Lack of a relationship between Lewis antigen expression and caga, CagA, vaca and VacA status of Irish Helicobacter pylori isolates

Transcription

1 FEMS Immunology and Medical Microbiology 24 (1999) 79^90 Lack of a relationship between Lewis antigen expression and caga, CagA, vaca and VacA status of Irish Helicobacter pylori isolates David G. Marshall a, Sean O. Hynes b, David C. Coleman c, Colm A. O'Morain d, Cyril J. Smyth a, Anthony P. Moran b; * a Department of Microbiology, Moyne Institute for Preventive Medicine, Trinity College, Dublin, Ireland b Department of Microbiology, National University of Ireland, Galway, Ireland c Department of Oral Medicine and Pathology, School of Dental Science, Trinity College, Dublin, Ireland d Department of Clinical Medicine and Gastroenterology, Meath/Adelaide Hospital, Dublin, Ireland Received 26 October 1998; received in revised form 25 January 1999; accepted 28 January 1999 Abstract The caga gene, vaca gene, CagA (cytotoxin-associated gene A product) and VacA (vacuolating cytotoxin) status of a collection of Helicobacter pylori isolates from the geographically distinct Irish population was determined, the potential association of these traits with Lewis (Le) antigen expression was assessed, and the relationship between these bacterial properties and the pathology associated with H. pylori infection was evaluated. Of the 57 isolates, a higher proportion from ulcer than from non-ulcer patients expressed VacA (71% vs. 53%). H. pylori isolates which were caga-positive were no more significantly associated with peptic ulcers than non-ulcer disease (71% vs. 67%, P = 0.775), nor were CagA-positive isolates (57% vs. 50%, P = 0.783), but 80% of the isolates from duodenal ulcer patients were caga-positive. Thirty-seven of the 57 isolates were tested for Le antigen expression. No statistically significant relationship (P s 0.05) was found between the occurrence and level of expression of Le x or Le y and caga, vaca, or VacA status. This lack of an association in the Irish H. pylori isolates contrasts with that previously reported for predominantly North American isolates, and may be attributable to the adaptation of H. pylori strains with differing attributes to different human populations. z 1999 Federation of European Microbiological Societies. Published by Elsevier Science B.V. All rights reserved. Keywords: Helicobacter pylori; caga; vaca; Lipopolysaccharide; Lewis antigens; Virulence 1. Introduction * Corresponding author. Tel.: +353 (91) , ext. 3163; Fax: +353 (91) ; anthony.moran@nuigalway.ie Helicobacter pylori is an important human gastric pathogen associated with a range of clinical pathologies [1]. In particular, H. pylori is the causal agent of active chronic gastritis and chronic infection with H. pylori is associated with the development of duodenal and gastric ulcers [2^4]. This bacterium is also associated with an increased risk for the development of both gastric adenocarcinoma and primary gastric lymphoma [5^7]. H. pylori isolates which express vacuolating cytotoxin, VacA, and another pro / 99 / $20.00 ß 1999 Federation of European Microbiological Societies. Published by Elsevier Science B.V. All rights reserved. PII: S (99)

2 80 D.G. Marshall et al. / FEMS Immunology and Medical Microbiology 24 (1999) 79^90 tein, CagA, have previously been associated with more aggressive pathologies, such as peptic ulcer disease [8^13]. The caga gene is found, along with genes associated with pilus or agellum assembly and secretion pathways within the cag pathogenicity island [14] and isolates carrying this island are considered more virulent [15,16]. Like the cell envelope of other Gram-negative bacteria, that of H. pylori contains lipopolysaccharides (LPSs). Fresh clinical isolates of H. pylori produce high molecular weight smooth-form LPSs, which consist of an O side chain, core oligosaccharide and lipid A [17,18]. Both chemical structural studies [19^21] and serological investigations [22^24] have shown that the O side chains of certain H. pylori strains exhibit mimicry of Lewis (Le) blood group determinants, in particular Le x and/or Le y determinants. The pathogenic relevance of Le antigen mimicry remains unclear, but this mimicry has been suggested to camou age the bacterium in the gastric mucosa, aid bacterial adhesion, contribute to the development of gastritis by the induction of autoantibodies, and in uence the in ammatory response in the gastric mucosa [25^27]. Moreover, it has been suggested that expression of host Le antigens by the bacterium could aid the persistence of H. pylori proin ammatory strains expressing the CagA protein [24,28]. Consistent with this, Wirth et al. [24] showed that H. pylori isolates positive for both Le x and Le y were predominantly caga-positive and that a caga-ablated strain had diminished expression of Le y. Although small numbers of isolates from di erent countries were examined in the latter study, predominantly isolates were from North America. Our preliminary investigation of H. pylori isolates from the geographically distinct Irish population has indicated the lack of an association between caga status and Le antigen expression [29]. The aim of the present study was to expand these investigations to establish whether a relationship exists between caga or vaca status, their encoded proteins CagA and VacA, and Lewis antigen expression in H. pylori strains isolated from Irish patients. Also, we attempted to ascertain the extent of any relationship between these bacterial properties and the pathologies observed. 