Mechanism of clarithromycin resistance in clinical isolates of Helicobacter pylori

MICROBIOLOGY LETTERS

ELSEVIER

FEMS Microbiology Letters 142 (1996) 3742

Mechanism of clarithromycin

resistance in clinical isolates of

Helicobacter pylori Y.J. Debets-Ossenkopp a, M. Sparrius c, J.G. Kusters c, J.J. Kolkman b, C.M.J.E. Vandenbroucke-Grauls a~cj* a Department

of Clinical Microbiology

and Infection Control, University Hospital, ‘Vroe Universiteit’, P.O. Box 7057, 1007 MB Amsterdam, The Netherlands

’ Department

of

Gastroenterology,

’ Department

of Medical

University Hospital ‘Vrge Microbiology,

Universiteit’, 1007 MB Amsterdam,

‘Vrije Universiteit’, 1007 MB Amsterdam,

The Netherlands

The Netherlands

Received 10 May 1996; revised 6 June 1996; accepted 6 June 1996

Abstract Seventy-three Helicobacterpylori-positive patients were treated with a combination of clarithromycin and ranitidine in order to eradicate the bacterium. Eradication was successful in 79.5%. In 15 patients eradication failed, and in 11 cases this was due to clarithromycin resistance. In one patient the infecting strain was resistant at the onset of treatment, while in the remaining 10 patients resistance developed during therapy. These isolates had also become resistant to various other antibiotics. Random amplified polymorphic DNA and restriction fragment end-labeling analysis of the isolates showed close genetic relatedness between pre- and post-treatment isolates, indicating that resistance was the result of selection of variants of the infecting strain rather then infection with an exogenous resistant strain. Nucleotide sequence comparisons revealed that all resistant isolates had a single base pair mutation in the 23s rRNA. Since this single point mutation results in co-resistance to various antibiotics at high frequencies, caution should be taken when using clarithromycin as a single antibiotic. Keywords:

Antibiotic resistance; Erythromycin resistance methylase; Macrolide antibiotic; 23s rRNA mutation

1. Introduction Helicobacter pylori is an important in chronic

active gastritis

Clarithromycin

and peptic

is frequently

etiologic

used in eradication

apy of H. pylori [2]. Development

resistance

during

therapy

[ 11. ther-

of clarithromycin is a predominant cause of

* Corresponding author. Tel.: +31 (20) 444 0485; Fax: +31 (20) 444 0473; E-mail: [email protected] 0378-1097

factor

ulcer disease

therapy failure. Resistance to clarithromycin and related antibiotics has been studied in various other bacteria and appears to be due to single nucleotide mutations [3] or post-transcriptional methylation of the 23s rRNA [4]. The aim of this study was to investigate whether therapy failure resulted from selection of spontaneous resistant variants of the infecting strain or reinfection with an exogenous strain. Therefore we determined the genetic relationship between pre- and post-treatment isolates and the mechanism whereby these strains became resistant.

/96/$12.00 Copyright 0 1996 Federation of European Microbiological Societies. Published by Elsevier Science B.V PIISO378-1097(96)00239-X

38

Y .I. Dehets-Ossenkopp

2. Materials

et ul. I FEMS

and methods

2.1. Patients and routine growth H. pylori isolates Seventy-three patients with peptic ulcer disease and gastric mucosal biopsy positive for H. pylori were treated for 2 weeks with a combination of ranitidine and clarithromycin (four daily dosages of 150 mg and 250 mg, respectively). All patients underwent gastroscopy with antral biopsies before and 4-6 weeks after completion of therapy. Cultures were performed on Columbia agar plates supplemented with 10% lysed horse blood (CA+ plates) and on Belohorizonte agar [5]. Plates were incubated for 72 h at 37’C in an atmosphere of O#Z02/Ns (5 : 10 : 85%). H. pylori was identified by typical Gram-stain morphology and biochemical tests positive for urease, oxidase and catalase. Eradication was defined as the absence of histological signs of H. pylori infection and negative gastric mucosal biopsy cultures. 2.2. Determination of minimal inhibitory concentration

were prepared from a 48 h grown H. pyon CA’ plates. The minimal inhibitory concentration (MIC) was determined on CA+ plates inoculated with lo7 cfu in 20 ~1 of 0.9% NaCl with the E-test as described by Glupczynski et al. [6]. Inocula

lori culture

2.3. Extraction of genomic DNA Bacteria were harvested from 1 to 2 freshly grown, confluent CA+ plates, washed twice in phosphatebuffered salt solution (PBS) and resuspended in 25 mM Tris-HCl, pH 8.0 and 10 mM EDTA. The bacteria were disrupted by two cycles of freeze-thawing and genomic DNA was purified as described by Ausubel et al. [7]. 2.4. Random ampll$ed polymorphic DNA analysis PCR amplifications were performed with the Primezym DNA polymerase Kit (Biometra, Gottingen, Germany) according to the manufacturer’s instructions. Either the 18-mer degenerate primer pair REPlR-Dt and REP2-Dt [8] or the erm primer

