N-Acylethanolamines in human reproductive fluids

N -Acylethanolamines (NAEs) are an important family of lipid-signaling molecules. Arachidonylethanolamide (anandamide) (AEA), palmitoylethanolamide (PEA), and oleoylethanolamide (OEA) are co-produced from similar phospholipid precursors when neurons are stimulated. AEA is an endogenous agonist (endocannabinoid) for cannabinoid receptors. It binds with higher affinity to type CB1 than to type CB2 cannabinoid receptors. PEA does not bind to CB1, while the hypothesis that it reacts with putati v e CB2-like receptors has been questioned. OEA does not acti v ate currently known cannabinoid receptors, but it mimics the effects of AEA and cannabinoids in reducing the fertilizing capacity of sea urchin sperm. OEA and PEA also act as entourage compounds by inhibiting the hydrolysis of AEA by fatty acid amide hydrolase. Cannabinoid receptors and/or AEA are present in mammalian reproducti v e organs including the testis, epididymis, prostate, o v ary, uterus, sperm, preimplantation embryo and placenta, as well as prostatic and mammary carcinomas. We now report that analysis by high-performance liquid chromatography/mass spectrometry (HPLC/MS) shows the presence of AEA, PEA, and OEA in human seminal plasma, mid-cycle o v iductal fluid, follicular fluid, amniotic fluid, milk, and fluids from malignant o v arian cysts. Pre v ious studies showed that AEA-signaling v ia cannabinoid receptors regulates capacitation and fertilizing potential of human sperm, early embryonic de v elopment and blastocyst implantation into the uterine mucosa of rodents, as well as proliferation of human mammary and prostatic carcinomas. Current results imply that NAEs also may modulate follicular maturation and o v ulation, normal and pathological o v arian function, placental and fetal physiology, lactation, infant physiology, and beha v ior. Collecti v ely, these findings suggest that NAEs in human reproducti v e fluids may help regulate multiple physiological and pathological processes in the reproducti v e system, and imply that exogenous cannabinoids deli v ered by marijuana smoke might impact these processes. This study has potential medical and public policy ramifications

Here we present evidence for the presence of AEA, PEA, and OEA in human seminal plasma, mid-cycle oviductal fluid, follicular fluid, amniotic fluid, milk, and fluids from malignant ovarian cysts.Preliminary accounts of this study have been reported previously (Giuffrida et al., 2001a,b).

Materials and methods
Protocols involving human subjects were performed in accordance with policies of the Institutional Review Board of the University at Buffalo and the University of California, Irvine.

Biological materials
Human semen samples were collected from healthy volunteer donors of proven fertility.Liquefied semen was centrifuged at 300)/g for 5 min.The supernatant (seminal plasma) was collected by aspiration.Mid-cycle human oviductal fluid was obtained from volunteers during elective surgery for tubal ligation (Lippes et al., 1972).Oviductal fluids were also collected from rabbits and goats at the time of ovulation.Human follicular fluid was retained following oocyte aspiration from women undergoing in vitro fertilization.Amniotic fluid was retained from women undergoing elective removal of conceptus.Fluids from malignant ovarian cysts were retained from women undergoing elective ovariectomy.Milk was collected from volunteer nursing mothers using breast pumps.Frozen samples were stored at (/ 20 8C.

High-performance liquid chromatography/ mass spectrometry
The presence of NAEs in reproductive fluids was determined by high-performance liquid chromatography/mass spectrometry (HPLC/MS; Giuffrida et al., 2000).Thawed samples were centrifuged at 16,000 )/g for 10 min at room temperature.The supernatants were recovered and spiked with [ 2 H 4 ]-labeled AEA, PEA, and OEA (500 pmol each) to act as internal standards.Proteins were precipitated with cold acetone and removed by centrifugation at 1000 )/g for 10 min.Residual acetone in supernatants was evaporated under a stream of nitrogen.Lipids contained in the supernatants were extracted with methanol/chloroform (1:2, v/v).Chloroform phases were recovered, evaporated to dryness under N 2 , reconstituted in a mixture of chloroform/methanol (1:3, 80 ml) and injected into a HPLC/MS (HP 1100 Series, Hewlett Packard) equipped with a Hewlett Packard octadecyl-silica Hypersil column (100 mm )/4.6 mm i.d., 5 mm).Fractionation of NAEs was carried out in reversed-phase using a gradient of methanol (B) in water (A) (25% A, 75% B for 2 min; 15% A, 85% B for 3 min; 5% A, 95% B for 20 min; 100% B for 5 min).Analyses were performed with an electrospray ion source set in the positive ionization mode.Quantitation of NAEs was carried out by isotope dilution, monitoring diagnostic ions (protonated molecule, [M'/H] ' , and sodium adducts of the molecular ions, [M'/Na] ' ) in the selected ion monitoring mode.

