Molecular and Cellular Biochemistry 228: 83–89, 2001. © 2001 Kluwer Academic Publishers. Printed in the Netherlands.

83

Inhibition of Helicobacter pylori adherence by a peptide derived from neuraminyl lactose binding adhesin G. Chaturvedi,1,3 R. Tewari,1 Mrigank,2 N. Agnihotri,3 R.A. Vishwakarma4 and N.K. Ganguly3 1

Department of Microbiology, Panjab University, Chandigarh; 2Bioinformatics Center, Institute of Microbial Technology, Chandigarh; 3Department of Experimental Medicine and Biotechnology, Post Graduate Institute of Medical Education and Research, Chandigarh; and 4Bio-organic Chemistry Laboratory, National Institute of Immunology, Aruna Asaf Ali Marg, New Delhi, India Received 1 February 2001; accepted 31 July 2001

Abstract Helicobacter pylori, like many other gut colonizing bacteria, binds to sialic acid rich macromolecules present on the gastric epithelium. NLBH (neuraminyl lactose binding haemagglutinin) a 32 kDa adhesin located on the surface of H. pylori has been shown to have specific affinity towards NeuAcα2,3Galβ1,4Gluc(3′SL). This sialic acid moiety is over-expressed in an atrophic stomach undergoing parietal cell depletion. Antibodies against a lysine rich peptide fragment of NLBH inhibit agglutination of human erythrocytes. This lysine rich sequence from NLBH was proposed to be the receptor-binding site. In order to elucidate the binding of NLBH to gastric epithelium, a peptide (D-P-K-R-T-I-Q-K-K-S) was synthesized. A series of experiments were performed involving adherence inhibition assays, 2D-NMR, molecular modelling and measurement of modulation in acid secretion. Results indicated that the peptide fragment could be involved in receptor recognition, which is important for the binding of H. pylori to gastric epithelium. The binding is possibly through hydrogen bonding. Two lysines and a threonine residue seem to be within the hydrogen bonding distance of NeuAcα2,3Galβ1,4Gluc. Further, in vitro assays were performed to evaluate the role of the peptide on acid secretion by parietal cells isolated from human fundal biopsies. Interestingly, the peptide increases acid secretion only in H. pylori negative and in treated patients but not in H. pylori positive patients. This highlights the role of NLBH in acid secretion and could be of some consequence in the prognosis of the disease. (Mol Cell Biochem 228: 83–89, 2001) Key words: H. pylori, NeuAcα2,3Galβ1,4Gluc, adhesin, parietal cells, 2D-NMR, molecular modelling Abbreviations: 3′SL – 3′sialyl lactose or NeuAcα2,3Galβ1,4Gluc; NLBH – neuraminyl lactose binding haemagglutinin; AMBER – assisted model building by energy refinement

Introduction Helicobacter pylori, a microaerophillic, Gram-negative spiral bacillus, colonizes the antrum of the stomach in humans [1]. It is one of the most common bacterial infections worldwide and is now being accepted as the cause of chronic ac-

tive type B gastritis [2]. It has been recognized to play a critical role in the etiology of gastric adenocarcinoma [3] and has also been defined as a class I carcinogen [4]. H. pylori specifically associate with gastric epithelium and colonize it by adhering to sialic acid rich macromolecules expressed on the membrane. Evans et al. [5] first described inhibition of H.

Address for offprints: N.K. Ganguly, Indian Council of Medical Research, Ansari Nagar, New Delhi – 110 029, India (E-mail: icmrhqds&sansad.nic.in)

