THE Jou~lva~ OF Ihmc~cn~, CI~.~TRY

Vol. 248, No. 11, Issue of June 10, pp. 3646-3652, 1971

Printed in U.S.A.

Affinity Labeling of Steroid Binding Sites

SYNTHESIS OF CORTISONE 21-IODOACETATE AND STUDY OF 20/hHYDROXYSTEROID DEHYDRO- GENASE*

(Received for publication, January 6, 1971)

MANIK GANGULY 1 AND JAMES C. WARRENS

From the Departments of Obstetrics-Gynecology and Biochemistry, University of Kansas School of Medicine, Kansas City, Kansas 66103

SUMMARY

To extend the application of affinity labeling for the charac- terization of macromolecular steroid binding sites, we syn- thesized two cortisone derivatives (cortisone Zl-iodoacetate and Zl-iododeoxycortisone) and studied their reactions with the use of 20/3-hydroxysteroid dehydrogenase from Strepto- myces hydrogenans in 0.05 M phosphate buffer, pH 7.0, as the model protein. Both derivatives are capable of the reversible binding step at the active site, as both serve as substrates. Cortisone 21 -iodoacetate inactivates the enzyme in a time-dependent and irreversible manner, cortisone slows the rate of this inactivation, and excess 2-mercaptoethanol stops it. Iodoacetic acid did not inactivate the enzyme (with or without added cortisone), nor did Zl-iododeoxycortisone. Nevertheless, both steroid derivatives easily react with model sulfhydryl compounds.

Radioactive cortisone Zl-iodoacetate was then synthesized with 2-3H-iodoacetic acid. Inactivation of the enzyme by this steroid was accompanied by radiolabeling; neither was seen with 2-3H-iodoacetic acid alone. After inactivation and acid hydrolysis, a single major fraction of radioactivity was seen on paper chromatography. Amino acid analysis of the hydrolysate revealed a major radioactive peak (containing 86% of total radioactivity) with mobility identical with stand- ard 1,3-dicarboxymethyl histidine and a minor peak (con- taining 5 % of total radioactivity) with mobility identical with standard 3-carboxymethyl histidine.

These observations are compatible with a mechanism whereby the steroid moiety of cortisone Zl-iodoacetate delivers the reagent group to the binding site of 20/Shydroxy- steroid dehydrogenase, where it reacts with a histidine residue at that site.

These derivatives are offered as compounds which may possibly be used to study certain steroid binding sites of high affinity present in receptor proteins of target organs.

* This research was supported by Research Grant AM-05546 from the National Institutes of Health, United States Public Health Service.

f Postdoctoral fellow, National Institute of Child Health and Human Development.

5 Career Development Awardee, National Institute of Child Health and Human Development. Present address, Departments of Obstetrics-Gynecology and Biochemistry, Washington Univer- sitv School of Medicine, St. Louis, Missouri 63110.

Previous effort in this laboratory effected the synthesis of two affinity labeling steroids (4-mercuri-17P-estradiol and 2- diazoestrone sulfate) and demonstrated that they did, in fact, affinity label various steroid binding sites. These demon- strations were based on spectral evidence, as well as on persistenl, biological activity (I-3).

The present investigation was carried out to synthesize derivatives of cortisone which can, under physiological condi- tions, utilize the steroid moiety to deliver the reagent at a steroid binding site where a reactive amino acid residue, at or near that site, will be favored in covalent bonding. Further, it was desired to effect a more direct demonstration that bonding occurred by isolation and identification of the amino acid derivative after hydrolysis of the enzyme.

EXPERIMEKTAL PROCEDURE

Materials-Glass-redistilled water was used for all solutions. Ethanol and spectroscopic grade acetone (Fisher) were used without further purification; other organic solvents were distilled prior to use. The concentration of all organic solvents is expressed as a percentage by volume. Bulk quantities of cortisone and cortisone acetate were obtained from the Stera- loids Company, Rawling, New York, and found to be homogc- neous by thin layer chromatography. Pyridine nucleotides (NAD+ and NADH) and 20fi-hydroxysteroid dehydrogenase from Xtreptomyces hydrogenans were purchased from Sigma. Norit A, reagent grade salts, silicic acid (100 mesh), 2-mcrcapto- ethanol, p-toluene sulfonyl chloride, and p-dimethylaminobenz- aldehyde were obtained from the Matheson Company, Inc., East Rutherford, New Jersey. The 2-mercaptoethanol was distilled under reduced pressure prior to use. Chloroacetic anhydride, Eastman thin layer chromatographic silica gel sheets (NO. 6060), reagent grade dichloromethane, pyridine, and other organic solvents were obtained from Fisher. Pyridine was distilled over solid KOH prior to use and stored over KOH in a colored bottle. Dicyclohexyl carbodiimide and 5,5’-dithiobis (2-nitro- benzoic acid) were from Aldrich. The latter compound was crystallized from alcohol before use. Histidine, lysine, cysteine, reduced glutathione, N-acetylhistidine, and N-acetyltryptophan were obtained from Calbiochem. Methionine and carboxy- methylcysteine were Mann assayed quality. The three car- boxymethyl derivatives (l-, 3-, and 1,3-dicarboxymethyl histidine) were synthesized according to the method of Crest-

