A method to estimate the percent loss of cytosine methyl groups at defined CpG sites in liver DNA from methyl-deficient rats

Abstract

Dietary methyl deficiency provides an ideal in vivo model system in which to study progressive alterations in DNA methylation patterns as they occur during multistage hepatocarcinogenesis. Weanling male F344 rats were given a semipurified diet deficient in the methyl-donors choline, methionine, and folic acid for a 36-week period with sampling intervals at 3, 9, 24, and 36 weeks. Using a genomic sequencing procedure based on the PCR amplification of bisulfite-modified DNA, the methylation status of individual CpG sites within exons 6 and 7 of the p53 gene in liver samples from control and deficient rats was assessed. Treatment of denatured nuclear DNA with sodium bisulfite converts unmethylated cytosine residues to uracil, which are then amplified as thymine in the PCR reaction. In contrast, methylated cytosines are resistant to bisulfite deamination under these reaction conditions and are amplified as cytosine. In this report, we describe a novel application of automated sequencing technology to estimate the proportion of methylated cytosines present at defined CpG sites within the total population of DNA molecules extracted. Using the bisulfite conversion-PCR genomic sequencing method, we demonstrate the validity of peak height analysis of co-eluting peaks in the autosequencer electrophoregram to estimate the percent methylation at a defined CpG site. The sensitivity of this method is demonstrated by the progressive loss of methyl groups at a defined CpG site in the methyl-deficient rats after 9, 24, and 36 weeks. The application of this sequence-specific technology will allow site-specific definition of the methylation status of each CpG site within a coding sequence or promoter region and should provide new insights into mechanisms and consequences of methylation dysregulation as a result of dietary deprivation of methyl donors.