点击显示 收起
Received 15 February 2003/ Returned for modification 27 March 2003/ Accepted 9 May 2003
ABSTRACT |
---|
Top Abstract Introduction Materials and Methods Results Discussion References |
---|
INTRODUCTION |
---|
Top Abstract Introduction Materials and Methods Results Discussion References |
---|
Identification of campylobacters and helicobacters to the exact species level using phenotypic tests is difficult, since these criteria are still insufficient for identification of all species adequately (7, 15). The use of nalidixic acid susceptibility in the identification schedule of campylobacters is hampered by acquired resistance to fluoroquinolones. Difficulties with correct identification are illustrated in literature by the use of terms as "atypical Campylobacter strain" and "Campylobacter-like organisms" in cases where phenotypic characteristics are insufficient for conclusive identification (6, 14). Several species that were formerly classified as Campylobacter have been reclassified as Helicobacter and Arcobacter (6, 13, 14, 22). Additionally, helicobacters are frequently misidentified as campylobacters (26).
Several molecular methods have been reported to reliably identify the clinically important species of Campylobacter and Helicobacter. In recent years, most published work has relied on 16S rRNA gene sequencing to identify the species and a few studies report on the usefulness of 23S rRNA sequences (2, 5, 7, 11, 27). In a previous study to prevalence of campylobacters among human immunodeficiency virus (HIV)-seropositive patients with diarrhea, we found variations of the region between helices 43 and 69 of the 23S ribosomal DNA (rDNA) gene and restriction patterns of a putative heme-copper oxidase domain useful for a correct identification of Campylobacter species (15). Restriction patterns of the heme-copper oxidase domain were specific for identification of the thermophilic species (10). Subsequently, these methods were introduced in our routine laboratory practice for identification of campylobacters that are difficult to identify with phenotypical tests.
The finding of a C. lari-like isolate resistant to erythromycin from the blood of a HIV-seropositive patient and the difficulties we encountered with a correct genetic identification of this isolate, prompted us to study other erythromycin-resistant campylobacter strains from our collection. Macrolides are the drugs of first choice for the treatment of Campylobacter infections. Macrolide resistance is rare in Campylobacter species but not in Helicobacter pylori. Five erythromycin-resistant fecal isolates were present in our collection of campylobacters stored at -70°C in the period 1995 to 2001. We subsequently genotyped the isolates by nucleotide sequence analysis of parts of the 23S and 16S rDNA to determine the species of the isolates. Furthermore, we performed PCR for a putative heme-copper oxidase domain and we analyzed the mechanisms underlying erythromycin resistance in these isolates.
MATERIALS AND METHODS |
---|
Top Abstract Introduction Materials and Methods Results Discussion References |
---|
|
Phenotypical identification. Campylobacter species were identified on the basis of colony morphology, the presence of curved gram-negative rods in a Gram stain and a positive oxidase reaction. All isolates were further tested using nitrate reduction; production of urease and H2S; hydrolysis of hippurate and indoxyl acetate; growth at 25, 35, and 42°C; and susceptibility to nalidixic acid and cephalothin (13, 23, 25).
Antimicrobial susceptibility testing. To determine susceptibility to nalidixic acid and cephalothin for identification purposes, a standard disk diffusion technique on 5% sheep blood agar plates was used. MICs of erythromycin were determined on Mueller-Hinton agar supplemented with 5% sheep blood by E-test (AB Biodisk, Solna, Sweden) in a microaerobic environment for 24 to 28 h at 37°C. Strains were considered resistant to erythromycin when the MIC of erythromycin was 8 µg/ml (24).
Genetic characterization. (i) DNA isolation. DNA was isolated from suspensions of bacterial cultures from solid media using the QIAmp DNA kit (Qiagen, Hilden, Germany) according to the manufacturer's instructions.
(ii) PCR amplification. (a) 23S rDNA amplification. A specific part of the 23S rDNA gene for identification of thermophilic Campylobacter sp. was amplified using Therm1 and Therm2 (2, 5, 14), and Bob1 and Bob2 primers (E. J. Kuijper, B. de Wever, F. Snijders, S. A. Danner, J. Dankert, Abstr. 98th Gen. Meet. Am. Soc. Microbiol., abstr. C-254, p. 173, 1998) (Table 2). PCR amplification was carried out in Qiagen PCR Mastermix with 10 pmol of each primer and using 5 µl of bacterial DNA. After an initial denaturation step of 3 min at 95°C, the protocol consisted of 35 cycles of 1 min at 95°C (denaturation), 1 min at 54°C (annealing), and 2 min at 72°C (elongation). The Bob primers and Therm primers yielded an amplified fragment of 290 and 330 bp, respectively. PCR products were investigated by nucleotides sequence analysis.
|
(c) 23S rDNA codon 714 mutation analysis. Single nucleotide polymorphisms at positions 2142 and 2143 of the 23S rDNA were investigated by amplification of adjacent 23S rDNA sequences using primer pair ar69fw and ery23sr (Table 2). PCR was performed in a volume of 50 µl containing 25 µl of HotStar PCR Mastermix (Qiagen), 25 pmol of each primer, and 5 µl of DNA isolate. Subsequently PCR products were sequenced (see below).