3646

by guest on January 29, 2020http://w

ww

.jbc.org/D

ownloaded from

Issue of ,Junc 10, 1071 -If. Ganyuly and J. C. Wtzwen 3647

field, Skin, and Moore (4). Ninhydrin and iodoacetic acid were obt~aincd from Eastman. The latter compound was decolorized by treating an acetone solution Iv&h Norit .\ and evaporating the acetone under vacuum. The 2-31-J-iodoscet,ic acid (44 mCi 11cr mmole), obtained from New England Nuclear was mkcd with cold iodoscetic acid and similarly decolorized. After mixing, the specific activity was found to bc 22 mCi per mmolo. 2,5-l~ipl~cn~~losazole and 1,4-bis[2-(5-p~lcn)llosnzo- lyl)]benzcnc were obtained from Amersham/Se:lr!c for scintil- lation counting, and Triton X-100 was from Packard Instru- ment Coml)any. Union Carbide dialysis tubing was purchasetl from Scielltific l’roducts. Elemental analysk was (lone ‘us Galbraith Laboratories, Knoxville, Tenuexsec.

Methods-Melting points were determined on n Fisher-,Johns apparatus and arc rcportcd without correction. Infmmcd spectra were. measured as IiBr pellets with a 13eclrmnn In-8 instrument, and optical rotation values were determined in chloroform solu- tion at 589 11m with a Jasco-Durrum spcctropolarimetcr. Ma.ss spectra were obtained with a Varian mass spectrometer (MAT CH 4I3), operating with an ionization energy of 70 C.V. and a direct inlet sysl;em. Ultraviolet absorption spectra were recorded on a Cary Model 14 spectrophotometer.

A mixtureof bcnzeneand ethanol (95:5) was used for chrotna- togrsphic separation of steroids by thin layer chromatography. Amino acids were separated by pspcr chromatography on a Whatman No. 3M1\‘I paper with butanol-acetic acid-water (200:30:75). Free amino acids were detected by spraying with ninhydrin (5) and N-a.cct~~~liltr~~ptol,han with lklich reagent (6). For separation of mixtures of nnlino acids and steroids or iodoacctic acid over thin layer silica gel, a solvent mixture of bulnnol-acetic acid-water (4: 1:2) was used. Rndio- labeled amino acid derivatives and iodoacetic acid were de- tected on paper chromatograrns with a Nuclear-Chicago Acti- graph III paper strip) counter (time constant, 20 s; chart speed, 60 cm per hour; full scale reading of 300 cpni; and 4?~ geometry),

Samples were I~ydrolyzecl in constant boiling 6 RI I-ICI in evacuated, senlcd lubes al, 110” for 24 hours. Amino acid analy- ses wcrc pcrrornlcd according to the proccdurc 01” Sparkman, Moore, :mtl Stcill (7) with a Spinco model 120 amino ncitl :tnaly- zer with 20-mm flow cells, operated at 60 1111 per hour. TllC

cflluent kern tl~c column aker ninhydrin reaction was collected at 1.5 ml l)er rnin per tube. Rliquots of 0.3 ml of this effiuent and 0.7 1111 of water were counted with 10 ml of phosphor in a Nuclear-Chicago scintillation counter. The phosphor was made with 0.5% 2,5-diphenyloxazole and 0.01% 1,4-bis[2-(5- phenylosazolyl)]benzcne in a mixture of Triton X-100 and toluenc (1:2). Quench correction was made with the internal standard technique.

Enzgrne Assuys-Enzyme assays were carried out at 25 =t 1” in a Beckman model DU spectrophotometer equipped with a Ciilford JJOWW supply and a Honeywell recorder. In all assays, enzyme activity was estimated from the initial linear decrease in absorbnncc at 340 nm. The BOO-hydrosysteroid dehydroge- nase used in these studies was found to be homogeneous by disc gel elcctrophoresis (8, 9). Steroid substrates were ncldcd in ethanol. The assay was carried out in 0.05 M phosphate buffer, ~11-1 6.5, in a fillal volume of 3.0 ml, with the use of 0.2 pmole of NAl)II as cofactor and, unless otherwise states, 0.3 ~molc of cortisone as substrate. The reaction was initiatctl by addition of approsimai.cly 750 ng of the enzyme.

