Experimental procedures

3.1. Food restriction schedule

Animals subjected to behavioral experiments were housed individually in the animal room and their baseline body weights were recorded after a habituation period of minimally 1 week. Animals were food-restricted by maintaining a restricted diet and kept at 80-85% of free-feeding body weight throughout the behavioral

experiments with daily weighed. According to their weights, food was put daily in their home cages at least an hour after they finished experiment. Water was available

ad libitum in their home cages throughout the behavioral experiments.

3.2. Locomotor activity before 5-CSRTT

Before the beginning the 5-CSRTT procedure, animals underwent two locomotor activity tasks. On the first day, each animal was placed in a clean transparent PVC cage (47.2 L × 25.3 W × 21 H cm³) containing a thin layer of clean bedding for 30 min. Animals were allowed to explore this new arena and their distance movements were recorded and analyzed by the EthoVision tracking system (Noldus, Wageningen, the Netherlands). On the second day, their locomotion was individually monitored in the 5-CSRTT apparatus. The 5-CSRTT apparatus was divided into three areas (i.e., aperture, middle, and magazine) as depicted in Figure 2.1. Each mouse was allowed

30

to explore and habituate this new experimental environment for 30 min. Their movements were recorded by videos, and analyzed off-line by a video tracking system, TopScan 2.0 (Cleaver System Incorporated, Reston, VA, USA). Their travel distances or duration in these 2 tasks were used as a measurement of locomotor activity.

3.3. 5-CSRTT procedures

Animals were trained and tested in a modified version of the 5-CSRTT procedure (Bari, Dalley, & Robbins, 2008; Debruin, Fransen, Duytschaever, Grantham, &

Megens, 2006; Humby, Wilkinson, & Dawson, 2005) and final trial sequences were represented in Figure 2.2. Animals conducted sessions of 25 min duration with 1 session per day and 6 days per week for approximately 3 months. The 5-CSRTT procedure consisted of 3 sequential phases, including shaping, learning, and testing phases.

3.3.1. 5-CSRTT shaping phase

Mice were first trained to operate the 5-CSRTT apparatus by a series of shaping stages. In each stage, each mouse was required to reach shaping criteria in 25 min, and then they moved to next shaping stages. Under stage 1~3, a trial started with the illumination of the house light, and ended after animals collected their earned reward

31

pellets following a new trial started automatically. Stage 1: Animals were required accumulating 10 nose pokes into either stimulus-response apertures or the food magazine. Each nose poke was followed by the delivery of a reward pellet and the illumination of the magazine light. Stage 2: Each mouse was required accumulating 5 nose pokes into one of the 5 stimulus-response apertures. Nose poke into the food magazine was still followed by the delivery of a reward. But after accumulating 5 nose pokes into the food magazine, no reward was delivered from the food magazine if the animal kept performing nose pokes into the food magazine. Stage 3: Each mouse was required to accumulate 10 nose pokes into one of the 5 stimulus-response apertures, and nose poking into the food magazine was no longer followed by any delivery of a reward. Each nose poke into stimulus-response apertures was followed by the delivery of a reward pellet and the illumination of the magazine light. Stage 4:

A trial started with the illumination of the house light, and then animals had to wait an intertrial interval (ITI) of 5 sec for the illumination of light stimuli. After the ITI, one of the 5 light stimuli illuminated for a stimulus duration (SD) of 32 sec. Animals were required to accumulate 15 nose pokes into the illuminated aperture within the SD or during the following 2 sec fixed limited hold (LH) after the light stimuli extinguished.

Each nose poke into the illuminated aperture within the SD or LH was followed by the delivery of a reward pellet and the illumination of the magazine light. Each trial

32

ended after animals collected their earned reward pellets or after the LH duration and then a well-prepared new trial started automatically. Stage 5: Daily sessions started with the illumination of the house light and the magazine light, and a trial started after mice nose poked into the magazine. One of the 5 light stimuli illuminated for a SD of 32 sec after the 5-sec ITI. Each mouse that nose poked into the illuminated aperture within the SD or the 2 sec LH earned a reward pellet followed the illumination of the magazine light. Animals had 5 sec to eat their food pellet after collecting it. Each mouse nose poked into the non-illuminated aperture within the SD was signaled a punishment by a 5 sec time-out (TO) with 1 sec successive on and off of the house light. Trials ended after the 5 sec for eating their earned reward pellets or after the LH and TO duration and then a new trial started automatically. Each mouse was required to accumulate 15 completed trials for 2 consecutive daily sections and then animals underwent the training phase.