Statistical analysis
Significance of data was evaluated by Student's t-test (Snedecor and Cochran, 1980).Data are presented as mean values9/S.E.M. for N number of trials.

Reproductive tract fluids
Freshly ejaculated sperm from human or other mammals, bathed in male secretions comprising the seminal plasma, are not yet capable of fertilizing an egg (Yanagimachi, 1994).Sperm acquire the capacity to fertilize eggs following removal from seminal plasma and exposure for several hours to female reproductive tract fluids in vivo, or by incubation in appropriate culture medium in vitro (Burkman, 1990(Burkman, , 1995;;Kopf et al., 1999;Meizel, 1985Meizel, , 1997;;Yanagimachi, 1994).''Capacitated'' sperm exhibit vigorous hyperactivated motility required for fertilization, and can undergo physiological acrosome reactions at the egg's surface (Burkman, 1990(Burkman, , 1995;;Wassarman, 1987;Yanagimachi, 1994).The acrosome is a secretory granule in the anterior region of the sperm head.Exocytosis of the acrosomal granule (acrosome reaction) is a ligand-stimulated event that enables sperm to penetrate the egg's investments (cumulus matrix and zona pellucida), and to fuse with the egg's plasma membrane during fertilization (Wassarman, 1987;Yanagimachi, 1994;Darszon et al., 1999).Mechanisms that modulate these processes within the female reproductive tract are poorly understood.
Transcripts of human brain CB1 are expressed in human testis (Gerard et al., 1991;Galiegue et al., 1995).This finding, together with our discovery of functional cannabinoid receptors in sea urchin sperm, led us to postulate that human sperm might contain cannabinoid receptors (Chang et al., 1993;Schuel et al., 1994).Subsequent studies on whole organs showed that AEA is synthesized in rodent testis, oviduct, and uterus (Paria et al., 1995;Sugiura et al., 1996;Schmid et al., 1997;Paria and Dey, 2000).The mouse uterus contains higher levels of AEA than any mammalian organ including the brain (Schmid et al., 1997).Recent radioligand binding experiments using the potent cannabinoid agonist [ 3 H]CP-55,940 (K D 9.719/1.04nM) confirmed our hypothesis by showing that human sperm contain functional cannabinoid receptors similar to those in sea urchin sperm and mammalian somatic tissues (Schuel et al., 2002).Together, these findings suggested that NAEs may be normal physiological constituents of fluids within human reproductive tracts.
We detected AEA, OEA, and PEA in human seminal plasma, mid-cycle oviductal fluid, and follicular fluid analyzed by HPLC/MS (Table 1).The levels of PEA and OEA were significantly higher than AEA in seminal plasma, oviductal fluid, and follicular fluids (P B/0.005 to B/0.0001).OEA was the most abundant NAE in human follicular fluid (P B/0.005).It is interesting to note that these results are very different from those obtained with the whole mouse uterus, where AEA represents up to 95% of total NAEs (Schmid et al., 1997).Furthermore, AEA, PEA, and OEA were also detected in oviductal fluids collected from rabbits and goats at the time of ovulation (data not shown).NAEs in human follicular fluid may be produced by granulosa in ovarian follicules and by granulosa cells in the cumulus matrix surrounding ovulated eggs, as previously shown for progesterone and prostaglandin E (Meizel et al., 1990;Schaefer et al., 1998).Since cannabinoid receptors are present in the ovary (Galiegue et al., 1995) and THC inhibits cAMP accumulation by cultured rat granulosa cells (Treinen et al., 1993), NAE-signaling may help regulate follicular maturation and development.These factors may account, in part, for the adverse effects of marijuana and cannabinoids on ovulation (Powell and Fuller, 1983).
Sperm are sequentially exposed to seminal plasma, oviductal fluid, follicular fluid, and secretions of granulosa cells in the cumulus matrix surrounding ovulated eggs as they move from the vagina to the site of fertilization in the oviductal ampulla (Burkman, 1990;Mortimer, 1995;Yanagimachi, 1994).The presence of NAEs in these locations and the detection of cannabinoid receptors in human sperm (Schuel et al., 2002) imply that NAE-signaling may regulate sperm capacita-tion and fertilizing potential within human reproductive tracts.
While the biological effects of PEA and OEA in reproductive tract fluids remain to be determined, evidence has been obtained suggesting that AEAsignaling via sperm cannabinoid receptors modulates sperm functions required for fertilization (Schuel et al., 2002).AEA is rapidly eliminated by carrier-mediated uptake into somatic cells (Beltramo et al., 1997), and hydrolyzed into arachidonic acid and ethanolamine by fatty acid ethanolamide amidase (Deutsch and Chin, 1993;Ueda et al., 2000;Giuffrida et al., 2001c).