84 pylori mediated haemagglutination by sialyl lactose derived from bovine milk. Bacterial binding to the murine adrenal cell line Y-1 was diminished after neuraminidase treatment and after pre-incubation of the bacteria with fetuin [6]. Several groups have shown that clinical isolates of H. pylori contain adhesins that bind to NeuAcα2,3Galβ glycoconjugates produced by several human gastrointestinal adenocarcinoma-derived cell lines. Subsequently, the gene for bacterial adhesin subunit (hpaA) of NLBH was cloned and its product HpaA was found to be specific for sialoglycoconjugates [7]. This adhesin contains a lysine rich sequence that is similar to the sialic acid binding motif of the sfas adhesin of S-fimbriated E. coli [8], the CFA/I of enterotoxigenic E. coli [9], and the sialylated ganglioside GM1-binding motif of Vibrio cholerae toxin B subunit [10]. The fibrillar NeuAcα2,3Galβ1,4Gluc binding hemaglutinin (NLBH) on H. pylori surface was identified as a putative colonization factor [7]. Recently, Syder et al. [11] showed that loss of parietal cells in patients with chronic atrophic gastritis is associated with the production of NeuAcα2,3Galβ epitopes. It was suggested that lineage ablation per se does not induce gastritis but hypochlorhydria accompanying the parietal cell loss does lead to severe alteration in the composition of the gastric microflora. Although several treatment regimens including the prevalent triple therapy suppress the infection, after cessation of treatment the stomach usually becomes recolonized by the H. pylori from sanctuary sites [12]. Thus the problem of reinfection is of equal concern as that of the primary infection. The present study aims at designing a synthetic analog that could prevent the recurrent infection. The data highlights the binding properties of the peptide with its receptor NeuAcα2,3Galβ using adherence inhibition assays, 2D-NMR spectroscopy, and homology based molecular modelling and also the role of the adhesin in disease progression in relation to parietal cell depletion.

Materials and methods Culture The bacterial strains used were the clinical isolates EMB70, EMB110 and a standard strain ATCC 43526. Bacteria were grown on blood agar consisting of Brucella broth base (Difco Laboratories, Detroit, MI, USA), 2% agar and 10% defibrinated sheep blood, for 48 h at 37°C in an atmosphere of 12% CO2 and 98% humidity [13]. HEp-2 cells (ATCC CCL-23) were maintained in minimal essential medium (MEM, HiMedia, India) with Earle’s salts containing 25 mM Hepes and 14% heat inactivated, mycoplasma free, fetal bovine serum (Gibco, Grand Island, NY, USA).

Peptide synthesis and purification The decapeptide (D-P-K-R-T-I-Q-K-K-S), which contained the proposed receptor-binding motif, was synthesized using standard FMOC procedure at 0.5 nmol scale over Wang resin (Bachem) (substitution 0.92 mmol/g) on automated peptide synthesiser (model ABI-430A). The peptide was isolated by using a cleavage mixture containing EDTA, thionisole, crystalline phenol and TFA. The peptide was purified on HPLC using Deltapak C18 prep column. Mass spectrometry The purity and the mass of the peptide were checked by ES/ MS. The spectrum was obtained using a VA platform II (Fisons Instruments, UK) quadruple mass spectrometer equipped with a mass LynxTM data system, DynoliteTM detection system and pneumatic nebulizer assisted electrospray LC/MS interface. The mass spectrometer was interfaced to a Jasco PV980 intelligent gradient HPLC system. The analyzer vacuum was maintained at 3 × 10–5 mbar and unit resolution was used for all measurements. Source temperature was maintained at 70°C. Nitrogen at a pressure of 60 psi was used for nebulization of carrier solvent (acetonitrile:water 1:1).

Raising antiserum Hyperimmune serum against the peptide was raised in rabbits (New Zealand white) using the Peptide–BSA conjugate prepared by the method of Walter et al. [14]. Peptide specific antibodies were purified using affinity chromatography by immobilizing the peptide on a CNBr (Cyanogen Bromide) activated Sepharose 4B column (Pharmacia). The titer was checked by ELISA.

