Unresponsiveness to Cannabinoids and Reduced Addictive Effects of Opiates in CB1 Receptor Knockout Mice

See allHide authors and affiliations

Science  15 Jan 1999:
Vol. 283, Issue 5400, pp. 401-404
DOI: 10.1126/science.283.5400.401


The function of the central cannabinoid receptor (CB1) was investigated by invalidating its gene. Mutant mice did not respond to cannabinoid drugs, demonstrating the exclusive role of the CB1 receptor in mediating analgesia, reinforcement, hypothermia, hypolocomotion, and hypotension. The acute effects of opiates were unaffected, but the reinforcing properties of morphine and the severity of the withdrawal syndrome were strongly reduced. These observations suggest that the CB1 receptor is involved in the motivational properties of opiates and in the development of physical dependence and extend the concept of an interconnected role of CB1 and opiate receptors in the brain areas mediating addictive behavior.

Marijuana and other derivatives ofCannabis sativa have been used for centuries for their therapeutic and mood-altering properties and are the most widely used recreational drugs today (1). The active compounds ofCannabis, including Δ9-tetrahydrocannabinol (Δ9-THC), as well as the endogenous cannabinoid anandamide, act through two G protein–coupled receptor subtypes. The CB1 receptor is abundant in the central and peripheral nervous systems but is also expressed in several peripheral organs, whereas CB2 receptor expression is essentially restricted to lymphoid organs (2). We investigated the in vivo function of the CB1 receptor by invalidating its gene in a mouse model (3). Northern (RNA) blotting demonstrated the absence of CB1 transcripts in brain and testis from knockout (CB1 −/−) mice, and binding assays confirmed the absence of binding sites for cannabinoid ligands (4). Histology of brain and other organs, body weight monitored over a 6-month period, and blood ionogram and cell count appeared to be unaffected by CB1 gene inactivation.

The consequences of CB1 receptor inactivation on spontaneous behavior were analyzed. A moderate increase in locomotor activity (5) was observed in CB1 −/− mice when newly exposed to the arena (119% of controls; P< 0.001, unpaired two-tailed Student's t test,n = 50), but not after an habituation period. Increased exploratory behavior was also found under the more stressful conditions of an open field (P < 0.01, t test,n = 15) (5) and in the spontaneous alternation test (total number of visits to the arms: CB1 +/+, 47.2 ± 1.6; CB1 −/−, 58.9 ± 2.2; P < 0.01, t test,n = 15) (6). Both groups of animals exhibited a rapid habituation to the open-field test (5). However, the time spent in exploring unknown objects placed in the field was significantly increased for the mutant mice (CB1 +/+, 0.66 ± 0.3 s; CB1 −/−, 5.33 ± 1.5 s; P < 0.01, t test , n = 15). Furthermore, a decrease in spontaneous alternation (6) was observed for the mutant mice in the Y maze (CB1 +/+, 61.4 ± 1.8%; CB1 −/−, 53.7 ± 1.9%; P< 0.01, t test, n = 15). In the elevated plus maze (5), the number of entries and time spent in the open arms were unaffected. Taken together, the data suggest that CB1 −/− mice present a mild impairment in the adaptation to new environment that could be related to changes in short-term memory or attention (or both).

The spontaneous nociceptive threshold (7) of wild-type and mutant naive mice was similar [not significant (NS),t test] in the hot-plate (jumping behavior: CB1 +/+, 52.0 ± 3.8 s; CB1 −/−, 46.6 ± 4.5 s; n = 10), tail-immersion (CB1 +/+, 0.97 ± 0.08 s; CB1 −/−, 1.04 ± 0.06 s; n = 10), writhing (CB1 +/+, 35.2 ± 1.9; CB1 −/−, 34.6 ± 2.0; n = 10), and tail-pressure tests (CB1 +/+, 7.0 ± 0.2 s; CB1 −/−, 7.2 ± 0.2 s; n= 20). These observations suggest that the endogenous activation of the CB1 receptor is not crucial for the control of pain or that other endogenous systems might compensate for the absence of this receptor (or both).

The role of the CB1 receptor in the central effects of cannabinoids was investigated by measuring the response of CB1 +/+ and CB1 −/− mice to Δ9-THC in different assays (Fig. 1). The antinociceptive properties of Δ9-THC were not observed for mutant mice in the hot-plate test and were strongly reduced in the tail-immersion test, in which a slight antinociceptive effect was observed for the highest dose (Fig. 1, A and B), possibly in line with the recent demonstration that CB2 receptors may regulate pain initiation at sites of tissue injury (8). Other classical effects of Δ9-THC, namely, the reduction of horizontal locomotor activity (5) and the decrease of rectal temperature, were observed in wild-type animals but not in mutant mice (Fig. 1, C and D). In an intravenous self-administration model (9), WIN55,212-2 was not self-administered by CB1 −/− mice, in contrast to wild-type animals (Fig. 1E). Dependence induced by Δ9-THC administration was also investigated in mutant mice (10). The selective CB1 receptor antagonist SR141,716A precipitated behavioral manifestations of abstinence in wild-type mice but not in mutant mice given long-term treatment with Δ9-THC (Fig. 1F). These results demonstrate that the main pharmacological responses to Δ9-THC, as well as the addictive properties of cannabinoids, are indeed mediated mostly, if not exclusively, by the CB1 receptor.

