Animal Models of Stomach Carcinogenesis

Abstract

Although incidences of stomach cancer have decreased over the past several decades, the disease remains an important public health problem. To identify pathological and molecular biochemical mechanisms, various experimental animal models have been established in rats and mice with chemical carcinogens including N-methyl-N′-nitro-N-nitrosoguanidine (MNNG) and N-methyl-N-nitrosourea (MNU). Helicobacter pylori (H. pylori) is one of the most important factors for human stomach disorders, including neoplasia, and the H. pylori-infected and carcinogen-treated Mongolian gerbil (MG) has proven very useful for analyses of underlying processes. The findings with this model support the hypothesis that intestinal metaplasia is important not as a precancerous lesion but rather as a paracancerous condition and that intestinalization of stomach cancer progresses with chronic inflammation. Furthermore, dose-dependent enhancing effects of salt on stomach carcinogenesis could be demonstrated in MGs treated with MNU and H. pylori modifying surface mucous gel layer. H. pylori itself only causes chronic inflammation and acts as a promoter of stomach carcinogenesis in experimental models. Based on the precise pathological diagnosis of stomach lesions such as noncancerous heterotopic proliferative glands (HPG) and adenocarcinomas, a basis for understanding mechanisms of carcinogenesis has been established on which chemoprevention can be modeled.

Introduction

Stomach cancer incidence and mortality have fallen dramatically in the United States and western countries over the past several decades. Nonetheless, it remains a major public health issue as the 4th most common cancer and the 2nd leading cause of cancer death worldwide (Crew and Neugut, 2004). In Japan, stomach cancer remains an important medical problem and its prevention is one of the most important aspects of cancer control strategy, despite its decreasing trend (Tsugane, 2005). Many pathological and biological analyses of stomach carcinomas, including precancerous lesions, have been performed with human samples and experimental animals. In this article, we review the history of stomach carcinoma research from the viewpoint of basic aspects, concentrating especial attention on pathological and biological findings and results for various important modifying factors.

History of Animal Models for Stomach Carcinogenesis

Rat Model

Attempts to experimentally induce stomach cancers in animal were performed by many researchers using several carcinogens such as benzo[a]pyrene, 3-methylcholanthrene, and 2-acethylaminofluorene starting the 1930s (Rusch, 1940; Wilson, 1941; Stewart and Snell, 1958). However, the incidences of experimentally induced stomach cancer were low, and it was only in 1967 that Sugimura and Fujimura were able to report good yields of adenocarcinomas in the glandular stomachs of rats treated with N-methyl-N′-nitro-N-nitrosoguanidine (MNNG) (Sugimura and Fujimura, 1967). In the pyloric mucosa of the rat glandular stomach exposed to MNNG, erosive lesions occur, and disordering of glandular structures and proliferation of pyloric mucosa are observed. Then, atypical glands and stomach cancer cells become detectable, and finally both differentiated and undifferentiated stomach carcinomas mimicking the histological types of human stomach cancer are induced in this model. The presence of surfactants, such as alkylbenzenesulfonate enhances the effects of carcinogens in the stomach of animals (Takahashi 1969, 1970).

Dog Model

Administration of MNNG to dogs in the drinking water was also found to result in a high incidence of tumors in the glandular stomach (Sugimura et al., 1971). With these animals, it was possible to perform endoscopic observation and take stomach biopsies on a sequential basis. Stomach cancers also occur in dogs given N-ethyl-N′-nitro-N-nitrosoguanidine (ENNG) with yields, however, being lower than with MNNG (Kurihara et al., 1974; Sugimura and Kawachi, 1976).

