Equilibrium in the hydrolysis and synthesis of cannabimimetic anandamide demonstrated by a purified enzyme

Biochimica et Biophysica Acta 1440 (1999) 205^214 www.elsevier.com/locate/bba

Equilibrium in the hydrolysis and synthesis of cannabimimetic anandamide demonstrated by a puri¢ed enzyme Kazuhisa Katayama b , Natsuo Ueda a , Itsuo Katoh b , Shozo Yamamoto

a;

*

a

b

Department of Biochemistry, Tokushima University, School of Medicine, Kuramoto-cho, Tokushima 770-8503, Japan Department of Cardiovascular Surgery, Tokushima University, School of Medicine, Kuramoto-cho, Tokushima 770-8503, Japan Received 10 May 1999; received in revised form 12 July 1999; accepted 15 July 1999

Abstract Anandamide, an endogenous ligand for cannabinoid receptors, loses its biological activities when it is hydrolyzed to arachidonic acid and ethanolamine by anandamide amidohydrolase. We overexpressed a recombinant rat enzyme with a hexahistidine tag in a baculovirus^insect cell expression system, and purified the enzyme with the aid of a Ni-charged resin to a specific activity as high as 5.7 Wmol/min/mg protein. The purified recombinant enzyme catalyzed not only the hydrolysis of anandamide and palmitoylethanolamide, but also their reverse synthetic reactions. In order to attain an equilibrium of the anandamide hydrolysis and its reverse reaction within 10 min, we utilized a large amount of the purified enzyme. The equilibrium constant ([arachidonic acid][ethanolamine])/([anandamide][water]) was calculated as 4U1033 (37³C, pH 9.0). These experimental results with a purified enzyme preparation quantitatively confirmed the reversibility of the enzyme reaction previously observed with crude enzyme preparations. ß 1999 Elsevier Science B.V. All rights reserved. Keywords: Anandamide; Arachidonic acid; Cannabinoid; Amidohydrolase; Palmitoylethanolamide ; Equilibrium constant

1. Introduction Cannabinoids, contained in the plant Cannabis sativa L., exert a variety of pharmacological actions through speci¢c seven-transmembrane G-proteincoupled receptors, CB1 and CB2 [1^3]. In 1992, an endogenous ligand for the cannabinoid receptors was isolated from porcine brain, identi¢ed as arachidonoylethanolamide, and referred to as anandamide [4]. In 1993, Deutsch and Chin demonstrated with the homogenates of rat brain and other mammalian tissues that anandamide lost its biological activities

Abbreviations: PMSF, phenylmethylsulfonyl £uoride; TLC, thin-layer chromatography * Corresponding author. Fax: +81-88-633-6409.

when it was enzymatically hydrolyzed to arachidonic acid and ethanolamine, and they termed the enzyme `anandamide amidase' [5]. Later, the enzyme was found in brain neurons [6], characterized with rat brain microsomes [7], and referred to as `anandamide amidohydrolase' [7]. We partially puri¢ed the enzyme from porcine brain microsomes, and found that the enzyme was active not only with anandamide, but also with other N-acylethanolamines [8]. This wide substrate speci¢city suggested that the enzyme was identical to `N-acylethanolamine amidohydrolase' described previously [9]. In 1996, a cDNA for `fatty-acid amide hydrolase' was cloned from rat liver [10]. This enzyme had been studied as a hydrolase for oleamide, a putative endogenous sleep-inducer. However, COS-7 cells transfected with this cDNA hydrolyzed anandamide as well as oleamide

1388-1981 / 99 / $ ^ see front matter ß 1999 Elsevier Science B.V. All rights reserved. PII: S 1 3 8 8 - 1 9 8 1 ( 9 9 ) 0 0 1 2 4 - 9

BBAMCB 55508 1-9-99

206

K. Katayama et al. / Biochimica et Biophysica Acta 1440 (1999) 205^214

[10]. All the enzymes mentioned above are now considered presumably to be identical. The enzymatic anandamide hydrolysis was reversible, and anandamide was synthesized from arachidonic acid and ethanolamine as demonstrated with the partially puri¢ed enzyme from porcine brain [8] and a recombinant rat liver enzyme overexpressed in COS-7 cells [11,12]. However, since Km values for ethanolamine in the synthetic reaction were reported to be extremely high, the enzyme was thought to function as a hydrolase rather than a synthase under the physiological conditions [8,11,12]. Recently, Schmid et al. showed that rat testis membrane generated anandamide from endogenous arachidonic acid in the presence of as low as 50 WM exogenous ethanolamine [13]. Moreover, Paria et al. reported that mouse uterus had an enzyme forming anandamide from arachidonic acid and ethanolamine, but presumably distinguished from the reverse reaction of anandamide amidohydrolase [14]. Thus, the contribution of anandamide amidohydrolase to the in vivo anandamide formation is still a debatable subject. In this study, we examined catalytic properties of the highly puri¢ed recombinant anandamide amidohydrolase of rat with special reference to the reversibility and equilibrium of the enzyme reactions. 2. Materials and methods 2.1. Materials [1-14 C]Arachidonic acid was purchased from Amersham (UK), [1-14 C]palmitic acid from Du Pont NEN (Boston, MA), arachidonic acid and palmitic acid from Nu-Chek-Prep (Elysian, MN), anandamide and palmitoylethanolamide from Cayman (Ann Arbor, MI), phenylmethylsulfonyl £uoride (PMSF) from Sigma (St. Louis, MO), protein assay dye reagent concentrate from Bio-Rad (Hercules, CA), and precoated silica gel 60 F254 glass plates for TLC (20U20 cm, 0.25 mm thickness) from Merck (Darmstadt, Germany). [Arachidonoyl1-14 C]anandamide was chemically prepared from [1-14 C]arachidonic acid and ethanolamine, and [palmitoyl-1-14 C]palmitoylethanolamide from [1-14 C]palmitic acid and ethanolamine as described previously [15]. The baculovirus expression system com-

