Sixty-five male C57BL/6J mice purchased from Jackson Laboratories at 6 weeks of age were used to test the effects of CIE on reversal learning and set-shifting. After arrival, mice were group-housed (4/cage) and allowed to acclimate to the colony room for at least two weeks. After acclimation, mice were brought to a laboratory testing room, singly housed, and maintained under a modified 12 hr light/dark cycle (lights on at 1400 hr). At all times, mice were maintained in a temperature and humidity controlled AAALAC-approved facility, with ad libitum access to water. Mice were food restricted to 85% of their free-feeding body weight by single daily feeding prior to the start of training and during discrimination training. Water was available at all times. In all respects, maintenance and treatment of mice were conducted within the guidelines for animal care as approved by the Medical University of South Carolina’s Institutional Animal Care and Use Committee and the NIH Guide for the Care and Use of Laboratory Animals (NIH Publication No.: 80-23, revised 1996).
The testing chamber for assessing performance in discrimination tasks was a rectangular box (30 × 20 × 20 cm) constructed of clear acrylic with a clear plastic lid. The chamber was divided into three separate compartments. Half of the chamber (15 × 20 × 20 cm) was the starting area in which mice were placed at the beginning of a trial. The remaining half of the chamber contained a piece of acrylic placed perpendicular to the shortest wall of the chamber to divide the front part of the apparatus into two separate choice areas (15 × 10 × 20 cm). The divider was used to deter mice from entering the alternate choice area after committing an error. During discrimination trials, a rodent food bowl (6 × 6 × 2 cm; PetSmart, Phoenix, AZ, USA) was placed on each side of the divider. Both bowls were filled with various digging materials and spices to provide specific textures and odors during the discrimination trials. A list of the different digging media and odors used are depicted in . Digging media were obtained from PetSmart (Phoenix, AZ, USA) or the Division of Laboratory Animal Research at the Medical University of South Carolina while all odors were obtained from Spice Place (Keedysville, MD, USA) or local grocery stores. In each discrimination trial, one of the bowls in the testing chamber was baited with a buried food reward (1/4th Honey Nut Cheerio Loop, General Mills, Minneapolis, MN). All trials were video taped using a mini digital camcorder (Insignia, Richfield, MN, USA).
Mice were habituated to materials used in the discrimination trials by placing one of the rodent food bowls filled with a small sample of each of the digging medias and odors listed in . To prompt mice to dig in the bowl, 1/4th of a Honey Nut Cheerio Loop was placed on top of the bowl. Sample bowls were placed inside the mouse home cage and remained there overnight the day before training began.
General Testing Procedure
Discrimination tasks were presented in the same order for all mice and each type of discrimination was conducted on separate days. All discrimination trials were conducted in the same manner except for differences in the placement of the food reward and the type of cues presented (). Each trial began by placing a mouse in the start area of the testing chamber and allowing mice to make a choice between two cued bowls, one of which contained the buried food reward. The mouse was allowed to dig in one of the food bowls to find the bait. The first four trials were exploratory and allowed the mouse to dig in either of the bowls. If the mouse dug in the unbaited bowl, the trial was recorded as an error but the mouse was still allowed to go to the other bowl, dig and obtain the bait. If the mouse dug in the baited bowl, the trial was recorded as a correct choice and was counted towards reaching criterion (i.e., 6 consecutive correct trials). Following the exploratory trials, mice were only allowed to dig in one bowl. If the mouse dug in the baited bowl first and obtained the food reward, the mouse was allowed to eat the bait for 10 sec, the trial was terminated and recorded as a correct choice. If the mouse dug in the unbaited bowl first, the mouse was quickly removed from the testing area and placed on top of the baited bowl for 10 sec. Mice were not allowed to dig in the baited bowl or obtain the bait at this time, and placement on top of the baited bowl was solely done to indicate an error was made. These trials were terminated and recorded as an error. At the end of each trial, mice were returned to the home cage to signify the end of the trial. Inter-trial intervals were 30 sec followed by the immediate start of the next trial. To keep the odor concentration constant between trials, extra odor was mixed into the digging medias after every few trials. Groups were counterbalanced for direction of set-shift by having some mice switch from odor to digging media discriminations while others switched from digging media to odor discriminations. Furthermore, no more than two consecutive trials had the same combination of exemplars. Switching the irrelevant exemplars verified that each mouse learned which dimension was irrelevant (e.g., neither alpha-dri or pet-bedding consistently signified the presence of reward when odor was the relevant dimension). Placement of the baited bowl in the testing apparatus was randomized to prevent a side bias. Trials continued until mice reached a criterion of 6 consecutive correct trials. Once mice achieved criterion performance, they were returned to their home cage and fed to maintain 85% of their body weight. For each discrimination task, the total number of trials required to reach criterion performance and the number of errors committed were recorded. The time required to obtain the bait during a given trial (latency) was also recorded to assess activity rates and motivation to find the reward.
