Prominent Role for the Gap Junction Protein Connexin37 in Hemostasis and Thrombosis.

Formation of platelet plug initiates hemostasis at sites of vascular injury, and triggers pathological thrombosis in ischemic tissue disease. Although various crucial molecules for platelet function have been identified in recent years, the mechanisms of inter- and intra-cellular signaling leading to the formation of a stable platelet plug are still poorly understood. Connexins form gap junctions, clusters of intercellular channels that are known to synchronize responses in multi-cellular organisms through the direct exchange of ions, small metabolites and other second messenger molecules between adjacent cells. Here, we report the expression of the gap junction protein connexin37 (Cx37) in mouse and human platelets. In addition, we observed functional gap junction communication between platelets during platelet aggregation in vitro, as assessed by microinjection of the gap junction-permeable tracer neurobiotin in platelets isolated from human or wild-type mice. In contrast, the tracer did not spread within the platelet aggregates obtained from Cx37-deficient mice (Cx37−/−, n=4). To assess hemostasis and thrombosis in mice in vivo, we have transsected the tail of Cx37−/− and control Cx37+/+ mice and obtained significant differences in the cessation of bleeding: bleeding time was 3.8±0.5 min for Cx37−/− mice and 8.8±1.4 min for Cx37+/+ controls (mean±SEM, n=20, P<0.005). In addition, platelet aggregation in response to collagen and ADP was increased in Cx37−/− when compared to Cx37+/+ mice (n=4, P<0.05). A similar effect on platelet aggregation was observed when murine wild-type platelets where pretreated with 200mM Cx37 mimetic blocking peptides. We then performed aggregation studies with human washed platelets with or without pre-incubation with the pharmacological gap junction blocker a-glycerrhetinic acid (aGA). Platelets were incubated 5 min with 50mM aGA or vehicle before adding an aggregating agent. Incubation with aGA increased significantly the platelet response with each agonist used: median maximal aggregation without or with aGA were 0% and 84% for arachidonic acid (P<0.05), 42% and 67% for ADP (P<0.05) and 17% and 59% for collagen (P<0.05). In agreement with these data, we also observed remarkable differences in the time of survival after inducing thromboembolism by injecting a mixture of collagen/epinephrine into the jugular vein of these two groups of mice: control mice survived 4.2±0.4 min and Cx37−/− mice only 1.9±0.3 min (mean±SEM, n=8, P<0.001). Finally, we recorded thrombus formation by intravital microscopy in the mesenteric arteries after vessel wall injury induced by FeCl3. We found that rhodamine 6G-labeled platelets recruitment at the site of the vascular injury was faster in Cx37−/− than in Cx37+/+ mice (n=4). Taken together, our results demonstrate that Cx37 plays a crucial role in the physiological and pathological signaling in platelets leading to hemostasis and thrombosis. We hypothesize that gap junctions synchronize responses in platelets brought in close contact during their activation, likely by transmitting intracellular signals to neighboring cells.