posed of fall armyworm ovarian cells Sf9, linearized Bac-N-Blue DNA of baculovirus, pBlueBac4.5, BacN-Blue Transfection Kit and ProBond Resin Column was purchased from Invitrogen (Carlsbad, CA). Grace's insect cell culture medium, yeastolate solution, lactalbumin hydrolysate solution and Pluronic F-68 were purchased from Gibco-BRL (Gaithersburg, MD) and fetal calf serum from PAA Laboratories (Linz, Austria). 2.2. Overexpression and puri¢cation of a recombinant enzyme The coding region of cDNA for rat anandamide amidohydrolase was prepared as we described previously [11]. In order to construct the enzyme with an additional carboxyl-terminal sequence Ala-(His)6 Ala, we prepared a cDNA fragment by PCR with an upstream primer 5P-GCCTGAAAGCTCTACTGTGTGAGC-3P and a downstream primer 5P-GCTCTAGATTAGGCGTGATGGTGATGGTGATGGGCCGATGGCTGCTTTTGAGGGGTC-3P utilizing pfu DNA polymerase (Stratagene, La Jolla, CA). The coding region of cDNA for the enzyme was used as a template. The digestion of the fragment with XbaI gave a 0.8-kilobase fragment containing nucleotides 1014^1789 (the numbers are due to [10]) and the sequence corresponding to the carboxyl-terminal hexahistidine tag. Then, the KpnI^XbaI fragment (nucleotides 41^1013) prepared as described before [11] and the above-mentioned XbaI^XbaI fragment were sequentially ligated to a baculovirus transfer vector pBlueBac4.5. Sf9 cells were cultured in Grace's medium supplemented with yeastolate, lactalbumin hydrolysate, 10% fetal calf serum and 0.1% Pluronic F-68. The cells were suspended in a glass £ask by continuous rotation and maintained at a density of 0.6^2.5U106 / ml in an air incubator at 26³C. According to the instruction for Bac-N-Blue Transfection Kit, recombinant baculoviruses were generated by cotransfection of Sf9 cells with the pBlueBac4.5 containing the enzyme cDNA and linearized pBac-N-Blue viral DNA. For the overexpression and puri¢cation of the enzyme, a high-titer recombinant viral stock (2.5 ml) was infected to 5U108 Sf9 cells in 250 ml of the medium. After 96-h culture the cells were harvested

BBAMCB 55508 1-9-99

K. Katayama et al. / Biochimica et Biophysica Acta 1440 (1999) 205^214

and stored as a cell pellet at 380³C until use. The following procedures were performed at 4³C. The 2U108 thawed cells were suspended in 5 ml of 50 mM Tris-HCl (pH 7.4), and sonicated with the aid of a Branson soni¢er model 185. The lysate was then centrifuged at 180 000Ug for 45 min. The resultant particulate fraction was suspended at a concentration of 5 mg protein/ml in 50 mM Tris-HCl bu¡er (pH 7.4) containing 0.1% Triton X-100, and the suspension was stirred for 12 h. The sample was centrifuged at 180 000Ug for 45 min. The clear supernatant obtained was used as a solubilized protein. The solubilized protein (about 30 mg protein) was loaded onto a Ni-charged resin column (ProBond Resin Column, a bed volume of 2 ml) which was pre-equilibrated with 20 mM Tris-HCl (pH 7.4) containing 1 mM CHAPS (bu¡er A). The column was washed with 200 ml of bu¡er A and then 10 ml of bu¡er A containing 20 mM imidazole. The enzyme adsorbed on the column was eluted with 15 ml of bu¡er A containing 50 mM imidazole. Active fractions could be stored at 4³C for at least 3 weeks without signi¢cant loss of the enzyme activity. Freezing and thawing of the puri¢ed enzyme caused a serious loss of the activity. Protein concentration was determined by the method of Bradford with bovine serum albumin as standard [16]. 2.3. Enzyme assay For the anandamide hydrolase assay, the enzyme was incubated with 100 WM [1-14 C]anandamide (10 000 cpm in 5 Wl dimethyl sulfoxide) at 37³C for 5 min in 200 Wl of 50 mM Tris-HCl (pH 9.0). The assay for the anandamide synthase activity was carried out by incubation of the enzyme with 250 WM [1-14 C]arachidonic acid (50 000 cpm in 5 Wl dimethyl sulfoxide) in 200 Wl of 250 mM ethanolamine-HCl (pH 9.0) at 37³C for 5 min. Termination of the reaction, separation of the product by TLC, and quanti¢cation of the radioactivity were carried out as described previously [17]. 2.4. Western blotting Antiserum for a peptide PLGLGTDIGGSIRFPS corresponding to amino acids 231^246 of rat anand-

