Mice That Lack RhoA In Their Platelets Have Normal Actin Dynamics, but Have Macrothrombocytopenia

Abstract 549

In response to agonist stimulation, platelets undergo a rapid reorganization of their actin cytoskeleton. This process involves simultaneous disassembly and assembly of filamentous actin, and is one of the earliest phenomena seen in platelet activation. Ex vivo flow models suggest that the platelet cytoskeleton is required for platelet adhesion that can withstand the shear conditions found within the arterial vascular system. The signaling pathways that link external stimuli with actin assembly are believed to include polyphosphoinositides, small GTP-binding proteins, and actin binding proteins. Extrapolations of data, mostly derived from tissue culture cell lines, suggest that a central component of this signaling cascade is the small GTP binding protein, RhoA. A few studies using a RhoA-specific pharmacologic inhibitor, C3 exotoxin, suggest that RhoA is essential for platelet spreading and focal adhesion formation. These findings support the hypothesis that RhoA within platelets is critical for the cytoskeletal dependent processes that contribute to platelet plug formation. To determine the true in vivo role of RhoA within platelets, we utilized a murine genetic approach. Mice were genetically modified to contain conditional RhoA null mutation by inserting LoxP sites flanking exon 3. This exon encodes the P-loop and Switch 1 domains within this protein. RhoA ^fl/fl mice were crossed with Platelet factor 4 (PF4) expressing Cre mice. The PF4 promotor leads to Cre expression exclusively in platelets and megakaryocytes, thereby producing homologous recombination at the LoxP sites, and deletion the critical exon only within these cells. This end result of this breeding strategy produced RhoA ^fl/fl PF4 Cre+ mice that specifically lacked RhoA only in their platelets and megakaryocytes. RhoA ^fl/fl PF4 Cre+ mice were compared with their RhoA ^fl/fl PF4 Cre- littermates. RhoA ^fl/fl PF4 Cre+ mice appeared normal, but had platelet counts that were 30% +/− 3% lower than normal. The mean platelet volume was also increased by 25 % +/− 7 % in the RhoA-null platelets. Review of the peripheral blood smears confirmed that the mice had macrothrombocytopenia, but did not reveal any abnormalities in the erythrocytes or leukocytes of the mice. Examination of the bone marrows from RhoA ^fl/fl PF4 Cre+ mice demonstrated that they had at least as many megakaryocytes as RhoA ^fl/fl PF4 Cre- mice. But compared to the control cells, the RhoA-null megakaryocytes were larger, more lobulated, and had more cytoplasm. Furthermore, the thromobocytopenia is probably not due to splenic sequestration because the spleens of RhoA ^fl/fl PF4 Cre- mice were only minimally larger (less than 10%) than those of the control mice. These results suggest that the mechanism for thrombocytopenia is due to peripheral destruction. Platelets derived from RhoA ^fl/fl PF4 Cre+ mice were studied ex vivo, and were found to undergo shape change and aggregate normally in response to thrombin, collagen, and the thromboxane A2 analog, U46619. Surprisingly, platelet adhesion and cell spreading was also unaffected by the loss of RhoA. It is also remarkable that total F-actin (as assessed by phalloidin staining) was identical in the platelets derived from RhoA ^fl/fl PF4 Cre+ and RhoA ^fl/fl PF4 Cre- mice. Our results definitively refute the model that RhoA is an essential component of platelet actin dynamics and platelet adhesion. Instead our findings surprisingly indicate that loss of the platelet RhoA causes macrothrombocytopenia. Our data suggests that the development of macrothrombocytopenia is due to an intrinsic platelet abnormality that leads to a shortened platelet lifespan.