Experimental Procedures

Materials

Capsaicin, capsazepine, resiniferatoxin, and (+)-WIN55212 were obtained from Tocris Cookson (St. Louis, MO) and SR141716A from SANOFI Research Center (Montpellier, France). [³H]CP55940 and [³H]resiniferatoxin were obtained from APBiotech (Little Chalfont, UK). PGE₂ethanolamide was obtained from Cayman Chemical (Ann Arbor, MI) and SC-51089 from BIOMOL Research Laboratories (Plymouth Meeting, PA). OL-093 is compound 53 in Boger et al. (2000) and was a gift from Dr. Boger (Scripps Research Institute, La Jolla, CA). Bovine serum albumin (BSA), cell culture medium, nonenzymatic cell dissociation solution, G418, l-glutamine, Krebs' salts, penicillin with streptomycin, PMSF, and Triton X-100 were all obtained from Sigma-Aldrich (St. Louis, MO). Rat VR1 transfected CHO cells were a gift from Novartis (London, UK). HEK cell membranes expressing hEP₁, hEP₃, and hEP₄ receptor subtypes were a gift from Merck Frosst (Montreal, QB, Canada) and hEP₂ was a gift from Allergan (Irvine, CA).

Cell Culture

VR1 Transfected CHO Cells.

rVR1 transfected CHO cells were maintained in minimum essential medium-α minus medium containing 2 mM l-glutamine supplemented with 10% Hyclone fetal bovine serum, 350 μg ml⁻¹ G418, 100 units ml⁻¹ penicillin, and 100 μg ml⁻¹ streptomycin. Cells were maintained in 5% CO₂ at 37°C. For the radioligand binding assay, cells were removed from flasks by scraping and then frozen as a pellet at −20°C for up to 1 month.

Primary Cultured DRG Neurons.

Primary cultures of DRG neurons were prepared following enzymatic (0.125% collagenase, 0.25% trypsin, and 1.6 units ml⁻¹ DNase) and mechanical dissociation of dorsal root ganglia from decapitated 2-day-old Sprague-Dawley rats. The sensory neurons were plated on laminin-polyornithine-coated coverslips and bathed in Dulbecco's modified Eagle's medium/F-12 Ham's medium supplemented with 10% fetal bovine serum, 5000 IU ml⁻¹ penicillin, 5000 mg ml⁻¹ streptomycin, and 20 ng ml⁻¹ nerve growth factor. The cultures were maintained for up to 2 weeks at 37°C in humidified air with 5% CO₂, and refed with fresh culture medium every 5 to 7 days.

Radioligand Binding Experiments

VR1 Cell Membranes.

Assays were performed in Dulbecco's modified Eagle's medium containing 25 mM HEPES and 0.25 mg ml⁻¹ BSA. The total assay volume was 500 μl containing 20 μg of cell membranes. Binding was initiated by the addition of the VR1 receptor agonist [³H]resiniferatoxin (100 pM). Assays were carried out at 37°C for 1 h, before termination by addition of ice-cold wash buffer (50 mM Tris buffer and 1 mg ml⁻¹ BSA, pH 7.4) and vacuum filtration using a 12-well sampling manifold (cell harvester; Brandel, Inc., Gaithersburg, MD) and GF/B filters (Whatman, Maidstone, UK) that had been soaked in wash buffer at 4°C for at least 24 h. Each reaction was washed nine times with a 1.5-ml aliquot of wash buffer. The filters were oven-dried for 60 min and then placed in 5 ml of scintillation fluid. Radioactivity was quantified by liquid scintillation spectrometry. Specific binding was determined in the presence of 1 μM unlabeled resiniferatoxin. Protein assays were performed using a Bio-Rad DC kit (Bio-Rad, Hercules, CA).

EP Receptor Membranes.

EP₁, EP₃, and EP₄ membranes were a gift from Merck Frosst, and EP₂ membranes were from Allergan. The transfection details can be found in Abramovitz et al. (2000). The assay conditions used in this study are identical to those used by Abramovitz et al. (2000) with the same membranes. Under these conditions it has been shown that the specific binding of PGE₂ 1) represents 75 to 95% of the total binding, 2) is linear with respect to the concentrations of radioligand and protein, and 3) has reached equilibrium within the stated incubation time and temperature. Assays were performed in 10 mM MES buffer containing 10 mM MgCl₂ and 1 mM EDTA, pH 6.0. The total assay volume was 200 μl containing 41, 40, 2.3, and 16.5 μg of cell membranes containing EP₁, EP₂, EP₃, and EP₄, respectively. Binding was initiated by the addition of 500 pM [³H]PGE₂. Assays were carried out at 30°C for 1 h, before termination by addition of ice-cold wash buffer (10 mM MES buffer, pH 6.0) and vacuum filtration using a 12-well sampling manifold (cell harvester; Brandel, Inc.) and GF/B filters (Whatman) that had been soaked in wash buffer at 4°C for at least 24 h. Each reaction was washed six times with a 1.5-ml aliquot of wash buffer. The filters were oven-dried for 60 min and then placed in 5 ml of scintillation fluid. Radioactivity was quantified by liquid scintillation spectrometry. Specific binding was determined in the presence of 1 μM unlabeled PGE₂. Protein assays were performed using a Bio-Rad Dc kit.

