Eur. J. Biochem. 267, 2991±2997 (2000) q FEBS 2000

Down-regulation of anandamide hydrolase in mouse uterusby sex hormones

Mauro Maccarrone,1 Massimo De Felici2, Monica Bari,1 Francesca Klinger,2 Gregorio Siracusa2 andAlessandro Finazzi-AgroÁ 1

1Department of Experimental Medicine and Biochemical Sciences, and 2Department of Public Health and Cell Biology,

University of Rome `Tor Vergata', Italy

Endocannabinoids are an emerging class of lipid mediators, which mimic several effects of cannabinoids.

Anandamide (arachidonoylethanolamide) is a major endocannabinoid, which has been shown to impair

pregnancy and embryo development. The activity of anandamide is controlled by cellular uptake through a

specific transporter and intracellular degradation by the enzyme anandamide hydrolase (fatty acid amide

hydrolase, FAAH). We characterized FAAH in mouse uterus by radiochromatographic and immunochemical

techniques, showing that the enzyme is confined to the epithelium and its activity decreases appreciably during

pregnancy or pseudopregnancy because of lower gene expression at the translational level. Ovariectomy

prevented the decrease in FAAH, and both progesterone and estrogen further reduced its basal levels, suggesting

hormonal control of the enzyme. Anandamide was shown to induce programmed cell death in mouse blastocysts,

through a pathway independent of type-1 cannabinoid receptor. Blastocysts, however, have a specific anandamide

transporter and FAAH, which scavenge this lipid. Taken together, these results provide evidence of an interplay

between endocannabinoids and sex hormones in pregnancy. These findings may also be relevant for human

fertility, as epithelial cells from healthy human uterus showed FAAH activity and expression, which in

adenocarcinoma cells was increased fivefold.

Keywords: anandamide hydrolase; apoptosis; development; endocannabinoids; sex hormones.

Endocannabinoids are an emerging class of lipid mediators,isolated from brain and peripheral tissues [1,2] and found alsoin chocolate [3] and milk [4]. They mimic the psychotropic,hypnotic, tranquilizing, antiemetic, anticonvulsive and analgesiceffects of cannabinoids [5,6]. These compounds, in particularD9-tetrahydrocannabinol, have been reported to have adverseeffects on reproductive functions, including retarded embryodevelopment, fetal loss and pregnancy failure [7,8]. A majorendocannabinoid, anandamide (arachidonoylethanolamide), hasbeen shown to impair pregnancy and embryo development [9].Down-regulation of anandamide levels in mouse uterus hasbeen associated with uterine receptivity, while up-regulationcorrelated with uterine refractoriness to embryo implantation[10]. Remarkably, mouse uterus contains the highest levels ofanandamide detected so far in mammalian tissues, and is theonly tissue in which anandamide is the main component (up to95%) of N-acylethanolamines [10]. Anandamide is an endo-genous ligand for both the brain-type (CB1-R) and spleen-type(CB2-R) cannabinoid receptors, mimicking several actions ofcannabinoids on the central nervous system and in peripheraltissues [11]. Mouse embryos express both CB1-R and CB2-RmRNA, the levels of the former being much higher than thosefound in brain [7,9,10]. CB1-R activation is detrimental for

mouse preimplantation development [10,12], but appears toaccelerate trophoblast differentiation [13].

The effect of anandamide through CB1-R and CB2-Rdepends on its concentration in the extracellular space, whichis controlled by a two-step process: (a) cellular uptake by aspecific anandamide transporter, and (b) intracellular degrada-tion by the enzyme anandamide hydrolase (fatty acid amidehydrolase, FAAH). Anandamide transporter and FAAH havebeen characterized in mammalian cell lines [14±16] and morerecently in human cells in culture and brain [17]. FAAH activityhas also been demonstrated in mouse uterus [9], although themethodology used was not suitable for accurate kinetic analy-sis. FAAH mRNA level has been measured in both mouseuterus and embryos during the peri-implantation period [18].Using a very sensitive radiochromatographic method [19], inthe present study we characterize FAAH activity in both mouseuterus and embryos during early pregnancy. FAAH expressionwas also measured at the protein level, showing that the enzymeis localized in the endometrial epithelium, and that sex hormonesdown-regulate FAAH activity and expression in early preg-nancy. Anandamide transporter and FAAH were also demon-strated and characterized in mouse embryos, suggesting thatanandamide degradation may be instrumental in preventinganandamide-induced programmed cell death (apoptosis) ofdeveloping embryos.

M A T E R I A L S A N D M E T H O D S

Materials

Chemicals were of the purest analytical grade. Anandamide,arachidonic acid, ethanolamine, minimal essential medium andM2 culture medium, phenylmethanesulfonyl fluoride, 33258

Correspondence to A. Finazzi-AgroÁ, Department of Experimental Medicine

and Biochemical Sciences, University of Rome `Tor Vergata',

Via di Tor Vergata 135, I-00133 Rome, Italy. Fax: 1 39 06 72596468,

Tel.: 1 39 06 72596468, E-mail: FINAZZI@UNIROMA2.IT

Abbreviations: FAAH, fatty acid amide hydrolase; CB1/2-R, type 1/2

cannabinoid receptor; AACOCF3, arachidonoyltrifluoromethylketone;

AM404, N-(4-hydroxyphenyl)-arachidonoylamide.

