Inhibition of Anandamide Hydrolysis by the Enantiomers of Ibuprofen, Ketorolac, and Flurbiprofen

Archives of Biochemistry and Biophysics Vol. 362, No. 2, February 15, pp. 191–196, 1999 Article ID abbi.1998.1025, available online at http://www.idealibrary.com on

Inhibition of Anandamide Hydrolysis by the Enantiomers of Ibuprofen, Ketorolac, and Flurbiprofen Christopher J. Fowler,* ,1 Ulrika Janson,* Randolph M. Johnson,† ,2 Go¨ran Wahlstro¨m,* Anders Stenstro¨m,* Åke Norstro¨m,‡ and Gunnar Tiger* *Department of Pharmacology and †Department of Clinical Pharmacology, Umeå University, SE-901 87 Umeå, Sweden; and ‡Department of Molecular and Cellular Biochemisitry, Center for Biological Research, Roche Bioscience, Palo Alto, California 94304

Received June 22, 1998

The endogenous cannabimimetic anandamide is hydrolyzed by a fatty acid amide hydrolase to yield arachidonic acid and ethanolamine. In the present study, the regional distribution of the activity and its sensitivity to inhibition by the enantiomers of ibuprofen, ketorolac, and flurbiprofen has been investigated. The rate of [ 3H]anandamide hydrolysis was found in both 7-week-old and 90-week-old rats to be in the order hippocampus > cerebral cortex > cerebellum > striatum ' midbrain, with higher rates of hydrolysis for the 7-week-old rats than for the 90-week-old rats. In whole brain (minus cerebellum), the R(2)-enantiomer of ibuprofen was a mixed-type inhibitor of anandamide hydrolysis and was ;2–3 times more potent than the S(1)-enantiomer, IC 50 values of 230 and 750 mM, respectively, being found. A similar pattern of inhibition of anandamide hydrolysis was seen when intact C6 rat glioma cells were used. Ketorolac inhibited rat brain anandamide hydrolysis, with IC 50 values of 50, 440, and 80 mM being found for the R-, S-, and R,Sforms, respectively. The IC 50 value for R-flurbiprofen (60 mM) was similar to the IC 50 value for the S-enantiomer (50 mM). These data demonstrate that there is no dramatic enantiomeric selectivity of NSAID compounds as inhibitors of fatty acid amide hydrolase enzyme(s) responsible for the hydrolysis of anandamide. The enantiomers of flurbiprofen and R-ketorolac are the most potent NSAID inhibitors of fatty acid amide hydrolase yet reported. © 1999 Academic Press Key Words: anandamide; fatty acid amide hydrolase; ibuprofen; ketorolac; flurbiprofen.

Anandamide (arachidonyl ethanolamide), an endogenous agonist at cannabinoid CB1 receptors (1) is metabolized to arachidonic acid by a fatty acid amide hydrolase (2, 3). The enzyme can hydrolyse several fatty acid amides, such as oleamide and palmitoylethanolamide (3, 4), and can potently be inhibited by the substrate analogue arachidonyl trifluoromethyl ketone (5) and by the serine protease inhibitor phenylmethylsulphonyl fluoride (PMSF) (2). The enzyme can differentiate between steric configurations of substrates. Thus, Lang et al. (6) found that the S-form of methanandamide was hydrolyzed by rat brain homogenates 10-fold faster than a related R-enantiomer. However, the ability of compounds to inhibit fatty acid amide hydrolase in a stereoselective manner has not previously been investigated. The nonsteroidal antiinflammatory drug (NSAID) ibuprofen has recently been shown to inhibit the hydrolysis of anandamide at pharmacologically relevant concentrations, whereas acetylsalicylic acid and acetaminophen were without effect (7, 8). Although NSAID compounds such as ketoprofen, fenoprofen, and sulindac have been evaluated with respect to their abilities to block anandamide metabolism, the effects of highly potent agents such as ketorolac and flurbiprofen are not known. In addition, ibuprofen, ketorolac, and flurbiprofen are optically active compounds, thus allowing an investigation into the steric requirements of the inhibition. In consequence, in the present study, the inhibitory effects of the enantiomers of these three NSAID compounds upon [ 3H]anandamide metabolism have been investigated. MATERIALS AND METHODS

1

To whom correspondence and reprint requests should be addressed. Fax: [46]90 785 2752. E-mail: [email protected]. 2 Present address: DURECT Corporation, 10240 Bubb Road, Cupertino, CA 95014. 0003-9861/99 $30.00 Copyright © 1999 by Academic Press All rights of reproduction in any form reserved.

