Typing methods for discriminating different bacterial isolates of the same species are essential epidemiological tools in infection prevention and control. Traditional typing systems based on phenotypes, such as serotype, biotype, phage-type, or antibiogram, have been used for many years. However, more recent methods that examine the relatedness of isolates at a molecular level have revolutionised our ability to differentiate among bacterial types and subtypes. Importantly, the development of molecular methods has provided new tools for enhanced surveillance and outbreak detection. This has resulted in better implementation of rational infection control programmes and efficient allocation of resources across Europe. The emergence of benchtop sequencers using next generation sequencing technology makes bacterial whole genome sequencing (WGS) feasible even in small research and clinical laboratories. WGS has already been used for the characterisation of bacterial isolates in several large outbreaks in Europe and, in the near future, is likely to replace currently used typing methodologies due to its ultimate resolution. However, WGS is still too laborious and time-consuming to obtain useful data in routine surveillance. Also, a largely unresolved question is how genome sequences must be examined for epidemiological characterisation. In the coming years, the lessons learnt from currently used molecular methods will allow us to condense the WGS data into epidemiologically useful information. On this basis, we have reviewed current and new molecular typing methods for outbreak detection and epidemiological surveillance of bacterial pathogens in clinical practice, aiming to give an overview of their specific advantages and disadvantages.

The multidrug-resistant Streptococcus pneumoniae Taiwan19F-14, or PMEN14, clone was first observed with a 19F serotype, which is targeted by the heptavalent polysaccharide conjugate vaccine (PCV7). However, “vaccine escape” PMEN14 isolates with a 19A serotype became an increasingly important cause of disease post-PCV7. Whole genome sequencing was used to characterize the recent evolution of 173 pneumococci of, or related to, PMEN14. This suggested that PMEN14 is a single lineage that originated in the late 1980s in parallel with the acquisition of multiple resistances by close relatives. One of the four detected serotype switches to 19A generated representatives of the sequence type (ST) 320 isolates that have been highly successful post-PCV7. A second produced an ST236 19A genotype with reduced resistance to β-lactams owing to alteration of pbp1a and pbp2x sequences through the same recombination that caused the change in serotype. A third, which generated a mosaic capsule biosynthesis locus, resulted in serotype 19A ST271 isolates. The rapid diversification through homologous recombination seen in the global collection was similarly observed in the absence of vaccination in a set of isolates from the Maela refugee camp in Thailand, a collection that also allowed variation to be observed within carriage through longitudinal sampling. This suggests that some pneumococcal genotypes generate a pool of standing variation that is sufficiently extensive to result in “soft” selective sweeps: The emergence of multiple mutants in parallel upon a change in selection pressure, such as vaccine introduction. The subsequent competition between these mutants makes this phenomenon difficult to detect without deep sampling of individual lineages.

No abstract

Real-time PCR targeting lytA (the major autolysin gene) and piaB (permease gene of the pia ABC transporter) are currently used as the gold-standard culture-independent assays for Streptococcus pneumoniae identification. We evaluated the performance of a new real-time PCR assay – targeting SP2020 (putative transcriptional regulator gene) – and compared its performance with the assays previously described. A collection of 150 pneumococci, 433 non-pneumococci and 240 polymicrobial samples (obtained from nasopharynx, oropharynx, and saliva; 80 from each site) was tested. SP2020 and lytA -CDC assays had the best performance (sensitivity of 100% for each compared to 95.3% for piaB ). The specificity for lytA and piaB was 99.5% and for SP2020 was 99.8%. Misidentifications occurred for the three genes: lytA , piaB and SP2020 were found in non-pneumococcal strains; piaB was absent in some pneumococci including a serotype 6B strain. Combining lytA and SP2020 assays resulted in no misidentifications. Most polymicrobial samples (88.8%) yielded concordant results for the three molecular targets. The remaining samples seemed to contain non-typeable pneumococci (0.8%), and non-pneumococci positive for lytA (1.7%) or SP2020 (8.7%). We propose that combined detection of both lytA -CDC and SP2020 is a powerful strategy for the identification of pneumococcus either in pure cultures or in polymicrobial samples.