Reversal of Behavioral Sensitization by Environmental Enrichment.

Adult C57BL6 mice were injected six times with cocaine (15 mg/kg) and developed behavioral sensitization (two-way ANOVA for repeated measures: Time effect, F_5,215 = 24.97, P < 0.0001) (Fig. 2A). After the last injection of cocaine, mice were randomly divided into two groups: one group was kept in standard housing conditions, whereas the other group was housed in an enriched environment; both groups remained in the animal facility until they were tested for expression of behavioral sensitization (Fig. 1A). At 1, 7, or 30 days after the last cocaine injection, the reactivity to cocaine (10 mg/kg) was measured in independent groups of mice. Behavioral sensitization to cocaine was long lasting (up to 30 days) in mice housed in the standard environment, as demonstrated by an enhanced behavioral response (approximately 200%) to cocaine (10 mg/kg) compared with cocaine-naïve control mice (Fig. 2 B and C) (Mann–Whitney test, P = 0.0003). In contrast, mice switched to enriched environments were sensitized to cocaine after 1 day of enrichment, but after 30 days their response to cocaine was reduced (approximately 65%) compared with sensitized mice housed in standard environments (Fig. 2B) (two-way ANOVA: Time effect, F_2,61 = 9.94, P < 0.0001; Time × Environment interaction, F_2,61 = 6.77, P < 0.01). Importantly, after 30 days of environmental enrichment, the reactivity to cocaine of sensitized mice was similar to that of the nonsensitized mice that were housed in standard or enriched environments (Fig. 2C).

To rule out the possibility that change in environments per se could be responsible for the reversal of behavioral sensitization, two groups of mice were sensitized to cocaine as described above and, after development of sensitization, were housed in either standard or isolated environments for 30 days. When challenged with cocaine (10 mg/kg), mice switched to isolated environments showed significantly higher locomotor activity than mice kept in standard environments (standard environments = 303 ± 39, isolated environments = 441 ± 55; Mann–Whitney test, P = 0.01).

Elimination of Established Conditioned Place Preferences by Environmental Enrichment.

Addiction is believed to result not only from a higher sensitivity to the effects of cocaine but also from a robust control of drug seeking by stimuli and contexts associated with the drug's effects (20–22). Therefore, we used a conditioned place preference paradigm (17) to assess whether the appetitive effects of cocaine and the incentive values acquired by contexts associated with cocaine would also be eliminated by environmental enrichment (Fig. 1B). Mice were conditioned to cocaine (10 mg/kg) and, at the end of the last session of conditioning, were randomly divided into standard and enriched housing conditions. The expression of conditioned place preferences was measured in independent groups of mice 1, 7, or 30 days after the last conditioning session. Mice housed in standard environments showed significant preferences for the compartment previously associated with cocaine after 1, 7, or 30 days (Fig. 3A) indicating that, as previously described (23, 24), the ability of contextual stimuli to elicit drug-seeking behavior is not lost spontaneously (without repeatedly exposing animals to the conditioned place preference apparatus in the absence of the drug) and persists for long periods. In contrast, the preferences for cocaine-paired compartments were completely abolished after 30 days of environmental enrichment (two-way ANOVA: Environment effect, F_1,66 = 6.53, P < 0.05) (Fig. 3A).

Although environmental enrichment is generally believed to increase learning and memory (6), we wanted to exclude the possibility that loss of conditioned place preference in mice housed in enriched environments might be due to memory impairment or to a general reduction in motivation. Thus, we conditioned mice to the aversive effects of lithium (3 mEq/kg), separated them randomly into standard or enriched housing conditions, and measured the expression of conditioned place aversion 1, 7, or 30 days after the last conditioning session. As shown in Fig. 3B, lithium induced long-lasting place aversions in mice housed in both standard and enriched conditions, indicating that memory of aversive motivational stimuli is preserved in mice housed in enriched environments.

Prevention of Cocaine-Induced Reinstatement by Environmental Enrichment.

One of the hallmarks of addiction and one of its most troubling aspects is relapse. Relapse can be studied in laboratory animals by reinstatement procedures that can be applied both to drug self-administration and conditioned place preference (18). In the latter case, conditioned place preferences can be eliminated by repeatedly exposing mice to the conditioning apparatus in the absence of drug injections and can then be reinstated by injections of the conditioning drug (24, 25). In our experiments (Fig. 1C) mice were conditioned to cocaine (10 mg/kg) and then tested for conditioned place preference. After this test, mice that showed a preference for the cocaine-paired compartment were assigned to either the enriched or standard housing condition, ensuring that similar levels of preferences were present in the two groups (Fig. 4A). Starting the next day, mice were exposed each day for 10 days to the place preference apparatus and tested for extinction. As previously described (24, 25), mice housed in standard environments lost their preference for the compartment previously paired with cocaine (Fig. 4A) (two-way ANOVA for repeated measures: Time effect, F_9,279 = 5.96, P < 0.0001). In addition, environmental enrichment did not significantly alter the rate of extinction (Fig. 4A). Control groups of mice that received only saline during conditioning showed no preference on the test day, and their preference score remained stable during the 10 days of extinction (Fig. 4A). After extinction, mice were injected with cocaine, and reinstatement of conditioned place preferences was measured. Although cocaine (10 mg/kg) readily reinstated extinguished conditioned place preferences in mice housed in standard environments after conditioning, mice housed in enriched environments after conditioning did not show any cocaine-induced reinstatement of preference for the cocaine-paired compartment (Fig. 4B) (three-way ANOVA: Environment effect, F_1,56 = 4.08, P < 0.05; Challenge effect, F_1,56 = 4.36, P < 0.05; Environment × Conditioning interaction, F_1,56 = 5.82, P < 0.05; Environment × Challenge interaction, F_1,56 = 4.07, P < 0.05; Conditioning × Challenge interaction, F_1,56 = 5.57, < 0.05; Environment × Conditioning × Challenge interaction, F_1,56 = 6.16, < 0.05).