Materials. Arachidonyl-ethanolamide-[1- 3H] ([ 3H]anandamide) specific activity 30 Ci/mmol, was custom synthesized by American Radiolabelled Chemicals Inc. (St. Louis, MO) RS-, R-, and S-ketorolac (5-benzoyl-2,3-dihydro-1H-pyrrolizine-1-carboxylic acid), and the 191

192

FOWLER ET AL.

two enantiomers of flurbiprofen (2-fluoro-a-methyl-4-biphenylacetic acid) were obtained from Roche Bioscience (Palo Alto, CA). The enantiomers of ibuprofen (a-methyl-4-(2-methylpropyl)benzeneacetic acid) were obtained from Research Biochemicals International (Natick, MA). Ketorolac, flurbiprofen, and ibuprofen were dissolved in ethanol prior to use. In general, stock solutions of 80 mM of the NSAIDs were prepared, and diluted in a mixture 1:1 (v:v) of ethanol and 10 mM Tris–HCl (pH 7.6) containing 1 mM EDTA on the day of assay. Hams F-10 medium, fetal bovine serum (FBS) and penicillinstreptomycin were obtained from Gibco BRL, Life Technologies (Sweden). 3

Assay of [ H]anandamide hydrolysis in rat brain homogenates. The Sprague–Dawley rats used in the present study were obtained from Mo¨llegaard Ll (Skensved, Denmark) (7-week-old and 90-weekold animals, data shown in Fig. 2 and Table I) and from the University of Umeå (adult rats , 1 year of age, other data). Either whole brain (minus cerebellum) or individual brain regions, dissected as described by Glowinski and Iversen (9), were used, as indicated. The experiments were approved by the regional animal research ethics committee. A method based upon that of Omeir et al. (10) was used to assay the hydrolysis of [ 3H]anandamide in the homogenates (8). Briefly, the brain samples were homogenised in 10 mM Tris–HCl (pH 7.6) containing 1 mM EDTA, and aliquots were stored frozen at 270 °C until used for assay. For the data shown in Fig. 1, 5–10 mg protein/ assay was used for 1–3 mM [ 3H]anandamide concentrations and 40 –50 mg protein/assay for 20 – 40 mM [ 3H]anandamide concentrations. For individual brain regions, protein concentrations in the ranges described in the legend to Table I were used. The samples were incubated with the test NSAID compound or with the same amount of ethanol carrier [at a concentration which did not affect enzyme activity] and [ 3H]anandamide. In some experiments, a preincubation period between ibuprofen enantiomers and the homogenates (0 –90 min at 37°C) was used prior to addition of [ 3H]anandamide. The preincubation times are indicated in these experiments. The assay volume was 200 ml (25 ml of the NSAID dissolved in ethanol/buffer, 25 ml [ 3H]anandamide, 150 ml homogenate). Reactions were stopped by the addition of chloroform:methanol (1:1 v/v). The samples were vortex mixed, and the aqueous layer containing the [ 3H]ethanolamine product was collected and assayed for tritium content by liquid scintillation spectroscopy with quench correction. Blanks were determined in the presence of 1.5 mM PMSF (Sigma, dissolved in either butanol or ethanol, essentially identical blanks being found with either solvent). The validity of the assay has been demonstrated elsewhere (8). Assay of [ 3H]anandamide hydrolysis in intact C6 glioma cells. C6 glioma cells were obtained from the American Type Culture Collection and used between passage 54 and 76. The cells (8 –11 3 10 6 cells) were cultured for 24 h in 75-cm 2 flasks in Ham‘s F10 medium supplemented with 25 mM Hepes buffer, 2 mM L-glutamine, 100 IU/ml penicillin 1 100 mg/ml streptomycin, and 10% fetal bovine serum. After the incubation, the cells were harvested (by trypsinization followed by centrifugation) and resuspended in serum-free F10 medium. The cells were again repelleted and carefully resuspended in 2 ml of a modified assay buffer (10 mM Tris–HCl containing 1 mM EDTA and 0.31 M sucrose [to preserve isotonicity], pH 7.6, buffer prewarmed to 37°C). After adjustment of the buffer volume to 11 ml, aliquots (150 ml) were incubated, when appropriate, for 60 min at 37°C with R(2)-, S(1)-ibuprofen, or the ethanol carrier. [ 3H]Anandamide (2 mM final concentration, in the modified assay buffer containing BSA) was then added, and the samples incubated for a further 15 min at 37°C, unless otherwise stated, after which the assay was completed as described for the rat brain samples. Determination of IC 50 values, K M and V max values and statistical evaluation of data. In order to determine IC 50 values for the inhibition curves, mean data (in the 10 –90% contiguous percentage activity remaining range, to avoid bias at high and low levels of inhibition) were plotted as log 10(inhibitor concentration) vs

