Abstract
Low doses of dopamine D(1) agonists improve working memory-related behavior, but high doses eliminate the improvement, thus yielding an 'inverted-U' dose-response curve. This dose-dependency also occurs at the single neuron level in the prefrontal cortex where the cellular basis of working memory is represented. Because signaling mechanisms are unclear, we examined this process at the neuron population level. Two D(1) agonists (2-methyldihydrexidine and CY208,243) having different signaling bias were tested in rats performing a spatial working memory-related T-maze task. 2-Methyldihydrexidine is slightly bias toward D(1)-mediated β-arrestin-related signaling as it is a full agonist at adenylate cyclase and a super-agonist at β-arrestin recruitment, whereas CY208,243 is slightly bias toward D(1)-mediated cAMP signaling as it has relatively high intrinsic activity at adenylate cyclase, but is a partial agonist at β-arrestin recruitment. Both compounds had the expected inverted U dose-dependency in modulating prefrontal neuronal activities, albeit with important differences. Although CY208,243 was superior in improving the strength of neuronal outcome sensitivity to the working memory-related choice behavior in the T-maze, 2-methyldihydrexidine better reduced neuron-to-neuron variation. Interestingly, at the neuron population level, both drugs affected the percentage, uniformity, and ensemble strength of neuronal sensitivity in a complicated dose-dependent fashion, but the overall effect suggested higher efficiency and potency of 2-methyldihydrexidine compared to CY208,243. The differences between 2-methyldihydrexidine and CY208,243 may be related to their specific D(1) signaling. These results suggest that D(1)-related dose-dependent regulation of working memory can be modified differentially by functionally selective ligands, theoretically increasing the balance between desired and undesired effects.