Sensory prediction errors (SPEs) are generated by comparing the sensory consequences of a motor command with the actual sensory feedback. Sensory prediction errors are hypothesized to be the critical error signals for model-based, implicit motor adaptation. However, error processing is not static. Instead, the CNS tunes error sensitivity and sensory feedback to match task demands. Also, updating of movements implies that information about errors occurring during a movement must persist to influence subsequent movements. One possibility is that information about motor errors is retained between movements. In spite of the importance of errors in motor control, very little is known about how SPEs are encoded and processed at the neuronal level. A long standing hypothesis is that the cerebellum implements a forward internal model that predicts the sensory consequences of a motor command and uses SPEs to control movements and update the model. For nearly 50 years, the accepted hypothesis has been that the low frequency complex spike (CS) discharge of Purkinje cells encodes motor errors. However, considerable evidence shows that error signaling in the cerebellum does not rely solely on CS discharge. Recently, our laboratory demonstrated that the high frequency simple spike (SS) discharge of Purkinje cells robustly encodes both predictions and sensory feedback about motor errors. Based on these novel observations, this proposal tests a series of hypotheses on the error signals encoded in the SS discharge of Purkinje cells in Rhesus monkeys during pseudo-random tracking tasks. By controlling the visual feedback available during tracking, Specific Aim 1 tests the hypotheses that the SS error signals at lag times are due to visual feedback and are used to compute SPEs. Furthermore, we will test whether the feedback error signals adapt to match the new feedback conditions.
Specific Aim 2 tests the hypotheses that the SS error signals at feedforward times are predictions of the upcoming errors and are also used to generate SPEs by introducing time delays in visual feedback.
Specific Aims 1 and 2 also examine how CS modulation changes with altering the visual feedback. We hypothesize that CS modulation is driven by an increase in SPEs.
Specific Aim 3 tests the hypothesis that the SS discharge encodes for a working memory of errors. In addition, Aim 3 tests a complementary hypothesis that the SS discharge predicts errors up to several seconds prior to movement.
Specific Aim 4 examines how motor errors are represented in the firing of cerebellar nuclear neurons, the target of Purkinje cells and the output stage of the cerebellum. Overall, the experiments will provide a comprehensive examination of error signaling in the firing of Purkinje cells and cerebellar nuclear neurons, including how predictive and feedback signals adapt to match changes in feedback and if these neurons encode a working memory of motor errors. The results have the potential to fundamentally change our view of how the cerebellum encodes and processes motor errors.
Project Relevance The cerebellum is crucial to our ability to make smooth, coordinated movements. Also, numerous disorders affect the cerebellum with devastating consequences. The results from these studies may lead to a better understanding of both normal and abnormal cerebellar function.
Showing the most recent 10 out of 59 publications