log 10((100% 2 % activity remaining)/% activity remaining). The regression lines where then used for calculation of IC 50 values. ANOVAs were conducted using the Statview computer programme (Abacus Concepts Inc., Berkeley, CA). K M and V max values for mean control data over a substrate concentration of 1–3 mM were calculated using the Direct Linear Plot analysis of Eisenthal and CornishBowden [11] using the Enzyme Kinetics v 1.4 software package, Trinity Software (Campton, NH).

RESULTS

Inhibition of [ 3H]Anandamide Hydrolysis by the Enantiomers of Ibuprofen in Rat Brain Homogenates. Concentration-response curves for the inhibition of 2 mM [ 3H]anandamide hydrolysis in homogenates of whole brain (minus cerebellum) by the two ibuprofen enantiomers are shown in Fig. 1A. The R(2)-enantiomer was ;3 times more potent than the S(1)-enantiomer with IC 50 values calculated by Hill analyses of the mean data of 230 and 750 mM, respectively, being found. These values are in line with the IC 50 value of 270 mM previously found for racemic ibuprofen at this substrate concentration (8). The observed IC 50 values were not affected by preincubation of the ibuprofen enantiomers with the whole brain (minus cerebellum) homogenates for up to 90 min at 37°C prior to addition of substrate. Thus, after preincubation times of 0, 30, 60, and 90 min, IC 50 values (calculated from analyses of mean data for three experiments with 200, 500, and 1000 mM inhibitor concentrations) were 320, 380, 360, and 290 mM, respectively, for R(2)ibuprofen. The corresponding values for S(1)ibuprofen were 580, 710, 450, and 610 mM, respectively. The mode of inhibition of anandamide hydrolysis by R(2)ibuprofen is shown in Fig. 1B. At low concentrations of substrate (1–3 mM), control samples metabolized [ 3H]anandamide with a K M value of 0.76 mM and a V max value of 1.7 nmol.(mg protein) 21.min 21. R(2)ibuprofen changed the substrate concentration– dependence curves at these low substrate concentrations from a saturable to a linear relation, consistent with an increased K M value. The separate experiments conducted at high concentrations of substrate (20 – 40 mM, shown in Fig. 1B) indicate a reduction in the V max value. These data would suggest that the inhibition produced by R(2)ibuprofen is of mixed-type. Regional Variation of [ 3H]Anandamide Hydrolysis across the Rat Brain The hydrolysis of 2 mM [ 3H]anandamide was determined at different protein concentrations in five regions of rat brain. In both 7- and 90-week-old rats, the specific activity was in the order hippocampus . cerebral cortex . cerebellum . striatum ' midbrain (Table I). The specific activity was, however, lower in the 90-week-old

NSAID INHIBITION OF ANANDAMIDE HYDROLYSIS

193

FIG. 1. (A) Inhibition of 2 mM [ 3H]anandamide hydrolysis by R(2)- (F, Œ) and S(1)- (E, ‚) ibuprofen. Homogenates of whole brain minus cerebellum were used. The triangles indicate data from separate experiments with a higher carrier (ethanol) concentration than used for the data presented by the circles. (B) [ 3H]Anandamide hydrolysis in the absence (E) and presence of 200 mM (Œ) and 500 mM (ƒ) R(2)-ibuprofen. The data points obtained at 20 – 40 mM [ 3H]anandamide are from a separate experiment to those for the 1–3 mM concentration range. Shown are means 6SEM (when not enclosed by the symbols), n 5 3– 4.

rats than in the 7-week-old rats (P , 0.05 for hippocampus, midbrain, and cerebellum; P , 0.06 for cortex and striatum). In hippocampal, cerebellar, and striatal regions from young rats, the specific activity did not vary with substrate concentration over a [ 3H]anandamide concentration range of 3–6 mM (Fig. 2). The effects of the enantiomers of ibuprofen upon the hydrolysis of 2 mM [ 3H]anandamide were determined in different regions of the rat brain. There were no dramatic regional differences in inhibitor sensitivities (Table II). TABLE I