2. Materials and methods 2.1. Bacterial strains and growth conditions H. pylori NCTC was obtained from the National Collection of Type Cultures (Colindale, London, UK). Clinical isolates were obtained from antral biopsies of patients undergoing upper gastrointestinal endoscopy at the Meath, Adelaide and St. James's hospitals in Dublin, Ireland. The clinical diagnoses of individuals from whom H. pylori isolates were recovered was established in a blinded manner, following histological analyses according to the Sydney system [30]. For the primary culture of H. pylori from biopsy samples, biopsy specimens (2^4 mm in diameter) were inoculated onto blood agar containing Skirrow's selective supplement (Oxoid, Basingstoke, UK), incubated microaerobically at 37³C for 5^7 days [31,32], and isolates were con rmed as H. pylori according to the criteria previously established [32]. H. pylori strains were grown routinely on blood agar under microaerobic conditions at 37³C for 48^ 72 h, and harvested in phosphate-bu ered saline, ph 7.4 (PBS) after centrifugation (12 000Ug, 5 min). Escherichia coli J5 was grown aerobically on nutrient agar (Oxoid) at 37³C for 24 h. Stock cultures were maintained at 370³C in 15% (v/v) glycerol-trypticase soy broth (BBL Microbiology systems, Cockeysville, MD) Isolation of H. pylori total genomic DNA Harvested bacterial pellet was resuspended in 567 Wl of TE bu er (l0 mm Tris-HCl ph 8.0 and 1 mm EDTA), and genomic DNA was extracted using an adaptation of the cetyltrimethyl ammonium bromide-nacl (CTAB) method of Ausubel et al. [33]. The DNA concentration was estimated by agarose gel electrophoresis against phage lambda DNA of known concentration DNA probes Plasmid pmc3, which contains a 3.6-kbp fragment of the caga gene of H. pylori strain [10], was cleaved with the restriction endonuclease

3 D.G. Marshall et al. / FEMS Immunology and Medical Microbiology 24 (1999) 79^90 81 HindIII. A 1.5-kb fragment was subsequently isolated from an agarose gel and labelled, and used in Southern hybridisations as a caga-speci c gene probe. The 2-kb EcoRI^BglII fragment of plasmid pws18, containing the vaca gene from H. pylori [34], was used as the probe for the presence of the vaca gene. The linearised DNA fragments to be used as DNA probes in hybridisation analysis were labelled with [K- 32 P]dATP to a speci c activity of s 10 6 dpm Wg 31 DNA by random hexanucleotide priming by means of the Prime-a-Gene kit (Promega, Madison, WI) Southern blot analysis for identi cation of caga and vaca genes The genomic DNA of bacterial strains was cleaved with restriction endonuclease HindIII, electrophoresed on 0.8% (w/v) agarose gels, transferred to nylon membranes, and subsequent prehybridisation and hybridisation of membranes were performed according to standard protocols [33]. Hybridised lters were exposed to Cronex 10 s lm (Dupont, Stevenage, UK) using two Quanta Rapid intensifying screens (Dupont) for between 4 and 72 h at 370³C, and subsequently developed according to the manufacturer's instructions. Bound probe was stripped from lters by immersion twice in a solution of 0.5% (w/v) SDS for 10 min, followed by a brief wash in distilled water Preparation of bacterial whole cell (WC) lysates and proteinase K (PKWC) digests Bacterial cells were washed twice in PBS containing protease inhibitors (Complete1, protease inhibitor cocktail tablets; Boehringer, Mannheim, Germany) and resuspended in PBS (200 Wl) to give 4 absorbance units at 600 nm. Subsequently, WC lysates were prepared from 30 Wl of this cell suspension using electrophoresis sample bu er as described previously [35]. For production of LPS mini-preparations, PKWC digests were prepared as described by Hitchcock and Brown [35] and stored at 370³C before electrophoretic analysis Antisera and monoclonal antibodies Anti-CagA antibodies were raised against an electroeluted 120-kDa CagA protein band of H. pylori strain ATCC (courtesy of A. Cockayne, University of Nottingham, UK). New Zealand white rabbits were immunised intramuscularly with the puri ed CagA protein in Freund's complete adjuvant on day 0, and in Freund's incomplete adjuvant on days 7, 14 and 21. When an adequate antibody titre was obtained, rabbits were exsanguinated and the antiserum was absorbed with WC lysates of H. pylori MI 119, a strain which is negative for the caga gene as determined by Southern hybridisation studies. Two di erent antisera against toxigenic VacA protein, K and L were used in these studies and were provided by D. Burroni, IRIS-Biocine, Sienna, Italy. The K- and L-antisera were raised against recombinant N-terminal fragments of the VacA protein, comprising amino acid residues 34^132 and 262^ 428, of the 1300 amino acid residues of the VacA protein, respectively [36]. These antisera have been shown not to detect VacA protein in cell extracts of cytotoxin-negative strains, broth culture supernatants from which failed to produce vacuolation of HeLa cells [36,37]. Murine monoclonal antibodies (Mabs) of the IgM class that were raised against Le a (clone T174), Le b (clone T128), Le x (clone P12), Le y (clone F3), sialyl- Le x (clone CSLEX1) and H type 1 chain (clone ) were obtained from Signet Laboratories (Dedham, MA) Sodium dodecyl-sulphate polyacrylamide gel electrophoresis (SDS-PAGE) and immunoblotting WC lysates and PKWC lysates of H. pylori strains were analysed by SDS-PAGE with the discontinuous bu er system described by Laemmli [38]. Electrophoresis was conducted using the gel systems and voltage conditions described previously [18,32]. Separated proteins of WC lysates and LPS of PKWC were electroblotted onto nitrocellulose membranes (BA85, 0.45-Wm pore, Schleicher and Schuell, Dassel, Germany) according to the procedure of Towbin et

