Significant outcomes
• We established a chronic oral regimen of carbamazepine (CBZ) that has behavioural mood stabilising-like effects in mice.
• We demonstrate that chronic oral administration of CBZ results in dose-relevant blood levels, anti-manic-like effects in the amphetamine-induced hyperactivity test and antidepressant-like effects in the forced swim test (FST) and the tail suspension test (TST).
Limitations
• The results are based on very few behavioural models, one for mania-like and two for depression-like behaviour. Additional models and tests are recommended.
• Detected blood levels are lower than clinically relevant levels in humans; however, this might be related to differences in kinetics of the drug between mice and humans.
Introduction
Despite the high prevalence of bipolar disorder (BPD) reaching lifetime prevalence of 1–3% in the adult general population, remarkably little is known regarding its pathophysiology or the means by which effective medications exert their therapeutic effects (Reference Belmaker1). Moreover, the classical mood-stabilising drugs including lithium salts and the anticonvulsants, CBZ and valproic acid, were discovered by serendipity and their therapeutic mechanisms of action are not yet understood (Reference Belmaker1).
One important approach to gain better understanding into the underlying pathophysiology of BPD and the mechanism of its treatment is to explore the biological targets of known medications utilising animal models (Reference Gould, Quiroz, Singh, Zarate and Manji2). A number of animal models and tests are used to study BPD and the effects of known and novel treatments (Reference Einat3); however, one limitation in this respect is the lack of an animal model that includes both the depression and the manic poles of the disorder and the oscillation between them (Reference Einat4). Therefore, the common practice is to use separate models for depression and for mania including the FST, the TST and the amphetamine-induced hyperactivity test that have been intensely utilised and were demonstrated to be useful in identifying possible mood-stabilising effects (Reference Cox, Harrison-Read, Steinberg and Tomkiewicz5–Reference Einat, Yuan, Szabo, Dogra and Manji7).
An attempt to understand the therapeutic mechanism of action of a drug requires that not only will the drug have an effect in a relevant animal model, but also that it will be administered in a way that is similar to its use in the clinic (Reference Gould, Quiroz, Singh, Zarate and Manji2,Reference Gould and Einat8). Hence, for mood stabilisers, it is important to establish a drug regimen in animals that will have behavioural effects following chronic oral administration. Such a regimen was established for the prototypic mood stabiliser lithium in mice (Reference O'Brien, Harper and Jove9). With this protocol, lithium is given in food starting with a 2% concentration for 5 days and followed by a 4% concentration for 10 days. This protocol, with some small variations in different laboratories, was demonstrated to result in clinically relevant lithium blood levels as well as therapeutic-like effects in behavioural models (Reference O'Brien, Harper and Jove9–Reference Kovacsics and Gould11). Yet, such protocols were not clearly established for other mood stabilisers.
The effect of the anticonvulsant mood stabiliser CBZ in animal models related to BPD was rarely examined, and past results are not highly consistent. For example, in the amphetamine-induced hyperactivity test, acute administration of CBZ reduced hyperactivity in mice in one study (Reference Arban, Maraia and Brackenborough12), but similar doses had no effect in mice in another one (Reference Kalinichev and Dawson13), and chronic CBZ regimen in rats had no effect as well (Reference Elphick14). In the FST, chronic administration of CBZ to rats resulted in a decrease of immobility time in one study (Reference Barros and Leite15), but not in another (Reference Kitamura, Akiyama and Kitagawa16), although both studies used similar doses, regimens and rats. We did not identify studies that tested acute or chronic CBZ administration in the TST.
In light of the scarcity of studies and contradicting results regarding the effect of chronic CBZ in behavioural tests relevant to BPD, the aim of the present study was to establish a chronic oral administration regimen of CBZ in mice that will elicit mood stabilising-like activity in models of depression and mania.
