Stroke is a common cause of disability, with the majority of stroke patients exhibiting chronic motor impairments. A commonly observed pattern of impairment is fragmented movement control (i.e. loss of smooth transitions during movements) and difficulty with sequences, even with physical rehabilitation therapy. We have poor understanding as to the neural circuit basis of these impairments. Current efforts in neuromodulation for motor recovery have resulted in variable effects, potentially due to nonspecific targeting of modulation. Work by Dr. Karunesh Ganguly's group, conducted in rats, has demonstrated that there is a characteristic low frequency [1.5-5Hz] quasi-oscillatory activity in low frequency local field potential (LFP) that occurs during a skilled reaching task in both the primary motor cortex (M1) and dorsolateral striatum (DLS). This phenomenon occurs when a group of neurons (ensemble) in each region fires in a synchronous manner, time-locked to the reach-to-grasp motion. In animals that have motor deficits due to stroke in M1, premotor cortex (M2) rearranges to recapitulate forelimb behavior representations that were lost; M2 is also connected to DLS. Additionally, electrical stimulation delivered therapeutically to M2 at the expected time onset of low frequency LFP power quasi-oscillations can improve accuracy in a skilled reaching task, but animals still display fragmented movement control ? characterized by a loss of smoothness that is present in both human stroke subjects and animal models. The central hypothesis is that similar to how coordinated LFP and ensemble spiking activity is a marker of M1-DLS coordination, increasing neural synchrony between M2 and DLS will result in increased recovery of skilled motor function after M1 stroke. In both aims, motor function will be assessed in rats with a reaching task to retrieve pellets. Behavioral outcomes such as reach time, kinematic trajectory, amplitude (distance paw traveled in a single reach), and successful retrieval will be assessed. In this proposal, my first aim is to characterize how coordinated neural activity in M2 and DLS evolves with recovery. In the second aim, I will assess how M2 ensemble spiking-dependent stimulation of DLS affects performance of the reach to grasp task in stroke animals with chronic motor deficits. In addition to my mentor's expertise on stroke rehabilitation, treatments, and neural interfaces, guidance from my thesis committee members Dr. Philip Starr (neurosurgery, deep brain stimulation, closed-loop stimulation) and Professor Michel Maharbiz (recording and stimulation electrode paradigms) support the feasibility of this study. We predict that increasing synchronizing M2-DLS activity post-stroke using population spike timing-dependent plasticity principles will improve behavioral outcomes after a chronic stroke plateau: reach time will decreased to baseline; reach trajectory will be smoother and more stereotyped, as assessed by a deep learning method that automatically captures movement parameters; and accuracy of obtaining pellets will increase. By accomplishing these aims, I will contribute to knowledge of how cortico-striatal connections play a role in motor recovery after stroke.
Stroke is a common cause of disability, resulting in many motor deficits such as poor coordination and smooth velocity regulation despite extensive physical rehabilitation, with little understanding of the neural circuit basis for these impairments. After primary motor cortex (M1) stroke, the perilesional premotor cortex (M2) rearranges to recapture movement representations that were lost, and both regions have monosynaptic projections to a deep brain structure in the striatum, the dorsolateral striatum (DLS) in rodents. This work will characterize the how M2 and DLS neural activity simultaneously evolve with stroke recovery, as well as assess whether closed- loop M2 ensemble spiking-dependent electrical stimulation of DLS will improve motor function after chronic stroke.
