Chemistry and Physics of 108 (2000) 107–121 www.elsevier.com/locate/chemphyslip

Review The fatty acid amide hydrolase (FAAH)

Natsuo Ueda a, Robyn A. Puffenbarger b, Shozo Yamamoto a, Dale G. Deutsch b,*

a Department of , School of Medicine, Uni6ersity of Tokushima, Kuramoto-cho, Tokushima 770-8503, Japan b Department of Biochemistry and Cell Biology, State Uni6ersity of New York at Stony Brook, Stony Brook, NY 11794-5215, USA Received 3 May 2000; received in revised form 5 June 2000; accepted 5 June 2000

Abstract

The topic of this review is fatty acid amide hydrolase (FAAH), one of the best-characterized involved in the hydrolysis of bioactive lipids such as , 2-arachidonoylglycerol (2-AG), and . Herein, we discuss the nomenclature, the various assays that have been developed, the relative activity of the various substrates and the reversibility of the reactions catalyzed by FAAH. We also describe the cloning of the enzyme from rat and subsequent cDNA isolation from mouse, human, and pig. The proteins and the mRNAs from different species are compared. Cloning the enzyme permitted the purification and characterization of recombinant FAAH. The conserved regions of FAAH are described in terms of sequence and function, including the amidase domain which contains the catalytic nucleophile, the hydrophobic domain important for self association, and the rich domain region, which may be important for subcellular localization. The distribution of FAAH in the major organs of the body is described as well as regional distribution in the brain and its correlation with receptors. Since FAAH is recognized as a drug target, a large number of inhibitors have been synthesized and tested since 1994 and these are reviewed in terms of reversibility, potency, and specificity for FAAH and cannabinoid receptors. © 2000 Elsevier Science Ireland Ltd. All rights reserved.

Keywords: Fatty acid amide hydrolase; Anandamide; 2-Arachidonoylglycerol; Oleamide; Amidase; Amidohydrolase

1. Introduction

The fatty acid amide hydrolase (FAAH) plays an important role in terminating the signaling of * Corresponding author. Tel.: +1-631-6328595; fax: +1- the endocannabinoids and oleamide in the central 631-6328575. E-mail address: [email protected] (D.G. nervous system and in peripheral tissues. For the Deutsch). endocannabinoids, the action of both anandamide

0009-3084/00/$ - see front matter © 2000 Elsevier Science Ireland Ltd. All rights reserved. PII: S0009-3084(00)00190-0 108 N. Ueda et al. / Chemistry and Physics of Lipids 108 (2000) 107–121 and 2-arachidonoylglycerol (2-AG) are terminated the remaining anandamide by thin-layer chro- by the hydrolysis of the amide or ester bonds after matography (Deutsch and Chin, 1993; Hillard et being taken up into the cell. This research area al., 1995; Ueda et al., 1995; Cravatt et al., 1996) has been covered to some extent as part of recent or high-performance liquid chromatography reviews on the endocannabinoids (Deutsch and (Maccarrone et al., 1999a). Alternatively, when Makriyannis, 1997; Di Marzo and Deutsch, 1998; [14C] or [3H]anandamide labeled on the Di Marzo et al., 1999b). Although the role of ethanolamine moiety was hydrolyzed by FAAH, FAAH in hydrolyzing anandamide was first re- the water-soluble radioactive ethanolamine ported in 1993, where it was called an amidase product was separated from anandamide by open (Deutsch and Chin, 1993), the enzyme has an bed column chromatography (Desarnaud et al., interesting history. An enzyme activity was de- 1995; Maurelli et al., 1995) or by chloroform scribed in the 1960s, which catalyzed the forma- extraction (Omeir et al., 1995). tion of fatty acid amides of ethanolamine and this When non-labeled anandamide was employed enzyme showed tissue specificity similar to FAAH as the substrate, the arachidonic acid product was since it was not present in heart muscle and separated and quantified by high-performance liq- skeletal muscle (Bachur and Udenfriend, 1966). In uid chromatography by monitoring absorption at the 1980s an enzyme, called an amidohydrolase, 204 or 205 nm (Lang et al., 1996; Goparaju et al., was well characterized in terms of the N- 1998) or by gas chromatography of its methyl acylethanolamine substrates hydrolyzed, pH ester (Watanabe et al., 1998). Alternatively, the profile, inhibition by detergents and sulfhydryl ethanolamine product was derivatized with o-ph- reagents, and reversibility (Schmid et al., 1985). thaldialdehyde, and the derivative was quantified Many breakthroughs and exciting discoveries by high-performance liquid chromatography have been reported since these early days. FAAH monitoring absorbance at 230 nm (Qin et al., was also shown to have an esterase activity with 1998). A fluorescence displacement assay has also the ability to hydrolyze 2-AG (Di Marzo et al., been reported (Thumser et al., 1997). 1998; Goparaju et al., 1998) and an amidase for oleamide (Maurelli et al., 1995). An important 2.2. Enzyme purification advance in our understanding of FAAH was the study where it was cloned and shown to belong to FAAH is a membrane-bound protein found a family of amidases with similar domains in the predominantly in microsomal and mitochondrial catalytic site (Cravatt et al., 1996). In this review, fractions (Schmid et al., 1985; Desarnaud et al., the first devoted solely to FAAH, we describe the 1995; Hillard et al., 1995; Ueda et al., 1995). The enzymological and molecular properties of the enzyme can be solubilized from the membrane enzyme, its distribution in organs and tissues, and with the aid of detergents such as sodium tau- the inhibitors that have been developed for rodeoxycholate (Schmid et al., 1985) and Triton FAAH. X-100 (Ueda et al., 1995; Cravatt et al., 1996). The enzyme solubilized from porcine brain micro- somes was purified 22-fold by hydrophobic chro- 2. Catalytic properties matography using a Phenyl-5PW column (Ueda et al., 1995). The enzyme solubilized from rat liver 2.1. Assays plasma membrane was purified approximately 20- fold by a combination of DEAE, organomercu- For the detection of the anandamide hydrolyz- rial, and heparin columns (Cravatt et al., 1996). ing activity, radioactive anandamide is generally The latter enzyme preparation was further used as substrate. When [arachidonoyl-1-14C] or purified by affinity chromatography using oleyl [arachidonoyl-5,6,8,9,11,12,14,15-3H] anandamide trifluoromethyl ketone as a , and was uti- was used as substrate, the radioactive arachidonic lized for sequencing of the enzyme acid produced by hydrolysis was separated from (Cravatt et al., 1996). N. Ueda et al. / Chemistry and Physics of Lipids 108 (2000) 107–121 109

