Role of Src Expression and Activation in Human Cancer

Rosalyn B Irby1 and Timothy J Yeatman*,1

1Department of Surgery, H. Lee Mott Cancer Center and Research Institute, University of South Florida, 12902 Magnolia Drive, Tampa, Florida, FL 33612, USA

Since the original identi®cation of a transmissible agent demonstrated that the combination of EGFR and Src responsible for the development of tumors in chickens, expression in ®broblasts leads to synergistic levels of now known to be a retrovirus encoding the v-src gene, tumorigenicity when compared with expression of signi®cant progress has been made in de®ning the EGFR or Src alone (Maa et al., 1995; Mao et al., potential functions of its human homolog, SRC. The 1997). Observations such as these provide strong product of the human SRC gene, c-Src, is found to be evidence that Src does not act alone but may require over-expressed and highly activated in a wide variety of partnerships for full expression of activity. Another human cancers. The relationship between Src activation potential mechanism for Src activation may relate to and cancer progression appears to be signi®cant. More- reductions in the levels or activity of Csk, the enzyme over, Src may have an in¯uence on the development of known to phosphorylate the negative regulatory tyrosine the metastatic phenotype. This review discusses the data residue of c-Src (Tyr 530) (Masaki et al., 1999). Similarly, supporting a role for c-Src as a critical component of the enhanced activity of phosphatases known to dephos- signal transduction pathways that control cancer cell phorylate the regulatory Tyr 530 may cause Src development and growth, and provides the rationale for activation (Bjorge et al., 1996; Egan et al., 1999; Peng targeting Src in drug discovery eorts. Oncogene (2000) and Cartwright, 1995). Recently, a mutational basis for 19, 5636 ± 5642. Src activation was identi®ed in a small subset of advanced human colon cancers, suggesting a potential Keywords: Src; oncogene; tyrosine kinase; human genetic link between Src activation and cancer develop- cancer; metastasis ment, progression, and metastasis (Irby et al., 1997). The eect of Src over-expression and activation appears to be pleiotropic. Numerous Src substrates are Introduction phosphorylated in tumor cells harboring activated forms of Src, and, many of these Src substrates have The c-Src proto-oncogene has been strongly implicated been linked to processes inducing tumorigenicity and in the development, growth, progression, and metastasis metastasis (Brown and Cooper, 1996). Furthermore, of a number of human cancers including those of the these dramatic alterations in signal transduction likely colon, breast, pancreas, and brain. Our working in¯uence multiple downstream transcriptional events. hypothesis is that Src may play a central role in these These events include alterations in homotypic adhe- processes aecting cancer outcome. The evidence under- sion, angiogenesis, tumor cell invasivity, tumor growth, lying this hypothesis is largely based on the observation and apoptosis, all of which play a signi®cant role in the that both Src protein levels and, to a greater degree, Src development of the metastatic phenotype. The exten- protein kinase activity, are frequently elevated in human sive presence of activated Src in human cancers and its neoplastic tissues when compared to adjacent normal potential role in their development and progression, tissues. Moreover, these levels appear to increase with makes Src an appealing target for drug discovery the stage of disease. Similarly, increased Src protein eorts. kinase activity has been observed in numerous human cancer cell lines derived from these tumors. Although the precise mechanism for the high levels of Activated Src is common in human tumors Src activation in human cancer has not been clearly elucidated, several mechanisms have been proposed, The eects of elevated Src kinase activity have been suggesting that the process is multifactorial (Bolen et al., extensively studied both in vitro using a variety of 1987b; Ottenho-Kal et al., 1992; Mao et al., 1997). human neoplastic cell lines (Budde et al., 1994; Rosen et The speci®c activity of Src protein kinase may be al., 1986; Biscardi et al., 1998; Bjorge et al., 1996; Bolen increased by direct or indirect interaction with receptor et al. 1987a; Cartwright et al., 1989; Weber et al., 1992; tyrosine kinases, such as epidermal growth factor Lutz et al., 1998) and in vivo with murine models receptor (EGFR) (Tice et al., 1999; Luttrell et al., 1994; (Biscardi et al., 1998; Irby et al., 1999; Muthuswamy et Mao et al., 1997) platelet derived growth factor receptor al., 1994; Staley et al., 1997; Wiener et al., 1999). Using (PDGFR) (Courtneidge et al., 1993) ®broblast growth these systems, investigators have studied modes of Src factor receptor (FGFR) (LaVallee et al., 1998), colony activation, eects of Src on tumor initiation and simulating factor-1 receptor (CSF-1R) (Courtneidge et progression, and eects of tyrosine kinase and antisense al., 1993; Levitzki, 1996), HER2/neu (Luttrell et al., inhibitors on cell behavior. These studies illustrate the 1994), and hepatocyte growth factor receptor (c-Met) complex network of Src interacting proteins that impact (Rahimi et al., 1998). For example, it has been on numerous signal transduction pathways. Since the early 1980s, increased Src kinase activity has been reported in various, apparently unrelated *Correspondence: TJ Yeatman human cancers. Elevated Src protein levels have been Src activation in human cancer RB Irby and TJ Yeatman 5637 found in many cancers, although the protein level may premalignant tissues (Cartwright et al., 1994) and not accurately re¯ect