JNJ-39220675, a novel selective histamine H3 receptor antagonist, reduces the abuse-related effects of alcohol in rats
- First Online:
- Cite this article as:
- Galici, R., Rezvani, A.H., Aluisio, L. et al. Psychopharmacology (2011) 214: 829. doi:10.1007/s00213-010-2092-4
- 235 Views
A few recent studies suggest that brain histamine levels and signaling via H3 receptors play an important role in modulation of alcohol stimulation and reward in rodents.
The present study characterized the effects of a novel, selective, and brain penetrant H3 receptor antagonist (JNJ-39220675) on the reinforcing effects of alcohol in rats.
The effect of JNJ-39220675 on alcohol intake and alcohol relapse-like behavior was evaluated in selectively bred alcohol-preferring (P) rats using the standard two-bottle choice method. The compound was also tested on operant alcohol self administration in non-dependent rats and on alcohol-induced ataxia using the rotarod apparatus. In addition, alcohol-induced dopamine release in the nucleus accumbens was tested in freely moving rats.
Subcutaneous administration of the selective H3 receptor antagonist dose-dependently reduced both alcohol intake and preference in alcohol-preferring rats. JNJ-39220675 also reduced alcohol preference in the same strain of rats following a 3-day alcohol deprivation. The compound significantly and dose-dependently reduced alcohol self-administration without changing saccharin self-administration in alcohol non-dependent rats. Furthermore, the compound did not change the ataxic effects of alcohol, alcohol elimination rate, nor alcohol-induced dopamine release in nucleus accumbens.
These results indicate that blockade of H3 receptor should be considered as a new attractive mechanism for the treatment of alcoholism.
KeywordsHistamineH3AlcoholismSelf administrationMicrodialysisRewardSelectively bred alcohol-preferring ratsDopamine
Alcoholism is a major health problem worldwide. Adequate treatments to help treat alcoholism are needed. Although significant progress in the pharmacotherapy of alcoholism has been achieved in the past decades, side effects and marginal efficacy of current medications limit their utility. New approaches for more effective treatment of alcoholism are necessary.
There are currently only three medications approved by the U.S. Food and Drug Administration for the treatment of alcohol abuse, alcohol dependence, or alcoholism: the aldehyde dehydrogenase blocker disulfiram, the opioid antagonist naltrexone, and acamprosate, a drug with a complex pharmacology (i.e., it blocks NMDA receptors and potentiates GABA-A function) (Garbutt 2009; Johnson 2008). Several other strategies are being evaluated in the clinic or in various preclinical models (Heilig and Egli 2006). However, modulation of the histamine system is not commonly listed in the emerging strategies for the treatment of alcoholism despite the fact that several studies have suggested a role for brain histamine in drug reward and addiction (Huston et al. 1997; Wagner et al. 1993). Emerging preclinical data suggest that H3 receptor antagonists might be useful for the treatment of alcoholism. For example, high alcohol preference and low alcohol sensitivity correlate with brain histamine and H3 receptor-mediated processes (Lintunen et al. 2001, 2002; Nuutinen et al. 2009, 2010). Furthermore, thioperamide and clobenpropit, two relatively selective imidazole-based H3 receptor antagonists, reduced operant responding for alcohol in alcohol-preferring rats (Lintunen et al. 2001). In contrast, R-α-methylhistamine, an H3 receptor agonist, increased operant responding for alcohol in alcohol-preferring rats (Lintunen et al. 2001). Histidine decarboxylase knockout mice showed stronger alcohol-induced conditioned place preference than wild-type mice (Nuutinen et al. 2010). Preliminary data indicate that mice lacking the H3 receptor showed lower alcohol preference ratios in a two-bottle choice compared to control mice (Panula et al. 2008). H3 receptors are abundant in cortico-mesolimbic areas of the brain in rodents and humans (Anichtchik et al. 2001; Pillot et al. 2002; Pollard et al. 1993). These are the same neuroanatomical areas on which alcohol and other drugs of abuse exert their reinforcing and subjective effects (George et al. 1998). In addition, H3 receptors are autoreceptors localized in non-histaminergic neurons throughout the brain, which modulate the activity and release of a variety of neurotransmitters. This heteromeric nature of H3 receptors allows for targeting multiple unbalanced systems simultaneously. This profile is particularly important given the complex and rich pharmacology of alcohol. For example, an increased acetylcholine and/or norepinephrine neurotransmission induced by H3 receptor antagonists may improve cognitive function (i.e., decision making and impulsivity) (Esbenshade et al. 2009; Galici et al. 2009; Witkin and Nelson 2004); whereas enhancing GABA, dopamine, and serotonin neurotransmission may prevent withdrawal signs and decrease the reinforcing and subjective effects of alcohol in dependent subjects.
