Functional and immunological relationships between metyrapone reductase from mouse liver microsomes and 3 alpha-hydroxysteroid dehydrogenase from Pseudomonas testosteroni.

3 Alpha-hydroxysteroid dehydrogenase (3 alpha-HSD) from Pseudomonas testosteroni was shown to reduce the xenobiotic carbonyl compound metyrapone (MPON). Reversely, MPON reductase purified from mouse liver microsomes and previously characterized as aldehyde reductase, was competitively inhibited by 3 alpha-HSD steroid substrates. For MPON reduction both enzymes can use either NADH or NADPH as co-substrate. Immunoblot analysis after native and SDS gel electrophoresis of 3 alpha-HSD gave a specific crossreaction with the antibodies against the microsomal mouse liver MPON reductase pointing to structural homologies between these enzymes. In conclusion, there seem to exist structural as well as functional relationships between a mammalian liver aldehyde reductase and prokaryotic 3 alpha-HSD. Moreover, based on the molecular weights and the co-substrate specificities microsomal mouse liver MPON reductase and Pseudomonas 3 alpha-HSD seem to be members of the short-chain alcohol dehydrogenase family.     

Functional expression, purification, and characterization of 3alpha-hydroxysteroid dehydrogenase/carbonyl reductase from Comamonas testosteroni

3alpha-Hydroxysteroid dehydrogenase (3alpha-HSD) catalyzes the oxidoreduction at carbon 3 of steroid hormones and is postulated to initiate the complete mineralization of the steroid nucleus to CO(2) and H(2)O in Comamonas testosteroni. By this activity, 3alpha-HSD provides the basis for C. testosteroni to grow on steroids as sole carbon and energy source. 3alpha-HSD was cloned and overexpressed in E. coli and purified to homogeneity by an affinity chromatography system as His-tagged protein. The recombinant enzyme was found to be functional as oxidoreductase toward a variety of steroid substrates, including androstanedione, 5alpha-dihydrotestosterone, androsterone, cholic acid, and the steroid antibiotic fusidic acid. The enzyme also catalyzes the carbonyl reduction of nonsteroidal aldehydes and ketones such as metyrapone, p-nitrobenzaldehyde and a novel insecticide (NKI 42255), and, based on this pluripotent substrate specificity, was named 3alpha-hydroxysteroid dehydrogenase/carbonyl reductase (3alpha-HSD/CR). It is suggested that 3alpha-HSD/CR contributes to important defense strategies of C. testosteroni against natural and synthetic toxicants. Antibodies were generated in rabbits against the entire 3alpha-HSD/CR protein, and may now be used for evaluating the pattern of steroid induction in C. testosteroni on the protein level. Upon gel permeation chromatography the purified enzyme elutes as a 49.4 kDa protein revealing for the first time the dimeric nature of 3alpha-HSD/CR of C. testosteroni.     

Lyophilized Clostridium perfringens 3 alpha- and Clostridium bifermentans 7 alpha-hydroxysteroid dehydrogenases: two new stable enzyme preparations for routine bile acid analysis

Preparations of 3 alpha-hydroxysteroid dehydrogenase (EC 1.1.1.50) from Clostridium perfringens were successfully lyophilized into a stable powder form. Purification of the enzyme was achieved using triazine dye affinity chromatography. C. perfringens 3 alpha-hydroxysteroid dehydrogenase was purified 24-fold using Reactive Red 120 (Procion Red) -cross-linked agarose (70% yield). Quantitative measurement of bile acids with the purified enzymes, 3 alpha-hydroxysteroid dehydrogenase and 7 alpha-hydroxysteroid dehydrogenase (EC 1.1.1.159) from Clostridium bifermentans (strain F-6), was achieved spectrophotometrically. Standard curves with chenodeoxycholic acid (CDC) and cholic acid were linear within a concentration range of 20-100 microM. Analysis of mixtures of ursodeoxycholic acid and CDC showed the additive nature of the 3 alpha-hydroxysteroid dehydrogenase and showed also that 7 alpha-hydroxyl groups were independently quantified by the 7 alpha-hydroxysteroid dehydrogenase. Bile acids in Folch extracts of human bile samples were measured using purified preparations of Pseudomonas testosteroni 3 alpha-hydroxysteroid dehydrogenase, C. perfringens 3 alpha-hydroxysteroid dehydrogenase, Escherichia coli 7 alpha-hydroxysteroid dehydrogenase and C. bifermentans (strain F-6) 7 alpha-hydroxysteroid dehydrogenase. Statistical comparison validated the use of C. perfringens 3 alpha- and C. bifermentans 7 alpha-hydroxysteroid dehydrogenases for the quantification of bile acids in bile.     