Recently, recombinant rat FAAH was highly anandamide (Cravatt et al., 1996; Giang and Cra- purified. The enzyme with a hexahistidine tag was vatt, 1997; Kurahashi et al., 1997; Goparaju et al., overexpressed in Escherichia coli (Patricelli et al., 1999). The enzyme can hydrolyze not only 1998a) or in a baculovirus-Sf9 insect cell system oleamide but also primary amides of other fatty (Katayama et al., 1999) and these enzyme prepa- acids (Cravatt et al., 1996; Giang and Cravatt, rations were purified by cobalt- or nickel-charged 1997). Their hydrolytic rates were in the order of \ \ resin. The kcat value of the recombinant enzyme oleamide myristamide palmitamide. Arachi- was 7.1 s−1 (E. coli, oleamide substrate) or 6.5 donamide (primary amide of arachidonic acid) s−1 (baculovirus system, anandamide substrate). was later shown to be hydrolyzed at the highest m The Km for anandamide was 2–67 M depending rate (Kurahashi et al., 1997; Lang et al., 1999). upon the enzyme preparations and assay condi- Various synthetic anandamide derivatives were tions (Desarnaud et al., 1995; Hillard et al., 1995; also tested as substrates for FAAH by reverse- Maurelli et al., 1995; Omeir et al., 1995; Ueda et phase high performance liquid chromatography al., 1995; Bisogno et al., 1997a; Katayama et al., (Lang et al., 1999). It should be noted that (R)- 1999; Maccarrone et al., 1999a). The optimum pH , an anandamide analogue resis- was 8.5–10 (Hillard et al., 1995; Maurelli et al., tant to the hydrolysis by FAAH, showed 1995; Omeir et al., 1995; Ueda et al., 1995; long-lasting biological activity in vivo (Romero et Bisogno et al., 1997a; Goparaju et al., 1998; al., 1996). Katayama et al., 1999). In addition to its amidase activity, FAAH also has an esterase activity for monoacylglycerols (Di 2.3. Substrate specificity Marzo et al., 1998; Goparaju et al., 1998, 1999) and methyl esters of fatty acids (Kurahashi et al., The relative reactivities of FAAH with various 1997; Goparaju et al., 1999; Patricelli and Cra- substrates are shown in Table 1. Although the vatt, 1999). 2-AG is another endogenous agonist first report of anandamide acting as a substrate for the cannabinoid (Mechoulam et al., for this enzyme was made in 1993 (Deutsch 1995; Sugiura et al., 1995) and was hydrolyzed at and Chin, 1993) there was an earlier report of a rate several-fold faster than anandamide by this activity breaking down other N-acyl- recombinant rat and porcine FAAH (Goparaju et ethanolamines (Schmid et al., 1985). As the chain al., 1998, 1999). 1(3)-Arachidonoylglycerol was as length of saturated fatty acids was varied between active as 2-AG while 1-oleoylglycerol was less

C12 and C18, the rat liver enzyme hydrolyzed active. 2-AG and racemic 1-arachidonoylglycerol N-acylethanolamines of shorter-chain fatty acids were also shown to be excellent FAAH substrates faster than those of longer-chain fatty acids. in rat brain microsomal fractions (Lang et al., Amides of propanolamine or higher homologs (up 1999) while 1,2-diacylglycerol was inactive (Go- to C6) were hydrolyzed at drastically slower rates paraju et al., 1998). than the amides of ethanolamine (Schmid et al., 1985). When the partially purified enzyme from 2.4. Re6ersibility of anandamide hydrolysis porcine brain was allowed to react with various N-acylethanolamines, anandamide was the best An energy-independent formation of anan- substrate followed by ethanolamides of linoleic, damide by the condensation of arachidonic acid oleic and palmitic acids (Ueda et al., 1995). with ethanolamine was reported previously Oleamide (primary amide of oleic acid) and (Deutsch and Chin, 1993; Devane and Axelrod, anandamide were shown to be hydrolyzed by a 1994; Kruszka and Gross, 1994). The formation single enzyme in mouse neuroblastoma cells of various other fatty acid ethanolamides, but not (Maurelli et al., 1995). This finding was confirmed anandamide, was described much earlier (Bachur with recombinant rat FAAH (Cravatt et al., and Udenfriend, 1966; Schmid et al., 1985). Ueda 1996). The rate of oleamide hydrolysis by the and colleagues revealed that the ‘anandamide syn- recombinant enzyme was 31–78% the rate for thase’ co-migrated with FAAH with hydrophobic 110 N. Ueda et al. / Chemistry and Physics of Lipids 108 (2000) 107–121 j – – – – – – – – – 31 recombinant Porcine j – 78 recombinant i recombinant h Humanrecombinant Rat g – – 912 – –– – – – 85 560 89 354 57 73 70 67 1033––– Rat Rat recombinant f ––––– ––––– ––––– –– – – – – – 11–––– 30–––– 33–––– 21–––– -oleoylethanolamine was considered to be 100%. RBL-2H3 (%) N e 2 TG – – – 93 58 18 N d – – – – – –– 27 Porcine brain c – – – – –– – –– – –– – ––– – – 42 – 35 45 100 100 Ratbrain Mouse b – –65––––– – – 100 100 100 100 100 100 –35– – –2 – – – – – 6275––––– – –– –––––– – 2465––– 43–10––––– 71 95 –––––– 100 119 a Rat liver -linolenoylethanol- g -linolenoylethanol- Schmid et al., 1985. Ueda et al., 1995. Giang and Cravatt, 1997. With each enzyme preparation the activity of anandamide or Cravatt et al., 1996. Desarnaud et al., 1995. Maurelli et al., 1995. Bisogno et al., 1997a. Kurahashi et al., 1997. Goparaju et al., 1999. g amine (C22:6) (C22:1) a b c d e f g h i j amine (C20:3) amine (C20:1) amine (C:20:2) (C18:2) amine (C18:3) (C18:0) (C18:1) (C16:0) (C12:0) (C14:0) -docosahexaenoylethanol- – -erucoylethanolamine-homo- 24 -eicosadienoylethanol- -eicosamonoenoylethanol- - -linoleoylethanolamine 44 -stearoylethanolamine – -oleoylethanolamine – – -palmitoylethanolamine-lauroylethanolamine 1 19 -myristoylethanolamine N Enzyme source Rat N Anandamide (C20:4)N 100 Table 1 Substrate specificity of FAAH N N N N N N N N Arachidonamide (C20:4) – N Linoleamide (C18:2) 64 Oleamide (C18:1) Methyl arachidonate – – 2-Arachidonoylglycerol – – – – – Myristamide (C14:0) Palmitamide (C16:0) – – N. Ueda et al. / Chemistry and Physics of Lipids 108 (2000) 107–121 111 and anion-exchange chromatography. Further- ingly, significant synthesis of anandamide was more, the synthase and hydrolase activities were reported in rat testes incubated with as little as observed essentially in parallel in pH profile, heat 10–50 mM ethanolamine (Schmid et al., 1998). inactivation, and effects of inhibitors. Thus, both the activities were attributable to a single enzyme 2.5. Another fatty acid amide hydrolase (Ueda et al., 1995). Additional evidence that the two activities are due to a single enzyme was As described below, the enzymatic hydrolysis of demonstrated in rat testes membranes (Schmid et anandamide occurs in a variety of mammalian al., 1998). The reversibility of the enzymatic anan- organs and cells. Since the organ distribution of damide hydrolysis was demonstrated with the re- FAAH mRNA in rat was similar to that of the combinant rat and porcine FAAH overexpressed anandamide hydrolyzing activity, the FAAH en- in COS-7 cells (Kurahashi et al., 1997; Arreaza zyme seemed to play a major role in anandamide and Deutsch, 1999; Goparaju et al., 1999). A hydrolysis (Katayama et al., 1997). However, re- recent experiment using a large amount of a cently Ueda and colleagues revealed that human purified recombinant rat enzyme demonstrated megakaryoblastic leukemia cells (CMK) have an clearly that the reverse reaction occurs. The equi- amidase that hydrolyzes anandamide but is cata- librium constant ([arachidonic acid] [ethanol- lytically distinct from FAAH. When the cell ho- amine])/([anandamide] [water]) was determined to mogenate was subjected to sequential centri- be 4×10−3 at pH 9 (37°C) (Katayama et al., fugation, the enzyme was recovered mainly in the 1999). The reversibility of the N-oleoylethanol- 12 000×g pellet. The enzyme was easily solubi- amine hydrolysis was also suggested previously lized by freezing and thawing without detergent. with the rat liver enzyme (Bachur and Uden- The solubilized enzyme showed an optimal pH friend, 1966; Schmid et al., 1985). near 5, and was almost inactive at alkaline pH,