Materials and methods
Male adult selectively bred alcohol-preferring (P) rats were used for the alcohol intake and alcohol relapse-like behavior experiments (Richard et al. 2006). The animals used in the study were male adult rats selected from the colony of P rats maintained at the Skipper Bowles Center for Alcohol Studies at the University of North Carolina at Chapel Hill, NC. For the pharmacokinetic and ex vivo receptor autoradiography studies, male Sprague–Dawley rats (Charles River), weighing 300–320 g, were used. For the alcohol elimination, operant self-administration, microdialysis, and rotarod experiments, male Wistar rats (Charles River), weighing 200–320 g, were used. P rats were housed under a reversed 12 h light–dark cycle (lights on at 18:00). Wistar rats were housed on a normal 12 h light–dark cycle (lights on at 06:00) with food and water provided ad libitum.
All methods complied with The Guide for the Care and Use of Laboratory Animals manual and were approved by the IACUC.
JNJ-39220675 (4-cyclobutyl-[1,4]diazepan-1-yl)-[6-(4-fluoro-phenoxy)-pyridin-3-yl]-methanone) was synthesized in house according to the method of Letavic and colleagues (Letavic et al. 2010), formulated in 5% DMSO and 20% Cremophor (RH-40) (operant self administration and rotarod studies), saline (pharmacokinetic and receptor occupancy studies), 20% (w/v) cyclodextrin (P rat study, alcohol elimination, and rotarod studies) and administered at 1 ml/kg to rats. For microdialysis studies, JNJ-39220675 was formulated in phosphate-buffered saline and administered at 2 ml/kg.
Naltrexone was purchased from Sigma-Aldrich (St Louis, MO, USA), dissolved in saline and administered at 1 ml/kg. Alcohol was purchased from EMD Chemicals Inc Gibbstown, NJ. The 10% (v/v) solutions of alcohol were prepared weekly by diluting a solution of 100% alcohol with tap water.
Pharmacokinetic and ex vivo H3 receptor autoradiography
Subcutaneous pharmacokinetic and ex vivo H3 receptor autoradiography studies were done as previously described (Barbier et al. 2007). The subcutaneous pharmacokinetic study was done after administration of a 3-mg/kg dose (n = 3). Plasma concentrations of JNJ-39220675 were determined using liquid chromatography tandem mass spectrometry assays. The lower limit of quantification was 5 ng/ml.
Ex vivo H3 receptor occupancy after subcutaneous administration of JNJ-39220675 was determined at various time points (n = 3 per time point) in rat striatum using [3H]α methylhistamine as tracer (Barbier et al. 2007).
An additional pharmacokinetic study was performed to compare the various vehicles used in the different models relevant to alcoholism. JNJ-39220675 was formulated in saline, 5% DMSO and 20% Cremophor (RH-40), 20% (w/v) cyclodextrin or phosphate-buffered saline and administered subcutaneously at a 3 mg/kg dose. Plasma concentrations of JNJ-39220675 were collected at the 1-h time point. Data were analyzed using a one-way ANOVA for vehicle followed by NK post hoc tests.
Alcohol intake and preference in selectively bred alcohol-preferring rats: acute administration
Male adult selectively bred alcohol-preferring rats were used. This particular strain of alcohol-drinking rat has been characterized and widely used to study the effects of different compounds on voluntary alcohol consumption (Overstreet et al. 1999). Alcohol intake was determined using the standard two-bottle choice method as previously described (Rezvani et al. 2007, 2000, 2009, 2010). Briefly, after establishment of a reliable baseline for alcohol and water intake, rats were administered systemically (subcutaneously) in the morning (10:00 am) either with the vehicle (20% (w/v) cyclodextrin), or one of the three doses of JNJ-39220675 (0.3, 3, and 10 mg/kg) or naltrexone (as a positive control, 5 mg/kg, s.c.). Nine animals were used for each group. A cross-over design with random assignment was used for each compound, i.e., each animal in each group received all doses of the test compound, one dose of naltrexone, and the vehicle following a random order. The interval between treatments was at least 3 days. Alcohol and water intake were recorded 1, 2, 4, 6, and 24 h after drug administration and the animal’s body weight was recorded the day of the treatment and the day after.
Alcohol intake and preference in selectively bred alcohol-preferring rats: post-deprivation effects
The rats that have been previously used for the establishment of a dose–response effect on alcohol intake were used. Upon the termination of the dose–response experiment, rats were put on water only for 12 consecutive days and then choice of alcohol and water for 35 consecutive days. This washout period allowed a total clearance of the compounds from the system and re-establishment of a stable baseline for alcohol intake.
