The organization of Helicobacter pylori cag-pathogenicity island (cagPAI) genes in multiracial population with histopathological changes of gastric mucosa

Background: Helicobacter pylori is a Gram-negative bacillus that colonises only the mucus layer of the human stomach and is implicated in gastric diseases. Virulent H. pylori harbouring cag-pathogenicity island (cagPAI) which encodes genes for type IV secretion system (T4SS) and CagA protein is one of the major virulence determinants involved in disease development. We examined the entire cagPAI genes in 95 H. pylori isolates from a multiracial population and examined the intactness of cagPAI region with histopathological scores of the gastric mucosa. Results: 95.8% of H. pylori isolates were cagPAI-positive with 23.2% having an intact cagPAI, whereas 72.6% had a partial/rearranged cagPAI. In our study, cag2 and cag4 were found to be significantly higher in H. pylori isolated from Malays, whereas cag4 was predominant in Chinese isolates. We also detected cag24 in significantly high proportion in isolates from the Malays and the Indians compared to the Chinese isolates. The intactness of cagPAI region showed an association with histopathological scores of the gastric mucosa. Significant association was observed between H. pylori harbouring partial cagPAI and higher density of H. pylori and neutrophil activity, whereas strains which lacked cagPAI was associated with higher inflammatory score. Conclusions: The screening of the entire cagPAI genes provides an accurate overview of the cagPAI organisation in H. pylori isolates in a multiracial population. The genotypes of H. pylori strains with various cagPAI rearrangement associated with patients’ ethnicities and histopathological scores might contribute to the pathogenesis of H. pylori infection in a multi-ethnic population.

causes gastrointestinal diseases such as chronic gastritis, peptic ulcer, gastric cancer and gastric mucosa-associated lymphoid tissue (MALT) lymphoma [2,3], although most infected patients appeared asymptomatic. Hence, H. pylori is also classed as type I carcinogen [4]. Factors that contribute to the infected patient's disease sequelae include environmental factors such as lifestyle and diet , host genetics, host immune responses and bacterial virulence factors [4][5][6].
Cytotoxin-associated gene pathogenicity island (cagPAI) is one of the major virulence factors associated with disease outcome in infected hosts. It is approximately 40 kb in size consisting of around 28 genes [7], encoding mainly CagA protein, type IV secretion system (T4SS) and other genes for induction of host's interleukin-8 (IL-8) [7,8]. Although the mechanisms resulting in severe disease development are poorly understood, a major factor is likely to be H. pylori-induced gastric injury and inflammation [9]. Studies show that intactness of cagPAI has a significant correlation with disease severity, whereas H. pylori strains with partial deletions within cagPAI region are significantly less-pathogenic in nature [10,11]. However, the rates of severe disease development vary between human populations, and differences in H. pylori genotypes may partially explained these differences [12,13].
Integrity of cagPAI seems to have an important role in the progress of the gastroduodenal disorders, so that intact cagPAI could be seen in H. pylori strains from countries with higher rate of gastric cancer [14]. This integrity also has important effect on the induction of inflammatory response in the gastric mucosa [15]. Several studies have investigated the association of H. pylori cagPAI and gastroduodenal diseases [14,16], however, knowledge about the relationship between H. pylori cagPAI intactness and changes of the infected gastric tissue is sparse. More than 90% of H. pylori strains in Malaysia are cagPAIpositive [17] and Malaysian population consists of multi-ethnic people, therefore the interaction of H. pylori strains with different genotype in various host genetics may have an impact on the differences in disease development.
The organisation of cagPAI genes in H. pylori in Malaysian population which has multiethnic groups of people has not been well studied. There is lack of comprehensive information with regards to abundance of intact versus rearranged cagPAI among H. pylori strains in this population. Hence, in this study, we sought to characterise the genes within cagPAI and to determine the association of various cagPAI structure in H. pylori isolates with histopathological changes of the infected gastric mucosa. The outcome of this study may provide valuable information in order to draw association between existence of cagPAI genes and its association with disease sequelae in strains from multi-ethnic population and also in strains isolated in different histopathological conditions.

