C19O2 and Some of Their Precursors in Blood from Normal Human Adrenals

Ralph G. Wieland, … , Antonia P. Zala, H. Hirschmann

J Clin Invest. 1965;44(1):159-168. https://doi.org/10.1172/JCI105122.

Research Article

Find the latest version: https://jci.me/105122/pdf Journal of Clinical Investigation Vol. 44, No. 1, 1965

C.902 Steroids and Some of Their Precursors in Blood from Normal Human Adrenals *

RALPH G. WIELAND, CONSTANCE DE COURCY, RICHARD P. LEVY, ANTONIA P. ZALA, AND H. HIRSCHMANN (From the Department of Medicine, Western Reserve University and University Hospitals, Cleveland, Ohio)

The ready response of the main urinary 11- Methods deoxy- 17- to adrenal suppression and Patients. The subjects studied were five women (I.A., adrenal stimulation leaves no doubt that a con- M.R., C.B., M.M., and R.R.) and three men (J.J., L.W., siderable fraction of urinary androsterone (3a- and F.B.). Their ages and serial numbers are given in hydroxy-5a-androstan-17-one), 5,8-androsterone Table III. R.R. had hirsutism but normal ovulatory (3ca-hydroxy-5/8-androstan-17-one), and andros- menses and F.B. had diabetes. There was no other evi- dence of endocrinologic disease or hepatic dysfunction in tenolone (3,8-hydroxyandrost-5-en-17-one, dehy- any of these subjects. The patients fasted on the morn- droepiandrosterone) must be derived from adrenal ing of the procedure. Secobarbital (100 mg orally) was precursors. However, when adrenal vein blood given at 8 a.m., and 50 mg meperidine hydrochloride was was examined for such compounds, no consistent given subcutaneously just before the catheterization,' pattern emerged (3-9) ; therefore the identity of which was usually done between 10 a.m. and noon es- sentially in the manner performed by Cranston (5). the precursors remained in doubt. The presence The location of the catheter was ascertained by veno- of (androst-4-ene-3,17-dione), gram. In most patients this produced sufficient pain to which was tentatively identified by Pincus and require additional meperidine. This was unnecessary Romanoff (3), was not confirmed by Lombardo, with L.W. and F.B., who had minimal discomfort. Blood and Hudson who examined 12 was collected after the iv injection of 100 mg of heparin. McMorris, (6), The adrenal blood sample (usually about 100 ml) was samples of human adrenal blood. In fact, these collected during a measured interval. This was followed investigators (6) obtained from only one indi- by the withdrawal of a similar volume of peripheral vidual a substance, androstenolone, that could be blood. Thereafter, corticotropin was administered to classed as an efficient precursor of the urinary R.R. and F.B., and a second sample was taken from the adrenal vein. In the case of R.R. this collection was be- 11-deoxy-17-ketosteroids. The scope of the in- gun 10 minutes after the completion of an iv infusion (11 vestigation widened when Baulieu (8) reported minutes) of 25 U of corticotropin and of F.B. 16 min- that in two instances, adrenal tumors secreted utes after the rapid injection of 40 U of corticotropin. but no free androstenolone. Radioactivity measurements. Radioactivity was de- androstenolone sulfate termined by scintillation counting with a two-channel The purpose of this communication is to document spectrometer. Free steroids were dissolved in 5 ml of and extend the findings of our preliminary re- scintillator solution (toluene containing 20 mg of 2,5- ports (1, 2) and to show that the normal human diphenyloxazole and 0.5 mg of 1,4-bis-2-(5-phenyloxa- zolyl)benzene), sulfates in 0.3 ml of methanol before 5 ml adrenal consistently secretes androstenolone sul- of the scintillator solution was added. When methanol fate, androstenolone, and androstenedione. These was present, corrections for the diminished efficiency of three compounds appear to be the principal pre- counting were determined by internal standards. cursors of the urinary 1 1-deoxy-17-ketosteroids. In calculating the isotope ratios C"2/HI,, we used the equation, C"2/H'L = [(N2/N1) - a] /[1 - (1/b) (N2/N1) ]. In this expression, the N values are the observed rates of * Submitted for publication July 13, 1964; accepted scintillation corrected for background; the subscripts 1 October 1, 1964. Supported by U. S. Public Health Service grant 1 A. more detailed description will be given elsewhere C-1679, a research fellowship under training grant by Levy and Katz. 2A-5293, and a research career award K6-AM-14,367. 2 Tricarb, Packard Instrument Co., La Grange, Ill. Preliminary accounts of some of our findings have Samples from Patients 1 to 5 were analyzed on a model been given (1, 2). 314 AX, the others on a model 314 EX. 159 160 WIELAND, DE COURCY, LEVY, ZALA, AND HIRSCHMANN