After re-establishment of a reliable stable baseline for alcohol and water intake, rats were withdrawn from only alcohol for 72 h (Friday 9 a. m. to Monday 9 a.m.), while they did have free access to food and water. After 72 h of alcohol deprivation, one group of rats (n = 9) was given an acute dose (subcutaneously) of 1 or 10 mg/kg of JNJ-39220675, vehicle or 5.0 mg/kg naltrexone 15 min before re-exposure to the alcohol solution. Alcohol and water intake were measured 1, 2, 4, 6, and 24 h after alcohol re-exposure. All animals in each group received all treatments following a random order design with a 1-week of washout period between injections.
Alcohol intake (grams per kilogram) was calculated by multiplying the volume of alcohol consumed in milliliter by 10% and 0.7893 (alcohol density)/body weight in kilogram. Alcohol preference, expressed as percentage, was calculated as follows (volume of alcohol consumed in milliliter/total fluid intake in milliliter) × 100 (Rezvani et al. 2010). Statistical differences between drug-treated and control groups are determined by using ANOVA and Dunnett’s two-tailed test for multiple comparison. All of the treatments groups were compared with the vehicle in the same analysis.
Oral alcohol and saccharin self-administration in alcohol non-dependent rats
Self-administration experiments were conducted in sound attenuated operant chambers (Med-Associates, St Albans, VT, USA) equipped with one house light, two retractable levers, two stimulus lights located above the levers, and two liquid hoppers.
Fourteen rats were used in this study. Sessions started with the illumination of the left and right light and lasted for up to 30 min or until 100 saccharin rewards were delivered, whichever occurred first. Initially, all rats were trained to lever press under a fixed ration (FR) 1 schedule of reinforcement. Pressing the active lever, which was assigned equally right and left, resulted in the delivery of 0.1 ml of a liquid solution of saccharin (0.1% v/v). Pressing the inactive lever had no programmed consequence but was recorded. When the number of saccharin deliveries was greater than 50, the FR was increased to 2 and, following the same criterion, to a final FR3. When the number of saccharin deliveries was stable for three consecutive sessions (did not change by more than 20%), half of the animals were switched to a mixed liquid solution of saccharin (0.1% v/v) and alcohol (8% v/v) and, upon reaching stability, to a 0.1-ml solution of alcohol (8% v/v) only. The remaining half continued to respond for saccharin (0.1% v/v) only. Tests began when stability was reached under the final schedule of reinforcement. JNJ-39220675 (30 min pretreatment time at 10 and 30 mg/kg, s.c.) or vehicle was administered using a latin-square design. In addition, animals received naltrexone (30 min pretreatment time at 1 mg/kg, s.c.) as a positive control.
Self-administration of alcohol and saccharin were measured as total number of lever presses. Statistical differences between drug-treated and control groups are determined by using a one-way ANOVA for dose, with repeated measures on dose, and Duncan post hoc tests when appropriate. One-tailed tests were used, since only decreases in operant responding would indicate efficacy. Data were analyzed using GraphPad Prism (GraphPad Software, San Diego, CA) and Suprnova (SAS Institute, Cary, NC).
Alcohol induced motor impairment
To examine the effect of alcohol on motor performance, the rats were trained to stay walking on a rotating rod for 60 s (Stoelting, Wood Dale, IL). Animals that reached the criteria (60 s) were tested 24 h later. A between-subject design was used for testing. Eight separate groups of animals (n = 6) received the following treatments: vehicle + saline, vehicle + alcohol 0.25 g/kg, vehicle + alcohol 1 g/kg, vehicle + alcohol 2.5 g/kg, JNJ-39220675 3 mg/kg + saline, JNJ-39220675 3 mg/kg + alcohol 1 g/kg, JNJ-39220675 3 mg/kg + alcohol 2.5 g/kg, JNJ-39220675 3 mg/kg + alcohol 0.25 g/kg. In a separate, follow-up study, 10 mg/kg JNJ-39220657 or vehicle, combined with 1 g/kg alcohol was tested (two more separate group of animals). Vehicle or JNJ-39220675 was given subcutaneously 30 min before alcohol, whereas saline or alcohol was given 15 min before the beginning of the test. Alcohol (20%) was administered intraperitoneally in a volume up to 15 ml/kg.
Motor performance was measured as latency to fall. Data were analyzed using GraphPad Prism (GraphPad Software, San Diego, CA). Latency to fall was analyzed with two-way analysis of variance (ANOVA). An effect was considered statistically significant when p < 0.05.
The effect of a subcutaneous dose of JNJ-39220675 (10 mg/kg) on alcohol elimination was also addressed by measuring alcohol plasma concentration after i.p. administration of a 2.5 g/kg dose of alcohol. Blood samples were collected 1, 3, and 6 h after alcohol administration. For the determination of plasma ethanol levels, 100 μl of plasma was mixed with 500 μl internal standard solution (n-propanol 300 mg/l in water) and vortexed. Of this solution, 200 μl was placed into a headspace vial. The vial was heated to 60°C for 5 min, and from the gasses in the vial, 250 μl was injected onto a DB-ALC-2 column (length, 30 m, internal diameter, 0.320 mm; film thickness, 1.2 μm; Agilent, Santa Clara, CA) using a headspace injector (Combi Pal, CTC Analytics, Switzerland) with an injection speed of 250 μl/s. Ethanol levels were determined using a gas chromatograph (7890 GC, Agilent, Santa Clara, CA) with elution times of 1.65 min for ethanol and 2.78 min for n-propanol.
