- Email: [email protected]
Distribution of vacA genotypes in Helicobacter pylori strains isolated from Brazilian adult patients with gastritis, duodenal ulcer or gastric carcinoma
FEMS Immunology and Medical Microbiology 33 (2002) 173^178
www.fems-microbiology.org
Distribution of vacA genotypes in Helicobacter pylori strains isolated from Brazilian adult patients with gastritis, duodenal ulcer or gastric carcinoma Abdussalam A.R. Ashour a;d , Paula P. Magalha‹es b;e;f , Edilberto N. Mendes a; , Guilherme B. Collares a , Valqu|¤ria R. de Gusma‹o a , Dulciene M.M. Queiroz b , Ana Margarida M.F. Nogueira c , Gifone A. Rocha b , Celso A. de Oliveira b b
a Laboratory of Molecular Biology, Faculdade de Medicina, Universidade Federal de Minas Gerais, 30130-100 Belo Horizonte, Brazil Laboratory of Research in Bacteriology, Faculdade de Medicina, Universidade Federal de Minas Gerais, 30130-100 Belo Horizonte, Brazil c Department of Pathology, Faculdade de Medicina, Universidade Federal de Minas Gerais, 30130-100 Belo Horizonte, Brazil d Department of Microbiology, Escola Paulista de Medicina, Universidade Federal de Sa‹o Paulo, 04023-062 Sa‹o Paulo, Brazil e Instituto de Microbiologia Prof. Paulo de Go¤es, Universidade Federal do Rio de Janeiro, 21941-590 Rio de Janeiro, Brazil f Faculdade de Fisioterapia, Universidade de Itau¤na, 35680-033 Itau¤na, Brazil
Received 4 December 2001 ; received in revised form 27 February 2002; accepted 14 March 2002 First published online 20 May 2002
Abstract This PCR-based analysis is the first molecular epidemiological study in Brazil testing Helicobacter pylori cagA and vacA distribution in adults with gastric complaints, that includes a large number of carcinoma patients. Multiple-strain infection was identified in 11/13.4% patients. vacA s1-m1 and cagAþ genotypes were the most common in patients with a non-mixed infection. All vacA s1 strains were s1b, so subtyping s1 strains was not useful. vacA s1b-m1 and cagAþ strains were associated with higher prevalence of peptic ulcer and gastric carcinoma than vacA s2-m2 and cagA3 ones. In conclusion, cagA and vacA genotyping may have clinical relevance in Brazil. 8 2002 Federation of European Microbiological Societies. Published by Elsevier Science B.V. All rights reserved. Keywords : vacA; cagA; Bacterial genetic diversity ; Helicobacter pylori
1. Introduction Most Helicobacter pylori infections are clinically silent. Only around 20% of H. pylori infected adults manifest one of the four severe outcomes associated with the bacterium: gastric ulcer, duodenal ulcer, gastric carcinoma or lymphoma [1^3]. It is still unknown why the organism is able to produce severe disease in certain hosts and be innocuous in others. Clinical outcome of the infection is determined by several factors, such as di¡erences in host responses to bacterium stimulation, speci¢c organism virulence factors and environmental in£uences, or a combination of them. Bacterium infection persists lifelong if un-
* Corresponding author. Rua Horte“ncia, 454 ^ apto 402, 30280-250, Belo Horizonte, Brazil. Tel.: +55 (31) 32 48 97 75; Fax : +55 (31) 32 48 97 82. E-mail address : [email protected] (E.N. Mendes).
treated [4], and its consequences are still a major public health problem in regions where H. pylori infection is frequent, such as in developing countries, where ulcer disease prevalence and gastric carcinoma morbidity and mortality are high. Several potential markers of pathogenicity have been described in H. pylori, and some of them seem to be associated with more severe clinical outcomes of the infection [5^9]. The bacterium is a genetically very heterogenous species [10,11] and, besides being associated with speci¢c diseases, certain genotypes are more frequently found in certain ethnicities or geographic regions of the world [8,11^14]. For example, cagA positivity has been detected in almost all H. pylori strains isolated from Eastern patients, despite clinical outcome [12,13]. Conversely, in the Western population, cagA strains are not so prevalent and are more frequently found in ulcer or gastric cancer patients [14,15]. With regard to vacA, some authors have proposed that gene mosaicism classi¢cation could be clin-
0928-8244 / 02 / $22.00 8 2002 Federation of European Microbiological Societies. Published by Elsevier Science B.V. All rights reserved. PII : S 0 9 2 8 - 8 2 4 4 ( 0 2 ) 0 0 3 1 1 - 5
FEMSIM 1412 24-6-02
174
A.A.R. Ashour et al. / FEMS Immunology and Medical Microbiology 33 (2002) 173^178
ically relevant, since strains harbouring speci¢c vacA genotypes are associated with di¡erent diseases and, for this reason, may predict the clinical outcome of the infection [6^8,16]. This ¢nding is particularly true for patients from Western countries, probably in consequence of aspects related to the geographic distribution of vacA alleles [6^ 8,13,16]. Despite several papers on vacA genotypes and their relationship to disease having been published, observations reported for one geographic region or ethnicity have not always been con¢rmed in other places [14,17^19]. Thus, studies on the genetic diversity of H. pylori may be important for predicting the clinical outcome of the infection and understanding the evolutionary origin of the organism and its global distribution. Such studies, however, are still sparse in Latin American countries. For these reasons, we undertook this study aiming to investigate H. pylori vacA genotype distribution in Brazilian adult patients with gastritis, duodenal ulcer or gastric carcinoma. Since it is unknown whether infection by multiple strains increases the risk for the development of more severe diseases, we also investigated the frequency of mixed infection by the bacterium. This condition has been little studied, especially in developing countries, where it is supposed to be more frequent in consequence of the higher prevalence of H. pylori infection. We also correlated the presence of cagA with speci¢c vacA genotypes. It may be important to provide a better comprehension of the subject, which could be relevant for our understanding of bacterium-to-host and bacterium-to-bacterium relationships, and which eventually could be useful in formulating strategies for organism eradication.
Endoscopies were performed at Dr. Celso A¡onso de Oliveira Endoscopy Service, when biopsy fragments from the gastric antrum and corpus were taken for microbiology and for histology. Twenty-¢ve of them (10 female and 15 male ; mean age 40.6 years; median 39.0 years; age range 18^61 years) had duodenal ulcer endoscopically proven, and 21 (11 female and 10 male; mean age 51.9 years; median 57.0 years; age range 20^78 years) did not have gastric or duodenal ulcerations at endoscopy. After each procedure, endoscopes were cleansed with detergent, disinfected by immersion in 2.0% glutaraldehyde for 20 min, and then thoroughly rinsed with water. The remaining 36 patients (12 female and 24 male ; mean age 59.1 years; median 58.0 years; age range 32^ 77 years) were selected among those who were submitted to gastric surgery to remove a distal gastric adenocarcinoma at the Surgical Clinic of University Hospital/UFMG, Luxemburgo Hospital, Santa Casa Hospital or the Ma¤rio Penna Hospital of Oncology. Specimens from the stomach of these patients were also taken from the antrum and body regions, at least 2 cm away from the tumours, for microbiological tests. Fragments from the tumour were also removed for histological testing. The materials used for sampling the stomach were cleansed in a way similar to that employed for endoscopes, but autoclaving them when possible. Gastric fragments obtained for microbiological tests were immediately placed in sodium thioglycolate broth (Difco, Detroit, MI, USA) at 4‡C, and processed within 2 h. Fragments from the antral and body mucosa of the stomach and from the tumour were ¢xed in formalin for histological examination. 2.2. Histological examination
2. Patients and methods This study was approved by the Ethics Committee of the University Hospital, Universidade Federal de Minas Gerais, Brazil. Informed consent to participate was obtained from all patients. 2.1. Patients and specimen collection The 82 H. pylori isolates were obtained between 1995 and 1997 from gastric biopsies of non-consecutive patients (33 female and 49 male ; mean age 51.6 years; median 54.0 years; age range 18^78 years) undergoing routine endoscopy (n = 46) for gastric complaints or undergoing gastric surgery to remove gastric carcinoma (n = 36). Most of them were from lower socioeconomic status, lived in an urban area, and were from the state of Minas Gerais, Brazil. All of them belonged to the Portuguese-speaking population, that is, at the present time, a mixed descent from Europeans, specially Iberians, native South American Indians, and Africans. No study patient could be assigned to any particular ancestral ethnic group.
