Introduction

Gastric cancer remains the world’s second commonest malignancy [1]. However, there is substantial variation in gastric cancer incidence with the highest rates reported from Korea, Japan and Eastern Europe, whereas Western Europe and the US have low incidence rates. The global burden of gastric cancer is shifting rapidly from the developed world to the developing world.

Recent evidence indicates that a large number of genetic and epigenetic alterations in oncogenes, tumor suppressor genes and genetic instability as well as telomerase activation determine a multi-step process of gastric carcinogenesis [1, 2]. However, these molecular events found in gastric cancer differ depending upon the two histological types, intestinal and diffuse types of gastric carcinoma, indicating that intestinal and diffuse type carcinomas have distinct genetic pathway [2]. The evolution of intestinal tumors is generally characterized as progressing through a number of sequential steps including gastritis, intestinal metaplasia, dysplasia and carcinoma, while no preceding steps are detected in the pathogenesis of diffuse carcinoma other than the obvious chronic gastritis.

The most significant advance in the pathogenesis of gastric cancer is involved with the recognition of the role of H. pylori as the most important etiological factor for gastric cancer [1]. H. pylori infection is associated with chronic inflammation and multistage process of gastric carcinogenesis [3].

Both the complex host/microbial interaction and growth factor/cytokine networks induced by H. pylori are mediated by the microenvironment that is an indispensable player in the multistage process of gastric carcinogenesis. This review provides an overview of H. pylori-induced inflammation and the molecular and cellular events of tumor microenvironment that underlie gastric cancer.

H. Pylori-Induced Inflammation Implicated in Gastric Carcinogenesis

There are several of H. pylori-virulence associated genes such as cagA, vacA, iceA and babA [1]. Among them, the cagA gene is localized at one end of the cag pathogenicity island (PAI) [4]. The cagPAI encodes components of a type IV secretion system, by which the cagA gene product, CagA, is delivered into gastric epithelial cells [5]. The cagA-positive H. pylori strains are associated with higher grade of gastric inflammation and higher risk of gastric cancer than the cagA negative strains [6]. Prinz et al. [7] reported that CagA+/VacAs1+ strains of H. pylori that are blood group antigen-binding adhesion (BabA2) positive are associated with activity and chronicity of gastritis. Adherence of H. pylori via BabA2 may play a key role for efficient delivery of VacA and CagA. Moreover, Hatakeyama [8] recently reported that CagA binds an Src homology 2 (SH2) containing tyrosine phosphatase SHP-2 in a tyrosine phosphorylation-dependent manner and activates the phosphatase activity of SHP-2. Deregulation of SHP2 by CagA is an important mechanism by which cagA positive H.pylori promotes gastric carcinogenesis. Moreover, Hatakeyama [8] found that East-Asian CagA shows stronger SHP2 binding and greater biological activity than Western CagA. Differences in the biological activity of Western and East-Asian CagA proteins, which are determined by variation in the typrosine phosphorylation sites, may underlie the different incidences of gastric cancer in these two geographic areas.

In addition to deregulation of SHP2 by CagA, H. pylori is a potent activating factor of NF-kB in gastric epithelial cells [9]. Activation of NF-kB by H. pylori infection induces a variety of gene expression including cytokines (IL-1, IL-6. IL-8, TNF alpha), VEGF, COX2, iNOS, cell-cycle regular, MMP2, MMP9 and adhesion molecules [10–12]. Successful eradication of H. pylori leads to down regulation of COX2 in the epithelial and stromal cells [2]. High expression of COX2 mRNA, protein, and enzymatic activity is detected in the tumor cells of gastric cancer [2]. Importantly, COX2 activity is induced by a variety of mediators including inflammatory cytokines such as TNF alpha, interferon-gamma and IL-1 [13]. COX2 facilitates tumor growth by inhibiting apoptosis, maintaining cell proliferation and stimulating angiogenesis within cancer cells [14].

H. pylori infection produces reactive oxygen and nitrogen species that cause DNA damage, followed by chronic gastritis and intestinal metaplasia [3]. Nitric oxide generated by iNOS is converted to reactive nitrogen species that bring about direct DNA mutation such as p53 and protein damage, inhibition of apoptosis, and promotion of angiogenesis [15]. Goto et al. [16] reported that the expression of iNOS and nitro tyrosine in the gastric mucosa was significantly high in H. pylori infected patients who developed gastric cancer at least two years after the initial biopsies.

