sign in sign up
<<
Home , ErbB, GRB10, Growth factor receptor, Janus kinase, Tyrosine kinase, Tyrosine phosphorylation

(1998) 16, 1343 ± 1352  1998 Stockton Press All rights reserved 0950 ± 9232/98 $12.00 http://www.stockton-press.co.uk/onc -activated pathways which lead to oncogenesis

Amy C Porter and Richard R Vaillancourt

Department of Pharmacology and Toxicology, College of Pharmacy, The University of Arizona, Tucson, Arizona 85721, USA

Oncogenesis is a complicated process involving signal proto- disrupt the regulated activity of the transduction pathways that mediate many di€erent normal or wild-type product and the activity of physiological events. Typically, oncogenes cause unregu- the mutated confers a selective advantage for lated and this phenotype has been attributed infected cells to survive over their non-infected to the growth-stimulating activity of oncogenes such as counterparts. As we identify and characterize more ras and src. In recent years, much research e€ort has proto-oncogenes and the they encode, we are focused on proteins that function downstream of Ras, gaining a better understanding of the mechanisms that leading to the identi®cation of the Ras/Raf/MAPK the wild-type proteins utilize to govern normal cellular pathway, because activation of this pathway leads to physiology and the mechanisms that the oncogenic cellular proliferation. Activated receptor tyrosine proteins utilize to ensure their evolutionary success. (RTKs) also utilize this pathway to mediate their growth- However, it is important to remember that onco- stimulating e€ects. However, RTKs activate many other proteins do not function alone in the cell. In other signaling proteins that are not involved in the cellular words, the cellular signaling pathways that are proliferation process, per se and we are learning that activated or inactivated by an onco-protein are equally these pathways also contribute to the oncogenic process. as important, as the onco-protein itself, for the In fact, RTKs and many of the proteins involved in oncogenic process to occur. RTK-dependent signal transduction can also function as Historically, the morphological transformation of oncogenes. For example, the catalytic subunit of the mouse ®broblast cell line NIH3T3 has been used as phosphoinositide 3-kinase (PI3-K) was recently identi®ed a screen for oncogenes. DNA derived from human as an oncogenic protein. The scope of pathways that are lung and colon tumors was transfected into NIH3T3 activated by oncogenic RTKs is expanding. Thus, not cells which led to the identi®cation of ras as a human only do RTKs activate Ras-dependent pathways that oncogene (Cooper, 1982; Weinberg, 1981). Further drive proliferation, RTKs activate PI3-K-dependent characterization of the ras oncogene determined that pathways which also contribute to the oncogenic point mutations resulted in the activation of Ras, even mechanism. PI3-K can initiate changes in gene tran- though the signal transduction pathway that involved scription, cytoskeletal changes through b-catenin, Ras had not been determined at the time (Reddy et al., changes in cell motility through the tumor suppressor, 1982; Tabin et al., 1982; Yuasa et al., 1983). It was adenomatous polyposis coli (APC), and over a decade later that Ras was identi®ed as an of BAD, a protein involved in apoptotic and anti- activator of Raf (Vojtek et al., 1993). These earlier apoptotic signaling. There is also cross-talk between studies demonstrated that oncogenes are the mutated RTKs and the oncostatin receptor which may forms of cellular proteins. Even today, a common positively and negatively in¯uence oncogenesis. For this experimental approach to identify oncogenic is to review, we will focus on oncogenic RTKs and the isolate DNA from tumor cells and compare the network of cellular proteins that are activated by RTKs nucleotide sequence relative to that of the proto- because multiple, divergent pathways are responsible for oncogene. This approach has led to the identi®cation oncogenesis. of receptor tyrosine kinases (RTKs) as proto-onco- genes. Keywords: phosphoinositide 3-kinase; cytoskeleton; In order for an oncogene to be tumorigenic, there are many events that must occur within the cell. A change in the proto-oncogene nucleotide sequence occurs and the resulting gene product is a protein whose activity is not regulated like its wild-type Introduction counterpart. has commonly been associated with the unregulated activity of oncogenes Through the evolutionary process, retroviruses have since cellular transformation is the phenotype that is targeted genes whose products are essential for a cell's observed in tumors. However, as we have learned more normal physiological role within an organism. The about signal transduction pathways and oncogenes, it targeted genes, referred to as proto-oncogenes, encode is fair to state that cellular proliferation may be proteins that are key regulators for cell growth, necessary, but it is not sucient, for a transformed di€erentiation, survival or motility. Mutations in cellular phenotype. Due to the pleiotropic nature of RTKs and the signaling proteins they activate, we are ®nding that maintenance of cell survival and motogenic pathways also play an important role in oncogenic Correspondence: RR Vaillancourt signaling. Oncogenic pathways AC Porter and RR Vaillancourt 1344 In the past, Ras has received much attention due The Ras/Raf/MAPK pathway has received much to the high frequency of mutations that occur in the attention over the last few years because it plays a ras gene that is isolated from a variety of tumors. major role in cellular proliferation. The link between Because of this observation, scientists focused on RTKs and Ras that Grb2 provides clearly demon- Ras-dependent pathways and found that many strates the importance of this pathway in RTK RTKs produce a mitogenic signal through Ras. signaling. However, we are learning more about the However, it has become apparent that tumors have network of signaling molecules that are activated by properties that make them unique and that these RTKs. As a consequence, the physiological events that properties are mediated by Ras-independent path- these pathways regulate also contribute to the ways. As an example, RTKs activate PI3-K whose oncogenic process. Accordingly, we need to character- lipid products function as second messengers to ize these pathways to gain a better understanding of activate Akt which phosphorylates BAD, producing the mechanisms that oncogenic RTKs utilize an anti-apoptotic signal for the cell. As a (Figure 1). consequence, RTK activation of Ras-dependent and PI3-K-dependent pathways may contribute equally to produce a proliferative and an anti-apoptotic RET response, respectively. As a result of activating these two divergent pathways, mutated cells are Ret is a proto-type RTK oncogene because numerous, allowed to selectively proliferate and those cells are di€erent mutations and gene rearrangements of the ret not removed by normal apoptotic mechanisms. proto-oncogene have been isolated from tumor cells. Therefore, multiple signaling pathways may be For this reason, we will discuss Ret to illustrate the necessary for cellular transformation and