WO2004084815A2 - Virulence genes of h. pylori in gastric cancer - Google Patents

Abstract

Gastric intestinal metaplasia and gastric cancer are associated with Helicobacter pylori, but the bacterium often is undetectable in these lesions. To unravel this apparent paradox, intestinal metaplasia, H. pylori presence and the expression of H. pylori virulence genes were quantified concurrently using histology, in situ hybridization and immunohistochemistry. H. pylori was detected inside metaplastic, dysplastic, and neoplastic epithelial cells, and cagA and babA2 expression was co-localized. Expression of cagA was significantly higher in patients with IM and adenocarcinoma than in controls. The preneoplastic 'acidic' MUC2 mucin was detected only in the presence of H. pylori, and MUC2 expression was higher in 1M, dysplasia and cancer. These findings are compatible with the view that all stages of gastric carcinogenesis are fostered by persistent intracellular expression of H. pylori virulence genes, and especially cagA, inside MUC2-producing precancerous gastric cells and pleomorphic cancer cells. The invention further relates to virulence genes of H. pylori such as, but not limited to, an invA or invE gene, portions thereof, and sequences that hybridize or 15 are complementary with, the invA or invE gene. The gene may be isolated or produced recombinantly and used to create expressed product for the treatment and/or prevention of H. pylori infection.

Description

VIRULENCE GENES OF H. PYLORI IN GASTRIC CANCER Rights in the Invention

This invention was made, in part, with United States government support under grant numbers R083GM and CA82312-03, both awarded by the National Institutes of Health, and the United States government may have certain rights in the invention. Reference to Related Applications

This application claims priority to U.S. Provisional Application No. 60/455,591 of the same title and filed March 19, 2003, the disclosure of which is hereby entirely incorporated by reference. Background of the Invention

1. Field of the Invention

This invention relates to tools and method for the detection and treatment of gastric cancer and ulcers, in particular, to virulence genes identified in H. pylori.

2. Description of the Background H. pylori is the main cause of distal gastric adenocarcinoma [1 ; 2]. The matched

Odds Ratio (OR) for the association between H. pylori infection and subsequent development of non-cardia gastric cancer is 3.0 [3], but the mechanism of this epidemiological association is unknown. The risk for cancer is greater in subjects infected by virulent strains carrying the cagA gene, the cag pathogenicity island (PAT) [4-6], and or the blood group antigen binding adhesin (babA2) gene that mediates the attachment of the microbe to the outside membrane of gastric epithelial cells [7; 8].

Even before the discovery of the role of H. pylori in gastric carcinogenesis, interpopulation comparisons demonstrated that increased prevalence of precursor lesions was associated with increased incidence of cancer [9]. The main precursor lesion is intestinal metaplasia (IM), defined as focal transformation of normal gastric epithelial cells into absorptive intestinal epithelial cells and goblet cells, and the switch from the expression of gastric-type MUC5AC to intestinal-type MUC2 apomucin. It is now established that H. pylori plays a major role in this preneoplastic transformation [10], but the mechanism of H. pylori involvement is unknown. The interpretation of epidemiological observations is complicated by the fact that many of the precancerous and cancerous biopsies are culture- negative for H. pylori and standard histology stains often do not demonstrate the presence of the bacterium in the lumen of glands with or in cancerous tissues [11; 12]. In addition, the OR for distal gastric cancer rises from 2.1 to 9.6 if tests are made on sera sampled 10-15 years before gastric cancer diagnosis instead of using samples obtained <5 years from diagnosis [3]. These findings have been interpreted as demonstrating that, although gastric carcinogenesis induced by H. pylori continues to progress, the microbe often disappears from the stomach before cancer is established [13].

It was determined the presence of intestinal metaplasia and of H. pylori in gastric biopsies from patients with adenocarcinoma and precancerous lesions using light microscopy, laser confocal microscopy, and transmission electron microscopy. In addition, in situ hybridization (ISΗ) and immunohistochemistry (IΗC) wre used to quantify mRNA and protein expression of bacterial virulence genes and of mucus genes associated with carcinogenesis. Using these highly specific and sensitive methods concurrently, it was observed that H. pylori expressing both cagA and babA2 were present within precancerous and cancerous epithelial cells, that the relative expression of cagA per bacterium was significantly higher in patients with IM and adenocarcinoma, and that MUC2 was expressed only if H. pylori was present. These observations demonstrate that H. pylori can colonize MUC2-producing gastric cells and pleomorphic cancer cells, and that the intracellular persistence of the bacteria may play a role throughout the stages of gastric carcinogenesis. Summary of the Invention

As embodied and broadly described herein, the present invention is directed to tools and method for the detection, treatment and prevention of gastric disorders such as ulcers and cancer, and, in particular, to virulence genes identified in Η. pylori, gene products derived therefrom, and antibodies and vaccines thereto. One embodiment of the invention is directed to isolated or recombinant virulence genes of Η. pylori and homologous and complementary sequences related thereto. Preferably the gene is member of the invA gene family or the invE gene family of an Η. pylori infection of a mammal. Sequences may consist of DNA, RNA or PNA, and may comprise all or only functionally active portions thereof. Another embodiment of the invention is directed to products expressed from Η. pylori virulence genes such as, for example, an invA or invE gene product. Products may be isolated or purified from infected cells and may comprise all or portions thereof such as functionally or antigenically active portions.

Another embodiment of the invention is directed to antibodies that are reactive to products of virulence genes of H. Pylori. Preferably antibodies, which they may be a monoclonal antibodies (e.g. IgA, IgD, IgE, IgG, IgM) or polyclonal antibodies, are specifically reactive to the invA or invE gene product.

Another embodiment of the invention is directed to a vaccine for the treatment or prevention of H pylori infection. The vaccine may be antigenically active portions of the invA or invE gene product, or antibodies directed thereto, and can be administered to patients with a pharmaceutically effective carrier.

Another embodiment of the invention is directed to the treatment or prevention of gastric disorders such as gastric ulcers or gastric cancer, by administering an agent which is effective against an H. pylori infection wherein the agent is reactive against an invA or invE gene product or a portion thereof. Agents include, but are not limited to, antibodies and pharmaceutical agents that functionally inactivate invA or invE protein. Agents may also include pharmaceutically acceptable carriers.

