Introduction

Helicobacter pylori is a Gram-negative, micro-aerophilic, curved bacterium, characterized by its capacity to establish close contact with the gastric mucosal surface cells, exhibit prominent tropism for this type of cells, and persistently colonize the human stomach. In a few patients it is an established cause of chronic active gastritis and peptic ulcer disease and a major risk factor for precursor lesions and the development of gastric cancer [1, 2].

Choreographed interactions exist among the micro-organism, the host, and the environment, in particular if considered with respect to social economic status, sanitary conditions, food, dietary habits, and lifestyle [3, 4]. All these interactions, and the changing balance among them, are the probable cause of the different typology and the low and in some cases geographically variable [5] incidence of the pathological events caused by H. pylori. They may also explain some of the conflicting bio-pathological peculiarities of the micro-organism, such as “the African enigma” [6], or the protective effects that it seems to produce, in children, against diarrheal diseases [7] and, in adults, in western countries, against gastro-esophageal reflux disease, Barrett’s esophagus, and adenocarcinoma of the esophagus [8, 9].

Eradication of the organism has been associated with the histological improvement of gastritis, a lower relapse rate, less risk of bleeding from a duodenal ulcer [10], and reduced development of gastric cancer, particularly in individuals without precancerous lesions belonging to high-risk populations [11]. Nevertheless, the real advantage of the microbial eradication still needs to be confirmed, and a better assessment of the micro-organism’s pathogenicity and of its interactions with the host and the environment is still required before global strategies can be launched to fight the micro-organism and its pathological effects.

It is now only possible to distinguish virulent H. pylori strains (i.e., those associated with duodenal ulceration or gastric carcinoma) from less virulent strains, and two candidate markers are generally considered useful for this purpose: polymorphism in vacuolating cytotoxin gene A (vacA) alleles and the presence or absence of the cytotoxin-associated gene A (cagA) [12, 13]. vacA is present in all strains and comprises two well-characterized variable parts: the s region, which exists as an s1 or s2 allele, and the m region, which occurs as an m1 or m2 allele. Strains carrying the s1m1 mosaic combination exhibit the highest levels of cytotoxic activity, while the s2m2 strains do not secrete the vacuolating cytotoxin [12]. A third recently identified vacA polymorphic site, the intermediate (I) region, has also been recognized to be present in the H. pylori strains as an i1 or i2 allele [14]; s1m1 and s2m2 strains are exclusively i1 and i2 respectively, whereas s1m2 strains are variable in their i-type; i1 confers toxicity and is defined as an independent marker of the VacA-associated pathogenicity. cag A encodes the high molecular weight, immunodominant antigen CagA, which is considered to be a marker for the presence of the H. pylori cag pathogenicity island (PAI); strains with genotype vacA s1/m1 are often CagA-positive, and strains cagA+/vacA s1m1 are often associated with extensive damage in gastric mucosa [15].

Many other H. pylori virulence factors have been described and considered to be involved in determining the severity or the different typologies and evolutions of the bacterial pathological manifestations [1]. For some of them, in particular the adhesins BabA, SabA, and the outer membrane proteins (OMPs) OipA and HopQ, but not the OMP HopZ, evidence has also been obtained of an association of the presence, the functional state, or the virulence-related alleles of their encoding genes with the presence of the cag PAI and, very often, of the vacA s1 allele [16–21]. If confirmed, these associations, in a bacterium such as H. pylori, characterized by a recombinational population structure [22], might be considered an example of linkage disequilibrium and, presumably, might indicate a selective advantage in strains with certain virulence genotypes [21]. With the lack of reliable information regarding the pathogenic role of the different virulence factors, their association could provide the potentially useful tool required for predicting the clinical outcome of infections.

At present, however, there is no unanimous evidence that all the described associations represent a general and clinically significant phenomenon [23–26], nor it is known if the associations, or at least some of them, are conditioned by the well-known differences in the geographical and ethnic prevalence of the H. pylori virulence-associated genotypes [27–29].

As it has been suggested that an analysis should be made of H. pylori strains from multiple populations of patients and from various parts of the world [21], an investigation was undertaken in our laboratory to determine the prevalence of specific genotypes of the micro-organism, and their association with the severity of current gastric pathology, in patients from western Sicily, Italy. The investigation was also considered interesting because Sicily, the largest and most populous island in the Mediterranean sea, is strategically located midway between Africa and Italy, and its inhabitants, for historical and present-day reasons, constitute one of the most complex mixtures of different ethnic elements in Europe.

