Dehydrogenase from Pseudomonas testosteroni

CopyrightC) 1975 American Society for Microbiology Printed in U.SA.

Enzymatic

Transformation of Morphine

by

Hydroxysteroid

Dehydrogenase from

Pseudomonas

testosteroni

PALOMA LIRAS,1 STEPHEN S. KASPARL4AN,2 AND WAYNE W. UMBREIT*

Department of Microbiology, Rutgers-The State University, New Brunswick,NewJersey08903 Received for publication 16June1975

Enzyme preparations from Pseudomonas testosteroni containing a- and hydroxysteroiddehydrogenasescatalyzed the oxidationofmorphineand codeine by nicotinamide adenine dinucleotide. Morphinewasconverted in relatively low

yieldinto14-hydroxymorphinone probablyviamorphinoneas anintermediate. Codeinewasconvertedtocodeinone and 14-hydroxycodeinone. Only the

conver-sionsatthe6-position werecarried outby the hydroxysteroid dehydrogenase. Hydroxylationatthe14-positiondidoccurspontaneously(orenzymatically with acontaminatingenzyme)after oxidationatthe6-position.

Wehave isolatedorganismsthat act on mor-phine. The organisms belongtothegenera Ar-throbacter (7) and Pseudomonas, and species identification shows themto be similarto the bacteria involvedinsteroid transformation. We thereforewondered whether the steroid-trans-forming organisms would alsoact onmorphine and its derivatives, and we used for this purpose the hydroxysteroid dehydrogenase (1.1.1.50and 1.1.1.51) preparations from Pseu-domonas testosteroni. Detailed analyses and characterization of this enzymearedescribedin the literature (1, 3, 8, 10-15).

MATERIALS AND METHODS

Determination of enzymeactivity.Steroid dehy-drogenaseactivity wasmeasured bythe spectropho-tometric assay ofMarcus andTalalay (8, 13, 15),

using a DB or aGilfordspectrophotometer (Gilford

410digital absorbance, model 222A).The activity of

the a- and P-hydroxysteroid dehydrogenases was obtained by using androsterone and testosterone,

respectively, assubstrates. Standard solutions con-taining 250

jtg

ofsteroidper mldissolved in

reagent-grade methanolwereprepared. Alternatively,

alka-loid drugs(morphineorcodeine)at 100- to250-tg/ml

concentration wereused as substrates. The reaction mixture consisted of0.1mlof a solution of nicotin-amide adenine dinucleotide (NADI) (30 mM) and 0.01 to 0.1 mlof the appropiate enzyme dilution. The volume was made up to 3.0 ml by the addition of 20 mMsodiumpyrophosphate buffer, pH 8.95. Unless otherwise indicated, the reaction was initiated by the addition ofenzyme. Measurements of theoptical

densityweremade againstablankcell containing all components except the appropriatesteroidor

al-kaloid. Oneunit of either a or , enzymeactivity was

IPresentaddress: Rosenstiel MedicalCenter, Brandeis

University,Waltham,Mass. 02154.

2Presentaddress:NewYork MedicalCollege,Valhalla,

N.Y. 10595.

definedas a change in optical density at 340 nm of 0.001per min in a 1-cm cell at 24 C and pH 8.95. The initialreaction rates are given in the kinetic experi-ments. Data on the enzyme activity will be in terms of the P enzyme only (using testosterone as the reference substrate).

Alternatively, the action of hydroxysteroid

dehy-drogenaseonalkaloiddrugswasfollowedby

study-ingthedecrease ofN-[methyl-'4Clmorphine. A typi-cal reaction mixture contained 1 mM NAD+, 100 to 250,ugof the alkaloid substrate per ml, 0.6 to 250 U

of enzyme per ml, 0.01 to 0.1

1&Ci

of N-[methyl-'4C]morphine,and 20 mMpyrophosphatebuffer, pH 8.95, tomake a total volume of 3 ml. The mixture wasplacedina50-ml Erlenmeyerflaskandallowed toshake at 200 rpm on a gyratory shaker, model G2 (New Brunswick Scientific Co., Inc., New Bruns-wick, N.J.), at room temperature. Unless otherwise indicated the reaction was initiated by the addition of the enzyme.

