Semaphorins are a large family of cell surface molecules best known for their ability to mediate communication between cells during neural development. We have recently shown that human platelets express the semaphorin family member, sema4D, and both of its known receptors, CD72 and plexin-B1 (Zhu, et al, PNAS, 2007). We have also shown that sema4D(−/−) mice have an impaired response to arterial injury in vivo and a selective defect in collagen- and convulxin-induced platelet aggregation in vitro. In the present studies we have sought the molecular basis for these defects, focusing on events downstream of glycoprotein VI (GPVI), which serves as a receptor for both collagen and convulxin. In normal platelets, GPVI signaling leads to the phosphorylation and activation of phospholipase Cγ2 (PLCγ2) through the formation of a signaling complex that includes SLP-76 and LAT. This complex is activated when GPVI-associated FcRγ is phosphorylated, allowing the tyrosine kinase, Syk, to bind. PLCγ2 activation results in phosphoinositide hydrolysis, an IP3-mediated increase in cytosolic Ca++, and activation of additional kinases, such as Akt. In theory, the absence of sema4D could affect any of these steps and by doing so impair collagen-induced platelet aggregation. Working backwards through the GPVI pathway, our results showed that compared to platelets from matched WT mice, sema4D(−/−) platelets have 1) a rightward-shift in the dose/response curve for collagen-induced Akt phosphorylation, 2) a 37% smaller increase in cytosolic Ca++, and 3) a 43% smaller increase in PLCγ2 phosphorylation. However, we found no defect in collagen-induced FcRγ phosphorylation, which is the earliest event in GPVI signaling. The defect in PLCγ2 phosphorylation was not limited to mouse platelets, but was also observed when human platelets were stimulated with collagen in the presence of an antibody directed towards the sema4D extracellular domain. Taken together, these results show that sema4D is needed for optimal activation of PLCγ2 by collagen downstream of the GPVI/FcRγ complex. Sema4D is believed to act in part through contact-dependent binding of sema4D to its receptors, CD72 and plexin-B1. Since these studies were performed under conditions in which platelet:platelet contacts can occur, the observed defect in collagen and convulxin responses could be due to impaired signaling by either of these receptors or, in theory, by retrograde signaling via sema4D. One candidate mechanism involves a regulatory complex between CD72 and the tyrosine phosphatase, SHP-1, which we have shown to occur in resting human platelets and to be lost when platelets are activated by agonists or stimulated by soluble sema4D. In theory, sema4D-dependent loss of the CD72/SHP-1 complex allows SHP-1 to relax into an inactive conformation, promoting protein tyrosine phosphorylation, which would not occur when sema4D is absent or blocked.

Author notes

Disclosure: No relevant conflicts of interest to declare.