Materials and methods

Patients and samples

This study involved 90 patients, with a mean age of 51.7 years (median, 53.5 years; range 27–76 years), who underwent upper gastrointestinal endoscopy for a variety of indications at the Endoscopy Services of the Ospedali Civili Riuniti in Sciacca (Agrigento) and of the Gastroenterology, Internal Medicine and Elderly Care, and Emergency Surgery Units of the University Hospital A.O.U.P. Paolo Giaccone in Palermo, Italy. Informed consent was obtained from all patients. During gastroscopy, biopsy samples were taken from the antrum and the body of the stomach for rapid urease testing, histology, DNA extraction, and, in some cases, culture, when the proper conditions for sample transport were realizable. Biopsy samples for DNA extraction were sent to the microbiology laboratory in dry tubes; biopsy samples for H. pylori culture were placed in 0.5 ml of Wilkins broth (Difco Laboratories, Detroit, MI, USA) containing 15% glycerol and transported in an ice bath to the microbiology laboratory where they were inoculated immediately onto culture medium or stored at −70°C. All patients were positive for H. pylori infection on rapid urease testing. Diagnoses, made on the basis of endoscopic and histological findings, were inactive chronic gastritis (CG), chronic gastritis, active (CGA), and chronic gastritis, active, with gastric or duodenal peptic ulcer (CGA+PU). Chronic inflammation was defined on the basis of the increase in lymphocytes and plasma cells in the lamina propria and activity when a neutrophilic infiltration of the lamina propria, pits or surface epithelium was found.

DNA isolation from gastric biopsies and H. pylori DNA detection by PCR methodology

In the microbiology laboratory, biopsy samples were immediately dissected and immersed in Eppendorf microtubes containing 300 μl of sterile extraction buffer (100 mM NaCl, 10 mM Tris-HCl pH 8, 25 mM EDTA, 0.5% SDS), and 5 μl of proteinase K(10 mg/ml), vortexed and incubated with moderate shaking for 18 h at 45°C. The DNA was extracted twice with an equal volume of a solution of phenol/chloroform/isoamylic alcohol (25:24:1) and precipitated at −20°C overnight by adding one-tenth volume of 3M sodium acetate (pH 5) and 2.5 volume of absolute cold ethanol. The extracted DNA was washed with 500 μl of 70% ethanol, air dried, and dissolved in 200 μl of sterile distilled water. The DNA (concentration) quality and quantity after each extraction was evaluated (spectrophotometrically) by horizontal electrophoresis in 1% agarose gel; the approximate quantity of DNA (for the purpose of PCR) was calculated according to a molecular weight standard (GeneRuler DNA ladder mix; MBI Fermentas, York, UK) always loaded on the gel together with the samples.

The presence of the conserved urease ureA gene, a marker of the presence of the H. pylori genome, was detected by a nested polymerase chain reaction in which, following Wang et al. [30] and Mravak-Stipetić et al. [31], the first two primer pairs shown in Table 1 were used.

Table 1 Oligonucleotide primers used in this studya
Full size table

In the first-round reaction, PCR was carried out in a 100-μl mixture containing 2–4 μl of each DNA solution, 10 mM Tris-HCl (pH 8.3), 50 mM KCl, 2.5 mM MgCl2, 200 μM dNTPs, 200 nM of each primer, and 2.5 U of Ampli Taq Gold polymerase (Applied Biosystems). The amplification was performed in a Perkin-Elmer ThermoCycler 2400 under the following conditions: at 95°C for 10 min, followed by 40 cycles at 94°C for 45 s, 56°C for 45 s, and 72°C for 45 s, and a final extension at 72°C for 5 min. In the second-round reaction, the same conditions were employed, except that 1 μl of the first PCR product was used as target DNA and that annealing was carried out at 55°C. Twenty microliters of each PCR product were analyzed by electrophoresis through a 2% (wt/vol) agarose gel containing 1 μg of ethidium bromide per millimeter, and the bands were visualized by excitation under UV light.

