In microbial taxonomy, one must first classify one’s unknown strains and determine whether they represent a new taxon.

One can then propose a name and formal description of the new taxon. One is then set to identify future unknowns to this new taxon. It cannot, however, be stressed enough that new names are only proposals, and their use is not mandatory. Any name published in the International Journal of Systemic and Evolutionary Microbiology (IJSEM) and in accordance with the International Code of Nomenclature of Prokaryotes (ICNP, formerly the Bacteriological Code) is validly published. There is no such thing as a ‘correct name’ for a bacterium; all validly published names are ‘correct’ and it is a grave mistake to think one has to meekly accept the latest proposal as soon it is published. Although it is generally advisable to adopt new names and to keep abreast of nomenclatural changes it is ultimately the scientific community that determines whether a new name comes into general acceptance; it is not automatic. As examples, numerical classifications based on phenotypic tests and later DNA–DNA hybridisation studies showed that, despite its motility, Enterobacter aerogenes should be placed in the genus Klebsiella. Since, at that time, there was already a ‘Klebsiella aerogenes’ the specific epithet also needed to change. The name Klebsiella mobilis was therefore proposed. However, the name K. mobilis has never been widely used and the organism is still generally referred to as Enterobacter aerogenes. The genus name Fluoribacter was proposed to accommodate certain species hitherto placed in the genus Legionella; however, Legionella bozemanae, L. dumoffii and L. gormanii are the names most widely used today; their names as placed in Fluoribacter have never found widespread usage.

Many bacteria were classified at a time when studies could only be carried out on morphology and behaviour (physiology, nutritional requirements, enzyme tests). As our abilities to study the genetic information in bacterial cells at increasingly higher levels have developed, so has bacterial classification had to change. As classification changes, so does the nomenclature of the organisms concerned. Many of the early classifications, based mainly on studies of morphology and behaviour, have stood the test of time. For some organisms, however, there have been major changes. The genus Flavobacterium for example, comprising yellow-pigmented oxidase-positive organisms, was found on the advent of mol percent guanine plus cytosine (mol % G+C) content determinations, to be divisible into high and low G+C content groups. In the 7th (1974) edition of Bergey’s Manual of Determinative Bacteriology, Flavobacterium was divided into two sections. Since the type species of the genus, F. aquatile, fell in the low G+C Section I, all the high G+C content species in Section II had to ultimately be moved into other genera.

The advent of DNA-DNA hybridization allowed a proper definition of a bacterial species (strains belonging to the same species should be 70-100% related, with less than 5% loss of thermal stability due to sequence divergence) and cast some surprises. Some ‘species’ proved to be merely biovars (biotypes or biochemical varieties) of a single genomic species. Thus ‘Klebsiella aerogenes’, ‘Klebsiella edwardsii’, Klebsiella ozaenae, Klebsiella pneumoniae and Klebsiella rhinoscleromatis are a single genomic species that should be regarded as a single species, K. pneumoniae. However, because K. ozaenae and K. rhinoscleromatis are associated with particular clinical conditions they continue to be accorded separate species status. ‘K. aerogenes’ and ‘K. edwardsii’, on the other hand, have lost standing in nomenclature as they were not included in the Approved Lists of Bacterial Names (it is an accepted convention to place such names in quotation marks) and such strains should now be called K. pneumoniae. Similarly, Escherichia coli and Shigella species are also a single genomic species, as are Neisseria gonorrhoeae and Neisseria meningitidis, but are afforded separate generic status for historical and medical reasons. In the case of Yersinia pestis and Yersinia pseudotuberculosis it was found that the two ‘species’ were so closely related that they should be regarded as subspecies of a single species with the name Y. pseudotuberculosis. However, the proposal, though scientifically valid, generated so much concern over possible confusion in the identification of plague organisms that the Judicial Commission formally rejected the name Y. pseudotuberculosis subspecies pestis and retained instead the name Y. pestis.