TABLE I and 17-hydroxypregnenolone-7-H3 (3,8,17-dihydroxypregn- Test for radiochemical purity of 5-en-20-one) were the same preparations as we described androstenolone-7-H3 sulfate* before (9) and were generally used in the same amounts. They were periodically tested for purity by paper chro- Fraction Solvent Weight matography and scanning4 for radioactivity. Cortexo- lone-1,2-H' (17,21-dihydroxypregn-4-ene-3,20-dione) was mg cpm/lmole a commercial sample -that was purified by Girard sepa- Mother liquor 4.8 1,375 ration and chromatography on paper in system 1. An- Methanol Crystals 8.2 1,318 drostenediol-4-C"4 (androst-5-ene-3,3,17p8-diol) was pre- pared from commercial -4-C" by a procedure Mother liquor 0.8 1,347 essentially like the one subsequently published by Gut and 95%aO Ethanol Crystals 6.8 1,305 Uskokovic (13). The resulting diace- tate melted at 161.5 to 1630 C; 5 the free diol melted at Sample solvolyzed 181 to 183° C and showed a single symmetrical peak when and recrystallized Methanol Crystals 0.3 1,391 chromatographed in system 7. Androstenolone-7-H3 sulfate 6 was prepared from the free (1.9 Ag, 1.1 X * Nonradioactive androstenolone sulfate (13 mg) was 106 cpm N3) in 1 ml of chloroform with 50 mg of pyridine- added before recrystallization. Samples were assayed for sulfur trioxide according to the method of McKenna and radioactivity and for steroid content by the Zimmermann Norymberski (15). The product was converted to the reaction. potassium salt by treatment with aqueous potassium hy- and 2 refer to the two channels of the spectrometer. droxide. The solution was then washed with ether and The symbol C"2, therefore, signifies the number of disinte- extracted with n-butanol, which was washed with water grations of C" that are registered in channel 2 per unit and taken to dryness in vacuo. The product was chro- time. The efficiency ratios, a and b, equal the N2/N, matographed on paper in system 6. The radioactive zone ratios of reference samples that contain only a single that had the same Rr as a recrystallized unlabeled refer- isotope, H' or C", respectively. Spectrometer 2 model ence sample (mp, 219 to 2210 C, mp in reference 16, 219 314 AX was operated at settings that resulted in efficiency to 2230 C) was eluted and used as tracer [7,000 cpm N. ratios a 0.2 and b 5.1; model 314 EX gave a 0.05 (corrected) /sample]. Its purity was tested by dilution of a sample with cold carrier, recrystallization, and sol- and b _- 11.8. Color reactions and spectroscopy. The Zimmermann volysis (16) followed by recrystallization. The results are reaction was carried out essentially according to the shown in Table I. At its last use as tracer the prepara- modification described by Wilson (10). In Cases 1 to 5, tion was still in its conjugated state, since only 1%o of the Porter-Silber reaction was done on the neutral water- the radioactivity could be extracted from water with soluble ketonic plasma fraction as described (9); in the benzene. others, on 5-ml samples of plasma by the method of The preparations of acetic anhydride were commercial Peterson, Karrer, and Guerra (11). Infrared spectra samples that had been diluted with benzene. Unless noted were measured with a double beam spectrometer 3 with otherwise the one containing H' introduced about 250 cpm beam condenser on samples (ca. 17o) embedded in KBr HW, into 1 ms.mole of a monohydroxysteroid; the one disks (1.5 mm). with C", about 180 cpm C"2 into 1 mumole. Precise Chromatography. Systems for chromatographic sepa- counts were determined by acetylation of reference com- rations were prepared from mixtures of the following pounds of known weight that contained the same amounts solvents in the volume ratios specified: 1) toluene, iso- of steroid tracer as the unknowns. These reference ace- 6ctane, n-heptane, methanol, and water, 10: 5: 5:16: 4; tates were prepared at the same time and purified in the 2) iso6ctene, methanol, and water, 5: 4: 1; 3) benzene, same manner as the acetates from blood. The amount of iso6ctane, methanol, and water, 10: 20: 12: 3; 4) ligroin, steroid in the test sample (s ,ug) was calculated by the methanol, and water, 10: 9:1; 5) isopropyl ether, t-bu- following equation, s = t [ (H31/C"2)unk. ( C142/H31) ref. tanol, and 2.8% aqueous ammonia, 3: 2: 5; 6) toluene, - 1], when t ,ug of a C"4-labeled steroid tracer had been n-butanol, and 2.8% aqueous ammonia, 1: 1: 2; and 7) added to the plasma or plasma fraction. The subscripts toluene, iso6ctane, methanol, and water, 15: 5: 16: 4. unk. and ref. relate to the tritiated acetates from blood We have previously described solvent systems 1 to 4 (9); and from the pure steroid tracer, respectively. When Schneider and Lewbart (12) described systems 5 and 6 tritiated steroid tracers of negligible weight were used, for the fractionation of steroid sulfates. Eluates from the same amount of radioactivity was added to the un- blank chromatograms were prepared for correcting all known solution containing s ug of steroid and to the spectrographic data except the infrared curves. unlabeled reference sample (r ,lg). In this case, the Materials. The radioactive tracers, androstenedione- 4Atomic Accessories, Bellerose, N. Y. 4-C", testosterone-4-C", 17-hydroxyprogesterone-4-C" 5All melting points reported are corrected. (17-hydroxypregn-4-ene-3,20-dione), androstenolone-7-H3, 6A procedure for the preparation of labeled sulfate 3 Perkin-Elmer, model 421, Perkin-Elmer Corp., Nor- with much lower specific activity has been reported previ- walk, Conn. ously (14). ADRENAL VEIN STEROIDS 161 unknown (unk.) and reference (ref.) acetates-C' are re- n-butanol, which was then washed with water and taken lated by the equation, s = r ( C42/H'1) up. (H31/C"42) ref.. to dryness in vacuo (sulfate fraction). Fractionation procedures. 