Purification and Properties of 3a-Hydroxysteroid Dehydrogenase from ...

Vol. 260, No. 28, Issue of December 5, pp. 15266-15272,1985 Printed in U.S.A.

THEJOURNAL OF BIOLOGICAL CHEMISTRY Q 1985 by The American Society of Biological Chemists, Inc.

Purification and Propertiesof 3a-Hydroxysteroid Dehydrogenase from Rat BrainCytosol INHIBITION BY NONSTEROIDAL ANTI-INFLAMMATORY DRUGS AND PROGESTINS* (Received for publication, April 5, 1985)

Trevor M. Penning$, Robert B. Sharp, andNeil R. Kriegers From Departmentof Pharmacology, University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania19104

The 3a-hydroxysteroiddehydrogenase(EC 1.1.1.50) of rat brain cytosol has been purified to apparent homogeneity. Thepurificationprocedureinvolvessix successive steps, includes one affinity chromatography, and yields enzyme which displays a 1,550-fold enhancement in specificactivity.The homogeneous enzyme has a K,,, of 8.0 p~ for 5a-dihydrotestosterone, a V,,, of 1.3 pmol of 3a-androstanediol formed per h/ mg of protein, and displays a preference for NADPH. It appears to be themajor activity responsible for the reduction of 5a-dihydrotestosterone in this tissue and metabolism. may playa pivotal role in brain androgen The homogeneous enzyme has several properties in common with the 3~hydroxysteroid dehydrogenase purified from rat liver cytosol (Penning, T. M., Mukharji, I., Barrows, s., and Talalay, P. (1984) Biochem. J. 222, 601-611). It is a monomer with a molecular weight of 31,000, it has a PI of 5.5, and it is potently inhibited bythe nonsteroidal anti-inflammatory drugs(IC6, value for indomethacin= 2.0 p ~ ) .The potency of inhibition observed for the brain enzyme parallels that observed for cyclooxygenase: indomethacin > fenamates > I-methylpyrrole acetic acids > arylpropionic acids > salicylates > acetaminophen. Examination of a variety of steroidal contraceptives as modulators of the dehydrogenase indicates that ethinylestradiol is a very poor inhibitor (IC60 = 100 p ~ ) , while 6-medroxyprogesterone acetate (Provera) is an extremely potent inhibitor (IC5o= 0.2 p ~ ) .The possibility exists that brain androgen metabolism may be altered by the nonsteroidal anti-inflammatory drugs and synthetic progestins.

Testosterone has diverse effects in the brain. It has been implicated in the organization of neuronal pathways in the neonate brain ( l ) , and such changes may account for differences in male and female behavior (2). In addition, circuits that mediate male sexual behavior, e.g. clasping in the frog, song in the bird, and mating in the rat,can be stimulated by testosterone and its metabolites (3-5). Testosterone is also

* These studies were supported by National Institutes of Health Grants GM33464, BRSG S07-RR-0783, and NIH 31820. This work was presented in part at the 69th Annual Meeting of the American Society of Biological Chemists, Anaheim, CA, April 22-26, 1985 (Penning, T. M., Sharp, R. B., and Krieger, N. R. (1985) Fed. Proc. 44, 846). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked ‘‘advertisement’’ in accordance with 18U.S.C. Section 1734 solelyto indicate this fact. $ To whom correspondence should be addressed. § Present address: Department of Anesthesia, Harvard Medical School, Brigham and Women’s Hospital, Boston, MA 02115.

involved in themediation of the release of luteinizing hormone releasing factor in thehypothalamus (6-8). It is thus apparent that within the central nervous system, testosterone may effect the organization of neuronal pathways, the activation of stereotyped behavior, and theregulation of neuroendocrine responses. On this basis, the central nervous system can be considered to be a true target tissue for this steroid hormone. Testosterone and its active metabolites formed within the brain regulate neuronal activity presumably by binding to soluble receptors and subsequently exerting their effects at the level of gene expression (9,lO). The presence of particular activating or converting enzymes in a brainregion (or neuron) and their absence in others will dictate which steroids are synthesized and which areinactivated in that region (or neuron). These transformations will regulate steroid hormone receptor occupancy and govern which neuronal circuits may be affected. However, the enzymes involved in these transformations have not been completely characterized, and their properties and localization are poorly understood. The major routes for testosterone transformation in the central nervous system appear to be itsaromatization to estradiol (11) and its reduction to Sa-dihydrotestosteronel (12, 13). Sa-Dihydrotestosterone can be subsequently converted by a brain 3a-hydroxysteroid dehydrogenase (3a-hydroxysteroidNAD(P)+ oxidoreductase (EC 1.1.1.50)) to 3aandrostanediol (14-17). It is uncertain whether testosterone or its metabolites are themost important mediators of androgen action within the brain. While 5a-dihydrotestosterone is known to be active in the brain (5), there is little evidence concerning the role of 3a-androstanediol. The work of Wilson and co-workers (18-20) has clearly established that in peripheral tissues, e.g. rat ventral prostate, testosterone is reduced by 5a-reductase to give the more potent androgen 5a-dihydrotestosterone. This steroid is reduced still further by prostatic 3a-hydroxysteroid dehydrogenase to yield 3a-androstanediol. Although the diol is a potent androgen (21), it can be readily conjugated and eliminated and is considered to be a primary androgen metabolite. Thus, in peripheral tissues, the dehydrogenase is believed to catalyze the first step in androgen inactivation. This paper describes the purification and properties of the 3a-hydroxysteroid dehydrogenase present in rat brain cytosol. The trivial names used are: 5a-dihydrotestosterone, 5a-androstan-17&01-3-one;3a-androstanediol, 5a-androstane-3a,l7B-diol;ethnorethininylestradiol, 1,3,5(10)-estratriene-l7a-ethinyl-3,17~-diol; drone, 17~-hydroxy-19-nor-l7a-pregn-4-en-20-yn-3-one; 6-medroxyprogesterone acetate(Provera), 17a-acetoxy-6-methylpregn-4-ene3,ZO-dione; indomethacin, l-(p-chlorobenzyl)-5-methoxy-2-methylindol-3-yl acetic acid; flufenamic acid, N-(a,a’,a”-trifluoro-m-to1yl)anthranilic acid tolmetin, 5-(4-methylbenzoyl)-l-methylpyrrol-2yl acetic acid; ibuprofen, 2-(p-isobutylpheny1)propionicacid.

