INFECTION AND IMMUNITY, July 2001, p. 4382–4389 0019-9567/01/$04.00⫹0 DOI: 10.1128/IAI.69.7.4382–4389.2001 Copyright © 2001, American Society for Microbiology. All Rights Reserved.
Vol. 69, No. 7
Type I Helicobacter pylori Lipopolysaccharide Stimulates Toll-Like Receptor 4 and Activates Mitogen Oxidase 1 in Gastric Pit Cells TSUKASA KAWAHARA,1 SHIGETADA TESHIMA,1 AYUKO OKA,1 TOSHIRO SUGIYAMA,2 KYOICHI KISHI,1 AND KAZUHITO ROKUTAN1* Department of Nutritional Physiology, School of Medicine, University of Tokushima, Tokushima 770-8503,1 and the Third Department of Internal Medicine, Hokkaido University School of Medicine, Sapporo 060-8638,2 Japan Received 16 April 2001/Accepted 21 April 2001
Guinea pig gastric pit cells express an isozyme of gp91-phox, mitogen oxidase 1 (Mox1), and essential components for the phagocyte NADPH oxidase (p67-, p47-, p40-, and p22-phox). Helicobacter pylori lipopolysaccharide (LPS) and Escherichia coli LPS have been shown to function as potent activators for the Mox1 oxidase. These cells spontaneously secreted about 10 nmol of superoxide anion (O2ⴚ)/mg of protein/h under LPS-free conditions. They expressed the mRNA and protein of Toll-like receptor 4 (TLR4) but not those of TLR2. LPS from type I H. pylori at 2.1 endotoxin units/ml or higher stimulated TLR4-mediated phosphorylations of transforming growth factor ␤-activated kinase 1 and its binding protein 1 induced TLR4 and p67-phox and up-regulated O2ⴚ production 10-fold. In contrast, none of these events occurred with H. pylori LPS from complete or partial deletion mutants of the cag pathogenicity island. Lipid A was confirmed to be a bioactive component for the priming effects, while removal of bisphosphates from lipid A completely eliminated the effects, suggesting the importance of the phosphorylation pattern besides the acylation pattern for the bioactivity. H. pylori LPS is generally accepted as having low toxicity; however, our results suggest that type I H. pylori lipid A may be a potent stimulator for innate immune responses of gastric mucosa by stimulating the TLR4 cascade and Mox1 oxidase in pit cells. Helicobacter pylori infection causes type B chronic gastritis, peptic ulcer diseases, and lymphomas of the mucosa-associated lymphoid tissue and is an important risk factor for gastric carcinoma (5, 27, 28). H. pylori strains are grouped into two families, type I and type II. Patients with the gastric lesions are most often infected by type I strains that are characterized by the presence of the cytotoxin-associated gene A (cagA) and the vacuolating cytotoxin gene A (vacA) (39). Type I strains have not only the cagA gene but also an insertion of approximately 40 kb of foreign DNA, named the cag pathogenicity island (PAI). These genes are now recognized as transmissible DNA that encodes virulence factors and maps to the chromosome of pathogenic organisms. The H. pylori cag PAI contains 31 genes, including cagA (8). Among the cag PAI genes, cagE (picB), cagG, cagH, cagI, cagL, and cagM are involved in the activation of nuclear factor B (NF-B) and stimulation of interleukin 8 (IL-8) secretion (16, 31). Six of the cag PAI genes code for the core subunits of the type IV export machinery that can transfer CagA protein into host epithelial cells, and translocated CagA has been shown to be tyrosine phosphorylated by host cells (3, 10, 25, 32). Deletion of the complete cag PAI, partial deletions, insertions, and rearrangements within the cag PAI have been proposed as the basis for modified or reduced virulence. Compared with these virulent factors, H. pylori lipopolysaccharide (LPS) has been believed to be less toxic (3, 21), since 1,000- to 10,000-times-higher concentrations of H. pylori LPS
are required for activation of host spleen cells or macrophages than of LPS from Salmonella enterica or Escherichia coli (20, 21, 29). In addition to having lower immunological activities, H. pylori LPS contains Lewis blood antigens, and the molecular mimicry has been proposed to camouflage and allow colonization to persist chronically. Recently, we showed that cultured guinea pig gastric pit cells express mitogen oxidase 1 (Mox1), a non-phagocyte-specific isozyme of gp91-phox, as well as p67-, p47-, p40-, and p22-phox, and that they spontaneously produce a large amount of superoxide anion (O2⫺) (35–37). H. pylori LPS markedly up-regulated the oxidase in association with the induction of Mox1, p67-phox, and p22-phox (36, 37). This enhanced O2⫺ production could activate NF-B in pit cells themselves (35, 36), suggesting that H. pylori LPS and Mox1 play important roles in the initiation of mucosal cell responses against H. pylori infection. Toll-like receptors (TLRs) have been characterized as a family of mammalian homologs of Drosophila Toll. Among the TLR family members, TLR4 confers responsiveness to LPS from gram-negative bacterial, while TLR2 responds to yeast or gram-positive bacterial cell wall components, such as lipoproteins (1, 34). The TLR4 mRNA is ubiquitously expressed in various types of cells and has been suggested to play an important role in various pathological conditions (1, 6, 7, 12, 13, 34). Using LPS-free cultures of guinea pig gastric mucosal cells, we found that H. pylori LPS stimulated distinct signaling pathways of TLR4 and activated Mox1, expressed in gastric pit cells. Our results also suggested that the cag PAI genes may be crucial for the synthesis of bioactive lipid A molecules that stimulate TLR4-mediated intracellular events in gastric pit cells.