H. pylori culture

Biopsy samples in transport medium were vortexed vigorously for 2 min, and 100 µl of the media were placed on Columbia agar (Oxoid, Basingstoke, Hampshire, UK) with the addition of 7% horse blood and 0.4% of selective supplement. The plates were incubated at 37°C under microaerobic conditions (CampyGen; Oxoid) for 3–6 days. H. pylori bacteria were identified on the basis of characteristic colony morphology, appearance on Gram staining, rapid urea hydrolysis, positive catalase, oxidase production, and the API Campy Kit (bioMérieux). H. pylori cells that grew out of one biopsy on the primary culture plate were subcultured, collected as a pooled population after 2–3 days of incubation, and preserved in sterile brain heart infusion (BHI) broth (Difco) with 15% glycerol at −70°C. All such cultures were analyzed, but, when multiple vacA or hopQ alleles were observed, five isolated clones were also obtained from each culture and all the assays were repeated on these clones.

Extraction of target bacterial genomic DNA

Suspensions of each bacterial strain were prepared in sterile distilled water (200 μl) by inoculation from single agar plate cultures with a standard loop. Samples were washed twice in sterile distilled water, boiled for 10 min, and centrifuged for 5 min at 14,000 × g. Supernatants were aliquoted and stored at + 4°C. At the testing time, the DNA concentration was evaluated spectrophotometrically by measuring optical density at 260 nm, and aliquots of 50 ng of the genomic DNA were used for PCRs.

Detection of cagA and babA2 and typing of vacA and hopQ homologs

A PCR-based methodology was applied to detect the presence, in some H. pylori strains, of the cagA and bab2 genes and to identify, in all the strains, the different families of the vacA and hopQ alleles. According to Tummuru et al. [32], Atherton et al. [12], Gerhard et al. [16], Zambon et al. [20], and Cao and Cover [21], the sets of primers shown in Table 1 were employed to specifically amplify the desired genomic fragments. The annealing temperatures during the amplification cycles were: 53°C, for cagA; 50°C, for vacAs1 and vacAs2; 53°C, for vacAm1 and vacAm2; 53°C, for bab2; 40°C, for type I hopQ, and type II hopQ. With respect to all other conditions, PCR amplifications were carried out with the same procedure employed in the first-round reaction for ureA detection. The expected bp length of the PCR products after electrophoresis on a 2% agarose gel is shown in Table 1.

Evaluation of the functional status of oipA, sabA, and hopZ

The oipA, sabA, and hopZ genes are characterized by calcitonin (CT) dinucleotide repeats in their 5’-coding regions, whose length is thought to determine whether or not the complete open reading frames are in frame, and full-length proteins are translated [19, 33–34]. A PCR sequencing-based methodology was applied to detect the number of the CT repeats in all these genes and, therefore, to attempt assessment of their expression status. According to de Jonge et al. [19] the sets of primers shown in Table 1 were employed to amplify the regions of the oipA, hopZ, and at least some of the sabA genes, containing the CT repeats. The PCR amplifications were performed using the same method described above, at the annealing temperatures of 55°C for oipA and hopZ, and 50°C for sabA. The resulting PCR products were subjected to gel electrophoresis in a 2% agarose gel and subsequently purified for sequencing using a GFX™ PCR DNA and a Gel Band Purification Kit (Amersham Biosciences). Sequences of purified amplicons were directly determined, with the same forward and reverse primers employed as in the previous PCR reactions, using the ABI PRISM BigDye Cycle Sequencing Ready Reaction kit (Applied Biosystems), according to the instructions supplied by the manufacturer, and a model 310 ABI DNA sequencer (Applied Biosystems), for the analysis of products.

Statistical methods

The Chi-squared (χ2) test with Yates’ continuity correction or the two-tailed Fisher’s exact test was used for statistical analysis. A p value of less than 0.05 was considered statistically significant.

Results

cagA, vacA, and babA2 genotypes in H. pylori-positive gastric biopsies

As revealed by the positivity on PCR urease amplification, H. pylori DNA was demonstrable in all 90 of the patients investigated: in both gastric body and antrum in 76 of them; in the body only in 11; in the antrum only in 3. All patients were affected by chronic gastritis, inactive (CG) in 37, active (CGA) in 26, and active and associated with a gastric or duodenal peptic ulcer (CGA+PU) in 6 and 21 respectively.

The H. pylori cagA, vacA, and babA2 genotypes of H. pylori were defined in all the 166 positive samples, with no evidence of multiple infections in patients in whom more than one biopsy was positive, except for one belonging to the CGA+PU group, who showed the presence of two different genotypes in his stomach: a cagA +/vacAs1m2/babA2 + genotype at the body level and a cagA -/ vacAs2m2/babA2 - genotype in the antrum.