15266

Brain

Rat

Sa-Hydroxysteroid Dehydrogenase

It has been previously shown that 3a-hydroxysteroid dehydrogenase from rat liver is a target for both steroidal and nonsteroidal anti-inflammatory drugs (22,23) as well as progestins (24). If the brain enzyme is also sensitive to these drugs, it would follow that brain androgen metabolism may be altered by these agents. It is demonstrated here that rat brain 3a-hydroxysteroid dehydrogenase is potently inhibited by nonsteroidal anti-inflammatory drugs such as indomethacin and contraceptive progestins such as Provera. This is the first accountof the purification to homogeneity of an enzyme that metabolizes 5a-dihydrotestosterone inmammalian brain. EXPERIMENTAL PROCEDURES

Materials All steroids were purchased from Sigma. [‘4C]5a-Dihydrotestosterone (57.8 mCi/mmol) was purchased from New England Nuclear. Glucose 6-phosphate (Torula yeast, 300-400 units/mg) was purchased from Sigma. Pyridine nucleotides and Agarose Blue were products of Pharmacia PL Biochemicals. Nonsteroidal anti-inflammatory drugs were obtained from the sources shown in parentheses: indomethacin (Sigma), flufenamate (Warner-Lambert Pharmaceutical Co., Morris Plains, NJ), tolmetin (Dr. Edward Muschek, McNeil Laboratory, Inc., Fort Washington, PA),and ibuprofen (The Upjohn Co.). DEAEcellulose (Whatman DE52 preswollen microgranular form) was purchased from Reeve-Angel (Clifton, NJ); Polybuffer Exchanger 94, Polybuffer 74, and Sephadex G-100 were products of Pharmacia. Enzyme-grade sucrose and ammonium sulfate were obtained from Schwarz/Mann, and glycerol was of spectroscopic grade (J. T. Baker Chemical Co., Phillipsburg, NJ). Assay For Rat Bruin 3a-Hydroxysteroid Dehydrogenase Assay systems contained a 20-pl reaction mixture (consisting of 500 mM potassium phosphate buffer, pH 7.4, 1 mM EDTA, 21 mM NADP+, and 1mM glucose 6-phosphate), 4 pl of glucose-6-phosphate dehydrogenase (3.4 units/ml), 2 pl of 1.75 mM 5a-dihydrotestosterone (containing 40,000 cpm of [“C]5a-dihydrotestosterone) and 0-50 p1 of enzyme. In every case, the final volume wasmade up to 100 p1 with distilled water. Assays were initiated by the addition of enzyme, and incubations were conducted for 30 min a t 37 “C. The reactions were quenched by the addition of400 pl of ethyl acetate, the resulting extracts were evaporated to dryness and redissolved in 40 pl of methanol, and 20-4 aliquots were applied to thin layer fiber glass sheets impregnated with polysilicic acid (Gelman ITLC sheets). Chromatograms were developed in ch1oroform:ethyl acetate (8:2, v/v), and the positions of the substrate(5a-dihydrotestosterone; RF = 0.44) and product (3a-androstanediol; RF = 0.25) were detected by spraying marker strips with p-anisaldehyde. These steroids were then quantified by placing appropriate strips in 5 ml of a toluene-based scintillation fluid (4 g of 2,5-diphenyloxazole and 200 mg of 1,4-bis[2-(4methyl-5-phenyloxazolyl)]benzene/literof toluene) and detecting “C radioactivity on a Tracor Model 43 scintillation counter whose machine efficiency was 98% for I4C. Using the specific radioactivity of the substrate, theproduct formed was expressed as nanomoles of 301androstanediol formed per h/mg of protein. One unit of enzyme activity is equivalent to 1 nmol of 3a-androstanediol formed per h/ mg of protein. The identity of 3a-androstanediol as the product of the enzyme-catalyzed reaction was confirmed by carrier recrystallization (see “Results”). Purification of Rat Brain3a-Hydroxysteroid Dehydrogenase Preliminary experiments indicate that thespecific activity of brain Sa-hydroxysteroid dehydrogenase is highest if extracts are prepared in buffers of pH 8.6 containing 0.1 mM EDTA and 1 mM 2-mercaptoethanol. Thus, initial specific activities ranged from 1.0 to 2.0 nmol of androstanediol produced per h/mg of protein. These activities are considerably lower than those originally reported by Krieger and Scott (17) and canbe attributed to thefact that, in the earlier work, homogenates were prepared in buffers containing NADPH, 5a-dihydrotestosterone, and glucose 6-phosphate. The presence of pyridine nucleotides plus steroidal substrate helps to stabilize enzyme activity. The expense of NADPH and thedifficulty in keeping 5a-dihydrotestosterone soluble in a variety of buffers precluded their use in the purification of the enzyme described below.