4 82 D.G. Marshall et al. / FEMS Immunology and Medical Microbiology 24 (1999) 79^90 al. [39]. Visualisation of nitrocellulose blots was performed with anti-caga or anti-vaca antisera or anti-lewis Mabs as primary antibody and the appropriate horse radish peroxidase-conjugated secondary antibody diluted in 0.5% (w/v) blocking reagent (Boehringer): for CagA, swine anti-rabbit IgG1 (Dako, Cambridge, UK) diluted 1:500; for VacA, rabbit anti-mouse IgG1 diluted 1:1000; for Le antigens, goat anti-mouse IgM (Sigma, St. Louis, MO) diluted 1:1000. Membranes were developed using the BM chemiluminescent blotting substrate (Boehringer) and exposed to Cronex 10 s lm at 20³C for 25^30 min or using a horse radish peroxidase substrate development kit (Bio-Rad). Immunoblotting of lysates from isolates which were CagA-negative or VacA-negative were repeated to con rm results. In addition, LPS fractionated by SDS-PAGE was detected by silver staining as described previously [40] Enzyme-linked immunosorbent assay (ELISA) for determination of Lewis antigen expression Protein concentrations of bacterial suspensions were determined using a Pierce protein assay detection system (Pierce, Rockford, IL). Flat-bottomed microtitre plates were coated overnight with 100 Wl of H. pylori cell suspensions, with a protein concentration of 60 Wg ml 31, in 0.05 M bicarbonate coating bu er ph 9.6 and blocked with 3% (w/v) bovine serum albumin (BSA) for 2 h at 20³C. Subsequently, the ELISA assay was performed by the procedure described by Wirth et al. [24]. The speci city of Mabs was validated by their ability to recognise the respective antigen from a panel of synthetic Lewis antigens (Isosep AB, Tullinge, Sweden and Dextra Laboratories, Reading, UK) and the LPSs of H. pylori strains NCTC 11637, P466 and MO19 whose O side chains contain polymeric Le x, polymeric Le x terminated with a Le y unit, and a Le y unit, respectively [19^21]. In addition to controls without primary or secondary antibody, wells coated with Escherichia coli J5 were included in order to determine the level of nonspeci c binding. The OD values of the wells coated with E. coli cells were subtracted from the values of the wells coated with equal protein concentrations of H. pylori whole cell preparations. Non-speci c binding values never exceeded an absorbance of 0.3 at 492 nm and, hence, absorbance values were considered positive for the presence of Le antigen if s 0.3. Assays were repeated in triplicate Statistics Statistical analysis was carried out using a twotailed Fisher's exact test when comparing the signi cance of proportions or a two-tailed Student's t-test when comparing the signi cance of mean levels of antigen expression. Both were performed with 95% con dence limits and a probability value P was considered signi cant. 3. Results 3.1. H. pylori strains Fifty-seven clinical isolates of H. pylori and the culture collection strain NCTC were examined in this study. Twenty-one clinical isolates were recovered from patients with peptic ulcer disease (PUD); nine isolates were from patients with duodenal ulcers, six from patients with gastric ulcers, and the remaining six from patients with both gastric and duodenal ulcers. Thirty-six isolates were recovered from patients with dyspeptic conditions with nonulcer associated disease (NUD); 15 from patients with asymptomatic gastritis, 11 from patients with non-ulcer dyspepsia, and ten from patients with re- ux oesophagitis. Table 1 Molecular typing a of 57 H. pylori clinical isolates and reference strain NCTC Strain type caga vaca CagA VacA Number of strains I b Ia Ib Ic Id II a Strain typing according to the scheme of Xiang et al. [37]. b H. pylori NCTC is included in this strain type.

5 D.G. Marshall et al. / FEMS Immunology and Medical Microbiology 24 (1999) 79^90 83 Fig. 1. Representative results of Southern hybridisation analyses of HindIII-digested DNA from H. pylori isolates with (a) vaca and (b) caga gene probes. Lanes: 1, MI 291; 2, MI 281; 3, MI 297; 4, MI 262; 5, MI 215; 6, MI 247; 7, MI 227; 8, MI 133; and 9, MI 146. Numbers on the left indicate molecular size markers (kbp) Detection of caga, CagA, vaca and VacA in isolates Using Southern blot analysis, 57 clinical isolates of H. pylori and NCTC were analysed for the presence of the caga and vaca genes (Table 1). HindIII-cleaved DNA of H. pylori NCTC was the positive control for detection of caga and vaca (Fig. 1). All isolates possessed the vaca gene, whereas 68% of the isolates possessed caga (Table 1). Expression of the CagA and VacA proteins was determined by immunoblotting with anti-caga antiserum and with two di erent antisera (K and L) against toxigenic VacA. For CagA studies, H. pylori MI 355 and MI 121 were the positive and negative controls, respectively (Fig. 2). Immunoblotting of WC lysates revealed that 53 and 60% of the isolates expressed the CagA and VacA proteins, respectively (Table 1). The K- and L-antisera had di erent speci cities; some isolates which were non-reactive with VacA K-antiserum reacted with L-antiserum, and vice versa (Fig. 2). Of the isolates, 39% expressed both CagA and VacA (Table 1), and were the largest group of strains (type I) in the typing system according to Xiang et al. [37]. Only 14% of the isolates were only vaca-positive (type II strains), and the remaining 47% were of intermediate strain types.

