Fatty Acid Amide Hydrolase Determines Anandamide- induced Cell ...

THE JOURNAL OF BIOLOGICAL CHEMISTRY VOL. 281, NO. 15, pp. 10431–10438, April 14, 2006 © 2006 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in the U.S.A.

Fatty Acid Amide Hydrolase Determines Anandamideinduced Cell Death in the Liver* Received for publication, September 2, 2005, and in revised form, December 5, 2005 Published, JBC Papers in Press, January 17, 2006, DOI 10.1074/jbc.M509706200

So¨ren V. Siegmund‡, Ekihiro Seki‡, Yosuke Osawa‡, Hiroshi Uchinami‡, Benjamin F. Cravatt§, and Robert F. Schwabe‡1 From the ‡Department of Medicine, Columbia University, College of Physicians and Surgeons, New York, New York 10032 and the § Skaggs Institute for Chemical Biology, Departments of Cell Biology and Chemistry, The Scripps Research Institute, La Jolla, California 92037 The endocannabinoid anandamide (AEA) induces cell death in many cell types, but determinants of AEA-induced cell death remain unknown. In this study, we investigated the role of the AEAdegrading enzyme fatty acid amide hydrolase (FAAH) in AEA-induced cell death in the liver. Primary hepatocytes expressed high levels of FAAH and were completely resistant to AEA-induced cell death, whereas primary hepatic stellate cells (HSCs) expressed low levels of FAAH and were highly sensitive to AEA-induced cell death. Hepatocytes that were pretreated with the FAAH inhibitor URB597 or isolated from FAAHⴚ/ⴚ mice displayed increased AEA-induced reactive oxygen species (ROS) formation and were susceptible to AEA-mediated death. Conversely, overexpression of FAAH in HSCs prevented AEA-induced death. Since FAAH inhibition conferred only partial AEA sensitivity in hepatocytes, we analyzed additional factors that might regulate AEA-induced death. Hepatocytes contained significantly higher levels of glutathione (GSH) than HSCs. Glutathione depletion by DL-buthionine-(S,R)-sulfoximine rendered hepatocytes susceptible to AEA-mediated ROS production and cell death, whereas GSH ethyl ester prevented ROS production and cell death in HSCs. FAAH inhibition and GSH depletion had additive effects on AEA-mediated hepatocyte cell death resulting in almost 70% death after 24 h at 50 ␮M AEA and lowering the threshold for cell death to 500 nM. Following bile duct ligation, FAAHⴚ/ⴚ mice displayed increased hepatocellular injury, suggesting that FAAH protects hepatocytes from AEA-induced cell death in vivo. In conclusion, FAAH and GSH are determinants of AEAmediated cell death in the liver.

Arachidonyl ethanolamide, also termed anandamide (AEA),2 is the main endogenous agonist among a recently discovered class of lipid mediators termed endocannabinoids. Endocannabinoids evoke a wide spectrum of physiological actions that are mostly mediated through the G-protein-coupled cannabinoid receptors CB1 and CB2 (1, 2) but can also be independently of these receptors (3–7) and through vanilloid receptor 1 (VR1) (8). Endocannabinoids were initially described to play

* This study was supported by a research scholar award from the American Gastroenterological Association (to R. F. S.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 1 To whom correspondence should be addressed: Columbia University, P&S 9-460, 630 West 168th St., New York, NY 10032. Tel.: 212-342-0288; Fax: 212-305-9822; E-mail: [email protected]. 2 The abbreviations used are: AEA, arachidonyl ethanolamide; HSC, hepatic stellate cell; FAAH, fatty acid amide hydrolase; LDH, lactate dehydrogenase; VR1, vanilloid receptor 1; BSO, DL-buthionine-(S,R)-sulfoximine; ROS, reactive oxygen species; JNK, c-Jun NH2-terminal kinase; TNF, tumor necrosis factor; PI, propidium iodide; Ad, adenovirus; PARP, poly(ADP-ribose) polymerase-2; GFP, green fluorescent protein; CM-H2DCFDA, 5-(and-6)-chloromethyl-2⬘,7⬘-dichlorodihydrofluorescein diacetate; MES, 4-morpholineethanesulfonic acid; Z, benzyloxycarbonyl; FMK, fluoromethyl ketone.

APRIL 14, 2006 • VOLUME 281 • NUMBER 15

major roles in the central nervous system, where they regulate food intake, pain perception, and sleep. More recent studies have shown that endocannabinoids are also involved in the regulation of immunity, cell growth, and inflammation (reviewed in Ref. 9). Anandamide evokes different cellular responses including proliferation, growth arrest, decreased cell contractility, cell death, and anti-inflammatory pathways (6, 10 –16). Anandamide has been shown to activate JNK, Erk, ROS, and Ca2⫹ (17–20). In the liver, AEA is elevated in diseases such as acute hepatitis, fatty liver, and compensated liver cirrhosis (3, 21, 22). We have demonstrated that AEA efficiently mediates cell death in primary hepatic stellate cells (HSCs), the main fibrogenic cell type in the liver, but not in primary hepatocytes (6). This different response to AEA renders AEA a candidate for antifibrotic therapy in the liver since elimination of activated HSC is a mechanism to terminate the fibrogenic response, whereas hepatocyte cell death is involved in promoting fibrogenesis (23). AEA-induced cell death in HSCs depends on the presence of membrane cholesterol and subsequent formation of ROS and intracellular Ca2⫹ (6). Although we reported a difference in membrane cholesterol content and AEA binding between hepatocytes and HSCs, these factors alone are unlikely to explain the remarkable difference in susceptibility to AEA-mediated death. In this study, we investigated intracellular factors that potentially modulate AEA-induced cell death. Our study focuses on two candidate systems that may contribute to the resistance to AEA: (i) the AEA degrading enzyme fatty acid amide hydrolase (FAAH), which is highly expressed and active in the liver and represents the main mechanism that determines the biological half-life of AEA (24 –26), and (ii) GSH, a major determinant of apoptotic and necrotic cell death in hepatocytes that renders hepatocytes resistant to TNF-induced cell death (27, 28). Our study shows that high expression levels of GSH and FAAH correlate with resistance to AEA-mediated cell death in primary hepatic cell populations and that FAAH is a determinant of cell death in the liver in vivo.

