Abstract
Aberrant DNA methylation of promoter-associated CpG islands is a frequent event in ALL. DNA methylation of specific subsets of genes is associated with poor prognosis and is stable in a majority of patients (pts) at the time of relapse (
Clinical Cancer Res 2002;8:1897
), and therefore tracking these epigenetic marks during remission may predict for relapse risk. Most commonly used methods to detect DNA methylation exploit the availability of sodium bisulfite that transforms unmethylated C into A/T leaving methylated C as such. This has allowed the development of several techniques that use conventional PCR or sequencing to detect methylated alleles. Although there are significant differences in the sensitivity and specificity of these assays, with bisulfite sequencing considered the gold standard, most of them are labor intensive and require several days to be performed. To circumvent some of these problems, we have developed a real-time bisulfite PCR assay to detect methylation of p57KIP2, p73, and p15 in samples obtained from bone marrows in pts with ALL at remission. Methylation of these genes has been shown to be common in ALL and to predict for poor prognosis at initial presentation (Blood 2003;101:4131
). This method allows for the simultaneous analysis of multiple samples in less than 24 hours, is quantitative, and requires less than 0.2μg of DNA for each target gene. To amplify bisulfite treated DNA, we designed primer sets and probes for all three genes in regions known to be inversely correlated with gene expression. To quantify methylation density, we used the interferon γ (INFG) gene as an internal control because it has very rare CpG sites, is a single copy gene and has no homology with other known genes. Methylation density is defined as: Methylation (%) = 2CT of target gene / 2 CT of EFM x 100. CT is the number of cycles threshold, and EFM the estimated 100% fully methylation (EFM) of the target gene in control experiments. Using DNA artificially methylated with SssI in dilution experiments with unmethylated DNA, p57, p15 and p73 methylation density could be detected at dilutions of 0.2%, 1.2% and 0.1%, respectively. We then studied the methylation status of 30 cell lines (22 hematopoietic and 8 no-hematopoietic). Methylation of p57, p15 and p73 gene was detected in 17(57%), 19(63%) and 16(53%) cell lines respectively. Methylation of 3 genes, 2 genes and 1 gene was observed in 9(30%), 8(27%) and 7(23%) cell lines respectively. DNA methylation of p15 and p73 was not detected in the marrow from 6 healthy volunteers but p15 was detected (0.1% methylation) in 1 out of 6 of these specimens. Subsequently, we studied 50 pts with ALL at the time of remission. We found the frequencies of p57, p15 and p73 gene to be 2(4%), 26(52%) and 10(20%), respectively. There was a trend towards a better overall survival for pts with methylation of 0 or 1 gene (209 weeks) compared with those with methylation of 2 or more genes (71 weeks), p=0.1. In conclusion, the real-time bisulfite PCR described here allows for the rapid detection of aberrant DNA methylation in samples obtained at the time of remission in pts with ALL and may have a role in the development of techniques to assess minimal residual disease in this group of pts.Author notes
Corresponding author
2005, The American Society of Hematology
2004
