Journal of Antimicrobial Chemotherapy (2002) 49, 675–678

JAC
Production of the RdxA protein in metronidazole-susceptible
and -resistant isolates of Helicobacter pylori cultured from
treated mice

Stephanie R. Lathama, Agnès Labigneb and Peter J. Jenksa,c*

a
Department of Medical Microbiology, Royal Free and University College Medical School,
Rowland Hill Street, London, UK; bUnité de Pathogénie Bactérienne des Muqueuses, Institut Pasteur,
28 Rue du Dr Roux, Paris, France; cInstitute of Infections and Immunity, University of Nottingham,
Nottingham NG7 2UH, UK

The objective of this study was to use immunoblotting with RdxA antisera to examine the
production of the RdxA protein in mouse-derived metronidazole-susceptible and -resistant
isolates of Helicobacter pylori. A 24 kDa immunoreactive band corresponding to RdxA was
observed in all 15 metronidazole-susceptible and five of 50 metronidazole-resistant isolates.
The rdxA gene of these five isolates contained missense mutations and transformation
experiments confirmed that these mutations were associated with inactivation of the rdxA
gene. No RdxA protein was produced in the other 45 metronidazole-resistant strains, including
one in which the nucleotide sequence of the rdxA gene was unchanged. These results
demonstrate a high correlation between production of the RdxA protein and susceptibility of
H. pylori to metronidazole. Testing for the absence of the RdxA protein identifies the majority
of strains that will respond poorly to metronidazole-containing eradication regimens.

Introduction Materials and methods

Up to 70% of clinical strains of Helicobacter pylori isolated
in western Europe are resistant to the 5-nitroimidazoles
and this prevalence is far higher in developing countries.1
The majority of resistance is associated with mutational
inactivation of the rdxA gene, which encodes an oxygen-
insensitive NADPH nitroreductase.2 Inactivation of other
reductase-encoding genes, including frxA and fdxB are also
associated with resistance and seem to be implicated in the
transition to high-level resistance.3–5 Because resistance is
associated with multiple changes within rdxA and possibly
other reductase-encoding genes, it has not been possible to
develop genotype-based assays capable of detecting metro-
nidazole resistance in this important gastric pathogen. We
have recently used immunoblot analysis with specific anti-
RdxA antibody to demonstrate that there is a high cor-
relation between production of the RdxA protein and
susceptibility of H. pylori to metronidazole.6
Bacterial strains, growth conditions and susceptibility
testing
Metronidazole-susceptible and -resistant derivatives of
SS1 (Table) were cultured from the stomachs of mice that
had been experimentally infected with the metronidazole-
susceptible SS1 strain and then treated orally with metro-
nidazole.7 H. pylori strains were routinely cultured on 10%
horse blood medium supplemented with appropriate anti-
biotics as previously described.6 Susceptibility to metro-
nidazole was assessed by agar dilution determination of the
MIC.7 The MIC was defined as the lowest concentration of
antibiotic inhibiting growth when the plates were read after
72 h incubation under microaerobic conditions at 37C.
Isolates were considered resistant to metronidazole if the
MIC was 8 mg/L.

*Correspondence address. Institute of Infections and Immunity, Floor C, West Block, University Hospital,
Queen’s Medical Centre, Nottingham NG7 2UH, UK. Tel: 44-115-924-9924 ext. 42457; Fax: 44-115-970-9923;
E-mail: Peter.Jenks@nottingham.ac.uk

675
© 2002 The British Society for Antimicrobial Chemotherapy
 S. R. Latham et al.

Metronidazole MIC (mg/L) DNA sequencing

Genomic DNA from individual H. pylori strains was ex-
tracted using the QIAamp Tissue Kit (Qiagen, Crawley,
UK) according to the manufacturer’s instructions. DNA
sequencing of the rdxA gene was carried out using two
0.0625

0.0625

8–64
pairs of oligonucleotide primers, rdxA-1 (CGTTAGG-
GATTTTATTGTATGCTAC) and rdxA-2 (CCCCACA-
1

16

8
16
8
16

16
16
32
16
32
GCGATATAGCATTGCTC), and rdxA-3 (GTTAGAG-
TGATCCCCTCTTTTGCTC) and rdxA-4 (CACCCC-
TAAAAGAGCGATTAAAACC). These were used to
amplify two overlapping PCR products that contained the
entire rdxA gene, and the nucleotide sequences of these

shift at 157, stop at 168

products were determined on both strands using the four
shift at 65, stop at 73
shift at 73, stop at 77
shift at 64, stop at 77
shift at 87, stop at 89
amino acid change

oligonucleotide primers described above.

