γ-Glutamyltranspeptidase and asparaginase have been shown to play important roles in Helicobacter pylori colonization and cell death induced by H. pylori infection. In this study, the association of γ-glutamyltranspeptidase and asparaginase was elucidated by comparing activities of both deamidases in H. pylori strains from patients with chronic gastritis, gastric and duodenal ulcers, and gastric cancer. γ-Glutamyltranspeptidase activities in H. pylori strains from patients with gastric cancer were significantly higher than in those from patients with chronic gastritis or gastric ulcers. There was a wide range of asparaginase activities in H. pylori strains from patients with gastric cancer and these were not significantly than those from patients with other diseases. To identify the contributions of γ-glutamyltranspeptidase and asparaginase to gastric cell inflammation, human gastric epithelial cells (AGS line) were infected with H. pylori wild-type and knockout strains and inflammatory responses evaluated by induction of interleukin-8 (IL-8). IL-8 response was significantly decreased by knockout of the γ-glutamyltranspeptidase-encoding gene but not by knockout of the asparaginase-encoding gene. Additionally, IL-8 induction by infection with the H. pylori wild-type strain was significantly decreased by adding glutamine during infection. These findings indicate that IL-8 induction caused by γ-glutamyltranspeptidase activity in H. pylori is mainly attributable to depletion of glutamine. These data suggest that γ-glutamyltranspeptidase plays a significant role in the chronic inflammation caused by H. pylori infection.
- H. pylori
optimal density at 600 nm
PBS with Tween 20
Helicobacter pylori is a pathogenic bacterium that inhabits the stomach . Chronic H. pylori infection has been proven to cause gastrointestinal diseases such as gastric and duodenal ulcers and gastric cancer [2, 3]. Gastric cancer is a common cause of cancer deaths worldwide, including in Japan, and it is well known that the prevalence of H. pylori infection does not always correspond to that of patients with gastric cancer . Therefore, other factors such as genetic factors, dietary habits (such as salt and alcohol intake) and differences in pathogenicity between H. pylori strains contribute to the development of gastric cancer in H. pylori -infected individuals. Many virulence factors have thus far been discovered in H. pylori, including CagA protein, which is injected by the Type IV secretion system to cause an inflammatory response  and VacA protein, which causes vacuolation of gastric cells . Polymorphisms of these proteins in H. pylori have been reported and shown to be associated not only with differences in inflammatory responses caused by H. pylori infection in vitro but also with differences in the clinical manifestations of H. pylori infection [6, 7]. Indeed, the prevalence of polymorphisms has been shown to differ geographically and to correlate with a high rate of gastric cancer in specific regions, such as Japan [8, 9]. However, these virulence factors do not sufficiently clarify why only a small percentage of H. pylori-infected individuals develop gastric cancer.
Both H. pylori and some other bacteria possess γ-glutamyltranspeptidase and asparaginase, which hydrolyse glutamine and asparagine, respectively, and produce ammonia, as well as glutamate or aspartate, respectively. Both deamidases are secretory proteins that are mainly present in the periplasm and are constitutively expressed to produce glutamate and aspartate, which undergo uptake via a transporter . It has been previously shown that H. pylori γ-glutamyltranspeptidase induces apoptosis  and inhibits proliferation of gastric cells and T cells [11, 12]. In addition, H. pylori asparaginase has been shown to inhibit T cell proliferation  and important roles in colonization by H. pylori of both γ-glutamyltranspeptidase and asparaginase have been demonstrated in an animal model [13-15]. Although these reports have demonstrated the importance of these deamidases in the pathogenicity of H. pylori infection, the mechanism of γ-glutamyltranspeptidase and asparaginase in the pathogenicity of this infection remains unclear. There are several possible explanations: one is that the strong activities of γ-glutamyltranspeptidase and asparaginase cause depletion of glutamine and asparagine, which are necessary for cell proliferation, respectively [10, 16, 17]. The other explanation is the production of ammonia by these deamidases, which results in damage to gastric cells . Glutathione is also a substrate of γ-glutamyltranspeptidase, and hydrogen peroxide produced by degradation of glutathione by γ-glutamyltranspeptidase has been suggested to cause gastric cell death [18, 19]. These explanations indicate the importance of γ-glutamyltranspeptidase and asparaginase in the pathogenicity of H. pylori, especially in the development of gastric cancer. However, the activities of these deamidases in clinical H. pylori strains from patients with gastric cancer have not yet been investigated.