Separation and purification of the enzymes. Thehydroxysteroid dehydrogenaseenzyme ofP. tes-tosteroni is anintracellular enzyme. Consequently, the enzyme had to be released by sonication when thegrade I enzyme (crude dried cells) was used. A preparation of the grade I enzyme suspended in pyrophosphate buffer (20 mM, pH 8.95) was soni-cated for 3 min atapproximately80 W in aSonifier celldisrupter, model W185D (Heat Systems-Ultra-sonic,Inc., Plainview,N.Y.). The beakercontaining the enzyme solution was surrounded by ice at all times.Thepreparation was thencentrifugedfor20 min at 15,000 rpm at 4C ina Sorvall Superspeed

RC2-B.The enzyme activity of the supernatantwas determined.

Topurify and separate thea-and

/3-hydroxyster-oiddehydrogenases, the methoddescribed by Mar-cusandTalalaywasfollowed(8):25mg of commer-cialhydroxysteroid dehydrogenasegradeII

(contain-ingapproximately50%activity each of theaanda enzymes)wasdissolvedat1-mg/mlconcentrationin

phosphatebuffer,30mM,pH 7.2, containing

ethyl-enediaminetetraacetic acid (EDTA) (0.15%) and 650

on April 23, 2021 by guest

http://aem.asm.org/

NADI(0.1mM), and ammonium sulfatewasadded

stepwise. The success of the purification depends upon the careful elimination of heavy metals

through theuseof deionizeddistilledwater, wash-ingall glassware with0.2%solutionsof EDTA, and using purified ammonium sulfate recrystallized from asaturated0.2% EDTA solution. The

ammo-niumsulfate precipitateswereresuspendedin5ml of the same buffer and dialyzed overnight. The enzyme is precipitated preferably between 35 and 40% ammonium sulfate saturation, and the a en-zymebetween55and60%.Thefinal preparation ofa

or enzyme used had no more than 4 to 8% of contamination of the other enzyme. Protein was

determined by the method of Warburg and Chris-tian (16).

Enzymesand reagents. Commerciallyavailable gradesI,II, andIIIofNAD+-dependent

hydroxyster-oiddehydrogenase preparationsfrom P. testosteroni

wereobtained fromSigma ChemicalCo.(St. Louis, Mo.),as wereU.S.P.-gradetestosterone, androster-one,NAD+, grade III,and NADphosphate.Reduced

NAD(NADH)wasobtained from P-LBiochemicals,

Inc. (Milwaukee, Wisc.), and flavin adenine dinu-cleotide, B grade, was obtained from Calbiochem (Los Angeles,Calif.).Alltheabove-mentioned cofac-torswere prepared by using pyrophosphate buffer

(NaH2PO4-Na4P207), 20mM, pH 8.95,unless other-wiseindicated. Reduced glutathionewaspurchased

from Nutritional Biochemical Corp. (Cleveland, Ohio). The source of the alkaloid drugs and N-[methyl-'4C]morphine has been already described (6, 7).

Determination ofproducts. Extraction of the al-kaloids andtheir determinationbygasorthin-layer

chromatography (TLC),orbyscintillationcounting

oflabeleddrugs,has beendescribed elsewhere (6, 7).