Kinetics of alcohol elimination were carried out using two-way non-repeated measures ANOVA for treatment × time.
Microdialysis experiments were performed as previously described (Dugovic et al. 2009) using a guide cannula (Eicom, Kyoto, Japan) in the nucleus accumbens core (incisor bar, −3.5 mm, +1.8 mm anterior, 1.6 mm lateral, and 6.0 mm ventral to Bregma) (Paxinos and Watson 1997). Microdialysis probes (artificial cellulose; molecular weight cut off, 50,000; 2 mm active membrane length, Eicom) were perfused with artificial cerebral spinal fluid (147 mM NaCl, 4.0 mM KCl, 0.85 mM MgCl2, 2.3 mM CaCl2) at a rate of 1 μl/min and implanted the afternoon prior to sample collection. The following morning, seven 10-min baseline samples were collected prior to drug injections. Animals were then pretreated (time 0) with JNJ-39220675 (3 mg/kg, s.c.) or vehicle, immediately before an injection of alcohol or its vehicle (n = 4 to 6). Alcohol was diluted to 20% with sterile water and delivered at 15.7 ml/kg (i.p.) for a dose of 2.5 g/kg. Samples were collected every 10 min for 160 min post-dosing into a 96-well plate maintained at 4°C containing 2.5 μl of antioxidant (0.1 M acetic acid, 1 mM oxalic acid, and 3 mM l-cysteine in ultra pure water). Microdialysis samples were analyzed for DA by HPLC with electrochemical detection (Eicom) as previously described (Barbier et al. 2007; Bonaventure et al. 2007). Following the studies, the animal’s brains were removed and sliced to confirm probe placement.
Microdialysis samples were analyzed as percentage of baseline with a three-way analysis of variance (ANOVA) for pretreatment (JNJ-39220675 or vehicle), treatment (vehicle or alcohol), and time, with repeated measures on time, followed by Newman–Keul’s post hoc tests when appropriate.
Subcutaneous pharmacokinetic and H3 receptor occupancy of JNJ-39220675 in rats
Pharmacokinetic parameters of subcutaneous administration of 3 mg/kg JNJ-39220675 in rat
0.25 ± 0.00
1.93 ± 0.10
AUCinf (h μmol/L)
2.31 ± 0.08
1.09 ± 0.12
The ex vivo receptor occupancy study showed that JNJ-39220675 crossed the blood brain barrier and occupied the H3 receptor after subcutaneous administration. Both the 3 and 10 mg/kg doses of JNJ-39220675 quickly reached maximal H3 receptor occupancy level (∼90%). This near-maximal level of H3 receptor occupancy measured after administration of 3 and 10 mg/kg doses of JNJ-39220675 was sustained for 4 h. At the 6-h time point, the 10 mg/kg dose still showed 75% of receptor occupancy (plasma concentration = 0.051 ± 0.017 μM; brain concentration = 0.193 ± 0.057 μM), whereas the 3 mg/kg dose was down to 55% receptor occupancy (plasma concentration = 0. 020 ± 0.011 μM; brain concentration = 0.115 ± 0.033 μM). The lower dose (0.3 mg/kg) reached 70–75% of H3 receptor occupancy up to 4 h (plasma concentration = 0.023 ± 0.003 μM; brain concentration = 0.098 ± 0.011 μM) and decreased to 20–25% receptor occupancy at the 6-h time point. For all three doses of JNJ-39220675, no significant H3 receptor occupancy level was measured at the 24-h time point. Similarly, no drug was detected in plasma or brain at the 24-h time point. The average brain to plasma ratio was 3.4.
