Neural Mechanisms of miss- and touch-guided sensorimotor corrections

PROJECT SUMMARY When you reach for an apple and miss, you immediately make corrective submovements to make contact. And after contact, precise touches at your fingers reshape your grasp. Unknowingly, our tongues move with similar sophistication during speech and as we use the tongue’s sense of touch to handle, chew, and swallow food. Tongue discoordination in systemic neurological diseases results in aspiration pneumonia, a leading cause of death in Parkinson’s and ALS. But because tongue movements are extremely fast and hard to measure, neural mechanisms of tongue control have been understudied. Here we combine high speed videography and deep learning guided segmentation to quantify tongue kinematics with millisecond timescale precision. Next we develop new behavioral paradigms inspired by primate target-jump reach-and-grasp tasks (Aim 1). As with the primate limb, when the mouse tongue undershoots an unexpectedly withdrawn target (water spout), it immediately produces corrective submovements to make contact. And as with a grasping hand, when the mouse nicks a moved spout with the left or right side of its tongue, it immediately redirects the lick left or right for better contact. To identify signals and circuits underlying these corrections, we combine brainwide neural recording and manipulation (Aim 2). So far, neural recordings identify representations of misses, nicks, and aimed corrections in single neurons and neural population dynamics in multiple brain regions. Miss-guided corrections are impaired by both cortical and cerebellar inactivations. But mechanisms of touch-guided corrections are fundamentally different. The ability to re-steer a tongue after a nick is impaired by inactivation of superior colliculus, but not tongue-jaw sensory cortex (TJS1), tongue jaw motor cortex (TJM1), orofacial premotor cortex (ALM), or a lick-associated region of cerebellum (fastigial nucleus). This double dissociation suggests that touch-guided tongue steering may use midbrain pathways associated with visually-guided orienting. To test this idea (Aim 3), we combine neural tracing, tongue surface receptive field mapping, and optical microstimulation across precise locations of the colliculus. Our pilot data support the existence of a topographically ordered tongue touch-to-tongue steering map on the surface of the colliculus which essentially re-purposes both the logic and the layout of more commonly studied visually-guided saccades. In sum, our goals are twofold: First our proposed research will finally clarify which aspects of the extended motor system are important for which aspects of sensorimotor tongue control. Second, by comparing our findings to what’s known sensorimotor processes underlying reaching and orienting, we aim to distinguish idiosyncratic solutions to narrower sensorimotor control problems from general principles that hold across effectors and species.