the speci®c protein kinase adenomatous polyps (Cartwright et al., 1990; Pena et activity. For this reason, reliable kinase assays using al., 1995). Activity is apparently high in malignant exogenous substrates were developed to determine the polyps and in benign polyps that contain villous speci®c activity of the protein. Results from numerous change or severe dysplasia, which are at greatest risk studies re¯ect the increase of Src speci®c activity in for developing cancer. Src activity was also found to be human tumors and in cell lines derived from those elevated in mildly dysplastic epithelia (6 ± 10-fold) when tumors (Cartwright et al., 1989, 1990; Muthuswamy et compared directly with adjacent non-dysplastic epithe- al., 1994; Budde et al., 1994; Jacobs and Rubsamen, lia in ulcerative colitis, increasing further in severely 1983; Mao et al., 1997; Masaki et al., 1998, 2000; dysplastic tissue which is at greatest risk of developing Muthuswamy and Muller, 1994; Rosen et al., 1986; cancer (Cartwright et al., 1994). Verbeek et al., 1996). A role for Src in tumor progression is poignantly illustrated by observations that Src activity increases with the progression of colon tumors, being higher in Breast cancer primary tumors than in polyps and higher yet in Src kinase activity, from 4 ± 20-fold higher than normal metastatic liver lesions (Talamonti et al., 1991). The tissues, has been found in human mammary carcino- trend is re¯ected in six paired samples of synchronous mas (Egan et al., 1999; Jacobs and Rubsamen, 1983; primary and metastatic lesions from the same patient. Muthuswamy and Muller, 1994; Muthuswamy et al., While Src protein levels vary widely between patients, the 1994; Ottenho-Kal et al., 1992; Rosen et al., 1986; activity level increases in liver metastases over that of Verbeek et al., 1996). Similarly, cell lines derived from synchronous primary tumors ± several fold greater than these tumors display up to a 30-fold elevation in Src the increases in Src protein levels. Further dierences activity. Recent data has suggested some of this exist in the levels of activated Src seen in colorectal activity may be attributable to the action of phospha- metastases to extrahepatic sites (Termuhlen et al., 1993). tases resulting in dephosphorylation of Tyr 530 (Egan In addition, colorectal metastases to abdomen, pelvis, et al., 1999). Rosen et al. (1986) reported an elevated and thorax showed a signi®cant increase in activity over Src kinase activity in breast tumors with relatively hepatic metastases. These data raise the question as to normal Src protein levels compared to normal tissue. whether the site of metastasis in¯uences the speci®c On the other hand, Verbeek et al. (1996) presented activity of Src or whether the speci®c activity of Src immunohistochemical evidence that a 4 ± 30-fold in- in¯uences the site of the metastatic lesion. crease in Src activity was accompanied by an increase The in¯uence of Src in colon cancer has also been in Src protein levels. Ottenho-Kal et al. (1992) found explored by examination of Src levels in colon tumors that 72/72 breast cancers showed an increase in of various dierentiation states. The reported results tyrosine kinase activity, 70% of which was attributed are interesting but not always intuitive. Weber et al. to c-Src or Src like kinases. (1992) reported that the highest levels of Src activity in Activated Src in mammary tumors has been well human colon tumors occurred in moderately to well studied in transgenic mice. Mice expressing viral dierentiated tumors, and the levels appear fairly polyoma middle T antigen under the control of the normal in poorly dierentiated colon tumors, a ®nding MMTV promoter produce highly metastatic mammary supported by various tumor cell lines. Park et al. tumors with elevated c-Src kinase activity (Guy et al., (1993) and Park and Cartwright (1995) reported an 1994). Muthuswamy et al. (1994) found that mice over- increase in Src as well as the Src family kinase, Yes, in expressing the neu oncogene also develop mammary both colonic cell lines and in primary colon cancers, tumors with 6 ± 8-fold higher c-Src kinase activity than but these studies showed a downregulation of Src adjacent normal tissue. These are two examples of kinases in fully dierentiated cells. These results, on the activated Src, one an example of a viral protein surface, are dicult to interpret knowing that poorly binding and activating Src, causing tumors, and the dierentiated tumors are biologically more aggressive other an example of a receptor tyrosine kinase causing than well-dierentiated tumors. The majority of color- Src activation and subsequent mammary tumors. Nude ectal liver metastases, however, are in fact well-to- mouse experiments (Biscardi et al., 1998) with breast moderately dierentiated and this prevalence may cancer cell lines over-expressing both c-Src and HER1 explain the observed results. (MDA-MB-468 and MDA-MB-231) compared to cell The role of Src in colon cancer has been recently lines over-expressing only c-Src (MCF7 and ZR-75-1) examined using the nude mouse model injected with revealed increased tumorigenicity in mice injected with various colon cancer cell lines (Irby et al., 1997; Staley et the MDA lines. This supports the hypothesis that Src al., 1997). Staley et al. (1997) transfected the colon activation can be mediated through HER1 interactions. cancer cell line HT 29 with an antisense vector designed to reduce c-Src expression but not c-Yes expression. When injected into nude mice, these cells formed slowly Colon cancer growing tumors with a proliferation rate retarded The c-src proto-oncogene has frequently been impli- further than the reduced proliferation rate of parental cated in the initiation and progression of human colon cells grown in culture. In contrast, cells stably cancer, and in resultant metastases (Bolen et al., 1987a; transfected with a sense vector showed no dierence in Cartwright et al., 1989, 1990, 1994; Talamonti et al., proliferation in culture nor in nude mice from wild-type 1991; Termuhlen et al., 1993; Weber et al., 1992). Src HT 29 cells. In a second study attempting to determine activity is increased 5 ± 8-fold in the majority of colon the phenotypic eects of wild-type c-Src over-expression tumors. This elevation of Src activity, and of the on human colon cancer cells, c-Src-transfected KM12C related Yes activity, is an early event, occurring even in colon cancer cells expressing up to 10-fold more c-Src