Effect of the various vehicles used in models relevant to alcoholism on plasma concentration of JNJ-39220675 after s.c. administration of a 3-mg/kg dose
Plasma concentration (μM)
0.74 ± 0.08
1.01 ± 0.15
0.60 ± 0.06
0.82 ± 0.06
Acute effect of JNJ-39220675 on alcohol intake, alcohol preference, water intake, and total fluid intake in alcohol-preferring rats
Effect of JNJ-39220675 (0.3, 3, and 10 mg/kg s.c.) and naltrexone (5 mg/kg s.c.) on water intake and total fluid intake at different time points in selectively bred alcohol-preferring P rats (mean ± SEM, n = 9)
Water intake (g/kg)
Total fluid intake (ml/kg)
4 ± 1.5
6 ± 1.8
10 ± 2.8
16 ± 2.6
26 ± 5.5
16 ± 2.6
23 ± 2
39 ± 3.1
50 ± 4.4
99 ± 5.0
JNJ-39220675 0.3 mg/kg
5 ± 1.4
7 ± 2.2
10 ± 3.5
14 ± 4.6
29 ± 7.1
10 ± 1.3
16 ± 2.3
30 ± 3.6
45 ± 4.9
99 ± 4.8
JNJ-39220675 3 mg/kg
8 ± 1.8
10 ± 1.9
16 ± 2.2
21 ± 3
29 ± 7.1
10 ± 1.5
17 ± 1.5
35 ± 2.1
48 ± 4.1
95 ± 4.8
JNJ-39220675 10 mg/kg
6 ± 1
12 ± 1.7
17 ± 2.6
23 ± 3.8
38 ± 6.1
8 ± 1.2
18 ± 2.3
34 ± 3.6
45 ± 4.1
94 ± 4.0
Naltrexone 5 mg/kg
4 ± 1.1
12.27 ± 2.1
23.2 ± 2.8
28.4 ± 4.1
6 ± 0.9
10 ± 1.1
20 ± 2.9
43 ± 6.1
90 ± 3.1
Compared with control vehicle, JNJ-39220675, when given systemically 15 min before alcohol exposure, significantly reduced alcohol intake. Analyses of data showed a significant main treatment effect for alcohol intake [F(4, 32) = 8.03, p = 0.0001] (Fig. 3a). The treatment × time interaction for alcohol intake was not significant (p = 0.6038).
An acute dose of naltrexone (5 mg/kg, s.c.) given to the same animals resulted in a significant treatment main effect on alcohol intake [F(1, 8) = 17.458, p = 0.003]. The treatment × time interaction was not significant (F(1, 8) = 1.085, p = 0.381).
Analysis of data showed that the treatment × time interaction was significant (p = 0.0027) for alcohol preference, indicating different effects at different time points. Analysis of the simple main effects of treatment at each time point showed that the low dose of JNJ-39220675 was not effective in reducing alcohol preference at any time point, while the medium dose was effective up to 4 h (p < 0.0018 and 0.02 at 2 and 4 h time points), and the high dose of 10 mg/kg was effective up to 6 h (p < 0.002, 0.0039, 0.025 at 2, 4 and 6 h, respectively) (Fig. 3b).
Naltrexone significantly reduced alcohol preference up to 6 h after the drug administration [F(1, 8) = 19.12, p = 0.002]. The treatment × time interaction was significant [F(1, 8) = 5.332, p = 0.002] (Fig. 3b), indicating different effect of the drug at different time points (P values at 1, 2, 4, and 6 h time points were <0.05).
Compared with control vehicle, JNJ-39220675 did not significantly increase water intake in the overall analysis [F(4, 32) = 2.411, p = 0.0695] (Table 3). However, a linear trend analysis over the treatment doses showed that the water intake followed a significant dose-dependent increase [F(1, 24) = 8.785, p < 0.01]. Acute administration of naltrexone also did not exert a significant effect on water intake [F(1, 8)1.212, p = 0.303] (Table 3).
Total fluid intake
An acute administration of JNJ-39220675 significantly decreased total fluid intake (Table 3). The main treatment effect was statistically significant [F(4, 32) = 8.867, p = 0.0001]. There was no significant treatment × time interaction [F(4, 32) = 0.94, p = 0.525]. Acute administration of naltrexone also significantly reduced total fluid intake [F(1, 8) = 22.74, p = 0.001] (Table 3). The treatment × time interaction was not significant [F(1, 8) = 2.083, p = 0.106].
Acute effect of JNJ-39220675 on alcohol intake, alcohol preference, water intake, and total fluid intake in alcohol-preferring rats following a 3-day alcohol deprivation
Effect of JNJ-39220675 (1 and 10 mg/kg s.c.) and naltrexone (5 mg/kg s.c.) on water intake and total fluid intake at different time points in selectively bred alcohol-preferring P rats withdrawn from alcohol for 72 h (mean ± SEM, n = 9)
Water intake (g/kg)
Total fluid intake (ml/kg)
1.3 ± 0.6
1.6 ± 0.7
1.6 ± 0.7
2.4 ± 1.1
6.8 ± 1.8
20 ± 1.7
23 ± 1.7
26 ± 1.8
33 ± 2.8
79 ± 4.3
JNJ-39220675 1 mg/kg
3.3 ± 1.04
4.4 ± 1.4
5 ± 1.5
6 ± 1.6
10 ± 2.3
12 ± 1.5
17 ± 0.9
30 ± 1.4
39 ± 2
82 ± 4.1
JNJ-39220675 10 mg/kg
5 ± 1.7
7.7 ± 2
9.9 ± 2.5
13.8 ± 3.6
19.1 ± 6.1
12 ± 1.1
17 ± 1.5
27 ± 2.7
43 ± 2.6
84 ± 5
Naltrexone 5 mg/kg
1.8 ± 0.5
3 ± 0.7
4.9 ± 1.1
7.2 ± 2.1
12.7 ± 3.4
5 ± 0.8
7 ± 0.7
13 ± 1.1
23 ± 2
61 ± 3.7
JNJ-39220675 did show a significant main effect by reducing alcohol intake [F(3, 32) = 10.707, p = 0.0001] but not a significant treatment × time interaction [F(12, 96) = 2.1, p = 0.235] (Fig. 4a).
