Food Intake Regulates Oleoylethanolamide Formation And

Food Intake Regulates Oleoylethanolamide Formation and Degradation in the Proximal Small Intestine* Received for publication, August 15, 2006, and in revised form, October 26, 2006 Published, JBC Papers in Press, November 22, 2006, DOI 10.1074/jbc.M607809200 Jin Fu‡1, Giuseppe Astarita‡1, Silvana Gaetani§, Janet Kim‡, Benjamin F. Cravatt¶, Ken Mackieʈ, and Daniele Piomelli‡2 From the ‡Department of Pharmacology and Center for Drug Discovery, University of California, Irvine, California 92697, §Department of Human Physiology and Pharmacology, University of Rome “La Sapienza”, Piazzale Aldo Moro 5, 00185 Roma, Italy, ¶Department of Cell Biology and Chemistry, The Scripps Research Institute, La Jolla, California 92037, and ʈDepartment of Anesthesiology, University of Washington, Seattle, Washington 98195

Oleoylethanolamide (OEA) is a lipid mediator that inhibits lin have attracted the most attention (1), but recent evidence food intake by activating the nuclear receptor peroxisome indicates that lipid-derived messengers such as oleoylethanol- Downloaded from proliferator-activated receptor-␣. In the rodent small intes- amide (OEA) may also be involved (2). OEA is synthesized in tine OEA levels decrease during food deprivation and the small intestine of various vertebrate species, where its increase upon refeeding, suggesting that endogenous OEA levels decrease during food deprivation and increase upon may participate in the regulation of satiety. Here we show that refeeding (3–5). The possibility that these fluctuations rep- feeding stimulates OEA mobilization in the mucosal layer of resent a satiety signal is suggested by experiments in rodents rat duodenum and jejunum but not in the serosal layer from which show that pharmacological administration of OEA www.jbc.org the same intestinal segments in other sections of the gastro- delays meal initiation and prolongs the interval between suc- intestinal tract (stomach, ileum, colon) or in a broad series of cessive meals, resulting in a persistent inhibition of food internal organs and tissues (e.g. liver, brain, heart, plasma). intake (3, 6–8). These anorexiant effects are strikingly dif- Feeding also increases the levels of other unsaturated fatty ferent from those elicited by traditional satiety factors such at The Scripps Research Institute on March 5, 2007 acid ethanolamides (FAEs) (e.g. linoleoylethanolamide) with- as CCK, which reduce meal size without affecting the inter- out affecting those of saturated FAEs (e.g. palmitoylethanol- val between meals (9). Moreover, the hypophagic actions of amide). Feeding-induced OEA mobilization is accompanied OEA differ from those exerted by glucagon-like peptide-1 by enhanced accumulation of OEA-generating N-acylphos- (10) and corticotropin-releasing factor (11) in that they are phatidylethanolamines (NAPEs) increased activity and not accompanied by behavioral signs of malaise and anxiety expression of the OEA-synthesizing enzyme NAPE-phospho- or by changes in circulating corticosterone levels (3, 12). lipase D, and decreased activity and expression of the OEA- The molecular mechanism through which OEA inhibits degrading enzyme fatty acid amide hydrolase. Immuno- feeding has been partially elucidated. In vitro studies have staining studies revealed that NAPE-phospholipase D and shown that OEA is a high affinity agonist of peroxisome fatty acid amide hydrolase are expressed in intestinal entero- proliferator-activated receptor-␣ (PPAR-␣) (13), a nuclear cytes and lamina propria cells. Collectively, these results receptor present in the small intestine and implicated in the indicate that nutrient availability controls OEA mobilization regulation of energy balance and lipid metabolism (14–17). in the mucosa of the proximal intestine through a concerted OEA binds to the purified ligand binding domain of PPAR-␣ regulation of OEA biosynthesis and degradation. ϳ with a KD of 40 nM and activates this receptor in cell-based assays with a half-maximal effective concentration (EC50)of 120 nM (13). Furthermore, in vivo experiments have revealed The gastrointestinal tract controls key aspects of feeding that genetic deletion of PPAR-␣ abrogates the anorexiant behavior through both neuronal and humoral mechanisms. effects of OEA without altering those evoked by CCK-8 and Among the regulatory signals released by the gut, peptides such 3 fenfluramine (13). Finally, it has been shown that synthetic as cholecystokinin (CCK), glucagon-like peptide-1, and ghre- PPAR-␣ agonists, but not agonists of PPAR-␦ and PPAR-␥, produce a hypophagic response that is behaviorally indistin- * This research was supported by grants from the National Institute on Drug guishable from that elicited by OEA and is absent in mutant Abuse (to K. M. and D. P.). The costs of publication of this article were mice lacking PPAR-␣ (13). A parsimonious interpretation of defrayed in part by the payment of page charges. This article must there- fore be hereby marked “advertisement” in accordance with 18 U.S.C. Sec- these results is that OEA regulates food intake in rodents by tion 1734 solely to indicate this fact. selectively activating intestinal PPAR-␣. Notably, OEA can 1 These authors contributed equally to this study. also activate the G protein-coupled receptor, GPR119 (8), 2 To whom correspondence should be addressed: Dept. of Pharmacology, 360 MSRII, University of California, Irvine, CA 92697-4625. Tel.: 949-824- and the vanilloid receptor channel, TRPV1 (18), albeit at 6180; Fax: 949-824-6305; E-mail: [email protected]. micromolar concentrations. However, the contribution of 3 The abbreviations used are: CCK, cholecystokinin; EC50, half-maximal effective concentration; FAE, fatty acid ethanolamide; FAAH, fatty-acid amide hydrolase; NAPE, N-acyl phosphatidylethanolamine; NAPE-PLD, NAPE-specific phospholipase D; OEA, oleoylethanolamide; PEA, palmi- some proliferator-activated receptor-␣; LC, liquid chromatography; MS, toylethanolamide; PBS, phosphate-buffered saline; PPAR-␣, peroxi- mass spectroscopy.