no one types of mutations that can occur in RTKs. The c-ret pathway, alone, may be sucient to produce an proto-oncogene encodes an RTK that also contains an oncogenic response. In the future, our knowledge of extracellular cadherin-like domain and an interrupted how these pathways are integrated may increase our domain, in addition to the aforemen- chances of developing successful therapeutic strate- tioned features of RTKs. The unglycosylated receptor gies to treat malignant , and possibly, allow is 120 kDa while neuroblastoma cells express a 150 and us to develop strategies to prevent them from 170 kDa form of the receptor (Takahashi et al., 1991, occurring. This review will focus on the mechanisms 1997). Ret expression occurs predominantly in neural and pathways that cause RTKs and their crest-derived cells. The biological function of Ret is not proteins to function as oncogenic proteins. well de®ned although expression studies suggest that Ret is involved in , proliferation, and di€erentiation (Pachnis et al., 1993; Tsuzuki et al., Receptor tyrosine kinases 1995). Ret knock-out mice die early during neonatal development and show severe renal abnormalities, Clearly, receptor tyrosine kinases can function as demonstrating the importance of the Ret protein for oncogenes. Over the past few years, numerous embryonic development (Schuchardt et al., 1994). excellent reviews have appeared on RTKs and the Missense, nonsense, and gene rearrangement of the `classical' SH2 (Src Homology 2) domain containing ret proto-oncogene are the types of mutations that we pathways they activate (Cohen et al., 1995; Ullrich and will discuss. When missense mutations occur, an amino Schlessinger, 1990; van der Geer et al., 1994). More acid substitution is introduced into the resulting gene recently, RTKs have been shown to utilize `novel' product which causes the protein to function in an mechanisms to generate signals (reviewed in Weiss et unregulated manner, typically producing constitutive al., 1997). As we learn more about RTK-dependent activity. In contrast, nonsense mutations introduce a signaling pathways, only time will allow us to stop codon within the coding sequence of the gene di€erentiate between classical and novel signaling causing incomplete translation of the protein, resulting pathways for RTKs. For purposes of this review, we in a loss-of-function mutation. A gene rearrangement will focus on RTKs and the signaling pathways they occurs when pieces of DNA, typically from di€erent utilize that allows them to function as oncogenes. Since , fuse to produce hybrid genes and the there are many oncogenic RTKs, we will discuss a resulting protein activity is not regulated in a manner select few in this review to highlight oncogenic that is similar to that of the wild-type proteins. mechanisms that involve RTKs. Although frameshift mutations occur in the ret proto- RTKs are composed of an extracellular oncogene, they will not be discussed in this review. binding domain, a transmembrane domain, an Mutations in the c-ret proto-oncogene result in two intracellular tyrosine kinase domain, and additional classes of genetic diseases (reviewed in Edery et al., sequences that function as regulatory 1997; Mak and Ponder, 1996). The developmental domains. Ligand binding induces receptor dimeriza- disorder, Hirschsprung's disease, is attributed to loss- tion and in a trans fashion, that of-function mutations, while activating point mutations is, one subunit of the dimer phosphorylates the are associated with autosomally dominant inherited opposing subunit. Phosphorylated function syndromes of three types, multiple endocrine to recruit intracellular signaling proteins, many via neoplasia (MEN) type 2A, MEN type 2B, and familial their SH2 domains. In addition, SH2-containing medullary thyroid carcinoma (FMTC). The extracel- proteins, such as Grb2, function as adaptor proteins lular domain of Ret is a hot spot for mutations. to recruit multiple proteins to RTKs, thus increasing Individuals with the MEN 2A and FMTC syndromes the repertoire of signaling proteins that are activated have mutations in a cluster of ®ve codons in by RTKs. the extracellular domain of ret, (C609?R/W/Y), Oncogenic pathways AC Porter and RR Vaillancourt 1345 (C611?Y/W), (C618?G/F/S/R/Y), (C620?G/F/S/R/ In addition to speci®c point mutations, the ret Y), (C634?G/F/S/R/Y/W) (Donis-Keller et al., 1993; proto-oncogene undergoes gene rearrangement. This Mulligan et al., 1993). The net result of each of these phenomena was observed when DNA from human T missense mutations is ligand-independent activation of cell lymphoma was transfected into NIH3T3 cells. Ret tyrosine kinase activity. Additional missense DNA from the 3' end of the ret gene fused to the 5' mutations occur at codons 768 and 804 which are end of unrelated DNA sequence, hence the name RET localized to the kinase domain and the kinase insert REarranged during Transfection, (Takahashi et al., domain of the receptor. These mutations have been 1985). Although these DNA rearrangements were due identi®ed from a few FMTC families demonstrating an to the experimental method, this ®nding suggested that oncogenic consequence of these mutations (Bolino et the ret gene has a tendency to undergo DNA al., 1995; Eng et al., 1995). However, the signaling rearrangement. In fact, rearrangement of the 3' pathways that are activated by these mutant receptors sequence of the ret proto-oncogene with the 5' have not been characterized. The MEN 2B syndrome, sequences of H4/D10S170 (Bongarzone et al., 1996; although less common than MEN 2A, is associated Grieco et al., 1990), the type Ia regulatory subunit of with a M?T mutation at codon 918 of the second part cyclic AMP-dependent (Bongarzone et of the tyrosine kinase domain (Carlson et al., 1994; al., 1993), and the ELE1 gene (Santoro et al., 1994a) Hofstra et al., 1994). Similarly, the signaling pathway have been detected from DNA extracts from papillary that mediates the oncogenic response for this receptor thyroid carcinomas (PTC). The resulting hybrid genes is not known, although constitutive activation of a (Ret/ptc1, Ret/ptc2, and Ret/ptc3) consist of the Ret signaling pathway probably mediates the oncogenic tyrosine kinase domain fused with genes that are activity of this receptor. capable of intracellular dimerization, thus producing

