Quantitative high-performance liquid chromatography/electrospray ionization tandem mass spectrometric analysis of 2- and 3-series prostaglandins in cultured tumor cells

This paper describes a rapid and simple technique for the simultaneous quantitative analysis of PGE(2), PGE(3), and other closely related prostaglandins from cultured cells using liquid chromatography/electrospray ionization tandem mass spectrometry. This method permits quantification of selected individual prostaglandins derived either from arachidonic acid (AA) or eicosapentaenoic acid (EPA) from cell extracts without tedious derivatization, lengthy sample preparation, and separation required by GC-MS- or HPLC-UV-based methods. The validation assessment showed that the quantitative determination is linear (r(2)>0.999) for both PGE(2) and PGE(3) in the range tested (1-500 ng/ml, 0.0028-1.4 microM) and a coefficient of variation lower than 10% was obtained for samples analyzed on 3 separate days. The detection limit was 2.5 pg for both PGE(2) and PGE(3). Extraction efficiency of PGE(2) and PGE(3) from cell suspensions ranged from 89.4 to 98.2%. As an application of the method, prostaglandins formed by EPA in human lung cancer A549 cells were determined. A 62% reduction of PGE(2) formation was noted when A549 cells were treated with 10 microM of EPA. Concomitantly, EPA increased formation of PGE(3) by 10-fold in A549 cells. This is the first report that unequivocally demonstrates that EPA can be converted to PGE(3) by cyclooxygenase in human cancer cells.     

Separation and quantification of PGE3 following derivatization with panacyl bromide by high pressure liquid chromatography with fluorometric detection.

Separation of prostaglandin E3 (PGE3) from prostaglandin E2 (PGE2) and prostaglandin E1 (PGE1) was achieved following derivatization with p-(9-anthroyloxy)phenacyl bromide (panacyl bromide). The eicosanoid esters were analysed by reverse phase high pressure liquid chromatography with fluorometric detection (excitation 360nm and emission 470nm). Human, rat and mouse adherent cells were incubated overnight and the culture medium extracted, derivatized and analysed for PG production. PGE2 was detected from biological samples of each species tested. PGE2 synthesis was reduced when cells were incubated overnight with 5 microM eicosapentaenoic acid. PGE3 was not detectable under these experimental conditions. Studies were also undertaken using adherent cells from mice, rats and human subjects given dietary fish oil supplements rich in EPA. PGE3 production by these cells was not detected although the dietary regimens yielded substantial incorporation of EPA into cell membranes and leukocyte LTB5 production was demonstrable.     

The measurement of E and 19-hydroxy E prostaglandins in human seminal plasma.

A method is described which measures the four main prostaglandins of human semen (PGE1, E2, 19-hydroxy PGE1, and 19-hydroxy PGE2). For routine measurements E1 and E2 are measured together as are 19-OH E1 and 19-OH E2. These are measured by forming oximes in aqueous solution extraction, methylation and trimethyl silylation followed by gas chromatography. The method has sufficient sensitivity to measure the levels found in the majority of semen samples. The normal range in men with proven fertility was 90 to 260 mug/ml of 19-hydroxy Es and 30-200 mug/ml of Es.     

High performance liquid chromotography of prostaglandins: biological applications.

Two procedures are described for separation and purification of prostaglandins by high performance liquid chromatography. Both systems show excellent resolution of PGA2, PGE2 and PGF2a. Peak definition on the micro-particle silicic acid system is particularly good with the PGs appearing in 2-3 ml of organs effluent. Studies on reproducibility showed that PGE2 and PGE2a could be recovered with a retention volume of 54.2+/-0.76 ml and 64+/-0.6 ml, respectively (n=7, mean +/-50) with good recovery. The column can be run in about one hour and can be regenerated indefinitely (greater than 200 times). The degree of purification is compatible with analysis by gas chromatography-mass spectrometry. Examples showing the application of this chromatographic method to human seminal fluid, human renal tissue, platelet rich plasma and human urine samples indicate that it makes possible analysis of these samples even at low levels.     

Prostaglandin moieties that determine receptor binding specificity in the bovine corpus luteum.

