Abstract
Cannabinoids evoke profound hypothermia in rats by activating central CB1 receptors. Nitric oxide (NO), a prominent second messenger in central and peripheral neurons, also plays a crucial role in thermoregulation, with previous studies suggesting pyretic and antipyretic functions. Dense nitric-oxide synthase (NOS) staining and CB1 receptor immunoreactivity have been detected in regions of the hypothalamus that regulate body temperature, suggesting that intimate NO-cannabinoid associations may exist in the central nervous system. The present study investigated the effect of Nω-nitro-l-arginine methyl ester (l-NAME), a NO synthase inhibitor, on the hypothermic response to WIN 55212-2 [4,5-dihydro-2-methyl-4(4-morpholinylmethyl)-1-(1-naphthalenylcarbonyl)-6H-pyrrolo[3,2,1ij]quinolin-6-one], a selective cannabinoid agonist, in rats. WIN 55212-2 (1–5 mg/kg, i.m.) produced dose-dependent hypothermia that peaked 45 to 90 min post-injection. l-NAME (10–100 mg/kg, i.m.) by itself did not significantly alter body temperature. However, a nonhypothermic dose of l-NAME (50 mg/kg) potentiated the hypothermia caused by WIN 55212-2 (0.5–5 mg/kg). The augmentation was strongly synergistic, indicated by a 2.5-fold increase in the relative potency of WIN 55212-2. The inactive enantiomer of WIN 55212-2, WIN 55212-3 [S-(–)-[2,3-dihydro-5-methyl-3-[(morpholinyl)methyl]pyrrolo[1,2,3-de]-1,4-benzoxazinyl]-(1-napthanlenyl) methanone mesylate] (5 mg/kg, i.m.), did not produce hypothermia in the absence or presence of l-NAME (50 mg/kg), confirming that cannabinoid receptors mediated the synergy. The present data are the first evidence that drug combinations of NOS blockers and cannabinoid agonists produce synergistic hypothermia. Thus, NO and cannabinoid systems may interact to induce superadditive hypothermia.
Cannabis and its derivative compounds, collectively called cannabinoids, evoke an array of pharmacological symptoms in rodents, including hypothermia, analgesia, catalepsy, and hypolocomotion. The discovery of endogenous ligands, such as anandamide and 2-arachidonyl glycerol, which act at cannabinoid receptors, has intensified the interest in the therapeutic potential of cannabinoids (Bisogno et al., 1997). CB1 receptors are located primarily in the central nervous system (Howlett, 1995), whereas CB2 receptors are expressed by peripheral immune cells (Dragic et al., 1996).
The development of cannabinoid agonists and antagonists has facilitated the characterization of cannabinoid receptor subtypes and their pharmacological profiles. One such agonist is the aminoalkylindole, WIN 55212-2, which exhibits high selectivity for cannabinoid receptors and interacts negligibly with other neurotransmitter systems and ion channels (Martin et al., 1991; Compton et al., 1992). Several reports indicate that WIN 55212-2 elicits hypothermia in rodents via a CB1 receptor mechanism (Compton et al., 1992; Fan et al., 1994; Fox et al., 2001). Recently, our laboratory confirmed those results by demonstrating that the systemic injection of WIN 55212-2 produces hypothermia that is dependent on CB1, but not CB2, receptors (Rawls et al., 2002a). Moreover, the injection of WIN 55212-2 into the preoptic anterior nucleus of the hypothalamus (POAH), which is thought to be the central site of thermoregulation, evoked CB1-sensitive hypothermia, indicating that intrahypothalamic CB1 receptors play a critical role in the hypothermic response to cannabinoids (Rawls et al., 2002a). CB1 receptor immunoreactivity, binding, and mRNA are also present in the POAH (Mailleux and Vanderhaeghen, 1992; Moldrich and Wenger, 2000), underscoring further the involvement of the cannabinoid system in thermoregulation.
