9-Hydroxyprostaglandin dehydrogenase in rat kidney cortex converts prostaglandin I2 into 15-keto-13,14-dihydro 6-ketoprostaglandin E1

15-Keto-13,14-dihydro 6-ketoprostaglandin E1 was positively identified by gas chromatography-mass spectrometry with negative-ion chemical ionisation detection from samples of rat kidney high-speed supernatant incubated with prostaglandin I2 in the presence of NAD+. A decreased formation of this product was observed when NAD+ was substituted with NADP+ and none was observed in the absence of nucleotide or substrate prostaglandin I2. Experiments with [9 beta-3H]prostaglandin I2 showed a time- and concentration-dependent loss of tritium which appeared as tritiated water, typical of reaction of [9 beta-3H]prostaglandin substrates with the enzyme, 9-hydroxyprostaglandin dehydrogenase. Time-course measurements of the appearance of tritiated water showed similar rates with 6-keto[9 beta-3H]prostaglandin F1 alpha and 15-keto-13,14-dihydro 6-keto[9 beta-3H]prostaglandin F1 alpha as substrates. These experiments suggest that the transformation of prostaglandin I2 and 6-ketoprostaglandin F1 alpha into the 15-keto-13,14-dihydro 6-ketoprostaglandin E1 catabolite occurs in this in vitro preparation via the corresponding 15-keto-13,14-dihydro catabolite of 6-ketoprostaglandin F1 alpha.     

Prostaglandin-E2 9-ketoreductase from human uterine decidua vera

Prostaglandin-E2 9-ketoreductase, the enzyme which catalyzes the reaction from prostaglandin E2 (PGE2) to prostaglandin F2 alpha (PGF2 alpha), has been purified 232-fold from human uterine decidua vera. The molecular mass of the enzyme, as estimated by fast protein liquid chromatography, was 29 kDa. Sodium dodecyl sulfate disc gel electrophoresis of the denatured enzyme revealed a molecular mass of 31 kDa. These data suggest that the enzyme consists of a single polypeptide chain. The rate equation of the enzyme reaction for two substrates was used for the determination of five kinetic constants. The equilibrium constant with respect to PGE2 was 83 microM, the Michaelis constant, Km, for PGE2 was 93 microM. For NADPH, the equilibrium constant was 1.0 microM and Km was 1.6 microM. The maximal velocity for the forward reaction was V1 = 217 pmol/min. The inhibition constants for the analgesic agents indomethacin and fentiazac were Ki = 850 microM and Ki = 450 microM and for the steroid progesterone Ki = 1.5 mM, respectively. Prostaglandin-E2 9-ketoreductase might be responsible for the control of the PGE2/PGF2 alpha ratio in human decidua vera. The enzyme, therefore, might be an important factor in the cascade of events leading to uterine contractions and parturition.     

Prostaglandin 15-hydroxy dehydrogenase from human placenta.

The enzyme system prostaglandin 15-hydroxy dehydrogenase, which catalyzes the inactivation of all biologically active prostaglandins, has been purified 1270-fold from human placenta. Kinetic studies on the enzyme have provided information on a well-organized control mechanism to avoid prostaglandin accumulation and for a fast prostaglandin degradation. 15-Ketoprostaglandin E2 and 13,14-dihydro-15-ketoprostaglandin E2 inhibit prostaglandin 15-hydroxy dehydrogenase non-competitively with respect to prostaglandin E2. The rate equation of enzyme reaction for two substrates was used for determination of the equilibrium constant and Michaelis constants of the enzyme. The following kinetic constants for prostaglandin 15-hydroxy dehydrogenase have been found. The equilibrium constant with repect to prostaglandin E2 is 18 muM, the Michaelis constant Km for prostaglandin E2 is 1 muM for NAD+ 44muM. The inhibition constants for 15-ketoprostaglandin E2 ar Ki(slope) = 70 muM, Ki(intercept) = 150 muM, and for 13,14-dihydro-15-ketoprostaglandin E2 Ki(slope) = 80 muM, and Ki(intercept) = 150 muM. The maximal velocity for the forward reaction is V1 = 0.45 mumol/min. These kinetic data exclude a random or ping-pong mechanism, and also a Theorell-Chance type as suggested by Braithwaite and Jarabak. We propose, therefore, a sequential ordered mechanism. The isoelectric point for prostaglandin 15-hydroxy dehydrogenase is at pH 5.35, judged by isoelectric focusing.     