Enzyme: anandamide hydrolase (EC 3.5.1.4).

Note: a web page is available at http://WWW.UNIROMA2.IT

(Received 19 January 2000, accepted 16 March 2000)

Hoechst, arachidonoyltrifluoromethylketone (AACOCF3) andN-(4-hydroxyphenyl)arachidonoylamide (AM404) were pur-chased from Sigma. N-Piperidino-5-(4-chlorophenyl)-1-(2,4-dichlorophenyl)-4-methyl-3-pyrazole-carboxamide (SR 141716)and N-{1(S )-endo-1,3,3-trimethylbicyclo[2.2.1] heptan-2-yl}-5-(4-chloro-3-methylphenyl)-1-(4-methylbenzyl)pyrazole-3-carbox-amide (SR 144528) were gifts from Sanofi Recherche(Montpelier, France). [3H]Anandamide (223 Ci´mmol21 �8251 GBq´mmol21) and [3H]arachidonic acid (100 Ci´mmol21

� 3700 GBq´mmol21) were purchased from NEN DuPont deNemours (Koln, Germany).

Biological material

CD-1 mice were purchased from Charles River Laboratories.Females were superovulated at 5±8 weeks of age by intraperi-toneal injection of 5 U pregnant mare's serum gonadotropin(Folligon, Intervet, Italy), followed 48 h later by 5 U humanchorionic gonadotropin (Corulon, Intervet, Italy). They werethen mated overnight, and the presence of a vaginal plug thenext morning was taken to indicate day 0.5 of pregnancy.FAAH activity and content (see below) were determined insamples of uteri obtained from nonpregnant, pregnant, pseudo-pregnant and ovariectomized pregnant mice. To obtain pseudo-pregnancy, normal females were mated with vasectomizedmales. Delayed implantation was obtained by surgically remov-ing the ovaries from pregnant mice on day 2.5 of pregnancy;ovariectomized mice were postoperatively subcutaneouslyinjected with 1 mg per 0.1 mL progesterone (Depo-provera;Upjohn, Puurs, Belgium). On day 5.5 of pregnancy somefemales were killed, and others were injected with 0.05 mg per0.1 mL 17b-estradiol (Benztrone, Amsa, Rome, Italy) toinduce implantation and killed the day after [20]. FAAHactivity and content were also evaluated in virgin femalestreated with the same concentrations of progesterone orestradiol and 3 days later uterine horns were excised andtested. Preimplantation embryos at the two-cell (1.5 days postcoitum) and blastocyst (3.5 days post coitum) stages werecollected by flushing the oviducts and uterus with M2 medium.Embryos were cultured in 0.5 mL minimal essential mediumsupplemented with 4 mg´mL21 crystallized BSA (Bayer,Milan, Italy) in four-well Nunclon dishes, in 5% CO2 at37 8C. All procedures adopted were approved by the AnimalCare Committee of University of Rome `Tor Vergata'. Humantissues were obtained from uteri of three normally cyclingpremenopausal women who underwent endometrial biopsy aspart of their infertility evaluation. These patients gave informedconsent to the study. Biopsy samples were taken using a pipetteendometrial suction cuvette (Veriurar Inc., Wilton, Wilts., UK),as described [21]. Tissues were removed and donated byM. Sbragia (Endocrinology and Reproduction Medical Center,Rome, Italy). The adenocarcinoma and the surroundingendometrial tissue (0.1 g each fresh tissue) were homogenizedand used to compare FAAH activity and expression in tumorand healthy uterine epithelial cells.

Assay of FAAH

Mouse uterus, blastocysts and human specimens were washedin NaCl/Pi, homogenized and assayed for FAAH activity byRP-HPLC [19]. FAAH activity was expressed as pmol arachi-donate released´min21´(mg protein)21. The apparent Km andVmax of anandamide hydrolysis by FAAH were calculated byLineweaver±Burk analysis, using [3H]anandamide in theconcentration range 0±25 mm. Anandamide synthase activitywas measured by following the formation of [3H]anandamide

from [3H]arachidonic acid and ethanolamine, as reported[17,22], and was expressed as pmol anandamide formed´min21´(mg protein)21.