Figure 1 Multiple pathways contribute to cellular transformation. Oncogenic receptor tyrosine kinases are pleiotropic, activating numerous intracellular signaling pathways. Grb2 and Shc (not shown) associate with activated receptors and participate in the activation of Ras, a small molecular weight GTP binding protein that functions upstream of sequential protein kinases of the MAPK pathway. Activation of MAPK is necessary for cellular proliferation, di€erentiation, and cell scattering. In addition to activation of MAPK, oncogenic Ras stimulates the release of growth factors such as IGF-1. Acting in an autocrine fashion, IGF-1 stimulates its tyrosine kinase receptor which recruits and activates PI3-K, via the receptor substrate 1 (IRS-1), to produce the phospholipids PI(3,4,5)P3 and PI(3,4)P2. When bound to the PH domain of Akt, these phospholipids activate Akt to phosphorylate BAD and GSK3. When phosphorylated, BAD is an anti-apoptotic signal in cells. Akt phosphorylation of GSK3 results in inactivation of GSK3. This prevents the association of GSK3, APC and free b-catenin which would normally result in degradation of b-catenin. Instead, b-catenin can go on to a€ect gene leading to proliferation and/or activation of anti-apoptotic pathways and APC can go on to play a role in cell motility. Oncogenic RTKs can also phosphorylate b-catenin, involved in cell-cell adhesion and the cytoskeleton, resulting in loss of the interaction of b-catenin with the actin cytoskeleton which could contribute to cell scattering. Thus, oncogenic RTKs are capable of using multiple, divergent mechanisms to undermine the normal physiological functions of cells Oncogenic pathways AC Porter and RR Vaillancourt 1346 hybrid proteins with a constitutively active Ret kinase. carcinoma cells (Weidner et al., 1990), is a potent The Ret/ptc1 gene product, unlike the wild-type angiogenic agent (Bussolino et al., 1992), and is receptor, is localized to the and constitu- tumorigenic (Rong et al., 1992). HGF, the ligand for tively phosphorylated on tyrosine (Ishizaka et al., the HGF-R, is also known as scatter factor because it 1992). stimulates cellular motility and functions as a Only recently has the Ret ligand been identi®ed as morphoregulatory agent. HGF is a heterodimer glial cell-derived neurotrophic factor (GDNF (Durbec composed of a 69 and a 34 kDa subunit which binds et al., 1996; Trupp et al., 1996)). Prior to this to the 190 kDa HGF/scatter factor (SF) receptor. discovery, a chimera of the extracellular ligand HGF-R is a pleiotropic, epithelial cell receptor, capable binding domain of the EGF-R and the intracellular of stimulating mitogenesis, motility, invasive growth domain of Ret was used to study Ret signaling and inducing branching tubulogenesis. pathways. Stimulation of cells expressing the EGF-R/ Following stimulation with HGF, many proteins Ret chimera resulted in phosphorylation of , a associate with the receptor, become phosphorylated cytoskeletal protein, and cellular transformation and/or become activated by the HGF-R. These (Romano et al., 1994; Santoro et al., 1994b). In proteins include PI3-K (Royal and Park, 1995), PLCg addition, stimulation of the EGF-R/Ret chimera in (Zhu et al., 1994), Grb2 (Zhu et al., 1994), Ras SK-N-MC neuroepithelioma cells leads to cell scatter- (Graziani et al., 1993), focal adhesion kinase ing, growth inhibition, and loss of anchorage- (Matsumoto et al., 1994), b-catenin, plakoglobin independent growth (van Puijenbroek et al., 1997). (Shibamoto et al., 1994) and Shc (Pelicci et al., 1995). Characterization of the -activated protein Many of these proteins are activated by mitogenic kinase (MAPK) pathway from these cells indicated receptors, such as the PDGF-R (Valius that ERK2 (extracellular signal-regulated kinase 2) was and Kazlauskas, 1993). Like many RTKs, tyrosine activated in a sustained fashion, suggesting that cell phosphorylation within the kinase domain is required scattering correlates with sustained ERK2 activation. for catalytic activity (Longati et al., 1994). In contrast, a mitogenic response from EGF in PC12 Little is known about the signaling pathways cells correlates with transient ERK activation (Heasley involving cell dissociation, motility, and morphogen- and Johnson, 1992). esis that are mediated by the HGF-R. The HGF-R Studies with the Ret/ptc2 oncogene have identi®ed recruits a number of e€ector proteins including Grb2, phospholipase Cg (PLCg), Grb10 and Shc as signaling Shc, the p85 regulatory subunit of PI3-K, PLCg and proteins with SH2 domains that associate with the phosphatase SHP2 by association with pTyr1356 phosphorylated tyrosines of Ret/ptc2. Phosphotyro- (Zhu et al., 1994). Grb2, Shc and PI3-K associate with sine (pTyr)539, which is Tyr761 on the Ret proto- pTyr489 (equivalent to Tyr1356 of Met or HGF-R) of oncogene, and pTyr429 are the docking sites for PLCg the Tpr-Met oncoprotein (Fixman et al., 1995, 1996). and Grb10, respectively (Borrello et al., 1996), while Functionally, Tyr1356 is also necessary for transforma- pTyr586 associates with Shc (Arighi et al., 1997). tion of ®broblasts mediated by Tpr-Met. Given our Although the biological role of Ret is not well de®ned, current understanding of the speci®city of SH2 the interaction of PLCg and Grb10 is not required for domains, it seems counter intuitive to have so many mitogenic signaling of Ret/ptc2. proteins bind to the same phosphotyrosine. Enigma is unique in that it associates with Ret/ptc2 More recently, Nguyen et al. (1997) have shown that via unphosphorylated Tyr586 demonstrating that phosphorylation of Tyr1356 is necessary for recruit- unphosphorylated RTKs are capable of recruiting ment and phosphorylation of Gab1. However, associa- signaling proteins (Durick et al., 1996). Additional tion of Gab1 with the HGF-R is not direct, but data to support this observation came from experi- mediated primarily by Grb2, where Grb2 functions as ments where kinase-inactive Ret/ptc2 was found to an adaptor between the two proteins. Thus, Gab1 may interact with Enigma in the yeast two-hybrid system, function as a docking site for the p85 subunit of PI3-K, demonstrating that the kinase activity of Ret/ptc2 is PLCg, and SHP2 through Tyr1356. Of these signaling not necessary for this interaction. However, the kinase proteins, PI3-K activity is essential for HGF-R- activity of Ret/ptc2, as well as an interaction between mediated cell dissociation and subsequent scattering Ret/ptc2 and Enigma, is necessary for Ret/ptc2 (Royal and Park, 1995). mitogenic signaling. The signaling pathways that are There is con¯icting data regarding the role of regulated by Ret are beginning to emerge. Now that Tyr1356 in HGF-R signaling. Madin-Darby canine many of the mutations in the ret proto-oncogene have kidney (MDCK) cells expressing the Y1356F mutant been identi®ed, a careful analysis of the signaling HGF-R retain their ability to scatter but fail to pathways that are activated by the mutant proteins stimulate tubulogenesis of tubular epithelial cells may shed on new oncogenic pathways or (Nguyen et al., 1997). In contrast, MDCK cells reemphasize the importance of oncogenic pathways expressing a receptor chimera composed of the CSF- that have already been de®ned. 1R (extracellular domain)/Met (transmembrane and intracellular domain) and treated with CSF-1 fail to scatter, undergo invasive cell growth, and form Hepatocyte branching tubules (Pasdar et al., 1997). In addition, stimulation of the CSF-Met receptor inhibits cell-cell Invasive cell growth is a component of the oncogenic contact and junction formation, possibly by regulating process that contributes to the malignancy of tumor E-cadherin and plakoglobin (reviewed in Daniel and cells. Stimulation of the Reynolds, 1997). E-cadherin may play a role in receptor (HGF-R) is particularly relevant to oncogen- invasive cell growth mediated by the HGF-R since it esis because this receptor stimulates invasive growth of is frequently down-regulated in metastatic tumors. Oncogenic pathways AC Porter and RR Vaillancourt 1347 Receptor over-expression dimers also correlated with the loss of autopho- sphorylation