Since the Km value for ethanolamine is ex- where FAAH was most active. The enzyme activ- tremely high with various enzyme preparations ity was stimulated by dithiothreitol, and showed (27–50 mM), most investigators have concluded low sensitivity to phenylmethylsulfonyl fluoride that anandamide is not produced in the reverse (PMSF) and methyl arachidonyl fluorophospho- reaction with FAAH under physiological condi- nate (MAFP) which are potent inhibitors for tions (Devane and Axelrod, 1994; Ueda et al., FAAH. Furthermore, N-palmitoylethanolamine 1995; Sugiura et al., 1996b; Kurahashi et al., was hydrolyzed about 1.5-fold faster than anan- 1997). On the basis of the equilibrium constant damide. These results demonstrate the presence of described above, if 100 mM arachidonic acid and another amidase distinguishable from FAAH 100 mM ethanolamine were present in the reaction (Ueda et al., 1999). An enzyme showing similar mixture, the anandamide/arachidonic acid ratio at catalytic properties was found in several rat or- equilibrium is calculated to be 0.05:99.95 and a gans such as lung and spleen (Ueda et al., unpub- value close to this ratio was obtained experimen- lished data). tally (Katayama et al., 1999). Therefore, anan- damide formation catalyzed by FAAH cannot be rigorously ruled out if localized cellular levels of 3. Molecular properties arachidonic acid and ethanolamine reach high concentrations. The equilibrium was dependent 3.1. Protein domains on pH; namely, acidic pH was unfavorable to anandamide formation. Similarly, oleamide (Sug- Studies using bacterial amidases defined a iura et al., 1996a; Bisogno et al., 1997b; Kura- highly conserved region termed the amidase signa- hashi et al., 1997), methyl arachidonate ture sequence (Mayaux et al., 1991). When the (Kurahashi et al., 1997), and monoarachidonoyl- first mammalian FAAH cDNA was cloned, it was glycerol (Goparaju et al., 1998) can be formed in also aligned to the bacterial amidases and shown the reverse reaction, but at low rates. Interest- to contain an amidase signature sequence (Cra- 112 N. Ueda et al. / Chemistry and Physics of Lipids 108 (2000) 107–121 vatt et al., 1996). A strong evolutionary relation- volved in catalysis, the genomic organization, and ship exists between the amidases from bacteria chromosomal location of FAAH. and yeast, and the mammalian FAAH enzymes in The human, porcine, mouse, and rat FAAH the amidase sequence. Since the cloning of the enzymes are 73% identical at the amino acid level first FAAH cDNA, many insights have followed overall and 90% identical in the amidase signature including the domain structure, amino acids in- sequence, amino acids 215–257 (Fig. 1). Although

Fig. 1. Alignment of mammalian FAAH enzymes. The amino acid alignment for human (GenBank U82535) (Giang and Cravatt, 1997) porcine (GenBank AB027132) (Goparaju et al., 1999), mouse (GenBank U82536) (Giang and Cravatt, 1997), and rat (GenBank U72497) (Cravatt et al., 1996) FAAH proteins is shown. The amidase signature sequence is boxed in gray (amino acids 215–257). Amino acids whose mutation caused significant catalytic defects are denoted by arrows. N. Ueda et al. / Chemistry and Physics of Lipids 108 (2000) 107–121 113 the mammalian FAAH mRNAs differ in size acids 307–315) that is homologous to the class II across species, the mammalian FAAH proteins SH3-binding domain (Arreaza and Deutsch, are all 579 amino acids in length. The mammalian 1999). When this region was deleted from FAAH FAAH proteins have several domains defined by (DPR), the resulting mutant had no measurable homology and functional studies. In the amino catalytic activity and abnormal subcellular distri- terminus, there is a predicted transmembrane do- bution. These studies suggest several independent main (Cravatt et al., 1996). Recombinant forms of functional domains exist in FAAH and further WT-FAAH (wild-type) and DTM-FAAH (1–30 research is needed to identify and further charac- amino acids deleted) exhibited essentially identical terize them. enzymatic properties and they are similar to that of the native enzyme from rat liver. However, 3.2. Mutation analysis WT-FAAH consistently behaved as a larger oligomer than DTM-FAAH. Sodium dodecylsul- Recent work by several groups has begun to fate-polyacrylamide gel electrophoresis (SDS- define the amino acids responsible for catalysis by PAGE) identified the presence of SDS-resistant FAAH. Since inhibitors used in early research of oligomers for WT-FAAH, but not for DTM- FAAH were known to inhibit serine and cystine FAAH. Self-association through FAAH’s proteases, it was suggested that FAAH had a transmembrane domain was further demonstrated serine or cystine nucleophile at its reaction center. by a WT-FAAH transmembrane domain-GST fu- This theory was tested in 1999 in three indepen- sion protein that formed dimers and large dent studies involving both the rat and porcine oligomeric assemblies in solution (Patricelli et al., FAAH enzymes (Goparaju et al., 1999; Omeir et 1998a). The DTM-FAAH deletion has also been al., 1999; Patricelli et al., 1999). Three used extensively to characterize biochemically the from the amidase signature sequence were mu- enzyme since it is both easier to purify and enzy- tated in all three studies. In each study, the serine matically indistinguishable from WT-FAAH. In a to mutation of S217, S218, and S241 similar study, a mutant lacking an N-terminal significantly reduced the catalytic power of hydrophobic domain missing amino acids 9 to 29 FAAH. These studies showed that the S218A (DHD1-FAAH) was studied (Arreaza and mutation was not as dramatic as the other two Deutsch, 1999). WT-FAAH and DHD1-FAAH serine mutations but a double S217A–S218A mu- were indistinguishable after immunostaining for tation had even greater catalytic defects than ei- their subcellular localization in COS-7 transfected ther of the single serine mutations alone cells. Furthermore, with DHD1-FAAH only a suggesting a role for both of these amino acids in small amount of enzymatic activity was observed catalysis. However, only S241 was affinity-labeled in the soluble fractions suggesting little release of demonstrating that it is the catalytic nucleophile the enzyme from the membranes. Therefore, it is (Patricelli et al., 1999). Several other amino acids likely that deletion of the N-terminal transmem- were also mutated in these studies including brane region is not sufficient to solubilize FAAH S152A, H184Q, H358A, and H449A. None of because other regions interact with the membrane these mutations had significant effects on FAAH (Patricelli et al., 1998a; Arreaza and Deutsch, activity suggesting that FAAH does not use a 1999). typical serine–histidine– triad found The next domain of interest is the amidase in serine hydrolytic enzymes for catalytic activity. signature sequence of FAAH, which was located The catalytic properties of one amino acid, C249, by homology to other known amidases remain in question. In the rat FAAH protein, (Kobayashi et al., 1997). This domain is serine conversion of C249A does not impair FAAH and rich and is approximately 50 amino activity (Omeir et al., 1999) while in porcine acids in length (amino acids 215–257 in mam- FAAH, the same mutation leads to an enzyme malian FAAH proteins). Finally, the third do- that is at least 1000-fold less active than the main of FAAH is a proline rich region (amino wild-type FAAH (Goparaju et al., 1999). In the 114 N. Ueda et al. / Chemistry and Physics of Lipids 108 (2000) 107–121

Table 2 Distribution of the FAAH activity in various organs and brain regions

Animal Substrate Organs or brain regions References

Rat N-oleoyl- Liver\brain\testis\kidney\lung\spleen\\heart Schmid et al., ethanolamine 1983 Rat Anandamide Liver\brain\other organs Desarnaud et al., 1995 Rata Anandamide Liver\cerebrum\cerebellum\testis\parotid gland\kidney Katayama et \submaxillarygland\small intestine\stomach=lung\spleen (esophagus, al., 1997 heart and skeletal muscle not detectable) Ratb Anandamide Liver\small intestine\testis\cerebrum \stomach Katayama et \colon\cerebellum\parotid gland=submaxillary al., 1997 gland\lung\kidney\spleen (esophagus, heart and skeletal muscle not detectable) Mouse Anandamide Liver\brain\testispancreas\spleen (heart and lung not detectable) Watanabe et al., 1998 Rat Anandamide Globus pallidus\hippocampus\substantia nigra\striatum Desarnaud et \thalamus\cerebellum\cortex\brain stem\medulla al., 1995 Rat Anandamide Hippocampus\cerebellum\cerebral cortex\hypothalamus\white Hillard et al., matter\striatum\brain stem 1995 RatOleamide Hippocampus\cortex\cerebellum\thalamus \olfactory bulb\ Thomas et al., striatum\hypothalamus\brain stem\pituitary 1997

a Native homogenates were assayed. b Acetone-treated homogenates were assayed.

