Methamphetamine (METH) abuse and dependence are serious public health problems worldwide with major medical, psychiatric, socioeconomic and legal consequences. However, a treatment for METH dependence has not been developed. The psychostimulative effects of METH are associated with the facilitation of the mesolimbic dopaminergic pathway through the exchange of the dopamine transporter and the activating vesicle monoamine transporter. Not only dopaminergic input, but also cholinergic input, modulates the activities of medium spiny neurons in the nucleus accumbens (NAc). Cholinergic cell ablation in the NAc enhances hyperlocomotion and places preference induced by cocaine (Hikida et al. 2001). Notably, systemic treatment with nicotine, a nicotinic acetylcholine receptor (nAChR) agonist, suppressed the reinstatement of METH‐seeking behavior in rats (Hiranita et al. 2006). Given these findings, the hyper cholinergic transmission and activation of nAChR may contribute to the suppression of METH dependence. Galantamine, a drug approved for the treatment of Alzheimer's disease, has a dual mechanism of action, serving as a potent allosteric potentiating ligand (APL) for nAChR and as a weak inhibitor of acetylcholinesterase (AChE) (Maelicke et al. 2001). However, whether galantamine has beneficial effects on METH‐induced addictive behavior, such as relapses in METH‐seeking behavior, remains uncertain. In the current study, we estimated the effect of galantamine on the reinstatement of METH‐seeking and food‐seeking behaviors induced by exposure to a cue‐light, which was associated with METH self‐administration and food consumption, as well as locomotor activity in mice.

We evaluated the effect of galantamine on the cue‐induced reinstatement of METH‐seeking behavior using a METH self‐administration paradigm. Mice were first subjected to food‐reinforced operant training under a fixed ratio (FR) 1 schedule. The mice were then taught to distinguish between active hole and inactive hole during a second food‐training session (Food2) (Fig. 1a, P < 0.01). After recovering from surgery for the implantation of a catheter, the mice were then subjected to METH self‐administration training. The mice were taught to distinguish between active hole and inactive hole during the last session of the FR1 schedule (L5/FR1) (Fig. 1a, P < 0.01). Under the FR2 schedule, active nose‐poking behavior entered a steady state and was increased during the last FR2 session (L3/FR2), compared with during the L5/FR1 session (Fig. 1a, P < 0.01). These results indicated that the mice had acquired METH self‐administration behavior. Once the METH self‐administration criterion was met, the mice were subjected to extinction training before the cue‐induced reinstatement test. With the extinction training, the number of active nose‐poke responses gradually decreased (<20 active nose‐poke responses) to a value less than that for the L3/FR2 session (Fig. 1a, P < 0.01).

Figure 1 Open in figure viewerPowerPoint Acquisition and extinction of methamphetamine (METH) self‐administration (a) and the effect of galantamine on cue‐induced reinstatement of METH‐seeking behavior (b) under the METH self‐administration paradigm. (a) Numbers of active and inactive nose‐pokes during food conditioning, the last five sessions (sessions L1‐L5/FR1) and the last three sessions (sessions L1‐L3/FR2) of METH self‐administration under schedules FR1 and FR2, and the last three sessions (L1‐L3/EXT) of the extinction training in C57BL/6J mice. Values indicate the mean ± standard error (SE) (n = 10). The results of a repeated‐measure analysis of variance (ANOVA) were as follows: session, F (12,216) = 11.41, P < 0.01; hole, F (1, 216) = 54.05, P < 0.01; interaction between session and hole, F (12, 216) = 8.18, P < 0.01. ##P < 0.01 compared with the number of inactive nose‐pokes during the same session. ¶¶P < 0.01 compared with the number of active nose‐pokes on the last day (L5) of the FR1 schedule. $$P < 0.01 compared with the number of active nose‐pokes on the last day (L3) of the FR2 schedule. (b) Nose‐poking responses during the last extinction session (L3) and the five sessions (SAL, GAL 0.1, 0.3, 1 mg/kg administration, and SAL re‐administration session) of the cue‐induced reinstatement of METH‐seeking behavior. Values indicate the mean ± SE (n = 10). The results of a repeated‐measure ANOVA were as follows: session, F (5,90) = 4.723, P < 0.01; hole, F (1, 90) = 15.838, P < 0.01; interaction between session and hole, F (5, 90) = 5.995, P < 0.01. #P < 0.05, ##P < 0.01 compared with the number of active nose‐pokes on the last day (L3) of extinction training. *P < 0.05 compared with the number of active nose‐pokes during the SAL administration session. ¶P < 0.05 compared with the number of active nose‐pokes during the GAL 1 administration session. EXT = extinction training, Food = food conditioning, FR1 = fixed ratio 1, FR2 = fixed ratio 2, GAL = galantamine, L = last, SAL = saline