3.3.2. 5-CSRTT learning phase

Each session began with the illumination of the house light and the magazine light. A nose poke into the magazine initiated a trial and extinguished the magazine light. A fixed ITI of 5 sec preceded the illumination of stimulus-response apertures.

After the ITI, one of the 5 light stimuli illuminated during the SD. The animal had to

33

respond with a nose poke into the illuminated aperture within the SD or during the following fixed 2 sec LH period after the light stimuli extinguished. A nose poke into the illuminated aperture resulted in the delivery of a reward food pellet following the illumination of the magazine light. The magazine light remained on until the animal collected its reward pellet and then the animal had fixed 5 sec duration to eat this pellet. A trial was finished after this 5 sec duration ended following a well-prepared new trial. The behavioral sequence described above was considered as a correct

response and recorded as a correct trial. If the animal did not follow such behavioral

sequence on any trial, the trial was considered as an improper trial and the response was recorded as an improper response. There were 4 types of improper responses and they were described as below. A premature response was recorded if the mouse did not nose poke into one of the 5 stimulus-response apertures during the ITI, before stimulus illuminated. An incorrect response was recorded if the mouse nose poked into one of the non-illuminated stimulus-response apertures during the SD or the LH period. An omission response was recorded if the mouse did not perform any nose poke into one of the 5 stimulus-response apertures during the SD or the LH period. A

perseverative response was recorded if the mouse repeatedly nose poked into one of

the 5 stimulus-response apertures, whether it illuminated or not, during the period of collecting reward pellet. Immediately after an improper response, a punishment was

34

signaled by a fixed 5 sec TO with 1 sec successive on and off of the house light. An improper trial finished after the TO period following a well-prepared new trial. The general parameters of the 5-CSRTT were set as ITI = 5 sec, TO = 5 sec, and LH = 2 sec. The learning phase of the 5-CSRTT was divided into 4 learning stages based on a scheduled stepped descending sequence of SD at 16, 8, 4 and finally 2 sec. Each mouse was required to reach the criteria ( ≥ 30 trials, ≥ 70% accuracy, ≤ 20%

omission) for 2 consecutive days in first three learning stages (SD = 16, 8, and 4 sec) before passing to the next training stage. Each mouse was considered to have acquired the task when they meet the baseline performances ( ≥ 30 trials, ≥ 80% accuracy, ≤ 20% omission) for 3 consecutive days in the final stage of SD = 2 sec and then the mice underwent the testing phase. The accuracy of responding expressed as a percentage was calculated using the following formula:

%

The percentage of other behavioral measures (including correct, incorrect, omissions, premature, and perseverative responses) were calculated along with the following formula:

The latency (in second) to correct nose poke after stimulus occurred (i.e., correct latency), and the latency (in second) to collect earned food pellets (i.e., reward latency)

35

were also calculated. According to our preliminary data, mice spent longer time to learn the final learning stage (SD = 2 s). Thus a mouse that did not reach the learning criteria for up to 18 days (i.e., 3 weeks of training; under SD = 16, 8, 4 s stages) and 24 days (under SD = 2 s stage) during the 4 learning stages was considered that it could not acquire this task. Consequently, the mouse was excluded from this

experiment. Daily sessions were accumulated until animals finished required criteria in one learning stage.

3.3.3. 5-CSRTT testing phase

Once baseline performance ( ≥ 30 trials, ≥ 80% accuracy, and ≤ 20% omission for 3 consecutive days of SD = 2 s) had been established in each mouse, the testing phase began. The testing phase consisted of 4 tests in which the difficulty of the 5-CSRTT was managed by manipulations of the 5-CSRTT parameters. Each mouse was sequentially tested in one of the 4 tests once it reached the baseline performance.

Each test lasted for 25 min or 50 trials whichever completed first. Each of the 4 tests consisted of 2 testing conditions that were applied in a pseudo-random order from trial to trial and they were described as below. (1) The alteration of the ITI duration:

The duration of the ITI was either 2 or 8 seconds. (2) The alteration of the SD

duration: The duration of the SD was reduced to either 1 or 0.5 second. (3) The

36

alteration of stimulus brightness: The brightness of the light stimuli was either

relatively lighter (i.e., enhanced approximately three times the illumination of daily training) or relatively darker (i.e., the same brightness of daily training). (4) The

distracting test: A distractor condition (i.e., a 100-dB, 0.5-sec on and off white noise)

or no distractor condition were used and played during the duration of the SD. After finishing one test, each mouse was re-trained to the baseline performance for 3

consecutive days and tested again until finished the 4 tests. A mouse that did not reach the baseline performance for up to 24 days before the 4 tests was also excluded from this experiment. The behavioral measures during the 4 tests were the same as

described above in the learning phase.