Factors Associated with Increased Incidence of Gastric Carcinoma

Three major factors, including environmental factors (diet, obesity, cigarette smoking), host factors (H. pylori strains), and genetic factors, cooperatively affect the genesis of gastric cancer [2]. Of these factors, host genetic factors play a critical role in susceptibility to gastric carcinogenesis.

El-Omar [17] reported that pro-inflammatory IL-1 gene cluster polymorphisms increase the risk of gastric cancer and its precursors in the presence of H. pyroli. Moreover, the pro-inflammatory IL-1 genotypes are associated with an increased risk of both intestinal and diffuse types of gastric cancer [18]. In addition, TNF alpha and IL-10 genotypes are viewed as independent risk factors for gastric cancer [18]. The combination of three or four pro-inflammatory cytokine polymorphisms affecting IL-1 beta (IL-1B-511T), IL-1 receptor antagonist (IL-1RN*2), TNF alpha (TNF-A-308A) and IL-10 (IL-10 haplotype ATA) results in a highly significant increased risk of development of gastric cancer [18]. However, we recently found that IL-10 haplotype (IL-10 GGCG) is associated with increased risk of gastric cancer and high levels of plasma IL-10 serum in atomic bomb survivors [19]. More recently, Ohyauchi et al. [20] reported that IL-8 polymorphism is also associated with increased risk of gastric cancer in the Japanese population. These host genetic factors as well as H. pylori strains may determine why some individual infected with H.pylori develop gastric cancer while others do not.

Stem Cell Hyperplasia Induced by H. pylori

H. pylori infection induces stem cell hyperplasia in chronic inflammation leading to increased mutation and DNA methylation. Stem cells in the gastric epithelium have yet to be precisely identified, but there are several proposed markers of normal human stem cells in the gastrointestinal tact including Musashi1 (Msi-1), Hes-1 and nuclear beta-catenin [21]. Msi-1 is an RNA binding protein that may be necessary for stem cell maintenance and/or asymmetric cell division. In human gastric epithelium, Msi-1 is expressed at the proliferative regions of the antrum and fundic glands but its expression is decreased in intestinal metaplasia suggesting that Msi-1 is a marker of gastric progenitor cells [22]. Interestingly, H. pylori infection strongly induces the expression of Msi-1 that is correlated with H. pylori density [23]. Ushijima group [24] recently reported that H. pylori infection potently induces methylation of CpG islands (LOX, HAND1, THBD and p41ARC), suggesting that H. pylori induces DNA methylation in stem cells. Cell division itself is a promoting factor for de novo DNA methylation.

The RNA component of telomerase, hTERC, and the catalytic subunit, hTERT, are essential and the minimum components that are required for telomerase activity [25]. Telomerase activity and TERT expression are necessary for stem cell function [26]. In fact, we found that low levels of telomerase activity exist in Ki-67-positive epithelial stem cells, which simultaneously express hTERT protein and hTERC protein in intestinal metaplasia of the stomach [27]. The levels of hTERT expression and telomerase activity are in parallel with degree of H. pylori infection, suggesting that telomerase competence of stem cell hyperplasia caused by H. pylori infection may be linked to “chronic mitogenesis” that can facilitate “increased mutagenesis”.

We previously reported that telomerase activity is present in a majority of gastric cancer [2]. The approach for cancer therapy targeting telomerase and/or telomeres has the potential for innovative anti-cancer drugs. Telomestatin, a G-quadruplex stabilizing agent found in natural products, is highly selective for telomere G-quadruplex structures formed by the 3’ overhang compared to other compounds [28]. Telomestatin preferentially induces apoptosis in cancer cells or not in normal fibroblasts and epithelial cells [29]. Moreover, Kondo et al. recently reported that loss of heterozygosity and histone hypoacetylation of the PINX1 gene, a possible telomerase inhibitor, are associated with reduced expression in gastric carcinoma [30]. They also revealed that telomerase activity is inhibited with trichostatin (TSA) and nicotinamide (NAM) in gastric cancer cells even if hTERT expression is not changed. These results provide a possibility for cancer therapy with HAM [30].

