Optical activation of midbrain dopamine neurons: do high and low stimulation frequencies produce functionally different effects?
Millard et al. propose1 that 20- and 50-Hz stimulation of optically excitable, midbrain dopamine neurons triggers functionally distinct effects: The "physiological" firing elicited by a 20-Hz train mimics a reward-prediction error (RPE) whereas the "supra-physiological" firing induced by a 50-Hz train creates "a sensory event that acts as a reward in its own right." Only the 50-Hz trains supported vigorous responding in the face of elevated response costs and sufficed to produce reward-specific Pavlovian-Instrumental Transfer (PIT). Nonetheless, the 20-Hz trains did produce unblocking. Here, we propose a more parsimonious account of these findings that is consistent with the results of prior experiments on operant responding for rewarding optical2-4 or electrical5-9 stimulation and with a new theory10,11 of the role of dopamine signaling in associative conditioning: 1) The effects of the 20- and 50-Hz trains differ quantitatively rather than qualitatively. They are mapped onto a single dimension, reward intensity, which reflects the aggregate release of dopamine in the critical terminal field(s). 2) The non-linear dependence of operant responding on both the aggregate dopamine release and the response cost3 can explain why responding for the 20- and 50 Hz trains diverged as response cost grew and predicts that this divergence can be eliminated by increasing the number of dopamine neurons stimulated. 3) Terminal-field dopamine concentration provides no information about pulse frequency per se. 4) Operant responding, PIT, and unblocking all depend on reward intensity, but the form and parameters of this dependence can differ, thus generating the observed contrasts in the impact of the 20- and 50-Hz trains. 5) The tested range of experimental conditions is too narrow to specify what constitutes "physiological" firing.