15267

Step 1: Preparation of the 30-75% Ammonium Sulfate Fractwn100 adult male Sprague-Dawley rat brains (117 g) were homogenized in a Polytron (Brinkmann Instruments) in3 volumes of 50 mM TrisHCl, pH 8.6, containing 0.1 mM EDTA, 1 mM 2-mercaptoethanol, and 250 mM sucrose a t 4 “C. The homogenate was centrifuged at 10,000 X g for 30 min a t 2 “C, and the supernatantwas collected by filtration and centrifuged a second time at 100,000 X g for 60 min. The clear red supernatant (cytosol) was then treated with solid ammonium sulfate to yield a fraction that was 30-75% saturated. The resulting precipitate was collected by centrifugation at 10,000 rpm for 30 min and dissolved in a minimal volume of 10 mM TrisHC1, pH 8.6, containing 0.1 mM EDTA and 1mM 2-mercaptoethanol. The material was then dialyzed overnight against three 2-liter changes of the same buffer. Step 2: DEAE-cellullose Column Chromatography-The dialyzed material (110 ml) was applied to a DE52 cellulose column, (26.3 X 2.3 cm) equilibrated in thedialysis buffer. The column was developed with a linear salt gradient of 0-250 mMNaC1. The active fractions were pooled, concentrated, and dialyzed against 25 mM imidazole HC1, pH 7.4, containing 0.1 mM EDTA, 1 mM 2-mercaptoethanol, and 20% glycerol. It should be emphasized that, in all subsequent steps, buffers contained 0.1 mM EDTA, 1 mM 2-mercaptoethanol, and 20% glycerol. Step 3: Chromatofocusing with Polybuffer Exchanger 94-The dialyzed material from the previous step (21.0 ml) was applied to a chromatofocusing column (25.0 X 1.0 cm) packed with Polybuffer Exchanger 94 and equilibrated with 25 mM imidazole HC1, pH 7.4. The column was developed with a linear pH gradient using Polybuffer 74, pH 5.0. The active enzyme fractions that eluted with a PI of 5.5 werepooled and dialyzed against three 1-liter changes of20mM potassium phosphate, pH 7.0. The dialyzed material (68.3 ml) was concentratedwith a Millipore CX-10 immersion filter to a final volume of 12.4 ml. Step 4: First Sephadex G-100 Column Chromatography-The concentrated sample was applied to a Sephadex G-100 column (92.0 x 1.6 cm) equilibrated in 20 mM potassium phosphate buffer, pH 7.0. At this stage, a single peak of enzyme activity eluted at 0.52bed volumes, and the active fractions (40.5 ml) were concentrated as described before to a final volume of 11.0 ml. Step 5: Agarose Blue Affinity Chromatography-The sample (11.0 ml) was applied to a 5-ml column of Agarose Blue equilibrated in 20 mM potassium phosphate buffer, pH 7.4. The column was washed in the equilibration buffer to remove nonabsorbed protein, and after raising the pH to 8.6, a second batch of nonspecific protein was washed from the column. The Sa-hydroxysteroid dehydrogenase was then eluted in5 mM NADP+ dissolved in 20 mM potassium phosphate buffer, pH 8.6. The active fractions (14.1 ml) were pooledand dialyzed against 20 mM potassium phosphate buffer, pH 7.0.

