Abstract
Pleckstrin makes up approximately one percent of total cellular protein within platelets and leukocytes, a protein best known for containing the two prototypic Pleckstrin Homology (PH) domains. Following platelet activation, PKC rapidly phosphorylates pleckstrin, inducing it to bind membrane bound phospholipids such as phosphatidylinositol 4,5 bisphosphate (PIP2). Platelets also contain a widely expressed paralog of pleckstrin, called pleckstrin-2. Although the activity of pleckstrin is regulated through protein phosphorylation, pleckstrin-2 is not a phosphoprotein, but is instead activated by binding a specific PI3K generated phospholipid, phosphatidylinositol 3,4 bisphosphate (PI3,4P2). To understand the true in vivo role of these two proteins, we genetically engineered mice to lack individual or both pleckstrin isoforms. Pleckstrinnull platelets exhibit mildly impaired aggregation in response to thrombin, but fail to aggregate in response to thrombin in the presence of PI3K inhibitors. This suggests that a PI3K-dependent signaling pathway compensates for the loss of pleckstrin. Platelets lacking pleckstrin exhibit a marked defect in the secretion of delta and alpha granules following exposure to the PKC stimulant, PMA. Although pleckstrin-null platelets centralized and merged their granules in response to stimulation of PKC, they failed to empty their contents into the open canalicular system. These results differ from that seen with platelets lacking the other pleckstrin isoform, pleckstrin-2. Platelets derived from pleckstrin-2 null mice secrete and aggregate normally in response to thrombin and PMA. In addition, unlike the effect seen on pleckstrin knockout platelets, inhibitors of PI3K had no effect on the aggregation or secretion of pleckstrin-2 knockout platelets. Also in contrast to pleckstrin knockout platelets, pleckstrin-2 null platelets fail to secrete in response to thrombin when they were exposed to inhibitors of either PLC or PKC. These data demonstrate that pleckstrin-2 knockout platelets compensate for their secretion defect by a pathway dependent on PLC and PKC. It is notable that PI3K or PKC inhibitors only minimally affected the thrombin-induced secretion of wild-type platelets unless both inhibitors were used together. Together, these results suggest that platelets utilize parallel signaling pathways, one dependent on PKC and pleckstrin, and the other on PI3K and pleckstrin-2. Studies in platelets and neuronal cells suggest that disassembly of the actin cytoskeleton is required for secretion. Since overexpression studies have suggested that both pleckstrin and pleckstrin-2 can modulate the actin cytoskeleton, we hypothesized that both pleckstrin isoforms affect secretion through an actin-dependent pathway. To test this hypothesis, we analyzed the effect of the pleckstrin and pleckstrin-2 null mutations on actin organization within platelets. When pleckstrin null platelets were allowed to adhere to immobilized fibrinogen, or when they were flowed over collagen-coated surfaces, they exhibited impaired adherence and spreading. Phalloidin staining indicated that they also assembled less F-actin than normal platelets. Similarly, platelets lacking pleckstrin-2 also adhered and spread poorly. Since we have shown that pleckstrin and pleckstrin-2 perform analogous roles in complementary signaling pathways, we bred mice to generate a murine lacking both pleckstrin isoforms. Platelets lacking both pleckstrin and pleckstrin-2 exhibited a marked spreading defect in response to PMA (0% of control) or thrombin (18% of control). Following stimulation with PMA, platelets containing the double null mutation also failed to increase in their F-actin content during the spreading process (8% of control). Electron micrographs of platelets lacking both pleckstrin and pleckstrin-2 revealed that the double null platelets fail to extend any broad lamellipodia, and instead, only extended small membrane blebs. These data show that pleckstrin and pleckstrin-2 are absolutely essential for the cytoskeletal organization that occurs during platelet adhesion. These data also demonstrate that adhesion-induced cytoskeletal changes within platelets can be mediated by one of two parallel pathways, the first involving PKC and pleckstrin, and the second involving PI3K and pleckstrin-2.
Cdc42p regulation of the yeast formin Bni1p mediated by the effector Gic2p