Microbiology

Letters

142 (I 994) 3742

pair El and E2 [9] were used in these PCR reactions. Amplification was performed in a total volume of 50 ul with 20 ng of template genomic DNA, and either 100 pmol of each REP primer, or 150 pmol and 200 pmol of the El and E2 primers, respectively. PCR amplifications were performed in an automated thermal cycler (Pharmacia, Uppsala, Sweden) with 35 cycles of a 30 s denaturation step at 95°C a 30 s annealing step at 45°C and 30 s of extension at 72°C. These cycles were followed by one final 10 min extension step at 72°C. Aliquots of the PCRamplified products (5 ul) were resolved in a 2% agarose gel, containing 0.5 X THE, stained with ethidium bromide and visualized under a short wave-length ultraviolet light source. 2.5. H. pylori 23s rRNA ampkjication and sequencing

The H. pylori 23s rRNA sequence from the Genbank Database [lo] was used to design the oligonucleotide primers 5’-GCG TTG AAT TGA AGC CCG AGT AAA C-3’ and 5’-TGT GTG CTA CCC AGC GAT GCT C-3’ that would allow amplification of an 850 bp fragment of the 23s rRNA gene. PCR amplification was as with RAPD except that only 30 cycles were performed and that the temperature for the annealing step was 59°C. The presence of a single 850 bp amplicon was verified by resolving a 5 ~1 aliquot of the PCR reaction on a 1% agarose gel. The remainder of the 850 bp fragment was purified from the PCR-amplification reaction using the Wizzard DNA Cleanup System (Promega, Madison, WI) and the purified DNA was then used as template in a sequencing reaction with the Sequenase 7-deaza-dGTP sequencing kit (Amersham, Buckinghamshire, UK). 2.6. Restriction fragment

end-labeling

One microgram genomic DNA was digested with restriction endonuclease EcoRI for 2 h and the resulting DNA fragments were end-labeled as described by van Steenbergen et al. [l 11. Excess unlabeled dATPs were added and the reaction was allowed to continue for an additional 15 min at room temperature. Upon incubation at 65°C for 10 min to inactivate the Klenow DNA polymerase, the

Y.J. Debets-Ossenkopp et al. I FEMS Microbiology Letters 142 (1996) 37-42

39

250

Fig. 1. Ethidium bromide-stained agarose gel showing REP-primer-derived RAPD fingerprints of paired isolates from 9 of the 10 patients (Roman numbers) with clarithromicin-sensitive pre-treatment (lane A) and resistant post-treatment isolates (lane B). Lane C: control, no template added. Lane M: molecular mass marker.

labeled EcoRI fragments were further digested with HindIII. Unincorporated label was removed by precipitation with ethanol, the labeled DNA fragments were separated by electrophoresis on a 6% polyacrylamide gel and visualized by autoradiography [l 11.

togramin type-B. Only one of the 73 infecting strains was resistant to clarithromycin at the onset of treatment, but 11 of the 15 post-treatment isolates were resistant. Thus, in 10 patients resistance developed during therapy. Susceptibility and resistance to clarithromycin always coincided with those for the other antibiotics.

3. Results 3.2. Genetic relationship 3.1. Clarithromycin susceptibility of pre- and post-treatment H. pylori isolates Eradication failed in 15 of 73 patients treated in our clinic with a combination of the antibiotic clarithromycin and the Hz-receptor antagonist ranitidine. Pre- and post-treatment mucosal gastric biopsies from all 73 patients were tested for the presence of H. pylori and its susceptibility to clarithromycin, azithromycin, erythromycin, clindamycin and strep-

To establish whether the resistance that had developed during treatment originated from clonal selection of a resistant variant of the endogenous infecting strain or from infection with an exogenous strain we determined the genetic relationship of the paired pre- and post-treatment isolates. In one patient this genetic relationship could not be determined because the frozen stock of the pre-treatment isolate was no longer viable. The genetic relationship of the other 9

40

Y.J. Deb&s-Ossenkopp

et al. I FEMS Microbiology Letters 142 11996) 37-42

paired isolates to RFEL analysis. Representative data of four paired isolates are presented in Fig. 2. There was a marked difference in banding pattern among isolates from different patients, while only minor differences were observed between pre- and post-treatment isolates from a single patient. 3.3. Molecular mechanisms underlying clarithromycin resistance Co-resistance to the tested antibiotics in other bacteria is generally due to methylation of a specific adenine residue of the 23s rRNA. We tested our strains for the presence of erythromycin resistance methylase (erm) genes using the erm-specific primers El and E2 [9]. No erm-specific amplification product was observed in any of these strains (data not shown). Next we investigated whether the resistance resulted from mutations in the clarithromycin-binding regions of the 23s rRNA. An 850 bp fragment of the 23s rRNA gene was PCR-amplified and the nucleotide sequence of the resulting was determined. Nucleotide sequence comparison of the paired isolates revealed that five out of the 10 post-treatment clarithromycin-resistant isolates contained an A to G transitional mutation at position 2514 (numbering according to ref. [lo]). The other five isolates all showed an A to G transitional mutation at position 2515.