RESULTS

Action of hydroxysteroid dehydrogenase

onmorphine. Whenmorphinewasincubatedin

the presence ofhydroxysteroid dehydrogenase

gradeIIfor3to6h inthe conditions indicated

above, a decrease in the amount ofmorphine

occurred. When the extractsof the incubation mixture were chromatographed in thin layer,

usingthe solvent system ethyl acetate-metha-nol-ammonium hydroxide (86:10:4), besides morphine (Rf=0.22), twonewsubstances with

Rf= 0.63(substance

S,)

and0.28(substance S2)

appeared. When N-[methyl-"'Clmorphine was

addedto the reactionmixture, all of the spots visualized in the I2 vapor were radioactive. A

substancerunningimmediatelybehind the

sol-ventfront didnotcontainradioactivityandwas

foundtobe duetotheenzymepreparation. The

relative amount of the two products formed (substances Sl and S2)varied accordingto the grade ofenzyme used andconditions ofassay.

Usually the amount ofsubstance

S,

was

be-tween two and three times higher than the

amountof substance S2. Upto20% ofthe

mor-phine added was transformedin 4 h. Various additions to and changes in the reaction mix-ture were attempted in order to optimize the formation oftransformation products. Neither NAD phosphate, NADH, nor flavin adenine dinucleotide could replaceNAD+ asthe cofactor for the enzyme. The enzyme was very labile in thepresenceof divalent metal ions(Fe2+,Cu2+, orZn2+)at 1 mMconcentration. As a resultof thisfinding,EDTA wasused for theremainder of the study as chelating agent. Talalay and Marcus(14)reported the necessity offree sulfhy-dryl groups for enzyme activity. However, the addition to the system ofsulfhydryl reducing agents as cysteine or glutathione failed to enhance enzyme activity. Substituting tris(hydroxymethyl)aminomethane buffer for pyrophosphate bufferalso inhibited the trans-formation ofmorphine.

Although determinations of nicotinamide-linked dehydrogenases are readily done by measuring the optical density at 340 nm, such spectrophotometric assays on extracts of crude cells ofsteroid dehydrogenase grade I showed very little activity. Kersters and DeLey (4) have pointed out that assays with NAD+ as cofactor cannot be carried out with crude ex-tractsduetothe usual presenceofhigh NADH oxidaseactivity. ThegradeIenzyme showed4 to 10times more NADH oxidase activity than didgrade III. Grades IIand IIIof the enzyme contained only small amounts of the NADH oxidaseactivity that didnotinterfere with the spectrophotometric assayand wereusedinthe rest ofthe study.

The influence ofvarious parameters on the activity of hydroxysteroid dehydrogenase on morphine was investigated. Controls lacking the morphine substrate or lacking enzyme failed to showanyreduction ofthe NAD+.The effect of temperatureonthehydroxysteroid

de-hydrogenase activity showed a plateau effect, with anoptimal temperature range of between 27and40 C (Fig. 1A). The effect of increasing enzyme concentration with respect to activity showedalinearrelationship, with a leveling off of activity at approximately 1,000 U/ml (Fig. 1B). The enzymeactivity wasmarkedly sensi-tivetochangesinpH,asalready pointedoutfor other reactions catalyzed by NAD+-requiring

enzymes(9). Increasing thepHof the solution favored higher transforming activity on mor-phine, with maximumactivityat pH 9.0(Fig. 1C).

After the addition ofincreasing amountsof NAD+, onewould expect to find a point where theenzyme activity levelsoff. Contrarytothis assumption, experimental data showed an in-hibitory effect at higher concentrations of 30,

on April 23, 2021 by guest

http://aem.asm.org/

20 10 0 TEMPERATURE (°C) 4 8 12 ENZYME (102units/ml) pH NAD (mM)

FIG. 1. Effect oftemperature (A),enzymeconcentration (B),pH (C), andNAD+concentration (D)onthe hydroxysteroid dehydrogenaseactivitywith respectto morphine. The velocity of the reactionwasmeasured spectrophotometrically, using80pgof morphinepermlassubstrate.

