In many organisms the primary DNA structure is covalently modified to regulate, for example, gene expression and genome structure. In eukaryotes, the dominant modification is methylcytosine, although others, such as hydroxymethylcytosine, have been detected. In bacteria, both methylcytosine and methyladenine are observed frequently. None of the currently available sequencing platforms can directly detect modified bases, and researchers rely on indirect methods such as bisulfite treatment, methylation-sensitive restriction enzyme mapping or affinity precipitation methods.

In a recent paper1 in Nature Methods, researchers at Pacific Biosciences (Menlo Park, CA, USA) have now shown that a single-molecule, real-time, sequencing-by-synthesis platform based on their zero-mode waveguide technology2 can distinguish methylcytosine, hydroxymethylcytosine and methyladenine from unmodified deoxynucleotides in sequences whose methylation patterns are known. As described previously2, the base sequence is determined by monitoring incorporation into the growing chain by a single DNA polymerase of nucleotides tagged with four different fluorescent colors. In the new study, the presence of covalent modifications in the template strand is identified through two kinetic parameters: the time interval between the addition of adjacent nucleotides and the length of each catalytic cycle (beginning with the binding of the fluorescent base to the enzyme and ending with the release of the fluorophore attached to the terminal phosphate of the nucleotide). Both parameters are influenced by the presence of methylcytosine, hydroxymethylcytosine and methyladenine, not only at positions opposite the incoming nucleotide but also at several adjacent positions. The authors use synthetic templates and DNA purified from Escherichia coli to define the kinetic signatures of specific modifications at a given position.

Although de novo determination of methylation patterns is not reported, the detection of methyladenine seems feasible using circular templates that allow the repeated interrogation of each base. Robust detection of methylcytosine and hydroxymethylcytosine and of multiple modified bases in close proximity (as in CpG islands) will require further optimization of the method.