This study provided a pharmacological evaluation of prostaglandin binding to bovine luteal plasma membrane. It was found that [3H]PGF2 alpha' [3H]PGE2' [3H]PGE1 and [3H]PGD2 all bound with high affinity to luteal plasma membrane but had different specificities. Binding of [3H]PGF2 alpha and [3H]PGD2 was inhibited by non-radioactive PGF2 alpha (IC50 values of 21 and 9 nmol l-1, respectively), PGD2 (35 and 21 nmol l-1), and PGE2 (223 and 81 nmol l-1), but not by PGE1 (> 10,000 and 5616 nmol l-1). In contrast, [3H]PGE1 was inhibited by non-radioactive PGE1 (14 nmol l-1) and PGE2 (7 nmol l-1), but minimally by PGD2 (2316 nmol l-1) and PGF2 alpha (595 nmol l-1). Binding of [3H]PGE2 was inhibited by all four prostaglandins, but slopes of the dissociation curves indicated two binding sites. Binding of [3H]PGE1 was inhibited, resulting in low IC50 values, by pharmacological agonists that are specific for EP3 receptor and possibly EP2 receptor. High affinity binding of [3H]PGF2 alpha required a C15 hydroxyl group and a C1 carboxylic acid that are present on all physiological prostaglandins. Specificity of binding for the FP receptor depended on the C9 hydroxyl group and the C5/C6 double bond. Alteration of the C11 position had little effect on affinity for the FP receptor. In conclusion, there is a luteal EP receptor with high affinity for PGE1' PGE2' agonists of EP3 receptors, and some agonists of EP2 receptors. The luteal FP receptor binds PGF2 alpha' PGD2 (high affinity), and PGE2 (moderate affinity) but not PGE1 due to affinity determination by the C9 and C5/C6 moieties, but not the C11 moiety.     

Prostaglandin-E2 and cyclic adenosine 3'-5' monophosphate levels in the hypertrophied rat heart.

To assess whether prostaglandin-E2 (PGE2) and cyclic adenosine 3'-5'-monophosphate (cAMP) are involved in the cardiac response to chronic pressure overload, we measured by specific radioimmunoassay method the cardiac tissue and plasma concentrations of PGE2 and cAMP in an animal model of left ventricular hypertrophy. The cardiac hypertrophy was accompanied by a significant increase in PGE2 content, and a significant decrease in cAMP content, in the heart. In addition, we found elevated PGE2 and cAMP levels in arterial plasma samples from the rats with hypertrophied hearts compared to normal rats. These findings suggest a link between cardiac and vascular PGE2 and cAMP generation and the hemodynamic stresses of advanced cardiac overload.     

In vitro response of prostaglandin E2 receptor (EP3) in the term pregnant rat uterus and cervix to misoprostol

We examined and compared the in vitro effects of misoprostol (synthetic prostaglandin E1 (PGE1) analogue) on prostaglandin E2 (PGE2) secretion and EP3 receptor mRNA expression in the pregnant rat myometrium and cervix at 19 days gestation. Myometrial and cervical tissue samples were exposed to media with or without misoprostol (50 or 100 pg/ml) and incubated for 15 and 30 min, and 1, 3, 6, 12, and 24 h. Media and tissue samples were collected for quantification of PGE2 and mRNA expression of rEP3alpha and rEP3beta receptor, respectively. PGE2 secretion increased (P < or = 0.05) in the myometrium exposed to 50 and 100 pg/ml misoprostol. Cervical PGE2 secretion increased following exposure to the 100 pg/ml dose only. In the myometrium, 50 and 100 pg/ml misoprostol induced elevations in rEP3alpha and rEP3beta receptor mRNA expression. rEP3alpha and rEP3beta receptor mRNA expression in the cervix was not different from controls. These data demonstrate that the EP3 receptor is differentially expressed in the myometrium and cervix in response to misoprostol. This may account for the ability of misoprostol to stimulate the myometrium when administered for cervical ripening.     

A specific radioimmunoassay for PGE2 using an antibody with high specificity and a sephadex LH-20 microcolumn for the separation of prostaglandins.

Antiserum against PGE2 was raised in rabbits following immunization with prostaglandin-hen-gamma-globulin conjugate. The antiserum exhibited 14% cross reactivity with PGE1 and far less cross-reaction with heterologous prostaglandins. A microcolumn of Sephadex LH-20 was used for a partial, but sufficient separation of PGE2 from PGE1 and a complete separation from heterologous prostaglandins to ensure a specific RIA for PGE2. The precision of the method in the rage 10-500 picograms showed a coefficient of variation varying between 4 and 13%. The detection limit was 10 picograms corresponding to 15 pg/ml of PGE2 in serum. In order to demonstrate the validity of the method values obtained for non-diuretic rat renal venous serum were compared with those obtained using the isotope derivative method of Bojesen & Buckhave (1972) on the same samples. The concentrations of PGE2 obtained were 239 +/- 25 pg/ml and 250 +/- 58 pg/ml, respectively.     