S~ptlmis of Cortisone dl-Cl~doroncelale (17~~ ,21 -Dihydrozy-.&

pregm-3,fl ,Z~-triol~e-Zl-c/~loroacelute)--Over a $ hour period, 3.6 g of chloroacetic anhydridc in 5 ml of acetone were added dropwise to ~LII ice-cold solution of 300 nlg of cortisone in 60 ml of acetone containing 0.6 ml 0r pyridinc. The m&lure was stirred at 0” for I hour alld kept overnight at 4’. The mistune was reduced to a volume of about 20 1111 under Nz and added clropwise to a well stincd solution of ice water (400 ml). After 1 hour, the resulting prccipitatc 1~:~s collected by filtration, washed thoroughly with c-old waler, dissolved in a minimum volume of acetone, and crystalliacd by slow addition of water. This dried product had a melting point of 228 to 230”. Miscd nicking l)oint with added (sort isonc 1~21% Ion-c%rctI al)out 20”. Total yield 300 mg (8OL;;), tE;l:,tt, = 1.55 X 10L, [a]S,$‘i,:; “” + 196”.

c2Jr?~o&l

Calcuhttctl: C 63.22, II 6.G4, Cl 8.13 Found : C 62.87, I1 0.80, Cl i.OS

Synthesis of Cortisone 9l-Iodoacetate (17~ ,21-Dihph-ozy-J- pregnevz-3,11 , WO-trione-dl-iodoacetale)-Cortisolle Pl-chloroace- tate (200 mg) and I<1 (300 mg) wcrc reflused in 30 ml of acetone for 4 hours. After the reaction, acetone was evaporated to a volume of 5 ml unclcr K2 and the mislurc I\-as adtlcd t.o 25 ml of an ice-cold 596 solution of sa&O3 in water. The mixture was stirred for 15 min in the cold, and the precipitate was re- moved by filtration. It was recr),stallized from acetone by addition of water as above. Yiclcl 160 mg (660/;,,), m.p. 180” (decomposed) tFon ’ > 2J8 nlil = 1.4 x 104, [a];;;;& 239 = f160”.

C23IJ2&061 Ca1cu1111cd: c 52.27, II 5.40, 124.05 I~‘oulld : C 51.65 11 5 74 I 25.08 , .,

In mass spectral analysis 0C the compound, the major molecu- lar ion value corresponded to tl~c mass number of cortisone 21- iodoacctate (528). Gkraviolct :tbsorl)lion, masinlum at 238 nm, indicated t IIC ~~rcscncc oC a ~i-3-lietone and infrared absorp- tion bands at Ii10 rn-* and 1670 ~n1-l indicated l)resclncc of ester and l&o groups, rcspcctivcly.

Complete hydrolysis of the estcll groul) of corl isone ‘21 -ioclo- acetate was carried out by mising 2 ml of I c; metha,nolic solu- tion oC cortisone 2I-iodoacetalc with 2 1~11 of 0.5 N 1u3&03. After 1 hour at 23”, the mixture was acidified with 1 N IICI and the methanol was evaporated uldcr St. The prccipilatecl solid was rcmovcd by ccntrifugnlion ant1 washed several times with cold water. It was finally dried in a vacuum tlcsiccator. Cortisone and cortisone ncctale wore Gnlilarly tre:ltctl. Infrared spectra and melting points of the three snmplca were dcttrmincd and found to be iclentical.

Synthesis of Rndioaclive Cortiso?le RI -Iodoacetnte-Iodoacetic- 2-% ncitl (10 mCi) was mixed with u~llabeled iodoacet.ic acid and specific activily of the mixture was determined. A solu- tion of 0.24 mmolc of iodoacetic acid (22 mCi per mmolc), 0.1 mnlole of cortisone, and 0.3 mmolc of pyridine in 4.5 ml of dry CH&l, was cooled to 0” and reacted with 0.25 mmole of clicyclo- hcsyl carbodiimidc in 0.1 ml of cold CH&12. Within 5 min, dicyclohexylurca precipitated out. The suspension was stirred at 0” for I hour and at 23” for another hour. Then, 25 ~1 of mlacial acetic acid were added t)o this susllension, and it was b stirred at 23’ for 15 min. Dichloromethane was evaporated under Nz. The solid residue was taken up in 3 ml of a.cetone and filtered. The filtrate was poured into 50 ml of a 2% N&+$03

by guest on January 29, 2020http://w

ww

.jbc.org/D

ownloaded from

3645 Affinity Labeling of Steroid Binding Sites Vol. 246, No. 11

TABLE I

E$ect of pH on hydrolysis of cortisone 21 -iodoacetafea

PH Time

min

7.0 30 60

180 300

8.0 30 60

180

- I

-

Original cortisone Zl-iodoacetate remaining

%

80 65 40 20

65 35 15

a Presence of 37.5 pg of enzyme did not alter rates of hydrolysis.

solution in water at 0”. The suspension was stirred for 15 min and then filtered. The precipitate was taken up with 5 ml of acetone and filtered to remove any remaining dicyclohexylurea. The filtrate was evaporated to a small volume under Nz and reprecipitated with cold mater. It was washed with a small volume of cold methanol and finally crystallized as white crystals from aqueous acetone. Yield 15 mg, m.p. 180” (decomposed). It gave a single spot on thin layer chromatography responding to Gmelin and Virtanen’s (10) iodine test and had a mobility identical with cortisone 21-iodoacetate prepared through chloro- acetic anhydride and KI.