Another embodiment of the invention is directed to kits and methods for the detection of H. pylori in a biological sample. Preferably, detection can distinguish a pathogenic vs. a non-pathogenic infection, as well as determine the potential severity of an infection.

Detection may involve detecting the invA or invE gene (e.g. PCR or RT-PCR), or the invA or invE gene product (e.g. ELISA using antibodies specific to characteristic portions of the invA or invE protein).

Other embodiments and advantages of the invention are set forth in part in the description, which follows, and in part, may be obvious from this description, or may be learned from the practice of the invention.

Description of the Figures

Figure 1. Luminal, intracellular and interstitial colonization of the gastric mucosa by H. pylori in the four groups of patients (bars: 10 μm). Bacteria are visualized: (A) in the lumen of antral gastric glands; (B) in close proximity to the luminal surface of columnar epithelial cells (insert: details of H. pylori); (C) within mucus-secreting cells; (D) in the areolar connective tissue ; (E) within goblet cells; and (F) in pleomorphic cells of gastric adenocarcinoma. Within each group of 4 pictures 1 : Genta stain; 2: Anti-H. pylori IΗC; 3: cagA expression by ISΗ; 4: babA2 expression by ISΗ. Figure 2. A. 1 & 3: dual-fluorescent localization of 16S rRN4-positive H. pylori (in red). In 2, the 16SrRNA- positive bacterial cluster expresses cagA (in green), whereas in 4, there is no co-localization of 16S rRNA and cagA. (bars: 10 μm). B. H. pylori and mucin detection using high iron diamine/Alcian blue/Steiner stain (bars: 10 μm) [23]. High iron diamine stains sulfomucin black, Alcian blue stains sialomucin blue, silver stains H. pylori. 1. Adhesion of one H. pylori to the apical membrane of a goblet cell in an area containing sulfomucin; 2. Four H. pylori sectioned under different angles and surrounded by a clear halo within grey- stained sulfomucin; 3. H. pylori surrounded by an area stained blue only (no sulfomucm) inside a goblet cell in the process of discharging mucus into the gastric lumen (note that the halo surrounding H. pylori stains blue); 4. One H. pylori sectioned transversely and surrounded by a clear halo within gray-stained sulfomucin. C. Apomucin expression in three serial sections of normal and precancerous mucosa by dual-fluorescence ISH (1-4) and immunoperoxidase IHC (5-8) (bars: 10 μm). In normal epithelial cells (1&2 and 5&6), MUC5AC is expressed (1 and 5, arrows) but MUC2 is not (2 and 6). In goblet cells (3 & 4 and 7 & 8), MUC5AC is not expressed (3 and 7) but MUC2 is expressed (4 and 8, arrows). D. MUC5AC and MUC2 expression in three serial sections of adenocarcinoma matrix by dual-fluorescence ISH (1 and 2) and immunoperoxidase IHC (3 and 4) (bars: 10 μm). Pictures 1 and 2 (serial section 2): dual-fluorescent reaction for MUC '5 AC (1, negative) and MUC2 (2, positive, arrow) expression by ISH; Picture 3 (serial section 1): &r x-MUC5AC IHC (negative); 4 (serial section 3): anti-MUC2 IHC (positive, arrow). Figure 3. FISH laser confocal microscopic illustration of H. pylori intracellular localization. On the right of the figure is a pseudo-3D reconstruction of 10 serial 0.3-:m scans recorded by observing a FISH preparation of H. ? /oπ-specific 16SrRNA at 1,000X (bar =10 μm). On the left is a schematic representation of cell contours, and of some of the most prominent intracellular clusters of H. pylori (A: absorptive columnar cell; G: goblet cell). Figure 4. Transmission electron micrographs of a metaplastic gland containing columnar absorptive cells with surface microvilli and associated terminal web. A. Cell containing intra- cytoplasmic H. pylori (arrows) that are not surrounded by specialized cellular interface (XI 7,000). B. Detail of the surface of a cell illustrating the presence of an H. pylori in the gastric gland lumen closely approximated with surface microvilli and associated terminal web. No specialized pedestal-like formation is seen. A low electron density circular spot is present in one of the extremities of the bacteria, suggesting the accumulation of polyphosphate granule (X33,000; bar =1 μm). Figure 5. Goblet cell mucus granules of different density creating a "checkerboard" appearance [29]. The long arrow points at a structure that closely resembles a late-stage coccoidal form of H. pylori [30]. Arrowheads point at three structures resembling flagellar filaments that are emerging from a single pole of the cell and are in close contact with the mucus granules (X24,000; bar =1 μm). Description of the Invention As embodied and broadly described herein, the present invention is directed to tools and method for the detection, treatment and prevention of gastric disorders such as ulcers and cancer, and, in particular, to virulence genes identified in Η. pylori, gene products derived therefrom, and antibodies reactive thereto.

H. pylori is present in the lumen of the stomach of over half of the world population. However, it is pathogenic in only a fraction - perhaps 10-15% - of the infected subjects. The cause of its pathogenicity is unknown, although multiple virulence genes have been described and characterized. H. pylori is mostly intraluminal and adherent to gastric surface epithelial cells, but it also invades gastric epithelial cells and the lamina propria. The vxvA gene family, which has been described in Yersinia, Rickettsia, and Salmonella, ^"3 and the invE gene family play pivotal roles in entry of enterobacteria into cells by interacting with heterodimeric cell adhesion receptors (integrals).

H. pylori invasion of epithelial and immune cells plays a pivotal role in pathogenicity of the bacterium and production of disease. The roles played by expression of the H. pylori invA and invE genes in tissue invasion, in the immune responses of the host, and in the production of disease, were determined. Subjects who have enterobacteria only in the lumen of their intestines (or H. pylori only in the lumen of the stomach) would have no disease, and those with bacterial invasion of the tissues would develop inflammatory bowel disease (EBD), gastro-duodenal ulcers, cancer, or other diseases. If one could prevent penetration of the bacteria into the tissues, one could prevent the inflammatory manifestations that are responsible for disease.

To demonstrate that such associations exist, invA expression in the stomach of patients with and without disease by using molecular methods [in situ hybridization (ISΗ), immunohistochemistry (IΗC), and RT-PCR] was explored. The presence of a cause and effect was evaluated by: (1) eradicating H. pylori infection in patients with precancerous lesions; and (2) determining appearance of inflammatory and precancerous lesions in animals inoculated with invA⁺, but not with invA^', strains. These observations can be used to justify the use of the invA gene as a target for diagnosis, treatment and immunization against H. pylori-related diseases.