Drug Additions.

PGE₂ and PGE₂ ethanolamide were stored as 10 mM top stocks in dimethyl sulfoxide (DMSO); for the radioligand binding assay they were diluted in assay buffer and the concentration of DMSO kept constant at 0.1% throughout.

Data Analysis.

The concentrations of competing ligands to produce 50% displacement of the radioligand (IC₅₀) from specific binding sites was calculated using GraphPad Prism (GraphPad Software, San Diego, CA).K_d values for [³H]PGE₂ and [³H]resiniferatoxin binding to EP receptor membranes and VR1 receptor membranes were obtained from Abramovitz et al. (2000) and Ross et al. (2001), respectively. Dissociation constant (K_i) values were calculated using the equation of Cheng and Prusoff (1973).

Effect of FAAH Inhibitors and PGE₂ Ethanolamide on Binding of Anandamide to Mouse Brain Membranes.

Binding assays were performed with the CB₁ receptor agonist [³H]CP55140 (0.5 nM), 1 mM MgCl₂, 1 mM EDTA, 1 mg ml⁻¹ BSA, and 50 mM Tris buffer, with a total assay volume of 500 μl. Binding was initiated by the addition of mouse brain membranes (50–75 μg). Assays were carried out at 37°C for 60 min before termination by addition of ice-cold wash buffer (50 mM Tris buffer and 1 mg ml⁻¹ BSA) and vacuum filtration using a 12-well sampling manifold (cell harvester; Brandel, Inc.) and GF/B (Whatman) glass fiber filters that had been soaked in wash buffer at 4°C for 24 h. Each reaction tube was washed five times with a 4-ml aliquot of buffer. The filters were oven-dried for 60 min and then placed in 5 ml of scintillation fluid (Ultima Gold XR; Packard BioScience, Meriden, CT), and radioactivity quantitated by liquid scintillation spectrometry. Specific binding was defined as the difference between the binding that occurred in the presence and absence of 1 μM unlabeled CP55940 and was 70 to 85% of the total binding. Membranes were preincubated with enzyme inhibitors (PMSF, OL-093, or PGE₂ ethanolamide) or vehicle (0.01% ethanol) for 30 min at 37°C before the addition of 100 nM anandamide or anandamide vehicle. The percentage of increase in the anandamide displacement of [³H]CP55940 due to the enzyme inhibitors was calculated as follows: [increase in displacement of [³H]CP55940 specific binding by 100 nM anandamide in the presence of enzyme inhibitor]/[increase in displacement of [³H]CP55940 specific binding by 100 nM anandamide in the presence of vehicle alone] × 100.

Isolated Tissue Experiments

Guinea Pig Vas Deferens.

Vasa deferentia were obtained from Dunkin-Hartley guinea pigs weighing 300 to 800 g. Each tissue was mounted in a 4-ml organ bath at an initial tension of 0.5 g. The baths contained Mg²⁺-free Krebs' solution, which was kept at 37°C and bubbled with 95% O₂ and 5% CO₂. The composition of the Krebs' solution was 118.2 mM NaCl, 4.75 mM KCl, 1.19 mM KH₂PO₄, 25.0 mM NaHCO₃, 11.0 mM glucose, and 2.54 mM CaCl₂ · 0.6H₂O. Isometric contractions were evoked by stimulation with supramaximal field stimulation (60 V, 1 ms, 10 Hz every 32 s) through a platinum electrode attached to the upper end and a stainless steel electrode attached to the lower end of each bath. Stimuli were generated by a Grass S48 stimulator then amplified (channel attenuator; Med-Lab, Reading, UK) and divided to yield separate outputs to four organ baths (StimuSplitter; Med-Lab). Contractions were monitored by computer (Apple Macintosh LCIII and Performa 475) using a data recording and analysis system (MacLab; ADI Instruments, Milford, MA) that was linked via preamplifiers (Macbridge) to either UF-1 transducers (Pioden Controls Ltd., Canterbury, UK) or model 1030 transducers (UFI, Morro Bay, CA).

