Fig. 1. The touchscreen Autoshaping task to assess Pavlovian approach behaviours toward reward-predicting stimuli.

a Layout of the Autoshaping touchscreen operant chamber depicting the two screens (left, CS-; right, CS+) and the reward magazine (RM) delivering strawberry milkshake reward (10 μl). Each chamber was equipped with a back infrared photobeam (BIR) to initiate trials, and two front infrared photobeams (FIR) on each side of the RM to record approaches to the CS screen. An infrared photobeam inside the RM (not displayed) recorded latency time to collect rewards. b Flowchart overview of the Autoshaping task during acquisition (left) and reversal (right) training sessions. (left) Following a variable ITI, a trial initiated after breaking the BIR followed by the presentation of the stimulus (CS+ or CS−) during 10 s. Upon CS+ offset a reward was delivered and a new ITI began once the mouse pulled away from the RM. Upon CS− offset, no reward was delivered, and a new ITI started. Within a single session, CS+ and CS− trials alternated pseudo-randomly. In total, each session ended after 20 CS+ and 20 CS- trials or after 60 min, whichever occurred first. (right) Following 10 acquisition sessions (1 session/day), mice undergo a total of 10 reversal sessions, in which the location of the CS+ and CS- were reversed. c (left) In contrast to the previous, both CS screens (left and right) had 50% of probability to deliver rewards in non-deterministic trials. Contingencies after CS+ or CS- remained similar as previously described. Within a single session a total of 20 CS+ and CS- trials were presented. (right) After 10 consecutive non-deterministic training sessions, mice followed 10 consecutive deterministic training sessions as described in (b). Figure 1a was created with BioRender.com.