Oncogene Src activation in human cancer RB Irby and TJ Yeatman 5638 than wild-type cells were injected subcutaneously and from experiments utilizing v-Src protein or other intrasplenically into nude mice (Irby et al., 1997). Cells activated forms of Src to induce cellular transforma- with a higher level of c-Src expression formed more tion, tumorigenicity, tumor progression, and metas- rapidly growing tumors than did wild-type cells, but did tasis. The v-src gene, encoded by Rous sarcoma not form liver metastases. Interestingly, transfected and virus, was the ®rst transmissible gene found to wild-type cells grown in vitro showed similar prolifera- induce tumors. It was originally discovered by tion rates. These two studies indicate that, ®rst, the level Peyton Rous in 1911 (Rous, 1911) as a transmissible of Src and its activity alters the rate of tumor growth agent that would produce tumors in chickens. The proportionally in vivo, and, second, growth rates in vitro protein product of this gene, v-Src, is a tyrosine do not necessarily re¯ect the growth rates of cells in vivo. kinase with a cellular homolog, c-Src, which is This suggests that tumor cell growth is greatly found in normal cells and presumed to be a proto- in¯uenced by the microenvironment, and perhaps oncogene (Czernilofsky et al., 1980) (Stehelin et al., indicates the potential for Src activity in the tumor cell 1976). The structure of the two proteins is similar, to aect the expression of tumor promoting proteins by but the regulatory carboxyl terminus of v-Src is the host. These studies also demonstrate that over- truncated (resulting in activation) and numerous expression of wild-type c-Src alone, while clearly dierences in amino acid sequence throughout the aecting tumor growth in vivo, may be insucient to protein exist (enhancing the activation) (Jove and induce the metastatic phenotype. Hanafusa, 1987). While the two proteins perform the same function, the kinase activity of v-Src and its capacity to transform are much higher than c-Src. Pancreatic cancer Moreover, cellular transformation by v-Src has been Src activity has recently been studied in pancreatic reported to produce a 10-fold elevation in protein cancer. Lutz et al. (1998) examined pancreatic ductal phosphotyrosine over that produced by c-Src. carcinomas as well as pancreatic cell lines for elevated Despite relatively modest protein expression levels Src protein levels and kinase activity. Src protein levels in cells, v-Src invariably results in high levels of were elevated in 13/13 pancreatic cancers and in 14/17 cellular transformation as distinguished by altered pancreatic cell lines. Kinase activity was only detectable cellular morphology, cytoskeletal reorganization, pro- in cancer cells and this activity did not correlate with liferation under low serum conditions, and ancho- either the c-Src or Csk protein levels. Further studies rage independent growth. The end result of these using the tyrosine kinase inhibitor, herbimycin A, changes in vitro is the enhanced tumorigenicity in indicated that the Src activity was instrumental in vivo. The phosphorylation of numerous cellular promoting the growth of pancreatic tumor cells. One substrates on tyrosine residues by v-Src is likely method by which Src increases pancreatic tumor growth responsible for the eects on cellular transformation. was suggested by Flossmann-Kast et al. (1998). This These phosphorylation events are presumed to aect group found that Src causes an increase in the number of proteins which regulate cell growth and dierentia- insulin-like growth factor receptor (IGF-R) molecules tion and include, among others, transcription factors, per cell, thus enhancing IGF-dependent growth. In a members of signal transduction cascades, and growth another study based on a rat model of pancreatic factor receptors (Brown and Cooper, 1996; Thomas carcinogenesis (Visser et al., 1996), an increase in Src and Brugge, 1997). kinase activity correlated positively with the number of Because oncogenes may, in some cases, represent lesions present in the pancreas. This increase in activity fetal genes inappropriately expressed in adult tissues, was accompanied by a relocalization of c-Src protein to further evidence for Src's role as an oncogene is the nucleus, suggesting a role for Src in gene regulation. derived from observations suggesting that it is involved in dierentiation of fetal tissues. This is especially true in neural tissues (Bjelfman et al., 1990), where Miscellaneous cancers dierentiating vertebrate neurons express high levels Elevated Src protein levels and/or kinase activity has of c-Src with elevated kinase activity. Similar events been reported in lung (50 ± 80%) (Mazurenko et al., occur in tumors of neuronal origin (Bjelfman et al., 1992), neural (23/27 neuroblastomas, 3/3 retinoblasto- 1990). An overproduction of c-Src by itself, however, mas) (Bjelfman et al., 1990; Bolen et al., 1985), ovarian is, at best, minimally