Y47H and P51L

Protein analysis by SDS–PAGE and

unknown

immunoblotting
rdxA allele

A67V
P51L
P51L
none

none
none

Immunoblotting with polyclonal rabbit antisera against

RdxA was carried out as previously described.6 Briefly, cell
–
Table. H. pylori strains used in this study

extracts prepared from 2 day cultures were analysed on

nucleotide change

deletion of 114 bp

T to C and C to T

slab gels, comprising a 4.5% acrylamide stacking gel and a

17.5% resolving gel. Proteins were transferred to nitro-
T and AT

cellulose membranes and reacted with RdxA antisera that

unknowna

had been diluted 1:100 in 50% Escherichia coli TG1

C to T
C to T
C to T

extract, 5% milk powder and 0.2% Tween in phosphate-

none

none
none

2T
A

A

G

buffered saline and incubated for 4 h at room temperature

to remove non-specific antibodies to E. coli. Immuno-
reactants were detected with anti-rabbit peroxidase-linked
mouse-derived metronidazole-susceptible H. pylori SS1
mouse-derived metronidazole-resistant H. pylori SS1
mouse-derived metronidazole-resistant H. pylori SS1
mouse-derived metronidazole-resistant H. pylori SS1
mouse-derived metronidazole-resistant H. pylori SS1
mouse-derived metronidazole-resistant H. pylori SS1
mouse-derived metronidazole-resistant H. pylori SS1
mouse-derived metronidazole-resistant H. pylori SS1
mouse-derived metronidazole-resistant H. pylori SS1
mouse-derived metronidazole-resistant H. pylori SS1
mouse-derived metronidazole-resistant H. pylori SS1
mouse-derived metronidazole-resistant H. pylori SS1

immunoglobulin (Amersham, Little Chalfont, UK) diluted

1:10 000 and reaction products were visualized on auto-
radiographic film by chemiluminescence using the ECL
H. pylori SS1 isogenic deletion mutant in rdxA

Western blotting detection system (Amersham).

Natural transformation
Characteristic

Metronidazole-susceptible H. pylori strain SS1 was trans-

formed using a specific PCR fragment generated from
resistant strains using the oligonucleotide primers rdxA-1
and rdxA-4. Bacteria were inoculated as 1 cm patches and
H. pylori wild-type strain

grown for 5 h before addition of 30 ng of the PCR fragment.

After further incubation for 18 h, the bacteria from each
individual patch were harvested and plated directly on to a
single plate containing metronidazole 0.5 g/mL. After
4 days incubation, single colonies were obtained, and
the MIC of metronidazole for three independent trans-
formants from each transformation was determined.
rdxA gene not sequenced.