Immunochemical analysis

SDS/PAGE (12% gel) was performed under reducing condi-tions in a Mini Protean II apparatus (Bio-Rad), with 0.75-mmspacer arms [23]. Rainbow molecular-mass markers (Amer-sham International) were phosphorylase b, BSA and ovalbumin(97.4, 66.0 and 46.0 kDa, respectively). Mouse uterus homo-genates (20 mg per lane) were subjected to SDS/PAGE, thenslab gels were electroblotted on to 0.45-mm nitrocellulosefilters (Bio-Rad), using a Mini Trans Blot apparatus (Bio-Rad)as reported [23]. Immunodetection of FAAH on nitrocellulosefilters was performed with specific FAAH polyclonal antibodies(diluted 1 : 200), elicited in rabbits against the conservedFAAH sequence VGYYETDNYTMPSPAMR [24], conjugatedto ovalbumin. This peptide antigen and the FAAH polyclonalantibodies were prepared by Primm (Milan, Italy). Goatanti-rabbit alkaline phosphatase conjugate (Bio-Rad), diluted1 : 2000, was used as secondary antibody, and immunoreactivebands were stained with the alkaline phosphatase stainingsolution according to the manufacturer's instructions (Bio-Rad).ELISA was performed by coating the plate with mouse uterushomogenates (20 mg per well), allowed to react with FAAHpolyclonal antibodies (diluted 1 : 300), and then with goatanti-rabbit alkaline phosphatase conjugate (diluted 1 : 2000).Color development of the alkaline phosphatase reaction wasmeasured at 405 nm, using p-nitrophenyl phosphate as sub-strate. For peptide competition experiments, the peptide antigenwas preincubated with a 1000-fold molar excess of FAAHpolyclonal antibodies for 30 min at room temperature, beforethe addition of the antibodies to the wells [17]. Controls werecarried out by using non-immune rabbit serum and includedwells coated with different amounts of BSA. For immuno-histochemistry, uterine horns were dissected from pregnant(5.5 days post coitum) and nonpregnant mice and fixed in 4%paraformaldehyde overnight. Tissues were repeatedly washedin NaCl/Pi containing 10 mg´mL21 BSA, kept overnight inNaCl/Pi containing 30% sucrose, incorporated in OCT com-pound (Sakura Finetek Inc., Torrance, CA, USA) and frozen inliquid nitrogen. Cryosections (8±10 mm) were collected onpoly-l-lysine-coated slides and incubated with 1 : 200 dilutedFAAH polyclonal antibodies for 2 h at 37 8C. Samples werewashed and labeled with 1 : 400 diluted Cy3-conjugated goatanti-(rabbit IgG) Ig (Chemicon International Inc., Ternecula,CA, USA) for 30 min at room temperature. Negative controlsincluded sections in which the primary antibody was replacedby non-immune rabbit serum.

Determination of anandamide uptake by blastocysts

The uptake of [3H]anandamide (223 Ci´mmol21) by mouseblastocysts (3.5 days post coitum) was studied essentially asdescribed previously [17]. Blastocysts (16/test) were collectedin M2 medium, washed in NaCl/Pi and transferred to 100 mLdrops of the same solution. They were then incubated for30 min at 37 8C with different amounts of [3H]anandamide, inthe range 0±1000 nm. Blastocysts were washed five times in100 mL drops of NaCl/Pi containing 10 mg´mL21 BSA andfinally resuspended in 200 mL NaCl/Pi. Membrane lipids wereextracted in methanol/chloroform (2 : 1, v/v) as reported [19],resuspended in 0.5 mL methanol, mixed with 3.5 mL Sigma-Fluor liquid-scintillation cocktail for non-aqueous samples(Sigma), and radioactivity was measured in an LKB 1214

2992 M. Maccarrone et al. (Eur. J. Biochem. 267) q FEBS 2000

Rackbeta scintillation counter (LKB, Stockholm, Sweden). Todistinguish non-protein-mediated from protein-mediated trans-port of anandamide into cell membranes, control experimentswere carried out at 4 8C [17]. The apparent Km and Vmax of theuptake were calculated by Lineweaver±Burk analysis (in thiscase, the uptake at 4 8C was subtracted from that at 37 8C). TheQ10 value was calculated as the ratio of anandamide uptake at30 8C and 20 8C [16]. Anandamide uptake was expressedas pmol´min21´(mg protein)21, and the effect of AM404determined by adding it directly to the incubation medium.

Evaluation of embryo development and apoptosis

Groups of 10±20 embryos were cultured in the presence of theindicated concentrations of freshly prepared anandamide, for48 h (blastocyst) or 72 h (two-cells). Embryo development wasmonitored and recorded every 12 h. At the end of the culturetime, the embryos were fixed in 4% paraformaldehyde andstained for 5 min with 33258 Hoechst (0.5 mg´mL21) todetermine the presence of apoptotic nuclei [25].

Statistical analysis

Data are reported as mean ^ SD from at least three inde-pendent determinations, each performed in duplicate. Statisticalanalysis was performed by Student's t-test, corroborated by thenon-parametric Mann±Whitney test. Experimental data wereelaborated using the instat program (GraphPad Software), anddifferences with P , 0.05 were considered significant.

R E S U LT S

FAAH activity and expression in mouse uterus during earlypregnancy

Pilot experiments indicated that mouse uterus FAAH activitywas linearly dependent on the amount of tissue homogenate (inthe range 0±30 mg protein) and the incubation time (in therange 0±30 min), whereas it depended on anandamide concen-tration according to Michaelis±Menten kinetics (Fig. 1A anddata not shown). It exhibited an apparent Km of 7.0 ^ 0.7 mmand Vmax of 320 ^ 30 pmol´min21´mg21. Its activity wasassayed in the pH range 5.0±11.0 and the temperature range20±65 8C. Under these experimental conditions, optimum pH

Fig. 1. FAAH activity and expression in

mouse uterus. (A) FAAH activity depended on

anandamide concentration. (B) Specific

antibodies to FAAH reacted with a single band

in mouse uterus homogenates (20 mg per lane).

FAAH activity (open bars) and content (hatched

bars) decreased during early pregnancy (C) or

pseudopregnancy (D). In (C,D),

100% � 160 ^ 15 pmol´min21´mg21 (activity)

or 0.330 ^ 0.030 A405 units (content). In (B),

molecular masses (kDa) are indicated on the

right.