NAD+(Fig. iD). The gradeIfl NAD+(obtained

from Sigma Chemical Co.) contains a minor

amountofanimpuritythat had been reported

(2, 17)to be acompetitive inhibitor in certain enzyme systems. The 1 mM concentration of NAD+used inthesestudieswaswellbelow the

criticalconcentration of4mMrequired for the inhibitoryeffect. Basedonthe abovefindings,a

typical reaction mixture consisted of 1 mM NAD+,250U ofenzyme perml, 100to250.ugof

morphineperml, andpyrophosphatebuffer,20

mM, pH 8.95, upto3ml.

TheMichaelisconstantsofmorphine and

co-deine were calculatedby using a

Lineweaver-Burkplot(Fig. 2). The Km of codeinewas

calcu-latedas 1.92 x 10-3 M. The calculation of the

Kmvalue formorphinewascomplicated by the

lowsolubilityof thisdrug (50 mg/ml in

metha-nol),butanapproximatevalue of 5.7 x 10-3M

wasfoundby extrapolation usingtwodifferent

Lineweaver-Burkplots fortwoenzyme

concen-trations (Fig. 2). TheKmfigures obtained can

be compared with the Km of 3-hydroxysteroid dehydrogenase for testosterone, 5.5 x 10-5 M (15), and that ofa-hydroxysteroid

dehydrogen-ase for androsterone, 1.5 x 10-3 M (12). The

enzyme had a higher Km values for alkaloids

thanfor steroids.

Reaction mixtures containing morphine and the hydroxysteroid dehydrogenase enzyme

showed a measurable decrease in morphine

within 4 h. There also occurredasimultaneous

increase in theamountof transformation prod-uctsdetected, namely, substancesS,andS2. A

comparison between the amount ofmorphine

presentand theamountof reducedNAD+over

61 B 4 '0 x 2 c E NIl 0 c

oI

o 12 z 10 w (I) W 8 z APPL. MICROBIOL. on April 23, 2021 by guest http://aem.asm.org/ Downloaded from

C\j I0 CODEINE KmCodeine 1.92x10 M KmMorphine 5.7x103M 1/s (mMM1)

FIG. 2. Lineweaver-Burkplot oftheeffect of sub-strate concentration on velocityofenzyme reaction. The velocity of the reaction wasmeasured

spectropho-tometrically, using 225 (0, 0)or80U (0)of hydrox-ysteroiddehydrogenase grade II.

aperiod of time showed the two substances to beinverselyproportionaltoeach other (Fig. 3). As the amount of morphine decreased with time, the absorption readings at 340 nm in-creased. No changes were notedinthe control

group lacking theenzyme. These data link the NAD+-dependent hydroxysteroid dehydrogen-ase with the decrease in the amount of mor-phine and the production of transformation

products.

Products formed byhydroxysteroid dehy-drogenase actingonmorphine and codeine. Theaction ofhydroxysteroid dehydrogenase on morphine has beendescribed as producing two major transformation products, one slow and onefastrunning. Similar results wereobtained

when codeine, the C-3 methyl ether of mor-phine, was used in place of morphine in the reaction mixture. Onceagain, two major spots werevisualizedintheI2vapor.

Theidentification of the morphine transfor-mation products was attempted by comparing

Rf

and

R,

values of theunknown products with known standards, using TLC and gas chroma-tography, respectively (6). Spot Sl was identi-fied as 14-hydroxymorphinone. It was postu-lated thatspotS2could beeither morphinone or

14-hydroxymorphine. Although morphinone was notavailable as astandard, the 14-hydroxy-morphine standard failed to match with spotS2

(inTLC) in the several solvent systems used, and in anexperiment in which

14-hydroxymor-phine wassubstitutedformorphine in the reac-tion mixture, theenzyme wasunable to act on thissubstrate and notransformationor degra-dation occurred. We therefore tentatively

con-cludedthatspot S2ismorphinone.

Since the enzyme acts on codeine, it was hoped that the identification of the codeine

transformation products might shed light on the reactioninvolvingmorphine.Byagain com-paringRfandRtvalues of the unknown

prod-ucts with known standards, the fast-running spot wasidentifiedas14-hydroxycodeinoneand theslow-running spot as codeinone.