transforming, suggesting that (Budde et al., 1994; Wiener et al., 1999), esophageal increased Src activity may play more of a role in tumor (3 ± 4-fold increases in activity in Barrett's esophagus progression than in tumorigenesis (Biscardi et al., and sixfold elevations in adenocarcinomas) (Kumble et 1999a). al., 1997) and gastric cancers (Takeshima et al., 1991), Until recently, the genetic evidence for Src as a as well as melanoma (Bjorge et al., 1996) and Kaposi's human oncogene has been lacking. While numerous sarcoma (Munshi et al., 2000). Src family kinases Lck, mutant forms of avian v-Src have been identi®ed and Lyn, and Fgr have similarly been shown to be multiple chicken c-Src mutants have been con- activated during leukemic cell growth (Abts et al., structed, none have been previously observed in 1991) (Dai et al., 1998; Danhauser-Riedl et al., 1996; humans. The recent observation that an activating Roginskaya et al., 1999). mutation (Src 531) exists in a small subset of advanced human colon cancers with high Src activity lends credence to the hypothesis that Src may have Src as a potential oncogene oncogeneic potential (Irby et al., 1999). Src 531 was shown to induce cellular transformation, and aug- The majority of the data supporting a role for Src ment tumor growth and metastatic potential in as a potential oncogene in human cancer is derived ®broblasts.

Oncogene Src activation in human cancer RB Irby and TJ Yeatman 5639 Mode of Src activation PCR based assay where the mutant allele is the target of ampli®cation. Because colon tumors tend to have large While elevation of Src protein levels is apparently amounts of normal in®ltrating cells such as lymphocytes common to a large number of cancers, this elevation is and monocytes, it is also critical that tumor specimens often modest when compared to the fold-increases of be carefully microdissected prior to analysis. Src kinase activity that have been observed. These data Another apparently unique mode of activation was suggest the relative importance of Src activation in found in neruroblastoma cells, which display a 20 ± 40- human tumor development and progression. Several fold increase in Src kinase activity. The cells with possible explanations exist for the activation of Src activated Src harbored a Src protein containing an kinase in cancer. First, Src can be activated by receptor additional tyrosine in the amino-terminal portion. tyrosine kinases such as EGFR and HGF receptor Phosphorylation of this tyrosine increased the activity (Maa et al., 1995; Mao et al., 1997). These receptors of c-Src in these cell lines (Bolen et al., 1985). A have been known to be active in the progression of neuronal form of Src has been subsequently identi®ed cancer, and, in turn, may activate Src. It has been in developing neuroblasts and neurons. This neuronal suggested that the association between c-Src and these Src, pp60c-srcN, was also found in neuroblastomas (9 receptor tyrosine kinases is instrumental in malignant of 12) and retinoblastomas, but not in other childhood transformation (Luttrell et al., 1994). tumors (Bjelfman et al., 1990). These data suggest that A second possibility is that Src is activated post- perhaps other forms of Src are present, but as yet translationally. Several modes of post-translational undetected or indistinguishable from the known Src activation exist: insucient Csk activity, increased family kinases by current detection methods. activity of Src phosphatases, or interaction with viral A ®nal vehicle for Src activation may relate to the or cellular proteins. Csk activity inactivates the human absence of regulatory proteins. The recently identi®ed Src protein by phosphorylation of the negative drs gene is capable of suppressing v-Src transformation regulatory tyrosine 530 (avian Tyr 527). A decrease of cells (Shimakage et al., 2000). This protein is in the level of Csk protein or production of a defective reduced in a variety of human cancer cell lines, Csk would permit a more active form of Src. While including colon, bladder, and ovary. While most studies of pancreatic carcinomas show no correlation normal tissues were found to express the drs gene, between Src activity levels and Csk levels (Lutz et al., drs mRNA was minimally detected in all human colon 1998), in hepatocarcinomas, activated Src has been adenocarcinoma cells tested. In another study, Brum- shown to be correlated with decreased Csk levels mer et al. (1995) found an association between Src and (Masaki et al., 1999). Although this mode of activation the biliary glycoprotein CD66. CD66 is downregulated is anticipated, it does not account for all of the in many colon cancers, perhaps resulting in the activated Src kinase in human tumors. Studies of disregulation of c-Src. In some cases, Src apparently human breast cancer cell lines demonstrated cells with causes the destruction of an anti-tumorigenic protein. elevated Src kinase activity, with or without increased Abi is a protein that interacts with Abl, inhibiting the Src protein levels. A set of these breast cancer cells oncogenic activity of Abl. Src triggers the destruction displayed active phosphatases that are believed to of Abi through the ubiquitin-proteosome pathway, activate the Src by dephosphorylating the regulatory thus allowing Abl to exert its oncogenic potential (Dai Tyr 530 (Egan et al., 1999). The mouse models of et al., 1998). Bcr-Abl positive leukemias are de®cient in mammary carcinoma indicated above are examples of Abi protein, and suggests a mechanism by which post-translational modi®cation by cellular or viral activated Src can in¯uence oncogenesis. proteins. In the event of post-translational modi®cation, the eects to the cell will be more limited to