of Tpr-Met and focus forming activity. One of the hallmarks of tumor cells is the over- Over-expression of p185erbB2 occurs in approxi- expression of growth factor receptors. The epidermal mately 33% of primary breast and ovarian tumors. A growth factor receptor (EGF-R) is not mutated nor point mutation (V664?E) in the putative transmem- does it undergo gene rearrangement. However, over- brane domain renders p185erbB2 constitutively active, expression of the EGF-R in many tumors, as well as leading to Shc phosphorylation and activation of association with the erbB-2 (Her-2/neu) oncogene and MAPK (Nowak et al., 1997). Analysis of the physical the c- proto-oncogene, likely contributes to the properties of this mutant receptor determined that it is oncogenic potential of the EGF-R as a stimulator of capable of forming dimers. Also, a chimeric p185erbB2 cell proliferation. In tumors, the EGF-R is frequently receptor containing the transmembrane domain of over-expressed which produces an autocrine/paracrine glycophorin A was capable of dimerization, suggesting loop such that the mitogenic potential of EGF is that amino acid sequences outside the transmembrane selectively enhanced in tumors as compared to normal domain are necessary for receptor dimerization (Burke cells. Similarly, the HGF-R or the Met protein is over- et al., 1997). Although the p185erbB2/glycophorin A expressed in tumors. chimera could form a dimer, it was unable to induce a When comparing normal epithelial cells to epithelial transformed phenotype in cells. Thus, dimerization tumor cells, the HGF-R is over-expressed, up to 100- may be necessary, but is not sucient for transforming fold in thyroid papillary carcinomas (Di Renzo et al., activity. These results suggest that constitutive or 1992; Prat et al., 1991). To investigate a possible unregulated kinase activity may be the driving force mechanism to explain this observation, Ivan et al. that determines the oncogenic potential of RTKs and (1997) expressed activated Ras under the control of an that dimerization may be an important feature of inducible promoter in the R18 human thyroid cell line regulated RTK activity. and found that HGF-R expression increased fourfold. In a paracrine fashion, the role of increased HGF-R expression would be to increase tumor sensitivity to HGF secreted from adjacent stromal cells. However, the authors found that this increased sensitivity to There is some indication that there is cross-talk HGF was not translated into a mitogenic response, as between the RTKs and the cytokine oncostatin M one would predict. In fact, the role of increased HGF- receptor (OSM-R) which may play a role in R expression may be to provide a motogenic signal to oncogenesis. Oncostatin M is a 28 kDa glycoprotein tumor cells, allowing for invasive cell growth (reviewed cytokine produced by activated T cells, macrophages in Jiang & Hiscox, 1997). and phorbol ester-treated U937 leukemia cells (Zarling et al., 1986). OSM appears to act through a two subunit , typical of dimerization receptors. The gp 130-signal transducing protein is a shared component of the Type I and Type II OSM-R. In order for a proto-oncoprotein to function as an The Type I receptor consists of gp 130 and the oncogene, a mutation must occur which changes the leukemia inhibitory factor receptor b (LIF-Rb) and regulated activity of the gene product. For RTKs, also binds leukemia inhibitory factor (LIF). The Type localization to the plasma membrane may serve two II receptor consists of gp 130 and the 180 kDa functions with respect to regulation of receptor b (OSM-Rb; gp 180) and does activity. One function would be to allow the receptor not bind LIF. It is the Type II receptor which appears to bind its respective growth factor, since the ligands to transduce most of the speci®c signals activated only do not di€use across the plasma membrane. A second by OSM and not by LIF (Thoma et al., 1994). The two function for localizing a RTK to the plasma membrane subunits of the receptor form a dimer upon ligand may be to prevent or minimize unregulated activation binding and are coupled to , of cytoplasmic proteins due to the intrinsic, basal although the receptor does not possess any endogenous tyrosine kinase activity of RTKs. Through mutations kinase activity (Ihle and Kerr, 1995). One of the many or the process of gene rearrangement, oncogenic RTKs e€ects of OSM is its ability to suppress growth of some bypass both of these functions. types of tumor cells (Liu et al., 1997a). However, OSM The receptor (NGF-R) under- is also able to stimulate growth of some types of goes a gene rearrangement with tropomyosin (Martin- cancer, such as Kaposi's sarcoma (Miles et al., 1992; Zanca et al., 1986), hence the name trk (tropomyosin Nair et al., 1992), while having no e€ect on other receptor kinase). The extracellular domain of the types, such as colon cancer (Horn et al., 1990). NGF-R is lost and tropomyosin is fused to the kinase The signal transduction pathways activated by OSM domain. Similarly, the HGF-R undergoes gene result in tyrosine phosphorylation of multiple proteins rearrangement when sequences from 1 (Auguste et al., 1997; Schieven et al., 1992), down- called tpr (translocated promoter region) fuse upstream regulation of EGF-R-mediated cellular proliferation with the HGF-R kinase domain (Cooper et al., 1984; (Spence et al., 1997) and a decrease in c- expression Park et al., 1986). The resulting Tpr-Met hybrid (Liu et al., 1997a). The OSM-R itself is also tyrosine protein lacks the extracellular and transmembrane phosphorylated on both the gp 130 and gp 180 domain of the HGF-R. It was determined from site- subunits (Auguste et al., 1997). However, this is not directed mutagenesis studies that the zipper autophosphorylation as cytokine receptors do not motif of Tpr was necessary for dimerization of Tpr- contain a catalytic domain (Ihle and Kerr, 1995; Met (Rodrigues and Park, 1993). The inability to form Taniguchi, 1995). Oncogenic pathways AC Porter and RR Vaillancourt 1348 Activation of the OSM-R leads to recruitment of the the catalytic subunit of PI3-K was cloned from the JAK/STAT (/signal transduction activa- genome of the avian sarcoma virus 16 (Chang et al., tion of transcription) pathway. JAK is a tyrosine 1997). The viral gene, termed v-p3k, has a 14 amino kinase and has been shown to activate Ras, leading to acid deletion at the amino terminus and is fused with Raf, MEK and ultimately MAPK activation in the retroviral gag gene product. In addition, v-p3k response to OSM stimulation of HeLa (Stancato et di€ers from c-p3k in that there are four missense al., 1997). In addition, OSM is also capable of mutations, outside the kinase domain. However, these activating PI3-K (Ihle and Kerr, 1995). MAPK mutations do not inhibit the catalytic activity of v-p3k. activation is responsible for phosphorylation of In fact, the focus forming activity of v-p3k is ten times the STAT proteins which then act as transcription more ecient than c-p3k. These data demonstrate that factors. However, it appears that STAT proteins PI3-K is an essential component of the cellular require both serine and tyrosine phosphorylation for machinery for cellular proliferation. activity (Wen et al., 1995). It is activation of the JAK/ Many of the RTKs and cytokine receptors that STAT pathway which appears to be responsible for the control cell growth also regulate PI3-K activity. growth inhibitory e€ects of OSM through changes in Mitogenic signaling from the bPDGF-R requires PI3- gene transcription (Stancato et al., 1997). It is not K activity in ®broblasts as determined by site-directed entirely clear what pathway is responsible for the mutagenesis of tyrosines that are known to associate growth-stimulatory e€ects of OSM but there is some with PI3-K (Valius and Kazlauskas, 1993). In MCF-7 evidence that the JNK (c-jun N-terminal kinase) human breast cancer cells, IGF-1-dependent mitogen- pathway is involved (Liu et al., 1997b). This divergent esis requires PI3-K and is independent of MAPK signaling by OSM may be the result of cell-type speci®c (Dufourny et al., 1997). PI3-K activity is also required expression of signal transduction proteins as well as for -2 (IL-2)-dependent control of lymphoid OSM receptor subunits. cell growth (Monfar et al., 1995). Recently, cross-talk between RTKs and cytokine The colony-stimulating factor receptor (CSF-1R) is receptors have been observed with the EGF-R and the encoded by the c-fms proto-oncogene and binds receptor, respectively. It has been macrophage-CSF. The CSF-1R is structurally related shown that growth hormone-stimulated activation of to the a and b platelet-derived growth factor receptors JAK is able to tyrosine phosphorylate the EGF-R (PDGF-R), and the c-ret proto-oncogene product, in leading to activation of the MAPK signaling pathway, that these receptors have a kinase domain that is independent of EGF-R kinase activity and EGF disconnected by a kinase insert domain. CSF-1 binding (Yamauchi et al., 1997). Phosphorylation of regulates macrophage proliferation, di€erentiation, the EGF-R results in an interaction between the and survival through the CSF-1R. Of the many signal receptor and Grb2 which initiates the MAPK transduction pathways that are activated by the CSF- signaling cascade, demonstrating an interaction be- 1R (reviewed in Hamilton, 1997), the PI3-K pathway tween cytokine receptors and RTKs. In those may be important in malignant cell growth. Evidence experiments, the IL-6 and LIF were unable for this comes from immunohistochemical studies of to stimulate EGF-R phosphorylation, while OSM was breast tumor specimens which demonstrated that 72% not tested (Yamauchi et al., 1997). However, co- of CSF-1R positive carcinomas had receptors that were administration of OSM is able to inhibit the phosphorylated at Tyr723, which is the PI3-K proliferation induced by EGF and basic ®broblast recognition site of this receptor (Flick et al., 1997). growth factor (bFGF) in some breast cancer cell lines (Liu et al., 1997a). This may be due to a dominant growth-inhibitory signal from the OSM-R, preventing Akt RTK-dependent signaling for cellular proliferation. It is interesting to speculate on the functional Akt is a proto-oncogene activated by phospholipids signi®cance of this cross-talk between cytokine produced PI3-K. The two SH2 domains of the 85 kDa receptors and RTKs. Both growth-stimulatory and subunit associate with speci®c phosphotyrosines on growth-inhibitory responses are seen with OSM. In RTKs which e€ectively brings the 110 kDa subunit in some cases, this cross-talk may serve as an inhibitory close proximity to the plasma membrane. There, it pathway to impede uncontrolled cellular proliferation. catalyzes the phosphorylation of membrane-bound However, it appears that during the oncogenic process, phosphoinositides at the D-3 hydroxyl of myo- constitutively activated RTKs may over-ride the inositol, producing the second messengers phosphati- growth-inhibitory mechanisms of oncostatins or con- dylinositol 3- [PIP], phosphatidylinositol 3,4-

versely, RTKs may utilize OSM-dependent pathways, bisphosphate [PI(3,4)P2], and phosphatidylinositol

such as the JAK/STAT pathway, as a synergistic 3,4,5-trisphosphate [PI(3,4,5)P3]. The target of the mechanism to promote unregulated cellular prolifera- latter two phospholipids is the proto-oncogene Akt tion. (Franke et al., 1997; Stokoe et al., 1997)]. Akt is also known at (PKB) and RACPK (Related to A and CProtein Kinase). Phosphoinositide 3-kinase v-Akt was cloned from the genome of the AKT8 virus (Bellacosa et al., 1991). Although the nucleotide Phosphoinositide 3-kinase (PI3-K) can now be added sequence for v-akt has ®ve G?A transitions, the to the list of proto-oncogenes. PI3-K is a heterodimer amino acid sequence is identical to that of c-Akt. The consisting of an 85 kDa regulatory subunit and a 110 v-akt oncogene is the product of a recombination kDa catalytic subunit that phosphorylates lipids between nucleotides from the viral gag gene, 785 (reviewed in Kapeller and Cantley, 1994). Recently, basepair (bp) downstream from the gag ATG codon, Oncogenic pathways AC Porter and RR Vaillancourt 1349 and 60 bp upstream of the c-akt ATG codon. An role casts b-catenins as structural components of the additional 3 bp were inserted between the gag and c- cell adhesion/actin cytoskeleton network (Daniel and Akt sequences producing a 63 nucleotide insert which, Reynolds, 1997). Another role casts free b-catenins as when translated, retains the c-Akt coding sequence. signaling molecules involved in gene transcription Thus, the v-Akt gene product is a hybrid protein (Korinek et al., 1997). Both of these functions may consisting of Gag/63 bp/c-Akt. This gene product was contribute to b-catenin involvement in the development immunoprecipitated from AKT8-transformed mink of cancer. The loss of cell/cell adhesion, mediated by b- cells, suggesting that v-akt is a cellular proliferation catenins, may be a factor in anchorage-independent cell signal (Bellacosa et al., 1991). As further evidence to growth and cancer metastasis. In addition, gene implicate v-akt in malignant cell growth, Akt2, a transcription mediated by b-catenins may turn on homologue of Akt, is over-expressed in some human genes required for cellular proliferation or genes with ovarian carcinoma cell lines and primary tumors an anti-apoptotic function (Peifer, 1997). RTKs may be (Cheng et al., 1992). directly or indirectly involved in both functions of b- c-Akt consists of 480 amino acids where amino acids catenins. 1 ± 147 compose the AH (Akt homology) domain and Cadherins are transmembrane glycoproteins which amino acids 1 ± 106 are referred to as the PH provide an anchor through cell-cell contact. Intracellu- (pleckstrin homology) domain. The AH and PH larly, cadherins interact directly with b-catenins which domains are involved in protein-protein interactions provide a link to the actin cytoskeleton through an (Datta et al., 1995) as well as the interaction with interaction with a-catenin. Loss of this link is thought phospholipids. The membrane-bound phospholipids, to be involved in the development of metastatic cancers

