PCR-Based Detection and Genotyping of Helicobacter pylori in Endoscopic Biopsy Samples from Brazilian Patients

Helicobacter pylori (H. pylori) is considered the second most prevalent infection in man. A precise diagnosis is important for treating patients with the indicative gastrointestinal symptoms. The present study analyzes the effectiveness of a molecular biology method (PCR) comparing the results obtained with the histology and with the rapid urease tests. PCR was used in the detection and genotyping of the H. pylori urease-C gene and the patterns which were obtained from the patients studied. 141 biopsy samples from 131 patients were evaluated. 59 paraffin biopsies samples were positive for H. pylori according to the histological examination. Of those, 59/12 (20.3%) were amplified using PCR. Of the 82 samples from the fresh biopsies, 64 were positive for H. pylori according to the rapid urease test (78%); there was an agreement of 100% with PCR. Sixty positive H. pylori samples were genotyped (58 samples of fresh biopsies and 2 samples of paraffin biopsies) using two restriction enzymes. The patterns observed were analyzed with the computational program BIO 1D; 11 patterns with the enzyme HhaI and 12 patterns with the enzyme MboI were found. However, it was not possible to find a statistically significant correlation between the specific genotypes and digestive pathologies. Accordingly, future research should be performed to confirm a statistically significant relationship between genotyping and gastrointestinal symptoms.


Introduction
e Helicobacter pylori (H. pylori) infection is currently endemic worldwide with high prevalence (up to 60%) in developing regions such as South America. e infection causes chronic Gastritis, gastric and duodenal ulcers, gastric adenocarcinoma, and mucosa-associated lymphoid tissue [1][2][3][4][5][6]. H. pylori is associated with several autoimmune diseases, including idiopathic thrombocytopenic purpura (ITP), Sjögren syndrome, systemic sclerosis [7], Graves' disease [8], and autoimmune pancreatitis [9]. As a result of this association with autoimmune diseases, we hypothesized that H. pylori might induce systemic immunological changes. Although the seroprevalence of H. pylori may be high in the normal population, a minority develops peptic ulcers [10,11]. Some possibilities could justify this data: genetic differences in the host's environmental factors and bacterial strains.
A variety of tests are now available to diagnose H. pylori infection. Histological examination of gastric tissue, bacterial cultures, rapid urease test, use of DNA probes, and PCR analysis, when used to test gastric tissue, all require endoscopy. In contrast, breath tests, serology, gastric juice PCR, and urinary excretion of N 15 ammonia are noninvasive tests that do not require endoscopy. PCR offers high sensitivity and speci�city as a technique for the detection of H. pylori although the accuracy of such techniques varies widely [12]. e aim of this work is to analyze the effectiveness of the molecular biology method PCR in the detection of H. pylori in patients with pylori were polymerase chain reaction (PCR) and PCR-RFLP for genotyping. ey were chosen in order to detect the bacterium and its subtypes in endoscopic biopsies of fresh tissues and paraffin tissues. e fresh biopsy samples were conserved in physiologic serum 0.9% until the DNA was extracted. At least two, 5 to 10 mm, ribbons of paraffin were collected from the paraffin tissue. In the fresh biopsies, at least two fragments were collected.

DNA Extraction-Gastric Paraffin
Biopsy. DNA extraction from the endoscopic biopsies fastened in paraffin followed the method described by Davis et al., 1995 [13], with some modi�cations. At least two, 5 to 10 mm, ribbons of paraffin were placed in a 1.5 mL Eppendorf tube. One mL of xylene was added to the samples. ey were shaken, allowed to rest for 3 to 5 minutes, and then centrifuged for 5 minutes, discarding the xylene aerwards. Aer three washes in 100%, 95%, and 70% ethanol, respectively, the samples were dried at room temperature. Next, the material was resuspended in a solution of Proteinase K, 50 mM Tris, 0.5% SDS, and sterile water. 430 L of phenol were added to the sample, which was homogenized and centrifuged for another 30 minutes at 14,000 RPM. e supernatant containing DNA was transferred to a new tube and 430 L of phenol/chloroform (1 : 1) were added and centrifuged again for 5 minutes at 14,000 RPM. Chloroform/isoamyl ethanol was added (24 : 1) to the supernatant, which was homogenized and centrifuged for another 30 minutes at 14,000 RPM. Aer the addition of 75 L of ammonium acetate and 750 L of 100%, ethanol samples were inverted several times and incubated overnight to −20 ∘ C. Aer centrifugation for 30 minutes at 12,000 RPM to −4 ∘ C, the supernatant was discarded. e precipitate was carefully washed with 500 L of chilled 70% ethanol, which was immediately discarded. e material was dried at room temperature and resuspended in a solution containing 50 mL of sterile water, 10 M of Tris (pH 8.0), and 1 mL of EDTA and stored at −20 ∘ C until its use.

