TEXT

Top Abstract Text References

Helicobacter pylori infection is one of the most common infections worldwide and is etiologically associated with chronic gastritis, duodenal ulcers, gastric ulcers, gastric adenocarcinoma, and primary gastric lymphoma (3, 12, 13). Clinical experience has demonstrated that the eradication of H. pylori from infected patients is not easy and that the difficulty is mostly due to the lack of patient compliance with drug regimens and the development of antibiotic-resistant H. pylori (4). Tetracyclines are a family of broad-spectrum antibiotics that have been widely used for the treatment of bacterial infections since the 1950s. Extensive and widespread therapeutic use of tetracyclines in human and veterinary medicine and their use as growth promoters in animal feeds have resulted in tetracycline resistance in almost all bacterial genera, which has reduced the therapeutic usefulness of the tetracyclines (2, 14, 15, 16). For H. pylori, two cases of tetracycline-resistant isolates have been reported (6, 9), but no further information is currently available.

In order to understand the prevalence of antibiotic resistance among clinical H. pylori isolates, we have performed surveillance of antibiotic resistance using H. pylori obtained from Korean (n = 460) and Japanese (n = 105) patients since 1994. The patients from Japan had not received any previous H. pylori therapies (i.e., pretreatment isolates), but the patients from Korea received dual or triple anti-H. pylori therapies (i.e., posttreatment isolates). H. pylori strains were routinely culture on brain heart infusion (BHI) (Difco, Detroit, Mich.) agar plates and maintained as described previously (7). MIC measurements for metronidazole, clarithromycin, tetracycline, and amoxicillin were performed as described earlier (7). The resistance breakpoints used for metronidazole and clarithromycin were MICs of >8 (11) and >1 µg/ml (8), respectively. Since the resistance breakpoints for tetracycline and amoxicillin are not established for H. pylori, we defined resistance as MICs of >2 µg/ml for tetracycline and >8 µg/ml for amoxicillin (9).

All the tetracycline-resistant isolates (7 of 105; 6.7%) from the Japanese patients and 20 of the tetracycline-resistant isolates (22 of 460; 4.9%) from the Korean patients remained stable during stability tests (Table 1). However, two Korean resistant strains reverted to sensitive, as was previously shown for other antibiotic resistances in H. pylori (6). The proportion of metronidazole-resistant H. pylori isolates was higher for Korean patients (40.4%) than for Japanese patients (23.8%), whereas clarithromycin resistance was more common in isolates from Japanese patients (15.2%) than in isolates from Korean patients (5.3%). Amoxicillin resistance was not seen in isolates from patients from either country. The higher frequency of metronidazole resistance and lower frequency of clarithromycin resistance in the isolates from Korean patients than in the isolates from Japanese patients probably reflect the relative use of those agents in the two countries. Of interest, all of the 29 tetracycline-resistant strains were also resistant to metronidazole (MIC  8 µg/ml) (Table 1). The cross-resistance of tetracycline and metronidazole was confirmed by culturing the cells on 5% horse blood BHI agar plates containing 2 to 8 µg of tetracycline per ml and 4 to 16 µg of metronidazole per ml for 3 days.

To examine whether the tetracycline resistance determinant(s) could be transferred among clinical H. pylori isolates, total genomic DNA purified from two tetracycline-resistant H. pylori strains (KH84B and KH179A) (MICs = 16 µg/ml) was introduced into two tetracycline-sensitive H. pylori type strains, ATCC 43629 and ATCC 700392 (MIC = 0.5 µg/ml), by natural transformation as described by Haas et al. (5). To avoid the selection of spontaneous tetracycline-resistant colonies, recipient cells for natural transformation were adjusted to approximately 10⁷ cells/ml with 1 to 3 µg of genomic DNA and spread on BHI agar plates containing 2 µg of tetracycline per ml. In parallel, a negative control was also included in the same experiment by using the same number of cells without adding DNA. The transformation frequencies were 1.2 × 10⁵ (ATCC 700392 with genomic DNA from KH84B), 0.7 × 10⁵ (ATCC 700392 with genomic DNA from KH179A), 0.2 × 10⁵ (ATCC 43629 with genomic DNA from KH84B), and 0.5 × 10⁵ (ATCC 43629 with genomic DNA from KH179A). However, the negative control did not show any colonies. The phenotypic nature of the transformed strains was analyzed using 12 colonies of each strain. Forty transformed colonies (10 colonies from each transformed ATCC 700392 and ATCC 43629 strain) were used to measure the MICs of tetracycline, metronidazole, amoxicillin, and clarithromycin. The tetracycline MICs for all of the transformed colonies were 16 µg/ml, which was identical to those for the parental strains. Importantly, the metronidazole resistance phenotype was also transferred to the two type strains (also requiring MICs of 16 µg/ml). There was no change in susceptibility to amoxicillin and clarithromycin compared to that of the parental strains (Table 2).

Patients from both countries had relatively high frequencies of tetracycline-resistant H. pylori (4.9 and 6.7% for Korean and Japanese patients, respectively). The MICs of tetracycline for the 29 strains with stable tetracycline resistance were 4 to 16 µg/ml. However, when the strains grew on low-level tetracycline-containing agar plates before the MICs were measured, the MICs increased more than twofold, suggesting that tetracycline-resistant strains would compromise therapies containing tetracycline. All the tetracycline-resistant strains showed cross-resistance to metronidazole, requiring MICs ranging from 8 to 32 µg of metronidazole per ml, and the metronidazole resistance remained stable. In addition, the tetracycline resistance from the clinical isolates was always transferred together with metronidazole resistance to the tetracycline-sensitive type strains and remained stable, as shown for the parental tetracycline-resistant strains. These observations imply that resistance to tetracycline and metronidazole from the tetracycline-resistant H. pylori clinical isolates is transferable to the sensitive H. pylori strains and also suggest that tetracycline resistance may be associated with metronidazole resistance but not vice versa. Interestingly, tetracycline-resistant H. pylori strains isolated by other investigators also showed cross-resistance to metronidazole (1, 9). The metronidazole resistance mechanism has been reported to occur with alterations in the rdxA, frxA, and/or fdxB gene of H. pylori (7). The mechanism of tetracycline resistance has not been reported for H. pylori, although the tetracycline resistance mechanism has been extensively studied for E. coli and other bacteria (2, 14, 15, 16). It is not clear whether the cross-resistance mechanism is due to a known metronidazole resistance mechanism and an unknown tetracycline resistance mechanism or if it is a part of multidrug resistance mechanisms, as in other gram-negative bacteria (10). In any case, the emergence of a new transferable antibiotic resistance among clinical isolates represents a major threat to current H. pylori eradication therapies. Currently, we are attempting to elucidate the possible mechanism(s) for the dual antibiotic resistance seen in these strains.