Infection with Helicobacter Pylori Presenting the vacA s2m2 haplotype is Strongly Associated with Protection Against Gastric Cancer


 Background: Infection with Helicobacter pylori is recognized as the main risk factor for gastric cancer (GC); the clinical outcome of this infection is variable and partially depends on the virulence of the infective strain. This study characterizes H. pylori virulence genes in patients with diverse gastric lesions, from preneoplasia to GC, from a South American region with high GC mortality rates.Methods: We studied the virulence profiles of H. pylori strains to colonize the antrum of 318 patients with non-atrophic gastritis (NAG), 58 patients with preneoplastic lesions (PN), and 90 with GC from Ibagué, Colombia. The presence of 16S rDNA, the cagA and cagE genes, and the vacA s1, s2, m1, and m2 alleles were determined by PCR.Results: H. pylori infection was detected in 44% of all patients, 41.2% in NAG, 43.1% in PN and 54.4% of GC patients (p= 0.0813). cagA and cagE genes were significantly more frequent in and GC than in NAG (p= <.0001). The vacA s1m1 haplotype was significantly more frequent in PN (68%) and GC (65.3%) than in NAG (37.4%). The frequency of vacA s2m2 haplotype decreased significantly from NAG (42.7%) to PN (12%) and this to GC (4.1%). A total of 23 different genotypes were identified, with cagA+/cagE+/vacA s1m1 (84/205) as the more frequent in PN and GC and cagA-/cagE-/vacA s2m2 in NAG (49/205).Conclusions: In the population studied, vacA s2m2 was identified as a significant marker for protection against PN and GC, and genotype cagA+/cagE+/vacA s1m1 as a marker for increased GC risk. We also found that patients with PN and GC had a higher frequency of cagA+/cagE+/vacA s1m1 H. pylori strains known to be aggressive.


Background
Gastric cancer (GC) is the third leading cause of cancer-related deaths worldwide. 1,2,3 In Colombia, it is the leading cause of cancer-related deaths in men and the fourth in women. 4,5 Globally, the GC incidence varies across different countries; in Latin America, an association between altitude and GC risk has been observed and Chile, Costa Rica, and Colombia are some of the countries with the highest mortality rates in the world. 4 The bacteria Helicobacter pylori (H. pylori) is classi ed as a class I human carcinogen and recognized as the main risk factor for GC. 6,7 This bacteria promotes in ammation of the gastric mucosa that in some patients lead to premalignant lesions such as chronic atrophic gastritis, intestinal metaplasia, and dysplasia that precedes the appearance of GC. [6][7][8][9][10][11][12][13] Worldwide, there is no correlation between the prevalence of H. pylori infection and GC incidence. 14 This bacterium colonizes the gastric mucosa of 50% of the worldwide population and 80% in developing countries; still, only 1-3% of infected people develop premalignant gastric pathologies. 9,15,16 In Colombia, whereas the prevalence of H. pylori infection varies widely between regions, 14,17−19 the mortality of GC as well as a higher prevalence of precancerous lesions are concentrated in the central part of the country that corresponds to the Andes Central Mountain Range, mainly in the Coffee axis departments (Caldas, Risaralda, and Quindío), Norte de Santander, Boyacá, Huila, Cauca, Tolima, and Caquetá. Furthermore, the lowest GC mortality rates are reported in regions with low altitudes but a high prevalence of infection, such as the Guajira, Chocó, Córdoba, Putumayo, and Sucre departments. 12 Several theories have been raised regarding the association between altitude and GC, but they have not been well elucidated yet. It is likely that those mountain regions cluster host genetic, dietary, and environmental factors as well as bacterial genotypes that promote GC development. 8,11-13,20−25 Among the best-studied H. pylori virulence factors are cytotoxin-associated gene A (cagA) and vacuolating cytotoxin gene A (vacA). More virulent strains carry the cagA gene, which encodes the oncoprotein CagA, which induces a myriad of changes in the gastric mucosa: an increase of in ammation, the loss of epithelial polarity, the disruption of intercellular junctions, increases proliferation, reduces apoptosis, and eventually promotes carcinogenicity. [26][27][28] The vacA gene, which is virtually present in all H. pylori strains, encodes the VacA protein, which is responsible for cytoplasmic vacuoles and pores in the membrane of gastric epithelial cells and apoptosis. [29][30][31] Some vacA alleles are known to increase the risk of developing peptic ulcer disease and GC. [32][33][34][35][36] Infection with H. pylori strains that possess the vacA s 1 m 1 genotype is associated with an increased risk of peptic ulceration and GC. 27,32,[37][38][39] However, little attention has been paid to evaluate the prognostic value of the vacA s 2 m 2 haplotype.
In the Tolima Department of Colombia, a central mountainous region with high GC mortality rates, H. pylori infection prevalence varies between 59% and 66% but the correlation between the bacterial genotypes and gastric lesions has not been properly studied. 32,40−42