Currently, a body of evidence indicates that bone marrow-derived stem cells can engraft and differentiate into nonhematopoietic cells of ectodermal, mesodermal and endodermal tissues other than hematopoietic tissues. These stem cells also contribute to cancer stroma in mouse models [21]. Recently, Taketo group [31] found that SMAD4-deficient mouse colorectal cancer cells produce CC chemokine ligand 9 (CCL9) and recruit the MMP-expressing immature myeloid cells (iMC) that express CC chemokine receptor 1 (CCR1) and that lack of CCR1 prevents the accumulation of MMP-expressing iMC at the invasion front and suppresses tumor invasion. These exciting results suggest that human gastric cancers also may have the possibility to exhibit interaction between cancer cells and bone marrow-derived stromal cells implicated in tumor invasion, as majority of human gastric cancer is associated with loss or dysfunction of TGF-beta signaling [2]. It should be examined whether or not bone marrow-derived cells in gastric cancer microenvironment produce MMP and then encourage tumor cells to invade into the stroma. Moreover, gastric cancer originating from bone marrow-derived cells has been reported in mouse models [32]. However, Cogle et al. [33] reported that bone marrow stem cells mimic cancer but do not initiate it in human cancers.

Molecular Mechanisms of Gastric Carcinogenesis

Multiple genetic and epigenetic alterations in oncogenes, tumor suppressor genes, cell cycle regulators, cell adhesion molecules and DNA repair genes, as well as genetic instability and telomerase activation are responsible for tumorigenesis and progression of gastric cancer [1, 2, 31]. Differences exist in the pathways leading to intestinal and diffuse types of gastric carcinoma.

Inactivation of various genes including p16. hMHL1, CDH1, RAR-beta2, pS2 and RUNX3 by DNA methylation is involved in two distinct major genetic pathways of gastric cancer [2, 3, 34]. Hypermethylation of the p16 and of hMLH1 promoters is preferentially associated with intestinal type gastric carcinoma, whereas concordant hypermethylation of the CDH1 and RAR-beta2 promoters is predominantly detected in diffuse type gastric carcinoma. Loss or reduction of RUNX3 and pS2 expression by promoter methylation is a common event in both types of gastric carcinoma. In addition to promoter methylation, acetylated histone H4 is reduced in the majority of both types of gastric carcinoma [2]. Histone H4 is progressively deacetylated from the early stage to the late stage of the multi-step carcinogenesis of the stomach. Moreover, histone H3 acetylation is associated with reduced p21 (WAF1/CIP1) expression by gastric carcinoma [35].

Among these epigenetic events, RUNX3, a Runt domain transcription factor involved in TGF-beta signaling, is silenced in 63% gastric cancer as well as in 73% of hepatocellular carcinoma, 70% of bile duct cancer, 75% of pancreas cancer, 62% of laryngeal cancer, 46% of lung cancer, 25% of breast cancer and 23% of prostate cancer and 5% of colon cancer. Interestingly, inactivation of RUNX 3 by promoter methylation is found in 8% of chronic gastritis, 28% of intestinal metaplasia, and 27% of gastric adenoma, but not in chronic hepatitis B [2]. These findings suggest that RUNX3 is a target for epigenetic gene silencing already at early stage of gastric carcinogenesis.

In the multi-step process of intestinal type gastric carcinoma, genetic instability and hyperplasia of hTERT positive stem cells precede replication error at the D1S191 locus, DNA hypermethylation at the D17S5 locus, pS2 loss, RAR-beta2 loss, RUNX3 loss, abnormal transcripts of CD44 and p53 mutation. All of these changes accumulate in at least 30% of incomplete intestinal metaplasia and are common events in intestinal type gastric cancer. An adenoma to carcinoma sequence is observed on around 20% of gastric adenoma with APC mutations. Molecular events associated with this sequence are loss of heterozygosity and mutation of p53, reduced p27 expression, loss of RUNX3, over-expression of cycline E and abnormal c-met transcription. The resulting advanced intestinal type gastric carcinomas frequently exhibit DCC loss, APC mutation, 1q LOH, loss of p27, reduced TGF beta-receptor expression, reduced nm23 and c-erbB2 gene amplification [1–3].