DNA–DNA hybridisation data, conversely, also showed some species to be heterogeneous and to comprise two or more different genomic groups, even though such species had proved relatively homogeneous in studies of their morphology and behaviour. In some cases, phenotypic characters were found to correlate with the new genomic data and this led to the description of many genomic groups as new species, often within existing genera. Thus, yellow-pigmented strains failing to ferment sorbitol and regarded as variants of Enterobacter cloacae proved to be a separate genomic group closely related to E. cloacae, and were proposed as a new taxon, Enterobacter sakazakii (and more recently as Cronobacter sakazakii). Where no phenotypic tests can be found to correlate with genomic differences then such separate genomic groups are retained, for the time being at least, in the original named species. Many new species have been determined and many new genera have been recognised as a result of DNA–DNA (and rRNA–DNA) hybridisation studies. The creation of a new genus has sometimes led to a proposal for the transfer of additional species from one genus to the new one, necessitating a change in the genus name of such species.

The advent of rRNA–DNA hybridisation and, later, 16S rRNA gene sequence comparisons, facilitated genomic studies at the genus and suprageneric levels. These studies have led to an ‘explosion’ of proposals of new genera in recent decades. It was known, for example, that the various rRNA homology groups of Pseudomonas, some as distantly related phylogenetically to each other as they were to E. coli, could be subdivided into several new genera. Pseudomonas maltophilia belonged to Pseudomonas rRNA homology group V, along with Xanthomonas, and given certain common features it was eventually proposed that the organism become Xanthomonas maltophilia. Further studies showed that despite their shared features, X. maltophilia nevertheless showed some significant differences to the other Xanthomonas species. It was therefore proposed that a new genus be created for it. This resulted in a second move, the latest proposed name being Stenotrophomonas maltophilia.


Importance to public health

A key aspect of medical, public health and diagnostic microbiology laboratories is the accurate and rapid reporting and communication regarding infectious agents of clinical significance. Microbial taxonomy in the age of molecular diagnostics and phylogenetics, however, has created changes in taxonomy at a rapid rate, further complicating this process. All the major pathogens of public health importance have been defined through various taxonomic studies over a long period of time. In foodstuffs we find Listeria monocytogenes, pathogenic E. coli and Salmonella species. In meat processing at slaughter there is frequent contamination with a number of pathogenic bacteria such as Shiga toxin-producing E. coli (STEC) O157. Other significant pathogens from food samples include Shigella species, enteroinvasive E. coli (EIEC) and enterohaemorrhagic E. coli (EHEC). Toxigenic Vibrio cholerae and Legionella pneumophila are significant waterborne pathogens. Staphylococcus aureus, especially clones that resist methicillin (MRSA) have caused a medical and public health problem worldwide.

Molecular studies have sometimes shown, as already mentioned, that certain species, considered separate in the past, are genomically a single species. Conversely, as already mentioned, molecular studies have sometimes shown the existence of multiple species within what was once regarded as a single species. It is important to discern these new species as they may have an association with certain infections or a predilection for particular patient groups. Thus the majority of C. sakazakii cases are adults, but low-birth-weight pre-term neonatal and older infants are at highest risk. The disease is associated with a rare cause of invasive infection in infants with historically high case fatality rates (40%–80%). Most neonatal C. sakazakii infections have been associated with the use of powdered infant formula. Whilst all Cronobacter species have been linked retrospectively to clinical cases of infection in either adults or infants, the species Cronobacter condimenti has not. In the case of the genus Obesumbacterium, a brewery contaminant not pathogenic to humans, there is but a single species (Obesumbacterium proteus), but it has two defined biogroups (one and two). These two biogroups are actually distinct species that are phenotypically different and only distantly related by DNA–DNA hybridisation. O. proteus biogroup one is actually a biogroup of Hafnia alvei, now referred to as H. alvei biogroup one.