Introduction

Abstract Introduction Materials & Methods Results Discussion References

Arachidonylethanolamide, more commonly known as anandamide, is an ethanolamine derivative of arachidonic acid which was first isolated in porcine brain (Devane et al., 1992). Several lines of evidence have suggested that anandamide may function as an endogenous ligand for the cannabinoid receptor. Anandamide competitively inhibited the specific binding of a radiolabeled cannabinoid probe to synaptosomal membranes, and it produced a dose-dependent inhibition of the electrically evoked twitch response in the mouse vas deferens (Devane et al., 1992). In preliminary behavioral studies, anandamide produced moderate effects similar to that of ⁹-THC (the prototypical psychoactive cannabinoid) after i.p. administration (Fride and Mechoulam, 1993). Subsequently, thorough dose-response analysis also indicated that anandamide produced effects similar to those of ⁹-THC in a tetrad of tests used to predict cannabimimetic activity, including antinociception (as determined in a latency to tail-flick evaluation), hypothermia, hypomotility and catalepsy in mice after i.v., i.p. and intrathecal administration (Smith et al., 1994). In general, the effects of anandamide occurred with a rapid onset and with a rather short duration of action. Also, it was 1.3 to 18 times less potent than ⁹-THC in all behavioral assays.

Binding studies have demonstrated that anandamide interacts with the CB1 cannabinoid receptor with a K_D of approximately 100 nM (Childers et al., 1994; Smith et al., 1994). Because anandamide can be hydrolytically cleaved by a membrane-bound enzyme in brain tissue to arachidonic acid and ethanolamine (Deutsch and Chin, 1993), PMSF (50 µM) has typically been included in the incubation medium of in vitro binding assays to obtain true estimates of receptor affinity for anandamide (Childers et al., 1994; Hillard et al., 1995; Smith et al., 1994) and related analogs (Adams et al., 1995). PMSF is a nonspecific inhibitor of various proteases and other enzymes, including acetylcholinesterase, palmityl coenzyme A deacylase, arylsulfatase A, chymotrypsin and trypsin (James, 1978; Moss and Fahrney, 1978). Generally, PMSF acts by sulfonylating the hydroxyl groups of active site serine residues of enzymes, which causes an irreversible inhibition. Although not yet identified, the enzyme responsible for the metabolism of anandamide would be considered an amidohydrolase. Metabolism of anandamide is greatest in the liver, also occurs to a significant degree in brain tissue, but only occurs to a much lesser extent in other tissues (Desarnaud et al., 1995). This amidohydrolase can also be inhibited by p-bromophenacyl bromide (a histidine-alkylating agent), but not by other nonselective peptidase inhibitors such as ethylenediaminetetraacetic acid, bacitracin and o-phenanthroline (Desarnaud et al., 1995). PMSF rapidly crosses the blood-brain barrier (Turini et al., 1969), so in vivo administration would be expected to inhibit anandamide metabolism in brain tissue as well as in the periphery.

Very few data are available concerning the in vivo metabolism of anandamide. It is clear that it possesses a very short duration of action on various pharmacological measures (Smith et al., 1994). This is assumed to be caused by rapid metabolic inactivation. If true, then it is also likely that the decreased in vivo potency of anandamide, relative to that of ⁹-THC, is also caused by rapid metabolism. However, in such a scenario the maximum in vivo effect that could be produced by anandamide should not change, and the interaction of anandamide with the CB1 receptor (as evidenced by the shape of the dose-response curve) should not be altered by manipulations of metabolism. This research evaluated the in vivo pharmacological effects of anandamide in the mouse model of cannabinoid pharmacological activity after the in vivo administration of PMSF. This agent should attenuate the metabolism of anandamide thereby providing indirect evidence as to whether or not the weak potency of anandamide is caused by rapid metabolic inactivation.

	Materials and Methods

Abstract Introduction Materials & Methods Results Discussion References

Male ICR mice (Harlan Laboratories, Dublin, VA) weighing 18 to 25 g were used in all experiments. The mice were maintained on a 14:10 hr light/dark cycle with free access to food and water. Anandamide, 2-methylarachidonly-(2-fluoroethyl)amide and SR141716A were obtained from Dr. Raj K. Razdan (Organix, Inc., Woburn, MA) and dissolved in 1:1:18 (emulphor/ethanol/saline) for in vivo administration. Emulphor (EL-620, a polyoxyethylated vegetable oil, GAF Corporation, Linden, NJ) is currently available as Alkmulphor. Anandamide and SR141716A injections were administered i.v. (tail vein) at a volume of 0.1 ml/10 g b.wt. PMSF was obtained from Sigma Chemical (St. Louis, MO), dissolved in sesame oil and administered i.p. at a volume of 0.1 ml/10 g b.wt. PMSF was always administered 10 min before i.v. anandamide or vehicle injections.