* Corresponding author. Mailing address: Department of Nutrition, School of Medicine, University of Tokushima, 3-18-15 Kuramoto-cho, Tokushima 770-8503, Japan. Phone: 81-88-633-9246. Fax: 81-88-6337086. E-mail: [email protected]
ACTIVATION OF Mox1 BY H. PYLORI
VOL. 69, 2001 MATERIALS AND METHODS Preparation and culture of gastric mucosal cells under LPS-free conditions. Gastric mucosal cells were isolated aseptically from guinea pig fundic glands, as previously described (36). In the present experiments, all reagents used for culture were free from detectable amounts of LPS by the Limulus amebocyte lysate assay (Endospecy; Seikagaku Kogyo Co., Tokyo, Japan). The isolated cells were cultured for 2 days in RPMI 1640 (GIBCO, Grand Island, N.Y.), containing 50 g of gentamicin per ml, 100 U of penicillin G per ml, and 10% fetal bovine serum (FBS). The FBS (ICN Biomedicals, Aurora, Ohio) contained ⬍0.01 endotoxin unit (EU) of LPS per ml. The complete culture medium contained less than 0.01 EU of LPS per ml. After 2 days of culture, growing cells consisted of pit cells (about 90%), pre-pit cells (about 5%), parietal cells (4 to 5%), mucous neck cells (less than 1%), and fibroblasts (less than 1%) (36). Mature pit cells were confirmed to be O2⫺-producing cells by nitroblue tetrazolium staining (36). The amount of O2 released was spectrophotometrically measured by the superoxide dismutase-inhibitable reduction of cytochrome c and expressed as nanomoles per milligram of protein per hour (36). Isolation and culture of clinical H. pylori strains. The present experiments were approved by the ethics committees of the Medical Faculty of Hokkaido University and the Medical Faculty of the University of Tokushima. H. pylori (NCTC 11637) was prepared as previously described (36). Clinical isolates of H. pylori were established from gastric biopsy specimens and were cultured on H. pylori-selective agar plates (Eiken Chemical Co., Tokyo, Japan) under microaerophilic conditions (12% CO2–5% O2–83% N2) for up to 5 days. The organisms were identified as H. pylori by Gram staining, colony morphology, and positive oxidase, catalase, and urease reactions. A single colony on the agar was collected and cultured in brucella broth (GIBCO) supplemented with 5% FBS and 10 g of vancomycin per ml. Determination of genotypes of clinical isolates. Bacterial genomic DNA was extracted, and PCR was performed using the following primer sets: vacA, 5⬘-A TGGAAATACAACAAACACA-3⬘ and 5⬘-CTCCAGAACCCACACGATT-3⬘ or 5⬘-TACAAACCTTATTGATTGATAGCC-3⬘ and 5⬘-AAGCTTGATTGATC ACTCC-3⬘; cagA, 5⬘-GGGGATCCATGACTAACGAAACC-3⬘ and 5⬘-GGCTT AAGTGATGGGACACCCAA-3⬘; cagE, 5⬘-GCTAGTTATAGAGCAAGAGG TTCAA-3⬘ and 5⬘-TAGTTGTTAGTAAGGATCACCCCAT-3⬘; and cagG, 5⬘CCCTAATATCGGTGGTAAAAA-3⬘ and 5⬘-CTATTTGCTTGGTGTCTTAT C-3⬘. The sequences of these primers corresponded to the cag PAI genes of H. pylori NCTC 11638. PCR was performed under the following conditions: 35 cycles of 1 min at 92°C, 1 min at 52°C, and 1 min at 72°C. For Southern blot analysis, 10 g of genomic DNA of H. pylori was digested by HaeIII, HindIII, or EcoRI (New England Biolabs, Beverly, Mass.), electrophoresed on a 1% agarose gel, and then transferred onto a nylon membrane. The cagA, cagE, cagG, and vacA probes prepared as described above were labeled with digoxigenin (DIG) by a PCR DIG probe synthesis kit (Roche, Basel, Switzerland). The membrane was hybridized with one of the labeled probes for 20 h at 42°C in DIG Easy Hyb (Roche). After being washed sequentially in 2⫻ standard saline citrate (SSC; 1⫻ SSC is 0.15 M NaCl plus 0.015 M sodium citrate) containing 0.1% sodium dodecyl sulfate (SDS) and 0.2⫻ SSC–0.1% SDS, the hybridized probes were detected using a DIG nucleic acid detection kit (Roche). Preparation and treatment of H. pylori LPS and lipid A. LPS was prepared from six clinical strains by the hot-phenol–water method of Westphal and Jann (38) and subsequently treated with DNase I, RNase A, and 100 g of proteinase K per ml as described by Moran et al. (19). The treated LPSs were ultracentrifuged and dialyzed against LPS-free water (Otsuka Pharmaceutical Co., Tokushima, Japan). For treatment with endo-␤-galactosidase, 2.1 ⫻ 103 EU of the prepared LPS or E. coli LPS from E. coli K-235 (Sigma Chemical Co., St. Louis, Mo.) was incubated with 20 mU of the enzyme per ml in 1 ml of 0.1 M acetate buffer (pH 5.8) at 37°C for 12 h and then boiled for 1 h. The ketosidic linkage between core oligosaccharide (OS) and lipid A was decomposed by boiling of LPS in 0.1 M acetate buffer (pH 6.5) for 1 h (4). Lipid A was pelleted by centrifugation at 3,000 ⫻ g for 30 min. OS and polysaccharide chain complexes in the supernatant were collected as a polysaccharide fraction. Precipitated lipid A was dissolved in LPS-free saline (Otsuka Pharmaceutical Co.). LPS and lipid A from H. pylori or E. coli were dephosphorylated by treatment with 48% hydrofluoric acid (HF) at 4°C for 48 h (19). After evaporation of aqueous HF, the dephosphorylated LPS and lipid A were dissolved in the saline. After the purified LPS and lipid A had been lyophilized, their dry weights and Limulus activities were measured using a Supermicro (model S4; Sartorius, Go ¨ttingen, Germany) and by the Limulus amebocyte lysate assay, respectively. Immunoblot analysis. Monoclonal antibodies against human TLR4 (HTA125 and HTA1216) and a polyclonal antibody against recombinant human p67-phox
