/mL) were coated on plastic Petri dishes for 24 hours at 4uC, and then blocked by 1% bovine serum albumin in phosphate buffer saline (PBS) before being used. The IRBC suspension was incubated with the immobilized receptor spots for 30 minutes, at 37uC. After that, unbound red cells were removed by gently washing with RPMI-HEPES. The adherent cells were fixed with glutaraldehyde, and stained with Giemsa. The number of IRBCs bound was counted per 100 high power fields or 1 mm 2 .

Factors reducing P. vivax-infected red cell adherence
Soluble CSA (50 ug/mL) and HA (50 ug/mL) were incubated with the IRBC suspensions for 15 minutes at 37uC and the red cells then resuspended before testing adherence. In order to characterize further the adherence properties of PvIRBC, the effects of heparin (Leo H 1-1000 units/mL) and EGTA (0.01-1 mM) on adherence were assessed by coincubation for 2 hours at 37uC before the adherence assays were performed. In separate experiments, parasite cultures were coincubated with trypsin (1 to 1000 ug/mL; Sigma) for 15 minutes at 37uC, then washed and resuspended in PV-MCM culture medium containing 10% human serum and tested in the adherence assay as described above. P.vivax adherence to HA and CSA was evaluated further by incubating the Petri dishes with chondroitinase ABC (5 units/mL) and testicular hyarulonidase (10 unit/mL) or without enzyme (control) at 37uC for 45 minutes. The number of bound IRBCs was counted per 100 high power fields or 1 mm 2 . The median and ranges of adherence cell numbers were calculated.

Stage of P. vivax development and adherence to CSA and HA
The relationship between P. vivax IRBC adherence and stage of parasite development was investigated in ten isolates during the 48 hour asexual blood stage cycle in vitro. Parasite ring stages (6 hour old) were cultured as described previously [13], in 5% CO 2 at 37 uC and 39uC in parallel. Thin blood films were prepared every 6 hours up to 42 hours of parasite development and stained with Field's stain. In the static adherence assay the number of CSA and HA adherent IRBCs per 1000 red cells was counted every six hours and parasite developmental stage evaluated in 100 infected cells using staging criteria described previously [13].

P. vivax IRBC adherence to fresh placental sections
The Stamper-Woodruff adherence assay [21] was performed with some modification for the assessment of placental adhesion. Freshly frozen normal placenta sections (10 mm thick) were placed on glass slides and fixed with 3% glutaraldehyde in PBS for 30 minutes, then washed with RPMI-1640, kept wet and then placed on a tray. The concentrated P.vivax infected red cell (P.v IRBC) suspension at 80% parasitaemia and 1% haematocrit was overlaid on the slides and incubated for 30 minutes at 37uC in 5%CO 2 . After incubation, non-adherent cells were removed by rinsing with RMPI-1640. The remaining adherent cells were fixed with 3% glutaraldehyde in PBS and stained with Giemsa. Placenta-adherent P.v IRBCs binding in the intervillous space and along the syncytiotrophoblast layer were counted under light microscopy from 50 high power fields. Uninfected red cell suspensions (1% haematocrit) overlaid on the placenta sections served as controls in each assay. The specific adhesion of the infected red cells on placenta was further investigated by incubating IRBC suspensions with soluble CSA (50 ug/mL) and HA (50 ug/mL) for 15 minutes at 37uC before testing adherence.

Pv IRBC adherence under laminar flow conditions
The strength of P.vivax infected red cell adhesion was assessed under laminar flow conditions. An infusion/withdrawal pump (Harvard II apparatusH,) infused the cell suspension through microslides (placed on the stage of the inverted microscope) at defined flow rates and hence, defined wall shear stresses (calculated from the dimensions of the microslides and the flow rate, the temperature and the viscosity of the suspension.) as described previously [22]. The microslides were coated with 5 mg/mL poly-L-lysine for 30 minutes, then coated overnight with CSA [22,23]. The coated slides were blocked with 1% bovine serum albumin in PBS before performing the adherence assays. PBS was used as the control. Red cell suspensions (1% albumaxH in RPMI-1640) containing 2610 8 red cells/mL and P.vivax trophozoites (,30 hours old) at 0.5% parasitaemia flowed through the microslides at a shear stress of 0.05 Pa, for 5 minutes. Nonadherent cells were removed by perfusion with 1% albumax in RPMI-1640 for 5 minutes. After removing the non-adherent cells, the adherent cells were counted per 120 mm 2 surface area. Six P.vivax isolates were evaluated. CSA-adherent P. falciparum parasites derived from clone A4, were used as positive controls. The force of adhesion was then assessed under increasing shear stresses (0.01 to 8 Pa.). The already adherent cells (10 infected cells per isolate) were exposed to step-wise increases in shear stress until the cells were detached.

