DNA methylation and hydroxymethylation of cytosine are not the same, and yet for many years, researchers have been unable to tell the difference. In fact, says Gary Hon from the University of California, San Diego's Ludwig Institute for Cancer Research, though researchers know that hydroxymethyl-ation of cytosine affects embryonic stem cell differentiation and plays a role in leukemia, they are still unsure of the mechanisms by which it works. "While we do know that 5-hydroxymethylation exists, we don't really know what it does," Hon says.
A big part of the problem is that 5-hydroxyl-methylcytosine — the so-called "sixth nucleotide" — is indistinguishable from 5-methylcytosine using current techniques, Hon adds. Bisulfite sequencing — the gold--standard method for the detection of DNA methylation — is actually detecting the sum of methylation and hydroxy-methylation. "If you see a peak, you don't know which one it really is. And those two marks mean totally different things," Hon says. "DNA methylation is a more repressive modification, whereas hydroxymethylation is thought to be more active. So if you can't tell an active from a repressive mark, you're in trouble."
Some researchers have devised affinity-based methods to detect hydroxy-methylation, but the problems with those approaches are that they generally can't precisely pinpoint which base is hydroxymethylated, nor can they pinpoint how much hydroxy-methylation there is, Hon says. These methods are also prone to sequence bias.
To address these problems, Hon, along with his UCSD colleagues, and their collaborators at the University of Chicago and Emory University devised a new method based on bisulfite sequencing that is able to tell the difference between methylation and hydroxymethylation. They described their method, called TAB-seq, in Cell in May. In bisulfite sequencing, says Chicago's Chuan He, 5-methyl-cytosine is distinguished from regular cytosine through the process of deamination, which turns all un-methylated cytosine into uracil.
TAB-seq protects the hydroxyl group with a sugar residue and then performs an oxidation step, so that all methyl-cytosine is converted to carboxylcytosine. This way, he adds, when deamination is performed, the newly converted carboxyl-cytosine is turned into uracil like the other cytosine, and all that is left is 5-hydroxymethyl-cytosine. "What the method provides is a way to carefully examine whether this is really methyl-C or hydroxymethyl-C, because they may have different implications for disease and treatment," He says. "The method is general. You can do it on the whole genome, or more likely people will use it in a loci-specific manner."
Hon adds that two groups of researchers — both those studying hydroxy-methylation and methylation — will be able to use TAB-seq: the first group to understand hydroxymethylation and the second group to make sure the purported methylation they are studying isn't also mixed in with hydroxymethylation. "This is the missing link we've needed to not only further dissect the function of 5-hydroxy-methylation but also to correctly interpret what we previously thought was DNA methylation," he adds.