NEW YORK (GenomeWeb News) – New research is providing a roadmap of the epigenetic DNA modifications that help orchestrate the gene expression of a commonly used plant model organism.
In a paper published in today’s issue of Cell, researchers from the Salk Institute for Biological Sciences and their collaborators used high-throughput sequencing to chart epigenetic modifications — specifically, cytosine methylation — throughout the Arabidopsis thaliana nuclear genome. By overlaying this information with messenger RNA and small RNA sequence data, the team was able to delve into the relationships between cytosine methylation, gene expression, and small RNAs.
“There were many small surprises which we found looking through the data,” co-lead author Ryan Lister, a postdoctoral researcher affiliated with the Salk Institute for Biological Studies’ Genomic Analysis Laboratory, told GenomeWeb Daily News today. “The overall message is that it’s now feasible to integrate all these levels of nucleotide information from the cell.”
Cytosine methylation is an important epigenetic DNA modification that alters the way the cell’s transcription machinery binds to specific regions of the genome. Such modifications have been implicated in everything from embryonic development to tumor formation. Enzymes called methyltransferases add methyl to cytosine, while others called demethylases remove them.
For this study, Lister and his colleagues used sequencing-by-synthesis with an Illumina Genome Analyzer to map the cytosine methylome, transcriptome, and small RNA transcriptome of A. thaliana, a model plant organism with a relatively small and well-defined genome.
To do this, they isolated genomic DNA from A. thaliana floral tissue and treated it with sodium bisulfite, which converts only unmethylated cytosines to uracil and leaves methylated cytosines unchanged. They then analyzed the DNA by high-throughput sequencing.
By incorporating this methylation data with gene expression and smRNA data gleaned from other sequencing experiments, the researchers could look at how the systems interacted. To assess this, Lister said, they relied heavily on a web-based genome browser called AnnoJ, developed by collaborators at the University of Western Australia’s ARC Center of Excellence in Plant Energy Biology. The program “allows you to dynamically zoom from megabase data down to an individual nucleotide,” Lister explained.
DNA methylation was more extensive than previously imagined, the authors reported. They found that more than five percent of the cytosines in the nuclear genome of A. thaliana floral bud tissue — some 2,267,447 bases — were methylated. They verified these findings using a ChIP-chip approach, hybridizing a chromatin immunoprecipitation with methylcytosine antibodies to whole-genome tiling arrays.
Though the regions identified by sequencing corresponded to those that fell out of their ChIP-chip analysis, sequencing pulled out nearly twice as many methylated cytosines as the ChIP-chip approach.
“[T]he demonstrated higher sensitivity, increased coverage, and reduced bias of the methylC-sequence approach allows for the discovery of a previously uncharted segment of the Arabidopsis DNA methylome,” the authors wrote.
Interestingly, about a third of the methylated cytosine residues in the genome seem to be associated with smRNAs — a pattern that may provide insight into how the molecules work together to control gene regulation. Specifically, the data suggest smRNAs help target DNA methylation.
“This may be a feedback loop established to reinforce silencing,” Lister explained.
The team also looked at the influence that local sequence had on cytosine methylation and compared epigenetic patterns in mutant A. thaliana strains lacking various methyltransferase or demethylase enzymes. Because these plants have methylation defects — resulting in less or more methylation than usual — they provide insights into the role of methylation within these cells.
For instance, in mutants lacking the methyltransferase met1, the researchers found 11,652 regions in which both DNA methylation and smRNA abundance differed from wild type. In more than 90 percent of these, mutants with lower levels of DNA methylation also had decreased smRNA density.
Altering methylation also changed the expression of hundreds of genes, transposons, and intergenic transcripts.
“Through the simultaneous study of these three interrelated phenomena in wild-type plants and in informative mutant backgrounds, we have helped to illuminate, genome-wide, the scope and sophistication of the interactions that exist between methylation and smRNA, and their ultimate effect on transcriptional regulation,” the authors concluded.