Gene therapy

Knocking certain noncoding DNA can lead to too many digits (red arrow) in mice.

M. Osterwalder/Lawrence Berkeley National Lab

Biologists identify the backup systems that ensure genes build limbs

COLD SPRING HARBOR, NEW YORK—Some aspects of our genomes would make a NASA engineer proud. Whereas spacecraft have many redundant systems that can kick in, such as when a thruster fails, cells have their own backup systems for regulating gene activity. But they are extraordinarily hard to understand, as was made clear last week when molecular biologists for the first time described the regulatory backups for two genes involved in mammalian limb formation. Understanding these redundancies, and how to bypass them, could be important for biomedical researchers wishing to manipulate gene activity to treat human diseases.

Genes may carry information for building proteins, but a host of other factors, including the DNA between genes that doesn’t encode proteins, tells them when to make their proteins. And this so-called noncoding DNA plays a role in some diseases, studies have shown. Noncoding DNA includes stretches called enhancers, short sequences that help control a target gene.  Geneticists have identified hundreds of thousands of potential enhancers in our genome, but verifying their role in DNA regulation and disease is a daunting proposition. “We still don’t have a good understanding of what [many do],” says Yang Li, a computational biologist at Stanford University in Palo Alto, California.

For the new study, presented last week at the Biology of Genomes meeting here, molecular biologist Marco Osterwalder of Lawrence Berkeley National Laboratory in Berkeley, California, and colleagues harnessed a powerful new gene-editing technique called CRISPR to figure out exactly how some of these candidate enhancers work. The method speeds up the development of "knockout" mice that lack a particular enhancer, helping reveal its function.

The lab already had preliminary evidence for more than 1200 enhancers. And in the new tests, Osterwalder used CRISPR to knock out 10 of the enhancers in different mouse embryos. These enhancers were located close to genes involved in limb development, so he and his colleagues expected the embryo's limbs to be defective in some way. But to their dismay, there were no such abnormalities, he reported last week. 

Wondering whether other enhancers pick up the slack—providing a backup system to keep limb development on track—Osterwalder and his colleagues bred mice lacking a pair of enhancers specifically implicated in digit formation.  Those mouse embryos developed extra digits, just as an embryo would have if the enhancers’ target gene, Gli3, itself had been defective, he reported at the meeting.

To further explore this redundancy, the team then knocked out one or both of those enhancers in mice in which they had also knocked out one of the pair of Gli3 genes that control digit number. Typically, with one copy of that gene out of commission, embryos make only half the normal amount of Gli3 protein—and an extra thumb forms. When scientists also knocked out just one of the enhancers in the same mice, the embryos again grew an extra thumb. But when scientists knocked out both enhancers in the mice with one missing Gli3, the mice grew several extra digits. Those results are the same if the enhancers are intact but both copies of the gene itself are defective, indicating that the amount of protein determines the number of digits. The enhancers “are redundant on an organismal level, but additive at a molecular level,” Osterwalder concluded. 

“It is really a wonderful experiment” says William Greenleaf, a biophysicist at Stanford who was not involved with the work. “It represents the logical next step in the mapping of regulatory elements.”

Three years ago, another team studying enhancers in the development of the mouse face and skull uncovered similar complexity. And studies in Drosophila suggest, too, that interactions of multiple enhancers are commonplace. 

Already, researchers have linked noncoding DNA to Crohn disease, heart disease, multiple sclerosis, and other disorders, suggesting that enhancers could be targets for medical manipulation. And “we are finding more associations all the time,” says Joseph Pickrell, an evolutionary geneticist at the New York Genome Center in New York City. But given the complexity revealed by the CRISPR experiments, understanding how enhancers interact with diseases “is going to be a marathon,” he predicts.