Compared with the control vehicle, administration of 5 mg/kg naltrexone significantly reduced alcohol intake (Fig. 4a). The main treatment effect was significant [F(1, 8) = 46.395, p = 0.001], while the treatment × time interaction was not statistically significant.
Alcohol preference was significantly reduced by JNJ-39220675 (Fig. 4b). Analysis of the results showed a significant main effect [F(3, 32) = 4.652, p < 0.0106], while the treatment × time interaction [F(12, 96) = 1.599, p < 0.1] was not significant.
Compared with the vehicle, naltrexone significantly reduced alcohol preference (Fig. 4b). The treatment × time interaction was statistically significant (p = 0.042). Analyses of the simple main effects of naltrexone at each time point showed that alcohol preference was significantly reduced up to 24 h after the administration of naltrexone (Fig. 4b) (p < 0.0002 at 1, 2, 4, and 6 h and p < 0.05 at 24 h).
Compared with the control vehicle, the compound significantly increased water intake (Table 4). The main effect of JNJ-39220675 on water intake was significant [F(3, 32) = 4.137 p < 0.0169]. The treatment × time interaction was not statistically significant [F(12, 96) = 1.7, p = 0.076].
Analysis of data showed that the main treatment effect for naltrexone on water intake was not significant. However, there was a statistically significant (p < 0.02) interaction between treatment and time (Table 4), indicating different effects of the drug at different time points.
Total fluid intake
Treatment × time interaction was statistically significant [F(12, 96) = 4.757, p = 0. 0001], indicating different effects of drug at different time points. P values for 1, 2, 4, 6, and 24 h for low dose of the compound were 0.0006, 0.02, 0.08, 0.02, and 0.33, respectively. The corresponding values for the high dose vs. control were 0.002, 0.03, 0.70, 0.0004, and 0.08 (Table 4).
Acute effect of JNJ-39220675 on alcohol and saccharin self-administration in alcohol non-dependent rats
A pretreatment time of 30 min for the administration of JNJ-39220675 was selected based on the receptor occupancy and pharmacokinetic profile of the compound.
To provide evidence against a rate-sensitive effect, an additional subgroup analysis was performed. In this subgroup analysis, the effects of JNJ-39220675 in the three subjects that showed the high baseline levels of ethanol self-administration was assessed. A similar analysis was performed with the three subjects that showed the lowest baseline levels of saccharine self-administration. The subgroup analysis was found to be consistent with the overall analysis (results not shown).
Effect of JNJ-39220675 on alcohol-induced ataxia and alcohol elimination rate
Two-way ANOVA indicated that there was a significant alcohol effect [F(3, 37) = 60.5, p < 0.05]. Alcohol dose-dependently reduced the latency to fall (Fig. 6a). A single administration of JNJ-39220675 (3 mg/kg) did not significantly change the ataxic effects of alcohol [F(1, 37) = 2.9, p > 0.05].
To confirm that JNJ-39220675 did not produce non-selective, sedative effects at the higher doses tested, the effect of 10 mg/kg JNJ-39220675 combined with 1 g/kg alcohol on rotarod performance was determined. The latency to fall for the animals treated with vehicle and 1 g/kg alcohol was 11.83 ± 8.01 s (N = 6), while the latency to fall for the animals treated with 10 mg/kg JNJ-39220675 and 1 g/kg alcohol was 32.20 ± 13.03 s (N = 5). Thus, even high doses JNJ-39220675 did not significantly change the ataxic effects of alcohol (p = 0.20, t test).
The effect of JNJ-39220675 (10 mg/kg, s.c.) on alcohol elimination rate was also determined (Fig. 6b, N = 6/group). Blood alcohol concentrations decreased over time (Fig. 6b, F(2, 30) = 251.44, p = 0.00005, main effect of time), but elimination rates were not affected by treatment (F(1, 30) = 1.68, p = 0.20, main effect of treatment, F(2, 30) = 0.03, p = 0.97, treatment × time interaction).
