Tyrosine kinases and gastric cancer

Tyrosine kinases and gastric cancer

Wen-chang Lin*,1,5, Hsiao-Wei Kao1, Daniel Robinson2, Hsing-Jien Kung2, Chew-Wun Wu3 andHua-Chien Chen*,4,5

1Institute of Biomedical Sciences, Academia Sinica, Taipei 115, Taiwan, Republic of China; 2Department of Biological Chemistry,UC Davis School of Medicine, UC Davis Cancer Center, Sacramento, California, CA 95817, USA; 3Department of Surgery,Veterans General Hospital-Taipei and School of Medicine, National Yang-Ming University, Taipei 112, Taiwan, Republic ofChina; 4Division of Biotechnology and Pharmaceutical Research, National Health Research Institutes, Taipei 115, Taiwan,Republic of China

Carcinoma of the stomach is one of the most prevalentcancer types in the world today. Two major forms ofgastric cancer are distinguished according to theirmorphological and clinicopathological classi®cations(well di�erentiated/intestinal type and poorly di�eren-tiated/di�use type), characteristics that could also beattributed to the altered expression of di�erent types ofoncogenes or tumor suppressor genes. Signi®cantdi�erences exist for gastric cancer incidence comparingpeople of di�erent ethnic origins, implicating variousgenetic and epigenetic factors for gastric oncogenesis.There are only a limited number of molecular markersavailable for gastric cancer detection and prognosticevaluation, among which are tyrosine kinases. There isconvincing evidence that tyrosine kinases are involved inoncogenesis and disease progression for many humancancers. Ampli®cations of certain tyrosine kinases (c-met, k-sam and erbB2/neu) have been associated withhuman gastric cancer progression. Alternatively splicedtranscripts and enhanced protein-expression levels forsome of these tyrosine kinases are correlated withclinical outcomes for gastric cancer patients. Withadvent of high throughput techniques, it is now possibleto detect nearly all expressed tyrosine kinases in a singlescreen. This increases the chance to identify additionaltyrosine kinases as predictive markers for gastriccancers. In this article, we will ®rst review the literaturedata concerning certain tyrosine kinases implicated ingastric carcinogenesis and then summarize more recentwork which provide comprehensive tyrosine kinasepro®les for gastric cancer specimens and cell lines.Two new gastric cancer molecular markers (tie-1 andmkk4) have been identi®ed through the use of thesepro®les and demonstrated e�ective as clinical prognosticindicators. Oncogene (2000) 19, 5680 ± 5689.

Keywords: gastric cancer; protein-tyrosine kinase pro-®le

Introduction

Gastric cancer remains one of the most commoncancer types in the world today, despite the fact thatits incidence has gradually declined in many countries(Neugut et al., 1996). Gastric cancer is more prevalentin the Andean region of South America and in the Far

East (including Taiwan) and it is still the leading causeof cancer death for several countries (e.g. Japan, Chile)(Parkin et al., 1988). In Taiwan, gastric cancer ranks asthe fourth-highest cause of cancer-related death, with amortality rate of 10.72 per 100 000 for the year 1999(Department of Health, 2000). In the United States, acountry with relatively low gastric cancer incidence, themortality rate for 1997 was 3.4 and 7.5 per 100 000 forCaucasians and African Americans, respectively (Rieset al., 2000). Therefore, given this regional variation, itseems reasonable to suggest that genetic variations/factors contribute to gastric cancer oncogenesis fordi�erent ethnicities. Although diet is an importantfactor in the development of gastric and colorectalcancers, the di�erence of colorectal cancer incidencefor Caucasians and African Americans is not assigni®cant as that of gastric cancer. Again, thissuggests that genetic factors may play a signi®cantrole in the development of gastric cancers, togetherwith well-known epigenetic factors, such as diet andlife style. Another intriguing observation is that femalegastric cancer incidence is as low as half that for males(Ries et al., 2000) ± noted both for Taiwan (43%), ahigh incidence region, and the United States (44%), alow incidence region.

The clinical outcome for gastric cancer patientsremains poor, with a 5-year survival rate of only about20% (all stages) (Ries et al., 2000), and a mere 40% ofall patients responding to surgical intervention (Ajaniet al., 1995). Surgical resection is the only e�ectivecurative modality for primary treatment of gastriccancer, with the overall cumulative 5-year survival rateimproving to 62.4% for 362 patients in one report (Wuet al., 2000b). Most gastric cancer patients arediagnosed with advanced diseases, however, resultingin a dismal prognosis and underscoring the importanceof the identi®cation of useful diagnostic and prognosticmarkers for gastric cancers. For this report, wesummarize data contributing to the identi®cation oftyrosine kinases as potential markers for gastriccancers. To provide background information, we alsobrie¯y summarize the current knowledge for reportedgenetic alterations of gastric cancers.

Known genetic alterations for gastric cancer

With the aid of continuing advances in molecularbiology, recent research e�orts have identi®ed anumber of genetic alterations associated with gastriccancer oncogenesis. These genetic changes include

Oncogene (2000) 19, 5680 ± 5689ã 2000 Macmillan Publishers Ltd All rights reserved 0950 ± 9232/00 $15.00

www.nature.com/onc

*Correspondence: W-c Lin and H-C Chen5W-c Lin and H-C Chen share the corresponding authorship

genetic instability represented by microsatellite instabil-ity (CA repeats), reactivation of telomerase activity,inactivation of tumor-suppressor genes (p53, APC,DCC, MCC), and activation of oncogenes (VEGF,cyclin E, K-ras, erbB2, K-sam, c-met, TGF-bRII).Tahara has summarized and discussed these geneticalterations in great detail in his review articles (Tahara,1995a,b; Yasui et al., 1999, 2000).

Although some genetic changes are common (micro-satellite instability, telomerase activity, CD44 splicingvariants, and p53 mutation status), it is interesting tonote that several di�erences for genetic alterations havebeen observed comparing the poorly di�erentiated(mostly di�use-type carcinoma) and well-di�erentiated(mostly intestinal-type carcinoma) gastric cancers(Ming, 1998). In well-di�erentiated gastric cancers,inactivation of APC and DCC tumor-suppressor genesby LOH (loss of heterozygocity) and mutation isprevalent, as well as aberrant expression patterns ofnm23 and bcl-2 genes. A previous investigation of eightcommonly studied tumor markers (EGF, EGFR, TGF-a, p53, erbB2, cripto, NM23), comparing British andJapanese gastric cancer samples, revealed that only twomarkers generated di�erential results. Immunoreactiv-ity was reportedly higher for erbB2 in British gastriccancer specimens, while NM23 immunoreactivity waspositive for the greater portion of Japanese samples(Livingstone et al., 1995). Since these markers arecommon components in cellular growth and signalpathways, it is perhaps not surprising that similarexpression pro®les have been revealed for gastriccancers comparing two di�erent ethnicities. In otherstudies, there is no di�erence in erbB2 expressionamong British or Japan gastric cancers (McCulloch etal., 1995, 1997). Loss of E-cadherin and catenin genes ispredominant, however, for the poorly di�erentiatedgastric cancers (Ming, 1998). This evidence stronglysupports the view that distinct genetic pathways areinvolved in the development of the two major types ofgastric cancers. Regarding oncogene activation forgastric cancer carcinogenesis, K-ras mutation has beendetermined for the well-di�erentiated gastric types,however, with a much lower frequency than forcolorectal cancers (Ming, 1998).