Mouse Model

The glandular stomach of mice has generally been found to be relatively resistant to MNNG action. Administration of MNNG in the drinking water to BRSUNT/NJms mice over the life span only resulted in adenomatous hyperplasia of stomach epithelium (Sugimura, 1973). Oral administration of 4-nitroquinoline 1-oxide (4-NQO) and 4-hydroxyaminoquinoline 1-oxide (4-HAQO) did, however, induce carcinomas in the stomach as well as various other tissues (Mori, 1967; Mori and Ohta, 1967). Then, Tatematsu et al. reported induction of good yields of adenocarcinomas (Figure 1A) in the glandular stomach of BALB/c (Tatematsu et al., 1992b) and C3H (Tatematsu et al., 1993b) mice treated with N-methyl-N-nitrosourea (MNU). The establishment of mouse models opened up new approaches using transgenic and knockout animals.

Mongolian Gerbil Model

Helicobacter pylori (H. pylori) is a major causative factor for stomach disorders and strong epidemiological evidence has accumulated indicating a significant relationship with active chronic gastritis, peptic ulcers, atrophic gastritis, intestinal metaplasia (IM), and malignant lymphoma as well as cancer development (Warren, 1983; Marshall, 1984; Forman et al., 1991; Nomura et al., 1991; Parsonnet et al., 1991; Craanen et al., 1992; Graham et al., 1992; Group, 1993; Parsonnet et al., 1994; Hu et al., 1995; Kuipers et al., 1995; Asaka et al., 1996; Huang et al., 1998). Based on the epidemiological findings, H. pylori was defined as a “definite biological carcinogen” by WHO/IARC in 1994 (IARC, 1994).

Many animals have been successfully infected with human H. pylori to study the pathogenetic background, but none of early models studied proved sufficiently similar to the situation with human H. pylori infection and pathology (Krakowka et al., 1987; Lee et al., 1990; Radin et al., 1990; Karita et al., 1991; Karita et al. 1994; Marchetti et al., 1995; Fox et al., 1996). In 1996, however, Hirayama et al. reported a Mongolian gerbil (MG) model of human H. pylori infection, with the bacteria detectable throughout a 12-month study period (Hirayama et al., 1996). The resultant chronic active gastritis, peptic ulcers, and IM resemble lesions apparent in man. Then, in 1998, Tatematsu et al. (1998) described establishment of an animal model of stomach carcinogenesis using MGs with MNU and MNNG as carcinogens. Furthermore, H. pylori infection was found to increase the incidence of both MNU- and MNNG-induced adenocarcinomas of all histological types in the MG glandular stomach (Figures 1B and 1C) (Sugiyama et al., 1998; Shimizu et al., 1999a, 1999b). This model has proved very useful for the analysis of stomach carcinogenesis.

Preneoplastic Lesions for Stomach Adenocarcinomas

Intestinal Metaplasia (IM)

IM has been extensively studied as a possible premalignant condition in human stomach (Morson, 1955; Stemmermann and Hayashi, 1968; Correa, 1992; You et al., 1993; Yuasa, 2003). However, many questions remain regarding its pathogenesis as well as the actual relationship to stomach cancers. The present widely applied classification of IM includes complete and incomplete types, the former featuring Paneth cells at the bottom of the glands (Kawachi et al., 1974; Matsukura et al., 1980). This classification is generally accepted, but is only based on intestinal properties (Tatematsu et al., 2003). We have, therefore, proposed a new IM classification based upon the cell differentiation status using both gastric and intestinal cell phenotypic markers, which take into account the gastric properties that are still preserved in association (Inada et al., 1997). Division is into 2 major types; a gastric-and-intestinal-mixed (GI) type, and a solely intestinal (I) type, which can be also applied to animal stomach lesions (Table 1).

