H. pylori CagL-Y58/E59 Prime Higher Integrin α5β1 in Adverse pH Condition to Enhance Hypochlorhydria Vicious Cycle for Gastric Carcinogenesis

Background/Aims H. pylori CagL amino acid polymorphisms such as Y58/E59 can increase integrin α5β1 expression and gastric cancer risk. Hypochlorhydria during chronic H. pylori infection promotes gastric carcinogenesis. The study test whether CagL-Y58/E59 isolates may regulate integrin α5β1 to translocate CagA via the type IV secretory system even under adverse pH conditions, and whether the integrin α5β1 expression primed by H. pylori is a pH-dependent process involving hypochlorhydria in a vicious cycle to promote gastric carcinogenesis. Methods The expressions of integrin α5 and β1, CagA phosphorylation, IL-8, FAK, EGFR, and AKT activation of AGS cells exposed to CagL-Y58/E59 H. pylori, isogenic mutants, and different H. pylori CagL amino acid replacement mutants under different pH values were determined. Differences in the pepsinogen I/II ratio (indirectly indicating gastric acidity) and gastric integrin α5β1 expression were compared among the 172 H. pylori-infected patients with different cancer risks. Results Even under adversely low pH condition, H. pylori CagL-Y58/E59 still keep active integrin β1 with stronger binding affinity, CagA translocation, IL-8, FAK, EGFR, and AKT activation than the other mutants (p<0.05). The in vitro assay revealed higher priming of integrin α5β1 by H. pylori under elevated pH as hypochlorhydria (p<0.05). In the H. pylori-infected patients, the gastric integrin α5β1 expressions were higher in those with pepsinogen I/II ratio <6 than in those without (p<0.05). Conclusions H. pylori CagL-Y58/E59 prime higher integrin under adverse pH and may involve to enhance hypochlorhydria vicious cycle for gastric carcinogenesis, and thus require an early eradication.


Introduction
Integrins are cell adhesion receptors that can be exploited during bacterial pathogenesis [1,2]. Integrin a5b1 is a gastric epithelial cell receptor which can bind with H. pylori cytotoxinassociated gene L protein (CagL) [3]. Such binding may lead into activation of integrin a5b1 receptors and further facilitate the delivery of oncoprotein cytotoxin-associated gene A protein (CagA) via the type-IV secretion system (T4SS) into gastric epithelial cells [3]. As a consequence of greater H. pylori CagA delivery via T4SS, the H. pylori infection can lead to the progression of gastric carcinogenesis [4][5][6][7], as shown by increased corpus inflammation in an animal study [8], and as linked with the formation of precancerous changes as intestinal metaplasia and even gastric cancers in human [9].
In general, the great majority of mammalian cells that initially have contact with pathogenic microorganisms do not readily present with adequate integrin receptors [10]. Such receptors become exposed to pathogens as a result of damage, generated either directly by such microorganisms or by associated pathophysiological changes to the infectious sites [10]. Therefore, the different abilities of H. pylori isolates and related injuries to change the gastric microenvironment may be involved in regulating the priming of integrin a5b1 for such microorganisms to interact with the target cells.
Our recent study revealed that H. pylori CagL amino acid polymorphisms such as Y58/E59 can exploit higher expressions of integrin a5b1 and gastritis in the upper stomach, and that this was associated with a 4.6-fold increase in the risk of gastric cancer [11]. H. pylori CagL-Y58/E59 isolates lead to higher corpus inflammation and integrin a5b1 expression in the upper stomach, where they commonly exist with the chief cells and acid output secretory cells in the mucosa. It is well known that the human gastric corpus reduces acid secretion after H. pylori infection, and, therefore, corpus inflammation and atrophy are two independent factors for hypoacidity in the stomach [12]. Moreover, hypochlorhydria has been found to increase gastritis and lead to the development of precancerous changes progressing into dysplasia or even gastric carcinoma [13][14][15][16]. It is therefore rational to propose that hypochlorhydria after chronic H. pylori infection may be a micro-environmental factor that regulates the expression of gastric integrin a5b1.
The aim of this study was to determine whether H. pylori isolates such as CagL-Y58/E59 can have a strong priming effect on gastric integrin a5b1, especially in the specific gastric microenvironment with adverse intragastric acidity, to promote gastric carcinogenesis. The findings confirmed that virulent H. pylori CagL-Y58/E59 lead to higher a5b1 integrin priming even under adverse low pH conditions, and that elevation of intragastric acidity during chronic H. pylori infection further primed integrin a5b1 in a vicious cycle to facilitate gastric carcinogenesis.