Results
Histopathological characteristics of the gastric mucosa in the studied populations Histopathological scores of the gastric mucosa among different ethnic groups showed that the Malays had higher mean scores for H. pylori density and neutrophil activity whereas the Chinese showed higher grade of inflammation (Table S1). Higher mean score for intestinal, metaplasia was observed among the Indians, while the atrophy of higher grade was observed in the Chinese. Patients of different ethnicities were grouped into different types of disease conditions based on the histopathological changes (Table S2), i.e chronic gastritis (CG) (n=20), chronic active gastritis (CAG) (n=44) and intestinal metaplasia/atrophy (IM/Atr) (n=28). There was a significant difference in the proportion of CG and CAG between the Chinese and the non-Chinese patients. CG was diagnosed more in the Chinese patients compared to the non-Chinese (p = 0.03), whereas CAG and IM/Atr were observed more in the non-Chinese than the Chinese (p = 0.042).
Distribution of the cagPAI genes in H. pylori isolates Detection of the cagPAI region in our clinical H. pylori isolates showed that 95.8% (n=91) of the isolates were cagPAI-positive. Four genes in the cagPAI region (cag1, cag6, cag8 and cag21) were detected in all isolates whereas 35.2% isolates were cag2 (n=32) and 52.7% cag14 (n=48) ( Table 1). Detection of other genes ranged from 69.2 -98.9%. The absence of cag2 was confirmed with 690 or 1100 bp amplicon using empty-site PCR as described by Schmidt et al., [21]. cag14 was detected using 4 sets of primer pair as described by Ta et al., [20].

Analysis of cagPAI intactness in H. pylori isolates
The cagPAI was defined as intact if all the gene sets of the cagPAI were present including strains lacking only the cag2 (HP0521). A previous systematic mutagenesis study showed that the HP0521 gene was not involved in the process of CagA translocation and IL-8 induction Fischer et al., [7]. In addition, NCBI database defined the HP0521 as a pseudogene (NCBI-Gene ID: 900040) (DBGET/LinkBD: an integrated database retrieval system, last accessed Oct 8, 2018). Partial cagPAI was defined when an isolate lacked one (other than HP0521) or more of the cagPAI genes, while negative/deleted cagPAI was defined if none of the genes were present and a product of approximately 650 bp with primers from the flanking regions was obtained. Among the 91 cagPAI-positive H. pylori strains, 24.2% (n=22) had intact cagPAI and 75.8% (n=69) exhibited partial (rearranged) cagPAI. Strains harbouring intact or partial cagPAI were not associated with patients' ethnicities (p > 0.05).
Association between cagPAI intactness and histopathological scores of the gastric mucosa are shown in Table 2. The presence of partial cagPAI was significantly related to the higher total score of H. pylori density (p = 0.036) and neutrophil activity (p = 0.03) compared to the intact cagPAI. H. pylori harbouring deleted cagPAI was significantly correlated with higher inflammatory score (mononuclear infiltration) compared to H. pylori with partial cagPAI (p = 0.002). The distribution of H. pylori with intact cagPAI was detected more in the gastric mucosa with IM/Atr, whereas partial cagPAI H. pylori was detected more in CAG, however the difference was not significant (Table 3).

