The ability to learn a language is a core human trait. Yet, the study of its neurobiological underpinnings has faced a major roadblock for hundreds of years - a tractable mammalian model system to study this function has never been established. The reason is that vocal learning abilities are remarkably sparse and out of over 5400 existing mammalian species only a select few possess vocal learning abilities, these are cetaceans, elephants and bats, with initial evidence of vocal development only recently emerging in non-human primates. Thus, all detailed neurobiological investigation of vocal learning focused on the avian brain. While this body of work produced invaluable insight, evolutionary divergence between avian and mammalian brains makes translating findings from birds to humans challenging. For example, the avian equivalent of cortex has a nuclear organization, which is very different from the six-layered cortex of mammals. A mammalian model system is thus imperatively needed to bridge this major gap. Therefore the goal of this proposal is to directly overcome this challenge by establishing bats as the first mammalian model system for studying the neural basis of vocal learning. Our investigations target key cortical areas that are involved in auditory perception and vocal production - two key requirements of vocal learning. These are the auditory and frontal cortices. Bats are further an attractive model system because many behavioral aspects of their vocal learning share many resemblances with those of human language learning. To facilitate our studies we advance three major aims that would enable detailed neurobiological investigation of vocal learning. First, we design a high-throughput, automated vocal-learning paradigm for bats. This system allows unprecedented control over bat vocal learning and targets core features of human vocal learning, such as vocal-motor plasticity at both the laryngeal source and vocal tract. We further use this paradigm in combination with immunohistochemistry to identify candidate brain areas that may participate in vocal learning. Second, we use our development of wireless electrophysiological methods to monitor at millisecond resolution the coding properties of mammalian cortical neurons during vocal learning. This approach will provide both the first description of mammalian cortical computations underlying vocal learning as well as enable a direct comparison to studies in birds and identify generalizable neural mechanisms between birds and mammals. Third, we use cellular-resolution calcium imaging to longitudinally monitor cortical networks throughout the vocal learning process. These experiments will facilitate the first description of the emergence of cortical responses and anatomical topographies during vocal learning. Ultimately, our research program will provide a detailed description of the mammalian neural computations that support our ability to learn language, but it should help us also understand the causes of disorders directly related to speech and language development and aid in the design of more effective therapeutic approaches that will be applicable to humans.
The ability to learn a language is a core human trait yet its detailed neurobiological underpinnings have never been studied in the mammalian cortex due to a lack of a tractable mammalian model system. Disruption of proper development of language and speech affects the welfare of both the individual and society and is associated with a variety of neurological disorders. Thus understanding the basic principles by which mammalian cortical neurons operate to facilitate language learning will provide critical insight to understanding this core human function and identify causes of neurological diseases.