Methods
Animals
Eight- to ten-week-old male ICR mice (Experiments 1 and 2) or C57bl/6 (Experiment 3) were group housed in standard mice cages in an animal colony room with a 12/12 h light/dark cycle, with constant temperature at 23 ± 1°C and ad libitum food and water. All experimental procedures followed the Israeli guidelines for the treatment and care of experimental animals and were approved by the Ben-Gurion University of the Negev IACUC (protocol #IL-56-08-2012).
Drugs
CBZ (Sigma, St Louis, MO, USA) was added to powdered food and thoroughly mixed using a Kitchen-Aid® mixer for at least an hour. For Experiment 1, CBZ was administered for 3 weeks at three dose regimens: (1) first week at 0.25% (w/w) and last 2 weeks at 0.5% (w/w); (2) 3 weeks at 0.5% w/w; and (3) first week at 0.5% (w/w) and last 2 weeks at 0.75% (w/w). The control group received regular powdered food. These doses were selected on the basis of previous work in different contexts and on a preliminary work in our lab (Reference Shaldubina, Einat, Szechtman, Shimon and Belmaker17). The 3-week schedule was selected based on clinical time span expected for the drug to elicit therapeutic response (Reference Belmaker1), animal results with similar mood stabilisers (Reference Chen, Cai, Cao, Zhang and Li18) and preliminary results from our lab. Mice were weighed every other day during the treatment period.
Experiment 1 presented us with two problems. (1) Mice did not float in the FST (see procedure below), and when not active they sank under water and were immediately rescued by the experimenter. (2) Mice treated with the highest CBZ dose [0.75% (w/w), namely, 7.5 g/kg food] showed an average weight loss of 6.5% after 3 days of treatment. To overcome these problems, in Experiment 2 we replaced the FST with the TST and used CBZ doses not higher than 0.5% (w/w). Furthermore, we added Experiment 3 to evaluate the effects of CBZ in the FST in a different strain of mice (C57bl/6). In these mice, there was no weight reduction in the 0.75% CBZ dose and mice did not sink in the FST. We therefore tested the C57bl/6 mice in the FST with the original dose range.
Behavioural tests
Spontaneous activity test
Mice were placed individually in the centre of a transparent Plexiglas box (37.5 × 37.5 × 45 cm) for 30 min and their behaviour was digitally recorded. Distance moved was analysed using an automated acquisition and analysis software (Ethovision, Noldus Inc., Wageningen, the Netherlands).
FST
Mice were placed for a 6-min session in a glass cylinder (22 cm diameter and 30 cm high) filled with water at a temperature of 22 ± 1°C such that the mouse cannot touch the bottom or climb out of the cylinder. Sessions were digitally recorded, and the duration of immobility during the last 4 min of the session was scored by an experimenter blind to the treatment groups. Immobility was defined as the time spent by a mouse floating or making only those movements necessary to keep its head above the water.
TST
Mice were suspended by the tail from a hook placed 30 cm above a padded table, using an adhesive tape for a 6-min session. Sessions were digitally recorded, and the duration of immobility during the last 4 min of the session was manually scored by a blind observer. Immobility was defined as the time spent by a mouse hanging passively and completely motionless.
Amphetamine-induced hyperactivity
Mice were injected with amphetamine (1.5 mg/kg dissolved in saline to a 10 ml/kg injection volume) or saline (control) and placed in the centre of a transparent Plexiglas box (37.5 × 37.5 × 45 cm) for a 30-min session. Their behaviour was recorded and analysed as described above for the spontaneous activity test.
Blood levels
At the end of Experiment 1, five mice from each group were euthanised using 20% isoflorane, and blood was withdrawn using a heart puncture. CBZ blood levels were measured by a fluorescence polarization immunoassay kit (Roche Diagnostics, Mannheim, Germany), according to the manufacturer's instructions, run on a Cobas 400 apparatus (Roche Diagnostics, Rotkreuz, Switzerland).