Radioimmunoassay of the A prostaglandins

Antibodies to the (PGA) prostaglandin A were produced in rabbits immunized with a conjugate of PGE2 covalently linked to (BSA) bovine serum albumin by reaction with carbodiimide reagent. A radioimmunoassay was developed using dextran-coated charcoal to separate the free from antibody bound PGA1-3H. The sensitivity of the method was found to be 100 picograms/ml of plasma. Ethyl acetate was used for extraction of plasma and the various classes of PGs were separated by silicic acid column chromatography. Recovery of PGA1-3H throughout the entire procedure was 65-75%. The antibody showed progressively decreasing affinity to PGA2, PGA1, PGE2, PGE1, PGB2, and PGF2alpha, respectively. The mean plasma PGA level in adult males (N=13) was found to be 1.39 + or - 0.55 ng/ml, and 1.62 + or - 0.52 ng/ml in adult females (N=7). Corresponding plasma and serum samples were found to give essentially similar results. Plasma PGA levels in adult males treated with indomethacin for rheumatoid arthritis were 0.18 + or - 0.15 ng/ml (P 0.001 in comparison with the normal adult males). This method is sufficiently sensitive, precise, and rapid to allow the routine estimation of the PGAs in biological samples.     

Prostaglandins and thromboxane outflow from the perfused mesenteric vascular bed in spontaneously hypertensive rats

To clarify the metabolism of PGE2, prostacyclin (PGI2) and thromboxane A2 (TxA2) in small vessels in spontaneously hypertensive rats (SHR), we removed superior mesenteric vascular beds from 10 week old SHR and age matched normotensive controls (WKY). The mesenteric artery was perfused with Krebs-Henseleit buffer and samples of effluent collected every 15 minutes during 3 hours perfusion for analysis of PGE2, 6-keto-PGF1 alpha (a stable metabolite of PGI2) and TxB2 (a stable metabolite of TxA2) by specific radioimmunoassays. Levels of all three arachidonic acid (AA) metabolites, PGE2, 6-keto-PGF1 alpha and TxB2, in the mesenteric effluent were significantly reduced in SHR as compared to WKY. TxB2 was detected in all samples throughout the perfusion. 6-keto-PGF1 alpha/PGE2 ratios and TxB2/PGE2 ratios were significantly increased in SHR. 6-keto-PGF1 alpha/TxB2 ratios in the first four samples were significantly decreased in SHR as compared to WKY. These data suggest that there may be reduced availability of PG precusor AA and unbalanced synthesis of PGs in small vessels in SHR. Both may have relevance to the development of hypertension in the animals.     

Redirection of eicosanoid metabolism in mPGES-1-deficient macrophages

Microsomal prostaglandin E synthase (mPGES)-1 is one of several prostaglandin E synthases involved in prostaglandin H2 (PGH2) metabolism. In the present report, we characterize the contribution of mPGES-1 to cellular PGH2 metabolism in murine macrophages by studying the synthesis of eicosanoids and expression of eicosanoid metabolism enzymes in wild type and mPGES-1-deficient macrophages. Thioglycollate-elicited macrophages isolated from mPGES-1-/- animals and genetically matched wild type controls were stimulated with diverse pro-inflammatory stimuli. Prostaglandins were released in the following order of decreasing abundance from wild type macrophages stimulated with lipopolysaccharide: prostaglandin E2 (PGE2)>thromboxane B2 (TxB2)>6-keto prostaglandin F1alpha (PGF1alpha), prostaglandin F(2alpha) (PGF2alpha), and prostaglandin D2 (PGD2). In contrast, we detected in mPGES-1-/- macrophages a >95% reduction in PGE2 production resulting in the following altered prostaglandin profile: TxB2>6-keto PGF1alpha and PGF2alpha>PGE2, despite the comparable release of total prostaglandins. No significant change in expression pattern of key prostaglandin-synthesizing enzymes was detected between the genotypes. We then further profiled genotype-related differences in the eicosanoid profile using macrophages pre-stimulated with lipopolysaccharide followed by a 10-min incubation with 10 microm [3H]arachidonic acid. Eicosanoid products were subsequently identified by reverse phase high pressure liquid chromatography. The dramatic reduction in [3H]PGE2 formation from mPGES-1-/- macrophages compared with controls resulted in TxB2 and 6-keto PGF1alpha becoming the two most abundant prostaglandins in these samples. Our results also suggest a 5-fold increase in 12-[3H]hydroxyheptadecatrienoic acid release in mPGES-1-/- samples. Our data support the hypothesis that mPGES-1 induction in response to an inflammatory stimulus is essential for PGE2 synthesis. The redirection of prostaglandin production in mPGES-1-/- cells provides novel insights into how a cell processes the unstable endoperoxide PGH2 during the inactivation of a major metabolic outlet.     