were kindly provided by K. Miyake (Saga Medical School, Saga, Japan) and B. M. Babior (The Scripps Research Institute, La Jolla, Calif.), respectively. A monoclonal antibody against actin was purchased from Oncogene Research Products (Cambridge, Mass.). The level of p67-phox was measured by immunoblot analysis as described previously (36, 37). A polyclonal antibody was made by immunizing rabbits with the synthetic peptide corresponding to residues 295 to 310 of human TLR2. The resultant serum was further purified by affinity chromatography with the synthetic peptide-conjugated agarose. A membrane fraction was prepared as previously described (36). The amounts of TLR2 and TLR4 in the membrane fraction were determined by Western blot analysis with the anti-TLR2 and HTA1216 or HTA125 antibodies, respectively. For detection of transforming growth factor ␤-activated kinase 1 (TAK1) and TAK1-binding protein 1 (TAB1), cellular proteins were prepared in the presence of inhibitors of proteases and phosphatases as described previously (30). Phosphorylation of the proteins was also confirmed by treatment with 5 U of bacterial alkaline phosphatase as previously described (30). Each sample was separated by SDSpolyacrylamide gel electrophoresis (PAGE) in a 7.5% polyacrylamide gel and transferred to a polyvinylidene difluoride filter. After blocking with 4% purified milk casein, the filter was incubated for 1 h at room temperature with a 1:500 dilution of polyclonal antibody against amino acid residues 554 to 579 of mouse TAK1 or residues 480 to 500 of human TAB1 (gifts from K. Matsumoto, Nagoya University, Nagoya, Japan). Bound antibodies were detected by an enhanced chemiluminescence system (Amersham Pharmacia, Piscataway, N.J.). Detection of TLR2 and TLR4 transcripts. Total RNA was isolated from guinea pig gastric mucosal cells and guinea pig peripheral blood lymphocytes (PBL) with an acid guanidinium thiocyanate-phenol-chloroform mixture (9). Reverse transcriptase (RT)-PCR was done to detect the TLR2 and TLR4 transcripts using the following PCR primer sets: TLR2, 5⬘-GTCCAGGAGCTG GAGAACT-3⬘ and 5⬘-GGAACCTAGGACTTTATCGCA-3⬘; TLR4, 5⬘-TC ACCTGATGCTTCTTGCTG-3⬘ and 5⬘-AGTCGTCTCCAGAAGATGTG-3⬘. The resultant PCR products separated on an agarose gel were purified, ligated into a pCR2.1-TOPO vector (Invitrogen, Carlsbad, Calif.), and transformed into JM109 cells. Transformed plasmids containing the appropriate insert DNA were selected and sequenced with a DNA sequencer (model ABI 377; PE Biosystems Japan, Tokyo, Japan). For measurement of the TLR4 mRNA level, total RNA (8 g per lane) was subjected to electrophoresis in a 1% agarose gel and transferred to a nylon filter. After prehybridization, the membrane was hybridized for 4 h at 60°C in a Rapid Hyb buffer (Amersham Pharmacia) containing the amplified TLR4 cDNA or a cDNA probe for human glyceraldehyde-3-phosphate dehydrogenase (GAPDH ATCC 57494; American Type Culture Collection, Rockville, Md.). These probes were prelabeled with [␣-32P]dCTP using a random primer kit (Amersham Pharmacia). The membrane was washed twice with 2⫻ SSC containing 0.5% SDS for 10 min at 65°C and then three times with 0.2⫻ SSC containing 1% SDS. Bound probes were autoradiographed by exposure to Kodak X-Omat film for an appropriate time at ⫺80°C.
RESULTS Effects of H. pylori on O2ⴚ production. Guinea pig gastric pit cells, cultured under the conditions used in our previous study, spontaneously released about 50 nmol of O2⫺/mg of protein/h (36). The basal rate of O2⫺ production decreased to 11 ⫾ 2 nmol/mg of protein/h (mean ⫾ standard deviation [SD], n ⫽ 12) in the LPS-free system used in the present study (Fig. 1A). According to the results of PCR and Southern blot analyses, H. pylori 1, 2, and 3 were vacA-positive and cag PAI-positive strains (type I). H. pylori 5 and 6 were determined to be vacApositive and cag PAI-negative strains (type II). H. pylori 4 was identified as a mutant with a partial deletion of cag PAI (Table 1). When gastric mucosal cells were cocultivated with NCTC 11637 or one of the clinical isolates, NCTC 11637, H. pylori 1, and H. pylori 3 significantly enhanced O2⫺ production 1.5-, 1.5-, and 4.7-fold, respectively, while H. pylori 2, 4, 5, and 6 had no effect on it (Fig. 1A). Next, the culture supernatant of each H. pylori strain was prepared and added to the cells. The supernatants of NCTC 11637, H. pylori 1, and H. pylori 3 increased the O2⫺ production
KAWAHARA ET AL.
FIG. 1. Effects of H. pylori on O2⫺ production by gastric mucosal cells. Gastric mucosal cells (105 cells) growing on 24-well culture plates were cocultivated with one of the live clinical strains of H. pylori (107 cells/ml) for 24 h in RPMI 1640 containing 10% FBS (A). The culture supernatant was collected after cultivation of H. pylori alone (107 cells/ml) in RPMI 1640 containing 10% FBS for 24 h. After filtration through a 0.2-mm-pore-size filter, gastric mucosal cells (105 cells) on 24-well culture plates were incubated with 1 ml of each supernatant for 24 h (B). The amounts of O2⫺ release were measured as described in Materials and Methods, and they are expressed as means ⫾ SD (n ⫽ 12). #, significant increase compared with untreated control cells (P ⬍ 0.05 by analysis of variance and Scheffe´’s test). (C) The concentrations of LPSs in the H. pylori culture supernatants were measured by the Limulus amebocyte lysate assay.