No published report was identified describing the inv gene. By searching GenBank, the sequences of a large number of invA or invE genes were found to have been submitted, and an invA gene was described in the genome of many enterobacteria. The web site http://www.ncbi.nlm.nih.gov/cgi-bin Entrez/tablik?gi^:=128 showed that 547 of the proteins expressed by H. pylori strain 26695 had sequence similarity to proteins of known 3-D structure. Among those proteins figured the invasion protein (InvA or InvE) with the number 15645842. The Protein Data Bank Database (http://www.rcsb.org/pdb/cgi/resultBrowser.cgi) yielded 1CWV, Deposited: 26-Aug-1999.⁴ The sequence of this protein was found on http://www.ncbi.nlm.nih.gov/Genbahk/. and listed as: NP_208020 (invasion protein (inv A) [Ηelicobacter pylori 26695]). This sequence was translated into the sequence of the invA gene. Using GenBank again, the invA gene sequence was found in the two strains that have been fully sequenced, i.e. J99 and 26695 (ΗP1228 according to TIGR nomenclature)⁵' ⁶.

Primers and probes for ISH and RT-PCR were designed and the expression of inv A in gastric tissues is determined. In addition, a monoclonal antibody against Salmonella typhimurium InvA was obtained for Foodbome Zoonoses.⁷ This mAb may be used to detect InvA protein and/or invE protein expression by immunohostochemistry. The H. pylori cagA virulence gene whose expression can be detected in the cytoplasm of pre-neoplastic goblet cells and pleomorphic cells of gastric adenocarcinoma when using concurrently standard light microscopy, ISH, IHC, confocal microscopy, and transmission electron microscopy. The number of intracellular and interstitial H. pylori and cagA expression are increased in IM and cancer compared to gastritis controls (Table 2) although, colonization of the lumen of gastric glands by H. pylori gradually decreases during progression of chronic gastritis to atrophic gastritis, IM, dysplasia, and gastric cancer. The bacterium can be present not only inside normal gastric epithelial cells and in the lamina propria [31-36], but also they can attach to goblet cells [37; 38]. Despite the relevance and importance of H. pylori invasiveness, this bacterial property is rarely considered to play a role in pathogenicity, and its potential involvement in gastric carcinogenesis has not been evaluated. This may be explained by the fact that intracellular and interstitial bacteria are difficult to detect by light microscopy because of their small size (3-5 :m), and because H. pylori staining (e.g. silver stain and Giemsa) also may stain many of the host's cellular organelles (e.g. cell membranes).

To circumvent the lack of specificity of standard light microscopy for intracellular detection of H. pylori, a combination of ISΗ, IΗC, confocal microscopy, and electron microscopy were used concurrently. Although the use of each technique separately might produce false positive reactions, their concurrent use in biopsies obtained in patients with and without IM and adenocarcinoma showed excellent concordance between the results, thus further supporting their validity. ISΗ was used because it permits the localization and relative quantisation of mRNA expression in single cells or in a population of cells. This technique was previously used to demonstrate co-localization of H. j^y/oπ^'-specific 16S rRNA and silver-stained bacteria in the gastric lumen of patients without precancerous lesions [39]. Here, the sensitivity and specificity of H. pylori detection was improved by combining IHC with ISH, by using specific probes for three different genes, and by systematically including negative controls. Using these techniques concurrently, it was possible to detect increased expression of the cagA gene and its protein inside precancerous and cancerous cells compared to normal epithelial cells (Table 2). Although cagA expression has not been previously detected within gastric epithelial cells, coculture of H. pylori and AGS cell was associated with delivery of the 145-kDa CagA protein into gastric adenocarcinoma cell lines by the cag type IV secretion system [40; 41], with resulting intracellular tyrosine phosphorylation, induction of signal transduction pathways [41], activation of mitogen-activated protein kinase cascades and expression of proto-oncogenes [42]. The data indicate that intracellular H. pylori also express cagA and babA2, and are likely to produce in vivo the cellular metabolic transformation shown to occur in vitro [41]. Therefore, gastric neoplastic transformation may be induced and sustained at least in part by intracellular expression of cagA in vivo. This effect plays a role not only during the early stages of gastric carcinogenesis, but also later in this process. The role of intracellular CagA in gastric carcinogenesis is supported by a recent study that showed that 20% of patients with non-cardia gastric cancer are CagA seropositive, even though they are H. pylori seronegative, and that the OR for distal gastric cancer rises from 2.2 for H. pylori+lH. pylori- to 27.7 for CagA+/CagA- [43]. In the present study, dual- fluorescent ISΗ demonstrated that, within the very same biopsy, some 16S rRNA-positive H. pylori expressed cagA (Fig 2-A, 1&2) whereas others did not (Fig 2-A, 3&4), which may reflect the recently reported in vivo occurrence of genetic divergence between subclones of H. pylori [44]. These findings also are compatible with the occurrence of co-infection by both cagA+ and cagA- H. pylori subclones or strains in patients with precancerous lesions, as was shown in most patients with non-ulcer dyspepsia [45; 46]. By exploring the association between H. pylori infection, cagA expression, and IM at the cellular and molecular level, it was confirmed that H. pylori attach to the surface of goblet cells only if sulfomucins were present (Figure 2-B, 1), as previously reported [23; 38]. The microbes were detected within sialomucin- as well as sulfomucin-producing goblet cells (Figure 2B, 2-4). Thus, Alcian-blue-stained sialomucins appear to allow H. pylori entry into goblet cells, but not their adhesion to the cell surface. Transmission electron microscopy showed the presence of structures compatible with H. pylori within the cytoplasm of normal, of metaplastic columnar absorptive cells (Figure 3A), and of goblet cell (Figure 4). These ultrastructural observations confirm and extend the previous reports that H. pylori are present within normal gastric epithelial cells [32; 34; 47; 48]. In addition, taken in the context of observations using light and confocal microscopy (Figures 1 and 2), they confirm that H. pylori can invade gastric metaplastic cells.