1518 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 282•NUMBER 2•JANUARY 12, 2007 Oleoylethanolamide Mobilization in Small Intestine

process have not been fully elucidated. In the present study we show that food intake promotes OEA mobilization in the mucosal layer of the rodent proximal intestine through a concerted regulation of NAPE biosynthesis, NAPE-PLD activity, and OEA hydrolysis. EXPERIMENTAL PROCEDURES Animals—Adult male Wistar rats (250–300 g) were purchased from Charles River (Wilmington, MA). FAAHϪ/Ϫ and NAPE-PLDϪ/Ϫ mice were generated and bred as described (23, 24). A 12-hour light/dark cycle was set with lights on at 5:30 a.m. Water and standard chow pellets (Pro- lab RMH 2500; PMI Nutrition International, Brentwood, MO) were available ad libitum, except when animals were food-deprived. Food deprivation was conducted for 24 h in

bottom-wired cages to prevent coprophagia. All procedures Downloaded from met the National Institutes of Health guidelines for the care and use of laboratory animals and were approved by the Institutional Animal Care and Use Committee of the Univer- sity of California, Irvine. Chemicals—Fatty acid chlorides were purchased from Nu-Chek Prep (Elysian, MN), 1,2-Dipalmitoyl-sn-glycero-3- www.jbc.org phosphoethanolamine was from Sigma-Aldrich, and tetrahydrolipstatin was from ChemPacific (Baltimore, MD). N-Cyclohexanecarbonylpentadecylamine was a kind gift at The Scripps Research Institute on March 5, 2007 from Dr. Natsuo Ueda (Kagawa University, Kagawa, Japan). Chemical Syntheses—URB597 was synthesized as described (25). OEA and other FAEs were prepared by the reaction of fatty acid chlorides with a 10-fold molar excess of ethanolamine (Sig- 2 ma-Aldrich) or [ H4]ethanolamine (Cambridge Isotope Labo- ratories, Andover, MA). Reactions were conducted in dichlo- romethane at 0–4 °C for 15 min with stirring. The products were washed with water, dehydrated over sodium sulfate, fil- FIGURE 1. Proposed route of OEA formation and degradation in verte- tered, and dried under N2. They were characterized by liquid brate tissues. PE, phosphatidylethanolamine; PC, phosphatidylcholine; PA, chromatography/mass spectrometry (LC/MS) and 1H nuclear phosphatidic acid; NAT, N-acyltransferase; NAAA, N-acylethanolamine-hydro- magnetic resonance spectroscopy. Purity was Ͼ98% by LC/MS. lyzing acid amidase. NAPEs were prepared by the reaction of fatty acid chlorides with a 2-fold molar excess of 1,2-dipalmitoyl-sn-glycero-3- these receptors to the satiety-inducing effects of OEA is still phosphoethanolamine. Reactions were conducted in dichlo- undefined. romethane at 25 °C for 2 h using triethylamine (Sigma-Aldrich) The enzyme pathway considered responsible for OEA for- as a catalyst. The products were washed with water, fraction- mation and deactivation in mammalian tissues is illustrated in ated by silica gel thin-layer chromatography (chloroform/ Fig. 1. Its first step is the transfer of a fatty acid residue from the methanol/ammonia, 80:20:2, v/v/v), and characterized by sn-1 position of phosphatidylcholine to the free amine group of LC/MS coupled to tandem mass spectrometry (LC/MSn). phosphatidylethanolamine. This reaction is catalyzed by an Purity was Ͼ95%. N-acyltransferase, which remains to be molecularly identified, Lipid Extractions—Rats were sacrificed with halothane, and yields a heterogeneous family of N-acyl phosphatidyletha- and tissues were rapidly collected and snap-frozen in liquid nolamine (NAPE) species. The second step is the cleavage of N2. Plasma was prepared as described (26). Frozen tissues NAPEs to produce other fatty acid ethanolamides (FAEs), were weighed and homogenized in methanol (1 ml/100 mg of 2 including OEA, a step catalyzed by NAPE-specific phospho- tissue) containing [ H4]FAEs and 1,2-dipalmitoyl-sn-glyc- lipase D (NAPE-PLD) (19, 20). OEA deactivation consists in the ero-3-phosphoethanolamine-N-heptadecanoyl as internal intracellular hydrolysis of this lipid amide to oleic acid and eth- standards. Lipids were extracted with chloroform (2 vol- anolamine, which is thought to be mediated by fatty-acid amide umes) and washed with water (1 volumes). Organic phases hydrolase (FAAH), an intracellular membrane-bound serine were collected and dried under N2. FAEs and NAPEs were hydrolase (21), and/or N-acylethanolamine-hydrolyzing acid fractionated by open-bed silica gel column chromatography amidase (NAAA), a lysosomal cysteine hydrolase (22) (Fig. 1). as described (27). Briefly, the lipids were reconstituted in Although previous reports have shown that feeding regulates chloroform and loaded onto small glass columns packed intestinal OEA levels (3–5, 13), the mechanistic details of this with Silica Gel G (60-Å 230–400 Mesh ASTM; Whatman,

JANUARY 12, 2007•VOLUME 282•NUMBER 2 JOURNAL OF BIOLOGICAL CHEMISTRY 1519 Oleoylethanolamide Mobilization in Small Intestine