Rhodococcus rhodochrous J1 amidase, conversion In summary, several amino acids involved in catal- of C203A, which corresponds to C249 in FAAH, ysis (S217 and S218), an active nucleophile (S241) resulted in an amidase with 10% of wild type and an amino acid responsible for substrate spe- activity (Kobayashi et al., 1997). The authors of cificity (L142) have been determined in FAAH this bacterial study concluded that C203 is not a proteins. part of the active site. Interestingly, FAAH has been shown to be inhibited by sulfhydryl reagents 3.3. FAAH mRNA and genomic loci (Schmid et al., 1985; Patterson et al., 1996; De Petrocellis et al., 1998). It has been postulated that The expression pattern of FAAH was initially the inhibition of FAAH by thiol reagents may be characterized in 1993 but the first cDNA was not caused by steric hinderance at the active site or by isolated from a rat liver library until several years disruption of disulfide bonds (Omeir et al., 1999). later due to the difficulty in purifying FAAH In addition, mutational analysis of porcine FAAH (Cravatt et al., 1996). Shortly thereafter, the hu- included D237E and D237N (Goparaju et al., man and mouse FAAH cDNAs were isolated 1999). Both of these mutations were shown to using the rat cDNA as a probe (Giang and Cra- decrease activity by more than 1000-fold over vatt, 1997). Each species varies slightly in the size wild-type FAAH. Interestingly, the D237E muta- of its FAAH mRNA with the human mRNA tion in rat FAAH had no effect on activity while measuring 2.1 kb, the mouse mRNA at 1.7 kb, the D237N mutation was so poorly expressed that and the rat mRNA at 2.5 kb. In addition, the its activity could not be determined (Deutsch lab- porcine FAAH cDNA has been reported recently oratory, unpublished data). In a separate muta- (Goparaju et al., 1999). The human and mouse tional analysis, the L142A mutation changed genomic loci have also been isolated and charac- FAAH’s substrate preference from amide\ester terized (Wan et al., 1998). The human FAAH gene into ester\amide (Patricelli and Cravatt, 1999). is located on 1p34–35 and the mouse FAAH gene N. Ueda et al. / Chemistry and Physics of Lipids 108 (2000) 107–121 115 is located on chromosome 4 in a region syntenic to immunoreactivity of FAAH was most prominent human chromosome 1p. The FAAH genomic loci in large ganglion cells with weaker immunoreactiv- from both species contain 15 exons with highly ity in several other types of cells (Yazulla et al., conserved intron–exon boundaries and the ami- 1999). dase signature sequence is located in exon 4 of both Regional distribution of the anandamide hy- species. drolyzing activity in rat brain was also examined in detail (Table 2). Hippocampus, cerebellum, and cerebral cortex showed a relatively high activity 4. Distribution (Hillard et al., 1995; Thomas et al., 1997). Northern and Western blotting revealed that the regional Organ distribution of the anandamide hydrolyz- distribution of FAAH mRNA and protein of ing activity has been examined in detail with rats FAAH was well correlated with that of the enzyme (Table 2). Among a variety of organs with the activity (Thomas et al., 1997; Egertova et al., 1998). enzyme activity, liver showed by far the highest In situ hybridization revealed a wide distribution of activity, followed by brain and testis (Desarnaud et FAAH mRNA in neuronal cells throughout the al., 1995; Katayama et al., 1997). A similar distri- central nervous system (Thomas et al., 1997). bution was earlier reported with N- Immunohistochemical study showed the strongest oleoylethanolamine as an amidase substrate FAAH immunoreactivity in large principal neu- (Schmid et al., 1983). These authors also reported rons, such as pyramidal cells in cerebral cortex and that FAAH activity was extremely low in the heart hippocampus, Purkinje cells in cerebellar cortex, and this has been confirmed by others (Deutsch and mitral cells in olfactory bulb (Tsou et al., and Chin, 1993; Cravatt et al., 1996). Any activity 1998). The regional distribution of FAAH corre- detected in the heart probably results from FAAH lated with that of CB1, but FAAH and CB1 were localized in endothelial cells lining the blood vessels expressed in different cell types in each brain region (Deutsch et al., 1997a; Maccarrone et al., 2000a). (Egertova et al., 1998). During development, The homogenate of small intestine showed a low FAAH mRNA in rat brain increased progressively anandamide hydrolyzing activity (Desarnaud et al., between embryonic day 14 and postnatal day 10, 1995; Katayama et al., 1997). However, removal of remained high until postnatal day 30, and then endogenous lipids by acetone precipitation remark- decreased to the adult level (Thomas et al., 1997). ably increased the activity (Katayama et al., 1997). The presence of FAAH in primary cells and cell As analyzed by Northern blotting, the distribution lines is not universal. For example, FAAH activity of FAAH mRNA in rat organs was almost consis- is not detectable in human epithelioid carcinoma, tent with that of the anandamide hydrolyzing HeLa, human larynx epiermoid carcinoma, Hep2, activity (Cravatt et al., 1996; Katayama et al., human hepatocellular carcinoma, HepG2, or hu- 1997). In human, FAAH mRNA showed a consid- man astrocytoma cells, CCF-STTG1 (Deutsch and erably different distribution; namely, the mRNA Chin, 1993; Piomelli et al., 1999). FAAH activity was abundant in pancreas, brain, kidney and skele- and/or FAAH mRNA distribution in isolated tal muscle, poor in liver and placenta, and unde- mammalian cells or established cell lines is summa- tectable in heart and lung (Giang and Cravatt, rized in Table 3. 1997). In addition to the organs listed in Table 2, the anandamide hydrolyzing activity was found in mouse uterus (Paria et al., 1996) and its expression 5. Inhibitors is regulated during pregnancy (Paria et al., 1999; Maccarrone et al., 2000b). The eye of several 5.1. Re6ersible inhibitors species also contains FAAH activity including the porcine retina, choroid, optic nerve, iris, and Analogs of anandamide representing three lacrimal glands (Matsuda et al., 1997) and the classes of putative transition-state inhibitors bovine retina (Bisogno et al., 1999). In rat retina, (trifluoromethyl ketones, a-keto esters, and a-keto 116 N. Ueda et al. / Chemistry and Physics of Lipids 108 (2000) 107–121 amides) were synthesized and tested as inhibitors of covalently the FAAH at concentrations 1000 times anandamide hydrolysis in vitro and as ligands for lower than those reported for phospholipaseA2 CB1 (Koutek et al., 1994). The trifluoromethyl inhibition and it was determined to be the most ketones and a-keto esters showed nearly 100% potent FAAH inhibitor so far described inhibition of anandamide hydrolysis in vitro in the (IC50 =1–3 nM, De Petrocellis et al., 1997a,b). low mM range. None of the compounds in the series ADMK was found to inhibit FAAH both re- of fatty acid derivatives bound appreciably to the versibly and irreversibly depending upon whether at 10 mM except the enzyme was membrane-associated or in a arachidonyltrifluoromethyl ketone (ATFMK). The solublilized form (Edgemond et al., 1998). In rat action of these inhibitors on FAAH is consistent brain membranes, ADMK inhibits the enzyme m with the studies showing that anandamide is with an IC50 of 0.5 M. At low concentrations, cleaved by a mechanism that involves an active-site ADMK reduces the Vmax and increases the Km of serine hydroxyl group in the enzyme (S241). In the enzyme for anandamide, and at higher con- related studies, the a-keto amides, a-keto ethyl centrations ADMK inhibition is completely non- esters, and trifluoromethyl ketones of oleic acid competitive. ADMK inhibition is irreversible were found to be the most potent inhibitors of with membrane-associated FAAH since hy- FAAH using oleamide as a substrate (Patterson et drolytic activity is not restored with extensive al., 1996). washing, dialysis of the membranes, or by anion Some very informative studies were conducted to exchange chromatography of the subsequently determine which classes of FAAH inhibitors solubilized enzyme. In contrast, ADMK inhibi- reacted reversibly. Three arachidonic acid analogs tion of detergent-solubilized enzyme exhibits were synthesized — arachidonoyldiazomethyl- competitive kinetics and is reversible upon ion ketone (ADMK), arachidonoylchloromethyketone exchange chromatography. Exposure of C6 (ACMK), and O-acetyl-arachidonoylhydroxamate glioma cells to ADMK resulted in concentration- (AcAHA). They were tested in mammalian brain related inhibition of FAAH activity in cellular m homogenates, mouse neuroblastoma cells, and rat membranes with an IC50 value of 0.3 M. Bro- basophilic leukemia cells and it was found that each moenol lactone, (E)-6-(bromomethylene) tetrahy- compound produced a significant inhibition, with dro-3-(1-naphthalenyl)-2H-pyran-2-one inhibited