Comparison of Initial Rates of [ 3H]Anandamide Hydrolysis in Different Brain Regions for 7-Week-Old and 90-Week-Old Rats 21

21

Specific activity (nmol[(mg protein) ]min ) Brain region

7-week-old rats

90-week-old rats

Cerebral cortex Hippocampus Striatum Midbrain Cerebellum

1.36 6 0.037 2.21 6 0.12 0.46 6 0.028 0.51 6 0.031 0.87 6 0.066

1.29 6 0.075 1.59 6 0.14 0.42 6 0.027 0.40 6 0.054 0.70 6 0.064

Note. Initial velocities were calculated from measurements over a range of protein concentrations (0 –3 mg protein/assay for hippocampus; 0 – 6 mg protein/assay for cerebral cortex and cerebellum; and 0 –12 mg protein/assay for striatum and midbrain). Data are means 6SEM, n 5 10 (7-week-old rats) or 8 (90-week-old rats). Two-way ANOVA for repeated measures indicated a significant contribution by protein content in all cases (F 3,48 . 157, P , 0.0001). The contributions by age (F 1,16) were: cortex, 4.3 (P 5 0.054); hippocampus, 13.4 (P , 0.01); striatum, 4.23 (P 5 0.056); midbrain, 16.6 (P , 0.001); cerebellum, 4.52 (P , 0.05).

Inhibition of Anandamide Hydrolysis by Ibuprofen Enantiomers in Intact C6 Glioma Cells Incubation of C6 cells with 2 mM [ 3H]anandamide resulted in a time-dependent production of [ 3H]ethanolamine (Fig. 3, inset). Cells were preincubated with the ibuprofen enantiomers for 60 min at 37°C prior to addition of [ 3H]anandamide and a further incubation for 15 min. The ibuprofen enantiomers both inhibited the metabolism of [ 3H]anandamide with IC 50 values of 230 and 340 mM for R(2)- and S(1)-ibuprofen, respectively, being found (Fig. 3). The cell viability (assessed by trypan blue exclusion) of C6 cells incubated in the assay buffer for 60 min was found in a single experiment to be 74%. Inhibition of Rat Brain Anandamide Hydrolysis by the Enantiomers of Ketorolac and Flurbiprofen Both ketorolac and flurbiprofen were found to inhibit the metabolism of anandamide in the rat brain (minus cerebellum) homogenates. The IC 50 value for the Risoform of ketorolac (50 mM) was ;9-fold lower than the corresponding IC 50 value for the S-isoform (440 mM) (Fig. 4A). The IC 50 value for the racemate was 80 mM. In contrast, the R- and S-isoforms of flurbiprofen had rather similar IC 50 values (60 and 50 mM, respectively) (Fig. 4B). DISCUSSION

The present study demonstrates that the two enantiomers of ibuprofen inhibit [ 3H]anandamide hydrolysis with slightly (but consistently) different potencies. In contrast to the clinical efficacy of the compounds and

194

FOWLER ET AL.

FIG. 2. Hydrolysis of 3– 6 mM [ 3H]anandamide by homogenates from hippocampus (HIPP), cerebellum (CERE), and striatum (STRI). Data are means 6SEM, n 5 3– 4. Three-way ANOVA indicated significant contributions of incubation time (F 2,107 5 94, P , 0.0001), brain region (F 2,107 5 97, P , 0.0001), and incubation time 3 brain region (F 4,107 5 5.1, P , 0.001) but not either [ 3H]anandamide concentration (F 3,107 5 0.71, P . 0.5) or any other interaction (P . 0.4).