Figure 1

Central effects of cannabinoids on CB1 +/+ (□, ▧) and CB1 −/− (▪, ▨) mice. For (A) to (D), an intraperitoneal injection of Δ9-THC (or vehicle alone) was made 20 min before measurements. (A) Latency for escape jumping in the hot-plate test (n = 10). (B) Latency for tail withdrawal in the tail-immersion test (n = 10). (C) Spontaneous activity in locomotor activity boxes (number of photocell counts within 10 min; n = 10). (D) Rectal temperature (n = 10). (E) Self-administration of WIN55,212-2 (9). Injection (inj) of agonist or vehicle to active (□, ▪) and passive (▧, ▨) mice was coupled to the nose-poke behavior of the active mouse (n = 8 for WIN55,212-2 and 4 for vehicle). (F) Signs reflecting Δ9THC withdrawal (10) were monitored (n = 5 to 15). The statistical significance [t test for (A) to (D) and (F) and Neuman-Keuls test for (E)] was measured between genotypes and against vehicle for drug-treated groups. Error bars: SEM.

Cannabinoids have been reported to elicit hypotension and bradycardia through peripheral CB1 receptors (11). Basal blood pressure and heart rate were measured in conscious mice (12) but were not significantly modified, suggesting that endogenous cannabinoids do not exert a tonic control on these parameters or that other systems may compensate for the absence of the CB1 receptor. Both anandamide and WIN55,212-2 promoted a sustained decrease in blood pressure and heart rate in CB1 +/+ mice, with a biphasic response to anandamide (Fig. 2), in agreement with previous reports (11). No significant hypotensive effect of either drug was observed after their administration to CB1 −/− mice, demonstrating that the CB1receptor is solely responsible for the cardiovascular effects of cannabinoids, including the two components of the response to anandamide.

Figure 2

Cardiovascular effects of cannabinoids on CB1 +/+ (○, □) and CB1 −/− (•, ▪) mice (12). Mean blood pressure (MBP) and heart rate were monitored for 60 min after administration of WIN55,212-2 (0.25 mg/kg; □, ▪), anandamide (2 mg/kg; ○, •), or vehicle (15). The transient drop in heart rate after injection in CB1 −/− mice was also observed after vehicle injection only. n = 9 to 11 for each group. Error bars: SEM. bpm, beats per minute.

An interaction between the opioid and cannabinoid systems has been proposed for the control of nociceptive responses (13). Opiate antagonists such as naloxone have been reported to inhibit cannabinoid agonist–induced dopamine release in the nucleus accumbens (14). Therefore, morphine-induced antinociception and hypothermia, as well as its reinforcing properties and the development of tolerance and physical dependence, were investigated in mutant mice. The antinociceptive effects of morphine in the tail-immersion (15) and the hot-plate (Fig. 3A) tests (7), as well as its hypothermic effects, were not modified in CB1 −/− mice. Furthermore, long-term morphine treatment (16) induced the development of tolerance to morphine antinociceptive effects in the hot-plate (Fig. 3A) and tail-immersion (15) tests in both genotypes. In an intravenous self-administration model (9,17), the number of nose pokes leading to morphine administration was much lower for CB1 −/− mice as compared with CB1 +/+ mice (Fig. 3B), suggesting a reduction of the reinforcing effects of the drug. The behavioral expression of naloxone-precipitated morphine withdrawal (18), shown to be critically dependent on the μ-opioid receptor (19), was also significantly decreased (seven of nine signs evaluated) in mutant mice (Fig. 4), suggesting that CB1receptors are required for the development of physical dependence or to obtain a complete manifestation of the somatic signs of opiate withdrawal. These findings are particularly important when one takes into account the proposed interaction between cannabinoids and opioid dependence (14, 20), which could influence the establishment of opiate addiction. Interestingly, our results show a dissociation between the development of opiate tolerance (unchanged) and dependence (decreased) in mutant mice, confirming that these two processes can be independently developed (21). The specific interactions between κ-opioid and cannabinoid receptors (22) were examined with the selective κ-opioid agonist U-50,488H (23). Antinociceptive responses and hypolocomotion induced by short-term U-50,488H administration were similar in mutant and wild-type mice. However, the dysphoric effects of this κ agonist in the conditioning place aversion paradigm (24) were observed in wild-type mice but not in mutants (Fig. 3C). Therefore, CB1 receptors seem to be involved in the behavioral manifestations of morphine physical dependence and the dysphoric properties of κ agonists but not in the acute effects induced by opioids in antinociception, body temperature, and locomotion. Cannabinoid agonists have been considered as therapeutics for their antiemetic, analgesic, anticonvulsant, and intraocular hypotensive effects (1). Long-term CB1 antagonist administration could also be considered for preventing the development of dependence on opiates and possibly other addictive drugs.

Figure 3

Central effects of opiates on CB1 +/+ (□, ▧) and CB1 −/− mice (▪, ▨). (A) Hot-plate test (jumping) after injection of morphine (or vehicle) to naı̈ve mice (short term) or mice treated for 6 days with morphine (long term), showing the development of tolerance (n = 8 to 19). Similar effects were obtained for the licking behavior, as well as in the tail-immersion test (15). (B) Self-administration of morphine (9). Injection of morphine or vehicle to active (□, ▪) and passive (▧, ▨) mice was controlled by nose pokes of the active mouse, and the number of nose pokes was recorded. n = 6 to 10 per group. (C) Place aversion test, with the κ agonist U-50,488H (24). n = 10 per group. The statistical significance [t test for (A) and (C) and Newman-Keuls test for (B)] was measured between genotypes and against vehicle for drug-treated groups. Error bars: SEM.

Figure 4

Morphine withdrawal syndrome on CB1 +/+ (□) and CB1 −/− mice (▪). Signs reflecting withdrawal were monitored after the long-term administration of morphine (Morph) followed by naloxone injection (18). Animals were observed for 30 min and scored.n = 9 to 10 per group. The t test was used. Error bars: SEM.

  • * To whom correspondence should be addressed. E-mail: mparment{at}


Stay Connected to Science

Navigate This Article