1.7-, 1.8-, and 9.3-fold, respectively, while those of H. pylori 2, 4, 5, and 6 had no effect on O2⫺ production (Fig. 1B). The stimulatory effects of supernatants could not be eliminated by boiling, and the amount of LPS contamination in each supernatant, except for that in the culture supernatant of H. pylori 4, roughly correlated with the magnitude of the stimulatory action (Fig. 1C), suggesting that LPS possibly produced by autolysis of the bacteria might be a crucial up-regulator for O2⫺ production. Effects of H. pylori LPS on O2ⴚ production. As listed in Table 1, the specific Limulus activities of LPSs from cag PAInegative strains (H. pylori 5 and 6) were 4,300 to 420,000 times lower than those from the type I H. pylori strains. Gastric mucosal cells were treated for 24 h with different concentrations of H. pylori 1 LPS (0.21 to 2,100 mEU/ml) or E. coli (0.03 to 344 mEU/ml). H. pylori 1 LPS at 2.1 EU/ml (19.3 ng/ml) or higher significantly enhanced O2⫺ production, and the 50% effective concentration (EC50) was calculated to be 8 EU/ml. In response to 21 EU of H. pylori 1 LPS per ml, O2⫺ production began to increase within 8 h and reached a maximum level of 105 ⫾ 2 nmol of O2⫺/mg of protein/h (mean ⫾ SD, n ⫽ 12) at 24 h. The pit cell oxidase was more sensitive to E. coli LPS (EC50, 0.3 EU/ml). E. coli LPS at 3.4 EU/ml (10 ng/ml) upregulated O2⫺ production in a similar time course, and the level had increased to 109 ⫾ 3 nmol of O2⫺/mg of protein/h (n ⫽ 12) at 24 h. The culture supernatant of H. pylori 4 contained 16.8 EU of LPS per ml, whereas the supernatant could not up-regulate the production of O2⫺. The specific Limulus activity of H. pylori 4 was similar to those of type 1 H. pylori LPSs (Table 1). However, the stimulatory actions of H. pylori 1 and H. pylori 4 LPSs on O2⫺ production were markedly different: the EC50 for H. pylori 4 LPS (210 EU/ml) was 26 times higher than that of H. pylori 1 LPS (Fig. 2A). Thus, the priming effect of H. pylori LPS did not simply correlate with its specific Limulus activity. Compared with H. pylori 1 and 3, live H. pylori 2 and its
culture supernatant exhibited lower priming effects on O2⫺ production. However, treatment with 21 EU of LPSs from all of the type I strains (H. pylori 1, 2, and 3) per ml similarly increased the O2⫺ production about 10-fold (Fig. 2B). On the other hand, LPS of H. pylori 5 or 6 increased the rate less than 2-fold, and H. pylori 4 had no effect (Fig. 2B). In the cases of H. pylori 5 and 6, much higher concentrations of H. pylori 5 LPS (15 mg/ml) and H. pylori 6 LPS (1.2 mg/ml) were added; therefore, they might have nonspecifically increased the production. The diversity of the bioactivity of each LPS was also confirmed by host cell response. Consistent with the results of a previous study (37), treatment with H. pylori 1 LPS significantly increased the levels of Mox1 and p67-phox 1.4- and 4.6-fold, respectively, while it did not change the p47-phox level (data not shown). The stimulatory action of each LPS on O2⫺ production coincided with its p67-phox-inducing capability; treatment with 21 EU of LPS from H. pylori 1, 2, or 3 per ml for 24 h significantly induced p67-phox in the cells, while H. pylori 4, 5, and 6 had no effect on the expression (Fig. 2C).
TABLE 1. Specific Limulus activities of LPS and lipid A of clinical isolatesa Organism
E. coli H. pylori H. pylori H. pylori H. pylori H. pylori H. pylori
1 2 3 4 5 6
vacA⫹ vacA⫹ vacA⫹ vacA⫹ vacA⫹ vacA⫹
cagA⫹ cagE⫹ cagG⫹ cagA⫹ cagE⫹ cagG⫹ cagA⫹ cagE⫹ cagG⫹ cagA⫹ cagE cagG cagA cagE cagG cagA cagE cagG
Limulus activity (EU/g) LPS
344 109 73.3 593 157 0.0014 0.017
40.0 22.1 62.8 109 87.1 0.00046 0.004
a H. pylori genotypes were determined by PCR and Southern blot analyses, and LPS and lipid A were purified from H. pylori 1 through 6, as described in Materials and Methods. Their dry weights and Limulus activities were measured.
VOL. 69, 2001
ACTIVATION OF Mox1 BY H. PYLORI
B). As shown in Fig. 3D, lipid A from H. pylori 1 or E. coli stimulated p67-phox induction. In contrast, their polysaccharide fractions did not change the O2⫺ production (Fig. 3C) or induce p67-phox (Fig. 3D), indicating that lipid A is a bioactive
FIG. 2. Effects of H. pylori LPS from clinical strains on O2⫺ production and p67-phox level in gastric mucosal cells. (A) Cells were treated with different concentrations of H. pylori 1 LPS (■) or H. pylori 4 (}) for 24 h in RPMI 1640 containing 10% FBS, (B) Cells were also incubated with 21 EU of LPS per ml from each clinical strain for 24 h. The amounts of O2⫺ release are expressed as means ⫾ SD (n ⫽ 12). #, significant increase compared with untreated control cells (P ⬍ 0.05 by analysis of variance and Scheffe´’s test). (C) After treatment with 21 EU of each LPS per ml for 24 h, whole-cell proteins were extracted from the cells, and samples of 20 g protein per lane were separated by SDS-PAGE in a 7.5% polyacrylamide gel and immunoblotted with anti-p67-phox antibody.