the effects upon prostaglandin synthesis in vivo, where the S(1)-form is regarded as the active isomer (see, e.g., Ref. (12)), the R(2)-form of ibuprofen was slightly more potent as an inhibitor of [ 3H]anandamide metabolism. The data investigating the inhibition of a range of [ 3H]anandamide concentrations by R(2)-ibuprofen suggested the inhibition to be of mixed-type, a finding consistent with the mixed-type inhibition seen with racemic ibuprofen, where K i and K9 i values of 82 and 1420 mM were found (8). A subsequent study undertaken using [ 3H]palmitoylethanolamide (concentration range 1–10 mM) as substrate, confirmed the mixedtype inhibition for R(2)-ibuprofen and gave K i and K9 i

values of 88 and 720 mM, respectively (G. Tiger and C. J. Fowler, unpublished results). This would suggest that the mode of inhibition found with the racemate is unlikely to be due to the presence of two enantiomers with different modes of inhibition. This is of particular relevance in vivo, where interconversion of enantiomeric forms (in the direction R(2)-ibuprofen 3 S(1)ibuprofen) occurs (13–15). The variation of [ 3H]anandamide hydrolysis across the rat brain found in the present study for both 7-week-old and 90-week-old rats is consistent with the data of Hillard et al. (16) who used an assay separating [ 14C]arachidonic acid from [ 14C]anandamide by thin layer chroma-

TABLE II

Inhibition of 2 mM [ 3H]Anandamide Hydrolysis by Ibuprofen Enantiomers in Different Regions of the Rat Brain Percentage activity remaining at ibuprofen concentration of: Enantiomer Cerebral cortex (2)Ibuprofen (1)Ibuprofen Hippocampus (2)Ibuprofen (1)Ibuprofen Striatum (2)Ibuprofen (1)Ibuprofen Midbrain (2)Ibuprofen (1)Ibuprofen Cerebellum (2)Ibuprofen (1)Ibuprofen

1000

IC 50 (mM)

50

100

200

300

500

92 6 5 90 6 6

77 6 4 80 6 7

42 6 7 60 6 8

47 6 6 48 6 4

18 6 6 28 6 2

13 17

6 4 6 4

210 270

98 6 4 88 6 6

83 6 6 81 6 6

57 6 6 76 6 7

39 6 8 69 6 4

40 6 5 43 6 9

17 29

6 10 6 5

290 470

90 6 8 108 6 11

79 6 5 93 6 1

56 6 3 78 6 6

41 6 6 67 6 5

28 6 5 72 6 19

6 14

6 4 6 3

250 490

84 6 7 93 6 7

69 6 8 81 6 9

48 6 7 44 6 7

25 6 7 31 6 12

12 6 7 18 6 12

6 6 4 0.9 6 5

160 200

94 6 6 92 6 7

80 6 5 104 6 11

65 6 5 85 6 10

41 6 5 59 6 5

34 6 5 47 6 6

6 7 6 3

290 450

18 19

Note. Data are means 6SEM, n 5 3– 8. Both young Mo¨llegaard and Umeå rats were used for these experiments.

NSAID INHIBITION OF ANANDAMIDE HYDROLYSIS

FIG. 3. Hydrolysis of 2 nM [ 3H]anandamide by intact C6 glioma cells. The inset shows a single experiment of the [ 3H]ethanolamine formed as percentage of added substrate. The main graph shows the percentage activity remaining in the presence of R(2)- (F) and S(1)- (E) ibuprofen (means 6SEM, n 5 3). IC 50 values calculated from the mean data were 230 and 340 mM for R(2)- and S(1)-ibuprofen, respectively.

tography. These authors found hydrolysis rates in the order hippocampus . cerebellum ' cerebral cortex . hypothamus . white matter ' striatum ' brain stem, a regional variation that correlated reasonably well with the cannabinoid receptor density measured with the radioligand [ 3H]CP55,940 (16). Desarnaud et al. (17), on the other hand, found a greater activity in the striatum than the cerebellum or cortex. The variation across the brain found in the present study is most likely the result of different fatty acid amide hydrolase concentrations in the different regions, rather than to the selective expression of different fatty acid amide hydrolase isoforms, since the properties of the [ 3H]anandamide hydrolyzing enzymes (potencies of inhibition by ibuprofen enantiomers, saturation at substrate concentrations $3 mM) were broadly similar in the different regions. Given, however, that fatty acid amide hydrolase is expressed in different sub-