6 84 D.G. Marshall et al. / FEMS Immunology and Medical Microbiology 24 (1999) 79^90 Fig. 2. Representative results of immunoblotting analyses of whole cell lysates from H. pylori isolates with (a) anti-vaca K-antiserum, (b) anti-vaca L-antiserum, and (c) anti-caga antiserum. Lanes: 1, MI 125; 2, MI 281; 3, MI 273; 4, MI 215; 5, MI 311; 6, MI 389; 7, MI 355; 8, MI 121; and 9, NCTC Numbers on the left indicate molecular mass markers (kda) Detection of Le antigens in isolates Expression of high molecular weight LPS in all H. pylori isolates was con rmed by SDS-PAGE of PKWC digests and silver staining (data not shown). Due to our inability to obtain su cient biomass from all the H. pylori isolates, 37 of the 57 isolates (63%) and H. pylori NCTC were subsequently tested for Le antigen expression in an ELISA system. As expected, H. pylori NCTC expressed Le x and Le y. The antigenic determinants Le a,le b, sialyl- Le x and H type 1 chain were not detected in the H. pylori isolates examined. However, expression of Le x and/or Le y epitopes were detected on 33 of the 37 (89%) isolates, with Le x expressed on 20 (54%) and Le y on 33 (89%) of the isolates. Immunoblotting demonstrated the binding of anti-le x and anti-le y Mabs to high molecular weight LPS in a similar manner to that described previously [17,18,22,24]. No H. pylori isolate was found which expressed Le x without the Le y determinant, and four (11%) of the isolates did not express either Le x or Le y Lack of association of Le expression with the caga gene The caga gene was detected in 68% of the isolates examined for Le status (Table 2). A higher proportion of these caga-positive isolates expressed Le x and

7 Table 2 Relationship between Le expression and the caga and vaca genotypic and phenotypic status of 37 H. pylori clinical isolates Genotypic/phenotypic status of isolates Number of isolates Lewis expression by isolates Le x -positive Two-tailed (n = 20) a P-value Le y -positive Two-tailed (n = 33) a P-value caga-positive caga-negative CagA-positive CagA-negative VacA-positive VacA-negative a Some isolates were both Le x - and Le y -positive. D.G. Marshall et al. / FEMS Immunology and Medical Microbiology 24 (1999) 79^90 85 Le y (60 and 96%, respectively) than those that were caga-negative (42 and 75%). However, the numbers of caga-positive isolates expressing or not expressing Le x did not di er signi cantly in a two-tailed Fisher's exact test from similar groups of caga-negative isolates. Likewise, using similar analysis, no signi cant di erence was observed in expression of Le y and caga status. Furthermore, when the numbers of Le x or Le y -positives/-negatives were compared, no signi cant di erence was discerned between these groups regarding CagA or VacA status. All isolates possessed the vaca gene and, thus, no comparison of Le status could be made. The mean level of expression of Le x (Fig. 3) did not di er signi cantly between isolates with or without the caga gene (P = 0.48, two-tailed Student's t test), nor was any signi cant association seen between the mean level of expression of Le x and isolates expressing or not expressing the CagA protein (P = 0.89) or VacA status (P = 0.13). Similarly, the mean level of expression of Le y (Fig. 4) did not di er signi cantly between isolates with or without the caga gene (P = 0.12), nor was any signi cant association observed between the mean level of expression of Le y and CagA (P = 0.59) or VacA status (P = 0.94). Fig. 3. Mean Le x expression ( þ S.D.) of 37 H. pylori clinical isolates of speci ed caga and vaca genotypic and phenotypic status. Fig. 4. Mean Le y expression ( þ S.D.) of 37 H. pylori clinical isolates of speci ed caga and vaca genotypic and phenotypic status.

8 86 D.G. Marshall et al. / FEMS Immunology and Medical Microbiology 24 (1999) 79^90 Table 3 Relationship between disease and the caga and vaca genotypic and phenotypic status of 37 H. pylori clinical isolates Genotypic/phenotypic status of isolates Number of isolates Associated disease Ulcer (n = 11) Non-ulcer (n = 26) Two-tailed P-value caga-positive caga-negative CagA-positive CagA-negative VacA-positive VacA-negative Le expression and associated pathologies The pathologies in patients infected with H. pylori were known for all the clinical isolates and isolates were divided into two groups; those associated with PUD (gastric and duodenal ulcers) and those associated with NUD (including asymptomatic gastritis, re ux oesophagitis and non-ulcer dyspepsia). Of the 57 H. pylori isolates available, 21 were associated with PUD and 15 with NUD. A higher proportion of isolates from ulcer patients expressed VacA compared with those from non-ulcer patients (71 vs. 53%, P = 0.263). Also, isolates of the type I strain type were more frequently recovered from patients with PUD than NUD (52 vs. 31%). H. pylori isolates which were caga-positive were no more signi cantly associated with PUD than NUD in patients (71 vs. 67%, P = 0.775), nor were CagA-positive isolates (57 vs. 50%, P = 0.783). However, 12 (80%) of the isolates from duodenal ulcer patients were caga-positive. Of the 37 H. pylori isolates examined for Le expression, a higher proportion (70%) of isolates was associated with NUD than PUD as shown in Table 3. Moreover, a higher proportion of the caga-positive isolates were associated with non-ulcer than ulcer patients, but no statistically signi cant di erence was observed between