Building a Better Mouse Burrow Takes Few Genes

Liz is a staff writer for Science.

  • Hole in one
    Credit: E. Pennisi

    Hole in one. Before studying the genetics of burrowing, researchers first had to characterize underground nests of deer mice.

  • Hands-on
    Credit: E. Pennisi

    Hands-on. The first step was removing any mice from the nest.

  • Injection
    Credit: E. Pennisi

    Injection. Through a tube pushed down the burrow, the researchers spray foam insulation into the tunnels.

  • Digging deep
    Credit: E. Pennisi

    Digging deep. After the foam hardens, they excavate the burrow.

  • Field work
    Credit: E. Pennisi

    Field work. Researchers measure and mark the foam cast of the burrow.

  • Escape hatch
    Credit: E. Pennisi

    Escape hatch. The L-shaped "arm" of the cast is the entrance tunnel; the central bulge is the nest; and the other "arm" is the escape tunnel.

The oldfield mouse digs a burrow that would make the gang from Ocean's Eleven proud. The 200-centimeter-long, two-tunnel design, complete with an escape hatch, is an architectural marvel compared with the short, single crawlway of the deer mouse. The difference, according to a new study, possibly involves just a few genes, suggesting that it doesn't take much to evolve a new complex behavior.

Animal behavior is determined to a large extent by genes. The trick is to figure out exactly how those pieces of DNA control animals' lives. Harvard University evolutionary biologist Hopi Hoekstra, and Jesse Weber, now at the University of Texas, Austin, started with a relatively simple question: How do genes influence the types of burrows mice build? They first studied the shape, depth, and size of oldfield mice (Peromyscus polionotus) burrows in several southeastern U.S. states, making foam casts of the nests (see slideshow). They found that while the depth of the burrow depended on the type of soil, the length of the tunnels and the presence of an escape route was consistent in all places examined. Next, they showed that deer mice and oldfield mice build the same kinds of burrows in the lab as in the wild—two very different environments. That indicated tunneling was not dependent on environmental factors, but genetically controlled.

To understand these genetics, Hoekstra and Weber crossed oldfield mice with deer mice, which are so closely related that they can interbreed. They then mated the offspring back with deer mice to do a genetic analysis of those young. They assessed tunnel characteristics in both generations by allowing the mice to build burrows in big sandboxes in the lab. Hoekstra, Weber, and Harvard evolutionary biologist Brant Peterson also developed a genetic technique to link aspects of tunnel design to specific locations in the genome of the second generation.

All of the first-generation mice made two-tunnel nests, suggesting that the genetic contribution of the oldfield mouse parent took precedence over that of the deer mouse; that is, it was genetically "dominant," the team reports online today in Nature. In the second generation, the mice built a variety of burrows—some with just one long tunnel, some with two short tunnels, and so on, indicating that tunnel traits were independently controlled by separate genes. The researchers pinpointed three gene regions underlying tunnel length, located on chromosomes 1, 2, and 20, and one, located on chromosome 5, that determined whether escape tunnels were made. The regions could have relevant genes, or just one.

"The idea that [burrowing] would break down into three regions and one region is just mind-boggling," says Cori Bargmann, a neurobiologist at Rockefeller University in New York City who was not involved with the work. Breaking down a behavior into quantifiable components and studying how those components were inherited is an approach other researchers can try, she says. "It's huge to show it's a tractable, manageable problem."

Still, there may be other gene regions involved that were not detected by the study, notes Catherine Peichel, an evolutionary biologist at the Fred Hutchinson Cancer Research Center in Seattle, Washington. And each region could contain several relevant genes. Even so, she says, "the paper provides a nice empirical example of how a complex behavior evolves on a genetic level."

Researchers have yet to tease apart the genetics underlying human behavior and to discern what is learned versus inherited. But "the [work] points to a general mechanism for generating complex behaviors" from simple ones, says Edvard Ingjald Moser, a neuroscientist at the Norwegian University of Science and Technology in Trondheim who was not involved with the work. If that mechanism holds true for humans, then it may one day be possible to find the human genes that help determine how we live our lives.

Posted in Evolution, Biology