R -Methanandamide (AM-356) is a metabolically stable substitute for AEA (Abadji et al., 1994;Khanolkar et al., 1996).The effects of AM-356 and THC on sperm capacitation and fertilizing potential were examined in vitro (Schuel et al., 2002).AM-356 elicited biphasic effects on the incidence of hyperactivated sperm motility (HA) between 1 and 6 h of culture: at 2.5 nM it inhibited HA, while at 0.25 nM it stimulated HA (P B/0.05; ANOVA).These data are consistent with results obtained by Burkman and colleagues on modulation of rabbit sperm HA in the oviduct that pointed to an unknown inhibitory substance in the isthmus which was negated upon 10-fold dilution (Johnson et al., 1981;Burkman et al., 1984).Factors in the isthmic environment provide regulatory control initially inducing sperm quiescence and, at the appropriate time, stimulating vigorous HA motility thus enabling sperm to escape and swim to the ampulla.There are regional differences in composition of fluids within the lumen of bovine and porcine oviducts (Anderson and Killian, 1994;Hunter, 1990).In addition, bovine isthmic and ampullary fluids produce different effects on motility, acrosome reaction, zona binding, and fertility of bull sperm (Grippo et al., 1995;Topper et al., 1999).
Acquisition of acrosomal competence during capacitation represents an intermediate state preparing sperm to respond to physiological stimuli at the egg's surface (Lee et al., 1987b;Kligman et al., 1991;Kim et al., 2001a,b).This process appears to be correlated with alteration of acrosomal caps of human, bovine, and rodent sperm in vitro, and/or in vivo within the isthmus, ampulla, and cumulus (Hunter et al., 1991;Kligman et al., 1991;Kim et al., 2001a,b;Lee et al., 1987a,b;Nolan et al., 1992;Yanagimachi and Phillips, 1984).Here, membrane destabilization and localized transient fusion events expose acrosomal matrix proteins required for sperm-zona binding and acrosomal exocytosis.Both AM-356 and THC inhibited morphological alterations over acrosomal caps between 2 and 6 h (IC 50 5.99/0.6 pM and 3.59/1.5nM, respectively), suggesting that they modulate this aspect of capacitation (Schuel et al., 2002).
Previous studies showed that AEA and cannabinoids reduce the fertilizing capacity of sea urchin sperm (Schuel et al., 1987(Schuel et al., , 1991(Schuel et al., , 1994;;Chang et al., 1993;Berdyshev, 1999).Ethical concerns preclude similar experiments on live human eggs.Nevertheless, fertilizing potential of human sperm can be determined on the basis of tight binding of capacitated sperm to the zona pellucida in the hemizona assay (Burkman et al., 1988;Burkman, 1995;Franken et al., 1989).This assay is highly predictive of sperm fertility during human in vitro fertilization.In this assay, one-half of a bisected zona was inseminated with sperm that had been capacitated in culture medium containing AM-356, and the matching half was inseminated with sperm incubated in medium containing vehicle to serve as an internal control.Tight binding of sperm was reduced 50% by 1 nM AM-356 under these conditions (Schuel et al., 2002).Our results provide the first evidence that endocannabinoids can directly modulate sperm capacitation and fertilization in human, and imply that exogenous cannabinoids delivered by marijuana smoke might impact these processes.
NAEs in reproductive tract fluids may also act via mechanisms that do not involve sperm cannabinoid receptors.For example, prostaglandins (derived from the seminal vesicles) regulate functions of epithelia and contractility of smooth muscle in the female reproductive tract via receptor-mediated mechanisms (Luke and Coffey, 1994).Further, semen is a potential vehicle for the introduction of infectious organisms into the female.As foreign cells, sperm are vulnerable to immunological attack in the female reproductive tract.Sperm are also very sensitive to damage by reactive oxygen radicals and lipid peroxidation (De Lamirande and Gagnon, 1999).PEA and OEA are potent anti-inflammatory, antioxidant, and antimicrobial agents (Schmid et al., 1990;Berdyshev et al., 1996;Mazzari et al., 1996;Calignano et al., 1998;Conti et al., 2002).These properties are shared with certain cannabinoids (Zimmerman et al., 1991;Hampson et al., 1998;Burstein, 1999;Cabral, 1999;Conti et al., 2002).These findings suggest that NAEs may perform multiple roles in reproductive tract fluids by modulating sperm capacitation, regulating reproductive tract function, protecting against infection, and maintaining sperm viability.