Clifton, NJ). FAEs and NAPEs were eluted with 9:1 and 5:5 chloroform/methanol (v/v), respectively. Eluates were dried under N2 and reconstituted in 0.1 ml of chloroform/methanol (1:4, v/v) for LC/MS and LC/MSn analyses. LC/MS Analyses—We used an 1100-LC system coupled to a 1946A-MS detector (Agilent Technologies, Inc., Palo Alto, CA) equipped with an electrospray ionization interface. FAEs were separated using a XDB Eclipse C18 column (50 ϫ 4.6-mm inner diameter, 1.8 ␮m, Zorbax) eluted with a gradient of methanol in water (from 85 to 90% methanol in 2.5 FIGURE 2. Feeding-induced OEA mobilization in the rat proximal small min) at a flow rate of 1.5 ml/min. Column temperature was intestine. Effect of free-feeding (FF), 24-h food deprivation (FD), and refeed- kept at 40 °C. MS detection was in the positive ionization ing (10, 30 and 60 min) on OEA levels in duodenum (A), jejunum (B), and ileum Ͻ Ͻ ϭ mode, capillary voltage was set at 3 kV, and fragmentor volt- (C). *, p 0.05; **, p 0.01 versus food deprivation, n 10–12. age was varied from 120 to 140 V. N2 was used as drying gas 2 at a flow rate of 13 liters/min and a temperature of 350 °C. methanol (2:1, v/v) containing [ H4]OEA as the internal ϫ Nebulizer pressure was set at 60 p.s.i. Quantifications were standard. After centrifugation at 1500 g at 4 °C for 5 min, the Downloaded from conducted using an isotope-dilution method, monitoring organic layers were collected and dried under N2. The residues Naϩ adducts of the molecular ions ([MϩNa]ϩ). Limit of were suspended in 50 ␮l of chloroform/methanol (1:3, v/v) and quantification for OEA was 0.25 pmol, as determined by analyzed by LC/MS. For quantification purposes, we monitored injecting into the LC/MS various amounts of OEA in the the [MϩNa]ϩ ions of m/z ϭ 334 for N-heptadecanoylethanol- 2 ϭ 2 presence of 100 pmol of [ H4]OEA standard. amide and m/z 352 for [ H4]OEA. LC/MSn Analyses—We used an 1100-LC system (Agilent FAAH Assays—Tissues homogenates were centrifuged at www.jbc.org Technologies) equipped with a Ion Trap XCT (Agilent Tech- 800 ϫ g for 15 min and then at 27,000 ϫ g for 30 min. The nologies). NAPEs were separated using a Poroshell 300SB 27,000 ϫ g pellet was suspended in phosphate-buffered ϫ ␮ C18 column (75 2.1-mm inner diameter, 5 m, Agilent) saline (PBS, pH 7.4). Reactions were conducted at 37 °C for at The Scripps Research Institute on March 5, 2007 maintained at 50 °C. A linear gradient of methanol (A) in 30 min in 0.5 ml of Tris buffer (50 mM, pH 8.0) containing water (B) containing 5 mM ammonium acetate and 0.25% fatty acid-free bovine serum albumin (0.05%), protein (50 acetic acid (from 85 to 100% of B in 4 min) was applied at a ␮g) and [ethanolamine-3H]anandamide (10,000 dpm, spe- flow rate of 1 ml/min. Detection was set in the negative cific activity 20 Ci/mmol; American Radiolabeled Chemi- mode, capillary voltage was 4.5 kV, skim1 was Ϫ40V, and cals, St. Louis, MO) or [ethanolamine-3H]OEA (10,000 dpm, Ϫ capillary exit was 151V. N2 was used as drying gas at a flow specific activity 15 Ci/mmol; American Radiolabeled Chem- rate of 12 liters/min, temperature of 350 °C, and nebulizer icals). After stopping the reaction with 1 ml of chloroform/ pressure of 80 p.s.i. Helium was used as the collision gas. methanol (1:1, v/v), we measured radioactivity in the aque- Tissue-derived NAPEs were identified by comparison of ous layers by liquid scintillation counting. their LC retention times and MSn fragmentation patterns Quantitative PCR—Total RNA was extracted from tissue with those of authentic standards, prepared as described with TRIzolTM (Invitrogen) and quantified with RibogreenTM above. We acquired full-scan MSn spectra of selected NAPE (Molecular Probes, Eugene, OR). cDNA was synthesized precursor ions by multiple reaction monitoring. Extracted from 2 ␮g of total RNA by using Superscript II RNase H ion chromatograms were used to quantify each NAPE precur- reverse transcriptase (Invitrogen) following the manufactur- sor ion by monitoring characteristic product ions as follows: er’s instructions. Real-time quantitative PCR was performed 1-stearoyl-2-arachidonoyl-sn-glycero-phosphoethanolamine- in an Mx 3000P system (Stratagene, La Jolla, CA). Primers N-oleoyl (m/z 1030.8 Ͼ 744.8), 1-stearoyl-2-arachidonoyl-sn- and fluorogenic probes were synthesized at TIB (Adelphia). glycero-phosphoethanolamine-N-palmitoyl (m/z 1004.8 Ͼ The primer/probe sequences were: NAPE-PLD, TGGCTG- 718.8) and 1,2-dipalmitoyl-sn-glycero-3-phosphoethanol- GGACACGCG (forward primer), GGGATCCGTGAGGA- amine-N-heptadecanoyl (m/z 942.8 Ͼ 704.8), which was used GGATG (reverse primer), CGCTGATGGTGGAAATGGA- as internal standard. Detection and analysis were controlled by CGAGC (probe); FAAH, GCCTCAAGGAATGCTTCAGC Agilent/Bruker Daltonics software version 5.2, and three-di- (forward primer), TGCCCTCATTCAGGCTCAAG (reverse mensional maps were generated using MS Processor from primer), ACAAGGGCCACGACTCCACACTGG (probe); Advanced Chemistry Development, Inc. (Toronto, Canada). glyceraldehyde 3-phosphate dehydrogenase, AAGTATGATG- NAPE-PLD Assays—Rats were anesthetized with halothane ACATCAAGAAGGTGGT (forward primer), AGCCCAGGA- and sacrificed by decapitation. Tissues were removed and TGCCCTTTAG (reverse primer), AAGCAGGCGGCCGA- homogenized in ice-cold Tris-HCl (20 mM, 10 volumes, pH 7.4) GGGC (probe). mRNA expression levels were normalized by containing 0.32 M sucrose. NAPE-PLD activity was measured at using glyceraldehyde 3-phosphate dehydrogenase as an inter- 37 °C for 30 min in Tris-HCl buffer (50 mM, pH 7.4) containing nal standard and were calculated as described (28). 0.1% Triton X-100, 1 mM phenylmethylsulfonyl fluoride, tissue Antibody Generation—We constructed a glutathione S- homogenate protein (100 ␮g), and 1,2-dipalmitoyl-sn-glycero- transferase-linked peptide comprising residues 38–53 of rat 3-phosphoethanolamine-N-heptadecenoyl (100 ␮M) as sub- NAPE-PLD. The DNA fragment was amplified using primers strate. The reactions were stopped by adding chloroform- incorporated with BamHI and XhoI restriction sites on the 5Ј

1520 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 282•NUMBER 2•JANUARY 12, 2007 Oleoylethanolamide Mobilization in Small Intestine

TABLE 1 Effects of 24-h food deprivation on OEA levels in various organs and tissues of rats OEA levels are expressed as pmol/g of tissue weight or pmol/ml of plasma. Results are expressed as mean Ϯ S.E., n ϭ 10–12. FF, free feeding; FD, 24-h food deprivation; WAT, water. Stomach Colon Kidney Heart Lung Muscle Liver Pancreas Spleen WAT Plasma FF 108.3 Ϯ 7.4 72.8 Ϯ 5.1 58.4 Ϯ 5.6 104.3 Ϯ 8.6 115.8 Ϯ 9.7 92.5 Ϯ 3.6 80.7 Ϯ 3.2 44.3 Ϯ 3.7 58.0 Ϯ 2.9 174.3 Ϯ 10.3 5.9 Ϯ 0.4 FD 102.1 Ϯ 9.1 68.4 Ϯ 2.3 62.1 Ϯ 2.9 97.9 Ϯ 10.6 111.6 Ϯ 8.2 89.1 Ϯ 7.2 136.3 Ϯ 6.5a 69.4 Ϯ 12.8a 66.6 Ϯ 6.6a 215.8 Ϯ 7.6a 7.9 Ϯ 0.5a a p Ͻ 0.01 vs. food deprivation (Student’s t test).