In vitro and Clinical Study Design
This study used the H. pylori strain (Hp1033) isolated from a patient with gastric cancer at the National Cheng Kung University Hospital, Tainan, Taiwan, and carrying CagL amino acid polymorphism as Y58/E59 [11]. H. pylori cultures were performed as described in previous articles [11,[17][18][19]. Human gastric cancer cell lines as AGS cells (Bioresource Collection and Research Center, BCRC 60102) were cultivated in F12 medium (Gibco, Invitrogen Corporation, Grand Island, NY, USA) supplemented with 10% heat-inactivated fetal bovine serum. H. pylori strains were grown on CDC plates at 37uC in 5% CO 2 for 24 h. AGS cells were seeded in 6-well plate tissue culture dishes (1610 6 /well) and co-cultivated with H. pylori at a multiplicity of infection (MOI) of 100 [20].
For assaying integrin a5b1 priming and CagA phosphorylation by H. pylori at different pH values, human gastric cancer cell lines as AGS cells (Bioresource Collection and Research Center, BCRC 60102) exposed to Hp1033 under the different pH ranges from 4.4, 5.4 to 7.4 in the culture mediums for 16 h. The cell lysates were collected to check the expression of integrin a5 or b1, and phosphorylated CagA by immunoblotting.
Besides in vitro assessments to test whether there were differences in the gastric integrin a5b1 expressions under different gastric acidities, the pepsinogen I/II ratio (to indirectly indicate the gastric acidity) and the gastric integrin a5b1 expression were compared among the 172 H. pylori-infected patients with different gastric cancer risks, including active duodenal ulcers (n = 36), chronic gastritis (n = 52), precancerous lesions as intestinal metaplasia (n = 46), and gastric cancers (n = 38). The institutional review board of National Cheng Kung University Hospital approved the study (certification code: HR-98-023), and each participant provided the written informed consent.

Construction of its cagL Isogenic Mutants and Amino Acid Replacement Mutants
The cagL isogenic mutant was constructed by insertion mutagenesis.

Immunoblotting for Integrin and CagA Phosphorylation
Cells were loaded on to SDS/polyacrylamide gels and blotted on to PVDF membranes. The blots were incubated with antibodies specific for CagA (Austral Biologicals, San Ramon, CA, USA), tyrosine phosphorylation (PY99, Santa Cruz Biotechnology, CA, USA), integrin a5, integrin b1, active form integrin b1, GAPDH, and actin (Chemicon, Temecula, CA, USA). The blots were then incubated with horseradish peroxidase-conjugated secondary antibodies. Images were recorded by enhanced chemiluminescence (ECL, Millipore) reagent with x-ray films. The intensities of the bands were measured with a digital image system (UVP Biospectrum AC Imaging Systems, UVP, Upland, CA, USA) [3,22,23].

Construct & Purify CagL Protein to Assess the Binding Affinity to Integrin a5b1
To validate the binding affinity of CagL to integrin, the study purified the CagL protein from Hp1033. The coding region corresponding to amino acid residues 21-237 of cagL from Hp1033 was engineered by PCR to add an NdeI site upstream of the start of ATG as well as a HindIII site. The dPCR products were cloned into NdeI-HindIII of pET28a (Novagen) carrying an Nterminal His-Tag and a thrombin cleavage site [3]. The resulting plasmids were designated pET28/CagL-YE. The pET28/CagL-DK was created by using site-directed mutagenesis following the protocol of the QuikChange II Site Directed Mutagenesis Kit (Stratagene, La Jolla, CA, USA). All the sequences were verified by DNA sequencing. E. coli BL21 (DE3) was transformed with pET28/CagL-YE and pET28/CagL-DK. Transformants were grown in LB broth to 1 OD 600 , and the target protein was induced by the addition of 0.5 mM (final concentration) IPTG. Following this induction, cells were collected by centrifugation and resuspended in B-PER Bacterial Protein Extraction Reagent (Thermo Fisher Scientific Inc, Rockford, IL, USA) and disrupted with two freeze/thaw cycles. The resulting cell extract was cleared by centrifugation at 23,000 g for 15 min as the soluble fraction to be purified for His-tagged CagL with the use of a B-PERH6xHis Fusion Protein Purification Kit (Thermo Fisher Scientific Inc, Rockford, IL, USA). The binding affinity between CagL-Y58E59 or CagL-D58K59 with integrin a5b1 was measured by a quartz crystal microbalance (QCM) (model ANTQ300, ANT Inc., Taipei, Taiwan) with a resonance frequency of 9 MHz. Human integrin a5b1 (100 mg/ml, Millipore, MA, USA) was coated onto the chip surfaces washed with PBS at a pH of 7.4 or 5.4. The masses deposited onto the QCM heads were calculated from the frequency shifts according to the Sauerbrey equation.