TABLEI Carrier recrystallization of the 3a-androstaanedwlproduct formed during the enzymatic reduction of 5a-dihydrotestosterone Assays (5 X 200 pl) each containing 40 p1 of the reaction mixture (see “Experimental Procedures”), 13.4 pM [“C]5a-dihydrotestosterone (44,000 cpm), 10 p1of glucose-6-phosphate dehydrogenase (3.4 units/ml), and 100 pl of 30-75% ammonium sulfate fraction of brain cytosol were incubated a t 37 “C for 60 min. Each assay mixture was extracted with 2 X 200 pl of ethyl acetate, and the resulting extract was evaporated to dryness and applied to silica gel plates (20 X 20 cm) developed in ch1oroform:ethyl acetate (4:1, v/v). 3a-Androstanediol and unreacted 5a-dihydrotestosterone were located on the plate by reference to standards run along each edge. These two spots accounted for all the radioactivity. The 3a-androstanediol was extracted from the silica with ethyl acetate, and the extracts from the five incubations were pooled. Unlabeled 3a-androstanediol (50 mg) was added as carrier, and the whole was evaporated to dryness. The counts/minute of steroid was determined, and the residue was subjected to threesuccessive recrystallizations using hexane:acetone (21) as solvent. At the completion of each recrystallization, the specific radioactivity of the product was determined. Sample

epm/mg

Extract plus 50 mg cold 3a-androstanediol 1st recrystallization 2nd recrystallization 3rd recrystallization

1480 1295 1142 1130

15268

3a-Hydroxysteroid Brain Rat

Dehydrogenase TABLEI1

Summary of the purification of 3a-hydroxysteroid dehydrogenase from rat brain cytosol UnitsVolume

Step

ml

Cytosol (NH4),SO430-75% fraction DEAE-cellulose Concentrate from DEAE Chromatofocusing 1st Sephadex G-100 Agarose Blue 2nd Sephadex G-100

Specific activity

Protein

nmollhlmg

mg

210 1.381628 138 909 112 6.2 225 21 225 68.3 25.0 27 40.5 4.7 14.1 0.42 31.2 1542.0 0.192

-7.5 -7.0 -6.5 -6.0 -5.5

2247 2028 85.71926 57.3 1186 36.3 823 37.7 846 21.8489 18.2409

2.67 8.56 5.73 35.0 180.7 1156.6 2129.2

Recovery

Purification

%

-fold

100 90.3

1 1.9

4.2 130.4 838.1

Step 6: Second Sephadex G-100 Chromatography-The 14.1-ml sample was reapplied to theSepahdex G-100 column (see Step 4). At this stage, a single major protein peak coeluted with enzyme activity, suggesting that the dehydrogenase was homogeneous. The active ; fractions were pooled (31.2 ml) and concentrated to a final volume of I 8.0 ml. The purified enzyme was stored as 0.4-ml aliquots (0.024 mg/ ml) a t -80 "C for future use. The storage buffer consisted of 20 mM .o = potassium phosphate buffer, pH 7.0, containing 0.1 M EDTA, 1 mM 2-mercaptoethanol, and 20% glycerol. In thisform, the enzyme retains 85-90% of its original activity over 3 months.

-5.0

Polyacrylamide Gel Electrophoresis 3 / l l % double-stack polyacrylamide gels were formed and run as described by Ornstein (25) and Davis (26). 3a-Hydroxysteroid dehydrogenase activity was detected on disc gels by linking the NADPdependent oxidation of androsterone to thereduction of tetrazolium blue as described by Schultz et al. (27). Sodium dodecyl sulfate slab gel electrophoresis was performed as described by Laemmli (28).

Fraction Number

FIG. 1. Chromatofocusing of rat brain 3a-hydroxysteroid dehydrogenase. A sample (21.0 ml) of enzyme obtained by a combination of ammonium sulfate fractionation and DE52 cellulose column chromatography was applied to a chromatofocusing column (26 X 1.1cm) packed with Polybuffer Exchanger 94 and equilibrated with 25 mM imidazole HC1, pH 7.4. Elution was achieved with Polybuffer 74, pH 5.0.

Protein Determinations Samples devoid of glycerol had their protein concentration determined by the method of Lowry et al. (29), while those containing glycerol had their protein concentration determined by the method of Bradford (30). In every instance, crystalline bovine serum albumin (Armour) was used as standard. RESULTS

Validation of the Assay for 3a-Hydroxysteroid Dehydrogenase-5a-Dihydrotestosterone is a 3-ketosteroid and can be reduced by both a3a- anda 30-hydroxysteroid dehydrogenase. I ' , I Using rat brain cytosol as a source of enzyme, all the radioactivity that was present at theend of the assay was accounted for by spots thatco-migrated on TLCwith either gar-dihydrotestosterone or 3a-androstanediol. Large-scale incubations in which the reaction was run toequilibrium followed by carrier 0 crystallization of the presumed Sa-androstanediol product I 0.2w indicated that all the radioactivity crystallized to a constant 0 specific radioactivity. Three recrystallizations were conducted W 4in all (Table I). These data suggest that the NADPH-linked reduction of 5a-dihydrotestosterone in rat brain is catalyzed 0.1by a 3a-hydroxysteroid dehydrogenase and that the product of the reaction is Sa-androstanediol. Purification of 3a-Hydroxysteroid Dehydrogenase-Homogeneous Sa-hydroxysteroid dehydrogenase can be obtained from 100 male Sprague-Dawley rat brains using a procedure that involves five successive column chromatography steps Fraction Number which include DEAE-cellulose, chromatofocusing, and SephFIG. 2. Affinity chromatography of rat brain 3 ~ h y d r o x y - adex G-100 and Agarose Blue affinity chromatography. Thus, steroid dehydrogenase. The active fractions from the chromato- 0.2 mg of purified enzyme can be obtained in an overall yield focusing step (Fig. 1)were pooled,concentrated, dialyzed, and further of 18.2%. This homogeneous enzyme represents a 1550-fold purified by Sephadex G-100 column chromatography. The resulting enhancement of the initial specific activity for the reduction pooled material (14.0 ml) was applied to a 5.0-ml column of Agarose of 5a-dihydrotestosterone present in whole rat brain cytosol Blue equilibrated in 20 mM potassium phosphate buffer, pH 7.4. The column was then washed successively with potassium phosphate at (Table 11).Examination of the elution profiles at each step of pH 7.4, pH 8.6, and pH 8.6 containing 5 mM NADP. All buffers the purification procedure along with estimates of enzyme contained 0.1 mM EDTA, 1mM 2-mercaptoethanol, and 20% glycerol. recovery indicate that the homogeneous protein represents I