Discussion
Racial differences in the prevalence of H. pylori infection and disease-related severity were observed among patients from multiracial ethnicities [22,23]. Bacterial virulence factor is one of the contributing factors to the development of severe H. pylori-related diseases. The diversity of cagPAI region in the H. pylori genome may have a modifying effect on the pathogenic potential of the infecting strain [24].
In this study, we comprehensively determined the presence of all cagPAI genes in 91/95 H. pylori isolates from Malaysian population which were isolated from patients of different ethnic groups. The results show that more than 95% of our H. pylori strains were cagPAIpositive where 24.2% of the isolates carry all cagPAI genes, 75.8% exhibited partial or rearrangement in the cagPAI genes. In our previous study, we detected only 3.2% of the isolates carrying all the selected cagPAI genes [17]. The low percentage of H. pylori isolates harbouring intact cagPAI genes in our previous study is because we analysed only a subset of the cagPAI genes (cag67, cag10, cag13, cagT, cagM and cagE) as these genes was shown to have linkage between certain genes in the cagPAI region and severe disease as described by earlier studies [25,26]. In contrast, high frequency of intact cagPAI and low frequency of partial cagPAI in H. pylori strains isolated from similar ethnic populations was reported by Schmidt et al., [21]. In their study, few cagPAI genes (cagE, cagL, cagT and HP521) were examined to detect the intactness of cagPAI region. Discordant in the frequency of cagPAI intactness in many reports was due to the difference cagPAI genes that being examined [14,27,28]. Thus, results of the present study indicate that deletions can occur in all parts of the cagPAI and screening the entire genes in the cagPAI is needed to determine the accurate organization of the cagPAI region. For comparison with our results, we reviewed only studies that screened all the cagPAI genes. A previous study observed complete cagPAI present in 82.6% of the strains, while a partially deleted cagPAI in 9.6% of the strains and 7.7% lacked the entire cagPAI in Indian population [11]. In Swedish population, 76% of the strains carried an intact cagPAI, 15% had partially deleted cagPAI and the cagPAI was lacked in 9% of the strains [10]. A study by Azuma et al., [29] showed that the complete cagPAI was identified in all 11  and cag21 were present in all isolates. These genes might represent core genes of the cagPAI region, however function of the cag1, cag6 and cag21 are still unknown [19]. cag8 (HP0528, cagX) is a component of T4SS (VirB9) encodes a membrane protein [19]. One The presence of specific genes in H. pylori isolates associated with different ethnicities (cag4 in the Chinese isolates and cag2, cag14 and cag24 in the non-Chinese isolates) might represent strain associated disease outcomes. The cagA (VirB1) is a component of T4SS, whereas the function is still unknown for cag2, cag14 and cag24 [19]. Although the difference was not statistically significant, high frequency of cag2 was detected in gastric mucosa with CAG and IM/Atr and reflects the presence of this gene in non-Chinese isolates. These observations require further investigation to decipher the role of these genes.
We found an association of cagPAI intactness with histopathological scores of the gastric mucosa. H. pylori harbouring partial cagPAI were associated with higher density of H. pylori and neutrophil activity, whereas H. pylori with deleted cagPAI causes increased in inflammatory score. The presence of neutrophil activity in the gastric mucosa is associated with CAG and this has been shown in our study that partial cagPAI H. pylori strains was detected more in CAG groups. As strains with deleted cagPAI only cause inflammation of the gastric mucosa, the presence of cagPAI proteins encoded by H. pylori strains is needed to cause more severe disease such as active gastritis and intestinal metaplasia. However, no specific gene could be identified that causes severe condition. A group of genes encoded T4SS and for induction of IL-8 secretion have been shown to involve in the process of disease development [7,21].

Conclusions
Results of the present study show that cagPAI organisation is diverse in isolates from different ethnicities. Comprehensive screening of the entire cagPAI genes provides a more accurate overview of the H. pylori cagPAI genotype and allows better identification of the virulence traits of the organisms in our multiracial population. H. pylori strains harbouring partial/rearrangement of the cagPAI genes associated with increased colonization and recruitment of neutrophil at the site of infection and further contribute to various disease outcomes caused by different genotypes of H. pylori strains.

Determination of cagPAI genes
The presence or absence of cagPAI in H. pylori strains was determined by PCR using primers for detection of the 5' and 3' flanking region of the cagPAI as described by Olbermann et al., [19]. The amplifications were carried out in 25 µl volume, each containing 12.5 µl mastermix (Lucigen, USA), 10 µl of each primers, 1 µl (25 ng) DNA and cagPAI region consisted of initial denaturation at 95°C for 3 min, followed by 30 cycles of 95° for 30 s, 50°C for 60 s, and 72°C for 45 s, ending with final extension at 72°C for 5 min. The amplifications were performed in a PCR thermal cycler T100 Series (Bio-Rad, USA). The products were run on 1.5% agarose gel and stained with FloroSafe DNA stain (1 st BASE Pte. Ltd, Singapore) and visualised with gel documentation (AlphaImager, Biosciences, CA). The cagPAI-positive isolates (n=91) were then subjected to subsequent PCRs for identification of all cagPAI genes using primers as described previously [19,20].
The deletion of HP0521 gene were confirmed using HP0521 empty site (ES) primer pair as described previously [21]. PCR amplification for cagPAI genes consisted of initial denaturation at 95°C for 3 min, followed by 30 cycles of 95° for 30s, annealing temperature for 60s (48C for cag11, 48.8C for cag3 and 55C for cag1, cag2, cag4, cag5, cag, cag6 to cag10, cag12 to cag26), and extension at 72°C for 45 s. A final extension at 72°C for 5 min was performed for each PCR run. Representative positive PCR products (n=28) were sent for sequencing and the nucleotide sequences were blasted against NCBI databases to confirm the gene identity.

Statistical analysis
Statistical analysis was performed using SPSS software version 23 (SPSS Inc, Chicago, IL, USA). Differences between groups were evaluated using Chi-square (χ 2 ) test, Yate's continuity correction and Fisher's exact probability test. Independent t-test was used to compared means between different groups of histopathological scores. Score was represented with mean standard error of mean (SE). Differences were considered significant when p value was <0.05.

Acknowledgments
We would like to thank to the Universiti Kebangsaan Malaysia for providing both the