Evaluation of 10 methods to distinguish epidemic-associated Campylobacter strains

We compared four phenotypic and six genotypic methods for distinguishing Campylobacter jejuni strains from animals and humans involved in four epidemics. Based on a comparison with epidemiologic data, the methods that correctly identified all strains in three milkborne outbreaks and one waterborne outbreak were heat-stable and heat-labile serotyping; multilocus enzyme electrophoresis (MEE); DNA restriction endonuclease analysis with BglII, XhoI, PvuII, or PstI; and Southern blot and hybridization of PvuII- and PstI-digested DNA with Escherichia coli 16S and 23S rRNA (ribotyping). Biotyping, phage typing, plasmid analysis, and probing of BglII and XhoI DNA digests with C. jejuni 16S rRNA genes failed to correctly separate one or more strains. MEE, restriction endonuclease analysis, and ribotyping were the most sensitive methods and identified nine types among the 22 strains. These methods were also capable of further distinguishing strains within the same serotype. Data from MEE were also analyzed to calculate genetic relatedness among strains. Serotyping was the most discriminating phenotypic method, with eight and seven types distinguished by the heat-stable and heat-labile methods, respectively. MEE and ribotyping had several advantages over the other methods because they measure relatively stable and significant chromosomal differences and are applicable to other species and genera. These methods, however, are complex and not easily quantified; they are currently limited to specialized laboratories. When antisera are available, serotyping appears to be an effective and more practical approach to the identification of epidemic-related strains.