Amniotic fluid
Marijuana smoke and THC adversely affect fetal growth, development, and the number of pregnancies carried to term (Harbison et al., 1972;Maykut, 1985;Asch and Smith, 1986;Smith and Asch, 1987).Cannabinoid receptors are present in the uterus and early embryo (Paria and Dey, 2000).Localized differences in AEA levels within the uterus determine implantation sites, while exogenous cannabinoid agonists such as THC and CP-55,940 inhibit embryonic development and implantation.In situ hybridization studies showed that mRNAs for CB1 and CB2 are expressed in the placental cone and smooth muscle of the uterus in rats, rodent embryo and fetus, and term human placenta, as well as in human BeWo choricocarcinoma cells (Buckley et al., 1998;Kenney et al., 1999).Together, these observations implied that human amniotic fluid might contain endocannabinoids.
Here we show that AEA, OEA, and PEA are constituents of human amniotic fluid (Table 2).The biological functions of these NAEs in amniotic fluids remain to be determined.However, previous studies using THC and other cannabinoids suggest functional roles for cannabinoid receptors in placental physiology, embryogenesis, and fetal development.THC crosses the placenta, accumulates in chorionic and amniotic fluids, and all fetal organs (Harbison et al., 1972;Martin et al., 1977;Asch and Smith, 1986;Bailey et al., 1987).Maternal plasma contains significant amounts of 11-nor-9-carboxy THC, a major metabolite of THC, but it is not detected in amniotic fluid or the fetus (Bailey et al., 1987).These data indicate that 11-nor-9-carboxy THC does not cross the placenta, and that the fetus cannot metabolize THC.The placenta of rhesus monkeys chronically exposed to THC during pregnancy shows gross morphological abnormalities and vascular infarctions (Sassenrath et al., 1979).Cannabinoid agonists such as THC and WIN55,212-2 inhibit uptake of amino acids (valine and alpha-amino isobutyric acid) by slices of term human placenta, and inhibit the serotonin transporter in human BeWo choricocarcinoma cells (Fisher et al., 1987;Kenney et al., 1999).These observations indicate that the human placenta is a direct target for cannabinoids during pregnancy.Furthermore, the embryo and fetus contain cannabinoid receptors.Indeed, CB1 mRNA is expressed in the neural tube, as well as in the fetal central nervous system, retina, autonomic ganglia, enteric ganglia, thyroid, and adrenal glands (Buckley et al., 1998).CB2 mRNA is expressed exclusively in the liver as early as embryonic stage-E13.The region-specific expression of CB1 and CB2 mRNAs suggests functional roles for these receptors during embryogenesis.Collectively, these findings indicate that endocannabinoid-signaling may directly mediate diverse functions from early embryonic development to birth.All these processes are potential targets for exogenous binoids, which may account for many adverse effects of marijuana and cannabinoids on pregnancy outcome.
The presence of AEA, PEA, OEA, 2-AG, and oleamide in milk implies potential effects on nursing infants such as feeding and sleep.These possibilities are suggested by the use of Cannabis to stimulate appetite and induce sleep, properties that have been known since ancient times (Abel, 1980;Mechoulam, 1986;Mechoulam and Hanus, 2001;Ahokas, 2002).Recent studies showed that AEA and THC produce similar effects (Holister, 1999;Jamshide and Taylor, 1999;Tassinari et al., 1999;Williams and Kirkham, 1999;Hao et al., 2000;Budney et al., 2001;Di Marzo et al., 2001a;Mechoulam and Fride, 2001).Withdrawal from marijuana smoke or