TABLE 2 Effects of 24-h food deprivation on OEA levels in various structures of the rat brain Results are expressed as the mean Ϯ S.E., n ϭ 10–12. FF, free-feeding; FD, 24-h food deprivation. OEA levels are expressed as pmol/g of tissue weight. Cortex Hippocampus Thalamus Striatum Hypothalamus Cerebellum Brainstem FF 162.5 Ϯ 13.3 137.3 Ϯ 12.8 264.1 Ϯ 12.6 216.8 Ϯ 11.1 139.5 Ϯ 8.8 195.1 Ϯ 11.8 439.3 Ϯ 38.2 FD 167.2 Ϯ 11.2 131.9 Ϯ 11.1 255.7 Ϯ 13.4 226.6 Ϯ 7.9 154.8 Ϯ 10.9 190.4 Ϯ 16.2 426.9 Ϯ 26.2

TABLE 3 Chemical structures and tissue levels (pmol/g tissue) of various FAEs in rat small intestine Downloaded from Values represent the means Ϯ S.E., n ϭ 5–6. FF, free-feeding; FD, 24-h food deprivation. n.d., non-detectable; tr, trace levels. www.jbc.org at The Scripps Research Institute on March 5, 2007

* p Ͻ 0.05 vs. food deprivation (Student’s t test). ** p Ͻ 0.01 vs. food deprivation (Student’s t test). *** p Ͻ 0.001 vs. food deprivation (Student’s t test). and 3Ј end of amplicon, respectively. After digestion with WI) and purified using glutathione-SepharoseTM 4B (Amer- BamHI and XhoI, amplicon was subcloned into vector pGEX- sham Biosciences) following the manufacturer’s instruction. 4T-1 (Amersham Biosciences). The fusion protein was The polyclonal NAPE-PLD antibody was raised in rabbits and expressed in Escherichia coli BL-21 cells (Novagen, Madison, characterized using standard procedures (29).

JANUARY 12, 2007•VOLUME 282•NUMBER 2 JOURNAL OF BIOLOGICAL CHEMISTRY 1521 Oleoylethanolamide Mobilization in Small Intestine

Immunohistochemistry—Mice were deeply anesthetized with halothane, and their tissues were fixed by cardiac per- fusion with saline followed by 4% paraformaldehyde in PBS (0.1 M, pH 7.2). The duodenum was collected and rinsed with PBS. 15-␮m-Thick transverse sections were cut with a Cry- ostat. The sections were rehydrated and treated with 0.3%

H2O2 in methanol for 30 min to inactivate endogenous per- oxidases, rinsed in PBS for 10 min, exposed to blocking serum (3% normal goat serum) for2hatroom temperature, rinsed again in PBS, and incubated overnight at 4 °C with the NAPE-PLD antibody (1:1000 dilution) or a commercial anti- FAAH antibody (1:500 dilution; Chemicon, Temecula, CA). After the incubation, the sections were rinsed in 0.1 M PBS for 10 min and exposed to anti-rabbit IgG HRP (1:1000 dilution, Vector Laboratories, Burlingame, CA) for 1 h. After

an additional 10-min rinse, the sections were treated with Downloaded from VectaStain௡ Elite ABC reagent (Vector) for 30 min and developed with the diaminobenzidine detection kit (Vector). The sections were counterstained by hematoxylin and mounted with Permount (Sigma). Protein Analysis—Rat jejunal mucosa was homogenized in Tris-HCl buffer (50 mM, pH 8.0) containing protease inhib- www.jbc.org itor mixture V (Calbiochem). 20 ␮g of homogenate were FIGURE 3. Effects of nutrient availability on PEA and anandamide levels in electrophoresed on 4–15% SDS-polyacrylamide gel (Invitro- the rat proximal small intestine. Effects of free-feeding (FF), 24-h food dep- gen) and transferred onto ImmobilonTM membranes (Milli- rivation (FD), and refeeding (10, 30, and 60 min) on the levels of PEA (A and B) and anandamide (AEA, C and D) in duodenum (left) and jejunum (right). *, p Ͻ at The Scripps Research Institute on March 5, 2007 pore, Billerica, MA). Western blot analyses were conducted 0.05, **, p Ͻ 0.01 versus food deprivation, n ϭ 10–12. using antibodies against NAPE-PLD (1:1,000), FAAH (1:500; Chemicon), and ␤-actin (1:10,000; Calbiochem). Bands were visualized with an electrochemiluminescence (ECL)-Plus kit Effects of Food Deprivation on Intestinal FAE Mobiliza- (Amersham Biosciences). tion—Because OEA is thought to share a common biosyn- Statistical Analyses—Results are expressed as the means Ϯ thetic pathway with other bioactive FAEs, we identified and S.E. Statistical significance was evaluated using Student’s t test quantified other members of this lipid family in the duode- or, when appropriate, one-way analysis of variance followed by num and jejunum of fasting and free-feeding rats. The results the Dunnett’s post hoc test. Analyses were conducted using of these analyses are reported in Table 3. Identifications were GraphPad Prism (GraphPad Software, San Diego, CA), and dif- based on the detection of sodium adduct ions ([MϩNa]ϩ)of ferences were considered significant if p Ͻ 0.05. appropriate m/z values and retention times and were confirmed by comparison with synthetic FAE standard. Food deprivation RESULTS decreased intestinal levels of OEA (18:1⌬9), linoleoylethanol- Effects of Nutrient Availability on OEA Mobilization in Rat amide (18:2⌬9,12), linolenoylethanolamide (18:3⌬9,12,15), and Tissue—We used an isotope-dilution LC/MS assay to meas- docosahexaenoylethanolamide (22:6⌬4,7,10,13,16,19) (Table 3). ure OEA levels in various organs and tissues of free-feeding By contrast, food deprivation did not change the duodenal or rats, 24-h food-deprived rats, or 24-h food-deprived rats that jejunal content of palmitoylethanolamide (PEA, 16:0) (Fig. 3, A had been reexposed to food for 10, 30, or 60 min. Confirming and B) and stearoylethanolamide (18:0) (Table 3). Moreover, as previous results (3, 4), we found that fasting decreases OEA previously reported (4, 30), food deprivation increased levels of content in the two most proximal segments of the small the endocannabinoid anandamide (eicosatetraenoylethanol- intestine, duodenum and jejunum, whereas refeeding rapidly amide, 20:4⌬5,8,11,14) (Fig. 3, C and D, and Table 3). The findings reverses this effect (Fig. 2, A and B). By contrast, no such a indicate that food intake (i) enhances the intestinal mobiliza- change was observed in other intestinal segments (ileum, tion of a subgroup of FAEs that include OEA, but not PEA and stomach, and colon) (Fig. 2C and Table 1), in a broad panel of stearoylethanolamide, and (ii) differentially regulates OEA and internal organs and tissues (kidney, heart, lung, skeletal mus- anandamide mobilization in the small intestine; the mechanism cle, liver, pancreas, spleen, epididymal white adipose pads, responsible for fasting-induced anandamide mobilization was plasma) (Table 1), or in brain structures involved in the con- not further investigated. trol of feeding (brainstem, hypothalamus, thalamus, stria- Localization of Feeding-dependent OEA Mobilization in the tum, and cortex) (Table 2). Indeed, several visceral organs, Small Intestine—To test whether nutrient availability differen- including liver and pancreas, responded to food deprivation tially affects OEA mobilization in the luminal and serosal por- with an increase rather than a decrease in OEA levels (Table tions of the small intestinal wall, we separated these structures 1), suggesting that nutrient availability regulates OEA mobi- with a blunt spatula (Fig. 4A) and measured OEA content in lization in a tissue-specific manner. each of them separately. Fasting and refeeding strongly altered

