Abstract

Dopamine (DA) is critical for reward-seeking. However, its specific role in reward has remained elusive. In the Intracranial Self-Stimulation (ICSS) paradigms, animals are trained to perform operant tasks to deliver trains of electrical or optical pulses to reward-related brain regions. The tight experimental control attainable by these paradigms makes them suitable for psychophysical and computational studies aiming at understanding how specific neural circuits and signals mediate reward seeking. However, common measurement methods in ICSS studies fail to distinguish between effects arising at different stages of reward processing. The reward-mountain model links the effects of experimental manipulations to specific stages of neural processing.
I used the reward-mountain model to distinguish between variables affecting the integration of reward intensity and other factors that influence ICSS. I administered a DA reuptake blocker or a DA receptor antagonist to animals working for electrical stimulation of the Medial Forebrain Bundle (MFB). The results show that DA signaling affects ICSS at a stage beyond the computation of reward intensity by modulating reward gain, costs, or the value of competing activities.
Optogenetics allows for direct optical stimulation of specific neural subtypes with tight temporal control. I present a technical development that proved critical in running long experimental sessions with rats. I used optogenetics to bypass the inputs to midbrain-DA neurons. I adapted the reward-mountain paradigm to rats working for optical stimulation of midbrain-DA neurons, allowing me to further dissect the role of DA in reward seeking. In these animals, DA reuptake blockade affected the integration of reward intensity. This contrasts with the effects produced by the same drug in rats working for electrical ICSS (eICSS). The drug also affected reward gain, cost, or the value of competing activities.
Overall, the results show that eICSS and optical ICSS (oICSS) recruit the reward system at different stages of neural processing. In eICSS, DA signaling affects reward-seeking at a stage beyond the computation of reward intensity, whereas in oICSS of midbrain-DA neurons, reward intensity is determined by a second reward-integrator downstream from the one that determines the reward intensity in eICSS.