Identification of two dihydrodiol dehydrogenases associated with 3(17)alpha-hydroxysteroid dehydrogenase activity in mouse kidney

Dihydrodiol dehydrogenase activity was detected in the cytosol of various mouse tissues, among which kidney exhibited high specific activity comparable to the value for liver. The enzyme activity in the kidney cytosol was resolved into one major and three minor peaks by Q-Sepharose chromatography: one minor form cross-reacted immunologically with hepatic 3 alpha-hydroxysteroid dehydrogenase and another with aldehyde reductase. The other minor form was partially purified and the major form was purified to homogeneity. These two forms, although different in their charges, were monomeric proteins with the same molecular weight of 39,000 and had similar catalytic properties. They oxidized cis-benzene dihydrodiol and alicyclic alcohols as well as trans-dihydrodiols of benzene and naphthalene in the presence of NADP+ or NAD+, and reduced several xenobiotic aldehydes and ketones with NAD(P)H as a cofactor. The enzymes also catalyzed the oxidation of 3 alpha-hydroxysteroids and epitestosterone, and the reduction of 3- and 17-ketosteroids, showing much lower Km values (10(-7)-10(-6) M) for the steroids than for the xenobiotic alcohols. The results of mixed substrate experiments, heat stability, and activity staining on polyacrylamide gel electrophoresis suggested that, in the two enzymes, both dihydrodiol dehydrogenase and 3(17)alpha-hydroxysteroid dehydrogenase activities reside on a single enzyme protein. Thus, dihydrodiol dehydrogenase existed in four forms in mouse kidney cytosol, and the two forms distinct from the hepatic enzymes may be identical to 3(17)alpha-hydroxysteroid dehydrogenases.     

The effects of metyrapone, chalcone epoxide, benzil, clotrimazole and related compounds on the activity of microsomal epoxide hydrolase in situ, in purified form and in reconstituted systems towards different substrates

The influence of metyrapone, chalcone epoxide, benzil and clotrimazole on the activity of microsomal epoxide hydrolase towards styrene oxide, benzo[a]pyrene 4,5-oxide, estroxide and androstene oxide was investigated. The studies were performed using liver microsomes from rats, rabbits, mice and humans; epoxide hydrolase purified from rat liver microsomes to apparent homogeneity; and the purified enzyme incorporated into liposomes composed of egg-yolk phosphatidylcholine or total rat liver microsomal lipids. All four effectors were found to activate the hydrolysis of styrene oxide by epoxide hydrolase in situ in rat liver microsomal membranes, in agreement with earlier findings. Epoxide hydrolase activity towards styrene oxide in liver microsomes from mouse, rabbit and man was also increased by all four effectors. The most striking effect was a 680% activation by clotrimazole in rat liver microsomes. However, none of the effectors activated microsomal epoxide hydrolase more than 50% when benzo[a]pyrene 4,5-oxide, estroxide or androstene oxide was used as substrate. Indeed, clotrimazole was found to inhibit microsomal epoxide hydrolase activity towards estroxide 30-50% and towards androstene oxide 60-90%. The effects of these four compounds were found to be virtually identical in the preparations from rats, rabbits, mice and humans. The effects of metyrapone, chalcone epoxide, benzil and clotrimazole on purified epoxide hydrolase were qualitatively the same as those on epoxide hydrolase in intact microsomes, but much smaller in magnitude. These effects were increased in magnitude only slightly by incorporation of the purified enzyme into liposomes made from egg-yolk phosphatidylcholine. However, when incorporation into liposomes composed of total microsomal lipids was performed, the effects seen were essentially of the same magnitude as with intact microsomes. When the extent of activation was plotted against effector concentration, three different patterns were found with different effectors. Activation of epoxide hydrolase activity towards styrene oxide by clotrimazole was found to be uncompetitive with the substrate and highly structure specific. On the other hand, inhibition of epoxide hydrolase activity towards androstene oxide by clotrimazole was found to be competitive in microsomes. It is concluded that the marked effects of these four modulators on microsomal epoxide hydrolase activity are due to an interaction with the enzyme protein itself, but that the presence of total microsomal phospholipids allows the maximal expression leading to similar degrees of modulation as those observed in intact microsomes.(ABSTRACT TRUNCATED AT 400 WORDS)     

Immunochemical characterization of NADPH-cytochrome P-450 reductase from Jerusalem artichoke and other higher plants.

