Project Summary
Hearing and speech are fundamental to human communication. Before infants are 12 months old, they
have learned the statistical regularities of speech sounds around them, which guides vocal learning throughout
early life and permanently alters auditory coding in cortical neurons. Deficits in cortical speech coding are
hallmarks of central auditory disorders such as auditory processing disorder (APD), specific language
impairment (SLI), and autism spectrum disorder (ASD), which together affect an estimated 7% of children.
These disorders are highly heritable and involve additive effects of multiple gene mutations. But, how genes
constrain the landscape through which auditory coding unfolds over development is entirely unknown. The
proposed aims test specific hypotheses regarding genetic constraints on auditory coding in songbirds, an
animal model of speech processing. Like humans and unlike other animals, songbirds learn to sing by learning
regularities in the vocal sounds of adults they interact with socially, and auditory cortical coding is permanently
altered by experience. The songbird secondary auditory cortex, caudomedial nidopallium (NCM) is required for
vocal learning, and contains neurons that are selectively tuned for acoustics of songs heard in early life.
Preliminary behavioral data indicate that juveniles raised by birds of a different species learn the morphologic
structure of their foster father’s song, but arrange their song with temporal structure typical of their genetic
relatives, despite never having heard those songs. In contrast, hybrid birds born of mixed-species parents copy
both the morphology and temporal structure of their father’s song, despite having only half the genetic makeup.
These data lead me to hypothesize that separate subpopulations of NCM neurons selectively code for
song morphology and temporal structure (Aim 1), and that temporal coding is constrained by genetics
(Aim 2). I propose to analyze singing behavior in songbirds with differing genetic backgrounds and specific
differences in singing behavior. I will compare electrophysiological responses of their NCM neurons to natural
songs, songs with altered temporal structure, and synthetic sounds that systematically vary in acoustics and
timing. I will use neural encoding analyses to measure which stimulus features drive responses, and decoding
analyses to measure what information neurons carry about stimuli. Expected outcomes will provide
fundamental insights into the genetic sources of limitations on cortical auditory coding capacity, and will inform
mechanism-based approaches to treating central auditory disorders such as APD, SLI, and ASD. The training
plan includes training in electrophysiology in awake animals, advanced neural data analyses, developing
expertise in behavioral genetics, and writing for journal publications, grants, and conference presentations.
Training will take place in the sponsor’s and co-sponsor’s labs, housed in Columbia University’s Zuckerman
Institute. The Institute is home to world-renowned neuroscientists and state-of-the-art facilities.