207

amide amidohydrolase was raised in rabbits by Yuko Kurahashi and Yoko Hayashi of our department, and diluted 1000-fold for Western blotting. Electrophoresis was performed with an 8.5% polyacrylamide gel in the presence of 0.1% sodium dodecylsulfate, followed by transfer of protein bands from the gel to an Immobilon-P membrane (Millipore, Bedford, MA). Immunocomplex on the membrane was visualized by a Vectastain ABC kit (Vector Laboratories, Burlingame, CA) and 3,3P-diaminobenzidine (Dojin, Kumamoto, Japan) as a peroxidase substrate. 3. Results 3.1. Functional expression and puri¢cation of recombinant anandamide amidohydrolase When the particulate fraction of the transfected Sf9 cells was allowed to react with [14 C]anandamide followed by the analysis using TLC, a radioactive band corresponding to arachidonic acid was detected as reported previously [17]. In contrast, control Sf9 cells did not produce arachidonic acid. The particulate fraction was solubilized with 0.1% Triton X-100, and the solubilized enzyme was puri¢ed with the aid of a Ni-charged resin. Starting from the particulate fraction, the recombinant anandamide amidohydrolase was ¢nally puri¢ed 16-fold to a speci¢c activity of 5.7 Wmol/min/mg protein at 37³C with an overall yield of 18%. The puri¢ed enzyme preparation gave a single protein band as analyzed by sodium dodecylsulfate^polyacrylamide gel electrophoresis followed by sliver staining (Fig. 1A). The apparent molecular mass was about 62 kDa. This band was immunostained with rabbit antiserum against a peptide corresponding to amino acids 231^246 of the rat enzyme (Fig. 1B). 3.2. Catalytic properties of the puri¢ed enzyme The puri¢ed enzyme catalyzed both the hydrolysis and synthesis of anandamide (Fig. 2). Vmax values of the hydrolase and synthase reactions were 5.7 þ 0.1 and 6.8 þ 0.1 Wmol/min/mg protein (mean þ S.D., n = 4), respectively, at saturating concentrations of the substrates. The corresponding values based on Lineweaver^Burk plot were 6.2 þ 0.1 and 7.8 þ 0.5

BBAMCB 55508 1-9-99

208

K. Katayama et al. / Biochimica et Biophysica Acta 1440 (1999) 205^214

values for palmitic acid and ethanolamine were 60 WM and 80 mM, respectively, and the Vmax was 2.2 Wmol/min/mg protein.

Fig. 1. Puri¢cation of the recombinant anandamide amidohydrolase. (A) The particulate fraction of the Sf9 cells infected with the recombinant baculovirus (8 Wg of protein, lane 1) and the puri¢ed enzyme (0.5 Wg of protein, lane 2) were subjected to sodium dodecylsulfate-polyacrylamide gel electrophoresis followed by staining with silver nitrate. Lane M, marker proteins. (B) The particulate fractions of the control Sf9 cells (5 Wg of protein, lane C) and the Sf9 cells infected with the recombinant baculovirus (5 Wg of protein, lane 1), and the puri¢ed enzyme (0.5 Wg of protein, lane 2) were applied to Western blotting using rabbit antiserum against a peptide corresponding to amino acids 231^246 of the rat enzyme.

Wmol/min/mg protein (n = 4). kcat values were calculated as 6.5 þ 0.1 s31 and 8.2 þ 0.6 s31 . These Vmax and kcat values for the hydrolase reaction and synthase reaction were signi¢cantly di¡erent (P 6 0.001). The Km value for anandamide in the hydrolytic reaction was about 30 WM (Fig. 2A), and those for arachidonic acid and ethanolamine in the synthetic reaction were approximately 70 WM and 40 mM, respectively (Fig. 2B,C). The anandamide hydrolase and synthase activities were high at alkaline pH, and their optimal pH values were around 9.0 (Fig. 3). We especially noted that the hydrolase activity was considerably higher than the synthase activity between pH 5.0 and 7.4. The puri¢ed enzyme also hydrolyzed palmitoylethanolamide, but at a lower rate than the anandamide hydrolysis (Fig. 2A). As estimated by Lineweaver^ Burk plot, the Km for palmitoylethanolamide was about 70 WM, and Vmax was 1.8 Wmol/min/mg protein. The palmitoylethanolamide hydrolysis by the enzyme was also reversible as shown in Fig. 2B,C. In the synthesis of palmitoylethanolamide, the Km