Polyclonal antibodies were prepared against NADPH-cytochrome P-450 reductase purified from Jerusalem artichoke. These antibodies inhibited efficiently the NADPH-cytochrome c reductase activity of the purified enzyme, as well as of Jerusalem artichoke microsomes. Likewise, microsomal NADPH-dependent cytochrome P-450 mono-oxygenases (cinnamate and laurate hydroxylases) were efficiently inhibited. The antibodies were only slightly inhibitory toward microsomal NADH-cytochrome c reductase activity, but lowered NADH-dependent cytochrome P-450 mono-oxygenase activities. The Jerusalem artichoke NADPH-cytochrome P-450 reductase is characterized by its high Mr (82,000) as compared with the enzyme from animals (76,000-78,000). Western blot analysis revealed cross-reactivity of the Jerusalem artichoke reductase antibodies with microsomes from plants belonging to different families (monocotyledons and dicotyledons). All of the proteins recognized by the antibodies had an Mr of approx. 82,000. No cross-reaction was observed with microsomes from rat liver or Locusta migratoria midgut. The cross-reactivity generally paralleled well the inhibition of reductase activity: the enzyme from most higher plants tested was inhibited by the antibodies; whereas Gingko biloba, Euglena gracilis, yeast, rat liver and insect midgut activities were insensitive to the antibodies. These results point to structural differences, particularly at the active site, between the reductases from higher plants and the enzymes from phylogenetically distant plants and from animals.     

3(20)alpha-hydroxysteroid dehydrogenase activity of monkey liver indanol dehydrogenase

Homogeneous indanol dehydrogenase from monkey liver catalyzed the reversible conversion of 3 alpha- or 20 alpha-hydroxy groups of several bile acids and 5 beta-pregnanes to the corresponding 3- or 20-ketosteroids. The kcat values for the steroids determined at pH 7.4 were low, but the kcat/Km values for the 3-ketosteroids were comparable to or exceeded those for 1-indanol and xenobiotic carbonyl substrates. The enzyme transferred the 4-pro-R-hydrogen atom of NADPH to the 3 beta- or 20 beta-face of the ketosteroid substrate. Competitive inhibition of the hydroxysteroid dehydrogenase activity of the enzyme by medroxyprogesterone acetate, hexestrol, and 1,10-phenanthroline suggests that both 1-indanol and hydroxysteroid are oxidized at the same active site on the enzyme. The specific inhibitor of the enzyme, 1,10-phenanthroline, suppressed the 3 alpha-hydroxysteroid dehydrogenase activity in the crude extract of monkey liver by 50%. The results strongly suggest that indanol dehydrogenase acts as a 3(20)alpha-hydroxysteroid dehydrogenase in the metabolism of certain steroid hormones and bile acids.     

Characterization of NADP+: 3 beta-hydroxysteroid dehydrogenase from microsomes of rat liver

A NADP(+)-dependent 3 beta-hydroxysteroid dehydrogenase activity was localized in the microsomal fraction of rat liver. This enzyme was solubilized and separated completely from 3 alpha-hydroxysteroid dehydrogenase by Matrex red A column chromatography. Partially purified 3 beta-hydroxysteroid dehydrogenase catalyzed the oxidation and reduction between the 3 beta-hydroxyl and 3-ketonic group of steroids or bile acids having no double bond in the A/B ring, but was inactive toward 3 alpha-hydroxyl group. The enzyme required NADP+ for oxidation and NADPH for reduction. The activity was inhibited by p-chloromercuribenzoic acid or p-chloromercuribenzenesulfonic acid at the concentration of 10(-4) M. The molecular weight of the enzyme was estimated to be about 43,000 by Sephadex G-200 column chromatography. From these results, it is concluded that the enzyme is a new type of microsomal NADP+:3 beta-hydroxysteroid dehydrogenase.     