FAK, EGFR and AKT Phosphorylations by Cell-based ELISA
The effects of H. pylori expressed CagL-Y58/E59 or CagL-D58/ K59 on the protein phosphorylations of FAK, EGFR and AKT were analyzed by Fast Activated Cell-based ELISA (FACE TM ) Kits (Active Motif) according to the manufacturer's protocol. A 96well plate was seeded with AGS cells under serum starvation for 24 h, and a quadruplicate time course to incubate AGS cells with Hp1033 CagL-Y58/E59 revertant or Y58D/E59K amino acid replacement mutant was performed. Cells were then formaldehyde fixed in separate wells to incubate with primary antibodies recognizing FAK phosphorylated at tyrosine 397, EGFR phosphorylated at tyrosine 845, AKT phosphorylated at serine 473, total FAK, total EGFR, and total AKT, followed by HRPconjugated secondary antibodies. The colorimetric absorbance was quantified using a plate reader.

Validation of Integrin a5b1 & Pepsinogen I/II Ratio in the H. pylori-infected Patients
Each enrolled patient was free from antisecretory agents for at least two weeks, and without previous history of H. pylori eradication. In each patient, the gastric mucosal biopsies (2 from the antrum and 2 from the corpus in non-cancerous sites) were obtained under gastroscopy for the immunohistochemistry of gastric integrin a5b1 that was performed by monoclonal antibodies of anti-human-integrin a5b1 (Chemicon International, Inc., Temecula, CA, USA) [11]. Each enrolled patient provided serum for pepsinogen I and II check-up by microplate-based quantitative enzyme linked immunosorbent assay (ELISA) using PG I and II kits (Biohit Oyj, Helsinki, Finland), respectively. The serum pepsinogen I/II (PG I/II) ratio, perhaps dropping to ,6, indirectly indicated gastric atrophy with chief cell loss to implicate low intra-gastric acidity in these H. pylori-infected subjects.

Gastric Integrin a5b1 Expressions Assessed by Immunohistochemistry
The same pathologist blinded to patients' background scored the intensity of integrin a5b1, and its specific supranuclear or apical location on the non-intestinal metaplasia gastric epithelium. Combining the positive staining locations on the supranuclear or apical surfaces and the percentage of positive integrin a5b1 staining epithelium cells [11], this study derived a modified intensity of a5b1 integrin score in a range from 1 to 4 (1, , 30% epithelium with positive staining but without apical or supranuclear staining; 2, $30% epithelium with positive staining but without apical or supranuclear staining; 3, , 30% epithelium with positive staining and with apical or supranuclear staining; 4, $30% epithelium with positive staining and with apical or supranuclear staining).

Statistical Analysis
The statistical analysis was performed with SPSS software (SPSS 13, Chicago, IL, USA). The Pearson's x 2 test, one-way ANOVA with Tukey's least significant difference, and Kruskal-Wallis oneway ANOVA by ranks and post hoc comparison by Mann-Whitney U test were used as appropriate. The Student's t-test, Pearson's x 2 test (with odd ration and 95% confidence intervals), and Mann-Whitney U test were conducted as appropriate to identify the statistical differences between the two comparison groups. All of the tests were two-tailed with the statistical significance defined as p,0.05.

Results
Up-regulation of Integrin a5 and CagA Phosphorylation by H. pylori is pH Dependent In the absence of H. pylori infection, there were no differences in the integrin a5 and b1 expressions of AGS cells among different pH values at 7.4, 5.4, or 4.4 ( Figure 1A). However, the integrin a5 expression, but not integrin b1 expression, was significantly increased in the AGS cells exposed to Hp1033 isolates, when the pH of the culture medium increased from 4.4 to 7.4 (p,0.05, Figure 1A). Moreover, as depicted in Figure 1B, the CagA phosphorylation of AGS cells co-cultivated with Hp1033 was lower at pH 4.4 than at either pH 7.4 or pH 5.4 (p = 0.08 or p,0.01, respectively).

H. pylori CagL-Y58/E59 Priming More Active Integrin b1 by Stronger Binding Affinity
To examine the relationship between higher pH values and integrin expression, we further determined the effect of H. pylori CagL amino acid replacement mutants of Y58/E59 on the integrin expression under different pH values. From Figure 2A, it can be seen that there were no differences in the integrin a5 or b1 expressions of AGS co-cultivated with wild-type Hp1033, its cagL isogenic mutant, and different mutants at pH 5.4 (p.0.05). Using immunoblotting with anti-active form and anti-total integrin b1 antibodies, the CagL-Y58/E59 infection preserved with higher active form of integrin b1 in AGS cells than CagL-Y58D/E59K infection (p,0.05, Figure 2B). Under Quartz crystal microbalance, the recombinant CagL-Y58/E59 had a higher affinity with the purified integrin a5b1 than CagL-D58/K59 did (dissociation constant [Kd]: 0.50 vs. 1.76 mM) ( Figure 2C).