Src- speci®c eects, not aecting parallel pathways. Rami®cations of Src activation Thirdly, there exists the possibility of activating mutations in the SRC gene. While a number of induced Src has been implicated in the progression and mutations in Src have been studied and found to be metastasis of many human cancers. Increased Src transforming, only one naturally occurring human activity has been shown to increase the growth rate mutation has been reported (Irby et al., 1999). Using a of cells (Mao et al., 1997), reduce adhesion between very sensitive, mutant allele-speci®c assay, a single base cells (Hamaguchi et al., 1993), and increase metastatic pair C-T transition was found in a subset of colon potential of cells (Irby et al., 1999). While the the cancer liver metastases and advanced colon cancer downstream targets of c-Src are still being de®ned, primary lesions (Irby et al., 1999). This mutation was Src has been shown to interact with regulatory identi®ed in codon 531 and resulted in truncation of the pathways in the cell cycle and signal transduction Src protein leading to Src activation. Src 531 was shown cascades involved with proliferation. For example, to be modestly transforming in focus formation assays Src interacts with Ras signaling pathways (Haas et and led to increased levels of metastatic lung coloniza- al., 2000; Machida et al., 2000; Odajima et al., 2000; tion. Recent attempts at replicating this work were not Tokumitsu et al., 2000) and Ras function has been successful possibly because either insensitive RFLP shown to be necessary for v-Src transformation of assays were utilized as the sole screening assay, or a human epithelial cells (Tokumitsu et al., 2000). Src preponderance of advanced lesions were excluded from activation provides a link to the Ras/MAPK cascade study, without the bene®t of microdissection to (Haas et al., 2000). Activated Src, either mutant Src eliminate the presence of normal contaminating tissues 531 or v-Src, has been shown to activate the or cellular in®ltrates. (Daigo et al., 1999; Wang et al., transcription factor Stat3 (Yu et al., 1995). Moreover, 2000). Detection of uncommon heterozygous mutations blocking Stat3 signaling has been shown to suppress likely requires a more sensitive mutant allele speci®c Src transformation (Bromberg et al., 1998; Turkson

Oncogene Src activation in human cancer RB Irby and TJ Yeatman 5640 et al., 1998; Wen et al., 1999). Each of these events VEGF appears to be somewhat reciprocal. VEGF alone may be capable of enhancing tumor growth induced activity is negated with the inhibition of Src (Bowman et al., 2000). A summary of these concepts activity. Src also appears to provide a link in VEGF linking Src activation to the development of the signaling of MAPK pathways that contributes to tumor metastatic phenotype is presented in Figure 1. proliferation and motility (Munshi et al., 2000). Activated Src also associates with the adhesion Src is capable of enhancing tumor growth through proteins, catenins and cadherins (Hamaguchi et al., the induction of Bcl-xL expression (Karni et al., 1999). 1993). Furthermore, Src appears to be activated by Bcl-xL is an anti-apoptotic protein that is transcrip- integrin-mediated substrate adhesion, which in turn, tionally regulated by Stat3 (Bromberg et al., 1999; causes inactivation and disruption of the cadherin/ Catlett-Falcone et al., 1999). Other Stat3-regulated catenin complex, and induces a metastatic phenotype genes implicated in cell cycle control and oncogenesis (Genda et al., 2000; Hamaguchi et al., 1993). are cyclin D1 (Bromberg et al., 1998) and c-Myc Src associates with EGFR, and is both activated by (Bromberg et al., 1999). Thus, Src activates Stat3, and activates the receptor in a synergistic fashion which in turn induces expression of genes that (Biscardi et al., 1999b; Tice et al., 1999). As many potentially contribute to oncogenesis by preventing human cancers are known to over-express both EGFR apoptosis and stimulating cell cycle progression. and c-Src, it is reasonable that they have coordinated functions. EGFR has been shown to activate c-Src kinase, and inhibition of EGFR reduces Src activity to Src as a target for anti-cancer drugs basal levels in colon cancer cell lines (Mao et al., 1997). Furthermore, association of c-Src with EGFR results The prevalence of activated Src in cancer indicates in phosphorylation of EGFR Tyr 845 and subsequent that this protein may play a signi®cant role in the activation of the receptor (Biscardi et al., 1999b). This progression of many cancers, and thus, is a likely results in a synergistic increase in EGF mediated DNA target for drug discovery eorts. Src activation has synthesis, growth in soft agar, and tumorigenesis over been shown to increase growth rates and invasion cells over-expressing either protein alone (Maa et al., characteristics of tumors, and to decrease apoptosis 1995). in tumor cells. Reduced expression of any of these Src induces vascular epithelial growth factor properties would potentially induce retarded tumor (VEGF), enhancing angiogenesis (Fleming et al., 1997; growth and reduction of metastatic tendencies. The Munshi et al., 2000). The relationship between Src and recent use of Src inhibitors or antisense therapy in nude mouse studies (Staley et al., 1997), pancreatic cell growth (Lutz et al., 1998), and leukemia cells (Roginskaya et al., 1999) supports the validity of this hypothesis. Numerous in vitro studies have been conducted to investigate the eects of Src speci®c inhibitors, general tyrosine kinase inhibitors, and growth factor receptor inhibitors. The challenge has been to develop inhibitors, which are not only speci®c for Src, but also stable in vivo and are bioavailable.