In this study, we analyzed clinical isolates from Japan to evaluate the relationship between the activities of γ-glutamyltranspeptidase and asparaginase in H. pylori and gastrointestinal diseases. In addition, we evaluated the mechanisms of these deamidase activities on inflammation of gastric cells and the effects of glutamine and asparagine on H. pylori infection of cells in vitro.
MATERIALS AND METHODS
Helicobacter pylori strains and culture conditions
Forty clinical strains were obtained from H. pylori -infected patients who had undergone investigative endoscopy from 2002 to 2008 at Tokyo Medical University Hospital and had diagnoses of gastric cancer (n = 10), gastric ulcer (n = 10), duodenal ulcer (n = 10) and chronic gastritis (n = 10). There were 16 women and 24 men of mean age 57.7 years (range, 24–90 years). All strains were isolated by a method described previously . ATCC700392 was used as a type strain of H. pylori, and the ggt gene encoding the γ-glutamyltranspeptidase knockout (Δggt) strain, the ansB gene encoding the asparaginase knockout (ΔansB) strain, and the double-knockout (Δggt/ansB) strain described previously [11, 13] were used for in vitro infection experiments. All strains were subcultured on Brucella agar containing 5% horse blood under microaerobic conditions (5% O2, 12% CO2, and 83% N2) with 95% humidity and stored in brucella broth containing 30% glycerol at −80°C until used.
γ-Glutamyltranspeptidase and asparaginase activities
γ-Glutamyltranspeptidase and asparaginase activities were determined by measuring the amount of ammonia produced by reactions with glutamine and asparagine, respectively, by phenol–hypochlorite reaction. After a 3 day incubation on blood plates, H. pylori cells were suspended in HEPES and the OD600 adjusted to 0.6. The suspension was mixed with 5 mM glutamine and 5 mM asparagine in HEPES to determine γ-glutamyltranspeptidase and asparaginase activity-ies, respectively. After a 2 hr incubation at 37°C, 20 µL suspension was mixed with 100 µL phenol nitroprusside solution (Sigma, St Louis, MO, USA) and 100 µL alkaline hypochlorite solution (Sigma). After a 30 min incubation at 37°C, the absorbance at 620 nm was measured. Each reaction was performed in duplicate. The standard curve of NH4Cl was measured in each experiment and the activity calculated as the concentration of NH4Cl. Statistical analyses were performed with GraphPad Prism 6 (GraphPad Software) software, using a one-way ANOVA test followed by Bonferroni's post hoc test and differences were determined to be significant when P < 0.05.
Expression of γ-glutamyltranspeptidase
Expression of γ-glutamyltranspeptidase was detected by western blotting using polyclonal anti-γ-glutamyltranspeptidase serum against a portion of γ-glutamyltranspeptidase from ATCC700392 as described in a previous study . H. pylori cells were grown in brucella broth containing 10% FBS to OD600 = 0.5–0.6. Cells were harvested and washed with HEPES buffer, and 50 µg protein per lane of whole cell lysate subjected to SDS–PAGE (12.5%). Proteins were transferred to PVDF membranes and the membranes blocked with PBS-T containing 5% non-fat milk for 1 hr. After washing three times with PBS-T, the membranes were incubated with anti-γ-glutamyltranspeptidase serum (1:1000) overnight at 4°C. The membranes were then washed three times with PBS-T and incubated for 1 hr with goat anti-rabbit horseradish peroxidase-conjugated antibody (Santa Cruz Biotechnology, Santa Cruz, CA, USA; 1:5000). Signals were visualized using ECL western blotting detection reagents (GE Healthcare, Little Chalfont, UK). Band densities were measured by ImageJ 1.45s (the sum of the gray values of all the pixels in the selection divided by the number of pixels) . Mean band densities were compared by Sudent t-test using GraghPad Prism 6 and differences defined as significant when P < 0.05. RNA polymerase beta expression was confirmed using mouse monoclonal anti-RNA polymerase beta antibody (Abcam, Cambridge, UK).
DNA sequencing of promoter regions of γ-glutamyltranspeptidase and asparaginase
The promoter regions of the ggt and ansB genes were amplified for DNA sequencing. Based on the genome sequence of ATCC700392 (GenBank: AE000511), primers ggt F (5′-AACACGGACGCTGAAAAATC-3′) and ggt R (5′TAGCTAGCGGGTGGCTAGAA-3′), and ansB F (5′GCGCTAATGACTGCCATGAT-3′) and ansB R (5′CATGTCTTGTGAGCCGATGT-3′) were designed to amplify the promoter regions of the ggt and ansB genes, respectively. Both strands were analyzed using BigDyeTerminator version 3.1 (Applied Biosystems, Foster, CA, USA) with the same primers used for amplification.
