The increasing use of Microbial Genotypic Techniques for the Identification of Pharmaceutical Isolates

The historical demarcation of prokaryotes has not been by way of a specific scientific-based concept, but has been defined by a more arbitrary, anthropocentric system, rooted in the practical necessity of the time of its inception, based on the infor­mation available at the time. Therefore, species are historically defined on the basis of the disease they cause.

The taxonomy of pathogens gave the microbiologist an implicit understanding of what that organism is likely to do, based on past experience. However, as the science of microbiology has spread significantly beyond the boundaries of human disease the traditional methods of species demarcation have proved increasingly problematic. Prokaryotic taxonomy is in a state of constant flux as our understanding of the phylogeny of the microorganisms increases, based on increasingly detailed analytical techniques. This inevitably leads to confusion and contradiction, particularly as the more recent genetic identification systems illustrate weaknesses and anomalies in the phenotypic taxonomic definitions.

It is widely acknowledged that the current identification of microorganisms by phenotypic methods is subject to a number of variables, many of which are beyond the control of the microbiologist. As a result, identification based on phenotypic tests does not always allow an unequivocal identification. The primary isolation and subculture for purity allows full phenotypic expression and growth of sufficient inoculum for the identification. At best, the first steps to identification (Gram stain, colonial morphology, oxidase/ catalse test, etc.) provide most of the information an experienced microbiologist needs to characterise an isolate, with the subsequent semi-automated profiling being confirmatory. At least, this ensures that the isolate being submitted for phenotyping is pure. Whatever the outcome, further identification to either confirm genus or assign species is an expectation for most isolates in the pharmaceutical industry.

Since the advent of techniques for nucleic acid manipulation and/or amplification, there has been significant growth in alternative microbial identification techniques. These methods are regarded as more accurate, reliable and reproducible than traditional phenotypic methods. In respect of semi-automated genotypic systems available to the pharmaceutical industry, those employing gene sequencing or ribotyping are most common. The 16S rRNA sequence of a species is a genotypic feature that allows the identification of bacteria to at least the species level and the similarity reflects phylogenetic relationships. If the goal is to identify an unknown organism on the basis of no prior knowledge, the 16S rRNA gene sequence is an excellent and extensively used choice. Ribotyping utilises the differences and similarities found in the 16S, 23S and 5S rRNA genes of microorganisms. These genes are highly conserved, but vary in number and position within the genome. The technique produces a stable fingerprint that is compared to other patterns differentiating isolates to the genus, species and sub-species level. Ribotyping has proved a powerful epidemiological tool when the traditional phenotypic methods fail to discriminate between strains.

Genotypic methods, while increasingly found in specialist and research laboratories, are yet to become well established in routine QC laboratories. There are a number of reasons for this, both real and perceived. One common argument against genotyping is the cost per test, which in reality is no longer reflected by either the initial capital outlay or ongoing reagent costs, particularly when compared to the cost of phenotypically identifying difficult environmental isolates, which may require a number of repeat tests before a satisfactory identification profile is achieved. A strong argument for genotyping is that it is increasingly regarded as the ‘gold standard’ for bacterial identification. However, there are a number of potential ‘pitfalls’ into which the inexperienced user can fall. While these pitfalls, either individual or collective, do not deflect from the value of genotyping, the users of genotyping systems or those responsible for interpreting results in the context of failures/investigations should be aware of them.

In initial employment of genotyping, the lack of familiarity with these techniques means they lend themselves to a high risk of, for example, nucleic acid contamination if they are not run in a suitable environment with controls in place. While this potential is largely eliminated with commercially available systems, there are still important new handling techniques to master. Furthermore, nucleic acid can be difficult to reliably extract from some organisms. In addition, the better discrimination provided by genotyping could introduce a communication difficulty, since there are potentially many more distinct sequences than names or phenotypic descriptions. In other words, it is possible to get a sequence for which there is no obvious phenotypic match. Also, gene sequencing (and any of the others) should not be seen as the sole criterion for assigning a cause in, for example, a sterility failure investigation; an isolate from an operator could possibly represent only one strain of mixed genetic population of the same species (e.g. Staphylococcus epidermidis). Therefore, if the isolate from the media fill failure does not match that isolated from the operator it should not necessarily rule out the operator as the cause.

Similar to commercially available phenotypic methods, genotypic systems rely on comparison of test results with a database. The database can be a key potential weakness of systems such as those for comparing gene sequence data. There are several reasons why sequence databases can vary and may not accurately link a name with a sequence, including: strains in certified collections being incorrectly named or classified by phenotypic means; the misplacement of species within a single genus, but actually found in many taxonomic groups. (e.g. Enterobacter, a polyphyletic genus); and the use of unverified public databases, which accept any linked name and sequence that is sent to them. Most system manufacturers provide validated databases of varying sizes, and these can only get bigger and more comprehensive over time.

If a laboratory is looking to invest in a new identification system, then the advantages of genotyping together with the easy of use of the available semi-automated systems make geno- typing an attractive alternative to phenotypic methods and, in the light of current regulatory expectations, perhaps the sensible option. However, the end-user together with the supporting quality organisation should be aware of the limitations. Genotyping is not necessarily a ‘one-size-fits-all’ technology and many laboratories may legitimately choose to remain faithful to phenotypic systems, with the contracting out of isolates for genotyping, where appropriate. Perhaps, in this time of rapidly changing technological capability, the pragmatic conclusion should be that the combination of all available stable markers, either phenotypic and/or genotypic is the best way to provide a satisfactory bacterial identification. However, the one thing that we must not forget is that in all cases good judgement by an experienced microbiologist is an absolute requirement in producing and analysing all the available information.