NEW YORK (GenomeWeb News) – Scientists yesterday unveiled a new, genome-wide approach for accurately mapping open chromatin — regions likely involved in gene regulation.
The researchers, who are members of the ENCODE Consortium and based at Duke and Boston Universities, the National Cancer Institute, and the National Human Genome Research Institute, used a novel combination of targeted DNA digestion, microarray analysis, and sequencing to identify 94,925 regions that are likely involved in some form of gene regulation. Their findings were published online yesterday in the journal Cell.
“As far as we’re aware, this is the first published set of a genome-wide [gene regulatory] annotation of this kind,” co-senior investigator Terrence Furey, a biostatistician at Duke University, told GenomeWeb Daily News today. “One of the things we like about this is that we think it provides a lot of important information for the genomics community as a whole.”
Most nuclear DNA is packaged and wrapped around protein complexes. But gene regulatory regions tend to be on open chromatin that is not assembled on these complexes, and these stretches of DNA are more easily chopped up by the enzyme DNase I.
For nearly 30 years, researchers have exploited this DNase I hypersensitivity to identify regulatory regions — from promoters and enhancers to gene silencers and insulators — for particular genes of interest. Furey and his colleagues took this approach several steps further, scaling up the process to look for DNase I sensitive, regulatory regions across the entire genome.
“Our goal was to identify the areas of DNA across the entire genome that are not packaged, because we know those are the regions that are important for regulating gene activity,” Gregory Crawford, the other senior author, also at Duke, explained in a statement. “We identified all unpackaged regions within the entire genome using two extremely efficient methodologies: microarrays and sequencing.”
After slicing and dicing the DNA from a healthy, primary blood cell line with DNase I, the team used size selection to pick up the smaller DNA fragments that had originated in hypersensitive, regulatory zones. Then they hybridized the fragments to whole-genome NimbleGen arrays or sequenced them using next-generation sequencing platforms from Illumina and 454 Life Sciences, which is owned by Roche.
Finally, they normalized the information they gleaned from the arrays and sequencing and mapped these regions to create what they call “a comprehensive and accurate genome-wide open chromatin map” based on the 94,925 DNase I hyper-sensitive sites they identified. They also assessed which potential regulatory regions fell in transcription sites, between transcription sites, or even within genes.
“It’s difficult to do that exactly, because gene annotation isn’t perfect,” Furey said. Even so, he said he was a bit surprised at how many apparent regulatory regions fell within putative genes. Only 16 percent to 21 percent of the sites were within promoters or first exons of genes.
Although these experiments were done in healthy cell lines, Furey and his colleagues hope to apply these studies to disease conditions as well. As such, their next step is using similar techniques to map DNase I sensitive regions for 15 to 20 additional cell lines — both healthy and cancerous — looking for patterns that are linked to disease states.
Furey said the genome-wide approach may also be used to look at gene regulation in an evolutionary way, for example by comparing the patterns in humans and other primates.