The DNA fragments in the agarose gel generated by PFGE are visualized by staining with ethidium bromide. Each lane on the gel represents the chromosomal pattern of one bacterial isolate. The migration of DNA fragments form patterns that determine chromosomal similarity and hence clonality of strains. The standardized recommendations (Tenover et al., 1995) for the interpretation of PFGE patterns of isolates linked to an epidemiological investigation are dependent upon the number of band differences on the gel. Those yielding the same pattern should be considered "indistinguishable," one to three band differences are "closely related," reflecting a single genetic change, four to six band differences are "possibly related," representing two independent genetic events, and six or more band differences represent three or more genetic changes and are considered "unrelated." However, comparisons of DNA fragment patterns present on multiple gels from large sets of isolates are technically difficult to interpret (Chung et al., 2000). There are variables that might alter fragment patterns and cause lack of interlaboratory reproducibility, such as type of PFGE instrumentation, protocols, or individual user techniques. Several commercially available software packages that provide computerized gel scanning and data analysis can compensate for these intra- and intergel variations (Duck et al., 2003). For example, the DNA patterns on the same or multiple gels can be more clearly represented as a dendrogram, showing the percent similarity obtained through Dice coefficients and the unweighted pair group method with arithmetic average. Through the use of computer-assisted analysis of DNA fragment polymorphism, investigators are able to create searchable databases of DNA patterns for multilaboratory comparison and for future strain comparisons.
Was this article helpful?