© BRYAN SATALINOMost DNA does not code for proteins. And figuring out how, when, and where this genomic “dark matter” plays a role in gene regulation is a huge undertaking. Now, scientists have developed a tool that could help. In a paper published today (April 3) in Nature Biotechnology, a team from Duke University in Durham, North Carolina, describes a high-throughput screening technique that uses CRISPR-Cas9 epigenome editing to identify regulatory elements in the genomes of human cells.
“It turns out that most of the genetic variation that’s responsible for more common complex diseases—like cardiovascular disease, diabetes, and neurological disorders—actually happens in this region in between genes,” said coauthor Charles Gersbach of Duke. “The exciting thing is having methods available to annotate the function of the noncoding genome,” he added.
“The noncoding genome is vast, and it can be challenging to identify which regions are important for modulating protein-coding genes,” Neville Sanjana of New York University, who was not involved in the study, wrote in an email to The Scientist. This study “harnesses CRISPR pooled screening technology to help us figure out where the functional regions in the noncoding genome [are],” he explained.
Gersbach and colleagues created lentiviral libraries of guide RNAs to target likely regulatory elements across several megabases of DNA surrounding two loci of interest: β-globin and HER2. They then generated cell lines with an integrated fluorescent protein to report on target gene activation.
The authors transduced their cell lines with one of two versions of the Cas9 protein with deactivated nuclease activity, dubbed dCas9. The repressor form of dCas9 recruits proteins that methylate lysine 9 on histone H3, leading to heterochromatin formation and gene repression in target sequences. The activator form of dCas9 binds to targeted DNA enhancers or promoters and facilitates the acetylation of lysine 27 on histone H3, which results in gene activation.
Next, the researchers transduced their cell lines with the libraries of guide RNAs at low levels to ensure that a single guide RNA would be present in each cell. They then sorted the cells based on fluorescence, and sequenced the guide RNA present in cells with especially high and low target gene expression.
Sequence identification confirmed known regulatory elements and revealed new roles for other DNA sequences. Many—though not all—of these sequences appeared in both activator and repressor screens. And some sequences seemed to play a regulatory role for a gene in one cell type, but not for the same gene in another cell type. Although the observed changes in gene expression were subtle, the authors validated the regulatory role of individual DNA sequences.
“We know that there are several aspects of the epigenome that correlate very strongly with changes in gene regulation, and it’s always been a struggle to turn those observations into a real understanding of how genes are actually regulated,” coauthor Tim Reddy of Duke told The Scientist.
“We don’t yet know if the particular epigenetic modification that we observe is the causal one,” Reddy said. “But we do know now from this, and from other work from our labs and others, that modifying the epigenome does indeed have a role in regulating some of these target genes.”
Richard Sherwood of the Brigham and Women’s Hospital in Boston, who did not participate in the study, said that this screening technique could help researchers interpret the large amounts of genomic data that has been generated as part of projects like ENCODE. “It’s exciting: the more tools we have in the toolbox, the more different questions we can approach,” Sherwood said.
Reddy, Gersbach, and their colleagues are working to scale up the screening technique, in order to examine candidate regulatory elements throughout the entire genome across different cell and tissue types. They also plan to leverage the technique in efforts to better understand certain diseases.
“There are so many places where we know gene regulation contributes to disease,” said Reddy. “Now we can start to get really clear new insights into the mechanisms that are causing those diseases, and those mechanisms themselves might be a pathway for therapy. Those mechanisms might help guide better diagnostics, they might help differentiate patients who do or do not respond to therapy, [and] that can inform better patient therapeutics.”
T.S. Klann et al., “CRISPR-Cas9 epigenome editing enables high-throughput screening for functional regulatory elements in the human genome,” Nature Biotechnol., doi:10.1038/nbt.3853, 2017.