Adherence inhibition assay The adherence inhibition assay was performed to see the effect of the peptide on H. pylori binding to HEp-2 cells, using the method of Evans and Evans [15]. Briefly, bacterial suspension made in MEM (Minimum essential medium) (2 × 109 CFUs/ml) was treated with varying concentrations of antipeptide antibody (0–120 ng of final protein concentration), anti-adhesin serum (0–150 ng of final protein concentration) and Fetuin (0–80 ng of final concentrations, Sigma Chemical Co., St. Louis, MO, USA). HEp-2 cells (4 × 106 CFUs/ml/ well) grown in 12 well tissue culture plates (CoStar) in a semiconfluent monolayer form were treated with varying concentrations of the peptide (0–1600 ng). After adding treated bacterial suspension to wells containing untreated HEp-2

85 cells and untreated bacterial suspension to peptide treated HEp-2 cells, the plates were incubated for 2 h at 37oC in an atmosphere of 12% CO2 and 98% humidity. Unbound bacteria were washed, the cells were fixed with 70% methanol and stained with Giemsa stain. Adherent bacteria were counted under oil immersion light microscopy.

Molecular modelling Prediction analysis for the secondary structure and other physicochemical properties of the adhesin was performed with the program Predict 7 [16], using the complete sequence of the adhesin. The region spanning residues 134–139 seemed to be the most likely region to serve as a receptor-binding site. This hexapeptide region with two flanking amino acids on each terminus (D-P-K-R-T-I-Q-K-K-S) was used as a query to search the Protein Database (Brookhaven). The structure of the most homologous sequence resulting from the search with its coordinates was utilized as a starting structure for energy minimization, by replacing the side chains of the nonidentical aminoacids. The protein backbone was optimised by energy minimization and hydrogen atoms were added. The force field employed was the all atom version of AMBER with a distance-dependent dielectric [17]. During the calculations of molecular dynamics, no non-bonded cut-offs were in effect and no restraints were applied. Extensive energy minimization was then performed by the successive application of steepest descent and conjugate gradient minimization modules. Minimization of the model was continued until the maximum derivation of energy fell below 1.0 kcal/mol (4.2 kJ/mol). The structure of the receptor (3′SL) was developed de novo using standard distance and angle parameters for sugars. The software used for energy minimization and molecular dynamics, AMBER 4.0 [17], was loaded on an Alfa DEC-3000 workstation from Digital Inc. The two structures were then docked using an in-house program Imf2K [18]. The complex so obtained was further subjected to energy minimization and hydration.

NMR The 1D 1H-NMR spectra, 2D 1H-1H double quantum filter COSY (correlation spectroscopy) [19], TOCSY (total correlation spectroscopy) [20] (MLEV-17, T-60 msec) and ROESY (rotating frame Overhauser spectroscopy) spectra were obtained in a 25 mM decapeptide in 10% 2H2O/90% H2O, with presaturation during relaxation delay, on a Bruker NMR spectrometer (Avance-DRX, 300 MHz for 1H). All the data were accumulated in the phase sensitive mode using time proportional phase incrementation (TPPI). The typical data set size was 2043 (t2) × 512 (t1) data points in the time domain with

32 and 64 scans/t1 experiment. All the spectra were recorded at a temperature of 300 K. To obtain all the above spectra for the peptide in its bound form, 3′SL was added in an equimolar concentration to the peptide solution. Chemical shifts are expressed in δ (ppm) relative to TMS (tetra methyl silane) as internal standard for 1H-NMR. The softwares used for data processing were XWIN-NMR and UX-NMR on an Aspect station with a UNIX operating system. The resonance assignments were made on the basis of phase sensitive DQF-COSY and TOCSY and the sequential assignments were made using ROESY spectra.

Measurement of acid secretion The measurement of acid secretion was done by the method of Mardh et al. [21]. To a suspension of isolated parietal cells in 1 ml HBSS (Hank’s balanced salt solution) containing 1% BSA, 4 µl of acridine orange (50 mM) was added. The fluorescence of this suspension was measured immediately after addition of acridine orange, at 493/530 nm (emission/excitation) wavelength (A). The cells (106 cells/ml) were then incubated with the peptide (30 µl of 50 mg/ml), for 5 min at 37°C and the fluorescence was measured again (B). Percent quenching of fluorescence was expressed as (A-B)/A × 100.

