Presence of the Genes cagA, cagE, virB11 and Allelic Variation of vacA of Helicobacter pylori Are Associated with the Activity of Gastritis

Non-atrophic active chronic gastritis (ACG) is characterized by the presence of H. pylori in the gastric epithelium, known to be one of the first steps that precede progression to gastric adenocarcinoma. Inactive chronic gastritis (ICG) suggests that the patient has H. pylori gastritis, but this diagnosis is rarely made in routine histopathology. Clinical manifestations associated with H. pylori infection are potentially due to differences in virulence between strains; however, it is unclear if the progression of ACG to ICG depends on the H. pylori strain. The aim of this study was to compare the prevalence of the virulence factors of H. pylori found in patients with ACG and ICG, and its influence on the development of ICG. A significant association was observed between H. pylori detection by histological examination and the activity of gastritis (p < 0.01). Long-term use of proton pump inhibitors (PPI) (>1 year) was reported by 28.6% of the ACG group and 42.5% of the ICG, while no evidence of association between long-term use of PPI and decreased inflammation was found in the patients studied. The genes cagA, cagE and virB11 were statistically associated with ACG (p = 0.01, p < 0.001 and p = 0.002, respectively). In the vacAs1 allele groups, ACG was associated with the most virulent group (p = 0.0015), while ICG was associated with the less virulent group (p < 0.001). The rate of co-infection was significantly higher in ICG than in ACG cases (p = 0.02). In conclusion, this study points to the role of virulent strains of H. pylori in the non-resolution of gastritis. These authors contributed equally to this work. P. P. de Negreiros Bessa et al. 348


Introduction
The presence of Helicobacter pylori in the gastric epithelium typically leads to an antral-predominant nonatrophic active chronic gastritis (ACG) lesion, known to be one of the first pathologic markers of H. pylori infection and one of the steps preceding progression to gastric adenocarcinoma [1].Inactive chronic gastritis (ICG) is infrequently observed in routine histopathology, and the association with positive H. pylori serology and post-eradication status is interpreted as an indicator of pre-existing H. pylori gastritis [2].Goldstain [3] (2002) found substantial rates of H. pylori infection in ICG patients.
Differences in the clinical manifestations associated with H. pylori infection are potentially due to differences in virulence between H. pylori strains [4].Two factors are well established in H. pylori virulence, namely the presence of the vacA s1m1 allele [5] and the cagA gene, although other virulence factors have been pointed out as potentially relevant to gastric cancer development.In particular, cagE and virB11 genes, also belonging to the cag pathogenicity island (cag-PAI), have been correlated with more severe lesions [6]- [10].Some authors even consider the cagE gene a better marker of cag-PAI than the cagA gene [11]- [13].
Taking into account the importance of H. pylori virulence in gastric lesions, it is unclear if the progression of ACG to ICG depends on the H. pylori strain or if the bacterium found in ICG is a casual finding or plays an important role in the pathogenesis of this lesion.There are no published studies that have performed H. pylori genetic analyses considering both lesions, ICG and ACG.Therefore, this study aimed to compare the frequency of the virulence factors in H. pylori strains detected in patients with ICG versus ACG, to see if the presence or absence of the virulence factors studied was relevant to the development of ICG.

Clinical Specimens
Samples were collected from 206 patients with dyspeptic symptoms who underwent endoscopy at two hospitals, Hospital Universitário Walter Cantídeo, and Hospital Geral de Fortaleza (HGF), both in Fortaleza, CE.All subjects signed an informed consent form before inclusion and a clinical and socioepidemiologic questionnaire was completed.Patients were excluded in the following cases: if duodenal, gastric or esophageal masses or Barrett esophagus lesions had been seen on endoscopy; previous gastric or duodenal surgery had been performed; intestinal metaplasia or atrophy were present in the gastric biopsy specimen; or if there had been recent use of nonsteroidal anti-inflammatory drugs.After these exclusions, 152 patients remained in the study group.H. pylori infection was confirmed by two tests, PCR and histopathological analysis.A total of eight biopsy specimens of gastric antrum and body were collected, in which four fragments were for histological examination and four fragments for DNA extraction and PCR assays.
Samples for PCR were immediately frozen at −80˚C.Samples for histopathological examination were preserved in 10% formaldehyde and submitted to routine processing.The levels of lymphoplasmacytic inflammation (chronic gastritis) and polymorphonuclear neutrophil inflammation (activity) were assessed using the updated Sydney gastritis classification system, with chronic gastritis being scored as normal, mild, moderate or marked.This study was approved by the Ethics Committee of the Federal University of Ceará (047.06.09) and by the Committee for Ethics and Research of HGF (070714/10).

