STUDYING MOTOR SEQUENCE LEARNING ACROSS MULTIPLE PHASES IN MICE

Complex motor skills require chaining elementary actions into rapid, precisely organized motor sequences with smooth transitions between successive components. Deficits in sequential motor control are common in neurological disorders such as Parkinson’s disease and stroke, even when the ability to execute simple, isolated movements is retained. However, the neural mechanisms that support the acquisition and refinement of motor sequences remain poorly understood. To address this gap, we developed a novel dexterous motor task in which mice learn to pull a lever along a trajectory composed of multiple defined turning points, thereby generating a structured motor sequence. Task acquisition proceeds through multiple training phases: mice are first shaped to perform straight pulls, after which turning elements are gradually introduced and temporal constraints on movement initiation and execution are progressively tightened. Across a four‑week training period, mice exhibited robust improvements in performance, with success rates approaching 100% and cue‑to‑reward latencies decreasing to under 1 second. Analysis of lever trajectories revealed increasing smoothness and trial‑to‑trial consistency over learning. Fine‑scale kinematics of the forelimb including elbow, paw, and digit positions were extracted using DeepLabCut, enabling high‑resolution characterization of evolving movement strategies. This behavioral paradigm can be integrated with neuronal recording approaches to dissect the neural dynamics within motor cortical circuits that support the acquisition and mastery of complex motor sequences.
This work is supported by HKRGC grants 14112423, 14113522, C4012-22G and CityUHK SGP grant 9380157.