Results Adherence inhibition assay revealed that all the four inhibitors, i.e. NLBH derived peptide, peptide specific antibodies, anti-adhesin antiserum and fetuin, were able to significantly inhibit the adherence of H. pylori to HEp-2 cell monolayers in a dose dependent manner (Fig. 1). Anti-peptide antibody was more effective in inhibiting the binding as compared to antiadhesin serum. The clinical isolates EMB70 and EMB110 showed more adherence in terms of number of bacteria per HEp-2 cell as compared to the standard strain (ATCC 43526). Moreover, the standard strain was less sensitive to inhibition by all four inhibitors. We performed prediction analysis on the complete sequence of the adhesin (NLBH) to look for potential binding sites. The analyses indicated that the peptide motif, proposed to be the receptor binding site, had significant degree of hydrophilicity, surface probability, flexibility and antigenicity but no glycosylation sites, which fulfil the basic requirements for a peptide fragment to be a binding site. The secondary structure prediction showed random coil conformation for the peptide motif. The results of prediction analyses led us to perform sequence homology based molecular modelling. The sequence homology search in Protein Database revealed that the pep-

86

Fig. 1. H. pylori adherence inhibition assay using (a) peptide (D-P-K-R-T-I-Q-K-K-S); (b) anti-peptide mono-specific antibodies; (c) anti-adhesin serum; (d) Fetuin. L (ATCC 43526); I (EMB70); G (EMB110). The values on X-axis are final concentrations of inhibitors. Data points are the mean of triplicate wells. Standard deviations were all within 15% of mean values. These results have been verified in triplicate assays.

tide exhibits 50–60% sequence homology to fragments of interleukin-1β human (PDB code-1|1B) and glutathione reductase (PDB code – 3 grs). After extensive energy minimization and 200 psec of molecular dynamics, 30% of φ and ψ values of the energy minimised peptide model were found to be in the most favoured regions and 70% in additionally allowed regions of the Ramachandran plot, which emphasizes the feasibility and conformational stability of the peptide model. Taking into consideration the nature of amino acid residues and the size of the peptide, it was not expected to have any defined conformation. To confer rigidity to the peptide, it was required to be in the receptor bound form. The peptide was docked to the receptor. The model of the complex is represented in stick form in Fig. 2. Two-dimensional NMR spectra showed a distinct change upon addition of 3′SL. The chemical shift assignments (Table 1) for all the protons of the peptide were made from double quantum filtered COSY (Fig. 3) and TOCSY data and the sequential assignments were made from the ROESY data.

Fig. 2. Stick model of the peptide-receptor complex. Peptide shown is a tetradecamer with two flanking amino acids on each terminus of the decapeptide. The decapeptide residues forming H bonds (shown by dotted lines) with 3′SL (receptor) are labeled.

87 Table 1. Chemical shift assignments of the peptide protons Residue 1

Asp Pro 2 Lys3 Arg 4 Thr5 Ile6 Gln7 Lys8 Lys9 Ser10

NH

Hα

Hβ1

Hβ2

Others

8.51 – 8.40 8.52 8.33 8.32 8.49 8.74 8.34 8.23

4.39 4.40 4.22 4.35 4.31 4.25 4.42 4.36 4.40 4.33

2.4 2.46 1.85 2.10 2.10 1.90 2.0 1.85 1.82 3.95

– 2.12 1.75 1.85 – 1.50 1.90 1.75 1.75 3.95

– Hγ1 2.10, Hγ2 2.10, CH2ε 3.41, 3.41 Hγ1 1.48, Hδ1 1.75 Hδ2 1.75, Hε1 3.05, NHε 7.5 Hγ1 1.70, Hγ2 1.70, Hδ 3.26 NHε 7.2, 6.7 CH3γ 1.25 CH2γ 1.50, 1.35 Hδ1 0.9, CH3γ 0.95 CH2γ 2.40, NH2δ1 6.68 , 7.60 Hγ1 1.48, Hδ1 1.75, Hδ2 1.75, Hε 3.05 , NHε 7.5 Hγ1 1.48, Hδ1 1.78 , Hδ2 1.75, Hε 3.05, NHε 7.5 –

(8.42) (8.29) (8.30)

(8.37) (8.15)

(4.27)

(4.35)

(4.45) (4.35)

The chemical shifts are given in δ (ppm). The change in chemical shift is shown in brackets.

