A reassessment of semiquantitative analytical procedures for DNA methylation: comparison of bisulfite- and HpaII polymerase-chain-reaction-based methods

Two techniques in particular are used to study site-specific DNA methylation: genomic sequencing after bisulfite modification and polymerase chain reaction after digestion by a methylation-sensitive endonuclease (usually HpaII). Only the former methodology assesses the methylation status of all the cytosine residues in the DNA sequence, but it is so complex and time consuming that the latter procedure, though limited to the restriction sites recognized by the endonuclease(s) used, is often preferred at least for a first analysis. In this work we investigate differences between these two techniques in the assessment of DNA methylation and offer some suggestions on how to avoid uncorrected results. Although there is substantial accordance in the results obtained using these different techniques, we observed a general overestimate for methylation levels above 30% and a general underestimate for methylation levels below this value using the HpaII/PCR technique in the study of methylation of the 5'-flanking region of the mouse myogenin gene in cultured muscle cells and mouse tissues.