Taxonomy of Campylobacter, Arcobacter, Helicobacter and related bacteria: current status, future prospects and immediate concerns.

Abstract

1. Summary, 1S 2. Introduction, 1S 3. Taxonomy of Campylobacter: a brief history, 2S 4. Current taxonomic status of ‘campylobacteria’, 2S 4.1 Campylobacter and Bacteroides ureolyticus, 2S 4.2 Arcobacter, 5S 4.3 Sulfurospirillum and ‘Dehalospirillum’, 6S 4.4 Helicobacter, ‘Gastrospirillum’ and ‘Flexispira’, 6S 4.5 Wolinella, Thiovulum and Thiomicrospira, 7S 4.6 Other taxa, 7S 5. Species diversity and misclassification: immediate problems for applied microbiologists, 8S 5.1 Identification of bacteria: a brief introduction, 8S 5.2 Phenotypic variation, 8S 5.3 Ribosomal RNA gene sequence variation by intervening sequence elements, 8S 5.4 Sequence variation by genetic drift, 9S 5.5 Ribosomal RNA gene sequence similarity between taxa, 9S 5.6 The dangers of misclassification, 10S 5.7 Misidentification: the human element, 10S 6. Future prospects, 10S 7. Acknowledgements, 12S 8. References, 12S The taxonomy of the genus Campylobacter has changed dramatically since its inception in 1963. At that time the genus comprised just two species. At present, taxa that were once assigned to Campylobacter may belong to one of over 50 species distributed among six genera. Most of these taxa belong to a phylogenetically distinct group referred to as either ribosomal RNA (rRNA) superfamily VI or the epsilon division of the class Proteobacteria. The taxonomic diversity of the group is matched by the diverse habitats in which they may be found, and by the wide range of diseases that they are associated with. Recognition of their clinical and economic importance has resulted in intense interest in the group, and the application of increasingly sophisticated isolation, detection and chemotaxonomic methods continues to elucidate new aspects of their biodiversity. However, despite the advances in new bacterial systematics, there remain a number of important issues concerning the classification of various campylobacterial taxa that require careful consideration. Ultimately, these issues are relevant to many working in the field of applied microbiology, including clinicians, veterinarians, epidemiologists and taxonomists. The purpose of this article is briefly to review the major developments in the taxonomy of Campylobacter from its inception to the present day; summarize the most recent changes in the field; analyse current topical issues of special relevance to applied microbiologists, including identification of the bacteria; and speculate on future prospects for campylobacterial taxonomy. The science of taxonomy comprises three principal areas: classification, identification and nomenclature, each linked to the other. Strains are ordered, or classified, into groups on the basis of some common feature(s), and traits allowing the group to be identified (i.e. discriminated from similar taxa) are defined. For practical purposes, the taxon must be named and, for the sake of clarity, the name must be formed with reference to rules of nomenclature. Taxonomy aims to provide a meaningful biological framework upon which a wide range of other sciences is built. However, the inaccurate delineation of taxa serves to confound workers in a variety of disciplines. The genus Campylobacter was formed in 1963 and its taxonomic structure has changed extensively since its inception. The use of increasingly sophisticated molecular methods has revealed fascinating aspects of campylobacterial biodiversity. Nonetheless, certain proposals have proven controversial, and several important problems require investigation. The aim of this paper is to provide a brief overview of the current status of the taxonomy of Campylobacter and related bacteria; analyse current topical issues of special relevance to applied microbiologists; and speculate on future developments in the field. Readers seeking more detailed appraisals of taxonomic developments concerning these taxa from the 1800s to the early 1990s are directed elsewhere (61). The genus Campylobacter was first proposed in 1963 by 45 and included just two species, Campylobacter fetus and ‘Campylobacter bubulus’ (now Campylobacter sputorum; see section 4.1). These taxa were formerly classified