Gastric and Intestinal Phenotypic Markers

With recent developments in mucin histochemistry and immunohistochemistry, intestinal metaplastic cells can now be readily classified into a gastric epithelial cell type, encompassing pyloric gland cells and surface mucous cells, and an intestinal epithelial cell type, like goblet and intestinal absorptive cells, on analysis of phenotypic expression, not only in humans but also in rodents (Tatematsu et al., 2003) (Table 2). Concerning gastric phenotypic markers, the surface mucous cell type contains galactose oxidase-Schiff (GOS) and sialidase-GOS reactive mucins, also being positive for MUC5AC and human gastric mucin (HGM). Cells of pyloric gland cell type contain class III mucin, positive for MUC6, and show pepsinogen reactivity. Regarding intestinal epithelial markers, the goblet cell type contains mucin that is GOS-negative and sialidase-GOS reactive, and exhibits sialyl-Tn antigen and small intestinal mucinous antigen (SIMA) as well as MUC2 core protein. Cells of intestinal absorptive cell type demonstrate sucrase and intestinal-type alkaline phosphatase activity (I-ALP), harboring CD10 as a surface marker and the structural protein villin.

Transcription Factors

Homeobox genes play an important role in developing and maintaining organ differentiation. Caudal-type homeobox gene (Cdx) 1 and Cdx2, mammalian members of the caudal-related homeobox gene family, are believed to be important in the early differentiation and maintenance of intestinal epithelial cells (Mallo et al., 1997; Soubeyran et al., 1999; Silberg et al., 2000, 2002). Nuclear staining of Cdx1 and Cdx2 can be detected in intestinal metaplastic cells, but not in normal gastric epithelial cells (Silberg et al., 1997; Mizoshita et al., 2001; Almeida et al., 2003; Mizoshita et al., 2004b). Cdx2 ectopic expression has been shown to induce IM in the glandular stomach of transgenic mice (Eda et al., 2002; Mutoh et al., 2002; Silberg et al., 2002).

With regard to the gastric phenotype, the cSox2 gene, a member of the transcription factor family containing an Srylike high mobility group (HMG) box, demonstrates localized expression in the chicken stomach (Ishii et al., 1998). Sox2 in fact may be a key molecule for gastric differentiation in the gastrointestinal tract, also in mammals (Yasugi, 2000). In humans, Sox2 is found localized in the nuclei of gastric foveolar cells and is decreased in IM (Tsukamoto et al., 2004).

Pancreatic-duodenal homeobox 1 (PDX1), a ParaHox gene that contributes to the genesis and development of the pancreas, duodenum, and antrum, has also been found to be frequently expressed in pseudopyloric glands and IM. MUC6 is more abundant than MUC5AC in pseudopyloric glands, while higher levels of MUC5AC than MUC6 are evident in IM. In carcinomas, PDX1 expression is closely associated with MUC6, whereas no link is apparent between PDX1 and MUC5AC reactivity. Thus, PDX1 may play an important role in the development of pseudopyloric glands and subsequent IM (Sakai et al., 2004; Faller and Kirchner, 2005).

Phenotypic Shift of Gastric Mucosa Toward GI and I Type IM

Experimentally, a phenotypic shift from GI-IM to I-IM could be clearly observed on sequential observation of rat stomachs treated with X-rays (Yuasa et al., 2002). Heterotopic proliferative glands (HPGs) frequently develop with H. pylori infection in the glandular stomach of infected MGs, with slightly dysplastic change of constituent cells (Figure 2) (Nozaki et al., 2002b). While they often resemble differentiated carcinomas, especially mucinous adenocarcinomas, they are not malignant in character. HPGs also show a phenotypic shift from G type to GI type or I type with appearance of Paneth cells during the overall course of H. pylori infection (Figure 3) (Nozaki et al., 2002b).

Pepsinogen-Altered Pyloric Glands (PAPG)

Normal rat glandular stomach possesses pepsinogen (Pg) isozymes, Pg1, Pg2, Pg3, and Pg4. Pg1 among the 3 isozymes (Pg1, 3 and 4) produced in the pyloric mucosa (Furihata et al., 1973, 1980) disappears or preferentially decreases during the early stages of MNNG-induced rat glandular stomach carcinogenesis before morphologically distinct changes appear (Furihata et al., 1975; Tatematsu et al., 1980b). This altered pepsinogen isozyme pattern has been also consistently observed in stomach tumors (Tatematsu et al., 1977). Immunohistochemical studies have demonstrated individual pyloric glands low in Pg1 (thus termed Pg1 altered pyloric glands, PAPG) in normal appearing pyloric mucosa after MNNG-treatment in a dose-dependent fashion (Tatematsu et al., 1986b, 1987) and the susceptibility of different rat strains to induction of stomach carcinomas by MNNG also correlates with that of PAPG, suggesting a preneoplastic nature (Tatematsu et al., 1988).

