1. Introduction to Sequence-based typing of micro-organisms

Typing of microbes is essential to detect outbreaks or to forecast future trends. Historically, microbial pathogens have been typed by phenotypic means. However, these methods frequently lack discriminatory power and often have a low reproducibility. Therefore, starting in the nineties molecular typing techniques came into widespread use. This resulted in a plethora of techniques and protocols for subtyping even the same species of microbes. Because each laboratory uses its own protocols for molecular typing and designations of patterns, the results cannot be compared with those of other laboratories, even if both laboratories have used essentially the same methods. This lack of comparability has greatly diminished the power of molecular subtyping methods.

At the moment, pulsed-field gel electrophoresis (PFGE) is regard as the uniform molecular typing "gold standard". Great efforts were made in the past to harmonise the protocols and to adopt a standardised nomenclature. The "glue" that linked such past initiatives was always special adapted software (e.g. GelCompar or BioNumerics software). Nevertheless, such projects were only partially successful when judged by speed and costs of analysis (e.g. the American PulseNet initiative; Swaminathan et al. Emerg. Infect. Dis. 7: 382, 2001; DGXII project HARMONY; Murchan et al. J. Clin. Microbiol. 41: 1574, 2003).

2. Advantages of sequence-based typing

Since the introduction of the multi locus sequence typing (MLST) technology in 1998, DNA sequence based molecular typing techniques are becoming more and more popular. Sequence data are in contrast to "band-based only" methods (e.g. PFGE or RFLP) proof-read bands (simultaneous bidirectional sequencing is possible) that translate into something biological meaningful, i.e. the universal genetic code (once translated the results are independent from sequencing technology or sequencing machines). A nucleotide base is the smallest and most accurate unit in biology. The advantages of sequencing are therefore, the

• speed of data generation from DNA (i.e., not living strains but DNA can be sent to reference laboratories),
• unambiguous and highly reproducible nature of data,
• additive nature of data (data for "eternity"),
• highly portable data (electronic intra- and inter-laboratory exchange),
• easy storage of data in searchable databases,
• easy usage of a standardised nomenclature (central expanding databases),
• straight-forward phylogenetic reconstructions from data.

3. Why don't we use MLST?

MLST is currently too expensive and laborious and not suited for outbreak detection. We decided therefore to use a single locus target, i.e. the Staphylococcus Protein A (spa), for this proof of principle proposal. We did so, because (i) S. aureus is the most common cause of nosocomial infections world-wide, (ii) methicillin resistant S. aureus (MRSA) are of special concern in many European countries, (iii) spa was recently shown to be a highly discriminatory and stable single-locus marker that reflects excellent macro- and microvariation in S. aureus populations (e.g. Shopsin et al. J. Clin. Microbiol. 37: 3556-3563, 1999 and Koreen et al. J. Clin. Microbiol. 42: 792-799, 2004), and (iv) a software is available for immediate use (Harmsen et al., J. Clin. Microbiol. 41: 5442-5448, 2003). This software is especially suited for this project, because synchronisation of the client's software with a central server is implemented, thereby ensuring a uniform nomenclature usage. Furthermore, the software has base quality calling algorithms implemented that directly correlates to an error-probability of the called bases. By having a quantitative measure of the quality of a sequence, quality-control and -assurance procedures can easily be set up.

4. Why spa-typing?

spa is a tandem repeat DNA element. These repetitive DNA elements were recently shown by genome sequencing projects to occur - to some surprise - very frequently in prokaryotic genomes. A repetitive DNA element subgroup, the variable number of tandem repeats (VNTR), is increasingly be used for microbial subtypings, e.g. for the bioterroristic B. anthracis and Y. pestis agents (LeFle`che et al. BMC Microbiol. 1: 2, 2001). It is therefore envisioned that the project is creating a technologically advanced European network of excellence laboratories that in addition also encompasses in future the typing of other micro-organisms, such as bioterroristic agents by DNA sequencing.