Food and water restriction lead to differential learning behaviors in a head-fixed two-choice visual discrimination task for mice

Abstract

Head-fixed behavioral tasks can provide important insights into cognitive processes in rodents. Despite the widespread use of this experimental approach, there is only limited knowledge of how differences in task parameters, such as motivational incentives, affect overall task performance. Here, we provide a detailed methodological description of the setup and procedures for training mice efficiently on a two-choice lick left/lick right visual discrimination task. We characterize the effects of two distinct restriction regimens, i.e. food and water restriction, on animal wellbeing, activity patterns, task acquisition, and performance. While we observed reduced behavioral activity during the period of food and water restriction, the average animal discomfort scores remained in the 'sub-threshold' and 'mild' categories throughout the experiment, irrespective of the restriction regimen. We found that the type of restriction significantly influenced specific aspects of task acquisition and engagement, i.e. the number of sessions until the learning criterion was reached and the number of trials performed per session, but it did not affect maximum learning curve performance. These results indicate that the choice of restriction paradigm does not strongly affect animal wellbeing, but it can have a significant effect on how mice perform in a task.

Fig 1. Behavioral apparatus and training protocol.

A. Setup used for head-fixed visual conditioning. Arrow ‘m’ points to a head-fixed mouse, resting on a Styrofoam ball, in front of a centrally positioned monitor and the two lick spouts (arrow ‘s’). Arrow ‘v’ indicates the pinch valves for reward delivery. B. Schematic of the behavioral setup as seen from behind. C. Lick detection. Top left: Position of dual lick spouts in front of the mouse. Bottom left: Photo of fully assembled 3D printed lick spout holder. Right: Electrical circuit for contact/lick detection on a single lick spout. D. 3D renderings of head-bar and head-bar holder. E. Temporal sequence of within-trial phases. Reward is delivered immediately upon the first (correct) lick in the response window. F. Overall experimental timeline depicting main experimental and training stages.

Fig 2. Animal weights and discomfort scores.

A. Mean (±SD) daily weight of each experimental group across the entire experiment (gray: non-implanted, n = 4; red: food-restricted, n = 6; blue: water-restricted, n = 6). B. Average weight, in percentage of reference weight, throughout the period of food or water restriction. C. Amount of supplemented food (red) or water (blue) given (average of entire training period). D. Amount of soymilk (red) or water (blue) earned during training (average of entire training period). E. Mean (±SD) daily score of food (red) and water (blue) restricted mice over the entire experiment. Circles indicate scores as judged by the animal welfare officer. F. Daily score averaged across the period of food/water restriction. Total score is the sum across all five individual scores (look/posture, urine/feces, body condition, dehydration signs, activity; MWU test, *p = 0.018, ** p = 0.002). G. Distribution of daily measured weight as a function of the daily determined discomfort score for food- and water-restricted mice (MWU test, *p = 0.016, ** p = 0.002). Numbers in distribution plots indicate n in individual daily measurements. H. Example photo of food-restricted mouse (discomfort score 0, ‘sub-threshold’). I. Example photo of water-restricted mouse (discomfort score, total = 2, ‘moderate’; look/posture = 1; dehydration signs = 1). All panels: grey crosses (non-implanted), red triangles (food-restricted) and blue squares (water-restricted) indicate averages for individual animals.

Fig 3. Continuous monitoring of physical activity in the home cage.

A. Heat maps depicting baseline-normalized physical activity per hour (x-axis) throughout the days of the experiment (y-axis) as measured in the home cage, averaged across the non-implanted and experimental groups separately. Arrows: ‘S’ indicates the two days on which surgeries were performed; ‘R’ the day on which food or water restriction started; ‘P’ start of the post-training period (note that the food-restricted group received ad libitum access to food from two days before this post-training period). Cage changes can be identified as single bright data points, weekly reoccurring at 08:00. B. Hourly averaged (±SEM) home cage activity for the first seven days of adaptation to the reversed day/night cycle. C. Hourly averaged (±SEM) activity centered on the day of head bar implantation (blue, experimental group) or a matched day for animals that did not receive a head bar (gray, non-implanted group). D. Six days of average hourly home cage activity (mean±SEM), starting one day before onset of food or water restriction. E. Average (±SEM) 24hr home cage activity pattern throughout the entire period of training. Data of the experimental groups during the period of training (14:00–18:00) were left out. F. Mean (±SEM) home cage activity in the (active) light-off period (training period excluded). ‘B’: Baseline period, day -14 to 0. ‘T’: Training period, day 1 to 66. ‘P’: Post-restriction period, day 67 to 85. Crosses, triangle and squares indicate data points from individual mice (* MWU test, p<0.02; ** MWU test, p = 0.003).

Fig 4. Visual discrimination in a head-fixed two-choice task.

A, B. Learning curves of food (A) and water (B) restricted mice. Red and blue lines show the day-to-day performance for each animal, starting at the first day of visual discrimination learning. Black curve is a sigmoidal fit to data from animals that reached criterion (>66% correct). Gray arrows indicate the day on which mice reached criterion. C. Mean (±SEM) learning curve of all food- (red) and water- (blue) restricted mice in the overall experimental timeline. ‘F’ and ‘W’ indicate start days of training food- and water-restricted mice. Gray bars denote days without training. D. Maximum learning curve performance, determined by the sigmoidal fit in the time period during which mice were trained (as shown in A and B). E. Average number of days until criterion (>66% correct) was reached (MWU test, *p = 0.038). F. Average number of trials that mice performed per day in each of the training stages (‘Pre-tr 1’: pre-training stage 1, 1 lick spout; ‘Pre-tr 2’: pre-training stage 2, 2 lick spouts; ‘Vis. Discr.’: visual discrimination task (training stage 3); MWU test, ns: not significant, *p = 0.023, **p = 0.003). The red/blue shaded area in bars of the visual discrimination stage indicates the fraction of rewarded (correct) trials.