Once the extinction criterion was achieved, cue‐induced reinstatement tests were performed 30 minutes after treatment with either saline or different doses of galantamine (0.1, 0.3 or 1 mg/kg, p.o.) for 3 hours. During the first saline treatment test (SAL), the mice exhibited a cue‐triggered reinstatement of the METH‐seeking behavior (Fig. 1b, P < 0.01). Low doses of galantamine (0.1 and 0.3 mg/kg) did not affect the increase in active nose‐poking responses induced by the METH‐associated cues (GAL 0.1, GAL 0.3), but a high dose of galantamine (1 mg/kg) significantly decreased the cue‐triggered reinstatement of METH‐seeking behavior (GAL 1) (Fig. 1b, P < 0.05). During the last saline treatment test (SAL), the mice once again showed the cue‐triggered reinstatement of METH‐seeking behavior (Fig. 1b, P < 0.05). The number of inactive nose‐poke responses did not change during the last extinction and the reinstatement test sessions (Fig. 1b).

To evaluate the effect of galantamine on the cue‐induced reinstatement of food‐seeking behavior, a food‐intake paradigm was used. The mice acquired food‐taking behavior under the FR1 and FR2 schedules and then showed an increase in active nose‐poking responses in the last session of the FR2 schedule (L3/FR2), compared with that in the last session of the FR1 schedule (L3/FR1) (Fig. 2a, P < 0.01). After the acquisition of food‐taking behavior, the mice were subjected to extinction training. With the extinction training, the number of active nose‐poke responses gradually decreased (<20 responses), and the number observed in the L3/FR2 session was less than that observed in the L3/EXT session (Fig. 2a, P < 0.01).

Figure 2 Open in figure viewerPowerPoint Acquisition and extinction of food self‐intake (a) and the effect of galantamine on the cue‐induced reinstatement of food‐seeking behavior (b) under the food self‐intake paradigm, and (c) the effect of galantamine on the locomotor activity in mice. (a) Number of active and inactive nose‐pokes during the last three sessions (sessions L1‐L3/FR1, FR2) of the food‐intake paradigm under schedules FR1 and FR2 and the last two sessions (L1‐L2/EXT) of the extinction training in C57BL/6J mice. Values indicate the mean ± standard error (SE) (n = 10). The results of a repeated‐measure analysis of variance (ANOVA) were as follows: session, F (7,126) = 40.82, P < 0.01; hole, F (1, 126) = 70.93, P < 0.01; interaction between session and hole, F (7, 126) = 38.35, P < 0.01. ¶¶P < 0.01 compared with the number of active nose‐pokes on the last day (L3) of the FR1 schedule. $$P < 0.01 compared with the number of active nose‐pokes on the last day (L3) of the FR2 schedule. (b) Number of nose‐poke responses during the last extinction session (L2) and two sessions (SAL, GAL 1 mg/kg administration) of the cue‐induced reinstatement of food‐seeking behavior. Values indicate the mean ± SE (n = 10). The results of a repeated‐measure ANOVA were as follows: session, F (3,54) = 9.738, P < 0.01; hole, F (1, 54) = 25.64, P < 0.01; interaction between session and hole, F (3, 54) = 12, P < 0.01. #P < 0.05, ##P < 0.01 compared with the number of active nose‐pokes on the last day (L2) of extinction training. (c) Locomotor activity after the administration of SAL or GAL 1 mg/kg for 3 hours. Values indicate the mean ± SE (n = 9). EXT = extinction training, FR1 = fixed ratio 1, FR2 = fixed ratio 2, GAL = galantamine, L = last, NS = not significant, SAL = saline

