Helicobacter pylori cagA Status and s and m Alleles of vacA in Isolates from Individuals with a Variety of H. pylori-Associated Gastric Diseases (original) (raw)

Abstract

The cagA gene was detected in 100% of 16 Helicobacter pylori isolates from patients with gastric carcinoma versus 78% of 18 isolates from patients with duodenal ulcers (P = 0.344) and only 64% of 22 isolates from patients with gastritis only (P = 0.005) in Brazil. Also, there was a significant association between isolation of cagA+ s1-type vacA H. pylori in cases of stomach cancer and ulcers as opposed to cases of gastritis only (P = 0.004), but this was not true in Houston (P = 0.238), where 94% of all isolates were cagA+.

As defined by Atherton et al. (1), Helicobacter pylori vacA genes exhibit allelic mosaicism such that each gene has one of two different signal sequences (s1 and s2) and also one of two alternative sequences in the middle region of the gene (m1 and m2), so that four different combinations are possible. The presence of cagA (cytotoxin-associated gene A) correlates with the presence of the s1 m1 vacA genotype, while the absence of cagA correlates with the s2 m2 vacA genotype; isolates with the s1 m2 vacA genotype are frequently but not always cagA+ (1). The significance of this relationship is that s1 m1 VacA is highly active, s2 m2 VacA is inactive, and s1 m2 VacA has no or only weak cytotoxic activity. Individuals with peptic ulcer disease, duodenal ulcer disease, functional dyspepsia, or gastric adenocarcinoma are more likely to be infected with CagA+ VacA+ (type I) H. pylori, while those with chronic active gastritis alone are more likely to carry CagA− VacA− (type II) H. pylori (2–4, 8, 19, 20). However, there seems to be no functional link between cagA and VacA (5, 6, 20), and it is likely that cagA is a genotypic marker for the presence and/or expression of other ulcer- or cancer-related virulence genes (6). Of potential significance in terms of H. pylori antigenic variation is the so-called “variable” region within cagA, the origin and function of which are obscure (7, 10). Repetition of 15- to 99-base nucleotide sequences and variation in the number of codons, particularly at the proximal end of the cagA variable region (12), result in CagA proteins which vary in both size and amino acid sequence (7, 12, 21).

The present study was prompted by observations that the proportion of H. pylori isolates which are cagA+ varies from one geographic region to another: it is reportedly 100% in Japan (13), Korea (14), and China (15), lower in the United States and Europe (3, 8, 17–19), and as low as 41% in Canada (16). The main goals were to compare the prevalence of cagA at two widely separated locations, to define the relationship between cagA status and the vacA s and m alleles in these isolates, and to compare H. pylori isolates from patients with gastritis only with those from subjects with ulcers or gastric carcinoma.

In this study, two collections of H. pylori isolates, one from Houston, Texas, and the other from the state of Minas Gerais in southern Brazil, were examined for cagA and vacA status. Test isolates originated as single colonies from antral biopsies. The ethnic backgrounds of the subjects in Houston were as follows: mixed European, 14; African, 10; Hispanic, 6; and Korean descent, 4. The subjects in Brazil all belonged to a non-Portuguese-speaking rural population of native Indian mixed with European and/or African descent. The procedures used to obtain gastric biopsies were approved by the Institutional Review Boards for Human Research of Baylor College of Medicine and the Veterans Affairs Medical Center, Houston, and Universidad Federal de Minas Gerais, Belo Horizonte, Brazil.

First, we tested purified DNA preparations from all isolates for reliability in PCR assays by performing PCR assays as follows: for napA and for hpaA with previously described primer sets (9, 11) and for a constant; i.e., nonvariable, region near the 3′ end of cagA with primers described by Bukanov and Berg (5). These forward and reverse primers, CagA/Con-F and CagA/Con-R in Table 1, amplify a 402-nucleotide (nt) fragment extending from nt 3568 to nt 3969. The variable region of cagA extends from nt 3171 to nt 3542 (7, 12), and the coding region of cagA ends at nt 3974. All 34 of the Houston samples and all 63 of the Brazil samples gave the expected hpaA PCR product. All of the Houston samples and 61 of 63 Brazil samples gave the expected napA PCR product. All except 2 (94%) Houston samples gave the expected cagA PCR product, as did 49 (78%) of 63 Brazil samples.

TABLE 1.