195

cellular fractions (4, 16, 17), and that membrane microenvironments are likely to be different in different subcellular and regional fractions, further work is required before any firm conclusions can be drawn as to the pattern of isoform expression across the brain. In addition to exploring the inhibitory properties of ibuprofen enantiomers, the present study provides data on the potencies of ketorolac and flurbiprofen as inhibitors of anandamide metabolism. The R-enantiomer of ketorolac was ;9 times more potent than the S-form, which is in contrast to the ability of the compound to inhibit cyclooxygenase and to exert antinociceptive effects, where the S-form is the most active (18, 19). The flurbiprofen enantiomers and R-ketorolac, with IC 50 values ;50 mM, are the most potent inhibitors of anandamide metabolism yet found among NSAID compounds, although these potencies are considerably lower than reported for anandamide analogues such as arachidonyl trifluoromethyl ketone (5). The therapeutic importance of the inhibition of anandamide metabolism by the NSAIDs is unclear. When compared with inhibitors of fatty acid amide hydrolase such as arachidonyl trifluoromethyl ketone (5), and with the potencies of NSAIDs as inhibitors of cyclooxygenase in intact cells (20), the inhibitory effects of ibuprofen, ketorolac, and flurbiprofen upon anandamide metabolism are rather modest. However, comparison of inhibitory potencies toward cyclooxygenase-2 in intact cells may be misleading since the anandamide data is undertaken using homogenates, and the cyclooxygenase-2 inhibitory potency of ibuprofen is lower in broken cell preparations (21). In order to investigate this possibility, anandamide metabolism was assayed in whole cells. C6 cells are known to express fatty acid amide hydrolase (2), and a previous study has used this cell line to investigate the inhibition in whole cells of the enzyme by diazomethylarachidonyl ketone (22). In

FIG. 4. Inhibition of 2 mM [ 3H]anandamide hydrolysis by (A) R(1)-ketorolac (), RS-ketorolac (E) and S(2)-ketorolac (‚); (B) (R)flurbiprofen () and (S)-flurbiprofen (‚). Data are means 6SEM (unless contained within the symbols), n 5 7 (A) or 4 (B).

196

FOWLER ET AL.

that study, the cells were preincubated with the inhibitor and membranes prepared thereafter and assayed for their ability to metabolise anandamide. While that approach is useful for irreversible inhibitors such as diazomethylarachidonyl ketone, it cannot be used for reversible inhibitors. In consequence, we assayed the metabolism of [ 3 H]anandamide in intact cells after a preincubation with the ibruprofen enantiomers, to allow their uptake into the cell. Under these conditions, the R(2)-form of ibuprofen was again slightly more potent as an inhibitor of [ 3 H]anandamide metabolism than the S(1)-form, but the potencies were very similar to those found for the rat brain homogenates. On the basis of our previous study using racemic ibuprofen, we suggested that anandamide metabolism may well be affected in vivo after drug administration, particularly in cases of rheumatoid arthritis, when high doses are often recommended (8). The finding in the present study that the “wrong” enantiomer is the more potent of the two, and that the potencies of the compounds are similar in intact cells to those for the cell-free preparations, would argue against this suggestion, at least for ibuprofen. With respect to ketorolac, it is unlikely that inhibition of anandamide metabolism is of any therapeutic relevance, since the peak plasma concentration of ketorolac after a therapeutic dose (;3 mM after 10 mg orally, (23)) is much lower than the IC 50 value for inhibition of anandamide metabolism, particularly for the S-enantiomer. On the other hand a peak plasma concentration of flurbiprofen of ;40 mM is reached following a single 200 mg sustained release capsule of this compound (24). Even though the high plasma protein binding of NSAID compounds is not taken into consideration (see 25), the possibility that inhibition of anandamide metabolism may contribute to the therapeutic efficacy of flurbiprofen should not be ruled out. Indeed, given the antinociceptive properties of cannabinoids (26), it is possible that a compound with dual actions upon cyclooxygenase-2 and upon fatty acid amide hydrolase may prove to be a useful therapeutic agent. ACKNOWLEDGMENTS The authors thank Ingrid Persson, Britt Jacobsson, Kerstin Wahlstro¨m, and Lena Gustafsson for excellent technical assistance and the Swedish Medical Research Foundation (Grant K98-03X-1254801A) and the Research Fund of the Medical Faculty, Umeå University, for financial support.

REFERENCES 1. Devane, W. A., Hanus, L., Breuer, A., Pertwee, R. G., Stevenson, L. A., Griffin, G., Gibson, D., Mandelbaum, A., Etinger, A., and Mechoulam, R. (1992) Science 258, 1946 –1949.

