NEW YORK (GenomeWeb News) – A team of researchers from China, Denmark, and the UK reported online last night in PLoS Biology that they have sequenced the methylome of an Asian individual to single base-pair resolution.
Using bisulfite sequencing, the researchers gleaned methylation data from both strands of DNA isolated from the peripheral blood mononuclear cells of an anonymous Han Chinese man participating in the BGI-led YanHuang Project — an effort to sequence the genomes of 100 Chinese individuals over three years. The team combined this data with previously published genome sequence information for the same individual.
"[O]ur work provides a first proof-of-concept for the importance of including [allele-specific methylation] in methylome analyses," co-corresponding author Xiuqing Zhang, vice president of BGI, and his co-authors wrote.
For instance, investigators were able to track down nearly 600 sites at which methylation patterns differed between alleles inherited from the man's mother and father. They also detected dozens of genes showing both allele-specific methylation and allele-specific expression, providing new clues about genetic imprinting processes in which a version of a gene inherited from one parent is preferentially expressed.
Zhang and his colleagues used the Illumina Genome Analyzer to do whole-genome bisulfite sequencing to generate methylation data for about 88 percent of cytosine guanine dinucleotide, or CpG, sites in the genome at nearly 25 times depth — about 12 fold coverage per DNA strand.
The researchers found methylation at more than 68 percent of the roughly 20 million CpG sites evaluated. In contrast, they noted, less than 0.2 percent of non-CpG sites appear to be methylated in the blood cells.
With the genome and methylome data in hand, the team was able to delve into the methylation patterns associated with coding and non-coding, regulatory, and repeat sequences in the genome.
For instance, the team explained, they typically saw lower methylation levels around the transcription start sites of expressed genes and higher methylation levels on expressed gene bodies. On the other hand, silent gene transcription start sites tended to be more highly methylated than start sites of expressed genes.
Non-coding RNA genes, meanwhile, showed methylation profiles that varied with RNA type, the researchers explained, while repeat elements had different levels of methylation depending on whether they were still active or not.
The team also compared methylation patterns between DNA strands, looking for allele-specific methylation differences at 1.17 CpG sites. In the process, they uncovered 599 so-called haploid differentially methylated regions corresponding to 287 genes.
Their data underscored the interplay between allele-specific expression and allele-specific methylation, since more than 80 percent of the genes with haploid differentially methylated regions within 2,000 bases of their transcriptional started sites also showed allele-specific expression.
"Our results show that [allele-specific methylation] is more frequent than can be accounted for by known imprinted loci and correlates very well with allele-specific expression for genes displaying [allele-specific methylation] in their promoter region," the researchers explained. "To further quantify this observation, additional methylomes will be required to increase the number of parental polymorphisms at imprinted genes."
Moreover, by comparing the mononuclear cell methylome with data from the first published human methylome study, which examined fetal lung fibroblast cells, the team identified hundreds of thousands of tissue-specific methylation differences.
Based on their results so far, the researchers argue that homing in on methylation at individual bases across the genome will lead to a better understanding of the methylation's role in the genome — particularly if such findings are tied to corresponding functional and clinical information.
"Our results demonstrate this methylome to be rich in biological information, compatible for integration with functional data," the team concluded, "and we [expect] it to form a lasting resource as part of the International Human Epigenome Project."