Statistical methods
Comparisons were performed using the Mann Whitney U test, proportions using the Chi-squared test and the correlation between the levels of adherence to different receptors was assessed using the Spearman Rank correlation coefficient. These were calculated using SPSS (version 11.5).

Results
In total 33 fresh P. vivax isolates were obtained from adults with acute vivax malaria. Admission parasitaemias ranged from 0.3-2.8%. P.falciparum coinfection was excluded by both microscopy and PCR genotyping in each case. The mean (SD) parasitaemia was 0.7 (0.3) %. Of 33 isolates 30 (93%) formed rosettes. The mean (SD) number of infected red cells (IRBCs) forming rosettes was 21 (17) per 100 infected red cells.

Binding to putative receptors
None of the P.vIRBCs adhered to immobilized CD36, ICAM-1, or thrombospondin, the putative vascular receptors for P.falciparum, whereas all 33 P.vivax isolates adhered to immobilized CSA and HA. The mean (SD) level of adhesion to CSA was 40 (25) and to HA was 33 (21) P.vivax infected red blood cells (P.v IRBCs) bound /mm 2 (Fig. 1). The levels of adherence to CSA and HA were correlated strongly (r s = 0.95; p,0.001). Bland -Altman analysis gave a mean (95% confidence interval) difference of 7.6 (4.65 to 10.54) P.v IRBCs bound /mm 2 from the average adhesion to the two receptors. None of the parasite isolates bound to human umbilical vein endothelial cells (data not shown), chondroitin sulphate-C, or dextran 500. There was no correlation between the number of P.v IRBC rosettes formed and adherence to CSA and HA. Adherence of P. vivax IRBCs to fresh placenta sections was assessed in five isolates, all of which showed attachment (Fig. 2). The mean (SD) number of bound P.v IRBCs per mm 2 was 52 (19).

Effects of parasite incubation temperature on adhesion properties
At an incubation temperature of 37uC adherence to CSA and HA began when the P.vivax parasites reached the large ring stage of development (approximately 12-16 hours after invasion). Adherence reached mean (SD) maximum values at 30 hours (n = 10); CSA 36 (10) and HA 41 (8) P.v IRBCs bound per mm 2 respectively. At 39uC adherence began earlier and reached peak values at 24 hours; CSA 45 (2) and HA 46 (3) per mm 2 respectively (Fig. 5A, B).

P.v IRBC adhesion under laminar flow conditions
In the in vitro laminar flow assay P.v IRBC adherence to CSA occurred only at low shear stresses (6 parasite isolates; 40 P.v IRBC each assessed). Maximum values were obtained at 0.01 Pa (Fig. 6). There was less adhesion at 0.1 Pa and none when shear stresses increased to 0.2 Pa. The mean (SD) shear stress reducing maximum attachment by 50% was 0.06 (0.02) Pa. However once they had adhered the P.v IRBCs were more resistant to mechanical detachment; most (34 out of 40) were detached after exposed to 1 Pa shear stresses (for 5 minutes), but 6 out of 40 P.v IRBCs resisted detachment by shear stresses of up to 5 Pa, for 5 minutes. Pre-incubation of Pv IRBCs with soluble CSA before perfusion (n = 3) totally inhibited parasite binding (p,0.01).

