Faculty Bibliography

Receptor protein tyrosine kinases (RTKs) transmit downstream signals via interactions with secondary signaling molecules containing SH2 domains. Although many SH2-phosphotyrosyl interactions have been defined in vitro, little is known about the physiological significance of specific RTK/SH2 interactions in vivo. Also, little is known about the mechanisms by which specific RTKs interact with and/or are regulated by specific protein tyrosine phosphatases (PTPs). To address such issue, we carried out a genetic analysis of the previously reported biochemical interaction between the RTK c-Kit, encoded at the W locus, and the SH2-containing non-transmembrane PTP SHP1, encoded at the motheaten (me) locus (1). Mice carrying a kinase-defective allele of c-Kit (Wv/+) were crossed with me/+ mice, which carry one effectively null allele of SHP1, and then backcrossed to generate all possible allelic combinations. Our results indicate strong intergenic complementation between these loci in hematopoietic progenitor cells. Compared to progenitors purified from normal mice, bone marrow progenitor cells (lin-) from me/me mice markedly hyper-proliferated in response to Kit ligand (KL). stimulation. Superimposition of the me/me genotype increased the number of one marrow-derived CFU-E from Wv/+ mice. Conversely, the presence of one or two copies of Wv decreased the number of macrophages and granulocytes in me/me lung, skin, peripheral blood and bone marrow, thereby decreasing the severity of the me/me phenotype. The decrease in dermal mast cells in Wv/Wv mice was rescued to levels found in Wv/+mice by superimposition of the me/me genotype. Surprisingly, however, the presence or absence of SHP1 had no effect on the proliferative response of bone marrow-derived cultured mast cells to KL or IL3 ex vivo. Nevertheless, the immediate-early response to KL stimulation, as measured by KL-induced tyrosyl phosphorylation, was substantially increased in mast cells from Wv/+:me/me compared to Wv/ +:+/+ mice, strongly suggesting that SHP1 directly dephosphorylates and regulates c-Kit. Taken together, our results establish that SHP1 negatively regulates signaling from c-Kit in vivo, but in a cell type-specific manner.

SHPTP1 (PTP1C, HCP, SHP) is an SH2 domain-containing tyrosine phosphatase expressed predominantly in hematopoietic cells. A frameshift mutation in the SHPTP1 gene causes the motheaten (me/me) mouse. These mice are essentially SHPTP1 null and display multiple hematopoietic abnormalities, most prominently hyperproliferation and inappropriate activation of granulocytes and macrophages. The me/me phenotype suggests that SHPTP1 negatively regulates macrophage proliferative pathways. Using primary bone marrow-derived macrophages from me/me mice and normal littermates, we examined the role of SHPTP1 in regulating signaling by the major macrophage mitogen colony-stimulating factor 1 (CSF-1) (also known as macrophage colony-stimulating factor). Macrophages from me/me mice hyperproliferate in response to CSF-1. In the absence of SHPTP1, the CSF-1 receptor (CSF-1R) is hyperphosphorylated upon CSF-1 stimulation, suggesting that SHPTP1 dephosphorylates the CSF-1R. At least some CSF-1R-associated proteins also are hyperactivated. SHPTP1 is associated constitutively, via its SH2 domains, with an unidentified 130-kDa phosphotyrosyl protein (P130). P130 and SHPTP1 are further tyrosyl phosphorylated upon CSF-1 stimulation. Tyrosyl-phosphorylated SHPTP1 binds to Grb2 via the Grb2 SH2 domain. Moreover, in me/me macrophages, Grb2 is associated, via its SH3 domains, with several tyrosyl phosphoproteins. These proteins are hyperphosphorylated on tyrosyl residues in me/me macrophages, suggesting that Grb2 may recruit substrates for SHPTP1. Our results indicate that SHPTP1 is a critical negative regulator of CSF-1 signaling in vivo and suggest a potential new function for Grb2.