Comparison of a New P2Y12 Receptor Specific Platelet Aggregation Test with Other Laboratory Methods in Stroke Patients on Clopidogrel Monotherapy

Background

Clinical studies suggest that 10-50% of patients are resistant to clopidogrel therapy. ADP induced platelet aggregation, a widely used test to monitor clopidogrel therapy, is affected by aspirin and is not specific for the P2Y12 receptor inhibited by clopidogrel.

Objectives

To develop a P2Y12-specific platelet aggregation test and to compare it with other methods used for monitoring clopidogrel therapy.

Patients/Methods

Study population included 111 patients with the history of ischemic stroke being on clopidogrel monotherapy and 140 controls. The effect of clopidogrel was tested by a newly developed ADP(PGE1) aggregation test in which prostaglandin E1 treated platelets are used. Results of conventional ADP induced platelet aggregation, VerifyNow P2Y12 assay and ADP(PGE1) aggregation were compared to those obtained by flow cytometric analysis of vasodilator stimulated phosphoprotein (VASP) phosphorylation. Reference intervals for all assays were determined according to the guidelines of Clinical Laboratory Standards Institute.

Results

The P2Y12-specificity of ADP(PGE1) test was proven by comparing it with ADP aggregation in the presence of P2Y1 antagonist, adenosine 3’, 5’-diphosphate. The method was not influenced by aspirin treatment. Approximately 50% of patients were clopidogrel resistant by conventional ADP aggregation and VerifyNow tests. The ADP(PGE1) method and the VASP phosphorylation assay identified 25.9% and 11.7% of patients as non-responders, respectively. ADP(PGE1) aggregation showed good correlation with VASP phosphorylation and had high diagnostic efficiency.

Conclusion

The new ADP(PGE1) method is a reliable test for monitoring P2Y12 receptor inhibition by platelet aggregation. As a subset of patients are non-responders, monitoring clopidogrel therapy by adequate methods is essential.     

Effects of microinjection of various prostaglandins into the 3rd ventricle, median eminence and pituitary on plasma LH in rats.

The effects of microinjection of PG'S (PGE1, E2, F2a) into the 3rd ventricle, median eminence (ME) and anterior pituitary on plasma LH in rats were investigated. Blood samples were obtained by jugular puncture before, and 10 and 45 min after the injection of PGS (50 or 100 mug), plasma LH was measured by radioimmunoassay. In the 3rd ventricle microinjection, PGE2 prodiced a significant rise in plasma LH. PGE1 and F2a did not significantly after plasma LH levels. In the median eminence, PGE2 and E1 produced a significant rise in plasma LH. PGF2a did not alter plasma LH levels. In the pituitary, PGE2 and E1 produced a significant rise in plasma LH. PGF2a did not alter plasma LH levels. These observations indicate that PGs act directly on the hypothalamic-pituitary axis, and that particular PG may be involved in the release of particular horomones from the hypothalamus and pituitary.     

Prostaglandin E2 mediates IL-1beta-related fibroblast mitogenic effects in acute lung injury through differential utilization of prostanoid receptors