6

I

Rat Brain 3a-Hydroxysteroid Dehydrogenase

1

2

3

4

5 6

7

. .

Ij z

..

0 ”

.” . -

----

-

”Purified

-

!

VW”

w

1MI)

15269

136,000 68,000 35,000 Enzyme

1

2

23,000

” 17,000

FIG. 3. Polyacrylamidegel electrophoreticanalysis of 3a-hydroxysteroiddehydrogenaseof rat brain cytosol. Left, sodium dodecyl sulfate-polyacrylamide slab gel electrophoretic analyses of the purification of 3ahydroxysteroid dehydrogenase: lune I , 30-75% ammonium sulfate fraction; lane 2, pooled active fractions from DE52 cellulose column chromatography; lane 3, pooled active fractions from chromatofocusing and subsequent Sephadex G-100 columnchromatography; lane 4, pooled activefractions from Agarose Blue affinity column chromatography; lane 5 , pooled active fractions from the final Sephadex G-100 chromatography; lune 6, molecular weight standards; lane 7, molecular weight standards plus purified enzyme. Molecular weight standards included bovine serum albumin dimer ( M , = 136,0001,bovine serum albumin monomer ( M , = 68,000), pepsin ( M , = 35,000), trypsin ( M , = 23,000), and myoglobin ( M , = 17,000). Right, polyacrylamide disc gel electrophoresis of purified 301hydroxysteroid dehydrogenase; lune 1, enzyme stained with Coomassie Brilliant Blue; lune 2, 3a-hydroxysteroid dehydrogenase activity detected by coupling the oxidation of androsterone to thereduction of tetrazolium blue in the presence of Phenazine methosulfate (27).

Fraction Number

FIG. 4. Sephadex G-100chromatography of rat brain 3ahydroxysteroid dehydrogenase. The activefractions from the affinitychromatography step were pooled, dialyzed, concentrated, and applied to a (92 X 1.6 cm) column of Sephadex G-100 equilibrated in 20 mM potassium phosphate buffer, pH 7.0, containing 1 mM EDTA, 0.1 mM 2-mercaptoethanol, and 20% glycerol.

the major 3a-hydroxysteroid dehydrogenase responsible for 5a-dihydrotestosterone metabolism in rat braincytosol. Table I1 indicates that themost effective purification steps are those that involve chromatofocusing and affinity chromatography (Figs. 1and 2), each of which increase the specific activity at least 7-fold over the previous step. The contribution of the various steps to thepurification is more dramatically illustrated by the sodium dodecyl sulfate-polyacrylamide gel electrophoretic analyses shown in Fig. 3 (left). Thus, lane 3 shows the result of subjecting the pooled fractions obtained from the DEAE-cellulose step to chromatofocusing and subsequent Sephadex G-100 column chromatography, while lane 4 shows the result of the affinity column chromatography step. Criteria for Purity-The rat brain 3a-hydroxysteroid dehydrogenase was judged homogeneous by the following criteria. The final Sephadex G-100 column chromatography gave a single peak of protein that was coincident with enzyme activity at anelution volume that corresponded to a molecular weight of 34,000 (Fig. 4). With double-stack disc polyacrylamide gels, the purified material gave one band that stained for both protein and enzyme activity (Fig. 3, right). Measurement of the mobility of the enzyme on theprotein and activity

Rat Brain 3a-Hydroxysteroid Dehydrogenase

15270

80

r 1.o I

. . -

10.0 , /

,

100

2

c

2: 1,000

10,000

INHlBlTOR(yM)

al

FIG. 6. Inhibition of purified rat brain 3a-hydroxysteroid dehydrogenase by the nonsteroidal anti-inflammatory drugs.

0

E

0.12

0.08

0.04

0

0.04

0.08

,0 12

\.