Microsomal cytochrome P450-dependent steroid metabolism in male sheep liver. Quantitative importance of 6 beta-hydroxylation and evidence for the involvement of a P450 from the IIIA subfamily in the pathway

The metabolism of testosterone (TEST), androstenedione (AD) and progesterone (PROG) was assessed in hepatic microsomal fractions from male sheep. Rates of total hydroxylation of each steroid were lower in sheep liver than in microsomes isolated from untreated male rat, guinea pig or human liver, 6 beta-Hydroxylation was the most important pathway of biotransformation of each of the three steroids (0.80, 0.89 and 0.43 nmol/min/mg protein for TEST, AD and PROG, respectively). Significant minor metabolites from TEST were the 2 beta-, 15 beta- and 15 alpha-alcohols (0.19, 0.22 and 0.17 nmol/min/mg microsomal protein, respectively). Apart from the 6 beta-hydroxysteroid, only the 21-hydroxy derivative was formed from PROG at a significant rate (0.27 nmol/min/mg protein). The 6 beta-alcohol was the only metabolite formed from AD at a rate greater than 0.1 nmol/min/mg protein. Antisera raised in rabbits to several rat hepatic microsomal P450s were assessed for their capacity to modulate sheep microsomal TEST hydroxylation. Anti-P450 IIIA isolated from phenobarbital-induced rat liver effectively inhibited TEST hydroxylation at the 2 beta-, 6 beta-, 15 alpha- and 15 beta-positions (by 31-56% when incubated with microsomes at a ratio of 5 mg IgG/mg protein). IgG raised against rat P450 IIC11 and IIB1 inhibited the formation of some of the minor hydroxysteroid metabolites but did not decrease the rate of TEST 6 beta-hydroxylation. Western immunoblot analysis confirmed the cross-reactivity of anti-rat P450 IIIA with an antigen in sheep hepatic microsomes; anti-IIC11 and anti-IIB1 exhibited only weak immunoreactivity with proteins in these fractions. Considered together, the present findings indicate that, as is the case in many mammalian species, 6 beta-hydroxylation is the principal steroid biotransformation pathway of male sheep liver. Evidence from immunoinhibition and Western immunoblot experiments strongly implicate the involvement of a P450 from the IIIA subfamily in ovine steroid 6 beta-hydroxylation.     

Purification of cholesterol 7 alpha-hydroxylase from human and rat liver and production of inhibiting polyclonal antibodies

Cholesterol 7 alpha-hydroxylase, the cytochrome P-450-dependent and rate-controlling enzyme of bile acid synthesis, was purified from rat and human liver microsomes. The purified fractions were assayed in a reconstituted system containing [4-14C]cholesterol, and cholesterol 7 alpha-hydroxylase activities in these fractions increased 500-600-fold relative to whole microsomes. Polyacrylamide gel electrophoresis of rat microsomes followed by immunoblotting with polyclonal rabbit antisera raised against purified cholesterol 7 alpha-hydroxylases revealed two peaks at molecular masses of 47,000 and 49,000 daltons for both rat and human fractions. Increasing amounts of rabbit anti-rat and anti-human antibodies progressively inhibited rat microsomal cholesterol 7 alpha-hydroxylase activity up to 80%. In contrast, monospecific antibodies raised against other purified cytochrome P-450 enzymes (P-450f, P-450g, and P-450j) did not inhibit rat or human cholesterol 7 alpha-hydroxylase activity. Immunoblots of rat microsomes with the rabbit anti-rat cholesterol 7 alpha-hydroxylase antibody demonstrated that the antibody reacted quantitatively with the rat microsomal enzyme. Microsomes from cholesterol-fed rats showed increased cholesterol 7 alpha-hydroxylase mass, whereas treatment with pravastatin, an inhibitor of hydroxy-methylglutaryl-coenzyme A reductase, reduced enzyme mass. Microsomes from starved rats contained slightly less cholesterol 7 alpha-hydroxylase protein than chow-fed control rats. These results indicate a similarity in molecular mass, structure, and antigenicity between rat and human cholesterol 7 alpha-hydroxylases; demonstrate the production of inhibiting anti-cholesterol 7 alpha-hydroxylase antibodies that can be used to measure the change in cholesterol 7 alpha-hydroxylase enzyme mass under various conditions; and emphasize the unique structure of cholesterol 7 alpha-hydroxylase with respect to other cytochrome P-450-dependent hydroxylases.     

Formation of (4R)- and (4S)-4-hydroxyochratoxin A from ochratoxin A by liver microsomes from various species.