Meanwhile, diffuse type gastric carcinogenesis involves LOH at chromosome 17p,LOH or mutation of p53, LUNX3 loss and mutation or loss of E-cadherin. Gene amplification of K-sam, c-met and cycline E confers progression and metastasis. Several of the above molecular events may be present in mixed gastric cancer that have both intestinal and diffuse components [1–3].

Meta-analysis of epidemiological studies and animal models shows that both intestinal and diffuse types of gastric cancer are equally associated with H. pylori infection [36, 37]. However, H. pylori infection may play a role only in the initial steps of gastric carcinogenesis. Importantly, younger H. pylori-infected patients have a higher relative risk for gastric cancer than older patients. In general, diffuse type gastric cancers occur mostly in younger patients. Hereditary diffuse gastric cancer caused by germ line E-cadherin mutations is also observed in younger age group [38]. We recently found that a low CagA IgG titer is a useful biomarker to identify a high-risk of gastric cancer and current smoking exhibits significantly higher risk for diffuse type than intestinal type gastric cancers, while radiation risk is significant only for nonsmokers and diffuse type gastric cancers [39]. Differences in H. pylori strain, patient age, exogenous and endogenous carcinogens and host genetic factors may be implicated in two distinct major genetic pathways for gastric carcinogenesis.

Abnormal Growth Factor/Cytokine Network in Gastric Cancer

Gastric cancer cells express a wide array of growth factors and cytokines that act via autocrine, paracrine and juxtacrine mechanisms in the tumor microenvironment [2, 3]. These complex interactions between tumor cells and stromal cells confer morphogenesis, angiogenesis, invasion and metastasis [Fig. 1]. Again the expression of these mediators varies depending on the histological subtype.

Fig. 1
figure 1

Tumor microenvironment of gastric cancer is composed of interaction between cancer cells and stromal cells via growth factor/cytokine network

Full size image

The EGF family including EGF, TGF alpha, cripto and amphiregulin (AR) are commonly overexpressed in intestinal type carcinoma. Meanwhile TGF beta, insulin-like growth factor II (IGF II) and bFGF are predominantly overexpressed in the diffuse type carcinoma. Co-expression of EGF/TGF alpha and EGFreceptor correlates well with the biological malignancy, as these factors induce metalloproteinase [3]. Overexpression of cripto is frequently associated with intestinal metaplasia and gastric adenoma [3]. Akagi et al. [40] reported that gastric cancer cells express neutrophilin-1 (NRP-1), a co-receptor for VEGF receptor 2 endothelial cells. EGF induces both NRP-1 and VEGF expression, suggesting that regulation of NRP-1 expression in gastric cancer is intimately associated with the EGF/EGFR system.

IL-1 alpha is produced not only by activated macrophages but also by gastric cancer cells. It acts as an autocrine growth factor for gastric carcinoma cells and plays a pivotal role as a trigger for the induction of EGF and EGF receptor [2, 3]. IL-6 also acts in an autocrine fashion to stimulate gastric cancer cells. IL-1 alpha and IL-6 both induces the expression of each other by tumor cells. Serum IL-6 in patients with gastric cancer is significantly correlated with tumor stage, depth of tumor invasion, lymphatic invasion, venous invasion and hepatic metastasis, suggesting that IL-6 may be useful for a prognostic factor of gastric cancer [41]. Interestingly, Lin et al. [42] recently found that IL-6 induces gastric cancer cell invasion via activation of the c-Src/RhoA/ROCK signaling pathway. Overexpression of these cytokines, IL-8 and growth factors by the tumor cells may be caused mainly by constitutive NF-kB activation due to constitutive activation of several upstream kinases [12].

More than 80% of primary gastric cancer express IL-8 and IL-8 receptor and this co-expression correlate directly with tumor vasularity and tumor invasion [43]. IL-8 enhances expression of EGF receptor, type IV collagenase (MMP9), VEGF and IL-8 mRNA itself by gastric cancer cells, while reducing E-cadherin mRNA expression [44]. Takehara et al. [45] recently reported that the COX-2 pathway might be also involved in IL-8 production in gastric cancer cells. In addition to IL-8, majority of gastric carcinoma express IL-11 and IL-11 receptor alpha and IL-11 promotes the migration of gastric cancer cells by the activation of the phosphatidylinositol-3 kinase pathway [46]. Moreover, recent studies show that expression of IL-12 and IL-18 also confers progression and metastasis [47, 48].