Considering all 91 different positive biopsies, the gene cagA, the alleles s1 and m1 of vacA, and the gene babA2 were present in the 53.8%, 60.4%, 35.2%, and 47.3% of the samples respectively (Table 2). The genotype cagA + appeared to be significantly associated (P < 0.005) with the s1 and m1 alleles of vacA alleles and with the babA2 gene; analogously, and always at high levels of significance (P < 0.005), the s1 and m1 alleles of vacA were associated with each other and with the genotype babA2 + (data not shown). Mixtures of the s1-s2 or m1-m2 alleles of vacA were seen in 4.4% and 7.7% of samples respectively, and in one sample, which was positive for a genotype cagA +/vacAs1/ babA2 +, the m region of vacA was untyped.

Table 2 Virulence-associated H. pylori genotypes in patients with different gastrointestinal diseases (all cases)
Full size table

Distribution and combined presence of virulence-associated cagA, vacA, and babA2 genotypes in different gastric diseases

With respect to their distribution in the three groups of patients with different gastrointestinal diseases, the cagA +, vacAs1, vacAm1, and babA2 + genotypes of H. pylori, as defined in all the 91 PCR-positive samples (Table 2), were more frequently associated with a high level of significance, with chronic gastritis in an active (CGA) rather than in an inactive (CG) form. The frequencies of cagA + and vacAs1 genotypes did not vary significantly in patients with an active form of chronic gastritis whether a gastric or duodenal ulcer was also present (CGA+PU) or not (CGA); on the contrary, vacAm1 and babA2 + genotypes were less frequently observed, with a low level of significance, in the CGA+PU than in the CGA group. The vacAm1 genotype was always more frequent in the CGA than in the CG group, but, unlike cagA +, vacAs1, and babA2 +, only with a low level of significance.

When considered in their associated combinations (Table 3), the vacAm1 allele and/or the babA2 gene, if present in addition to the cagA gene and the vacAs1 allele, did not allow a better differentiation of the CGA, and in particular of the CGA+PU group, from the group of the inactive forms of chronic gastritis (CG); therefore, they did not appear to increase the value of the infecting H. pylori genotype as a predictor of the worsening of the gastritis. Although present in a few samples, only genotypes including cagA + and vacAs1m2 combinations were more frequently observed in patients with a gastric or duodenal ulcer (CGA+PU group) than in patients of both the CG and CGA groups.

Table 3 Prevalence of combined H. pylori cagA, vacA, and babA genotypes in the different disease groups
Full size table

Genotypes of H. pylori strains isolated from patients with different gastric diseases

H. pylori isolation was carried out on the biopsy samples from 52 out of the 90 patients investigated, the only samples that were sent to the microbiology laboratory in a proper form for culture. Antrum and body biopsy samples from 32 patients were inoculated immediately and H. pylori growth was observed in 24 of the cultures; samples from 20 patients were inoculated after a freezing time of not more than 7 days and H. pylori growth was observed in 6 of the cultures. Isolates were obtained from both gastric antrum and body in 19 patients; from the body only in 5; from the antrum only in 6. Also, on genotyping these isolates, no evidence of multiple infections was observed, except for the same patient in the CGA+PU group, in whom the presence of two different genotypes had been already observed by analyses on biopsy DNA extracts. Concerning the 29 patients from whom one or only identical isolates were obtained, 12, 9, and 8 of them belonged to the CG, CGA, and CGA+PU groups respectively.

Genotyping of the 31 different H. pylori strains considered not only the cagA, vacA, and babA2 genes, but also the allelic forms of hopQ and the functional status of oipA, hopZ, and sabA, which were not efficiently detectable in the biopsy DNA.

In all cases, even when mixtures of vacAs1-s2 or vacAm1-m2 alleles had been found, and their presence validated by the assays on the isolated clones, results concerning cagA, vacA, and babA2 genes confirmed those previously obtained by the DNA analyses from biopsies, and analogously, in spite of the lower numbers, a confirmation was obtained (Table 4) of the significantly higher presence of the cagA +, vacAs1, vacAm1, and babA2 + genotypes in patients of the CGA and CGA+PU groups than in the CG group of patients.

