The CD40/CD40 Ligand (CD40L) Pathway Plays a Role in Primary Hemostasis and Platelet Function.

Previous studies have suggested a role for platelet CD40L in thrombosis and atherosclerosis. However, there are contradictory reports on the biologic activity of its soluble variant (sCD40L) and the involved receptor signaling pathways (CD40 vs αIIbβ3). Furthermore, CD40L mAb-associated thromboembolic complications in recent human and animal studies have raised additional questions about the pro-thrombotic properties of this molecule. This study was conducted to further investigate the function of the CD40/CD40L dyad in primary hemostasis and platelet function. CD40−/− and CD40L−/− mice and mice deficient for both genes (“double knock-out”) showed prolonged tail vein bleeding and platelet function analyzer (PFA-100) closure times as compared to their wild-type littermates, indicating an inherited defect in platelet function. Recombinant human sCD40L (rsCD40L), chemical cross-linking of which yielded a single reaction product compatible with a trimeric structure of the protein in solution, bound to CD40 on resting platelets and induced CD62P (P-selectin) expression in a concentration-dependent manner (0–5 μg/ml) from 1±1 to 23±5% positivity (means±SD, n=4–8; P<0.01). This response was completely abolished by CD40 mAb M3 and not affected by blocking the β3 integrin (CD61) with abciximab. In contrast, CD40 mAb G28-5 significantly enhanced rsCD40L-induced CD62P expression to 51±5%. This agonistic effect was strongest when G28-5 was added to the platelets after rsCD40L. Pre-incubation of rsCD40L with CD40L mAb M90 also had a potentiating effect, showing an inverted V-shaped dose response curve with maximum levels of CD62P+ platelets (60–95%) at a molar M90/rsCD40L ratio of 1/3. G28-5 and M90 alone had no effect, but their combination in the presence of rsCD40L proved synergistic. Experiments with corresponding F(ab’) fragments and the FcγRII-inhibitory mAb IV.3 demonstrated that G28-5- and M90-mediated additional platelet activation resulted from signaling through FcγRII cross-linked to CD40 and rsCD40L bound to CD40 on the platelet surface, respectively. The CD40L mAb TRAP-1 showed similar characteristics to M90, but its synergistic effect with rsCD40L was less pronounced. Platelet activation by rsCD40L in combination with G28-5 and/or M90 also induced morphological shape changes, fibrinogen binding, dense granule release, microparticle generation, and monocyte-platelet-conjugate formation. Interestingly, consistent with their platelet function-modulating effects, M3 but not G28-5 prolonged PFA-100 closure times of normal human blood. This work provides genetic evidence for a role of CD40 and CD40L in primary hemostasis. Our mechanistic studies further suggest that CD40-mediated platelet activation by CD40L, in its membrane and/or soluble form, may be at least partially involved in this process. The results also offer a potential explanation for the unexpected high incidence of CD40L mAb-associated thrombotic events in patients with systemic lupus erythematosus, an inflammatory autoimmune disease characterized by elevated levels of circulating sCD40L. The underlying pathomechanism, in which sCD40L as a platelet-derived, homotrimeric protein causes multimeric clustering of bivalent antibodies on the platelet surface with subsequent FcγRII-mediated platelet activation, resembles other known immune thrombophilias such as the heparin-induced thrombocytopenia (HIT) syndrome.