the numbers of caga-positive or -negative isolates in either of these patient groups in a two-tailed Fisher's exact test. Similarly, no signi cant relationship was observed between PUD or NUD and CagA or VacA status. However, comparison of disease groups and Le x status showed that the number of isolates expressing Le x was signi cantly di erent (P ) in ulcer patients to the other groups (Table 4). In both ulcer and non-ulcer patients, Le y -positive isolates predominated over Le y -negative isolates, but no signi cant association between Le y status and disease was observed. The mean level of expression of Le x on H. pylori isolates associated with PUD was higher than that of isolates from non-ulcer patients (0.65 vs. 0.41, absorbance units at 492 nm), but was not signi cantly di erent (P = 0.19) in a Student's t-test. Sim- Table 4 Relationship between disease and Le expression of 37 H. pylori clinical isolates Lewis status of isolates Associated disease Ulcer (n = 11) Non-ulcer (n = 26) Two-tailed P-value Le x -positive 9 11 Le x -negative Le y -positive Le y -negative

9 D.G. Marshall et al. / FEMS Immunology and Medical Microbiology 24 (1999) 79^90 87 ilarly, the mean level of expression of Le y on H. pylori isolates associated with PUD or NUD was not signi cantly di erent (1.11 vs absorbance units, P = 0.13). 4. Discussion In the present study H. pylori clinical isolates were characterised rstly, on the basis of caga and vaca gene carriage and VacA and CagA expression and, secondly, a proportion of these isolates for which biomass was available were characterised for Le antigen expression. On the basis of the former genotypic and phenotypic characteristics, isolates were grouped into distinct strain types as de ned by Xiang et al. [37]. Of the 57 isolates examined in this study, type I was the largest group of strains and these isolates were more frequently isolated from patients with PUD than NUD, in agreement with previous reports [9,41^43]. Moreover, H. pylori isolates expressing VacA were more predominantly associated with PUD than NUD, which is in agreement with previous studies of other populations in Italy, Belgium, Sweden, Australia and USA [11,13,41,44^47]. However, the association of VacA with PUD did not reach statistical signi cance. The caga-positive Irish H. pylori isolates were no more frequently associated with PUD than NUD which contrasts with studies in other populations where such strains have been associated with increased risk of peptic ulceration [9,42,48,49]. On the other hand, caga-positive strains have been observed in European non-ulcer patients previously, but have more active gastritis characterised by increased polymorphonuclear cell in ltration into the gastric epithelium [48,50] and a greater likelihood of developing intestinal metaplasia [50]. However, equally high prevalences of caga-positive isolates from PUD and NUD patients have been reported also in Chinese patients [51] and this trait is considered not a useful marker of enhanced virulence in the Chinese and Japanese populations [51^53]. Of the isolates from Irish duodenal ulcer patients, 80% were caga-positive, which is comparable to those of about 80^100% that have been reported for di erent populations worldwide [42,43,54,55]. A higher proportion of Irish H. pylori isolates were of the intermediate strain types Ia to Id than was previously observed in a study with strains predominantly from Italy, but also from England, France, and USA (47 vs. 28%) [37]. Thus, these intermediate types are more common in the Irish population than in the previously screened populations. Consistent with our previous studies [17,18], all clinical isolates examined in this study produced high molecular weight LPS. Of the 57 Irish isolates, biomass was available from 37 for detection of Le antigen expression. Le x and/or Le y epitopes were detected on 89% of the H. pylori isolates which is comparable to that observed on isolates from other parts of the world [23,24]. Of potential interest, none of the Irish isolates expressed Le a,le b, sialyl-le x and H type 1 chain which have been detected on H. pylori isolates from other geographical populations [23,24,56]. Of the isolates examined, four were nontypeable with the panel of antibodies used. This lack of typeability is attributable to the presence of an, as yet, unde ned structure rather than the loss of the O side chain from H. pylori LPS during cultivation [17,18,32] since production of high rather than low molecular weight LPS was con rmed by SDS-PAGE with silver staining and immunoblotting. The results of the present study, which are consistent with those of our preliminary investigations [29], show the absence of a statistically signi cant relationship between the occurrence of Le x or Le y and caga, vaca, and CagA status of the Irish isolates. Moreover, no signi cant association between the mean level of expression of Le x or Le y and caga, vaca, and CagA status of the H. pylori isolates was observed, in contrast to the report of Wirth et al. [24]. Although small numbers of isolates from di erent countries were examined in the latter study, predominantly isolates were from North America. Whether the contrasting lack of an association between Le antigen expression and caga, CagA or VacA status in the Irish H. pylori isolates could be attributable to the adaptation of H. pylori strains with di ering attributes to di erent human populations remains an open question. Supporting such a view, ethnic tropism by H. pylori strains has been demonstrated [54]. Furthermore, using the same study population, Wirth et al. reported that Le x and Le y expression by H. pylori is related to Le expression in the host gastric epithelium [57]. Again, in