Fragments were processed routinely. Adjacent sections were placed on slides, stained with haematoxylin and eosin, and examined under a light microscope to con¢rm the diagnosis of gastritis and of gastric adenocarcinoma. All H. pylori-positive patients presented some degree of gastritis ranging from mild to severe. Only patients with a histologically con¢rmed diagnosis of gastric adenocarcinoma were included in the group of carcinoma patients. 2.3. H. pylori isolation and identi¢cation Fragments were ground separately in a tissue homogenizer and processed as previously described [15]. Bacterial growth was presumptively identi¢ed as H. pylori by a rapidly positive urease test and by microscopic morphology [16]. Colonies were then removed with a cotton swab and transferred to a microcentrifuge tube containing 500 Wl of sterile distilled water. Bacterium cells were pelleted at 6000Ug for 5 min and maintained at 380‡C until used for DNA extraction. The identi¢cation of H. pylori was later con¢rmed by
FEMSIM 1412 24-6-02
A.A.R. Ashour et al. / FEMS Immunology and Medical Microbiology 33 (2002) 173^178
the detection of ureA, a gene that is present in all H. pylori strains. 2.4. DNA isolation and polymerase chain reactions All bacterial pellets were subjected to genomic DNA extraction as previously described by Fox et al. [20]. After extraction, DNA was pelleted by centrifugation at 12 000Ug at 4‡C for 30 min, allowed to dry in air, and suspended in sterile distilled water. DNA was quanti¢ed by measuring optical density at 260 nm and used for investigation of ureA, cagA and of vacA genotypes. All PCR reaction mixtures were prepared in a dedicated PCR chamber (Plaslabs, Lansing, MI, USA). Reaction mixtures (as described below) were cycled in an MJ Research thermal cycler (MJ Research, Watertown, MA, USA). Amplicons were checked in a separate area to prevent contamination. All ampli¢ed PCR products were resolved in 5% polyacrylamide gels stained with silver nitrate as described by Bassam et al. [24] and modi¢ed by Rocha [25]. Standards of 50 bp and 100 bp (Life Technologies) were used as molecular size markers. 2.5. ureA detection Genomic DNA from all bacterial samples was used to detect ureA by employing oligonucleotide primers and the methodology described by Clayton et al. [21], that amplify a gene fragment of 411 bp. An Escherichia coli (human isolate) was used as negative control, H. pylori ATCC 49503 was used as positive control, and distilled water as an internal-reaction negative control. Only ureA positive samples were employed for further study. 2.6. vacA genotyping Ampli¢cation of vacA signal sequences and midregions was performed by PCR with synthetic oligonucleotide primers described by Atherton et al. [6,22]. They were initially identi¢ed as s1 or s2 types [6] and m1, m2 or m hybrid types [22]. Detailed characterization of the signal
175
region into s1a, s1b [6] or s1c variants [17] was done for all s1 H. pylori samples. For each sample, the PCR mixture contained 25 pmol of each primer, approximately 50 ng of genomic DNA, each deoxynucleotide (Life Technologies, Gaithersburg, MD, USA) at 100 WM concentration, and 1 U of Taq DNA polymerase. All reactions were performed according to methodologies previously described by Atherton et al. [6,22] and by van Doorn et al. [17]. H. pylori strain ATCC 49503 (s1a-m1), strain 435-95 (identi¢ed previously in the Laboratory of Molecular Biology as s1b-m2), and strain Tx30A (s2-m2) were used as controls for vacA allele identi¢cation. Sterile distilled water was used as an internal-reaction negative control. 2.7. cagA detection PCR ampli¢cation of cagA was performed with synthetic oligonucleotides and using the methodology reported by Peek et al. [23]. H. pylori strain Tx30A was used as negative control, H. pylori strain ATCC 49503 was employed as positive control, and distilled water was used as an internal-reaction negative control. Positive reactions yielded a 349-bp ¢nal product. 2.8. Sequencing of s region Three bacterial samples, isolated from patients presenting with carcinoma, ulcer disease, and gastritis only, were selected for s region sequencing among 14 H. pylori samples that reacted with the sets of primers designed to identify s1a and s1b variants of the s1 allele [6]. This vacA region was PCR ampli¢ed with primers VA1-F and VA1-R described by Atherton et al. [6], and ampli¢ed products were directly sequenced, using the DYEnamic ET Terminator Cycle Sequencing Kit (Amersham Pharmacia Biotech, Little Chalfont, Buckinghamshire, UK) and the same primers as above [6]. Sequences were determined in an ABI Prism 377 DNA Sequencer (PE Applied Biosystems, Foster City, CA, USA) and compared to those published by Atherton et al. [6].
Table 1 vacA genotypes in H. pylori strains isolated from patients with gastritis only, duodenal ulcer and gastric carcinoma Genotype
Single-strain infection s1b-m1 s1b-m2 s2-m2 Multiple-strain infection s1a-m1+s1b-m1 s1a-m2+s1b-m2+s2-m2 s1b-m2+s2-m2 Total
Number of strains obtained from patients with
Total
gastritis
ulcer
carcinoma
18 8 2 8 3 0 0 3 21
20 18 0 2 5 3 1 1 25
33 31 0 2 3 3 0 0 36
FEMSIM 1412 24-6-02
71 57 2 12 11 6 1 4 82
176
A.A.R. Ashour et al. / FEMS Immunology and Medical Microbiology 33 (2002) 173^178
2.9. Statistical analysis Two tailed M2 test with Yate’s correction or Fisher’s test was used for statistical analysis. P values were determined by the Epi-Info 2000 software program. Values of less than 0.05 were taken as signi¢cant.
Table 2 cagA status of H. pylori strains isolated from patients with gastritis only, duodenal ulcer and gastric carcinoma cagA
Gastritis (No. (%)) Ulcer (No. (%)) Carcinoma (No. (%))
Positive Negative Total
11 (52.4) 10 (47.6) 21
22 (88.0) 3 (12.0) 25
34 (94.4) 2 (5.6) 36
3. Results Typing by the strategies of Atherton et al. [6,22] was possible in almost all isolates. Alleles s and m were identi¢ed in H. pylori samples obtained from all patients studied. Seventy-one of 82 (86.6%) patients were colonized by H. pylori strains harbouring single vacA genotype (nonmixed infection). Among them, s1 strains were detected in 59 (83.1%) patients. The allele was associated with gastric carcinoma (P 6 0.002, OR = 12.40, 95% CI = 1.91^ 130.51) and with ulcer (P 6 0.03, OR = 7.20, 95% CI = 1.07^78.11). When these strains were further characterized into s1a, s1b, and s1c variants, only s1b ones were identi¢ed. Strains s2 were detected in 12 (16.9%) patients, most of them with gastritis only (Table 1). Strains m1 were detected in 57 of 71 (80.3%) patients with a non-mixed infection. The allele was frequently associated with carcinoma (P 6 0.0002, OR = 19.38, 95% CI = 3.01^200.23) and with ulcer (P 6 0.008, OR = 11.25, 95% CI = 1.69^119.84). H. pylori strains harbouring m2 allele were identi¢ed in 14 (19.7%) patients, mainly in the gastritis only group (Table 1). In sum, the most frequently found genotype among patients with a non-mixed infection was s1b-m1 (57 out of 71 patients; 80.3%). It was signi¢cantly associated with carcinoma (P 6 0.0002, OR = 19.38, 95% CI = 3.01^ 200.23) and with ulcer (P 6 0.008, OR = 11.25, 95% CI = 1.69^119.84). Genotype s1b-m2 was found only in two (2.8%) gastritis-only patients. Pro¢le s2-m2 was detected in 12 (16.9%) patients, and was associated with gastritis only (P 6 0.0011; OR = 9.80, 95% CI = 2.05^ 51.38) (Table 1). H. pylori s1b-m1 strains were distributed across the age cohorts of the patients we studied with frequencies in each age group higher than 72%. No di¡erence in age distribution of vacA genotypes was observed in the group of patients with a non-mixed infection whether carcinoma patients were included (P s 0.90) in the analysis or not (P s 0.19). Nor did di¡erences exist between male and female patients with regard to relative frequencies of vacA genotypes (P = 0.54 for carcinoma patients, P = 0.49 for the ulcer group, and P = 1.0 for gastritis subjects) (data not shown). The three strains selected for s region sequencing were identi¢ed as s1b strains since they had more than 99.5% homology with the s1b sequence published by Atherton et al. [6]. For this reason, the three patients from whom they
were isolated were included among those with a nonmixed infection. Multiple-strain infection (colonization of the same patient by H. pylori strains harbouring more than one vacA genotype) was identi¢ed in 11 (13.4%) patients. The most frequently found association was s1a-m1 plus s1b-m1 (six of 11, 54.5%) (Table 1). No association between multiplestrain infection and patient disease (P s 0.38 and P s 0.70 for gastric carcinoma and ulcer disease, respectively) was observed. Also, mixed infection was not associated with sex (P s 0.70). Although multiple-strain infections were more common in patients younger than 40 years (four out of eight, 50.0%) when carcinoma patients were excluded from the analysis, no statistical association of this condition with patient age cohort was observed (P s 0.40) (data not shown). The gene cagA was detected in 67 (81.7%) patients. The presence of the gene was associated with duodenal ulceration (P 6 0.02, OR = 6.67, 95% CI = 1.30^43.47) and with gastric carcinoma (P 6 0.0003, OR = 15.45, 95% CI = 2.57^ 156.71) (Table 2). No association between cagA and sex (P s 0.49) or age (P s 0.25) was observed. We also found that the s1b-m1 vacA genotype was strongly associated with the presence of cagA (P 6 1037 , OR = unde¢ned, 95% CI = unde¢ned).