Identification of bioactive component of H. pylori LPS. H. pylori LPS consists of a lipid A region, an OS region, and polysaccharide chains that are also known as O antigen chains (4). Breakage of lactosaminoglycan chains in the polysaccharides of H. pylori 1 and E. coli LPS by endo-␤-galactosidase did not affect the priming effect (data not shown). On the other hand, polymyxin B, which is known to bind to the lipid moiety of LPS and inactivates its activity, inhibited the LPS action in a dose-dependent manner; the 50% inhibitory concentrations for H. pylori 1 and E. coli LPS were determined to be 5.0 ⫻ 10⫺5 and 2.0 ⫻ 10⫺6 g/ml, respectively. LPS was separated into free lipid A and a polysaccharide fraction by boiling in acetate buffer. The specific Limulus activity of each lipid A is listed in Table 1. The EC50s for the priming effect of lipid A from H. pylori 1 and E. coli were calculated to be 0.9 and 0.01 EU/ml, respectively (Fig. 3A and
FIG. 3. Effects of lipid A and polysaccharides from H. pylori 1 or E. coli LPS on O2⫺ production and p67-phox induction. Lipid A and polysaccharides (PS) were separated from H. pylori 1 or E. coli LPS by boiling in 0.1 M acetate buffer (pH 6.5) for 1 h as described in Materials and Methods. Gastric mucosal cells were treated with H. pylori 1 (A) or E. coli (B) lipid A at the indicated concentrations for 24 h in RPMI 1640 containing 10% FBS. (C) Cells were treated for 24 h with PS from H. pylori 1 (21 EU/ml) or from E. coli (3.4 EU/ml), lipid A from H. pylori 1 (20 EU/ml) or from E. coli (0.5 EU/ml), and the respective concentrations of LPS from H. pylori 1 or from E. coli. The amounts of O2⫺ release are expressed as means ⫾ SD (n ⫽ 12). #, significant increase compared with untreated control cells (P ⬍ 0.05 by analysis of variance and Scheffe´’s test). (D) After treatment with LPS, PS, or lipid A for 24 h, the amounts of p67-phox were measured by immunoblot analysis as described in Materials and Methods.
KAWAHARA ET AL.
component for the priming effect. We also confirmed that excess amounts of lipid A from H. pylori 5 (4.5 mg/ml) and H. pylori 6 (0.52 mg/ml) neither increased the O2⫺ production nor induced p67-phox (data not shown). Treatment with lipid A (20 EU/ml) from H. pylori 1 for 24 h increased the O2⫺ release from 10 ⫾ 3 to 107 ⫾ 9 nmol/mg of protein/h (mean ⫾ SD, n ⫽ 12). Removal of bisphosphates from lipid A by treatment with HF for 48 h at 4°C completely abolished the priming effect, and the O2⫺ production remained at the basal level (16 ⫾ 4 nmol/mg of protein/h, n ⫽ 12). We also confirmed that treatments of H. pylori 1 LPS, E. coli LPS, and E. coli lipid A with HF completely block their priming effects (data not shown). Expression of TLR4 in guinea pig gastric mucosal cells. As shown in Fig. 4, RT-PCR amplified a transcript corresponding to the TLR4 mRNA in guinea pig gastric mucosal cells as well as guinea pig PBL. DNA sequencing showed that the amplified product of gastric mucosal cells had 90% DNA sequence identity to the human TLR4 cDNA (bp 2132 to 2596, GenBank accession number U88880). Northern blot analysis showed that gastric mucosal cells expressed a significant amount of the TLR4 mRNA with a molecular size of 5 kbp and that H. pylori 1 LPS stimulated the mRNA expression within 4 h (Fig. 4B). The increased TLR4 mRNA expression resulted in the accumulation of TLR4 protein (Fig. 4C). We also examined whether gastric mucosal cells expressed another candidate for an LPS receptor, TLR2 (40). As shown in Fig. 5, RT-PCR analysis did not amplify the TLR2 transcript, and TLR2 protein was not detected by Western blot analysis. These results suggested the importance of TLR4 in the cellular responses to H. pylori LPS. Phosphorylation of TAK1 and TAB1 in guinea pig gastric mucosal cells. To confirm that TLR4-mediated signals were evoked by H. pylori LPS, we examined whether H. pylori LPS could induce phosphorylation of TAK1 and TAB1, which are known to be common signal transmission molecules for TLR and IL-1 receptor signal cascades (17, 24). As shown in Fig. 6A and B, H. pylori 1 LPS phosphorylated TAK1 and TAB1. Furthermore, the stimulatory action of each LPS on O2⫺ production coincided with its TAK1- and TAB1-phosphorylating activities; H. pylori 1 and E. coli LPSs could induce their phosphorylations (Fig. 6A and B), while LPS from H. pylori 5 and 6 could not (Fig. 6A and B), suggesting that H. pylori LPS may stimulate O2⫺ production by activating TLR4-mediated signal pathways. DISCUSSION Recently, several novel isozymes of gp91-phox expressed in nonphagocytic cells, including Mox1 (33), Renox (15), and Thox proteins (11), have been molecularly identified. Among these nonphagocytic oxidases, the pit cell oxidase Mox1, whose O2⫺-producing capacity is equivalent to that of macrophages (35–37), is the most potent one. Mox1-derived O2⫺ and related oxygen intermediates are thought to play crucial roles in the initiation of inflammatory and immune responses as well as in the regulation of cell growth (33, 35, 36). It is particularly important to identify the regulator(s) of Mox1 activity. H. pylori does not attach to guinea pig gastric mucosal cells; therefore, our system is an excellent model for studying the contact-
FIG. 4. Detection of TLR4 mRNA and protein in gastric mucosal cells. (A) Total RNA was isolated from cultured guinea pig gastric mucosal cells (GMC) and guinea pig PBL, and RT-PCR was performed, as described in Materials and Methods. RT-PCR products from PBL (5 g in lane 2 and 1.5 g in lane 3) and GMC (5 g in lane 4 and 1.5 g in lane 5) were subjected to electrophoresis in a 6% polyacrylamide gel. Lane 1 shows molecular weight (MW) standard markers, and lane 6 contains PCR products from GMC without RT reaction as a negative control (NC). (B) After cells had been treated with 21 EU of H. pylori 1 LPS per ml for the indicated times, total RNA (8 g per lane) was separated in a 1% agarose gel. Northern hybridization with the cDNA probe for TLR4 or GAPDH was performed as described in Materials and Methods. (C) Membrane proteins were prepared from GMC and guinea pig PBL, as described in Materials and Methods, and immunoblot analysis with an antibody against TLR4 was performed. The bound antibodies were then removed by rinsing the membranes for 15 min at 50°C in 60 mM Tris-HCl buffer containing 0.1 mM 2-mercaptoethanol and 2% SDS. The membrane was again subjected to immunoblotting with an antibody against actin. The levels of TLR4 and membrane-associated actin were quantified by laser densitometry, and the TLR4/actin ratios are expressed as means ⫾ SD (n ⫽ 4). #, significant increase compared with untreated control cells (P ⬍ 0.05 by analysis of variance and Scheffe´’s test).