Tyrosine kinases and gastric cancer

Tyrosine-kinase receptors are key molecules in signal-ing pathways leading to growth and di�erentiation ofnormal cells (Marshall, 1995). Aberrant expression oftyrosine kinases as re¯ected by the aberrant tyrosinephosphorylation in gastric cancer cells was reported(Takeshima et al., 1991). Three receptor-tyrosinekinases, c-met, k-sam, and erbB2/neu, which arefrequently implicated in epithelial cancers were mostextensively studied in gastric carcinomas. While c-met(HGF receptor) is over-expressed/ampli®ed for ad-vanced gastric cancers, for both well-di�erentiated andpoorly di�erentiated gastric carcinomas, k-sam (FGFreceptor 2) activation is more speci®c for the poorlydi�erentiated variants. By contrast, ampli®cation of theerbB2/neu gene is often revealed for well-di�erentiatedgastric carcinomas (Ming, 1998). Thus, subversion ofdi�erent signal pathways may contribute to theprogression of di�erent types of gastric cancers.

This review consists of two parts. In the ®rst, theinvolvement of the c-met, k-sam and erbB-2 in gastriccancers as reported in the literature will be reviewed.The discussion includes the characterization of thesegenes in gastric cancers, in terms of gene ampli®cationat the DNA level, alternatively spliced transcripts,protein expression as studied by immunohistochemis-try, and biological activities of these receptor tyrosinekinases. In the second part, the discussion will beextended to the entire collection of tyrosine kinasesexpressed in gastric cancers, i.e., the tyrosine kinasepro®les of gastric cancers. The cited literature is notmeant to be inclusive, but rather, representative.

HGF receptor tyrosine kinase

DNA amplification

The c-met receptor tyrosine kinase was ®rst isolatedfrom an osteogenic sarcoma cell line (Cooper et al.,1984), and subsequently identi®ed as the receptor forhepatocyte growth factor (HGF) (see, Furge et al. thisissue). The c-met functional unit consists of anextracellular alpha subunit and a transmembrane betasubunit. Ampli®cation of the c-met gene has beendemonstrated for gastric cancer tissues (Kuniyasu etal., 1992), especially for the scirrhous type. Co-ampli®cation of c-met and c-myc oncogenes have alsobeen reported for gastric cancers (Seruca et al., 1995).

Alternatively spliced transcripts

Rearrangement fusion of the tpr-met gene is one of theactivation mechanisms reported to produce aberranttranscripts (Park et al., 1986). In addition, isoforms ofthe c-met transcripts, generated by alternative splicing,have been observed (Rodrigues et al., 1991), and aspecial 6.0 Kb transcript, demonstrated for gastriccancer cells, has been correlated with cancer progression(Kuniyasu et al., 1993). The biological activities fordi�erent isoforms, especially alterations in the extra-cellular ligand-binding region, have not been elucidatedas yet. In KATO III gastric cancer cells, alternativelyspliced transcripts, generating in-frame deletion for thec-met related receptor tyrosine kinase (ron), have alsobeen demonstrated for the extracellular domain (Collesiet al., 1996; Okino et al., 1999). It is likely that thesealterations may alter ligand-binding a�nity, and aresubsequently involved in oncogenic activation.

Protein localization by immunohistochemistry

It has been demonstrated that the numbers offunctional c-met receptors are more important thanoverall numbers of receptors for the growth of gastriccancer cells (Liu et al., 1997). Nevertheless, a highpositive rate (40 ± 80%) for c-met immunoreactivity hasbeen observed in several reports (Park et al., 2000;Taniguchi et al., 1998; Wu et al., 1998). Simultaneousdetection of HGF expression and c-met immunoreac-tivity has also been noted for the majority of gastriccancer specimens (Park et al., 2000; Wu et al., 1998),indicating a possible autocrine or paracrine mechanism.This proposition is supported by growth inhibition forgastric cancer cell lines with anti-sense oligonucleotides

Oncogene

Tyrosine kinases in human gastric cancer

W-c Lin et al

5681

blocking c-met gene expression (Kaji et al., 1996). Forgastric cancer patients, co-expression of c-met anderbB2 proteins have been associated with poorersurvival rates compared with over-expression of eitherone (Nakajima et al., 1999). Co-ampli®cation of c-met,k-sam or erbB2 genes at the genomic level, however,has not been observed for most of the gastric cancerspecimens examined (Hara et al., 1998; Tsugawa et al.,1998; Tsujimoto et al., 1997). This indicates thatactivations of c-met, k-sam and erbB2 are likely to beindependently regulated. As stated previously, ampli-®cation of these receptor tyrosine kinases wasobserved separately for well-di�erentiated (k-sam)and poorly di�erentiated (erbB2) gastric cancers(Ming, 1998).

FGF receptor tyrosine kinase

DNA amplification

The k-sam receptor tyrosine kinase was ®rstidenti®ed from the KATO III gastric cancer cellline, and determined as the type 2 receptor for the®broblast growth factor (FGF) or keratinocytegrowth factor (KGF) (Hattori et al., 1990).Ampli®cation of the k-sam gene has been observedmainly for poorly di�erentiated gastric cancer tissues(Ming, 1998), correlating with poor prognosis(Hattori et al., 1996).


At least four major alternatively spliced k-samtranscripts (k-sam I ± IV) have been previously ob-served (Katoh et al., 1992). Alterations in theextracellular ligand binding region for k-sam I and k-sam II transcripts confer ligand speci®city betweenbasic FGF and KGF (Ishii et al., 1994; Miki et al.,1992). The coding sequences of K-sam III and k-samIV transcripts lack the transmembrane domain andtherefore encode for soluble forms of the FGFreceptor, with k-sam III transcript still retaining thesequence encoding the kinase domain (Katoh et al.,1992). Soluble k-sam protein has indeed been detectedin the culture medium of KATO III cells (Kishi et al.,1994). For KATO III gastric cancer cells, a majortranscript (k-sam IIC3) encoding a truncated carboxyl-terminus FGFR 2 protein has been demonstrated to betransforming (Itoh et al., 1994). This protein isgenerated by an alternative splicing mechanism andpossibly loses a regulatory tyrosine residue at the Y769

position due to truncation, similar to src oncogeneactivation. Recently, several additional novel k-samtranscripts, derived from alternative-splicing (Ueda etal., 1999a) or short homology-mediated recombination(Ueda et al., 1999b) have been discovered forscirrhous-type gastric cancer cell lines. Such truncatedforms of the k-sam receptor alter the induced-di�erentiation pathways for transfected cells (Ishii etal., 1995), indicating that alternatively spliced k-samproducts may have physiologically distinct functionsand be involved in the regulation of cellular di�er-entiation.

It has also been demonstrated that FGFR 4, anothermember of the FGFR tyrosine kinases, expressed

alternatively spliced transcripts that generate a solubleform of FGFR 4 for human gastric cancer cells(Takaishi et al., 2000). The biological signi®cance forthese soluble forms of receptor tyrosine kinases stillneeds further elucidation. It is possible that thesesoluble forms regulate membrane-type receptor func-tions by competing with ligand-binding activity, ormay serve as a ligand reservoir as seen for cytokinereceptors (Jung et al., 1999; Saily et al., 1999; Sayani etal., 2000; Scott and Weiss, 2000).


As articulated through immunohistochemistry, k-samprotein is detected in about 50% of poorly di�er-entiated gastric cancers, but not in well-di�erentiatedgastric cancers (Hattori et al., 1996). It has beendetermined that k-sam protein expression is associatedwith the malignant progression of gastric cancer,however, no k-sam autocrine/paracrine mechanism forgastric cancers has been observed. The protein-expression pattern for basic FGF and bek/k-sam doesnot correlate with tumor stages of gastric cancers(Takahashi et al., 1996). Since the FGFs and theirreceptors belong to large gene families (Gospodaro-wicz, 1989; Jaye et al., 1992), detailed expressionpro®les for all FGFs and FGFRs are needed toachieve a full appreciation about their involvement ingastric cancers.

