Dopamine and acetylcholine correlations in the nucleus accumbens depend on behavioral task states.

Dopamine release in the nucleus accumbens (NAcc) changes quickly in response to errors in predicting events like reward delivery(1)(,)(2)(,)(3) but also slowly ramps up when animals are moving toward a goal.(4)(,)(5)(,)(6)(,)(7)(,)(8)(,)(9)(,)(10) This ramping has attracted much recent attention, as there is controversy regarding its computational role(5)(,)(7)(,)(9)(,)(11) and whether they are driven by dopamine neuron firing(7)(,)(8)(,)(9) or local circuit mechanisms.(5)(,)(6) If the latter is true, cholinergic transmission would be a prime candidate mechanism,(12)(,)(13)(,)(14) and acetylcholine and dopamine signals should be positively correlated during behavior, particularly during motivated approach. However, in the dorsal striatum, striatal cholinergic interneurons typically "dip" their activity when reward or associated cues are presented, in opposition to dopamine,(15)(,)(16)(,)(17)(,)(18) and acetylcholine and dopamine release is generally anti-correlated in vivo.(19)(,)(20) Furthermore, acetylcholine and dopamine have opposing effects on downstream striatal projection neurons (SPNs),(21)(,)(22) which suggests that cholinergic dips create a permissive window for dopamine to drive plasticity.(23) These studies therefore suggest that dopamine and acetylcholine should be anti-correlated during behavior. We tested between these hypotheses by simultaneously recording accumbal dopamine and acetylcholine signals in rats executing a task involving motivated approach. We found that dopamine ramps were not coincidental with changes in acetylcholine. Instead, acetylcholine was positively, negatively, or uncorrelated with dopamine depending on the task phase. Our results suggest that accumbal dopamine and acetylcholine dynamics are largely independent but may combine to engage different postsynaptic mechanisms depending on task demands.