Abnormal Dopamine Reward Prediction Errors in a Mouse Model of Chronic Dopamine D2 Receptor Over-Expression in the Striatum

BELLO, EP; GOLDMAN, O; SIMPSON, EH; JEONG, N

Berlin

Congreso; 11th FENS Forum of Neuroscience; 2018

Federation of European Neuroscience Societies

Abnormal Dopamine Reward Prediction Errors in a Mouse Model of Chronic Dopamine D2 Receptor Over-Expression in the StriatumEstefanía P. Bello 1,4, Nuri Jeong 2, Olivia Goldman 2, Eleanor H. Simpson 2,31. Department of Neuroscience, Columbia University, New York, NY10032, USA2. New York State Psychiatric Institute, New York, NY10032, USA3. Department of Psychiatry, Columbia University, New York, NY10032, USA4. IFIBIO Houssay, University of Buenos Aires-CONICET, Buenos Aires, C1121ABG, ArgentinaAims: Imaging studies support the hypothesis that the symptoms of schizophrenia are related to alterations in dopamine. Aberrant dopamine-reward processing has been implicated in both positive (hallucinations) and negative (avoilition) symptomatology. Midbrain dopamine neurons compute reward prediction errors (RPEs) that, by signaling discrepancies between expected and received outcomes, are proposed to serve as a basic process underlying associative learning. Patients with high negative symptoms show a reduced capacity for learning frompositive outcomes, but do not differ from healthy controls in learning from negative outcomes. However, it is unclear whether such impairments reflect impaired encoding of positive RPEs per se, or a deficit in representing the expected reward value of choices themselves.Methods: To understand if aberrant dopamine encoding of RPEs could explain the observed learning deficits, we studied RPE encoding in a mouse model of the increased striatal D2 receptor activity observed in patients with schizophrenia (D2R-OE mice). D2R-OE and control mice were trained in a probabilistic learning task while phasic dopamine release in the ventral striatum (nucleus accumbens) was monitored in real-time from chronically implanted carbon fiber microelectrodes using fast scan cyclic voltammetry.Results: The dopamine encoding of positive RPEs is altered in D2R-OE mice while negative RPE signaling is normal.Conclusions: The valence specific deficit in dopamine encoding of RPEs in D2R-OE mice is consistent with the finding that patients learn normally from negative, but not positive outcomes. This finding suggests that reward learning deficits in patients may result from a valence specific deficit in RPE encoding.