Fig. 2. The hydrolase and synthase activities depending on substrate concentration. The puri¢ed enzyme (0.2 Wg of protein for the hydrolysis and synthesis of anandamide or 0.8 Wg of protein for the hydrolysis and synthesis of palmitoylethanolamide) was allowed to react under the standard conditions with various concentrations of the following substrates. (A) [14 C]Anandamide (closed circles) and [14 C]palmitoylethanolamide (closed triangles); (B) [14 C]arachidonic acid (open circles) and [14 C]palmitic acid (open triangles) in the presence of 250 mM ethanolamine; and (C) ethanolamine in the presence of 250 WM [14 C]arachidonic acid (open circles) or [14 C]palmitic acid (open triangles). Mean values þ S.D. are shown (n = 4).

BBAMCB 55508 1-9-99

K. Katayama et al. / Biochimica et Biophysica Acta 1440 (1999) 205^214

209

3.3. Equilibrium in anandamide hydrolytic reaction

Fig. 3. pH dependence of the anandamide hydrolase and synthase reactions. The puri¢ed enzyme (0.2 Wg of protein) was assayed for the anandamide hydrolase activity (closed circles) and synthase activity (open circles) under the standard conditions at various pH values. The bu¡ers (50 mM) used were: glycineHCl at pH 3.0, citrate-sodium phosphate at pH 4.0, 5.0, 6.0 and 7.0, Tris-HCl at 7.4, 8.0 and 9.0, and NaHCO3 -NaOH at pH 10.0 and 11.0. Mean values þ S.D. are shown (n = 4).

In an attempt to determine the equilibrium constant in the reversible anandamide hydrolysis, the reaction time course was examined with a large amount (5 Wg) of the puri¢ed enzyme (Fig. 4A). In the absence of ethanolamine, [14 C]anandamide was rapidly hydrolyzed, and almost all the radioactivity was detected as [14 C]arachidonic acid after 20 min. However, in the presence of 250 mM ethanolamine, only about 45% of anandamide was converted to arachidonic acid within 10 min, and the apparent hydrolytic reaction did not proceed any more. In the presence of 250 mM ethanolamine and 100 WM arachidonic acid, the anandamide synthetic reaction apparently ceased after about 55% conversion of arachidonic acid to anandamide. Thus, the anandamide hydrolysis and synthesis reached an equilibrium. Based on the concentrations of anandamide and arachidonic acid at equilibrium, the equilibrium constant ([arachidonic acid][ethanolamine])/([anandamide][water]) was calculated as 4U1033 . Even with ethanolamine at as low as 100 WM, a small but detectable amount (0.05^0.1 WM) of anandamide was produced from 100 WM arachidonic acid with

Fig. 4. Equilibrium in the hydrolytic and synthetic reactions of anandamide examined with the puri¢ed enzyme. (A,C,D) The puri¢ed enzyme (5 Wg of protein) was allowed to react for the indicated time periods with 100 WM [14 C]anandamide in the presence of 250 mM ethanolamine (closed circles), 100 WM [14 C]anandamide in the absence of ethanolamine (closed triangles), and 100 WM [14 C]arachidonic acid in the presence of 250 mM ethanolamine (open circles). (B) The puri¢ed enzyme (5 Wg of protein) was allowed to react for the indicated time periods with 100 WM [14 C]arachidonic acid in the presence of 100 WM ethanolamine (open circles). The bu¡er (50 mM) used was Tris-HCl at pH 9.0 (A,B), Tris-HCl at pH 7.4 (C) or citrate-sodium phosphate at pH 6.0 (D). Mean values þ S.D. are shown (n = 3).

BBAMCB 55508 1-9-99

210

K. Katayama et al. / Biochimica et Biophysica Acta 1440 (1999) 205^214

the puri¢ed enzyme (Fig. 4B). Furthermore, when the concentration of anandamide or arachidonic acid was changed between 5 and 400 WM and the ethanolamine concentration was ¢xed at 250 mM in the hydrolytic or synthetic reaction, the ratios of the product and the remaining substrate were almost constant at the equilibrium points (Fig. 5A). On the other hand, the increase in the concentration of ethanolamine (10^500 mM) resulted in the increase in a ratio of anandamide to arachidonic acid (Fig. 5B). These results were consistent with the values calculated on the basis of the equilibrium constant. The equilibrium point was a¡ected by pH of the reaction mixture. When the reaction reached an equilibrium at pH 7.4, anandamide and arachidonic acid were found in a ratio of 38:62 (Fig. 4C). The ratio was 14:86 at pH 6.0 (Fig. 4D). The equilibrium constants for the reactions were calculated at 8U1033

Fig. 6. Equilibrium in the hydrolytic and synthetic reactions of palmitoylethanolamide. The puri¢ed enzyme (5 Wg of protein) was allowed to react for the indicated time periods with 100 WM [14 C]palmitoylethanolamide (closed triangles) and 100 WM [14 C]palmitic acid (open triangles) both in the presence of 250 mM ethanolamine. The bu¡er used was 50 mM Tris-HCl at pH 9.0. Mean values þ S.D. are shown (n = 3).