Effect of JNJ-39220675 on alcohol-induced dopamine release in the nucleus accumbens
The ANOVA revealed a main effect of alcohol treatment (F(1, 15) = 8.79, p = 0.009), a main effect of time (F(23, 345) = 9.41, p = 0.0001), and a time × treatment interaction (F(23, 345) = 5.20, p = 0.00001). However, JNJ-39220675 was without effect (main effect of pretreatment, F(1, 15) = 0.30, p = 0.59; pretreatment × treatment interaction, F(1, 15) = 0.12, p = 0.73; pretreatment × time interaction, F(23, 345) = 0.84, p = 0.86; pretreatment × treatment × time interaction, F(23, 345) = 0.41, p = 0.99).
Newman–Keuls post hoc tests on the treatment × time interaction revealed that alcohol significantly stimulated dopamine release compared to vehicle at the 60 (p = 0.03), 80 (p = 0.02), 90 (p = 0.04), 100 (p = 0.002), 110 (p = 0.03), 120 (p = 0.004), and 130 (p = 0.03) minute time points (Fig. 7).
JNJ-39220675, a selective H3 receptor antagonist, significantly reduced both alcohol intake and preference in alcohol-preferring (P) rats. Furthermore, this compound also significantly reduced both alcohol intake and preference in alcohol-preferring rats following a 3-day alcohol deprivation when the urge for drinking is enhanced (Rezvani et al. 2009). JNJ-39220675 significantly and dose-dependently reduced alcohol self-administration without changing saccharin self-administration in alcohol non-dependent rats. JNJ-39220675 at the dose range effective in reducing alcohol intake did not affect the ataxic effects of alcohol nor alcohol elimination rate. Neurochemical experiments also demonstrated that JNJ-39220675 did not affect alcohol-induced dopamine release in nucleus accumbens.
JNJ-39220675 has low nanomolar affinity and excellent selectivity for the rat and human H3 receptor (Letavic et al. 2010). However, unlike thioperamide, JNJ-39220675 has no affinity for the H4 receptor. JNJ-39220675 displays physical properties consistent with good absorption and distribution. After systemic administration in rat, JNJ-39220675 occupied the H3 receptor in brain [see Fig. 2 for subcutaneous administration and (Letavic et al. 2010) for oral administration]. In a previous study, it was demonstrated that the compound increased extracellular histamine release in frontal cortex, and it induced wake-promoting effects in a rat EEG study (Letavic et al. 2010). Therefore, this compound is suitable for in vivo pharmacological studies.
In the present study, we first investigated the effect of JNJ-39220675 on alcohol intake and preference in a genetic animal model of human alcoholism, the P rats (Overstreet et al. 1999; Rezvani et al. 2010; Richard et al. 2006). Similar to naltrexone, the H3 antagonist, when given acutely, significantly reduced both alcohol intake and alcohol preference in P rats (Fig. 3). The medium dose of 3 mg/kg significantly reduced alcohol preference up to 4 h and a larger dose of 10 mg/kg up to 6 h. These results indicate that near-maximal level of H3 receptor occupancy is required for efficacy in this model (Fig. 2a). The low dose of JNJ-39220675 was not effective in reducing alcohol preference at any time point despite showing ∼75% H3 receptor occupancy. Noteworthy, different vehicles were used for the pharmacokinetic/receptor occupancy study and alcohol intake/preference study (saline vs. cyclodextrin). However, the data presented in Table 2 indicate that the plasma exposure measured at the 1-h time point was not statistically different between the two vehicles.
There was no significant change in water intake. Although the main effect on total fluid intake was significant, direct comparison between vehicle and treatment at each time point showed no significant effect on total fluid intake. The profile of the effect of this compound is comparable with naltrexone, as naltrexone also significantly reduced alcohol intake and alcohol preference and significantly reduced total fluid intake. However, there was a non-significant trend for the water to be increased with JNJ-39220675.
In alcohol post-deprivation experiments, rats were withdrawn from alcohol for 3 consecutive days and then were given a dose of JNJ-39220675, naltrexone, or vehicle before re-exposure to alcohol. The H3 antagonist when tested in this model of “relapse” did show a significant main effect but not a significant treatment × time interaction. The reduction in alcohol intake was accompanied with a significant increase in water intake which resulted in a significant reduction in alcohol preference. Compared to the naltrexone treatment, the JNJ-39220675 seems to have a better profile in this animal model of “relapse” because it increased water intake while decreasing alcohol consumption, whereas naltrexone reduced alcohol intake without changing water intake resulting in more total fluid reduction.