The EGF/ErbB receptor family of tyrosine kinases

DNA amplification

Among receptor tyrosine kinases, the EGF receptorgene family including EGFR and erbB2, is the mostfrequently implicated in human cancers. Ampli®cationof EGFR and erbB2 genes for human gastric cancershas been determined at around 3*5% and 10*20%,respectively (Albino et al., 1995; Hirono et al., 1995;Sato et al., 1997; Tahara, 1995a; Tsugawa et al., 1993;Yoshida et al., 1989; Hung and Lau, 1999). All erbB2-ampli®ed gastric cancer cells reveal enhanced erbB2protein expression patterns (Ooi et al., 1998). Co-ampli®cation of gastrin and erbB2 has been reportedfor intestinal-type gastric cancers (Vidgren et al., 1999),probably because they are located close together atchromosome 17q.


In situ hybridization detection of the EGFR messagerevealed that its high expression is associated with ahigher frequency of gastric-cancer recurrence (Anzai etal., 1998). Activated erbB family receptor tyrosinekinases as well as the soluble form of the receptorproteins, can be generated by alternatively splicedmessages for all erbB family members (EGFR, erbB2,erbB3, and erbB4) (Flickinger et al., 1992; Gebhardt etal., 1998; Ilekis et al., 1995; Katoh et al., 1993; Kwongand Hung, 1998; Lee and Maihle, 1998; Sawyer et al.,1998; Scott et al., 1993). Secreted erbB2 and erbB3splicing variants have been identi®ed for human gastriccancer cells (Katoh et al., 1993; Scott et al., 1993).Elevated level of soluble EGFR and erbB2 proteins,


W-c Lin et al

5682

Oncogene

possibly resulting from alternatively spliced transcriptsor proteolytic cleavage, are detectable in the serum ofgastric cancer patients (Choi et al., 1997; Kono et al.,2000; Perng et al., 1994; Molina et al., 1997). Asdiscussed above, soluble tyrosine kinase receptors mayplay a regulatory role for receptor signaling. On theother hand, the elevated level of soluble receptorscould be a result of shedding from tumor cells over-expressing these receptors. In either case, they may bepro®tably used as markers for gastric cancer cells.


Immunoreactivity to EGFR antibodies is demonstratedfor 30 ± 50% of human gastric cancers (Jonjic et al.,1997; Yonemura et al., 1989) and correlated with poorprognosis, however, EGFR immunoreactivity is not anindependent prognostic factor according to multi-variate analysis (Santoro et al., 1997). For erbB2protein, positive rates for erbB2 immunoreactivityrange from 10 to 50% for the human gastric cancersthat have been studied, and is higher for well-di�erentiated gastric cancers (Kameda et al., 1990;Lee et al., 1996; Motojima et al., 1994; Orita et al.,1997; Uchino et al., 1993; Yonemura et al., 1991a). Theexpression of erbB2 protein is a signi®cant independentrisk factor for long term survival, disease recurrenceand lymph-node metastasis (Lin et al., 1995; Motojimaet al., 1994; Orita et al., 1997; Sanz-Ortega et al., 2000;Yonemura et al., 1991a).

Since homo-dimerization or hetero-dimerization be-tween members of the EGFR family is required foractivation, the presence of other erbB family members inthe same cell can profoundly in¯uence the outcome ofthe signals. The expression pattern of erbB familymembers (EGFR, erbB2, erbB3 and erbB4) has beenexamined in several studies. Co-expression of EGFR,erbB2 and erbB3 all three receptors has been demon-strated in a signi®cant portion of gastric cancers but notin chronic gastritis (Slesak et al., 1998). Expression ofErbB4 is less common and has been noted for gastriccancers in one study (Kataoka et al., 1998).

Autocrine and paracrine pathways have beendemonstrated for the erbB family. The presence ofthe EGFR ligands (EGF and TGF-a) and the erbB2/erbB3 ligand (heregulin-a), with their respectivereceptors, have been reported for the human gastricmucosa (Beauchamp et al., 1989) and gastric cancertissues (Tokunaga et al., 1995). Heregulin-a isexpressed in ®broblasts associated with stomach tissue,but is capable of inducing proliferation of gastriccancer cells (Noguchi et al., 1999). Proliferation index,tumor invasion and metastasis frequency are higher forgastric cancer patients expressing both EGFR andEGF/TGF-a in cancer tissues (Borlinghaus et al., 1993;Filipe et al., 1995; Karameris et al., 1993; Yasui et al.,1988; Yonemura et al., 1991b, 1992; Yoshida et al.,1990). Although both EGF binding activity and EGFRimmunoreactivity are important parameters for humangastric cancer progression, there is no signi®cantcorrelation between these two elements. Likewise, theexpression level of functionally activated receptorsshould be a more reliable biomarker than the totalnumbers of receptors (Turner et al., 2000) and higherEGFR kinase activity has indeed been observed forgastric cancer tissues compared to adjacent mucosa

(Kameda et al., 1992; Masaki et al., 2000). HigherEGF-binding activity also correlates with potential ofgrowth for gastric tumor cells in nude mice (Kiyama etal., 1992). This indicates that activation of expressedEGFR family receptors is one of the critical elementsin the growth of gastric cancers.

Blocking EGFR expression with antisense vectorsinhibits the growth of gastric cancer cells (Hirao et al.,1999). Antibodies against erbB2/neu have been demon-strated e�ective for reducing the tumor burden andprolonging survival of treated animals with gastriccancers (Karameris et al., 1993; Ohnishi et al., 1995;Tokuda et al., 1996). Targeting erbB2-expressing cellswith conjugated immunoliposomes containing che-motherapeutic reagents has also proven useful (Suzukiet al., 1995). Recently, the recombinant humanizedanti-HER2 antibody, herceptin, has been adopted forclinical use for treatment of several human cancers(Agus et al., 1999; Baselga et al., 1998; Stebbing et al.,2000). Its e�cacy in gastric cancers has yet to bedetermined.

Other protein tyrosine kinases in gastric cancer

There are several other protein tyrosine kinases thathave, in a limited fashion, been previously examined forhuman gastric cancers. These include src, vascularendothelial growth factor receptor (KDR), hek5 (erk),abl, nerve growth factor receptor (trkA), and stem cellfactor receptor (c-kit). An autocrine/paracrine mechan-ism has been proposed for the VEGF/KDR and SCF/c-kit ligand receptor systems for human gastric cancers(Hassan et al., 1998; Takahashi et al., 1996; Yonemura etal., 1999). Mutation of the c-kit tyrosine kinase gene hasbeen reported for gastrointestinal stromal tumors, andassociated with poor clinical prognosis (Ernst et al.,1998; Sakurai et al., 1999). The trkA and hek5 receptortyrosine kinases are expressed more frequently for well-di�erentiated gastric cancers than for the poorlydi�erentiated variants (Kiyokawa et al., 1994; Koizumiet al., 1998). While abl and src non-receptor tyrosinekinases are expressed in all gastric cell lines and cancertissues examined, the src protein levels do not re¯ectactivation status (kinase activity) (O'Neill et al., 1997;Takaishi et al., 2000). It is worth noting thatH. pylori, aninfectious pathogen, can up-regulate tyrosine kinasesignal transduction pathways including src, EGFR,erbB2, MAPK and MKK4, following infection (Elitsuret al., 1999; Haruma et al., 2000; Meyer-Ter-Vehn et al.,2000; Naumann et al., 1999; Romano et al., 1998).Recently, H. pylori infection has been linked to pepticulcer disease and gastric adenocarcinoma formation(EUROGAST Study Group, 1993; Segal et al., 1997;Kuniyasu et al., 2000). Reduced EGF and EGFR levelsfor stomach tissues have been observed following theeradication of H. pylori (Coyle et al., 1999), however, H.pylori status does not correlate with c-met and erbB2immunoreactivities in clinicopathological analysis (Wuet al., 1997).

