Search

To inhibit pathological thrombus formation without impairing hemostasis is the holy grail of anticoagulant therapy. Recent data from animal models have indicated that factor XII (FXII) may be a promising new antithrombotic target that is particularly intriguing due to the longstanding clinical observation that severe congenital FXII deficiency is not associated with a bleeding diathesis in humans. FXII is thought to participate in thrombus formation after being activated in high shear arterial environments. FXIIa then initiates downstream activation of the contact pathway, culminating in thrombin generation. However, the relevant cell surface for FXII activation remains unclear. Here we compare the role of platelets versus endothelial cells in FXII activation and study the function of surface phospholipids in this process. To explore the effect of FXII inhibition on thrombus formation in vivo, we used antibody X210-C01, a novel human IgG1 developed using phage-display technology that blocks both mouse and human FXIIa. Using a mouse laser injury model of arterial thrombosis, we showed that X210-C01 inhibited both fibrin formation and platelet accumulation at sites of vascular injury. Plasma removed from animals after completion of these experiments was used to quantify the concentrations of X210-C01 achieved in vivo at a given dose. FXII inhibition was somewhat more potent in preventing platelet accumulation (IC50 dose = 27 mg/kg, R2=0.93) than fibrin formation (IC50 dose = 43 mg/kg, R2=0.95). Importantly, treatment with X210-C01 at 100 mg/kg did not prolong bleeding times or increase total blood loss in a tail bleeding assay. To evaluate the mechanism underlying our in vivo observations, we studied the differential role of FXII in thrombin generation by stimulated platelets and endothelium. X210-C01 did not globally inhibit SFLLRN-induced platelet aggregation or granule release. We next performed a fluorogenic thrombin generation assay (TGA) using human platelets treated with the peptide agonist SFLLRN. X210-C01 inhibited platelet-based thrombin generation in a dose-dependent fashion, whereas anti-tissue factor (TF) and anti-factor VIIa (FVIIa) antibodies did not. By contrast, in a similar TGA using SFLLRN-stimulated endothelial cells, X210-C01 had no effect, while anti-TF antibodies abrogated thrombin generation. These results indicate that stimulated endothelium generates thrombin by a mechanism distinct from that of platelets. FXII is known to be activated in vitro by anionic surfaces. Because phosphatidylserine (PS) is a negatively-charged phospholipid expressed on the surface of stimulated platelets, we reasoned that PS may serve as the platelet-based activator of FXII. To test this hypothesis, we used lactadherin, a potent and specific inhibitor of PS, in the platelet-based TGA and showed that PS blockade inhibited platelet-based thrombin generation at concentrations as low as 10 nM. We then used a chromogenic FXIIa activity assay to test the ability of PS-containing liposomes to activate FXII. Liposomes containing 80% phosphatidylcholine (PC) and 20% PS (PC-PS 80/20) failed to activate FXII at concentrations as high as 100 µM. In this assay, SFLLRN-stimulated platelets led to significantly greater FXII activation than either resting platelets or the PC-PS 80/20 liposomes tested. In summary, we have made the unexpected finding that thrombin generation on the surface of stimulated platelets proceeds by a FXIIa-dependent pathway and does not require FVIIa or TF. By contrast, thrombin generation on endothelium requires TF but not FXIIa. Additionally, surface PS is necessary but not sufficient for platelet-based FXII activation and thrombin generation, pointing to the involvement of a second platelet component. Further studies will be directed towards investigating the in vivo role of platelet-based FXII activation in arterial thrombus formation. Disclosures Mason: Shire Pharmaceuticals: Employment. Kenniston:Shire Pharmaceuticals: Employment.

