Serotyping

Methods

Serotyping has traditionally played an essential role in determining species and subspecies, and has been used for epidemiologic classification down to subspecies level. The method is based on serological typing of the cell surface antigens of the bacteria. A serotype (syn. serovar: International Code of Nomenclature of Prokaryotes, 2019, p134) corresponds to the combination of surface structures or antigens. This has been proved to be an effective way of discriminating groups of bacteria, with some serotypes being host specific and others associated with virulence intensity (high/mild) (CDC). Moreover, further research experiments have associated serotypes to particular features such as pathological properties, susceptibility to antimicrobials and niche distribution (eg. Ørskov, F. and Ørskov, I. 1992, Yang, X. et al. 2015, Lee, S. et al. 2018, Zoz, F. et al. 2017).

Traditional serotyping (wet lab) relies on the detection of cell surface antigens by agglutination assays using species-specific antibodies (see the Species specific sections for which ones). Serotyping has been used in epidemiology since 1960 in the wet-lab to detect Salmonella outbreaks, and as per today more than 2,500 have been described in this species. As the expression of surface structures of each pathogen are coded in their genome, molecular typing methods based on the detection/amplification of certain alleles or DNA sequences have been developed (eg. Beaubrun et al. 2012, Borucki et al. 2003, Iguchi et al. 2015). The advent of WGS technologies brought a higher discriminatory power for genetic clustering without losing the ability to link genomic information to previously available knowledge. This has led to the implementation of a gradual technological transition and to the need of using WGS data for serotyping. Some studies have compared traditional serotyping with WGS-based serotyping (eg: Salmonella, Escherichia coli, Listeria monocytogenes).

Whole-genome based serotyping can be done in different ways:

  • By emulating laboratory methods such as presence/absence DNA-fragments detection PCR (see ePCR section)
  • By detecting alleles within a set of antigen-genes loci (serotyping performed similarly as MLST finding, see MLST method description)
  • By specific genes identification, by either mapping or BLAST searches towards serotype determinants (similarly to AMR and virulence finding.)

In many cases, the tools used for MLST and AMR finding can also be used for serotype finding. When marker genes are associated to at set specific antigens are defined, it is possible to determine serogroups based on the presence/ absence of the PCR amplification pattern of a combination of specific gene markers (eg. Doumith et al. 2004). PCR based serotyping can be bioinformatically emulated (ePCR, in silico PCR), and serotypes can thus be inferred from whole genome sequencing data. In such cases presence / absence of genes can be detected using blast in conjunction with a set of rules determining when to consider a match presence or not (Eg. as performed on assemblies in LisSero for ePCR serotyping of Listeria monocytogenes, or In Silico PCR for fliC and filB alleles of the H antigens for Salmonella serotyping with assemblies in SeqSero1.