VOL. 69, 2001
ACTIVATION OF Mox1 BY H. PYLORI
FIG. 5. Detection of TLR2 mRNA and protein. (A) Total RNA was isolated from cultured guinea pig gastric mucosal cells (GMC) and guinea pig and human PBL, and RT-PCR was performed, as described in Materials and Methods. (B) Membrane fractions of these cells (20 g per lane) were subjected to immunoblot analysis with the anti-TLR2 antibody.
independent interactions between gastric epithelial cells and the gram-negative bacterium. We had studied the effects of growth factors (epidermal growth growth factor and transforming growth factor ␤), cytokines (gamma interferon, tumor necrosis factor alpha, IL-1, IL-3, and IL-6), histamine, carbacol, and phorbol 12-myristate 13-acetate on the oxidase activity; however, none of them were able to up-regulate O2⫺ production (reference 36 and data not shown). Finally, we had found that guinea pig gastric mucosal cells were sensitive to H. pylori LPS, as was found in the case of E. coli LPS, and that they increased the rate of O2⫺ production in association with the induction of Mox1, p67-phox, and p22-phox (37). Although quiescent gastric mucosal cells maintained in an LPS-free culture system were more sensitive to E. coli LPS than to an H. pylori one, H. pylori LPS appears to be a more important up-regulator relevant to gastric pathophysiology. H. pylori frequently changes its LPS molecule depending on the culture conditions (20) and passages (19). The complete genome sequencing of H. pylori provided a genetic basis for understanding the biological processes. It is known that H. pylori genomes, especially those encoding the outer membrane proteins or enzymes for LPS synthesis, are diverse (14). However, this information is limited to the biosynthesis of Lewis antigens in the O-polysaccharide regions, such as the ␣1,3fucosyltransferase gene (2). Lipid A, but not polysaccharides, was determined to be a bioactive component on Mox1, while the molecular basis for biosynthesis of H. pylori lipid A is not fully understood. According to the information on synthetic E. coli-like lipid A, phosphate patterns, the numbers of acyl chains, and fatty acid compositions are important for the full expression of a range of biological activities (20). For example, lipid A composed of bisphosphates and hexaacyl chains is more toxic than that composed of monophosphate and tetraacyl chains (18). H. pylori synthesizes two types of lipid A molecules: hexaacyl- and tetraacyl-lipid A (19, 20). Hexaacyl-lipid A has two phosphorylates or phosphorylethanolamines on the lipid A disaccharide backbone, while tetraacyl-lipid A contains only one phosphate. The toxicity of H. pylori tetraacyl-lipid A on human monocytes is about fourfold lower than that of the
FIG. 6. Phosphorylation of TAK1 (A) and TAB1 (B) by LPS. Cultured cells were treated with 21 EU of H. pylori 1 LPS per ml for the indicated times or with 21 EU of LPS from three H. pylori clinical strains (H. pylori 1, 5, and 6) per ml or 3.4 EU of E. coli per ml for 60 min. Protein samples were extracted from these cells. Protein extracts, prepared from the cells after treatment with 21 EU of H. pylori 1 LPS per ml for 2 h, were treated with bacterial alkaline phosphatase (BAP) (lanes 7), as described in Materials and Methods. Immunoblot analysis with an antibody against TAK1 or TAB1 was performed as described in Materials and Methods. p-TAK1, phosphorylated TAK1; p-TAB1, phosphorylated TAB1. Similar results were obtained in three separate experiments.
KAWAHARA ET AL.
hexaacyl form (26). It has been reported that dephosphorylation of H. pylori LPS did not alter the priming activity on neutrophils (23), while our results suggest that the phosphorylation pattern of lipid A, rather than its acylation pattern, may be more important for the priming activity on gastric mucosal cells. The present novel approach has clearly demonstrated that the presence of cag PAI genes is crucial for the stimulatory actions of H. pylori LPS of the Mox1 oxidase. The proteins encoded by the cag PAI genes contain motifs found in bacterial proteins, such as translocases, sensors, permeases, flagellumassembling proteins, and components of the type IV secretion machinery (14). At present, there is no evidence that the cag PAI genes directly participate in the synthesis of bioactive lipid A molecules. Furthermore, multiple genes and environmental factors appear to be involved in the synthesis of H. pylori LPS; therefore, the genomic and molecular bases for the linkage between the cag PAI genes and synthesis of bioactive lipid A remain to be elucidated. Consistent with the results of other studies (20, 21, 29), H. pylori 1 LPS at concentrations up to 210 EU/ml did not enhance O2⫺ production from murine peritoneal macrophages stimulated by phorbol 12-myristate 13-acetate (data not shown). Therefore, we examined whether H. pylori LPS could actually stimulate TLR4 on gastric mucosal cells. TLR2 has been suggested to be another possible candidate for an LPS receptor (40), but the TLR2 mRNA and its protein were not detected in gastric mucosal cells. Furthermore, a typical TLR2 ligand, Staphylococcus aureus peptidoglycan (34), did not enhance the O2⫺ production or TAK1 phosphorylation (data not shown). On the other hand, type I H. pylori LPS could stimulate the TLR4 mRNA expression in gastric mucosal cells, as was observed in TLR4-expressing cells exposed to bacterial LPS (13, 22). In addition, type I H. pylori LPS, but not type II H. pylori LPS, could activate TAK1 and TAB1 and up-regulate the expression of Mox1 and p67-phox (Fig. 2C), suggesting that TLR4-dependent pathways may, at least partially, play a crucial role in the up-regulation of the pit cell oxidase. The present results suggest that TLR4 and Mox1 oxidase in pit cells may mediate the interactions between the gram-negative bacterium and host epithelial cells, initiating inflammatory and immune responses against H. pylori infection. ACKNOWLEDGMENT This work was supported by a Grant-in Aid for Science Research from the Japanese Ministry of Education, Science and Culture (to K.R.). REFERENCES 1. Aliprantis, A. O., R. Yang, M. R. Mark, S. Suggett, B. Devaux, J. D. Radolf, G. R. Klimpel, P. Godowski, and A. Zychlinsky. 1999. Cell activation and apoptosis by bacterial lipoproteins through Toll-like receptor-2. Science 285: 736–739. 2. Appelmelk, B. J., S. L. Martin, M. A. Monteiro, C. A. Clayton, A. A. McColm, P. Zheng, T. Verboom, J. J. Maaskant, D. H. van den Eijnden, C. H. Hokke, M. B. Perry, C. M. J. E. Vandenbroucke-Grauls, and J. G. Kusters. 1999. Phase variation in Helicobacter pylori lipopolysaccharide due to changes in the lengths of poly(C) tracts in ␣,3-fucosyltransferase genes. Infect. Immun. 67:5361–5366. 3. Asahi, M., T. Azuma, S. Ito, Y. Ito, H. Suto, Y. Nagai, M. Tsubokawa, S. Maeda, M. Omata, T. Suzuki, and C. Sasakawa. 2000. Helicobacter pylori CagA protein can be tyrosine phosphorylated in gastric epithelial cells. J. Exp. Med. 191:593–602. 4. Aspinall, G. O., M. A. Monteiro, H. Pang, E. J. Walsh, and A. P. Moran.