DNA Extraction: Fresh
Biopsy. Firstly, a fresh 3 to 7 mm biopsy section was placed in a 1.5 mL sterile tube with 190 L of a solution that contained 0.1 M of Tris HCl (pH 7.5) and 1% of SDS. Secondly, 10 L of proteinase K were added (10 mg/mL) to the solution. e sample was macerated and incubated overnight at 55 ∘ C. Aer that, 200 L of phenol and 200 L of both chloroform and isoamyl alcohol (24 : 1) were added. e solution was then homogenized and centrifuged for one minute. Next, the supernatant was removed, and 200 L of chloroform/isoamyl alcohol were added, homogenized, and centrifuged for 1 minute. Next, the supernatant was removed again, and 25 L of sodium acetate 3 M and 900 L of 100% ethanol at −20 ∘ C were added; aer vortexing the mixture, it was incubated for 30 minutes at −70 ∘ C. e samples were centrifuged for 15 minutes at 15,000 RPM. e supernatant was discarded. Lastly, the DNA was resuspended in 25 L of distilled and sterile water. [14].
For each ampli�cation reaction, 0.5 to 0.7 L of the DNA under investigation were used, for a total reaction volume of 20.0 L. e reaction buffer contained 50 mM KCL, 10 mM Tris-HCL pH 8.4, 2.5 mM MgCL 2 , 2.0 pmol of each "primer, " 200 M of each deoxynucleotide triphosphate (dATP, dCTP, dGTP, and dTTP), 2.5 units of Taq polymerase (Gibco-BRL), and sufficient water to give the total volume of 20.0 L. e reaction mixture was covered with 100 L of mineral oil and the tubes were placed in a DNA ermal Cycler (Perkin-Elmer).
e reactions that followed were found to be optimal. e samples were heated to 94 ∘ C for 60 s to denature the DNA, cooled to 57 ∘ C for 90 s to allow the primers and the DNA to reanneal, and then heated to 72 ∘ C for 120 s for primer extension. By the �nal cycle, the extension period was 7 min. A total of 40 cycles were performed. e ampli�ed product was detected by direct gel analysis. 5 L of the reaction mixture were subjected to electrophoresis with 2% agarose minigel, and the DNA was visualized using UV �uorescence aer staining with ethidium bromide. Molecular F 1: Automated sequencing (Abi Prism 377) for the ureC region of H. pylori. Positive samples for H. pylori infection obtained from PCR in order to prove that the sequences being ampli�ed did belong to the genetic sequence of a DNA segment of the urease-C area of H. pylori (enzymes HhaI and MboI). All DNA samples taken from the patients presented genetic sequences similar to the one of H. pylori, as described for the urease-C area of the Gene Bank (I Square 1).
weight markers were included in each gel. An 820 base-pair band was seen when samples were ampli�ed using primers P1 and P2 to detect H. pylori (Table 1 and Figure 1).

PCR-Based Restriction Fragment Length Polymorphism
Typing of Helicobacter pylori (RFLP). A�er the ampli�cation was con�rmed, the PCR product was submitted to digestion with the restriction enzymes HhaI and MboI for fragmentation of Urease-C [16].
e fragments which were produced were submitted to electrophoresis in a 2% gel agarose 1000 (Gibco-BRL), stained with ethidium bromide, visualized under ultraviolet light  (Figures 2 and 3). Approximately 10 L of the ampli�ed product were used for the digestion process which also contained 2.0 L of the corresponding enzyme. �ater was added to �ll 20.0 L and the mixture was placed in a 37 ∘ C bath overnight.

Automatic Sequencing.
Automatic sequencing was performed using the program Abi Prism, model 377, version 3.4, and Abi 100, version 3.2. Sequencing allowed for the identi�cation of the studied DNA region (Primers P1 and P2). Figure 1 shows the automatic sequencing, proving that the sequence is Helicobacter pylori.