Fig. 2. Effect of sex hormones on mouse uterus FAAH. (A) The decrease

in FAAH activity (open bars) and content (hatched bars) from day 0 to

day 5.5 of pregnancy was smaller in ovariectomized mice (ovari.) and

larger in the same animals treated with estrogen (ovari. 1 estro.),

100% � 25 ^ 3 pmol´min21´mg21 (activity) or 0.130 ^ 0.014 A405

units (content) or semi-ovariectomized (Emi-ovari). (B) Treatment of

virgin females with sex hormones decreased FAAH specific

activity (open bars) and total activity (hatched bars).

100% � 180 ^ 20 pmol´min21´mg21 (specific activity) or

35 000 ^ 4000 pmol´min21´mg21 (total activity). In (A), *P , 0.01

compared with 5.5 day control (n � 8), **P , 0.01 compared with

ovariectomized (n � 8). In (B), *P , 0.01 compared with control (n � 8).

q FEBS 2000 Anandamide hydrolase in pregnancy and embryo development (Eur. J. Biochem. 267) 2993

and temperature were 9.0 and 37 8C, respectively. Westernblotting showed that FAAH polyclonal antibodies specificallyrecognized a single immunoreactive band in uterus homo-genates, corresponding to a molecular mass of < 67 kDa (Fig.1B). The FAAH polyclonal antibodies were used to quantitateFAAH expression in mouse uterus by ELISA, validated byantigen competition experiments [17]. It was found that FAAHactivity and expression decreased during early pregnancy in atime-dependent manner, down to 10% (activity) and 40%(content) of the control after 5.5 days (Fig. 1C). The obser-vation that FAAH activity decreased much more than enzymecontent might be explained by the formation of a solubleinhibitor of FAAH. Kinetic analysis showed that FAAH activitytowards anandamide had apparent Vmax values of 320 ^ 30,280 ^ 30, 150 ^ 15, and 50 ^ 5 pmol´min21´mg21, in homo-genates of mouse uterus at 0, 1.5, 3.5, and 5.5 days of preg-nancy, whereas apparent Km was unchanged (7.0 ^ 0.7 mm inall cases).

FAAH activity and expression during pseudopregnancy anddelayed implantation

FAAH protein decreased in the pseudopregnant uterus follow-ing the same trend as observed during normal pregnancy(Fig. 1D), while enzyme activity decreased but to a signifi-cantly (P , 0.05, n � 8) slower rate (45% at day 5.5 in thepseudopregnant uterus vs. 10% in the pregnant uterus). Wefound that the fall in FAAH activity and expression wassignificantly (P , 0.01, n � 8) lower in ovariectomizedanimals than controls, and that estrogen treatment reversedthis effect (Fig. 2A). On the other hand, a decrease inFAAH activity and expression similar to that observed innormal pregnancy was present in sham-operated or semi-ovariectomized mice (Fig. 2A). Progesterone and estrogensignificantly (P , 0.01, n � 8) decreased both specific andtotal activity of FAAH compared with untreated controls(Fig. 2B). Moreover, the decreased FAAH activity wasparalleled by decreased enzyme content, which leveled offat 60% of the untreated control (P , 0.05, n � 8) on treatmentof mouse uterus with either hormone (not shown). Anandamidesynthase activity was also present in the uterus of virginfemales (40 ^ 4 pmol´min21´mg21), which was approximatelyhalved by progesterone (20 ^ 2 pmol´min21´mg21) orestrogen (25 ^ 3 pmol´min21´ mg21).

FAAH localization in mouse uterus and FAAH activity inhuman uterus

Antibodies to FAAH were used to immunolocalize FAAHin the luminal (Fig. 3A) and glandular (Fig. 3B) epithelia ofnonpregnant mouse uterus. In keeping with the activity andELISA data, FAAH was hardly detectable in the uterus of5.5-day-pregnant mice (not shown).

FAAH activity was also analyzed in human uterine epithelialcells, where it showed an apparent Km and a Vmax for anand-amide of 4.0 ^ 0.5 mm and 1000 ^ 100 pmol´min21´mg21,respectively. Remarkably, FAAH activity in human adeno-carcinoma cells was more than five times higher than in healthycontrols (Fig. 4). Quantitation of FAAH by ELISA, validatedby antigen competition experiments, showed that FAAHexpression paralleled the activity (Fig. 4).

Fig. 3. Immunolocalization of FAAH and

anandamide-induced apoptosis. FAAH was

immunolocalized in the luminal (A) and

glandular (B) epithelia of nonpregnant mouse

uterus. Control mouse blastocysts (C) showed

increased numbers of apoptotic cells on

treatment with 40 nm anandamide (D). Original

magnification: (A) (B) � �350; (C,D) � �175.

Bar represents 10 mm (A,B) or 5 mm (C,D).

Arrowheads indicate autofluorescence of

unidentified cells (A,B) or apoptotic cells (D).

Fig. 4. FAAH activity and expression in human uterus. FAAH activity

(open bars) and content (hatched bars) was significantly increased in

adenocarcinoma cells compared with healthy epithelial cells. Antigen

competition ELISA (closed bars) validated the specificity of FAAH

quantitation. *P , 0.01 compared with control (n � 3). 100% � 600 ^

50 pmol´min21´mg21 (activity) or 0.500 ^ 0.050 A405 units (content).

