Small variations in the DNA sequence of a gene, called polymorphisms, are known to have an impact on normal and disease phenotypes. Some polymorphisms change the length or makeup of the resulting protein, whereas others interfere with binding sites for molecules that regulate the gene’s expression. Andrew Chess and his colleagues at Massachusetts General Hospital are exploring a new way by which polymorphisms can control gene regulation - by influencing the addition of methyl groups to the DNA strand, impeding its expression. Most studies of DNA methylation have focused on discrete clusters at the beginning of genes, but Chess and colleagues are looking more broadly at methylation near polymorphisms throughout the genome. The researchers plan to explore whether and how a mechanism involving polymorphisms and methylation might be at work in autism. To analyze methylation across the genome, the researchers digest the entire genome using special enzymes that break the DNA strand at all potential methylation sites, except those that are actually methylated. The remaining DNA fragments can then be analyzed to determine which polymorphisms they represent. Chess and his team plan to use this technique to look for differences in methylation patterns between individuals with autism and healthy controls. They also plan to analyze identical twin pairs in which only one twin has autism in the hopes of uncovering a reason for the difference. The findings may lead to the discovery of genes that cause autism when their expression is altered by methylation. Researchers have already linked polymorphisms to methylation in DNA from unaffected individuals, with the interesting qualification that the variations specify only a greater propensity toward methylation and not necessarily constant methylation. This highlights the importance of analyzing methylation status, as looking at genetic sequence alone may not provide the complete picture on heritable DNA alterations. Another useful feature of the methylation sites they are examining is that the patterns are similar regardless of whether the genes are turned on or off, so their impact on autism can be studied using easily accessible cells such as blood cells rather than requiring brain samples. The researchers’ new technique may supplement the usual genetic analyses in studying the complex heritability of autism.