1) Free ketonic steroids. The sodium bicarbonate fraction and the washings Our basic procedure for the separation of free ketonic (case F.B., sample A2) were combined, acidified (pH 1), steroids of human plasma has been described in detail and extracted with n-butanol. The extract was washed before (9) and entails the following steps: 1) precipi- with water and taken to dryness in vacuo. The residue tation of proteins by the addition of a mixture (3: 1, vol: was dissolved in 0.5 M sodium acetate buffer (6 ml, pH vol) of ethanol and ethyl ether; 2) evaporation of the 5.0) and incubated with 10 ml of 8-glucuronidase (Keto- filtrate in vacuo; 3) partition of the residue between dase, 50,000 U) at 38° C for 24 hours. The mixture methylene chloride and water; 4) removal of lipid im- was extracted with ether, which was washed with hydro- purities by extracting the residue of the organic phase chloric acid, sodium hydroxide, and water. The residue with methanol, by precipitation with water, and by wash- contained 2% of the C"2 and 4% of the H'1 counts added ing of the resulting supernatant solution in dilute methanol to the plasma. Labeled androstenolone, androstenedione, with petroleum ether; 5) Girard separation and removal and testosterone were added, and the mixture was frac- of acidic compounds; 6) partition of the neutral ketonic tionated by Girard separation and partition column and fraction between benzene and water; 7) partition chro- paper chromatography as described for the free steroids. matography on powdered cellulose in system 3 of the The sulfate fraction was applied to a partition column fraction soluble in benzene (the efficacy of this step for containing 12 g cellulose and 6 ml of the lower phase of the separation of androstenedione, androstenolone, 17- system 5 and was eluted with the upper phase in 5-ml hydroxyprogesterone, and 17-hydroxypregnenolone is il- portions. Labeled androstenolone sulfate was usually lustrated in reference 9); 8) chromatography, if appropri- found in fractions 25 to 37. These were combined and ate, on alumina; chromatography on paper for the fur- chromatographed on paper in system 6. The area con- ther purification of these steroids and of testosterone; taining radioactivity was extracted; the eluate was as- acetylation and recrystallization with carrier when ap- sayed for tritium and by the Zimmermann reaction and plicable. Full details for each compound are given in then solvolyzed. The neutral reaction product was chro- reference 9. matographed on paper in system 1. The eluate from the This procedure is an improvement over one initially radioactive zone was again assayed for tritium and 17- used (Cases 1 to 3), in which step 7 was omitted. The content. In patient J.J. (2) the isolation of method was modified further as we gained experience and the sulfates was omitted, and the aqueous fraction con- expanded our objectives. Specifically, in Cases 6 to 8 we taining the conjugated steroids and androstenolone-H3 omitted the benzene-water partition (step 6) and the sulfate was solvolyzed and after the addition of 17-hydroxy- chromatography on alumina of step 8 since these proce- pregnenolone-H' chromatographed on cellulose in system dures were no longer needed. Finally when steroid sul- 3 (9). The two fractions containing peaks of radioac- fates were investigated (Cases 6 to 8), a twofold par- tivity were chromatographed individually on paper in tition between equal volumes (30 to 75 ml) of ethyl system 1. Androstenolone was measured both by the acetate and 5% sodium bicarbonate at 4° C 7 was sub- Zimmermann reaction and by forming an acetate with stituted for the distribution between methylene chloride acetic-C1' anhydride (30 IA) in pyridine (40 Id); after and water. This change became necessary because we the addition of unlabeled carrier (20 mg) the product found that in the presence of blood lipids, water removed was recrystallized until the ratio N2/N1 for crystals and less than 6%o of labeled androstenolone sulfate from the mother liquors agreed within 2%o. 17-Hydroxypreg- organic phase.8 nenolone was acetylated in the same manner. The prod- 2) Conjugated steroids. The combined aqueous phases of the ethyl acetate-bicarbonate distribution removed all TABLE II tritiated androstenolone sulfate from the organic phases Recrystallization of adrenal 17-hydroxy- but carried about 4%o of the C"4-labeled neutral steroids.9 pregnenolone 3-acetate* The aqueous phases were combined and extracted with Sample A Sample B 7This precaution was taken because cleavage of the Fraction N2 N1 N2/Ni N2 N1 N2/NT dihydroxyacetone side chain to 17-ketosteroids has been reported to occur in alkaline solution (17). cpm cpm cpm cpm Mother liquor 340 1,049 0.324 8 After a twofold distribution of the pure tracer be- n-1 201 121 1.65 tween 4 vol of methylene chloride and 1 vol of water, Crystals 93%o of the radioactivity was in the aqueous phase. Bur- Mother liquor 128 77 1.67 260 804 0.323 n stein and Dorfman (18) observed the anomalous behavior Crystals 193 115 1.67 116 360 0.321 of this sulfate on distribution in the presence of tissue lipids. In their case, lowering of the ionic strength was * Sample A was obtained from the free steroids of J.J, sample B from the fraction that had been solvolyzed with ethyl acetate. Both con- found to decrease the amount of sulfate soluble in tained tritiated tracer, were Acetylated with acetic-C" anhydride, and recrystallized after the addition of unlabeled carrier from methanol, toluene. n = 7 times (A) and n = 5 times (B), respectively. The minimal net 9 The volume used in the distribution had no discernible count per channel was 17,000 per sample. N = observed rates of scintillation corrected for background; the effect on this. subscripts 1 and 2 refer to the two channels of the spectrometer. 162 WIELAND, DE COURCY, LEVY, ZALA, AND HIRSCHMANN