and 3U1032 , respectively. These values were higher than that at pH 9.0 (4U1033 ) as described above. Thus, acidic pH seems to be unfavorable to the anandamide formation. This is consistent with the ¢nding that the hydrolase activity was considerably higher than the synthase activity at acidic pH (Fig. 3). We also examined equilibrium in the palmitoylethanolamide hydrolysis at pH 9.0 in the presence of 250 mM ethanolamine (Fig. 6). When the reaction reached an equilibrium, palmitoylethanolamide and palmitic acid were present in a ratio of 92:8. The equilibrium constant for the reaction was determined as 4U1034 , which was considerably lower than that of the anandamide hydrolysis. 3.4. Inhibition of the puri¢ed enzyme by fatty acids and PMSF

Fig. 5. E¡ect of the substrate concentrations on the equilibrium of the hydrolytic and synthetic reactions of anandamide. The puri¢ed enzyme (5 Wg of protein) was allowed to react for 20 min at pH 9.0 with various concentrations of the following substrates: (A) [14 C]anandamide (closed circles) or 14 [ C]arachidonic acid (open circles) both in the presence of 250 mM ethanolamine; and (B) ethanolamine in the presence of 100 WM [14 C]anandamide (closed circles) or 100 WM [14 C]arachidonic acid (open circles). Mean values þ S.D. are shown (n = 3).

When we incubated 100 WM [14 C]anandamide in the presence of various concentrations of non-radioactive arachidonic acid and palmitic acid, the anandamide hydrolysis was inhibited dose-dependently by arachidonic acid and palmitic acid (Fig. 7A). Their IC50 values were about 80 WM. Furthermore, the puri¢ed enzyme was incubated with various concentrations of [14 C]anandamide in the presence of 80 WM non-radioactive arachidonic acid or in its absence, and the result was shown by Lineweaver^Burk plot (Fig. 7B). Judging from this ¢gure, the inhibition was

BBAMCB 55508 1-9-99

K. Katayama et al. / Biochimica et Biophysica Acta 1440 (1999) 205^214

Fig. 7. Inhibition of the anandamide hydrolysis by arachidonic acid and palmitic acid. (A) The puri¢ed enzyme (0.2 Wg of protein) was allowed to react with 100 WM [14 C]anandamide in the presence of di¡erent concentrations of non-radioactive arachidonic acid (closed circles) and palmitic acid (open triangles). The compounds to be tested and [14 C]anandamide were simultaneously added to the reaction mixture. (B) The puri¢ed enzyme (0.2 Wg of protein) was allowed to react under the standard conditions with various concentrations of [14 C]anandamide in the presence of 80 WM non-radioactive arachidonic acid (open circles) or in its absence (closed circles). The values were shown by Lineweaver^Burk plot. Mean values þ S.D. are shown (n = 4).

of a non-competitive type. When the puri¢ed enzyme was pretreated with PMSF (a serine hydrolase inhibitor) for 5 min, both the anandamide hydrolase and synthase activities were inhibited in parallel with IC50 values of about 20 WM. 4. Discussion Previously, we partially puri¢ed anandamide amidohydrolase from porcine brain microsomes [8]. However, the speci¢c activity was not so high (0.37 Wmol/min/mg protein), and further attempts to purify the enzyme were unsuccessful probably due to its

211

hydrophobicity. Later we prepared the particulate fraction of COS-7 cells overexpressing the recombinant enzyme of rat liver with a speci¢c activity of 0.13 Wmol/min/mg protein [11]. In the present study, we overexpressed the recombinant enzyme with a hexahistidine tag in a baculovirus^insect cell expression system, and puri¢ed the enzyme by the use of a Ni-charged resin with a high a¤nity for the hexahistidine tag. The speci¢c activity of the particulate fraction of the transfected Sf9 cells was 0.4 Wmol/ min/mg protein, which was three-times higher than that of the enzyme expressed in COS-7 cells. The puri¢ed recombinant enzyme showed a speci¢c activity as high as 5.7 Wmol/min/mg protein. By one column chromatography according to the thus established puri¢cation procedure, we could easily and reproducibly prepare about 0.8 mg of an apparently homogenous enzyme which enabled us to perform about 4000 standard assays. When we were preparing this manuscript, Patricelli et al. reported the puri¢cation of rat fatty acid amide hydrolase overexpressed in Escherichia coli [18]. Using oleamide as substrate, kcat of the enzyme was 7.1 s31 . However, anandamide hydrolase and synthase activities of their puri¢ed enzyme were not described. The enzymatic anandamide formation by the condensation of arachidonic acid with ethanolamine was earlier reported [5,19,20]. By the use of the partially puri¢ed enzyme, we suggested that this `anandamide synthase' activity was attributable to the reverse reaction of anandamide amidohydrolase [8]. This reversibility was also shown with a recombinant rat enzyme expressed in COS-7 cells [11,12], and was ¢nally con¢rmed with the highly puri¢ed recombinant enzyme in this study. The Km value for anandamide (30 WM) with the puri¢ed enzyme was similar to what we reported previously with the partially puri¢ed enzyme of porcine brain (60 WM) [8], but was higher than those for the crude enzyme preparations including membranes (3^12 WM) [7,21]. When we tested the particulate fraction of Sf9 cells expressing the recombinant enzyme, its Km for anandamide was about 6 WM (data not shown). Thus, membrane components such as lipid or the detergent contained in the puri¢ed enzyme preparation may a¡ect the a¤nity of anandamide for the enzyme. The Km value for ethanolamine was as high as 40 mM. Such high Km values were also reported with crude enzyme