PI(3,4)P2 and PI(3,4,5)P3, bind to the pleckstrin (reviewed in Daniel and Reynolds, 1997). Activation of homology (PH) domain of Akt (Figure 1), thus RTKs, particularly by EGF and HGF/SF, has been recruiting Akt to the plasma membrane and activating shown to induce scattering of human cancer cells via its kinase activity (Franke et al., 1997; Klippel et al., tyrosine phosphorylation of b-catenin (Shibamoto et 1997; Stokoe et al., 1997). When Akt binds these al., 1994). Whether the phosphorylation is mediated by inositol , it becomes phosphorylated, by an the RTK, or through activation of a soluble protein unknown kinase, and Akt protein kinase activity tyrosine kinase by the RTK, is not yet clear. However, increases (Stokoe et al., 1997). Thus, Akt is the target a direct interaction has been shown between the EGF- of phospholipid second messengers. Randb-catenin in vitro and EGF induces immediate PDGF, EGF and bFGF are potent activators of b-catenin phosphorylation (Hoschuetzky et al., 1994). PI3-K and activation of their respective receptors It is possible that phosphorylation of b-catenin severs increases Akt kinase activity (Burgering and Co€er, its interaction with the actin cytoskeleton (Daniel and 1995; Franke et al., 1995). Thus, PI3-K functions to Reynolds, 1997). It is also possible that phosphoryla- link growth factor receptors with Akt. So then, what tion of b-catenin increases the pool of free b-catenin in does Akt phosphorylate? Recently, the apoptotic/anti- the cell. This may be the link between activation of apoptotic protein, BAD, has been identi®ed as an RTKs and cellular growth, that is characterized by the Akt substrate where phosphorylation of BAD is loss of contact inhibition, which is common in regulated by the growth factors PDGF and IGF-1 metastatic tumor formation. (Datta et al., 1997), as well as the cytokines, IL-2 In addition to its role in cell-cell interactions, free b- (Ahmed et al., 1997) and IL-3 (del Peso et al., 1997). catenin plays a role in cellular signal transduction that The signi®cance of these ®ndings is that when has only recently been identi®ed (Peifer, 1997). Free b- phosphorylated, BAD functions as an anti-apoptotic catenin is capable of forming a complex with DNA protein, promoting cell survival (Zha et al., 1996). binding proteins of the factor (Tcf) and Thus, growth factors and cytokines mediate prolif- lymphoid enhancer factor (Lef) families. Tcf and Lef erative as well as anti-apoptotic signals. At this time, interact with b-catenin and the subsequent regulation it has not been determined whether v-Akt phosphor- of gene transcription is thought to play an important ylates BAD. However, it is tempting to speculate that role in tumorigenesis by activating genes involved in BAD will be a substrate for v-Akt. As a universal proliferation and/or genes with anti-apoptotic func- mechanism, maintenance of cell survival by phos- tions. phorylation of BAD would serve to insure the The levels of free b-catenin, as well as its stability, transforming potential of oncogenes. can be increased in some forms of cancer, leading to Glycogen synthase kinase 3 (GSK3) is another unregulated gene transcription. Some colon cancer cell substrate for Akt phosphorylation. Serine phosphor- lines are marked by the lack of tumor suppressor, ylation of GSK3 results in inactivation of GSK3 adenomatous polyposis coli (APC), which leads to the (Sutherland et al., 1993) which may disrupt the constitutive activation of b-catenin-mediated signal degradation pathway for free b-catenin in the cell. transduction (Korinek et al., 1997). In addition, This result is signi®cant because excess b-catenin has deletions, truncations, and point mutations have been been shown to play a role in tumorigenesis due to its found in both APC and b-catenin in colon cancer cell transcriptional activity (Rubinfeld et al., 1997). lines (Morin et al., 1997). The mutations in APC disrupt interactions with b-catenin and lead to constitutive transcriptional activity of b-catenin. b-catenins Deletions or point mutations in b-catenin either remove or change serines to other amino RTKs may also transduce oncogenic signals through b- acids. Interestingly, these serines are potential phos- catenins (Figure 1). Evidence suggests that there are phorylation sites for GSK3 (Morin et al., 1997). As a two di€erent roles played by b-catenins in the cell. One functional consequence, loss of serines in b-catenin Oncogenic pathways AC Porter and RR Vaillancourt 1350 leads to increased transcriptional activity and insensi- of the PDGF B chain, is an oncogene. Inappropriate tivity to APC-mediated down-regulation of b-catenin. expression of v-sis e€ectively activates the PDGF-R to In melanoma cell lines, very high levels of free b- produce a persistent mitogenic signal in cells. catenin are often present. As in colon cancer cell lines, The inappropriate expression of a growth factor can point mutations resulting in the loss of critical serines function as an oncogenic mechanism to induce cell or truncations of the amino terminal region of b- growth or cell survival. Aside from promoting cell catenin have been detected. In addition, mutations and proliferation and di€erentiation, oncogenic Ras can loss of APC were found in melanoma cell lines. A also induce over-expression of the growth factor IGF-1 single point mutation, S37?F, appears to be (Dawson et al., 1995). One of the major functions of responsible for the stabilization of b-catenin and a IGF-1 is to function as a survival factor in serum, greater than tenfold increase in the half life of the protecting cells from apoptosis. The IGF-1R activates protein in the cell (Rubinfeld et al., 1997). The loss of PI3-K whose phospholipid products bind to the PH serines reduced phosphorylation of b-catenin, produ- domain of Akt, producing an anti-apoptotic signal. In cing higher levels of free b-catenin which is capable of the cascade of signals, Akt binds PI3-K-generated regulating transcriptional activity. phospholipids which activate Akt to phosphorylate RTKs regulate the activity of b-catenin by two BAD, which is an anti-apoptotic signal in an IL-3- mechanisms. First RTKs may be responsible for direct dependent hematopoietic cell line (Zha et al., 1996). phosphorylation of b-catenin which results in a loss of Therefore, activated Ras may function to e€ect contact with the actin cytoskeleton and ultimately a autocrine loops that produce growth factors, such as loss of cell/cell adhesion (Daniel and Reynolds, 1997). IGF-1, whose function is to maintain cell survival by This may contribute to cancer metastasis. Secondly, inhibiting apoptosis, thus allowing mutated cells to RTKs may regulate the transcriptional activity of b- proliferate (Figure 1). catenin via a PI3-K dependent pathway. RTKs activate PI3-K to produce phospholipid second messengers which are involved in activation of Concluding remarks Akt (Franke et al., 1997). Akt then phosphorylates and inhibits the activity of GSK3. GSK3b has been shown In conclusion, it is not only RTKs and their growth to phosphorylate APC (Rubinfeld et al., 1996) and b- factor ligands that function as oncogenes. Many of the catenin (Peifer et al., 1994). Inhibition of GSK3 proteins involved in RTK signal transduction are also activity prevents phosphorylation of APC. Formation capable of acting as oncogenes, independent of RTK of a complex between b-catenin, APC, and GSK3 is stimulation. RTKs have their e€ect on cellular dependent upon phosphorylation