4. Discussion This PCR-based analysis is the ¢rst molecular epidemiological study in Brazil testing H. pylori vacA genotype distribution in a series of individuals presenting with gastric complaints, that includes a large number of gastric carcinoma patients. The reasons for the di¡erences in the prevalence of vacA genotypes around the world are not yet completely understood. Considerable heterogeneity in vacA genotypes of H. pylori has been demonstrated by several authors employing di¡erent methods [6^8,11,12,18,26^29]. It has also been demonstrated that the geographic distribution of bacterial strains and the correlation of organism genotypes with clinical outcomes are uneven around the world [12,13,18,19,26^30]. The genotype s1-m1 was the one most frequently found in our population, in agreement with the results we have recently reported for Brazilian children [8] and adults [9].
FEMSIM 1412 24-6-02
A.A.R. Ashour et al. / FEMS Immunology and Medical Microbiology 33 (2002) 173^178
Also, as has been demonstrated for other Western countries, the genotype is signi¢cantly associated with duodenal ulcer and gastric carcinoma [9,26^29]. A possible explanation for this ¢nding is that most s1 strains are associated with higher gastric mucosa in£ammation scores, neutrophil in¢ltration, glandular atrophy, intestinal metaplasia, and epithelial damage [27], possibly because s1 strains are the ones producing cytotoxin [6,27]. However, since cagA and s1 vacA genotypes are strongly associated, the correlation between more severe H. pylori-associated diseases may be due to cagA or to other uncontrolled confounding variables. The ¢nding could also re£ect the high prevalence of the genotype in speci¢c locales or ethnicities as we have recently suggested for the iceA2 genotype in Brazil [30]. Genotype s1b was the only s1 variant found in H. pylori strains isolated from patients with a non-mixed infection. This variant is also more frequent in strains obtained from patients from Portugal and Spain [26,27], countries with historic cultural and economic relationships with Brazil [26], from Mexican [29], and from Colombian patients [27]. Similar distribution of this genotype in the Iberian Peninsula [26] could help us explain why it is so frequent in our population. Our results are also in agreement with data reported by Kidd et al. [28], who demonstrated that the s1b variant was associated with the presence of gastric carcinoma in South African patients. We propose that this association may only be due to the high prevalence of the s1b allele in our H. pylori strains, as suggested above for the s1 genotype. Our hypothesis is supported by the association of s1 strains with duodenal ulcer in some places or ethnicities [6^9,28,29] but not in others [18,27]. No di¡erence was found in vacA genotype distribution with regard to sex and among diverse age cohorts, irrespective of the disease presented by the patient. The genotype vacA s1b-m1, the most common one in the population we studied, was widely distributed across all age groups, from 18 to 78 years. This result probably indicates that the proportion of s1b-m1 strains infecting people in our region has remained constant almost the whole last century, since the infection is acquired in childhood, especially in developing countries. No increasing proportion of any genotype was seen in the younger groups, which probably means that the circulating bacterial strains remained the same in the second half of the 20th century. In conclusion, the s1b variant of vacA is the most prevalent genotype in H. pylori strains isolated from Brazilians, which is in contrast to the results reported for patients from the USA and other Western developed countries. The strong association between vacA s1 strains, peptic ulcer, and gastric carcinoma makes it possible to use vacA genotyping to assess which patients are prone to develop more severe H. pylori-associated diseases in our population, and might help us de¢ne a treatment or prevention strategies for H. pylori infection, as proposed by van Doorn et al. [31]. It should be pointed out that only by sequencing the s1
177
region it was possible to identify strains harbouring the s1b variant. This technique should be employed to identify untypi¢able strains until primers for accurate genotyping of bacterial samples from di¡erent geographical settings are available. Since H. pylori carries only a single copy of vacA [6], detection of multiple genotypes indicates the presence of multiple strains in the same sample. In total, we detected mixed infection in 11 out of 82 patients (13.4%), a condition that was found to be independent of patient disease. This result was similar to our recent report on Brazilian children [8]. It is possible that di¡erent strains acquired in childhood may coexist in the patient and persist lifelong if untreated. Surprisingly, the prevalence of mixed infection we found is lower than that detected by other authors, by employing a similar methodology [26,29]. The frequency of mixed infection in our population could even be overestimated, since sequencing of the s1 region was shown to be necessary for identi¢cation of the s1b variant. For example, three strains that cross-reacted with the sets of primers designed to detect s1a and s1b were identi¢ed as s1b ones by sequencing. However, we did not sequence all such strains. On the other hand, mixed infection prevalence could be underestimated because of the low number of mosaics of any combination of signal and midregion alleles and because multiple strains that have identical vacA and cagA genotypes are not detected by the method we employed. In this study we employed only a pair of primers, which does not anneal all cagA sequences because of gene heterogeneity. Thus, the cagA negative strains that we detected may include false negative ones. Despite this, the presence of cagA was signi¢cantly associated with gastric carcinoma and duodenal ulcer disease, as previously described for patients from di¡erent Western countries [9,12,14,17,19,27^29,31], and from Brazil [16]. No di¡erence was found in cagA distribution according to sex and among diverse age cohorts in ulcer and gastritis-only patients. This ¢nding reinforces the possibility that the circulating H. pylori strains remained the same in the last decades of the past century.
Acknowledgements This work was supported by FAPEMIG and CNPqBrazil. We acknowledge the generosity of Dr. E. Kalapothakis for providing the Taq DNA polymerase, and Ms. Anna Christina Matos Salim, from the Laborato¤rio de Gene¤tica do Ca“ncer^Ludwig Institute for Cancer Research, Brazil, for help with sequencing.