Nitric oxide (NO) has captured the attention of neuroscientists because of its role as a prominent second messenger in the central and peripheral nervous systems (Breder and Saper, 1996). The enzyme nitric-oxide synthase (NOS) catalyzes the production of NO and l-citrulline from the substrate l-arginine. Three isoforms of NO have been discovered. The neuronal and endothelial forms are constitutive, and the third form is inducible (Lowenstein and Snyder, 1992). An accumulating body of evidence suggests that NO participates in thermoregulation. Some studies provide strong evidence that NO production is involved in the hyperthermia evoked by prostaglandin E2, lipopolysaccharide, and morphine (Amir et al., 1991; Minano et al., 1997; Benamar et al., 2001). Furthermore, the injection of l-NAME, a NOS inhibitor, abolishes the fever produced by interleukin-1β and lipopolysaccharide (Roth et al., 1998a). Other studies, however, suggest that NO has an antipyretic function (Gourine, 1995) and participates in hypothermia (Steiner et al., 1998; Almeida and Branco, 2001; Benamar et al., 2002).
Although the hypothermic effects of cannabinoids have been investigated extensively, the contribution of other neurotransmitter systems to this hypothermia is not entirely clear. Recent evidence suggests that glutamatergic, dopaminergic, and opioidergic systems modulate cannabinoidevoked hypothermia (Ledent et al., 1999; Rawls et al., 2002b). The aim of the present study was to ascertain the role of NO in cannabinoid-evoked hypothermia. We investigated the hypothermic effects of WIN 55212-2 by itself and in combination with l-NAME.
Materials and Methods
Animals. All animal use procedures were conducted in strict accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals and were approved by the Temple University Animal Care and Use Committee. Male Sprague-Dawley rats (Zivic-Miller Laboratories, Zelienople, PA) weighing 200 to 250 g were housed two per cage for a minimum of 5 days before experimental use. Rats were maintained on a 12-h light/dark cycle and fed rat chow and water ad libitum.
Drug Preparation and Administration. WIN 55212-2, WIN 55212-3, and l-NAME were purchased from Sigma-Aldrich (St. Louis, MO). All drugs and drug combinations were dissolved in a 10% cremophor/saline solution and injected intramuscularly into the right thigh.
Experimental Protocol. The present research project was conducted over a 3-month period. In all cases, rats were used in only one experiment and received only one injection. No more than 12 rats were tested per day. Body temperature measurements were recorded from each rat over time. On the morning of the experiment, rats were weighed and placed one per cage into an environmental room, which was maintained at a constant temperature of 21 ± 0.3°C and relative humidity of 52 ± 2%. Body temperature experiments were always started between 9 and 10 AM. All rats were allowed to acclimate to their environment for 60 min prior to measuring the first body temperature value. Baseline temperature measurements were taken for 60 min at 30-min intervals (–60, –30, and 0 min) using a thermistor probe (YSI series 400; YSI Inc., Yellow Springs, OH), which was lubricated and inserted approximately 7 cm into the colon. A digital thermometer (model 49 TA; YSI Inc.) was used to record body temperature. Rats were unrestrained during the temperature readings, with only the tail being held gently between two fingers. To allow adaptation to the technique, the first body temperature value was discarded in each rat. Following the baseline interval, rats were administered a drug or drug combination at 0 min. Thereafter, body temperature was recorded at 15, 30, 45, 60, 90, 120, 180, 240, and 300 min post-injection.
Effect of WIN 55212-2 on Body Temperature. The initial set of experiments used a total of 23 rats to determine the effect of WIN 55212-2 alone on body temperature. Different rats were used for each treatment group. A maximum of 12 rats was included in a single experiment, and rats received an injection of either vehicle or one of three doses of WIN 55212-2 (1, 2.5, or 5 mg/kg). Following a 60-min baseline interval, we injected rats at 0 min with vehicle (n = 5), 1 mg/kg WIN 55212-2 (n = 6), 2.5 mg/kg WIN 55212-2 (n = 6), or 5 mg/kg WIN 55212-2 (n = 6). Body temperature was recorded from each rat for 300 min. These doses of WIN 55212-2 have been reported to produce hypothermia in rats (Fox et al., 2001; Rawls et al., 2002 a,b).