as Vibrio spp. until Sebald and Véron applied Hugh and Leifson’s test for fermentative metabolism, and DNA base composition, to distinguish them from ‘true’Vibrio spp. However, most of the scientific community continued to refer to these taxa as ‘Vibrio fetus’ and ‘Vibrio bubulus’, until a more extensive investigation of the genus was made (66). This study used serological and biochemical analyses, and DNA base composition, to study relationships between ‘V. fetus’, ‘V. bubulus’ and several other misclassified vibrios described by various workers but not included in the original generic proposal, namely ‘Vibrio jejuni’, ‘Vibrio coli’, ‘Vibrio sputorum’ and ‘Vibrio faecalis’. Although the taxonomic ranks of ‘C. bubulus’, C. sputorum and ‘Campylobacter faecalis’ would subsequently change or be refined (see section 4.1), Campylobacter was established as a distinct, recognizable genus. By coincidence, the study of 4 greatly increased interest in Campylobacter by indicating their high prevalence in human diarrhoea. Since C. fetus was already recognized as an important animal pathogen as a cause of abortion and infectious infertility, the identification of Campylobacter as a cause of enteric disease in humans ensured the attention of both clinicians and veterinarians alike. As a consequence, and aided by a greatly improved understanding in growth characteristics and isolation methods, 12 new species or subspecies were discovered in a range of different diseases and habitats from 1974 to 1988 (reviewed by 61). The taxonomic status of many of these taxa would change (42; 16; 65; 37) owing to the wider application of advanced taxonomic methods. Indeed, the potential of the 16S rRNA gene for determining phylogenetic relationships among all living organisms had attracted much interest (69) and would play a major role in an extensive rearrangement of Campylobacter taxonomy. Numerical comparison of partial 16S rRNA gene sequences identified several distinct clades within Campylobacter (41). Of these, ‘Campylobacter pylori’ and ‘Campylobacter mustelae’ (from human and ferret gastric mucosa respectively) were reclassified into a new genus, Helicobacter, to reconcile major differences from other Campylobacter spp. in flagellar structure, fatty acid and menaquinone composition, and 16S rRNA gene sequences (16). 41 also identified ‘Campylobacter cryaerophila’ and ‘Campylobacter nitrofigilis’ as related, yet distinct from other Campylobacter and Helicobacter spp., and several Wolinella and Bacteroides spp. appeared to be affiliated to either Campylobacter (‘Wolinella curva’, ‘Wolinella recta’, ‘Bacteroides gracilis’, Bacteroides ureolyticus) or Helicobacter (Wolinella succinogenes). Even Thiovulum, an uncultivable marine bacterium, appeared to be distantly related to these taxa. These issues were resolved in an extensive polyphasic taxonomic study of the entire Campylobacter complex (60). By use of DNA–rRNA hybridizations, relative phylogenetic positions were determined and data cross-referenced with other phenotypic and genetic data. This study provided the basis of the taxonomic structure used at present, and delineated Campylobacter spp. as a diverse yet phylogenetically distinct group, rRNA superfamily VI (also known as the ɛ-division of the Proteobacteria;54). This comprised rRNA homology groups I (Campylobacter and B. ureolyticus), II (Arcobacter) and III (Helicobacter and W. succinogenes). The close relatedness of rRNA homology groups I and II was noted and a subsequent paper proposed these be included in the family Campylobacteraceae (58). 60 also examined two saprophytic, sulphur-reducing strains described as ‘Campylobacter-like’ or ‘Spirillum sp.’ that appeared to be related, yet separate, from the other rRNA homology groups of rRNA superfamily VI. These strains would later form a new genus, Sulphurospirillum (44). At present, the genus Campylobacter contains 16 species and six subspecies (Fig. 1). Campylobacter jejuni subsp. jejuni, C. jejuni subsp. doylei, Campylobacter coli, Campylobacter lari, Campylobacter upsaliensis and Campylobacter helveticus form a genetically close group of species which (C. jejuni subsp. doylei aside) are the most commonly isolated from human and animal diarrhoea. ‘Campylobacter hyoilei’, isolated from lesions