Step I: DNA Glucosylation Reaction with T4 β-glucosyltransferase (T4-BGT) Genomic DNA of interest is treated with T4-BGT, adding a glucose moeity to 5-hydroxymethylcytosine. This reaction is sequence-independent - therefore all 5-hmC will be glucosylated, unmodified or 5-mC containing DNA will not be
Step II: Restriction Endonuclease Digestion
MspI and HpaII recognize the same sequence (CCGG) but are sensitive to different methylation status. HpaII cleaves only a completely unmodified site: any modification (5-mC, 5-hmC or 5-ghmC) at either cytosine blocks cleavage. MspI will recognize and cleave 5-mC and 5-hmC, but not 5-ghmC.
Step III: Interrogation of the Locus by PCR as little as 20 ng of input DNA can be used. Amplification of the experimental (glucosylated and digested) and control (mock glucosylated and digested) target DNA with primers flanking a CCGG site of interest (100–200 bp) is performed. If the CpG site contains 5-hydroxymethylcytosine, a band is detected after glucosylation and digestion, but not in the non-glucosylated control reaction (see Figure 2). Real time PCR will give an approximation of how much hydroxymethylcytosine is in this particular site.
Figure 1a: Experimental Overview
The DNA of interest is treated with T4 β-Glucosyltransferase (T4-BGT) and UDP-Glucose (UDP-Glc). T4-BGT transfers glucose from UDP-Glc onto 5-hydroxymethylcytosine (generating glucosylated 5-hydroxymethylcytosine [5-ghmC]). MspI cuts DNA containing 5-hmC, but does not cut 5-ghmC containing sites; in contrast, HpaII is blocked by any of these modifications. Presence of 5-hmC and 5-mC can be determined by PCR analysis.
Figure 1b: Experimental Overview
The DNA of interest is digested following a control reaction with UDP-Glucose (UDP-Glc) and no T4 β-Glucosyltransferase (T4-BGT), leaving 5-hmC unmodified. MspI cleaves unmodified, 5-mC and 5-hmC DNA, while HpaII cleaves only unmodified DNA.
Figure 2: Comparison of 5-hydroxymethylcytosine amounts at locus 12 in different mouse Balb/C tissue samples. (A) End-point PCR. (B) Real time PCR.
DNA from four mouse tissues was analyzed. For comparative purposes, real time PCR data were normalized to uncut DNA. A standard curve was used to determine copy number. The samples could be normalized by dividing the copy number of samples No 1-6 by the copy number of the control that is undigested (No 5). Boxed gel lane shows variation in 5-hmC present.
Figure 3: High sensitivity 5-hydroxymethylcytosine detection achieved by the EpiMark kit.
100 bp unmodified, 5-mC, and 5-hmC control DNAs were mixed in different ratios (blue bars), and then measured with the EpiMark hydroxymethylated DNA detection kit (orange bars). Error bars represent the standard deviation of four independent experiments.