Tyrosine kinase pro®les for human gastric cancer

As has been discussed, about 15 ± 20 tyrosine kinaseshave been previously investigated for human gastric

Oncogene


W-c Lin et al

5683

cancers, with only c-met, k-sam, EGFR and erbB2studied in greater detail. The past e�orts wereprimarily directed toward selective types of tyrosinekinases, based on investigators' insights, expertise andreagents available. Considering the fact that tyrosinekinases often cross-talk and the activation of one maya�ect the other, there is a need to `see' the entirecollection of expressed tyrosine kinases, if one is tofully appreciate the detailed oncogenic pathway(s) or tohave the best chance of identifying tyrosine kinase-based biomarkers. Several high-through-put ap-proaches including cDNA microarray (DeRisi et al.,1996; Moch et al., 1999), Serial Analysis of GeneExpression (Velculescu et al., 1995) and kinetic PCR(Kang et al., 1997) have been developed to study globalgene expression pattern; the one that will be high-lighted here is RT ±PCR based tyrosine kinasepro®ling approach, using degenerate primers thatpermit the ampli®cation of all tyrosine kinases whichhave these highly conserved motif in the catalyticdomain (Robinson et al., 1996). In a single reaction,most, if not all expressed tyrosine kinases can beidenti®ed. In this section, the results of a comparisonof two matched pairs of gastric cancer and normalspecimen are reviewed. In a later section, the results ofa comparative study of four gastric cancer cell lines willbe summarized. The latter results were obtained usingan improved RT±PCR based approach, tyrosinekinase display, which utilizes restriction digestions inlieu of cloning and sequencing, providing a simpler andmore quantitative way of identi®cation of tyrosinekinases. Both approaches together have identi®ed atotal of 44 tyrosine kinases expressed in various gastriccancers. Some of them have been previously reportedto be expressed in gastric cancers, while others aresomewhat unexpected to be found in gastric tissues. Atleast two of them were shown to be of prognosticvalues. As these primers also identify dual and a subsetof serine/threonine kinases involved in signaling andcell cycle, information concerning their expressions isalso included in the data summary.

It is estimated that there are 1000 ± 2000 di�erentprotein kinases encoded by the human genome andabout 10% of them represent tyrosine kinases (Hunter,1987). A close scrutiny of the human genome databasereveals the existence of close to 95 tyrosine kinases (seeRobinson et al., this issue). With this number, it isreasonable to assume that about 30 ± 50 tyrosinekinases are expressed in a given cell, a number largeenough to give a `®ngerprint' of a cancer cell, but smallenough to be identi®ed in a simple screen. RT ±PCRapproaches using degenerate primers have beenemployed by several laboratories to identify newtyrosine kinases for various human cells (for examples,Schultz and Nigg, 1993; Takahashi et al., 1995; Wilks,1989), including gastric cancer cells (Kim et al., 1995).Robinson et al. (1996) designed a set of primers basedon motifs in subdomain VII and IX of the catalyticdomain of tyrosine kinases, and optimized theconditions, such that the reaction is of high ®delity(more than 90% of the amplicons are kinases) andbroad speci®city (more than 90% of tyrosine kinasescan be primed). Wu et al. (2000a) utilized a slightmodi®cation of this approach (Lin et al., 1998, 1999ato compare two pairs of gastric cancers. The datasummarized in Figure 1a reveals the expression of 32

di�erent tyrosine kinases and a few dual kinases(mkk3, mkk4 etc) in these tissues, including 13 non-receptor tyrosine kinases, 19 receptor tyrosine kinases.These studies con®rmed the earlier reports that c-met,k-sam and erbB2 are di�erentially expressed in gastriccancer cells, but also identi®ed a number of otherdi�erentially expressed tyrosine kinases that werepreviously unrecognized. The latter include fyn, itk,tyk2, c-fms, c-kit, cak, tie-1, mkk4which are morehighly expressed in cancer tissues. The presence of c-fms, c-kit and tie-1 is likely due to contaminatinghematopoietic and vascular cells in the tissue speci-mens. Further immunohistochemical study howevershowed that tie-1, a receptor tyrosine kinase thought tohave a restricted expression in endothelial andhematopoietic cells (Hashiyama et al., 1996; Sato etal., 1993), is actually expressed in gastric adenocarci-noma cells as well (Figure 1b). The aberrantlyexpressed tie-1 protein is localized in the cytoplasmicregion of the cancer cells which are located on theluminal side of the adenocarcinoma tissues (Lin et al.,1999; Wu et al., 2000a). Moreover, the aberrant

a

b

Figure 1 (a) Tyrosine kinase expression pro®les for humangastric cancer tissues. Two di�erent pairs of gastric cancersurgical specimens were used for RT±PCR ampli®cation oftyrosine kinases. Individual clones were randomly selected andtheir insert sequences were determined. A total of 243 tyrosinekinase clones were analysed from gastric cancer tissues and 129clones from their matching normal gastric mucosa. The height ofthe bar indicates the number of clones identi®ed and sequenced.A total of 32 di�erent tyrosine were identi®ed. (b) Immunohis-tochemical staining of tie-1 tyrosine kinase in human gastricadenocarcinoma sections. Anti-tie-1 protein antibody (Santa CruzBiotech, Santa Cruz, CA, USA) was used for immunohistochem-ical staining of human gastric cancer tissues by ABC methods(Lin et al., 1999) (original magni®cation 6400)


W-c Lin et al

5684

Oncogene

expression of tie-1 in gastric cancer cells was shown tobe associated with reduced survival of the patients andserve as an independent predictor for survival (Lin etal., 1999). Likewise, expression of the mkk4 protein (astress-induced dual kinase that phsophorylates theMAPK family of kinases) was observed in 44.8%(43/96) of gastric adenocarcinoma patients. Based onmultivariate analysis, mkk4 was also shown to be anindependent prognostic factor (relative risk=2.1) forgastric cancer progression (Wu et al., 2000a). Asigni®cant role may be played by mkk4 kinase duringthe later stages of gastric cancer development.

The tyrosine kinase pro®ling approach describedabove is e�ective for the identi®cation of nearly alltyrosine kinases expressed for particular cells ortissues. Yet, the cloning and sequencing steps remaintedious and the introduction of these steps compro-mises the quantitative aspects of this approach. Avariation of this procedure utilizing restriction

digestion analysis of the amplicon instead of cloningand sequencing was recently developed. This method,referred to as tyrosine kinase display in this review,combines the special features of the RT ±PCR/degenerate-primers approach described above andthe di�erential display approach which separateindividual genes by sizing on a sequencing gel.Brie¯y, by radio-labeling the 5' end of forwardprimers, subjecting the resulting amplicon to restric-tion digestions, and resolving the 5' end fragments ina sequencing gel, individual tyrosine kinases can beidenti®ed by the characteristic sizes of their 5' endfragments. Since each molecule is labeled only at the5' end, the relative intensity of the displayedfragments re¯ects the approximate molar ratio ofthe tyrosine kinase transcript copies. This approachwhich does away with cloning and sequencing shouldgreatly improve the quantitative aspects of theprocedure.