After assignments were made, 3′SL was added to the peptide solution in an equimolar ratio. The DQF-COSY spectra revealed a change in chemical shift assignments of the peptide protons (in brackets, Table 1). The chemical shift perturbation was observed in the case of four of the residues in the hexapeptide fragment. Lys3, Thr5, Ile6, Lys9 and Ser10 seemed to be involved in making contact with the trisaccharide receptor. In addition, an attempt was made to monitor the effect of the NLBH derived peptide on acid secretion by isolated parietal cells. The results revealed that the peptide induces an increase in acid secretion by parietal cells isolated from the human fundal biopsies of H. pylori negative patients in comparison to H. pylori positive patients (Fig. 4). On an average

the number of parietal cells recovered from H. pylori positive patients was significantly less as compared to those from negative or treated patients.

Discussion Earlier inhibition experiments with oligosaccharides derived from laminin indicated that H. pylori binding is specific for the 3' sialyllactose not the 6'-sialyl lactose isomer [22]. Inhibition of laminin binding by fetuin and reduced bacterial binding to periodate or sialidase treated laminin indicated that sialylation was important for laminin binding [23]. The ad-

Fig. 3. DQF-COSY spectra. Comparison of DQF-COSY spectra of the peptide to 3′SL (a) and peptide alone (b). One cross peak (7.8, 3.8) in (a) could not be assigned to any residue.

88

Fig. 4. Acid secretion profile. Response of peptide treatment to parietal cells isolated from three categories of patients, i.e. (A) for H. pylori –ve, (B) for H. pylori +ve and (C) for treated patients. All the categories had 10 patients and each data point represents value of a single patient.

herence inhibition assay results showed significant inhibition of binding which further reflects the specificity of NLBH towards NeuAcα2,3Galβ1,4Gluc receptor on Hep-2 cells Use of extensively passaged strains can result in the failure in localization of NLBH on the bacterial surface and reduced adherence to their specific receptors [24–26]. Our results with standard strain ATCC 43526 showed lesser binding or reduced sensitivity to various inhibitors used and thereby support the role of N-acetyl neuraminyl lactose (3′SL) in bacterial adhesin recognition for binding. The analysis of the docking experiment indicated that the hexapeptide (3rd–8th residue, i.e. K-R-T-I-Q-K), of the tetradecamer peptide, had maximum interaction with the receptor. The binding appeared to be primarily due to hydrogen bonding and van der Waal’s forces and resulted in the formation of three strong hydrogen bonds. Lys3, Thr5 and Lys9 made hydrogen bonds with 3′SL. The NMR and Modelling results substantiated the role of lysine and isoleucine. Arginine did not play any role in the interaction, since there was

no change in the chemical shift assignment of this residue after the addition of the receptor, as indicated by the 2D-NMR spectroscopy. In fact the arginine side chain was observed to be facing away from the binding pocket as shown by the modelling studies. The acid secretion results suggest that gastric epithelial cells and parietal cells share this receptor molecule as also reported by Guruge et al. [27]. Bacterial ligands bind to these receptors and trigger acid secretion and apoptosis, leading to parietal cell ablation, which enhances the expression of 3′SL receptor on gastric epithelial cells [11]. It is becoming increasingly evident that during the progression of inflammation, the bacteria are able to maintain an acid-base balance in their vicinity. Acid neutralizing substances like ammonia and certain fatty acids [28] are also produced by H. pylori to maintain the acid-base balance, which presumably is important for its survival. It appears that as a result of increased expression of 3′SL, which is due to parietal cell ablation [11], the bacteria bind more strongly to the epithelium leading to increased inflammation and ulcerogenesis. Two main constraints with the present treatment regimens are lack of compliance with the triple therapy, and the resistance of H. pylori to metronidazole [29, 30]. Several patients experience pseudomembranous colitis and various other side effects. Resistance to metronidazole is appreciable in western countries at about 10–20% of subjects and tends to be higher in women, possibly because of previous use of this drug for pelvic infections [30]. There is a dire need to develop newer strategies to deal with the problem of infection and reinfection by H. pylori. Our strategy differs from that used by Simon et al. [22] in that we have used the peptide to block the receptor instead of using the receptor to block the peptide, which would be a better approach for achieving the inhibition of NLBH mediated adherence of H. pylori to gastric mucosa, as the receptor forms the immobile phase on the gastric mucosa. In light of the above results and extensive distribution of NLBH on the surface of H. pylori [7], the peptide could be a suitable lead candidate for a potential therapeutic molecule for preventing the colonization of this organism in the stomach. We propose that by elucidating the conformation of the peptide in its receptor-bound form, a nonpeptidic synthetic analog can be designed with the help of combinatorial chemistry and molecular modelling. This analog can then be chemically modified, so as to have acid resistance and a greater binding affinity to the receptor than the bacterial adhesin. Such a molecule can be modified to be toxicologically inert so as to be included as a dietary supplement to continuously keep the gastric wall off the bacterial load.