2994 M. Maccarrone et al. (Eur. J. Biochem. 267) q FEBS 2000

FAAH activity and anandamide uptake by mouse blastocysts

FAAH activity showed Michaelis±Menten kinetics with appa-rent Km and Vmax towards anandamide of 4.0 ^ 0.4 mm and200 ^ 20 pmol´min21´mg21, respectively (Fig. 5A). Specificinhibitors of FAAH activity, such as phenylmethanesulfonylfluoride [14,17] and AACOCF3 [26], almost completelyinhibited the hydrolysis of anandamide by mouse blastocysts(10% and 5% residual activity), when used at 100 mm and10 mm, respectively. Blastocysts also showed anandamidesynthase activity (25 ^ 3 pmol´min21´mg21), which was reducedto 15% and 10% of the control by 100 mm phenylmethane-sulfonyl fluoride and 10 mm AACOCF3, respectively.

Blastocysts were able to accumulate [3H]anandamide ina temperature-dependent (Q10 � 1.5), time-dependent(t1/2 � 5 min) and concentration-dependent way (Fig. 5Band data not shown). [3H]Anandamide uptake wassaturable (apparent Km � 160 ^ 20 nm andVmax � 0.32 ^ 0.04 pmol´min21´ mg21, respectively) andwas almost completely inhibited by 10 mm AM404 (Fig. 5B),a selective inhibitor of the anandamide transporter [15,27].

Impairment of hatching and development of mouseblastocysts and induction of apoptosis by anandamide

Nanomolar concentrations of anandamide inhibited develop-ment of two-cell embryos to blastocysts and zona-hatchingof blastocysts developed from eight-cell embryos in vitro(Fig. 5C). Both detrimental effects of 40 nm anandamide couldbe reversed by equimolar concentrations of SR 141716, aCB1-R antagonist, but not by the CB2-R antagonist SR 144528(up to 1 mm). At the same concentration, anandamide wasfound to induce programmed cell death in mouse blastocysts,40 nm anandamide yielding a fourfold increase in apoptoticcells as compared with the untreated control (Figs 5D and3C,D). Anandamide-induced apoptosis could not be preventedby coadministration of either SR 141716 or SR 144528 (up to1 mm).

D I S C U S S I O N

In this study we characterized FAAH activity and expression inboth mouse uterus and embryos during early pregnancy. FAAH

was localized in the endometrial epithelium and was down-regulated by sex hormones. Under our experimental conditions,the optimum pH and temperature of mouse uterus FAAH weredetermined, and the kinetic constants of enzyme activity (Fig.1A) were calculated. Remarkably, the observed apparent Km

and Vmax were one order of magnitude smaller than thosecalculated with column filtration procedures [9], far fromthe resolution and reproducibility of HPLC. Moreover, theobserved molecular mass of FAAH (Fig. 1B) was in agreementwith that expected from the size of FAAH mRNA in mouseuterus [18]. Activity and expression of FAAH decreased duringearly pregnancy (Fig. 1C), an observation that, together withkinetic analysis of the enzyme at the different stages ofpregnancy, suggests lower expression of the same gene, ratherthan the presence of FAAH isozymes with different kineticproperties. Recent observations on FAAH mRNA accumulationin pregnant mice corroborate this concept [18].

Despite the growing evidence that anandamide adverselyaffects uterine receptivity and embryo implantation [7,10,12,13]and that anandamide degradation by FAAH may have physio-logical significance in these processes [9,18], regulation ofFAAH during early pregnancy is still obscure. In this study, weused two different methods of manipulating pregnancy-relateduterine changes (pseudopregnancy and ovariectomy-induceddelayed implantation) as tools to understand the relative rolesplayed by the embryo and sex hormones in modulating theabove changes in FAAH activity and expression in the pregnantmouse. Pseudopregnancy occurs in female mice and rats matedto a sterile male: the uterus undergoes all the normal changesthat prepare it for implantation, but no embryos are present inthe uterine lumen [28]. Therefore, down-regulation of FAAHexpression in pseudopregnant mice (Fig. 1D) was independentof the presence of embryos in the uterus. However, the decreasein FAAH activity was more marked in the presence of embryo(compare Fig. 1C and D), suggesting that the latter mayrelease an enzyme inhibitor, the molecular identity of which isas yet unknown. On the other hand, delayed implantation is aphenomenon that occurs when a pregnant female mouse isovariectomized before implantation: the preparation of theuterus for implantation is arrested and the blastocyst remains ina state of suspended development [29]. Hormonal administra-tion is rapidly followed by uterine changes that make itreceptive to the implanting blastocyst. The observation that the

Fig. 5. FAAH activity and anandamide

transporter in mouse blastocysts. FAAH

activity (A) and anandamide uptake (B)

depended on anandamide concentration.