a b

Figure 2 (a) Tyrosine kinase display for the human gastric cancer cell line, AGS. Human gastric-cancer cell AGS cDNA wasampli®ed with degenerated primers (33P-label at the 5' end of the sense primer). Equal amounts of PCR amplicons (100 000 c.p.m.)were further digested with speci®c restriction endonucleases as indicated. The digested tyrosine kinase amplicon fragments wereseparated on a 7% denaturing polyacrylamide gel. Distinct bands representing di�erent kinases were identi®ed according to a pre-established restriction-fragment database. (b) Tyrosine kinase expression pro®les for the four di�erent human gastric cancer celllines. The cDNA samples from four human gastric cancer cell lines (AGS, AZ521, HR, and KATO III) were subject to tyrosinekinase display analysis. Distinct tyrosine kinase bands were assigned by the sizes of their speci®c restriction fragments, and theexpression level (radiation dose unit PSL) was measured with a Fuji BAS phospho-imager. The height of each bar indicates the PSLor the intensity of each band as measured by the phospho-imager. The upper closely-spaced bands are the intact ampliconsbelonging to kinases, which do not have the particular restriction site present in the ampli®ed region

Oncogene


W-c Lin et al

5685

A representative tyrosine kinase display example isshown in Figure 2a, the amplicons from a humangastric cancer cell line (AGS) were digested with sevendi�erent restriction enzymes and separated on adenaturing sequencing gel. Di�erent tyrosine kinasesand dual kinases were identi®ed according to theirspeci®c restriction-fragment sizes, when compared to apre-established database. For example, the amplicon ofhek5 (eph-B2), but not of any other tyrosine kinase,after Aci I digestion is predicted to yield a 5' endfragment of 74 bp in size and is identi®ed as such.Similarly, expressions of erbB2 and c-met genes in AGScells are readily observed by Bsp12886 and Alu Idigestion (Figure 2a). In some cases, fragments of asimilar size represent a number of di�erent kinases(labeled as `many'). Therefore, alternative restrictionenzymes are necessary to ensure identi®cation ofunique gene-speci®c restriction fragments. The intensityof individual bands visible in the phosphoimager-processed ®les can be used to quantify the expressionlevel of the original transcripts. Thus, using a singleRT ±PCR reaction and using a single gel, a tyrosinekinase pro®le can be obtained. The results aresummarized in Figure 2b, with the height of the bardepicting the intensity of the band, and hence theexpression level. In this ®gure, four di�erent humangastric cancer cell lines (two American gastric cancercell lines, AGS (Barranco et al., 1983) and HR(Shimizu et al., 1991), and two Japanese gastric cancercell lines, AZ521 (Taki et al., 1985) and KATO III(Sekiguchi et al., 1978)) are compared. Thirty tyrosinekinases were identi®ed from these cell lines including 10non-receptor tyrosine kinases and 20 receptor tyrosinekinases. The AGS line expressed the highest number oftyrosine kinases (25 di�erent tyrosine kinases). All ofthe cell lines studied express yes, c-met, EGFR, erbB2and k-sam at varying degrees, consistent with earlierreports. AZ521 express all four members of the erbBfamily of receptors. KATOIII expresses three of them(i.e., EGFR, erbB2 and erbB3) while the other two celllines only harbor EGFR and erbB2. Ron, a receptorrelated to c-met, is expressed in two of the cell lines(AGS and KATO III). It is interesting to note thatboth members of the abl family of kinases, abl and arg,are expressed in all four cell lines. Most of thesetyrosine kinases are expressed at moderate level (200 ±5000 radiation dose units) with the exception of k-sam/

FGFR2 in KATO III and c-yes in HR, which areexpressed 10 ± 100-fold higher than other kinases(above 20 000 radiation dose units). This high level ofexpression usually results from gene ampli®cation. Thisis indeed the case for k-sam in KATO III (Hara et al.,1998). Whether yes is ampli®ed at the genomic level inHR cells remains to be determined. Both HR andKATO III cells were from the metastases of gastriccancers. It has been demonstrated that k-sam is animportant indicator for poorly di�erentiated gastriccancer (Hattori et al., 1990). Whether yes or its relatedkinases involved in the progression of gastric cancermerits further investigation. Ampli®cation of yesoncogene was reported in a primary gastric carcinoma(Seki et al., 1985). Many tyrosine kinases areubiquitously expressed in the various gastric cancercell lines. The greatest variation in terms of theexpression patterns seems to be associated with theeph/hek family of tyrosine kinases. For instance, hek5is found to be expressed in all four cell lines withvarying degree, consistent with its role in tumorigenesis(Kiyokawa et al., 1994). Additional two eph receptortyrosine kinases were noted in the pro®le, with hek8expressed for AGS and HR cells, while AZ521 and HRcells expressed hek11. For KATO III cells, neither hek8nor hek11 were detected. This variation could meanthat eph/hek family of kinases serve redundantfunctions or they may endow the di�erent cell lineswith di�erent migratory and metastatic properties, asthese kinases generally function in cell ± cell adhesionand communication (see Dodelet and Pasquale, thisissue). The development of these tyrosine kinasepro®les provides a good framework to study thefunctions of these kinases in gastric carcinogenesis aswell as to identify potential predictive markers andintervention targets.

AcknowledgmentsThis study was supported in part by grants from AcademiaSinica Medical Biotechnology Research and from theNational Science Council to W-c Lin (NSC 90-2314-B-001-007-M58) and US. NIH and DOD grants to HJ Kung.We would like to thank Mr Jyh-Shi Lin for his excellenttechnical assistance.

References

Agus DB, Scher HI, Higgins B, Fox WD, Heller G, FazzariM, Cordon-Cardo C and Golde DW. (1999). Cancer Res.,59, 4761 ± 4764.

Ajani JA, Mans®eld PF and Ota DM. (1995).World J. Surg.,19, 216 ± 220.

Albino AP, Jaehne J, Altorki N, Blundell M, Urmacher C,Lauwers G, Niedzwiecki D and Kelsen DP. (1995). Eur. J.Surg. Oncol., 21, 56 ± 60.

Anzai H, Kitadai Y, Bucana CD, Sanchez R, Omoto R andFidler IJ. (1998). Eur. J. Cancer, 34, 558 ± 565.

Barranco SC, Townsend Jr CM, Casartelli C, Macik BG,Burger NL, Boerwinkle WR and Gourley WK. (1983).Cancer Res., 43, 1703 ± 1709.

Baselga J, Norton L, Albanell J, Kim YM andMendelsohn J.(1998). Cancer Res., 58, 2825 ± 2831.

Beauchamp RD, Barnard JA, McCutchen CM, Cherner JAand Co�ey Jr RJ. (1989). J. Clin. Invest, 84, 1017 ± 1023.

Borlinghaus P, Wieser S and Lamerz R. (1993). Clin.Investig., 71, 903 ± 907.

Choi JH, Oh JY, Ryu SK, Kim SJ, Lee NY, Kim YS, Yi SY,Shim KS and Han WS. (1997). Cancer, 79, 1879 ± 1883.

Collesi C, Santoro MM, Gaudino G and Comoglio PM.(1996). Mol. Cell. Biol., 16, 5518 ± 5526.

Cooper CS, Park M, Blair DG, Tainsky MA, Huebner K,Croce CM and Vande Woude GF. (1984). Nature, 311,29 ± 33.

Coyle WJ, Sedlack RE, Nemec R, Peterson R, DuntemannT, Murphy M and Lawson JM. (1999). Am. J. Gastro-enterol., 94, 2885 ± 2889.

Department of Health. (2000). Leading Causes of Death fromCancer of Taiwan-Fuchien Area, The Republic of China.Department of Health, Taipei.


W-c Lin et al

5686

Oncogene

DeRisi J, Penland L, Brown PO, Bittner ML, Meltzer PS,Ray M, Chen Y, Su YA and Trent JM. (1996).Nat. Genet.,14, 457 ± 460.