2. Deutsch, D. G., and Chin, S. A. (1993) Biochem. Pharmacol. 46, 791–796. 3. Cravatt, B. F., Giang, D. K., Mayfield, S. P., Boger, D. L., Lerner, R. A., and Gilula, N. B. (1996) Nature 384, 83– 87. 4. Maurelli, S., Bisogno, T., De Petrocellis, L., Luccia, A. D., Marino, G., and Di Marzo, V. (1995) FEBS Letts. 377, 82– 86. 5. Koutek, B., Prestwich, G. D., Howlett, A. C., Chin, S. A., Salehani, D., Akhavan, N., and Deutsch, D. G. (1994) J. Biol. Chem. 269, 22937–22940. 6. Lang, W., Qin, C., Hill, W. A. G., Lin, S., Khanolkar, A. D., and Makriyannis, A. (1996) Analyt. Biochem. 238, 40 – 45. 7. Fowler, C. J., Stenstro¨m, A., and Tiger, G. (1997) Pharmacol. Toxicol. 80, 103–107. 8. Fowler, C. J., Tiger, G., and Stenstro¨m, A. (1997) J. Pharmacol. Exp. Ther. 283, 729 –734. 9. Glowinski, J., and Iversen, L. L. (1966) J. Neurochem. 13, 655– 669. 10. Omeir, R. L., Chin, S., Hong, Y., Ahern, D. G., and Deutsch, D. G. (1995) Life Sci. 56, 1999 –2005. 11. Eisenthal, R., and Cornish-Bowden, A. (1974) Biochem. J. 139, 715–720. 12. Neupert, W., Brugger, R., Euchenhofer, C., Brune, K., and Geisslinger, G. (1997) Br. J. Pharmacol. 122, 487– 492. 13. Kaiser, D. G., Vangiessen, G. J., Reischer, R. J., and Wechter, W. J. (1976) J. Pharm. Sci. 65, 269 –273. 14. Lee, E. J., Williams, K., Day, R., Graham, G., and Champion, D. (1985) Br. J. Clin. Pharmacol. 19, 669 – 674. 15. Kantoci. D., and Wechter, W. J. (1996) J. Clin. Pharmacol. 36, 500 –504. 16. Hillard, C. J., Wilkison, D. M., Edgemond, W. S., and Campbell, W. B. (1995) Biochim. Biophys. Acta 1257, 249 –256. 17. Desarnaud, F., Cadas, H., and Piomelli, D. (1995) J. Biol. Chem. 270, 6030 – 6035. 18. Guzman, A., Yuste, F., Toscano, R. A., Young, J. M., Van Horn, A. R., and Muchowski, J. M. (1986) J. Med. Chem. 29, 589 –591. 19. Carabaza, A., Cabre, F., Rotllan, E., Gomez, M., Gutierrez, M., Garcia, M. L., and Mauleon, D. (1996) J. Clin. Pharmacol. 36, 505–512. 20. Riendeau, D., Percival, M. D., Boyce, S., Brideau, C., Charleson, S., Cromlish, W., Ethier, D., Evans, J., Falgueyret, J.-P., FordHutchinson, A. W., Gordon, R., Greig, G., Gresser, M., Guay, J., Kargman, S., Le´ger, S., Mancini, J. A., O’Neill, G., Ouellet, M., Rodger, I. W., The´rien, M., Wang, Z., Webb, J. K., Wong, E., Xu, L., Young, R. N., Zamboni, R., Prasit, P., and Chan, C.-C. (1997) Br. J. Pharmacol. 121, 105–117. 21. Mitchell, J. A., Akarasereenont, P., Thiemermann, C., Flower, R. J., and Vane, J. R. (1994) Proc. Natl. Acad. Sci. USA 90, 11693–11697. 22. Edgemond, W. S., Greenberg, M. J., McGinley, P. J., Muthian, S., Campbell, W. B., and Hillard, C. J. (1998) J. Pharmacol. Exp. Ther. 286, 184 –190. 23. Jung, D., Mroszczak, E., and Bynum, L. (1988) Eur. J. Clin. Pharmacol. 35, 423– 425. 24. Pargal, A., Kelkar, M. G., and Nayak, P. J. (1996) Biopharm. Drug. Dispos. 17, 511–519. 25. Benet, L. Z., Øie, S., and Schwartz, J. B. (1996) in Goodman & Gilman’s The Pharmacological Basis of Therapeutics, (Hardman, J. G., Goodman Gilman, A., and Limbard, L. E., Eds.), 9th ed., pp. 1707–1792, McGraw-Hill, New York. 26. Dewey, W. L. (1986) Pharmacol. Rev. 38, 151–178.