Guinea Pig Trachea.

Trachea were obtained from Dunkin-Hartley guinea pigs weighing 300 to 800 g. Animals were stunned, exsanguinated, and the lungs with attached trachea were quickly removed. The main trachea was dissected and 3-mm rings prepared. Each ring was mounted in a 4-ml organ bath at an initial tension of 0.5 g. The baths contained Krebs' solution, which was kept at 37°C and bubbled with 95% O₂ and 5% CO₂. The composition of the Krebs' solution was 1.29 mM MgSO₄ · 7H₂O, 118.2 mM NaCl, 4.75 mM KCl, 1.19 mM KH₂PO₄, 25.0 mM NaHCO₃, 11.0 mM glucose, and 2.54 mM CaCl₂ · 6H₂O. It also contained 10 μM indomethacin. At the end of each experiment the tissue was exposed to 100 μM histamine and the contractions for each compound expressed as a percentage of this response. Contractions were monitored by computer (Apple Macintosh LCIII and Performa 475) using a data recording and analysis system (MacLab) that was linked to either UF-1 transducers (Pioden Controls) or model 1030 transducers (UFI).

In experiments investigating the relaxation of the guinea pig trachea, the tissue was preincubated with the EP₁ receptor antagonist SC-51089 (10 μM). The tissues were then contracted with a concentration of 1 μM histamine and exposed to each concentration of agonist for 5 min. The relaxation of the tissue was expressed as a percentage of the inhibition of the histamine-induced contraction.

Rabbit Jugular Vein.

Jugular veins were obtained from New Zealand White rabbits weighting 2 to 3 kg that had been anesthetized with 30 mg ml⁻¹ Sagatal injected into the marginal ear vein. The jugular veins were cut into rings of 3 to 5 mm wide. Each ring was mounted between two stainless steal hooks in a 4-ml organ bath at an initial tension of 0.75 g. The baths contained Krebs' solution (as for the trachea) containing 10 μM indomethacin, which was kept at 37°C and bubbled with 95% O₂and 5% CO₂. Tissues were exposed to an initial concentration of 100 nM phenylephrine and allowed to recover before the addition of 1 μM phenylephrine. Prostanoids were diluted in ethanol from a top stock of 10 mM and were added cumulatively with no washout between additions. The vehicle did not significantly relax the tissue at the concentrations used. Contractions were monitored by computer (Apple Macintosh LCIII and Performa 475) using a data recording and analysis system (MacLab) that was linked to either UF1 transducers (Pioden Controls) or model 1030 transducers (UFI, California).

Drug Additions.

All agonist additions were made cumulatively without washout in a volume of 10 μl. Top stocks of drugs were 10 mM in DMSO except indomethacin, which was a 20-mg ml⁻¹ stock in ethanol. For isolated tissue experiments, PGE₂ and PGE₂ethanolamide were serially diluted in saline, SR141716A was diluted in a DMSO/saline (50:50) mixture, and capsazepine was diluted in neat DMSO. In control experiments, the appropriate vehicle was added instead of agonist or antagonist, and it had no effect when added alone in either the trachea (contraction or relaxation) or the vas deferens (n = 4, data not shown). Antagonists/vehicle were added 30 min before the addition of agonists.

Analysis of Data.

Values have been expressed as means and variability as S.E.M. or as 95% confidence limits. The values for pEC₅₀ (−log EC₅₀) are defined as the effective concentration producing 50% of the maximum response inducible by that compound. EC₅₀and maximal effects (E_max), and the S.E.M. or 95% confidence limits of these values have been calculated by nonlinear regression analysis using the equation for a sigmoid concentration-response curve (GraphPad Prism).

Calcium Imaging Experiments

Calcium Imaging.

Cultured DRG neurons were incubated for 1 h in NaCl-based extracellular solution (130 mM NaCl, 3.0 mM KCl, 0.6 mM MgCl₂, 2.0 mM CaCl₂, 1.0 mM MgHCO₃, 10.0 mM HEPES, and 5.0 mM glucose; pH 7.4, 310–320 mOsM) containing 10 μM Fura-2 AM. The cells were constantly perfused with NaCl-based solution (1–2 ml/min) and viewed under an inverted BX50WI microscope with a KAI-1001 S/N 5B7890-4201 camera attached (Olympus, Tokyo, Japan). The fluorescence ratiometric images from data obtained at excitation wavelengths of 340 and 380 nm were viewed and analyzed using UltraView (Merlin morphometry). The DRG neurons were exposed to 3 μM PGE₂ ethanolamide for 3 min followed by 10 min of washout with NaCl solution. This was followed by exposure to 100 nM capsaicin for 30 s and washout for 5 min and finally exposure to 30 mM KCl for 30 s. The Ca²⁺ transient (fluorescence ratio after background subtraction) generated by each compound was measured. All data are expressed as means ± S.E.M.