DNA Extraction
Genomic DNA was extracted from two frozen fragments according to the lesion location, one obtained from the antrum and the other from the body, using cetyltrimethyl ammonium bromide (CTAB), adapted from the method of Foster and Twell [14] with some modifications.DNA extraction quality was analyzed by 1% agarose gel electrophoresis with ethidium bromide staining, and the DNA amount was determined using the NanoDropTM 3300 fluorospectrometer.

H. pylori and cagA, cagE, virB11 and vacA Arrangement Gene Detection
The H. pylori infection was detected by PCR amplification of the urease C gene using primers described by Lage [15].For the H. pylori-positive samples, the presence of the vacA alleles, cagA, cagE and virB11 genes were identified using primers described in the literature (Table 1).PCR mixtures, for amplification of ureC, cagE and virB11 genes, were performed using Green Master Mix ® 2× (Promega ® , Madison, USA), according to the manufacturer's instructions, with addition of 0.8% Tween 20 mix for cagE, and water with 0.1% of BSA for virB11.The concentration of each primer was 0.48 μM for ureC and 0.4 μM for both virB11 and cagE.DNA sample was used at a concentration of 100 ng.For the detection of cagA, vacA s1/s2, vacA m1 and vacAm2 genes, the PCR mixture consisted of 10× PCR buffer Invitrogen ® , a final concentration of 1× taq buffer and 2U of enzyme (Invitrogen Platinum ® Taq DNA Polymerase), BSA, 0.2 mM dNTP, 1.5 mM MgCl 2 , 0.4 μM primer and 100 ng DNA.
Amplified products were electrophoresed in a 6% polyacrylamide gel (Figure 1 and Figure 2), silver staining, for the genes ureC, cagA and vacA, and in 1% agarose gel (Figure 3) with ethidium bromide staining, for the genes virB11 and cagE.The size of the amplification product (Table 1) was used to confirm the identity of the PCR product.

Statistical Analysis
The statistical analyses were conducted using the EPINFO ® 6.0 and SPSS ® 15.0 versions of the statistical software programs (SPSS, Chicago, IL, USA).Statistically significant differences were evaluated by the Chi-square test (χ 2 ) and Fisher's exact test.A p-value lower than 0.05 was considered as statistically significant.

Results
Among the samples analyzed, 40 cases of ICG (26.3%) and 112 of ACG (73.7%) were identified.Clinical and epidemiological data, such as the predominant location of the injury evaluation, were compared between patients with ICG or ACG, as shown in Table 2.No statistically significant differences were found between the groups.
In relation to the intensity of the inflammatory process, 80% (32/40) of the ICG group showed mild gastritis and 20% had moderate gastritis.No intense inflammation was found among the ICG patients.When the CGA group was evaluated, 50.9% (57/112) had mild, 37.7% (40/112) moderate and 11.4% (15/112) intense gastritis.Long-term use of proton pump inhibitors (PPI) (>1 year) was reported by 28.6% (32/112) of the patients diagnosed with ACG and 42.5% (17/40) of the ICG group.Although a higher number of cases of ICG were associated with treatment, no evidence of association between long-term use of PPI and decrease in polymorphonuclear infiltrate was found in the sample studied (p = 0.15).
H. pylori infection was positive in 98.2% (110/112) of patients with ACG and in 92.5% (37/40) with ICG, according to PCR detection.Histological examination showed a significantly lower frequency of H. pylori infection in the ICG group than in the ACG group (Table 3).A significant association between H. pylori detection by histological examination and the presence of polymorfonuclear infiltrate in the gastric mucosa was also observed (p < 0.01).Genotyping results for the H. pylori virulence genes studied are summarized in Table 4.There was a significantly higher frequency of H. pylori with cagA, cagE and virB11 genes in the ACG group.The allelic combination of vacA s1m1 was the most frequent in both groups, and none of the allelic combinations showed a statistically significant association with gastritis classification.The analysis of vacA allelic variability revealed the presence of a co-infection, characterized by the detection of strains with both vacA alleles.This was significantly more frequent among the ICG cases (86.5%) than the ACG cases (64.5%) (p = 0.02).Regardless of the presence of co-infection, it was observed that strains with a high virulence gene profile, concerning cag-PAI genes, were more frequent in ACG.
Based on the importance of the vacA s1 allele, the H. pylori strains were divided into two groups (I and II), according to allele presence.Within these groups, the strains were also grouped into four subgroups (A, B, C or D), in an attempt to assess the integrity of the cag-PAI island, as shown in Table 5.The strains with a high virulence gene profile (IA group) were strongly associated with ACG cases (p = 0.0015), while the ID group was associated with ICG (p < 0.001).Strains with the vacA s2 phenotype were not associated with any of the two groups analyzed.
The virulence of H. pylori strains was categorized according to the presence of vacA alleles and the cag-PAI genes.Strains with vacAs1 allele and all cag-PAI genes studied (IA) were considered the most virulent followed by the ones which had at least one right marker (cagA/cagE) and one left marker (virb11).

