Reward-related activity in the human motor cortex

Abstract

The human primary motor cortex (M1) participates in motor learning and response selection, functions that rely on feedback on the success of behavior (i.e. reward). To investigate the possibility that behavioral contingencies alter M1 activity in humans, we tested intracortical inhibition with single and paired (subthreshold/suprathreshold) transcranial magnetic stimulation during a slot machine simulation that delivered variable money rewards for three-way matches and required no movement. A two-way match before the third barrel had stopped (increased reward expectation) was associated with more paired-pulse inhibition than no match. Receiving a large reward on the preceding trial augmented this effect. A control task that manipulated attention to the same stimuli produced no changes in excitability. The origin of this reward-related activity is not clear, although dopaminergic ventral tegmental area neurons project to M1, where they are thought to inhibit output neurons and could be the source of the finding. Transcranial magnetic stimulation of M1 may be useful as a quantitative measure of reward-related activity.

Fig. 1

Experimental design for Experiments 1 and 2.

Fig. 2

In Experiment 1, the condition of two-way match associated with reward expectation significantly decreased the conditioned MEP ratio, both when the finger flexors (F = 8.8, P = 0.003) and the wrist extensor (F = 10.4, P = 0.002) were tested. Conditioned MEP ratios were larger during the experimental tasks compared with the pre-task baseline (for both muscles combined, F = 1.0, P = 0.002; for the flexors, F = 11.3, P = 0.001; for the extensor, F = 1.8, P = 0.178). Left panel: Conditioned MEP ratio by match condition and pre-task baseline for finger flexors. Right panel: Conditioned MEP ratio by match condition and pre-task baseline for wrist extensors. The columns represent means and the error bars their SEs.

Fig. 3

In Experiment 2, attention engagement to the two-way match condition without association with reward expectation does not change the conditioned MEP ratio (F = 1.157, P = 0.283). Conditioned MEP ratio by match condition and pre-task baseline. The columns represent means and the error bars their SEs.

Fig. 4

In Experiment 1 for finger flexors, the condition of two-way match associated with reward expectation decreased the conditioned MEP ratio for 2-and 10-ms ISI following large ($1 or $5) but not small ($0.25 or no reward) rewards. Two-way matches produced significantly more inhibition when they followed large ($1 or $5) compared with small ($0.25 or no reward) rewards (for large rewards, two-way matches vs. no matches, F = 8.0, P = 0.005; for small rewards, two-way matches vs. no matches, F = 2.0, P = 0.159; for two-way matches, large rewards vs. small rewards; F = 6.6, P = 0.011). The columns represent means and the error bars their SEs.