Mice were acclimated to the evaluation room overnight without interruption of food or water. After i.v. anandamide or vehicle administration each animal was evaluated as follows: tail-flick latency (antinociception) response at 5 min and spontaneous (locomotor) activity at 5 to 15 min; or core (rectal) temperature at 5 min and ring-immobility (catalepsy) at 5 to 10 min, as described elsewhere (Smith et al., 1994). Where indicated, the cannabinoid antagonist SR141716A was administered (i.v.) 1 min before the i.p. administration of PMSF.

Spontaneous activity. Inhibition of locomotor activity was accomplished by placing mice in individual activity cages (6.5 × 11 inches) and recording interruptions of the photocell beams (16 beams per chamber) for a 10-min period with a Digiscan Animal Activity Monitor (Omnitech Electronics Inc., Columbus, OH). Activity in the chamber was expressed as the total number of beam interruptions.

Tail-flick latency. Antinociception was assessed by the tail-flick procedure. The heat lamp of the tail-flick apparatus was maintained at an intensity sufficient to produce control latencies of 2 to 3 sec. Control values for each animal were determined before drug administration. Mice were then re-evaluated after drug administration and latency (sec) to tail-flick response was recorded. A 10-sec maximum was imposed to prevent tissue damage. The degree of antinociception was expressed as the % MPE which was calculated as:

(1)

Core temperature. Hypothermia was assessed by first measuring base-line core temperatures before drug treatment with a telethermometer (Yellow Springs Instrument Co., Yellow Springs, OH) and a rectal thermistor probe inserted to a depth of 25 mm. Rectal temperatures were then measured after drug administration so that the temperature difference (°C) between values could be calculated for each animal.

Immobility. Catalepsy was determined by a ring-immobility procedure. At 90 min after injection, mice were placed on a metal ring (5.5 cm in diameter) that was attached to a stand at a height of 16 cm. The amount of time (sec) that the mouse spent motionless during a 5-min test session was recorded. The criterion for immobility was the absence of all voluntary movements (excluding respiration, but including whisker movement). The immobility index was calculated as: Mice that fell or actively jumped from the ring were allowed five such escapes. After the fifth escape, the test for that animal was terminated and immobility was calculated as a percentage of time that it remained on the ring before being discontinued. Data from mice failing to remain on the ring at least 2.5 min were not included.

Statistical analysis. Statistical analysis of all in vivo data was performed by ANOVA with Bonferroni/Dunn post hoc for comparison with vehicle with use of the StatView statistical package (Brainpower, Inc., Agoura Hills, CA). Differences were considered significant at the P < .05 level. ED₅₀ values were determined by ALLFIT, a program for the simultaneous curve fitting of a family of sigmoidal curves.

	Results

Abstract Introduction Materials & Methods Results Discussion References

Agonist effects of PMSF. Although the primary objective was to determine the influence of PMSF pretreatment on anandamide's pharmacological potency, it was necessary to determine whether PMSF exhibited agonist effects. To mimic the treatment regimen involving the dual injections of PMSF and anandamide, evaluation of the effects of PMSF in the absence of anandamide was conducted by the dual-injection approach. All animals received an i.p. injection of PMSF followed by a 10-min period before the i.v. administration of vehicle. Thus, the pharmacological effects observed in these studies (figs. 1, 2, 3, table 1) correspond exactly to what would be the anticipated contribution of PMSF to the total response observed in the subsequent PMSF plus anandamide studies.

View larger version (17K): [in this window] [in a new window]   Fig. 1.   The antinociceptive effects of PMSF administered i.p. to mice and tested 15 min later for tail-flick responsiveness. The results are presented as means ± S.E.M. for at least six mice per group.

View larger version (13K): [in this window] [in a new window]   Fig. 2.   The hypothermic effects of PMSF were determined 15 min after i.p. administration to mice. The results are presented as means ± S.E.M. for at least six mice per group.