of porcine proliferative enteritis, was later identified by a wide range of phenotypic and genotypic methods as a strain of C. coli (64). It was, however, suggested that these strains may represent a pathogenic variant (pathovar) of C. coli and the identification of a genetic marker that is evidently specific to ‘C. hyoilei’, and the observation of some metabolic differences (6), makes this prospect feasible. Dendrogram showing relationships between 62 strains of Campylobacter, Arcobacter, Helicobacter, Sulfurospirillum, Wolinella, Thiovulum, Lawsonia, Sutterella, Anaerobiospirillum, Bacteroides ureolyticus, ‘Dehalospirillum multivorans’ and Thiomicrospira denitrificans, as inferred by comparison of 16S rRNA gene sequences and neighbour-joining clustering. The scale bar indicates 19% sequence dissimilarity. Sequences were obtained via GenBank, and a multiple alignment obtained using BioNumerics software (Applied Maths, Kortrijk, Belgium). Gaps and unknown bases were not considered in the analysis The hydrogen-requiring species Campylobacter concisus, Campylobacter showae, Campylobacter curvus, Campylobacter rectus, Campylobacter gracilis, C. sputorum and C. hominis appear to be closely related phylogenetically. Most occur in the human oral cavity, although C. hominis has been found only in the human lower intestine and C. sputorum is also found in the enteric and reproductive tracts of various production animals. Campylobacter sputorum comprises three biovars, revised by 37, defined on the ability of a given strain to produce catalase or urease; bv. sputorum (negative in both tests), bv. faecalis (catalase positive, urease negative), and bv. paraureolyticus (urease positive). Strains described previously as bv. bubulus were reclassified as bv. sputorum, since the tests differentiating these two taxa (tolerance to 1% bile and 3·5% NaCl) were not reproducible. Campylobacter hominis was initially found to be uncultivable using standard methods, and was first designated Candidatus C. hominis in view of its presumptive taxonomic status (28). This species has now been successfully cultured under anaerobic conditions, primarily by use of an immunomagnetic separation method to separate bacterial cells from faeces (29). Campylobacter mucosalis is, by DNA–DNA hybridization studies, most similar to C. sputorum (42), with which it shares a highly similar phenotype and a common source (pig intestine). Campylobacter fetus subsp. fetus, C. fetus subsp. venerealis, Campylobacter hyointestinalis subsp. hyointestinalis and C. hyointestinalis subsp. lawsonii are also similar by phenotype and genotype. Campylobacter fetus is mainly found in bovine reproductive disorders, whereas C. hyointestinalis subsp. hyointestinalis is mainly enteric in origin and C. hyointestinalis subsp. lawsonii occurs in the pig stomach. Campylobacter hyointestinalis shows remarkable diversity at the 16S rRNA genetic level (18), a finding with implications for Campylobacter lanienae, from abbatoir workers (32), as discussed later in this section. A number of taxonomic problems in Campylobacter remain. Campylobacter concisus was first isolated from the oral cavity but there have been several reports of strains associated with diarrhoea, often found with a frequency equal to that of C. jejuni (9). However, some strains from diarrhoea show as little as 46% relatedness to the type strain (of oral origin) in DNA–DNA hybridization experiments (59). By current definition, strains belonging to a given species should exhibit approximately 70% DNA–DNA relatedness to each other (49). Thus, the C. concisus-type strain and these diarrhoal isolates represent genetically distinct species, yet there are no other phenotypic or genotypic criteria to distinguish them, and C. concisus is regarded as a ‘complex’. The presence of C. concisus in both diarrhoea and normal faeces has been used to argue that its presence in diarrhoea is simply coincidental (56), or at best opportunistic (9). However, at present it cannot be determined if all the diarrhoeal strains represent a single species – distinct from, but closely related to, the oral C. concisus strains. There is a need to clarify this issue and develop methods for effective and rapid differentiation of C. concisus genomospecies. In this