10 88 D.G. Marshall et al. / FEMS Immunology and Medical Microbiology 24 (1999) 79^90 contrast, with isolates from an Irish population, we were unable to establish this association [58], emphasising di erences in the interactions between H. pylori isolates and the di ering host populations. Of the 37 H. pylori isolates examined for Le expression, no signi cant relationship was observed between PUD or NUD and caga, CagA or VacA status. Nevertheless, comparison of disease groups and Le status showed a signi cant association between the number of isolates expressing Le x and PUD. Wirth et al. also found that H. pylori isolates from ulcer patients were more likely to express Le antigens than those from non-ulcer patients [24]. Furthermore, we found that the mean level of expression of Le x on H. pylori isolates associated with PUD was higher than that of isolates from non-ulcer patients, but this was not statistically signi cant. Assuming an association between the putative virulence factors of H. pylori and the pathologies observed in certain populations, the lack of a relationship between caga, CagA or VacA expression and PUD in this survey of Irish isolates may help explain their further lack of relationship with Le expression. It can be speculated that if the isolates with CagA and VacA do not induce aggressive pathologies in this population, then the need for an association with counterbalancing host Le antigen expression would not be required. Acknowledgments We thank M. Tuummuru, Vanderbilt University School of Medicine, Nashville, USA for providing plasmid pmc3, R. Haas, Max Planck Institut fuër Biologie, Tubingen, Germany for plasmid pws18, A. Cockayne, University of Nottingham, UK for electroeluted CagA protein, and D. Burroni, IRIS- Biocine, Sienna, Italy for anti-vaca antisera. This work was supported by grants from the Irish Health Research Board (to C.J.S and to A.P.M.). References [1] Dunn, B.E., Cohen, H. and Blaser, M.J. (1997) Helicobacter pylori. Clin. Microbiol. Rev. 10, 720^741. [2] Marshall, B.J. and Warren, J.R. (1984) Unidenti ed curved bacilli in the stomach of patients with gastritis and peptic ulceration. Lancet i, 1311^1315. [3] Graham, D.Y., Lew, G.M., Klein, P.D., Evans, D.G., Evans, D.J., Saeed, Z.A. and Malaty, H.M. (1992) E ect of treatment of Helicobacter pylori infection on the long-term recurrence of gastric or duodenal ulcer. Ann. Intern. Med. 116, 705^708. [4] NIH Consensus Development Panel on Helicobacter pylori in Peptic Ulcer Disease (1994) Helicobacter pylori in peptic ulcer disease. J. Am. Med. Assoc. 272, 65^69. [5] Forman, D., Newell, D.G., Fullerton, F., Yarnell, J.W., Stacey, A.R., Wald, N. and Sitas, F. (1991) Association between infection with Helicobacter pylori and risk of gastric cancer: evidence from a prospective investigation. Br. Med. J. 302, 1302^1305. [6] Parsonnet, J., Hansen, S., Rodriguez, L., Gelb, A.B., Warnke, R.A., Jelum, E., Orentreich, N., Vogelman, J.H. and Freidman, G.D. (1994) Helicobacter pylori infection and gastric cancer lymphoma. New Engl. J. Med. 330, 1267^1271. [7] International Agency for Research on Cancer (1994) Schistosomes, liver ukes and Helicobacter pylori. IARC Monogr. Carcinog. Risks Human 61, 218^220. [8] Crabtree, J.E., Taylor, J.D., Wyatt, J.I., Heatley, R.V., Shallcross, T.M., Tompkins, D.S. and Rathbone, B.J. (1991) Mucosal IgA recognition of Helicobacter pylori 120 kda protein, peptic ulceration, and gastric pathology. Lancet 338, 332^ 335. [9] Covacci, A., Censini, S., Bugnoli, M., Petracca, R., Burroni, D., Macchia, G., Massone, A., Papini, E., Xiang, Z., Figura, N. and Rappuoli, R. (1993) Molecular characterization of the 128-kDa immunodominant antigen of Helicobacter pylori with cytotoxicity and duodenal ulcer. Proc. Natl. Acad. Sci. USA 90, 5791^5795. [10] Tummuru, M.K.R., Cover, T.L. and Baser, M.J. (1993) Cloning and expression of a high-molecular mass major antigen of Helicobacter pylori: evidence of linkage to cytotoxin production. Infect. Immun. 61, 1799^1809. [11] Blaser, M.J. (1994) Helicobacter pylori phenotypes associated with peptic ulceration. Scand. J. Gastroenterol. 29 (Suppl. 205), 1^5. [12] Atherton, J.C., Cao, P., Peek, R.M., Tummuru, M.K.R., Blaser, M.J. and Cover, T.L. (1995) Mosaicism in vacuolating cytotoxin alleles of Helicobacter pylori: association of speci c vaca types with cytotoxin production and peptic ulceration. J. Biol. Chem. 270, 17771^ [13] Atherton, J.C. (1997) vaca genotypes and the virulence of Helicobacter pylori. In: Pathogenesis and Host Response in Helicobacter pylori Infections (Moran, A.P. and O'Morain, C.A., Eds.), pp. 76^87. Normed Verlag, Bad Homburg, Germany. [14] Censini, S., Lange, C., Xiang, Z., Crabtree, J.E., Ghiara, P., Borodovsky, M., Rappuoli, R. and Covacci, A. (1996) cag, a pathogenicity island of Helicobacter pylori, encodes type I-speci c and disease-associated virulence factors. Proc. Natl. Acad. Sci. USA 93, 14648^ [15] Covacci, A., Falkow, S., Berg, D.E. and Rappuoli, R. (1997) Did the inheritance of a pathogenicity island modify the virulence of Helicobacter pylori? Trends Microbiol. 5, 205^208.