EXPERIMENTAL PROCEDURES Isolation of Primary Hepatocyte and Hepatic Stellate Cells—Primary rat hepatocytes were isolated from male Sprague-Dawley rats (225–250 g, n ⫽ 15) by collagenase perfusion and cultured as described previously (6). Primary mouse hepatocytes were isolated from male FAAH⫺/⫺ mice or C57BL/6 FAAH⫹/⫹ controls (24) as described previously (29). Primary HSCs were isolated by a two-step Pronase-collagenase perfusion from livers of male Sprague-Dawley rats (300 – 450 g, n ⫽ 20) followed by a Nycodenz (Axis Shields, Oslo, Norway) two-layer discontinuous density gradient centrifugation as described previously (6). Purity of rat HSCs was consistently higher than 95%, as determined by vitamin A fluorescence 2 days after isolation. Hepatic stellate cells were cultured on uncoated plastic tissue culture dishes and considered cul-

JOURNAL OF BIOLOGICAL CHEMISTRY

10431

FAAH Prevents Anandamide-induced Cell Death ture-activated between day 7 and 14 after isolation. All animal procedures were approved by the Columbia University Institutional Animal Care and Use Committee and are in accordance with those set by the National Institutes of Health. Liver Injury Induction—To induce liver injury, 10-week-old male FAAH⫺/⫺ mice and FAAH⫹/⫹ mice underwent double ligation of the common bile duct. Five days after bile duct ligation, mice were sacrificed, serum was collected, and aspartate aminotransferase and alanine aminotransferase levels were determined (performed by Antech Diagnostic, Lake Success, NY). Some mice were sacrificed 7 and 21 days after bile duct ligation to measure FAAH mRNA expression and activity. All surgical procedures were approved by the Columbia University Institutional Animal Care and Use Committee and are in accordance with those set by the National Institutes of Health. Cell Treatment and Detection of Cell Death—Hepatocytes were kept in serum-free hormonally defined medium (30) for 12 h before experiments. HSCs were serum-starved with Dulbecco’s modified Eagle’s medium containing 0.5% fetal calf serum for 12 h. Cells were treated either with AEA (Sigma), methanandamide (Cayman Chemicals, Ann Arbor, MI), or vehicle (ethanol; 0.1% final concentration) or with actinomycin D (Sigma) plus murine TNF␣ (R&D Systems, Minneapolis, MN). Where indicated, cells were pretreated with GSH ethyl ester (Sigma), the FAAH inhibitors URB597 or oleoyl ethyl amide (both from Cayman Chemicals), the ␥-glutamyl cysteine synthase inhibitor DL-buthionine-(S,R)-sulfoximine (BSO; Sigma), CB1 antagonist SR141716, CB2 antagonist SR144528 (Sanofi-Synthe´labo, Montpellier, France), VR1 antagonist capsazepine (Sigma), or the pan-caspase inhibitor Z-VAD-FMK (R&D). Cell death in HSCs was measured by lactate dehydrogenase (LDH) release into the culture medium according to the manufacturer’s instructions (Roche Applied Science) and by propidium iodide (PI) fluorescence (Sigma) as described previously (6). Apoptosis was visualized by fluorescent microscopy using an annexin V/PI staining kit (Roche Applied Science) according to the manufacturer’s instructions. DNA laddering was performed as described previously (30). Adenovirus Construction and Infection—Mouse FAAH (31) was excised from pcDNA3.0 and inserted into AdTrack using XbaI and KpnI sites. AdTrack was linearized with PmeI and electroporated into AdEasy1 containing competent BJ 5183-AD1 bacteria (Stratagene, La Jolla, CA). Recombinants were screened using BSTX1 and, after linearization with PacI, transfected into HEK293 cells. After 10 days, 293 cells were lysed, and another round of HEK293 cells was infected to obtain cell lysates that were purified on a CsCl density gradient. Rat HSCs or FAAH⫺/⫺ mouse hepatocytes were infected with the FAAH-expressing adenoviruses (AdFAAH) or an AdTrack-based GFP-expressing control virus (Ad5GFP) at a multiplicity of infection of 50, achieving transduction rates of at least 90%. FAAH Activity Assay—FAAH activity was measured according to the method described by Patricelli and Cravatt (32). In brief, serum-starved activated rat HSCs (1.5 ⫻ 105 cells/well), rat hepatocytes (4 ⫻ 105 cells/ well), or whole liver tissue were lysed in 20 mM Hepes, pH 7.8, containing 10% glycerol, 150 mM NaCl, and 1% Triton X-100 at 4 °C. After centrifugation, 20 ␮l of the supernatant were added to 175 ␮l of reaction buffer (125 mM Tris, pH 9.0, and 1 mM EDTA) containing 1 ␮M FAAH substrate decanoyl m-nitroaniline (Cayman Chemicals), measured at 390 nm in a multiwell platereader (Fluostar Optima, BMG, Offenburg, Germany), and normalized to protein content as determined by Bradford assay. Detection of Reactive Oxygen Species—Serum-starved HSCs (2 ⫻ 104 cells/well) or hepatocytes (5 ⫻ 104 cells/well) were loaded with the redox-