The fibroproliferative response to acute lung injury (ALI) results in severe, persistent respiratory dysfunction. We have reported that IL-1beta is elevated in pulmonary edema fluid in those with ALI and mediates an autocrine-acting, fibroblast mitogenic pathway. In this study, we examine the role of IL-1beta-mediated induction of cyclooxygenase-2 and PGE2, and evaluate the significance of individual E prostanoid (EP) receptors in mediating the fibroproliferative effects of IL-1beta in ALI. Blocking studies on human lung fibroblasts indicate that IL-1beta is the major cyclooxygenase-2 mRNA and PGE2-inducing factor in pulmonary edema fluid and accounts for the differential PGE2 induction noted in samples from ALI patients. Surprisingly, we found that PGE2 produced by IL-1beta-stimulated fibroblasts enhances fibroblast proliferation. Further studies revealed that the effect of fibroblast proliferation is biphasic, with the promitogenic effect of PGE2 noted at concentrations close to that detected in pulmonary edema fluid from ALI patients. The suppressive effects of PGE2 were mimicked by the EP2-selective receptor agonist, butaprost, by cAMP activation, and were lost in murine lung fibroblasts that lack EP2. Conversely, the promitogenic effects of mid-range concentrations of PGE2 were mimicked by the EP3-selective agent, sulprostone, by cAMP reduction, and lost upon inhibition of Gi-mediated signaling with pertussis toxin. Taken together, these data demonstrate that PGE2 can stimulate or inhibit fibroblast proliferation at clinically relevant concentrations, via preferential signaling through EP3 or EP2 receptors, respectively. Such mechanisms may drive the fibroproliferative response to ALI.     

Direct evidence for the presence of selective binding sites for (3H) prostaglandin E2 on rat peritoneal macrophages

A method is presented which provides for a simple and rapid determination of PGE2 receptors on viable peritoneal macrophages. Incubation of the harvested cells with (3H)PGE2 revealed specific binding of (3H)PGE2 by use of the Millipore filter assay system. Maximum binding was attained in the presence of 1 mM EDTA. Specific binding was saturable at 65 fmol/mg protein with an equilibrium dissociation constant (Kd) of 3.2 X 10(-8)M. Inhibition of (3H)PGE2 binding with unlabelled prostaglandins revealed a potency series of PGE2 greater than PGE2 greater than PGI2. The PGE2 concentration which displaced 50% of the labelled ligand was 10(-7)M. Comparable kinetic data were obtained for adenylate cyclase stimulation, since the concentration which showed a halfmaximal stimulation of cAMP production was 2 X 10(-7)M of PGE2. Since PGE1 and PGI2 compete with (3H)PGE2 binding in a non-parallel manner compared to PGE2 itself, it is proposed that macrophages possess different types of PG receptors.     

Binding of (3H)prostaglandin E1 to putative receptors linked to adenylate cyclase of cultured cell clones.

A method for assessing the binding of 3H-labeled prostaglandin E1 ([3H]PGE1) to cell membranes has been developed and used to study the interaction of [3H]PGE1 with membranes from cultured mammalian cells. Receptor sites were identified by correlation of the potency of a series of compounds to compete for [3H]PGE1 binding sites and to stimulate adenylate cyclase activity, by correlation of rates of binding and change in enzyme activity, and by the correspondence of [3H]PGE1-binding activity with the presence or absence of PGE1-sensitive adenylate cyclase in several clones. In clone B82, a murine L-cell, [3H]PGE1 binds with an activation energy of 14 kcal/mol to a class of sites with an affinity of 0.5 X 10(8) M-1 and a capacity of 150 fmol/mg of protein. Concentration dependence of adenylate cyclase activation by PGE1 (KD =30 nM) and kinetic analysis of [3H]PGE1 binding (k1 = 4 X 10(6) liters/mol/min, k-1 0.15/min) verify this affinity. Concentration dependence and specificity of binding and activation of adenylate cyclases in neuroblastoma clone N4TG1 and N18TG2 substantiate the method. In several clones that lack PGE1-responsive adenylate cyclase, no specific [3H]PGE1 binding is detectable.     

Predominance of locally-produced prostaglandin E2 over prostacyclin in skin incisions in man

Blood was collected from skin incisions made by the 'Simplate' technique in 8 healthy men. Prostaglandin (PG) E2, but not 6-oxo-PGF1 alpha, the stable hydrolysis product of prostacyclin (PGI2), was tentatively identified using capillary column gas chromatography/electron capture mass spectrometry. It was not possible to quantify PGE2 because of the small volumes of blood generated by this method. Larger blood samples were then collected from 22 skin incisions in 13 patients undergoing cardiothoracic surgery. Concentrations of PGE2 were substantially greater than 6-oxo-PGF1 alpha in every sample. 13,14-Dihydro-15-oxo-PGF2 alpha, a pulmonary metabolite of PGE2, was not elevated, indicating that the PGE2 was synthesised locally at the site of incision. In 6 further patients undergoing cardiac surgery, blood sampled from an antecubital vein before and during operation contained little or no PGE2. We conclude that a substantial proportion of the PGE2 in blood emerging from skin incisions may be formed locally by the traumatised microvessels, consistent with the hypothesis that PGE2 is the principle prostaglandin synthesised by human cutaneous microvessels in vivo.     