Conclusions

The role of c-Src in human cancer has evolved from a curious avian viral homolog with high transformation capacity to a candidate oncogene with a critical position in the signal transduction cascades that control cell growth. Clearly, Src does not act alone but works in concert with a large number of substrates, which orchestrate the processes of human cancer development and tumor progression. Moreover, there is a signi®cant amount of data supporting the in¯uence Figure 1 The role of Src activation in promoting metastasis. Src of Src on the development of the metastatic phenotype. activation can be viewed as central to the multiple signal Collectively, these observations linking Src to multiple transduction pathways necessary or sucient to induce the metastatic phenotype. Src does not act alone, but rather processes that determine the clinical outcome of a cooperates with other proteins and substrates whose interactions tumor make it a promising target for drug discovery lead to transcription of metastasis promoting genes. Src activation and targeting. can occur through a variety of mechanisms including interactions with receptor tyrosine kinases and integrin receptors. Some of these interactions require partnerships such as those between EGFR and Her2/neu or between intergrins, Fak, and Cas. Src activation in turn can lead to the activation of signal transduction pathways and to the activation of individual substrates, which may also impact on signal transduction pathways. The end result Acknowledgments is the production of factors that enable the metastatic process We thank R Jove for his kind reading of the manuscript. such as those involved with growth, survival, invasion, migration, This work was supported by a grant from the American adhesion, and angiogenesis Cancer Society ACS RPG-99-099-01.

Oncogene Src activation in human cancer RB Irby and TJ Yeatman 5641 References

Abts H, Jucker M, Diehl V and Tesch H. (1991). Leuk. Res., Irby R, Mao W, Coppola D, Jove R, Gamero A, 15, 987 ± 997. Cuthbertson D, Fujita DJ and Yeatman TJ. (1997). Cell Biscardi J, Tice D and Parsons S. (1999a). Advances in Growth Dier., 8, 1287 ± 1295. Cancer Research, 76, 61 ± 119. Irby RB, Mao W, Coppola D, Kang J, Loubeau JM, Biscardi JS, Belsches AP and Parsons SJ. (1998). Mol. TrudeauW,KariR,FujitaDJ,JoveRandYeatmanTJ. Carcinog., 21, 261 ± 272. (1999). Nat. Genet., 21, 187 ± 190. BiscardiJS,MaaMC,TiceDA,CoxME,LeuTHand Jacobs C and Rubsamen H. (1983). Cancer Res., 43, 1696 ± Parsons SJ. (1999b). J. Biol. Chem., 274, 8335 ± 8343. 1702. Bjelfman C, Hedborg F, Johansson I, Nordenskjold M and Jove R and Hanafusa H. (1987). Annu.Rev.CellBiol.,3, 31 ± Pahlman S. (1990). Cancer Res., 50, 6908 ± 6914. 56. Bjorge JD, O'Connor TJ and Fujita DJ. (1996). Biochem. Karni R, Jove R and Levitzki A. (1999). Oncogene, 18, Cell Biol., 74, 477 ± 484. 