The negative growth factor TGF beta is frequently overexpressed in gastric carcinoma, particularly in diffuse type carcinoma with productive fibrosis [2]. Hawinkels et al. [49] reported that active TGF beta 1 is present in the tumor cells and fibroblasts, and high tumor TGF beta1 activity are significantly associated with worse survival of the patients. More importantly, recent in vivo and in vitro studies of Mishra group [50, 51] demonstrate that inactivation of ELF/TGF-beta signaling plays a key role in the development of gastric carcinoma. Embryonic Liver Fodrin (ELF) is a beta-spectrin, an adaptor protein that plays an essential role in the propagation of TGF beta signaling. She found that loss of ELF and reduced Smad4 expression are observed in human gastric cancer.

Angiogenic factors such as VEGF, bFGF and IL-8 are produced by tumor cells and results neovacularisation within gastric carcinoma [2]. VEGF promote mainly angiogenesis and progression of intestinal type gastric cancer while bFGF has a stronger association with diffuse type gastric cancer. VEGF-C produced by tumor cells participates in the development of lymph node metastasis.

Stromal cells, especially fibroblasts stimulated by growth factors and cytokines such as IL-1 alpha, TGF alpha and TGF beta secrete hepatocyte growth factor/scatter factor (HGF/SF), which functions in a paracrine manner as a morphogen or motogen [2, 3]. HGF/SF from stromal cells binds to c-met on the tumor cells leading to enhanced progression via activation of c-met pathways. Moreover, keratinocyte growth factor (KGF) produced by gastric fibroblasts binds to KGF receptor, K-sam on tumor cells, resulting in the development of scirrhous type gastric cancer [52].

Osteopontin (OPN), also termed Eta-1 (early T-lymphocyte activation-1), which is a reported protein ligand of CD44, is overexpressed in 73% of gastric cancer [2, 3]. The co-expression of OPN and CD44 v9 in tumor cells correlates with the nodal metastasis in diffuse type gastric cancer. Wu et al. [53] recently reported that elevated plasma OPN level is significantly associated with gastric cancer development, invasive phenotype and survival, suggesting that plasma OPN might have potential usefulness as a diagnostic and prognostic factor for gastric cancer.

Regenerating islet-derived family, member 4 (Reg IV) is expressed in gastric cancer, colon cancer and pancreas cancer [54]. High Reg IV expression is associated with 5-FU resistance in gastric and colon cancer cells by inhibiting the mitochondrial apoptotic pathway, suggesting that expression of Reg IV is a marker for prediction of resistance to 5-FU chemotherapy in patients with gastric cancer [55]. Reg IV also induces phosphorylation of EGF receptor at Tyr992 and expression of Bcl2. Excitingly we found that Reg IV as well as MMP-10 are able to be useful for biomarkers to detect early stage of gastric cancer [56]. Moreover, Interferon inducible gene 6-16, GIP3 is also expressed in gastric cancers and inhibits mitochondrial-mediated apoptosis in gastric cancer cells, suggesting that 6-16 may be a new target for cancer therapy [57].

Wnt-5a is a representative of Wnt proteins that activate the beta-catenin-independent pathway. However, the functions of Wnt-5a in human cancer are controversial. Wnt-5a inhibits proliferation, migration, and invasiveness in thyroid tumor and colorectal cancer cell lines [58, 59], whereas Wnt-5a expression is correlated with cell motility and invasiveness in melanoma cells and breast cancer cells with tumor -associated macrophages [60, 61]. Kurayoshi et al. [62] showed that expression of Wnt-5a is correlated with aggressiveness of gastric cancer by stimulating cell migration and invasion. Wnt-5a activates PKC and JNK, thereby leading to the activation of FAK and paxillin. Wnt-5a and extracellular matrix bind to Frizzled (Wnt-5a receptor) and integrin, and cooperatively activate a signaling cascade to stimulate cell migration. Wnt-5a may be a novel biomarker of prognostic factor and also a therapeutic target for gastric cancer [62].

These events of abnormal growth factor/cytokine networks in the tumor microenvironment will no doubt provide a great deal of information on new approaches for biomarkers and effective therapeutic targeting of gastric cancer.