Prevalence of the Helicobacter pylori babA2 Gene in Children Mainly Depends on the PCR Primer Set Used

Various polymerase chain reaction- (PCR-) based methods with varying positivity rates were designed to detect the Helicobacter pylori babA2 gene. To compare different primer sets, babA2 prevalence was determined in 279 H. pylori-positive pediatric samples using the 832 bp, 139 bp, and 271 bp PCR primer sets, resulting in 34.0%, 51.3%, and 79.6% prevalence of the babA2 gene, respectively. The babA2 status determined using the 832 bp and 139 bp PCR primer sets significantly correlated with bacterial density and activity of inflammation, whereas no such correlations were found using the 271 bp PCR primer set. The 139 and 832 bp PCR primer sets concordantly detected the babA2 gene in 93 cases; however, in comparison to the 832 bp PCR primer set, the 139 bp PCR primer set detected additional 50 babA2 cases, whereas only two 832 bp positive cases were missed. The 271 bp PCR primer set missed 32 babA2 cases that were 832 bp and/or 139 bp PCR positive, but tested solely positive in 109 cases. Interestingly, cloning of a subset of 271 bp PCR positive samples revealed amplification of the babA/B gene chimera. Hence, in our opinion, the 271 bp PCR protocol is not a reliable diagnostic tool for detecting the babA2 gene in children. Our results reaffirm previous observations that the use of certain babA2 PCR primer sets can significantly impact estimation of the prevalence and clinical relevance of the H. pylori babA2 gene in children, suggesting babA2 detection methods should be carefully selected.


Introduction
Adherence factors of Helicobacter pylori are considered one of the most important virulence factors, which enable longterm persistence of the bacteria in the human stomach [1,2]. H. pylori genome encodes a variety of outer membrane proteins (OMPs), including blood group antigen binding adhesion (BabA) proteins. ree bab paralogues have been described thus far: BabA, BabB, and BabC [3]. BabA is one of the most studied H. pylori adhesins, capable of binding to Lewis b (Le b ) and related ABO antigens on gastric epithelial cells [1], which in turn results in increased pathogenicity of H. pylori and may play a crucial role in the development of H. pylori related gastric pathology such as severe gastritis, peptic ulcers, and gastric adenocarcinoma [4][5][6][7]. In contrast, the role of BabB and BabC as adhesins has not been demonstrated yet.
e BabA protein is encoded by the babA2 gene, whereas the babA1 gene harbors a 10 bp deletion, resulting in protein's inability to interact with Le b antigens [1]. Interestingly, it has been shown that the presence of the babA2 gene may not uniformly reflect the functional status of the gene [3,8]. Based on the level of BabA protein production, strains can be divided into BabA high producers (Le b binding activity, contain babA2 gene), BabA low producers (lacking Le b binding activity but contain babA2 gene), and BabA nonproducers (lacking Le b binding activity, no detectable babA2 gene) [8].
babA and babB exhibit substantial N-and C-terminal identity [1,9]. Recombination events between babA and babB appear to be relatively common and have previously been demonstrated for both East Asian and Western strains [10]. Moreover, these intragenomic translocations between the babA and babB genes are thought to be the main mechanism of BabA expression regulation [3]. Replacement of first 56 bp in the 5′ region of the babB with babA gene (e.g., babA/babB chimera) was first identified in clinical isolates [10] and later also reported during experimental H. pylori infection in Rhesus macaques [11]. Somewhat less common are babB/A chimera. An analysis of the chimeric babB/A gene revealed that the first 47 bp were specific for babB, followed by a 66 bp fragment that was shared between both babA and babB, while the 3′ region showed higher sequence similarity with babA. Interestingly, this chimeric BabB/A protein was shown to have the Le b binding capacity [12]. e plasticity of OMPs, not only across different strains but also within a strain colonizing an individual patient, can be regarded as an evolutionary asset, enabling better adaptation of H. pylori to the harsh milieu of the human stomach [13].
Various polymerase chain reaction-(PCR-) based methods have been used to detect the babA2 gene [8]. e majority of studies evaluating the BabA status used PCR primers, designed by Gerhard et al. [4], which aim to detect the 10 bp deletion that distinguishes between the babA1 and babA2 genes. ese primers target the highly polymorphic zone of the babA2 gene, amplifying an 832 bp long fragment. Another commonly used PCR primer set was designed by Sheu et al., targeting the promoter region and thus amplifying a 271 bp fragment of the babA2 gene [14].
In our previous study, we evaluated the babA2 gene prevalence and clinical relevance in Slovenian children using newly developed primers designed by our research group [15]. Interestingly, while additionally testing a subset of samples with the 832 bp PCR primer set, we noticed considerable differences in the prevalence of the babA2 gene depending on the type of PCR primer set used [15]. Hence, in order to determine whether the use of a specific PCR primer set influences the babA2 gene prevalence and clinical relevance in a pediatric population, we performed a head-to-head comparison of babA2-specific PCR primer set targeting the 139 bp of the babA2 gene designed by our research group [15] with two widely used babA2 PCR primer sets targeting 832 bp [4] and 271 bp segments [14] of the babA2 gene.

