Src family kinases (SFKs) have been implicated as important regulators of ligand-induced cellular responses including proliferation, survival, adhesion and migration. Analysis of SFK function has been impeded by extensive redundancy between family members. We have generated mouse embryos harboring functional null mutations of the ubiquitously expressed SFKs Src, Yes and Fyn. This triple mutation leads to severe developmental defects and lethality by E9.5. To elucidate the molecular mechanisms underlying this phenotype, SYF cells (deficient for Src, Yes and Fyn) were derived and tested for their ability to respond to growth factors or plating on extracellular matrix. Our studies reveal that while Src, Yes and Fyn are largely dispensable for platelet-derived growth factor (PDGF)-induced signaling, they are absolutely required to mediate specific functions regulated by extracellular matrix proteins. Fibronectin-induced tyrosine phosphorylation of focal adhesion proteins, including the focal adhesion kinase FAK, was nearly eliminated in the absence of Src, Yes and Fyn. Furthermore, consistent with previous reports demonstrating the importance of FAK for cell migration, SYF cells displayed reduced motility in vitro. These results demonstrate that SFK activity is essential during embryogenesis and suggest that defects observed in SYF triple mutant embryos may be linked to deficiencies in signaling by extracellular matrix-coupled receptors.

Receptor tyrosine kinases (RTKs) direct diverse cellular and developmental responses by stimulating a relatively small number of overlapping signaling pathways. Specificity may be determined by RTK expression patterns or by differential activation of individual signaling pathways. To address this issue we generated knock-in mice in which the extracellular domain of the mouse platelet-derived growth factor alpha receptor (PDGF␣R) is fused to the cytosolic domain of Drosophila Torso (␣ Tor ) or the mouse fibroblast growth factor receptor 1 (␣ FR ). ␣ Tor homozygous embryos exhibit significant rescue of neural crest and angiogenesis defects normally found in PDGF␣R-null embryos yet fail to rescue skeletal or extraembryonic defects. This phenotype was associated with the ability of ␣ Tor to stimulate the mitogen-activated protein (MAP) kinase pathway to near wildtype levels but failure to completely activate other pathways, such as phosphatidylinositol (PI) 3-kinase. The ␣ FR chimeric receptor fails to rescue any aspect of the PDGF␣R-null phenotype. Instead, ␣ FR expression leads to a gain-of-function phenotype highlighted by ectopic bone development. The ␣ FR phenotype was associated with a failure to limit MAP kinase signaling and to engage significant PI3-kinase response. These results suggest that precise regulation of divergent downstream signaling pathways is critical for specification of RTK function.Evolutionary conservation of a gene requires that it contributes specific functions that enhance the fitness of the organism in which it is expressed. Receptor tyrosine kinases (RTKs) represent a large family of genes that have been conserved due to their ability to control multiple fundamental cellular processes. Upon binding to specific extracellular ligands, activated RTKs regulate cell proliferation, survival, migration, differentiation, and metabolism. The conservation of over 50 RTKs in mammals suggests that each executes critical specific cellular functions. How functional specificity is generated remains a matter of controversy. A vast amount of research has focused on understanding the molecular mechanisms that underlie the ability of these receptors to mediate diverse cellular functions. Surprisingly, studies have shown that even divergent RTKs activate highly redundant array of downstream signaling proteins, such as mitogen-activated protein (MAP) kinases, Src family kinases, phospholipase C (PLC)-␥, and phosphatidylinositol (PI) 3-kinase. Several models have been proposed to explain how RTKs achieve functional specificity at the cellular level while at the molecular level appearing to activate redundant signaling pathways (12,27).One model proposes that specific responses to RTKs are a result of differential activation of downstream pathways or biochemical differences in RTK signaling. A great deal of evidence based on experiments performed in cultured cell lines to assess individual contributions of downstream signaling pathways supports this idea (14,26). Differences in the strength and/or duration ...

A central issue in signal transduction is the physiological contribution of different growth factor-initiated signaling pathways. We have generated knockin mice harboring mutations in the PDGFalpha receptor (PDGFalphaR) that selectively eliminate its capacity to activate PI3 kinase (alpha(PI3K)) or Src family kinases (alpha(Src)). The alpha(PI3K) mutation leads to neonatal lethality due to impaired signaling in many cell types, but the alpha(Src) mutation only affects oligodendrocyte development. A third knockin line containing mutations that eliminate multiple docking sites does not increase the severity of the alpha(PI3K) mutation. However, embryos with mutations in the PI3K binding sites of both PDGFRs (alpha and beta) recapitulate the PDGFalphaR null phenotype. Our results indicate that PI3K has a predominant role in PDGFalphaR signaling in vivo and that RTK-activated signaling pathways execute both specific and overlapping functions during mammalian development.

Eph receptor tyrosine kinases and their corresponding surface-bound ligands, the ephrins, provide cues to the migration of cells and growth cones during embryonic development. Here we show that ephrin-A5, which is attached to the outer leaflet of the plasma membrane by a glycosyl-phosphatidylinositol-anchor, induces compartmentalized signaling within a caveolae-like membrane microdomain when bound to the extracellular domain of its cognate Eph receptor. The physiological response induced by this signaling event is concomitant with a change in the cellular architecture and adhesion of the ephrin-A5-expressing cells and requires the activity of the Fyn protein tyrosine kinase. This study stresses the relevance of bidirectional signaling involving the ephrins and Eph receptors during brain development.