Two metabolic products were formed from ochratoxin A by human, pig, and rat liver microsomal fractions in the presence of reduced nicotinamide adenine dinucleotide phosphate. They were isolated from the incubation mixture in the presence of pig liver microsomes by extraction, thin-layer chromatography, and high-pressure liquid chromatography Their structures are suggested to be (4R)- and (4S)-4-hydroxyochratoxin A on the basis of mass and nuclear magnetic resonance spectroscopy. Km and the maximum velocity for the formation of the two metabolites by human, pig, and rat microsomes were determined. Their formation was inhibited by carbon monoxide and metyrapone. The results indicate that the microsomal hydroxylation system is a cytochrome P-450 and that different species are involved in the formation of the two epimeric forms of 4-hydroxyochratoxin A.     

3alpha-, 7alpha- and 12alpha-hydroxysteroid dehydrogenase activities from Clostridium perfringens.

25 strains of Clostridium perfringens were screened for hydroxysteroid dehydrogenase activity; 19 contained NADP-dependent 3alpha-hydroxysteroid dehydrogenase and eight contained NAD-dependent 12alpha-hydroxysteroid dehydrogenase active against conjugated and unconjugated bile salts. All strains containing 12alpha-hydroxysteroid dehydrogenase also contained 3alpha-hydroxysteroid dehydrogenase although 12alpha-hydroxysteroid dehydrogenase was invariably in lesser quantity than the 3alpha-hydroxysteroid dehydrogenase. In addition, 7alpha-hydroxysteroid dehydrogenase activity was evident only when 3alpha, 7alpha, 12alpha-trihydroxy-5beta-cholanoate was substrate but notably absent when 3alpha, 7alpha-dihydroxy-5beta-cholanoate was substrate. The oxidation product 12alpha-hydroxy-3, 7-diketo-5beta-cholanoate is rapidly further degraded to an unknown compound devoid of either 3alpha- or 7alpha-OH groups. Group specificity of these enzymes was confirmed by thin-layer chromatography studies of the oxidation products. These enzyme systems appear to be constitutive rather than inducible. In contrast to C. perfringens. Clostridium paraputrificum (five strains tested) contained no measurable hydroxysteroid dehydrogenase activity. pH studies of the C. perfringens enzymes revealed a sharp pH optimum at pH 11.3 and 10.5 for the 3alpha-OH- and 12alpha-OH-oriented activities, respectively. Kinetic studies gave Km estimates of approx. 5 X 10(-5) and 8 X 10(-4) M with 3alpha, 7a-dihydroxy-5beta-cholanoate and 3alpha, 12alpha-dihydroxy-5beta-cholanoate as substrates for two respective enzymes. 3alpha-hydroxysteroid dehydrogenase was active against 3alpha-OH-containing steroids such as androsterone regardless of the sterochemistry of the 5H (Both A/B cis and A/B trans steroides were substrates). There was no activity against 3beta-OH-containing steroids. The 3alpha- and 12alpha-hydroxysteroid dehydrogenase activities, although differing in cofactor requirements cannot be distinguished by their appearance in the growth curve, their mobility on disc gel electrophoresis, elution volume on passage through Sephadex G-200 or heat inactivation studies.     

Strain differences in purified rat hepatic 3 alpha-hydroxysteroid UDP-glucuronosyltransferase.

Qualitative and quantitative differences of purified hepatic 3 alpha-hydroxysteroid UDP-glucuronosyltransferase were investigated in Wistar and Sprague-Dawley rats. Individual differences in the glucuronidation rate of androsterone and chenodeoxycholic acid were observed in hepatic microsomal fractions from Wistar but not Sprague-Dawley rats. No individual variation was observed in the glucuronidation of testosterone, p-nitrophenol or oestrone. The 3 alpha-hydroxysteroid UDP-glucuronosyltransferases from livers of Wistar and Sprague-Dawley rats were isolated and highly purified by using Chromatofocusing and affinity chromatography. The amount of 3 alpha-hydroxysteroid UDP-glucuronosyltransferase in the liver of Wistar rats exhibiting low rates for androsterone glucuronidation is about 10% or less than that found in hepatic microsomal fractions obtained from Wistar rats having high rates for androsterone glucuronidation. The apparent Km for androsterone with purified 3 alpha-hydroxysteroid UDP-glucuronosyltransferase from Wistar rats with high glucuronidation activity (6 microM) was not different from that observed for the enzyme purified from Sprague-Dawley animals, whereas that for the enzyme purified from Wistar rats with low glucuronidation activity was substantially higher (120 microM). Despite the differences in apparent Km values for androsterone, the apparent Km for UDP-glucuronic acid (0.3 mM) was not different in the different populations of rats.     

Hormonal regulation of male-specific 20beta-hydroxysteroid dehydrogenase with carbonyl reductase-like activity present in kidney microsomes of rats.