These in vivo observations are further supported by the recent demonstration that H. pylori can enter and survive within multivesicular vacuoles of gastric adenocarcinoma epithelial cells in vitro [49]. The entry process involves specific phagocytic mechanisms that are dependent on host cell determinants, as demonstrated by the observation that entry is more extensive in AGS cells (derived from gastric cancer cells) than in ΗeLa or ΗEp-2 cells (derived from non gastric cancer cells) [50]. Such differences could be ascribed to different expression of dynamin isoforms in the various cell types [51]. An intact f-actin cytoskeleton appears to be required for intracellular penetration in vitro [49] and a similar requirement is likely to occur in vivo. Entry of H. pylori into cultured gastric adenocarcinoma cells also depends on active processes of bacterial motility [52] and on other yet undefined bacterial factors. One of these factors could be the homolog of Yersinia, Rickettsia, Salmonella inv A gene [53-55] that is also expressed in H. pylori [56]. The present study does not show that H. pylori actually divide within gastric goblet cells, but the fact that they express IδSrRNA, cagA, and babA2 indicate that they are alive. This possibility is further supported by the recent in vitro demonstration that the bacteria can survive for up to 48 hours within cytoplasmic vacuoles of gastric adenocarcinoma cells, that they can move within the vacuoles with the help of flagellar activity, and that intravacuolar replication may occur [49]. The observation that H. pylori can be released with mucus into the lumen of metaplastic glands (Figures l.E.2. and 2.B.3.) and sometimes adhere to goblet cells (Figure 2.B.1.) is compatible with the hypothesis that coccoid bacteria can restart growth and division, and then re-enter other metaplastic cells. This process would be similar to the bacterial movement recently observed in vitro [49]. At this time, the mechanism of passage of H. pylori into the lamina propria or directly between cells in infected patients remains unknown.

It was also examined whether the intensity of intracellular and interstitial colonization and expression of virulence genes were related to the stage of gastric carcinogenesis. As shown in Table 2, H. pylori colonization was greater in patients with a history of IM compared to Group V controls, even if precancerous lesions were not present in the biopsy that was examined. Colonization was even greater in neoplastic cells. Expression of cagA, but not babA2, was higher in biopsies with IM compared to biopsies without IM and to controls, and the absolute level of cagA and babA2 expression was greater in high grade dysplasia and adenocarcinoma. In addition, the level of cagA (but not babA2) expression relative to the number of bacteria present in adenocarcinoma was significantly higher than in dysplasia. The clinical relevance of these observations is unclear, but they suggest that gastric carcinogenesis is related to the bacterial load and/or to the intracellular expression of virulence genes, and that high expression of cagA appears directly related to IM and intestinal-type adenocarcinoma, whereas other factors may be responsible for high grade dysplasia. In addition, the observation that babA2 expression remained unchanged in the presence of an increase in H. pylori density may reflect a down regulation of this adhesion molecule when more bacteria invade epithelial cells.

The observation that 6/41 patients had developed gastric dysplasia and/or gastric cancer since the endoscopy at which IM had been diagnosed is consistent with d e precancerous nature of this lesion. The hallmark of gastric IM is the presence of MUC2, an apomucin normally present in the intestine, but not in the stomach [57]. Similar to aberrant mucin expression/structure reported in other pathological conditions including neoplasia [58], expression of MUC2 gene was found to be increased in gastric regenerative, metaplastic and neoplastic epithelia [17], and aberrant MUC2 apomucin synthesis by gastric cells is considered pre-neoplastic [59]. Co-regulation of fucosyltransferase (FUTl, FUT2, or FUT3) determines the differential expression of MUC5AC and MUC6, which is lost in gastric cancer, leading to metaplastic production of MUC2 [60]. In addition, overexpression of MUC2 in colon cancer can be induced by nuclear factor kappaB (NF-6B) and require the presence of Ras, Raf and MEK signaling proteins [61]. These latter factors are persistently overexpressed during H. pylori infection [62]. Here, the MUC2 gene was expressed in goblet cells, as reported [57], but also in normal epithelial cells of H. pylori+ subjects, and that the levels of MUC2 and cagA expression were significantly correlated in biopsies with IM. In contrast, MUC5AC expression was decreased significantly if the patient had a history of IM, which was associated to greater H. pylori colonization and cagA expression (Table 2). In addition, MUC2 expression was increased in the connective tissue and in the extracellular matrix of intestinal type gastric adenocarcinoma as previously reported [63], but there was no detectable MUC5AC expression (Fig 2 D). However, the present data do not show that gastric cancer originates from goblet cells, and the malignant cells may have evolved from other cells. In addition, the number of patients with cancer is relatively low, although the present observations are in agreement with epidemiological data suggesting that the risk of gastric cancer may be increased in the presence of long term and/or heavy colonization by vimlent H. pylori.

The present observations demonstrate that H. pylori is an invasive organism that expresses virulence factors within precancerous and cancerous cells, hi addition, cagA expression within superficial and foveolar mucus-secreting cells is associated with aberrant expression of the MUC2 gene and Alcian blue-positive mucus production. Although a causal relationship between these two findings is not proven by these novel observations, they are compatible with a possible role for intracellular H. pylori in the carcinogenesis process. In addition, they are in agreement with the recent finding that H. pylori eradication can induce regression of cancer precursor lesions [27; 64-66], and may prevent relapse of early gastric carcinoma [2]. They also support the proposed hypothesis that H. pylori continues to play a role in the advanced stages of gastric carcinogenesis [23].

The following examples illustrate embodiments of the invention, but should not be viewed as limiting the scope of the invention. Examples

Patients. H. pylori-X)θsiX ve patients previously diagnosed with IM by gastroscopy (N =43, age 72±1 years, mean ± SEM) were re-endoscoped, and two antral gastric biopsies were harvested. The protocol was approved by the Institutional Review Boards of the VA NY Harbor Health Care System and of USUHS, and written informed consent was obtained from all patients before study entry. Controls consisted of archival biopsies harvested in 8 younger H. pylori-positive (age 57±2 years) patients with no known gastric disease and no prior record of IM (Group V).