1522 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 282•NUMBER 2•JANUARY 12, 2007 Oleoylethanolamide Mobilization in Small Intestine Downloaded from www.jbc.org at The Scripps Research Institute on March 5, 2007

FIGURE 4. Feeding-induced OEA mobilization is restricted to the lumen of the rat proximal small intestine. A, hematoxylin staining of a section of rat duodenum before (left) or after (right) removal of the lumen by scraping with a blunt spatula. Scale bar, 100 ␮m. Effects of free-feeding (FF), 24-h food deprivation (FD) and refeeding (10, 30 and 60 min) on OEA contents in luminal (B and D) and serosal fraction (C and E) of duodenum (B and C) and jejunum (D and E). *, p Ͻ 0.05, **, p Ͻ 0.01, ***, p Ͻ 0.001 versus food deprivation, n ϭ 10–12.

OEA mobilization in luminal fractions of duodenum or jejunum (Fig. 4, B and D) but had no effect in corresponding serosal fractions (Fig. 4, C and E). Thus, feeding-induced OEA mobilization appears to be restricted to anatomical structures located in the innermost aspect, most likely the mucosal layer, of the duodenal and jejunal wall. Identification of OEA-generating NAPEs in the Small Intestine—The biosynthesis of NAPEs, catalyzed by N-acyltransferase, is thought to be the first committed step in OEA mobilization (Fig. 1). We used LC/MSn to identify NAPE species that may serve as OEA precursors in the small intestine. An example of our approach is illustrated in Fig. 5, 1 3 which shows the elution profile and mass spectra (MS -MS ) FIGURE 5. Identification of NAPE species in luminal layer of rat jejunum generated by the endogenous NAPE, 1-stearoyl-2-arachido- by LC/MSn. A, representative LC/MS3 extracted ion chromatogram for ϭ endogenous 1-stearoyl,2-arachidonoyl-sn-glycero-3-phosphoethanol- noyl-sn-glycero-phosphoethanolamine-N-oleoyl (m/z amine-N-oleyl (m/z 1030.8 Ͼ 744.8 Ͼ 460.3). Chromatographic conditions 1030.8) (Fig. 5, A and B), along with the fragmentation pat- are described under “Experimental Procedures.” B, top, negative ion elec- tern deduced from these spectra (Fig. 5C). Notable in this trospray mass spectra of the precursor-ion (m/z 1030.8) corresponding to the deprotonated NAPE; center, collision-induced dissociation product- pattern is the sequential loss of the sn-2 and sn-1 fatty-acid ions of m/z 1030.8; bottom, product-ions for CID of the fragment with m/z substituents, with eventual generation of a fragment con- 744.8. C, proposed fragmentation pattern for 1-stearoyl,2-arachidonoyl- taining the N-acyl moiety (Fig. 5C). Fig. 6A depicts a bidi- sn-glycero-3-phosphoethanolamine-N-oleoyl. mensional LC/MS tracing (LC elution time versus m/z)of NAPE species present in the luminal fraction from the jeju- the unnatural synthetic NAPE, 1,2-dipalmitoyl-sn-glycero- num of refed rats, and Fig. 6B reports the LC/MSn identifi- 3-phosphoethanolamine-N-heptadecanoyl, as an internal cation and relative quantification of OEA-generating NAPEs standard. These analyses revealed an unexpected paucity of within this fraction. Quantifications were conducted using NAPEs containing ether- or vinyl ether-linked fatty acids at

JANUARY 12, 2007•VOLUME 282•NUMBER 2 JOURNAL OF BIOLOGICAL CHEMISTRY 1523 Oleoylethanolamide Mobilization in Small Intestine

PLD. The antibody specifically stained cellular elements in the duodenal epithelial layer and lamina propria from wild-type mice (Fig. 9C). Only background staining was observed in corresponding structures of NAPE-PLDϪ/Ϫ mice (Fig. 9D). The results indicate that (i) both NAPE-PLD activity and expression levels are regulated by nutrient availability, albeit with dif- ference time-courses, and (ii) NAPE-PLD is expressed in intestinal enterocytes and lamina propria FIGURE 6. Molecular composition and relative quantification of OEA-generating NAPE species in the cells. luminal layer of rat jejunum. A, representative bidimensional LC/MS contour map. The first dimension is elution time, and the second is m/z. The colored density distribution represents relative intensity of the signal. Effects of Nutrient Availability on Downloaded from Numbers denote individual pseudo-molecular ions [M-H]Ϫ for OEA-generating NAPEs (1–8). The precursor ion Intestinal FAAH Activity—FAAH n Ͼ Ͼ labeled as 9 is the most abundant PEA-generating, which was identified by LC/MS (1004.8 718.8 434.3). catalyzes the hydrolytic cleavage B, electrospray ionization-MSn identification, sn-1 and sn-2 fatty acid composition, and relative amounts (%) of the main OEA-generating NAPE species in jejunal luminal fraction. of OEA and other FAEs (Fig. 1). We determined the effects of nutrient availability on FAAH the sn-1 position (Ͻ2%). By contrast, sn-1 ether-containing activity in membrane preparations of mucosal tissue from www.jbc.org NAPEs, which are readily identifiable by their distinctive rat jejunum using either [3H]OEA or [3H]anandamide as a fragmentation pattern (Fig. 7, A and B), were abundant in the substrate. Irrespective of the substrate used, fasting pro- jejunal serosal fraction (Fig. 7, C–E). The functional signifi-