Prostaglandin synthesis and catabolism in the gastric mucosa: studies in normal rabbits and rabbits immunized with prostaglandin E2

Antral and fundic mucosal homogenates obtained from prostaglandin E2-immunized rabbits converted 14C-arachidonic acid to prostaglandin E2, 6-keto prostaglandin F1 alpha, prostaglandin F2 alpha, and prostaglandin D2. Percentage conversion of 14C-arachidonic acid to these prostaglandin products was not significantly different in prostaglandin E2-immunized rabbits compared with control rabbits (thyroglobulin-immunized and unimmunized rabbits combined). Synthesis of 6-keto prostaglandin F1 alpha, prostaglandin E2 and 13,14-dihydro 15-keto prostaglandin E2 from endogenous arachidonic acid after vortex mixing fundic mucosal homogenates was similar in prostaglandin E2 immunized rabbits and control rabbits. Both in prostaglandin E2-immunized rabbits and controls, 3H-prostaglandin E2 was catabolized extensively by the fundic mucosa, whereas 3H-6-keto prostaglandin F1 alpha, 3H-prostaglandin F2 alpha, and 3H-prostaglandin D2 were not catabolized to any appreciable extent. The rate of catabolism of PGs was not significantly different in prostaglandin E2-immunized rabbits and control rabbits, with the exception of prostaglandin F2 alpha which was catabolized slightly more rapidly in prostaglandin E2-immunized rabbits. These results indicate that development of gastric ulcers in prostaglandin E2-immunized rabbits is not associated with an alteration in the capacity of the gastric mucosa to synthesize or catabolize prostaglandins.     

Plasma prostaglandin E2 level in Kawasaki disease

Plasma levels of prostaglandin E2 and prostaglandin F2 alpha were determined in 15 patients in the acute and recovery stages of Kawasaki disease, 10 patients with anaphylactoid purpura, 16 with bacterial and viral infections and 10 healthy children. Plasma levels of prostaglandin E2 were markedly increased in the acute stage of Kawasaki disease, and these levels were decreased in the recovery stage. The prostaglandin F2 alpha/prostaglandin E2 ratio in the acute stage of Kawasaki disease was markedly decreased. Plasma levels of prostaglandin E2 in patients with anaphylactoid purpura, bacterial and viral infections were within the normal range. In Kawasaki disease which is associated with systemic vasculitis with a severe inflammatory reaction, prostaglandin E2 is considered to be more selectively produced and released than prostaglandin F2 alpha, suggesting that prostaglandin E2 plays an important role in the immunological and inflammatory reaction.     

Prostaglandin 9-hydroxydehydrogenase activity in the adult rat kidney. Identification, assay, pathway, and some enzyme properties.

Prostaglandin F2alpha is converted to 15-keto-13,14-dihydroprostaglandin E2 by adult rat kidney homogenates. A variety of substrates labeled as either the 9beta position alone or at several other positions in the prostaglandin molecule were used to define the step at which the crossover from the F type to the E type prostaglandins takes place. Time course studies further confirmed that 15-keto-13,14-dihydroprostaglandin F2alpha is the immediate substrate for this enzyme which we have termed prostaglandin 9-hydroxydehydrogenase. An assay system based on specific loss of tritium from 9beta-tritiated prostaglandin F2alpha is described. Enzyme activity with prostaglandin F2alpha as substrate is linear with time up to 10 min, stimulated by NAD+, saturable at low concentrations of substrate, stable to storage at minus 25 degrees in phosphate buffer (up to 3 weeks), and has a broad pH optimum around 7.5. The product, 15-keto,13,14-dihydroprostaglandin E2 was identified by mass spectrometry through a sodium borohydride-sodium borodeuteride reduction method.     

Metabolism of prostaglandins, prostaglandin analogs and thromboxane B2 by lung and liver microsomes from pregnant rabbits

Liver microsomes from pregnant rabbits converted prostaglandins F2 alpha, E1, and E2 to their 20-hydroxy metabolites along with smaller amounts of the corresponding 19-hydroxy compounds. Prostaglandins E1 and E2 were also reduced to prostaglandins F1 alpha and F2 alpha, respectively, and prostaglandin E1 was isomerized to 8-isoprostaglandin E1. The above products were also identified after incubation of prostaglandins with liver microsomes from non-pregnant rabbits. In this case, the yield of 20-hydroxy metabolites was much lower. Thromboxane B2 and a number of prostaglandin F2 alpha analogs were also hydroxylated by lung and liver microsomes from pregnant rabbits. The relative rates of hydroxylation by lung microsomes were: prostaglandin E2 approximately prostaglandin F2 alpha approximately 16,16-dimethylprostaglandin F2 alpha approximately 13,14-didehydroprostaglandin F2 alpha greater than thromboxane B2 greater than 15-methylprostaglandin F2 alpha approximately 17-phenyl-18,19,-20-trinorprostaglandin F2 alpha approximately ent-13,14-didehydro-15-epiprostaglandin F2 alpha. Similar results were obtained with liver microsomes except that thromboxane B2 was a relatively poorer substrate for hydroxylation.     