References [1] NIH Consensus Development Panel on Helicobacter pylori in peptic
FEMSIM 1412 24-6-02
178
[2]
[3] [4] [5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
A.A.R. Ashour et al. / FEMS Immunology and Medical Microbiology 33 (2002) 173^178 ulcer disease (1994) Helicobacter pylori in peptic ulcer disease. J. Am. Med. Assoc. 272, 65^69. International Agency for Research on Cancer (1994) Infection with Helicobacter pylori. IARC Monogr. Eval. Carcinog. Risks Hum. 61, 177^240. Wotherspoon, A.C. (1998) Helicobacter pylori infection and gastric lymphoma. Br. Med. Bull. 54, 79^85. Taylor, D.N. and Blaser, M.J. (1991) The epidemiology of Helicobacter pylori infection. Epidemiol. Res. 13, 42^59. Cover, T.L., Glupczynski, Y., Lage, A.P., Burette, A., Tummuru, M.K.R., Perez-Perez, G.I. and Blaser, M.J. (1995) Serological detection of infection with cagAþ Helicobacter pylori strains. J. Clin. Microbiol. 33, 1496^1500. Atherton, J.C., Cao, P., Peek, R.M., Tummuru, M.K.R., Blaser, M.J. and Cover, T.L. (1995) Mosaicism in vacuolating cytotoxin alleles of Helicobacter pylori: association of speci¢c vacA types with cytotoxin production and peptic ulceration. J. Biol. Chem. 270, 17771^17777. Atherton, J.C., Peek, R.M., Tham, K.T., Cover, T.L. and Blaser, M.J. (1997) Clinical and pathological importance of heterogeneity in vacA, the vacuolating cytotoxin gene of Helicobacter pylori. Gastroenterology 112, 92^99. Gusma‹o, V.R., Mendes, E.N., Queiroz, D.M.M., Rocha, G.A., Rocha, A.M.C., Ashour, A.A.R. and Carvalho, A.S.T. (2000) vacA genotypes in Helicobacter pylori strains isolated from children with and without duodenal ulcer in Brazil. J. Clin. Microbiol. 38, 2853^2857. Evans, D.G., Queiroz, D.M.M., Mendes Jr., E.N. and Evans, D.E. (1998) Helicobacter pylori cagA status and s and m alleles of vacA in isolates from individuals with a variety of H. pylori-associated gastric diseases. J. Clin. Microbiol. 36, 3435^3437. Suerbaum, S., Smith, J.M., Bapumia, K., Morelli, G., Smith, N.H., Kunstmann, E., Dyrek, I. and Achtman, M. (1998) Free recombination within Helicobacter pylori. Proc. Natl. Acad. Sci. USA 95, 12619^12624. Achtman, M., Azuma, T., Berg, D.E., Ito, Y., Morelli, G., Pan, Z.J., Suerbaum, S., Thompson, S.A., van der Ende, A. and van Doorn, L.J. (1999) Recombination and clonal groupings within Helicobacter pylori from di¡erent geographical regions. Mol. Microbiol. 32, 459^ 470. van der Ende, A.Z., Pan, J., Bart, A., van der Hulst, R.W., Feller, M., Xiao, S.D., Tytgat, G.N. and Dankert, J. (1998) cagA-positive Helicobacter pylori populations in China and The Netherlands are distinct. Infect. Immun. 66, 1822^1826. Ito, Y., Azuma, T., Ito, S., Miyaji, H., Hirai, M., Yamazaki, Y., Sato, F., Kato, T., Kohli, Y. and Kuriyama, M. (1997) Analysis and typing of the vacA gene from cagA-positive strains of Helicobacter pylori isolated in Japan. J. Clin. Microbiol. 35, 1710^1714. Miehlke, S., Kibler, K., Kim, J.G., Figura, N., Small, S.M., Graham, D.Y. and Go, M.F. (1996) Allelic variation in the cagA gene of Helicobacter pylori obtained from Korea compared to the United States. Am. J. Gastroenterol. 17, 1322^1325. Queiroz, D.M.M., Mendes, E.N. and Rocha, G.A. (1987) Indicator medium for isolation of Campylobacter pylori. J. Clin. Microbiol. 25, 2378^2379. Queiroz, D.M.M., Mendes, E.N., Rocha, G.A., Oliveira, A.M.R., Oliveira, C.A., Magalha‹es, P.P., Moura, S.B., Cabral, M.M.D.A. and Nogueira, A.M.M.F. (1998) cagA-positive Helicobacter pylori and risk for developing gastric carcinoma in Brazil. Int. J. Cancer 78, 135^139. van Doorn, L.-J., Figueiredo, C., Sanna, R., Plaisier, A., Schneeberger, P., de Boer, W. and Quint, W. (1998) Clinical relevance of cagA,
[18]
[19]
[20]
[21]
[22]
[23]
[24]
[25]
[26]
[27]
[28]
[29]
[30]
[31]
vacA, and iceA status of Helicobacter pylori. Gastroenterology 115, 58^66. Go, M.F., Cissell, L. and Graham, D.Y. (1998) Failure to con¢rm association of vacA gene mosaicism with duodenal ulcer disease. Scand. J. Gastroenterol. 33, 132^136. Yamaoka, Y., Kodama, T., Gutierrez, O., Kim, J.G., Kashima, K. and Graham, D.Y. (1999) Relationship between Helicobacter pylori iceA, vacA, and cagA status and clinical outcome: studies in four di¡erent countries. J. Clin. Microbiol. 37, 2274^2279. Fox, J.G., Dewhirst, F.E., Fraser, G.J., Paster, B.J., Shames, B. and Murphy, J.C. (1994) Intracellular Campylobacter-like organism from ferrets and hamsters with proliferative bowel disease is a Desulfovibrio sp.. J. Clin. Microbiol. 32, 1229^1237. Clayton, C.L., Kleanthous, H., Coates, P.J., Morgan, D.D. and Tabaqchali, S. (1992) Sensitive detection of Helicobacter pylori by using polymerase chain reaction. J. Clin. Microbiol. 30, 192^200. Atherton, J.C., Cover, T.L., Wells, R.J., Morales, M.R., Hawkey, C.J. and Blaser, M.J. (1999) Simple and accurate PCR-based system for typing vacuolating cytotoxin alleles of Helicobacter pylori. J. Clin. Microbiol. 37, 2979^2982. Peek, R.M., Miller, G.G., Tham, K.T., Perez-Perez, G.I., Zhao, X., Atherton, J.C. and Blaser, M.J. (1995) Heightened in£ammatory response and cytokine expression in vivo to cagAþ Helicobacter pylori. Lab. Invest. 71, 760^770. Bassam, B.J., Caetano-Anlolle¤s, G. and Gressho¡, P.M. (1991) Fast and sensitive silver staining of DNA in polyacrylamide gels. Anal. Biochem. 196, 80^83. Rocha, J.L.F. (1995) O receptor V2 de vasopressina e o canal de a¤gua aquaporina-2 no diabetes insipidus nefroge“nico. Master’s Thesis. Instituto de Cie“ncias da Biolo¤gicas da UFMG, Belo Horizonte, Brasil. van Doorn, L.-J., Figueiredo, C., Me¤graud, F., Pena, S., Midolo, P., Queiroz, D.M.M., Carneiro, F., Vanderborght, B., Pegado, M.G.F., Sanna, R., de Boer, W., Schneeberger, P.M., Correa, P., Ng, E.K.W., Atherton, J., Blaser, M.J. and Quint, W.G.V. (1999) Geographic distribution of vacA allelic types of Helicobacter pylori. Gastroenterology 116, 823^830. Nogueira, C., Figueiredo, C., Carneiro, F., Gomes, A.T., Barreira, R., Figueira, P., Salgado, C., Belo, L., Peixoto, A., Bravo, J.C., Bravo, L.E., Realpe, J.L., Plaisier, A.P., Quint, W.G., Riz, B., Correa, P. and van Doorn, L.-J. (2001) Helicobacter pylori genotypes may determine gastric histopathology. Am. J. Pathol. 158, 647^654. Kidd, M., Lastovica, A.J., Atherton, J.C. and Louw, J.A. (1999) Heterogeneity in the Helicobacter pylori vacA and cagA genes: association with gastroduodenal disease in South Africa? Gut 45, 499^ 502. Morales-Espinosa, R., Castillo-Rojas, G., Gonzalez-Valencia, G., Ponce de Leo¤n, S., Cravioto, A., Atherton, J.C. and Lo¤pez-Vidal, Y. (1999) Colonization of Mexican patients by multiple Helicobacter pylori strains with di¡erent vacA and cagA genotypes. J. Clin. Microbiol. 37, 3001^3004. Ashour, A.A.R., Collares, G.B., Mendes, E.N., Gusma‹o, V.R., Queiroz, D.M.M., Magalha‹es, P.P., Carvalho, A.S.T., Oliveira, C.A., Nogueira, A.M.M.F., Rocha, G.A. and Rocha, A.M.C. (2001) iceA genotypes of Helicobacter pylori strains isolated from Brazilian children and adults. J. Clin. Microbiol. 39, 1746^1750. van Doorn, L.-J., Schneeberger, P.M., Nouhan, N., Plaisier, A.P., Quint, W.G.V. and de Boer, W.A. (2000) Importance of Helicobacter pylori cagA and vacA for the e⁄cacy of antibiotic treatment. Gut 46, 321^326.