BBAMCB 55508 1-9-99

212

K. Katayama et al. / Biochimica et Biophysica Acta 1440 (1999) 205^214

preparations by us [11] and other investigators [9,20,22]. By the use of the puri¢ed enzyme, we attempted to demonstrate the equilibrium of the reversible anandamide hydrolysis. When the enzyme reaction with anandamide alone reached an equilibrium, almost all the added anandamide was hydrolyzed, and it was technically di¤cult to determine an exact equilibrium constant. Therefore, we determined the constant in the presence of high concentrations (5^400 mM) of ethanolamine, and a value of about 4U1033 was obtained. Calculation on the basis of this constant indicates that 0.05% of 100 WM arachidonic acid should have been converted to anandamide in the presence of 100 WM ethanolamine when the reaction reaches an equilibrium. Experimentally, we detected 0.05^0.1% conversion of 100 WM arachidonic acid to anandamide for up to 20 min (Fig. 4B). The results showed that the anandamide formation by the reverse reaction was minor, but signi¢cantly detectable, even if only low concentrations of ethanolamine were supplied. If free arachidonic acid and ethanolamine are released by phospholipases A2 and D in the cells expressing anandamide amidohydrolase, the anandamide formation could occur at a signi¢cant level. Our results may explain a recent report that anandamide was formed by the condensation reaction in rat testis utilizing a low concentration of exogenous ethanolamine [13]. Although anandamide is known to be synthesized predominantly through the N-acylation-phosphodiesterase pathway [6,22^24], our results do not rule out a minor in vivo formation of anandamide by the reverse reaction of anandamide amidohydrolase. Palmitoylethanolamide is almost always detected in higher amounts than anandamide in various organs and cells [2], and is known to show anti-in£ammatory e¡ects [25] although it could bind to neither CB1 nor CB2 [26,27]. Our puri¢ed enzyme hydrolyzed palmitoylethanolamide at a much lower rate than anandamide as reported previously with crude enzyme preparations [7,8,28,29]. The lower rate of the enzyme with palmitoylethanolamide is consistent with the fact that this compound is more abundant than anandamide in animal tissues and cells [2]. The palmitoylethanolamide synthesis was reported to occur at similar rates to [8,22] or at lower rates than [19,30,31] that of the anandamide synthesis. In

our study, the puri¢ed enzyme synthesized palmitoylethanolamide at a lower rate than the anandamide synthesis (Fig. 2B). Interestingly, the equilibrium constant in the palmitoylethanolamide hydrolysis was much lower than that in the anandamide hydrolysis. Thus, if su¤cient concentrations of palmitic acid and ethanolamine are supplied, a considerable amount of palmitic acid will be converted to palmitoylethanolamide by the reverse reaction of the enzyme. The product inhibition of the anandamide hydrolysis by arachidonic acid was reported with crude enzymes [21,32,33]. Our puri¢ed enzyme was also inhibited dose-dependently by arachidonic acid and palmitic acid with an IC50 value of 80 WM. Arachidonic acid was shown by Lineweaver^Burk plot to be a non-competitive inhibitor rather than a competitive inhibitor. This may not be a conclusive characterization of the inhibition because of possible nonspeci¢c detergent activity of unsaturated fatty acid. The anandamide hydrolysis is also inhibited by PMSF, a non-speci¢c serine hydrolase inhibitor [5]. However, it was reported previously with crude enzyme preparations that PMSF showed no or less inhibitory e¡ect on the anandamide synthesis from arachidonic acid and ethanolamine [5,8,14,34]. In contrast, our present work showed that the anandamide hydrolase and synthase activities of the puri¢ed enzyme were inhibited in parallel by PMSF. This discrepancy may be interpreted in a way that there is another enzyme distinguishable from anandamide amidohydrolase or the sulfonation of a single enzyme a¡ected the catalytic activities for the forward and reverse reactions in di¡erent manners [34]. In summary, these quantitative experimental results with the puri¢ed enzyme con¢rmed our previous ¢nding of the reversibility of the enzymatic anandamide hydrolysis. Acknowledgements We are grateful to Dr. Akira Shibata, Faculty of Pharmaceutical Sciences, Tokushima University, Dr. Kenji Aki, Tokushima Bunri University, and Drs. Hiroyuki Kagamiyama and Hideyuki Hayashi, Osaka Medical College, for helpful discussion on the equilibrium analysis of our enzyme. This work was

BBAMCB 55508 1-9-99

K. Katayama et al. / Biochimica et Biophysica Acta 1440 (1999) 205^214

supported by grants-in-aid for scienti¢c research from the Ministry of Education, Science, Sports and Culture of Japan, the Human Frontier Science Program, the Japanese Foundation of Metabolism and Disease, the Uehara Memorial Foundation, the Japan Foundation for Applied Enzymology, Ono Pharmaceutical Co., Kissei Pharmaceutical Co., Sankyo Co., Japan Tobacco Inc., and Takeda Pharmaceutical Industry.