Methylation of the Pg1 gene observed early in the carcinogenesis process in fact occurs progressively with tumor development (Tatematsu et al., 1993a). PAPG are also detectable immunohistochemically in MNU-treated mice, suggesting possible general use as a preneoplastic lesion for stomach carcinogenesis (Figure 4) (Yamamoto et al., 1997), independent of the strain (Yamamoto et al., 2002). Thus, PAPG can be regarded as a common change in rodents, acting as a precursor for a variety of adenocarcinoma types (Table 1).

Characteristics and Differentiation of Stomach Cancer Cells

The Histological Classification of Stomach Carcinomas

Human stomach cancers histologically can be divided into 2 major groups, the “intestinal” and “diffuse” types of Lauren (1965), which, respectively, nearly correspond to the “differentiated” and “undifferentiated” types (Nakamura et al., 1968; Sugano et al., 1982; Japanese Gastric Cancer Association, 1998). Although these classifications have been widely used, they are inadequate for studies of histogenesis of stomach carcinomas and phenotype expression at the cellular level, because of the confusion of intestinal phenotypic cancer cells with “diffuse” structure and the presence of a gastric phenotype with the “intestinal” type of Lauren (Tatematsu et al., 2003).

Thus, we tried to classify rodent stomach cancers morphologically into the following: (i) differentiated adenocarcinoma forming glandular (Figure 5A) or papillary structures; (ii) mucinous adenocarcinoma producing mucous lakes, in which carcinoma cells appear to float (Figure 5B); (iii) poorly differentiated lesions, growing solidly or diffusely (Figure 5C, left side); and (iv) signet-ring cell carcinomas possessing foamy mucus-filled cytoplasm and proliferating diffusely, often with submucosal invasion (Figure 5C, right side and Figure 5D).

Gastric and Intestinal Phenotypic Classification of Stomach Cancers

The phenotypic expression of malignant cells is widely thought to resemble that of the tissue of origin. Using gastric and intestinal epithelial cell markers, it is possible to analyze the phenotypic expression of each stomach cancer cell, independent of the histological type (Tatematsu et al., 1986a, 1990a, 1992a, 1997; Koseki et al., 2000; Bamba et al., 2001; Kawachi et al., 2003; Mizoshita et al., 2003, 2004a, 2004b). Stomach cancers comprising epithelial elements presenting only gastric or intestinal phenotypic expression are classified as of gastric (G) or intestinal (I) phenotype, respectively. Those expressing both types are classified as a gastric-and-intestinal-mixed phenotype (GI), while the remainder exhibiting neither are grouped as the null (N) type (Tatematsu et al., 1990a; Yoshikawa et al., 1998; Tajima et al., 2001; Kawachi et al., 2003; Mizoshita et al., 2003).

