- Email: [email protected]
Separation of the major adrenal steroids by reversed-phase high-performance liquid chromatography
ANALYTICAL
BIOCHEMISTRY
Separation
147, 374-381 (1985)
of the Major Adrenal Steroids by Reversed-Phase High-Performance Liquid Chromatography’ MICHAEL
W. CAPP AND MICHAEL
H. SIMONIAN~
Department o/Physiology and Biophysics, The University oflowa, Iowa City, Iowa 52242 Received June 20. 1984 The major adrenal steroids were separated by multistep gradient elution with a reversedphase high-performance liquid chromatography system, employing water and I-propanol as solvents. With this solvent system, a wide range of 2 I 5-ene-3@-ol and 4-ene-3-one steroids can be resolved in a single chromatogram, which was not possible with previously published gradient solvent systems. In particular. intermediate steroids of the biosynthetic pathway, 17ahydroxyprogesterone, 17a-hydroxypregnenolone, and dehydroepiandrosterone, were separated with baseline or sufficient resolution to allow accurate quantitation. Using the l-propanolwater gradient, the separations of 5-ene and 4-ene steroids were compared on different octadecylsilyl packings. Optimum resolution was obtained with a fully covered, spherical particle. The I-propanol-water gradient was compared to a previously published methanolwater gradient in the analysis of steroidogenesis by adrenocortical cell cultures. HPLC analysis of the steroid production was quantitatively the same with both gradient solvent systems. However, qualitatively, the methanol-water gradient system did not resolve the abovementioned intermediate steroids. Q 1985 Academic press. IX. KEY WORDS: 5-ene steroids; 4-ene steroids; adrenal; HPLC; binary gradient.
The major adrenal steroids have been separated by gas chromatography (1) and thinlayer chromatography (2). In the last decade, high-performance liquid chromatography has eclipsed these methods with both normalphase (3,4) and reversed-phase chromatography (5) being used with isocratic elution and gradient elution, which has enabled a wider range of steroids to be separated efficiently. Schoneshoffer and co-workers (6,7) used normal-phase gradient systems to separate a range of circulating steroids in human serum. O’Hare and co-workers (5,8) published a number of binary-solvent reversed-phase gradient systems utilizing solvent selectivity to obtain separation of a complex mixture of adrenal and testicular steroids from cell cultures and intact tissue. Huang et al. (9) ’ This work was supported by NIH research Grant HD15882 from the Institute of Child Health and Human Development. 2 To whom correspondence should be addressed. 0003-2697185 $3.00 Copyright 0 1985 by Academic Press. Inc. All rights of reproductmn in any form reserved.
374
used a multiple-solvent reversed-phase gradient system to separate 15 assorted steroids found in carp. None of these systems has allowed a complete separation of all of the major human adrenal steroids, including 5ene and 4-ene steroids, without using a second system with different selectivity to resolve pairs of steroids which coelute on the first system. This study demonstrates a binarysolvent gradient system, using I-propanol in water, which separates the major adrenal steroids in a single HPLC analysis. MATERIALS
AND
METHODS
HPLC methods. HPLC was performed on a Beckman 324 chromatograph (Beckman Instruments, Berkeley, Calif) equipped with an LDC Spectromonitor III variable-wavelength uv detector (Laboratory Data Control, Riviera Beach, Fla.). All chromatography was performed at 45°C and at flow rates of 1 ml/ min. Columns (150 X 4.6 mm) were labo-
CHROMATOGRAPHIC
SEPARATION
OF ADRENAL
STEROIDS
375
ratory packed with 5 pm Apex ODS (Jones 6 X lo4 cells/well) that were coated with 2 Chromatography), LiChrosorb RP 18 (E. pg/cm2 of human fibronectin. For experiMerck), or Spherisorb SS-ODS (Phase Sep) ments with unlabeled pregnenolone, cells were allowed 24 h for attachment and cultures using a Shandon slurry-packing instrument were changed to a serum-free, defined meall from Jones Chromatography (Columbus, dium described previously (10). After 24 h Ohio). Solvents were reagent-grade methylene chloride and HPLC-grade methanol, aceto- the medium was replenished, now containing 5j~M pregnenolone, and the cultures were nitrile (Fisher Scientific), and 1-propanol incubated for 24 h. At the end of the incu(Burdick and Jackson, Muskegon, Mich.). bation with exogenous steroid, the medium Steroid standards were from Sigma Chemical Company (St. Louis, MO.) with the excep- was collected and stored at -20°C and the cell number was quantitated with a Coulter tion of I7a-hydroxy-20a-dihydroprogesterone counter as published (12). and 17a-hydroxy-20a-dihydropregnenolone which were from Steraloids, inc. (Wilton, N. For experiments with radiolabeled pregH.). All steroid standards