Synthesis of 21-Iododeoxycortisone (21-Iodo-17cr-hydroxy+ pregnen-J, 11,20-trione)-The compound was synthesized by refluxing acetone solution of NaI and cortisone tosylate which was prepared by reacting 1 g of cortisone in 10 ml of dry pyridine and 900 mg of p-toluene sulfonyl chloride in 9 ml of dry CI-IJ& at -25” for 40 hours, similar to the method of Borrevang (11). The product gave a single spot, on thin layer chromatography with positive test to Gmelin’s reagent. It had a melting point (155’, decomposed) identical with that reported by Borrevang

(11).

RESULTS

The stability of cortisone 21.iodoacetate was evaluated by placing 0.3 pmole (in 0.3 ml of alcohol) in 4.7-ml solutions of 0.05 M phosphate buffer at pH 7.0 and 8.0, 23”. Steroids were extracted with chloroform and separated by thin layer chroma- tography. A new product with the RF of cortisone was seen to increase as a function of time. The approximation of remaining cortisone 21.iodoacetate was made by the intensity of absor- bance of light at 240 nm. As shown in Table I, hydrolysis does occur much more rapidly at pH 8.0. Presence of 37.5 pg of 20&hydroxysteroid dehydrogenase caused no appreciable difference. The decision was made to carry out all further studies at pH 7.0.

Alkylation of Reduced Glutathione and 6,5’-Dithiobis(2-nitro- benzoic acid) with Cortisone 21 -Iodoacetate, dl-Iododeoxycortisone, and Iodoacetic Acid-The colored anion was formed by reacting 0.2 Hmole of 5,5’-dithiobis(2-nitrobenzoic acid) and 0.2 pmole of GSH in 2.8 ml of 0.05 M phosphate buffer, pH 7.0. Then, the appropriate iodo compound (steroid or iodoacetic acid) in 0.2 ml of alcohol was added, and absorbance at 412 nm was followed at different intervals of time. In another experiment, 0.2 pmole of reduced glutathione was similarly reacted with the iodo compounds, and the remaining free thiol group was deter- mined with 5,5’-dithiobis(2-nitrobenzoic acid) solution (12).

1.. . . . . . I. 24 48 72

TIME hinutes)

FIG. 1. R.eaction of iodo compounds with colored anionic form of 5,5’-dithiobis(2-nitrobenzoic acid). Iodo compounds in the amounts shown were added separately in 0.20 ml of alcohol to the previously formed color of reduced glutathione and 5,5’-dithiobis- (Z-nitrobenzoic acid) (0.2 pmole each) in 2.80 ml of 0.05 M phosphate buffer, pH 7.0, 23”.

The results are shown in Figs. 1 and 2. It can bc seen that both 21-iododeoxycortisone and cortisone 21-iodoacetate react rapidly to alkylate the model sulfhydryl compounds The former, in both cases, reacts more rapidly than the latter, and both rea& more rapidly than iodoacetic acid.

Evidence That Cortisone 2l-Iodoacetate and M-lododeoxy- cortisone Bind at Catalytic Site of 2Op-Hydroxysteroid Dehydroge- nase-Assays, carried out as described above with 0.20 pmole of NADH and the steroids in a total volume of 3.0 ml, revealed that both are substrates. Both must bind at the enzyme active site, as the catalytic step can be carried out. Further, apparent initial velocities were obtained during these 3-min assays. The Michaelis constants of cortisone 21-iodoacetate and 21-iodode- oxycortisone, estimated by the method of Lineweaver and Burk (13), are shown to be 1.0 x 10m4 and 1.4 x 10e4 M, respectively (Fig. 3). Under similar conditions, the Michaelis constant of cortisone was 5.1 X 10e5 M.

Inactivation of 2O&Hydroxysteroid Dehydrogenase-The en- zyme (37.5 yg) was incubated in phosphate buffer, pH 7.0, at 23”, in the dark in the presence of various reagents. At various times, aliquots were assayed. The results are shown in Fig. 4. It can be seen that 6 X lop5 M 21-iododeoxycortisone, 1.2 X 1.0v4 M iodoacetic acid, 1.2 X 10e4 M iodoacetic acid with 6 X 1O-5 M cortisone, and even 6 x lop4 M iodoacetic acid failed to inactivate the enzyme. Additional experiments revealed that 1 X 10-h M 21-iododeoxycortisone also failed to inactivate it. On the other hand, 6 x 10u5 M cortisone 21-iodoacetate effected a progressive inactivation of the enzyme over the period of observation. Further, presence of 6 X 10e5 M cortisone tlefi- nitely slowed the cortisone 21-iodoacetate-induced inactivation. These results were obtained in several repetitive experiments, of which Fig. 4 illustrates a typical example.