oral THC in human is associated with decreased appetite and sleep (Haney et al., 1999;Budney et al., 2001).The CB1 antagonist SR141716A decreases food uptake and body weight, indicative of regulation by endocannabinoid-signaling (Arnone et al., 1997;Rowland et al., 2001).SR141716A also reduces AEAinduced food intake and sleep (Columbo et al., 1998;Jamshide and Taylor, 1999;Williams and Kirkham, 1999;Murillo-Rodriguez et al., 2001).Administration of OEA to rats causes potent and persistent decrease in food intake (Rodriguez de Fonseca, 2001).PEA also produces anorexic effects but is less potent than OEA.The inhibitory effect of OEA on feeding is not blocked by CB1and CB2-specific antagonists (SR141716 and SR144528, respectively), suggesting that it does not act via currently known cannabinoid receptors.Oleamide is a sleep inducing agent (Cravatt et al., 1995), is an effective appetite suppressant in rats (Rodriguez de Fonseca et al., 2001), exhibits some cannabimimetic activity, and binds with low affinity to CB1 and CB2 (Lambert and Di Marzo, 1999).SR141716A blocks the soporific action of oleamide suggesting a possible role for cannabinergic pathways (Mendelson and Basile, 2001).The physiological interactions of NAEs, 2-AG, and oleamide appear to be quite complex and are poorly understood.Since NAEs and oleamide are substrates for fatty acid amide hydrolase, substances such as PEA and OEA may function as entourage compounds that enhance the actions of endocannabinoids by preventing their inactivation (Mechoulam et al., 1998;Lambert and Di Marzo, 1999;Di Marzo et al., 2001b).Furthermore, endocannabinoids and oleamide may affect appetite and sleep indirectly via interactions with other signaling systems, e.g.leptin, dopamine, serotonin acetylcholine, etc. (Chaperon and Thiebot, 1999;Kenney et al., 1999;Lambert and Di Marzo, 1999;Di Marzo et al., 2001a).
Endocannabinoid-signaling has a critical role in nursing (Fride et al., 2001).Injection of SR141716A into newly born mouse pups produces devastating effects on suckling and milk ingestion, followed by death within 4 Á/8 days.Oral administration of AEA and 2-AG to mice can produce psychotropic effects, suggesting that these endocannabinoids reach the brain (Di Marzo et al., 1998).Co-administration of THC or 2-AG protects pups from the adverse effects of SR141716 on suckling and survival, suggesting the involvement of CB1 receptors (Fride et al., 2001).These findings indicate that milk-derived endocannabinoids may influence infant physiology and behavior, a notion supported by common parental experience that nursing babies get sleepy as they become satiated.
Studies on human, monkeys, and rodents showed that THC administered to nursing mothers accumulates in milk, and is transferred to their offspring (Chao et al., 1976;Dalterio and Bartke, 1979;Frischknecht et al., 1980;Perez-Reyes and Wall, 1982;Asch and Smith, 1986;Astley and Little, 1990).Dairy animals, grazing in regions where Cannabis is part of the natural vegetation, produce milk contaminated with cannabinoids that can be transferred to human infants (Ahmad and Ahmad, 1990).Postnatal growth, motor development, and behavior are adversely affected in babies fed THC-contaminated milk (Dalterio et al., 1984;Frischknecht et al., 1980;Astley and Little, 1990).Furthermore, THC inhibits suckling-induced milk ejection in lactating rats (Tyrey and Murphy, 1988).Collectively, these observations suggest that THC in milk can directly affect newborns, and that it also can produce secondary effects in infants by reducing maternal milk availability.