Results
A total of 141 endoscopic biopsy samples from 131 patients were studied for H. pylori infections with PCR and the results were compared with Urease and Histology tests. 82/64 (78%) fresh samples had a positive Urease test for H. pylori. A PCR test detected all of the 64 positive samples identi�ed by the Urease test (100%) ( Table 2).
Fiy-nine paraffin biopsies, all found to be positive through a histological examination, were submitted to the DNA extraction procedure and Beta-Globin PCR to prove the quality and the presence of DNA in the extracted samples. Only 14/59 (23.7%) samples were positive for the Beta Globin gene, but in two of them H. pylori was not ampli�ed by PCR, even though they had a positive Histology test (Table 3). In the other 45 samples, it was impossible to detect Beta Globin in the DNA using PCR, primarily because of the low amount of paraffin samples and/or because the reaction was inhibited due to paraffin and xilol in the extraction procedures. No contamination occurred and the samples were tested two times.
Among the 141 fresh endoscopic biopsy samples, 58 were tested using the RFLP technique to detect the different H. pylori strainswith the restriction enzymes HhaI and MboI.

Discussion
Gastric cancer is one of neoplasms that cause the majority of deaths not only in Brazil but all over the world. e type of cancer caused by H. pylori could be linked to gastric chronic. Differences in the degree of virulence between strains have lead to an increased risk of developing gastric diseases [17]. e H. pylori infection is distributed in a cosmopolitan way, reaching mainly the adult population of low socioeconomic levels in developing countries. e discharge infection rate is correlated with bacterial virulence and inherent factors of the particular host, mainly with respect to the immune system [18].
It should be noticed that the route of fecal-oral transmission appears to be the biggest problem in the prevalence of infection, making H. pylori a serious public health problem in both developed and developing countries [19].
e present study analyzes the effectiveness of the molecular biology method PCR in the detection of H. pylori in patients with gastrointestinal symptoms, comparing it with the histology and rapid urease test.
e polymerase chain reaction (PCR) for the diagnosis of H. pylori is a very sensitive and speci�c method [20], providing fast and safe diagnosis. Many results indicate that PCR sensitivity is close to that of culture tests [21], but for verifying the eradication of H. pylori the effectiveness of PCR can be markedly superior [22,23]. e methodology used in other studies to distinguish the different H. pylori subtypes has been PCR-RFLP [24] that through analyzis of the PCR product with restriction endonucleases that resulting fragments of different sizes and the digestion pro�le is decisive to de�ne the strains. e restriction enzymes HhaI and MboI were used for the Urease-C area [16]. e extreme degree of variability observed among the strains of H. pylori became an important focus of scienti�c attention, as the investigators recognized the signi�cant impact that this phenomenon can have on several research areas, such as the development of vaccines, the development of resistance to antimicrobial agents, and the study of the pathogen-host interaction [25,26]. Considering that 10 to 20% of people infected with H. pylori develop obvious diseases, the reliable identi�cation of the lineages could actually be very bene�cial [27]. Previous studies that have used several techniques characterized H. pylori as a highly variable species that presents countless lineages, each one with its own and different genotype [28][29][30][31].
e genotyping of H. pylori is important for characterizing the most pathogenic genotype and the most frequent strain. is information can be used for clinical and epidemic studies. Even if many infections are clinically silent, the organism infected with H. pylori presents increased morbidity and mortality [5,32,33].
In the present study we standardized PCR with material obtained from the fresh endoscopic biopsies samples of patients attending Gastrocentro (the Center of Digestive Tract Studies), Medical School, UNICAMP. Some of the gastric biopsy samples were collected in paraffin and some were not. With regard to the standardization of the DNA extraction technique from the paraffin biopsy samples, several difficulties were found, because the samples contained a small amount of tissue fragments and many of the paraffin samples did not amplify the -Globin gene, demonstrating degraded DNA of poor or inhibited quality.
PCR was used because it is more speci�c and faster when compared to other methods; the product of PCR can be processed with restriction enzymes to verify H. Pylori strains. Besides, starting with the PCR, DNA sequencing can be made to verify mutations, which no other technique is capable of doing.
As an internal control of the reaction was used in all samples (human -Globin gene), in the fresh-air biopsy samples positive for H. pylori, we had 100% PCR ampli�cation. However, in several paraffin samples, the -Globin did not amplify, indicating an inefficient DNA extraction of the samples.