Effect of l-NAME on Body Temperature. The second set of experiments used a total of 31 rats to determine the effect of l-NAME alone on body temperature. Different rats were used for each treatment group. A maximum of 12 rats was included in a single experiment, and rats received an injection of either vehicle or one of three doses of l-NAME (10, 50, or 100 mg/kg). Following a 60-min baseline interval, we injected rats at 0 min with vehicle (n = 8), 10 mg/kg l-NAME (n = 7), 50 mg/kg l-NAME (n = 9), or 100 mg/kg l-NAME (n = 7). Body temperature was recorded from each rat for 300 min. We selected doses of l-NAME based on previous studies, which have mostly reported that l-NAME by itself does not produce significant hypothermia (Thorat and Bhargava, 1994; Spina et al., 1998; Almeida and Branco, 2001; Benamar et al., 2001, 2002; Kamerman et al., 2002).
Effect of WIN 55212-2 and l-NAME on Body Temperature. For l-NAME, none of the doses (10–100 mg/kg) used in the present study produced significant alterations in body temperature as compared with vehicle. l-NAME was considered, therefore, to be inactive across the entire dose range used here, and 50 mg/kg was chosen as an inactive and intermediate dose for combination with WIN 55212-2. Thus, in a third set of experiments, we used a total of 97 rats to determine the effect of an inactive dose of l-NAME (50 mg/kg) on the hypothermia produced by various doses of WIN 55212-2 (0.5, 1, 2, 2.5, or 5 mg/kg). Different rats were used for each treatment group. A maximum of 12 rats was included in a single experiment. Following a 60-min baseline interval, rats received an injection of l-NAME (50 mg/kg) alone, WIN 55212-2 (0.5, 1, 2, 2.5, or 5 mg/kg) alone, or a combination of l-NAME (50 mg/kg) plus a single dose of WIN 55212-2 (0.5, 1, 2, 2.5, or 5 mg/kg). Body temperature was recorded from each rat for 300 min.
Effect of WIN 55212-3 and l-NAME on Body Temperature. To verify that cannabinoid receptors mediated the hypothermic actions of WIN 55212-2, we investigated whether WIN 55212-3, the inactive enantiomer of WIN 55212-2, altered body temperature when administered by itself or in combination with l-NAME. A total of 24 rats were used in this final set of experiments. Different rats were used for each treatment group, and a maximum of 12 rats was included in a single experiment. Following a 60-min baseline interval, we injected rats at 0 min with vehicle (n = 6), 5 mg/kg WIN 55212-3 (n = 6), 50 mg/kg l-NAME (n = 6), or a combination of 50 mg/kg l-NAME plus 5 mg/kg WIN 55212-3 (n = 6).
Data Analysis. To allow adaptation to the experimental technique, the first body temperature value for each rat was discarded. Two consecutive body temperature readings were then recorded and averaged to establish a baseline temperature prior to drug injection. Data were calculated as the mean ± S.E. of body temperature. The data were analyzed by either a one-way repeated-measures analysis of variance (ANOVA) followed by a Dunnett's post hoc test or a two-way (group, time) mixed-model ANOVA with repeated measures on time. Comparisons between treatment groups in the two-way ANOVA were conducted using least-squares means on the main effect of group.
The analysis of drug combinations to distinguish synergism from simple additivity followed the procedure described previously (Tallarida, 2001). In that procedure, the graded dose-effect data of the individual drugs are first analyzed to determine an effect level that is reached by both. These equieffective doses (isoboles) are then used to determine the proportions of the combination for testing and to calculate the expected (additive) total dose of the combination needed to attain the specified effect level. This calculated quantity is then statistically compared with the total dose of the combination that produced the specified effect. If the combination total dose is less than the calculated additive total dose, the interaction is synergistic; equality means simple additivity. In cases in which one of the two drugs is inactive, its presence in a simply additive combination has no effect on the dose-effect curve of the active drug. Therefore, the analysis becomes simply one in which the dose-effect curve of the active drug is statistically compared before and after the addition of the inactive agent. The relative potency was determined by using the values of the individual rats in each experimental group (WIN 55212-2 or WIN 55212-2 + l-NAME) at the 60-min time point. The 60-min time point was chosen because WIN 55212-2 by itself produces maximal hypothermia 60 to 90 min post-injection.