In (B), anandamide uptake by blastocysts was

measured at 37 8C, in the absence (X) or

presence (asterisk) of 10 mm AM404, or at 4 8C

(O). (C) Anandamide impaired embryo

development from two cells to blastocyst

(open bars) and blastocyst hatching (hatched

bars). (D) Anandamide induced programmed

cell death in mouse blastocysts.

q FEBS 2000 Anandamide hydrolase in pregnancy and embryo development (Eur. J. Biochem. 267) 2995

fall in FAAH activity and expression was significantly lower inovariectomized animals than controls, and that estrogen treat-ment reversed this effect (Fig. 2A), suggests that sex hormonesmay regulate FAAH activity by modulating gene expression atthe translational level. The results of the treatment of virginfemales with progesterone or estrogen (Fig. 2B) corroboratedthis hypothesis. It should be stressed that hormone treatmentmay increase protein synthesis in the uterus [30,31]. Thus, achange in FAAH specific activity (i.e. activity per mg protein)may be seen in hormone-treated animals rather than a truedecrease. To check this possibility, we also measured the totalactivity of FAAH (i.e. activity per uterine horn). In Fig. 2B(hatched bars), total activity showed the same trend as specificactivity, ruling out the possibility that the observed decreasewas apparent. Therefore, it can be concluded that sex hormonesdown-regulate FAAH activity by reducing gene expression atthe level of protein synthesis. This is a demonstration of adirect interplay between endocannabinoid degradation andsex hormones in mammals. Also anandamide synthase activitywas measured in mouse uterus, and was found to respond to sexhormones in the same way as FAAH. Although it is still underdebate whether or not anandamide hydrolase and synthaseactivities belong to the same [22,32] or different [9,18]enzymes, these data demonstrate that the two activities areunder the same hormonal control.

FAAH was localized in the luminal (Fig. 3A) and glandular(Fig. 3B) epithelia of nonpregnant mouse uterus. In situhybridization consistently detected FAAH mRNA primarily inuterine luminal and glandular epithelial cells [18]. Humanuterine epithelial cells also showed appreciable FAAH activity,which was more than five times higher in human adeno-carcinoma cells (Fig. 4). These findings are consistent with anepithelial localization for FAAH in the human endometriumalso. In this context, it is noteworthy that the Km values forFAAH from mouse and human uterus were comparable withthose recently reported for human brain, human neuroblastomaand lymphoma cell lines, whereas Vmax values varied [17].Therefore, it can be proposed that the same enzyme isdifferently expressed in various species and in different tissuesof the same species. Sequence homology between rat, mouseand human FAAH genes [24] suggests that the FAAH gene isindeed highly conserved. Therefore, the observations reportedhere on the hormonal regulation of FAAH in mouse uterus mayalso hold true for the human counterpart.

It can be proposed that down-regulation of FAAH duringearly pregnancy may allow higher local levels of anandamide,which indeed have been shown to increase with pregnancy inthe mouse uterus [10]. In turn, anandamide may play a role inthe endometrial changes associated with pregnancy, forinstance through inhibition of gap junctions and intercellularcalcium signaling [33,34]. Interestingly, anandamide has beenrecently reported to mobilize intracellular calcium [35], thusalso perturbing intracellular signaling.

FAAH activity was demonstrated and characterized in mouseblastocysts (Fig. 5A). To be hydrolyzed by FAAH, anandamidemust be transported into the cell. Recent experiments per-formed on rat neuronal and leukemia cells [15] and humanneuronal and immune cells [17] clearly showed the presence ofa high-affinity anandamide transporter in the cell outer mem-branes. A similar transporter was found in mouse blastocysts(Fig. 5B). The affinity of this transporter was comparable withthat of the anandamide carrier in rat astrocytes (Km � 320 nm)[15] and human cells (Km � 130±200 nm) [17]. The anand-amide transporter and FAAH of blastocysts may play a criticalrole, because nanomolar concentrations of anandamide inhibited

embryo development and blastocysts hatching in vitro (Fig. 5C).Both detrimental effects of anandamide were inhibited by aCB1-R antagonist, in line with the hypothesis that they weremediated by the type-1 cannabinoid receptor [10]. Interestingly,anandamide-induced apoptosis (Fig. 5D) was not prevented byCB1-R or CB2-R antagonists, ruling out the involvement ofeither cannabinoid receptor in the induction of programmed celldeath. Therefore, the arrest of embryo development and blasto-cyst hatching by anandamide did not involve the deploymentof apoptotic programs [36,37]. As yet, only one report hasdemonstrated the ability of anandamide to inhibit cancer cellproliferation [38], and another gave preliminary evidence on itsability to induce apoptosis in lymphocytes [39]. It is noteworthythat D9-tetrahydrocannabinol has been recently shown to pro-mote apoptosis in glioma cells, through a CB1-R independentmechanism [40].

Collectively, our findings lead to a general picture suggestingthat decreased FAAH activity in the mouse uterus during earlypregnancy may allow higher levels of anandamide to accumu-late, which may be instrumental in modifying the endometriumduring pregnancy. The detrimental effects of anandamide onthe blastocysts are prevented by the presence of an efficientanandamide transporter and FAAH in these cells, which rapidlyscavenge this lipid. These events are under hormonal control,showing an interplay between endocannabinoids and sexhormones in regulating fertility in mammals.