The fact thatcodeine wastransformedto co-deinone by the enzyme, together with the en-zyme's inability to act on14-hydroxymorphine, suggestedthatmorphine,inbeingtransformed

to14-hydroxymorphinone, wasfirstchangedto morphinone and that spotS2wasindeed

morphi-none.

The transformation ofmorphine by the hy-droxysteroid dehydrogenase enzyme into two products, one involving dehydrogenation and

the other involving hydroxylation, seemed to contradict the notion ofhighspecificity in the enzymesystem. It appearsthat the

hydroxyster-oid dehydrogenase first acted upon the

mor-E_E C OL 0c OQ i-25 20 15 Q U) 0 :3 3 z 0 I 0 3 Q Time (hours)

FIG. 3. Comparation between the amountof

mor-phineremaining andtheamountofNAD+reduced over a period oftime. Flasks contained in a3-ml volume: 250 Uof enzyme gradeII;250pg(0.1

A&Ci)

of

morphine per ml;1 mMNAD+;and20 mMpyrophos-phate buffer, pH895.

["4C]morphine,

;NADH

formed,- --,control,0;hydroxysteroid

dehydrogen-ase,0.

VOL.30,1975

on April 23, 2021 by guest

http://aem.asm.org/

phine to formmorphinone (Fig. 4). The subse-quenttransformation ofthemorphinone to 14-hydroxymorphinone could be due either tothe action of acontaminating enzymeorto normal chemical oxidation of morphinone in the pres-ence ofair and high pH, since chemically the conjugated double-bond system found in mor-phinonemade the C-14positionevenmore sus-ceptible toattack.

Purification of the a and I8 forms of hy-droxysteroid dehydrogenase and their ac-tion on morphine.The a-

and,8-hydroxyster-oid dehydrogenases were separated and puri-fiedbyammoniumsulfatefractionation accord-ing to the method described by Marcus and Talalay (8). Each collected fraction was then assayed foraand ( activity usingandrosterone andtestosterone, respectively,assubstrates.

Reaction mixtures containing either thea-or

3-hydroxysteroid dehydrogenase

were

assayed

by using morphineassubstrateinthe spectro-photometer. Only the ( enzyme produced in-creased optical absorbance values at 340 nm, indicating that the

P

andnottheaenzymewas responsible for the morphine transformation (Fig. 5). Similarexperiments showeddecreases

N-CH3 MORPHINE H OH | BETA-HYDROXYSTEROID DEHYDROGENASE N-CH3 /

MORPHINONE

N-CH3

(-.f~-\) 14-HYDROXY <?\ < b

MORPHINONE

FIG. 4. Hypothesizedsequence of the

transforma-tion of morphine by hydroxysteroid dehydrogenase.

TIME (hours)

FIG. 5. Reduction ofNAD+by purified aand ,B enzymefractions of hydroxysteroid dehydrogenase,

in the presence ofmorphine. Sixty units ofaor ,B

purified enzyme was used. Other conditions as in

Fig.3.

intheamountofmorphineandincreasesinthe

amountoftransformationproductswhen the ( enzyme was incubated for 4 h with the drug,

but no changes occurred in the morphine in incubationswith the a enzyme.

Further evidencethat the (8enzyme was in-volved in morphine transformation was pro-vided by studies ofcompetition. The possible

influenceofandrosteroneortestosteroneonthe action ofhydroxysteroid dehydrogenaseon mor-phine was examined in an effort to evaluate whether the a and/or (8 enzymes were responsi-ble for the transformation products of

mor-phine.Since theCrhydroxylgroupofmorphine is in the aposition,onemightexpectthe andros-terone to act as a competitive inhibitor with the morphine. The androsterone, despite its very much higher affinity for the a