Patients.
e study protocol has been described in detail previously [15][16][17]. Briefly, gastric biopsy samples were obtained from 279 consecutive H. pylori-positive children (107 male, 172 female; age range, 5 months-18 years) who had undergone upper endoscopy in the University Children's Hospital in Ljubljana, Slovenia, during the period 1999-2011. Exclusion criteria included previous antibiotic treatment of H. pylori infection and usage of nonsteroidal anti-inflammatory drugs, proton pump inhibitors, or H2 receptor blockers within 30 days prior to the endoscopic examination.

Histology.
At endoscopy, four biopsy samples were obtained from the gastric antrum and corpus from each patient. Two biopsy samples were used for rapid urease testing (H. pylori Quick Test; Biohit Diagnostics, Helsinki, Finland) and the rest were subjected to histopathological examination [16,17]. According to the updated Sydney classification, two experienced pathologists evaluated the following histological features: density of H. pylori colonization, activity of inflammation, chronic inflammation, atrophy, and intestinal metaplasia [18].

H. pylori DNA Extraction and babA2
Genotyping. As described in detail previously [16,17], H. pylori DNA was extracted from the biopsy samples used for the rapid urease testing using the QIAamp DNA Mini kit (Qiagen GmbH, Hilden, Germany). e detection of the babA2 gene was performed in all H. pylori samples using a PCR primer set targeting a 139 bp fragment of the central region of babA2 that was designed by our research group [15] and two widely used PCRs, targeting 832 bp [4] and 271 bp [14] fragments of a 5′ part of babA2 (Table 1). e 139 bp babA2 PCR protocol was performed using a HotStarTaq Plus DNA Polymerase kit (Qiagen, Hilden, Germany), as described previously [15]. e 832 and 271 bp PCR protocols were performed using AmpliTaq Gold DNA Polymerase kit (Applied Biosystems, Foster City, CA, USA) and FastStart Taq DNA Polymerase kit (Roche, Basel, Switzerland) as described previously [4,14]. e resulting 139 bp and 832 bp amplicons were analyzed on a 2% agarose gel and only a subset of amplicons (15 PCR products per each PCR protocol) was sequenced with the same primers as those used for initial PCR [19] because both PCR protocols have already been extensively evaluated in previous studies [4,15].
Sequencing of 13 randomly selected solely 271 bp PCR positive samples and two samples that tested positive using all three PCRs was attempted; however, we were not able to obtain the respective sequences due to technical difficulties. Hence, these samples were purified using QIAquick PCR Purification kit (Qiagen) and subsequently cloned, using CloneJET PCR Cloning Kit ( ermo Fisher Scientific, Vilnius, Lithuania). A sticky-end cloning protocol was used and the overnight transformation was performed using One Shot ® TOP10 Chemically Competent E. coli (Invitrogen, Carlsbad, USA), according to the manufacturer's instructions. Recombinant plasmid clones from the transformed bacterial culture were verified for the corresponding H. pylori inserts with colony PCR and sequencing of PCR products performed using pJET1.2 sequencing primers (pJET1.2 F 5′-CGACTCACTATAGGGAGAGCGGC-3′ and pJET1.2 R 5′-AAGAACATCGATTTTCCATGGCAG-3′), according to the manufacturer's instructions. e obtained sequences were annotated by BLAST nucleotide search against the current NCBI database. e sequences were aligned and edited using BioEdit ClustalW multiple sequence alignment together with babA2 and babB reference sequences obtained from publically available NCBI database. Phylogenetic analysis was performed using PhyML (v3.0) integrated in the Phylogeny.fr platform, including HKY85 as a substitution model with 4 gamma categories.
e proportion of invariable sites and gamma shape parameter were estimated directly from the data. e obtained phylogenetic tree was graphically presented with Figtree version 1.4.3 (http://tree.bio.ed.ac.uk/software/figtree/) and phylogenetic relationships were assessed according to the approximate likelihood ratio test (aLRT) branch support values [20].
In addition, because some of the sequences obtained exhibited mismatch with the babA2 reference sequences, amplification of babA/B chimera was suspected. e presence of chimeric babA/B gene was studied by performing bootscanning method available in Simplot, version 3.5.1., with 500 bootstrap replicates on a varying window sizes (120-60 bp) with a 10 bp sliding window [21]. Two separate slices of alignments were generated and phylogenetic analysis was performed, as described above.