View larger version (17K): [in this window] [in a new window]   Fig. 3.   Immobility produced by PMSF was determined for the 15- to 20-min period after i.p. administration to mice. The results are presented as means ± S.E.M. for at least six mice per group.

Effect of SR141716A on PMSF. To determine whether the pharmacological effects of PMSF could be prevented by the cannabinoid receptor antagonist SR141716A, mice were treated i.v. with this antagonist 10 min before the i.p. administration of PMSF (table 2). The dose of SR141716A chosen was known to be completely effective in antagonizing the effects of either WIN-55,212-2 or ⁹-THC by this same protocol. The cannabinoid antagonist failed to reduce the pharmacological effect of 100 mg/kg PMSF.

View this table: [in this window] [in a new window]   TABLE 2 Effect of SR141716A on the pharmacological activity of PMSF Mice were injected i.v. with SR 141716A 10 min before the i.p. administration of either PMSF or its vehicle (sesame oil) before testing as described under "Materials and Methods." Values (mean ± S.E.M., n = 6 per group) for tail-flick antinociception (% MPE), inhibition of locomotor activity (% inhibition), decrease in core temperature (°C) and ring-immobility catalepsy (% immobility) after i.p. administration of PMSF. PMSF alone (in the absence of antagonist) was statistically different (P < .05) from both vehicle control (1:1:18 vehicle + sesame oil) (data not shown) and the drug control (SR141716A + sesame oil) but not from the SR 141716A/PMSF combination.

Effect of PMSF on anandamide. To evaluate the effect of PMSF on the pharmacological response to anandamide, the PMSF pretreatment dose chosen was 30 mg/kg. This dose of PMSF failed to produce a statistically significant effect on any measure, is much less than either the 100 or 200 mg/kg doses required to produce any pharmacological action and is sufficiently small that even in the tail-flick antinociception measure (see fig. 1) the response would be minimal (<20%). Thus, groups of mice were pretreated with either PMSF (30 mg/kg) or vehicle (1:1:18) before treatment with various doses of anandamide. The dose-response curves demonstrating the effect of PMSF on anandamide-mediated responses are demonstrated in figures 4, 5, 6 and summarized in table 3.

View larger version (18K): [in this window] [in a new window]   Fig. 4.   Enhancement of anandamide-induced antinociception by a 10-min i.p. pretreatment with PMSF (30 mg/kg). The results are presented as means ± S.E.M. for at least six mice per group.

View larger version (20K): [in this window] [in a new window]   Fig. 5.   Enhancement of anandamide-induced hypomotility by a 10-min i.p. pretreatment with PMSF (30 mg/kg). The results are presented as means ± S.E.M. for at least six mice per group.

View larger version (20K): [in this window] [in a new window]   Fig. 6.   Enhancement of anandamide-induced immobility by a 10-min i.p. pretreatment with PMSF (30 mg/kg). The results are presented as means ± S.E.M. for at least six mice per group.

Effect of PMSF on 2-methylarachidonly-(2-fluoroethyl)amide.

View larger version (13K): [in this window] [in a new window]   Fig. 7.   Failure of PMSF to enhance the pharmacological activity of the metabolically stable 2-methylarachidonly-(2-fluoroethyl)amide for inhibition of spontaneous activity (A) and antinocicepton (B). Mice were pretreated i.p. with either vehicle or PMSF (30 mg/kg) 10 min before the administration of the analog. The animals were tested for tail-flick response 5 min after the injection of the analog and spontaneous activity for 5 to 15 min. The results are presented as means ± S.E.M for at least six mice per group.