Results

Radioligand Binding Studies

EP Receptor Membranes.

Figure 1shows the displacement of 0.5 nM [³H]PGE₂ from membranes expressing EP1 to 4 receptors by PGE₂ and PGE₂ ethanolamide. PGE₂ethanolamide had significantly lower affinity than PGE₂ for all the hEP receptor subtypes investigated in the radioligand binding assay (Fig. 1). At each receptor, the pK_i values for PGE₂ ethanolamide (Table1) are significantly higher than for PGE₂ (P < 0.01, one-way ANOVA). The level of specific binding for [³H]PGE₂ observed in membranes was 0.9 ± 0.1 pmol mg⁻¹ for EP₁, 0.48 ± 0.05 pmol mg⁻¹ for EP₂, 17.3 ± 4.08 pmol mg⁻¹ for EP₃, and 2.45 ± 0.19 pmol mg⁻¹ for EP₄ (all n = 6). The specific binding represented 81.98 ± 3.87, 66.55 ± 0.91, 93.90 ± 2.39, and 95.50 ± 0.29% (all n = 6) of the total binding for EP₁, EP₂, EP₃, and EP₄, respectively. These data are in line with those published by Abramovitz et al. (2000)using the same membranes under identical conditions.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Displacement of 0.5 nM [³H]PGE₂ by PGE₂ and PGE₂ ethanolamide (PGE₂Eth) from membranes of HEK cells transfected with human EP₁ receptors, where the specific binding of PGE₂ was 0.9 ± 0.1 pmol mg⁻¹, representing 81.98 ± 3.87% of total binding (a); with EP₂ receptors, where the specific binding of PGE₂ was 0.48 ± 0.05 pmol mg⁻¹, representing 66.55 ± 0.91% of the total binding (b); with EP₃ receptors, where the specific binding of PGE₂ was 17.3 ± 4.08 pmol mg⁻¹, representing 93.90 ± 2.39% of the total binding (c); and with EP₄ receptors, where the specific binding of PGE₂ was 2.45 ± 0.19 pmol mg⁻¹, representing 95.50 ± 0.29% of the total binding (d). Each symbol represents the mean percentage of displacement ± S.E.M. (n = 3–6).

It is possible that, in the human embryonic kidney cell membranes, PGE₂ ethanolamide is undergoing hydrolysis to be converted to PGE₂. To further investigate this possibility experiments with the EP₃ membranes were carried out in the presence of the nonselective amidase inhibitor PMSF. In the presence of 200 μM PMSF the pK_i values for PGE₂ and PGE₂ ethanolamide in the EP₃ membranes were 9.52 ± 0.08 (n = 3) and 6.64 ± 0.26 (n = 3). These values were not significantly different from the values in the absence of PMSF (P > 0.05, unpaired ttest).

VR1 Receptor Membranes.

In cell membranes from rat VR1 transfected CHO cells, PGE₂ ethanolamide produced only a modest displacement (23.58 ± 5.36% at 10 μM,n = 4) of specifically bound [³H]resiniferatoxin.

Effect of FAAH Inhibitors and PGE₂ Ethanolamide on Binding of Anandamide to Mouse Brain Membranes.

In mouse brain membranes the level of inhibition of specific binding of the CB₁ receptor agonist [³H]CP55940 by 100 nM anandamide (32.54 ± 4.15%, n = 9) was not enhanced by PGE₂ ethanolamide at concentrations up to 10 μM (Fig. 2). In contrast, PMSF caused a concentration-related enhancement of the inhibition of specific binding of [³H]CP55940 by 100 nM anandamide. The recently developed potent FAAH inhibitor OL-093 (compound 53 in Boger et al., 2000) also enhanced the activity of anandamide. The EC₅₀ values for the percentage of increase in displacement of [³H]CP55940 by 100 nM anandamide were 535 nM (confidence limits, 170-1681) for PMSF and 0.22 nM (confidence limits, 0.065–0.71) for OL-093. There is evidence that the inhibition of FAAH is pH-sensitive and the inhibitory action of PMSF has been shown to be greater at low pH (Holt et al., 2001). However at pH 5, PGE₂ ethanolamide failed to alter the level of inhibition of specific binding of [³H]CP55940 induced by 100 nM anandamide (31.55 ± 2.13%, n = 3). PGE₂ ethanolamide and OL-093 alone had no effect on the specific binding of [³H]CP55940 to mouse brain membranes at concentrations up to 10 μM (n = 3 for each, data not shown) at either pH 7.4 or 5.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

The percentage of increase in the displacement of 0.5 nM [³H]CP55940 binding to mouse brain membranes by 100 nM anandamide in the presence of increasing concentrations of PMSF, OL-093, and PGE₂ ethanolamide (PGE₂Eth). Each symbol represents the mean percentage of increase in displacement of [³H]CP55940 by anandamide ± S.E.M. (n = 6).