Elitsur Y, Majumdar AP, Tureaud J, Dosescu J, Neace C,Velusamy L and Moshier JA. (1999). Life Sci., 65, 1373 ±1380.

Ernst SI, Hubbs AE, Przygodzki RM, Emory TS, Sobin LHand O'Leary TJ. (1998). Lab. Invest., 78, 1633 ± 1636.

EUROGAST Study Group. (1993). Lancet, 341, 1359 ± 1362.Filipe MI, Osborn M, Linehan J, Sanidas E, Brito MJ and

Jankowski J. (1995). Br. J. Cancer, 71, 30 ± 36.Flickinger TW, Maihle NJ and Kung HJ. (1992). Mol. Cell.

Biol., 12, 883 ± 893.Gebhardt F, Zanker KS and Brandt B. (1998). Biochem.

Biophys. Res. Commun., 247, 319 ± 323.Gospodarowicz D. (1989). Crit. Rev. Oncog., 1, 1 ± 26.Hara T, Ooi A, Kobayashi M, Mai M, Yanagihara K and

Nakanishi I. (1998). Lab. Invest., 78, 1143 ± 1153.Haruma K, Ito M, Kohmoto K, Kamada T, Kitada Y, Yasui

W, Tahara E and Kajiyama G. (2000). Int. J. Mol. Med., 5,185 ± 190.

Hashiyama M, Iwama A, Ohshiro K, Kurozumi K,Yasunaga K, Shimizu Y, Masuho Y, Matsuda I,Yamaguchi N and Suda T. (1996). Blood, 87, 93 ± 101.

Hassan S, Kinoshita Y, Kawanami C, Kishi K, MatsushimaY, Ohashi A, Funasaka Y, Okada A, Maekawa T, He-YaoW and Chiba T. (1998). Dig. Dis. Sci., 43, 8 ± 14.

Hattori Y, Odagiri H, Nakatani H, Miyagawa K, Naito K,Sakamoto H, Katoh O, Yoshida T, Sugimura T andTerada M. (1990). Proc. Natl. Acad. Sci. USA, 87, 5983 ±5987.

Hattori Y, Itoh H, Uchino S, Hosokawa K, Ochiai A, Ino Y,Ishii H, Sakamoto H, Yamaguchi N, Yanagihara K,Hirohashi S, Sugimura T and Terada M. (1996). Clin.Cancer Res., 2, 1373 ± 1381.

Hirao T, Sawada H, Koyama F, Watanabe A, Yamada Y,Sakaguchi T, Tatsumi M, Fujimoto H, Emoto K,Narikiyo M, Oridate N and Nakano H. (1999). CancerGene Ther., 6, 423 ± 427.

Hirono Y, Tsugawa K, Fushida S, Ninomiya I, YonemuraY, Miyazaki I, Endou Y, Tanaka M and Sasaki T. (1995).Oncology, 52, 182 ± 188.

Hung MC and Lau YK. (1999). Semin. Oncol., 26, 51 ± 59.Hunter T. (1987). Cell, 50, 823 ± 829.Ilekis JV, Stark BC and Scoccia B. (1995).Mol. Reprod. Dev.,

41, 149 ± 156.Ishii H, Hattori Y, Itoh H, Kishi T, Yoshida T, Sakamoto H,

Oh H, Yoshida S, Sugimura T and Terada M. (1994).Cancer Res., 54, 518 ± 522.

Ishii H, Yoshida T, Oh H, Yoshida S and Terada M. (1995).Mol. Cell. Biol., 15, 3664 ± 3671.

Itoh H, Hattori Y, Sakamoto H, Ishii H, Kishi T, Sasaki H,Yoshida T, Koono M, Sugimura T and Terada M. (1994).Cancer Res., 54, 3237 ± 3241.

Jaye M, Schlessinger J and Dionne CA. (1992). Biochim.Biophys. Acta., 1135, 185 ± 199.

Jonjic N, Kovac K, Krasevic M, Valkovic T, Ernjak N, SassoF and Melato M. (1997). Anticancer Res., 17, 3883 ± 3888.

Jung T, Wagner K, Neumann C and Heusser CH. (1999).Eur. J. Immunol., 29, 864 ± 871.

Kaji M, Yonemura Y, Harada S, Liu X, Terada I andYamamoto H. (1996). Cancer Gene Ther., 3, 393 ± 404.

Kameda T, Yasui W, Yoshida K, Tsujino T, Nakayama H,Ito M, Ito H and Tahara E. (1990). Cancer Res., 50, 8002 ±8009.

Kameda T, Yasui W, Tsujino T, Nakayama H, Ayhan A, ItoH and Tahara E. (1992). Pathol. Res. Pract., 188, 37 ± 43.

Kang JJ, Kaysen GA, Jones Jr H and Holland MJ. (1997).Methods, 13, 437 ± 443.

Karameris A, Kanavaros P, Aninos D, Gorgoulis V, MikouG, Rokas T, Niotis M and Kalogeropoulos N. (1993).Pathol. Res. Pract., 189, 133 ± 137.

Kataoka H, Joh T, Kasugai K, Okayama N, Moriyama A,Asai K and Kato T. (1998). Life Sci., 63, 553 ± 564.

Katoh M, Hattori Y, Sasaki H, Tanaka M, Sugano K,Yazaki Y, Sugimura T and Terada M. (1992). Proc. Natl.Acad. Sci. USA, 89, 2960 ± 2964.

Katoh M, Yazaki Y, Sugimura T and Terada M. (1993).Biochem. Biophys. Res. Commun., 192, 1189 ± 1197.

Kim SJ, Sasaki H, Takahashi A, Katoh M, Kakizoe T, KimYK, Sugimura T and Terada M. (1995). Biochem. Biophys.Res. Commun., 214, 60 ± 68.

Kishi T, Yoshida T and Terada M. (1994). Biochem. Biophys.Res. Commun., 202, 1387 ± 1394.

Kiyama T, Onda M, Tokunaga A, Fujita I, Okuda T,Mizutani T, Yoshiyuki T, Shimizu Y, Nishi K, MatsukuraN, Tanaka N, Todome Y, Ohkuni H and Asano G. (1992).Gastroenterol. Jpn., 27, 459 ± 465.

Kiyokawa E, Takai S, Tanaka M, Iwase T, Suzuki M, XiangYY, Naito Y, Yamada K, Sugimura H and Kino I. (1994).Cancer Res., 54, 3645 ± 3650.

Koizumi H, Morita M, Mikami S, Shibayama E andUchikoshi T. (1998). Pathol. Int., 48, 93 ± 101.

Kono K, Naganuma H, Sekikawa T, Ameiya H, TakahashiA, Iizuka H and Matsumoto Y. (2000). Tumour Biol., 21,139 ± 144.

Kuniyasu H, Yasui W, Kitadai Y, Yokozaki H, Ito H andTahara E. (1992). Biochem. Biophys. Res. Commun., 189,227 ± 232.

Kuniyasu H, Yasui W, Yokozaki H, Kitadai Y and TaharaE. (1993). Int. J. Cancer, 55, 72 ± 75.

Kuniyasu H, Yasui W, Yokozaki H and Tahara E. (2000).Langenbecks Arch. Surg., 385, 69 ± 74.

Kwong KY and Hung MC. (1998). Mol. Carcinog., 23, 62 ±68.

Lee H and Maihle NJ. (1998). Oncogene, 16, 3243 ± 3252.Lee HR, Kim JH, Uhm HD, Ahn JB, Rha SY, Cho JY, Lee

JI, Lee KH, Chung HC, Roh JK, Min JS, Lee KS, ShinDH, Kim BS, Hong SW and Choi JH. (1996). Oncology,53, 192 ± 197.