ADMK being the most potent (IC50 =3, 2, and 6 FAAH in rat brain microsomes with an IC50 of m m M) and AcAHA the weakest (IC50 =34, 15, and 0.8 M. Kinetic and dialysis experiments indi- 25 mM) inhibitors. Inhibition by these three cated that this effect was non-competitive and compounds was shown to be reversible by irreversible (Beltramo et al., 1997). Other inter- anion-exchange and furthermore, Lineweaver– esting inhibitors have also been described. For Burk profiles indicated competitive inhibition for example, 2-octyl g-bromoacetoacetate, an en- each compound. Conversely, MAFP inhibited dogenous compound originally isolated from

Table 3 Mammalian cells showing the anandamide hydrolyzing activity or expressing FAAH mRNA

Animal Cell

Human H358 bronchioalveolar non-small cell lung carcinoma (Deutsch and Chin, 1993) EFM-19 breast carcinoma cell (Bisogno et al., 1998) U937 monocytic cell (Maccarrone et al., 1998) CHP100 neuroblastoma cell (Maccarrone et al., 1998) Platelet (Maccarrone et al., 1999b) Rat C6 glioma cell (Deutsch and Chin, 1993) RBL-2H3 basophilic leukemia cell (Bisogno et al., 1997a) RBL-1 basophilic leukemia cell (Bisogno et al., 1997a) PC-12 adrenal pheochromocytoma cell (Bisogno et al., 1998) renal endothelial cell (Deutsch et al., 1997a) mesangial cell (Deutsch et al., 1997a) macrophage (Di Marzo et al., 1999a) Mouse N18TG2 cell (Deutsch and Chin, 1993; Maurelli et al., 1995) N. Ueda et al. / Chemistry and Physics of Lipids 108 (2000) 107–121 117 human cerebrospinal fluid was found to competi- lfluorides had high selectivity for the amidase tively inhibit FAAH activity with an IC50 and Ki versus CB1 and are potentially useful as selective of 2.6 and 0.8 mM, respectively. When synthetic FAAH inhibitors (Deutsch et al., 1997b). When analogs of 2-octyl g-bromoacetoacetate were as- MAFP was tested as an inhibitor of FAAH, it was sayed at pH 7.0, the g-halo b-keto ester inhibitors 1000 times more potent than ATFMK in rat brain proved to be significantly more potent than the homogenates and neuroblastoma cells in vitro. trifluoromethyl ketone of oleic acid (Patricelli et MAFP demonstrated selectivity towards FAAH al., 1998). for which it was approximately 3000 and 30 000- fold more potent than it was towards chy- 5.2. Irre6ersible inhibitors motrypsin and trypsin, respectively (Deutsch et al., 1997c). PMSF was discovered serendipitously to be an inhibitor of FAAH. When PMSF was added to 5.3. Other inhibitors brain and cell culture homogenates, it was expected to protect FAAH against proteolytic degradation, In another study, it was found that FAAH instead it inhibited the enzyme (Deutsch and Chin, activity in mouse brain microsome was inhibited 1993). Similarly, when PMSF was added to rat by D9- (THC), , brain membranes as a general serine protease and at high micromolar concentrations inhibitor, it protected anandamide against degra- (Watanabe et al., 1996, 1998). Additionally, dation in a cannabinoid receptor binding assay FAAH was studied in the presence of a select (Childers et al., 1994). In rat brain homogenate group of cannabimimetics to identify inhibitors. Of FAAH was found to be inhibited by PMSF, these, (−)D8-THC and SR141716A, a CB1 antag- diisopropylfluorophosphate (DFP), and onist, were found to inhibit enzymatic activity by thimerosal (Hillard et al., 1995). FAAH partially approximately 23 and 30% at 30 mM (Lang et al., purified from porcine brain was inhibited by 1999). Ibuprofen and other non-steroidal anti-infl- ATFMK, p-chloromercuribenzoic acid, DFP, and ammatory drugs (NSAID) also inhibit FAAH. In PMSF (Ueda et al., 1995). DFP has been shown to rat brain homogenates, at 2 mM anandamide, the m inhibit FAAH in COS-7 cells that express recombi- IC50 were as follows — suprofen, 170 M; ibupro- nant FAAH (Omeir et al., 1999). fen, 270 mM; fenoprofen, 480 mM; naproxen, 550 Recognizing that PMSF and DFP are nonspe- mM; ketoprofen, 650 mM; and diclofenac, approx- cific inhibitors of FAAH, a series of studies were imately 1000 mM. It was concluded that following undertaken where various fatty acids incorporated therapeutic doses of ibuprofen, the breakdown of a sulfonyl fluoride or a fluorophosphate moiety to anandamide may be reduced (Fowler et al., 1997). increase the specificity of the inhibitors for FAAH. The same group also showed that there was no

A series of fatty acids (C12 –C20) sulfonyl fluorides dramatic enantiomeric selectivity of NSAID com- were evaluated as inhibitors of anandamide degra- pounds as inhibitors of FAAH. The enantiomers dation and as ligands for CB1. AM374 (palmityl- of flurbiprofen and R-ketorolac are the most po- sulfonyl fluoride, C16) was approximately 20 times tent NSAID inhibitors yet reported to date more potent than PMSF and 50 times more potent (Fowler et al., 1999). Recently, four novel in- than ATMFK in preventing the hydrolysis of hibitors selective for FAAH have been described anandamide in brain homogenates. AM374 was — malhamensilipin, grenadadiene, arachi- over a 1000-fold more effective than PMSF in donoylethylene glycol and arachidonoyl- inhibiting the amidase in cultured cells with (AA-5-HT). Using mouse neuroblastoma FAAH nanomolar IC50 values. These compounds gener- preparations, the IC50s for these compounds ally had a decreasing affinity for CB1 as the chain ranged from 12.0 to 26 mM, with the most active length increased; thus, C12 sulfonylfluoride had an compound being AA-5-HT. AA-5-HT did not IC50 of 18 nM and C20 sulfonylfluoride had an IC50 interfere with cytosolic phospholipaseA2-mediated, m of 78 M for CB1. The C14,C16, and C18 sulfony- ionomycin, or antigen-induced release of 118 N. Ueda et al. / Chemistry and Physics of Lipids 108 (2000) 107–121

