Recombinant human thrombomodulin(csa+): a tool for analyzing Plasmodium falciparum adhesion to chondroitin-4-sulfate.

The proteoglycan thrombomodulin has been shown to be involved, via its chondroitin-sulfate moiety, in the cytoadhesion of chondroitin-4-sulfate-binding-Plasmodium falciparum-infected erythrocytes to endothelial cells and syncytiotrophoblasts. We cloned and expressed in CHO and COS-7 cells a gene encoding soluble human recombinant thrombomodulin, with a chondroitin-4-sulfate moiety. This system is complementary to the in vitro cell models currently used to study the chondroitin-4-sulfate-binding phenotype. It also provides a means of overcoming the lack of specificity observed in interactions of infected erythrocytes with modified chondroitin-4-sulfate. This thrombomodulin displayed normal activity in coagulation, indicating that it was in a functional conformation. The recombinant protein, whether produced in CHO or COS-7 cells, inhibited cytoadhesion to Saimiri brain microvascular endothelial cells 1D infected with Palo-Alto(FUP)1 parasites selected for chondroitin-4-sulfate receptor preference. Thus, the recombinant protein was produced with a chondroitin-sulfate moiety, identified as a chondroitin-4-sulfate, in both cell types. In both cases, the recombinant protein bound to the chondroitin-4-sulfate phenotype, but not to CD36- and ICAM-1-binding parasites. The chondroitin-4-sulfate was 36 kDa in size for CHO and 17.5 kDa for COS-7 cells. There was, however, no difference in the capacities of the recombinant proteins produced by the two cell types to inhibit the cytoadhesion of infected erythrocytes. Thrombomodulin immobilized on plastic or coupled to Dynabeads was used to purify specifically the infected erythrocytes that bind to chondroitin-4-sulfate. These infected erythrocytes were cultured to establish parasite lines of this phenotype. We then showed that the thrombomodulin, labeled with FITC, could be used to detect this phenotype in blood samples. Finally, the direct binding of infected erythrocytes to immobilized thrombomodulin was used to screen for anti-chondroitin-4-sulfate-binding antibodies.

Characterisation of the chondroitin sulphate of Saimiri brain microvascular endothelial cells involved in Plasmodium falciparum cytoadhesion.

Cytoadhesion of Plasmodium falciparum-infected erythrocytes (IRBC) to chondroitin-4-sulphate (CSA) is inhibited by soluble CSA in vitro on Saimiri brain microvascular endothelial cells (SBEC) and in vivo in P. falciparum-infected Saimiri monkeys. We tested whether the SBEC model was appropriate for studying CSA-binding IRBC using four cell lines. All SBEC expressed a chondroitin sulphate (CS), with a composition of CSA. The mean sizes of these CSA were 20.5, 22, 23, 32.5 and 36 kDa for SBEC 3A and C2, CHO, SBEC 1D and 17, respectively. We found that cytoadhesion of the Palo-Alto (FUP)1 CSA-binding phenotype, selected by panning on SBEC 17, was specifically inhibited in a dose-dependent manner by all the purified CSA. The extent of inhibition depended on the cellular origin of the tested CSA. SBEC 17 CSA was 33 times more efficient than CHO-CSA and 21 times more efficient than the 50 kDa commercial bovine trachaea CSA. Dynabeads coated with a total extract of SBEC 1D CS-proteoglycans interacted with CSA- but not with CD36- or ICAM-1-binding IRBC. These Dynabeads also interacted specifically with the PfEMP1 DBL-3 domain, on the surface of CHO transfectants, but not with the CIDR-1 domain. Thrombomodulin was involved in IRBC adhesion to all SBEC whereas CD44 was only expressed by SBEC 1D and 17. These two CSA-proteoglycans have also been detected at the surface of human endothelial cells. Thus, the two homologous models, SBEC/Saimiri sciureus, are useful and reliable tools for the evaluation of new anti-CSA adhesion treatments and anti-disease vaccines for pregnant women.