Results
SS1-26 to SS1-65
SS1-1 to SS1-15

Immunoblotting of metronidazole-susceptible and

SS1 rdxA–

-resistant isolates of H. pylori with anti-RdxA antibody

SS1-16
SS1-17
SS1-18
SS1-19
SS1-20
SS1-21
SS1-22
SS1-23
SS1-24
SS1-25
Strain

Immunoblotting with RdxA antisera was carried out on

SS1

whole bacterial cell lysates of 15 metronidazole-susceptible

a

676
 Production of RdxA in H. pylori
Figure. Immunoblot analysis of metronidazole-resistant isolates
of H. pylori SS1 using polyclonal antiserum to H. pylori RdxA.
Lane 1, H. pylori strain SS1. Lane 2, isogenic rdxA mutant in
strain SS1. Lanes 3–7, H. pylori strains SS1-22, SS1-21, SS1-20,
SS1-19 and SS1-18, respectively.
and 50 metronidazole-resistant isolates. The metronida-
zole-resistant strains included 10 in which the rdxA gene
had previously been sequenced: in nine the rdxA gene
contained one or more mutations (SS1-17 to SS1-25),
and in one the gene sequence was identical to that of the
parental SS1 strain (SS1-16).8 In all of the 15 metronidazole-
susceptible strains a 24 kDa immunoreactive band cor-
responding to the RdxA protein was observed. The 24 kDa
protein was also produced by five of the 50 metronidazole-
resistant H. pylori isolates: SS1-17, SS1-18, SS1-19, SS1-52
and SS1-56 (Figure). In three of these isolates, which were
isolated from the stomachs of different mice, the mutations
in the rdxA gene had been identified in a previous study:
SS1-17 contained Y47H and P51L amino acid substitutions;
SS1-18 and SS1-19 contained P51L amino acid substitu-
tions (Table).8 The nucleotide sequence of the rdxA gene
of the other two isolates, SS1-52 (isolated from the same
mouse as SS1-19) and SS1-56 (isolated from the same
mouse as SS1-17), was determined and found to contain the
missense point mutation (C to T at position 153) that
resulted in the P51L amino acid substitution. In the remain-
ing 45 metronidazole-resistant strains, including strain
SS1-16, which had an intact rdxA gene, no equivalent
immunoreactive band corresponding to the RdxA protein
was observed.
Further analysis of metronidazole-resistant H. pylori
isolates that produced the RdxA protein
In order to determine whether the mutations within the
rdxA gene of the metronidazole-resistant H. pylori isolates
that produced the RdxA protein were associated with
inactivation of rdxA, the rdxA gene of SS1-17, SS1-18,
SS1-19, SS1-52 and SS1-56 was amplified and transformed
into the metronidazole-susceptible, parental SS1 strain
(MIC 0.0625 mg/L). PCR fragments amplified from SS1
and SS1 rdxA mutant (MIC 1.0 mg/L) were used as con-
trols. The MIC of metronidazole for colonies isolated after
transformation with the rdxA allele of the SS1 isogenic
677
deletion mutant in rdxA (SS1 rdxA–) and strains SS1-17,
SS1-18, SS1-19, SS1-52 and SS1-56 was 0.5–2.0 mg/L. No
colonies were isolated on plates containing metronidazole
0.5 mg/L after transformation with the rdxA allele of strain
SS1.
Discussion
We have recently reported a high correlation between pro-
duction of the RdxA protein and susceptibility of H. pylori
to metronidazole in clinical strains isolated from different
geographical regions.6 In order to examine the factors that
determine production of RdxA, we used a series of well-
characterized metronidazole-resistant isolates cultured
from mice that were experimentally infected with the
metronidazole-susceptible H. pylori strain SS1 and then
treated orally with metronidazole.7 We found that the
RdxA protein is produced in all mouse-derived as well as
clinical metronidazole-susceptible strains so far tested. As
in our previous study, the RdxA protein was not produced
by the majority (90%) of resistant strains. Sequence analy-
sis of the rdxA gene of the five metronidazole-resistant
strains that did produce the 24 kDa RdxA protein revealed
that four contained missense mutations resulting in a P51L
amino acid substitution, and the fifth contained two muta-
tions giving a Y47H as well as the P51L substitution. The
finding that four isolates contained the same P51L amino
acid substitution is not explained by clonal expansion
of a single mutant, since these strains were isolated from
the stomachs of three different mice. Transformation of
metronidazole-susceptible H. pylori SS1 (MIC 0.0625 mg/L)
with the rdxA allele of these isolates generated trans-
formants with MICs similar to that of an SS1 isogenic dele-
tion mutant in rdxA (MIC 1.0 mg/L),4,9 indicating that the
rdxA gene was inactivated in these strains. This confirmed
that these mutant alleles were associated with inactivation
of rdxA, and indicates that these strains produced a full-
sized, but functionally inactive RdxA enzyme.
The rdxA gene of many metronidazole-resistant