Oligonucleotide primers used in this work

Designation Sequence
CagA/Con-Fa 5′-GTGCCTGCTAGTTTGTCAGCG-3′
CagA/Con-Ra 5′-TTGGAAACCACCTTTTGTATTAGC-3′
CAGVar-1F 5′-GCATCAAAAGGGAATTGTCTG-3′
CAGVar-1R 5′-GCGTGTGTGGCTGTTAGTAG-3′
CAGVar-2F 5′-AACAAGGATTTGAAAGACTTTAGT-3′
CAGVar-2R 5′-GATCGCTCCATTTTGGATATTG-3′

Because of the potential importance of the cagA variable region, PCR assays were performed on DNA from all Houston isolates with two different sets of primers encompassing this region of the gene. The first set, forward and reverse primers CAGVar-1F and CAGVar-1R (Table 1), was designed to amplify a sequence from nt 2566 to nt 3624, with products of at least 1,059 nt. The second set, forward and reverse primers CAGVar-2F and CAGVar-2R (Table 1), was designed to amplify a sequence from nt 2604 to nt 3654, with products of at least 1,051 nt. The oligonucleotide primers were synthesized and purified at the Molecular Genetics Core Facility of Baylor College of Medicine. Of 34 Houston samples, 20 were positive with both sets of CAGVar primers, 2 were negative in both assays, 7 were positive with the CAGVar-1 primers and negative with the CAGVar-2 primers, and 5 were negative with the CAGVar-1 primers and positive with the CAGVar-2 primers. In summary, 32 (94%) of 34 Houston H. pylori isolates were found to be cagA+.

By random selection, 34 of the 63 H. pylori isolates from Brazil were assayed with the CAGVar-1 primers and 29 were assayed with the CAGVar-2 primers. Of the 49 isolates which were positive for the constant region of cagA, 15 were PCR positive and 5 were PCR negative with the CAGVar-1 primer set; 27 were PCR-positive and 2 were PCR-negative with the CAGVar-2 primer set. By chance all 14 Brazil isolates which were negative for the constant region of cagA were PCR assayed with the CAGVar-1 primers, and all of them were also negative in the assay.

PCR assays were performed to distinguish between the s1 and s2 alleles of vacA and to distinguish between the m1 and m2 alleles, using the allele-specific primers described by Atherton et al. (1). Although seven Brazil samples could not be assayed for vacA alleles, all of the vacA assays produced unambiguous results. Table 2 presents the cagA and vacA genotypes of the isolates according to source and gastric disease type. H. pylori isolated in Houston is predominately cagA+, with the gene being as prevalent among isolates from patients with gastritis only as among isolates from patients with gastric or duodenal ulcers. Of the 32 cagA+ Houston isolates, 23 were s1 m1, 8 were s1 m2, and only 1 was s2 m2, showing the expected relationship between the vacA s1 allele and cagA positivity. Both of the _cagA_-negative Houston isolates have the s2 m2 allelic type of vacA. Thus, these results confirm the cagA-vacA allele relationship (2).

TABLE 2.

cagA status versus vacA s and m alleles in H. pylori isolated from patients with various H. pylori associated diseases

Source Disease Total no. of isolates cagA status (no.)a No. of isolates with:
s1 m1 s1 m2 s2 m1 s2 m2
Brazil Gastric carcinoma 16 + (16) 14 2 0 0
− (0) 0 0 0 0
Duodenal ulcer 18 + (14) 13 1 0 0
− (4) 3 0 0 1
Gastritis 22 + (14) 7 5 1 1
− (8) 0 0 0 8b
Houston Gastric ulcer 8 + (7) 6 1 0 0
− (1) 0 0 0 1
Duodenal ulcer 13 + (13) 8 5 0 0
− (0) 0 0 0 0
Gastritis 13 + (12) 9 2 0 1
− (1) 0 0 0 1

In summary, the cagA status of H. pylori isolates in Brazil correlates with gastric disease type, since 100% of 16 isolates from gastric carcinoma patients, 78% of 18 isolates from duodenal ulcer patients, and 64% of 22 isolates from patients with gastritis only are cagA+ (Table 2). Also, all but 2 of 44 cagA+ Brazil isolates have the s1 allele of vacA, predominately s1 m1. The two exceptions were isolated from patients with gastritis only, and thus, type I H. pylori was isolated from only 55% of those cases. In Houston, 31 of 32 cagA+ isolates had the s1 allele of vacA. These data indicate that the cagA status of H. pylori does not perfectly correlate with disease type in either Brazil or Houston, but possession of cagA does show a significant (P = 0.005) correlation with severe gastric disease outcome, as opposed to gastritis only, in Brazil. Finally, these data (Table 2) indicate that the s2 allele of vacA is carried by the more benign strains of H. pylori, since 12 of 14 (86%) isolates carrying the allele, including 2 cagA+ isolates, were from patients with gastritis only.

Acknowledgments

This work was supported in part by U.S. Public Health Service grant AI40506 (D.G.E.) from the National Institutes of Health and by funding from the U.S. Department of Veterans Affairs (D.J.E.) and from CNPq and FAPEMIG/Brazil (D.M.M.Q. and E.N.M.).

REFERENCES