4. Discussion

Fig. 2. Autoradiograph of polyacrylamide gel electrophoresis of restriction endonuclease-digested [a-32P]dATP end-labeled genomic DNA fragments of paired H. pyhi isolates from four representative patients (Roman numbers) with Cla’ pretreatment (lane A) and ClaR post-treatment (lane B) H. pylori isolates.

paired isolates was determined by RAPD analysis using two different primer sets. Representative results obtained with the REP-primers are shown in Fig. 1. Data obtained with erm primers are not shown. To obtain further evidence for the genetic relatedness we subjected the genomic DNA of all

Even in the acidic environment of the stomach the semi-synthetic macrolide clarithromycin is extremely active against H. pylori [12]. For this reason clarithromycin in combination with ranitidine is often used as an eradication regimen for H. pylori infections. However, development of resistance to clarithromycin during therapy occurs at high frequency and is the predominant reason for therapy failure. In a variety of other microorganisms a common mechanism for resistance to clarithromycin and related antibiotics is methylation of a specific adenine residue in 23s rRNA by so-called erythromycin resistance methylases (41. Methylation of this adenine residue also confers resistance to lincosamide and streptogramin type-B antibiotics. This is commonly

Y.J. Debets-Ossenkopp et al. IFEMS Microbiology Letters I42 (1996) 37-42

referred to as the MLS-phenotype resistance. Although all 10 of our clarithromycin-resistant posttreatment isolates showed this MLS phenotype, no erm gene-specific amplification product was observed. From this we concluded that it is unlikely that erm genes are present in these isolates. From the combined RAPD data of both primers sets and RFEL analysis we conclude that clarithromytin resistance resulted from clonal selection of a resistant variant rather then from reinfection with an exogenous clarithromycin-resistant strain. There are two domains on 23s rRNA involved in binding of macrolide antibiotics [13], and resistance seems confined to alterations by methylation or transitional mutations in one of these two regions [3,14,15]. Nucleotide sequence analysis of the 23s rRNA of our clarithromytin-resistant H. pylori isolates revealed that: (1) transitional mutation was confined to one region only, and (2) transitions were always from A to G. This is in accordance with the findings of Versalovic et al. [16] who recently showed that in 7 patients with clarithromycin-resistant H. pylori, clarithromycin resistance was also due to mutations in the same 23s rRNA region. It is striking that in all 10 of our clarithromytin-resistant isolates and the 7 isolates from the study of Versalovic et al. the mutations were always from A to G and never from A to U or C. Our post-treatment isolates were obtained 4-6 weeks after clarithromycin treatment and strain isolation did not involve the use of clarithromycin. Since our sequence data of the 850 bp 23s rRNA amplicons did not show any indications for the presence of the wild-type allele we presume that there is no significant reversion to the wild-type allele. The observed in vivo stability of the clarithromycin resistance is in contrast with the findings of Hua-Xiang Xia et al. [17] who found that 9 out of 20 clarithromycin-resistant H. pylori isolates lost their clarithromycin resistance in vitro as a result of repetitive subculture, unless the four isolates that we found sensitive both before and after treatment represent isolates that reverted due to instability of the resistance. In this study the prevalence of clarithromycin resistance in H. pylori in pre-treatment isolates was low (l/73; i.e. < 1.5%). However, because the new clarithromycin and related antibiotics like azithromycin are prescribed with increasing frequency for

41

respiratory tract infections and sexually transmitted diseases we expect an increase of clarithromycin-resistant H. pylori isolates. Therefore we would like to stress the importance of performing a sensitivity assay prior to starting therapy. The finding that the clarithromycin resistance is confined to a simple base mutation in one of two adjacent nucleotides allows for a rapid PCR-based detection of clarithromycin resistance in H. pylori. Since a single point mutation confers resistance to clarithromycin it is not surprising that resistance develops at such high frequencies (lo/73 = 13.7%). Therefore one should be careful when using clarithromycin as a single antibiotic in eradication therapy for H. pylori infection.

Acknowledgments Dr. S.G.M. Meuwissen is acknowledged for critical reading of the manuscript. J.G. Kusters and J.J. Kolkman were supported by a fellowship from the Royal Netherlands Academy of Arts and Sciences and Glaxo-Wellcome, The Netherlands, respectively.