FEMSIM 1412 24-6-02
www.fems-microbiology.org
Distribution of vacA genotypes in Helicobacter pylori strains isolated from Brazilian adult patients with gastritis, duodenal ulcer or gastric carcinoma Abdussalam A.R. Ashour a;d , Paula P. Magalha‹es b;e;f , Edilberto N. Mendes a; , Guilherme B. Collares a , Valqu|¤ria R. de Gusma‹o a , Dulciene M.M. Queiroz b , Ana Margarida M.F. Nogueira c , Gifone A. Rocha b , Celso A. de Oliveira b b
a Laboratory of Molecular Biology, Faculdade de Medicina, Universidade Federal de Minas Gerais, 30130-100 Belo Horizonte, Brazil Laboratory of Research in Bacteriology, Faculdade de Medicina, Universidade Federal de Minas Gerais, 30130-100 Belo Horizonte, Brazil c Department of Pathology, Faculdade de Medicina, Universidade Federal de Minas Gerais, 30130-100 Belo Horizonte, Brazil d Department of Microbiology, Escola Paulista de Medicina, Universidade Federal de Sa‹o Paulo, 04023-062 Sa‹o Paulo, Brazil e Instituto de Microbiologia Prof. Paulo de Go¤es, Universidade Federal do Rio de Janeiro, 21941-590 Rio de Janeiro, Brazil f Faculdade de Fisioterapia, Universidade de Itau¤na, 35680-033 Itau¤na, Brazil
Received 4 December 2001 ; received in revised form 27 February 2002; accepted 14 March 2002 First published online 20 May 2002
Abstract This PCR-based analysis is the first molecular epidemiological study in Brazil testing Helicobacter pylori cagA and vacA distribution in adults with gastric complaints, that includes a large number of carcinoma patients. Multiple-strain infection was identified in 11/13.4% patients. vacA s1-m1 and cagAþ genotypes were the most common in patients with a non-mixed infection. All vacA s1 strains were s1b, so subtyping s1 strains was not useful. vacA s1b-m1 and cagAþ strains were associated with higher prevalence of peptic ulcer and gastric carcinoma than vacA s2-m2 and cagA3 ones. In conclusion, cagA and vacA genotyping may have clinical relevance in Brazil. 8 2002 Federation of European Microbiological Societies. Published by Elsevier Science B.V. All rights reserved. Keywords : vacA; cagA; Bacterial genetic diversity ; Helicobacter pylori
1. Introduction Most Helicobacter pylori infections are clinically silent. Only around 20% of H. pylori infected adults manifest one of the four severe outcomes associated with the bacterium: gastric ulcer, duodenal ulcer, gastric carcinoma or lymphoma [1^3]. It is still unknown why the organism is able to produce severe disease in certain hosts and be innocuous in others. Clinical outcome of the infection is determined by several factors, such as di¡erences in host responses to bacterium stimulation, speci¢c organism virulence factors and environmental in£uences, or a combination of them. Bacterium infection persists lifelong if un-
* Corresponding author. Rua Horte“ncia, 454 ^ apto 402, 30280-250, Belo Horizonte, Brazil. Tel.: +55 (31) 32 48 97 75; Fax : +55 (31) 32 48 97 82. E-mail address : [email protected] (E.N. Mendes).
treated [4], and its consequences are still a major public health problem in regions where H. pylori infection is frequent, such as in developing countries, where ulcer disease prevalence and gastric carcinoma morbidity and mortality are high. Several potential markers of pathogenicity have been described in H. pylori, and some of them seem to be associated with more severe clinical outcomes of the infection [5^9]. The bacterium is a genetically very heterogenous species [10,11] and, besides being associated with speci¢c diseases, certain genotypes are more frequently found in certain ethnicities or geographic regions of the world [8,11^14]. For example, cagA positivity has been detected in almost all H. pylori strains isolated from Eastern patients, despite clinical outcome [12,13]. Conversely, in the Western population, cagA strains are not so prevalent and are more frequently found in ulcer or gastric cancer patients [14,15]. With regard to vacA, some authors have proposed that gene mosaicism classi¢cation could be clin-
0928-8244 / 02 / $22.00 8 2002 Federation of European Microbiological Societies. Published by Elsevier Science B.V. All rights reserved. PII : S 0 9 2 8 - 8 2 4 4 ( 0 2 ) 0 0 3 1 1 - 5
FEMSIM 1412 24-6-02
174
A.A.R. Ashour et al. / FEMS Immunology and Medical Microbiology 33 (2002) 173^178
ically relevant, since strains harbouring speci¢c vacA genotypes are associated with di¡erent diseases and, for this reason, may predict the clinical outcome of the infection [6^8,16]. This ¢nding is particularly true for patients from Western countries, probably in consequence of aspects related to the geographic distribution of vacA alleles [6^ 8,13,16]. Despite several papers on vacA genotypes and their relationship to disease having been published, observations reported for one geographic region or ethnicity have not always been con¢rmed in other places [14,17^19]. Thus, studies on the genetic diversity of H. pylori may be important for predicting the clinical outcome of the infection and understanding the evolutionary origin of the organism and its global distribution. Such studies, however, are still sparse in Latin American countries. For these reasons, we undertook this study aiming to investigate H. pylori vacA genotype distribution in Brazilian adult patients with gastritis, duodenal ulcer or gastric carcinoma. Since it is unknown whether infection by multiple strains increases the risk for the development of more severe diseases, we also investigated the frequency of mixed infection by the bacterium. This condition has been little studied, especially in developing countries, where it is supposed to be more frequent in consequence of the higher prevalence of H. pylori infection. We also correlated the presence of cagA with speci¢c vacA genotypes. It may be important to provide a better comprehension of the subject, which could be relevant for our understanding of bacterium-to-host and bacterium-to-bacterium relationships, and which eventually could be useful in formulating strategies for organism eradication.
Endoscopies were performed at Dr. Celso A¡onso de Oliveira Endoscopy Service, when biopsy fragments from the gastric antrum and corpus were taken for microbiology and for histology. Twenty-¢ve of them (10 female and 15 male ; mean age 40.6 years; median 39.0 years; age range 18^61 years) had duodenal ulcer endoscopically proven, and 21 (11 female and 10 male; mean age 51.9 years; median 57.0 years; age range 20^78 years) did not have gastric or duodenal ulcerations at endoscopy. After each procedure, endoscopes were cleansed with detergent, disinfected by immersion in 2.0% glutaraldehyde for 20 min, and then thoroughly rinsed with water. The remaining 36 patients (12 female and 24 male ; mean age 59.1 years; median 58.0 years; age range 32^ 77 years) were selected among those who were submitted to gastric surgery to remove a distal gastric adenocarcinoma at the Surgical Clinic of University Hospital/UFMG, Luxemburgo Hospital, Santa Casa Hospital or the Ma¤rio Penna Hospital of Oncology. Specimens from the stomach of these patients were also taken from the antrum and body regions, at least 2 cm away from the tumours, for microbiological tests. Fragments from the tumour were also removed for histological testing. The materials used for sampling the stomach were cleansed in a way similar to that employed for endoscopes, but autoclaving them when possible. Gastric fragments obtained for microbiological tests were immediately placed in sodium thioglycolate broth (Difco, Detroit, MI, USA) at 4‡C, and processed within 2 h. Fragments from the antral and body mucosa of the stomach and from the tumour were ¢xed in formalin for histological examination. 2.2. Histological examination
2. Patients and methods This study was approved by the Ethics Committee of the University Hospital, Universidade Federal de Minas Gerais, Brazil. Informed consent to participate was obtained from all patients. 2.1. Patients and specimen collection The 82 H. pylori isolates were obtained between 1995 and 1997 from gastric biopsies of non-consecutive patients (33 female and 49 male ; mean age 51.6 years; median 54.0 years; age range 18^78 years) undergoing routine endoscopy (n = 46) for gastric complaints or undergoing gastric surgery to remove gastric carcinoma (n = 36). Most of them were from lower socioeconomic status, lived in an urban area, and were from the state of Minas Gerais, Brazil. All of them belonged to the Portuguese-speaking population, that is, at the present time, a mixed descent from Europeans, specially Iberians, native South American Indians, and Africans. No study patient could be assigned to any particular ancestral ethnic group.
