By releasing acetylcholine (Ach), the basal forebrain cholinergic neurons and their widespread projections to the cortical mantle play a key role in the control of cognitive functions, like sensory processing, attention, arousal and reinforcement expectation. Several age-related changes were observed in these neurons, such as degeneration of dendrites, axons and synapsis; nonetheless the connection between cholinergic activity during learning and the age-related neurodegeneration is still not completely clear.

To address this, we combined fluorescent in vivo recording techniques and optogenetic manipulations in headfixed mice during an auditory cued Pavlovian conditioning task. By using fiber photometry, we measured ACh release in the basolateral nucleus of the amygdala (BLA) using a newly developed ACh sensor. As another approach, we optogenetically inhibited the cholinergic neurons of the horizontal diagonal band of Broca (HDB) during the presentation of conditioned stimuli (CS).

Our results suggest that cholinergic cells respond to the reward-predicting CS and to the unconditioned stimuli with an increase of activity and ACh release. It seems that acetylcholine release in the BLA occurs only during the reward-predicting but not the punishment-predicting sensory stimuli. In old mice, ACh levels in the BLA showed no response to the CS.

In the optogenetic manipulation experiments, we found that the animals, which received the inhibitory virus in their HDB, had difficulties learning the task compared to the animals, which received the control virus. According to our results, ACh release from basal forebrain neurons is required during the acquisition of the CS-US association during Pavlovian conditioning.

1. Figure (A,B) Ach4.3 sensor photometry activity of a young animal. When aligned to the auditory cue, BLA showed an increased on the ACh flow related to the reward-predicting cue (A), but not to the punishment one (B).(C,D) Ach4.3 sensor photometry activity of an old animal. When aligned to the auditory cue Ach BLA levels showed no response either to the reward-predicting cue (C), nor to the punishment one (D). We can see a bit sharper reaction in case of the arriving of the punishment compared to the arriving of the reward.

2. Figure (A) Learning curves of a young ArchT inhibited animal (up) and young control animal (down). In shades of yellow-brown are represented the different learning stages. Green line reflects licks to reward predicting cues among all possible reward predicting cue trials. In red the same but for the punishment predicting cue. Inhibited animals showed lower levels of lick rate and discrimination.(B) Anticipatory lick rate of control animals (red) and ArchT animals (blue). Control animals showed significant higher levels of anticipatory lick rate.