NEW YORK (GenomeWeb News) – In a paper appearing online today in Nature, American and Australian researchers reported that they have mapped cytosine methylation across the genomes of two human cell lines.
"This paper documents the first complete mapping of the methylome, a subset of the entire epigenome, of two types of human cells — an embryonic stem cell and a human fibroblast line," Linda Birnbaum, director of the National Institute of Environmental Health Sciences, who was not involved in the study, said in a statement. "This will help us better understand how a diseased cell differs from a normal cell, which will enhance our understanding of the pathways of various diseases."
The work was done as part of a National Institutes of Health Roadmap Project on epigenetics.
The team mapped methylated cytosines to single base resolution across the genomes of an embryonic stem cell line and a differentiated lung fibroblast line, also incorporating information on messenger and small RNA transcripts and chromatin marks in the cells. When they compared the methylomes, the researchers found a slew of differences, including cell specific cytosine methylation patterns.
The study also opens the door for more extensive characterization of human epigenomes, senior author Joseph Ecker, a plant biology researcher at the Salk Institute for Biological Studies and director of the Salk Institute Genomic Analysis Lab, told GenomeWeb Daily News. "The real goal is to compare differentiating cells," he said.
Last spring, he and his co-workers reported that they had used a combination of sodium bisulfite sequencing and other high-throughput sequencing methods to characterize the methylome, transcriptome and small RNA transcriptome of the model plant Arabidopsis.
For the current study, Ecker and his team applied a similar approach to tackle the human epigenome, using bisulfite sequencing with the Illumina Genome Analyzer II to map cytosine methylation patterns across the genome in two human cell lines — the differentiated fetal lung fibroblast line IMR-9 and the human embryonic stem cell line H1.
Nearly all of the methylation in the differentiated fibroblast cells, was "CG methylation," occurring at sites at which cytosine is followed by guanine.
In contrast, roughly a quarter of the cytosine methylation in the stem cell was not at sites where cytosine was followed directly by guanine. This "non-CG methylation" was known to exist in stem cells, Ecker explained, but its prevalence was hard to judge in past studies that looked at only small portions of the genome.
"Non-CG methylation is not completely unheard of — people have seen it in dribs and drabs, even in stem cells. But nobody expected that it would be so extensive," co-lead author Mattia Pelizzola, a post-doctoral researcher in Ecker's lab, said in a statement. "[N]on-CG methylation was often considered a technical artifact."
By looking at the number of methyl-cytosine reads at each site, the researchers were also able to map methylation levels across the genome. For instance, the team found that differentiated fibroblast cells contained partially methylated areas that frequently corresponded with decreased gene expression, Ecker noted.
They also reported that non-CG methylation in stem cells was often depleted at transcription start sites as well as enhancer and transcription factor binding sites. In contrast, the researchers did not see this periodicity in the differentiated cells, Ecker said.
When they targeted some of the non-CG methylation loci with bisulfite sequencing in another human embryonic stem cell line called H9, the researchers found similar non-CG methylation patterns at conserved positions.
"The exclusivity of non-CG methylation in stem cells, probably maintained by continual de novo methyltransferase activity and not observed in differentiated cells, suggests that it may have a role in the origin and maintenance of this pluripotent state," the team concluded. "Essential future studies will need to explore the prevalence of non-CG methylation in diverse cell types, including variation throughout differentiation and its potential re-establishment in induced pluripotent states."
Consistent with the potential link between non-CG methylation and pluripotency, the same sites were non-CG methylated in the team's pilot experiments looking at an induced pluripotent stem cell line created by reprogramming fibroblast cells.
In the future, Ecker said, the team hopes to track changes in the epigenome, including genome-wide chromatin marks, methylation, and more, as they coax embryonic stem cells into a variety of differentiated cell types.
The single-base resolution human methylome data is available online through the Human DNA Methylome web site.