Biopsies. Biopsies were immediately fixed in Z-fix (10% paraformaldehyde + 1% ionized Zn, Anatech LTD, Battle Creek, MI), dehydrated, and embedded in paraffin. Serial 5 μm sections were stained with Hematoxylin-Eosin or Genta stain [14] for grading of gastritis and atrophy according to the Sydney system [15] (0 =none, 1 =mild, 2 =moderate, and 3 =marked). In biopsies with high-grade dysplasia or adenocarcinoma, no grading was made. Additional serial sections were prepared for in situ hybridization or immunohistochemistry as described below. For detection of mucin expression, sections # 1 and #3 were processed by IHC for MUC5AC and MUC2, respectively, while section #2 was processed for MUC5AC and MUC2 expression using dual-fluorescence ISH.

Oligonucleotide probes for H. pylori and mucins. (Table 1): Specific RNA and cDNA probes were designed using a sequence database (Genbarik) or, in the case of cagA, babA2, and MUC2, probes were selected as published [8; 16; 17]. All probes were synthesized in the Synthesis and Sequencing Facility, Biomedical Instrumentation Center, USUHS. The 5' end of the oligonucleotides was labeled with either biotin or digoxigenin-3-0-methylcarbonyl-ε- aminocaproic acid-N-hydroxy-succinimide ester (DIG-NHS ester) (Roche Diagnostics, Indianapolis, IN). In initial experiments, both RNA and cDNA probes were used concurrently, and confirmed earlier observations that the two probes can detect the same structures. In addition, it was observed (see controls, below) that detection was abolished by RNase, but not DNase, further demonstrating that these probes were specifically targeting mRNA. Having validated the cDNA antisense probes in this system, it was subsequently included RNase and DNase controls with each run, and discontinued the use of RNA probes, because they are more susceptible to degradation and more delicate and costlier to prepare. In situ hybridization. All procedures were performed as described for paraffin-embedded tissues [18] and at room temperature except as specified.

Section pretreatment: Sections were first deparaffinized in xylene, rehydrated by a series of washes in graded ethanol (95% X 2, 80%, 70% and 50%), and finally washed in DEPC- treated water and PBS. To improve subsequent probe penetration, sections were treated with proteinase K (Roche Diagnostic, Indianapolis, IN; 10 :g/ml in 100 mM Tris-HCl buffer, pH 8.0, containing 5 mM EDTA) for 10 min, and then washed twice for 5 min each with IX standard saline citrate (SSC, Quality Biological, fric, Gaithersburg, MD). Pre-hvbridization: Sections were covered for 5 min with hybridization mixture prepared as follows: 200 :1 Denhardt's solution [2% bovine serum albumin (BSA), 2% Ficoll and 2% polyvinylpyrrolidone (PVP) in water], 5 ml formamide (Roche Diagnostic, Indianapolis, IN), 600 :1 5M NaCl, 2ml dextran sulfate (0.8g/8ml) heated at 65EC for 20 min, 20Φ1 EDTA (0.29g/10ml), 2.5 ml hybridization solution (phenol-chloroform extracted salmon testis DNA and SSC), 20 :1 1.0 M Tris-HCl buffer, pH 7.5 and 660 :1 DEPC-treated water]. Sections were then covered for 20 min with blocking solution [0.2 g BSA fraction V, 0.5 ml normal horse serum (Jackson ImmunoResearch, Inc.), and 10 ml of immuno-Tris buffer (50 ml of IM Tris-HCl buffer pH 7.5 and 25 ml of 5M NaCl in 1,000 ml water, pH 7.5). Hybridization: The probes listed in Table 1 were denatured by heating for 10 min at 65EC and subsequent chilling on ice for 3 min. Each unstained section was layered with 70-100 :1 of the probe solution (4 pmol/:l), covershpped, and placed horizontally on filter paper soaked with water in an air-tight box wrapped with aluminum foil. Separate boxes were used for different probes and for negative controls. Each box was incubated overnight at 37 °C. Post-hybridization: After 18 h of incubation, unbound probe was removed by successive washes in decreasing concentrations of SSC [2X SSC 30 min, IX SSC 10 min, 0.5X SSC 10 min and 0.1 X SSC 15 min at 60°C].

Detection of H. pylori gene expression: 1. Bright-field ISH: Anti-digoxigenin antibody- conjugated or streptavidin-conjugated alkaline phosphatase (Roche Diagnostic, Indianapolis, IN) was used to detect the digoxigenin-labeled (cagA) or the biotin-labeled (babA2) probe, respectively (Table 1). Sections were incubated for 2 hrs with 1:500 dilution of blocking immuno-Tris buffer. Unbound alkaline phosphatase was washed gently with immuno-Tris buffer (3 times, 5 min each), using a different Coplin jar for each probe and negative controls. Bound alkaline phosphatase was detected by covering the slides with a chromogenic substrate (nitro-blue-tetrazolium, NBT/BCIP kit; Vector Labs, Burlingame, CA) and keeping them in the dark for 20 min. After washing in water for 3 min, slides were counterstained with Nuclear Fast Red for 15 min, washed in water, dehydrated, cleared in xylene, air dried and mounted with permount.

2. Fluorescence ISH (FISH) dual localization [19]. The hybridization mixture contained both a biotin-labeled probe recognizing H. pylori 16S rRNA and a digoxigenin-labeled probe recognizing cagA (Table 1). After incubation as described above, the slides were covered with the following three layers: 1^st layer: 1 Φl avidin-Texas red + 1.5 Φl mouse monoclonal anti-digoxigenin; 2^nd layer: 10 :1 biotin-anti-avidin + .1 :1 rabbit anti-mouse-IgG-fluorescein isothiocyanate (FITC) conjugated; 3^rd layer: 1 :1 avidin-Texas red + 10 :1 monoclonal anti- rabbit-FITC conjugated. All antibodies were prepared in 1 ml of blocking solution. After each antibody layer, slides were incubated for 30 min at 37°C in a humid chamber and washed 3 times 5 min each. After air-drying, the sections were mounted in Vectashield (Vector Labs, Burlingame, CA).

Detection of apomucin gene expression: The method described for fluorescence microscopy dual-localization of H. pylori was used in serial section #2. MUC2 and MUC5AC expression was detected using specific probes (Table 1), FITC and Texas red labels, respectively.