Isolated Tissue Experiments

Guinea Pig Vas Deferens.

PGE₂ethanolamide inhibited electrically evoked contractions of the vas deferens, with a pEC₅₀ of 7.38 ± 0.09 (n = 8) compared with that of 9.09 ± 0.06 (n = 6) for PGE₂ (Fig.3a). In the presence of PMSF, the potency of both PGE₂ and PGE₂ethanolamide was reduced by around 4-fold; however, the relative potency of the compounds was unchanged (data not shown). The VR1 receptor antagonist capsazepine (10 μM) caused a slight rightward shift in the log concentration-response curve for PGE₂ ethanolamide, but the pEC₅₀ values (8.27 ± 0.29 with vehicle and 7.80 ± 0.06 with capsazepine) were not significantly different (P > 0.05, unpaired t test) (Fig. 3b). The inhibitory action of PGE₂ ethanolamide was not significantly affected by SR141716A (P > 0.05, unpaired t test) (Fig. 3c). The pEC₅₀values were 7.63 ± 0.12 (n = 6) in the presence of vehicle and 7.57 ± 0.11 (n = 6) in the presence of 100 nM SR141716A. The CB₁ receptor agonist CP55940 also inhibited the electrically evoked contractions, and this effect was antagonized by the CB₁receptor antagonist SR141716A (Fig. 3d). The pEC₅₀ value of CP55940 was 7.22 ± 0.12 (n = 6) in the presence of vehicle and 5.57 ± 0.13 (n = 6) in the presence of 100 nM SR141716A. Thus, the K_B value for SR141716A was 2.17 ± 0.71 nM (n = 6).

• Download figure
• Open in new tab
• Download powerpoint

Figure 3

Inhibition of electrically evoked contractions of the guinea pig vas deferens by PGE₂ and PGE₂ethanolamide (PGE₂Eth) (a), PGE₂ ethanolamide in the presence of vehicle (vh) and 10 μM capsazepine (CPZ) (b), PGE₂ ethanolamide in the presence of vehicle and SR141716A (SR141) (c), and CP55940 in the presence of vehicle and 100 nM SR141716A (d). Each symbol represents the mean percentage of inhibition ± S.E.M. (n = 5–6).

Guinea Pig Trachea.

In the trachea, the behavior of PGE₂ ethanolamide was similar to that of PGE₂, producing contractions at lower concentrations followed by relaxation at higher concentrations. Thus, the concentration-response curve for both compounds was bell-shaped (Fig. 4). Maximum contractions were produced by PGE₂ ethanolamide at 1 μM and by PGE₂ at 100 nM. These were 38.9 ± 5.6% and 51.8 ± 10.6% (n = 6), respectively. The EP₁ receptor antagonist SC-51089 (10 μM) abolished the contractions induced by both PGE₂and PGE₂ ethanolamide (Fig.5a). In the presence of 10 μM SC-51089, the relaxant effects of these compounds may be indicative of EP₂ receptor activation. Under these circumstances, both PGE₂ and PGE₂ ethanolamide caused a concentration-related relaxation of histamine-induced contractions of this tissue, the pEC₅₀ values being 8.29 ± 0.17 (n = 6) and 7.11 ± 0.18 (n = 6), respectively (Fig. 5b).

• Download figure
• Open in new tab
• Download powerpoint

Figure 4

Histogram showing the bell-shaped concentration-response relationship for contraction of the guinea pig trachea by PGE₂ and PGE₂ ethanolamide (PGE₂Eth). Each column represents the mean percentage of contraction (expressed as a percentage of the histamineE_max) ± S.E.M. (n= 4–6).

• Download figure
• Open in new tab
• Download powerpoint

Figure 5

a, contraction of the guinea pig trachea by PGE₂ and PGE₂ ethanolamide (PGE₂Eth) in the presence of vehicle and the EP₁ receptor antagonist SC-51089 (SC) (10 μM). Each symbol represents the mean percentage of contraction (expressed as a percentage of the histamine maximum) ± S.E.M. (n = 4–6). b, relaxation by PGE₂ and PGE₂ ethanolamide of guinea pig trachea precontracted with 1 μM histamine in the presence of 10 μM SC-51089. Each symbol represents the mean percentage of inhibition ± S.E.M. (n = 4–6).