FIG. 5. K , determination for purified rat brain 3a-hydroxysteroid dehydrogenase. The regular assay for the enzyme was performed with different concentrations of 5a-dihydrotestosterone (5-40 p ~ )In. every instance, 40,000 cpm of ['4C]5a-dihydrotestosterone was added to theassays. Reactions were initiated by the addition of 0.32 pg ofhomogeneous enzyme and terminated a t regular intervals over 30 min. The number of nanomoles of 3a-androstanediol formed at each time pointwas computed from the final specific radioactivity of 5a-dihydrotestosterone used in the assay. Plots of nanomoles of 3a-androstanediol formed uersus time gave the initial velocities. The resulting hyperbolic velocity uersus substrate curve obtained with the purified enzyme is shown as the inset on the double reciprocal plot ( A ) .An identical experiment was performed using 32 pg of a 35-70% ammonium sulfate fraction of rat brain cytosol as a source of enzyme and theresults are shown in B. In each case, the best fitted line was computed by the Wilkinson hyperbolic method (44).

gels relative to the tracking dye bromphenol blue gave RF values of 0.45 and 0.476, respectively. Complete alignment of the bands on the two gels is not possible due to the swelling and shrinking that occurs during the different development procedures. Androsterone (a 3a-hydroxysteroid) was used as substrate in these zymographic experiments, providing further evidence that this enzyme catalyzes the stereospecific and reversible oxidoreduction of 3a-hydroxysteroids to 3ketosteroids. Addition of the denatured enzyme to sodium dodecyl sulfate polyacrylamide slab gels also gave one band of protein of M , = 31,000 (Fig. 3, Zeft). From these data, it is concluded that the rat brainSa-hydroxysteroid dehydrogenase is both homogeneous and a monomer. Initial Velocity Measurements-Using homogeneous enzyme, the initial velocity of 5a-dihydrotestosterone reduction was measured in a series of incubations in which the concentration of steroid was varied in the presence of a fixed and saturating amount of NADPH (Fig. 5A). Transformation of the datagave a K , of 8.27 f 2.0 PM for 5a-dihydrotestosterone and a V, value of 1.3 pmolof 3a-androstanediol formed per h/mg of protein. An almost identical estimate of the K , ( K ,

The formation of 3a-androstanediol was followed over 30min using the regular assay procedure. Each assay contained 0.32 pg of purified enzyme and increasing amounts of representative nonsteroidal antiinflammatory drugs dissolved in 6% organic solvent. The velocity observed in the presence of drug was expressed .as the percentage inhibition of the control velocity. The resulting dose-response curves are shown for the following drugs:indomethacin (INDO),flufenamic acid (FLU),tolmetin (TOL),ibuprofen (IBU), salicylate (SAL),acetaminophen (ACEZ'), and aspirin (ASP).

TABLE 111 Inhibition of Sa-hydroxysteroid dehydrogenase of rut brain cytosol by steroid hormones IC5ovalues for a number of contraceptive steroids were determined as described in the legend to Fig. 6. Steroid hormone

IC.. BM

Estrogens 178-Estradiol Ethinylestradiol Androgens Testosterone 3a-Androstanediol Progestins Norethindrone Progesterone Medroxyprogesterone acetate

200 200 25 >>loo 9 7 0.2

= 14.0 PM; Fig. 5B) was obtained using the 30-75% ammonium sulfate fractionof rat braincytosol, demonstrating that crude preparations of the enzyme display a similar affinity for 5a-dihydrotestosterone. Examination of the pyridine nucleotide specificity of the purified enzyme indicated that it displayed adistinct preference for the triphosphopyridine nucleotide. Thus, the specific activity for 5a-dihydrotestosterone reduction was &fold higher in the presence of NADPH as compared to NADH. Inhibitor Studies-3a-Hydroxysteroid dehydrogenase has been recently purified to homogeneity from rat liver cytosol (22,23); an unanticipated property of this enzyme is its potent inhibition by the nonsteroidal and steroidal anti-inflammatory drugs (e.g. indomethacin, aspirin, dexamethasone, and prednisone). values for indomethacin were as low as 0.6 P"

Brain 3a-hydroxysteroid dehydrogenase is similarly sensitive to inhibition by the nonsteroidal anti-inflammatory drugs. Thus, indomethacin inhibited 5a-dihydrotestosterone reduction with an value of 2.0 p~ (Fig. 6). Since this

Brain

Rat

3a-Hydroxysteroid Dehydrogenase

15271

value was generated a t a substrate concentration that was twice the K, value, the Kivalue for indomethacin is believed to be substantially lower. The brain enzyme, like the liver isoenzyme, is inhibited by the major classes of nonsteroidal anti-inflammatory agents in a rankorder that parallels their potency as inhibitors of cyclooxygenase (31-33): thus, indomethacin > fenamates > I-meihylpyrrole acetic acids > arylpropionic acids > salicylates > acetaminophen (Fig. 6). Under physiological conditions, the 3a-hydroxysteroid dehydrogenase of rat brain is exposed to circulating steroid hormones; it was thus of particular interest to evaluate a number of natural andsynthetic steroids as enzyme inhibitors (Table 111).Progestins were among the most potent inhibitors studied. These included medroxyprogesterone acetate (ICso= 0.2 PM), progesterone (IC50= 7.0 PM) and norethindrone (IC5o = 0.9 PM). Estrogens were relatively ineffective inhibitors; thus, estradiol and ethinylestradiol gave ICs0values between 100 and 200 PM. Such high values suggest that estrogens play an insignificant role in regulating the activity of this enzyme. Androgens were of intermediateinhibitory potency; thus, testosterone gave an IC50 value of 25 @A, while 3a-androstanediol, the product of the reaction, showed no inhibitory activity.