[13]

[14]

[15]

References [1] R.G. Pertwee, Pharmacology of cannabinoid CB1 and CB2 receptors, Pharmacol. Ther. 74 (1997) 129^180. [2] V. Di Marzo, `Endocannabinoids' and other fatty acid derivatives with cannabimimetic properties: biochemistry and possible physiopathological relevance, Biochim. Biophys. Acta 1392 (1998) 153^175. [3] C.C. Felder, Cannabinoid receptors and their endogenous agonists, Annu. Rev. Pharmacol. Toxicol. 38 (1998) 179^ 200. [4] W.A. Devane, L. Hanus­, A. Breuer, R.G. Pertwee, L.A. Stevenson, G. Gri¤n, D. Gibson, A. Mandelbaum, A. Etinger, R. Mechoulam, Isolation and structure of a brain constituent that binds to the cannabinoid receptor, Science 258 (1992) 1946^1949. [5] D.G. Deutsch, S.A. Chin, Enzymatic synthesis and degradation of anandamide, a cannabinoid receptor agonist, Biochem. Pharmacol. 46 (1993) 791^796. [6] V. Di Marzo, A. Fontana, H. Cadas, S. Schinelli, G. Cimino, J.-C. Schwartz, D. Piomelli, Formation and inactivation of endogenous cannabinoid anandamide in central neurons, Nature 372 (1994) 686^691. [7] F. Desarnaud, H. Cadas, D. Piomelli, Anandamide amidohydrolase activity in rat brain microsomes. Identi¢cation and partial characterization, J. Biol. Chem. 270 (1995) 6030^ 6035. [8] N. Ueda, Y. Kurahashi, S. Yamamoto, T. Tokunaga, Partial puri¢cation and characterization of the porcine brain enzyme hydrolyzing and synthesizing anandamide, J. Biol. Chem. 270 (1995) 23823^23827. [9] P.C. Schmid, M.L. Zuzarte-Augustin, H.H.O. Schmid, Properties of rat liver N-acylethanolamine amidohydrolase, J. Biol. Chem. 260 (1985) 14145^14149. [10] B.F. Cravatt, D.K. Giang, S.P. May¢eld, D.L. Boger, R.A. Lerner, N.B. Gilula, Molecular characterization of an enzyme that degrades neuromodulatory fatty-acid amides, Nature 384 (1996) 83^87. [11] Y. Kurahashi, N. Ueda, H. Suzuki, M. Suzuki, S. Yamamoto, Reversible hydrolysis and synthesis of anandamide demonstrated by recombinant rat fatty-acid amide hydrolase, Biochem. Biophys. Res. Commun. 237 (1997) 512^515. [12] G. Arreaza, W.A. Devane, R.L. Omeir, G. Sajnani, J. Kunz,

[16]

[17]

[18]

[19]

[20]

[21]

[22]

[23]

[24]