Statistical Analysis.
e data were analyzed using the SPSS 11.0 statistical package (SPSS GmbH Software, Munich, Germany). Statistical significance of differences between evaluated PCR primer sets was evaluated using the t-test and Chi square test with a p value of less than 0.05 considered as statistically significant.
As shown in Figure 1, the 139 bp and 832 bp babA2 PCR primer sets concordantly detected babA2 in 93 cases; however, in comparison to the 832 bp PCR primer set, the 139 bp PCR primer set additionally detected 50 babA2positive samples while missing two samples that were babA2 positive according to the 832 bp PCR primer set. In total, 109 samples were concordantly negative using the 832 bp and 139 bp PCR primer sets but tested positive using the 271 bp PCR primer set. Furthermore, the 271 bp PCR primer set missed 32 H. pylori samples that tested positive using the 139 bp and/or 832 bp PCR primer sets.

Association of babA2 Genotypes with Histological Parameters in Gastric Mucosa.
In order to evaluate to what extent a certain babA2 PCR protocol affects result interpretation on the clinical relevance of the babA2 gene, we assessed potential correlations between the presence of the babA2 gene detected by three different PCR protocols and five histological parameters. As shown in Table 2, H. pylori strains determined as babA2 positive by the 139 bp PCR protocol were strongly associated with two out of five histological parameters studied: H. pylori density score (p � 0.007) and activity of inflammation (p � 0.003), while the association with chronic inflammation of the gastric mucosa was borderline (p � 0.063). Similarly, when using the 832 bp PCR primer set, a strong association was observed between the H. pylori strains determined as babA2 positive and H. pylori density score (p � 0.015) and activity of inflammation (p � 0.016) ( Table 2). In contrast, babA2 positive strains confirmed by the 271 bp PCR showed no statistically significant associations with any of the five histological parameters evaluated ( Table 2).   e obtained phylogenetic tree of the first section of the alignment displayed phylogenetic clustering of all 15 generated sequences with the babA2 reference sequences, although with less significant aLRT value (0.787) (Figure 4). On the other hand, the second part of the alignment displayed the same phylogenetic relationship, as did the previous phylogenetic analysis performed on the complete alignment ( Figure 2). erefore, our results indicate that the 271 bp primers set amplified chimeric babA/B gene.