Methods
In this study, we characterized H. pylori virulence genes in patients with NAG, PN, or GC from Tolima, with a particular interest in the study of vacA s 2 m 2 haplotype.

Patients and sampling
A total of 466 patients with gastric pathologies were recruited during the period 2010-2019 in Ibagué, including 90 patients with GC who had undergone upper gastrectomy at the Federico Lleras Acosta Hospital and 376 patients who had undergone upper gastrointestinal endoscopy as part of the dyspepsia study at the Javeriano Medical Center. A gastric biopsy from the pyloric antrum region was taken from each patient, placed in 70% alcohol solution, and stored at -20°C until studied. Biopsies from antrum and tumor lesions were placed in paraformaldehyde for histology studies. All patients were informed about the study and if willing to participate they were asked to sign an informed consent letter. The research protocol used in the study was approved by the University of Tolima Ethics Committee and adhered to the Helsinki Declaration.

Histopathological examinations
Biopsies in paraformaldehyde were embedded in para n for histopathological diagnosis. Antral and tumor sections were stained with hematoxylin and eosin and evaluated independently by three surgical pathologists to establish the diagnosis in each case. Patients were assigned to one of the three groups: 1. chronic non-atrophic gastritis (NAG); 2. PN, which included patients with chronic and atrophic gastritis, intestinal metaplasia, and dysplasia; and 3. GC. Molecular identi cation of H. pylori DNA was extracted from the gastric antrum biopsy specimens with the DNeasy Blood and Tissue Kit (QIAGEN, USA) following the manufacturer's instructions. DNA was quanti ed using a Nanodrop ND™ 1000 UV-Vis spectrophotometer of Thermo Scienti c and quality was evaluated by using the ratio of absorbance at 260 and 280 nm (A260/A280). To test for the presence of H. pylori, a 537-bp fragment of the Sub unit 16 of ribosomal DNA (16Ss rDNA) gene was ampli ed using polymerase chain reaction (PCR) with the primers ACT-1 and ACT-2, as previously described (ref). DNA extracted from the H. pylori NCTC 11638 strain (donated by the National Institute of Cancerology, Bogotá, Colombia) was used such as positive control.
Ultrapure water instead of DNA was used such as negative control. The nal volume of 25 µl contained 9.5 µl of ultrapure water, 1 µl of each primer, 12.5 µl of the BIOLINE MyTaq™ Extract from the PCR kit, and 1.5 µl of DNA. The thermocycling program was previously described by López et al. 40 Ampli cation and typing of the cagA, cagE, and vacA genes To determine the genotype of H. pylori virulence genes the samples were subjected to PCR for the cagA and cagE genes and the signal regions (s 1 and s 2 alleles) and mid-region (m 1 and m 2 alleles) of vacA using primers and conditions previously described. 33,40,43 As a control for one possible inhibition of the reaction, a fragment from the human β-globin gene was ampli ed. All PCRs were performed in a Bio-Rad Dual-Touch 1000 thermocycler. The ampli ed products were visualized on a Thermo Fisher™ ultraviolet light transilluminator on 1.5% agarose gels at 100 volts for 60 minutes using ethidium bromide (0.4%). All primers used in this study are listed in Supplementary table 1, Additional File 1.

Statistical methods
The chi-square (X 2 ) independence tests were performed to evaluate differences between the epidemiological variables, the presence of H. pylori infection, and the bacterial genotypes, and the gastric pathology of the patients. A two-tailed p-value (α) < 0.05 was considered statistically signi cant. The data were processed with R software version 3.6.1.

