Supervised mechanisms for locomotor learning in the cerebellum

Motor learning is essential to move in a continuously changing environment, but the underlying neural circuit mechanisms are still poorly understood.

This is critical for locomotion, a fundamental but complex behavior, which requires precise control of whole-body movements at the same time.

The cerebellum plays a key role in motor learning, supposedly generating corrective movements through supervised error-based mechanisms in response to perturbation.

For simple tasks, climbing fibers originating in the Inferior olive drive supervised learning, generating changes of cerebellar activity and corrective responses.

During locomotion, perturbations cause gait asymmetries, which can be externally induced in a split-belt treadmill with belts running at different speeds.

The host laboratory developed a split-belt treadmill to study locomotor adaptation in mice, showing that it is similar to humans, occurs through motor corrections with complex spatiotemporal dynamics, and is driven by the cerebellum. However, the underlying neural bases are unknown.

SuperLoco aims to identify supervised mechanisms for locomotor learning, exploiting the synergy between cutting-edge technologies to interact with neural circuits and computational neuroscience tools.

Specifically, we will: (i) record climbing fiber and cerebellar activity during locomotor learning in mice with high-yield electrophysiology, (ii) simulate locomotor learning in a bioinspired spiking neural network of the cerebellum with climbing fiber-supervised plasticity, (iii) optogenetically stimulate climbing fibers to induce locomotor learning based on model predictions.

Our main hypothesis is that the timing of climbing fiber signals determines changes of cerebellar activity through supervised plasticity driving locomotor learning.

SuperLoco outcomes will shed light on neural mechanisms for complex whole-body movements, fundamental in neuroscience and crucial for treatment of movement disorders.