10432 JOURNAL OF BIOLOGICAL CHEMISTRY

sensitive dye 5-(and-6)-chloromethyl-2⬘,7⬘-dichlorodihydrofluorescein diacetate (CM-H2DCFDA; Molecular Probes, Eugene, OR) for 30 min at 37 °C, washed, and stimulated with AEA. Reactive oxygen species formation was measured for the indicated time in a multiwell fluorescence platereader (Fluostar Optima, BMG) using excitation and emission filters of 485 and 535 nm, respectively. [3H]Anandamide Binding Assay—Activated rat HSCs (0.4 ⫻ 105/ well) or hepatocytes (1 ⫻ 105/well) were plated in 12-well dishes and serum-starved for 12 h. The cells were incubated with [3H]AEA (205 Ci/mmol; PerkinElmer Life Sciences) in triplicate for 10 min and washed extensively with phosphate-buffered saline at 4 °C. Extracts were prepared with 0.5 N NaOH containing 0.1% SDS and measured in a scintillation counter (PerkinElmer Life Sciences). Quantitative Real-time-PCR Analysis—RNA was isolated from serum-starved activated rat HSCs (day 7) and rat hepatocytes using the TRIzol method (Invitrogen). After DNase treatment, RNA was reverse-transcribed using random hexamer primers. Real-time PCR was performed for 40 cycles of 15 s at 95 °C and 60 s at 60 °C using an ABI 7000 sequence detection system (Applied Biosystems, Foster City, CA). Each sample was measured in duplicate, and quantification was performed by comparing the threshold cycle C⫹ values of each sample to a standard curve. Probes and primers for rat FAAH, CB1, CB2, VR1, and 18s were designed by ABI. FAAH, CB1, CB2, and VR1 levels were normalized to 18s and are expressed as fold induction in comparison with respective controls. Western Blot Analysis—Electrophoresis of protein extracts and subsequent blotting were performed as described (33). Blots were incubated with rabbit anti-FAAH (Cayman Chemical), anti-caspase-3, or anti-PARP (both from Cell Signaling Technologies, Beverly, MA) antibodies at a dilution of 1:1000 overnight at 4 °C. After incubation with secondary horseradish peroxidase-conjugated antibody (Santa Cruz Biotechnology, Santa Cruz, CA), the bands were visualized by the enhanced chemiluminescence light method (Amersham Biosciences) and exposed to a chemiluminescence imager (Image Station 2000R, Eastman Kodak Co.). Blots were reprobed with anti-actin mouse antibody (MP Biomedicals) to demonstrate equal loading. Measurement of Cellular Glutathione Levels—Serum-starved activated rat HSCs (1 ⫻ 106/well) and rat hepatocytes (4 ⫻ 106/well) were washed three times with ice-cold phosphate-buffered saline and scraped into 1⫻ MES lysis buffer. Total intracellular GSH levels were determined by the 5,5⬘-dithiobis(nitrobenzoic acid) method according to the manufacturer’s instructions (Cayman Chemical) and measured at 405 nm using a multiwell platereader (Fluostar Optima, BMG). Statistical Analysis—All data represent the mean of three independent experiments ⫾ S.E., if not otherwise stated. Unpaired two-tailed Student’s t tests or, where indicated, Mann-Whitney U tests, were performed to determine statistical significance using SigmaStat (SPSS, Chicago, IL). p values of ⬍0.05 were considered to be statistically significant.

RESULTS Anandamide Induces ROS Production and Cell Death in HSCs but Not in Hepatocytes—Anandamide induced an extremely rapid onset of cell death in primary rat HSCs with more than 40% cell death occurring after 2 h at 25 and 50 ␮M (Fig. 1A). After 24 h, more than 80% of HSCs underwent cell death and were PI-positive (Fig. 1, A and B), confirming our previous results that HSCs are extremely sensitive to undergoing AEA-induced necrosis (6). Hepatocytes, on the other hand, showed no increase in cell death after AEA treatment even at the highest doses of 100 ␮M (Fig. 1A), neither was there an increase in PI-positive cells (Fig.

VOLUME 281 • NUMBER 15 • APRIL 14, 2006

FAAH Prevents Anandamide-induced Cell Death

FIGURE 1. Different susceptibility to AEA-mediated death in primary hepatocytes and primary HSCs. A, serum-starved primary rat hepatocytes (left panel) and primary rat HSCs (right panel) were treated with the indicated concentrations of AEA or vehicle (⫺) for the indicated times or the positive control TNF␣ (30 ng/ml) plus actinomycin D (ActD) (0.2 ␮g/ml). Cell death was determined by LDH release. B, primary rat hepatocytes were treated with 100 ␮M AEA for 24 h, and primary rat HSCs were treated with 25 ␮M AEA for 4 h. Cell death is visualized by red fluorescence of nuclei stained by PI. C, primary rat hepatocytes (left panel) and HSCs (right panel) were loaded with CM-H2DCFDA (5 ␮M) for 30 min and exposed to AEA (50 ␮M), vehicle, or H2O2 (1 mM). Reactive oxygen species formation was measured as described. All figures are representative of at least three independent experiments.

1B). We had previously shown that ROS contribute to AEA-mediated death in HSCs (6). We therefore investigated whether ROS production after AEA differed between HSCs and hepatocytes. Indeed, hepatocytes showed a much lower production of ROS after AEA than HSCs, whereas H2O2 induced a comparable increase in CM-H2DCFDA-fluorescence in both cell types (Fig. 1C). FAAH Determines Resistance to AEA-mediated Death in Hepatocytes—Anandamide is inactivated by the cell membrane enzyme FAAH, and it has previously been shown that the liver is the organ with the highest expression and activity levels of FAAH (24 – 26). We compared the levels of FAAH in rat hepatocytes and rat HSCs to investigate whether a differential expression of FAAH may contribute to the different AEA response in these two cell types. Indeed, rat hepatocytes expressed 70-fold higher levels of FAAH than rat HSCs as determined by quantitative real-time PCR (Fig. 2A, p ⬍ 0.01). To confirm these data, we performed Western blot analysis to detect FAAH protein expression. Although FAAH was robustly expressed in rat hepatocytes, almost no FAAH was detectable in rat HSCs by Western blot (Fig. 2B). Accordingly, no FAAH activity was detectable in rat HSCs, whereas rat hepatocytes exhib-