Order of Discrimination Tasks
Chronic Intermittent Ethanol Exposure
Ethanol was administered via the inhalation route using a well-established dependence model (Becker and Lopez, 2004
; Griffin et al., 2009a
; Griffin et al., 2009b
; Lopez and Becker, 2005
). Briefly, one group of mice (CIE group) received vapor exposure in inhalation chambers (16 hr/day for 4 consecutive days) while the remaining mice were similarly handled, but maintained in control (air) inhalation chambers. At the end of the 4th
day, mice were given three days of abstinence before beginning the next cycle of CIE exposure. This pattern of CIE (or air) vapor exposure was repeated for a total of three consecutive weekly cycles. Ethanol (95%) was volatilized by passing air through a submerged airstone. Ethanol vapor was mixed with fresh air and delivered to Plexiglas inhalation chambers at a rate of 5 L/min to maintain consistent ethanol concentrations (17 – 21 mg/L air) in the chamber. This yielded blood ethanol concentrations (BEC) in the range of 150 – 250 mg/dl. Prior to entry into the ethanol chambers, CIE mice were administered ethanol (1.6 g/kg; 8% w/v) and the alcohol dehydrogenase inhibitor pyrazole (1 mmol/kg) by intraperitoneal injection (20 ml/kg body weight) to maintain stable intoxication. Control mice received injections of saline and pyrazole before being placed in air chambers. The housing conditions were identical to those in the colony room. Chamber ethanol concentrations were monitored daily and air flow was adjusted to maintain concentrations within the specified range. Additionally, blood samples were collected from all animals to monitor BECs during the course of inhalation exposure. Chamber ethanol concentrations and BEC levels were determined as previously described (Griffin et al., 2009a
). All mice resumed discrimination training at 72 hr (Experiment 1 and 2) or 10 days (Experiment 2) following the final cycle of CIE (or air) exposure.
Experiment 1: Effects of CIE on reversal learning and set-shifting
To examine the effects of CIE exposure on reversal learning and set-shifting, mice were trained to complete a series of discrimination tasks. Each discrimination task differed on the perceptual dimension that signaled reward (i.e., odor vs. digging media) and on the specific type of odor and digging media used (e.g., garlic vs. cinnamon or gravel vs. confetti). The experimental design is depicted in . Ethanol-naïve mice, maintained at 85% ad libitum body weight, were first trained to perform 4 tasks including simple discrimination (SD), compound discrimination (CD), and two intradimensional shifts (ID1 and ID2). During the simple discrimination task, mice learned to associate the presence of a buried food reward (1/4th Honey Nut Cheerio Loop) in one of two cued bowls. For example, if odor was the relevant dimension, one bowl contained cloves/gravel while the other bowl contained smoked paprika/gravel. If digging media was the relevant dimension, one bowl contained cloves/gravel while the other bowl contained cloves/confetti. For compound discrimination, the same relevant dimension from simple discrimination was used, however, an irrelevant dimension (e.g., novel odor or digging media) was added making the bowls differ on both odor and digging medium (e.g., cloves/gravel vs. smoked paprika/confetti). The ability to shift attention within the same perceptual dimension was then assessed by changing the type of relevant cue used (i.e., cloves/smoked paprika changed to garlic/cinnamon). Each intradimensional shift was essentially performed in the same manner as a compound discrimination but had novel odors and digging medias. Following the second intradimensional shift (ID2), mice were divided into treatment groups (CIE or control) and matched according to their SD performance. During the treatment phase, mice were exposed to 3 consecutive weeks of CIE vapor (or air) inhalation. At 72 hr following the final CIE exposure, mice were re-tested for performance on an intradimentional shift (ID3 and ID4) in order to reorient mice to the task. The ability to reverse the association between odor and the presence of a food reward was measured during reversal learning (RevID4). The same odors and digging medias were used from the previous intradimensional shift but the food reward was buried in the previously unbaited bowl. Another intradimensional shift was performed (ID5) followed by an extradimensional shift (ED) that evaluated the ability to shift attention to another perceptual dimension in order to receive the food reward. Novel odors and digging medias were used for extradimensional shift but the food reward was paired with another perceptual dimension (e.g., switched from odor to digging media).