Discussion
In this study, we performed a head-to-head comparison of babA2 gene-specific PCR primer set designed by our research group [15] with two previously established babA2 PCR protocols in order to assess the babA2 prevalence using the largest collection of H. pylori positive biopsies obtained from children (n � 279). Interestingly, the prevalence of the babA2 gene in our study varied significantly according to the PCR primer set used; the babA2 positivity rate using the 271 bp PCR primer set was more than double of that obtained by the 832 bp PCR primer set (79.6% versus 34.0%), whereas the   In majority of previous studies in adults and children [4,6,[22][23][24][25][26][27][28][29][30], the babA2 status was determined using the 832 bp PCR primer set, originally designed by Gerhard et al. [4]. According to the literature data, the prevalence of the babA2 gene in adults determined by using these primers is highly variable, ranging from 32 to 72% in Western countries and from 64 to 85% in Asian countries [8], while in children, the babA2 gene positivity ranges from 17 to 84% [22][23][24][25][26][27][28][29][30]. Sheu et al. [14] designed a second widely used PCR primer set, which amplifies a 271 bp fragment of the babA2 gene. In comparison to the 832 bp PCR primer set, the babA2 prevalence determined using the 271 bp PCR primer set in adults was generally higher, ranging from 41 to 100% [7,14,31,32]. To the best of our knowledge, only one previous study used the 271 bp PCR primer set for detecting the babA2 gene in children, yielding a 41% prevalence among Turkish pediatric population [33]. Adherence of H. pylori to gastric epithelial cells is one of the most important contributing factors to the pathogenicity of the bacteria [34]. e babA2 gene encodes a protein that enables H. pylori attachment to Le b epitopes on human epithelial cells and thus the delivery of bacterial virulence factors into host target cells, resulting in gastric tissue damage [35]. In this study, we also aimed to evaluate whether the use of certain babA2 PCR primer set influences the clinical relevance of the babA2 gene. Interestingly, we found a strong correlation between babA2 positive status and the density of H. pylori colonization, as well as the degree of active inflammation determined by the 139 and 832 bp PCR primer sets (Table 2). However, no significant associations were found between the babA2 status and H. pylori density score, degree of activity, or chronic inflammation of the gastric mucosa when using the 271 bp PCR primer sets. In this study, babA2 positive strains (determined with any of the three PCR methods used) were not found more frequently in patients with gastric atrophy or intestinal metaplasia, suggesting involvement of other virulence or host factors in the disease progression. Nevertheless, data regarding the babA2 gene prevalence and clinical significance should be interpreted with caution, since the use of a particular babA2 PCR primer set results in significant differences in the babA2 gene detection rates. Moreover, conflicting results on the association between the presence of the babA2 gene and gastric cancer obtained in previous studies could be due to using only a single PCR primer set for detection of the babA2 gene, an approach that may not reflect the actual status of the BabA protein expression [35]. Chang et al. [36] even proposed using multiple pairs of PCR primers to provide a more reliable estimation of the BabA status and its influence on the risk of gastric cancer, especially in isolates from East Asia with nearly 100% prevalence of the babA2 gene [14,36]. e babA gene was initially cloned from H. pylori strain CCUG17875, which contains a silent (nonfunctional) babA1 gene and an expressed babA2 gene. e sequences of these two genes differ only by the presence of a 10 bp deletion in the signal peptide sequence of babA1, which eliminates its translational initiation codon. However, later analysis showed that naturally occurring babA1 sequences are very rare, suggesting PCR-based methods designed to detect the 10 bp deletion (e.g., the 832 bp PCR primer set) do not reliably reflect the BabA status [8,13]. As shown in Figure 1, the 139 bp PCR primer set detected 50 additional babA2 positive cases compared to the widely used 832 bp PCR primer set, suggesting superior sensitivity of the 139 bp PCR primer set for detecting the babA2 gene. us, our data support previous observations of ineffectiveness and underestimation of the babA2 gene prevalence when using the 832 bp PCR primer set [6,8,22]. Although the forward babA2F primer is located within a relatively conserved 5′ region, the presence of single nucleotide polymorphisms may nonetheless impede effective amplification. Moreover, because the reverse babA2R primer was designed to anneal within a highly variable region in order to reliably discriminate between various bab homologues, this may invariably lead to underestimation of the babA2 prevalence [6]. According to Fujimoto et al. [8], the sensitivity of the babA2 PCR primer sets that were designed to discriminate between the babA2 and babA1 genes was shown to be only approximately 36-49%, thus missing a significant proportion of the BabA high and low producers. Nevertheless, significantly higher prevalence of the babA2 gene (e.g., >70%) obtained from Asian isolates using the same primer set highlights the significance of geographical variability among H. pylori strains [6,8,22]. e issue of nucleotide mismatching in the primer binding region of the 832 bp PCR primer set was circumvented when designing primers for the 139 bp PCR protocol by systematic investigation of the babA2 genomic diversity of geographically distinct H. pylori samples and the consequent addition of more babA2 specific primers to cover all variations of the babA2 gene known at the time of primer design. In specific, three forward and one reverse primer were designed on the basis of multiple alignments of 94 babA2 and 24 babB sequences [15]. To avoid amplifying nearly identical 5′ and 3′ regions of the babA and its paralogue babB, a set of three forward