Once the extinction criterion was achieved, cue‐induced reinstatement tests were performed 30 minutes after treatment with either saline or the high dose of galantamine (1 mg/kg) that had suppressed METH‐seeking behavior. During the first saline treatment test (SAL), the mice exhibited the cue‐triggered reinstatement of food‐seeking behavior (Fig. 2b, P < 0.01). Unlike the METH‐seeking behavior, galantamine did not suppress the cue‐induced reinstatement of food‐seeking behavior (GAL 1) (Fig. 2b). During the last saline treatment test (SAL), the mice once again exhibited the cue‐triggered reinstatement of food‐seeking behavior (Fig. 2b, P < 0.05). The number of inactive nose‐poke responses did not change during the last extinction and the reinstatement test sessions (Fig. 2b). The activation of cholinergic transmission suppresses drug‐reward and drug‐seeking behaviors (Hikida et al. 2001; Hiranita et al. 2006; Takamatsu et al. 2006). However, the systemic injection of mecamylamine, a nAChR antagonist (Liu et al. 2007), and the microinjection of scopolamine, a muscarinic acetylcholine receptor antagonist, into the NAc (Pratt & Blackstone 2009) did not potentiate food‐seeking behavior. In agreement with these previous studies, our results showed that galantamine suppressed METH‐seeking, but not food‐seeking behavior.

Galantamine (3 mg/kg, i.p.) depresses locomotor activity and rearing behavior in a novel environment (Myhrer, Enger & Aas 2010). To exclude the possibility that the suppression of motor activity may be responsible for the attenuation of the cue‐induced METH‐seeking behavior, we also investigated the effect of galantamine on locomotor activity. Galantamine (1 mg/kg) did not suppress the locomotor activity of the mice (Fig. 2c). Furthermore, the effective dose of galantamine did not affect the nose‐poke responses during sessions of cue‐induced food‐seeking behavior (Fig. 2b). These results suggest that galantamine suppresses the reinstatement of METH‐seeking behavior without impairing motor function or natural cravings.

Donepezil, another AChE inhibitor, also suppresses the cue‐induced reinstatement of METH‐seeking behavior in rats through its effect on nAChR (Hiranita et al. 2006). Conversely, donepezil suppresses the place preference and hyperlocomotion induced by cocaine, but not METH (Takamatsu et al. 2006). In addition, donepezil failed to change the cocaine‐use behavior of abusers (Winhusen et al. 2005). Unlike donepezil, galantamine acts on nAChR as an APL, suggesting that it strongly facilitates cholinergic function through nAChR. Nicotine attenuates the METH‐priming‐induced and cue‐induced reinstatement of METH‐seeking behavior (Hiranita et al. 2006), indicating that nAChR activation by APL may play a role in the suppression of METH‐seeking behavior. Furthermore, galantamine treatment has been shown to ameliorate the impairment of recognition memory in METH‐treated mice through its effect on nAChR (Noda et al. 2010). Thus, of all the known AChE inhibitors, galantamine might be the most effective for the alleviation of METH cravings.

A major clinical problem in treating drug dependence is the high rate of relapse even after long periods of abstinence, and drug‐associated environmental stimuli are important determinants of drug‐seeking behavior that resemble relapsing behavior in humans (Kalivas & Volkow 2005). Therefore, recent data suggest that galantamine may be a useful therapeutic agent for METH dependence, such as the treatment of relapses of METH‐seeking behavior.