In the rat glandular stomach, experimentally induced adenocarcinomas consist mainly of G type cancer cells, with I type cancer cells appearing later in larger tumors (Tatematsu et al., 1980a, 1983, 1984, 1990b; Yuasa et al., 1994). In the H. pylori infected MG model, a series of stomach cancers could be divided phenotypically into 21 G, 24 GI, 4 I and 1 N types, with 90.0% of the lesions harboring gastric elements and 56.0% demonstrating intestinal phenotypic expression. All 6 lesions arising in non-infected MG were classified as G type. There was no clear correlation with the presence of IM in surrounding mucosa. It might thus be proposed that IM is a paracancerous phenomenon rather than a premalignant condition. H. pylori infection may trigger intestinalization of both stomach cancers and non-neoplastic mucosa (Figure 6) (Mizoshita et al., 2006), the same phenotypic shift occurring in accordance with increasing depth of invasion in human signet ring cell carcinomas and with progression in differentiated cancers (Tatematsu et al., 1986a, 1992a; Yamachika et al., 1997; Yoshikawa et al., 1998; Bamba et al., 2001). The incidence of gastric cancer cells with intestinal phenotypic expression in early differentiated cases is higher than in undifferentiated cases, suggesting that the former may be more prone to intestinalization (Tatematsu et al., 1992a; Yoshikawa et al., 1998; Mizoshita et al., 2004b).

Modifying factors for Stomach Carcinogenesis (Table 3)

Helicobacter pylori

From epidemiological findings, there is little a room for doubt that H. pylori infection has a “positive correlation” with stomach cancer development (Forman et al., 1991; Nomura et al., 1991; Parsonnet et al., 1991; Asaka et al., 1992; IARC, 1994; Asaka et al., 1996; Uemura et al., 2001). Indeed, 2 reports concluded that H. pylori infection alone could induce well-differentiated adenocarcinomas at very high incidences in the glandular stomach of MGs (Honda et al., 1998; Watanabe et al., 1998). However, our studies based on histopathology showed development of adenoarcinomas if treated with carcinogens but no carcinomas in animals exposed only to H. pylori infection (Figure 7) (Sugiyama et al., 1998; Tatematsu et al., 1998; Shimizu et al., 1999a; Shimizu et al., 1999b; Shimizu et al., 2000). Another study resulted in only one poorly differentiated adenocarcinoma (Hirayama et al., 1999a). It should be stressed that the incidences and histological patterns of the lesions differed greatly among these papers.

After H. pylori infection, glands in the stomach of MGs start to proliferate into the submucosa, disrupting the lamina muscularis mucosa. Thus submucosal heterotopic proliferating glands (HPGs) develop in the glandular stomach of MGs with H. pylori infection alone, often resembling differentiated carcinomas (Nozaki et al., 2002b). The characteristics of the HPGs include: (1) organized polarity of their component cells; (2) differentiation from G type HPGs into I type HPGs with mature Paneth cells; (3) formation of large cystic dilatations containing mucin, often with calcification; (4) shedding of epithelial cells and necrosis at the tips of lesions; (5) high-grade inflammation with infiltration of inflammatory cells; and (6) organized polarity of proliferating zones (Table 4, Figure 7). These characteristics are quite different from those of well-differentiated adenocarcinomas, which are characterized by obvious cellular atypia. After eradication, HPGs become obviously reduced, and stomach lesions in the mucosa also disappear with few remnants. Thus, distinguishing reversible lesions from true neoplasms is necessary in investigating the relationship of H. pylori infection with stomach carcinogenesis in the MG model (Nozaki et al., 2002b). Taking into account all the available data, we conclude that H. pylori is a strong promoter of stomach carcinogenesis rather than an initiator.

To evaluate variation in susceptibility to stomach carcinogenesis in relation to age of acquisition of H. pylori infection, Cao et al. designed an experiment involving inoculation of H. pylori followed by MNU exposure at different time points in the MG lifespan (Cao et al., 2002). Early acquisition of H. pylori significantly increases stomach chemical carcinogenesis with MNU, as compared to the case with later infection, possibly because of differences in host stomach mucosal factors and immunologic responses (Figure 7) (Cao et al., 2002). This would imply that childhood H. pylori infection must not be overlooked in approaches to the prevention of stomach cancer in adult life (Correa et al., 2000; Malaty et al., 2002).

