Search

Summary Background Phosphatase and tensin homolog (PTEN) mediates many of its effects on proliferation, growth, survival, and migration through its PtdIns(3,4,5)P 3 lipid phosphatase activity, suppressing phosphoinositide 3-kinase (PI3K)-dependent signaling pathways. PTEN also possesses a protein phosphatase activity, the role of which is less well characterized. Results We have investigated the role of PTEN in the control of cell migration of mesoderm cells ingressing through the primitive streak in the chick embryo. Overexpression of PTEN strongly inhibits the epithelial-to-mesenchymal transition (EMT) of mesoderm cells ingressing through the anterior and middle primitive streak, but it does not affect EMT of cells located in the posterior streak. The inhibitory activity on EMT is completely dependent on targeting PTEN through its C-terminal PDZ binding site, but can be achieved by a PTEN mutant (PTEN G129E) with only protein phosphatase activity. Expression either of PTEN lacking the PDZ binding site or of the PTEN C2 domain, or inhibition of PI3K through specific inhibitors, does not inhibit EMT, but results in a loss of both cell polarity and directional migration of mesoderm cells. The PTEN-related protein TPTE, which normally lacks any detectable lipid and protein phosphatase activity, can be reactivated through mutation, and only this reactivated mutant leads to nondirectional migration of these cells in vivo. Conclusions PTEN modulates cell migration of mesoderm cells in the chick embryo through at least two distinct mechanisms: controlling EMT, which involves its protein phosphatase activity; and controlling the directional motility of mesoderm cells, through its lipid phosphatase activity.

Background: The minor membrane phospholipid phosphatidylinositol 4,5-bisphosphate (PIP 2 ) has been implicated in the control of a number of cellular processes. Efficient synthesis of this lipid from phosphatidylinositol has been proposed to require the presence of a phosphatidylinositol/phosphatidylcholine transfer protein (PITP), which transfers phosphatidylinositol and phosphatidylcholine between membranes, but the mechanism by which PITP exerts its effects is currently unknown. The simplest hypothesis is that PITP replenishes agonist-sensitive pools of inositol lipids by transferring phosphatidylinositol from its site of synthesis to sites of consumption. Recent cellular studies, however, led to the proposal that PITP may play a more active role as a co-factor which stimulates the activity of phosphoinositide kinases and phospholipase C (PLC) by presenting protein-bound lipid substrates to these enzymes. We have exploited turkey erythrocyte membranes as a model system in which it has proved possible to distinguish between the above hypotheses of PITP function. Results: In turkey erythrocyte ghosts, agonist-stimulated PIP 2 hydrolysis is initially rapid, but it declines and reaches a plateau when ∼15% of the phosphatidylinositol has been consumed. PITP did not affect the initial rate of PIP 2 hydrolysis, but greatly prolonged the linear phase of PLC activity until at least 70% of phosphatidylinositol was consumed. PITP did not enhance the initial rate of phosphatidylinositol 4-kinase activity but did increase the unstimulated steady-state levels of both phosphatidylinositol 4-phosphate and PIP 2 by a catalytic mechanism, because the amount of polyphosphoinositides synthesized greatly exceeded the molar amount of PITP in the assay. Furthermore, when polyphosphoinositide synthesis was allowed to proceed in the presence of exogenous PITP, after washing ghosts to remove PITP before activation of PLC, enhanced inositol phosphate production was observed, whether or not PITP was present in the subsequent PLC assay. Conclusion: PITP acts by catalytically transferring phosphatidylinositol down a chemical gradient which is created as a result of the depletion of phosphatidylinositol at its site of use by the concerted actions of the phosphoinositide kinases and PLC. PITP is therefore not a co-factor for the phosphoinositide-metabolizing enzymes present in turkey erythrocyte ghosts.

Background: Protein kinase B (PKB) is activated by phosphorylation of Thr308 and of Ser473. Thr308 is phosphorylated by the 3-phosphoinositide-dependent protein kinase-1 (PDK1) but the identity of the kinase that phosphorylates Ser473 (provisionally termed PDK2) is unknown. Results: The kinase domain of PDK1 interacts with a region of protein kinase C-related kinase-2 (PRK2), termed the PDK1-interacting fragment (PIF). PIF is situated carboxy-terminal to the kinase domain of PRK2, and contains a consensus motif for phosphorylation by PDK2 similar to that found in PKBα, except that the residue equivalent to Ser473 is aspartic acid. Mutation of any of the conserved residues in the PDK2 motif of PIF prevented interaction of PIF with PDK1. Remarkably, interaction of PDK1 with PIF, or with a synthetic peptide encompassing the PDK2 consensus sequence of PIF, converted PDK1 from an enzyme that could phosphorylate only Thr308 of PKBα to one that phosphorylates both Thr308 and Ser473 of PKBα in a manner dependent on phosphatidylinositol (3,4,5) trisphosphate (PtdIns(3,4,5)P 3 ). Furthermore, the interaction of PIF with PDK1 converted the PDK1 from a form that is not directly activated by PtdIns(3,4,5)P 3 to a form that is activated threefold by PtdIns(3,4,5)P 3 . We have partially purified a kinase from brain extract that phosphorylates Ser473 of PKBα in a PtdIns(3,4,5)P 3 -dependent manner and that is immunoprecipitated with PDK1 antibodies. Conclusions: PDK1 and PDK2 might be the same enzyme, the substrate specificity and activity of PDK1 being regulated through its interaction with another protein(s). PRK2 is a probable substrate for PDK1.