a) WU)1

i C- .= 0 C-I-4 0 Ca 040 3

0 - U) -V 0

cr) d _. ' - C V Vt~

0

(U2 0 m 0 -1 C14 C' -0 C) I 4) Ca ._ 40 W= = eop W= W-= W-= t- ~00 %0 0 cl 0 44 OC 000C . - "-I ~~~~~~~~~C-I

.5a) c) ds4 00 0 6 C'cr ._ 0 tt adCa

.3---I---I.3--I-- .a) .4 ._. $- (, C .40 -~ 14 C-4 ._ 0 3 4) 3) ct 04V4 0

CODesoeoeo oo i-- W t CN0O 1 0 d O Ui) 0 cb 0 O O r- C-d M0 C- ,-vV , 0 -40 _.4 .0 . C'sC Ca

4u a). * CS4 4) * C_ _- O - m) 0\'-4 el-I0C 1-Ut ta t--.I -o 00 O Un C 1 m ell °#- - C-I cq - Qa N0° - C-I U) - U) cOC St ¢ a) C U) *C 4 r) 0 '0o d a-t-. 0 m)) . q ) a- Ca)Q4-5

*- £ 0oo C6 U) 3-.- .10.00 .

.~ a 4)40 0> \0 C-I C) ')44 4-r 3 -o -

4i I' . m j4 U*) \0 *a-- 00 :Fz <¢. 061.# 4 C))=I- V-. Ui ~~~~~9 * .3- 9.: ADRENAL VEIN -STEROIDS 163 uct was chromatographed in system 1. The eluate from protein precipitation and methylene chloride-water par- the radioactive zone was diluted with nonradioactive car- tition in the usual manner (9) contained 83 and 81%, rier and recrystallized as shown in Table II. respectively, of the radioactivity added. Nonketonic free steroids. The neutral nonketonic To test whether our measurements of androstenolone fraction of the free steroids was adsorbed on alumina sulfate determine an artifact as Oertel (19) has sug- (1.4 g) and eluted with benzene containing decreasing gested for many procedures, we compared our method amounts of petroleum ether (75 to 0%) and benzene with one that has been reported to minimize the alleged with 0.1 %o of ethanol. The radioactive fractions were artifactual formation of the sulfate from a more com- combined, chromatographed on paper in system 1 or 7, plex lipid in peripheral blood that Oertel presumes to be a and acetylated in pyridine with acetic-H3 anhydride. The sulfatide. Tritiated androstenolone sulfate (14 X 103 cpm product was added to carrier androstenediol diacetate and N1) was added to 70 ml of plasma from peripheral blood. recrystallized from methanol until the ratio of isotopes One-half was fractionated as described for the isolation became constant. of the conjugate and was found to contain as the sulfate, Tests of procedure. Cortexolone (13 Mug, 38 X 103 cpm 126 lug of androstenolone per 100 ml of plasma. The H31) and androstenedione (2.4 ,ug, 13 X 103 cpm C1'2) were other half was extracted by method D of Oertel (19) added to 30 ml of plasma. Androstenedione was isolated and fractionated (19) on a DEAE-Sephadex (20) col- by the modified procedure that was used in the last umn. The first two fractions were solvolyzed and found three cases. The dione contained 1.2% of the H' and to contain about 2%o of the tritium added to the plasma 45% of the C" that had been added to the plasma. sample and 19 Iug per 100 ml of plasma of a 17-ketosteroid When unlabeled cortexolone (10.3 and 100 ug, re- that traveled with the Rf of androstenolone in system 1. spectively) and 1 MAg of androstenedione-C" were chro- Fraction 6 (19) contained the bulk of the tracer. The matographed on 1 g acid-washed, nearly neutral alumina, androstenolone was obtained by solvolysis and assayed in the eluates containing androstenedione-C" were found to the Zimmermann reaction after chromatography. The contain 1.1 and 4.6 Mg, respectively, of unlabeled ma- measurement after correction for losses of isotope indicated terial. Although in Cases 1 to 3 fractions that still con- that 100 ml of plasma contained 129 ug of androstenolone tained some cortexolone were chromatographed on alu- as the sulfate. As we obtained consistent and rather high mina, we estimated that cleavage to androstenedione values for the sulfate by both procedures, we did not would have yielded about 0.5 Mug (C.B.) or less per 100 pursue the matter further. ml of plasma. To test the recovery of androstenolone from blood, a Results tritiated sample was added with stirring to each of two 30-ml samples of heparinized blood. One was centri- Our results on eight samples of plasma from fuged immediately, the other after it had stood for 30 apparently normal adrenals are summarized in minutes at 0° C. The steroid fraction obtained after Table III. As this presentation implies that all

FIG. 1. INFRARED SPECTRA OF SYNTHETIC (UPPER CURVE) AND OF ISO- LATED (F.B., SAMPLE A2) ANDROSTENOLONE SULFATE. Both samples were chromatographed on paper in system 6, eluted with methanol, and embedded in potassium bromide without further purification. The samples presum- ably represent ammonium salts. A somewhat different reference curve has been published previously (21) for this compound. (The positive and the negative peaks of the upper curve near 667 cm' are instrumental artifacts.) 164 WIELAND, DE COURCY, LEVY, ZALA, AND HIRSCHMANN