References [l] Marshall, B.J., Warren, J.R., Blincow, E.D., Philips, M., Goodwin, C.S., Murray, R., Blackboum, S.J., Waters, T.E. and Sanderson, C.R. (1988) Prospective double-blind trial of duodenal relapse after eradication of Campylobacter pylori. Lancet. ii, 1437-1441. [2] Walsh, J.H. and Peterson, W.L. (1995) The treatment of Helicobacter pylori infection in the management of peptic ulcer disease. N. Engl. J. Med. 333, 984991. [3] Vester, B. and Garrett, R.A. (1987) A plasmid-coded and sitedirected mutation in Escherichia coli 23s. RNA that confers resistance to erythromycin: implications for the mechanism of action of erythromycin. Biochemistry 69, 891-900. [4] Arthur, M., Brisson-Noel, A. and Courvalin, P. (1987) Origin and evolution of genes specifying resistance to macrolide, lincosamide and streptogramin antibiotics: data and hypotheses. J. Antimicrobial. Chemother. 20, 783-802. [5] Queiroz, D.M.M., Mendes, E.N.M. and Rocha, G.A. (1987) Indicator medium for isolation of Campylobacter pylori. J. Clin. Microbial. 25, 2378-2379. [6] Glupczynski, Y., Labbe, M., Hansen, W., Crokaert, F. and Yourassowsky, E. (1991) Evaluation of the E Test for quantitative antimicrobial susceptibility testing of Helicobacter pylori. J. Clin. Microbial. 29, 207222075. [7] Ausubel, F.M., Brent, R., Kingston, R.E., Moore, D.D., Seid-

42

[8]

[9]

[lo] [l l]

[12]

Y.J. Deb&-Ossenkopp

et al. I FEMS

man, J.G., Smith, J.A. and Struhl, K.S. (1989) Current Protocols in Molecular Biology. Greene Publ. and Wiley-Interscience, New York. Go, M.F., Chan, K.Y.. Versalovic, J., Koeuth. T., Graham, D.Y. and Lupski, J.R. (1995) Cluster analysis of Helicobacter pylori genomic DNA fingerprints suggests gastroduodenal disease-specific associations. Stand. J. Gastroenterol. 30, 64& 646. Arthur, M., Mohnas, C., Mabilat, C. and Courvalin, P. (1990) Detection of erythromycin resistance by the polymerase chain reaction using primers in conserved regions of erm rRNA methylase genes. Antimicrob. Agents Chemother. 34, 2024 2026. Hiratsuka, K. (1995) Helicobacter pylori 23s and 5S ribosomal RNA genes, complete sequence. Genbank accession U27270. Van Steenbergen, M.T.J., Colloms, S.D., Hermans, P.W.M., de Graaff, J. and Plasterk, R.H.A. (1995) Genomic DNA fingerprinting by restriction fragment end labeling. Proc. Natl. Acad. Sci. USA 92, 5572-5576. Debets-Ossenkopp, Y.J., Namavar, F. and Maclaren, D.M. (1995) Effect of an acidic environment on the susceptibility of Helicobacter pylori to Trospectomycin and other antimicrobial agents, Eur. J. Clin. Microbial. Infect. Dis. 14. 353-355.

Microbiology

Letters 142 (1996)

37-42

[13] Moazed, D. and Noller, H.F. (1987) Chloramphenicol, erythromycin, carbomycin and vemamycin B protect overlapping sites in the peptidyl transferase region of 23s ribosomal RNA. Biochimie 69, 8799884. [14] Meier, A., Kirschner, P., Springer, B., Steingrube, V.A. Brown, B.A., Wallace Jr., R.J. and BBttger, E.C. (1994) Identification of mutations in 23s rRNA gene of clarithromycin resistant Mycobacterium intracellulare. Antimicrob. Agents Chemother. 38, 381-384. [15] Nash, K.A. and Inderlied, C.B. (1995) Genetic basis of macrolide resistance in Mycobacterium avium isolated from patients with disseminated disease. Antimicrob. Agents Chemother. 39, 262552630. [16] Versalovic, J., Shortridge, D., Kibler, K., Griffie, M.V., Beyer, J., Flamm, R.K., Tanaka, S.K., Graham, D.Y. and Go, M.F. (1996) Mutations in 23s rRNA are associated with clarithromycin resistance in Helicobacrer pylori. Antimicrob. Agents Chemother. 40, 477480. [171 Xia, H., Buckley, M., Keane, C.T. and O’Morain, CA. (1996) Clarithromycin resistance in Helicobacter pylori: prevalence in untreated dyspectic patients and stability in vitro. J. Antimicrab. Chemother. 37, 473481.