Results
A total of 466 patients with different gastric pathologies were included in this study. There were 285 women (61.2%) and 181 men (38.8%), with an average age of 52.1 ± 16 years. A 2:1 ratio of women to men was observed in patients with premalignant pathologies, while in patients with GC, this proportion was reversed The molecular diagnosis tests showed that 44% (205/466) of the patients were infected with H. pylori, 59.5% (122/205) were women and 40.5% (83/205) were men. The H. pylori prevalence was lower than 50% in most gastric histological lesions, except in GC (54.4%). The infection status was not associated with the sex, age, or gastric pathology of the patients. The molecular diagnosis results are presented in Table 1. In all H. pylori-positive biopsies, the cagA, cagE, and vacA genes were ampli ed in 59.5%, 66.3%, and 94.6% of cases, respectively (Table 2). Strains with the cagA and cagE genes were signi cantly more frequent in PN and GC than in NAG (p-value 0.004 and <0.0001, respectively), and OR values showed that the presence of these genes increased around 10 times the risk for GC. As the severity of gastric pathology increased, the proportion of strains with cagA and cagE also increased ( Figure 2). These genes were present in almost 90% of H. pylori strains in GC, regardless of the subtype of cancer (intestinal or diffuse). PCR ampli cation of the signal regions (s 1 and s 2 alleles) and mid-regions (m 1 and m 2 alleles) of vacA showed different allelic combinations (Supplementary table 2, Additional File 1) with a predominance of s 1 m 1 (47.8%) and s 2 m 2 (29.8%). The s 1 m 1 genotype was signi cantly more frequent in patients with PN lesions (68%, p-value 0.001) or GC (65.3%, p <0.0001) than in patients with NAG (37.4%), and OR values indicated a risk for GC of 18 times higher when s 1 m 1 is present. Besides, the s 2 m 2 haplotype was frequent in NAG (42.7%) but drastically decreased in PN (12%,) with an OR value of 0.1544 and was almost absent in GC cases (4.1%) with an OR value as low as 0.0547 ( Figure 2 and Table 2).  cagA-/cagE-/vacA s2m2 (24%) genotypes. Patients with NAG presented the highest variation in genotypes (21 different genotypes), followed by patients with PN (10 genotypes), patients with GCI (9 genotypes), and nally, patients with GCD and GCM (5 genotypes each); there seems to be a selection of genotypes as the disease progress. In patients with NAG, the genotypes cagA-/cagE-/vacA s2m2 and cagA+/cagE+/vacA s 1 m 1 were the most frequent (