The Integrin a5b1 and PG I/II Ratio Expression in the Clinical Patient Groups
As evidenced from the in vitro assays, the decrease of gastric acidity with a higher pH value may up-regulate the integrin a5b1. We further validated that such a finding could be translated to the clinical patients. The integrin a5b1 was in general stained on the basolateral membrane of the gastric epithelial cells in the duodenal ulcer group ( Figure 4A) and gastritis group ( Figure 4B), but could be stained on the supranuclear or apical surfaces in the intestinal metaplasia group ( Figure 4C) and gastric cancer group ( Figure 4D). In Table 1, it can be seen that the rates of patients with PG I/II ratio ,6.0 increased in order in the following ranking: duodenal ulcer (8.3%), gastritis (19.2%), intestinal metaplasia (19.6%), and gastric cancer group (39.5%) (p = 0.01 by linear-by-linear association). The rates of supranuclear or apical integrin a5b1 expression on the gastric epithelium also increased in the same order for patients with duodenal ulcer, gastritis, intestinal metaplasia, and gastric cancer group (respectively for antrum: 22.2%, 34.6%, 56.5%, and 63.2%, p = 0.001; respectively for corpus: 16.7%, 28.8%, 34.8%, and 50%, p = 0.02).
We thus further investigated whether the patients with low PG I/II ratio ,6 as implicated in their lower gastric acidity could have higher gastric integrin a5b1 expression. In Table 2, the rate of supranuclear or apical expression of integrin a5b1 on the gastric epithelium was significantly higher in the H. pylori-infected patients with PG I/II ratio ,6.0 than in those with PG I/II ratio $6.0 on either antrum or corpus (p,0.05). In addition, the modified intensity of the H. pylori-infected patients with PG I/II ratio ,6.0 was higher than that of patients with PG I/II ratio $6.0 in corpus (p,0.05, by Mann-Whitney test).