Table 4 Virulence-associated H. pylori genotypes in patients with different gastrointestinal diseases (cases of bacterial isolation)
Full size table

The presence of hopQ in its homologous form “type I” is considered to be a virulence-associated characteristic of H. pylori [21] and, indeed, a significantly higher presence of type I hopQ, either alone or more frequently, in association with type II, also characterized the H. pylori strains isolated from our worst cases of gastritis (groups CGA and CGA+PU; Table 4).

All cagA +, vacAs1, vacAm1, babA2 +, and type I hopQ genotypes were significantly correlated with each other (in all cases with P ≤ 0.01).

The functional status of oipA, hopZ, and sabA in H. pylori isolates does not correlate with the severity of the gastric disease nor, in most cases, with the cagA, vacA, babA2, and hopQ genotypes

The oipA, hopZ, and sabA genes are described as other virulence-associated characteristics when they have “on” status, i.e., when they are expressing their whole coding potentialities. Unlike the cagA, vacA, babA2, and hopQ genotypes, the “on” status of all these three genes, evaluated on the basis of the sequencing detection of the CT repeats in the PCR-amplified DNAs, was observed in our H. pylori strains with no significant difference with respect to their clinical provenience (Table 4). In addition, no significant relationship could be demonstrated regarding the functional status of the three oipA, hopZ, and sabA genes, nor between the cagA, vacA, babA2, and hopQ genotypes and the functional status of hopZ and sabA; with regard to oipA, a correlation was only observed with the babA2 (P = 0.007) and hopQ (P = 0.022) genotypes.

Discussion

Considerable geographical variations have been described with respect to the distribution, among the Helicobacter pylori strains, of many virulence-associated genes, as well as to the effective role of the virulence-associated H. pylori genotypes in the outcome, in the patient populations, of gastric histopathology and diseases such as chronic gastritis, peptic ulceration, gastric adenocarcinoma, and gastric mucosa-associated lymphoid tissue (MALT) lymphoma. In particular, for instance, considering comparable groups of patients with or without different gastric pathologies, the cagA + status was observed, in the isolated strains, in percentages of more than 90% in China [35], Hong Kong [36], and Colombia [37], of 73–75% in Turkey [38] and Brazil [39], of 62.9% in Padua, Italy [20], of 50–60% in France [40] and Portugal [37], and of 48 and 22%, in Kuwait, from patients of Middle Eastern or African origin respectively [41]. The vacAs1 allele was observed, in a study by van Doorn et al. [5] that considered strains from Asia, Central and South America, Europe, North America, Australia, and North Africa (Egypt), in percentages of 95, 71, 67, 62, 57, and 36% respectively. The vacAm1 allele is usually observed in percentages lower than those of the s1 allele [20, 35, 38, 40], except in the Iberian peninsula and in Central and South America [5]. The babA2 genotype was present in all the isolates in Taiwan [42], in 56% and 86% of gastric biopsy specimens from routine gastroscopy or from gastric cancer patients respectively, in Korea [43], in 69% of the strains in Brazil [39], in 54% in France [40] and Turkey [38], and in 34% in Padua, Italy [20]. The presence of mixed combinations of vacA s1-s2 or m1-m2 alleles is often observed and also in this case with considerable geographical variations (i.e., 10% of the isolates in Kuwait [41], 37% in Portugal [37], and 45 and 52% in two different areas of China [35]); as each H. pylori strain contains only one vacA s and vacA m region [44] both the mixed s1-s2 and m1-m2 combinations must be considered to be a strong indication of the presence of multiple strains. Multiple infections are more frequent in areas of high H. pylori prevalence and may reflect the adaptation of multiple bacterial genotypes to different, non-overlapping microniches in the same stomach [45].

The cagA +, vacAs1, vacAm1, and babA2 genotypes are usually significantly associated with each other and are involved in the establishment of different types of histopathological lesions and gastric pathologies. However, as indicated at the beginning of this discussion, in the evaluation of the clinical involvement of the H. pylori virulence-associated genotypes, geographical differences must also be considered. This mainly concerns areas where the putative virulence factors are predominant and where their role in the clinical outcome cannot be demonstrated and cannot be helpful in predicting the patient’s disease status [46]; in addition, if western and East Asian groups of H. pylori are differentiated on the basis of the vacA and cagA sequencing, only the western group has been reported to be associated with peptic ulcer disease [47].