Low intensity pulsed ultrasound exposure increases prostaglandin E2 production via the induction of cyclooxygenase-2 mRNA in mouse osteoblasts

Both prostaglandin E2 (PGE2) and low intensity pulsed ultrasound (U/S) exposure have been reported to accelerate fracture repair. We hypothesized that these two pathways are interactive. To verify this hypothesis, we examined the regulation of PGE2 production by U/S exposure (sine wave of 1.5MHz repeating at 1kHz, 30mW/cm2, 20 minutes) in mouse osteoblastic cell line, MC3T3-E1. The production of PGE2 in osteoblasts was augmented by U/S, which was threefold at 60 min. in comparison with unexposed samples. Then we evaluated the expression of cyclooxygenase-2 (COX-2) mRNA, which is a critical enzyme for PGE2 production. U/S rapidly up-regulated the expression of COX-2 mRNA in a time dependent manner. In addition, PGE2 production by U/S was drastically suppressed by a selective inhibitor of COX-2. These results provide strong evidence that PGE2 production in osteoblasts is dependent upon the induction of COX-2 mRNA expression by U/S and offer a mechanistic insight into how U/S accelerates fracture repair.     

Increased and intermittent prostaglandin release from amnion detected by a new superfusion technique for full thickness fetal membrane

Prostaglandin E2 (PGE) and F2 alpha (PGF) release by the intact fetal membranes is described using a novel superfusion technique allowing for the independent assessment of prostaglandin release from the amnion and chorio-decidua whilst maintaining the anatomical integrity of the fetal membranes. The effect of labour on prostaglandin release is described. Using this system it was confirmed that the amnion is a major site of prostaglandin release and possibly production. Labour resulted in a significant increase of both PGE and PGF release from the amnion side only (Pre-labour: PGE 918 pg/cm2/3h, PGF 370 pg/cm2/3h; Labour: PGE 2993 pg/cm2/3h, PGF 662 pg/cm2/3h). No change in either PGE or PGF release from the chorio-decidual side was observed in relation to labour. In addition a change in the pattern of prostaglandin release from the amnion was observed in tissues obtained after the onset of labour. In 6 of 8 samples obtained after spontaneous labour an intermittent or pulsatile release of both PGE and PGF was observed from the amnion side as compared to the steady state of prostaglandin release from all 10 samples obtained before labour.     

Prostaglandin E receptor subtypes involved in stimulation of gastroduodenal bicarbonate secretion in rats and mice.

We investigated prostaglandin E (EP) receptor subtypes responsible for the HCO3- stimulatory action of prostaglandin E2 (PGE2) in the gastroduodental mucosa, by examining the effects of various prostanoids with subtype specific EP receptor agonists in rats and those of PGE2 in knockout mice lacking EP1 or EP3 receptors. In rats, gastric HCO3- secretion was stimulated by i.v. administration of PGE2, 17-phenyl PGE2 the selective EP1 agonist as well as sulprostone the EP1 and EP3 agonist, but was not affected by other EP agonists such as butaprost the selective EP2 agonist, ONO-NT-012 the selective EP3 agonist or 11-deoxy PGE1 the EP3 and EP4 agonist. In contrast, the HCO3- secretion in rat duodenums was stimulated by PGE2, sulprostone, ONO-NT-012 as well as 11-deoxy PGE1 but not affected by either 17-phenyl PGE2 or butaprost. The HCO stimulatory effect of sulprostone in the stomach was significantly inhibited by ONO-AE-829, the selective EP1 antagonist. On the other hand, PGE2 applied topically to the mucosa for 10 min caused a dose-dependent increase of HCO3- secretion in both the stomach and duodenum of wild-type mice. The HCO3- stimulatory action of PGE2 in the stomach was also observed dose-dependently in knockout mice lacking EP3-receptors but was absent in EP1-receptor knockout mice, while the stimulatory effect in the duodenum was observed in EP1-receptor knockout mice, similar to wild-type animals, but not in knockout mice lacking EP3-receptors. These results indicate that PGE2 stimulates HCO3- secretion via different EP receptor subtypes in the stomach and duodenum; the former is mediated by EP1-receptors, while the latter mediated by EP3-receptors.