2 orders of magnitude, and it can oxidoreduce carbonyls on a variety of nonsteroidal substrates, e.g. quinones and aromatic aldehydes and ketones. By contrast, the rat brain enzyme appears to be devoid of quinone reductase activity? This observation clearly distinguishes the rat brain enzyme from the carbonyl reductase recently purified from human brain (40). Previous studies have shown that the cytosolic 3a-hydroxysteroid dehydrogenase is widely distributed inrat tissues and that it is uniformly inhibited by micromolar concentrations of indomethacin (41). This is true of the enzyme present in tissues that are targets for androgens (prostate, seminal vesicle, and testis), tissues thatmetabolize androgens (liver), as well as tissues not usually associated with androgen action (lung and heart). The brain 3a-hydroxysteroid dehydrogenase can now be added to thelist of indomethacin-sensitive activities. The brain enzyme is in fact inhibited by all the major classes of nonsteroidal anti-inflammatory drugs and by the same low concentrations that inhibit cyclooxygenase (31-33). Certain sex steroids are also inhibitors of the brain 3ahydroxysteroid dehydrogenase. Contraceptive progestins, e.g. 6-medroxyprogesterone acetate (Provera), areparticularly effective, yielding IC50 values of 0.2 ,UM.This potency is comparable to thatobserved for this drug during its inhibition of DISCUSSION the Sa-hydroxysteroid dehydrogenase of testis, ovary, and epididymis (Ki = 0.42 PM) (42,43). By contrast, contraceptive This paper describes the purification of a steroid hormoneestrogens, e.g. ethinylestradiol, are not inhibitors of the entransforming enzyme from mammalian brain, namely the 3azyme. Thesedata suggest that androgen transformation hydroxysteroid dehydrogenase. This enzyme was obtained in within the brain may be affected by circulating steroids. We homogeneous form from rat brain cytosol with a 1550-fold speculate that pharmacologically relevant concentrations of enhancement in specific activity and 20% overall yield. This the nonsteroidal anti-inflammatory drugs and long-lasting enzyme is the major activityin rat brain cytosol for the progestins such as Provera via inhibition of the 3a-hydroxymetabolism of the androgen Sa-dihydrotestosterone. It is steroid dehydrogenase may regulate Sa-dihydrotestosterone interesting to speculate that this may be its physiological metabolism and hence the action of this androgen within the function in the male rat brain. The same enzyme may play brain. More generally, regulation of Sa-hydroxysteroid dehyan important role in female brain in reducing Sa-dihydroprodrogenase may be an important physiological determinant of Progesterone metabogesterone to 5a-pregnan-3a-ol-2O-one. the effects of androgen and progestin action on brain tissue. lism is one route by which sexual behavior and luteinizing hormone release are regulated in the female brain (2, 34-36). Acknowledgment-Initial rate datawere analyzed at a PROPHET Both membrane-bound and soluble 3a-hydroxysteroid de- computer terminal. PROPHET is a National Institutes of Health hydrogenase activities have been described in the brain (35, research source. 36) and peripheral tissues (37,38).The relative contributions REFERENCES of the soluble and the membrane-bound isoenzymes with 1. McEwen, B. S. (1982) in Molecular Approaches to Neurobiology respect to either androgen or progestin metabolism will depend on their relative V,,,/K, ratios for each substrate. Data 2. (Brown, I., ed) pp. 159-219, Academic Press, New York Rainbow, T. C., and McEwen, B. S. (1984) Handbook of Neurofor the brain dehydrogenase suggest that the cytosolic form chemistry (Lajtha, A., ed) 8th Ed., pp. 29-46, Plenum Press, reduces progestins much more efficiently than does its microNew York somal counterpart (34, 35). The corresponding data for an3. Erulkar, S. D., Kelley, D. B., Jurman, M.E., Zemaln, F. P., Schneider, G., and Krieger, N. R. (1981) Proc: Natl. Acad. Sci: drogens remain to be determined. U. S. A. 78, 5876-5880 Previous studies on the distribution of testosterone 5aF. (1980) Prog. Psychobiol. Physiol. Psychol. 9, 85Nottebohm, 4. reductase and 3a-hydroxysteroid dehydrogenase in rat brain 124 indicate that discrete patterns of localization exist for the two 5. Goy, R. W., and McEwen, B. S. (1980) Sexual Differentiation of enzymes. In addition, the specific activity for 3a-hydroxysteThe Brain,pp. 131-139, MIT Press, Cambridge, MA roid dehydrogenase exceeds that for 5a-reductase by 100-fold 6. Martini, L. (1982) Endocr. Rev. 3, 1-25 7. Schally, A. V., Redding, T. W., and Arimura, A. (1973) Endocriin every structure studied (17, 39). Thus, in vivo 3a-hydroxnology 93, 893-902 ysteroid dehydrogenase could serve to rapidly deplete the 8. Schally, A. V., Arimura, A., and Kastin, A. J. (1973) Science 179, accumulation of free 5a-dihydrotestosterone. The latter ste341-350 roid could be protected against such metabolism when it is 9. O'Malley, B. W., and Means, A. R. (1974) Science 183, 610-620 bound to thesoluble androgen receptor. 10. McEwen, B. S., Davis, P. G., Parsons, B., and Pfaff, D. W. (1979) The brain enzyme described here displays many properties Annu. Rev. Neurosci. 2, 65-112 in common with the isofunctional rat liver enzyme (22, 23). 11. Weisz, J., and Gibbs, C. (1974) Endocrinolom 94, 616-620 Both enzymes are monomers with molecular weights of ap- 12. Jaffe, R. B. (1969) Steroids 14, 483-498 13. Massa, R., Stupnicka, E., Kniewald, Z., and Martini, L. (1972) J. proximately 31,000; they have PI values in the range of 5.9Steroid Biochem. 3, 385-399 5.5, and they show a preference for NADPH over NADH as 14. Denef, C., Magnus, C. and McEwen, B. S. (1973) J. Enclocrinol. pyridine nucleotide. Both catalyze the reversible oxidoreduc59,605-621 tion of 3-ketosteroids to 3a-hydroxysteroids. However, the liver enzyme has a specific activity that is greater by at least R. B. Sharp andT. M. Penning, unpublished observations. "