respect, the determination of amplified fragment length polymorphisms (AFLP) shows potential (38). Campylobacter lari is also known to be diverse. Three phenotypically distinct variants are currently known, namely the classic nalidixic acid-resistant biotype, a nalidixic acid-sensitive variant and a urease-positive variety. Various DNA–DNA hybridization (35) and protein profile analyses (8) show the latter two groups are closely related to C. lari sensu strictu, but to date there has been no concerted effort to define the taxonomic status of these biotypes properly. Furthermore, strains from Dutch shellfish were found to be extremely heterogeneous by biotype, and protein and Randomly Amplified Polymorphic DNA (RAPD) profiles, forming several distinct groups (8). The taxonomic status of these different biotypes and protein groups requires further study, especially in view of the wide distribution (birds, animals, water, shellfish) of these bacteria. A recent study demonstrated that classical and urease-positive strains form closely related but distinct clusters in a numerical analysis of AFLP profiles (38), suggesting a subspecies-level relationship. The subspecies division of C. fetus poses a special problem for veterinary laboratories. Effective differentiation of C. fetus subsp. fetus (a cause of abortion in cattle and sheep) from C. fetus subsp. venerealis (the cause of infectious infertility) is essential since the latter disease can be economically crippling, especially where it is endemic. The subspecies designation is unique within the genus in that it is based on these different pathogenic associations. Although several phenotypic and genotypic methods are useful for discriminating the two subspecies (39), none of these directly relates to the strain pathotype. Improved classification and identification of C. fetus subspecies can only be attained through elucidating the relationship between genotype and pathotype. Improved knowledge of the diversity in the 16S rRNA gene for some taxa (see section 5.4) has possible implications for the taxonomic status of a new species, C. lanienae, which clusters closely with strains of the existing C. hyointestinalis subspecies (Fig. 2). Indeed, C. hyointestinalis subsp. lawsonii shares 16S rRNA gene sequence similarities of 97·2–97·7% with C. lanienae in this analysis. At this level of similarity (and arguably below; see section 5.4), 16S rRNA sequence analyses lack sufficient resolution to determine if taxa represent the same, or distinct, species and DNA–DNA hybridization experiments are required to clarify the taxonomic position of such strains. However, such experiments between the latter two taxa have not yet been performed and the source of C. lanienae (abbatoir personnel) is in keeping with our knowledge of the habitats of C. hyointestinalis. Further studies will be necessary to determine if C. lanienae is deserving of species status, or if it represents a novel subspecies (or an existing one) of C. hyointestinalis. Dendrogram showing relationships between Campylobacter hyointestinalis subspecies, C. lanienae, related Campylobacter spp. and Bacteroides ureolyticus, as inferred by comparison of 16S rRNA gene sequences and neighbour-joining clustering. The scale bar indicates 9% sequence dissimilarity. Sequences were obtained via GenBank, and a multiple alignment obtained using BioNumerics software (Applied Maths, Belgium). Gaps and unknown bases were not considered in the analysis The taxonomic position of the generically misnamed B. ureolyticus requires clarification. Comparisons of 16S rRNA gene sequences consistently place this species as an outlier to the main Campylobacter group (57; see Fig. 2) and several notable phenotypic differences between B. ureolyticus and Campylobacter spp. have been noted, including fatty acid content and proteolytic metabolism (57). Nonetheless, the overall phenotype of this species is certainly comparable to that of many anaerobic campylobacters, as illustrated by a numerical analysis of 67 phenotypic traits of 367 campylobacterial strains (40). However, the prospect of transferring B. ureolyticus to the genus Campylobacter cannot be undertaken realistically before further studies on its diversity are undertaken, as suggested previously (57). Initially, the genus Arcobacter comprised Arcobacter cryaerophilus (found in animal abortions and human enteric disease) and Arcobacter nitrofigilis (isolated from plant roots), reclassified Campylobacter spp. capable of growth under aerobic conditions, and at 15°C (60). Two species have been added to the genus since its inception. Arcobacter butzleri was first described as a Campylobacter sp. (26), but an extensive polyphasic taxonomic study of aerotolerant Campylobacter-like organisms clearly assigned ‘Campylobacter butzleri’ to Arcobacter, mainly on the basis of DNA–rRNA hybridization results (65). Arcobacter skirrowii was described in the same study and comprised 18 strains from various veterinary specimens. Both studies by 26 and 65 showed that two subgroups of A. cryaerophilus could be observed on the basis of DNA–DNA hybridization values, fatty acid composition and ribotype patterns. It is likely that these subgroups deserve subspecies status, but the lack of unequivocal biochemical features to differentiate them has so far hampered a formal description. Thus, Arcobacter contains four species at present (Fig. 1). However, investigations of bacterial communities of oil fields and a hypersaline pond (53; 67) have revealed the presence of Arcobacter-like organisms that share less than 89% 16S rRNA gene sequence similarity with existing taxa, thus probably representing novel species. Although 60 described the taxonomic position of two free-living ‘Campylobacter-like’ strains (Spirillum 5175 and DSM 806) as distinct within rRNA superfamily VI, they were not formally classified until 44 compared the growth factor requirements, DNA base composition, and DNA hybridization values of Spirillum 5175 with C. sputorum and W. succinogenes. As a consequence of the differences observed, the genus Sulfurospirillum was proposed to contain both Spirillum 5175 and DSM 806, although Sulfospirillum deleyianum comprised the former strain only; the exact taxonomic status of DSM 806 remains undetermined at this time. A further study of environmental, microaerophilic, sulphur-reducing curved bacilli revealed 16S rRNA sequences of the above two strains formed a distinct clade within rRNA superfamily VI and classified two other strains into Sulfurospirillum. Of these, one was named Sulfurospirullum arcachonense (13) which was found to have a fatty acid and respiratory menaquinone composition characteristic for members of rRNA superfamily VI. The other strain (initially named ‘Geospirillum barnesii’) was later named Sulfurospirullum barnesii by 50, who also described Sulfurospirullum arsenophilum, represented by a single isolate from arsenic-contaminated watershed sediments. 16S rRNA sequence comparisons identified one other Sulfurospirillum strain (SM-5) but this was not formally classified by 50. The distinct phylogenetic positions of SM-5 and DSM 806 suggest that these probably represent novel species. ‘Dehalospirillum multivorans’ (43) was recovered from activated sludge and has the ability to reduce tetrachloroethene, a volatile man-made pollutant. This unusual characteristic, and its distinct position in a phylogenetic tree, were key arguments for its proposal as a representative of a new genus and species belonging to rRNA superfamily VI. However, other Sulfurospirillum spp. were not included for comparison. Comparison of 16S rRNA gene sequences clearly places ‘D. multivorans’ in the Sulfurospirillum clade (Fig. 1) and the organism shares many features (source of isolation; temperature range; optimum pH; metabolism of formate, lactate and pyruvate; DNA base composition) with Sulfurospirillum spp. These data suggest that ‘D. multivorans’ belongs to the genus Sulfurospirillum. The intense interest in H. pylori as a frequent cause of human gastritis (and related disorders, including gastric cancer) has inspired many investigators to examine a variety of animals for similar bacteria. The realization that some species (initially Helicobacter cinaedi and Helicobacter fennelliae) were associated with lower intestinal disorders also contributed to the wide attention to the genus. Interest continues to be fuelled by the zoonotic potential of some species (e.g. Helicobacter felis, Helicobacter bizzozeronii, Helicobacter canis, Helicobacter pullorum) and