of APC (Rubinfeld et physiology via a number of di€erent mechanisms, al., 1996). It is the formation of this complex which is including increased proliferation, disruption of apop- responsible for degradation of excess free b-catenin. tosis, disruption of normal cytoskeletal interactions, Thus, activation of RTKs disrupts degradation of b- increased cell motility, and invasive cell growth. RTK catenin. This pool of free b-catenin is then able to and Ras-mediated production of growth factors can function as a transcriptional activator and turn on activate a number of these pathways due to autocrine genes involved in proliferation and/or inhibition of and paracrine loops. In addition, activation of PI3-K apoptosis. by oncogenic RTKs can lead to changes in gene Because the GSK3/APC/b-catenin complex is not transcription, cell motility, cytoskeletal changes, and formed due to inactivation of GSK3 in RTK anti-apoptotic signals; all of which could contribute to stimulated cells, APC may then be capable of playing oncogenesis. Moreover, there is evidence that there a role in cell migration and motility. Nathke et al., may be cross-talk between the RTKs and the (1996) have shown that APC localizes to the ends of oncostatin cytokine receptor which could in¯uence microtubules in actively migrating areas of epithelial proliferation. The picture of RTK signal transduction cell membranes. This localization and stimulation with which leads to oncogenesis is far from complete. HGF indicate that APC plays an active role in cell However, as we learn more about oncogenes and migration. In addition, APC is involved in normal cell signal transduction pathways, we are beginning to migration from the crypt in the intestine. Together appreciate that oncogenes utilize multiple pathways for these data indicate that APC is a key component of cell their selective growth advantages. De®ning these migration. Then it is possible that, in RTK stimulated pathways and mechanisms will ultimately lead to a cells in which GSK3 is inactivated by Akt phosphor- better understanding of the oncogenic process, which ylation, APC may be involved in aberrant migratory may provide more selective methods for the treatment activity associated with tumor progression. of cancers.

Autocrine/paracrine loops Acknowledgements In addition to receptor tyrosine kinases, growth factors The authors are supported, in part, by grants from the may also function as oncogenes. Retroviruses contain American Cancer Society (IRG 110T), the Southwest within their genome oncogenic forms of growth Environmental Health Sciences Center (ES 06694), and a factors. For example, although the PDGF receptors grant from the Arizona Disease Control Research Com- are not proto-oncogenes, v-sis, which is a homologue mission. Oncogenic pathways AC Porter and RR Vaillancourt 1351 References

Ahmed NN, Grimes HL, Bellacosa A, Chan TO and Tsichlis Durick K, Wu RY, Gill GN and Taylor SS. (1996). J. Biol. PN. (1997). Proc. Natl. Acad. Sci USA, 94, 3627 ± 3632. Chem., 271, 12691 ± 12694. Arighi E, Alberti L, Torriti F, Ghizzoni S, Rizzetti MG, Edery P, Eng C, Munnich A and Lyonnet S. (1997). Pelicci G, Pasini B, Bongarzone I, Piutti C, Pierotti MA Bioessays, 19, 389 ± 395. and Borrello MG. (1997). Oncogene, 14, 773 ± 782. Eng C, Smith DP, Mulligan LM, Healey CS, Zvelebil MJ, Auguste P, Guillet C, Fourcin M, Olivier C, Veziers J, Stonehouse TJ, Ponder MA, Jackson CE, Water®eld MD Pouplard-Barthelaix A and Gascan H. (1997). J. Biol. and Ponder BA. (1995). Oncogene, 10, 509 ± 513. Chem., 272, 15760 ± 15764. Fixman ED, Fournier TM, Kamikura DM, Naujokas MA Bellacosa A, Testa JR, Staal SP and Tsichlis PN. (1991). and Park M. (1996). J. Biol. Chem., 271, 13116 ± 13122. Science, 254, 274 ± 277. Fixman ED, Naujokas MA, Rodrigues GA, Moran MF and Bolino A, Schu€enecker I, Luo Y, Seri M, Silengo M, Tocco Park M. (1995). Oncogene, 10, 237 ± 249. T, Chabrier G, Houdent C, Murat A, Schlumberger M, Flick MB, Sapi E, Perrotta PL, Maher MG, Halaban R, Tourniaire J, Lenoir GM and Romeo G. (1995). Carter D and Kacinski BM. (1997). Oncogene, 14, 2553 ± Oncogene, 10, 2415 ± 2419. 2561. Bongarzone I, Fugazzola L, Vigneri P, Mariani L, Franke TF, Kaplan DR, Cantley LC and Toker A. (1997). Mondellini P, Pacini F, Basolo F, Pinchera A, Pilotti S Science, 275, 665 ± 668. and Pierotti MA. (1996). J. Clin. Endocrinol. Metab., 81, Franke TF, Yang SI, Chan TO, Datta K, Kazlauskas A, 2006 ± 2009. Morrison DK, Kaplan DR and Tsichlis PN. (1995). Cell, Bongarzone I, Monzini N, Borrello MG, Carcano C, 81, 727 ± 736. Ferraresi G, Arighi E, Mondellini P, Della Porta G and Graziani A, Gramaglia D, dalla Zonca P and Comoglio PM. Pierotti MA. (1993). Mol. Cell. Biol., 13, 358 ± 366. (1993). J. Biol. Chem., 268, 9165 ± 9168. Borrello MG, Alberti L, Arighi E, Bongarzone I, Battistini Grieco M, Santoro M, Berlingieri MT, Melillo RM, Donghi C, Bardelli A, Pasini B, Piutti C, Rizzetti MG, Mondellini R, Bongarzone I, Pierotti MA, Della Porta G, Fusco A P, Radice MT and Pierotti MA. (1996). Mol. Cell. Biol., and Vecchio G. (1990). Cell, 60, 557 ± 563. 16, 2151 ± 2163. Hamilton JA. (1997). J. Leukoc. Biol., 62, 145 ± 155. Burgering BM and Co€er PJ. (1995). Nature, 376, 599 ± 602. Heasley LE and Johnson GL. (1992). Mol. Biol. Cell, 3, 545 ± Burke CL, Lemmon MA, Coren BA, Engelman DM and 553. Stern DF. (1997). Oncogene, 14, 687 ± 696. Hofstra RM, Landsvater RM, Ceccherini I, Stulp RP, Bussolino F, Di Renzo MF, Ziche M, Bocchietto E, Olivero Stelwagen T, Luo Y, Pasini B, Hoppener JW, van Amstel M, Naldini L, Gaudino G, Tamagnone L, Co€er A and HK, Romeo G, Lips CJM and Buys CHCM. (1994). Comoglio PM. (1992). J. Cell Biol., 119, 629 ± 641. Nature, 367, 375 ± 376. CarlsonKM,DouS,ChiD,ScavardaN,ToshimaK, Horn D, Fitzpatrick WC, Gompper PT, Ochs V, Bolton- JacksonCE,WellsJrSA,GoodfellowPJandDonis- Hansen M, Zarling J, Malik N, Todaro GJ and Linsley PS. Keller H. (1994). Proc. Natl. Acad. Sci USA, 91, 1579 ± (1990). Growth Factors, 2, 157 ± 165. 1583. Hoschuetzky H, Aberle H and Kemler R. (1994). J. Cell Chang HW, Aoki M, Fruman D, Auger KR, Bellacosa A, Biol., 127, 1375 ± 1380. Tsichlis PN, Cantley LC, Roberts TM and Vogt PK. Ihle JN and Kerr IM. (1995). Trends Genet., 11, 69 ± 74. (1997). Science, 276, 1848 ± 1850. Ishizaka Y, Shima H, Sugimura T and Nagao M. (1992). Cheng JQ, Godwin AK, Bellacosa A, Taguchi T, Franke TF, Oncogene, 7, 1441 ± 1444. Hamilton TC, Tsichlis PN and Testa JR. (1992). Proc. Ivan M, Bond JA, Prat M, Comoglio PM and Wynford- Natl. Acad. Sci USA, 89, 9267 ± 9271. Thomas D. (1997). Oncogene, 14, 2417 ± 2423. Cohen GB, Ren R and Baltimore D. (1995). Cell, 80, 237 ± Jiang WG and Hiscox S. (1997). Histol. and Histopath., 12, 248. 537 ± 555. CooperCS,ParkM,BlairDG,TainskyMA,HuebnerK, Klippel A, Kavanaugh WM, Pot D and Williams LT. (1997). Croce CM and Vande Woude GF. (1984). Nature, 311, Mol. Cell. Biol., 17, 338 ± 344. 29 ± 33. Korinek V, Barker N, Morin PJ, van Wichen D, de Weger R, Cooper GM. (1982). Science, 217, 801 ± 806. Kinzler KW, Vogelstein B and Clevers H. (1997). Science, Daniel JM and Reynolds AB. (1997). Bioessays, 19, 883 ± 891. 