4654 ± 4662. Bolen J, Veillette A, Schwartz A, DeSeau V and Rosen N. Kumble S, Omary MB, Cartwright CA and Triada®lopoulos (1987a). Proc. Natl. Acad. Sci. USA, 84, 2251 ± 2255. G. (1997). Gastroenterology, 112, 348 ± 356. Bolen J, Veillette A, Schwartz A, Deseau V and Rosen N. LaVallee TM, Prudovsky IA, McMahon GA, Hu X and (1987b). Oncogene Res., 1, 149 ± 168. Maciag T. (1998). J. Cell Biol., 141, 1647 ± 1658. Bolen JB, Rosen N and Israel MA. (1985). Proc. Natl. Acad. Levitzki A. (1996). Anticancer Drug Des., 11, 175 ± 182. Sci. USA, 82, 7275 ± 7279. LuttrellD,LeeA,LansingT,CrosbyR,JungK,WillardD, Bowman T, Garcia R, Turkson J and Jove R. (2000). Luther M, Rodriguez M, Berman J and Gilmer T. (1994). Oncogene, 19, 2474 ± 2488. Proc. Natl. Acad. Sci. USA, 91, 83 ± 87. Bromberg JF, Horvath CM, Besser D, Lathem WW and Lutz MP, Esser IB, Flossmann-Kast BB, Vogelmann R, Darnell Jr JE. (1998). Mol. Cell Biol., 18, 2553 ± 2558. Luhrs H, Friess H, Buchler MW and Adler G. (1998). Bromberg JF, Wrzeszczynska MH, Devgan G, Zhao Y, Biochem. Biophys. Res. Commun., 243, 503 ± 508. Pestell RG, Albanese C and Darnell Jr JE. (1999). Cell, 98, Maa M, Leu T, McCarley D, Schatzman R and Parsons S. 295 ± 303. (1995). Proc. Natl. Acad. Sci. USA, 92, 6981 ± 6985. Brown M and Cooper J. (1996). Biochim. Biophys. Acta., Machida K, Matsuda S, Yamaki K, Senga T, Thant AA, 1287, 121 ± 149. Kurata H, Miyazaki K, Hayashi K, Okuda T, Kitamura Brummer J, Neumaier M, Gopfert C and Wagener C. (1995). T, Hayakawa T and Hamaguchi M. (2000). Oncogene, 19, Oncogene, 11, 1649 ± 1655. 1710 ± 1718. Budde RJ, Ke S and Levin VA. (1994). Cancer Biochem. Mao W, Irby R, Copolla D, Fu L, Wloch M, Turner J, Yu H, Biophys., 14, 171 ± 175. Garcia R, Jove R and Yeatman T. (1997). Oncogene, 15, Cartwright C, Coad C and Egbert B. (1994). J. Clin. Invest., 3083 ± 3090. 93, 509 ± 515. Masaki T, Okada M, Shiratori Y, Rengifo W, Matsumoto K, Cartwright C, Kamps M, Meisler A, Pipas J and Eckhart W. MaedaS,KatoN,KanaiF,KomatsuY,NishiokaMand (1989). J. Clin. Invest., 83, 2025 ± 2033. Omata M. (1998). Hepatology, 27, 1257 ± 1264. Cartwright C, Meisler A and Eckhart W. (1990). Proc. Natl. Masaki T, Okada M, Tokuda M, Shiratori Y, Hatase O, Acad. Sci. USA, 87, 558 ± 562. Shirai M, Nishioka M and Omata M. (1999). Hepatology, Catlett-Falcone R, Landowski TH, Oshiro MM, Turkson J, 29, 379 ± 384. Levitzki A, Savino R, Ciliberto G, Moscinski L, Masaki T, Shiratori Y, Okada H, Nishioka M, Taniguchi K, Fernandez-Luna JL, Nunez G, Dalton WS and Jove R. Hatanaka Y and Omata M. (2000). Am. J. Gastroenterol., (1999). Immunity, 10, 105 ± 115. 95, 837 ± 838. Courtneidge SA, Fumagalli S, Koegl M, Superti-Furga G Mazurenko NN, Kogan EA, Zborovskaya IB and Kisseljov and Twamley-Stein GM. (1993). Dev. Suppl., 57 ± 64. FL. (1992). Eur. J. Cancer, 28, 372 ± 377. Czernilofsky AP, Levinson AD, Varmus HE, Bishop JM, Munshi N, Groopman JE, Gill PS and Ganju RK. (2000). J. Tischer E and Goodman HM. (1980). Nature, 287, 198 ± Immunol., 164, 1169 ± 1174. 203. Muthuswamy S and Muller W. (1994). Adv. Cancer Res., 64, Dai Z, Quackenbush RC, Courtney KD, Grove M, Cortez 111 ± 123. D, Reuther GW and Pendergast AM. (1998). Genes Dev., Muthuswamy S, Siegel P, Dankort D, Webster M and Muller 12, 1415 ± 1424. W. (1994). Mol. Cell. Biol., 14, 735 ± 743. Daigo Y, Furukawa Y, Kawasoe T, Ishiguro H, Fujita M, Odajima J, Matsumura I, Sonoyama J, Daino H, Kawasaki Sugai S, Nakamori S, Liefers GJ, Tollenaar RA, van de A, Tanaka H, Inohara N, Kitamura T, Downward J, Velde CJ and Nakamura Y. (1999). Cancer Res., 59, Nakajima K, Hirano T and Kanakura