duced a remarkable increase in FAAH activity, which was at The Scripps Research Institute on March 5, 2007 cance of this heterogeneity is unclear at present but presum- completely reversed within 60 min of reexposure to food ably reflects the different functions served by mucosal and (Fig. 10A and data not shown). Quantitative RT-PCR and serosal cell layers in intestinal physiology. Western blot analyses revealed parallel, but slower changes Effects of Nutrient Availability on Intestinal NAPE Bio- in jejunal FAAH mRNA (Fig. 10B) and protein levels (Fig. synthesis—Next, we determined whether nutrient availabil- 10B, inset). Beside FAAH, intestinal tissue contains other ity regulates intestinal NAPE levels in the luminal fraction of lipid hydrolases, including pancreatic lipase (31) and n rat jejunum. To this end, we quantified by LC/MS the most N-acylethanolamine-hydrolyzing acid amidase (22), which abundant intestinal NAPE precursors for OEA and PEA; might contribute to the changes in OEA hydrolysis elicited respectively, 1-stearoyl-2-arachidonoyl-sn-glycero-3-phos- by fasting and refeeding. Thus, to determine the role of 5 phoethanolamine-N-oleoyl (Fig. 6, ) and 1-stearoyl-2-arachi- FAAH in this response, we incubated jejunal membrane donoyl-sn-glycero-3-phosphoethanolamine-N-palmitoyl (Fig. preparations from the mucosal layer of food-deprived rats 6, 9). The results show that fasting decreased jejunal levels of with the selective FAAH inhibitor URB597 (1 ␮M) (32), the the OEA precursor 5, whereas refeeding produced an oppo- pancreatic lipase inhibitor tetrahydrolipstatin (1 ␮M) (33), or site effect (Fig. 8A). By contrast, levels of the PEA precursor the N-acylethanolamine-hydrolyzing acid amidase inhibitor 9 remained unchanged (Fig. 8B). This response mirrors the N-cyclohexanecarbonylpentadecylamine (100 ␮M) (34). one previously observed with OEA and PEA (Fig. 2), further URB597 inhibited [3H]OEA hydrolysis by ϳ50%, whereas underscoring the selective nature of feeding-induced OEA mobilization. tetrahydrolipstatin and N-cyclohexanecarbonylpentade- Effects of Nutrient Availability on Intestinal NAPE-PLD cylamine had little or no effect (Fig. 10C). To confirm the Activity—NAPE-PLD catalyzes the conversion of NAPEs to presence of FAAH in the small intestine and determine its OEA and other FAEs (Fig. 1). Using an LC/MS-based assay, we cellular localization, we immunostained sections of duode- Ϫ/Ϫ measured NAPE-PLD activity in homogenates of jejunal tissue nal tissue from wild-type and FAAH mice with a com- obtained from free-feeding, food-deprived, or food-deprived/ mercial polyclonal antibody. The antibody specifically refed rats. The results show that fasting decreases tissue NAPE- stained cellular elements in the epithelial layer and, to a PLD activity, whereas refeeding rapidly reverses this effect (Fig. minor extent, lamina propria of wild-type mouse duodenum 9A). Quantitative RT-PCR and Western blot analyses revealed (Fig. 10D). By contrast, only background staining was Ϫ/Ϫ a parallel, albeit slower change in jejunal NAPE-PLD mRNA observed in corresponding structures from FAAH mice (Fig. 9B) and protein levels (Fig. 9B, inset). To confirm the pres- (Fig. 10E). The findings outlined above suggest that (i) food ence of NAPE-PLD in the small intestine and determine its deprivation markedly increases FAAH activity and expres- cellular localization, we immunostained sections of duodenal sion, (ii) refeeding decreases FAAH activity, presumably tissue from wild-type and NAPE-PLD-deficient (NAPE- through a post-translational mechanism, (iii) refeeding also PLDϪ/Ϫ) mice (24) using a selective polyclonal antibody raised reduces FAAH expression; (iv) like NAPE-PLD, FAAH is against the N-terminal sequence of rat recombinant NAPE- expressed in intestinal enterocytes and lamina propria cells,

1524 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 282•NUMBER 2•JANUARY 12, 2007 Oleoylethanolamide Mobilization in Small Intestine Downloaded from www.jbc.org at The Scripps Research Institute on March 5, 2007

FIGURE 7. Molecular composition and relative quantification of OEA-generating NAPE species in the serosal layer of rat jejunum. MS3 fragmentation of representative sn-1 alkenyl NAPE (A) and sn-1 alkyl NAPE (B) identified in jejunal serosal layer. C, mirrored display of averaged MS1 spectra at retention time from 2.7 to 3.7 min of serosal (top) and luminal (bottom) NAPEs. Arrows denote alkyl or alkenyl NAPE species. Numbers refer to the identification of those species, reported in panel E. D, representative bidimensional LC/MS contour map of NAPEs from jejunal serosal fraction. The first dimension is elution time, and the second is m/z. The colored density distribution represents relative intensity of the signal. Each number denotes an individual pseudo-molecular ion for 1-alkenyl (alkyl), sn-2-acyl NAPE species (10–17). E, electrospray ionization-MSn identification, sn-1 and sn-2 fatty acid composition, and relative amounts (%) of the main OEA-generating ether-containing NAPEs in jejunal serosal fraction. and (v) other hydrolase activities beside FAAH may contrib- and absorption of nutrients (10). For example, fat and pro- ute to feeding-regulated OEA hydrolysis. tein in the chyme interact with endocrine I cells in the duodenal and jejunal mucosa, causing the release of CCK-33 and DISCUSSION CCK-8 into the bloodstream. These peptides engage type-1 Intrinsic signals generated by the gut play critical roles in CCK receptors located on vagal sensory fibers and visceral the regulation of feeding behavior and digestion. As ingested organs, initiating a coordinated physiological response that food transits through the small intestine, it stimulates nutri- shortens the duration of an ongoing meal, modulates gastro- ent-sensing endocrine cells to secrete biologically active intestinal motility, and induces gallbladder contraction and peptides, which modulate feeding and aid in the digestion pancreatic secretion (9). In addition to its well recognized

JANUARY 12, 2007•VOLUME 282•NUMBER 2 JOURNAL OF BIOLOGICAL CHEMISTRY 1525 Oleoylethanolamide Mobilization in Small Intestine ability to stimulate the release of regulatory peptides, food The present results demonstrate that feeding-dependent intake may also control the mobilization of lipid mediators OEA mobilization is confined to the mucosal layer of the such as OEA, the intestinal levels of which decrease during duodenal and jejunal wall, where the OEA-metabolizing food deprivation and increase upon refeeding (3–5). How- enzymes NAPE-PLD and FAAH are localized. Our study fur- ever, despite the potential role of OEA as a lipid-derived ther shows that food intake, in addition to mobilizing OEA, satiety factor (3, 13), the molecular mechanisms underlying also enhances the formation of OEA-generating NAPEs, OEA release in the gut remain largely undefined. increases NAPE-PLD activity, and decreases FAAH activity. These changes in enzyme activity are likely to be caused by a combination of post-translational modifications, which appear to occur within minutes of food presentation, and more prolonged alterations in gene expression. Thus, nutrient availability may control OEA signaling in duodenal and jejunal mucosa by reciprocally regulating OEA biosynthesis through activation of the N-acyltransferase/NAPE-PLD pathway and by OEA hydrolysis through down-regulation of