Shimizu et al. have provided direct evidence that H. pylori eradication may be useful as a prevention approach against stomach cancer (Shimizu et al., 2000). In H. pylori-infected MGs treated with MNU, the incidences of stomach cancers after curative treatment for H. pylori were thus significantly lower than without H. pylori eradication. For further evaluation, an experimental model with eradication in the early, middle, late period was studied using H. pylori-infected and MNU-treated MGs (Nozaki et al., 2003). H. pylori infection was found to strongly enhance stomach carcinogenesis initiated with the chemical carcinogen, and following eradication at an early period this effect was effectively reduced (Figure 7). However, after complete clearance of the bacteria, reflux esophagitis often occurs (Labenz et al., 1997), and this side effect is thought to be an important risk factor for esophageal adenocarcinoma development (Naef et al., 1975). Therefore, establishment of criteria for H. pylori eradication is now a top priority.

High-Salt Diet

Salt and salted foods are probable risk factors for stomach cancer, based on evidence from a large number of case-control and ecological studies (Tajima and Tominaga, 1985; Tsugane et al., 1992; Joossens et al., 1996; Kono and Hirohata, 1996). In experimental animals, Tatematsu et al. (1975) found sodium chloride to enhance the carcinogenic effects of MNNG and 4-NQO in the rat glandular stomach. Sodium chloride possibly decreases the viscosity of the gastric mucin and may reduce the protective mucous barrier. When given alone, it has no apparent carcinogenicity in rats, but when administered with MNNG or 4-NQO, it promotes stomach carcinogenesis (Tatematsu et al., 1975) in a dose-dependent fashion (Takahashi et al., 1994). A high concentration of sodium chloride causes initial tissue damage and consequent regenerative cell proliferation (Furihata et al., 1996).

Furthermore, a high-salt diet enhances the effects of H. pylori infection on stomach carcinogenesis, and these 2 factors act synergistically to promote the development of stomach cancers in the MG model (Nozaki et al., 2002a) in a dose-dependent fashion (Figure 7) (Kato et al., 2006). High salt diet upregulates the amount of surface mucous cell mucin, suitable for H. pylori colonization, despite the lack of any increment in MUC5AC mRNA, while H. pylori infection itself has an opposite effect, stimulating transcription of MUC6 and increasing the amount of gland mucous cell mucin (GMCM). A high salt diet, in turn, decreases the amount of GMCM, which acts against H. pylori infection (Kato et al., 2006). The available data from experimental animal models clearly supports the concept that salt-preserved foods and salt itself increase the risk of stomach cancer in man.

Cyclooxygenase-2 (COX-2) Inhibitors

The effects of the selective COX-2 inhibitor, etodolac, on MNU initiated and H. pylori-infected stomach carcinogenesis have been investigated in MGs, dose-dependent inhibition being observed with effective suppression at a dose of 30mg/kg/day. Etodolac did not affect the extent of inflammatory cell infiltration or oxidative DNA damage, but significantly inhibited mucosal cell proliferation and development of IM. COX-2 may thus be a key molecule in H. pylori associated stomach carcinogenesis (Figure 7) (Magari et al., 2005). A mouse model has been also used to assess effect of nimesulide, another COX-2 inhibitor, which substantially reduced H. pylori-associated stomach tumorigenesis with induction of apoptosis (Nam et al., 2004). In contrast, COX-2 and microsomal prostaglandin E synthase (mPGES)-1 double transgenic mice developed metaplasia, hyperplasia, and tumorous growths in the glandular stomach with heavy macrophage infiltration upon Helicobacter infection (Oshima et al., 2004), through a tumor necrosis factor-α (TNF-α) dependent pathway (Oshima et al., 2005). Thus, COX-2 activity may be a prime target for chemoprevention of stomach cancer.

Peroxisome Proliferator-Activated Receptor γ(PPARγ) Ligand

When PPARγwild-type (+/+) and heterozygous-deficient (+/−) mice were administered MNU followed by treatment with a PPARγligand, troglitazone, significant reduction of the incidence of stomach cancer was observed in PPARγ(+/+) mice but not in (+/−) mice suggesting a potential chemopreventive effect (Lu et al., 2005).