ited high FAAH activity, which was completely blocked by preincubation with URB597, a highly selective FAAH inhibitor (Fig. 2C) (34). Next, we investigated whether FAAH contributes to the resistance to AEA-mediated cell death by pretreating hepatocytes with URB597. URB597 sensitized hepatocytes to the effects of AEA with more than 40% cell death (Fig. 2D, left panel) and strong PI uptake (Fig. 2D, middle panel), suggesting that FAAH is indeed critically involved in the resistance to AEA. Similar data were obtained using another FAAH inhibitor, oleoyl ethyl amide (data not shown). Pretreatment with URB597 also increased AEA-induced ROS production in hepatocytes, whereas URB597 alone did not increase ROS (Fig. 2D, right panel). In contrast to AEA, methanandamide, an AEA analogue that is resistant to FAAH-mediated degradation, was capable of inducing cell death in rat hepatocytes (Fig. 2E, p ⬍ 0.05). To further investigate the role of FAAH in AEA-induced cell death, we constructed an FAAH-expressing adenovirus (AdFAAH), allowing us to overexpress FAAH in HSCs. Infection with AdFAAH resulted in a strong expression of FAAH and high levels of FAAH activity in HSCs (Fig. 2F). AdFAAH induced almost complete resistance to AEA-mediated death up to doses of 25 and 50 ␮M with less than 20% cell death and no increase in PI uptake (Fig. 2G). At 100 ␮M AEA, AdFAAH conferred partial protection to HSCs. In contrast, HSCs infected with the GFP-expressing control virus (AdGFP) remained sensitive to AEA-induced cell death (Fig. 2G, left panel) and showed a strong uptake of PI and leakage of GFP as signs of necrotic cell death (Fig. 2G, right panel). FAAH⫺/⫺ Hepatocytes Are Sensitive to AEA-mediated Death—To further confirm the role of FAAH in AEA-mediated death in hepatocytes, we isolated hepatocytes from FAAH⫺/⫺ and FAAH⫹/⫹ mice. FAAH⫺/⫺ hepatocytes were sensitive to AEA-mediated cell death, whereas FAAH⫹/⫹ hepatocytes did not undergo AEA-induced death (Fig. 3A, p ⬍ 0.05). We observed almost 40% cell death after 24 h of 50 ␮M AEA, which is comparable with the amount of cell death induced by AEA in the presence of FAAH inhibitors. When we infected FAAH⫺/⫺ hepatocytes with AdFAAH, resistance to AEAmediated death was completely restored, whereas hepatocytes infected with AdGFP showed strong uptake of PI and GFP leakage as signs of necrosis (Fig. 3B). FAAH Does Not Determine AEA Cellular Uptake—Fatty acid amide hydrolase has not only been implicated in the degradation of AEA but also as a determinant of AEA cellular uptake (35). To investigate whether FAAH was involved in the uptake of AEA in hepatocytes and HSCs, we analyzed AEA cellular uptake in FAAH⫺/⫺ hepatocytes that had been infected with AdFAAH or AdGFP. Hepatocytes that overexpressed FAAH showed no different rate of AEA uptake from the GFPoverexpressing FAAH⫺/⫺ mice, indicating that FAAH is not involved in AEA uptake in hepatocytes (Fig. 3C, left panel). Similarly, FAAH overexpression in HSCs, which do not express significant amounts of FAAH, did not increase AEA uptake (Fig. 3C, right panel). GSH Levels Determine Sensitivity to AEA-mediated Death—Although pharmacological or genetic inactivation of FAAH resulted in a sensitization to AEA-mediated death, only about 40% of hepatocytes died after 24 h of AEA treatment, whereas HSCs showed more than 80% cell death after 24 h. These data suggest the presence of additional mechanisms that render hepatocytes resistant to the effects of AEA. Because ROS formation is crucial for the cytotoxic effects of AEA, we determined whether hepatocytes possess more efficient antioxidant defenses than HSCs. GSH is a major cellular defense system against ROS and protects hepatocytes from undergoing necrotic and apoptotic cell death (27, 28). Measurement of


10433


FIGURE 2. FAAH determines AEA-mediated death in primary hepatocytes (Hep) and HSCs. A, mRNA expression in rat HSCs and rat hepatocytes was determined by quantitative real-time PCR (n ⫽ 3; **, p ⬍ 0.001). B, FAAH protein expression was analyzed in rat hepatocytes and rat HSCs by Western blotting. C, FAAH activity was analyzed in hepatocytes that were pretreated with the FAAH inhibitor URB597 (10 ␮M) or vehicle (0.1% Me2SO) for 1 h and in HSCs (**, p ⬍ 0.001). D, rat hepatocytes were pretreated with either vehicle (Me2SO) or URB597 (10 ␮M) for 1 h followed by AEA (100 ␮M) or vehicle for 24 h. Cell death was determined by LDH release (left panel, **, p ⬍ 0.001) and visualized by nuclear uptake of PI (middle panel). ROS formation was measured by CM-H2DCFDA fluorescence in vehicle- or URB597-pretreated cells after the addition of 100 ␮M AEA. (right panel). E, rat hepatocytes were treated with vehicle, methanandamide (Meth-AEA), or AEA (100 ␮M) for 24 h. Cell death was measured by LDH release (*, p ⬍ 0.05). F, expression of FAAH protein (left panel) and FAAH activity (right panel, **, p ⬍ 0.001) in rat HSCs infected with AdGFP or AdFAAH was performed as described under “Experimental Procedures.” G, rat HSCs were infected with AdGFP or AdFAAH and treated with AEA (5–100 ␮M) for 4 h. Cell death was determined by LDH release (left panel, *, p ⬍ 0.05 versus Ad5GFP) and visualized by cellular uptake of PI (right panel). All figures are representative of three independent experiments.