(d) PCR for the putative heme-copper oxidase domain of Campylobacter spp. A PCR specific for thermophilic Campylobacter spp. was performed according to the method of Jackson et al. (10), targeting a gene encoding a putative heme-copper oxidase domain, downstream from a Campylobacter two-component regulator gene (Table 2 shows the oligonucleotide sequences). After digestion of the PCR products with AluI, DdeI, and DraI, the restriction fragment length polymorphisms were analyzed by electrophoresis on 2% NuSieve agarose.
(e) erm(B) and mef(A) analysis. PCR assays for the erm(B) gene, encoding a 23S rDNA methylase and the mef(A) gene, encoding a macrolide efflux protein were performed as described previously (1, 12, 17) (Table 2 shows oligonucleotide sequences).
(iii) Nucleotide sequence analysis. Nucleotide sequence analysis of PCR products was performed using an ABI PRISM BigDye Terminator Cycle Sequencing kit (version 3; Applied Biosystems). PCR products were sequenced in forward and reverse orientation using the same primers as used for PCR amplification. Sequence reaction mixtures were applied to an ABI PRISM 310 Genetic Analyzer. Nucleotide sequences were edited and aligned using Vector NTI Suite 6 (InforMax Inc.) and compared against GenBank DNA sequences using Basic Logic Alignment Tool (BLAST), which is available at http://www.ncbi.nlm.nih.gov.
RESULTS |
---|
Top Abstract Introduction Materials and Methods Results Discussion References |
---|
Genotypic identification by 23S and 16S rDNA sequence analysis. All six clinical isolates reacted with the Therm1-Therm2 primer pair. Of the reference strains, C. jejuni, C. lari, and H. cinaedi (LMG 7543) reacted with Therm1-Therm2, but H. fennelliae (Z6) showed a negative reaction. 23S rDNA sequence analysis of PCR products obtained with primer pair Bob1-Bob2 showed strongest homology of all isolates with Wolinella succinogenes (Table 3). Reference strains C. lari Lior 34 and C. jejuni were correctly identified based on their 23S rDNA sequence (similarity with GenBank sequences of C. lari and C. jejuni of 100%), but H. cinaedi reference strain LMG 7543 was incorrectly identified as W. succinogenes with 97.9% similarity. Since 23S rDNA sequence analysis was poorly informative and homology with Campylobacter species was low, we sequenced 16S rDNA using primer pairs P1-P2 and P3-P4, covering more than 40% of the gene, including the most discriminative 5' end. Sequence analysis using P1-P2 enabled identification of isolates D19B1, D19G1, D23G3, and 208445J as Helicobacter cinaedi, with similarity percentages of 99.2 to 100%. Strains C14D7 and C14E3 showed strongest homology with H. pullorum. Sequence analysis using primer pair P3-P4 confirmed the identification of these strains as H. pullorum with a similarity of 99.8 and 99.3%, respectively, whereas the similarity to H. cinaedi was 98.2 and 97.6% (Table 3).
|
Genetic mechanisms underlying erythromycin resistance in H. cinaedi and H. pullorum. The putative mechanisms that may lead to erythromycin resistance in the Helicobacter strains was investigated by a PCR for the erm(B) and mef(A) genes. None of the isolates were positive in erm(B) PCR. Control S. pyogenes strain 59C1133 was positive for erm(B) DNA in this experiment. All isolates were negative for the mef(A) gene as well, while control S. pyogenes strains 59C1139 was positive for mef(A) DNA (Table 4).
|
DISCUSSION |
---|
Top Abstract Introduction Materials and Methods Results Discussion References |
---|
The nucleotide sequences of 23S rDNA fragment between helices 42 and 63 in 23S rDNA have been described as one of the most variable regions of the 23S rDNA of campylobacters. This fragment has been used to develop specific primers for thermophilic Campylobacter species (5). Although the primer set was used on a large number of reference strains (including H. cinaedi but not H. pullorum) and clinical isolates, only visible bands corresponding to 290 bp were observed for the thermophilic Campylobacter species. In contrast with these findings, we found a reference strain of H. cinaedi (LMG 7543) and clinical erythromycin-resistant isolates of H. pullorum and H. cinaedi also positive with these primers. This finding has important implications, since we introduced this method in our routine laboratory practice and misidentified clinical isolates as erythromycin-resistant C. lari-like strains (15).
