Neural Circuits for Evaluating Complex Motor Sequences in Female Songbirds

PROJECT SUMMARY
In social interactions, we are constantly evaluating others based on their behaviors, such as posture or tone.
This is a highly complex task requiring the brain to integrate prior knowledge with incoming sensory information.
Here, I propose to examine the neural computations underlying social evaluation in zebra finches, a songbird
species in which females choose males based on the quality of their courtship songs. Females prefer stereotyped
over variable songs, where evaluating stereotypy requires estimating variability across many song renditions.
While decades of research have revealed the neural circuitry for song production in males (the ‘song system’),
the female brain has been largely overlooked. Remarkably, although females do not sing, they possess an intact
song system, and behavioral studies suggest this circuit plays an important role in song preference. This raises
the exciting possibility that while the male song system evolved for song production, the female song system
evolved for song evaluation. I hypothesize that the female song system is responsible for the evaluation of male
courtship song; specifically, the sensorimotor nucleus HVC (proper name) encodes an internal representation of
song, while the higher auditory region caudomedial mesopallium (CMM) computes stereotypy across renditions.
This project will be the first to manipulate and record neural activity in female zebra finches during
courtship to investigate the neural basis of social evaluation. My novel approach will integrate large-
scale neural recordings with advanced computational methods to reveal how the female songbird brain
evaluates complex motor sequences. In Aim 1, I will investigate how the female brain encodes an internal
representation of song, by examining time-locked firing patterns in HVC. In Aim 2, I will investigate how the
female brain computes the stereotypy of song across renditions, by examining prediction-error responses in
CMM. Furthermore, I will determine how the anatomical architecture of the system supports song evaluation,
and I will model how neural networks integrate top-down inputs (internal representations) with bottom-up inputs
(sensory information) for predictive coding. This project will establish the female songbird as a novel model for
the neural basis of social evaluation, and provide an unprecedented framework for how the brain integrates
internal representations with external sensory information for social interactions. As avian and mammalian
circuits are evolutionarily conserved, the results can offer critical insight into many human disorders
characterized by social impairments. My research plan will be conducted at Columbia University’s Zuckerman
Institute, which hosts a world-renowned research community with expertise in experimental and theoretical
neuroscience, and a highly collaborative environment for cutting-edge multidisciplinary projects. With the
guidance of my sponsor, Prof. Vikram Gadagkar, and co-sponsor, Prof. Larry Abbott, this plan will provide me
with the best training to achieve my goal of conducting transformative disease-targeted neuroscience research
as an independent principal investigator.

Grant Number: 1F31NS143402-01
NIH Institute/Center: NIH
Principal Investigator: Hannah Chen