GSH levels in hepatocytes and HSCs showed that hepatocytes contained 10-fold higher amounts of GSH than HSCs (Fig. 4A, p ⬍ 0.05). Pretreatment with the ␥-glutamyl cysteine synthase inhibitor BSO efficiently depleted GSH in hepatocytes (Fig. 4A) and rendered hepa-


tocytes sensitive to AEA-mediated death (Fig. 4B, p ⬍ 0.05). Conversely, GSH ethyl ester pretreatment protected HSCs from undergoing cell death after AEA (Fig. 4B, p ⬍ 0.05). Accordingly, AEAinduced intracellular ROS production was increased in hepatocytes



FIGURE 3. FAAH-deficient hepatocytes are susceptible to AEA-mediated death. A, in the left panel, primary hepatocytes from wild type (WT) or FAAH⫺/⫺ mice were treated with AEA (5–100 ␮M) or ActD plus TNF␣ for 24 h. Cell death was measured by LDH release into the media (*, p ⬍ 0.05 versus wild type). The right panel shows phase contrast and PI pictures of FAAH⫺/⫺ mouse hepatocytes treated with AEA (10 –100 ␮M) for 24 h. B, FAAH⫺/⫺ hepatocytes were infected with either Ad5GFP or AdFAAH followed by treatment with AEA (100 ␮M) for 24 h. Expression of FAAH protein in FAAH⫺/⫺ hepatocytes after infection was determined by Western blot (left panel, inset). Cell death was determined by LDH assay (left panel, *, p ⬍ 0.05 versus AdGFP) and visualized by nuclear uptake of PI (right panel). Veh, vehicle. C, FAAH⫺/⫺ hepatocytes and rat HSCs were infected with either Ad5GFP or AdFAAH. FAAH⫺/⫺ hepatocytes (left panel) and rat HSCs (right panel) were exposed to the indicated concentrations of [3H]AEA for 10 min and washed and lysed as described. [3H]AEA uptake was measured in a scintillation counter and normalized to the number of cells per well. All figures are representative of three independent experiments.

after GSH depletion (Fig. 4A, right panel), whereas GSH ethyl ester pretreatment of HSCs reduced ROS production (Fig. 4B, right panel). Although HSCs did not contain large amounts of GSH, pretreatment with BSO also sensitized HSCs to AEA-mediated cell death and lowered the threshold for AEA-induced cell death to 1 ␮M after 4 h of AEA (Fig. 4C). When hepatocytes were pretreated with a combination of BSO and URB597, AEA induced almost 70% cell death at a concentration of 50 ␮M, and the threshold for AEA-induced cell death was lowered to 500 nM (Fig. 4D). Thus, FAAH and GSH are main determinants of AEA-induced cell death in the liver. AEA Induces Necrosis in Hepatocytes after GSH Depletion or FAAH Inhibition—We had previously shown that AEA induces necrosis and not apoptosis in HSCs. However, in hepatoma cell lines, we find that AEA induces apoptosis.3 To determine whether necrosis or apoptosis was the predominant form of AEA-induced death in hepatocytes after GSH depletion or FAAH inhibition, we tested whether AEA induced caspase-3 cleavage and the 85-kDa PARP cleavage product, two hallmarks of apoptosis. In contrast to hepatocytes that were treated with TNF␣ plus actinomycin D, we did not observe caspase-3 or PARP cleav3

S. V. Siegmund and R. F. Schwabe, unpublished observation.


age in hepatocytes treated with combinations of BSO, URB597, and AEA (Fig. 5A). Additionally, AEA-treated hepatocytes were not annexin V-positive but instead showed strong uptake of PI (Fig. 5B). Although TNF␣ plus actinomycin D induced DNA ladder formation, another typical feature of apoptosis, combinations of BSO, URB597, and AEA, did not induce DNA ladder formation (Fig. 5C). Pretreatment with the pan-caspase inhibitor Z-VAD-FMK blocked TNF␣ but not AEA-induced cell death (Fig. 5D). Altogether, these data suggest that after GSH depletion and FAAH inhibition, AEA-mediated death in hepatocytes is purely necrotic. Next, we determined whether AEA-mediated cell death in hepatocytes is mediated by CB1, CB2, or VR1 receptors. Quantitative real-time PCR showed that cultured primary rat hepatocytes did not express significant amounts of either CB1 or CB2 in comparison with brain or spleen, respectively, whereas VR1 mRNA was about 10% of the levels found in brain (Fig. 5E). Accordingly, the CB1 antagonist SR141716 and the CB2 antagonist SR144528 did not protect hepatocytes sensitized with BSO and URB597 from AEA-induced cell death (Fig. 5F). Moreover, the VR1 antagonist capsazepine did not show protective effects (Fig. 5F), suggesting that AEA-mediated cell death in hepatocytes is independent of CB1, CB2, and VR1. FAAH Determines Hepatocellular Injury in Vivo—To characterize the regulation of FAAH during liver injury, we measured the levels of FAAH expression and activity after bile duct ligation. In this model of hepatic injury, FAAH mRNA expression and activity significantly decreased (Fig. 6, A and B). To follow up on our in vitro results that showed a crucial role for FAAH in protecting hepatocytes from AEAinduced cell death, we performed bile duct ligation in FAAH⫺/⫺ and FAAH⫹/⫹ mice and sacrificed mice after 5 days to determine hepatic injury. After bile duct ligation, serum levels of aspartate aminotransferase and alanine aminotransferase, two indicators of hepatocellular injury, were 6-fold higher in FAAH⫺/⫺ mice in comparison with FAAH⫹/⫹ mice (Fig. 6C, p ⬍ 0.05). Consistent with these measurements, we detected a higher number of necrotic areas in FAAH⫺/⫺ mice after bile duct ligation (Fig. 6D), suggesting that FAAH is indeed required to protect hepatocytes from anandamide-induced cell death in vivo.

DISCUSSION The ability to induce cell death in a wide range of non-transformed and cancer cells has made AEA an interesting target for cancer and inflammatory and degenerative diseases (9, 37). Knowledge about mechanisms that determine AEA sensitivity in primary cells may be useful for therapeutic exploitation of the endocannabinoid system. We have recently described that the endocannabinoid AEA selectively induces cell death in primary HSCs, the main fibrogenic cell type of the liver, but does not induce death in primary hepatocytes (6). A selective induction of cell death in HSCs has been linked to the resolution of liver fibrosis, whereas cell death in hepatocytes worsens liver function and promotes fibrogenesis (23). These properties suggest that modulation of the endocannabinoid system may be useful for the treatment of hepatic injury and fibrosis. Fatty acid amide hydrolase is the main AEA-degrading enzyme. Among all organs, the liver has the highest FAAH expression and activity levels (24 –26). However, the function of FAAH in the liver has not been determined. Although previous studies have shown that FAAH is involved in limiting the effects of AEA on nociception, anxiolysis, and hemodynamics (24, 36), the role of FAAH in AEA-mediated cell death remains unclear. One study reported that an increase of FAAH activity induced by follicle-stimulating hormone correlated with reduced AEAinduced cell death in Sertoli cells (37). We now present several lines of


