Quantitative DNA methylation profiling on a high-density oligonucleotide microarray.

Recently, the analysis and functional elucidation of CpG island methylation has become a focus area of genomic research. Deviations from the normal parental imprinting pattern have been shown to cause developmental defects associated with serious symptoms. Aberrant DNA methylation of tumor suppressor and other functional genes, especially when found in 5' untranslated regions and early exons, has been associated with tumorigenesis. In the context of applying DNA methylation analysis for the molecular characterization of cancer and other diseases, standardized protocols enabling parallel genome-wide methylation profiling of numerous samples are required. DNA methylation profiling is described using a CpG island microarray representing more than 50,000 CpG-rich DNA fragments. Fragments were selected to represent the vast majority of known 5'-untranslated regions as well as the first exons of thousands of genes. Measurement probes were designed to represent these fragments were displayed on an Affymetrix custom array. A modified procedure for differential methylation hybridization (DMH) is described for methylation enrichment. Application of a novel signal normalization concept enables accurate and reproducible measurements using a single fluorescence channel. The use of defined calibrator material allows quantification of DNA methylation patterns by DMH in a massively parallel fashion.

Tumour class prediction and discovery by microarray-based DNA methylation analysis.

Aberrant DNA methylation of CpG sites is among the earliest and most frequent alterations in cancer. Several studies suggest that aberrant methylation occurs in a tumour type-specific manner. However, large-scale analysis of candidate genes has so far been hampered by the lack of high throughput assays for methylation detection. We have developed the first microarray-based technique which allows genome-wide assessment of selected CpG dinucleotides as well as quantification of methylation at each site. Several hundred CpG sites were screened in 76 samples from four different human tumour types and corresponding healthy controls. Discriminative CpG dinucleotides were identified for different tissue type distinctions and used to predict the tumour class of as yet unknown samples with high accuracy using machine learning techniques. Some CpG dinucleotides correlate with progression to malignancy, whereas others are methylated in a tissue-specific manner independent of malignancy. Our results demonstrate that genome-wide analysis of methylation patterns combined with supervised and unsupervised machine learning techniques constitute a powerful novel tool to classify human cancers.