Results
WIN 55212-2 Produces Hypothermia.Figure 1 illustrates the effects of WIN 55212-2 (1–5 mg/kg) on body temperature. A two-way ANOVA revealed a significant effect of time [F(3,14) = 194, P < 0.0001], group [F(5,45) = 15.56, P < 0.0001], and group × time [F(27,156) = 5.7, P < 0.0001] (Fig. 1, inset). Doses of 2.5 and 5 mg/kg WIN 55212-2 produced significant hypothermia relative to vehicle (P < 0.0001). Conversely, a lower dose, 1 mg/kg, was ineffective. Consistent with previous studies, the onset of hypothermia was rapid, with a reduction in body temperature observed 15 min post-injection. The hypothermia peaked 45 to 90 min post-injection, with body temperature returning to predrug values thereafter. A dose of 2.5 mg/kg evoked a maximal hypothermia of 2.1 ± 0.2°C 90 min post-injection, whereas 5 mg/kg produced a peak drop in body temperature of 3.1 ± 0.3°C 60 min post-injection.
l-NAME Does Not Produce Significant Hypothermia.Figure 2 illustrates that l-NAME (10–100 mg/kg) does not alter body temperature significantly. A two-way ANOVA revealed that there was not an effect of group, time, or group × time on body temperature (Fig. 2, inset). The median dose of l-NAME, 50 mg/kg, was chosen for drug combination experiments with various doses of WIN 55212-2.
Drug Combinations of WIN 55212-2 and l-NAME Produce Synergistic Hypothermia. Because a dose of 50 mg/kg l-NAME by itself did not alter body temperature significantly, this dose of l-NAME was used in combination experiments with five doses of WIN 55212-2 (0.5, 1, 2, 2.5, or 5 mg/kg). The experimental design, in which one or both of the agents is used in a dose that is devoid of activity, allows a clear analysis of the combination. Temporal profiles of the hypothermia produced by the WIN 55212-2/l-NAME combinations are shown in Fig. 3, A, B, C, D, and E. l-NAME (50 mg/kg) significantly potentiated the actions of all doses of WIN 55212-2 except the highest dose, 5 mg/kg. Thus, we compared the dose-response effect of the active agent, WIN 55212-2, and the dose-response effect of that agent (five doses) in combination with the inactive agent, l-NAME (Fig. 4). The depression in body temperature at 60 min was used as the effect. The two dose-response regression lines (effect on log dose) shown in Fig. 4 were generated from the temporal profiles in Fig. 1 and Fig. 3 using the body temperature drop at 60 min. A simply additive interaction would lead to the same dose-response relation, whereas a significant shift in the combination curve means that an interaction has occurred (Tallarida, 2001). It is seen that there is a pronounced leftward shift in the regression line of the combination. As these lines did not differ significantly in slope (F = 7.29, P < 0.05), it was possible to express this shift in terms of relative potency, R, a value computed with the assistance of Pharm Tools Pro (The McCary Group, Elkins Park, PA). R was found to have a mean of 2.51, with 95% confidence limits (1.31–4.79). This significant leftward shift in the regression line of WIN 55212-2 means that WIN 55212-2 was 2.51-fold more potent in the presence of this nonhypothermic dose of l-NAME, a factor that quantitates the synergism. This value of R, significantly greater than unity, indicates enhanced potency and, thus, synergy for the interaction.
WIN 55212-3 Does Not Produce Significant Hypothermia. WIN 55212-3 (5 mg/kg) did not cause hypothermia as compared with 10% cremophor/saline (Fig. 5). Moreover, the combination of WIN 55212-3 (5 mg/kg) and l-NAME (50 mg/kg) was without effect on body temperature. The lack of effect of WIN 55212-3 confirms that cannabinoid receptors mediated the synergistic hypothermia caused by the combination of WIN 55212-2 and l-NAME.
Discussion
The major finding in the present study is that drug combinations of nonhypothermic doses of l-NAME and WIN 55212-2 produce synergistic hypothermia. The drug interaction was strongly synergistic, with l-NAME increasing the relative potency of WIN 55212-2 by 2.5-fold. Moreover, the inactive enantiomer of WIN 55212-2, WIN 55212-3, did not alter body temperature significantly by itself or in combination with l-NAME. The lack of effect of WIN 55212-3 confirms the synergy caused by the drug combination of WIN 55212-2 and l-NAME and strengthens our observation that inhibiting NOS activity enhances CB1 receptor-mediated hypothermia (Iadecola et al., 1994; Traystman et al., 1995; Roth et al., 1998b).
