Phosphatidylinositol (P1) 3-kinase activity underlies key biochemical events in signal transduction elicited by hormones and growth factors, such as activation of the Akt/PKB pathway. P1 3-kinases are also crucial in the regulation of endosomal trafficking, which in turn affects signal transduction by regulating receptor internalization, recycling and degradation. Because of the importance of these enzymes in fundamental aspects of hormone action, our goal has been to elucidate the molecular basis of PI 3-kinase function by identifying and characterizing direct effectors of the 3'-phosphoinositides produced by these kinases. In the previous funding period, we demonstrated that a specific protein motif, the FYVE domain, binds to Ptdlns(3)P, the major product of PI 3-kinase activity in mammalian cells, with high affinity and specificity. Approximately 40 mammalian proteins contain FYVE domains. Several are involved in endosome trafficking, but others have been implicated in signal transduction. One example is SARA (Smad Anchor for Receptor Activation), a 140 kD protein necessary for the phosphorylation and nuclear translocation of the transcription factor Smad2 in response to TGF-beta signaling. Preliminary work shown here indicates that, like other EWE-domain containing proteins, SARA is localized on early endosomes and disruption of this endosomal localization impairs TGF-beta-stimulated Smad2 nuclear translocation. These results suggest that the assembly of certain signal transduction pathways, such as those directing transcriptional activation, may occur on early endosomes, and the sensitivity of these pathways to P1 3-kinase inhibitors may result from the impairment in the binding of EWE domain-containing proteins to Ptdlns(3)P. Our current goal is to further understand the structural basis for the interaction of EWE domain-containing proteins with Ptdlns(3)P and with endosomes, and to test the hypothesis that these interactions are requisite in signal transduction to transcriptional activation. We will use methodology that includes chemically induced dimerization of EWE domains to test the hypothesis, derived from crystallization studies, that these domains function as obligatory oligomers. To test the hypothesis that signal transduction to nuclear import requires endosomal localization, we will study two paradigms; the TGFbeta to Smad2 pathway, and the insulin to p42/44 MAP kinase pathway.