Rabbit Jugular Vein.

The rabbit jugular vein is thought to contain EP₄ receptors that mediate relaxation (Coleman et al., 1994; Milne et al., 1995). In this tissue both PGE₂ and PGE₂ ethanolamide caused concentration-related relaxation of the rabbit jugular vein precontracted with phenylephrine (Fig.6). However, PGE₂ethanolamide was 200-fold less potent than PGE₂. The pEC₅₀ values were 9.35 ± 0.25 (n = 5) and 7.05 ± 0.4 (n = 6) for PGE₂ and PGE₂ethanolamide, respectively.

• Download figure
• Open in new tab
• Download powerpoint

Figure 6

Relaxation by PGE₂ and PGE₂ethanolamide (PGE₂Eth) of rabbit isolated jugular vein precontracted with 1 μM phenylephrine in the presence of 10 μM indomethacin. Each symbol represents the mean percentage of inhibition ± S.E.M. (n = 4–6).

Calcium Imaging.

Of a total of 252 DRG neurons from 15 separate cultures there were three groups of responses observed (Fig.7). In all the cells analyzed, high K⁺ (30 mM) evoked an increase in [Ca²⁺]_i, thus confirming the neuronal phenotype. One group (group 1) of neurons (160/252) responded only to 30 mM KCl; thus, this group constituted 63.5% of the total population. The second group (group 2) (92/252) responded to 100 nM capsaicin and constituted 36.5% of the total population. A third group (group 3) of cells responded to PGE₂ethanolamide and this constituted 21% of the capsaicin-sensitive population (19/92). PGE₂ ethanolamide did not elicit a Ca²⁺ transient in any capsaicin-insensitive cells. The Ca²⁺ transient evoked by 3 μM PGE₂ ethanolamide was not significantly different (P > 0.05, one-way ANOVA) from that of evoked by 100 nM capsaicin in the same group of cells, the change in fluorescence being 0.55 ± 0.09 for PGE₂ ethanolamide and 0.75 ± 0.10 for capsaicin (Fig. 7c). In cells that did not respond to PGE₂ ethanolamide, the response to capsaicin (0.95 ± 0.09) was not significantly different (P> 0.05, one-way ANOVA) from that obtained in those that responded to PGE₂ ethanolamide (Fig. 7c). The response to 30 mM KCl was not significantly different in any of the three groups. The area (μm²) of the DRG neurons that responded to KCl only (group 1) (334.64 ± 13.88 μm²) was significantly greater than those that responded to capsaicin (group 2) (239.19 ± 10.77 μm²) (P < 0.01, one-way ANOVA) and those that responded to PGE₂ ethanolamide and capsaicin (group 3) (212 ± 18.6 μm²) (P < 0.001, one-way ANOVA) (Fig. 7d). The VR1 receptor antagonist capsazepine did not significantly attenuate the response to PGE₂ ethanolamide. Experiments were conducted in which the cells were exposed to 3 μM PGE₂ethanolamide followed by 15 min washout and then treatment for 5 min with either capsazepine or vehicle (0.1% DMSO) before a second exposure to 3 μM PGE₂ ethanolamide. The change in fluorescence in response to PGE₂ ethanolamide was 0.240 ± 0.043 and 0.206 ± 0.052 (n = 3) (P > 0.05, one-way ANOVA) before and after pretreatment with 10 μM capsazepine, respectively. The change in fluorescence in response to PGE₂ ethanolamide was 0.277 ± 0.05 and 0.280 ± 0.033 (n = 3) (P > 0.05, one-way ANOVA) before and after pretreatment with the vehicle, respectively.

• Download figure
• Open in new tab
• Download powerpoint

Figure 7

Effect of PGE₂ ethanolamide (PGE₂Eth) on [Ca²⁺]_i in single DRG neurons. The traces (a and b) show examples of [Ca²⁺]_i responses (F₃₄₀/F₃₈₀ ratio) evoked by 3 μM PGE₂ ethanolamide, 100 nM capsaicin, and KCl in DRG neurons from two separate cultures. The agonists were applied for the periods shown by the horizontal lines. The dotted lines show a DRG neuron that only responds to KCl (group 1), the gray line shows a DRG neuron that responds to both capsaicin and KCl (group 2), and the black line shows a DRG neuron that responds to PGE₂ ethanolamide, capsaicin, and KCl (group 3). The histogram (c) shows the [Ca²⁺]_i evoked in 252 DRG neurons obtained from 15 separate cultures. The data represent the mean change in fluorescence ratio ± S.E.M. The histogram (d) shows the area (μm²) of the DRG neurons in each of the three groups ± S.E.M. The area of the DRG neurons that responded to KCl only (group 1) was significantly (P < 0.01, one-way ANOVA) greater than those that responded to capsaicin (group 2) and those that responded to PGE₂ ethanolamide and capsaicin (group 3) (P < 0.001, one-way ANOVA). The area of the DRG neurons that responded to PGE₂ ethanolamide (group 3) was not significantly different (P > 0.05, one-way ANOVA) from those that responded to capsaicin alone (group 2).