Discussion
This study demonstrates that integrin expression and CagA translocation induced by H. pylori were up-regulated under a neutral pH condition. H. pylori CagL-Y58/E59, the gastric cancer strain inducing gastric epithelial cells to have greater CagA translocation, IL-8 secretion, as well as to have higher integrin b1, FAK, EGFR and AKT activation than H. pylori CagL-Y58D/ E59K did. Moreover, we showed that CagL-Y58/E59 had a stronger binding affinity to integrin a5b1 as compared to CagL-Y58D/E59K, which was also found to be primed and increased to express at the supranuclear or apical surface of the superficial epithelial cells in patients with precancerous or cancer lesions and extended to deep glands of gastric mucosa in patients with gastric cancers. These findings are particularly important to indicate that integrin a5b1 activation and expression primed by H. pylori CagL amino acid polymorphisms exist in a hypochlorhydria vicious cycle that promotes gastric carcinogenesis.
Integrin a5b1 is a receptor of gastric epithelial cells for H. pylori binding, which increases the risk of gastric cancer [3,8,9]. Because H. pylori CagL-Y58/E59 has a 4.6-fold risk in the development of gastric cancers and exploits higher integrin a5b1 in corpus [9], we tested whether such an isolate exerted a unique priming on gastric integrin a5b1 expression or activation. It can be observed in Figure 2B that while H. pylori CagL-Y58/E59 triggered higher integrin b1 activation than H. pylori CagL-Y58D/E59K did (p,0.05), it did not do so for integrin expression (p.0.05, Figure 2A). This suggests that integrin a5b1 expression might not be primed by H. pylori directly, and that some cytokines released from activated immune cells may be possibly involved to upregulate integrin a5b1 expression [27,28]. It is thus worth further study to determine whether H. pylori CagL-Y58/E59 attracts immune cells and stimulates cytokine secretion to prime integrin expression in gastric epithelial cells. Integrin clustering, which drives integrin b1 activation [29], is a vitally necessary step for CagA translocation [30]. Given that H. pylori carrying the CagL- Y58/E59 infection triggers higher integrin b1 activation, it seems that this interaction may trigger more integrin clustering, and have higher ability to transport CagA. Indeed, H. pylori carrying the CagL-Y58/E59 infection was found to have higher CagA translocation (p,0.05, Figure 2D) and higher IL-8 secretion than H. pylori CagL-Y58D/E59K did (p,0.05, Figure 2E). H. pylori CagL-Y58/E59 infection can thus contribute to more severe gastric inflammations.
The interaction of CagL and integrin a5b1 can activate FAK [3], and also activate metalloprotease ADAM-17 with subsequent increase of the active EGFR [26]. AKT is the downstream of integrin and EGFR, and is activated under the H. pylori cagdependent manner [31][32][33]. AKT has been observed to participate as the regulator of tumorigenesis [34][35][36]. FAK and EGFR can also mediate cell cycle progression and survival via AKT activation [37]. In addition, EGFR activation, resulting from ADAM17 dissociation, has been shown to contribute hypochlorhydria [38]. Here, we demonstrated CagL-Y58/E59 H. pylori induced higher phosphorylation levels of FAK, EGFR, and AKT than CagL-Y58D/E59K H. pylori did (p,0.05, Figure 3). These data thus suggest H. pylori CagL-Y58/E59 is predisposed to lead into hypochlorhydria in clinical settings.
According to the 3D structure model of CagL predicted by the SWISS-MODEL program and by using the TraC structure as template, the solvent-exposed RGD motif is located at the C terminus of the a1 helix [39]. The exposed residue 58 and 59 is also located at the C terminus of the a1 helix, with 16 amino acid residues apart from RGD motif. To test whether a conservative change of CagL from a negatively-charged aspartic acid (D) to tyrosine (Y) at position 58, and from a positively-charged lysine (K) to a negatively-charged glutamic acid (E) at position 58 may affect binding affinity to integrin a5b1, we performed QCM and confirmed that the affinity of CagL-Y58/E59 with integrin a5b1 was stronger than that for CagL-D58/K59 ( Figure 2C). Given this high affinity, there could be more CagA phosphorylation with downstream carcinogenetic effects.
Extracellular pH can be considered as a factor to affect the activation and functioning of integrin [40,41]. Besides to the CagL polymorphisms, the integrin a5 expression of gastric epithelial cells and CagA translocation induced by H. pylori can be pH dependent with higher expression in pH 7.4 than in pH 4.4 (p,0.05, Figure 1). As blockade of integrin a5 expression decreases PI3K/AKT activity and tumor invasion [42], it is rational to see the increase in integrin a5 expression under hypochlorhydria may result into a vicious cycle to enhance CagL-integrin interaction for gastric carcinogenesis.
This study provided evidence that gastric integrin a5b1 is mainly restricted to the basolateral membrane of gastric epithelial  cells in non-cancer patients. The location of integrin a5b1 can be shifted to the supranuclear or apical surfaces in the patients with precancerous lesions or gastric cancers. On the basis that lower intragastric acidity is indirectly implied by the decrease in the pepsinogen I/II ratio in patients with precancerous lesions or cancers ( Table 1), we revealed that the intensity and the supranuclear or apical locations of integrin a5b1 increased in parallel to the risk of cancer in patients with duodenal ulcer, gastritis, precancerous lesions as intestinal metaplasia, and the gastric cancer. As shown in Table 1 and 2, our data confirms that the lower the intragastric acidity (as indicated by a low pepsinogen I/II ratio), the stronger the priming of integrin expression in the stomach (p,0.05). These clinical data suggest that during chronic H. pylori infection, the priming of integrin as correlating with hypochlorhydria is involved in a vicious cycle to promote gastric carcinogenesis.
Experiments using a transformed cell line may be not adequate to mimic the real gastric physiological state. However, we studied patients' gastric tissues from the H. pylori-infected diseases with compatible evidence to support our in vitro findings. The study illustrated H. pylori CagL-Y58/E59 can not only trigger higher EGFR activation, but also point out its role on the suppression of acid secretion. At the very least, our data suggest such isolates exert virulence under adverse pH conditions to initiate and possibly intensify the consequent vicious cycle. In addition, CagL can activate EGFR via integrin avb5 and ILK signaling pathways to alter the level of gastrin [43]. A promising line future work is to validate the role of H. pylori CagL-Y58/E59 on such pathways.
In summary, even at a lower intragastric pH, H. pylori CagL-Y58/E59 still exploited integrin a5b1 to result in processes leading to gastric carcinogenesis. The intragastric pH elevation (hypochlorhydria) can enhance the integrin priming by H. pylori. A vicious cycle shall have existed within the H. pylori CagL, priming integrin and hypochlorhydria for gastric carcinogenesis during chronic H. pylori sequels ( Figure 5). It is promising to indicate the need of early eradication of H. pylori CagL-Y58/E59 before hypochlorhydria for the control of gastric carcinogenesis.