Profiling DNA methylation patterns using genomic tiling microarrays

The epigenetic inheritance or 'memory' of transcriptional activity is regulated through the covalent modification of chromatin and the binding of specific chromosomal proteins. Epigenetic modifications include the methylation of cytosines on DNA(1) as well as the acetylation, methylation, phosphorylation and ubiquitylation of the N-terminal tails of the core histone proteins H2A, H2B, H3 and H4 (ref. 2). Densely stained regions of chromosomes, known as heterochromatin, are large components of eukaryotic genomes that comprise primarily repetitive DNA sequences such as satellite repeats and transposons. These regions are heritably silenced and enriched in DNA methylation and methylation of histone H3 Lys9-two evolutionarily conserved modifications involved in transcriptional repression. In some cases, heterochromatic modifications can affect the activity of euchromatic genes that influence phenotype(1). Furthermore, aberrant DNA methylation is associated with some human diseases and is a hallmark of cancer progression(3). Therefore, mapping the epigenetic modifications of entire genomes of various organisms to decipher the so-called 'epigenome' will further our understanding of these processes. This protocol, modified from ref. 4, provides a description of the use of methylation-sensitive restriction enzyme digestion and microarrays to profile DNA methylation patterns (Fig. 1). In this method the restriction enzyme McrBC, which prefers methylated DNA as a substrates, is used to selectively exclude the methylated fraction of DNA from a genotype of interest. This sample is then compared to matched 'untreated' genomic DNA of the same genotype in a microarray hybridization. Genomic tiling microarrays, which represent contiguous stretches of chromosomes without bias toward coding sequences, allow DNA methylation patterns of all sequence types to be assayed simultaneously at high resolution. This combination of toots has been used successfully in plants(4). This protocol has been optimized for Arabidopsis thaliana and should provide a reliable and robust way to determine DNA methylation patterns in more complex genomes, such as the human genome. A protocol that outlines the profiting of histone methylation patterns at similar high resolution has also been developed(6).