9. 10. 11.
13. 14. 15. 16. 17. 18.
19. 20. 21. 22.
23. 24. 25. 26.
1996. Lipopolysaccharide of the Helicobacter pylori type strain NCTC 11637 (ATCC 43504): structure of the O antigen chain and core oligosaccharide regions. Biochemistry 35:2489–2497. Blaser, M. J. 1990. Helicobacter pylori and the pathogenesis of gastroduodenal inflammation. J. Infect. Dis. 161:626–633. Cario, E., I. M. Rosenberg, S. L. Brandwein, P. L. Beck, H. C. Reinecker, and D. K. Podolsky. 2000. Lipopolysaccharide activates distinct signaling pathways in intestinal epithelial cell lines expressing Toll-like receptors. J. Immunol. 164:966–972. Cario, E., and D. K. Podolsky. 2000. Differential alteration in intestinal epithelial cell expression of Toll-like receptor 3 (TLR3) and TLR4 in inflammatory bowel disease. Infect. Immun. 68:7010–7017. Censini, S., C. Lange, Z. Xiang, J. E. Crabtree, P. Ghira, M. Borodovsky, R. Rappuoli, and A. Covacci. 1996. cag, a pathogenicity island of Helicobacter pylori, encodes type I-specific and disease-associated virulence factors. Proc. Natl. Acad. Sci. USA 93:14648–14653. Chomezynski, P., and N. Sacchi. 1987. Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction. Anal. Biochem. 162:156–159. Covacci, A., J. L. Telford, G. D. Giudice, J. Parsonnet, and R. Rappuoli. 1999. Helicobacter pylori virulence and genetic geography. Science 284:1328– 1333. De Deken, X., D. Wang, M. C. Many, S. Costagliola, S. Libert, G. Vassart, J. E. Dumont, and F. Miot. 2000. Cloning of two human thyroid cDNAs encoding new members of the NADPH oxidase family. J. Biol. Chem. 275: 23227–23233. Faure, E., O. Equils, P. A. Sieling, L. Thomas, F. X. Zhang, C. J. Kirschning, N. Polentarutti, M. Mizio, and M. Arditi. 2000. Bacterial lipopolysaccharide activates NF-B through Toll-like receptor 4 (TLR-4) in cultured human dermal endothelial cells. J. Biol. Chem. 275:11058–11063. Franz, S., L. Kobzik, Y. D. Kim, R. Fukazawa, R. Medzhitov, R. T. Lee, and R. A. Kelly. 1999. Toll4 (TLR4) expression in cardiac myocytes in normal and failing myocardium. J. Clin. Invest. 104:271–280. Ge. Z., and D. E. Taylor. 1999. Contributions of genome sequencing to understanding the biology of Helicobacter pylori. Annu. Rev. Microbiol. 53: 353–387. Geiszt, M., J. B. Kopp, P. Varnai, and T. L. Leto. 2000. Identification of Renox, an NAD(P)H oxidase in kidney. Proc. Natl. Acad. Sci. USA 97:8010– 8014. Glocker, E., C. Lange, A. Covacci, S. Bereswill, M. Kist, and H. L. Pahl. 1998. Proteins encoded by the cag pathogenicity island of Helicobacter pylori are required for NF-B activation. Infect. Immun. 66:2346–2348. Irie, T., T. Muta, and K. Takeshige. 2000. TAK1 mediates an activation signal from toll-like receptor(s) to nuclear factor-B in lipopolysaccharidestimulated macrophages. FEBS Lett. 467:160–164. Kotani, S., II. Takeda, M. Tsujimoto, T. Ogawa, I. Takahashi, T. Ikeda, K. Otsuka, H. Shimauchi, N. Kasai, J. Mashimo, S. Nagao, A. Tanaka, S. Tanaka, K. Harada, K. Nagaki, H. Kitamura, T. Shiba, S. Kusumoto, M. Imoto, and H. Yoshimura. 1985. Synthetic lipid A with endotoxic and related biological activities comparable to those of a natural lipid A from an Escherichia coli Re-mutant. Infect. Immun. 49:225–237. Moran, A. P., I. M. Helander, and T. U. Kosunen. 1992. Compositional analysis of Helicobacter pylori rough-form lipopolysaccharides. J. Bacteriol. 174:1370–1377. Moran, A. P., B. Lindner, and E. J. Walsh. 1997. Structural characterization of the lipid A component of Helicobacter pylori rough- and smooth-form lipopolysaccharides. J. Bacteriol. 179:6453–6463. Muotiala, A., I. M. Helander, L. Pyhala, T. U. Kosunen, and A. P. Moran. 1992. Low biological activity of Helicobacter pylori lipopolysaccharide. Infect. Immun. 60:1714–1716. Muzio, M., D. Bosisio, N. Polentarutti, G. D’amico, A. Stoppacciaro, R. Mancinelli, C. V. Veer, G. Penton-Rol, L. P. Ruco, P. Allavena, and A. Mantovani. 2000. Differential expression and regulation of Toll-like receptors (TLR) in human leukocytes: selective expression of TLR3 in dendritic cells. J. Immunol. 