A C K N O W L E D G E M E N T S

The authors wish to thank Dr M. Sbragia (Endocrinology and Reproductive

Medical Center, Rome, Italy) for kindly donating human specimens, and

Drs M. Bari and R. Agostinetto for their skilful assistance. This

investigation was supported by Istituto Superiore di SanitaÁ (II AIDS

Programme) and Ministero dell'UniversitaÁ e della Ricerca Scientifica e

Tecnologica (MURST-PRIN 1997 and 1998).

R E F E R E N C E S

1. Devane, W.A., Hannus, L., Breuer, A., Pertwee, R.G., Stevenson, L.A.,

Griffin, G., Gibson, D., Mandelbaum, A., Etinger, A. & Mechoulam,

R. (1992) Isolation and structure of a brain constituent that binds to

the cannabinoid receptor. Science 258, 1946±1949.

2. Mechoulam, R., Fride, E. & Di Marzo, V. (1998) Endocannabinoids.

Eur. J. Pharmacol. 359, 1±18.

3. di Tomaso, E., Beltramo, M. & Piomelli, D. (1996) Brain cannabinoids

in chocolate. Nature (London) 382, 677±678.

4. Di Marzo, V., Sepe, N., De Petrocellis, L., Berger, A., Crozier, G.,

Fride, E. & Mechoulam, R. (1998) Trick or treat from food

endocannabinoids? Nature (London) 396, 636.

5. Calignano, A., La Rana, G., Giuffrida, A. & Piomelli, D. (1998)

Control of pain initiation by endogenous cannabinoids. Nature

(London) 394, 277±281.

6. Meng, I.D., Manning, B.H., Martin, W.J. & Fields, H.L. (1998) An

analgesia circuit activated by cannabinoids. Nature (London) 395,

381±383.

7. Das, S.K., Paria, B.C., Chakraborty, I. & Dey, S.K. (1995) Cannabinoid

ligand-receptor signaling in the mouse uterus. Proc. Natl Acad. Sci.

USA 92, 4332±4336.

8. Ness, R.B., Grisso, J.A., Hirschinger, N., Markovic, N., Shaw, L.M.,

Day, N.L. & Kline, J. (1999) Cocaine and tobacco use and the risk of

spontaneous abortion. N. Engl. J. Med. 340, 333±339.

9. Paria, B.C., Deutsch, D.D. & Dey, S.K. (1996) The uterus is a potential

site for anandamide synthesis and hydrolysis. differential profiles of

anandamide synthase and hydrolase activities in the mouse uterus

during the periimplantation period. Mol. Rep. Dev. 45, 183±192.

10. Schmid, P.C., Paria, B.C., Krebsbach, R.J., Schmid, H.H.O. & Dey,

S.K. (1997) Changes in anandamide levels in mouse uterus are

2996 M. Maccarrone et al. (Eur. J. Biochem. 267) q FEBS 2000

associated with uterine receptivity for embryo implantation. Proc.

Natl Acad. Sci. USA 94, 4188±4192.

11. Di Marzo, V. (1998) `̀ Endocannabinoids'' and other fatty acid deriv-

atives with cannabimimetic properties. biochemistry and possible

physiopathological relevance. Biochim. Biophys. Acta 1392, 153±175.

12. Yang, Z.-M., Paria, B.C. & Dey, S.K. (1996) Activation of brain-type

cannabinoid receptors interferes with preimplantation mouse embryo

development. Biol. Reprod. 55, 756±761.

13. Wang, J., Paria, B.C., Dey, S.K. & Armant, D.R. (1999) Stage-specific

excitation of cannabinoid receptor exhibits differential effects on

mouse embryonic development. Biol. Reprod. 60, 839±844.

14. Di Marzo, V., Bisogno, T., De Petrocellis, L., Melck, D., Orlando, P.,

Wagner, J.A. & Kunos, G. (1999) Biosynthesis and inactivation of the

endocannabinoid 2-arachido-noylglycerol in circulating and tumoral

macrophages. Eur. J. Biochem. 264, 258±267.

15. Beltramo, M., Stella, N., Calignano, A., Lin, S.Y., Makriyannis, A. &

Piomelli, D. (1997) Functional role of high-affinity anandamide

transport, as revealed by selective inhibition. Science 277, 1094±1097.

16. Hillard, C.J., Edgemond, W.S., Jarrahian, A. & Campbell, W.B. (1997)

Accumulation of N-arachidonoylethanolamide (anandamide) into

cerebellar granule cells occurs via facilitated diffusion. J. Neurochem.

69, 631±638.

17. Maccarrone, M., van der Stelt, M., Rossi, A., Veldink, G.A.,

Vliegenthart, J.F.G. & Finazzi-AgroÁ, A. (1998) Anandamide hydro-

lysis by human cells in culture and brain. J. Biol. Chem. 273,

32332±32339.

18. Paria, B.C., Zhao, X., Wang, J., Das, S.K. & Dey, S.K. (1999) Fatty-

acid amide hydrolase is expressed in the mouse uterus and embryo

during the periimplantation period. Biol. Reprod. 60, 1151±1157.

19. Maccarrone, M., Bari, M. & Finazzi-AgroÁ, A. (1999) A sensitive and

specific radiochromatographic assay of fatty acid amide hydrolase.

Anal. Biochem. 267, 314±318.

20. Bergstrom, S. (1978) Experimentally delayed implantation. In Methods

in Mammalian Reproduction (Daniel, J.C., ed.), pp. 419±435.

