ABSTRACT
Although the role of 5-methylcytosine has been well studied, the biological role of 5-hydroxymethylcytosine still remains unclear because of the limited methods available for single-base detection of 5-hydroxymethylcytosine (5hmC). Here, we present mirror bisulfite sequencing for 5hmC detection at a single CpG site by synthesizing a DNA strand to mirror the parental strand. This semiconservative duplex is sequentially treated with ß-glucosyltransferase and M.SssI methylase. The glucosyl-5hmCpG in the parental strand inhibits methylation of the mirroring CpG site, and after bisulfite conversion, a thymine in the mirroring strand indicates a 5hmCpG site in the parental strand, whereas a cytosine indicates a non-5hmC site. Using this method, the 5hmC levels of various human tissues and paired liver tissues were mapped genomewide.
Subject(s)
5-Methylcytosine/analogs & derivatives , DNA/chemistry , Sequence Analysis, DNA/methods , Sulfites/chemistry , 5-Methylcytosine/analysis , Base Sequence , DNA-Cytosine Methylases/chemistry , Gene Library , Glucosyltransferases/chemistry , Humans , Reproducibility of Results , Sensitivity and SpecificityABSTRACT
5-hydroxymethylcytosine (5hmC) is an epigenetic modification, which has been associated with gene expression in many biological contexts. Reduced representation hydroxymethylation profiling was developed as an enzymatic-based method for genome-wide 5hmC detection. It exploits ß-glucosyltransferase to inhibit enzymatic cleavage of adapters ligated to a genomic library, allowing only fragments with glucosylated 5hmC residues at adapter junctions to be amplified and sequenced. The simple workflow and high sensitivity make it an efficient assay for 5hmC mapping. In this review, we discuss some technical consideration in applying reduced representation hydroxymethylation profiling, such as the use of alternative restriction enzymes for increased genomic coverage in different species, application of control libraries and specifications for multiplexing, data processing and normalization.
Subject(s)
Cytosine/analogs & derivatives , DNA Methylation , Epigenesis, Genetic , Epigenomics/methods , Gene Expression Profiling/methods , 5-Methylcytosine/analogs & derivatives , Animals , Computational Biology/methods , Cytosine/metabolism , Gene Library , Guidelines as Topic , High-Throughput Nucleotide Sequencing , HumansABSTRACT
Current methods for genomic mapping of 5-hydroxymethylcytosine (5hmC) have been limited by either costly sequencing depth, high DNA input, or lack of single-base resolution. We present an approach called Reduced Representation 5-Hydroxymethylcytosine Profiling (RRHP) to map 5hmC sites at single-base resolution by exploiting the use of beta-glucosyltransferase to inhibit enzymatic digestion at the junction where adapters are ligated to a genomic library. Therefore, only library fragments presenting glucosylated 5hmC residues at the junction are sequenced. RRHP can detect sites with low 5hmC abundance, and when combined with RRBS data, 5-methylcytosine and 5-hydroxymethylcytosine can be compared at a specific site.
Subject(s)
Cytosine/analogs & derivatives , 5-Methylcytosine/analogs & derivatives , Cytosine/physiology , DNA Methylation , Gene Library , High-Throughput Nucleotide Sequencing , Humans , Liver Neoplasms/genetics , Molecular Sequence Annotation , Sequence Analysis, DNAABSTRACT
We developed a novel approach, J-binding protein 1 sequencing (JBP1-seq), that combines the benefits of an improved recombinant JBP1 protein, Nextera-based library construction, and next-generation sequencing (NGS) for genome-wide profiling of 5-hydroxymethylcytosine (5hmC). Compared with the original JBP1, this new recombinant JBP1 was biotinylated in vivo and conjugated to magnetic beads via biotin-streptavidin interactions. These modifications allowed a more efficient and consistent pull-down of ß-glucosyl-5-hydroxymethylcytosine (ß-glu-5hmC), and sequence-ready libraries can be generated within 4.5h from DNA inputs as low as 50ng. 5hmC enrichment of human brain DNA using the new JBP1 resulted in over 25,000 peaks called, which is significantly higher than the 4003 peaks enriched using the old JBP1. Comparison of the technical duplicates and validations with other platforms indicated the results are reproducible and reliable. Thus, JBP1-seq provides a fast, efficient, and cost-effective method for accurate 5hmC genome-wide profiling.
