AN AUDITORY VIRTUAL ENVIRONMENT TO STUDY NEURONAL PLASTICITY IN THE HIPPOCAMPUS

Many forms of learning are driven by mismatches between predicted and actual sensory input. Investigating such learning processes therefore requires experimental paradigms that directly couple behavior with sensory feedback in a controlled manner, while monitoring neuronal activity and neuronal plasticity.
In this project, we developed a low-latency, auditory virtual environment based on capacitive coupling of mouse front paws with a sensor antenna. The auditory feedback is generated with low latency using an Arduino-based system. In this behavioral paradigm, animals learn to continuously generate soundscapes through their own movements, which provides precise control over sensorimotor contingencies and prediction-error signals.
We combined this closed-loop auditory virtual environment with longitudinal two-photon calcium imaging to investigate how mismatches between expected and actual auditory feedback shape neuronal representations in hippocampal CA1 of head-fixed mice.
As learning-related plasticity may occur in both somatic and dendritic compartments, we developed an optical approach that facilitates simultaneous calcium imaging from both somata and dendrites. To align the imaging plane with the pyramidal cell layer or individual dendrites, we adjust the orientation of the scan plane using a combination of scan-field rotation using a K-mirror and axial focus modulation using an electrically tunable lens. This strategy allows us to precisely align the imaging plane with hippocampal anatomy or neuronal morphology.
Together, these methods will enable direct investigation of how closed-loop learning and prediction errors shape non-spatial hippocampal representations.