Discussion

It would seem that PGE₂ ethanolamide has a similar profile of action to that of PGE₂ in that it binds to EP₁, EP₂, EP₃, and EP₄ receptors. The affinity of PGE₂ ethanolamide for EP receptor subtypes is significantly lower than that of PGE₂, being 500-, 500-, 440-, and 651-fold lower than that of PGE₂ for hEP₁, hEP₂, hEP₃, and hEP₄ receptors, respectively. The inclusion of the FAAH inhibitor PMSF did not alter theK_i values for either PGE₂ or PGE₂ ethanolamide. This suggests that PGE₂ ethanolamide is acting on the receptor directly rather than being cleaved to form PGE₂. It is notable that Kozak et al. (2001) have demonstrated that PGE₂ ethanolamide is subject to little or no hydrolysis in rat or human blood. It has been previously demonstrated that anandamide and other ethanolamides have significant affinity for the vanilloid VR1 receptor (Smart et al., 2000; Ross et al., 2001). However in this study, PGE₂ethanolamide seemed to have limited affinity for the VR1 receptor with a K_i value of >10 μM, compared with a value of 1.66 μM for anandamide in the same cell line (Ross et al., 2001).

PGE₂ ethanolamide does not seem to interact with FAAH. At pH 7.4, this compound did not enhance the ability of anandamide to displace [³H]CP55940 from mouse brain membranes. Although this is indicative of a lack of modulation of FAAH by this compound, conclusive data can only be obtained with a more specific and direct measure of FAAH activity. In contrast to PGE₂ ethanolamide, the established FAAH inhibitors PMSF and OL-093 (compound 53 in Boger et al., 2000) produce a marked enhancement of the activity of anandamide in mouse brain membranes. This is the first demonstration that OL-093 enhances the activity of anandamide with high potency, presumably by inhibition of FAAH. It has recently been demonstrated that the pharmacological properties of FAAH are highly pH-dependent (Holt et al., 2001) and PMSF has been shown to be almost 60-fold more potent as an inhibitor or FAAH at pH 5.28 than at pH 8.37. However at pH 5, PGE₂ethanolamide did not enhance the activity of anandamide at concentrations up to 10 μM.

In the guinea pig vas deferens preparation, which is reported to contain EP₃ receptors, PGE₂ethanolamide was 45-fold less potent than PGE₂, the EC₅₀ values being 37 and 0.82 nM, respectively. In the presence of PMSF the relative potency of these compounds was unaltered, indicating that PGE₂ethanolamide is not inhibiting the twitch response via conversion to PGE₂. The affinity of PGE₂for hEP₃ receptors (K_i = 0.48 nM) is similar to its potency in the guinea pig vas deferens (EC₅₀ = 0.82 nM). In contrast, the potency of PGE₂ethanolamide is higher than predicted from the low affinity of this compound for the hEP₃ receptor (K_i = 250 nM). At present, EP₃ receptor antagonists are not available but we have excluded the possibility that this compound is interacting with CB₁ or VR1 receptors in this tissue.