The irreversibility of the inactivation of 20@-hydroxysteroid dehydrogenase was ascertained as shown in Fig. 5. It can be seen that the mixing of a 4-fold excess of 2-mercaptocthanol with cortisone 21-iodoacetate 30 min before its addition to

by guest on January 29, 2020http://w

ww

.jbc.org/D

ownloaded from

Issue of June 10, 1971 M. Ganguly and J. C. Warren

b

20 40 60

TIME (minutes)

FIG. 2. Reaction of iodo compounds with reduced glutathione. Reduced glutathione (0.2 pmole) was reacted with iodo compounds in 0.05 M phosphate buffer, pH 7.0, 23”, for times shown. Un- reacted glutathione determined by reaction with 5,5’-dithiobis- (2-nitrobenzoic acid) which develops full absorbance in 1 min.

1.2 1

-2 2 4 6 5 IO 12

! XlC4 m

FIG. 3. Double reciprocal plot of reduction of 21-iododeoxy- cortisone and cortisone 21-iodoacetate by 20p-hydroxysteroid dehydrogenase with assay conditions described in the text. V, velocity in nmoles of NAD+ generated per min. [S], molar con- centration of substrate. Each point is the mean of five deter- minations.

20&hydroxysteroid dehydrogenase renders the steroid incapable of inactivating the enzyme. Addition of the same quantity of 2-mercaptoethanol after partial inactivation by cortisone 21- iodoacetate stops (but does not reverse) the inactivation. Simi- lar lack of reversibility was seen on dialysis.

Treatment of f?O@Hydroxysteroid Dehydrogenase with Radio- active Cortisone 21 -Iodoacetate and Iodoacetic Acid and Subsequent Analysis-To 10 mg of enzyme (lo-’ mole) in 9.4 ml of 0.05 M phosphate buffer, pH 7.0, 0.5 ml of ethanol containing 0.55 nmole of tritiated cortisone 21-iodoacetate was added. Six hours later, 0.1 ml of ethanol containing 0.55 pmole of tritiated cortisone 21-iodoacetate was added, and the incubation con-

60 120 18(

TIME (minutes)

FIG. 4. Effect of iodo compounds on the activity of 2Op-hy- droxysteroid dehydrogenase. Incubations conducted with 37.5 pg of enzyme in 4.7 ml of 0.05 M phosphate buffer, pH 7.0, 23”, to which was added at zero time: (a) 0.3 ml of alcohol; alcohol containing 0.3 rmole of 21-iododeoxycort,isone; alcohol containing 0.6 rmole of iodoacetic acid and 0.3 pmole of cortisone; alcohol containing 4.0 Mmoles of iodoacetic acid; alcohol containing 0.3 pmole of cortisone (all indicated by +--+); (5) 0.3 ml of alcohol containing 0.3 rmole of cortisone 21-iodoacetate (A-A) ; and (c) 0.3 ml of alcohol containing 0.3 rmole of cortisone 21-iodoace- tate and 0.3 pmole of cortisone (W--m). At times indicated, 0.1 ml of this solution was assayed as described in the text with -aturation concentrations of substrate and cofactor. Values are

he means of five assays.

g 25.

iti P

I I 1 1 2 3 4

TIME (hours)

FIG. 5. Irreversibility of inactivation of 20fi-hydroxysteroid dehydrogenase by cortisone 21.iodoacetate. Prior incubations, set up as described in Fig. 4, with addition of 0.3 ml of alcohol containing: (a) nothing or 1.2 pmoles of 2-mercaptoethanol or 0.3 pmole of cortisone 21-iodoacetate and 1.2 rmoles of 2-mercapto- ethanol which had been mixed 30 min before addition (all indicated by 0); and (5) 0.3 pmole of cortisone 21-iodoacetate (A, A). At times indicated, O.l-ml aliquots were assayed as described in Fig. 4. After assay at 3 hours, 1.2 pmole of 2-mercaptoethanol was added to one reaction vial containing cortisone 2Kodoacetate (A). Values are the means of four assays.

tinued for a further 18 (hours (all at 23”). To a second sample of enzyme, two additions (0.55 pmole each) of tritiated iodo- acetic acid were similarly made. To a third sample, only alcohol was added. (To conserve enzyme, this third sample was, in

by guest on January 29, 2020http://w

ww

.jbc.org/D

ownloaded from

3650 Afinity Labeling of Steroid Binding Sites Vol. 246, No. 11

FIG. 6. Distribution of radioactivity on paper chromatography of the hydrolysates of BO/%hydroxysteroid dehydrogenase after t,reat- ment with tritiat’ed cortisone 21-iodoacetate (upper scan) and tritiated iodoacetic acid (lower scan). Positions of standard amino acids on the chromatogram are gicen in the center. Origin on the left.

all respects, performed on a one-tenth scale.) Serial assays of diluted aliquots of all samples were carried out in quadruplicate. At the conclusion of the 24-hour incubation, cortisone 21-iodo- acetate had induced a 57% loss of enzyme activity, whereas control and iodoacetic acid samples still retained 96% of the original activity.