Platelet granule secretion is important not only for hemostasis and thrombosis, but also for a variety of physiological processes including inflammation, angiogenesis and malignancy. Vesicle Associated Membrane Proteins (VAMPs) are a group of v-SNARE proteins resident on the platelet granule surface that participate in granule secretion. Platelets contain several VAMP isoforms including VAMP-2, VAMP-3, VAMP-7, and VAMP-8. VAMP-7 is unique in that it contains an N-terminal profilin-like longin domain. Previous work by our group demonstrated spatial segregation of granules expressing different VAMPs during platelet spreading. Granules expressing VAMP-3 and VAMP-8 localized to the granulomere of spreading platelets, while those expressing VAMP-7 moved towards the periphery. Based on this observation, we proposed that VAMP-7+ granules move to the periphery of the spreading platelet to add membrane to growing actin structures. To assess this hypothesis, platelets from VAMP-7 null mice were used to analyze the role of VAMP-7 in platelet spreading, aggregation and secretion. VAMP-7 null platelets were normal in size, shape, and number. When compared to wild-type platelets, VAMP-7 null platelets did not show any defects in aggregation upon exposure to increasing doses of the PAR4 agonist peptide, AYPGKF, or collagen. In contrast, the surface area of VAMP-7 null platelets following 15 min of spreading on poly-L-lysine was only 51% that of wild-type of platelets (P < 0.05). To assess mechanisms of the movement of VAMP-7 to the platelet periphery, the association of VAMP-7 to the Triton X-100-insoluble platelet cytoskeleton was evaluated and results showed that VAMP-7 associated with the actin cytoskeleton. Moreover, VAMP-7 null platelets showed impaired P-selectin surface expression and PF4 secretion at low concentrations of AYPGKF. TIMP-2 and VEGF localize to VAMP-7 expressing granules in the periphery of spread platelets. We therefore evaluated the secretion of TIMP-2 and VEGF from VAMP-7 null platelets. Secretion of TIMP-2 and VEGF was reduced even at saturating doses of agonist (300 mM AYPGKF). To examine the role of VAMP-7 in a-granule exocytosis during platelet activation in vivo, PF4 release was monitored following laser-induced injury of cremaster arterioles. Platelet accumulation at sites of laser injury was identical in wild-type and VAMP-7 null mice. In wild-type mice, PF4 was secreted by activated platelets and bound back to activated endothelium and platelets producing a localized concentration of PF4 that accumulated over 15 min following injury. PF4 release from platelets lacking VAMP-7 was decreased to 47% of that of control. These results demonstrate that VAMP-7 interacts with the actin cytoskeleton and functions selectively in a-granule exocytosis. VAMP-7 associates with the actin cytoskeleton and functions during platelet spreading, adding further support to the premise that membrane fusion occurring during granule secretion is an essential component of normal platelet spreading. This VAMP-7 mediated, actin-dependent mechanism of secretion is not important for platelet thrombus formation, but rather functions in the release of particular granular contents, such as PF4, at sites of vascular injury. Disclosures: No relevant conflicts of interest to declare.

Abstract 361 Platelet granule secretion serves a central role in hemostasis and thrombosis. During platelet secretion, fusion of granule membranes with those of the plasma membrane results in the release of granule contents. Recently electrochemical techniques using single-cell amperometry have shown that platelet membrane fusion results in the formation of a fusion pore. The fusion pore subsequently expands to enable the complete extrusion of granule contents. However, the molecular mechanisms that control platelet fusion pore formation and expansion are not known. To discover novel components of the platelet secretory machinery, we tested >300,000 compounds in a forward chemical genetic screen designed to identify inhibitors of dense granule secretion. A compound, ML160, was found that blocked dense granule release with an IC50 of approximately 0.5 μM. ML160 was also identified in an unrelated high throughput screen designed to detect inhibitors of dynamin-related protein-1 (Drp-1). Although best known as mediators of membrane fission, dynamins also contribute to granule exocytosis by controlling fusion pore expansion. Immunoblot analysis of platelet pellets and supernatants confirmed the presence of Drp-1 in platelets and demonstrated nearly equal distribution between platelet membranes and cytosol. mDivi-1, a well-characterized small molecule inhibitor of Drp-1 that acts outside of the GTP binding site, blocked PAR1-mediated platelet dense granule and α-granule release with an IC50 of approximately 20 μM. mDivi-1 also inhibited granule release induced by the thromboxane receptor agonist U46619, PMA, or Ca2+ ionophore, indicating that Drp-1 acts distally in the secretory pathway. To assess whether Drp-1 functions in platelet fusion pore dynamics, we tested the effect of mDivi-1 on the release of dense granules from rabbit platelets using single-cell amperometry. This technique monitors the release of serotonin from single granules in real-time with sub-millisecond temporal resolution. mDivi-1 exposure (10 μM) retarded each release event, resulting in a prolonged spike width of 23.00 ± 1.702 msec compared to the control value of 14.71 ± 1.194 msec. Although this concentration of mDivi-1 did not change the overall percentage of the fusion pore events or the amount of serotonin released through the fusion pore, it showed a distinct effect on the transition from stable fusion pore to maximal fusion pore dilation (% foot= 17.46 ± 1.809%, 9.464 ± 2.014% for control and mDivi-1 conditions, respectively). Evaluation of fluorescein-dextran incorporation into activated platelets by fluorescence microscopy enabled visualization of fusion pore dynamics and confirmed the effect of mDivi-1 on fusion pore expansion. To assess whether Drp-1 participates in platelet function in vivo, we determined the effect of mDivi-1 on thrombus formation following laser-induced injury of mouse cremaster arterioles. mDivi-1 inhibited platelet accumulation at the site of vascular injury by 74%. In contrast, mDivi-1 had no significant effect on fibrin formation under the same conditions. These results identify Drp-1 in platelets, demonstrate a role for Drp-1 in fusion pore dynamics, and indicate that pharmacological regulation of platelet fusion pore expansion can be used to control thrombus formation in vivo. Disclosures: No relevant conflicts of interest to declare.