Cannabinoid agonists produce marked hypothermia by activating CB1 receptors (Compton et al., 1996; Costa et al., 1999; Ledent et al., 1999; Rawls et al., 2002a). The majority of previous studies have demonstrated that l-NAME does not alter body temperature in rodents (Thorat and Bhargava, 1994; Spina et al., 1998; Almeida and Branco, 2001; Benamar et al., 2001, 2002; Kamerman et al., 2002), although hypothermic responses have been reported (Scammell et al., 1996; Kamerman et al., 2002). In our hands, 10 to 100 mg/kg of l-NAME by itself did not affect body temperature. NO does, however, modulate the thermoregulatory actions of other neurotransmitters. l-NAME attenuates the hypothermia produced by insulin, κ-opioids, hypoxia, and arginine vaso-pressin, suggesting that NO production is necessary for the development of hypothermia (Steiner et al., 1998; Almeida and Branco, 2001; Benamar et al., 2002). Conversely, NO generation facilitates the hyperthermic actions of morphine, prostaglandin E2, and lipopolysaccharide (Amir et al., 1991; Lin and Lin, 1996; Minano et al., 1997; Benamar et al., 2001).
The effects of NO-cannabinoid interactions on body temperature have not been investigated extensively. NG-monomethyl-l-arginine, a competitive inhibitor of NOS, did not affect Δ9-tetrahydrocannabinol-evoked hypothermia in mice (Thorat and Bhargava, 1994). Similarly, the acute injection of l-NAME did not alter the hypothermic, analgesic, or cataleptic response to WIN 55212-2 in mice (Spina et al., 1998). Of considerable importance, however, is that the authors reported the hypothermic effect of only one drug combination of l-NAME and WIN 55212-2. Moreover, the dose of WIN 55212, 8 mg/kg, was the highest dose of cannabinoid reported in the Spina et al. (1998) study. Although conjectural, the combination of lower, submaximal doses of WIN 55212-2 with l-NAME may have yielded greater-than-expected hypothermia in that study. It should also be noted that in mice that were tolerant to the pharmacological effects of WIN 55212-2, l-NAME injected once daily 20 min before WIN 55212-2 blocked the development of tolerance to the hypothermic and cataleptic actions but not to the analgesic effect of WIN 55212-2 (Spina et al., 1998). A more recent study demonstrated that Δ9-tetrahydrocannabinol produced antinociception, but not hypothermia or hypolocomotion, in mice lacking neuronal NOS (Azad et al., 2001). It is unclear why Azad et al. (2001) demonstrated that cannabinoids require NO production to produce hypothermia, whereas we observed synergistic hypothermia following the injection of l-NAME and WIN 55212-2. One possibility is that cannabinoids exert distinct effects on different pharmacological end-points, including hypothermia, in animals that are devoid of NOS versus animals in which NO production is blocked acutely (Thorat and Bhargava, 1994; Spina et al., 1998). A combination of factors, such as species of animal, route of injection, type of cannabinoid agonist, method of altering NO production, or dose of cannabinoid agonist, may account for the discrepancies.
Anatomical evidence supports the existence of synergistic interactions between cannabinoid systems and NO. The hypothalamus, particularly the POAH, is a central site of thermoregulation (Boulant et al., 1981). CB1 receptor immunoreactivity is present in the lateral hypothalamic area and the POAH, indicating that intraPOAH neurons express the CB1 receptor protein (Mailleux and Vanderhaeghen, 1992; Tsou et al., 1998; Moldrich and Wenger, 2000). Because hypothalamic deafferentation did not alter cannabinoid receptor binding, it is thought that cannabinoid receptor-expressing neurons are primarily intrinsic to the hypothalamus (Romero et al., 1998). NOS-positive neurons in the ventromedial hypothalamus, another hypothalamic nucleus involved in the regulation of body temperature, also express CB1 receptors (Bredt and Snyder, 1992; Azad et al., 2001). Those data suggest that NO and CB1 systems are functionally linked.