[3H]arachidonic acid from RBL-2H3 cells. In addi- Bisogno, T., Delton-Vandenbroucke, I., Milone, A., Lagarde, tion, AA-5-HT did not inhibit cytosolic M., Di Marzo, V., 1999. Biosynthesis and inactivation of N-arachidonoylethanolamine (anandamide) and N-docosa- phospholipaseA2- activity in cell-free experiments hexaenoylethanolamine in bovine retina. Arch. Biochem. and it did not activate CB1 (Bisogno et al., 1998). Biophys. 370, 300–307. FAAH in rat brain microsomes was inhibited by Bisogno, T., Maurelli, S., Melck, D., De Petrocellis, L., Di p-bromophenacyl bromide, a histidine-alkylating Marzo, V., 1997a. Biosynthesis, uptake, and degradation of anandamide and in leukocytes. reagent while N-ethylmaleimide and various nonse- J. Biol. Chem. 272, 3315–3323. lective peptidase inhibitors (EDTA, o-phenanthro- Bisogno, T., Sepe, N., De Petrocellis, L., Mechoulam, R., Di line, bacitracin) had no effect (Desarnaud et al., Marzo, V., 1997b. The sleep inducing factor oleamide is 1995). Other protease inhibitors such as aprotinin, produced by mouse neuroblastoma cells. Biochem. Bio- phys. Res. Commun. 239, 473–479. benzamidine, leupeptin, chymostatin, and pep- Bisogno, T., Melck, D., De Petrocellis, L., Bobrov, M., Gret- statin also have no effect on FAAH activity skaya, N.M., Bezuglov, V.V., Sitachitta, N., Gerwick, (Deutsch and Chin, 1993). This is consistent with W.H., Di Marzo, V., 1998. Arachidonoylserotonin and the classification of FAAH as a member of the other novel inhibitors of fatty acid amide hydrolase. Biochem. Biophys. Res. Commun. 248, 515–522. amidase family that is distinct from other serine Childers, S.R., Sexton, T., Roy, M.B., 1994. Effects of anan- proteases (Cravatt et al., 1996). damide on cannabinoid receptors in rat brain membranes. Biochem. Pharmacol. 47, 711–715. Cravatt, B.F., Giang, D.K., Mayfield, S.P., Boger, D.L., Lerner, R.A., Gilula, N.B., 1996. Molecular characteriza- 6. Conclusion tion of an enzyme that degrades neuromodulatory fatty- acid amides. Nature 384, 83–87. In conclusion FAAH is a well-characterized De Petrocellis, L., Melck, D., Ueda, N., Kurahashi, Y., enzyme responsible for the hydrolysis of biologi- Bisogno, T., Yamamoto, S., Di Marzo, V., 1997a. Brain and peripheral anandamide amidohydrolase and its inhibi- cally active lipids including anandamide, 2-AG, tion by synthetic arachidonate analogues. Adv. Exp. Med. and oleamide. While these studies have contributed Biol. 433, 259–263. greatly to comprehending the mechanism of FAAH De Petrocellis, L., Melck, D., Ueda, N., Maurelli, S., Kura- catalysis, more research into the molecular proper- hashi, Y., Yamamoto, S., Marino, G., Di Marzo, V., ties of FAAH is needed. Further studies of the 1997b. Novel inhibitors of brain, neuronal, and basophilic anandamide amidohydrolase. Biochem. Biophys. Res. amino acids involved in catalysis may explain the Commun. 231, 82–88. differences between the activity of C249 in the rat De Petrocellis, L., Melck, D., Palmisano, A., Bisogno, T., and porcine FAAH enzymes. A crystallographic Laezza, C., Bifulco, M., Di Marzo, V., 1998. The endoge- analysis of FAAH would enhance the understand- nous cannabinoid anandamide inhibits human breast can- cer cell proliferation. Proc. Natl. Acad. Sci. USA 95, ing of substrate specificity and could allow for the 8375–8380. rational design of specific FAAH inhibitors. A Desarnaud, F., Cadas, H., Piomelli, D., 1995. Anandamide genetic knockout of FAAH could also contribute amidohydrolase activity in rat brain microsomes. Identifi- significantly to the understanding of the endoge- cation and partial characterization. J. Biol. Chem. 270, 6030–6035. nous cannabinoid signaling pathways. Deutsch, D.G., Chin, S.A., 1993. Enzymatic synthesis and degradation of anandamide, a cannabinoid receptor ago- nist. Biochem. Pharmacol. 46, 791–796. References Deutsch, D.G., Makriyannis, A., 1997. Inhibitors of anan- damide breakdown. NIDA Res. Monogr. 173, 65–84. Deutsch, D.G., Goligorsky, M.S., Schmid, P.C., Krebsbach, Arreaza, G., Deutsch, D.G., 1999. Deletion of a proline-rich R.J., Schmid, H.H., Das, S.K., Dey, S.K., Arreaza, G., region and a transmembrane domain in fatty acid amide Thorup, C., Stefano, G., Moore, L.C., 1997a. Production hydrolase. FEBS Lett. 454, 57–60. and physiological actions of anandamide in the vasculature Bachur, N.R., Udenfriend, S., 1966. Microsomal synthesis of of the rat kidney. J. Clin. Invest. 100, 1538–1546. fatty acid amides. J. Biol. Chem. 241, 1308–1313. Deutsch, D.G., Lin, S., Hill, W.A., Morse, K.L., Salehani, D., Beltramo, M., Di Tomaso, E., Piomelli, D., 1997. Inhibition of Arreaza, G., Omeir, R.L., Makriyannis, A., 1997b. Fatty anandamide hydrolysis in rat brain tissue by (E)-6-(bro- acid sulfonyl fluorides inhibit anandamide metabolism and momethylene) tetrahydro-3-(1-naphthalenyl)-2H-pyran-2- bind to the cannabinoid receptor. Biochem. Biophys. Res. one. FEBS Lett. 403, 263–267. Commun. 231, 217–221. N. Ueda et al. / Chemistry and Physics of Lipids 108 (2000) 107–121 119