Catabolism of an isolated, purified intermediate of prostaglandin biosynthesis by regions of the adult rat kidney.

1. A heat labile, cold-stable, stannous chloride-reducible intermediate of prostaglandin biosynthesis was formed in good yield (greater than 60%) from 3H-labeled arachidonic acid during brief incubations (30--90 s, 37 degrees C) with sheep seminal vesicle microsomes in the presence of p-hydroxymercuribenzoate (4 mM). This intermediate appears to have properties similar to one of the endoperoxides (15-hydroxyprostaglandin-9,11-endoperoxide) recently isolated by Hamberg and Samuelsson (Proc. Natl. Acad. Sci. U.S. (1973) 70, 889-903) AND Nugteren and Hazelhof (Biochem. Biophys. Acta. (1973) 326, 448-461). 2. Treatment of the purified intermediate with homogenates of rat kidney cortex, medulla and papilla resulted within 2 min (37 degrees C) in complete conversion into several compounds including prostaglandins E2 and F2alpha. The main product (40-50% yield formed by papilla homogenates was prostaglandin E2. The conversion into prostaglandin E2 was largely abolished by previous bo9ling of the homogenate whereas the conversion into prostaglandin F2alpha was not. The intermediate was stable in buffer for the same period of incubation. 3. The ratio of tritiated prostaglandins E2: F2alpha obtained were: papilla, 1.90; medulla, 0.76; cortex, 0.48. 4. These observations indicate that both types of prostaglandins can be formed by all three regions of the rat kidney and that regional differences exist in the proportion of E2 : F2alpha that is formed. Whereas prostaglandin E2 is mostly formed by an enzymatic process, prostaglandin F2alpha is not.     

Properties of prostaglandin synthetase of rabbit kidney medulla.

The formation in vitro of prostaglandins E2, D2, and F2alpha from arachidonic acid by rabbit kidney medulla homogenate or microsomal fraction is markedly affected by the composition of the incubation medium employed. Optimal biosynthesis is obtained in 0.1 M potassium phosphate buffer, with the optimum pH being 8.0--8.8. Under these conditions prostaglandin formation is linear up to arachidonic acid concentration of 30 muM. The initial rate of formation of prostaglandin E2 + prostaglandin D2 is 3--4 times higher than that of prostaglandin F2alpha. Reduced glutathione (1 mM) did not affect the biosynthesis by medulla homogenate and produced only small stimulation of the biosynthesis by microsomal powder. Hydroquinone produced a small stimulation at a low concentration of 0.005 mM, and a strong inhibition at concentrations of 0.1 mM or higher. Addition of bovine serum albumin (0.1%) reduced the microsomal biosynthesis of prostaglandins by approximately 80%. Addition of boiled homogenate or boiled 140 000 X g supernatant produced small stimulation of microsomal biosynthesis while 140 000 X g supernatant (not boiled) caused small inhibition which was not dose-related. It appears that rabbit kidney prostaglandin-synthetase converts arachidonic acid to prostaglandins E2 and F2alpha in comparable amounts, without apparent need for a cytoplasmic soluble cofactor or specific reducing agents.     

On the metabolism of prostaglandins by human gastric fundus mucosa.

1.Specific radioimmunoassays for the prostaglandins E2, F2alpha and A2 and the metabolites 13,14-dihydro-15-keto-prostaglandin E2, 15-keto-prostaglandin F2alpha and 13,14-dihydro-15-keto-prostaglandin F2alpha were used to study the metabolism of prostaglandins by gastroscopically obtained small biopsy specimens of human gastric fundus mucosa. 2.Three prostaglandin-metabolizing enzymes were found in the 100 000 X g supernatant of human gastric fundus mucosa, 15-hydroxy-prostaglandin-dehydrogenase, delta13-reductase and delta9-reductase. The specific activity was highest for 15-hydroxy-prostaglandin-dehydrogenase and lowest for delta9-reductase. 3.Formation of prostaglandin A2 (or B2) was not observed under the same conditions. 4.None of the three enzyme activities detected in the 100 000 X g supernatant was found in the 10 000 X g and 100 000 X g pellets of human gastric fundus mucosa. 5.The results indicate that high speed supernatant derived from human gastric mucosa can rapidly metabolize prostaglandin E2 and prostaglandin F2alpha to the 15-keto and 13,14-dihydro-15-keto-derivatives. Furthermore, prostaglandin E2 can be converted to prostaglandin F2alpha, the biological activity of which, on gastric functions, differs from that of prostaglandin E2.     

