CpG island methylation and associated gene silencing has been recognized as a very common event in human neoplasia, with potential to functionally contribute to tumor formation and progression. It has recently become apparent that CpG island methylation events are not random. Rather, there are substantial differences in methylation patterns between different tumors, as well as evidence of hypermethylator phenotypes in subsets of cases. These methylation differences are associated with distinct environmental exposures, distinct gene mutation patterns, and potentially important differences in progression and survival following standard therapies. Methylation profiling of human cancers may therefore prove useful both in etiologic studies of cancer, and in efforts to relate molecular changes to prognosis and treatment decisions. We have recently developed a new technique termed methylated CpG island amplification (MCA), which allows for the rapid methylation profiling of primary tumors. MCA is based on sequential restriction enzyme digestion with methylation-sensitive/insensitive isoschizomers, adaptor ligation and whole-methylated-genome PCR. In pilot studies, MCA appeared to be accurate, relatively high-throughput, and led to the identification of a new methylator phenotype in colorectal cancer. Here, we propose to continue developing MCA by simplifying the procedure to make it readily applicable to the analysis of primary tumors, and by developing tumor-specific CpG island arrays that can be used for profiling most of the major malignancies. Using MCA, the methylated CpG island pools of a given primary tumor can be labeled and hybridized to pre-made filters or DNA chips containing the CpG islands critical to the tumor type under study. This procedure could then reveal the methylation profiles of hundreds of CpG islands in a given tumor in less than 72 hours, without requiring sophisticated reagents or gel electrophoresis. None of the currently existing technologies for methylation analysis can achieve this rapidly or cost-effectively. Ultimately, methylation profiling may prove very valuable in identifying subsets of patients with distinct clinical courses and response to specific therapeutic interventions, including using methylation inhibitors.
