5-Hydroxymethylcytosine (5hmC) has always been a bear to sequence, mostly because it’s virtually indistinguishable from its close relative, 5-methylcytosine (5mC) using classic bisulfite sequencing techniques. Now a new sample preparation method has hit the scene that can discriminate between cytosine, 5mC and 5hmC at single base resolution, called oxidative bisulfite sequencing (oxBS-Seq) that essentially smashes the 5hmC-sequencing barrier.
Scientists from the Babraham Institute and University of Cambridge solved the 5hmC sequencing problem by implementing a new twist on the familiar bisulfite conversion method. Normally, bisulfite conversion doesn’t produce a difference in readout between 5mC and 5hmC, but the researchers found that treating DNA with potassium perruthenate (KRuO4) selectively oxidizes 5hmC into 5-formylcytosine (5fC), which after bisulfite conversion is read as uracil, just like unmethylated cytosine. So all you have to do is subtract out an oxBS-Seq data set from regular bisulfite sequencing (BS-Seq) results and, presto! You’ve got a high-resolution map of 5hmC locations for your sample.
The Cambridge-Babraham crew put the oxBS-Seq approach to work by testing it out on mouse embryonic stem cell DNA. They combined oxBS-Seq with reduced representation bisulfite sequencing (RRBS) in order to look more closely at CpG islands (they covered and estimated 55% of CGIs in the mouse genome). Here are a few things they uncovered:
- 800 CpG islands had 5hmC
- Those CGIs averaged 3.3% hydroxymethylation
- The highest 5hmC levels were found in transcriptional regulator and LINE1 element CGIs, hinting to researchers that these regions undergo epigenetic reprogramming in ES cells.
The oxBS-Seq inventors think their innovation will enable more in depth investigation into 5hmC dynamics and the epigenetic role of TETs.
Get the full oxBS-Seq experience at Science, April 2012.