Turning Up The Heat On DNA Methylation

Although methyl groups are uncharged and relatively small, they exert enormous influence over genomes. When enzymes add a methyl group to the DNA base cytosine, the modification tends to repress gene expression. Meanwhile, undermethylated genes are often highly expressed. It's no wonder then that researchers are keen to study this simple modification.

But for most molecular techniques, methylated DNA bases look just like their unmethylated counterparts. Now a study finds that researchers can detect methylation using a simple melting test (Anal. Chem., DOI: 10.1021/ac1024057).

Currently, biologists have three ways to detect DNA methylation: They can probe the DNA with restriction enzymes sensitive to methylation, enrich methylated DNA fractions using antibodies that bind methylcytosine, or treat the DNA with sodium bisulfite, which converts cytosine into uracil, but leaves methylcytosine alone.

The latter method allows scientists to pinpoint sequence-specific methylation patterns mainly through sequencing or high-resolution melting (HRM). This latter technique measures how the oligonucleotide's two strands unwind and pull apart during heating—a process called melting. (The bonding of G-C pairs requires more heat to break than those of the bisulfite-converted A-T base pair.)

But bisulfite chemistry is tedious and error-prone, because it requires 100% cytosine conversion. So, researchers have long sought something better.

Mike Wilkinson, a plant geneticist at Aberystwyth University in the U.K., and his team recently stumbled upon a simpler detection method while investigating whether epigenetic changes, such as DNA methylation, make it physically easier or harder to unwind DNA during transcription.

Basically, the researchers found that the bisulfite treatment step is unnecessary. They examined a series of untreated, long DNA oligonucleotides: some fully methylated, some fully unmethylated, and some methylated on only one strand. They then measured the oligonucleotides' melting properties with HRM, which involves adding a fluorescent dye that binds between base pairs, and slowly heating the DNA. As the temperatures increase, the molecules unwind and release the bound dyes, quenching their fluorescence.

Methylation led to greater thermal stability, with the DNA melting at higher temperatures. The technique was sensitive enough to distinguish between unmethylated molecules and DNA methylated at 1% of its base pairs, and to differentiate molecules containing no methyl groups from those containing one modification per strand.

"It's very quick and straightforward," Wilkinson says: The assay takes about 15 minutes and requires only a fluorescent dye and an HRM machine.

But researchers shouldn't toss out their bisulfite kits just yet, says Wilkinson. The technique is not sensitive enough to probe sequence-specific methylation, an issue that Wilkinson's team is working to address. Currently, the method works best when studying genome-scale methylation changes, such as screening for methylation pathway mutants or comparing tissue culture clones to their parental strains for epigenetic changes.

Rama Badugu, a technical services consultant at Roche Applied Science, which markets an HRM instrument and assay, says that the study represents "a good start." But he cautions that the researchers need to run more optimizations and controls to ensure that the technique can handle complex, variably methylated, real-world genomic samples.