	Discussion

Abstract Introduction Materials & Methods Results Discussion References

The data presented in this manuscript are consistent with the hypothesis that the in vivo administration of amidohydrolase inhibitors like PMSF can prevent the rapid in vivo metabolism of anandamide. The result is a parallel shift to the left in the dose-response curves for anandamide on pharmacological measures such as antinociception, the inhibition of locomotor activity and the production of catalepsy. PMSF administration had no effect on the hypothermic effects of anandamide, however, which perhaps suggests that hypothermia, unlike the other measures of this tetrad of pharmacological measures, is not produced by anandamide directly, or perhaps that anandamide is still rapidly metabolized within the brain region that provides the neural substrates for this response. This could be caused by either incomplete inhibition of the unidentified amidohydrolase or by the existence of an isozyme or novel enzyme that can also metabolize anandamide, but there are no data to support this contention. The shift in potency of anandamide by PMSF pretreatment varied slightly, but was 5 to 10 times that of anandamide alone. The result was that the ED₅₀ values obtained for anandamide fell into a 1 to 4 mg/kg range, almost exactly that observed for ⁹-THC. The notion that PMSF enhances the potency of anandamide by diminishing its metabolism was supported by the findings that PMSF had little effect on the potency of 2-methylarachidonly-(2-fluoroethyl)amide. Although PMSF has a dramatic effect on the binding affinity of anandamide in receptor binding assays, it has little influence on that of 2-methylarachidonly-(2-fluoroethyl)amide (Adams et al., 1995).

The dose of PMSF used in these studies to manipulate the pharmacological effect of anandamide was 30 mg/kg. Although PMSF produced effects on all four pharmacological tests, it did so only at doses of 100 mg/kg or greater. The only measure at which PMSF produced dose-responsive effects below 100 mg/kg was in antinociception, which necessitated the choice of the lower dose of 30 mg/kg for interaction studies with anandamide. These results are similar to the general depression, impairment of righting reflexes and unresponsiveness to noxious stimuli described by others (Turini et al., 1969).

Additionally, the agonist activity of PMSF is not likely caused by actions at the cannabinoid receptor, because the antagonist SR141716A did not alter the PMSF responses. Thus, it seems unlikely that PMSF administration is either increasing levels of endogenous anandamide or another as-yet-unidentified endogenous cannabinoid ligand, or acting directly on the receptors. On the other hand, recent studies in our laboratory have failed to demonstrate SR 141716A-blockade of some of anandamide's pharmacological effects in mice (Adams, I. B., Compton, D. R. and Martin, B. R., unpublished observations). Therefore, it will be important to actually measure endogenous levels of anandamide in blood and brain after PMSF administration before a definitive conclusion can be reached. Such an endeavor is beyond the scope of the present investigation.

A possibility that cannot be ruled out is that PMSF enhancement of anandamide is caused by the inhibition of other enzymes. Inhibition of esterases, thus increasing levels of endogenous acetylcholine, could be responsible for some the effects observed after the administration of PMSF alone. However, a PMSF dose of 85 mg/kg produces only about 30% inhibition of brain acetylcholinesterase activity, and this effect occurs 18 hours after drug administration (Moss et al., 1985). Thus, it seems unlikely the interaction of 30 mg/kg of PMSF with anandamide is compromised by changes in esterase activity, but it does seem plausible that the effects observed at 100 to 300 mg/kg of PMSF could be caused by esterase inhibition.

Although the primary emphasis of the present investigation was directed toward inactivation of anandamide via hydrolysis, both inactivation and activation via other metabolic pathways is a possibility. Relatively little is known regarding the metabolic profile of anandamide. One report revealed that anandamide can serve as a substrate for brain lipoxygenase with the resultant product being 12-hydroxyanandamide (Hampson et al., 1995). Studies in our own laboratories have demonstrated that there is a discordance between the time courses of brain levels of exogenously administered anandamide and pharmacological effects (Willoughby et al., 1997). Essentially, brain levels of anandamide fell despite the persistence of pharmacological effects implying the formation of active metabolites. One cannot rule out the possibility that PMSF-blockade of anandamide results in increased formation of active metabolites, the actions of which may or may not be antagonized by SR 141716A.

In conclusion, the pharmacological effects of the enzyme inhibitor PMSF generally occur in the range of 100 mg/kg of drug or higher, with the exception of the antinociceptive properties which occur at lower doses. These effects cannot be attributed to direct actions at the cannabinoid receptor. An inactive dose of PMSF significantly increased the pharmacological activity of anandamide and produced parallel shifts to the left in the dose-response curves for all anandamide-mediated measures evaluated, except for hypothermia. The use of PMSF or other specific amidohydrolase inhibitors should be used to evaluate the in vivo potencies of anandamide analogs, just as for in vitro studies, to establish an appropriate correlation between in vivo potency and in vitro affinity to the cannabinoid receptor. It seems likely that the correlation between in vivo potency and receptor affinity for a variety of anandamide analogs could be enhanced with the inclusion of PMSF (Adams et al., 1995). These studies underscore the importance of metabolism in the expression of pharmacological actions of anandamide.