Deutsch, D.G., Omeir, R., Arreaza, G., Salehani, D., Prest- Hillard, C.J., Wilkison, D.M., Edgemond, W.S., Campbell, wich, G.D., Huang, Z., Howlett, A., 1997c. Methyl arachi- W.B., 1995. Characterization of the kinetics and distribu- donyl fluorophosphonate: a potent irreversible inhibitor of tion of N-arachidonylethanolamine (anandamide) hydroly- anandamide amidase. Biochem. Pharmacol. 53, 255–260. sis by rat brain. Biochim. Biophys. Acta 1257, 249–256. Devane, W.A., Axelrod, J., 1994. Enzymatic synthesis of anan- Katayama, K., Ueda, N., Katoh, I., Yamamoto, S., 1999. damide, an endogenous ligand for the cannabinoid recep- Equilibrium in the hydrolysis and synthesis of can- tor, by brain membranes. Proc. Natl. Acad. Sci. USA 91, nabimimetic anandamide demonstrated by a purified en- 6698–6701. zyme. Biochim. Biophys. Acta 1440, 205–214. Di Marzo, V., Bisogno, T., Sugiura, T., Melck, D., De Petro- Katayama, K., Ueda, N., Kurahashi, Y., Suzuki, H., Ya- cellis, L., 1998. The novel endogenous cannabinoid 2- mamoto, S., Kato, I., 1997. Distribution of anandamide arachidonoylglycerol is inactivated by neuronal- and amidohydrolase in rat tissues with special reference to basophil-like cells: connections with anandamide. Biochem. small intestine. Biochim. Biophys. Acta 1347, 212–218. J. 331, 15–19. Kobayashi, M., Fujiwara, Y., Goda, M., Komeda, H., Di Marzo, V., Bisogno, T., De Petrocellis, L., Melck, D., Shimizu, S., 1997. Identification of active sites in amidase: Orlando, P., Wagner, J.A., Kunos, G., 1999a. Biosynthesis evolutionary relationship between amide bond- and pep- and inactivation of the endocannabinoid 2-ararchidonoyl- tide bond-cleaving enzymes. Proc. Natl. Acad. Sci. USA glycerol in circulating and tumoral macrophages. Eur. J. 94, 11986–11991. Biochem. 264, 258–267. Koutek, B., Prestwich, G.D., Howlett, A.C., Chin, S.A., Sale- Di Marzo, V., De Petrocellis, L., Bisogno, T., Melck, D., hani, D., Akhavan, N., Deutsch, D.G., 1994. Inhibitors of 1999b. Metabolism of anandamide and 2-arachidonoyl- arachidonoyl ethanolamide hydrolysis. J. Biol. Chem. 269, glycerol: an historical overview and some recent develop- 22937–22940. ments. Lipids 34, S319–S325. Kruszka, K.K., Gross, R.W., 1994. The ATP- and CoA-inde- Di Marzo, V., Deutsch, D.G., 1998. Biochemistry of the pendent synthesis of arachidonoylethanolamide. A novel endogenous ligands of cannabinoid receptors. Neurobiol. mechanism underlying the synthesis of the endogenous Dis. Exp. Neurol. B 5, 386–404. ligand of the cannabinoid receptor. J. Biol. Chem. 269, Edgemond, W.S., Greenberg, M.J., Mcginley, P.J., Muthian, 14345–14348. S., Campbell, W.B., Hillard, C.J., 1998. Synthesis and Kurahashi, Y., Ueda, N., Suzuki, H., Suzuki, M., Yamamoto, characterization of diazomethylarachidonyl ketone: an ir- S., 1997. Reversible hydrolysis and synthesis of anan- reversible inhibitor of N-arachidonylethanolamine ami- damide demonstrated by recombinant rat fatty-acid amide dohydrolase. J. Pharmacol. Exp. Ther. 286, 184–190. hydrolase. Biochem. Biophys. Res. Commun. 237, 512– Egertova, M., Giang, D.K., Cravatt, B.F., Elphick, M.R., 515. 1998. A new perspective on cannabinoid signalling: com- Lang, W., Qin, C., Hill, W.A., Lin, S., Khanolkar, A.D., plementary localization of fatty acid amide hydrolase and Makriyannis, A., 1996. High-performance liquid chro- the CB1 receptor in rat brain. Proc. R. Soc. London B matographic determination of anandamide amidase activ- Biol. Sci. 265, 2081–2085. ity in rat brain microsomes. Anal. Biochem. 238, 40–45 Fowler, C.J., Janson, U., Johnson, R.M., Wahlstrom, G., 241, 274 (erratum). Stenstrom, A., Norstrom, K., Tiger, G., 1999. Inhibition of Lang, W., Qin, C., Lin, S., Khanolkar, A.D., Goutopoulos, anandamide hydrolysis by the enantiomers of ibuprofen, A., Fan, P., Abouzid, K., Meng, Z., Biegel, D., Makriyan- ketorolac, and flurbiprofen. Arch. Biochem. Biophys. 362, nis, A., 1999. Substrate specificity and stereoselectivity of 191–196. rat brain microsomal anandamide amidohydrolase. J. Fowler, C.J., Tiger, G., Stenstrom, A., 1997. Ibuprofen in- Med. Chem. 42, 896–902, 1682 (erratum). hibits rat brain deamidation of anandamide at pharmaco- Maccarrone, M., van del Stelt, M., Rossi, A., Veldink, G.A., logically relevant concentrations. Mode of inhibition and Vliegenthart, J.F.G., Agro, A.F., 1998. Anandamide hy- structure-activity relationship. J. Pharmacol. Exp. Ther. drolysis by human cells in culture and brain. J. Biol. Chem. 283, 729–734. 273, 32332–32339. Giang, D.K., Cravatt, B.F., 1997. Molecular characterization Maccarrone, M., Bari, M., Agro, A.F., 1999a. A sensitive and of human and mouse fatty acid amide hydrolases. Proc. specific radiochromatographic assay of fatty acid amide Natl. Acad. Sci. USA 94, 2238–2242. hydrolase activity. Anal. Biochem. 267, 314–318. Goparaju, S.K., Kurahashi, Y., Suzuki, H., Ueda, N., Ya- Maccarrone, M., Bari, M., Menichelli, A., Del Principe, D., mamoto, S., 1999. Anandamide amidohydrolase of porcine Agro, A.F., 1999b. Anandamide activates human platelets brain: cDNA cloning, functional expression and site-di- through a pathway independent of the arachidonate cas- rected mutagenesis. Biochim. Biophys. Acta 1441, 77–84. cade. FEBS Lett. 447, 277–282. Goparaju, S.K., Ueda, N., Yamaguchi, H., Yamamoto, S., Maccarrone, M., Bari, M., Lorenzon, T., Bisogno, T., Di 1998. Anandamide amidohydrolase reacting with 2-arachi- Marzo, V., Finazzi-Agro, A., 2000a. Anandamide uptake donoylglycerol, another cannabinoid receptor ligand. by human endothelial cells and its regulation by nitric FEBS Lett. 422, 69–73. oxide. J. Biol. Chem. 275, 13484–13492. 120 N. Ueda et al. / Chemistry and Physics of Lipids 108 (2000) 107–121