observations indicating an involvement with hepatitis (reviewed by 14). As a consequence, the number of Helicobacter spp. validly described (i.e. published in the International Journal of Systematic and Evolutionary Microbiology (IJSEM), formerly the International Journal of Systematic Bacteriology (IJSB)) since 1991 has increased dramatically from four to 18 as of 31 March 2000. There remains an ecological division and taxa are usually gastric or lower enteric in origin, although a few enteric species can colonize the stomach under certain conditions and are those most often associated with hepatic disease. The gastric species were isolated from studies of various animals and include H. felis (cats and dogs), Helicobacter nemestrinae (monkey), Helicobacter muridarum (mouse), Helicobacter acinonychis (cheetahs), H. bizzozeronii (cats) and Helicobacter salomonis (cats). The enteric species include H. canis (dogs), Helicobacter hepaticus (mice), Helicobacter pametensis (birds), H. pullorum (chickens), Helicobacter bilis (mice), Helicobacter trogontum (mice), Helicobacter cholecystus (mice) and Helicobacter rodentium (rats) (listed by 7). Two as yet uncultured taxa have been proposed as provisional, or Candidatus spp: Candidatus H. bovis, from cattle, and Candidatus H. suis, from pigs (listed by 7). Moreover, several putative novel species have been named in journals other than IJSEM/IJSB, and consequently have no official standing in nomenclature at present. These include the uncultivated ‘Helicobacter heilmannii’ (synonym, ‘Gastrospirillum hominis’, discussed later in this section), ‘Helicobacter rappini’ (synonym, ‘Flexispira rappini’, discussed later in this section), ‘Helicobacter suncus’ and ‘Helicobacter typhlonicus’ (listed by 7). In addition to the above, a range of as yet unnamed taxa have been described in various reports that are presumed (principally by their position in phylogenetic trees) to represent novel Helicobacter spp. There are several outstanding taxonomic issues to be resolved in Helicobacter. A number of taxa are recognized mainly by their characteristic cell morphology as large, tightly coiled helical rods and have been referred to as ‘gastrospirilla’ after ‘G. hominis’, the generic name first suggested for strains seen in some human gastritis cases (33). Similar organisms have been seen in a wide range of other animals, including domestic pets, monkeys and various captive exotic carnivores, with ‘Gastrospirillum suis’ and ‘Gastrospirillum lemur’ suggested for bacteria in pig and lemur stomachs respectively (see 36). 16S rRNA gene sequence analyses confirm all known ‘gastrospirilla’ as Helicobacter spp. ‘Gastrospirilla’ have proven extremely difficult to culture in vitro and this has greatly hindered taxonomic studies to determine accurately the relationship among strains found in various hosts. However, a few ‘gastrospirilla’ have been cultured from cats and dogs and three distinct species are known: H. felis, H. bizzozeronii and H. salomonis (23). Pig strains now constitute a provisional species, Candidatus H. suis (5). The human strains (referred to as either ‘G. hominis’ or ‘H. heilmannii’) (33; 48) are divided into two distinct sequence types. Type 1 strains form a clade with Candidatus H. suis, while type 2 strains are affiliated to H. felis, H. bizzozeronii and H. salomonis. Intraspecific and infraspecific variability in the 16S rRNA gene of these species is such that unequivocal differentiation cannot be achieved (23). The taxonomic status of the human strains in relation to the H. felis–bizzozeronii–salomonis complex and to Candidatus H. suis has therefore been impossible to assess accurately. However, a polyphasic taxonomic analysis has identified a cultured human ‘H. heilmannii type 2’ strain as H. bizzozeronii (24). This does not mean that all human ‘gastrospirilla’ are H. bizzozeronii; the 16S rRNA gene sequence of ‘H. heilmannii’ type 1 is more similar to that of Candidatus H. suis (Fig. 1). Available data suggest the two sequence types of human ‘H. heilmannii’ represent at least two distinct, extant species. This area requires further investigation, especially to clarify the zoonotic potential of these bacteria. The performance of polyphasic taxonomic studies on ‘gastrospirilla’ will