Experimental design for the reversal learning and set-shifting experiments
Experiment 2: Effects of CIE exposure and protracted abstinence on reversal learning
To examine the possibility that deficits in cognitive performance could reverse over a more extended period of abstinence, Experiment 1 was repeated but included three procedural changes. First, only reversal learning was tested since this task was the only one that showed a change in CIE treated animals. Specifically, reversal learning of a simple discrimination (rather than an intradimensional shift) was incorporated in this study design because we hypothesized that simplifying the task might reveal an even greater deficit in reversal learning for CIE exposed mice. Second, performance during reversal of a simple discrimination task was measured prior to and following the CIE exposure. This was conducted to verify reversal of a simple discrimination task could be established in mice prior to CIE exposure, and provides a measure for baseline performance for comparison to post-CIE exposure training. Third, mice were tested for the effects of CIE on reversal learning at two time-points during abstinence: 72 hours and 10 days. Briefly, mice were first trained to complete a SD and reversal learning discrimination task. Mice were then divided into treatment groups (CIE or control) and matched according to their SD performance. As in Experiment 1, mice were exposed to 3 consecutive weeks of CIE vapor (or air) inhalation during the treatment phase. To determine how long the effects of CIE exposure on reversal learning persist, each treatment group was further divided into two separate abstinent groups (between-subjects design). Half of the mice from each treatment group were tested at 72 hrs following the last CIE (or air) exposure. The remaining mice from both treatment groups were tested 10 days following the last CIE (or air) exposure. To allow for similar intoxication levels between each abstinent group (72 hrs vs. 10 days), group assignments were additionally matched according to BECs. The experimental design for this study is depicted in . Simple discrimination and reversal learning tasks were performed in the same manner during baseline (SD1 and RevSD1) and post-treatment testing (SD2 and RevSD2), except different (novel) odors and digging media were used ().
For Experiments 1 and 2, separate analyses of variance (ANOVA) were used to analyze the number of trials to reach criterion, the number of errors committed, and latency (seconds) to obtain the food reward for each experiment. For Experiment 1, separate two-way mixed factor ANOVA with Treatment as the between-subjects factor and Task as the repeated measure were used to analyze trials, errors and latency before and after ethanol exposure. As expected, there were no differences between CIE and control mice in performance prior to CIE treatment and thus, data were collapsed across treatment groups. For experiment 2, a three-way mixed factor ANOVA with Treatment and Time of testing as between-subject factors and Task as the repeated measure was used to analyze trials, errors and latency before and after ethanol exposure. Again, as expected, there were no group differences in performance prior to CIE treatment and this warranted collapsing pretreatment data across Treatment and Time groups. For all analyses, post-hoc comparisons were performed by isolating simple effects and by using Fisher’s Least Significant Difference (LSD) test (significance level set at p< 0.05).
To compare performance between Experiments 1 and 2, all three measures (trials, errors, latency), were normalized to SD performance by using the overall SD mean obtained from all mice [(score/mean of SD)*100]. The overall SD mean included scores from both CIE and control mice given that there were no initial group differences in SD performance (see Results section for details). Separate group means were used for Experiments 1 and 2. Scores were normalized to SD since this was the first discrimination task mice were exposed to and thus provided a baseline measure of discrimination learning. Expressing scores as percent of SD also allowed the magnitude of CIE effects to be easily compared between Experiments 1 and 2.