Restiction patterns of the putative heme-copper oxidase domain PCR product has also been reported as a tool for accurate specification of thermophilic Campylobacter spp. (10). Unfortunately, we found this gene also present in an H. cinaedi isolate. H. cinaedi was not included in the manuscript of Jackson et al. in which the specificity of the primers for the heme-copper oxidase domain were tested on all campylobacters of RNA homology group 1, all Arcobacter species and Helicobacter pylori species (10). The target sequence of the PCR is highly conserved and under conditions of low stringency, specific 256-bp products can be amplified from DNA of all members of the genus Campylobacter. However, we applied similar stringent criteria as described by Jackson et al. and found other Helicobacter cinaedi isolates, H. pullorum and H. pylori strains negative.
Of six clinical isolates, four showed homology of 99.2 to 99.8% to Flexispira rappini in 16S rDNA. F. rappini is a provisional name to a helicobacter with spiral periplasmic fibers and bipolar tufts of sheathed flagella (27). Considerable diversity in the 16S rRNA genes of 35 F. rappini strains has been observed so that the International Subcommittee on the Taxonomy of Campylobacter and Helicobacter decided that F. rappini is not a well-defined species but encompasses multiple Helicobacter species (3, 27). Therefore, the several interesting reports of bacteremia due to Helicobacter (Flexispira) rappini in HIV-seropositive patients and in a patient with X-linked hypogammaglobulinemia should be interpreted with care (8, 20, 21, 30).
Macrolides are the drugs of first choice for the treatment of Campylobacter and Helicobacter infections, except for H. pylori. Three point mutations in the 23S rRNA gene of H. pylori were found to be associated with macrolide resistance (19). The mechanism of erythromycin resistance we found in four H. cinaedi and two H. pullorum strains seems to resemble macrolides resistance in H. pylori and is due to due to an mutation of one of two adenines (A2142G or A2143G) in the 23S rRNA at the erythromycin-binding site (28, 29). Recently, sequencing of the 23S rRNA genes from 22 erythromycin-resistant Campylobacter spp. identified mutations at these same sites, which are most probably responsible for resistance (24; C. A. Trieber and D. E. Taylor, Abstr. Final Prog. 10th Int. Workshop Campylobacter Helicobacter Rel. Organisms, abstr. CA6, p. 3, 1999). In one study high levels of erythromycin resistance in 34 H. cinaedi strains (MIC 128 mg/liter) has been reported, but the mechanism of erythromycin resistance was not investigated further (11). Identification of these 34 H. cinaedi strains was performed by total DNA-DNA hybridization but discrepancies of DNA-DNA hybridization have been observed with 16S rRNA sequencing (30). Neither in vitro susceptibility testing data nor treatment recommendations are available for infections caused by H. pullorum, but it seems reasonable to follow the same guidelines as for other helicobacters and to take macrolides resistance into account. Of special interest is the observation of Tee et al. that two strains fulfilling the morphological criteria of Helicobacter (Flexispira) rappini were also resistant to erythromycin (20). Whether erythromycin-resistant Helicobacter species other than H. pylori are an increasing clinical problem should be the scope of a larger study on this subject.
This study clearly demonstrates the shortcomings of identification of erythromycin-resistant Campylobacter-like strains using phenotypical tests, available 23S rRNA sequence data and the gene encoding a putative heme-copper oxidase domain. The strains were finally identified using the sequence of the most informative part of 16S rDNA, as recently recommended by the International Subcommittee on the Taxonomy of Campylobacter and Helicobacter (4). Therefore, we recommend that erythromycin-resistant campylobacter-like strains and campylobacter-like strains isolated from blood cultures be investigated further by sequencing of 16S rDNA. We realize that this technology is not available for routine microbiology laboratories, but reference laboratories should overcome this problem until commercial available reagents and procedures are introduced in routine practice. Additionally, this is also the first report of H. cinaedi and H. pullorum strains resistant to erythromycin with mutations in the 23S rDNA gene similar to those observed in H. pylori.
ACKNOWLEDGMENTS |
---|
REFERENCES |
---|
Top Abstract Introduction Materials and Methods Results Discussion References |
---|