The enzyme samples incubated with cortisone 21-iodoacetate and iodoacetic acid were then separately dialyzed against glass- redistilled water with frequent changes, until the external com- partment contained only insignificant quantities of radioactivity. Contents of the bags were lyophiliaed, and 3.0-mg samples were hydrolyzed separately in 4.0 ml of 6 N HCI. The hydrol- ysate were evaporated to dryness and freed of HCl by repeated evaporation of the aqueous solution. The residues were dis- solved in 0.20 ml of water.

Then, 20-~1 samples of each were applied to Whatman No. 3MM: paper, dried, and developed with butanol-acetic acid- water for 22 hours at room temperature. The chromatogram was dried at 50” in a vacuum chamber. Amino acids were detected with ninhydrin spray. After drying, the chromatogram was cut, and iodoacetic acid and cortisone 21-iodoacetate- treated samples were scanned on the strip scanner. Results are shown in Fig. 6. It can be seen that a single peak of radioactiv- ity is present in the hydrolysate of the cortisone al-iodoacetate- treated enzyme, which has an RF in the range of, but not identi- cal with, standard phenylalanine. No such labeling is present in the hydrolysate of iodoacetic acid-treated enzyme.

The remaining portion of the hydrolysates was evaporated to dryness. The dried residues were dissolved in 4.0 ml of 0.2 M sodium citrate buffer, pH 3.25. Aliquots of each solution (1 ml) were analyzed on the amino acid analyzer. No peaks of radioactivity were present in the case of iodoacetate treatment. In the case of cortisone 21-iodoacetate treatment, a major radio- active fraction was seen (Fig. 7). Of known carboxymethyl derivatives, it had a mobility resembling only 1 ,3-dicarboxy- methyl histidine. It was clearly not carboxymethylcysteine which consistently eluted at 48 ml of effluent.

Synthesis of carboxymethyl histidines was carried out by reacting 225 mg of N-acetylhistidine and 270 mg of freshly decolorized iodoacetic acid according to the method of Crest- field et al. (4). After acid hydrolysis, as with the enzyme, the products were evaporated to dryness and freed of HCl. The residue was dissolved in 10 ml of water, and 0.1 ml was mixed with another l.O-ml aliquot of the 0.2 M citrate solution con- taining the hydrolysate of cortisone 21-iodoacetate-treated enzyme for cochromatography on the amino acid analyzer. As shown in Fig. 7, the radioactivity had a mobilit,y identical with standard 1,3-dicarbosymethyl histidine.

Labeled cortisone 21-iodoacetate-treated enzyme hydrolysate (1 ml) was mixed with 0.01 ml of the carboxymethyl histidine derivatives, and the mixture was lyophilized. The residue was extracted with 1 ml of methanol, and the methanol solution was evaporated to dryness. The residue was taken in 50 ~1 of methanol and applied to Whatman No. 3MM paper. Then, 0.01

by guest on January 29, 2020http://w

ww

.jbc.org/D

ownloaded from

Issue of June 10, 1971 M. Ganguly and J. C. Warren 3651

FIG. 7. Amino acid analysis of synthetic carboxymethyl histidines (lower tracing) and hydrolysate of radioactive cortisone 21.iodo- acetate-treated enzyme after addition of synthetic carboxymethyl histidines (upper tracing). Solid lines represent ninhydrin absorb- ance at 440 nm. Broken line indicates the only major radioactive peak which contained 86% of total radioactivity.

ml of the carboxymethyl histidines in water was mixed with 1 ml of 0.2 M sodium citrate buffer, pH 3.25, worked up similarly and applied to the origin of the same chromatogram. The developed chromntogram, after drying, was sprayed with nin- hydrin and divided. The chromatogram of the radiolabeled enzyme hydrolysate was cut into l-cm pieces which were eluted with 1 ml of water and counted with 10 ml of phosphor. The mobility of the radioactive peak was coincident with the fastest moving carboxymethyl histidine derivative (RF = 0.57).

