Project Summary: Whether to laugh at a joke or to engage in a lively debate, we flexibly modify our vocalizations based upon social contexts. Such adaptive behavior requires real-time adjustments of motor outputs in response to rapidly changing sensory inputs. How does the brain accomplish this sensorimotor feat? Pioneering studies have characterized the brain areas responsible for sound production in many species (e.g., drosophila, zebra finches, marmosets, mice), but the neural circuits that generate vocal flexibility remain poorly understood. Vocal flexibility, such as during a conversation, requires voluntary, context-dependent control over sound production. In mammals, based on human brain lesions, gene expression profiles, and neurophysiology data in primates, cortical control has been proposed to exert volitional control over sound production. However, direct evidence for this idea is scarce and the neural circuit-level mechanisms underlying vocal flexibility, especially in mammals, remain largely unknown. Finding an appropriate rodent model would complement prior work in the primates and would permit circuit-level mechanisms to be deciphered. Alston’s singing mice (S. teguina), a highly vocal rodent from the cloud forests of Central America, are ideally suited to study flexible vocal behaviors. Singing mice show remarkable vocal flexibility, switching between variable, ultrasonic vocalizations (USVs) and stereotyped, human-audible songs depending upon social context. In contrast, most rodents including lab mice (M. musculus) produce only USVs and are not known to participate in vocal interactions. Singing mice and lab mice are roughly the same body size, and brain slices of S. teguina at a first glimpse is indistinguishable from those of M. musculus. Neural circuit differences underlying such drastic behavioral divergence are unknown. Here we propose to test whether the ability of the singing mice to apply vocalizations flexibly within a social context, and lack thereof in most other rodent species, is dependent upon motor cortical function, acting via downstream vocal production circuits. Using chronic electrophysiology (Aim 1), single-cell comparative connectomics (Aim 2), we will determine the role of motor cortex during natural vocal behaviors and compare cortical connectivity and function between two species. In Aim 3, we will manipulate the circuit to determine their causal role in various vocal behaviors in each species. By mapping, measuring and manipulating cortical circuits, we will learn how motor cortex modulates behavioral flexibility in service of social communication. More broadly, these experiments will provide a systems-level framework to study hierarchical motor control circuits – for e.g., how high-level (cortical) control can inform low-level controllers (subcortical pattern-generators) to generate appropriate motor commands – a challenge faced by biological and artificial agents moving through the world.

Project Narrative Loss of control over interactive vocalizations is associated with many different serious health conditions in humans such as stroke-induced aphasia, movement disorders such as Parkinson’s, and neurodevelopmental disorders such as autism. Using chronic electrophysiology, optogenetic manipulations, and comparative connectomics with single-cell resolution, we will compare the role of motor cortex during vocal behaviors in animals that can interact vocally (i.e., singing mice) versus those that do not (i.e., lab mice). This contribution is significant because the insight we stand to gain on how motor cortex controls ethologically-relevant vocal behaviors in mammals is likely to advance the field toward a better understanding of what happens when these systems go awry in humans.