~

-835

15272

3a-Hydroxysteroid Dehydrogenase Brain Rat

15. Kao, L.W., Lloret, A. P., and Weisz, J. (1977)J. Steroid Biochem. glandins & Cancer First International Conference (Powles, T., 8,1109-1115 Bockman, R. S., V., Honn, K. and Ramwell, P., eds) pp. 21-45, Alan R. Liss, Inc., New York 16. Lieberburg, I., Maclusky, N. J., and McEwen, B. S. (1977) E n docrinology 100, 598-607 32. Ferriera, S. H., and Vane, (1974) J. R. Annu.Pharmacol. Reu. 17. Krieger, N. R., and Scott, R. G.(1984) J. Neurochem. 42,887-14,57-73 890 33. Dembinska-Kiec, A., Zmuda. A., and Kupinska, J. (1976) Adu. 18. Bruchovsky, N., and Wilson, J. D. (1968) J. Biol. Chem. 243, Prostaglandin Thromboxane Res. 1,99-103 2012-2021 34. Garfias-Kubli, C., and Whalen, R. E. (1977) Hormones & Behau19. Frederiksen, D. W., and Wilson, J. D. (1971) J. B i d . Chem. 246, ior 9,380-386 2584-2593 35. Krause, J. E., and Karavolas, H. J. (1980) J. B i d . Chem. 2 5 5 , 11807-11814 20. Moore, R. J., and Wilson, J. D. (1974) Biochemistry 13,450-455 21. Walsh, P. C., and Wilson, J. D. (1976) J. Clin. Inuest. 57, 1093- 36. Krause, J. E., and Karavolas, H. J. (1981) J . Steroid Biochem. 14. RR-69 1097 22. Penning, T. M., and Talalay, P. (1983) Proc. Natl.Acad.Sci. 37. Verhoeven, G., Heyns, W., and De Moor, P. (1976) Eur. J. U. S. A. 80,4504-4508 Biochem. 65,565-576 23. Penning, T. M., Mukharji, I., Barrows, S., and Talalay, P. (1984) 38. Ghraf, R., Deutsch, H. J., Lax, R., and Schriefers, H. (1978) J. Biochem. J. 222,601-611 Biochem. Steroid 11, 1259-1264 24. Worner, W., and Oesch, F. (1984) FEBS Lett. 170, 263-267 39. Krieger, N. R., Scott, R. G., and Jurnam, M. E. (1983) J. Neu25. Ornstein, L. (1964) Ann. N . Y.Acad. Sci. 121, 321-349 rochem. 40, 1460-1464 26. Davis, B. J. (1964) Ann. N . Y. Acad.Sci. 121,404-427 40. Wermuth, B.(1981) J. Biol. Chem. 256,1206-1213 27. Schultz, R. M., Groman, E. V., and Engel, L. L. (1977) J. B i d . 41. Smithgall, T. E., and Penning, T. M. (1985) Biochem. Pharmacol. Chem. 252,3775-3783 28. Laemmli, U. K. (1970) Nature 227,680-685 42. Altman, K., Gordon, C. G., Southern, A. L., Vittek, J., and Wilher, 29. Lowry, 0. H., Rosebrough, N. J., Farr, A. L., and Randall,R. J. S. (1972) Endocrinology 90, 1252-1260 (1951) J. Bid. Chem. 193, 265-275 43. Wentworth, B. M., and Terner, C. (1981) Biol. Reprod. 25, 93Biochem. 72, 248-254 102 30. Bradford,M.M. (1976) Anal. 31. Vane, J. R., Flower, R. J., and Salmon, J. A. (1982) in Prosta- 44. Willtinson, G. N. (1961) Biochem. J. 80, 324-332