Controls: H. pylori pure cultures were streaked onto precleaned microscope slides and used as positive controls. E. coli and S.flexneri cultures and gastric biopsies from a 62 year-old H. pylori-negative patient with dyspepsia were used as negative controls. Control for nonspecific binding was performed by using: 1. sense instead of antisense probe; 2. hybridization buffer instead of antisense probe; 3. unlabeled antisense probe; 4. digoxigenin or biotin-labeled probe for scorpion Buthus martensi Karsch neurotoxin sequence [5'-GGC CAC GCG TCG ACT AGT AC-3' SEQ ID NO 1] [20]; 5. RNase A pretreatment (Roche); 6. DNase I pretreatment (Roche); and 7. RNase + DNase I pretreatment. Immunohistochemistry. IΗC was performed using polyclonal rabbit anti-H. pylori (Biomeda Corp. Foster City, CA), dilution 1:50; anti-H. pylori CagA antigen IgG fraction monoclonal antibody (Austral Biologicals, San Ramon, CA), dilution 1:100; and anti-MUC2 (LUM2-3) or anti-MUC5AC (LUM5-1) polyclonal antibodies (dilution of 1:1,000) raised using keyhole limpet hemocyanin-conjugated peptides with the sequences NGLQPVRVEDPDGC (SEQ ID NO 2)and RNQDQQGPFKMC (SEQ ED NO 3), respectively [21; 22]. Detection of bound primary antibody was performed using the StreptABComplex/ΗRP Duet kit (Dako) followed by incubation with DAB (0.6 mg/ml) in Tris-buffered saline containing 0.03% hydrogen peroxide for 15 min. Sections were counterstained with Harris' (for H. pylori and cagA detection) or Meyer's (for apomucin detection) hematoxylin. Histo chemistry. Alcian blue staining [14] was used to detect the presence of Alcian blue- positive mucins in the gastric mucosa, using sections adjacent to those used for ISH (i.e. section # 4). Additional serial sections were stained with the high iron diamine/Alcian blue/Steiner method [23] to concurrently detect H. pylori and to examine the type of intracellular mucin (sulfo- or sialomucin).

Data analysis. An Eclipse Nikon microscope was used to examine the sections at 100X, 400X and 1,000X. The number of silver-stained bacterial clusters and the mRNA expression was quantified at 400X magnification according to a modification of the point-counting stereological method and using an intraocular reticle of 27 mm diameter, covering 3,578 :m², i.e. 17,892 :m³ for 5 :m-thick sections (No. Kr409, Klarman Rulings, Inc. Litchfield, NH) [24]. Two different morphometric analysis were performed: (1) number of H. pylori clusters identified either as black structures after Genta stain, as cagA and babA2 blue structures after NBT-detecting alkaline phosphatase-ISH, or as IβSrRNA and cagA fluorescent structures during dual fluorescence ISH (N/10,000:m³) and (2) volumetric density of apomucin (Vvi/10,000:m³). Counts were made in 3 randomly selected entire field-of-view. Values were expressed as mean ± SEM and statistical significance was determined using ANOVA. For reproduction purpose, images were digitized using a CCD color camera (Hitachi model HV- C20C20M) interfaced with a Sun X-20 workstation. FISH laser confocal microscopy. The cellular localization of H. pylori was examined in selected 4-:m FISH preparations at 1 ,000X using an MRC600 laser scanning confocal microscope (interfaced with a Zeiss Axiovert 35 Inverted microscope, Bioinstrumentation Center, USUHS). Series often 0.3-:m scans were obtained and 3D images were constructed. Transmission electron microscopy (TEM). [25; 26]. Selected gastric biopsies were fixed overnight in 2.5 % glutaraldehyde in 0.1M phosphate buffer (pH 7.3, 4°C), post-fixed in 1% phosphate-buffered-osmium tetroxide pH 7.3 for 60 min at 4°C, dehydrated in a series of increasing ethanol concentrations, embedded using flat molds in Spurr low viscosity epoxy resin (Polysciences, Warrington, PA), and polymerized at 60°C for 18 h. Semi-thin sections (0.5 :m) stained with toluidine blue were used to identify areas of the gastric biopsy with recognizable H. /^/on-positive metaplastic cells (goblet cells and absorptive cells). Ultra- thin sections (500 Δ) of these areas were cut in a MT 6000 Sorvall ultramicrotome, mounted on 400 mesh cooper hexagonal grids (diameter: 3.05mm) and stained with saturated aqueous solution of uranyl acetate and lead citrate. The grids were examined using a 100 CX JEOL (Peabody, MA) or a CM 100 Phillips electron microscope at 80kv. Based on histological criteria [15], biopsies from patients with prior diagnosis of antral were classified as HVI absent (Group I, N =23, age 72 ± 1 years, mean±SEM); JJVI present (Group II, N=12, 72 ± 3 years); dysplasia (Group HI, N=3, 77 ± 7 years); and differentiated intestinal-type adenocarcinoma (Group TV, N =5, 80 ± 1.5 years). It is important to note that the criterion to select patients in groups I-TV was the presence of IM at a previous endoscopy, and that the absence of DVI in biopsies used for the present study is due to the fact that it is a focal lesion that can be missed in non-targeted endoscopic biopsies. Furthermore, spontaneous regression of IM has not been reported, and it has been observed only in long-term studies after eradication of H pylori associated with administration of micronutrients [27]. Archival biopsies harvested in 8 H. /jy/orz^'-positive patients with no known gastric disease and no prior record of IM served as controls (Group V; age 57±2 years).

H. pylori were present in all biopsies by Genta stain and IΗC. In Groups I and V, H. pylori were observed in the lumen of antral gastric glands (Figure A-l & 2), close to the luminal surface of columnar epithelial cells (Figure B-l & 2), within gastric epithelial cells (figure C-l & 2), and in the areolar connective tissue (Figure D-l & 2). H. pylori were seldom observed free-floating in the lumen of gastric glands with IM, but they occasionally were detected attached to, or within the cytoplasm of, goblet cells in Group II (Figure 1.E and 2.B.), in dysplastic cells in Group III, and in pleomorphic cells of gastric adenocarcinoma in Group rV (figure l.F-1 & 2). cagA and babA2 genes were expressed in the lumen of glands and inside epithelial cells of H. rμ/oπ-positive (but not H. jτj /oπ-negative) biopsies (Figure 1-A to 1-E, pictures 3 and 4), and expression was absent in RNase-treated, but not DNase- treated, sections. The CagA product was detected by IΗC in the same locations. All positive controls were positive, and all negative controls were negative for both cagA and babA2 expression. Dual-fluorescent ISΗ demonstrated that, within the very same biopsy, some 16S rRNA- positive H. pylori expressed cagA (Fig 2-A, 1 & 2) whereas others did not (Fig 2-A 3 & 4), and the same was true for babA2 expression.