Prostaglandin biosynthesis and catabolism in the lamb ductus arteriosus, aorta and pulmonary artery.

Homogenates of tissues from fetal and neonatal lamb ductus arteriosus, aorta and pulmonary artery have the capacity to convert arachidonic acid as well as the intermediate prostaglandin endoperoxide, prostaglandin H2, into three products: prostaglandins E2, F2alpha and a major product 6-ketoprostaglandin F1alpha. The three tissues also displayed prostaglandin 15-hydroxydehydrogenase and 13-reductase catabolic activities. The catabolishing system showed considerable substrate specificity: prostaglandin E1 was a good substrate whereas prostaglandins F1alpha and F2alpha were completely devoid of catabolism. The complete system was observed in immature as well as mature arterial vessels, in the fetus as well as the neonate (up to 7 days old). These experiments demonstrate the presence of several components of the prostaglandin system in these tissues and offer biochemical evidence for the implication of prostaglandins E2 and I2 in the maintenance of the ductus and neighboring vessels in a relaxed state in the fetus.     

Rat renal medulla possess high capacity to catabolize prostaglandins

Prostaglandin E2 is converted to 15-keto-13,14 dihydro prostaglandin E2,15-keto-prostaglandin F2 alpha and 15-keto-13,14 dihydro prostaglandin F2 alpha, by supernatants from rat kidney medulla. The main pathway for prostaglandin E2 inactivation is the combined action of 15 hydroxy dehydrogenase and delta 13 reductase enzymes. 9-Keto-reductase route constitutes a minor pathway. Prostaglandin F2 alpha is converted into 15-keto-prostaglandin F2 alpha, 15-keto-13, 14 dihydro prostaglandin F2 alpha and 15-keto-dihydro prostaglandin E2. Enzyme activities are time and substrate-concentration dependent. In the presence of an excess of substrate, rat renal medulla inactivates 40 and 56 times more prostaglandin E2 and prostaglandin F2 alpha, respectively, than the amount which is released under basal conditions. These results are in contrast to the generally accepted concept that the kidney cortex is the sole site of renal prostaglandin catabolism, and suggest, for the first time, that rat renal medulla may be a key site for the modulation of prostaglandin levels in the kidney.     

Uterine vein prostaglandin levels in late pregnant dogs.

We studied the uterine venous plasma concentrations of prostaglandins E2, F2 alpha, 15 keto 13,14 dihydro E2 and 15 keto 13,14 dihydro F2 alpha in late pregnant dogs in order to evaluate the rates of production and metabolism of prostaglandin E2 and F2 alpha in pregnancy in vivo. We used a very specific and sensitive gas chromatography-mass spectrometry assay to measure these prostaglandins. The uterine venous concentrations of prostaglandin E2 and 15 keto 13,14 dihydro E2 were 1.35 +/- .27 ng/ml and 1.89 +/- .37 ng/ml, respectively; however, we could not find any prostaglandin F2 alpha and very little of its plasma metabolite in uterine venous plasma. Since uterine microsomes can generate prostaglandin F2 alpha and E2 from endoperoxides, prostaglandin F2 alpha production in vivo must be regulated through an enzymatic step after endoperoxide formation. Prostaglandin E2 is produced by pregnant canine uterus in quantities high enough to have a biological effect in late pregnancy; however, prostaglandin F2 alpha does not appear to play a role at this stage of pregnancy.     

Prostaglandin receptor-adenylate cyclase system in plasma membranes of rat liver and ascites hepatomas, and the effect of GTP upon it.