Oncogenes and Tumor Suppressor Genes in Rodent Stomach Cancers

p53 Tumor Suppressor Gene

Mutations of the p53 tumor suppressor gene constitute 1 of the most frequent molecular changes in a wide variety of human cancers but Hirayama et al. reported that MNNG induced rat stomach adenocarcinomas had a mutation of the p53 gene, at the second position of codon 171 (Val –> Glu), in only one of 10 cases (Hirayama et al., 1999b). Furthermore, Furihata et al. found no mutations in exons 5, 6, 7, and 8 in a total of 30 stomach tumors in MNU treated BALB/c mice (Furihata et al., 1997). We have investigated the susceptibility of p53 nullizygote (−/−), heterozygote (+/−) and wild-type (+/+) mice to MNU stomach carcinogenesis and in a 15-week experiment, adenomas and a well-differentiated adenocarcinoma were observed in p53 (−/−) animals. However, after 40 weeks, there were no significant difference in the incidences of stomach tumors between p53 (+/+) and (+/−) mice and there were no mutations in p53 genes (Ohgaki et al., 2000; Yamamoto et al., 2000). p53 may not be a direct target of chemical carcinogens but rather play an important role as a gatekeeper in rodent stomach carcinogenesis (Tsukamoto et al., 2005).

Ras Oncogenes

Ras is one of the most frequently activated oncogenes in human cancers. However, none of 10 stomach cancers had mutations in codons 12, 13, or 61 of Ki-ras gene in one rat series (Hirayama et al., 1999b). In mouse, one mutation of GGT to AGT at K-ras codon 12 was found in an MNU induced adenocarcinoma from a total of 19 specimens (Furihata et al., 1997). Moreover, no mutations were detected in H- and N-ras oncogenes at exons 1 (codons 12 and 13) and 2 (codon 61) in a total of 26 specimens (Furihata et al., 1997).

β-Catenin Oncogene

Aberrant Wnt/β-catenin signaling caused by mutations in exon 3 of the β-catenin gene has been identified in a number of human malignancies, including stomach cancers. Among 22 MNNG-induced rat differentiated adenocarcinomas, 4 tumors (18.2%) contained dysplastic regions that displayed nuclear β-catenin localization featuring mutations in exon 3, at glycine 34, threonine 41, and serine 45, which affected phosphorylation sites. β-Catenin mutations appear to be associated with the late progression stage of adenocarcinoma development in rat stomach carcinogenesis (Tsukamoto et al., 2003). Nuclear localization of β-catenin in stomach cancers is more frequently observed in mouse cases (Takasu et al., manuscript in preparation) (Figure 8).

In MGs, examination of 45 stomach adenocarcinomas induced with MNU plus H. pylori infection and 7 induced with MNU alone revealed only 1 stomach cancer in the MNU + H. pylori group (2.2%) to display nuclear β-catenin localization with a missense mutation at codon 34, 3 showing cytoplasmic distribution in local regions, and 41 demonstrating cell membrane localization. Tumors induced by MNU alone showed only membranous β-catenin localization (7/7) (Cao et al., 2004).

Other Gene Alterations

Hirayama et al. reported no amplification of K-sam or c-erbB-2 in 7 rat adenocarcinomas (Hirayama et al., 1999b) and no microsatellite alterations in 12 loci in nine cases (Hirayama et al., 1999b).

Conclusions

Many animal models for stomach cancers have been established using rats, dogs, and mice. However, since the discovery of H. pylori, the MG has become one of the most important model animals for analysis of stomach carcinogenesis and trials of chemoprevention. It should be noted that H. pylori itself only causes chronic inflammation and acts as promoter in stomach carcinogenesis, as revealed by the experimental models described above. With a firm base of precise pathological diagnosis of stomach lesions such as HPGs and adenocarcinomas, further analyses can now be conducted to determine mechanisms of carcinogenesis and contribute to chemopreventive methods.