Our results, obtained by analysis of all the specimens collected, showed that cagA +, vacAs1, vacAm1, and babA2 genotypes were present in percentages of 53.8, 60.4, 35.2, and 47.3% respectively. Two different H. pylori strains were isolated by only one patient and mixed infections, revealed by a combined presence, in a single strain, of vacA s1-s2 or vacA m1-m2 alleles, were observed in 12.1% of the patients. On the basis of these results, and with respect to the cagA, vacA, and babA genotypes, the H. pylori strains circulating in western Sicily, southern Italy, although exhibiting, as expected, the general characteristics of the strains present in the European and Mediterranean areas, appeared to be more similar to those circulating in northern Italy [20] and France [40] than in Spain [5], owing to the lower prevalence in Sicily of the vacAm1 allele, or in Portugal [37], owing to the lower number of mixed infections, or in Turkey [38] or North Africa [5, 41], owing to the generally lower or higher prevalence respectively, again in Sicily, of the virulence-associated genotypes.

According to the more frequent observations in western countries, and again considering all the specimens we collected, vacAs1 and vacAm1 alleles and the babA2 genotype were significantly associated with the cagA + status, and all cagA +, vacAs1, vacAm1, and babA2 + genotypes were significantly more prevalent in specimens collected from patients with chronic gastritis in an active form or with peptic ulcer disease than in those from patients with an inactive form of gastritis. On the contrary, no combination of well-characterized virulence-associated genotypes allowed better differentiations among the different groups of pathologies, nor did a single virulence-associated genotype characterize, with a higher prevalence, the pathological condition of the peptic ulcer disease. Although in contrast with some data in the literature [20, 21, 48] these last observations would appear to support the conclusions of Backert et al. [26] who say, at least with respect to the cagPAI functions, that they are important for the pathogenicity of H. pylori, but not sufficiently important to explain the development of the different clinical outcomes, the determination of which is highly complex and involves multiple bacterial and/or host factors. A new determinant of the H. pylori vacuolating cytotoxin, the type 1 of the vacA i region, has been recently described [14] and assessed to be associated with the risk of peptic ulcer [49]. Its possible presence in the vacAs1m2 strains, if confirmed, could give pathogenic significance to the higher number of cagA + vacAs1m2 genotypes observed in our patients with peptic ulcer than in those without. Anyway, even if the cagA + vacAs1m2 genotype was considered, no differences were observed between biopsy-extracted or microbial H. pylori DNA obtained from patients with gastric or duodenal peptic ulcers, and peptic ulcers were considered as a whole.

The other hopQ, oipA, hopZ, and sabA virulence-associated genotypes were assayed only on the 31 H. pylori strains isolated from 30 of the 90 patients.

With respect to the OMP 27 encoding hopQ locus, 29.0% of our strains exhibited only type I allele and 38.7% both type I and type II alleles. The latter percentage was somewhat higher than the 23.33% originally observed by Cao and Cover [21], but in any case, as they reported, the presence of type I allele, either alone or in a mixed combination with type II allele, was significantly associated with the cagA + status and with both active chronic gastritis and peptic ulcer disease.

On the contrary, with respect to the oipA, hopZ, and sabA genotypes, no significant association was observed between their status and the clinical outcome in patients from whom the related strains had been isolated and, in most cases, between their “on” or “off” statuses and the presence of any other virulence-associated genotype. The results in these cases conflict with data in the literature [1], mainly concerning the oipA gene, encoding an OMP usually considered to be involved in some crucial aspects of H. pylori pathogenicity and useful as a marker to discriminate duodenal ulcer from gastritis [50]. Because we evaluated the oipA functional switch status using DNA sequencing, on the basis of the number of the CT dinucleotide repeats in the 5’ region of the gene [28], the effective production of the OipA protein needs to be tested [51] and the possibility has to be excluded that in our strains mutations in the promoter region prevented the switch effect of the CT repeats on the expression of a functional open reading frame.

In conclusion, although the need for these latter evaluations remains, it was possible to define the prevalence of many virulence-associated genotypes in our Sicilian H. pylori isolates and their association with the clinical outcome in the infected patients. It is interesting to note that the prevalence values of the genotypes assayed were more similar in our isolates to values observed in the north of Italy and France than to values observed in Spain or other Mediterranean countries that are nearer and climatically more similar to western Sicily.