10435


FIGURE 4. GSH determines AEA-mediated cell death. A, intracellular GSH levels were determined in primary rat hepatocytes (Hep) and primary rat HSCs with or without BSO pretreatment (100 ␮M for 1 h). *, p ⬍ 0.05 versus untreated, #, p ⬍ 0.05 versus HSCs. B, cell death was determined by LDH release (left panel, **, p ⬍ 0.001) or PI uptake (middle panel) in rat hepatocytes after pretreatment with BSO (100 ␮M) or vehicle followed by AEA (100 ␮M) for 24 h. ROS formation was measured by CM-H2DCFDA fluorescence (right panel). C, rat HSCs were pretreated with GSH ethyl ester (GSH-EE) (4 mM, upper left panel), BSO (100 ␮M, lower left panel), or vehicle prior to exposure to AEA for 4 h. Cell death was determined by PI assay (*, p ⬍ 0.05, **, p ⬍ 0.001) and visualized by nuclear uptake of PI (middle panel). AEA-induced ROS formation in rat HSCs after GSH ethyl ester pretreatment was analyzed as described above (right panel). D, rat hepatocytes were pretreated with vehicle, URB597 (10 ␮M), BSO (100 ␮M), or URB597 plus BSO for 1 h followed by AEA (100 ␮M) for 24 h (left panel, **, p ⬍ 0.001) or exposed to increasing concentrations of AEA (0.5–100 ␮M, right panel). Cell death was analyzed by LDH release.

evidence that FAAH is an important determinant of AEA-induced cell death in vitro as well as in vivo. (i) Low FAAH expression in HSCs correlates with AEA susceptibility, whereas high AEA expression in


hepatocytes correlates with resistance to AEA. (ii) Pretreatment with the FAAH inhibitor URB597 restores AEA-induced ROS generation and cell death in hepatocytes. (iii) The AEA analogue methanandamide,



FIGURE 5. FAAH inhibition and GSH depletion induce necrosis, but not apoptosis, in hepatocytes after AEA treatment. A–C, rat hepatocytes were pretreated with BSO (100 ␮M), URB597 (10 ␮M), or both for 1 h and then exposed to vehicle or AEA (100 ␮M) for 24 h or less where indicated or treated with actinomycin D (ActD) plus TNF␣. A, Western blotting was performed with antibodies directed against caspase-3, PARP, and actin. B, apoptotic cell death is indicated by green fluorescence of annexin V (Ann V), and necrotic cell death is indicated by red fluorescence PI. C, DNA was extracted and analyzed on a 1.8% agarose gel. D, BSO- or URB597-pretreated rat hepatocytes were exposed to the pan-caspase inhibitor Z-VAD-FMK (20 ␮M) for 30 min followed either by incubation with AEA (100 ␮M) or by treatment with actinomycin plus TNF␣ for 16 h. Cell death was measured by LDH assay (*, p ⬍ 0.05). E, mRNA expression of CB1, CB2, and VR1 receptors in primary rat hepatocytes (rHep, n ⫽ 3) and control tissue (rat brain for CB1 and VR1, rat spleen for CB2) was determined by quantitative real-time PCR. F, primary rat hepatocytes were incubated with BSO (100 ␮M), URB597 (10 ␮M), SR141716 (SR1; 5 ␮M), SR144528 (SR2; 5 ␮M), or capsazepine (CPZ; 10 ␮M) for 1 h followed by treatment with vehicle or AEA (25 ␮M) for 24 h. Cell death was analyzed by LDH release.

FIGURE 6. FAAH determines liver injury after bile duct ligation. A and B, liver injury was induced by double ligation of the common bile duct. Hepatic FAAH mRNA levels were determined by quantitative real-time PCR and normalized to 18s levels (A). *, p ⬍ 0.05. FAAH activity was determined in hepatic extracts as described under “Experimental Procedures” (B). *, p ⬍ 0.05. C and D, hepatic injury was induced by bile duct ligation (BDL) in FAAH⫺/⫺ mice and FAAH⫹/⫹ control mice. Five days after bile duct ligation, mice were sacrificed, and serum aspartate aminotransferase and alanine aminotransferase levels were measured (C). *, p ⬍ 0.05, determined by Mann-Whitney U test. Paraffin sections of liver after days of bile duct ligation were stained with hematoxylin and eosin for the detection of hepatocellular necrosis. Representative fields are shown at ⫻4 and ⫻10 magnification, and arrows indicate necrotic areas (D).

which is resistant to FAAH-mediated degradation, induces cell death in hepatocytes. (iv) FAAH⫺/⫺ hepatocytes are susceptible to AEA-mediated cell death. (v) Overexpression of FAAH in HSCs and FAAH⫺/⫺ hepatocytes almost completely prevents AEA-induced cell death. (vi) FAAH⫺/⫺ mice show increased hepatocellular injury following bile duct ligation. Our results were consistent with the hypothesis that FAAH is crucial in preventing AEA-induced cytotoxicity. After pharmacological or genetic inactivation of FAAH, AEA-mediated cell death was only partial with about 40% of hepatocytes undergoing cell death, whereas HSC showed almost 90% cell death after AEA. Therefore, additional mechanisms must contribute to the resistance of hepatocytes to AEA. Glutathione is a major determinant of apoptotic and necrotic cell death in the liver and acts as a scavenger for ROS (27, 28, 38). Moreover, ROS generation is required for AEA-mediated cell

death in HSCs but is almost absent in primary hepatocytes after AEA. Our data shows that the cellular GSH content is an additional factor that determines the response to AEA-induced cytotoxicity. Hepatocytes contain high levels of GSH and thus are able to block ROS generation evoked by AEA. When hepatocytes are depleted of GSH, they not only show an enhanced rate of ROS formation but also become susceptible to AEA-mediated cell death. Conversely, the low GSH levels in HSCs render them more susceptible to AEA, and pretreatment with GSH ethyl ester confers resistance to AEA in HSCs. Interestingly, the effects of FAAH inhibition and GSH depletion in hepatocytes are additive, demonstrating that resistance to AEA indeed occurs at different levels. The combination of FAAH inhibition and GSH depletion lowers the threshold for AEA-induced cell death in hepatocytes to 500 nM, a concentration that is close to the range of AEA levels in liver disease (3, 21,