Discussion
Plasmodium vivax is generally considered to produce debilitating but non-fatal relapsing infections. Recent case series suggest this relatively benign reputation may need some revision [6]. Low birth weight, malnutrition, developmental retardation, severe anaemia, acute pulmonary oedema and encephalopathy are all well documented with vivax malaria [6,[24][25][26]. P.vivax has not previously been considered to sequester in the microcirculation through cytoadherence of infected erythrocytes, although recent physiological studies of pulmonary gas transfer in vivax malaria might be explained by pulmonary sequestration [27]. Consequently severe manifestations have been generally ascribed to other pathological processes, such as high systemic concentrations of pro-inflammatory cytokines. This study conducted in Thailand confirms that Plasmodium vivax infected erythrocytes do not cytoadhere to the principal vascular endothelial ligands responsible for P.falciparum sequestration in the microvasculature, or to  umbilical vein endothelial cells, but they do cytoadhere to both CSA and HA, the placental syncytiotrophoblast glycosaminoglycans responsible for P.falciparum sequestration in the placenta [1][2][3][4][5][28][29][30]. Similar findings have recently been reported from South America [9]. P.v IRBC also adhered to fresh placental cells, although binding was not intense. Malaria reduces birthweight and this reduces infant survival [11,31]. In P.falciparum infections this is associated with placental sequestration [1,2]. The data presented here suggest that the adverse effects of vivax malaria on intrauterine development may have a similar pathogenesis, and raise the possibility that pathology in other organs (such as lung capillaries and venules in patients who develop pulmonary  All isolates bound both CSA and HA to some extent although the degree of adhesion to the two receptors was different with some isolates. The specificity of adherence to these glycosaminoglycan receptors is supported by the inhibition by pretreatment with hyaluronidase and chondroitinase and competitive blocking by the soluble purified glycosaminoglycans under static and flow conditions [28][29][30]. Soluble CSA blocked binding to placental sections supporting a primary role for this molecule in placental cytoadherence. Commercial preparations of HA have been contaminated by CSA so the precise nature of the interaction with these two glycosaminoglycans remains to be elucidated. P.vIRBC adhesion to CSA and HA was inhibited by heparin, but was relatively resistant to cation depletion (EGTA), suggesting charge might contribute to non-specific inhibitory effects. P.falciparum cytoadherence is prevented by trypsin. PvIRBC adherence to CSA and HA was also largely abolished by trypsin indicating a red cell surface protein -receptor interaction [1][2][3]6], but the identity of this protein remains to be identified. The different sensitivities to trypsin among the different isolates may indicate different levels of adhesive surface protein expression, different adhesins, or multiple domains of the same adhesive molecules, as observed for P. falciparum binding to both CD36 and ICAM-1 [32].
P.v IRBC cytoadherence had other properties which are similar to that of P.falciparum. The IRBCs began to cytoadhere at the end of the first third of the asexual cycle, at the large ring to trophozoite stage, and this was accelerated by incubation at a ''febrile'' temperature (39uC) [33]. Cytoadherence did not occur at high shear stresses, but once the IRBCs had adhered to CSA or HA they were relatively resistant to increases in shear stresses up to 5 Pa. P.vivax reduces birthweight but the causative mechanisms have been unclear. P.vivax forms rosettes more than P.falciparum, and this might also contribute to pathological processes in the placenta. There is very limited information on placental pathology in acute vivax malaria so the extent of sequestration in-vivo is not known [34,35]. Adhesion of Plasmodium vivax infected red cells occurred at the low shear stresses (0.01-0.06 Pa) similar to those encountered in the intervillous spaces of the placenta. Higher stresses prevail in the systemic circulation. Given the physical characteristics of the placental circulation with its large vascular spaces, mechanical obstruction seems unlikely to account for placental dysfunction. Inflammatory processes with secondary interference with nutrient transfer may be more likely. Whichever the mechanism the evidence from P. falciparum infections that glycosaminoglycans are an important contributor to placental cytoadherence, and that antibodies which block this adherence are associated with protection against the adverse effect of birth weight, strongly suggests a pathological role [36]. The Duffy binding ligand of P.falciparum PfEMP1 that mediates cytoadherence to CSA is being developed as a possible vaccine against pregnancy malaria. A similar strategy might prevent the adverse effects of vivax malaria in pregnancy.