One of the primary disabling features of Parkinson's disease is the inability to properly start and stop action sequences, often manifested as “freezing” or “festination” in gait. And as movement disorders represent disease processes in the basal ganglia, any description of the neural circuity during normal physiology is extremely helpful in unraveling the electrophysiology of the diseased state. Jin and Costa have now published a study of single unit potential firing related to the initiation and cessation of motor action sequences in a rodent operant task (Nature 2010;466:457-462). In this work, mice learned a sequence of actions (lever presses) to obtain a sugar solution reward. Concurrent extracellular recordings in both striatum and substantia nigra revealed neurons specifically active during starting and stopping the sequence of learned lever presses, as opposed to cellular activity more broadly sensitive to motor function.
The mice in this study learned a particular sequence of activity involving a fixed-ratio schedule, whereby 8 successive lever pushes resulted in a sucrose solution reward. After several days of training, the mice demonstrated consistent cycles of 8 lever pushes, showing decreases in sequence lengths, inter-sequence interval times, and variability in within-sequence press rates (Figure, b). The authors discovered in 3 different neuronal populations (medium spiny neurons in the striatum, substantia nigra GABAergic neurons, and substantia nigra dopaminergic neurons) selective increases in firing just before the onset of the sequence of 8 lever presses. This increase in firing rate was statistically different from any firing rate changes occuring between lever presses in the sequence, and furthermore many of these neurons showed significant firing rate changes just before the last lever press (Figure, f-k).
Perhaps most interesting was that while there were no time variations for the proportion of lever-press related neurons during the sequence training, the percentage of neurons demonstrating sequence start/stop-related activity increased. This increase in activity occurred over the same time course of sequence learning itself (days 1-6 of training). Additionally, in a 2-lever version of the task, where 1 lever was associated with a smaller reward, the authors did find significantly higher firing rates in dopaminergic neurons when the larger reward lever was pressed (likely associated with the higher reward expectation). However, sequence start/stop activity was independent of this reward expectation, and occurred for both small and large rewards.
The NMDA receptor was also removed using the Cre/loxP method, to produce a strain of mice with specific NMDAR1 absence in striatum. These mice demonstrated impaired sequence learning and significantly lower proportions of neurons with sequence start/stop activity, implying the importance of this receptor in the plasticity of this sequence learning system. In summary, these authors have excitingly demonstrated sequence start/stop specific activity in a set of neurons in the nigrostriatal circuitry. It is likely that the activity of this sequence system is impaired in patients with Parkinson's disease, given the difficulties with initiation and termination of movement. This may also have implications for therapeutic drug delivery or stimulation systems, which in the future for maximum efficacy, might require precise dosage timing with respect to motor sequence activation.