89

Acknowledgements We thank Dr. V.S. Chauhan at ICGEB, New Delhi, India for providing the peptide, Dr. D.G. Evans at Veterans Affairs Center, Houston, TX, USA for providing antiadhesin serum, and Dr. Douglas E. Berg at Department of Medicine, Washington State University, St. Louis, MO, USA, for providing the standard strain of Helicobacter pylori (ATCC 43526)

References 1 . Kobayashi Y, Okazaki K-I, Murukami K: Adhesion of Helicobacter pylori to gastric epithelial cells in primary cultures obtained from the stomachs of various animals. Infect Immunol 61: 4058– 4063, 1993 2 . National Institute of Health: Helicobacter pylori in peptic ulcer disease NIH Consensus statement. 12(1), Bethesda, MD. USA, 1994 3 . Talley NJ, Zinseimer AR, Weaver A, DiMagno EP, Carpenter HA, Perez-Perez GI, Blaser MJ: Gastric adenocarcinoma and Helicobacter pylori infection. J Natl Cancer Inst 83: 1734–1739, 1991 4 . Logan RP: Helicobacter pylori and gastric cancer. Lancet 344: 1078–1079, 1994 5 . Evans DG, Evans DJ, Moulds JJ, Graham DY: N-Acetyl neuraminyl lactose binding fibrillar hemagglutinin of Campylobacter pylori: A putative colonizing factor antigen. Infect Immunol 56: 2896–2906, 1988 6 . Evans DG, Evans DJ, Graham DY: Receptor mediated adherence of Campylobacter pylori to mouse Y-1 adrenal cell monolayers Infect Immunol 57: 2272–2278, 1989 7 . Evans DG, Karjalainen TK, Evans DJ, Graham DY, Lee CH: Cloning and nucleotide sequence and expression of a gene encoding an adhesin subunit protein of Helicobacter pylori. J Bacteriol 175: 674–680, 1993 8 . Parkkinen J, Rogers GN, Korhonen T, Dahr W, Finne J: Identification of the O-linked sialyl oligosaccharides of glycophorin A as the erythrocyte receptors for s fimbriated Escherichia coli. Infect Immunol 54: 37–43, 1986 9 . Karjalainen TK, Evans DG, So M, Lee CH: Molecular cloning and nucleotide sequence of the colonizing factor antigen I of Escherichia coli. Infect Immunol 57: 1126–1130, 1989 10. Finkelstein RA, Burks MF, Zupan A, Dallas WS, Jacob CO, Ludwig DS: Epitopes of the cholera family of enterotoxins. Rev Infect Dis 9: 544–561, 1987 11. Syder AJ, Guruge JL, Li Q, Hu Y, Olekseiwicz CM, Lorenz RG, Karam SM, Falk PG, Gordon JI: Helicobacter pylori attaches to Neu-Ac alpha2, 3 gal beta 1,4 glycoconjugates produced in the stomach of mice lacking parietal cells. Mol Cell 3: 263–274, 1999 12. Tytgat GNJ, Noach L, Rauws EAJ: Helicobacter pylori. Scand J Gastroenterol 26: 1–8, 1991 13. Graham DY, Klein PD, Evans DJ Jr, Evans DG, Alpert LC, Opekun AR, Boutton TW: Campylobacter pylori detected noninvasively by the 13C urea breath test. Lancet ii: 1174–1181, 1989