11 D.G. Marshall et al. / FEMS Immunology and Medical Microbiology 24 (1999) 79^90 89 [16] Covacci, A., Falkow, S., Censini, S., Lange, C., Segal, E., Salama, N., Tompkins, L.S., Guidotti, S., Gori, M., Marchetti, M. and Rappuoli, R. (1997) The cag pathogenicity island of Helicobacter pylori: origin, molecular biology and relevance to virulence. In: Pathogenesis and Host Response in Helicobacter pylori Infections (Moran, A.P. and O'Morain, C.A., Eds.), pp. 88^100. Normed Verlag, Bad Homburg, Germany. [17] Moran, A.P. (1995) Cell surface characteristics of Helicobacter pylori. FEMS Immunol. Med. Microbiol. 10, 271^280. [18] Moran, A.P., Helander, I.M., Kosunen, T.U. (1992) Compositional analysis of Helicobacter pylori rough-form lipopolysaccharides. J. Bacteriol. 174, 1370^1377. [19] Aspinall, G.O., Monteiro, M.A., Pang, H., Walsh, E.J. and Moran, A.P. (1994) O antigen chains in the lipopolysaccharide of Helicobacter pylori NCTC Carbohydr. Lett. 1, 151^ 156. [20] Aspinall, G.O., Monteiro, M.A., Pang, H., Walsh, E.J. and Moran, A.P. (1996) Lipopolysaccharide of the Helicobacter pylori type strain NCTC (ATCC 43504): structure of the O antigen chain and core oligosaccharide regions. Biochemistry 35, 2489^2497. [21] Aspinall, G.O. and Monteiro, M.A. (1996) Lipopolysaccharides of Helicobacter pylori strains P466 and MO19: structures of the O antigen and core oligosaccharide regions. Biochemistry 35, 2498^2504. [22] Sherburne, R. and Taylor, D.E. (1995) Helicobacter pylori expresses a complex surface carbohydrate, Lewis X. Infect. Immun. 63, 4564^4568. [23] Simoons-Smit, I.M., Appelmelk, B.J., Verboom, T., Negrini, R., Penner, J.L., Aspinall, G.O., Moran, A.P., Fei Fei, S., Bi- Shan, S., Rudnica, W., Savio, A. and De Graa, J. (1996) Typing of Helicobacter pylori with monoclonal antibodies against Lewis antigens in lipopolysaccharide. J. Clin. Microbiol. 34, 2196^2200. [24] Wirth, H.P., Yang, M.Q., Karita, M. and Blaser, M.J. (1996) Expression of the human cell-surface glycoconjugates Lewis X and Lewis Y by Helicobacter pylori isolates is related to caga status. Infect. Immun. 64, 4598^4605. [25] Moran, A.P. (1996) Pathogenic properties of Helicobacter pylori Scand. J. Gastroenterol. 31 (Suppl. 215), 22^31. [26] Appelmelk, B.J., Simmons-Smit, I., Negrini, R., Moran, A.P., Aspinall, G.O., Forte, J.G., De Vries, T., Quan, H., Verboom, T., Maaskant, J.J., Ghiara, P., Kuipers, E.J., Bloemena, E., Tadema, T.M., Townsend, R.R., Tyagarajan, K., Crothers, J.M., Monteiro, M.A., Savio, A. and De Gra, J. (1996) Potential role of molecular mimicry between Helicobacter pylori lipopolysaccharide and host Lewis blood group antigens in autoimmunity. Infect. Immun. 64, 2031^2040. [27] Moran, A.P., Appelmelk, B.J. and Aspinall, G.O. (1996) Molecular mimicry of host structures by lipopolysaccharides of Campylobacter and Helicobacter spp.: implications in pathogenesis. J. Endotoxin Res. 3, 521^531. [28] Blaser, M.J. (1997) Heterogeneity of Helicobacter pylori. Eur. J. Gastroenterol. Hepatol. 9 (Suppl. 1), S3^S7. [29] Marshall, D.G., Hynes, S.O., Coleman, D.C., O'Morain, C.A., Smyth, C.J. and Moran, A.P. (1998) Lack of relationship between caga status and expression of Lewis antigens in Irish Helicobacter pylori isolates. Gut 43 (Suppl. 2), A18^A19. [30] Sipponen, P., Kekki, M. and Siurala, M. (1991) The Sydney system: epidemiology and natural history of chronic gastritis. J. Gastroenterol. Hepatol. 6, 244^251. [31] Heneghan, M.A., Moran, A.P., Feeley, K.M., Egan, E.L., Goulding, J., Connolly, C.E. and McCarthy, C.F. (1998) Effect of host Lewis and ABO blood group antigen expression on Helicobacter pylori colonisation density and the consequent in ammatory response. FEMS Immunol. Med. Microbiol. 20, 257^266. [32] Walsh, E.J. and Moran, A.P. (1997) In uence of medium composition on the growth and antigen expression of Helicobacter pylori. J. Appl. Microbiol. 83, 67^75. [33] Ausubel, F.M., Brent, R., Kingston, R.E., Moore, D.D., Seidman, J.G., Smith, J.A. and Struhl, K. (1987) Current Protocols in Molecular Biology. John Wiley and Sons, Chichester, UK. [34] Schmitt, W. and Haas, R. (1994) Genetic analysis of the Helicobacter pylori vacuolating cytotoxin: structural similarities with the IgA protease type of exported protein. Mol. Microbiol. 12, 307^319. [35] Hitchcock, P.J. and Brown, T.M. (1983) Morphological heterogeneity among Salmonella lipopolysaccharide chemotypes in silver-stained polyacrylamide gels. J. Bacteriol. 154, 269^ 277. [36] Telford, J.L., Ghiara, P., Dell'Orco, M., Comanducci, M., Burroni, D., Bugnoli, M., Tecce, M.F., Censini, S., Covacci, A., Xiang, Z., Papini, E., Montecucco, C., Parente, L. and Rappuoli, R. (1994) Gene structure of the Helicobacter pylori cytotoxin and its key role in gastric disease. J. Exp. Med. 179, 1653^1658. [37] Xiang, Z., Censini, S., Bayeli, P.F., Telford, J.L., Figura, N., Rappouli, R. and Covacci, A. (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. [38] Laemmli, U.K. (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227, 680^685. [39] Towbin, H., Staehelin, T. and Gordon, J. (1979) Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedure and some applications. Proc. Natl. Acad. Sci. USA 76, 4350^4354. [40] Tsai, C.