FAAH and possibly other as-yet uncharacterized amidases. Downloaded from The physiological mechanism underlying these regulatory events remains to be determined, although its rapid time- FIGURE 8. Feeding stimulates NAPE accumulation in the luminal layer of course suggests the participation of neural or neurohumoral rat jejunum. Effects of free feeding (FF), 24-h food deprivation (FD), and factors. refeeding (10, 30, and 60 min) on the levels of 1-stearoyl,2-arachidonoyl-sn- glycero-3-phosphoethanolamine-N-oleyl (N-oleoyl PE, 5)(A) and 1-stearoyl,2- In a recent report, Petersen et al. (4) confirmed our earlier arachidonoyl-sn-glycero-3-phosphoethanolamine-N-pamitoyl (N-palmitoyl finding that food deprivation and refeeding regulate OEA www.jbc.org PE, 9)(B), as assessed by LC/MSn.*,p Ͻ 0.05, **, p Ͻ 0.01 versus food deprivation, n ϭ 8–10. and NAPE levels in the small intestine (3, 13) but failed to observe any detectable change in NAPE-PLD and FAAH activities at The Scripps Research Institute on March 5, 2007 under those conditions. This dis- crepancy is likely due to method- ological differences between studies. In particular, Petersen et al. (4) conducted their analyses on membranes prepared from the entire small intestine, whereas we used either total homogenates or luminal fractions from duodenum and jejunum, which are enriched in OEA-metabolizing enzymes. Nev- ertheless, future experiments using genetically modified mice and selective enzyme inhibitors will be neces- sary to establish the role of NAPE- PLD and FAAH in feeding- dependent OEA mobilization. Beyond the plausible implica- tion of the NAPE-PLD/FAAH pathway, suggested by our bio- chemical and immunostaining experiments, several aspects of feeding-dependent OEA mobilization remain to be elucidated. Three points appear to be particu- larly relevant. The first pertains to the selectivity with which food intake mobilizes OEA and various FIGURE 9. Feeding increases NAPE-PLD activity and expressions in rat jejunum. Effects of free feeding polyunsaturated FAEs compared (FF), 24-h food deprivation (FD), and refeeding (10, 30, 60, and 120 min) on NAPE-PLD activity (A), NAPE- with their saturated counterparts PLD mRNA expression (B), and NAPE-PLD protein levels (B inset). *, p Ͻ 0.05, **, p Ͻ 0.01 versus food PEA (16:0) and stearoylethanol- deprivation, n ϭ 6. C, representative immunostaining showing the localization of NAPE-PLD in the duodenal epithelium and lamina propria of wild-type C57/BL6 mice. D, absence of immunoreactive NAPE-PLD amide (18:0). The finding of a par- in corresponding structures of NAPE-PLDϪ/Ϫ mice; scale bar,50␮m. allel chemoselectivity in the bio-

1526 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 282•NUMBER 2•JANUARY 12, 2007 Oleoylethanolamide Mobilization in Small Intestine

consistent with the possibility that OEA may act as a paracrine signal rather than as a blood-borne hormone. Previous work has shown that pharmacological administration of OEA delays meal initiation and prolongs the interval between successive meals rather than reducing meal size (6, 13). We report here that feeding-dependent OEA mobilization is confined to the innermost layers of the gut wall and is not accompanied by significant changes in circulating OEA

levels. Thus, unlike classical sati- Downloaded from ety factors such as CCK, which are released into the bloodstream to terminate an ongoing meal, OEA may be produced and act within the gut to prolong the interval between successive meals. Such an www.jbc.org action would be consistent with FIGURE 10. Feeding decreases FAAH activity and mRNA expression in rat jejunum. Effects of free feeding the pattern of OEA mobilization, (FF), 24-h food deprivation (FD), and refeeding (10, 30, 60, and 120 min) on FAAH activity (A), FAAH mRNA which occurs after rather than expression (B), and FAAH protein levels (B inset).C, effects of vehicle (food deprivation, 0.1% Me SO), URB597 at The Scripps Research Institute on March 5, 2007 2 during a meal, and is reminiscent (URB, 0.1 ␮M), tetrahydrolipstatin (THL,1␮M), and carbonyl cyanide p-chlorophenylhydrazone (CCP,5␮M)on FAAH activity in intestinal mucosal layer of food-deprived rats. *, p Ͻ 0.05; **, p Ͻ 0.01 versus food deprivation, of the physiological role proposed n ϭ 6. D, representative immunostaining showing the localization of FAAH in the duodenal epithelium and for PYY , a member of the pan- lamina propria of wild-type C57/BL6 mice. E, absence of immunoreactive FAAH in corresponding structures of 3–36 FAAHϪ/Ϫ mice; scale bar,50␮m. creatic family of polypeptides that is released postprandially and sup- synthesis of NAPEs suggests the existence of a regulatory presses food intake for an extended time period (38). step located upstream of NAPE-PLD; however, the identity of this hypothetical step remains unknown. The second Acknowledgments—We thank Kevin Eng for help with initial experi- point, related to the first, concerns the possible functions ments, Dr. Faizy Ahmed for discussion on chromatographic separa- served by linoleoylethanolamide (18:2⌬9–12) and other poly- tions, Dr. Gary Schwartz for critical discussion, and Dr. Robert unsaturated FAEs that are released together with OEA. Evi- Edwards for expert advice on immunostaining. Lipid analyses were dence indicates that these compounds may not affect food conducted at the Agilent/University of California, Irvine (UCI) Ana- intake (4), raising the intriguing possibility that they might lytical Discovery Facility, Center for Drug Discovery at UCI. contribute to distinct aspects of intestinal OEA signaling, such as the regulation of fatty acid intake (35). The final point refers to the anatomical specificity of OEA signaling. REFERENCES The ability of food deprivation to inhibit OEA mobilization 1. Broberger, C. (2005) J. Intern. Med. 258, 301–327 in the proximal intestine and enhance it in liver and spleen 2. Lo Verme, J., Gaetani, S., Fu, J., Oveisi, F., Burton, K., and Piomelli, D. (2005) Cell. Mol. Life Sci. 62, 708–716 implies that OEA may be released in a tissue-restricted and 3. Rodrı´guez de Fonseca, F., Navarro, M., Go´mez, R., Escuredo, L., Nava, F., stimulus-specific manner. Two additional results support Fu, J., Murillo-Rodrı´guez, E., Giuffrida, A., LoVerme, J., Gaetani, S., Kathu- this view. First, acute cold exposure stimulates OEA mobili- ria, S., Gall, C., and Piomelli, D. (2001) Nature 414, 209–212 zation in rodent white adipose tissue but has no such effect 4. Petersen, G., Sorensen, C., Schmid, P. C., Artmann, A., Tang-Christensen, in small intestine, liver, or skeletal muscle (36). This M., Hansen, S. H., Larsen, P. J., Schmid, H. H., and Hansen, H. S. (2006) response is prevented in vivo by the ␤-antagonist proprano- Biochim. Biophys. Acta 1761, 143–150 ␤ 5. Astarita, G., Rourke, B. C., Andersen, J. B., Fu, J., Kim, J. H., Bennett, A. F., lol and is reproduced in vitro by noradrenaline and the -ag- Hicks, J. W., and Piomelli, D. (2006) Am. J. Physiol. Regul. Integr. Comp. onist isoproterenol, suggesting that it may be mediated by Physiol. 290, 1407–1412 sympathetic activation of ␤-adrenoreceptors on adipose 6. Gaetani, S., Oveisi, F., and Piomelli, D. (2003) Neuropsychopharmacology. cells (36). Second, food deprivation and refeeding only mod- 28, 1311–1316 estly affect plasma OEA concentrations (Table 1), which at 7. Nielsen, M. J., Petersen, G., Astrup, A., and Hansen, H. S. (2004) J. Lipid their highest level are 15–400 times lower than those Res. 45, 1027–1029 ␣ 8. Overton, H. A., Babbs, A. J., Doel, S. M., Fyfe, M. C., Gardner, L. S., Griffin, expected to produce significant activation of PPAR- G., Jackson, H. C., Procter, M. J., Rasamison, C. M., Tang-Christensen, M., Ϸ Ϸ ␮ (EC50 120 nM) (13), GPR119 (EC50 3.2 M) (8), or Widdowson, P. S., Williams, G. M., and Reynet, C. (2006) Cell Metab. 3, Ϸ ␮ TRPV1 (EC50 2 M) (37). Together, these observations are 167–175

