The nucleus accumbens (NAc) has long been thought of as a limbic-motor interface. Despite behavioral and anatomical evidence in favor of this idea, little is known about how NAc neurons encode information about motivationally relevant environmental cues and use this information to affect motor action. We therefore investigated the firing of these neurons during the performance of a discriminative stimulus (DS) task using simultaneous multiple single-unit recordings in rats. In this task, two stimuli are randomly presented to the animal: a DS, which signals the availability of a sucrose reward contingent on an operant response, and a similar but nonrewarded stimulus (NS). Subpopulations of NAc neurons increased or decreased their firing in association with several distinct components of the task. In this paper, we investigate cue- and operant-responsive neurons. Neurons excited and inhibited by cues showed larger firing changes in response to the DS than the NS and larger changes when the animal made an operant response to the cue than when the animal failed to respond. Excitations during operant responding were not modulated by the information contained by the cue, whereas inhibitions during operant responding were somewhat larger if the operant response occurred during the DS and somewhat smaller if they occurred in the absence of a cue. These results are consistent with the hypothesis that the firing of subpopulations of NAc neurons encode both the predictive value of environmental stimuli and the specific motor behaviors required to respond to them.

Reward-predictive cues exert powerful control over behavioral choice and may be a critical factor in drug addiction. Reward-seeking elicited by predictive cues is facilitated by the release of dopamine in the nucleus accumbens (NAc), yet the contribution of dopamine to the specific NAc firing patterns that underlie goal-directed behavior has remained elusive. We present evidence that subpopulations of NAc neurons that respond to predictive cues require the dopaminergic projection from the ventral tegmental area (VTA) to promote reward-seeking behavior. Rats trained to perform an operant response to a cue to obtain a sucrose reward were implanted with both multiunit recording electrodes in the NAc and microinjection cannulas in the VTA. Both the behavioral response to cues and the cue-evoked firing of NAc neurons were blocked by injection of the GABA B agonist baclofen into the VTA. An additional group of rats was trained on the same task and then implanted with microinjection cannulas in the NAc. Like VTA baclofen injection, injection of dopamine receptor antagonists into the NAc profoundly reduced cue-elicited reward seeking. Together, these results support the conclusion that both the behavioral response to the cue and the specific NAc neuronal firing that promotes the response depend on dopamine release within the NAc. Our findings suggest a neural mechanism by which the dopamine-dependent firing of NAc neurons mediates goaldirected behavior.

. Firing of nucleus accumbens neurons during the consummatory phase of a discriminative stimulus task depends on previous reward predictive cues . J Neurophysiol 91: 1866 -1882, 2004. First published November 26, 2003 10.1152/jn.00658.2003. The nucleus accumbens (NAc) plays an important role in both appetitive and consummatory behavior. To examine how NAc neurons encode information during reward consumption, we recorded the firing activity of rat NAc neurons during the performance of a discriminative stimulus task. In this task, the animal must make an operant response to an intermittently presented cue to obtain a sucrose reward delivered in a reward receptacle. Uncued entries to the receptacle were not rewarded. Both excitations and inhibitions during reward consumption were observed, but substantially more neurons were inhibited than excited. These excitations and inhibitions began when the animal entered the reward receptacle and ended when the animal exited the receptacle. Both excitations and inhibitions were much smaller or nonexistent when the animal made uncued entries into the reward receptacle. In one set of experiments, we randomly withheld the reward in some cued trials that would otherwise have been rewarded. Excitations and inhibitions were of similar magnitude whether or not the reward was delivered. This indicates that the sensory stimulus of reward does not drive these phasic responses; instead, the rewardassociated responses may be driven by the conditioned stimuli associated with reward, or they may encode information about consummatory motor activity. Another population of NAc neurons was excited on exit from the reward receptacle. Many of these excitations persisted for tens of seconds after the receptacle exit and showed a significant inverse correlation with the rate of uncued operant responding. These findings are consistent with a contribution of NAc neurons to both reward consummatory and reward seeking behavior.