Lin JS, Lu CW, Huang CJ, Wu PF, Robinson D, Kung HJ,Chi CW, Wu CW, Yang WK, Whang-Peng JJ and LinWC. (1998). J. Biomed. Sci., 5, 101 ± 110.

Lin JT, Wu MS, Shun CT, Lee WJ, Sheu JC and Wang TH.(1995). Cancer Res., 55, 1428 ± 1430.

Lin WC, Li AF, Chi CW, Chung WW, Huang CL, Lui WY,Kung HJ and Wu CW. (1999). Clin. Cancer Res., 5, 1745 ±1751.

Liu SI, Lui WY, Mok KT, Wu CW and Chi CW. (1997).Anticancer Res., 17, 3575 ± 3580.

Livingstone JI, Yasui W, Tahara E andWastell C. (1995). Br.J. Cancer, 72, 976 ± 980.

Marshall CJ. (1995). Cell, 80, 179 ± 185.Masaki T, Hatanaka Y, Nishioka M, Tokuda M, Shiratori

Y, Reginfo W and Omata M. (2000). Am. J. Gastroenter-ol., 95, 2135 ± 2136.

McCulloch PG, Ochiai A, O'Dowd GM, Nash JR, Sasako Mand Hirohashi S. (1995). Cancer, 75, 920 ± 925.

McCulloch P, Taggart T, Ochiai A, O'Dowd G, Nash J andSasako M. (1997). Eur. J. Surg. Oncol., 23, 304 ± 309.

Meyer-Ter-Vehn T, Covacci A, Kist M and Pahl HL. (2000).J. Biol. Chem., 275, 16064 ± 16072.

Miki T, Bottaro DP, Fleming TP, Smith CL, Burgess WH,Chan AM and Aaronson SA. (1992). Proc. Natl. Acad.Sci. USA, 89, 246 ± 250.

Ming S-C. (1998). Gastric Cancer, 1, 31 ± 50.Moch H, Schraml P, Bubendorf L, Mirlacher M, Kononen J,

Gasser T, Mihatsch MJ, Kallioniemi OP and Sauter G.(1999). Am. J. Pathol., 154, 981 ± 986.

Molina R, Jo J, Filella X, Bruix J, Castells A, Hague M andBallesta AM. (1997). Tumour Biol., 18, 188 ± 196.

Motojima K, Furui J, Kohara N, Izawa K, Kanematsu T andShiku H. (1994). Surgery, 115, 349 ± 354.

Oncogene


W-c Lin et al

5687

Nakajima M, Sawada H, Yamada Y, Watanabe A, TatsumiM, Yamashita J, Matsuda M, Sakaguchi T, Hirao T andNakano H. (1999). Cancer, 85, 1894 ± 1902.

Naumann M, Wessler S, Bartsch C, Wieland B, Covacci A,Haas R, and Meyer TF. (1999). J. Biol. Chem., 274,31655 ± 31662.

Neugut AI, Hayek M and Howe G. (1996). Semin. Oncol.,23, 281 ± 291.

Noguchi H, Sakamoto C, Wada K, Akamatsu T, Uchida T,Tatsuguchi A, Matsui H, Fukui H, Fujimori T andKasuga M. (1999). Gastroenterology, 117, 1119 ± 1127.

Okino T, Egami H, Ohmachi H, Takai E, Tamori Y,Nakagawa K, Nakano S, Akagi J, Sakamoto O, Suda Tand Ogawa M. (1999). Int. J. Oncol., 15, 709 ± 714.

O'Neill AJ, Cotter TG, Russell JM and Ga�ney EF. (1997).J. Pathol., 183, 325 ± 329.

Ohnishi Y, Nakamura H, Yoshimura M, Tokuda Y,Iwasawa M, Ueyama Y, Tamaoki N and Shimamura K.(1995). Br. J. Cancer, 71, 969 ± 973.

Ooi A, Kobayashi M, Mai M and Nakanishi I. (1998). Lab.Invest., 78, 345 ± 351.

Orita H, Maehara Y, Emi Y, Kakeji Y, Baba H, Korenaga Dand Sugimachi K. (1997). Hepatogastroenterology, 44,294 ± 298.

Park M, Dean M, Cooper CS, Schmidt M, O'Brien SJ, BlairDG and Vande Woude GF. (1986). Cell, 45, 895 ± 904.

Park WS, Oh RR, Kim YS, Park JY, Shin MS, Lee HK, LeeSH, Yoo NJ and Lee JY. (2000). Apmis, 108, 195 ± 200.

Parkin DM, Larra E and Muir CS. (1988). Int. J. Cancer, 41,184 ± 197.

Perng CL, Lin HJ and Lee SD. (1994). Chung. Hua. I. Hsueh.Tsa. Chih. (Taipei), 54, 82 ± 86.

Ries LAG, Eisner MP, Kosary CL, HAnkey BF, MIller BA,Clegg L and Edwards BK. (2000). SEER cancer statisticsreview, 1973 ± 1997. National Cancer Institute, Bethesda,MD, USA.

Robinson D, He F, Pretlow T and Kung HJ. (1996). Proc.Natl. Acad. Sci. USA, 93, 5958 ± 5962.

Rodrigues GA, Naujokas MA and Park M. (1991). Mol.Cell. Biol., 11, 2962 ± 2970.

Romano M, Ricci V, Di Popolo A, Sommi P, Del VecchioBlanco C, Bruni CB, Ventura U, Cover TL, Blaser MJ,Co�ey RJ and Zarrilli R. (1998). J. Clin. Invest., 101,1604 ± 1613.

Saily M, Koistinen P and Savolainen ER. (1999). Ann.Hematol., 78, 173 ± 179.

Santoro E, Carboni M, Catarci M, Carlini M, Carboni F,Zupi G, Vecchione A, D'Agnano I, Giannarelli D,Santoro R and Garofalo A. (1997). Hepatogastroenterol-ogy, 44, 826 ± 837.

Sanz-Ortega J, Steinberg SM, Moro E, Saez M, Lopez JA,Sierra E, Sanz-Esponera J and Merino MJ. (2000). Histo.Hitopathol., 15, 455 ± 462.

Sakurai S, Fukasawa T, Chong JM, Tanaka A andFukayama M. (1999). Jpn. J. Cancer Res., 90, 1321 ± 1328.

Sato TN, Qin Y, Kozak CA and Audus KL. (1993). Proc.Natl. Acad. Sci. USA, 90, 9355 ± 9358.

Sato T, Abe K, Kurose A, Uesugi N, Todoroki T and SasakiK. (1997). Pathol. Int., 47, 179 ± 182.

Sawyer C, Hiles I, Page M, CromptonM and Dean C. (1998).Oncogene, 17, 919 ± 924.

Sayani F, Montero-Julian FA, Ranchin V, Prevost JM,Flavetta S, Zhu W, Woodman RC, Brailly H and BrownCB. (2000). Blood, 95, 461 ± 469.

Schultz SJ and Nigg EA. (1993). Cell. Growth Di�er., 4, 821 ±830.

Scott CD and Weiss J. (2000). J. Cell. Physiol., 182, 62 ± 68.Scott GK, Robles R, Park JW, Montgomery PA, Daniel J,

Holmes WE, Lee J, Keller GA, Li WL, Fendly BM, WoodWI, Shepard HM and Benz CC. (1993). Mol. Cell. Biol.,13, 2247 ± 2257.

Segal ED, Lange C, Covacci A, Tompkins LS and Falkow S.(1997). Proc. Natl. Acad. Sci. USA, 94, 7595 ± 7599.

Seki T, Fujii G, Mori S, Tamaoki N and Shibuya M. (1985).Jpn. J. Cancer Res., 76, 907 ± 910.