The data obtained in alcohol-preferring rat lines should be interpreted with caution. For example, alcohol-preferring rats (AA rats), a line which was not used in these studies, have an altered histaminergic system (Lintunen et al. 2001). Therefore, the effects of H3 antagonists may not necessarily translate into humans who might not have the same alterations in their histaminergic system. Noteworthy, it is not known if P rats (the line of rats we used in our studies) also have an altered histaminergic system. A major limitation of the genetic model of high alcohol preference is that alcohol preference alone does not necessarily indicate uncontrolled alcohol-drinking behavior but often reflects controlled alcohol consumption (Spanagel 2003). Thus, we also evaluated the effects of the JNJ-39220675 on operant alcohol self-administration in alcohol non-dependent rats. Interestingly, the data indicate that the H3 antagonist dose-dependently reduced alcohol-reinforced lever presses, but did not alter saccharin self-administration. These results suggest that the effects of the compound were selective for alcohol, unlike naltrexone, which non-selectively decreases alcohol and saccharin intake (data not shown). Thioperamide and clobenpropit have been shown to reduce operant responding for alcohol in alcohol-preferring rats (Lintunen et al. 2001). These results are consistent with our studies, indicating that JNJ-39220675 also decreases alcohol self-administration in Wistar rats. It is not clear why larger doses (i.e., 10 mg/kg and above) were required to observe a significant decrease in alcohol self-administration in Wistar rats compared to the effects observed in P rats. However, it is important to highlight that large doses of JNJ-39220675 did not modify saccharin self-administration in Wistar rats, demonstrating that the effects of the H3 receptor antagonist were selective and specific for alcohol. The pharmacokinetic bridging study with the various vehicle indicated that the DMSO/Cremophor vehicle used for the self-administration experiments resulted in lower plasma exposure than cyclodextrin (Table 2). However, it should be noted that this difference in plasma exposure is relatively small and not likely to explain the different doses needed in alcohol self-administration experiments vs. the P rats.
We then tested whether JNJ-39220675 reduced the reinforcing effects of alcohol by changing dopamine release in the brain. In the striatum, presynaptic H3 receptors have been identified on dopaminergic terminals, where they seem to exert an inhibitory role on dopamine release (Doreulee et al. 2000; Schlicker et al. 1993). Here, we found that the H3 antagonist by itself (in the absence of alcohol) had no effect on dopamine release in the nucleus accumbens. In addition, alcohol-induced dopamine release in the nucleus accumbens was not blocked by JNJ-39220675. Further studies will address the interaction of H3 antagonists with other neurotransmitter systems especially glutamate. For example, activation of H3 receptors has been shown to inhibit glutamate release from rat striatal synaptosomes (Molina-Hernandez et al. 2001). H3 receptors are also expressed at postsynaptic sites in the striatum (Haas and Panula 2003), but their function is not fully understood. Anatomical studies have demonstrated that the majority of H3 receptors in the striatum co-localized with dopamine D1 receptors in the striato nigral direct pathway and with D2 receptors in the indirect pathway (Garcia et al. 1997; Ryu et al. 1994). A recent study also suggests that H3 receptors coupled to dopamine D2 receptors, decreasing the affinity of D2 receptors for its agonists (Ferrada et al. 2008). Therefore, H3 receptor-mediated change in alcohol stimulation might result from an interaction between H3 receptor and D2 or D1 receptor signaling.
Lastly, we also studied the effect of the H3 antagonist on alcohol-induced ataxia. JNJ-39220675 had no effect on alcohol-induced ataxia. These findings are in line with the data obtained in histidine decarboxylase knockout mice, where histamine deficiency did not lead to a change in alcohol-evoked impairment of motor task, suggesting that histamine is not critical for the mediation of alcohol’s motor impairing effects (Lintunen et al. 2001). We also demonstrated that JNJ-39220675 had no effect on alcohol elimination. There is no indication that the animals showed nausea, since saccharin intake was unaltered. The compound was tested in the Irwin screen test with no effects at the doses used in the current study (unpublished results). In addition, the animals do not move less as indicated by the lack of change in saccharin operant responding. Unpublished data also confirmed that JNJ-39220675 had no effect on locomotor activity up to 30 mg/kg s.c.
Overall, our results confirmed and extend previous findings obtained with relatively unselective imidazole containing H3 antagonists and H3 receptor knockout mice (Lintunen et al. 2001; Panula et al. 2008). These results indicate, for the first time, that selective blockade of H3 receptors by non-imidazole H3 receptor antagonists is an attractive and alternative mechanism for the treatment of alcoholism. This is particularly important given that alcoholism and drug abuse and dependence is often associated with cognitive related deficits, especially decision making and impulsivity (Vocci et al. 2005). Thus, the positive effect of H3 antagonists on cognitive function (Esbenshade et al. 2009; Galici et al. 2009; Witkin and Nelson 2004) may also represent an additional benefit for the treatment of alcoholism. Further preclinical work is needed to address the effects of H3 antagonists on craving and relapse and to better understand their exact mechanism of action. Future studies will also determine if the effect of the H3 antagonists in model relevant to alcoholism are mediated by histamine via the H1 receptor.
The assistance of Dr. Kevin Sharp and his staff at Johnson & Johnson Pharmaceutical Research & Development L.L.C. (San Diego, CA) is gratefully acknowledged.