10437

FAAH Prevents Anandamide-induced Cell Death 22). It is known that GSH levels decrease in animal models of bile duct ligation (39). Therefore, it is likely that the high degree of liver injury in bile duct-ligated FAAH⫺/⫺ mice is caused by the combination of absent FAAH activity and lower GSH levels. In agreement with previous studies that demonstrated elevated levels of AEA in several organs of FAAH⫺/⫺ mice including liver (24, 26), we found about 8-fold higher hepatic levels of AEA in FAAH⫺/⫺ mice than in wild type mice (data not shown). Since FAAH is predominantly located at the cell membrane, our in vitro and in vivo results are consistent with the notion that FAAH deficiency increases the amount of AEA that cells are exposed to and thus predisposes them to cell death. Our study demonstrates that FAAH activity levels are decreased after bile duct ligation and after administration of CCl4 (data not shown). Since GSH depletion occurs in several other pathological states of the liver and is especially remarkable in alcoholic liver disease (38), it is conceivable that hepatic AEA signaling and cell death are modulated under pathological conditions such as alcoholic liver disease or biliary obstruction due to low hepatic GSH and FAAH levels. Our study did not detect considerable amounts of CB1 and CB2 mRNA in three different preparations of rat hepatocytes. Although this result is in contrast with a recent study that demonstrates CB1 expression in zone III hepatocytes in mouse liver (21), it is possible that this is due to species differences or that CB1 expression disappears when hepatocytes are cultured. In hepatocytes, AEA-induced cell death was independent of CB1, CB2, or VR1 as judged by the use of receptorspecific antagonists. It remains to be determined whether FAAH prevents CB1-, CB2-, or VR1-dependent cell death in other cell types in a similar manner. Despite the majority of authors demonstrating that AEA induces apoptosis and not necrosis (40), we could not detect any signs of apoptosis in AEA-treated HSCs or hepatocytes. Future studies should address whether FAAH and GSH regulate AEA-induced apoptosis in a similar fashion and whether they play a role in determining AEA-induced cell death in other tissues besides the liver and in cancer. In addition to its effects on AEA degradation, FAAH has also been suggested to determine the cellular uptake of AEA (35). We overexpressed FAAH in FAAH⫺/⫺ hepatocytes and in HSCs and compared the rates of AEA cellular uptake. Our results show that FAAH overexpression causes a small decrease in AEA cellular uptake that is most likely due to the rapid degradation of AEA. These results are in contrast to the hypothesis of Glaser et al. (35), who suggested that the intracellular hydrolysis of AEA by FAAH creates a concentration gradient across the cell membrane and thus promotes AEA uptake. Several recent studies have come to the conclusion that AEA cellular uptake is not mediated by FAAH (41, 42), but there is still considerable debate regarding this issue. Our results support the hypothesis that FAAH primarily serves to protect target cells from the effects of AEA. Previous studies have shown that hepatic FAAH expression and activity are the highest in the body and even exceed those of the brain (24 –26). It remains to be speculated whether hepatic FAAH serves purposes in addition to protecting hepatocytes from AEA-induced cell death. REFERENCES 1. Devane, W. A., Hanus, L., Breuer, A., Pertwee, R. G., Stevenson, L. A., Griffin, G., Gibson, D., Mandelbaum, A., Etinger, A., and Mechoulam, R. (1992) Science 258, 1946 –1949 2. Di Marzo, V., Fontana, A., Cadas, H., Schinelli, S., Cimino, G., Schwartz, J. C., and Piomelli, D. e. (1994) Nature 372, 686 – 691 3. Biswas, K. K., Sarker, K. P., Abeyama, K., Kawahara, K., Iino, S., Otsubo, Y., Saigo, K., Izumi, H., Hashiguchi, T., Yamakuchi, M., Yamaji, K., Endo, R., Suzuki, K., Imaizumi, H., and Maruyama, I. (2003) Hepatology 38, 1167–1177 4. Movsesyan, V. A., Stoica, B. A., Yakovlev, A. G., Knoblach, S. M., Lea, P. M. t., Cernak, I., Vink, R., and Faden, A. I. (2004) Cell Death Differ. 11, 1121–1132