enzyme,

had noinhibitory effecton the transformation ofmorphine by the crude enzyme (containing approximately 50% a and 50%,8activity). Tes-tosterone (which is the substrateforthe (8 en-zyme) was highly inhibitory to the morphine-transforming reaction (Fig. 6). The additionof aslittle as 0.25,gof testosterone per ml in the presence of 250 ,ug of morphine per ml signifi-cantly inhibited the action of the enzyme on morphine. In the presence of5 ug or more of testosterone per ml,notransformationproducts could be visualizedinthel2vapor,andonlyan insignificant amount ofradioactivity was de-tected in the appropriate regions of the TLC plate. Atconcentrations of less than0.1

mg

of testosterone per ml, morphine transformation was not inhibited. This inhibition of the

mor-APPL.MICROBIOL.

on April 23, 2021 by guest

http://aem.asm.org/

phine dehydrogenation by testosterone could not be due to competition for NAD+ of two differentdehydrogenasesinthe preparation. In that case, androsterone wouldproduce a simi-lar inhibition. Thus it appears that the 13 en-zyme wasresponsible for the transformation of morphine.

DISCUSSION

This study has demonstrated the ability of the hydroxysteroid dehydrogenase enzyme from P. testosteroni to transform morphine. Conditions favoring morphine transformation included a high pH of 8.95 and a temperature range of 27 to 40 C (Fig. 1). The

K,.

values of morphine and codeine were calculated to'be 1.9 and5.7 x 10-3 M, respectively, well above the

K.

forandrosterone or testosterone (12, 15). Morphine is transformed by the enzyme to 14-hydroxymorphine, probably via a none intermediate. The production of morphi-none from morphine has been attributed to the hydroxysteroid dehydrogenase enzyme, whereasthe subsequent transformation to 14-hydroxymorphinone is the result of either a contaminating enzyme present in the original hydroxysteroid dehydrogenase extract or the result of chemical transformation. The oxygen of the14-hydroxylgroupof the 14-hydroxymor-phinone appears to originate from molecular oxygen, since in the absence of air this sub-stance was notfound(unpublished data).

The a-hydroxysteroid dehydrogenase was found incapable of reacting with the morphine, andactivity was dependent upon the(8enzyme (Fig. 5).

-Hydroxysteroid

dehydrogenase oxi-dizes three

3-hydroxysteroids

of the C19 and C21 series, 17

3-hydroxysteroids

of the

C18, Cl,,,

and

C21series, and certain 16

(3-hydroxysteroids

(13, 15). Therefore, it is less specific in its action thantheaenzyme,which is only able to oxidize the threea-hydroxysteroidsoftheC19, C21,and

C24

series (15). However, since

(-hydroxyster-oid dehydrogenase has been found to act only

onthe

(-hydroxyl

groups,itappears that the

(-hydroxysteroid dehydrogenase enzyme must at-tack the morphine molecule from the 'back"

side,sothat theC6hydroxylgroup is nowinthe

,3

position. Thiswould explain the lack of action of the a enzyme on 14-hydroxymorphine, in which the 14-hydroxyl group may notallowthe binding of the enzyme. Addition of testosterone to the reactionmixture, inquantities as small as1:1,000 (testosteronetomorphine),was capa-ble of inhibiting the transformation of mor-phine by the enzyme. On the other hand, the addition of androsterone to preparations con-taining both a-and

(3-hydroxysteroid

dehydro-100 z -- -0 O-Z

0r-0 QZi. 0 cr 0 Oen z I Ln 7-Z 80o 60 40 20 0 1 2 37- 5 STEROID (jpg/ml)

FIG. 6. Inhibition of the transformation of

mor-phine by hydroxysteroid dehydrogenase, by andros-terone(-) andtestosterone(0). ConditionsasinFig.

3.

genaseswas unable to cause asimilar

inhibi-toryeffect (Fig. 6).