Src family kinases (SFKs) are crucial for signaling through a variety of cell surface receptors, including integrins. There is evidence that integrin activation induces focal adhesion kinase (FAK) autophosphorylation at Y397 and that Src binds to and is activated by FAK to carry out subsequent phosphorylation events. However, it has also been suggested that Src functions as a scaffolding molecule through its SH2 and SH3 domains and that its kinase activity is not necessary. To examine the role of SFKs in integrin signaling, we have expressed various Src molecules in fibroblasts lacking other SFKs. In cells plated on fibronectin, FAK could indeed autophosphorylate at Y397 independently of Src but with lower efficiency than when Src was present. This step was promoted by kinase-inactive Src, but Src kinase activity was required for full rescue. Src kinase activity was also required for phosphorylation of additional sites on FAK and for other integrin-directed functions, including cell migration and spreading on fibronectin. In contrast, Src mutations in the SH2 or SH3 domain greatly reduced binding to FAK, Cas, and paxillin but had little effect on tyrosine phosphorylation or biological assays. Furthermore, our indirect evidence indicates that Src kinase activity does not need to be regulated to promote cell migration and FAK phosphorylation. Although Src clearly plays important roles in integrin signaling, it was not concentrated in focal adhesions. These results indicate that the primary role of Src in integrin signaling is as a kinase. Indirect models for Src function are proposed.

PDGF induces distinct waves of PI 3-kinase activity. The second wave is required for PDGF-dependent DNA synthesis, whereas the initial wave is not. One of the ways in which cells use PI 3-kinase to mediate distinct cellular responses seems to be by regulating when its products accumulate.

Gene targeting studies have indicated that the two receptors for PDGF, alpha and beta, direct unique functions during development. Distinct ligand affinities, patterns of gene expression, and/or mechanisms of signal relay may account for functional specificity of the two PDGF receptor isoforms. To distinguish between these factors, we have created two complementary lines of knockin mice in which the intracellular signaling domains of one PDGFR have been removed and replaced by those of the other PDGFR. While both lines demonstrated substantial rescue of normal development, substitution of the PDGFbetaR signaling domains with those of the PDGFalphaR resulted in varying degrees of vascular disease. This observation provides a framework for discussing the evolution of receptor tyrosine kinase functional specificity.

Because the protein-tyrosine phosphatase (PTP) Syp associates with the tyrosine-phosphorylated platelet-derived growth factor ␤ receptor (␤PDGFR), the ␤PDGFR is a likely Syp substrate. We tested this hypothesis by determining whether recombinant Syp (rSyp) and a control PTP, recombinant PTP1B (rPTP1B), were able to dephosphorylate the ␤PDGFR. The ␤PDGFR was phosphorylated at multiple tyrosine residues in an in vitro kinase assay and then incubated with increasing concentrations of rSyp or rPTP1B. While the receptor was nearly completely dephosphorylated by high concentrations of rPTP1B, receptor dephosphorylation by rSyp plateaued at approximately 50%. Two-dimensional phosphopeptide maps of the ␤PDGFR demonstrated that rSyp displayed a clear preference for certain receptor phosphorylation sites; the most efficiently dephosphorylated sites were phosphotyrosines (Tyr(P))-771 and -751, followed by Tyr(P)-740, while Tyr(P)-1021 and Tyr(P)-1009 were very poor substrates. In contrast, rPTP1B displayed no selectivity for the various ␤PDGFR tyrosine phosphorylation sites and dephosphorylated all of them with comparable efficiency. A Syp construct that lacked the SH2 domains was still able to discriminate between the various receptor phosphorylation sites, although less effectively than full-length Syp.These in vitro studies predicted that Syp can dephosphorylate the receptor in vivo. Indeed, we found that a ␤PDGFR mutant (F1009) that associates poorly with Syp, had a much slower in vivo rate of receptor dephosphorylation than the wild type receptor. In addition, the GTPase-activating protein of Ras (GAP) and phosphatidylinositol 3-kinase were less stably associated with the wild type ␤PDGFR than with the F1009 receptor. These findings are consistent with the in vitro experiments showing that Syp prefers to dephosphorylate sites on the ␤PDGFR, that are important for binding phosphatidylinositol 3-kinase (Tyr(P)-740 and Tyr(P)-751) and GAP (Tyr(P)-771). These studies reveal that Syp is a substrate-selective PTP and that both the catalytic domain and the SH2 domains contribute to Syp's ability to choose substrates. Furthermore, it appears that Syp plays a role in PDGF-dependent intracellular signal relay by selectively dephosphorylating the ␤PDGFR and thereby regulating the binding of a distinct group of receptor-associated signal relay enzymes.Tyrosine phosphorylation modulates the activity of many enzymes involved in regulation of a wide variety of cellular events and is a reversible process governed by the opposing effects of protein-tyrosine kinases and protein-tyrosine phosphatases (PTPs) 1 (1-3). In addition to a conserved catalytic domain, most PTPs also contain highly variable noncatalytic sequences, which can regulate the activity and/or subcellular location of the PTP (1, 3). The initially purified PTPs failed to show a high degree of substrate specificity (4), and this led to the hypothesis that subcellular targeting is the primary determinant in selection of a substrate. More recent evidence indicates that an i...