213

B.F. Cravatt, D.G. Deutsch, The cloned rat hydrolytic enzyme responsible for the breakdown of anandamide also catalyzes its formation via the condensation of arachidonic acid and ethanolamine, Neurosci. Lett. 234 (1997) 59^62. P.C. Schmid, D. Schwindenhammer, R.J. Krebsbach, H.H.O. Schmid, Alternative pathways of anandamide biosynthesis in rat testes, Chem. Phys. Lipids 92 (1998) 27^35. B.C. Paria, D.G. Deutsch, S.K. Dey, The uterus is a potential site for anandamide synthesis and hydrolysis: di¡erential pro¢les of anandamide synthase and hydrolase activities in the mouse uterus during the periimplantation period, Mol. Reprod. Dev. 45 (1996) 183^192. N. Ueda, K. Yamamoto, S. Yamamoto, T. Tokunaga, E. Shirakawa, H. Shinkai, M. Ogawa, T. Sato, I. Kudo, K. Inoue, H. Takizawa, T. Nagano, M. Hirobe, N. Matsuki, H. Saito, Lipoxygenase-catalyzed oxygenation of arachidonylethanolamide, a cannabinoid receptor agonist, Biochim. Biophys. Acta 1254 (1995) 127^134. M.M. Bradford, A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding, Anal. Biochem. 72 (1976) 248^ 254. K. Katayama, N. Ueda, Y. Kurahashi, H. Suzuki, S. Yamamoto, I. Kato, Distribution of anandamide amidohydrolase in rat tissues with special reference with small intestine, Biochim. Biophys. Acta 1347 (1997) 212^218. M.P. Patricelli, H.A. Lashuel, D.K. Giang, J.W. Kelly, B.F. Cravatt, Comparative characterization of a wild type and transmembrane domain-deleted fatty acid amide hydrolase: identi¢cation of the transmembrane domain as a site for oligomerization, Biochemistry 37 (1998) 15177^15187. K.K. Kruszka, R.W. Gross, The ATP- and CoA-independent synthesis of arachidonoylethanolamide, J. Biol. Chem. 269 (1994) 14345^14348. W.A. Devane, J. Axelrod, Enzymatic synthesis of anandamide, an endogenous ligand for the cannabinoid receptor, by brain membranes, Proc. Natl. Acad. Sci. USA 91 (1994) 6698^6701. C.J. Hillard, D.M. Wilkison, W.S. Edgemond, W.B. Campbell, Characterization of the kinetics and distribution of Narachidonylethanolamine (anandamide) hydrolysis by rat brain, Biochim. Biophys. Acta 1257 (1995) 249^256. T. Sugiura, S. Kondo, A. Sukagawa, T. Tonegawa, S. Nakane, A. Yamashita, Y. Ishima, K. Waku, Transacylasemediated and phosphodiesterase-mediated synthesis of Narachidonoylethanolamine, an endogenous cannabinoid-receptor ligand, in rat brain microsomes, Eur. J. Biochem. 240 (1996) 53^62. H.H.O. Schmid, P.C. Schmid, V. Natarajan, The N-acylation-phosphodiesterase pathway and cell signalling, Chem. Phys. Lipids 80 (1996) 133^142. S. Kondo, T. Sugiura, T. Kodaka, N. Kudo, K. Waku, A. Tokumura, Accumulation of various N-acylethanolamines including N-arachidonoylethanolamine (anandamide) in cadmium chloride-administered rat testis, Arch. Biochem. Biophys. 354 (1998) 303^310.

BBAMCB 55508 1-9-99

214

K. Katayama et al. / Biochimica et Biophysica Acta 1440 (1999) 205^214

[25] H.H.O. Schmid, P.C. Schmid, V. Natarajan, N-Acylated glycerophospholipids and their derivatives, Prog. Lipid Res. 29 (1990) 1^43. [26] V.M. Showalter, D.R. Compton, B.R. Martin, M.E. Abood, Evaluation of binding in a transfected cell line expressing a peripheral cannabinoid receptor (CB2): identi¢cation of cannabinoid receptor subtype selective ligands, J. Pharmacol. Exp. Ther. 278 (1996) 989^999. [27] T. Sheskin, L. Hanus­, J. Slager, Z. Vogel, R. Mechoulam, Structural requirements for binding of anandamide-type compounds to the brain cannabinoid receptor, J. Med. Chem. 40 (1997) 659^667. [28] S. Maurelli, T. Bisogno, L. De Petrocellis, A. Di Luccia, G. Marino, V. Di Marzo, Two novel classes of neuroactive fatty acid amides are substrates for mouse neuroblastoma `anandamide amidohydrolase', FEBS Lett. 377 (1995) 82^86. [29] T. Bisogno, S. Maurelli, D. Melck, L. De Petrocellis, V. Di Marzo, Biosynthesis, uptake, and degradation of anandamide and palmitoylethanolamide in leukocytes, J. Biol. Chem. 272 (1997) 3315^3323. [30] J. Priller, E.M. Briley, J. Mansouri, W.A. Devane, K. Mack-

[31]

[32]

[33]

[34]

ie, C.C. Felder, Mead ethanolamide, a novel eicosanoid, is an agonist for the central (CB1) and peripheral (CB2) cannabinoid receptors, Mol. Pharmacol. 48 (1995) 288^292. S. Matsuda, N. Kanemitsu, A. Nakamura, Y. Mimura, N. Ueda, Y. Kurahashi, S. Yamamoto, Metabolism of anandamide, an endogenous cannabinoid receptor ligand, in porcine ocular tissues, Exp. Eye Res. 64 (1997) 707^711. C. Qin, S. Lin, W. Lang, A. Goutopoulos, S. Pavlopoulos, F. Mauri, A. Makriyannis, Determination of anandamide amidase activity using ultraviolet-active amine derivatives and reverse-phase high-performance liquid chromatography, Anal. Biochem. 261 (1998) 8^15. S.K. Goparaju, N. Ueda, K. Taniguchi, S. Yamamoto, Enzymes of porcine brain hydrolyzing 2-arachidonoylglycerol, an endogenous ligand of cannabinoid receptors, Biochem. Pharmacol. 57 (1999) 417^423. D.G. Deutsch, R. Omeir, G. Arreaza, D. Salehani, G.D. Prestwich, Z. Huang, A. Howlett, Methyl arachidonyl £uorophosphonate: a potent irreversible inhibitor of anandamide amidase, Biochem. Pharmacol. 53 (1997) 255^260.

BBAMCB 55508 1-9-99