Quantification of Genta stains and bright field ISΗ (Table 2, columns 1-3) demonstrated thatH. pylori density and the expression of cagA and babA2 were significantly greater in biopsies from patients with a history of DVI compared to controls, even if no IM was present in the biopsy examined (Groups I and II, vs. Group V). In addition, expression of cagA, but not babA2 was significantly higher in biopsies with IM (Group II vs Group I; p <0.05). In high grade dysplasia (Group HI), H. pylori density was significantly greater than in Groups I and II, and cagA expression was lower than babA2 expression (p <0.05). In adenocarcinoma (Group IV), H. pylori density was also greater, but cagA expression was significantly higher than babA2 expression (p <0.05). In addition, cagA was expressed inside a significantly greater fraction of cells than babA2 (Group II, goblet cells: 30±5 % vs 19±2 %; Group EQ, dysplastic cells: 36±7% vs 16±1%; Group IV, pleomorphic cells: 37±4% vs 15±l%; p <0.05). Similar differences were observed using dual FISH (Table 2, columns 4-5), but

16SrRNA density was generally lower than H. pylori density by Genta stain. This difference could reflect the lack of specificity of silver staining. However, individual values obtained by the two methods were highly correlated (p <0.001), suggesting that only a fraction of silver- stained H. pylori do express 16SrRNA, or that fluorescence faded during the counting of fluorescent clusters. This latter interpretation is supported by the observation that cagA expression was consistently lower when determined by FISH as compared to bright field ISH and that the two were significantly correlated (p <0.0001). In addition, the ratio cagAllβSrRNA was significantly higher in Groups II and IV than in Group I (Table 2, column 6; p <0.05). Interestingly, inflammation score tended to be higher in Group II than in Group I (1.8 ± 0.4 vs 1.0 ± 0.2; N.S.), and was significantly higher than in controls (Group V; 0.4 ± 0.3; p <0.05). Atrophy scores did not differ among groups.

To further investigate the interactions between H. pylori and goblet cells, the high iron diamine/Alcian blue/Steiner stain was used to examine whether the type of intracellular mucin (sulfo- or sialomucin) influenced the presence of H. pylori within those cells. H. pylori did not attach to the surface of goblet cells, except when those cells contained sulfomucins (Figure 2-B, 1), as previously reported [23]. In addition, H. pylori present inside goblet cells were surrounded by either Alcian blue-positive sialomucin granules (Figure 2-B, 2) or by high iron diamine-stained sulfomucins (Figure 2-B, 3 & 4). ISΗ and IΗC performed in subsequent serial sections confirmed that H. pylori were present in those cells. The association between H. pylori infection, cagA expression, and IM at the cellular and molecular level was explored. MUC5AC was expressed in normal epithelial cells (Figure 2-C, 1 & 5) whereas MUC2 expression was very low or absent (Figure 2-C, 2 & 6). The opposite was observed in goblet cells (Figure 2-C, 3 & 7 vs. 4 & 8). In gastric adenocarcinoma, MUC5AC expression was low (Fig 2-D, 1 & 3), whereas MUC2 mRNA and MUC2 apomucin were detected (along with cαg4-positive H. pylori) in the connective tissue and in the extracellular matrix (Fig 2-D, 2 & 4). By mo hometric analysis, MUC5AC expression was similar in patients with no history of gastric disease, whether they were H. rμ/oπ-positive or not (Table 2), and was significantly decreased in all patients with prior history of EM (Table 2, Groups I-IV). MUC2 expression was low in the absence of LM (Groups I and V), and was significantly higher in biopsies with EVI, with dysplasia and with cancer compared to biopsies without EM (Table 2). In addition, MUC2 and cagA expression were significantly correlated in biopsies with EM (r =0.897; p <0.05), but not in those without EM (r =0.315).

Confocal microscopy demonstrated that H. pylori-specific 16SrRNA was expressed within the cytoplasm of normal and metaplastic (Figure 3) gastric mucosal cells.

Transmission electron microscopy also demonstrated the presence of structures compatible with H. pylori in the cytoplasm of normal gastric mucosal cells and of metaplastic columnar absorptive cells (Figure 4A). H. pylori were observed in the lumen of gastric gland and also close approximation with surface microvilli of metaplastic epithelial absorptive cell (Figure 4B). A low electron density circular spot was occasionally visible in one of the extremities of the bacteria (Figure 4B), suggesting the accumulation of polyphosphate granule, as previously described in vitro [28]. Structures suggestive of late stage coccoidal forms of H. pylori were observed within the cytoplasm of goblet cells (Figure 5). These structures were tightly packed between mucin droplets of different density typical of the previously reported "checkerboard appearance" [29], and up to three flagellar filaments attached to one pole of H. pylori were visible between mucin granules (Figure 5). The presence of flagellar filaments have been observed in coccoid forms of H pylori [30].

Apoptosis is a physiological process that is enhanced by intracellular bacteria, perhaps through the expression of bacterial invasin (InvA) protein and of transmembrane receptors such as βl -integrin. H. pylori is known to cause chronic gastritis, peptic ulcer diseases and gastric adenocarcinoma, but the pathogenic role of cell invasion (J Infect Dis 2003;187: 1165- 77) is unknown. The goals of the present study were to determine the expression of H pylori inv A gene and of rhesus monkeys βl -integrin by host cells in relation to apoptosis. Gastroscopies were performed in 18 monkeys naturally infected with H. pylori. Biopsies were cultured or fixed for histology or transmission electron microscopy (TEM). An intraocular grid was used to quantify mRNA expression of 16S rRNA, cagA and inv A by in situ hybridization (ISΗ), βl -integrin expression by immunohistochemistry, and apoptosis by the TUNEL method (confirmed by TEM). Gastritis score was determined using the Sydney System. Infection was cured with quadruple therapy, the animals were re-endoscoped, and repeat gastric biopsies were harvested. In infected animals, 16S rRNA and cagA were expressed by H pylori both in the gastric lumen or when intracellular, whereas invA was expressed only by intracellular H. pylori or by those adherent to surface epithelial cells. Following treatment, gastritis was abolished and cultures became negative for H. pylori. 16S rRNA, cagA and invA expression were markedly suppressed (Table 3), but a few isolated bacteria remained detectable within gastric epithelial cells. Co-localization of these genes by confocal microscopy demonstrated that these bacteria were H. pylori. Concurrent with H. pylori suppression, βl -integrin and apoptosis also were suppressed (Table 3). β 1 -integrin expression was significantly correlated with inv A expression (r =0.84), and apoptosis was significantly correlated with invA and /31-integrin expression (r =0.68 and 0.76, respectively). These observations indicate that apoptosis is mediated by intracellular expression of H. pylori invA gene, perhaps through interaction with the host's βl -integrin.