The mechanism for the NO-cannabinoid synergy is unknown. Because the CB1 receptor is located predominantly in the CNS and mediates cannabinoid-evoked hypothermia, cannabinoids appear to suppress body temperature by acting centrally (Fitton and Pertwee, 1982; Compton et al., 1992; Fan et al., 1994; Howlett, 1995; Ovadia et al., 1995; Rawls et al., 2002). The fact that NOS staining occurs on neurons that coexpress CB1 receptors (Bredt and Snyder, 1992; Azad et al., 2001) also suggests a central locus for the synergy between NO and cannabinoid systems. Although speculative, NO release may increase in response to cannabinoid-induced hypothermia, possibly as a compensatory reaction to the marked reduction in body temperature. In this model, WIN 55212-2 evokes hypothermia by activating CB1 receptors, which causes an increase in NO release in brain regions that regulate body temperature. The elevated NO levels attempt to spawn a hyperthermia that counteracts the CB1-mediated hypothermia. Our data suggest that inhibition of NO production by l-NAME abolishes the compensatory effects of NO-evoked hyperthermia, resulting in a cumulative augmentation in the hypothermic response to cannabinoids. Indeed, endogenous cannabinoids stimulate NO release in rat kidneys, invertebrate nerve ganglia, and human immune tissue (Deutsch et al., 1997; Stefano et al., 1997, 2000). Also consistent with our hypothesis, l-NAME potentiates anandamide-induced inhibition of contractile responses in rats, prompting the authors to suggest a compensatory role for endocannabinoids in vascular function in situations where NO synthesis is chronically impaired (Mendizabal et al., 2001).
Another explanation is that CB1 receptor activation by WIN 55212-2 reduces NO release in CNS regions that regulate body temperature. This would lead to a decline in cumulative NO levels and removal of a hyperthermic tone mediated by endogenous NO. The presence of l-NAME may potentiate the hypothermic action of WIN 55212-2 by further diminishing NO production and transmission. The fact that WIN 55212-2 suppresses potassium-evoked neuronal NOS in cerebellar granule cells suggests that CB1 receptor activation attenuates the activation of neuronal NOS and supports the hypothesis that WIN 55212-2 inhibits NO levels (Hillard et al., 1999). Yet another possibility is that l-NAME potentiated cannabinoid-evoked hypothermia by acting at sites outside of the CNS (Nagashima et al., 1994). The sites of action of l-NAME are distributed throughout the body, including brown adipose tissue, where they are responsible for heat production, and vascular smooth muscle, where they promote heat conservation.
In conclusion, we have shown that a cannabinoid agonist and NOS inhibitor interact to evoke synergistic hypothermia. The use of body temperature, a precise metric, provides an analysis that demonstrates synergism and its statistical confirmation. These results may provide the first step in elucidating a mechanism, since the same drug combination may apply to endpoints other than body temperature.
Acknowledgments
We thank Andy Baron and David Kon for assistance in the recording of body temperature data.
Footnotes
-
This study was supported by National Institute on Drug Abuse Grants DA09793, DA00376, DA13429, and T32 DA07237.
-
ABBREVIATIONS: CB1 and CB2, cannabinoid CB1 and CB2 subtypes; NO, nitric oxide; NOS, nitric-oxide synthase; POAH, preoptic anterior nucleus of the hypothalamus; WIN 55212-2, (+)-WIN 55,212-2 or 4,5-dihydro-2-methyl-4(4-morpholinylmethyl)-1-(1-naphthalenyl-carbonyl)-6H-pyrrolo[3,2,1ij]quinolin-6-one; l-NAME, Nω-nitro-l-arginine methyl ester; WIN 55212-3, S-(–)-[2,3-dihydro-5-methyl-3-[(morpholinyl)methyl]pyrrolo[1,2,3-de]-1,4-benzoxazinyl]-(1-napthanlenyl) methanone mesylate; ANOVA, analysis of variance; CNS, central nervous system.
-
DOI: 10.1124/jpet.103.054668.
- Received May 16, 2003.
- Accepted October 24, 2003.
- The American Society for Pharmacology and Experimental Therapeutics