TABLE IV mental section, on countercurrent distribution Recrystallization of adrenal (Table V). Androstenedione also accompanied androstenolone acetate* the tracer through two chromatographic systems. Fraction N2 N1 N2/Ni It gave the same spectrum as the reference com- pound in the ultraviolet and after the Zimmermann cPm cPm Mother liquor no. 4 1,219 654 1.87 reaction in the visible. Estimates based on these Mother liquor no. 5 1,091 584 1.87 two sets of data were in reasonable agreement (25 Crystals 430 232 1.85 and 28 Ag, respectively). A sample derivatized * Sample from J.J. containing tritiated tracer was puri- with tritiated dinitrophenylhydrazine could not be fied, acetylated with acetic-C'4 anhydride, and recrystallized separated from the corresponding hydrazone of from methanol after the addition of unlabeled androsteno- lone acetate. the dione tracer by chromatography on alumina or by recrystallization. compounds measured were adequately purified and Less detailed characterization was used to es- correctly identified, the following observations are tablish the presence of 1 7-hydroxyprogesterone, offered as evidence. The conjugated 17-ketoster- which has been found repeatedly in blood from oid from adrenal vein blood that was found to ac- normal adrenals (6), and of testosterone, which company a tracer of androstenolone sulfate through is a normal constituent of the peripheral plasma two solvent partitions and on fractionation in two of either sex (22). In the case of 17-hydroxy- chromatographic systems gave an infrared spec- progesterone, the material that had followed the trum which agreed closely with that of a reference tracer through two chromatographic separations specimen of androstenolone sulfate (Figure 1). was found to give the same spectrum as the ref- The isolated compound was cleaved by a proce- erence compound in the Zimmermann reaction, dure that has been found effective only with sul- and at least fairly consistent estimates by this fates (2). The product was chromatographed method (21 ,ug) and ultraviolet absorption ( 16 and the free steroid with the same Rf as andros- MAg) were obtained. Testosterone after the same tenolone measured. In general the estimates were chromatographic purification could not be sepa- in accord with those made on the sulfates, and no rated from the tracer by recrystallization of the systematic deviations were observed. As the acetate. mean of the ratios of the two determinations was The plasma fraction that was hydrolyzed with 1.02, one may conclude that no further purification fl-glucuronidase was examined in one instance had occurred. The free androstenolone was simi- (sample A, of F.B.). No androstenolone was larly identified. Again a sample that had accom- panied the tracer in two chromatographic systems TABLE V Countercurrent distribution of gave an infrared spectrum in good agreement with androstenediol diacetate* that of a reference specimen that had been chro- matographed in the same manner. Acetylation Tube no. N2 N1 with acetic-C14 anhydride gave a product that N21N could not be separated from the tracer by crystal- cPm cpm 7 146 75 1.96 lization (Table IV). Moreover the isotope ratio 8 276 143 1.93 signified an amount of adrenal androstenolone 9 435 227 1.91 10 599 311 1.93 that agreed within 6% with that determined colori- 1 1 735 381 1.93 metrically on the unacetylated steroid (2). In 12 747 386 1.94 13 658 342 1.92 the case of free and of conjugated 17-hydroxypreg- 14 483 247 1.96 nenolone, three chromatographic systems (two 15 288 149 1.94 before, one after acetylation) were used, and 16 160 86 1.87 again no separation of steroid and tracer occurred * Sample from adrenal blood of C.B. and 2.4 ,ug of andro- during these steps or on recrystallization (Table stenediol-C'4 (15,800 cpm N2) were purified, acetylated with acetic-H3 anhydride, and recrystallized with 12 mg II). The identity of androstenediol (androst-5- carrier as described in Methods. The solvents used for rests on the behavior a the 24-fold distribution in the Craig apparatus were 380 ene-3,3,17,73-diol) of sample ml 95% ethanol, 20 ml water, and 400 ml petroleum ether. that had been purified as described in the experi- The N2/N, ratio for the acetate of the pure tracer was 2.33. ADRENAL VEIN STEROIDS 165 found, and the amounts of androstenedione (1.1 though there may be no need, therefore, to ques- ug per 100 ml of plasma) and testosterone (0.2) tion our sampling method on the basis of our were too small to warrant any conclusions, par- findings, it would seem advisable not to empha- ticularly in view of the imperfect separations of size absolute amounts but to compare the ratios free and conjugated steroids. at which the various steroids were found in our samples of adrenal venous blood. Such ratios, of Discussion course, would be expected not to be constants characteristic of an individual but to vary with blood was a Adrenal venous sampled through the extent of adrenal stimulation. This view is catheter in the hope that this method would allow supported by data obtained on F.B., which indi- examination of glandular secretion under more cate that exogenous corticotropin stimulated the conditions than would be possible by nearly basal secretion of various steroids to a different extent. at The measured rate of cannulation surgery. Our measurements show quite consistently the secretion of 17-hydroxycorticoids ranged from of androstenolone sulfate, androsteno- adrenal secretion 0.03 (F.B) to 1.2 mg per hour per lone, and androstenedione. Although we failed It is difficult to attribute this dif- (C.B.). large to find the dione twice and androstenolone once ference to variations in basal rates of dif- merely when the 17-hydroxycorticoid values were very individuals. values seem if ferent High explicable low (< 30 per 100 ml of plasma), corticotropin of suffered on the ,g the varying degree pain injec- caused the secretion of the free steroids in both tion of the dye caused the release of radiopaque these These findings agree with those of in some Very low secre- patients. corticotropin patients.'0 who isolated both free from tion rates were attributed at first to incomplete col- Short (7), compounds that were obtained at sur- lections of the adrenal secretion and very low adrenal blood samples from three women who had received cortico- concentrations to dilution of the adrenal blood. gery In of consistent it is Numerous samples were discarded on this basis tropin. view these findings difficult understand other investigators although the X-ray films indicated correct place- to why have to obtain similar results. As the ment of the catheter in the adrenal vein and only failed it seemed to show minor dilution by collateral circulation. How- methodology varied, important our results were caused artifactual ever, the dramatic effect of corticotropin on the that not by 17-hydroxycorticoid secretion of F.B. suggested formation of the free steroids during fractionation Hydrolysis of androstenolone sulfate that incomplete collection and dilution may not procedures. always be major factors in such cases and that to the free compound was considered. This pos- remote we there may be much greater variations in secretion sibility seemed very because consistently blood 11 during short intervals than are apparent from pub- failed to find androstenolone in peripheral lished ranges of the mean secretion rates of corti- although it is known to contain appreciable sol that are measured over several hours (23). amounts of the sulfate. The hydrolysis of andros- In some instances (24) the decline of peripheral tenolone sulfate was definitely eliminated as the 17-hydroxycorticoids after 8 a.m. was at a rate source of free androstenolone obtained from sub- close to that observed after the iv administration jects J.J. and F.B. because we found that the of cortisol (25). In such situations the rate of tritium that had been added as androstenolone-H3 cortisol secretion must be very small. Moreover, sulfate to their blood or plasma was removed from Lombardo and associates (6), who obtained their the fraction that subsequently yielded the free samples at surgery, observed as large a variation steroid in large amounts. Numerous conditions in cortisol concentration as we did with the con- exist that can convert steroids with the dihydroxy- centrations of the free 17-hydroxycorticoids. Al- acetone side chain to 17-keto compounds. Con- trol experiments showed, however, that only neg- 10 Data obtained on J.J. suggested that his adrenals were stimulated because they produced at least 15 mg 11 The sensitivity of the colorimetric method does not per day of C19 precursors of 11-deoxy-17-ketosteroids if exceed 1 Asg per 100 ml of plasma. The presence of such the observed rates of production were maintained. On amounts of this or other steroids, as were recently an- the day before catheterization the total urinary 17-keto- nounced by Gandy and Peterson (26), is therefore not in steroids amounted to 8 mg. conflict with our findings. 166 WIELAND, DE COURCY, LEVY, ZALA, AND HIRSCHMANN ligible amounts could have formed during our method for measuring the in vivo production of procedure. We, therefore, regard androstenolone C,9s steroids by measurements of the specific ac- and androstenedione as genuine secretion prod- tivities of selected metabolites in the urine of sub- ucts of the human adrenal. jects who had been injected with labeled speci- The differences in adrenal and peripheral an- mens of the presumed precursors. Since these drostenolone sulfate exceeded those of the free methods are intended to measure total production, compound, both when stimulated and nonstimu- their results (28) are not directly comparable to lated adrenals were examined. Our series of ob- those reported here for adrenal secretion. These servations that demonstrate the secretion of the workers deduced the following production rates sulfate by the normal adrenal is enlarged by find- for a normal woman: androstenolone sulfate, 7.7; ings of Baulieu (27), who also observed higher androstenolone, 0.7; androstenedione, 3.8; and levels of androstenolone sulfate in adrenal than testosterone, 0.7 mg per day; but more nearly in peripheral blood samples of three normal comparable rates of androstenolone and its sul- subj ects.12 fate for two normal men. Although Vande Our data do not establish that there is secretion Wiele and associates took great pains to point of androstenediol or testosterone 13 by the normal out that their calculations are based on the as- human adrenal but show that such secretion, if sumption that their model is representative of the it occurs, must be quite small compared to that physiological situation, others have taken their of the corresponding ketones. Short (7) also results as proof that the compounds for which failed to find testosterone in his samples by meth- production rates were computed must be actual ods he estimates to be capable of detecting 2 Mug secretion products (33). Although Vande Wiele per 100 ml of plasma. These results could not be and his colleagues' measurements were capable of foreseen from in vitro studies in which the ratio detecting erroneous assumptions through incon- of yields of testosterone and androstenedione from sistent results (28), they evidently have not yet progesterone was 0.5 (30). If our findings are uncovered all differences between the model and typical of the normal adrenal, its contribution to the actual situation (34, 35). It would seem that peripheral testosterone levels may be more through direct examination of the blood from endocrine peripheral metabolism of its secretion products glands is necessary to prove secretion of a com- (31) than by the secretion of testosterone itself. pound and that it, therefore, is an essential step After the completion of our studies, Burger, Kent, in the development of indirect methods for the and Kellie (32) reported differences for testoster- estimation of secretion rates. one between adrenal and peripheral plasma of hir- Our other measurements were done on com- of sute women. Other C19 steroids were not meas- pounds believed to be the adrenal precursors ured. They found rather large values in two pa- the adrenal 1 7-ketosteroids. They demonstrate tients (7.76 and 0.85 Mtg per 100 ml) but differ- for the first time the capacity of the normal hu- to