Y. (2000). J. Biol. 4222 ± 4224. Chem., (in press). Danhauser-Riedl S, Warmuth M, Druker BJ, Emmerich B Ottenho-Kal A, Rijksen G, van BE, Hennipman A, and Hallek M. (1996). Cancer Res., 56, 3589 ± 3596. Michels A and Staal G. (1992). Cancer Res., 52, 4773 ± Egan C, Pang A, Durda D, Cheng HC, Wang JH and Fujita 4778. DJ. (1999). Oncogene, 18, 1227 ± 1237. Park J and Cartwright C. (1995). Mol. Cell. Biol., 15, 2374 ± Fleming RY, Ellis LM, Parikh NU, Liu W, Staley CA and 2382. Gallick GE. (1997). Surgery, 122, 501 ± 507. Park J, Meisler A and Cartwright C. (1993). Oncogene, 8, Flossmann-Kast BB, Jehle PM, Hoe¯ich A, Adler G and 2627 ± 2635. Lutz MP. (1998). Cancer Res., 58, 3551 ± 3554. Pena SV, Melhem MF, Meisler AI and Cartwright CA. Genda T, Sakamoto M, Ichida T, Asakura H and Hirohashi (1995). Gastroenterology, 108, 117 ± 124. S. (2000). Lab. Invest., 80, 387 ± 394. Peng Z and Cartwright C. (1995). Oncogene, 11, 1955 ± 1962. GuyC,MuthuswamyS,CardiR,SorianoPandMullerW. Rahimi N, Hung W, Tremblay E, Saulnier R and Elliott B. (1994). Genes Dev., 8, 23 ± 32. (1998). J. Biol. Chem., 273, 33714 ± 33721. Haas M, Askari A and Xie Z. (2000). J. Biol. Chem. (in Roginskaya V, Zuo S, Caudell E, Nambudiri G, Kraker AJ press). and Corey SJ. (1999). Leukemia, 13, 855 ± 861. HamaguchiM,MatsuyoshiN,OhnishiY,GotohB, Takeichi M and Nagai Y. (1993). EMBO. J., 12, 307 ± 314.

Oncogene Src activation in human cancer RB Irby and TJ Yeatman 5642 RosenN,BolenJB,SchwartzAM,CohenP,DeSeauVand Tokumitsu Y, Nakano S, Ueno H and Niho Y. (2000). J. Israel MA. (1986). J. Biol. Chem., 261, 13754 ± 13759. Cell. Physiol., 183, 221 ± 227. Rous P. (1911). J. Exp. Med., 13, 397 ± 411. Turkson J, Bowman T, Garcia R, Caldenhoven E, De Groot Shimakage M, Kawahara K, Kikkawa N, Sasagawa T, RP and Jove R. (1998). Mol. Cell. Biol., 18, 2545 ± 2552. Yutsudo M and Inoue H. (2000). Int. J. Cancer, 87, 5 ± 11. Verbeek BS, Vroom TM, Adriaansen-Slot SS, Ottenho- Staley C, Parikh N and Gallick G. (1997). Cell Growth & Kal AE, Geertzema JG, Hennipman A and Rijksen G. Dier., 8, 269 ± 274. (1996). J. Pathol., 180, 383 ± 388. Stehelin D, Varmus H, Bishop J and Vogt P. (1976). Nature, Visser CJ, Rijksen G, Woutersen RA and De Weger RA. 260, 170 ± 173. (1996). Lab. Invest., 74, 2 ± 11. Takeshima E, Hamaguchi M, Watanabe T, Akiyama S, Wang NM, Yeh KT, Tsai CH, Chen SJ and Chang JG. Kataoka M, Ohnishi Y, Xiao HY, Nagai Y and Takagi H. (2000). Cancer Lett., 150, 201 ± 204. (1991). Jpn. J. Cancer Res., 82, 1428 ± 1435. Weber T, Steele G and Summerhayes I. (1992). J. Clin. Talamonti MA, Roh MS, Curley SA and Gallick GE. (1991). Invest., 90, 815 ± 821. J. Clin. Invest., 91, 53. Wen X, Lin HH, Shih HM, Kung HJ and Ann DK. (1999). J. Termuhlen PM, Curley SA, Talamonti MS, Saboorian MH Biol. Chem., 274, 38204 ± 38210. and Gallick GE. (1993). J. Surg. Res., 54. Wiener JR, Nakano K, Kruzelock RP, Bucana CD, Bast Jr Thomas SM and Brugge JS. (1997). Annu. Rev. Cell. Dev. RC and Gallick GE. (1999). Clin. Cancer Res., 5, 2164 ± Biol., 13, 513 ± 609. 2170. Tice DA, Biscardi JS, Nickles AL and Parsons SJ. (1999). YuCL,MeyerDJ,CampbellGS,LarnerAC,Carter-SuC, Proc. Natl. Acad. Sci. USA, 96, 1415 ± 1420. Schwartz J and Jove R. (1995). Science, 269, 81 ± 83.

Oncogene