SPLIT-WHEEL TREADMILL: A BEHAVIORAL TOOL TO STUDY LOCOMOTION ADAPTATION IN HEAD-FIXED MICE

Walking is a fundamental motor act. The process of maintaining a stable and efficient gait when external perturbations are introduced is called locomotion adaptation. This phenomenon has been demonstrated in both humans and animals using a split-belt treadmill that imposes unequal speeds on the two sides of the body independently (Reisman et al., 2007; Darmohray et al., 2019). Locomotion adaptation depends on the cerebellum and is characterized by the gradual reduction of sensorimotor errors, the difference between predicted and actual movements. Previous studies explored the behavioural but not the neuronal component of this motor learning task. Thus, the cerebellar mechanisms underlying precise locomotion adaptation in space and time are still poorly understood. Here, we present a versatile motorized treadmill that enables measurement and imaging of behaviour and neural activity during locomotion adaptation at high temporal resolution in head-fixed mice. The treadmill consists of two transparent wheels independently controlled by the servo motors, a mirror positioned at 45-degree angle between the wheels to capture the bottom view of the animal’s limbs, and a compact head-fixation system allowing stable imaging of neuronal activity. In addition, we developed a Python-based analysis tool for extracting and characterizing the treadmill-derived data as well as gait-related behavioral components. In preliminary experiments, we performed two-photon calcium imaging of Purkinje cell (PC) dendrites in head-fixed L7-Cre mice expressing Cre-dependent GCaMP6f in cerebellar lobules IV/V. We are currently exploring the relationship between PC dendritic spiking activity and inter-/intralimb parameters during locomotion adaptation.