1. Adenylate cyclase in plasma membranes from rat liver was stimulated by prostaglandin E1, and to a lesser extent by prostaglandin E2. Prostaglandin F1alpha and A1 did not stimulate the cyclase. The prostaglandin E1-mediated activation was found to require GTP when the substrate ATP concentration was reduced from 3 mM to 0.3 mM in the reaction mixture. Adenylate cyclase of the plasma membranes from rat ascites hepatomas AH-130 and AH-7974 was not stimulated by prostaglandin E1 in the presence or the absence of GTP, although the basal activity of adenylate cyclase as well as its stimulation by GTP alone were similar to normal liver plasma membranes. 2. Liver plasma membranes were found to have two specific binders for [3H] prostaglandin E1 with dissociation constants of 17.6-10(-9) M and 13.6-10(8) M (37 degrees C) and one specific binder for [3H]prostaglandin F2alpha with a dissociation constant of 2.31-10(8) M (37 degrees C). The specific binders for prostaglandin E1 could not be detected in the hepatoma plasma membranes. 3. Binding of [3H] prostaglandin E1 to the liver plasma membranes was exchange by, GTP dGPT, GDP, ATP and GMP-P(N)P, but not by GMP, CGMP, DTTP, UTP or CTP. The increase in the binding of [3H] prostaglandin E1 was found to be due to the increased affinity of the specific binders to prostaglandin F2alpha was not affected by GTP. 4. GTP alone was found to increase V of adenylate cyclase of liver plasma membranes, while GTP plus prostaglandin E1 was found to decrease Km of adenylate cyclase in addition to the increase of V to a further extent.     

Identification of metabolites from peroxisomal beta-oxidation of prostaglandins

We have recently shown that isolated rat liver peroxisomes can chain-shorten prostaglandin F2 alpha and prostaglandin E2 to tetranor-metabolites. In the present report dinor-metabolites of these two prostaglandins were also identified, suggesting that the peroxisomal chain-shortening reaction of prostaglandins is a beta-oxidation reaction. Furthermore, an intermediate containing an extra double bond was isolated from incubates of prostaglandin F2 alpha with peroxisomes. This intermediate was tentatively assigned the structure 2,3-dehydroprostaglandin F2 alpha. Prostaglandin E1 and a major circulating prostaglandin F2 alpha metabolite were also metabolized to chain-shortened products by peroxisomes. The accumulation of the 2,3-dehydro-metabolite and the dinor-metabolites suggest that the peroxisomal beta-oxidation sequence is not tightly coupled, in contrast to mitochondrial fatty acid oxidation.     

Altered glomerular eicosanoid biosynthesis in uranyl nitrate-induced acute renal failure

The present studies were designed (1) to examine the pattern of changes in eicosanoid biosynthesis in isolated rat glomeruli, and (2) to correlate these changes with the previously observed alterations in renal perfusion and glomerular filtration rate which occur after uranyl nitrate administration, a model of toxin-induced acute renal failure. In the first part of this study, the in vitro and the in vivo effects of two cyclooxygenase inhibitors were examined for their ability to inhibit rat glomerular eicosanoid biosynthesis. Inhibition of prostaglandin E2 and prostaglandin F2 alpha generation by 1 mM aspirin in vitro was 76 and 82%, respectively. Similar inhibitions of 85 and 72% of biosynthesis of the above-mentioned lipids by 0.1 mM indomethacin were also noted. Intraperitoneal administration of aspirin (150 mg/kg) resulted in a significant inhibition of 88% or greater of prostaglandin E2, prostaglandin F2 alpha, 6-keto-prostaglandin F2 alpha, and thromboxane B2 biosynthesis. These results indicated that the expected alterations produced under in vivo conditions were detectable by in vitro techniques used in this study. 24 h after the administration of uranyl nitrate (25 mg/kg), significant increases in the biosynthesis of prostaglandin E2 (124%) and prostaglandin F2 alpha (88%) were observed when compared to the control values. No significant changes in prostacyclin or thromboxane formation were noted at this time. A further increase in the biosynthesis of prostaglandin E2 (248%), prostaglandin F2 alpha (262%), and a significant increase in prostacyclin (120%), measured as 6-keto-prostaglandin F1 alpha, were noted at 48 h. No changes in thromboxane B2 biosynthesis were noted. It is concluded that these data are consistent with the hypothesis that the increased glomerular biosynthesis of vasodilator eicosanoids (i.e., prostaglandin E2 and prostacyclin) may play a significant role in the homeostatic regulation of renal perfusion and glomerular filtration after acute toxic injury to the kidney.     

Interactions between prostaglandin analogues and a receptor in bovine Corpora lutea. Correlation of dissociation constants with luteolytic potencies in hamsters.