164:5998–6004. Nielsen, H., S. Birkholz, L. P. Andersen, and A. P. Moran. 1994. Neutrophil activation by Helicobacter pylori lipopolysaccharides. J. Infect. Dis. 170:135– 139. Ninomiya-Tsuji, J., K. Kishimoto, A. Hiyama, J. Inoue, Z. Cao, and K. Matsumoto. 1999. The kinase TAK1 can activate the NIK-IB as well as the MAP kinase cascade in the IL-1 signalling pathway. Nature 398:252–256. Odenbreit, S., J. Pu ¨ls, B. Sedlmaier, E. Gerland, W. Fisher, and R. Hass. 2000. Translocation of Helicobacter pylori CagA into gastric epithelial cells by type IV secretion. Science 287:1497–1500. Ogawa, T., Y. Suda, W. Kashihara, T. Hayashi, T. Shimoyama, S. Kusumoto, and T. Tamura. 1997. Immunobiological activities of chemically defined lipid A from Helicobacter pylori LPS in comparison with Porphyromonas gingivalis lipid A and Escherichia coli-type synthetic lipid A (compound 506). Vaccine 15:1598–1605. Parsonnet, J., G. D. Friedman, D. P. Vandersteen, Y. Chang, J. H. Vogelman, N. Orentreich, and R. K. Sibley. 1991. Helicobacter pylori infection and the risk of gastric carcinoma. N. Engl. J. Med. 325:1127–1131.
ACTIVATION OF Mox1 BY H. PYLORI
VOL. 69, 2001 28. Parsonnet, J., S. Hansen, L. Rodriguez, A. B. Gelb, R. A. Warnke, E. Jellum, N. Orentreich, J. H. Vogelman, and G. D. Friedman. 1994. Helicobacter pylori and gastric lymphoma. N. Engl. J. Med. 330:1267–1271. 29. Perez-Perez, G. I., V. L. Shepherd, J. D. Marrow, and M. J. Blaser. 1995. Activation of human THP -1 cells by Helicobacter pylori lipopolysaccharide. Infect. Immun. 63:1183–1187. 30. Rokutan, K., M. Miyoshi, S. Teshima, T. Kawai, T. Kawahara, and K. Kishi. 2000. Phenylarsin oxide inhibits heat shock protein 70 induction in cultured guinea pig gastric mucosal cells. Am. J. Physiol. Cell Physiol. 279:C1506– C1515. 31. Sharma, S. A., M. K. R. Tummuru, M. J. Blaser, and L. D. Kerr. 1998. Activation of IL-8 gene expression by Helicobacter pylori is regulated by transcription factor nuclear factor-B in gastric epithelial cells. J. Immunol. 160:2401–2407. 32. Stein, M., R. Rappuoli, and A. Covacci. 2000. Tyrosine phosphorylation of the Helicobacter pylori CagA antigen after cag-driven host cell translocation. Proc. Natl. Acad. Sci. USA 97:1263–1268. 33. Suh, Y., R. S. Arnold, B. Lassegue, J. Shi, X. Xu, D. Sorescu, A. B. Chung, K. K. Griendling, and J. D. Lambeth. 1999. Cell transformation by the superoxide-generating oxidase Mox1. Nature 401:79–82. 34. Takeuchi, O., K. Hoshino, T. Kawai, H. Sanjo, H. Takeda, T. Ogawa, K. Takeda, and S. Akira. 1999. Differential roles of TLR2 and TLR4 in recog-
Editor: W. A. Petri, Jr.
35. 36. 37. 38. 39.
nition of Gram-negative and Gram-positive bacterial cell wall components. Immunity 11:443–451. Teshima, S., H. Kutsumi, T. Kawahara, K. Kishi, and K. Rokutan. 2000. Regulation of growth and apoptosis of cultured guinea pig gastric mucosal cells by mitogen oxidase 1. Am. J. Physiol. 279:G1169–G1176. Teshima, S., K. Rokutan, T. Nikawa, and K. Kishi. 1998. Guinea pig gastric mucosal cells produce abundant superoxide anion through an NADPH oxidase-like system. Gastroenterology 115:1186–1196. Teshima, S., S. Tsunawaki, and K. Rokutan. 1999. Helicobacter pylori lipopolysaccharide enhances the expression of NADPH oxidase components in cultured guinea pig gastric mucosal cells. FEBS Lett. 452:243–246. Westphal, O., and K. Jann. 1965. Bacterial lipopolysaccharides: extraction with phenol-water and further application of the procedure. Methods Carbohydr. Chem. 5:83–91. Xiang, Z. Y., S. Censini, P. F. Bayeli, J. L. Telford, N. Figura, R. Rappuoli, and A. Covacci. 1995. Analysis of expression of CagA and VacA virulence factors in 43 strains of Helicobacter pylori reveals that clinical isolates can be divided into two major types and that CagA is not necessary for expression of the vacuolating cytotoxin. Infect. Immun. 63:94–98. Yang, R. B., M. R. Mark, A. Gray, A. Huang, M. H. Xie, M. Zhang, A. Goddard, W. I. Wood, A. L. Gurney, and P. J. Godowski. 1998. Toll-like receptor-2 mediates lipopolysaccharide-induced cellular signaling. Nature 395:284–288.