Insulin receptor kinase-associated phosphotyrosine phosphatases in hepatic endosomes: assessing the role of phosphotyrosine phosphatase-1B.

Previous work has shown that bisperoxo(1,10-phenanthroline)-oxovanadate(v) anion [bpV(phen)] induces potent insulin-mimicking effects in the rat, selectively activates the endosomal (EN) insulin receptor kinase (IRK) in liver, and markedly abolishes endosomal IRK-associated phosphotyrosine phosphatase (PTP) activity while reducing that of total ENs by approximately 30%. In this study we examined the relatively selective effect of bpv(phen) on endosomal PTP activities for the purpose of defining IRK-associated PTP(s). Using an in-gel PTP assay, we detected multiple (approximately 20) species of endosomal PTP (30 to >220 kDa), with five that were markedly inhibited after in vivo bpV(phen) administration. Using a combination of Mono Q anionic exchange chromatography and immunoblotting, we demonstrated that LAR (leukocyte common antigen-related), PTP-alpha, and PTP-1B were present in endosomal subfractions not significantly inhibited by bpv(phen). PTP-1B activity was assayed in immunoprecipitates from hepatic ENs of control and bpV(phen)-treated rats and was found to be inhibited by approximately 30% after bpv(phen) treatment. To clarify the role of PTP-1B in dephosphorylating IRK, we prepared hepatic ENs from wild-type and PTP-1B-null mice. We found that the phosphotyrosine content of IRK was similar in these two types of ENs, and that IRK dephosphorylation was not affected in ENs from PTP-1B-null mice compared with that in ENs from wild-type mice. These data suggest that LAR , PTP-alpha, and PTP-1B are not candidates for the IRK-associated PTP in hepatic ENs, and that IRK dephosphorylation in ENs may result from the concerted actions of several PTPs.

Gab2 tyrosine phosphorylation by a pleckstrin homology domain-independent mechanism: role in epidermal growth factor-induced mitogenesis.

In primary rat hepatocyte cultures, activation of phosphatidylinositol 3-kinase is both necessary and sufficient to account for epidermal growth factor (EGF)-induced DNA synthesis. In these cells, three major p85-containing complexes were formed after EGF treatment: ErbB3-p85, Shc-p85, and a multimeric Gab2-Grb2-SHP2-p85, which accounted for more than 80% of total EGF-induced PI3K activity (Kong, M., C. Mounier, J. Wu, and B. I. Posner, J Biol Chem, 2000, 275:36035-36042). More recently, we found that EGF-dependent tyrosine phosphorylation of endogenous Gab2 is essential for EGF-induced DNA synthesis in rat hepatocytes. Here we show that, after EGF treatment, ErbB3-p85 and Shc-p85 complexes were localized to plasma membrane and endosomes, whereas the multimeric Gab2-Grb2-SHP2-p85 complex was formed rapidly (peak at 30 sec) and exclusively in cytosol. Western blotting of subcellular fractions from intact liver and immunofluorescence analyses in cultured hepatocytes demonstrated that EGF did not promote the association of cytosolic Gab2 with cell membranes. These observations prompted us to evaluate the role of the PH domain of Gab2 in regulating its function. Overexpression of the PH domain of Gab2 did not affect EGF-induced Gab2 phosphorylation, PI3K activation, and DNA synthesis. Overexpressed Gab2 lacking the PH domain (DeltaPHGab2) was comparable to wild-type Gab2 in respect to EGF-induced tyrosine phosphorylation, recruitment of p85, and DNA synthesis. In summary, after EGF stimulation, ErbB3, Shc, and Gab2 are differentially compartmentalized in rat liver, where they associate with and activate PI3K. Our data demonstrate that Gab2 mediates EGF-induced PI3K activation and DNA synthesis in a PH domain-independent manner.