Fragments were processed routinely. Adjacent sections were placed on slides, stained with haematoxylin and eosin, and examined under a light microscope to con¢rm the diagnosis of gastritis and of gastric adenocarcinoma. All H. pylori-positive patients presented some degree of gastritis ranging from mild to severe. Only patients with a histologically con¢rmed diagnosis of gastric adenocarcinoma were included in the group of carcinoma patients. 2.3. H. pylori isolation and identi¢cation Fragments were ground separately in a tissue homogenizer and processed as previously described [15]. Bacterial growth was presumptively identi¢ed as H. pylori by a rapidly positive urease test and by microscopic morphology [16]. Colonies were then removed with a cotton swab and transferred to a microcentrifuge tube containing 500 Wl of sterile distilled water. Bacterium cells were pelleted at 6000Ug for 5 min and maintained at 380‡C until used for DNA extraction. The identi¢cation of H. pylori was later con¢rmed by
FEMSIM 1412 24-6-02
A.A.R. Ashour et al. / FEMS Immunology and Medical Microbiology 33 (2002) 173^178
the detection of ureA, a gene that is present in all H. pylori strains. 2.4. DNA isolation and polymerase chain reactions All bacterial pellets were subjected to genomic DNA extraction as previously described by Fox et al. [20]. After extraction, DNA was pelleted by centrifugation at 12 000Ug at 4‡C for 30 min, allowed to dry in air, and suspended in sterile distilled water. DNA was quanti¢ed by measuring optical density at 260 nm and used for investigation of ureA, cagA and of vacA genotypes. All PCR reaction mixtures were prepared in a dedicated PCR chamber (Plaslabs, Lansing, MI, USA). Reaction mixtures (as described below) were cycled in an MJ Research thermal cycler (MJ Research, Watertown, MA, USA). Amplicons were checked in a separate area to prevent contamination. All ampli¢ed PCR products were resolved in 5% polyacrylamide gels stained with silver nitrate as described by Bassam et al. [24] and modi¢ed by Rocha [25]. Standards of 50 bp and 100 bp (Life Technologies) were used as molecular size markers. 2.5. ureA detection Genomic DNA from all bacterial samples was used to detect ureA by employing oligonucleotide primers and the methodology described by Clayton et al. [21], that amplify a gene fragment of 411 bp. An Escherichia coli (human isolate) was used as negative control, H. pylori ATCC 49503 was used as positive control, and distilled water as an internal-reaction negative control. Only ureA positive samples were employed for further study. 2.6. vacA genotyping Ampli¢cation of vacA signal sequences and midregions was performed by PCR with synthetic oligonucleotide primers described by Atherton et al. [6,22]. They were initially identi¢ed as s1 or s2 types [6] and m1, m2 or m hybrid types [22]. Detailed characterization of the signal
175
region into s1a, s1b [6] or s1c variants [17] was done for all s1 H. pylori samples. For each sample, the PCR mixture contained 25 pmol of each primer, approximately 50 ng of genomic DNA, each deoxynucleotide (Life Technologies, Gaithersburg, MD, USA) at 100 WM concentration, and 1 U of Taq DNA polymerase. All reactions were performed according to methodologies previously described by Atherton et al. [6,22] and by van Doorn et al. [17]. H. pylori strain ATCC 49503 (s1a-m1), strain 435-95 (identi¢ed previously in the Laboratory of Molecular Biology as s1b-m2), and strain Tx30A (s2-m2) were used as controls for vacA allele identi¢cation. Sterile distilled water was used as an internal-reaction negative control. 2.7. cagA detection PCR ampli¢cation of cagA was performed with synthetic oligonucleotides and using the methodology reported by Peek et al. [23]. H. pylori strain Tx30A was used as negative control, H. pylori strain ATCC 49503 was employed as positive control, and distilled water was used as an internal-reaction negative control. Positive reactions yielded a 349-bp ¢nal product. 2.8. Sequencing of s region Three bacterial samples, isolated from patients presenting with carcinoma, ulcer disease, and gastritis only, were selected for s region sequencing among 14 H. pylori samples that reacted with the sets of primers designed to identify s1a and s1b variants of the s1 allele [6]. This vacA region was PCR ampli¢ed with primers VA1-F and VA1-R described by Atherton et al. [6], and ampli¢ed products were directly sequenced, using the DYEnamic ET Terminator Cycle Sequencing Kit (Amersham Pharmacia Biotech, Little Chalfont, Buckinghamshire, UK) and the same primers as above [6]. Sequences were determined in an ABI Prism 377 DNA Sequencer (PE Applied Biosystems, Foster City, CA, USA) and compared to those published by Atherton et al. [6].
Table 1 vacA genotypes in H. pylori strains isolated from patients with gastritis only, duodenal ulcer and gastric carcinoma Genotype
Single-strain infection s1b-m1 s1b-m2 s2-m2 Multiple-strain infection s1a-m1+s1b-m1 s1a-m2+s1b-m2+s2-m2 s1b-m2+s2-m2 Total
Number of strains obtained from patients with
Total
gastritis
ulcer
carcinoma
18 8 2 8 3 0 0 3 21
20 18 0 2 5 3 1 1 25
33 31 0 2 3 3 0 0 36
FEMSIM 1412 24-6-02
71 57 2 12 11 6 1 4 82
176
A.A.R. Ashour et al. / FEMS Immunology and Medical Microbiology 33 (2002) 173^178
2.9. Statistical analysis Two tailed M2 test with Yate’s correction or Fisher’s test was used for statistical analysis. P values were determined by the Epi-Info 2000 software program. Values of less than 0.05 were taken as signi¢cant.
Table 2 cagA status of H. pylori strains isolated from patients with gastritis only, duodenal ulcer and gastric carcinoma cagA
Gastritis (No. (%)) Ulcer (No. (%)) Carcinoma (No. (%))
Positive Negative Total
11 (52.4) 10 (47.6) 21
22 (88.0) 3 (12.0) 25
34 (94.4) 2 (5.6) 36
3. Results Typing by the strategies of Atherton et al. [6,22] was possible in almost all isolates. Alleles s and m were identi¢ed in H. pylori samples obtained from all patients studied. Seventy-one of 82 (86.6%) patients were colonized by H. pylori strains harbouring single vacA genotype (nonmixed infection). Among them, s1 strains were detected in 59 (83.1%) patients. The allele was associated with gastric carcinoma (P 6 0.002, OR = 12.40, 95% CI = 1.91^ 130.51) and with ulcer (P 6 0.03, OR = 7.20, 95% CI = 1.07^78.11). When these strains were further characterized into s1a, s1b, and s1c variants, only s1b ones were identi¢ed. Strains s2 were detected in 12 (16.9%) patients, most of them with gastritis only (Table 1). Strains m1 were detected in 57 of 71 (80.3%) patients with a non-mixed infection. The allele was frequently associated with carcinoma (P 6 0.0002, OR = 19.38, 95% CI = 3.01^200.23) and with ulcer (P 6 0.008, OR = 11.25, 95% CI = 1.69^119.84). H. pylori strains harbouring m2 allele were identi¢ed in 14 (19.7%) patients, mainly in the gastritis only group (Table 1). In sum, the most frequently found genotype among patients with a non-mixed infection was s1b-m1 (57 out of 71 patients; 80.3%). It was signi¢cantly associated with carcinoma (P 6 0.0002, OR = 19.38, 95% CI = 3.01^ 200.23) and with ulcer (P 6 0.008, OR = 11.25, 95% CI = 1.69^119.84). Genotype s1b-m2 was found only in two (2.8%) gastritis-only patients. Pro¢le s2-m2 was detected in 12 (16.9%) patients, and was associated with gastritis only (P 6 0.0011; OR = 9.80, 95% CI = 2.05^ 51.38) (Table 1). H. pylori s1b-m1 strains were distributed across the age cohorts of the patients we studied with frequencies in each age group higher than 72%. No di¡erence in age distribution of vacA genotypes was observed in the group of patients with a non-mixed infection whether carcinoma patients were included (P s 0.90) in the analysis or not (P s 0.19). Nor did di¡erences exist between male and female patients with regard to relative frequencies of vacA genotypes (P = 0.54 for carcinoma patients, P = 0.49 for the ulcer group, and P = 1.0 for gastritis subjects) (data not shown). The three strains selected for s region sequencing were identi¢ed as s1b strains since they had more than 99.5% homology with the s1b sequence published by Atherton et al. [6]. For this reason, the three patients from whom they
were isolated were included among those with a nonmixed infection. Multiple-strain infection (colonization of the same patient by H. pylori strains harbouring more than one vacA genotype) was identi¢ed in 11 (13.4%) patients. The most frequently found association was s1a-m1 plus s1b-m1 (six of 11, 54.5%) (Table 1). No association between multiplestrain infection and patient disease (P s 0.38 and P s 0.70 for gastric carcinoma and ulcer disease, respectively) was observed. Also, mixed infection was not associated with sex (P s 0.70). Although multiple-strain infections were more common in patients younger than 40 years (four out of eight, 50.0%) when carcinoma patients were excluded from the analysis, no statistical association of this condition with patient age cohort was observed (P s 0.40) (data not shown). The gene cagA was detected in 67 (81.7%) patients. The presence of the gene was associated with duodenal ulceration (P 6 0.02, OR = 6.67, 95% CI = 1.30^43.47) and with gastric carcinoma (P 6 0.0003, OR = 15.45, 95% CI = 2.57^ 156.71) (Table 2). No association between cagA and sex (P s 0.49) or age (P s 0.25) was observed. We also found that the s1b-m1 vacA genotype was strongly associated with the presence of cagA (P 6 1037 , OR = unde¢ned, 95% CI = unde¢ned).