In the guinea pig trachea, PGE₂ ethanolamide exhibited a bell-shaped concentration-response relationship, as has previously been demonstrated for PGE₂ (Dong et al., 1986). PGE₂ and PGE₂ethanolamide produced a maximal contraction of the preparation at 100 nM and 1 μM, respectively, followed by a relaxation of the tissue at higher concentrations. The contractile action of both compounds was abolished by SC-51089, indicating that they are contracting the tissue via an interaction with the EP₁ receptor. When the tracheal preparations were precontracted with histamine, in the presence of SC-51089, both PGE₂ and PGE₂ ethanolamide produced concentration-related relaxation of the tissue. PGE₂ ethanolamide was only 15-fold less potent than PGE₂, the EC₅₀ values being 76.9 and 5.19 nM, respectively. Previous pharmacological analysis indicates that this relaxant action of PGE₂ is mediated by the EP₂ receptor subtype (Coleman et al., 1990). The relatively high potency of PGE₂ ethanolamide at the EP₂ receptor in the trachea is not in line with the low affinity of the compound for the EP₂receptor subtype. As with the EP₃ receptor, selective EP₂ receptor antagonists are not yet available. Thus, the potency of PGE₂ ethanolamide in both the vas deferens and the trachea (relaxation) is higher than predicted from the binding assays using human receptors. This may be accounted for by species differences. However, it is also possible that this compound may be interacting with other, as yet uncharacterized receptors for prostaglandin ethanolamides. Interestingly, the pharmacology of PGF_2α ethanolamide (prostamide F_2α) suggests the existence of a novel prostamide receptor. Thus, PGF_2α ethanolamide contracts the cat iris sphincter with potent activity that is not exhibited in other preparations that respond to PGF_2α (Chen et al., 2001). Furthermore, this compound has little affinity for the recombinant cat or human PGF_2α (FP) receptor. In contrast to PGE₂, PGE₂ethanolamide is not significantly hydrolyzed in plasma, is stable in cerebrospinal fluid, and is oxidized less efficiently than PGE₂ (Kozak et al., 2001). This raises the possibility that the higher relative potency of this compound in some tissues may be due to its resistance to metabolism by enzymes that are responsible for the rapid inactivation PGE₂.

In the rabbit jugular vein, which is thought to contain EP₄ receptors mediating relaxation (Coleman et al., 1994; Milne et al., 1995), PGE₂ ethanolamide is around 200-fold less potent than PGE₂.

Prostanoids are known both to directly activate sensory neurons and to sensitize sensory neurons to other potent nociceptive agents such as bradykinin. PGE₂ has been shown to activate a subpopulation of small-diameter capsaicin-sensitive DRG neurons and to potentiate the bradykinin-evoked increases in [Ca²⁺]_i. Both of these actions of PGE₂ seem to involve protein kinase A-dependent mechanisms (Smith et al., 2000). Smith et al. (2000)found that 1 μM PGE₂ evoked an increase in [Ca²⁺]_i in 16% of capsaicin-sensitive DRG neurons. PGI₂ and PGF_2α (1 μM) also evoked calcium transients in 26 and 29% of DRG neurons, respectively. Similarly, in this study we found that 3 μM PGE₂ ethanolamide evoked an increase in [Ca²⁺]_i in 21% of small-diameter capsaicin-sensitive DRG neurons. The possibility that PGE₂ ethanolamide shares the ability of PGE₂ to sensitize DRG neurons to capsaicin (Lopshire and Nicol, 1997) and bradykinin (Smith et al., 2000) is the subject of ongoing investigations. One would expect that activation of cyclic AMP-dependent kinase by PGE₂ ethanolamide may enhance vanilloid receptor-mediated responses in DRG neurons (De Petrocellis et al., 2001). The receptor mechanisms underlying these actions of the prostanoids have not yet been investigated and it remains to be established whether the prostanoids are acting through the same or distinct sites of action to activate DRG neurons.

The physiological significance of the conversion of anandamide to PGE₂ ethanolamide has yet to be established. It may be that the prostaglandin ethanolamides are a new class of mediator; alternatively, it could be speculated that by competing with arachidonic acid for COX-2, increasing levels of anandamide might modulate the local production of prostanoids by this enzyme. This, in turn, would result in less activation of EP receptors because the alternative product, PGE₂ ethanolamide, has lower potency at these receptors than PGE₂. There is strong evidence that anandamide is metabolized by COX-2 to produce PGE₂ ethanolamide in physiologically relevant environments (Yu et al., 1997; Burstein et al., 2000). However, it has yet to be established whether the levels of PGE₂ethanolamide synthesized by COX-2 metabolism of anandamide in vivo are sufficient to activate EP receptors. Potent FAAH inhibitors have recently been synthesized (Boger et al., 2000), which enhance the levels of anandamide significantly, and these compounds may be of considerable therapeutic benefit. Acute and chronic peripheral inflammation, interleukins, and spinal cord injury increase the expression of COX-2 (Vanegas and Schaible, 2001) in the spinal cord and DRG neurons. In the event of inhibition of FAAH metabolism of anandamide, increased levels of endogenous anandamide in combination with an up-regulation of COX-2 (inflammation, injury) may lead to the production of significant levels of the prostaglandin ethanolamides. Recently, it has been shown that PGE₂ethanolamide is significantly more resistant to metabolism than PGE₂, being detectable in rat plasma 2 h after administration (Kozak et al., 2001), thus raising the possibility that this compound may act systemically. In this study we have shown for the first time that PGE₂ ethanolamide is indeed pharmacologically active in some tissues at relatively low concentrations. The physiological consequence and relevance of these actions remain to be established.