Using transient overexpression and microinjection approaches, we examined SHPTP2's function in growth factor signaling. Overexpression of catalytically inactive SHPTP2 (PTP2CS) but not catalytically inactive SHPTP1, inhibited mitogen-activated protein (MAP) kinase activation and Elk-1 transactivation following epidermal growth factor (EGF) stimulation of 293 cells. An SHPTP2 mutant with both C-terminal tyrosyl phosphorylation sites converted to phenylalanine (PTP2YF) was also without effect; moreover, PTP2YF rescued PTP2CS-induced inhibition of EGF-induced Elk-1 transactivation. PTP2CS did not inhibit transactivation by activated Ras, suggesting that SHPTP2 acts upstream of or parallel to Ras. Neither PTP2CS nor PTP2YF inhibited platelet-derived growth factor (PDGF)-induced Elk-1 transactivation. Thus, protein-tyrosine phosphatase activity, but not tyrosyl phosphorylation of SHPTP2, is required for the immediate-early responses to EGF but not to PDGF. To determine whether SHPTP2 is required later in the cell cycle, we assessed S-phase entry in NIH 3T3 cells microinjected with anti-SHPTP2 antibodies or with a glutathione S-transferase (GST) fusion protein encoding both SH2 domains (GST-SH2). Microinjection of anti-SHPTP2 antibodies prior to stimulation inhibited EGF- but no PDGF- or serum-induced S-phase entry. Anti-SHPTP2 antibodies or GST-SH2 fusion protein could inhibit EGF-induced S-phase entry for up to 8 h after EGF addition. Although MAP kinase activation was detected shortly after EGF stimulation, no MAP kinase activation was detected around the restriction point. Therefore, SHPTP2 is absolutely required for immediate-early and late events induced by some, but not all, growth factors, and the immediate-early and late signal transduction pathways regulated by SHPTP2 are distinguishable.

In hunting for unknown genes on the human X chromosome, we identified a cDNA in Xq28 encoding a transmembrane protein (SEX) of 1871 amino acids. SEX shares significant homology with the extracellular domain of the receptors encoded by the oncogenes MET, RON, and SEA [hepatocyte growth factor (HGF) receptor family]. Further screenings of cDNA libraries identified three additional sequences closely related to SEX: these were named SEP, OCT, and NOV and were located on human chromosomes 3p, 1, and 3q, respectively. The proteins encoded by these genes contain large cytoplasmic domains characterized by a distinctive highly conserved sequence (SEX domain). Northern blot analysis revealed different expression of the SEX family of genes in fetal tissues, with SEX, OCT, and NOV predominantly expressed in brain, and SEP expressed at highest levels in kidney. In situ hybridization analysis revealed that SEX has a distinctive pattern of expression in the developing nervous system of the mouse, where it is found in postmitotic neurons from the first stages of neuronal differentiation (9.5 day postcoitus). The SEX protein (220 kDa) is glycosylated and exposed at the cell surface. Unlike the receptors of the HGF family, p220SEX, a MET-SEX chimera or a constitutively dimerized TPR-SEX does not show tyrosine kinase activity. These data define a gene family (SEX family) involved in the development of neural and epithelial tissues, which encodes putative receptors with unexpected enzymatic or binding properties.

Shaggy is a downstream component of the wingless and Notch signaling pathways which operate during Drosophila development. To address the role of glycogen synthase kinase 3 beta (GSK3 beta), a mammalian homologue of Shaggy, in vertebrate embryogenesis, it was overexpressed in Xenopus embryos. Microinjection of rat GSK3 beta mRNA into animal ventral blastomeres of 8-cell-stage embryos triggered development of ectopic cement glands with an adjacent anterior neural tissue as evidenced by in situ hybridization with Xotx2, a fore/midbrain marker, and NCAM, a pan-neural marker. In contrast, animal dorsal injection of the same dose of GSK3 beta mRNA caused eye deficiencies, whereas vegetal injections had no pronounced effects on normal development. Using several mutated forms of rat GSK3 beta, we demonstrate that the observed phenotypes are dose-dependent and tightly correlate with GSK3 beta enzymatic activity. Lineage tracing experiments showed that the effects of GSK3 beta are cell autonomous and that ectopic cement glands and eye deficiencies arose directly from cells containing GSK3 beta mRNA. Molecular marker analysis of ectodermal explants overexpressing GSK3 beta has revealed activation of Xotx2 and of cement gland marker XAG-1, but expression of NCAM and XIF-3 was not detected. Phenotypic effects of mRNA encoding a Xenopus homologue of GSK3 beta were identical to those of rat GSK3 beta mRNA. We hypothesize that GSK3 beta mediates the initial steps of neural tissue specification and modulates anteroposterior ectodermal patterning via activation of Otx2 transcription. Our observations implicate GSK3 beta in signaling pathways operating during neural tissue development and during specification of anterior ectodermal cell fates.