Other embodiments and uses of the invention will be apparent to those skilled in the art from consideration of the specification and practice of the invention disclosed herein. All references cited herein, including all publications, U.S. and foreign patents and patent applications, are specifically and entirely incorporated by reference. It is intended that the specification and examples be considered exemplary only with the true scope and spirit of the invention indicated by the following claims.

Table 1 Sequences for H. pylori and apomucin DNA probes (RNA probes were identical except for substitution of U for T) 1. H. pylori

16S rRNA: Antisense: 5'-TACCTC TCC CACACTCTAGAATAGTAGTTTCAAATGC-3' (SEQ

ID NO 4) sense: 5'-CTA TGA CGG GTA TCC GGC-3' (SEQ ID NO 5) cagA [16] Antisense: 5'-CTG CAA AAG ATT GTT TGG CAG A-3' (SEQ ID NO 6) sense: 5' GAT AAC AGG CAA GCT TTT GAG G-3' (SEQ ID NO 7)

BabA2 [8] Antisense: 5'-AAT CCA AAA AGG AGA AAA AGT ATG AAA-3' (SEQ ID NO 8) sense: 5'-TGT TAG TGA TTT CGG TGT AGG ACA-3' (SEQ ID NO 9)

2. Apomucins

MUC5AC Antisense: 5*-AGT TGT GCT CGT TGT GGG AGC AGA GGT TGT GC-3' (SEQ ID

NO 10) sense: 5'-CGG GGA CAA GGA AAC CTA-3' (SEQ ID NO 11)

MUC2 Antisense: 5-TCC AAT GGG AAC ATC ACG ATA CAT GGT GGC-3' (SEQ ID NO [17] 12) sense: 5'-CCA TTC TCA ACG ACA ACC CCT ACT ACC CC-3' (SEQ ID NO 13)

Table 2 Count of H. pylori clusters (Genta stain =silver-stained bacteria, cagA and bάbA2 = virulence mRNA expression) and determination of volumetric density (Vvi) of apomucin mRNA expression per field of view (at X400) in gastric biopsies (Mean ± SEM; * p <0.05 vs Group V). ha the absence of H. pylori infection, all values were 0, except for MUC5AC expression that was 296 Vvi/10,000 :m³.

H. pylori (bright field) H. pylori (fluorescence) Ratio cagA/ Apomucin Volumetric (N/l 0,000 :m³) (N/10,000 :m^J) 16SrRNA Density (Vvi/10,000 :m³) (fluorescence)

Genta stain cagA babA2 16SrRNA CagA MUC5AC MUC2

Group I 47.9 ±3.6* 27.3 ±3.7* 26.6 ±2.6* 32.6 ±3.3* 19.4 ±2.0* 0.60 ±0.06* 148 ±19* 80 ±9

(IM absent; n=23)

46.1 ±1.7* 39.8 ±2.5* 27.3 ±5.7* 41.3 ±3.1* 31.9 ±3.1* 0.76 ±0.03* 155 ±8* 260 ±12*

Group π

(TM present; n=12)

60.9±3.2* 21.1 ±1.1* 31.6 ±0.6* 44.3±10.9* 20.6±4.9* 0.50±0.10 46±7* 116±2*

Group m (Dysplasia; n =3)

69.7 ±9.5* 55.2 ±11.9* 15.0 ±3.8* 67.0 ±10.2* 55.2 ±10.3* 0.81 ±0.04* 109 ±42* 336 ±55*

Group IV (Adenocarcinoma; n=5)

12.3 ±1.5 4.5 ±2.5 4.6 ±1.3 5.7 ±1.8 2.3 ±1.0 0.35 ±0.08 481 ± 14 60 ±41

Group V

(No history of IM; n=8)

Claims

Claims

1. An isolated or recombinant virulence gene of H. pylori.

2. A sequence that is complementary to or that hybridize with the gene of claim 1.

3. The gene of claim 1 which is an invA or invE gene or a portion thereof.

4. The gene portion of claim 3 which consists essentially of a conserved region of the invA or invE gene of a plurality of different strains of H. pylori.

5. The promoter of the invA or invE gene of claim 3 which is functionally linked to a gene, that encodes a cytotoxic agent.

6. A protein product, or portion thereof, expressed or derived from the gene of claim 1.

7. The protein or protein portion of claim 6 which is antigenic.

8. A vaccine for the treatment of a gastric disorder comprising the antigenic portion of the protein or protein portion of claim 7.

9. An antibody reactive against the protein product, or portion thereof, of claim 6.

10. The antibody, of claim 9 with an isotype selected from the group consisting of IgA, IgD, IgE, IgG, IgM, and combinations thereof.

11. The antibody of claim 9 which is a polyclonal or a monoclonal antibody.

12. A hybridoma which expresses the monoclonal antibody of claim 11.

13. A vaccine for the treatment of a gastric disorder comprising the antibody of claim 9.

14. A method for the treatment or prevention of a gastric disorder comprising administering to a patient a vaccine comprising an antigenic portion of an invA or invE protein to a patient.

15. The method of claim 14 wherein the gastric disorder is gastric cancer or an ulcer.

16. The method of claim 14 wherein administering comprises delivering an effective amount of an oral, intravenous or intramuscular dose of said vaccine to said patient.

17. The method of claim 16 wherein the effective, amount is that amount which generates an immune response in the patient 18. The method of claim 14 wherein the patient is a human or other mammal.