4. Discussion This PCR-based analysis is the ¢rst molecular epidemiological study in Brazil testing H. pylori vacA genotype distribution in a series of individuals presenting with gastric complaints, that includes a large number of gastric carcinoma patients. The reasons for the di¡erences in the prevalence of vacA genotypes around the world are not yet completely understood. Considerable heterogeneity in vacA genotypes of H. pylori has been demonstrated by several authors employing di¡erent methods [6^8,11,12,18,26^29]. It has also been demonstrated that the geographic distribution of bacterial strains and the correlation of organism genotypes with clinical outcomes are uneven around the world [12,13,18,19,26^30]. The genotype s1-m1 was the one most frequently found in our population, in agreement with the results we have recently reported for Brazilian children [8] and adults [9].
FEMSIM 1412 24-6-02
A.A.R. Ashour et al. / FEMS Immunology and Medical Microbiology 33 (2002) 173^178
Also, as has been demonstrated for other Western countries, the genotype is signi¢cantly associated with duodenal ulcer and gastric carcinoma [9,26^29]. A possible explanation for this ¢nding is that most s1 strains are associated with higher gastric mucosa in£ammation scores, neutrophil in¢ltration, glandular atrophy, intestinal metaplasia, and epithelial damage [27], possibly because s1 strains are the ones producing cytotoxin [6,27]. However, since cagA and s1 vacA genotypes are strongly associated, the correlation between more severe H. pylori-associated diseases may be due to cagA or to other uncontrolled confounding variables. The ¢nding could also re£ect the high prevalence of the genotype in speci¢c locales or ethnicities as we have recently suggested for the iceA2 genotype in Brazil [30]. Genotype s1b was the only s1 variant found in H. pylori strains isolated from patients with a non-mixed infection. This variant is also more frequent in strains obtained from patients from Portugal and Spain [26,27], countries with historic cultural and economic relationships with Brazil [26], from Mexican [29], and from Colombian patients [27]. Similar distribution of this genotype in the Iberian Peninsula [26] could help us explain why it is so frequent in our population. Our results are also in agreement with data reported by Kidd et al. [28], who demonstrated that the s1b variant was associated with the presence of gastric carcinoma in South African patients. We propose that this association may only be due to the high prevalence of the s1b allele in our H. pylori strains, as suggested above for the s1 genotype. Our hypothesis is supported by the association of s1 strains with duodenal ulcer in some places or ethnicities [6^9,28,29] but not in others [18,27]. No di¡erence was found in vacA genotype distribution with regard to sex and among diverse age cohorts, irrespective of the disease presented by the patient. The genotype vacA s1b-m1, the most common one in the population we studied, was widely distributed across all age groups, from 18 to 78 years. This result probably indicates that the proportion of s1b-m1 strains infecting people in our region has remained constant almost the whole last century, since the infection is acquired in childhood, especially in developing countries. No increasing proportion of any genotype was seen in the younger groups, which probably means that the circulating bacterial strains remained the same in the second half of the 20th century. In conclusion, the s1b variant of vacA is the most prevalent genotype in H. pylori strains isolated from Brazilians, which is in contrast to the results reported for patients from the USA and other Western developed countries. The strong association between vacA s1 strains, peptic ulcer, and gastric carcinoma makes it possible to use vacA genotyping to assess which patients are prone to develop more severe H. pylori-associated diseases in our population, and might help us de¢ne a treatment or prevention strategies for H. pylori infection, as proposed by van Doorn et al. [31]. It should be pointed out that only by sequencing the s1
177
region it was possible to identify strains harbouring the s1b variant. This technique should be employed to identify untypi¢able strains until primers for accurate genotyping of bacterial samples from di¡erent geographical settings are available. Since H. pylori carries only a single copy of vacA [6], detection of multiple genotypes indicates the presence of multiple strains in the same sample. In total, we detected mixed infection in 11 out of 82 patients (13.4%), a condition that was found to be independent of patient disease. This result was similar to our recent report on Brazilian children [8]. It is possible that di¡erent strains acquired in childhood may coexist in the patient and persist lifelong if untreated. Surprisingly, the prevalence of mixed infection we found is lower than that detected by other authors, by employing a similar methodology [26,29]. The frequency of mixed infection in our population could even be overestimated, since sequencing of the s1 region was shown to be necessary for identi¢cation of the s1b variant. For example, three strains that cross-reacted with the sets of primers designed to detect s1a and s1b were identi¢ed as s1b ones by sequencing. However, we did not sequence all such strains. On the other hand, mixed infection prevalence could be underestimated because of the low number of mosaics of any combination of signal and midregion alleles and because multiple strains that have identical vacA and cagA genotypes are not detected by the method we employed. In this study we employed only a pair of primers, which does not anneal all cagA sequences because of gene heterogeneity. Thus, the cagA negative strains that we detected may include false negative ones. Despite this, the presence of cagA was signi¢cantly associated with gastric carcinoma and duodenal ulcer disease, as previously described for patients from di¡erent Western countries [9,12,14,17,19,27^29,31], and from Brazil [16]. No di¡erence was found in cagA distribution according to sex and among diverse age cohorts in ulcer and gastritis-only patients. This ¢nding reinforces the possibility that the circulating H. pylori strains remained the same in the last decades of the past century.
Acknowledgements This work was supported by FAPEMIG and CNPqBrazil. We acknowledge the generosity of Dr. E. Kalapothakis for providing the Taq DNA polymerase, and Ms. Anna Christina Matos Salim, from the Laborato¤rio de Gene¤tica do Ca“ncer^Ludwig Institute for Cancer Research, Brazil, for help with sequencing.
