JOURNAL OF CLINICAL MICROBIOLOGY, July 2001, p. 2751–2752 0095-1137/01/$04.00⫹0 DOI: 10.1128/JCM.39.7.2751–2752.2001
Vol. 39, No. 7
Misidentifying Helicobacter cinaedi in order to clear the situation, each culture collection should trace the origin of its corresponding type strain and probably carry out additional tests. In this context, we would like to stress the fact that the quality of the strains and services provided by the major culture collections is undoubtedly of a very high standard.
In their paper Vandamme et al. (1) report on the problems associated with the proper identification of helicobacters, citing as an example Helicobacter cinaedi. This species is very heterogeneous, and they propose that several groups of helicobacters formerly thought to represent separate species, such as Helicobacter westmeadii and Helicobacter sp. Mainz, should be classified as H. cinaedi. In the course of our own work on helicobacters, we recently realized another possible source for misidentification of H. cinaedi. By genetic analysis of several type strains of helicobacters, we found a discrepancy between the type strain of H. cinaedi from the Culture Collection of the University of Go ¨teborg (CCUG) (CCUG 18818T) and the one from the American Type Culture Collection (ATCC) (ATCC 35683T). The analysis consisted of sequencing of the 16S rRNA gene (rrs) (lengths sequenced were 1,330 bp and 1,681 bp with intervening sequence [IVS]) as well as sequencing of 480 bp of the gene for the -subunit of the RNA polymerase (rpoB). The strains analyzed and the results of sequence comparisons are listed in Table 1. We found a discrepancy between the H.
REFERENCE 1. Vandamme, P., C. S. Harrington, K. Jalava, and S. L. W. On. 2000. Misidentifying helicobacters: the Helicobacter cinaedi example. J. Clin. Microbiol. 38:2261–2266.
Peter Kuhnert* Andre´ P. Burnens Institute of Veterinary Bacteriology La ¨nggass-Str. 122 CH-3012 Bern, Switzerland *Phone: 41-31-631485 Fax: 41-31-6312634 E-mail:
[email protected]
TABLE 1. Comparison of rrs and rpoB sequences from H. cinaedi and H. fennelliae type strains Strain
Identity (%) with H. cinaedi ATCC 35683T rrs
rpoB
H. cinaedi CCUG18818T
96a
75
H. fennelliae NCTC 11612T LMG 7546T
100b 100b
100 100
a b
Authors’ Reply Dr. Kuhnert and Dr. Burnens highlight another potential source of error in the identification of H. cinaedi and of bacteria in general. Since the identification process involves matching data sets obtained from an unknown strain with those of previously described taxa, any mislabeling of the latter can result in unknown strains being misidentified. The experiences of Dr. Kuhnert and Dr. Burnens suggest that the type strain of H. cinaedi currently held at the ATCC may be a strain of H. fennelliae. We urge the authors to contact the ATCC directly to bring this to the attention of the authorities at that institution. In this context, mislabeled strains are not the only source of potential error in identification studies. Incorrectly labeled sequences held in public databases can also result in misidentifications. The mislabeling of a 16S rRNA gene sequence (GenBank accession number M88137) originally attributed to “Flexispira (Helicobacter) rappini” but subsequently corrected by Dr. F. E. Dewhirst has been presumed to explain the identification of a strain of this taxon which bears a poor resemblance to other “F. rappini” strains (1). Furthermore, poorly classified strains can also serve to confuse other workers (1), and we recommend caution where strains are identified as species not validated by publication in the International Journal of Systematic and Evolutionary Microbiology. For these reasons, the practice of polyphasic strategies to classify and identify new bacterial taxa is recommended (2), since this allows the most accurate assignment of taxonomic relationships. Furthermore, any incongruent data on a strain can be checked and potential errors determined before the information is made public. The identification of Helicobacter, Campylobacter, Arcobacter, and species of related genera is known to be a difficult task.
Comparison started after the IVS. Including 350-bp IVS.
cinaedi type strains from ATCC and CCUG of 4% for the rrs sequences and 25% for the rpoB sequences. Interestingly, the ATCC type strain showed sequence identity with the H. fennelliae type strains from the National Collection of Type Cultures, the Laboratory for Microbiology at the University of Ghent (LMG), and CCUG, also containing an IVS of about 350 bp in the rrs gene (Table 1 and GenBank accession no. M88154). In order to exclude our own laboratory artifacts due to a mix-up of long-term stored strains, the two type strains for H. cinaedi were ordered separately again from ATCC and CCUG and the genetic analysis was repeated. The same sequences were obtained from the new batches and from the long-term stock cultures of the H. cinaedi type strains. The sequences determined for the comparison are available under GenBank accession no. AF348752 to AF348752. Our results show that, besides biological problems in identifying and characterizing H. cinaedi isolates, the reference strain to which results are compared might also add to confusion. Based on our data, one could argue that the ATCC type strain of H. cinaedi is most probably a H. fennelliae isolate, but 2751
2752
LETTERS TO THE EDITOR
We hope that with the publicizing of the detrimental influence of mislabeling and misclassification on the identification process, such sources of error will greatly diminish. REFERENCES 1. On, S. L. W. Taxonomy of Campylobacter, Arcobacter, Helicobacter and related bacteria: current status, future prospects, and immediate concerns. J. Appl. Microbiol., in press. 2. Vandamme, P., B. Pot, M. Gillis, P. De Vos, K. Kersters, and J. Swings. 1996. Polyphasic taxonomy, a concensus approach to bacterial systematics. Microbiol. Rev. 60:407–438.
Stephen L. W. On* Danish Veterinary Laboratory Bu ¨lowsvej 27 DK-1790 Copenhagen V, Denmark *Phone: (45) 35 30 02 59 Fax: (45) 35 30 01 20 E-mail:
[email protected] Peter A. R. Vandamme* Laboratorium voor Microbiologie Universiteit Gent Ledeganckstraat 35 B-9000 Gent, Belgium *Phone: (32) 9 264 5113 Fax: (32) 9 264 5092 E-mail:
[email protected]
J. CLIN. MICROBIOL.