Maccarrone, M., Valensise, H., Bari, M., Lazzarin, N., Ro- Patterson, J.E., Ollmann, I.R., Cravatt, B.F., Boger, D.L., manini, C., Finazzi-Agro, A., 2000b. Relation between Wong, C.H., Lerner, R.A., 1996. Inhibition of oleamide decreased anandamide hydrolase concentrations in human hydrolase catalyzed hydrolysis of the endogenous sleep-in- lymphocytes and miscarriage. Lancet 355, 1326–1329. ducing cis-9-octadecenamide. J. Am. Chem. Soc. 118, Matsuda, S., Kanemitsu, N., Nakamura, A., Mimura, Y., Ueda, 5938–5945. N., Kurahashi, Y., Yamamoto, S., 1997. Metabolism of Piomelli, D., Beltrano, M., Glasnapp, S., Lin, S.Y., Goutopou- anandamide, an endogenous cannabinoid receptor ligand, in los, A., Xie, X.-Q., Makriyannis, A., 1999. Structural porcine ocular tissues. Exp. Eye Res. 64, 707–711. determinants for recognition and translocation by the Maurelli, S., Bisogno, T., De Petrocellis, L., Di Luccia, A., anadamide transporter. Proc. Natl. Acad. USA 96, 5802– Marino, G., Di Marzo, V., 1995. Two novel classes of 5807. neuroactive fatty acid amides are substrates for mouse Qin, C., Lin, S., Lang, W., Goutopoulos, A., Pavlopoulos, S., neuroblastoma ‘anandamide amidohydrolase’. FEBS Lett. Mauri, F., Makriyannis, A., 1998. Determination of anan- 377, 82–86. damide amidase activity using ultraviolet-active amine Mayaux, J.F., Cerbelaud, E., Soubrier, F., Yeh, P., Blanche, F., derivatives and reverse-phase high-performance liquid chro- Petre, D., 1991. Purification, cloning, and primary structure matography. Anal. Biochem. 261, 8–15. of a new enantiomer — selective amidase from a Rhodococ- Romero, J., Garcia-Palomero, E., Lin, S.Y., Ramos, J.A., cus strain: structural evidence for a conserved genetic Makriyannis, A., Fernandez-Ruiz, J.J., 1996. Extrapyrami- coupling with nitrile hydratase. J. Bacteriol. 173, 6694–6704. dal effects of methanandamide, an analog of anandamide, Mechoulam, R., Ben-Shabat, S., Hanus, L., Ligumsky, M., the endogenous CB1 receptor ligand. Life Sci. 58, 1249– Kaminski, N.E., Schatz, A.R., Gopher, A., Almog, S., 1257. Martin, B.R., Compton, D.R., et al., 1995. Identification of Schmid, P.C., Reddy, P.V., Natarajan, V., Schmid, H.H.O., 983. an endogenous 2-monoglyceride, present in canine gut, that Metabolism of N-acylethanolamine phospholipids by a binds to cannabinoid receptors. Biochem. Pharmacol. 50, mammalian phosphodiesterase of the phospholipase D type. 83–90. J. Biol. Chem. 258, 9302–9306. Omeir, R.L., Chin, S., Hong, Y., Ahern, D.G., Deutsch, D.G., Schmid, P.C., Schwindenhammer, D., Krebsbach, R.J., Schmid, 1995. Arachidonoyl ethanolamide-[1,2-14C] as a substrate H.H.O., 1998. Alternative pathways of anandamide biosyn- for anandamide amidase. Life Sci. 56, 1999–2005. thesis in rat testes. Chem. Phys. Lipids 92, 27–35. Omeir, R.L., Arreaza, G., Deutsch, D.G., 1999. Identification Schmid, P.C., Zuzarte-Augustin, M.L., Schmid, H.H.O., 1985. of two serine residues involved in catalysis by fatty acid Properties of rat liver N-acylethanolamine amidohydrolase. amide hydrolase. Biochem. Biophys. Res. Commun. 264, J. Biol. Chem. 260, 14145–14149. 316–320. Sugiura, T., Kondo, S., Sukagawa, A., Nakane, S., Shinoda, Paria, B.C., Deutsch, D.D., Dey, S.K., 1996. The uterus is a A., Itoh, K., Yamashita, A., Waku, K., 1995. 2-Arachi- potential site for anandamide synthesis and hydrolysis: donoylglycerol: a possible endogenous cannabinoid receptor differential profiles of anandamide synthase and hydrolase ligand in brain. Biochem. Biophys. Res. Commun. 215, activities in the mouse uterus during the periimplantation 89–97. period. Mol. Reprod. Dev. 45, 183–192. Sugiura, T., Kondo, S., Kodaka, T., Tonegawa, T., Nakane, S., Paria, B.C., Zhao, X., Wang, J., Das, S.K., Dey, S.K., 1999. Yamashita, A., Ishima, Y., Waku, K., 1996a. Enzymatic Fatty-acid amide hydrolase is expressed in the mouse uterus synthesis of oleamide (cis-9,10-octadecenoamide), an en- and embryo during the periimplantation period. Biol. Re- dogenous sleep-inducing lipid, by rat brain microsomes. prod. 60, 1151–1157. Biochem. Mol. Biol. Int. 40, 931–938. Patricelli, M.P., Lashuel, H.A., Giang, D.K., Kelly, J.W., Sugiura, T., Kondo, S., Sukagawa, A., Tonegawa, T., Nakane, Cravatt, B.F., 1998a. Comparative characterization of a wild S., Yamashita, A., Ishima, Y., Waku, K., 1996b. Transacy- type and transmembrane domain-deleted fatty acid amide lase-mediated and phosphodiesterase-mediated synthesis of hydrolase: identification of the transmembrane domain as a N-arachidonoylethanolamine, an endogenous cannabinoid- site for oligomerization. Biochemistry 37, 15177–15187. receptor ligand, in rat brain microsomes. Comparison with Patricelli, M.P., Patterson, J.E., Boger, D.L., Cravatt, B.F., synthesis from free arachidonic acid and ethanolamine. Eur. 1998b. An endogenous sleep-inducing compound is a novel J. Biochem. 240, 53–62. competitive inhibitor of fatty acid amide hydrolase. Bioorg. Thomas, E.A., Cravatt, B.F., Danielson, P.E., Gilula, N.B., Med. Chem. Lett. 8, 613–618. Sutcliffe, J.G., 1997. Fatty acid amide hydrolase, the degra- Patricelli, M.P., Cravatt, B.F., 1999. Fatty acid amide hydrolase dative enzyme for anandamide and oleamide, has selective competitively degrades bioactive amides and esters through distribution in neurons within the rat central nervous system. a nonconventional catalytic mechanism. Biochemistry 38, J. Neurosci. Res. 50, 1047–1052. 14125–14130. Thumser, A.E., Voysey, J., Wilton, D.C., 1997. A fluorescence Patricelli, M.P., Lovato, M.A., Cravatt, B.F., 1999. Chemical displacement assay for the measurement of arachidonoyl and mutagenic investigations of fatty acid amide hydrolase: ethanolamide (anandamide) and oleoyl amide (octade- evidence for a family of serine hydrolases with distinct cenoamide) hydrolysis. Biochem. Pharmacol. 53, 433– catalytic properties. Biochemistry 38, 9804–9812. 435. N. Ueda et al. / Chemistry and Physics of Lipids 108 (2000) 107–121 121

Tsou, K., Nogueron, M.I., Muthian, S., Sanudo-Pena, M.C., ture of human and mouse fatty-acid amide hydrolase genes Hillard, C.J., Deutsch, D.G., Walker, J.M., 1998. Fatty and evaluation of clasper as a candidate neurological mu- acid amide hydrolase is located preferentially in large tation. Genomics 54, 408–414. neurons in the rat central nervous system as revealed by Watanabe, K., Kayano, Y., Matsunaga, T., Yamamoto, I., immunohistochemistry. Neurosci. Lett. 254, 137–140. Yoshimura, H., 1996. Inhibition of anandamide amidase Ueda, N., Kurahashi, Y., Yamamoto, S., Tokunaga, T., 1995. activity in mouse brain microsomes by . Biol. Partial purification and characterization of the porcine Pharm. Bull. 19, 1109–1111. brain enzyme hydrolyzing and synthesizing anandamide. J. Watanabe, K., Ogi, H., Nakamura, S., Kayano, Y., Mat- Biol. Chem. 270, 23823–23827. sunaga, T., Yoshimura, H., Yamamoto, I., 1998. Distribu- Ueda, N., Yamanaka, K., Terasawa, Y., Yamamoto, S., 1999. tion and characterization of anandamide amidohydrolase An acid amidase hydrolyzing anandamide as an endoge- in mouse brain and liver. Life Sci. 62, 1223–1229. nous ligand for cannabinoid receptors. FEBS Lett. 454, Yazulla, S., Studholme, K.E., Mcintosh, H.H., Deutsch, D.G., 267–270. 1999. Immunocytochemical localization of cannabinoid Wan, M., Cravatt, B.F., Ring, H.Z., Zhang, X., Francke, U., CB1 receptor and fatty acid amide hydrolase in rat retina. 1998. Conserved chromosomal location and genomic struc- J. Comp. Neurol. 415, 80–90.

.