Cell culture, Helicobacter pylori infection, and measurement of interleukin-8
Human gastric epithelial cell line AGS cells (ATCC CRL 1739) were obtained directly from the ATCC (Manassas, VA, USA) and maintained in Ham's F-12 medium (Sigma) containing 1 mM L-glutamine and 10% FBS at 37°C with 5% CO2. To assay IL-8 production, AGS cells were seeded at 1 × 106 cells/mL in 24-well tissue culture dishes and H. pylori cells, after culturing for ∼48 hr on plates, were scraped and resuspended in sterile Ham's F-12 medium to a concentration of 1 × 108 to create an multiplicity of infection of 100. AGS cells were infected by ATCC700392, the Δggt strain, the ΔansB strain, or the Δggt/ansB strain. To determine the effects of glutamine, asparagine, and glutathione, L-glutamine (5–50 mM), asparagine (5 mM), or glutathione (10 mM) was added to the medium when the cells had been infected. After 18-hr incubation in 5% CO2, the culture supernatants were stored until use. IL-8 in the culture supernatant was measured by ELISA using the Human CXCL8/IL-8 Quantikine ELISA Kit (R&D Systems, Minneapolis, MN, USA). Reactions were performed in duplicate. Statistical analyses were performed with GraphPad Prism 6 software, using the one-way ANOVA test followed by Bonferroni's post hoc test and differences determined to be significant when P < 0.05.
γ-Glutamyltranspeptidase and asparaginase activities in clinical Helicobacter pylori isolates
To compare the activities of γ-glutamyltranspeptidase and asparaginase in clinical isolates, 10 isolates each from patients with chronic gastritis, gastric ulcers, duodenal ulcers and gastric cancer were analyzed. The findings are shown in Figure 1. γ-Glutamyltranspeptidase activity of isolates from gastric cancer patients (mean value of ammonia production 1.60 mM) was significantly higher than that from those with chronic gastritis (1.23 mM, P < 0.05) and duodenal ulcers (1.30 mM, P < 0.05). Asparaginase activity was not significantly different between diseases, although the activities of the isolates from patients with gastric cancer (0.64 mM) were higher than those of isolates from patients with chronic gastritis (0.54 mM), gastric ulcers (0.54 mM) and duodenal ulcers (0.49 mM).
γ-Glutamyltranspeptidase expression in clinical Helicobacter pylori isolates
γ-Glutamyltranspeptidase expression was compared by western blot assay using polyclonal anti-γ-glutamyltranspeptidase serum. Seven isolates with high γ-glutamyltranspeptidase activities (range, 1.998–1.710 mM), and seven with low γ-glutamyltranspeptidase activities (range, 1.015–0.717 mM), were used for this experiment. As shown in Figure 2a, some isolates with high γ-glutamyltranspeptidase activities represented strong expression (TS1774 and TS1354) and some with low γ-glutamyltranspeptidase activities represented weak expression (S1903 and S1975). Mean band density from H. pylori isolates with high γ-glutamyltranspeptidase activity (mean ± SD, 57.73 ± 11.12) was significantly higher than that from H. pylori isolates with low γ-glutamyltranspeptidase activity (mean ± SD, 40.07 ± 7.78, P < 0.01), as shown in Figure 2b.
DNA sequencing of the promoter region of the ggt and ansB genes
The promoter regions of γ-glutamyltranspeptidase and asparaginase were compared between clinical isolates. Fourteen isolates (seven with strong activity and seven with weak activity), which were the same strains as were used for western blot assay for γ-glutamyltranspeptidase activity, were sequenced to compare the promoter region of the ggt gene. Another 14 isolates, which included seven with high asparaginase activity (range, 1.198–0.779 mM) and seven with low asparaginase activity (range, 0.322–0.185 mM), were used to compare the promoter region of the ansB gene. As indicated in Figure 3a, even though there was high diversity in the promoter region of the ggt gene between isolates, no specific polymorphism was related to either high or low γ-glutamyltranspeptidase activity.
