Thrombospondin mediates adherence of CD36+ sickle reticulocytes to endothelial cells

Abstract

Abstract Initiation of vasocclusion in sickle disease pathophysiology may involve abnormal red blood cell (RBC) adhesivity to endothelium, a phenomenon influenced by both RBC and plasma factors. Using human umbilical vein endothelial cells and a gravity sedimentation adherence assay, we have examined thrombospondin (TSP) as a plasma factor in this adhesive event. The already-abnormal adherence of sickle RBCs in buffer/albumin is significantly augmented (P < .001) by the addition of TSP, with half-maximal effect at about 0.3 microgram/mL. This effect is abolished by antibodies to either TSP or glycoprotein (GP) IV (CD36), as well as peptides RGDS and CSVTCG. The even greater adherence (P < .005) of sickle RBCs in autologous platelet-rich plasma (without added TSP) is dramatically inhibited by alpha CD36 antibodies (OKM5 and alpha GPIV) and significantly diminished by alpha TSP, by peptides RGDS and CSVTCG, and by two antibodies to the vitronectin receptor (7E3 and LM609). Studies of density-separated subpopulations and of RBC adhesion to immobilized proteins, as well as analysis of sickle RBCs using fluorescence-activated cell sorting and single cell microfluorometry, show that TSP responsiveness is a feature of the immature sickle “stress” reticulocytes, which carry CD36 (and not GPIIbIIIa-like receptors) as the TSP-receptive moiety. The endothelial cell's participation in this phenomenon appears to be more complex, and the data are consistent with the notion that it involves TSP interaction with other plasma proteins and/or multiple receptor structures. Other potential adhesogenic proteins (plasma von Willebrand factor, vitronectin, fibrinogen, and fibronectin) neither exhibited an affinity for reticulocytes nor supported increased sickle RBC adherence when added to buffer/albumin in these assay systems. In aggregate, our results indicate that TSP may be the major promoter of RBC adhesivity in plasma, and they suggest that therapeutic benefit might derive from interference with sickle reticulocyte CD36, as achieved by antibodies and CSVTCG in these studies.