Statistics
When more than one dose of CBZ was tested, the data were analysed using a one-way analysis of variance (ANOVA). When ANOVA showed a significant effect, it was followed by post-hoc LSD tests. When only one dose of CBZ was tested and compared with control, data were analysed using a Student's t-test. The amphetamine-induced hyperactivity test was analysed using a two-way ANOVA with amphetamine and CBZ as main factors. The ANOVA analysis was followed by post-hoc LSD tests.
Results
General observations of mice under the experimental treatment regimen did not indicate any gross sickness behaviour, changes in home cage behaviour or in fur conditions.
Experiment 1
CBZ at the 0.25% or 0.5% doses had no effects on spontaneous activity, not for the total activity in 30 min and not for analysis across the session in 10-min time bins [distance in cm in a 30-min session – regular food: 9134 ± 582; 0.25% CBZ: 7615 ± 766; 0.5% CBZ: 9807 ± 913; repeated measures ANOVA across three 10-min time bins: CBZ effect: F(3, 16) = 1.22, p = 0.33, NS; time effect: F(3, 32) = 59.2, p < 0.001; CBZ × time interaction: F(6, 32) = 0.61, p = 0.72, NS]. Data from additional tests were not available for Experiment 1 because of the issues that arose from the FST (see the Methods section for details).
Experiment 2
Chronic oral administration of 0.5% CBZ resulted in a 27%, statistically significant reduction in immobility time in the TST [Fig. 1a; t(33) = 2.31, p = 0.027] and a 28%, statistically significant reduction in amphetamine-induced hyperactivity [Fig. 1b; two-way ANOVA with amphetamine and CBZ as main factors: amphetamine effect – F(1, 36) = 16.8, p < 0.001; CBZ effect – F(1, 36) = 1.83, p = 0.18; amphetamine × CBZ interaction – F(1, 36) = 4.9, p = 0.033; post hoc: control food/amphetamine group different from all other groups]. Further analysis of the amphetamine groups across time (in 10-min time bins) shows that the effects of CBZ does not change across the session [repeated measures ANOVA, CBZ × time interaction: F(2, 36) = 0.47, p = 0.63, NS].
Chronic oral administration of 0.5% carbamazepine (CBZ) in food resulted in a significant reduction in tail suspension test immobility time (a) and in amphetamine-induced hyperactivity (b) in ICR mice. Results are presented as means ± SEM, Cont = control group; amph = amphetamine. *p < 0.05 compared with Cont (a) and compared with all other groups (b).
Experiment 3
Treatment of C57bl/6 mice with 0.75% but not lower CBZ doses resulted in a 26%, statistically significant reduction in FST immobility [Fig. 2; ANOVA F(2, 13) = 5.46, p = 0.02; post-hoc LSD: CBZ 0.75% different from other groups, p < 0.05].
Chronic oral administration of 0.75% carbamazepine (CBZ) in food resulted in a significant reduction in forced swim test immobility time in C57bl/6 mice. Results are presented as means ± SEM. *p < 0.05 compared with other groups.
Blood levels
Oral administration of all doses of CBZ resulted in detectable blood levels of the drug that were significantly higher compared with control [Fig. 3; ANOVA: F(3, 16) = 18.03, p < 0.0001, post-hoc LSD, all CBZ groups different from control] and with a significant dose/concentration correlation (r = 0.86, p < 0.0001).
Chronic oral administration of carbamazepine (CBZ) resulted in dose-relevant blood levels. *Significant difference from control; #significant difference from all other groups.
Discussion
The objective of this work was to establish a chronic oral administration regimen of CBZ that elicits mood stabilising-like effect in mice models of depression and mania. The objective was achieved as chronic oral administration of 0.5% (w/w) CBZ in food to ICR mice resulted in a mood stabilising-like effect, including an anti-manic-like effect in the amphetamine-induced hyperactivity test and in an antidepressant-like effect in the TST.