Malignant ovarian cysts
NAEs and cannabinoids inhibit proliferation of human prostatic and mammary carcinoma, as well as tumors of non-reproductive organs (De Petrocellis et al., 2000;Di Marzo et al., 2001b).THC induces apoptosis in glioma and prostatic carcinoma cells, by mechanisms that do not depend on currently known cannabinoid receptors (Ruiz et al., 1999;Sanchez et al., 1998).Significantly, the activity of the enzyme responsible for removal of NAEs (fatty acid amide hydrolase) is fivefold higher in human adenocarcinoma than in normal epithelial cells in a healthy uterus (Maccarrone et al., 2000).These results imply that NAE-signaling normally regulates cell division, differentiation, and survival.We detected AEA, OEA, and PEA in fluids collected from human malignant ovarian cysts (Table 4).The concentrations of PEA and OEA in fluids from malignant ovarian cysts are significantly lower than those in follicular fluid.The possible physiological significance of these data remains to be determined.

Conclusions
The discovery of functional cannabinoid receptors and their endogenous ligands has stimulated an explosive expansion of research during the past decade on their functions in reproduction and other physiological processes.Many obvious questions need to be answered.For example, are there other endogenous ligands?Do additional subtypes of cannabinoid receptors exist?Are there endocannabinoid receptors that do not react with classical cannabinoids?How does NAE-signaling affect other signal systems to modulate cellular responses to stimulation?How do exogenous cannabinoids delivered via smoking marijuana or by other modalities impact these processes?What are the medical consequences of marijuana abuse?Does marijuana smoke reduce human fertility significantly, and are the adverse effects of cannabinoids on reproductive functions reversible?Should public policies regulating the use of marijuana and synthetic cannabinoids be modified?Contradictory findings in the literature need to be reconciled.Finally, the effects of orally administered cannabinoids or endocannabinoids need to be explained by demonstrating their absorption and tissue distribution.Do these compounds reach the brain and other target organs unchanged and in sufficient amounts?Additional research is required to resolve these issues.
Normal operation of an endogenous signaling system requires regulated rapid release and removal of endogenous agonists.A classic example of this phenomenon is the hydrolysis of acetylcholine by acetylcholine esterase immediately after its release at synaptic endings.A similar situation applies to signals mediated by AEA, other NAEs, 2-AG, and oleamide which are hydrolyzed by fatty acid amide hydrolase (Piomelli et al., 1998;Giuffrida et al., 2001c;Hillard, 2000).2-AG is also degraded by monoacylglycerol lipase (Schmid, 2000).By contrast, drugs such as THC flood endocannabinoid signal systems because they are slowly metabolized, accumulate in fat stores, and produce persistent effects with potentially damaging consequences, especially in chronic marijuana smokers (Nahas et al., 2002).For example, AEA and CB1 receptors are found in the testis, suggesting possible roles for AEA-signaling in modulation of sperm production and maturation (Gerard et al., 1991;Galiegue et al., 1995;Sugiura et al., 1996;Wenger et al., 2001).These factors may account, at least in part, for the adverse effects of marijuana smoke and cannabinoids on sperm production in human and laboratory animals (see Schuel et al., 1999 and references cited therein).
Children born to parents who are chronic marijuana users are vulnerable to cumulative effects of cannabinoids.Exposure begins during maturation of eggs and sperm within their parents' gonads.It continues during sperm capacitation and fertilization, preimplantation development of the embryo within the oviduct and uterus, and during pregnancy via THC transit across the placenta (Harbison et al., 1972;Martin et al., 1977;Asch and Smith, 1986;Paria and Dey, 2000;Paria et al., 2001;Schuel et al., 2002).Newborn humans, that had been exposed to marijuana prenatally, exhibit increased tremors, exaggerated responses to stimulation, and spend less time sleeping quietly (Fried and Smith, 2001).Nursing infants ingest THC in mother's milk (Astley and Little, 1990), and are also continuously exposed to secondary marijuana smoke at home.Together, these factors may be responsible for the adverse effects of marijuana and cannabinoids on growth and behavior in offspring of human and laboratory animals (Vardaris et al., 1976;Dalterio and Bartke, 1979;Frischknecht et al., 1980;Morrill et al., 1983;Dalterio et al., 1984;Asch and Smith, 1986;Fisher et al., 1987;Astley and Little, 1990;Ahmad and Ahmad, 1990;Fried and Smith, 2001).
Prior to the discovery of cannabinoid receptors and their endogenous ligands, it was generally believed that the peripheral pharmacological effects of cannabinoids resulted from their actions within the central nervous system.Direct druginduced effects on non-neural cells outside the central nervous system were considered to be unlikely.These opinions reflected the primary focus of investigators on the psychoactive properties of marijuana, and its classification as a dangerous illicit substance.However, recent studies show that cannabinoid receptors are expressed by epithelial, smooth muscle, and immune cells in peripheral organs, as well as by sperm and preimplantation embryos (Deutsch et al., 1997;Schuel et al., 1999Schuel et al., , 2002;;Berdyshev, 2000;De Petrocellis et al., 2000;Paria et al., 2001).Endocannabinoids are likewise produced locally by these cells.With the wisdom of hindsight, it is now clear that marijuana's peripheral effects were red flags announcing the presence of NAE-signaling systems outside the nervous system.Understanding these phenomena will expand our knowledge of human biology and pathology, and also may provide the basis for the future development of novel drugs for medicinal uses.

Table 4
NAEs in fluids from human malignant ovarian cysts and fluids from normal ovarian follicles P -values compare significance of levels for each NAE in malignant ovarian cysts fluids to those in follicular fluids; values are given in nM.