5. Sarker, K. P., and Maruyama, I. (2003) Cell Mol. Life Sci. 60, 1200 –1208 6. Siegmund, S. V., Uchinami, H., Osawa, Y., Brenner, D. A., and Schwabe, R. F. (2005) Hepatology 41, 1085–1095 7. Begg, M., Pacher, P., Batkai, S., Osei-Hyiaman, D., Offertaler, L., Mo, F. M., Liu, J., and Kunos, G. (2005) Pharmacol. Ther. 106, 133–145 8. Smart, D., Gunthorpe, M. J., Jerman, J. C., Nasir, S., Gray, J., Muir, A. I., Chambers, J. K., Randall, A. D., and Davis, J. B. (2000) Br. J. Pharmacol. 129, 227–230 9. Di Marzo, V., Bifulco, M., and De Petrocellis, L. (2004) Nat. Rev. Drug Discov. 3, 771–784 10. Walker, J. M., Huang, S. M., Strangman, N. M., Tsou, K., and Sanudo-Pena, M. C. (1999) Proc. Natl. Acad. Sci. U. S. A. 96, 12198 –12203 11. Fischer, R., Cariers, A., Reinehr, R., and Haussinger, D. (2002) Gastroenterology 123, 845– 861 12. Valk, P., Verbakel, S., Vankan, Y., Hol, S., Mancham, S., Ploemacher, R., Mayen, A., Lowenberg, B., and Delwel, R. (1997) Blood 90, 1448 –1457 13. Ligresti, A., Bisogno, T., Matias, I., De Petrocellis, L., Cascio, M. G., Cosenza, V., D’Argenio, G., Scaglione, G., Bifulco, M., Sorrentini, I., and Di Marzo, V. (2003) Gastroenterology 125, 677– 687 14. McKallip, R. J., Lombard, C., Fisher, M., Martin, B. R., Ryu, S., Grant, S., Nagarkatti, P. S., and Nagarkatti, M. (2002) Blood 100, 627– 634 15. Maccarrone, M., Lorenzon, T., Bari, M., Melino, G., and Finazzi-Agro, A. (2000) J. Biol. Chem. 275, 31938 –31945 16. Batkai, S., Jarai, Z., Wagner, J. A., Goparaju, S. K., Varga, K., Liu, J., Wang, L., Mirshahi, F., Khanolkar, A. D., Makriyannis, A., Urbaschek, R., Garcia, N., Jr., Sanyal, A. J., and Kunos, G. (2001) Nat. Med. 7, 827– 832 17. Rueda, D., Galve-Roperh, I., Haro, A., and Guzman, M. (2000) Mol. Pharmacol. 58, 814 – 820 18. Sarker, K. P., Biswas, K. K., Yamakuchi, M., Lee, K. Y., Hahiguchi, T., Kracht, M., Kitajima, I., and Maruyama, I. (2003) J. Neurochem. 85, 50 – 61 19. Berdyshev, E. V., Schmid, P. C., Krebsbach, R. J., Hillard, C. J., Huang, C., Chen, N., Dong, Z., and Schmid, H. H. (2001) Biochem. J. 360, 67–75 20. Derkinderen, P., Valjent, E., Toutant, M., Corvol, J. C., Enslen, H., Ledent, C., Trzaskos, J., Caboche, J., and Girault, J. A. (2003) J. Neurosci. 23, 2371–2382 21. Osei-Hyiaman, D., Depetrillo, M., Pacher, P., Liu, J., Radaeva, S., Batkai, S., HarveyWhite, J., Mackie, K., Offertaler, L., Wang, L., and Kunos, G. (2005) J. Clin. Investig. 115, 1298 –1305 22. Fernandez-Rodriguez, C. M., Romero, J., Petros, T. J., Bradshaw, H., Gasalla, J. M., Gutierrez, M. L., Lledo, J. L., Santander, C., Fernandez, T. P., Tomas, E., Cacho, G., and Walker, J. M. (2004) Liver Int. 24, 477– 483 23. Bataller, R., and Brenner, D. A. (2005) J. Clin. Investig. 115, 209 –218 24. Cravatt, B. F., Demarest, K., Patricelli, M. P., Bracey, M. H., Giang, D. K., Martin, B. R., and Lichtman, A. H. (2001) Proc. Natl. Acad. Sci. U. S. A. 98, 9371–9376 25. Cravatt, B. F., Giang, D. K., Mayfield, S. P., Boger, D. L., Lerner, R. A., and Gilula, N. B. (1996) Nature 384, 83– 87 26. Cravatt, B. F., Saghatelian, A., Hawkins, E. G., Clement, A. B., Bracey, M. H., and Lichtman, A. H. (2004) Proc. Natl. Acad. Sci. U. S. A. 101, 10821–10826 27. Colell, A., Garcia-Ruiz, C., Miranda, M., Ardite, E., Mari, M., Morales, A., Corrales, F., Kaplowitz, N., and Fernandez-Checa, J. C. (1998) Gastroenterology 115, 1541–1551 28. Nagai, H., Matsumaru, K., Feng, G., and Kaplowitz, N. (2002) Hepatology 36, 55– 64 29. Schwabe, R. F., Uchinami, H., Qian, T., Bennett, B. L., Lemasters, J. J., and Brenner, D. A. (2004) FASEB J. 18, 720 –722 30. Schwabe, R. F., Bennett, B. L., Manning, A. M., and Brenner, D. A. (2001) Hepatology 33, 81–90 31. Giang, D. K., and Cravatt, B. F. (1997) Proc. Natl. Acad. Sci. U. S. A. 94, 2238 –2242 32. Patricelli, M. P., and Cravatt, B. F. (2001) Biochemistry 40, 6107– 6115 33. Schwabe, R. F., Schnabl, B., Kweon, Y. O., and Brenner, D. A. (2001) J. Immunol. 166, 6812– 6819 34. 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., Palmery, M., Cuomo, V., and Piomelli, D. (2003) Nat. Med. 9, 76 – 81 35. Glaser, S. T., Abumrad, N. A., Fatade, F., Kaczocha, M., Studholme, K. M., and Deutsch, D. G. (2003) Proc. Natl. Acad. Sci. U. S. A. 100, 4269 – 4274 36. Pacher, P., Batkai, S., Osei-Hyiaman, D., Offertaler, L., Liu, J., Harvey-White, J., Brassai, A., Jarai, Z., Cravatt, B. F., and Kunos, G. (2005) Am. J. Physiol. 289, H533–H541 37. Maccarrone, M., Cecconi, S., Rossi, G., Battista, N., Pauselli, R., and Finazzi-Agro, A. (2003) Endocrinology 144, 20 –28 38. Fernandez-Checa, J. C., and Kaplowitz, N. (2005) Toxicol. Appl. Pharmacol. 204, 263–273 39. Krahenbuhl, S., Talos, C., Lauterburg, B. H., and Reichen, J. (1995) Hepatology 22, 607– 612 40. Maccarrone, M., and Finazzi-Agro, A. (2003) Cell Death Differ. 10, 946 –955 41. Ligresti, A., Morera, E., Van Der Stelt, M., Monory, K., Lutz, B., Ortar, G., and Di Marzo, V. (2004) Biochem. J. 380, 265–272 42. Fegley, D., Kathuria, S., Mercier, R., Li, C., Goutopoulos, A., Makriyannis, A., and Piomelli, D. (2004) Proc. Natl. Acad. Sci. U. S. A. 101, 8756 – 8761
