OVERTRAINING OF DYT-TOR1A MICE TO MODEL TASK-SPECIFIC DYSTONIA

Dystonia is a movement disorder characterized by involuntary muscle contractions causing abnormal movements and postures. DYT-TOR1A is the most common genetic form of isolated dystonia, but only ~30% of carriers develop symptoms, suggesting the need for additional “second hits,” like environmental factors. Task-specific dystonia is a particular form that typically emerges in contexts where highly skilled actions are extensively repeated. However, its underlying mechanisms remain poorly defined, highlighting the need for a mouse model that recapitulates this phenotype and its experience dependence.
We aim to create a task-specific dystonia mouse model by combining a DYT1 dystonia mouse model (DYTKI mice) with continuous, high-intensity forelimb skill overtraining. To achieve this, DYT1KI mice and WT littermates are group-housed in a new, home-cage-integrated system for automated, reach-to-grasp training to obtain water from an RFID-gated chamber. We hypothesize that this paradigm will selectively induce a task-specific dystonia-like phenotype in DYT1KI mice.
We implemented a high-speed videography system coupled to a DeepLabCut-based pose-estimation workflow to enable quantitative and kinematic profiling of reaching behavior (like forelimb trajectories, joint angles, movement smoothness) to identify an emerging dystonia-like phenotype. A 29-hour pilot study validated the setup’s stability for high-throughput, unsupervised training, with three mice collectively executing 4,776 reaches (>1,600 per animal) and consistently yielding stable forelimb trajectory extraction.
We developed an automated reach-to-grasp training system to maximize skilled forelimb use. Using this novel paradigm, we aim to establish a task-specific dystonia model and explore how genetic susceptibility and motor overuse interact to cause maladaptive motor phenotypes.