Reaction between N-Acetylhistidine, N-Acetyltryptophan, and Iodoacetic Acid-N-Acetylhistidine (25 pmoles) was incubated with 250 pmoles of labeled iodoacetic acid in 5.0 ml of 0.05 M

phosphate buffer, pH 7.0, at 23” for 48 hours in the dark. A similar incubation was carried out with N-acetyltryptophan (25 pmoles) replacing N-acetylhistidine. Finally, a control was pre- pared with labeled iodoacetic acid alone. The mixtures were lyophilized and extracted with 2 ml of alcohol. Aliquots of the alcoholic solutions (25 ~1) were applied on Whatman No. 3MM paper and developed as usual. The dried chromatograms were scanned for radioactivity. In the case of iodoacetate and N- acetylhistidine, two new peaks of radioactivity (RF = 0.25 and 0.38) were noted, in addition to the iodoacetic acid peak (RF = 0.88). In the case of iodoacetic acid alone and iodoacetic acid with N-acetyltryptophan, the mobility of the radio peak was unaltered (RF = 0.88). Visualization of N-acetyltryptophan with Ehrlich’s reagent revealed that its mobility (RF = 0.77) was unchanged by mixing with iodoacetic acid. These observa- tions indicate that, whereas carboxymethylation of N-acetyl- histidine occurs under conditions used for treatment of the en- zyme, no reaction occurs with N-acetyltryptophan.

DISCUSSIOiX

The synthesis of cortisone 21-iodoacetate followed basic chemical principles. The product shows ultraviolet absorbance

at 238 nm (A4-3-ketone), gives chemical, infrared, and mass spectral analyses which agree with the chemical structure, and on hydrolysis in 0.5 N Na&Os yields a compound which has a melt- ing point and infrared spectrum identical with that of cortisone. The identity and chemical structure of 21Gododesoxycortisone were established by comparison of properties with those listed in the literature (11).

Both steroids react rapidly with model sulfhydryl compounds. Unfortunately, cortisone 21-iodoacetate undergoes considerable spontaneous hydrolysis in aqueous solut,ions under physiological conditions so that any such solution will ultimately cont.ain mixtures of cortisone 21-iodoacetate, cortisone, and iodoacetic acid.

That cortisone 21-iodoacetate and 21-iododeoxycortisone can carry out the reversible binding step at the active site of 20& hydroxysteroid dehydrogenase is shown by the fact that they are substrates for the enzyme. Further, within the limitations of using K, values to approximate Kd values, they seem to have approximately equal affinity for the site. Finally, during these assays, conducted over a 3- to 5-min period and initiated by addition of enzyme, rates were linear, indicating that, if an ir- reversible covalent bonding occurs, it is either minimal during this time period or does not prohibit catalytic activity.

Inactivation of the enzyme of cortisone 21-iodoacetate on more prolonged periods of exposure indicates the irreversible step does occur. On the other hand, no such inactivation was induced by similar concentrations of 21-iododeoxycortisone, suggesting that some enzyme group, other than a sulfhydryl, may be participat- ing in the affinity labeling process. No inactivation by similar concentrations of iodoacetic acid (and, indeed, even concentra- tions 10 times in excess) is noted. Further, cortisone does not facilitate inactivation by iodoacetate. These observations are all compatible with the conclusion that inactivation of 2Op- hydroxysteroid dehydrogenase by cortisone 21-iodoacetate is an

by guest on January 29, 2020http://w

ww

.jbc.org/D

ownloaded from

Ajinity Labeling of Steroid Binding Sites Vol. 246, No. 11

affinity labeling process by which a steroid carboxymethylation is effected with some amino acid, probably other than cysteine. This being the case, cortisone would be expected to occupy the site, exclude cortisone 2l-iodoacetate, and slow the rate of the inactivation. No inactivation is seen when cortisone 21-iodo- acetate is reacted with excess 2-mercaptoethanol prior to placing it with the enzyme because it alkylates 2-mercaptoethanol and is no longer capable of reacting with an amino acid residue in the enzyme. Finally, the inactivation induced by cortisone 21- iodoacetate is irreversible, as shown by failure of recovery on addition of 2-mercaptoethanol or dialysis. The irreversibility is compatible with covalent bonding at or near the active site.

As shown by chromatography (Fig. 6), hydrolysates of enzyme treated with tritiated cortisone 21-iodoacetate contain a single major radioactive peak, suggesting covalent bonding of the steroid iodoacetate to a single amino acid residue. No radio- active peak is seen with hydrolysates of enzyme treated with iodoacetate. Thus, iodoacetate, which does not inactivate the enzyme, also fails to carboxymethylate it.

On the ammo acid analyzer, the carboxymethyl derivative is clearly shown not to be carboxymethylcysteine, but to have a mobility suggesting its identity as 1,3-dicarboxymethyl histi- dine. When standard 1,3-dicarboxymethyl histidine is co- chromatographed with enzyme hydrolysate, the tritiated deriva- tive is shown to have a mobility that is identical with it. On paper chromatography of the hydrolysate and st,andard carboxy- methyl histidines, mobility is again identica1 with the fast moving dicarboxymethyl derivative.

Because tryptophan tolerates the acid hydrolysis procedure poorly and gives rise to artifact formation, it was considered that the radioactive carboxymethyl derivative might be such an artifact, which by coincidence, was identical in mobility with 1,3-dicarboxymethyl histidine. An intensive literature search revealed no reports of carboxymethylation of tryptophan (14, 15). Further evidence that this is unlikely is supplied by failure of carboxymethylation of N-acetyltryptophan under the condi- tions of enzyme inactivation, whereas N-acetylhistidine is easily carboxymethylated.

