The back-and-forth songs of Alston’s singing mouse (Scotinomys teguina) could share mechanisms with human conversation.

Christopher Auger-Dominguez

This singing mouse’s brain could reveal keys to snappy conversation

If there were a rodent opera, Alston’s singing mouse would be the star. This unassuming brown mouse native to Central American cloud forests rears up on its hind legs and belts out long, intricate trills. Each animal’s performance appears to be unique, says Michael Long, a neuroscientist at the New York University School of Medicine in New York City. “I can recognize this one particular song and say, ‘Ah, that’s Ralph.’”

Long and his colleagues have now found that the species (Scotinomys teguina) does something that many lab animals don’t: It takes turns singing. And these rapid-fire duets are giving researchers a new model to study how the brain controls conversation. The mice might even inform our understanding of what goes wrong in disorders that affect communication, such as autism.

A human conversation has a lot of moving parts. While we listen, we plan our words, adjust them on the fly, and direct our vocal muscles to spit them out at socially appropriate moments. Most lab animals lack this verbal complexity, Long says. The most common lab mouse (Mus musculus) has a relatively disorganized and unpredictable song—and it doesn’t tend to take turns crooning. Even marmosets, primates with back-and-forth calls that make them popular in neuroscience research, pause for several seconds between responses, which means they may not rely on the same neural machinery that drives the split-second responses in a human repartee, Long says. “Imagine a conversation between two people where there’s a 5-second pregnant pause between every exchange. I think I would be going crazy.”

That’s why he and other researchers are interested in Alston’s singing mouse. The social functions of its distinctive songs aren’t totally clear yet. (The performance seems to be part of a territorial display, though it may not always be adversarial.) The species, a relative newcomer to neuroscience, is a bit of a diva in the lab, requiring a spacious terrarium, a specialized diet, and exercise equipment to thrive. But Long and his team have discovered an intriguing tendency for turn-taking in the mice. When a male mouse enters a chamber adjacent to another male and can hear its neighbor sing, it precisely times its own songs to avoid overlap with the neighbor; it starts about half a second after the other mouse finishes. (In human conversation, the lag is even shorter—roughly 200 milliseconds, on average.) The mouse with a neighbor also sings four times as much as when it is alone, the researchers found.

The researchers wanted to know how the mouse’s brain controlled these carefully timed exchanges. Many studies have shown that animal vocalizations originate in deep, evolutionarily ancient parts of the brain—so-called subcortical structures. But they wondered whether, in this singing mouse, a separate structure in a higher region called the motor cortex acted like an orchestra conductor, turning the songs on and off based on social cues.

They found a variety of evidence supporting this hunch. They first identified a part of the brain called the orofacial motor cortex (OMC) that, when stimulated, caused the mouse to flex its vocal muscles. When they put a cooling device over that region to slow down its neural activity, the mouse took longer to reach the end of its song.

What’s more, the mice could still sing when the scientists gave them a drug that completely inactivated the OMC—the vocalizations were evidently produced elsewhere in the brain—but hearing another mouse’s song no longer increased their own singing. And the mice no longer launched a prompt “counter song” in response, the researchers report online today in Science. They conclude that to carry out vocal turn-taking, the mouse brain divides labor between a basic song generator (still to be identified in the subcortical brain) and a higher-level conductor.

“I think it’s beautiful, the combination of kind of methods they were applying,” says Julia Fischer, an ethologist who studies animal social behavior and cognition at the University of Göttingen in Germany. “This a breakthrough, in terms of the fine details … they were able to work out.”

Alston’s singing mouse is “a new, potentially interesting model for vocal communication,” says Karel Svoboda, a neuroscientist at the Howard Hughes Medical Institute's Janelia Research Campus in Ashburn, Virginia. Still, he says, “We need to know a lot more.” This study doesn’t sort out how the OMC influences activity in the lower brain, he notes, or how circuits there actually make the muscles move to produce song.

Human speech is vastly more complex than these mouse duets, but Long’s team now wants to look to the human brain for a comparable timing mechanism in the human motor cortex, known to be involved in controlling speech. He and his colleagues are designing experiments that record brain activity during conversational tasks, such as responding quickly to another voice.

Long also sees this mouse as a potential way to study autism, a disorder that can limit a person’s ability to communicate. Researchers could manipulate genes in the mouse OMC that are implicated in autism to observe how they affect brain activity and behavior. The turn-taking divas, Long says, have really “broken open a brand-new slice of biology for us.”