JANUARY 12, 2007•VOLUME 282•NUMBER 2 JOURNAL OF BIOLOGICAL CHEMISTRY 1527 Oleoylethanolamide Mobilization in Small Intestine

9. Little, T. J., Horowitz, M., and Feinle-Bisset, C. (2005) Obes. Rev. 6, 24. Leung, D., Saghatelian, A., Simon, G. M., and Cravatt, B. F. (2006) Bio- 297–306 chemistry 45, 4720–4726 10. Strader, A. D., and Woods, S. C. (2005) Gastroenterology 128, 175–191 25. Mor, M., Rivara, S., Lodola, A., Plazzi, P. V., Tarzia, G., Duranti, A., Ton- 11. Bale, T. L., and Vale, W. W. (2004) Annu. Rev. Pharmacol. Toxicol. 44, tini, A., Piersanti, G., Kathuria, S., and Piomelli, D. (2004) J. Med. Chem. 525–557 47, 4998–5008 12. Proulx, K., Cota, D., Castaneda, T. R., Tschop, M. H., D’Alessio, D. A., Tso, 26. Giuffrida, A., Rodriguez de Fonseca, F., and Piomelli, D. (2000) Anal. Bio- P., Woods, S. C., and Seeley, R. J. (2005) Am. J. Physiol. Regul. Integr. Comp. chem. 280, 87–93 Physiol. 289, 729–737 27. Cadas, H., di Tomaso, E., and Piomelli, D. (1997) J. Neurosci. 17, 13. Fu, J., Gaetani, S., Oveisi, F., Lo Verme, J., Serrano, A., Rodriguez de Fon- 1226–1242 seca, F., Rosengarth, A., Luecke, H., Di Giacomo, B., Tarzia, G., and Pio- 28. Schmittgen, T. D., Zakrajsek, B. A., Mills, A. G., Gorn, V., Singer, M. J., and melli, D. (2003) Nature 425, 90–93 Reed, M. W. (2000) Anal. Biochem. 285, 194–204 14. Desvergne, B., and Wahli, W. (1999) Endocr. Rev. 20, 649–688 29. Twitchell, W., Brown, S., and Mackie, K. (1997) J. Neurophysiol. 78, 43–50 15. Kersten, S., and Wahli, W. (2000) Experientia (Basel) 89, 141–151 30. Gomez, R., Navarro, M., Ferrer, B., Trigo, J. M., Bilbao, A., Del Arco, I., 16. Berger, J., and Moller, D. E. (2002) Annu. Rev. Med. 53, 409–435 Cippitelli, A., Nava, F., Piomelli, D., and Rodriguez de Fonseca, F. (2002) 17. Bocher, V., Pineda-Torra, I., Fruchart, J. C., and Staels, B. (2002) Ann. N. Y. J. Neurosci. 22, 9612–9617 Acad. Sci. 967, 7–18 31. Lowe, M. E. (1997) Annu. Rev. Nutr. 17, 141–158 18. Wang, X., Miyares, R. L., and Ahern, G. P. (2005) J. Physiol. (Lond.) 564, 541–547 32. Kathuria, S., Gaetani, S., Fegley, D., Valino, F., Duranti, A., Tontini, A., Mor, M., Tarzia, G., La Rana, G., Calignano, A., Giustino, A., Tattoli, M., 19. Okamoto, Y., Morishita, J., Tsuboi, K., Tonai, T., and Ueda, N. (2004) Downloaded from J. Biol. Chem. 279, 5298–5305 Palmery, M., Cuomo, V., and Piomelli, D. (2003) Nat. Med. 9, 76–81 20. Wang, J., Okamoto, Y., Morishita, J., Tsuboi, K., Miyatake, A., and Ueda, 33. Shi, Y., and Burn, P. (2004) Nat. Rev. Drug Discov. 3, 695–710 N. (2006) J. Biol. Chem. 281, 12325–12335 34. Tsuboi, K., Hilligsmann, C., Vandevoorde, S., Lambert, D. M., and Ueda, 21. McKinney, M. K., and Cravatt, B. F. (2005) Annu. Rev. Biochem. 74, N. (2004) Biochem. J. 379, 99–106 411–432 35. Yang, Y., Chen, M., Georgeson, K. E., and Harmon, C. M. (2006) Am. J. 22. Tsuboi, K., Sun, Y. X., Okamoto, Y., Araki, N., Tonai, T., and Ueda, N. Physiol. Regul. Integr. Comp. Physiol., in press www.jbc.org (2005) J. Biol. Chem. 280, 11082–11092 36. Loverme, J., Guzman, M., Gaetani, S., and Piomelli, D. (2006) J. Biol. Chem. 23. Cravatt, B. F., Demarest, K., Patricelli, M. P., Bracey, M. H., Giang, D. K., 281, 22815–22818 Martin, B. R., and Lichtman, A. H. (2001) Proc. Natl. Acad. Sci. U. S. A. 98, 37. Ahern, G. P. (2003) J. Biol. Chem. 278, 30429–30434 9371–9376 38. Onaga, T., Zabielski, R., and Kato, S. (2002) Peptides 23, 279–290 at The Scripps Research Institute on March 5, 2007

1528 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 282•NUMBER 2•JANUARY 12, 2007