It is interesting that a dicarboxymethyl derivative is formed on affinity labeling of 20&hydroxysteroid dehydrogenase by corti- sone 21-iodoacetate. A small radioactive peak with the mo- bility of 3-carboxymethyl histidine appears to be present on characterization of the hydrolysate of inactivated enzyme in the amino acid analyzer. It accounted for only 5% of the radio- activity in the major peak. This suggests that the first reaction is formation of 3-steroid carboxymethyl histidine, a step which requires the steroid to deliver the reagent group. Subsequent reaction at carbon atom 1 of histidine may be effected by corti- sone 21-iodoacetate, or it is possible that the first reaction exposes the histidine to such a degree that it is susceptible to a second

carboxymethylation by iodoacetic acid, which is always present after a few minutes of incubation as a result of hydrolysis of the steroid iodoacetate. The hydrolysis step removes the steroid so that one is unable to differentiate between these possibilities.

Cortisone 21-iodoacetate clearly carboxymethylates cysteine and histidine. Reports of carboxymethylation of lysine, methi- onine, and homoserine by various iodoacetate reagents (16) sug- gest that it may well (particularly if concentrated by the reversi- ble binding step at a binding site) react with these residues if they are present. 21-Iododeoxycortisone clearly alkylntes cysteine. Reports of alkylation of tyrosine and tryptophan liy various alkyl halides (17) suggest that it might also react with these residues.

Therefore, these compounds may be useful to affinity label steroid receptor proteins in various target organs if these recep- tors have present any of the above amino acid residues in the D ring area of their binding sites. Such affinity labeling may allow not only the characterization of the binding site but also the decision as to the obligatory role of these receptors in the mech- anism of steroid action.

Acknowledgment-We thank Dr. Joe R. Kimmel for amino acid analyses and valuable discussion.

1.

2.

3. 4.

5.

G. 7.

8.

9.

10.

11. 12. 13.

14.

15. 16.

17.

REFERENCES CHIN, C-C., AND WARREN, J. C., J. Biol. Chem., 243, 5056

(1968). MULDOON, T. G., AND WARREN, J. C., J. Biol. Chem., 244,

5430 (1969). CHIN, C-C., AND WARREN, J. C., Biochemistry, 9, 1917 (1970). CRESTFIELD, A. M., STEIN, W. I-I., AND MOORE, S. M., J. Biol.

Chem., 238, 2413 (1963). ZWEIG, G., AND WHITAICER, J. R., Paper chromatography,

Vol. 1, Academic Press, New York, 1967, pp. 54, 96. SMITH, I., Nature, 171, 43 (1953). SPACKMAN, D. II., MOORE, S., AND STEIN, W. II., Anal. Chem.,

30, 1190 (1958). BETZ, G., AND WARREN, J. C., Arch. Biochem.. Biophys., 128,

745 (1968). BETZ, G., Ph.D. thesis, University of Kansas Medical Center,

1968. GMELIN, R., AND VIRTANEN, A. I., Acta Chem. &and., 13, 1469

(1959). BORREVANO, P., Acta Chem. &and., 9, 587 (1955). ELLMAN, G. L., Arch. Biochem. Biophys., 82, 70 (1959). LINEWEAVER, I-I., AND BURK, D., J. Amer. Chem. Sot., 66,

658 (1934). VALLEE, B. L., AND RIORDAN, J. F., Annu. Rev. Biochem., 38,

739 (1969). GLAZER, A. N., Annu. Rev. Biochem., 39, 119 (1970). GURD. F. R. N.. in S. P. COLOWICX AND N. 0. KAPLAN (Etl-

itors), Methods in enzymology, Vol. XI, Academic Press, New York, 1967, p. 532.

HORTON, H. R., AND KORHLAND, D. E., in 6. P. COI,OWICIC AND N. 0. KAPLAN (Editors), Methods in enzymology, Vol. XT, Academic Press, New York, 1967, p. 556.

by guest on January 29, 2020http://w

ww

.jbc.org/D

ownloaded from

Manik Ganguly and James C. WarrenDEHYDROGENASE

-HYDROXYSTEROIDβ21-IODOACETATE AND STUDY OF 20Affinity Labeling of Steroid Binding Sites: SYNTHESIS OF CORTISONE

1971, 246:3646-3652.J. Biol. Chem. 

  http://www.jbc.org/content/246/11/3646Access the most updated version of this article at

 Alerts:

  When a correction for this article is posted• 

When this article is cited• 

to choose from all of JBC's e-mail alertsClick here

  http://www.jbc.org/content/246/11/3646.full.html#ref-list-1

This article cites 0 references, 0 of which can be accessed free at

by guest on January 29, 2020http://w

ww

.jbc.org/D

ownloaded from