Substantia nigral activity in self-timed movements

Abstract

The basal ganglia (BG) are a set of subcortical nuclei involved in movement and other aspects of behavioral control. Diseases of the basal ganglia cause difficulty in the initiation and execution of movements. The main output nuclei – the substantia nigra pars reticulata (SNr) and globus pallidus interna (GPi) – are composed of tonically-active GABAergic neurons that project to movement-related brain areas including the superior colliculus and thalamus. For skeletomotor movements, previous studies reported that a majority of SNr/GPi neurons exhibit increased spiking activity during movement, while a smaller proportion exhibit decreased spiking. To reconcile this finding with the inhibitory nature of SNr/GPi output, it has been proposed that the BG output could be involved in action-selection, whereby some output neurons turn off to disinhibit desired movements, while others turn on to inhibit unwanted movements. Recent structural studies on the spatial organization of parallel cortico-BG-thalamic loops in rodents, as well as their similarities to somatotopic maps in primates, may provide insights for dissecting functionally divergent BG outputs. In this study we performed extracellular recordings from single SNr neurons in mice, while the animals performed two types of self-timed movements, reaching and licking, in alternating blocks of trials. We chose self-timed movements based on clinical findings suggesting that self-initiated movements are more severely affected in Parkinson’s disease than movements that are made in rapid reaction to external sensory cues. Similar to previous studies in monkeys, we found that a majority of SNr units increased firing during pre- and peri-movement periods of the self-timed movement task. We hypothesized that the sign of the movement-related responses could be related to the movement-selectivity of the unit under study, but we found that the only a small subset of the SNr units distinguished between reaching and licking while most units showed congruent responses in both tasks. However, we found that many SNr neurons showed modulated activity hundreds to thousands of ms before movement onset. It had been previously argued that GPi neurons in monkeys were activated too late to play a role in movement initiation, but our findings suggest that BG output neurons could play a role in movement initiation when animals make movements “on their own”, without abrupt external prompting. We also examine the nature of striatal input to SNr neurons, by recording SNr responses evoked by spatially-selective optogenetic activation of neurons of the striatal direct pathway (dSPNs) and indirect pathway (iSPNs). The “Standard Model” posits that dSPNs should inhibit SNr, and thus facilitate movement, whereas iSPNs should excite SNr, thus inhibiting movement. However, we found that striatal stimulation of dSPNs or iSPNs often produced robust “non-canonical” responses in SNr neurons, with signs opposite to those predicted from the Standard Model. We were unable to relate the sign of stimulation-evoked responses to the sign of movement-related activity in SNr units, nor did stimulation-evoked responses conform to a strong prediction of a prominent action-selection model. Taken together, our results invite a re-consideration of some longstanding assumptions about the basal ganglia’s role in movement control, and suggest a more complex, divergent pattern of functional connectivity in the striatonigral circuit.