The dissociation constants for the interactions between some prostaglandin analogues and a prostaglandin F2 receptor in bovine corpora lutea were determined. These values were compared to the antifertility potencies of these compounds in hamsters and the rates of metabolism by 15-hydro-syprostaglandin dehydrogenase. The most active analogues with regard to both affinity for the receptor and luteolytic potency were 17-phenyl-18, 19, 20-trinorprostaglandin F2alpha and 15-methylprostaglandin F2alpha. The alkyl side chain of prostaglandins could be modified considerably without altering the affinity for the receptor. In this way metabolism by 15-hydroxyprostaglandin dehydrogenase could be blocked. Some of these compounds -ad greatly increased luteolytic effects. Substitution of a phenyl group for the 3 terminal carbon units of the alkyl side chain of prostaglandins increased both the affinity for the receptor and the luteolytic activity in vivo. 7-oxa-13-prostynoic acid, an antagonist of the luteolytic effect of prostaglandin F2alpha in vivo was a weak competitive inhibitor of the interation between prostaglandin F2alpha and the receptor.     

Radioimmunologic identification of prostaglandins produced by serum-stimulated mouse embryo palate mesenchyme cells

High-performance liquid chromatography and radioimmunoassay were used to identify the prostaglandins synthesized by mouse embryo palate mesenchyme cells. Serum stimulated the release of several different metabolites of arachidonic acid including 6-ketoprostaglandin F1 alpha (the stable product of prostacyclin, prostaglandin I2), prostaglandin E2 and prostaglandin F2 alpha. Compared to control cells, the serum-stimulated cells produce elevated levels of prostaglandin E2 (36-fold), 6-ketoprostaglandin F1 alpha (15-fold) and prostaglandin F2 alpha (7-fold). The acetylenic analogue of arachidonic acid, 5,8,11,14-eicosatetraynoic acid prevented this accelerated synthesis.     

Formation of 20-hydroxyprostaglandins by lungs of pregnant rabbits

Homogenates or particulate fractions (1,000 to 100,000 X g) from lungs of pregnant rabbits were incubated with prostaglandins or prostaglandin metabolites and the products were purified by chromatography and identified by gas chromatography-mass spectrometry. In the presence of NADPH, particulate fractions from pregnant rabbit lungs converted prostaglandins E1, E2, and F2alpha as well as 13,14-dihydro-15-oxoprostaglandin E2 and 13, 14-dihydro-15-oxoprostaglandin F2alpha to their 20-hydroxy derivatives. In the cases of the 3 primary prostaglandins, the corresponding omega-carboxylic acids were also isolated. The omega-hydroxylation reaction occurred in the presence of the microsomal fraction. The mitochondrial fraction was much less active whereas the cytosol fraction converted prostaglandins to their 13, 14-dihydro-15-oxo derivatives. When prostaglandin F2alpha was incubated with homogenates of lungs from pregnant rabbits, omega-oxidation was combined with oxidation of the 15-hydroxyl group and reduction of the 13, 14-double bond to give 13, 14-dihydro-20-hydroxy-15-oxoprostaglandin F2alpha as well as the corresponding derivative with an omega-carboxylic acid group. Lungs from nonpregnant rabbits were much less active than lungs from pregnant rabbits in the omega-oxidation of prostaglandins.     

Metabolism of prostaglandins D2 and F2 alpha in primary cultures of rat hepatocytes

3H-Labeled prostaglandins D2 and F2 alpha rapidly degraded to more-polar metabolites in primary cultured rat hepatocytes. The metabolites of prostaglandins D2 and F2 alpha accumulated in the culture medium. The metabolites extracted by ethyl acetate at pH 3 were purified by silicic acid column and thin-layer chromatography of silica gel, and were analysed by gas chromatography-mass spectrometry. The major metabolites from prostaglandin D2 were identified as dinor-prostaglandin D1 (7 alpha,13-dihydroxy-9-ketodinorprost-11-enoic acid) and tetranor-prostaglandin D1 (5 alpha,11- dihydroxy-7-ketotetranorprost-9-enoic acid). Those from prostaglandin F2 alpha were identified as dinor-prostaglandin F1 alpha (7 alpha,9 alpha,13-trihydroxydinorprost-11-enoic acid), tetranor-prostaglandin F1 alpha (5 alpha,7 alpha,11-trihydroxytetranorprost-9-enoic acid) and 9 alpha,11 alpha,15-trihydroxyprost-13-ene-1,20-dioic acid. These data indicate that prostaglandins D2 and F2 alpha mainly degraded by beta-oxidation, which is the same process as reported earlier for prostaglandins E1 and E2, and that prostaglandin F2 alpha was also subjected to omega-oxidation.