Analysis of rdxA and involvement of additional genes encoding NAD(P)H flavin oxidoreductase (FrxA) and ferredoxin-like protein (FdxB) in metronidazole resistance of Helicobacter pylori.

Metronidazole (Mtz) is a critical ingredient of modern multidrug therapies for Helicobacter pylori infection. Mtz resistance reduces the effectiveness of these combinations. Although null mutations in a rdxA gene that encodes oxygen-insensitive NAD(P)H nitroreductase was reported in Mtz-resistant H. pylori, an intact rdxA gene has also been reported in Mtz-resistant H. pylori, suggesting that additional Mtz resistance mechanisms exist in H. pylori. We explored the nature of Mtz resistance among 544 clinical H. pylori isolates to clarify the role of rdxA inactivation in Mtz resistance and to identify another gene(s) responsible for Mtz resistance in H. pylori. Mtz resistance was present in 33% (181 of 544) of the clinical isolates. There was marked heterogeneity of resistance, with Mtz MICs ranging from 8 to >/=256 microg/ml. rdxA inactivation resulted in Mtz MICs of up to 32 microg/ml for 6 Mtz-sensitive H. pylori strains and 128 microg/ml for one Mtz-sensitive strain. Single or dual (with rdxA) inactivation of genes that encode ferredoxin-like protein (designated fdxB) and NAD(P)H flavin oxidoreductase (frxA) also increased the MICs of Mtz for sensitive and resistant strains with low to moderate levels of Mtz resistance. fdxB inactivation resulted in a lower level of resistance than that from rdxA inactivation, whereas frxA inactivation resulted in MICs similar to those seen with rdxA inactivation. Further evidence for involvement of the frxA gene in Mtz resistance included the finding of a naturally inactivated frxA but an intact rdxA in an Mtz-resistant strain, complementation of Mtz sensitivity from an Mtz-sensitive strain to an Mtz-resistant strain or vice versa by use of naturally inactivated or functional frxA genes, respectively, and transformation of an Mtz-resistant Escherichia coli strain to an Mtz sensitive strain by a naturally functional frxA gene but not an inactivated frxA gene. These results are consistent with the hypothesis that null mutations in fdxB, frxA, or rdxA may be involved in Mtz resistance.