The neural code has evolved to produce behavior in complex and unpredictable environments. The space of all possible neural codes is ridiculously large, but the search can be narrowed by using behavior to provide insight into what stimuli are being encoded, at what timescales and with what precision. Neuroethology, the study of the neural basis of natural behavior, provides a way to further narrow the search space. Studying natural behavior allows us to take advantage of the fact that each species has evolved to solve a specific set of social and environmental problems. I am studying mouse vocal and social behavior to establish a variety of behavioral paradigms in which to study information processing in the nervous system.
An ideal neurobiological system has good sensory input (easily manipulable stimuli and a large space of possible inputs) and good motor output (easily measurable, reliably elicited behavior). Vocal behavior is a beautiful example of this. Vocalizations are both highly salient to the animal and easily recorded, elicited, and modified in a laboratory setting. Vocal behavior is the product of a complex interaction between the neural circuits that underlie auditory perception and those that underlie vocal production. In addition, vocal signals are often elicited by, or modulated by, social context. Vocal signals are a privileged acoustic stimulus, one the nervous system is optimized to process.
Mice, members of the large and diverse order Rodentia, display many neurobiologically accessible behaviors (vocalizations, nest building, mating, dominance interactions, communal pup care, territorial marking and defense, and grooming). I use a variety of approaches to describe and quantify these behaviors.
Although wild mice display many interesting behaviors, the variety of behaviors of interest to neurobiological investigation have often not been maintained in inbred lines. Inbred mice often have sensory or motor deficits, as well as differences in less defined areas, such as aggressiveness or temperament. I have identified nine inbred strains that vary in "wildness" (escape behavior, aggression, biting), hearing, and exploratory behavior. I am characterizing their vocal and social behavior, looking for variations between strains. One important question is the degree to which the structure of mouse vocalizations depends on auditory feedback. I will be looking to see whether there are consistent differences between the vocalizations of mice with good hearing and those of mice with poor hearing.
Recording Vocalizations in Groups of Freely Behaving Mice
Not surprisingly, mouse social behavior and vocalizations become more complex in more complex social environments. However, identifying vocalizations from specific individuals then becomes difficult, even when only two mice are present. I am working on a microphone array and video system to track individual mice living in large (~20 individuals) social groups. The goal is to noninvasively generate continuous position information and high-quality vocal recordings of all members of the group.
The Function of Male Mouse "Song"
Male mice produce relatively stereotyped ultrasonic vocalizations during courtship. Female mice, given a choice between a muted male and a vocalizing male, spend more time with a vocalizing male. What features of male vocalizations are attractive to females? Are there vocal features that are inherently attractive, or is attraction based on familiarity (or lack thereof) or on recognition of dominance roles? Male mice are extremely aggressive and compete fiercely for territorial dominance. Do male mice change their vocalizations when they are competing for females?
Variation in Behavior Based on Social and Environmental Experience
There is widespread evidence that both short- and long-term changes in the gross architecture of the nervous system can be controlled by the structural or social complexity of the living environment. In collaboration with Christine Portfors (Washington State University), I am comparing the social behavior and vocalizations of mice born and raised in large, multigenerational cages with those of mice raised in standard cages.
Jennifer Linden (University College London) is currently pursuing behaviorally relevant auditory cortical recordings in mice. Her laboratory addresses questions specific to auditory cortical circuits, both during development and in adulthood. We will be collaborating on cortical responses to mouse vocalizations.
Effect of Potassium Channel Knockouts on Vocal Behavior
Len Kaczmarek (Yale University) is maintaining knockout mice with interesting behavioral phenotypes. These include Kv1.3–/–, Kv3.1–/–, Kv3.3–/–, and Kv3.1/Kv3.3 double-knockout animals. One knockout in particular, Kv3.1, appears to be uninterested in mating or in caring for pups. Kv3.1 is a potassium channel recently shown to be modulated in auditory brainstem nuclei in response to auditory stimulation. We are investigating whether these mice (either homozygous or heterozygous) produce vocalizations at all and, if they do, whether they differ from wild-type mice.
Production Mechanism for Ultrasonic Vocalizations
Experiments in the 1970s suggested that mouse ultrasonic vocalizations are produced by a whistle mechanism, rather than by the vibration of the vocal folds as in most mammalian vocal production. Mouse audible vocalizations, however, are probably produced more conventionally. This suggests that there are two separate motor control circuits—one for audible vocalizations and one for ultrasonic vocalizations. I am interested in pursuing the motor side of vocal behavior, starting with the biomechanics of the production mechanism itself.
As of April 15, 2009