Biomathematics & Statistics Scotland
The bacterium Escherichia coli O157 presents a serious public health problem in Scotland: shed asymptomatically by infected cattle, it has the potential to cause serious illness, renal failure and even death in humans. Different populations of E. coli O157 present in field samples can be distinguished using Pulsed-Field Gel Electrophoresis (PFGE), a subtyping technique which uses a restriction enzyme to “chop-up” the DNA into pieces of different sizes. We have been studying a strain (EDL933) of E. coli O157: H7 whose PFGE pattern, obtained using the restriction enzyme Xba1, differs from the pattern predicted from its DNA sequence.
Having found no errors in the reported DNA sequence in the regions near the predicted restriction sites, we sought to to establish another explanation for the discrepancy. The distribution of dam methylation sites within the genome was investigated in silico and a new PFGE pattern was predicted by assuming that, whenever the restriction enzyme Xba1 was expected to cut the DNA at a dam methylation site, no such cut occurred. The new predicted pattern agrees closely with that observed in practice. This work is important in enhancing our understanding of the relationship between PFGE banding patterns and the underlying DNA sequences of pathogens.
Left: Digitised image of a PFGE image for strain EDL933 after Xba1 digestion, showing only five clear bands between 200 and 300 kb; the arrow indicates where an additional band is sometimes present.
Right: Summary PFGE banding patterns for strain EDL933 after Xba1 digestion. The “Observed” column shows mean band lengths that were present on each of six separate gels; the remaining columns show the predicted band lengths depending on whether methylated restriction sites for Xba1 were cut or not.
Further details from: Iain McKendrick
Article date 2007
