1b, and data not shown). The D values of EHEC O26 and O111 were comparable to the D value of EHEC O157 that was already proven to be useful in epidemiological analyses (14); the findings of this study suggest a sufficient discriminating power of the MLVA system. In the present study, the new MLVA system was also useful for detecting outbreak-related isolates, and this LY294002 mouse is one of the most prioritized objectives of genotyping (Fig. 3; Table 2). Most of the outbreak-related isolates did not exhibit any, or exhibited only single-locus, variations within each outbreak (Table 2). The cluster analysis based on the MLVA profiles revealed that each outbreak could

form a unique cluster. This was also true for the cluster analysis based on the PFGE profiles. Further, consistent results were obtained check details by both these methods (Figs 3, 4). However, the relationships between the clusters observed in one method differed from those observed in the other method because of the differences in the two methods with regard to the targets; MLVA discriminates isolates by repeat copy numbers of specific loci, whereas PFGE differentiates them by restriction fragment length polymorphisms of the entire DNA. Moreover, either PFGE or MLVA can be superior to the other method for discriminating isolates in some outbreaks. These results indicate that MLVA can complement

PFGE analysis. Considering that the procedure of MLVA is simpler and more rapid than that of PFGE, MLVA can be applied for the first screening of isolates in outbreak investigations before the results can be confirmed by PFGE. PFGE analysis is currently the golden method for subtyping bacterial pathogens. (13). Other researchers reported that subtyping methods, such as AFLP, rep-PCR and MLST, could be useful

for analyzing EHEC O157, but PFGE was the best method to discriminate isolates, for example, in outbreak investigations (17, 18). In this study, the results of MLVA were similar to those of PFGE analysis in outbreak investigations; this suggests that Edoxaban the discriminating power of MLVA is greater than that of the above-mentioned methods, although it might be necessary to evaluate the discriminating power of them for EHEC non-O157 strains, as described below. Furthermore, other methods are more time-consuming than MLVA. The results of the other methods, except MLST, are deduced from anonymous banding patterns, which can lead to ambiguous typing, whereas the results of MLVA are deduced from known loci and can be controlled by direct sequencing of the amplified products (19). Recently, infection with EHEC serogroups other than O157 has raised concerns not only in Japan but also in other countries: EHEC O26:[H11], O103:H2, O111:[H8], and O145:[H28] are frequently associated with HC and HUS (20). Although PFGE is the first line of choice for subtyping, most of the methods mentioned above have not yet been evaluated for analyzing EHEC non-O157 strains.