require significant advances in culture methods. The name ‘Flexispira rappini’ was used to describe isolates from aborted lambs, dog and human faeces, and later pig intestines (2). The ultrastructure of this organism is highly distinctive: the cell body is cigar shaped, encircled by a complex array of periplasmic fibrils and has bipolar tufts of sheathed flagella. Although several studies clearly assign strains meeting this description to Helicobacter (60; Fig. 1), it is far from certain that all such strains belong to a single species. A comparative 16S rDNA sequence analysis of several Helicobacter spp. that included two ‘H. rappini’ strains of human and porcine origin delineated them to two markedly different clades (46). Moreover, both H. bilis and H. trogontum are considered distinct species (cf. 7), although these too exhibit the typical ultrastructure of ‘H. rappini’. 25 described ‘H. rappini’-like strains from dog stomachs that were phenotypically and protein-electrophoretically different from a human isolate. The limited resolution of the 16S rRNA gene at the species level (see sections 5.4 and 5.5) make it impossible to determine whether strains from each of the sources described (or strains of a particular sequence type) represent distinct or a single species. There is a need for additional chemotaxonomic information (notably DNA–DNA hybridization data) to investigate relationships among all strains exhibiting the ‘H. rappini’ cell morphology. Since the revision of Wolinella by 60, no changes have been made and W. succinogenes, a commensal of the bovine rumen, remains the only member species. 16S rRNA gene sequence comparisons have showed Thiovulum to be affiliated to rRNA superfamily VI (41; Fig. 1) although, unusually, the bacterium is not available in pure culture. According to the Approved List of Bacterial Names, the genus contains one species, Thiovulum majus, and is marine in origin. The genus Thiomicrospira contains seven species, of which six are unrelated to rRNA superfamily VI. However, Thiomicrospira denitrificans is generically misnamed since 16S rRNA gene sequence analyses show it is affiliated to Thiovulum and rRNA superfamily VI generally (Fig. 1). The environmental origin of Thiomicrospira denitrificans is a trait shared by Sulfurospirillum, Thiovulum and some Arcobacter spp., but further work is necessary to confirm its taxonomic position within the group. Several taxa that were first classified as Campylobacter spp. have subsequently been found to represent species in genera that are phylogenetically distinct from the entire rRNA superfamily VI. Lawsonia intracellularis, now widely accepted as the causal agent of porcine proliferative enteropathy, was first named ‘Campylobacter intracellulare’ (34). 16S rRNA sequence comparisons showed the organism to be more closely related to Desulfovibrio; several studies have identified similar bacteria in various other animals including horses, hamsters and ostriches (27). A group of bile-resistant isolates mainly from intestinal tract infections and first described as ‘C. gracilis-like’ have now been named Sutterella wadsworthensis, whose closest phylogenetic neighbours include Alcaligenes and Bordetella, aerobic β-division Proteobacteria (68). The genus Anaerobiospirillum (containing two species, An. succiniciproducens and An. thomasii) is noteworthy since cultures of these large spiral bacteria closely resemble those of enteric campylobacters and are similarly associated with enteritic and septicaemic infections (51). Unlike campylobacters, acid production from sugar fermentation is readily detected in Anaerobiospirillum spp. (36) and 16S rRNA sequence analysis delineates An. succiniciproducens to the δ-division of the Proteobacteria (51). For most applied microbiologists, the identification of a bacterial strain forms their most common involvement with taxonomy. Accurate identification of a strain is essential to facilitate effective clinical management, estimate prevalence of a given species, or as a precursor to the effective performance and evaluation of various analyses. Identification involves the comparison of data obtained for an unknown strain with those of known taxa. Strains are identified when two data sets (unknown with known) match to an acce