-M. and Frasch, C.E. (1982) A sensitive silver stain for detecting lipopolysaccharides in polyacrylamide gels. Anal. Biochem. 119, 115^119. [41] Figura, N., Guglielmetti, P., Rossolini, A., Barberi, A., Cusi, G., Musmanno, R.A., Russi, M. and Quaranta, S. (1989) Cytotoxin production by Campylobacter pylori strains isolated from patients with peptic ulcers and from patients with chronic gastritis only. J. Clin. Microbiol. 27, 225^226. [42] Weel, J.F., van der Hulst, R.W.M., Gerrits, Y., Roorda, P., Feller, M., Dankert, J., Tytgat, G.N. and van der Ende, A. (1996) The interrelationship between cytotoxin-associated

12 90 D.G. Marshall et al. / FEMS Immunology and Medical Microbiology 24 (1999) 79^90 gene A, vacuolating cytotoxin, and Helicobacter pylori-related diseases. J. Infect. Dis. 173, 1771^1775. [43] Takata, T., Fujimoto, S., Anzai, K., Shirotani, T., Okada, M., Sawae, Y. and Ono, J. (1998) Analysis of the expression of CagA and VacA and the vacuolating activity in 167 isolates from patients with either peptic ulcers or non-ulcer dyspepsia. Am. J. Gastroenterol. 93, 30^34. [44] Goossens, H., Glupczynski, Y., Burette, A., Lambert, J.P., Vlaes, L., Butzler, J.P. (1992) Role of the vacuolating toxin from Helicobacter pylori in the pathogenesis of duodenal and gastric ulcer. Med. Microbiol. Lett. 1, 153^159. [45] Rautelin, H., Blomberg, B., Fredlund, H., Jaërnerot, G. and Danielsson, D. (1993) Incidence of Helicobacter pylori strains activating neutrophils in patients with peptic ulcer disease. Gut 34, 599^604. [46] Cover, T.L., Cao, P., Lind, C.D., Tham, K.T. and Blaser, M.J. (1993) Correlation between vacuolating cytotoxin production by Helicobacter pylori in vitro and in vivo. Infect Immun. 61, 5008^5012. [47] Tee, W., Lambert, J.R., Dwyer, B. (1995) Cytotoxin production by Helicobacter pylori from patients with upper gastrointestinal tract diseases. J. Clin. Microbiol. 33, 1203^1205. [48] Crabtree, J.E., Taylor, J.E., Wyatt, J.I., Heatley, R.V., Shallcross, T.M., Tompkins, D.S. and Rathbone, B.J. (1991) Mucosal recognition of Helicobacter pylori 120 kda protein, peptic ulceration and gastric pathology. Lancet 338, 332^335. [49] Xiang, Z., Bugnoli, M., Rappuoli, R., Covacci, A., Ponzetto, A. and Crabtree, J.E. (1993) Helicobacter pylori: host responses in peptic ulceration. Lancet 341, 900. [50] Crabtree, J.E., Wyatt, J.I., Perry, S., Davies, G.R., Covacci, A., Morgan, A.G. (1996) CagA seropositive Helicobacter pylori infected non-ulcer patients have increased frequency of intestinal metaplasia. Gastroenterology 110 (Suppl.), A85. [51] Pan, Z.-J., van der Hulst, R.W.M., Feller, M., Xiao, S.-D., Tytgat, G.N., Dankert, J. and van der Ende, A. (1997) Equally high prevalences of infection with caga-positive Helicobacter pylori in Chinese patients with peptic ulcer disease and those with chronic gastritis-associated dyspepsia. J. Clin. Microbiol. 35, 1344^1347. [52] Shimoyama, T., Fukuda, S., Tanaka, M., Mikami, T., Saito, Y. and Munakata, A. (1997) High prevalence of the cagapositive Helicobacter pylori strains in Japanese asymptomatic patients and gastric cancer patients. Scand. J. Gastroenterol. 32, 465^468. [53] Moran, A.P. and Wadstroëm, T. (1998) Pathogenesis of Helicobacter pylori. Curr. Opin. Gastroenterol. 14 (Suppl. 1), S9^ S14. [54] Campbell, S., Fraser, A., Holliss, B., Schmid, J. and O'Toole, P.W. (1997) Evidence for ethnic tropism of Helicobacter pylori. Infect. Immun. 65, 3708^3712. [55] Warburton, V.J., Everett, S., Mapstone, M.P., Axon, A.T.R., Hawkey, P., Dixon, M.F. (1998) Clinical and histological association of caga and vaca genotypes in Helicobacter pylori gastritis. J. Clin. Pathol. 51, 55^61. [56] Monteiro, M.A., Chan, K.H.N., Rasko, D.A., Taylor, D.E., Zheng, P.Y., Appelmelk, B.J., Wirth, H.P., Yang, M., Blaser, M.J., Hynes, S.O., Moran, A.P. and Perry, M.B. (1998) Simultaneous expression of type 1 and type 2 Lewis blood group antigens by Helicobacter pylori lipopolysaccharides. Molecular mimicry between H. pylori lipopolysaccharides and human gastric epithelial cell surface glycoforms. J. Biol. Chem. 19, 11533^ [57] Wirth, H.P., Yang, M., Peek, R.M., Tham, K.T. and Blaser, M.J. (1997) Helicobacter pylori Lewis expression is related to the host Lewis phenotype. Gastroenterology 113, 1091^1098. [58] Heneghan, M.A., McCarthy, C.F. and Moran, A.P. (1998) Is colonization of gastric mucosa by Helicobacter pylori strains expressing Lewis X and Y epitopes dependent on host secretor status? In: Campylobacter, Helicobacter and Related Organisms (Lastovica, A.J., Newell, D.G. and Lastovica, E.E., Ed.), pp. 473^477. Institute of Child Health and University of Cape Town, Cape Town, South Africa.