H. pylori strains contain frameshift mutations that result in
the creation of a translational stop codon in the region
immediately downstream of the mutation, and such strains
would be predicted to produce a truncated RdxA pro-
tein.2,8,10 However, this study confirmed that production of
the RdxA protein is completely abrogated in the majority
of resistant strains, including five (SS1-21 to SS1-25) that
contained such frameshift mutations. The finding that the
RdxA protein was not produced by strain SS1-16 was par-
ticularly interesting, since this strain is one of a handful of
metronidazole-resistant strains that have been reported to
contain an intact rdxA gene.8 This implies that this isolate
contained a mutation within the promoter region of
the rdxA gene and is the first good evidence that muta-
tions within this region are associated with the resistant
phenotype.
 S. R. Latham et al.
The exact contribution of other genes to the acquisition
of metronidazole resistance by H. pylori is currently un-
clear. However, regardless of whether other resistance
mechanisms are present or not, our results demonstrate
that the vast majority of resistant strains contain mutations
within the rdxA gene or its promoter that prevent
production of the RdxA protein or result in production of
an abnormal polypeptide that is subsequently degraded.
This is an important finding, since it indicates that testing
for the absence of the RdxA protein would identify the
majority of isolates that will respond poorly to metronida-
zole-containing eradication regimens and has implications
for the development of assays capable of detecting metro-
nidazole resistance in H. pylori. One major advantage of
this approach is that it identifies all resistant strains carry-
ing mutations that affect expression of the rdxA gene,
including those not yet identified by nucleotide sequence
analysis.
Acknowledgement
Peter J. Jenks is supported by an Advanced Fellowship
for Medical, Dental and Veterinary Graduates from the
Wellcome Trust, UK (Ref. 061599).
References
1. Dunn, B. E., Cohen, H. & Blaser, M. J. (1997). Helicobacter
pylori. Clinical Microbiology Reviews 10, 720–41.
2. Goodwin, A., Kersulyte, D., Sisson, G., Veldhuyzen van Zanten,
S. J., Berg, D. E. & Hoffman, P. S. (1998). Metronidazole resistance
in Helicobacter pylori is due to null mutations in a gene (rdxA) that
encodes an oxygen-insensitive NADPH nitroreductase. Molecular
Microbiology 28, 383–93.
678
3. Jeong, J. Y., Mukhopadhyay, A. K., Dailidiene, D., Wang, Y.,
Velapatino, B., Gilman, R. H. et al. (2000). Sequential inactivation of
rdxA (HP0954) and frxA (HP0642) nitroreductase genes causes
moderate and high-level metronidazole resistance in Helicobacter
pylori. Journal of Bacteriology 182, 5082–90.
4. Jeong, J. Y. & Berg, D. E. (2000). Mouse-colonizing Helicobacter
pylori SS1 is unusually susceptible to metronidazole due to two
complementary reductase activities. Antimicrobial Agents and
Chemotherapy 44, 3127–32.
5. Kwon, D. H., El-Zaatari, F. A., Kato, M., Osato, M. S., Reddy, R.,
Yamaoka, Y. et al. (2000). Analysis of rdxA and involvement of addi-
tional genes encoding NAD(P)H flavin oxidoreductase (FrxA) and
ferredoxin-like protein (FdxB) in metronidazole resistance of Helico-
bacter pylori. Antimicrobial Agents and Chemotherapy 44, 2133–42.
6. Latham, S. R., Owen, R. J., Elviss, N. C., Labigne, A. & Jenks, P.
J. (2001). Differentiation of metronidazole-sensitive and -resistant
Helicobacter pylori by immunoblotting with antisera to the RdxA
protein. Journal of Clinical Microbiology 39, 3052–5.
7. Jenks, P. J., Labigne, A. & Ferrero, R. L. (1999). Exposure to
metronidazole in vivo readily induces resistance in Helicobacter
pylori and reduces the efficacy of eradication therapy in mice.
Antimicrobial Agents and Chemotherapy 43, 777–81.
8. Jenks, P. J., Ferrero, R. L. & Labigne, A. (1999). The role of the
rdxA gene in the evolution of metronidazole resistance in Helico-
bacter pylori. Journal of Antimicrobial Chemotherapy 43, 753–8.
9. Jenks, P. J., Ferrero, R. L., Tankovic, J., Thiberge, J. M. &
Labigne, A. (2000). Evaluation of nitrofurantoin combination therapy
of metronidazole-sensitive and -resistant Helicobacter pylori infec-
tions in mice. Antimicrobial Agents and Chemotherapy 44, 2623–9.
10. Tankovic, J., Lamarque, D., Delchier, J. C., Soussy, C. J.,
Labigne, A. & Jenks, P. J. (2000). Frequent association between
alteration of the rdxA gene and metronidazole resistance in French
and North African isolates of Helicobacter pylori. Antimicrobial
Agents and Chemotherapy 44, 608–13.
Received 13 August 2001; returned 6 December 2001; revised
2 January 2002; accepted 11 January 2002