those reported in Table III man adrenal to secrete 17-hydroxypregnenolone ences comparable since it in others. and a conjugate believed to be a sulfate, acetate solvolysis to 17-hy- work was in progress Vande Wiele was cleaved by ethyl While this An attempt to isolate the and co-workers (28) considerably refined their droxypregnenolone. compound directly was unsuccessful, since the 12 One of these observations was mentioned in a discus- tracer 1 7-hydroxypregnenolone-Hi3 3-sulfate was sion remark to a paper by Vande Wiele, MacDonald, Gur- extensively rearranged during chromatography in pide, and Lieberman (28), which was published after sub- The of the Dr. an alkaline system.'4 analysis pre- mission of our preliminary report (2). We thank a results. sumed sulfate in the blood of J.J. represents Baulieu for permission to cite his new free 13 The specific activity of the acetic anhydride was too minimal figure, since the tracer was the low for reliable measurements of the small differences steroid that had to be added at a stage in the between adrenal and peripheral testosterone that we ob- purification where partial decomposition of the served but would have been suitable for measuring the natural compound could have occurred. In this larger amounts of testosterone that were expected from the estimate that normal adrenals secrete about 500 ,ug 14 Calvin and Lieberman (36) recently reported the of testosterone per day (29). successful purification of this sulfate in such a system. ADRENAL VEIN STEROIDS 167 case at least, the three 17-ketosteroids appear to Biochemistry for the use of a scintillation counter during be accompanied by the corresponding 17-hydroxy- the earlier phases of this work. 20-oxopregnenes, and in this series, too, the free A5-unsaturated compound predominated over the References afl-unsaturated ketone. This predominance is in 1. Hirschmann, H., C. de Courcy, R. P. Levy, and contrast to the situation encountered with the K. L. Miller. Adrenal precursors of urinary 17- plasma derived from an adrenal tumor that had ketosteroids. J. biol. Chem. 1960, 235, PC48. 2. Wieland, R. G., R. P. Levy, D. Katz, and H. Hirsch- yielded more 17-hydroxyprogesterone than free mann. Evidence for secretion of 3,f-hydroxyan- 17-hydroxypregnenolone (9). The earlier re- drost-5-en-17-one sulfate by measurement in nor- port discusses this problem more fully and also mal human adrenal venous blood. Biochim. bio- makes reference to the isolation of 17-hydroxy- phys. Acta (Amst.) 1963, 78, 566. pregnenolone from human urine and from canine 3. Pincus, G., and E. B. Romanoff. The synthesis of corticosteroids by the human adrenal cortex in adrenal blood. The Human Adrenal Cortex. Boston, Little, Calvin and associates (36, 37) have presented Brown, 1955, p. 97. evidence for the existence of enzyme systems that 4. Bush, I. E., J. Swale, and J. Patterson. Analysis of can effect the conversions, pregnenolone (3,8- steroids in adrenal venous blood and in urine in sulfate -> a case of virilism. Biochem. J. 1956, 62, 16 P. hydroxypregn-5-en-20-one) 17-hydroxy- 5. I. and B. Mahesh. Adrenocortical -> Bush, E., V. hy- pregnenolone sulfate androstenolone sulfate. perfunction with sudden onset of hirsutism. J. Our probable identification of 17-hydroxypreg- Endocr. 1959, 18, 1. nenolone sulfate in adrenal vein blood indicates 6. Lombardo, M. E., C. McMorris, and P. B. Hudson. that the substrate of the cleavage reaction occurs The isolation of steroidal substances from human in the human adrenal. As was discussed previ- adrenal vein blood. Endocrinology 1959, 65, 426. 7. Short, R. V. The secretion of sex hormones by the ously (36) such a demonstration is needed if the adrenal gland in The Biosynthesis and Secretion reaction is to be regarded as an actual rather of Adrenocortical Steroids, F. Clark and J. K. than merely a potential pathway to androstenolone Grant, Eds. Cambridge, Cambridge University sulfate. Press, 1960, p. 59. 8. Baulieu, E.-E. Studies of conjugated 17-ketosteroids Summary in a case of adrenal tumor. J. dlin. Endocr. 1962, 22, 501. Steroids were measured in peripheral and in 9. Wieland, R. G., C. de Courcy, and H. Hirschmann. adrenal venous blood that was obtained by cathe- Detection of 3p,17-dihydroxypregn-5-en-20-one in terization. The main C10O2 steroids secreted by blood from a human adrenal tumor. Steroids the adrenal were found to be 1963, 2, 61. 3,/-hydroxyandrost- 10. Wilson, H. Chromogenic values of various keto- 5-en-17-one sulfate, 3,8-hydroxyandrost-5-en-17- steroids in a micro modification of the Zimmermann one, and androst-4-ene-3,17-dione. Androst-5- reaction: comparison with the macro procedure. ene-3,8,177/-diol was found only in small amounts Arch. Biochem. 1954, 52, 217. in adrenal blood. The secretion by the normal 11. Peterson, R. E., A. Karrer, and S. L. Guerra. adrenal of this steroid and of testosterone, which Evaluation of Silber-Porter procedure for deter- mination of plasma hydrocortisone. Analyt. was not definitely established, must have been very Chem. 1957, 29, 144. much smaller than the output of the correspond- 12. Schneider, J. J., and M. L. Lewbart. Fractionation ing 17-ketones. 3,8,17-Dihydroxypregn-5-en-20- and isolation of steroid conjugates. Recent one was identified as a secretion product of the Progr. Hormone Res. 1959, 15, 201. normal human adrenal, and evidence for the pres- 13. Gut, M., and M. Uskokovic. Synthesis of radio- active . J. org. Chem. ence of its sulfate was obtained. 1959, 24, 673. 14. Roberts, K. D., R. L. Vande Wiele, and S. Lieber- Acknowledgments man. The conversion in vivo of dehydroisoandros- terone sulfate to androsterone and etiocholanolone The authors wish to express their indebtedness to Drs. glucuronidates. J. biol. Chem. 1961, 236, 2213. Kenneth L. Miller, David Katz, and James J. Roda of 15. McKenna, J., and J. K. Norymberski. Steroid the Department of Radiology and to Dr. Charles A. sulphates. Part I. Some solvolytic reactions of Hubay of the Department of Surgery for their assistance the salts of steroid sulphates. J. chem. Soc. 1957, with adrenal catheterizations and to the Department of 3889. 168 WIELAND, DE COURCY, LEVY, ZALA, AND HIRSCHMANN