Comparison between isolates of the promoter regions of the ansB gene is shown in Figure 3b. A 161 bp insertion was observed in TS1531, which had low asparaginase activity. Asparaginase of H. pylori is known to be a periplasmic asparaginase and to possess a signal peptide in the N-terminal region to be secreted in the periplasm . The insertion observed in TS1531 contained the translation start site of the secreted form of asparaginase. Specific polymorphism related to either high or low activity of asparaginase was not observed in the promoter region of the ansB gene.
Interleukin-8 production in AGS cells infected by Helicobacter pylori and mutant strains
AGS cells were infected with either H. pylori ATCC700392 wild-type, the Δggt strain, ΔansB strain or Δggt/ansB strain and IL-8 production measured by ELISA 18 hr after induction of infection. As shown in Figure 4a, IL-8 production by infection of the Δggt strain (mean ± SD, 576 ± 3 pg/mL) was significantly lower than that by wild-type infection (1177 ± 17 pg/mL, P < 0.05). IL-8 production by ΔansB infection (1140 ± 44 pg/mL) was similar to that by wild-type infection. Infection by the double-knockout strain Δggt/ansB led to the same level of IL-8 production (792 ± 48 pg/mL) as did infection by the Δggt strain. When glutamine was supplemented during wild-type strain infection, IL-8 production was significantly reduced (from 1177 ± 17 to 785 ± 87 pg/mL, P < 0.05), whereas no reduction was observed in infection by the Δggt strain (from 576 ± to 641 ± 45 pg/mL). IL-8 production was not changed by asparagine supplementation in any strain. Addition of both glutamine and asparagine led to IL-8 production similar to that seen after addition of glutamine alone in all strains.
Several glutamine concentrations (from 5–50 mM) were tested during infection by the wild-type and Δggt strains (Fig. 4b). Addition of 5 mM glutamine significantly reduced IL-8 production in wild-type strain infection (from 2478 to 1447 pg/mL, P < 0.05). When the concentration of glutamine was increased to 20 mM, the difference in IL-8 production between wild-type and Δggt strain infections was diminished. IL-8 production was significantly reduced when glutathione was added during wild-type strain infection (from 3411 to 1822 pg/mL, P < 0.05), whereas this reduction was not observed in Δggt strain infection (Fig. 4c).
The importance of γ-glutamyltranspeptidase and asparaginase in the pathogenesis of H. pylori infection has been described in both in vitro and in vivo studies of H. pylori infection; however, there have been few assessments of the activities of these deamidases in clinical H. pylori isolates. In this study, we compared the activities of deamidases in clinical isolates from patients with different gastro-intestinal diseases and found that γ-glutamyltranspeptidase activities in H. pylori isolated from patients with gastric cancer were significantly higher than in that from patients with duodenal ulcer sand chronic gastritis. We also analyzed γ-glutamyltranspeptidase expression in clinical isolates and demonstrated strong expression of γ-glutamyltranspeptidase in some isolates; this represented high γ-glutamyltranspeptidase activities. However, because some isolates with high γ-glutamyltranspeptidase activities represented moderate γ-glutamyltranspeptidase expression, high γ-glutamyltranspeptidase activities may be caused by other factors such as protein stability and localization. We compared the promoter sequence of the ggt gene in the clinical isolates and found no significant polymorphism related to high or low activities of γ-glutamyltranspeptidase. The mechanism of high γ-glutamyltranspeptidase activity requires further investigation. Previous reports have shown higher γ-glutamyltranspeptidase activities of H. pylori isolates from patients with peptic ulcers than in those from patients with non-ulcer-related dyspepsia . In this study, the activities of γ-glutamyltranspeptidase were not significantly different between patients with peptic ulcers and those with chronic gastritis. Since non-ulcer dyspepsia is not equivalent to chronic gastritis, these two studies cannot simply be compared. However, there was a wide range of γ-glutamyltranspeptidase activity in the strains from patients with gastric ulcers and some of these strains had γ-glutamyltranspeptidase activity that was as high as that observed in the strains from patients with gastric cancer.
Using both gastric and non-gastric cells, Scotti et al. showed that asparaginase inhibits the cell cycle . We have previously demonstrated the cytotoxic activity of asparaginase against a histiocytic lymphoma cell line . In this study we found that the activities of asparaginase in H. pylori isolates from patients with gastric cancer ranged widely compared with those from other diseases and, although high, were not significantly higher than those from other diseases. An insertion sequence was observed in the ansB gene of strain TS1531, which had low asparaginase activity, and this sequence was inserted into a signal peptide that leads AnsB protein to the periplasm. Therefore, the low asparaginase activity in this strain could have been attributable to this insertion. However, because this insertion was not observed in the other strains, which also had low asparaginase activity, another mechanism must be involved in the difference in asparaginase activities between strains.