Interferons (IFNs) induce early-response genes by stimulating Janus family (Jak) tyrosine kinases, leading to tyrosine phosphorylation of Stat transcription factors. Previous studies implicated protein-tyrosine phosphatase (PTP) activity in the control of IFN-regulated Jak/Stat signaling, but the specific PTPs responsible remained unidentified. We have found that SH2 domain-containing PTP1 (SHPTP1; also called PTP1C, HCP, or SHP) reversibly associates with the IFN-alpha receptor complex upon IFN addition. Compared with macrophages from normal littermate controls, macrophages from motheaten mice, which lack SHPTP1, show dramatically increased Jak1 and Stat1 alpha tyrosine phosphorylation, whereas Tyk2 and Stat2 activation is largely unaffected. These findings correlate with selectively increased complex formation on a gamma response element, but not an IFN-stimulated response element, in motheaten macrophages. Our results establish that SHPTP1 selectively regulates distinct components of Jak/Stat signal transduction pathways in vivo.

Retinoic acid (RA) is required for normal airway epithelial cell growth and differentiation both in vivo and in vitro. One of the earliest events following the exposure of bronchial epithelial cells to RA is the strong induction of RA receptor beta (RAR beta) mRNA. Previous work established that many lung cancer cell lines and primary tumors display abnormal RAR beta mRNA expression, most often absence or weak expression of the RAR beta 2 isoform, even after RA treatment. Restoration of RAR beta 2 into RAR beta-negative lung cancer cell lines has been reported to inhibit tumorigenicity. Since RAR beta 2 inactivation may contribute to lung cancer, we have investigated the molecular mechanism of defective RAR beta 2 expression. Nuclear run-on assays and transient transfections with RAR beta 2 promoter constructs indicate the presence of trans-acting transcriptional defects in most lung cancer cell lines, which map to the RA response element (RARE). These defects cannot be complemented by RAR-retinoid X receptor cotransfection and can be separated into two types: (i) one affecting transcription from direct repeat RAREs, but not palindromic RAREs, and (ii) another affecting transcription from both types of RARE. Studies using chimeras between RAR alpha, TR alpha, and other transcription factors suggest the existence of novel RAR-thyroid hormone receptor AF-2-specific cofactors, which are necessary for high levels of transcription. Furthermore, these factors may be frequently inactivated in human lung cancer.

Insulin activates hexose transport via at least two mechanisms: a p21ras-dependent pathway, leading to an increase in the amount of cell surface GLUT1; and a metabolic, p21ras-independent pathway, leading to translocation of the insulin-responsive transporter GLUT4 to the cell surface. Following insulin stimulation, SHPTP2, a non-transmembrane protein-tyrosine phosphatase, associates with insulin receptor substrate 1 via its Src homology 2 (SH2) domains. Microinjection of a glutathione S-transferase fusion protein encoding the N- and C-terminal SH2 domains of SHPTP2 (GST-NC-SH2) or anti-SHPTP2 antibodies into NIH-3T3 fibroblasts overexpressing the insulin receptor blocks insulin-induced DNA synthesis. Microinjection of either GST-NC-SH2 or anti-SHPTP2 antibodies into 3T3-L1 adipocytes inhibited the insulin-stimulated increase in expression of GLUT1. In contrast, translocation of GLUT4 to the cell surface was unaffected by either GST-NC-SH2 or anti-SHPTP2 antibodies. These data confirm a role for SHPTP2 in insulin-stimulated mitogenesis and indicate that whereas SHPTP2 is necessary for insulin-stimulated expression of GLUT1, it is not required for activation of the metabolic pathway leading to GLUT4 translocation.

The binding of erythropoietin (EPO) to its receptor (EPO-R) activates the protein tyrosine kinase JAK2. The mechanism of JAK2 inactivation has been unclear. We show that the hematopoietic protein tyrosine phosphatase SH-PTP1 (also called HCP and PTP1C) associates via its SH2 domains with the tyrosine-phosphorylated EPO-R. In vitro binding studies suggest that Y429 in the cytoplasmic domain of the EPO-R is the binding site for SH-PTP1. Mutant EPO-Rs lacking Y429 are unable to bind SH-PTP1; cells expressing such mutants are hypersensitive to EPO and display prolonged EPO-induced autophosphorylation of JAK2. Our results suggest that activation of SH-PTP1 by binding to the EPO-R plays a major role in terminating proliferative signals.