Discussion
In this study, high rates of H. pylori infection determined by PCR detection were found in both gastritis groups, ACG and ICG.The rate of H. pylori detection found was higher than that reported in other studies [18] [19], probably because this study was conducted in a population with a low socioeconomic status and deficient sanitary conditions [20].Histological staining showed statistically lower rates of H. pylori detection when compared with PCR, mainly in the ICG group in which false negatives made up 75% of cases.The low frequency of H. pylori in the histological specimens could be explained by two studies that showed that stress conditions, such as use of PPI and antibiotics, may inhibit bacterial growth cause changes in morphology from the spiral to coccoid form, making the bacteria difficult to identify in routine staining for H. pylori [3] [21].Most studies determining the presence of H. pylori in ICG have used histological analysis for bacterial identification, explaining the low detection rates found in these studies.In addition, PCR is a technique with higher sensitivity and specificity compared histological assays [22] [23].Untreated H. pylori-associated gastritis is usually antral-predominant and ACG [1].PPI intake and H. pylori eradication therapy rapidly result in a reduction or absence of polymorphonuclear infiltrate in the antrum and a slower long-term decrease in lymphoplasmacytic inflammation in this region [24] [25].Although the literature shows an association between inactive chronic gastritis and previous H. pylori eradication regime and PPI intake [2] [3], in the present study no significant association was found between chronic uptake of PPI and inactive gastritis.A higher number of cases of ICG were associated with treatment.It is important to note that in other published studies, the association between PPI uptake and ICG could only be demonstrated in some of the cases studied, and therefore, other causative factors may be involved in the development of ICG.
Genotyping results for H. pylori virulence factors showed a statistically significant association between the cag-PAI genes and the presence of active gastritis.Several genes from H. pylori cag-PAI, such as cagA, cagE and virB11, can induce the expression of cytokines in the gastric epithelium, especially IL-1B and IL-8 [8] [26] [27].As IL-8 is an important neutrophil chemotactic and activating factor [28], the cag-PAI genes seem to have an important role in the development of polymorphonuclear infiltrate, which characterizes ACG.This is supported by the results obtained when the genotypes were grouped in an attempt to assess the integrity of the cag-PAI.Strains with a complete island were associated with activity, while the absence of the island showed an association with lack of polymorphonuclear infiltrate.vacA alleles were not associated with the development of activity.Nevertheless, it seemed to be related to the intensity of the inflammation response.It has been shown that the vacA toxin can act as an immunomodulator by interfering with the IL-2 signaling pathway in T-cells [29] or interfering with antigen presentation mediated by MHC class II cells [30].In the present study, regardless of the high prevalence of vacA s1 strains in the population analyzed, i.e., 85.7% of all samples, no association was found between vacA genotypes and the development of gastritis activity or the intensity of gastric inflammation in the samples studied.

Conclusion
In conclusion, this study showed that the genes cagA, cagE and virB11 and vacAs1 allele were statistically associated with ACG.These data point to the fact that the non-resolution of gastritis may be due to the bacterial genotype.There have been no reported studies comparing H. pylori strains in ACG and IGG considering cag-PAI genes.Additionally, the present study corroborates others in demonstrating that PCR is better than histological analysis for H. pylori detection.In contrast to reports in the literature taking into account the use of PPI, no association was found between the use of this medication and the decrease in polymorphonuclear infiltrate.

Figure 1 .
Figure 1.Gel 6% polyacrylamide stained with silver nitrate, showing the detection of specific fragment of 294 pb (arrow) from the amplification reaction of the ureC gene.

Figure 2 .
Figure 2. Polyacrylamide gels to 6% by detecting specific fragments of 297 pb for cagA gene (a), and 290 bp for 192 pb alleles of vacA m1 and m2 (b) and 259 pb and 286 pb allele of vacA s1 and s2, respectively (c).

Figure 3 .
Figure 3. Agarose gels 2% detecting fragments of 509 bp generated from the amplification of the cagE gene (a) and 491 pb to virB11 (b).The molecular weight marker used in the gels of 2% agarose was 1 kb to cagE, virB11 (Invitrogen).

Table 1 .
PCR primers used for genotyping Helicobacter pylori.

Table 2 .
Clinical epidemiological data of surveyed population: comparison between ACG and ICG.

Table 3 .
Comparison of PCR and histological methods used for diagnosis of Helicobacter pylori infection.

Table 4 .
Frequencies distribution of the H. pylori genes according to active chronic gastritis and Inactive chronic gastritis in the population studied.

Table 5 .
Distribution of strains of H. pylori according to virulence genes in consonance to activity of gastritis (chronic active gastritis and chronic inactive gastritis) in the population studied.