Discussion
GC is a multifactorial disease associated with genetic, environmental, and infectious factors, with H. pylori infection as the more important risk factor. GC associated with H. pylori infection is the result of long-term chronic in ammation in the gastric mucosa that when unregulated may lead to tissue damage, which may progress to atrophic gastritis, intestinal metaplasia, dysplasia, and eventually GC. 44,45 Bacterial genotypes play an important role in clinical outcomes, particularly when they are associated with an increased in ammatory response; however, this association may vary between and within different populations. 46 In Colombia, the prevalence of H. pylori infection and the risk for GC vary among the different departments. 10,58,60 The association between gastric lesions and H. pylori virulence factors have been evaluated in Colombian departments with a similar prevalence of infection but contrasting GC risk and higher frequencies of cagA-positive and vacA s 1 m 1 genotypes were found in populations with a higher risk of GC than in low-risk areas. [47][48][49] In Ibagué, located in the Andean mountain region with high GC mortality rates, 12,13,50,51 only studies with small samples size have been carried out, analyzing cagA and the s and m regions of vacA. [40][41][42]52 In the present study we aimed to do a more comprehensive analysis of genotypes in virulence genes in larger groups of patients.
In our studied population, the most frequent type of GC was GCI (44.4%), similar to previous works, [55][56][57][58] while the frequency of GCM (27.8%) was higher than that reported in other countries. 70,71 Of note, 20% of the GC patients had an average age of diagnosis of 42 years, which shows the need for the implementation of early detection programs to identify and treat the disease.
The genotyping of virulence genes showed that in our population the frequency of the cagA gene was lower (59.5%) than that reported in previous investigations in Colombian mountain regions 33,48,64,65 and other Latin American countries, with 67 to 80% reported in patients with premalignant lesions. 34,66−68 The presence of the cagE gene has been suggested as a better marker than the cagA gene about the integrity of the Cag Pathogenicity Island (cag PAI), and it has been observed with a frequency greater than 80% in patients with gastrointestinal diseases in other countries. 46,69 In the present study, 66.3% of infected patients were positive for cagE, which was lower than in previous reports in Colombian strains. 33,54 However, the frequency of these genes was signi cantly higher in patients with severe lesions, and cagA was present in 76% of PN and 89.8% in GC, whereas cagE was in 80% of PN and 91.8% in GC. These results are in agreement with previous studies in Colombia 33,48 and con rm the value of using these genes as markers for PN and GC.
Analysis of the vacA gene alleles showed that in the group of patients with PN and GC, the most frequent combination was s 1 m 1 , whereas the s 2 m 2 genotype was rare; in contrast, in patients with NAG, the most frequent combination was s 2 m 2 ( Table 2). These results are consistent with those of previous reports, where the s 1 m 1 genotype has been widely associated with the presence of chronic in ammation and the development of severe gastric pathologies. 70 while the vacA s 2 m 2 nontoxic strains are more frequent in patients with non-ulcer dyspepsia and mild gastritis. 35,71−74 The distributions of the tested genes strengthen the observed association of the genes with the disease; thus, the percentage of strains harboring the cagA and cagE genes and the vacA s 1 m 1 haplotype increased as the severity of the gastric pathology increased (Fig. 2). A similar trend was previously reported by Cittely et al. 48 in Colombian patients from Bogotá and in patients with gastroduodenal disorders in other countries. 36 However, little attention has been paid to the value of vacA s 2 m 2 as a potential marker of disease risk; our results show a strongly signi cant protective factor of s 2 m 2 for PN (OR 0.1544) and even stronger for GC (OR 0.0547) where this vacA haplotype is almost absent. Thus, carrying strains with the vacA s 2 m 2 haplotype signi cantly reduces the risk for gastric cancer and to determine vacA alleles there is no need to isolate and culture H. pylori, the test can be done in DNA from biopsies. These make PCR ampli cation of vacA alleles a useful and accessible biomarker to assess the risk for gastric cancer.
In this study, we analyzed genotype pro les using results from all genes and alleles studied and found as many as 23 genotypes; two of them commonly reported and another 21 different combinations that are less commonly found ( Supplementary Fig. 1, Additional File 1). We aimed to see if building a genotype would improve the value of these tests as markers for disease risk. The most frequent genotypes were cagA+/cagE+/vacA s 1 m 1 (41%) and cagA-/cagE-/vacA s 2 m 2 (23.9%). A positive association was found between the severity of these lesions and the cagA+/cagE+/vacA s 1 m 1 H. pylori genotype, while the cagA-/cagE-/vacA s 2 m 2 genotype was rare in patients with PN and GC and was more frequent in patients with NAG, which supports the relationship between histological features and H. pylori genotypes reported in previous studies. 35,71,74,75

Conclusions
In conclusion, the present study shows that the cagA+/cagE+/vacA s 1 m 1 genotype is a marker for increased GC risk, whereas cagA-/cagE-/vacA s 2 m 2 is a marker for decreased risk. Of importance, our results suggest that the determination of vacA alleles is useful enough to evaluate the risk for disease and the presence of s 2 m 2 is a strong marker for protection against gastric cancer.

Declarations
Ethics approval and consent to participate.
All subjects included in the study provided his/her written informed consent for study participation.
Similarly, the study protocol was developed in concordance with The Declaration of Helsinki (subsection of ethics approval and consent to participate). Consent for publication Not applicable Availability of data and materials The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Competing interests
None declared. The authors declare that they have no competing interests related to the subject matter or materials discussed in this article. Authors' contributions AGT: conceptualization, data curation, formal analysis, funding acquisition, investigation, methodology, project administration, resources, software, supervision, validation, visualization, writing -original draft, writing -review & editing FCV: data curation, formal analysis, investigation, methodology, validation, visualization, writing -review & editing.
All authors read and approved the nal manuscript.