16. Burstein, S., and S. Lieberman. Hydrolysis of ke- 28. Vande Wiele, R. L., P. C. MacDonald, E. Gurpide, tosteroid hydrogen sulfates by solvolysis proce- and S. Lieberman. Studies on the secretion and dures. J. biol. Chem. 1958, 233, 331. interconversion of the androgens. Recent Progr. 17. Wendler, N. L., and R. P. Graber. Alkaline degra- Hormone Res. 1963, 19, 275. dation of the cortical side chain. Chem. and Ind. 29. Dorf man, R. I. Steroid hormone biosynthesis. 1956, 549. Proceedings of the Fif th International Congress 18. Burstein, S., and R. I. Dorfman. Determination of of Biochemistry, 1963, 7, 335. mammalian steroid sulfatase with 7a-H3-3,B-hy- 30. Ichii, S., E. Forchielli, C. E. Cassidy, C. B. Rosoff, droxyandrost-5-en-17-one sulfate. J. biol. Chem. and R. I. Dorfman. Biosynthesis of androgens by 1963, 238, 1656. homogenates of normal and abnormal human 19. Oertel, G. W. On steroid conjugates in plasma adrenal glands. Biochem. biophys. Res. Commun. VIII. Extraction of 17-ketosteroid conjugates. 1962, 9, 344. Clin. chim. Acta 1963, 8, 154. 31. Mahesh, V. B., and R. B. Greenblatt. The in, vivo 20. Oertel, G. W., and E. Kaiser. tber Steroid-con- conversion of dehydroepiandrosterone and an- jugate in Plasma IV. Chromatographie von drostenedione to testosterone in the human. Acta Steroid-conjugaten an Ionenaustauscher und An- endocr. (Kbh.) 1962, 41, 400. wendung bei Isolierung von 17-Ketosteroid-con- 32. Burger, H. G., J. R. Kent, and A. E. Kellie. De- jugaten aus peripherem Plasma. Biochem. Z. termination of testosterone in human peripheral 1962, 336, 10. and adrenal venous plasma. J. clin. Endocr. 1964, 21. Wallace, E. Z., and S. Lieberman. Biosynthesis of 24, 432. dehydroisoandrosterone sulfate by human adreno- 33. Siiteri, P. K., and P. C. MacDonald. The utilization cortical tissue. J. clin. Endocr. 1963, 23, 90. of circulating dehydroisoandrosterone sulfate for 22. Forchielli, E., G. Sorcini, M. S. Nightingale, N. estrogen synthesis during human pregnancy. Ster- Brust, R. I. Dorfman, W. H. Perloff, and G. oids 1963, 2, 713. Jacobson. Testosterone in human plasma. Analyt. 34. Korenman, S. G., and M. B. Lipsett. Physiologically Biochem. 1963, 5, 416. inactive testosterone: derivation from androstene- 23. Peterson, R. E., and J. B. Wyngaarden. The miscible dione with immediate conversion to the glucurono- pool and turnover rate of hydrocortisone in man. side. The Endocrine Society, Program of the J. clin. Invest. 1956, 35, 552. Forty-sixth Meeting, 1964, p. 67. 24. Bliss, E. L., A. A. Sandberg, D. H. Nelson, and K. 35. Baulieu, E.-E., and P. Mauvais-Jarvis. Studies on Eik-Nes. The normal levels of the 17-hydroxy- testosterone metabolism II. Metabolism of tes- corticosteroids in the peripheral blood of man. J. tosterone-4-'4C and androst-4-ene-3,17-dione- 1,2- clin. Invest. 1953, 32, 818. 3H. J. biol. Chem. 1964, 239, 1578. 25. Peterson, R. E., C. E. Pierce, J. B. Wyngaarden, 36. Calvin, H. I., and S. Lieberman. Evidence that J. J. Bunim, and B. B. Brodie. The physiological steroid sulfates serve as biosynthetic intermediates. disposition and metabolic fate of cortisone in man. II. In vitro conversion of pregnenolone-3H sul- J. clin. Invest. 1957, 36, 1301. fate-_5S to 17a-hydroxypregnenolone-3H sulf ate- 26. Gandy, H. M., and R. E. Peterson. Plasma levels 3"S. Biochemistry 1964, 3, 259. of unconjugated A4-androstene-3,17-dione, dehydro- 37. Calvin, H. I., R. L. Vande Wiele, and S. Lieberman. isoandrosterone, etiocholanolone, androsterone, and Evidence that steroid sulfates serve as biosyn- testosterone in man. The Endocrine Society, thetic intermediates: in, vivo conversion of preg- Program of the Forty-sixth Meeting, 1964, p. 54. nenolone-sulfate-S35 to dehydroisoandrosterone sul- 27. Baulieu, E.-E. Personal communication. fate-S3". Biochemistry 1963, 2, 648.