Our initial plan was to study mice in the FST; however, unfortunately, and for an unknown reason, ICR mice treated chronically with CBZ did not float well and showed a tendency to sink when not active. We therefore switched from the FST to the TST. The limited floating ability of ICR mice treated with CBZ cannot be attributed to generalised motor impairment, as these mice were not different from treated control mice in the spontaneous activity, and showed reduced immobility (increased activity) compared with controls in the TST, which is also a motor-dependent test. The floating impairment of these mice in the FST is therefore an open question at this time. CBZ administration to rats was not reported to induce similar problems (Reference Kitamura, Akiyama and Kitagawa16), and acute CBZ was not reported to affect FST behaviour in Swiss mice (Reference Szymczyk and Zebrowska-Lupina19). We did not find any previous reports on the effects of CBZ on swimming in ICR (CD-1) mice. We therefore added an FST experiment with C57bl/6 mice (Experiment 3) where mice did not sink, and the results also show that chronic oral administration of CBZ results in an antidepressant-like effect in the FST (Fig. 2). There is a difference in the effective concentration of CBZ between ICR and C57bl/6 mice; however, such differences in dosing between diverse strains are not uncommon.
The regimen established in the present study has three advantages: (1) it is closely related to the human administration paradigm being oral and chronic; (2) it is technically simple as the drug is administered in food; and (3) it is significantly less stressful to animals, compared with chronic administration via injections or via oral gavage. It is important to note that our experiments attempted to mimic the prophylactic effects of CBZ, as chronic treatment preceded the induction of manic-like and depression-like behaviours. This is important because the effect of CBZ in patients is well established for acute mania but not so much for acute bipolar depression. Yet, the drug's prophylactic effect is established for both poles of the disorder (Reference Post, Ketter, Uhde and Ballenger20,Reference Fountoulakis, Grunze, Panagiotidis and Kaprinis21).
Interestingly, blood level of CBZ after treatment with the dose that was effective in the behavioural experiments (0.5%) was lower than the human therapeutic range. Whereas the therapeutic range for patients is 4–12 μg/ml, the mean blood level in our 0.5% group was only 2.3 μg/ml and only the 0.75% group was close to the human therapeutic levels, with a 3.7 μg/ml concentration. The relatively low blood levels in the 0.5% group may cast some doubt on the results, but these results are not unique. For example, in a study related to the anticonvulsant effects of CBZ in rats, chronic administration of 30 mg/kg/day of the drug resulted in blood levels between 2 and 3 μg/ml, but in a significant behavioural effect reducing seizure frequency in over 50% (Reference Ali, Dua, Constance, Franklin and Dudek22). Therefore, it is possible that the steady state concentrations of CBZ that are sufficient to induce behavioural changes in rodent models are lower than the doses used in the clinic.
A large body of work demonstrates the effects of lithium in a variety of doses and regimens in the FST and the amphetamine-induced hyperactivity test. Most importantly, some of these studies are based on chronic oral administration, which is the most relevant to the clinical realm (Reference O'Brien, Harper and Jove9,Reference Bersudsky, Shaldubina and Belmaker10). We now present a comparable method of clinically relevant administration of CBZ to mice and suggest that it will now be possible to use this method to gain better understanding of the therapeutically relevant biological effects of CBZ.
Acknowledgements
The study was partially supported by a US–Israel Binational Science Foundation (BSF) grant to HE (grant #2011313).
Authors contributions
Ms. Kara coordinated the experiments, took part in the execution of the experiments and led the analysis and the writing of the initial draft of the manuscript. Ms. Toker and Ms. Karpel took part of the initial planning and the execution of the experiments, and were involved in the analysis of the data and the writing of the initial draft. Profs Agam and Belmaker were responsible for the initial planning of the experiments and their execution and were involved in data analysis and in writing the manuscript. Prof. Einat was involved in the execution of the experiments and was in charge of data analysis and of the final writing, editing and submission of the manuscript.
Statement of Interest
None.
Ethical Standards
The authors assert that all procedures contributing to this work comply with the ethical standards of the relevant national and institutional guides on the care and use of laboratory animals. All experimental procedures followed the Israeli guidelines for treatment and care of experimental animals and were approved by the Ben-Gurion University of the Negev IACUC (protocol # IL-56-08-2012).