References [1] NIH Consensus Development Panel on Helicobacter pylori in peptic
FEMSIM 1412 24-6-02
178
[2]
[3] [4] [5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
A.A.R. Ashour et al. / FEMS Immunology and Medical Microbiology 33 (2002) 173^178 ulcer disease (1994) Helicobacter pylori in peptic ulcer disease. J. Am. Med. Assoc. 272, 65^69. International Agency for Research on Cancer (1994) Infection with Helicobacter pylori. IARC Monogr. Eval. Carcinog. Risks Hum. 61, 177^240. Wotherspoon, A.C. (1998) Helicobacter pylori infection and gastric lymphoma. Br. Med. Bull. 54, 79^85. Taylor, D.N. and Blaser, M.J. (1991) The epidemiology of Helicobacter pylori infection. Epidemiol. Res. 13, 42^59. Cover, T.L., Glupczynski, Y., Lage, A.P., Burette, A., Tummuru, M.K.R., Perez-Perez, G.I. and Blaser, M.J. (1995) Serological detection of infection with cagAþ Helicobacter pylori strains. J. Clin. Microbiol. 33, 1496^1500. Atherton, J.C., Cao, P., Peek, R.M., Tummuru, M.K.R., Blaser, M.J. and Cover, T.L. (1995) Mosaicism in vacuolating cytotoxin alleles of Helicobacter pylori: association of speci¢c vacA types with cytotoxin production and peptic ulceration. J. Biol. Chem. 270, 17771^17777. Atherton, J.C., Peek, R.M., Tham, K.T., Cover, T.L. and Blaser, M.J. (1997) Clinical and pathological importance of heterogeneity in vacA, the vacuolating cytotoxin gene of Helicobacter pylori. Gastroenterology 112, 92^99. Gusma‹o, V.R., Mendes, E.N., Queiroz, D.M.M., Rocha, G.A., Rocha, A.M.C., Ashour, A.A.R. and Carvalho, A.S.T. (2000) vacA genotypes in Helicobacter pylori strains isolated from children with and without duodenal ulcer in Brazil. J. Clin. Microbiol. 38, 2853^2857. Evans, D.G., Queiroz, D.M.M., Mendes Jr., E.N. and Evans, D.E. (1998) Helicobacter pylori cagA status and s and m alleles of vacA in isolates from individuals with a variety of H. pylori-associated gastric diseases. J. Clin. Microbiol. 36, 3435^3437. Suerbaum, S., Smith, J.M., Bapumia, K., Morelli, G., Smith, N.H., Kunstmann, E., Dyrek, I. and Achtman, M. (1998) Free recombination within Helicobacter pylori. Proc. Natl. Acad. Sci. USA 95, 12619^12624. Achtman, M., Azuma, T., Berg, D.E., Ito, Y., Morelli, G., Pan, Z.J., Suerbaum, S., Thompson, S.A., van der Ende, A. and van Doorn, L.J. (1999) Recombination and clonal groupings within Helicobacter pylori from di¡erent geographical regions. Mol. Microbiol. 32, 459^ 470. van der Ende, A.Z., Pan, J., Bart, A., van der Hulst, R.W., Feller, M., Xiao, S.D., Tytgat, G.N. and Dankert, J. (1998) cagA-positive Helicobacter pylori populations in China and The Netherlands are distinct. Infect. Immun. 66, 1822^1826. Ito, Y., Azuma, T., Ito, S., Miyaji, H., Hirai, M., Yamazaki, Y., Sato, F., Kato, T., Kohli, Y. and Kuriyama, M. (1997) Analysis and typing of the vacA gene from cagA-positive strains of Helicobacter pylori isolated in Japan. J. Clin. Microbiol. 35, 1710^1714. Miehlke, S., Kibler, K., Kim, J.G., Figura, N., Small, S.M., Graham, D.Y. and Go, M.F. (1996) Allelic variation in the cagA gene of Helicobacter pylori obtained from Korea compared to the United States. Am. J. Gastroenterol. 17, 1322^1325. Queiroz, D.M.M., Mendes, E.N. and Rocha, G.A. (1987) Indicator medium for isolation of Campylobacter pylori. J. Clin. Microbiol. 25, 2378^2379. Queiroz, D.M.M., Mendes, E.N., Rocha, G.A., Oliveira, A.M.R., Oliveira, C.A., Magalha‹es, P.P., Moura, S.B., Cabral, M.M.D.A. and Nogueira, A.M.M.F. (1998) cagA-positive Helicobacter pylori and risk for developing gastric carcinoma in Brazil. Int. J. Cancer 78, 135^139. van Doorn, L.-J., Figueiredo, C., Sanna, R., Plaisier, A., Schneeberger, P., de Boer, W. and Quint, W. (1998) Clinical relevance of cagA,
[18]
[19]
[20]
[21]
[22]
[23]
[24]
[25]
[26]
[27]
[28]
[29]
[30]
[31]
vacA, and iceA status of Helicobacter pylori. Gastroenterology 115, 58^66. Go, M.F., Cissell, L. and Graham, D.Y. (1998) Failure to con¢rm association of vacA gene mosaicism with duodenal ulcer disease. Scand. J. Gastroenterol. 33, 132^136. Yamaoka, Y., Kodama, T., Gutierrez, O., Kim, J.G., Kashima, K. and Graham, D.Y. (1999) Relationship between Helicobacter pylori iceA, vacA, and cagA status and clinical outcome: studies in four di¡erent countries. J. Clin. Microbiol. 37, 2274^2279. Fox, J.G., Dewhirst, F.E., Fraser, G.J., Paster, B.J., Shames, B. and Murphy, J.C. (1994) Intracellular Campylobacter-like organism from ferrets and hamsters with proliferative bowel disease is a Desulfovibrio sp.. J. Clin. Microbiol. 32, 1229^1237. Clayton, C.L., Kleanthous, H., Coates, P.J., Morgan, D.D. and Tabaqchali, S. (1992) Sensitive detection of Helicobacter pylori by using polymerase chain reaction. J. Clin. Microbiol. 30, 192^200. Atherton, J.C., Cover, T.L., Wells, R.J., Morales, M.R., Hawkey, C.J. and Blaser, M.J. (1999) Simple and accurate PCR-based system for typing vacuolating cytotoxin alleles of Helicobacter pylori. J. Clin. Microbiol. 37, 2979^2982. Peek, R.M., Miller, G.G., Tham, K.T., Perez-Perez, G.I., Zhao, X., Atherton, J.C. and Blaser, M.J. (1995) Heightened in£ammatory response and cytokine expression in vivo to cagAþ Helicobacter pylori. Lab. Invest. 71, 760^770. Bassam, B.J., Caetano-Anlolle¤s, G. and Gressho¡, P.M. (1991) Fast and sensitive silver staining of DNA in polyacrylamide gels. Anal. Biochem. 196, 80^83. Rocha, J.L.F. (1995) O receptor V2 de vasopressina e o canal de a¤gua aquaporina-2 no diabetes insipidus nefroge“nico. Master’s Thesis. Instituto de Cie“ncias da Biolo¤gicas da UFMG, Belo Horizonte, Brasil. van Doorn, L.-J., Figueiredo, C., Me¤graud, F., Pena, S., Midolo, P., Queiroz, D.M.M., Carneiro, F., Vanderborght, B., Pegado, M.G.F., Sanna, R., de Boer, W., Schneeberger, P.M., Correa, P., Ng, E.K.W., Atherton, J., Blaser, M.J. and Quint, W.G.V. (1999) Geographic distribution of vacA allelic types of Helicobacter pylori. Gastroenterology 116, 823^830. Nogueira, C., Figueiredo, C., Carneiro, F., Gomes, A.T., Barreira, R., Figueira, P., Salgado, C., Belo, L., Peixoto, A., Bravo, J.C., Bravo, L.E., Realpe, J.L., Plaisier, A.P., Quint, W.G., Riz, B., Correa, P. and van Doorn, L.-J. (2001) Helicobacter pylori genotypes may determine gastric histopathology. Am. J. Pathol. 158, 647^654. Kidd, M., Lastovica, A.J., Atherton, J.C. and Louw, J.A. (1999) Heterogeneity in the Helicobacter pylori vacA and cagA genes: association with gastroduodenal disease in South Africa? Gut 45, 499^ 502. Morales-Espinosa, R., Castillo-Rojas, G., Gonzalez-Valencia, G., Ponce de Leo¤n, S., Cravioto, A., Atherton, J.C. and Lo¤pez-Vidal, Y. (1999) Colonization of Mexican patients by multiple Helicobacter pylori strains with di¡erent vacA and cagA genotypes. J. Clin. Microbiol. 37, 3001^3004. Ashour, A.A.R., Collares, G.B., Mendes, E.N., Gusma‹o, V.R., Queiroz, D.M.M., Magalha‹es, P.P., Carvalho, A.S.T., Oliveira, C.A., Nogueira, A.M.M.F., Rocha, G.A. and Rocha, A.M.C. (2001) iceA genotypes of Helicobacter pylori strains isolated from Brazilian children and adults. J. Clin. Microbiol. 39, 1746^1750. van Doorn, L.-J., Schneeberger, P.M., Nouhan, N., Plaisier, A.P., Quint, W.G.V. and de Boer, W.A. (2000) Importance of Helicobacter pylori cagA and vacA for the e⁄cacy of antibiotic treatment. Gut 46, 321^326.
FEMSIM 1412 24-6-02