275, 1784 ± 1787. Datta K, Franke TF, Chan TO, Makris A, Yang SI, Kaplan Liu J, Spence MJ, Wallace PM, Forcier K, Hellstrom I and DR, Morrison DK, Golemis EA and Tsichlis PN. (1995). Vestal RE. (1997a). Cell Growth and Di€., 8, 667 ± 676. Mol. Cell. Biol., 15, 2304 ± 2310. Liu ZY, Ganju RK, Wang JF, Ona MA, Hatch WC, Zheng Datta SR, Dudek H, Tao X, Masters S, Fu H, Gotoh Y and T, Avraham S, Gill P and Groopman JE. (1997b). J. Clin. Greenberg ME. (1997). Cell, 91, 231 ± 241. Invest., 99, 1798 ± 1804. Dawson TP, Radulescu A and Wynford-Thomas D. (1995). Longati P, Bardelli A, Ponzetto C, Naldini L and Comoglio Cancer Res., 55, 915 ± 920. PM. (1994). Oncogene, 9, 49 ± 57. delPesoL,Gonza lez-Garcia M, Page C, Herrera R and Mak YF and Ponder BA. (1996). Curr. Opin. Genet. and NunÄ ez G. (1997). Science, 278, 687 ± 689. Dev., 6, 82 ± 86. Di Renzo MF, Olivero M, Ferro S, Prat M, Bongarzone I, Martin-Zanca D, Hughes SH and Barbacid M. (1986). Pilotti S, Bel®ore A, Costantino A, Vigneri R, Pierotti MA Nature, 319, 743 ± 748. et al. (1997). Oncogene, 7, 2549 ± 2553. Matsumoto K, Matsumoto K, Nakamura T and Kramer Donis-KellerH,DouS,ChiD,CarlsonKM,ToshimaK, RH. (1994). J. Biol. Chem., 269, 31807 ± 31813. Lairmore TC, Howe JR, Moley JF, Goodfellow P and Miles SA, Martinez-Maza O, Rezai A, Magpantay L, Wells Jr SA. (1993). Hum. Mol. Genet., 2, 851 ± 856. Kishimoto T, Nakamura S, Radka SF and Linsley PS. Dufourny B, Alblas J, van Tee€elen HAAM, van Schaik (1992). Science, 255, 1432 ± 1434. FMA, van der Burg B, Steenbergh PH and Sussenbach JS. Monfar M, Lemon KP, Grammer TC, Cheatham L, Chung (1997). J. Biol. Chem., 272, 31163 ± 31171. J, Vlahos CJ and Blenis J. (1995). Mol. Cell. Biol., 15, Durbec P, Marcos-Gutierrez CV, Kilkenny C, Grigoriou M, 326 ± 337. Wartiowaara K, Suvanto P, Smith D, Ponder B, Morin PJ, Sparks AB, Korinek V, Barker N, Clevers H, Costantini F, Saarma M, Sariola H and Pachnis V. Vogelstein B and Kinzler KW. (1997). Science, 275, 1787 ± (1996). Nature, 381, 789 ± 793. 1790. Oncogenic pathways AC Porter and RR Vaillancourt 1352 Mulligan LM, Kwok JB, Healey CS, Elsdon MJ, Eng C, Stancato LF, Sakatsume M, David M, Dent P, Dong F, Gardner E, Love DR, Mole SE, Moore JK, Papi L, Petricoin EF, Krolewski JJ, Silvennoinen O, Saharinen P, Ponder MA, Telenius H, Tunnacli€e A and Ponder BAJ. Pierce J, Marshall CJ, Sturgill T, Finbloom DS and Larner (1993). Nature, 363, 458 ± 460. AC. (1997). Mol. Cell. Biol., 17, 3833 ± 3840. Nair BC, DeVico AL, Nakamura S, Copeland TD, Chen Y, Stokoe D, Stephens LR, Copeland T, Ga€ney PR, Reese CB, Patel A, O'Niel T, Oroszlan S, Gallo RC and Sarngadhar- Painter GF, Holmes AB, McCormick F and Hawkins PT. an MG. (1992). Science, 255, 1430 ± 1432. (1997). Science, 277, 567 ± 570. Nathke IS, Adams CL, Polakis P, Sellin JH and Nelson WJ. Sutherland C, Leighton IA and Cohen P. (1993). Biochem. J., (1996). J. Cell. Biol., 134, 165 ± 179. 296, 15 ± 19. Nguyen L, Holgado-Madruga M, Maroun C, Fixman ED, Tabin CJ, Bradley SM, Bargmann CI, Weinberg RA, Kamikura D, Fournier T, Charest A, Tremblay ML, Papageorge AG, Scolnick EM, Dhar R, Lowy DR and Wong AJ and Park M. (1997). J. Biol. Chem., 272, 20811 ± Chang EH. (1982). Nature, 300, 143 ± 149. 20819. Takahashi A, Goldschmidt-Clermont PJ, Alnemri ES, Nowak F, Jacquemin-Sablon A and Pierre J. (1997). Exp. Fernandes-Alnemri T, Yoshizawa-Kumagaya K, Naka- Cell Res., 231, 251 ± 259. jima K, Sasada M, Poirier GG and Earnshaw WC. (1997). Pachnis V, Mankoo B and Costantini F. (1993). Develop- Exp. Cell Res., 231, 123 ± 131. ment, 119, 1005 ± 1017. Takahashi M, Buma Y and Taniguchi M. (1991). Oncogene, Park M, Dean M, Cooper CS, Schmidt M, O'Brien SJ, Blair 6, 297 ± 301. DG and Vande Woude GF. (1986). Cell, 45, 895 ± 904. Takahashi M, Ritz J and Cooper GM. (1985). Cell, 42, 581 ± Pasdar M, Li Z, Marreli M, Nguyen BT, Park M and Wong 588. K. (1997). Cell Growth and Di€., 8, 451 ± 462. Taniguchi T. (1995). Science, 268, 251 ± 255. Peifer M. (1997). Science, 275, 1752 ± 1753. Thoma B, Bird TA, Friend DJ, Gearing DP and Dower SK. Peifer M, Pai LM and Casey M. (1994). Developmental (1994). J. Biol. Chem., 269, 6215 ± 6222. Biology, 166, 543 ± 556. Trupp M, Arenas E, Fainzilber M, Nilsson AS, Sieber BA, Pelicci G, Giordano S, Zhen Z, Salcini AE, Lanfrancone L, Grigoriou M, Kilkenny C, Salazar-Grueso E, Pachnis V, Bardelli A, Panayotou G, Water®eld MD, Ponzetto C, Arumae U, Sariola H, Saarma M and Iba nÄ ez CF. (1996). Pelicci PG and Comoglio PM. (1995). Oncogene, 10, Nature, 381, 785 ± 788. 1631 ± 1638. Tsuzuki T, Takahashi M, Asai N, Iwashita T, Matsuyama M PratM,NarsimhanRP,CrepaldiT,NicotraMR,NataliPG and Asai J. (1995). Oncogene, 10, 191 ± 198. and Comoglio PM. (1991). Int. J. Cancer, 49, 323 ± 328. Ullrich A and Schlessinger J. (1990). Cell, 61, 203 ± 212. Reddy EP, Reynolds RK, Santos E and Barbacid M. (1982). Valius M and Kazlauskas A. (1993). Cell, 73, 321 ± 334. Nature, 300, 149 ± 152. van der Geer P, Hunter T and Lindberg RA. (1994). Annu. Rodrigues GA and Park M. (1993). Mol. Cell. Biol., 13, Rev. Cell Biol., 10, 251 ± 337. 6711 ± 6722. van Puijenbroek AA, van Weering DH, van den Brink CE, Romano A, Wong WT, Santoro M, Wirth PJ, Thorgeirsson Bos JL, van der Saag PT, de Laat SW and den Hertog J. SS and Di Fiore PP. (1994). Oncogene, 9, 2923 ± 2933. (1997). Oncogene, 14, 1147 ± 1157. Rong S, Bodescot M, Blair D, Dunn J, Nakamura T, Mizuno Vojtek AB, Hollenberg SM and Cooper JA. (1993). Cell, 74, K, Park M, Chan A, Aaronson S and Vande Woude GF. 205 ± 214. (1992). Mol. Cell. Biol., 12, 5152 ± 5158. Weidner KM, Behrens J, Vandekerckhove J and Birchmeier Royal I and Park M. (1995). J. Biol. Chem., 270, 27780 ± W. (1990). J. Cell Biol., 111, 2097 ± 2108. 27787. Weinberg RA. (1981). Biochim. Biophys. Acta, 651, 25 ± 35. Rubinfeld B, Albert I, Por®ri E, Fiol C, Munemitsu S and Weiss FU, Daub H and Ullrich A. (1997). Curr. Opin. Genet. Polakis P. (1996). Science, 272, 1023 ± 1026. and Dev., 7, 80 ± 86. Rubinfeld B, Robbins P, El-Gamil M, Albert I, Por®ri E and Wen Z, Zhong Z and Darnell Jr JE. (1995). Cell, 82, 241 ± Polakis P. (1997). Science, 275, 1790 ± 1792. 250. Santoro M, Dathan NA, Berlingieri MT, Bongarzone I, Yamauchi T, Ueki K, Tobe K, Tamemoto H, Sekine N, Paulin C, Grieco M, Pierotti MA, Vecchio G and Fusco A. Wada M, Honjo M, Takahashi M, Takahashi T, Hirai H, (1994a). Oncogene, 9, 509 ± 516. TushimaT,AkanumaY,FujitaT,KomuroI,YazakiY Santoro M, Wong WT, Aroca P, Santos E, Matoskova B, and Kadowaki T. (1997). Nature, 390, 91 ± 96. Grieco M, Fusco A and di Fiore PP. (1994b). Mol. Cell. Yuasa Y, Srivastava SK, Dunn CY, Rhim JS, Reddy EP and Biol., 14, 663 ± 675. Aaronson SA. (1983). Nature, 303, 775 ± 779. Schieven GL, Kallestad JC, Brown TJ, Ledbetter JA and Zarling JM, Shoyab M, Marquardt H, Hanson MB, Lioubin Linsley PS. (1992). J. Immunol., 149, 1676 ± 1682. MN and Todaro GJ. (1986). Proc. Natl. Acad. Sci USA, Schuchardt A, V, DA, Larsson-Blomberg L, Costantini F 83, 9739 ± 9743. and Pachnis V. (1994). Nature, 367, 380 ± 383. Zha J, Harada H, Yang E, Jockel J and Korsmeyer SJ. ShibamotoS,HayakawaM,TakeuchiK,HoriT,OkuN, (1996). Cell, 87, 619 ± 628. Miyazawa K, Kitamura N, Takeichi M and Ito F. (1994). Zhu H, Naujokas MA, Fixman ED, Torossian K and Park Cell Adhesion and Commun., 1, 295 ± 305. M. (1994). J. Biol. Chem., 269, 29943 ± 29948. Spence MJ, Vestal RE and Liu J. (1997). Cancer Res., 57, 2223 ± 2228.