Academic Press, New York.

21. Hill, G.A., Herbert, C.M., Parker, R.A. & Werz, A.C. (1989)

Comparison of late luteal phase endometria utilizing the Nowak

cuvette or pipette endometrial suction cuvette. Obstet. Gynecol. 73,

443±450.

22. Ueda, N., Kurahashi, Y., Yamamoto, S. & Tokunaga, T.J. (1995) Partial

purification and characterization of the porcine brain enzyme

hydrolyzing and synthesizing anandamide. J. Biol. Chem. 270,

23823±23827.

23. Maccarrone, M., Veldink, G.A. & Vliegenthart, J.F.G. (1991) An

investigation on the quinoprotein nature of some fungal and plant

oxidoreductases. J. Biol. Chem. 266, 21014±21017.

24. Giang, D.K. & Cravatt, B.F. (1997) Molecular characterization of

human and mouse fatty acid amide hydrolase. Proc. Natl Acad. Sci.

USA 94, 2238±2242.

25. Brison, D.R. & Schultz, R.M. (1997) Apoptosis during mouse

blastocyst formation. evidence for a role for survival factors including

transforming growth factor a. Biol. Reprod. 56, 1088±1096.

26. Koutek, B., Prestwich, G.D., Howlett, A.C., Chin, S.A., Salehani, D.,

Akhavan, N. & Deutsch, D.G. (1994) Inhibitors of arachidonoyl

ethanolamide hydrolysis. J. Biol. Chem. 269, 22937±22940.

27. Piomelli, D., Beltramo, M., Glasnapp, S., Lin, S., Goutopoulos, A., Xie,

X.Q. & Makriyannis, A. (1999) Structural determinants for recog-

nition and translocation by the anandamide transporter. Proc. Natl

Acad. Sci. USA 96, 5802±5807.

28. Everett, J.W. (1969) Neuroendocrine aspects of mammalian repro-

duction. Annu. Rev. Physiol. 31, 383±416.

29. Flint, A.P.F., Renfree, M.B. & Weir, B.J. (1980) Embryonic diapause in

mammals. J. Reprod. Fertil. S29.

30. Coulson, P.B. & Pavlik, E.J. (1977) Effects of estrogen and

progesterone on cytoplasmic estrogen receptor and rates of protein

synthesis in rat uterus. J. Steroid Biochem. 8, 205±212.

31. DeSombre, E.R. & Kuivanen, P.C. (1985) Progestin modulation of

estrogen-dependent marker protein synthesis in the endometrium.

Semin. Oncol. 12, 6±11.

32. Kurahashi, Y., Ueda, N., Suzuki, H., Suzuki, M. & Yamamoto, S.

(1997) Reversible hydrolysis and synthesis of anandamide demon-

strated by recombinant rat fatty acid amide hydrolase. Biochem.

Biophys. Res. Commun. 237, 512±515.

33. Venance, L., Piomelli, D., Glowinski, J. & Giaume, C. (1995)

Inhibition by anandamide of gap junctions and inercellular calcium

signalling in striatal astrocytes. Nature (London) 376, 590±594.

34. Boger, D.L., Patterson, J.E., Guan, X., Cravatt, B.F., Lerner, R.A. &

Gilula, N.B. (1998) Chemical requirements for inhibition of gap

junction communication by the biologically active lipid oleamide.

Proc. Natl Acad. Sci. USA 95, 4810±4815.

35. Maccarrone, M., Bari, M., Menichelli, A., Del Principe, D. & Finazzi-

AgroÁ, A. (1999) Anandamide activates human platelets through a

pathway independent of the arachidonate cascade. FEBS Lett. 447,

277±282.

36. Afford, S.C., Randhawa, S., Eliopoulos, A.G., Hubscher, S.G., Young,

L.S. & Adams, D.H. (1999) CD40 activation induces apoptosis in

cultured human hepatocytes via induction of cell surface fas ligand

expression and amplifies fas-mediated hepatocyte death during

allograft rejection. J. Exp. Med. 189, 441±446.

37. Tonnetti, L., Veri, M.C., Bonvini, E. & D'Adamio, L. (1999) A role for

neutral sphingomyelinase-mediated ceramide production in T cell

receptor-induced apoptosis and mitogen-activated protein kinase-

mediated signal transduction. J. Exp. Med. 189, 1581±1589.

38. De Petrocellis, L., Melck, D., Palmisano, A., Bisogno, T., Laezza, C.,

Bifulco, M. & Di Marzo, V. (1998) The endogenous cannabinoid

anandamide inhibits human breast cancer cell proliferation. Proc.

Natl Acad. Sci. USA 95, 8375±8380.

39. Schwarz, H., Blanco, F.J. & Lotz, M. (1994) Anandamide, an

endogenous cannabinoid receptor agonist, inhibits lymphocyte

proliferation and induces apoptosis. J. Neuroimmunol. 55,

107±115.

40. SaÁnchez, C., Galve-Roperh, I., Canova, C., Brachet, P. & Guzman, M.

(1998) D9-Tetrahydrocannabinol induces apoptosis in C6 glioma

cells. FEBS Lett. 436, 6±10.

q FEBS 2000 Anandamide hydrolase in pregnancy and embryo development (Eur. J. Biochem. 267) 2997