The mechanisms by which these deamidases contribute to the development of gastric cancer are controversial. Chronic inflammation of gastric cells caused by H. pylori infection is known to be a trigger of gastric cancer and it has been proposed that IL-8 contributes to chronic inflammation and cancer. We investigated the proinflammatory cytokine response by measuring the IL-8 production induced by H. pylori infection in vitro and found that the inflammatory response was significantly reduced by ggt gene knockout but not by ansB gene knockout. Moreover, IL-8 induction was significantly reduced by the addition of glutamine. Several roles of γ-glutamyltranspeptidase and asparaginase in the pathogenesis of H. pylori infection have been suggested; one possible route is the depletion of glutamine and asparagine caused by γ-glutamyltranspeptidase and asparaginase, respectively [10, 16, 17]. However, we demonstrated in this study that depletion of asparagine by asparaginase contributed little to induction of the inflammatory response. This result corresponds to the findings of a previous study, which found that AnsB protein is not cytotoxic to gastric cell lines . Taken together, the depletion of glutamine, but not of asparagine, caused by high γ-glutamyltranspeptidase activity could be responsible for the strong inflammatory response caused by H. pylori infection. This is consistent with the results of this study, which demonstrated significantly high γ-glutamyltranspeptidase activities in clinical isolates from patients with gastric cancer. Therefore, infection with H. pylori strains that possess high γ-glutamyltranspeptidase activity would increase the risk of developing gastric cancer because of depletion of glutamine in gastric cells.
Ammonia has been demonstrated to induce cell death in an in vitro H. pylori infection assay  and Leduc et al. demonstrated the significant role of toxic ammonia produced by deamidases in the pathogenesis of H. pylori . We measured the concentrations of ammonia in culture supernatants and found no significant difference between the wild-type strain (2.99 mM) and several knockout strains (Δggt, 2.49 mM; ΔansB, 2.55 mM; Fig. S1). Moreover, ammonia concentrations increased 2.8-fold (8.35 mM) and 2.0-fold (5.94 mM) with the addition of glutamine and asparagine, respectively, when cells were infected by H. pylori wild-type strain (Fig. S1), whereas IL-8 induction was decreased when glutamine was added. Therefore, toxic ammonia may not contribute substantially to inflammatory responses to H. pylori infection. The increase in oxidative cell damage caused by hydrogen peroxide, which is produced by the degradation of glutathione by γ-glutamyltranspeptidase, has also been suggested as a mechanism for γ-glutamyltranspeptidase-related H. pylori pathogenicity [18, 19]. However, we observed a significant reduction in IL-8 production after adding glutathione during wild-type strain infection; however, we did not observe this in Δggt strain infection. These observations suggest that degradation of glutathione is not responsible for the inflammatory response. Taken together, depletion of glutamine could be mainly responsible for the inflammatory response caused by high γ-glutamyltranspeptidase activity of H. pylori. The beneficial effects of supplemental glutamine have been reported in animal models [25, 26]; however, further analysis is needed to elucidate whether supplemental glutamine could reduce the risk of developing gastric cancer caused by chronic H. pylori infection.
In conclusion, in this study we measured γ-glutamyltranspeptidase and asparaginase activities in H. pylori strains from patients gastrointestinal-related disease and demonstrated significantly higher γ-glutamyltranspeptidase activities in H. pylori strains from patients with gastric cancer than in those from patients with chronic gastritis and duodenal ulcers. We suggest that the inflammatory response caused by γ-glutamyltranspeptidase activity of H. pylori is attributable to depletion of glutamine. These data suggest a significant role of γ-glutamyltranspeptidase in the chronic inflammation caused by H. pylori infection.
We thank Drs. Norihisa Noguchi, Takashi Kawai and Masanori Sasatsu for providing the H. pylori isolates used in this study. This study was supported by a Grant-in-Aid for Scientific Research (C) from the Japan Society for the Promotion of Science (No. 22590410) and a grant from the Ministry of Health, Labour and Welfare of Japan (H21-Shinkou-Ippan-008).
The authors declare no conflicts of interest.