Sekiguchi M, Sakakibara K and Fujii G. (1978). Jpn. J. Exp.Med., 48, 61 ± 68.

Seruca R, Suijkerbuijk RF, Gartner F, Criado B, Veiga I,Olde-Weghuis D, David L, Castedo S and Sobrinho-Simoes M. (1995). Cancer Genet. Cytogenet., 82, 140 ± 145.

Shimizu Y, Weidmann E, Iwatsuki S, Herberman RB andWhiteside TL. (1991). Cancer Res., 51, 6153 ± 6162.

Slesak B, Harlozinska A, Porebska I, Bojarowski T,Lapinska J, Rzeszutko M and Wojnar A. (1998). Anti-cancer Res., 18, 2727 ± 2732.

Stebbing J, Copson E and O'Reilly S. (2000). Cancer Treat.Rev., 26, 287 ± 290.

Suzuki S, Uno S, Fukuda Y, Aoki Y, Masuko T andHashimoto Y. (1995). Br. J. Cancer, 72, 663 ± 668.

Tahara E. (1995a). Cancer, 75, 1410 ± 1417.Tahara E. (1995b). World J. Surg., 19, 484 ± 488.Takahashi T, Shirasawa T, Miyake K, Yahagi Y, Maruyama

N, Kasahara N, Kawamura T, Matsumura O, Mitarai Tand Sakai O. (1995). Biochem. Biophys. Res. Commun.,209, 218 ± 226.

Takahashi Y, Cleary KR, Mai M, Kitadai Y, Bucana CDand Ellis LM. (1996). Clin. Cancer Res., 2, 1679 ± 1684.

Takaishi S, Sawada M, Morita Y, Seno H, Fukuzawa H andChiba T. (2000). Biochem. Biophys. Res. Commun., 267,658 ± 662.

Takeshima E, Hamaguchi M, Watanabe T, Akiyama S,Kataoka M, Ohnishi Y, Xiao HY, Nagai Y and Takagi H.(1991). Jpn. J. Cancer Res., 82, 1428 ± 1435.

Taki T, Ishikawa H, Imai K, Yachi A and Matsumoto M.(1985). J. Biochem. (Tokyo), 98, 887 ± 895.

Taniguchi K, Yonemura Y, Nojima N, Hirono Y, Fushida S,Fujimura T, Miwa K, Endo Y, Yamamoto H andWatanabe H. (1998). Cancer, 82, 2112 ± 2122.

Tokuda Y, Ohnishi Y, Shimamura K, Iwasawa M,Yoshimura M, Ueyama Y, Tamaoki N, Tajima T andMitomi T. (1996). Br. J. Cancer, 73, 1362 ± 1365.

Tokunaga A, Onda M, Okuda T, Teramoto T, Fujita I,Mizutani T, Kiyama T, Yoshiyuki T, Nishi K andMatsukura N. (1995). Cancer, 75, 1418 ± 1425.

Tsugawa K, Fushida S and Yonemura Y. (1993). Oncology,50, 418 ± 425.

Tsugawa K, Yonemura Y, Hirono Y, Fushida S, Kaji M,Miwa K, Miyazaki I and Yamamoto H. (1998). Oncology,55, 475 ± 481.

Tsujimoto H, Sugihara H, Hagiwara A and Hattori T.(1997). Virchows Arch., 431, 383 ± 389.

Turner JR, Liu L, Fligiel SE, Jaszewski R and MajumdarAP. (2000). Am. J. Physiol. Gastrointest. Liver Physiol.,278, G805 ±G810.

Uchino S, Tsuda H, Maruyama K, Kinoshita T, Sasako M,Saito T, Kobayashi M and Hirohashi S. (1993). Cancer,72, 3179 ± 3184.

Ueda T, Sasaki H, Aoyagi K, Narikiyo M, Tsubosa Y,Kuwahara Y, Sakamoto H, Mafune K, Yoshida T,Makuuchi M and Terada M. (1999a). Biochem. Biophys.Res. Commun., 265, 739 ± 745.

Ueda T, Sasaki H, Kuwahara Y, Nezu M, Shibuya T,Sakamoto H, Ishii H, Yanagihara K, Mafune K,Makuuchi M and Terada M. (1999b). Cancer Res., 59,6080 ± 6086.

Velculescu VE, Zhang L, Velgostein B and Kinzler KW.(1995). Science, 270, 484 ± 487.

Vidgren V, Varis A, Kokkola A, Monni O, Puolakkainen P,Nordling S, Forozan F, Kallioniemi A, Vakkari ML,Kivilaakso E and Knuutila S. (1999). Genes ChromosomesCancer, 24, 24 ± 29.

Wilks AF. (1989). Proc. Natl. Acad. Sci., 86, 1603 ± 1607.


W-c Lin et al

5688

Oncogene

Wu CW, Li AFY, Chi CW, Chung WW, Liu TY, Lui WYand P'Eng FK. (1998). Oncol. Rep., 5, 817 ± 822.

Wu CW, Li AF, Chi CW, Huang CL, Shen KH, Liu WY andLin W. (2000a). Am. J. Pathol., 156, 2007 ± 2015.

Wu CW, Lo SS, Shen KH, Hsieh MC, Lui WY and P'EngFK. (2000b). World J. Surg., 24, 465 ± 472.

Wu MS, Shun CT, Wang HP, Sheu JC, Lee WJ, Wang THand Lin JT. (1997). Gastroenterology, 112, 1457 ± 1465.

Yasui W, Hata J, Yokozaki H, Nakatani H, Ochiai A, Ito Hand Tahara E. (1988). Int. J. Cancer, 41, 211 ± 217.

Yasui W, Yokozaki H, Shimamoto F, Tahara H and TaharaE. (1999). Pathol. Int., 49, 763 ± 774.

Yasui W, Yokozaki H, Fujimoto J, Naka K, Kuniyasu Hand Tahara E. (2000). J. Gastroenterol., 35 (suppl 12),111 ± 115.

Yonemura Y, Sugiyama K, Fujimura T, Kamata T, FushidaS, Yamaguchi A, De Aretxebala X, Miwa K and MiyazakiI. (1989). Oncology, 46, 158 ± 161.

Yonemura Y, Ninomiya I, Yamaguchi A, Fushida S,Kimura H, Ohoyama S, Miyazaki I, Endou Y, TanakaM and Sasaki T. (1991a). Cancer Res., 51, 1034 ± 1038.

Yonemura Y, Sugiyama K, Fushida S, Kamata T, OhoyamaS, Kimura H, Yamaguchi A and Miyazaki I. (1991b).Anal. Cell. Pathol., 3, 343 ± 350.

Yonemura Y, Takamura H, Ninomiya I, Fushida S,Tsugawa K, Kaji M, Nakai Y, Ohoyama S, YamaguchiA and Miyazaki I. (1992). Oncology, 49, 157 ± 161.

Yonemura Y, Endo Y, Fujita H, Fushida S, Ninomiya I,Bandou E, Taniguchi K, Miwa K, Ohoyama S, SugiyamaK and Sasaki T. (1999). Clin. Cancer Res., 5, 1823 ± 1829.

Yoshida K, Tsuda T, Matsumura T, Tsujino T, Hattori T,Ito H and Tahara E. (1989). Virchows Arch. B. Cell.Pathol. Incl. Mol. Pathol., 57, 285 ± 290.

Yoshida K, Kyo E, Tsujino T, Sano T, Niimoto M andTahara E. (1990). Jpn. J. Cancer Res., 81, 43 ± 51.

Oncogene


W-c Lin et al

5689

Documents

Tyrosine kinases and gastric cancer