e efficiency of H. pylori detection PCR in fresh samples was superior to that in the paraffin samples. We suspect that PCR inhibition may have occurred due to the method used in DNA extraction from paraffin or the fact that that the samples were insufficient.
e extraction of DNA from fresh samples had excellent results. Among the 82 analyzed samples, 64 were positive and 18 negative, with 100% in agreement with PCR.
In the present study, we used the PCR-RFLP method for the differentiation of H. pylori strains from specimens obtained from gastric biopsies taken from Brazilian patients. Using this methodology we observed that 12 and 11 patterns were produced, respectively, by the two restriction enzymes MboI and HhaI from 58 specimens obtained from gastric biopsies. Two were samples of biopsies in paraffin and 58 were samples of nonfastened gastric biopsies (Table 1).
is data suggests that genotyping using PCR-RFLP can be useful as a fast procedure for the speci�c identi�cation of H. pylori lineages in gastric biopsies specimens [16]. Several protocols of genotyping analysis were proposed for distinguishing the lineages of clinically isolated H. pylori [34][35][36][37][38]. Several primer pairs were described for detection and the typing of H. pylori was based on the ampli�cation of the ureA [34], ureA plus ureB [35], and ureC genes [36,38]. ese results demonstrate great diversity in the urease genes in clinical H. pylori samples. Li et al. [16] found 3, 11, and 6 different patterns which were produced by 19 clinically isolated samples, respectively, digested by the restriction enzymes HhaI, MboI and AluI. Foxall et al. [35] found 10 different patterns which were produced by 22 clinically isolated samples, when the restriction enzyme HaeIII digested the PCR product of 2.4 Kb which had been ampli�ed by the ureA and ureB genes. Lopez et al. [37] found that the patterns generated by the digestion of PCR products with the HaeIII enzyme, starting from ureA and ureB, were almost as different as the standard HaeIII. Akopyanz et al. [28] found 18 MboI and 27 HaeIII RFLP patterns, PCR products ampli�ed by ureA and ureB genes of 2.4 Kb of 60 H. pylori lineages, and that the patterns distinguished 44 separate groups. Each isolated group did not differ from the other ones in the RFLP analyses of ureA and ureB products, but differed in MboI digestion of the 1.7 Kb ureC and ureD segments. Such a fact indicates that PCR-RFLP analyses of ureC genes can produce a great number of standard RFLPs.
Several studies have con�rmed that PCR-RFLP analysis of the ureC gene can differentiate clinically isolated H. pylori. Using restriction endonucleases, Moore et al. [38] analyzed the 1.1 Kb portion of the ureC gene ampli�ed by the �PCR� of 21 clinically isolated H. pylori. e samples were divided into four groups aer digestion with the enzyme HindIII, while the lineages were divided into 15 groups aer they were digested with the enzymes AluI and PvuI. Fujimoto et al. [36] demonstrated that the digestion of 820 bp of the H. pylori ureC gene with the restriction enzymes HhaI, MboI, and MseI resulted in 10, 10, and 11 different patterns, respectively. Dooley et al. [3] used three types of enzymes in the PCR product of a 1.179 bp portion of the H. pylori ureC gene. Eleven, 10, and 6 digestion patterns were produced by the HhaI, MboI, and AluI enzymes, respectively.
In our study we used two types of restriction enzymes in the ampli�ed products of 820 pb of the H. pylori ureC gene. We obtained 11 and 12 different patterns, respectively, from the 58 clinically isolated samples which were studied. Our results suggest that the PCR-RFLP analysis of this portion of the H. pylori ureC gene is a reliable method, and that genotyping of PCR in this area can be used for epidemic studies and for the differentiation of isolated H. pylori strands.
is study also revealed a high level of genetic diversity isolated in the different H. pylori positive patients studied in Brazil.
e obtained genotyping patterns were compared using the computational program Bio 1D. We found 11 patterns with the HhaI enzyme and 12 patterns with the MboI enzyme. e reason for the small number of studied samples was due to the fact that it was not possible to establish a signi�cant statistical correlation between speci�c digestive pathologies and standard genotypes.
We believe that the genotyping of H. pylori can contribute to the study of the microorganism's characteristics, facilitating the detection of pathogenic or nonpathogenic strains and, in turn, providing a better understanding of the several virulence factors that the bacterium uses to cause diseases.

Statistics.
Percentage agreement was calculated to compare H. pylori genotypes obtained from PCR performed directly on gastric biopsies, with the genotypes obtained from the PCR of DNA extracted from paraffin and fresh samples, as well as histology and urease tests.

�on��ct of �nterests
Authors have no con�ict to declare.