Environmental Enrichment Decreases Activation of Several Brain Areas Involved in Cocaine-Induced Reinstatement.

Because relapse represents one of the most important and intriguing aspects of drug addiction, we investigated the mechanisms underlying the ability of environmental enrichment to prevent cocaine-induced reinstatement. Thus, in brain tissues obtained from animals used in the reinstatement experiments (Fig. 4B), we evaluated the degree of c-FOS expression (19) in several brain areas involved in drug addiction and cocaine-induced reinstatement [see supporting information (SI)] (26–28). In cocaine-naïve mice, injections of cocaine (10 mg/kg i.p.) induced significant increases in c-FOS expression only in the orbitofrontal cortex and did so to a similar extent in mice housed in standard or enriched environments (see SI). This indicates that, as previously reported (29), this dose of cocaine is low and may be insufficient to induce c-FOS expression in other areas, such as the striatum of cocaine-naïve mice. On the other hand, in mice conditioned to cocaine, challenge injections of cocaine induced similar significant increases in c-FOS expression in mice housed in standard or enriched environments in the dorsal caudate putamen, the central amygdala, the ventral pallidum, the prelimbic cortex, the anterior cingulate cortex, and the orbitofrontal cortex. However, in mice conditioned to cocaine and then housed in enriched environments, specific reductions in cocaine-induced c-FOS expression were found in the infralimbic cortex, the shell and core of the nucleus accumbens, the basolateral amygdala, and the ventral tegmental area (Fig. 5), suggesting that these regions may be involved in the protective effects of environmental enrichment on cocaine-induced reinstatement. Interestingly, given the known effects of environmental enrichment on the hippocampus (6), mice housed in enriched environments showed higher levels of c-FOS expression in the dentate gyrus than mice housed in standard environments regardless of the treatment, with cocaine inducing similar increases in c-FOS expression when basal levels are taken into account (Fig. 5). However, the activation of the dentate gyrus does not seem to be specific to the reinstatement condition, suggesting that this region does not play a direct role in the protective effects of environmental enrichment on cocaine-induced reinstatement. Analysis of locomotor activity in the conditioned place preference apparatus during reinstatement tests (see SI) demonstrated that cocaine-induced increases in locomotor activity did not differ significantly between mice housed in enriched and standard environments, ruling out the possibility that differences in motor behaviors could account for the differences in c-FOS expression.

Subjects.

Adult male C57/BL6 mice (Elevage Janvier), 10–12 weeks old, experimentally naïve at the start of the study, were housed in a temperature- and humidity-controlled room and maintained on a 12 h light/dark cycle with the lights on from 7:00 a.m. to 7:00 p.m. and had ad libitum access to food and water. All experimentation was conducted during the light period. Experiments were carried out in accordance with the European Communities Council Directive of November 24, 1986 (86/609/EEC) for the care of laboratory animals.

Housing Environmental Conditions.

After arrival and during the development of behavioral sensitization or conditioned place preference (see following sections), all mice were housed in standard environments (i.e., were housed in groups of four in common housing cages, 25 × 20 × 15 cm). After being subjected to experimental manipulations, half of the mice were kept in standard environments, whereas the other half was switched to enriched environments, which consisted of larger (60 × 38 × 20 cm) cages containing constantly a running wheel and a small house and four to five toys that were changed once per week with new toys of different shapes and colors. To avoid stress, all mice in each cage were assigned to either standard environment or enriched environment, and littermates were not changed during the course of the study.

Behavioral Procedures.

Behavioral sensitization and conditioned place preference or conditioned place aversion were conducted as previously described (13). At the end of the last sensitization or conditioning session, half the mice were kept in standard environments, and the other half were switched to enriched environments before being tested for expression of addiction-related behaviors (for detailed description, see SI).

For the reinstatement procedure, mice were assigned to the enriched environment or standard environment condition after being tested for expression of conditioned place preference. The next day and for 10 subsequent days, conditioned place preferences were extinguished by running 30-min test sessions similar to the initial test session. On the 11th day, mice were injected with either saline or 10 mg/kg of cocaine, and reinstatement of extinguished conditioned place preference was studied.

Immunohistochemistry.

Fifty-two mice from experiment 3 (n = 6–8 per group) were used for immunohistochemistry analysis. General procedures for c-FOS immunostaining and counting were based on previous publications (33). Twelve regions were quantified for c-FOS expression: (i) prelimbic cortex cortex, (ii) infralimbic cortex, (iii) orbitofrontal cortex, (iv) anterior cingulate cortex, (v) the dorsolateral caudate putamen, (vi) the shell of the nucleus accumbens, (vii) the core of the nucleus accumbens, (viii) the ventral pallidum, (ix) the dentate gyrus of the hippocampus, (x) the central amygdala, (xi) the basolateral amygdala, and (xii) the ventral tegmental area.

Drugs.

Cocaine HCl was obtained from the Research Triangle Institute. Lithium chloride was purchased from Sigma-Aldrich (France). Drugs were dissolved in sterile saline solutions (0.9%) and administered i.p. in a volume of 1 ml/100 g.

Statistics.

All results are presented as group means (± SEM). Differences in locomotor activity or preference scores between standard environment and enriched environment groups were assessed by two-way or three-way ANOVA. Results showing significant overall changes were subjected to Student–Newman–Keuls post hoc test. For comparison between two sets of values, a Mann–Whitney nonparametric test was used. Differences were considered significant when P < 0.05.