Progesterone, 17alpha-hydroxyprogesterone, cortisone and cortisol, which are C(21)-steroids with a ketone group at the 20-position, potently inhibited the activity of enzyme acetohexamide reductase (AHR) responsible for the reductive metabolism of acetohexamide in kidney microsomes of male rats. Furthermore, progesterone was a competitive inhibitor of AHR. In the case of progesterone usage as the substrate, 20beta-hydroxysteroid dehydrogenase (20beta-HSD) activity was much higher than 20alpha-hydroxysteroid dehydrogenase (20alpha-HSD) activity in kidney microsomes of male rats. These results indicate that AHR present in kidney microsomes of male rats, functions as 20beta-HSD with carbonyl reductase-like activity. In male rats, both testectomy and hypophysectomy decreased the renal microsomal 20beta-HSD activity, but the decreased enzyme activities were increased by the treatment with testosterone propionate (TP). We propose the possibility that TP treatment regulates the renal microsomal 20beta-HSD activity by acting directly on the kidney of male rats. This is supported from the fact that when TP was given to ovariectomized and hypophysectomized female rats, the male-specific 20beta-HSD activity was detected in their kidney microsomes.     

Inhibition of microsomal NAD(P)H oxidation by Triton X-100

The non-ionic detergent Triton X-100 is shown to inhibit the spontaneous oxidation of NAD(P)H associated with rat liver microsomes. Advantage of this observation is taken to measure different microsomal NAD(P)H-dependent oxidoreductase activities such as 3-alpha-hydroxysteroid dehydrogenase, dihydrodiol dehydrogenase and various xenobiotic oxidoreductases. This inhibition provides an easy method for the screening of the under-investigated microsomal oxidoreductive metabolism of xenobiotics.     

Evidence for glucose-6-phosphate transport in rat liver microsomes

The existence of glucose-6-phosphate transport across the liver microsomal membrane is still controversial. In this paper, we show that S3483, a chlorogenic acid derivative known to inhibit glucose-6-phosphatase in intact microsomes, caused the intravesicular accumulation of glucose-6-phosphate when the latter was produced by glucose-6-phosphatase from glucose and carbamoyl-phosphate. S3483 also inhibited the conversion of glucose-6-phosphate to 6-phosphogluconate occurring inside microsomes in the presence of electron acceptors (NADP or metyrapone). These data indicate that liver microsomal membranes contain a reversible glucose-6-phosphate transporter, which furnishes substrate not only to glucose-6-phosphatase, but also to hexose-6-phosphate dehydrogenase.     

Factors responsible for the formation of different N-alkylated porphyrins in rat liver microsomal systems exposed to norethindrone. The role of 3 alpha-hydroxysteroid dehydrogenase.

Incubation of rat liver microsomes with norethindrone and a NADPH-generating system leads to the formation of one N-alkylated porphyrin (green pigment, GP1). Administration of this steroid to male rats in vivo results in the formation of three more-polar green pigments (GP2, 3 and 4). A cytosolic protein (green-pigment converting protein) has been purified from rat liver that, when added to liver microsomal mixtures containing norethindrone (0.03 mM) and a NADPH-generating system, results in the formation of all four green pigments (GP1, 2, 3 and 4). Field-desorption mass spectrometry of the purified green pigments gave protonated molecules, [M + H]+, at m/z 905 for GP1, m/z 909 for GP2, m/z 925 for GP3 and 4. The Mr of the purified cytosolic protein on SDS/polyacrylamide-gel electrophoresis or gel filtration was 37000. Polyacrylamide-gel isoelectric focusing gave a pI value of 5.9. Antibodies raised in rabbits against this protein, after preincubation with rat liver cytosol, subsequently prevented the formation of the more-polar norethindrone-induced green pigments (GP2, 3 and 4). The purified protein in the presence of either NADH or NADPH catalysed the reduction of delta 4-ring-reduced norethindrone, 5 alpha-oestran-17 alpha-ethynyl-17 beta-ol-3-one and, with the appropriate cofactor, the oxidation and reduction of steroids lacking the ethynyl function, e.g. androsterone or dihydrotestosterone. Indomethacin inhibited the reduction of dihydrotestosterone by this protein with an I50 (concn. causing 50% inhibition) value of 4.9 microM. From its physical and enzymic properties it is concluded that green-pigment converting protein is the same as 3 alpha-hydroxysteroid dehydrogenase (EC 1.1.1.50).