Abstract
Pleckstrin-2, a widely expressed paralog of pleckstrin, is composed of two Pleckstrin Homology (PH) domains and Disheveled-Egl 10-Pleckstrin (DEP) domain. Although the activity of pleckstrin is regulated by its’ phosphorylation state, pleckstrin-2 is not a phospho-protein suggesting that it possesses a different mechanism of regulation. Previous reports have shown that many PH domains mediate protein binding to inositol phosphates and phospholipids, thus regulating protein function. Therefore, we speculated that localized production of specific polyphosphatidylinositols might bind, and activate, pleckstrin-2. Using a lipid-binding assay, we found that pleckstrin-2 bound with greatest affinity to the products of phosphatidylinositol 3-kinase and phosphatidylinositol 5-kinase. The individual PH domains of pleckstrin-2 bound the same products but with lower affinity, implying that both PH domains cooperate for maximal lipid affinity of the full-length protein. GFP-tagged pleckstrin-2 had a cytoplasmic distribution in non-adherent Jurkat cells, but through a pathway dependent on the phospholipid-binding pocket of its PH domains, rapidly moved to the cell membrane following adhesion to immobilized fibronectin. Once bound to the cell membrane, pleckstrin-2 enhanced Jurkat cell spreading 2-fold and increased membrane ruffling. The membrane association of pleckstrin-2, and its resultant cell spreading, were dependent on D3-phosphoinositides since these effects were disrupted by pharmacologic inhibition of PI3K with either wortmannin or LY294002. To investigate the role of this protein within platelets, we generated mice containing a null mutation within the pleckstrin-2 gene. Pleckstrin-2 null mice were born at the expected frequency, and had no overt spontaneous hemorrhagic events. Mice lacking pleckstrin-2 had normal platelet counts and morphologic appearance of their megakaryocytes. Following stimulation of the PAR-4 (thrombin) receptor, pleckstrin-2 knockout platelets displayed normal assembly of filamentous actin. However, pleckstrin-2 null platelets had impaired aggregation following stimulation by collagen or submaximal doses of the PAR-4 activating peptide. Since pleckstrin-2 deficient platelets aggregated normally in response to ADP, these results suggested that these platelets might have an impaired ability to secrete dense granules. Accordingly, we found that pleckstrin-2 null platelets had a defect in their ability to secrete ATP in response to stimulation by 5μM collagen or 200μM of the PAR-4 activating peptide. However, pleckstrin-2 knockout platelets did incorporate 14C-serotonin as efficiently as wild type platelets. This latter observation suggested that the secretion defect in pleckstrin-2 null platelets was not attributable to a deficiency of dense granules, but instead is due to a defect in exocytosis of granules. Together, these data suggest that the PH domains of pleckstrin-2 cooperatively bind PI3K generated phospholipids on the cell membrane, and help mediate platelet secretion.
Author notes
Corresponding author