Thus, itappearsthat the highly specific

ste-roid transformingenzyme(3-hydroxysteroid

de-hydrogenaseisabletoactonanalkaloid. Still,

the preparation of highly purified (3hydroxy-steroid dehydrogenaseisnecessarytostudyits actiononmorphine. The transformation of the

morphine alkaloids by the steroidenzyme

sys-tem opens a new area in which to direct the search for new and better morphine-related

compounds. The results of this study suggest

that in addition to searching for new

drug-transforming enzymes, one might also utilize

previously isolated and documented enzymes.

Inthe quest fora betteranalgesic, theuse of

established steroidenzymes seemsto bea

rea-sonableapproach.

LITERATURE CITED

1. Boyer, J.,D. N.Baron,and P.Talalay.1962. Purifica-tion andpropertiesofa3a-hydroxysteroid

dehydro-genasefrom Pseudomona testosteroni. Biochemistry

4:1825-1833.

2. Dalziel, K.1963. Thepurificationof nicotinamide ade-ninedinucleotide and the kinetic effects of nucleotide impurities.J. Biol. Chem. 238:1538-1543.

3. Delin, S., P. G.Squire, andJ. Porath. 1964. Purifica-tion of steroid-induced enzymesfrom Pseudomonas testoseroni. Biochim.Biophys. Acta 89:398-408. 4. Kersters, K., and J.DeLey. 1971. Enzymictests with

restingcells and cell-free extracts,p.33-52 In J. R. Norris and D. W. Ribbons(ed.),Methods in

microbiol-ogy,vol. 6A. Academic PressInc.,New York. 5. Krueger,M. 1955. Narcotics andanalgesics,p.1-77. In

R. F. Manska(ed.),Thealkaloids, vol. 5. Academic PressInc.,New York.

6. Liras, P. 1975. Thin layer and gas chromatography parametersofmorphineand codeine derivatives. J. Chromatogr. 106:238-240.

7. Liras, P.,andW. W.Umbreit. 1975. Transformationof morphine by restingcells and cell-freesystem of

Ar-Testosterone Androsterone 30,1975 on April 23, 2021 by guest http://aem.asm.org/ Downloaded from

throbacter sp.Appl. Microbiol. 30:262-266. 8. Marcus, P. I.,and P. Talalay. 1956. Induction and

purifi-cation of aandP-hydroxysteroiddehydrogenases. J. Biol. Chem. 218:661-674.

9. Racker, E. 1950. Crystalline alcohol dehydrogenase from baker's yeast. J.Biol. Chem. 184:313-319. 10. Roe, C. R., and N. 0. Kaplan. 1969. Purification and

substratespecificities of bacterial hydroxysteroid de-hydrogenases. Biochemistry 8:5093-5103.

11. Squire, P. G., S. Delin, and J. Porath. 1964. Physical and chemicalcharacterization ofhydroxysteroid dehy-drogenases from Pseudomonas testosteroni. Biochim. Biophys. Acta89:409-421.

12. Talalay, P. 1963. Enzymic analysis of steroid hormones, p. 119-143.In D. Glick (ed.), Methods of biochemical analysis, vol. 8. Interscience Publishers, Inc., New

York.

13. Talalay, P., and M. M.Dobson. 1953. Purification and properties of P-hydroxysteroid dehydrogenase. J. Biol. Chem. 205:823-838.

14.Talalay, P., and P. I. Marcus. 1954.Enzymic formation of 3 a-hydroxysteroids. Nature (London) 173:1189-1190.

15. Talalay, P., and P. I. Marcus. 1956.§pecificity,kinetics and inhibition ofaandP-hydroxysteroid dehydrogen-ases. J.Biol. Chem. 218:675-692.

16. Warburg, O.,and W.Christian. 1942. Isolierungund Kristallisation des Garungsferments Enolase. Bio-chem.Z. 310:385-521.

17. Winer, A. D. 1964.Crystallizationofnicotinamide ade-nine dinucleotide. J. Biol. Chem. 239:PC3598-PC3600.

on April 23, 2021 by guest

http://aem.asm.org/