A critical component of many human brain diseases may be changes to associative learning functions. For example, much of the aberrant behavior that characterizes drug addiction is driven by environmental cues, which the addict has learned to associate with their drug of choice. Understanding the processes and brain circuits governing such associative learning could generate novel approaches for treating these aspects of addiction. Associative learning is supported by the ability to recognize errors between expected and actual outcomes. Evidence suggests that dopaminergic neurons in ventral tegmental area (VTA) signal these prediction errors through fluctuations in phasic activity. Elevated activity in VTA neurons has been shown to signal unpredicted reward while a decline in activity signals the omission of a predicted reward. However, generating such prediction errors presumably requires comparison of the actual outcome to an a priori expectation for reward. Perhaps the best applicant for generating such outcome expectancies is the orbito frontal cortex (OFC). Not only does OFC send projections to VTA, OFC neurons also fire in anticipation of and during expected outcomes. This activity develops with learning and reflects animals' preferences for different outcomes. This role of expectancy signaling in OFC is also supported by lesion studies, which demonstrate that OFC is critical to behavior guided by outcome expectancies and for new learning in the face of unexpected outcomes. This evidence suggests a model in which OFC signaling provides critical information that leads to the generation or suppression of temporal prediction error signals in VTA to facilitate learning. This proposal will test this hypothesis, using inactivation and single-unit recording to ask how OFC is involved in learning and in the calculation of prediction errors in dopaminergic VTA neurons. For this, we will use a Pavlovian blocking task, in which associative learning for a novel cue is prevented by the simultaneous presentation of a second previously-conditioned cue. Based on our hypothesis, we predict 1) that OFC signaling will correlate with the effectiveness of blocking, 2) that OFC inactivation will impair blocking, 3) that suppression of VTA signaling during blocking (previously observed in primates) will be impaired by OFC inactivation, and 4) that contralateral lesions, which disconnect OFC and VTA, will impair the effectiveness of blocking. The results, whether they confirm or reject our hypothesis, will greatly enhance our understanding of the role OFC and VTA play in simple associative processes that underlie normal and pathological learning. ? ? ?
| Roesch, Matthew R; Calu, Donna J; Esber, Guillem R et al. (2010) All that glitters ... dissociating attention and outcome expectancy from prediction errors signals. J Neurophysiol 104:587-95 |
| Roesch, Matthew R; Calu, Donna J; Esber, Guillem R et al. (2010) Neural correlates of variations in event processing during learning in basolateral amygdala. J Neurosci 30:2464-71 |
| Calu, Donna J; Roesch, Matthew R; Haney, Richard Z et al. (2010) Neural correlates of variations in event processing during learning in central nucleus of amygdala. Neuron 68:991-1001 |
| Calu, Donna J; Schoenbaum, Geoffrey (2008) Cocaine-paired cues activate aversive representations in accumbens neurons. Neuron 57:633 |