and one reverse primer was selected in the central region of the babA2 gene which target conserved regions of the babA2, yielding a 139 bp babA2 gene-specific amplicon. However, despite our best efforts to design a PCR primer set that would amplify all babA2 gene variants, the 139 bp PCR primer set failed to detect two babA2 positive cases determined by the 832 bp PCR primer set, suggesting addition of more primers or use of a combined approach (e.g., concurrent amplification with the 139 and 832 bp PCR primer sets) may be used in order to increase the sensitivity of the PCR. Nevertheless, as shown by Pride et al. [37], sequencing of the babA and babB genes from H. pylori strains from around the world showed high sequence variability in different strains, thus making it difficult to construct a universally applicable primer set. e 271 bp PCR primer set uses a forward bab7-F primer that is located in the promoter region of the babA2 gene in combination with a unique reverse primer from the 5′ region of the babA2 gene. A study by Fujimoto et al. [8] showed that the 271 bp PCR primer set originally developed by Sheu et al. [14] is not a reliable molecular tool for detecting the babA2 gene in adults, because it frequently amplifies also the babB gene sequences and, consequently, produces babA2 falsepositive results. Interestingly, the reverse bab7-R primer sequence matches exactly both the babA2 and babB gene sequences, which may be the reason for the occasional nonspecific babB gene amplification. Our study showed that the inadequacy of the 271 bp PCR for detecting the babA2 gene also exists in pediatric population. In this study, Sheu's primers generated not only the predicted 271 bp amplicons, but also amplicons of various sizes, ranging from approximately 240 to 300 bp, which was already noted previously [32]. Unlike the study by Mattar et al. [32], in which the two randomly sequenced PCR amplicons were babA2 in origin, 6/15 cloned sequences from our study were shown to belong to the babB cluster, whereas the rest of the sequences showed higher similarity to the babA2 gene sequences. As shown in Figure 2, significant aLRT branch support values were observed, thus clearly demonstrating clustering of the strains into two distinct groups: the babA2 and babB gene group. To our surprise, sequencing of one out of two cloned 271 bp amplicons that tested positive using all three PCR protocols yielded babB-like sequences. Indeed, further analysis revealed that the 271 bp PCR primer set in fact detected babA/B gene chimeras (Figure 3), which appear to be relatively common among bacterial isolates [38]. It has previously been shown that, in isolates lacking the babA gene (11/34; 33%), the chimeric babA/B gene which results in loss of Le b activity due to inhibition of the BabA expression predominates chimeric babB/A gene that otherwise subjects protein expression to phase variation [38]. Furthermore, 32 H. pylori samples that were positive with the 139 bp and/or 832 bp PCR primer sets tested negative with the 271 bp PCR primer set, indicating that this PCR protocol may also generate babA2 false-negative results (Figure 1). is may be due to significant variations in babA2 gene sequences among the isolates.
To the best of our knowledge, no PCR-based method can be considered as "gold standard" for detecting the babA2 gene. Moreover, PCR-based determination of the functional status of the gene is importantly hampered by several factors, namely varying sensitivity and specificity of different babA2 PCR primer sets, also reflecting a high degree of heterogeneity among different H. pylori isolates, as well as the presence of BabA low producers: H. pylori strains that in spite of detectable babA2 gene lack Le b binding activity [3,8]. Despite being relatively inexpensive and, most importantly, easy to perform and to interpret, PCR-based methods will probably be soon replaced by whole-genome sequencing. It has already been shown that whole-genome sequencing can be used for precise and fast characterization of H. pylori virulence genes, including babA2 [39].
ere are some limitations to this study, including the relatively low number of sequenced 271 bp PCR amplicons. In addition, determining which bab gene occupies which genomic locus would also provide additional information regarding the usefulness of respective babA2 PCR primer sets. Because PCRs were performed using tissue biopsies and not H. pylori isolates, it is likely that H. pylori subpopulations Canadian Journal of Infectious Diseases and Medical Microbiology were not adequately evaluated. Hence, future studies evaluating these PCR primer sets should be performed on multiple H. pylori isolates obtained from various gastric biopsies.

Conclusions
is study showed that the estimation of the prevalence and clinical relevance of the H. pylori babA2 gene in children mainly depends on the PCR primer set used. e 139 bp PCR primer set exhibited superior sensitivity than the previously most frequently used PCR primer set targeting 832 bp fragment of the babA2 gene. Similar to the 832 bp PCR primer set, the babA2 status determined by the 139 bp PCR primer set statistically significantly correlated with two out of five histological parameters evaluated, whereas such correlation was not evident when the 271 bp PCR primer set was used. Comparison of three different babA2 PCR primer sets on the largest number of pediatric H. pylori samples to date confirmed previous observations that the 271 bp PCR protocol is a not reliable diagnostic tool for the detection of the babA2 gene and should thus be avoided.

Abbreviations
PCR: Polymerase chain reaction BabA: Blood group antigen binding adhesion Le b : Lewis b aLRT: Approximate likelihood ratio test bp: Base pairs.

Data Availability
e data used to support the findings of this study are included within the article.
Ethical Approval e study was approved by the National Medical Ethics Committee at the Slovenian Ministry of Health.

Consent
Informed consent was obtained from the parents of all the children.

Conflicts of Interest
e authors declare no conflicts of interest.