Direct comparison of pulsed-field gel electrophoresis and multilocus sequence typing
Abstract number: 1733_1348
Dobay O., Rozgonyi F., Walsh F., Diggle M., Nagy K., Amyes S.
Objectives: Pulsed-field Gel Electrophoresis (PFGE) and Multilocus Sequence Typing (MLST) are thought to have similar discriminatory power. However, although the results from MLST may be more transportable which may be useful in global epidemiological studies, PFGE is more sensitive to smaller genomic arrangements and potentially more discriminatory. In this study, we compare the two methods directly with Streptococcus pneumoniae.
Methods: PFGE was performed on 30 routine Scottish clinical isolates of serotype 14 with identical MLST pattern (sequence type, ST). MLST was performed on 26 Hungarian isolates of very similar PFGE patterns, 12 of serotype 6A and 14 of 23F. Serotyping was performed with the MAST antisera.
Results: PFGE examination of the 30 isolates with identical ST revealed small differences in the banding pattern indicating very little diversity in this group. On the other hand, MLST of the 26 strains with similar PFGE pattern resulted in two main STs (i.e. differences in 6 loci out of 7). One ST comprised the 12 isolates of serotype 6A with identical PFGE type. The other ST contained 11 isolates of serotype 23F, but these had 13 PFGE band differences. The remaining 3 strains (23F) belonged to two further STs, but these had also only 13 band differences from the others.
Conclusions: Strains identified as a single ST, did show some diversity when tested by the more sensitive discriminator of PFGE. Also in the other way around, strains with 13 PFGE band differences, resulted mostly in identical ST. Our results suggest that for more diverse bacteria such as pneumococci, MLST categorises them into sharply distinct groups, as it is based on sequences, while PFGE allows a bit more genetic variation, being based on fragment differences. Therefore, PFGE seems to be better at identifying smaller changes that occur at a certain geographical region within a shorter time period, while MLST is probably more useful for international comparisons. Interestingly clustering by both methods showed little diversity within individual serotypes. However, although the two techniques are very different in principle, they are equally important in epidemiology.
|Session name:||European Society of Clinical Microbiology and Infectious Diseases|
|Location:||ICC, Munich, Germany|
|Back to top|