14. Walter G, Scheidtmann KH, Canbone A, Laudano AP, Dolittle RF: Antibodies specific for the carboxy and amino terminal regions of simian virus 40 large tumor antigen. Proc Natl Acad Sci USA 77: 5197–5200, 1980 15. Evans DG, Evans DJ Jr: Adhesion properties of Helicobacter pylori. Meth Enzymol 253: 336–360, 1995 16. Garnier J, Osguthorpe DJ, Robson BJ: Analysis of the accuracy and implications of simple methods for predicting the secondary structure of globular proteins. Mol Biol 120: 97–120, 1978 17. Pearlman DA, Case DA, Caldwell JC, Seibel GL, Singh UC, Weiner P, Kollman PA: AMBER Version 4.0, UCSF, USA, 1991 18. Kothekar V, Mrigank: Computer aided geometry optimization with energy minimization technique. J Phys Educ 5: 169–174, 1988 19. Rance M, Sorensen OW, Bodenhousen G, Wagner G, Ernst RR, Wuthrich K: Improved spectral resolution in COSY 1H NMR spectra of proteins via double quantum filtering. Biochem Biophys Res Commun 117: 479–485, 1983 20. Kessler H, Seip S: NMR of peptides. In: W.R. Croasmun, R.M.K. Carlson (eds). NMR Spectroscopy. Applications for Chemists and Biochemists NMR Spectroscopy. VCH, 1994, pp 619–650 21. Mardh S, Norberg L, Ljungstrom M, Wallert S, Nyren O, Gustavson S: A method for in vitro studies on acid formation in human parietal cells. Stimulation by histamine, pentagastrin and carbachol. Physiol Scand 123: 349–354, 1985 22. Simon PM, Goode PL, Mobasseri A, Zopf D: Inhibition of Helicobacter pylori binding to gastrointestinal epithelial cells by sialic-acid containing oligosaccharides. Infect Immunol 65: 757, 1997 23. Valkonen KH, Wadstrom T, Moran AP: Identification of N-acetylneuraminyllactose specfic laminin binding protein of Helicobacter pylori. Infect Immunol 65: 916–923, 1997 24. Boren T, Falk P, Roth KA, Larson G, Normark S: Attachment of Helicobacter pylori to human gastric epithelium mediated by blood group antigens. Science 262: 1892–1895, 1993 25. Fauchere JL, Blaser MJ: Adherence of Helicobacter pylori cells and their components to HeLa cell membranes. Microb Pathog 9: 427–439, 1990 26. O’Toole PW, Janson L, Doig P, Huang J, Kostzynska M, Trust TH: The putative neuraminyllactose binding hemagglutinin HpaA of Helicobacter pylorus CCUG 17874 is a lipoprotein. J Bacteriol 177: 6049–6057, 1995 27. Guruge JL, Falk PG, Lorenz RG, Dans M, Wirth HP, Blaser MJ, Berg DE, Gordon JI: Epithelial attachment alters the outcome of Helicobacter pylori infection. Proc Natl Acad Sci USA 95: 3925– 3930, 1998 28. Smout AJ: Is the sensitivity to gastric acid inhibition Helicobacter pylori status-dependent? Scand J Gastroenterol Suppl 225: 32– 35, 1998 29. Graham DY, Go MF, Evans DJ: Urease, gastric ammonium, ammonia and Helicobacter pylori – the past, the present and recommendations for future research. Aliment Pharmacol Ther 6: 659– 669, 1992 30. Graham DY, Lew GM, Malaty HM, Evans DG, Evans DJ Jr, Klein PD, Alpert LC, Genta RM: Factors influencing eradication of Helicobacter pylori with triple therapy. Gastroenterology 102: 493–496, 1992

90