Methylation-free site patterns along a 1-Mb locus on Chr19 in cancerous and normal cells are similar. A new fast approach for analyzing unmethylated CCGG sites distribution

We describe a newly developed technique for rapid identification of positions of genomic DNA breaks, preexisting or introduced by specific digestion, in particular, by restriction endonucleases (RIDGES). We applied RIDGES in analyzing unmethylated CCGG sites distribution along a 1-Mb long genome region (D19S208-COX7A1 on chromosome 19) in cancerous and normal lung tissues. Both tissues were characterized by a profoundly uneven density of unmethylated sites along the fragment. Interestingly, the distribution of hypomethylated regions did not correlate with gene locations within the fragment, and one of the most hypomethylated areas contained practically no genes. We also demonstrated that the methylation pattern of a long genome DNA fragment was rather stable and practically unchanged in human lung cancer tissue as compared with its normal counterpart, in accordance with the suggestion (Ross et al. in Nat Genet 24:227-235, 2000) that cell lines of common origin have typically similar transcription profiles. An analogous suggestion might probably be made for global methylation patterns of genomic DNA.