were dissolved in nenolone, cells were allowed to attach for absolute ethanol and 10 ~1, containing 100 48 h and cultures were changed to Ham’s Fng of each 4-ene-3-one steroid and 10 pg of 12 medium supplemented with 10% horse each 5-ene-3/3-ol steroid, was injected into serum and 50 rig/ml fibroblast growth factor, the chromatograph. A 50-min exponential which was replenished every 2 days. After 7 concave gradient (y = x2) of 40-100% methdays in culture, the cell cultures were incuanol in water or 20-100% acetonitrile in bated in the serum-free medium without and water was used as described by O’Hare and with 10 nM adrenocorticotropin (ACTH lNice (8). The multistep propanol-water gra- 24, synthetic) for 48 h with a complete dient consisted of 13, 15, 2 1, 28, 39, 39, and medium change at 24 h. After the 48 h 100% 1-propanol at time 0, 6, 22, 30, 38, treatment, cell cultures were incubated for 8 39, and 44 min, respectively, of the chro- h in the serum-free medium without ACTH3 matogram. The viscosity of propanol neces- and with 5j~M [3H]pregnenolone (sp act 1 sitated a dynamic mixer (Beckman InstruCi/mmol; Amersham, Arlington Heights, Ill.). ments) with a dead volume of 4-5 ml for At the end of the incubation, the medium generating the gradient with water. When was collected and stored and the cell number mixers with a lower dead volume were used, quantitated as described above. the mixing of propanol with water was inFor HPLC and analysis, each medium adequate and interfered with uv-absorbance sample (0.5 ml) was diluted with 1 ml water, detection, especially at high sensitivity. Steroid extracted with 3 X 6 ml methylene chloride, standards were detected by absorbance at and the combined extracts for each sample 280 nm with a sensitivity of 8-12 rig/peak were washed with 0.1 N NaOH. Equal alifor 4-ene-3-one steroids and 400-800 ng/ quots (9 ml) of each extract were evaporated peak for 5-ene-3P-ol steroids. With absor- and redissolved in 20 ~1 60% water and 40% bance at 246 nm, the 4-ene-3-one steroids of either propanol or methanol, depending are detected with an increased sensitivity of on the gradient system to be used for the 1-2 rig/peak. HPLC analysis as described above. Detection Cell culture methods. The dissection, prep- was by absorbance at 246 nm for culture aration, and treatment of definitive zone cell samples incubated with unlabeled pregnenocultures from human fetal adrenal glands lone or by liquid scintillation counting of 0.2 (13-l 8 weeks gestation) has been described (10,ll). Cell suspensions in Ham’s F- 12 me3 Abbreviations used: ACTH. corticotropin (adrenodium supplemented with 10% newborn bo- corticotropin, adrenocorticotropic hormone); RP, reversed vine serum were plated into 2-cm2 wells (3- phase.
376
CAPP AND SIMONIAN
or 1.0 ml fractions from chromatograms for [3H]pregnenolone-incubated samples (10). The 0.2-ml fractions were collected at the retention times corresponding to 17cY-hydroxyprogesterone, 17a-hydroxypregnenolone, and dehydroepiandrosterone. The identities of the radioactive peaks were determined by comparison to the retention times of authentic unlabeled 5-ene and 4-ene standards, which were detected by absorbance at 280 nm. The recovery of [3H]pregnenolone added to samples incubated with unlabeled pregnenolone was >98%. RESULTS
A gradient system of propanol in water was developed for separation of both 5-ene3&ol and 4-ene-3-one adrenal steroids. Using a fully covered, spherical packing, Apex ODS, this solvent system separated 17a-hydroxyprogesterone, 17+hydroxypregnenolone, and dehydroepiandrosterone sufficiently for quantitation while maintaining the complete separation of the major adrenal 4-ene-3-one steroids (Fig. 1). The dropping baseline with increasing propanol concentration shown in this figure was also observed in blank chromatograms without steroid standards. The shape of the earlier part of the gradient (O22 min) was critical to the separation of some of the later-eluting steroids, particularly the separation of 17a-hydroxypregnenolone and dehydroepiandrosterone. Either a shallower or a steeper early gradient reduced the resolution of these two steroids. In addition, this solvent system resolved 17cY-hydroxy20adihydroprogesterone and 20adihydroprogesterone from other steroids, which are produced by bovine adrenals (12,13). The retention times of 25 steroid standards on this solvent system are compared to those of the methanol-water system using an Apex ODS column (Table 1). A separation of 20.5 min between steroids was required for accurate quantitation of each steroid. Unlike the propanol-water system, the methanol-water system demonstrated poor resolution of de-
w
Retention
time
(mid
FIG. I. HPLC of S-ene and 4-ene steroids by a lpropanol-water gradient. A mixture of 17 steroid standards was separated with a gradient of I-propanol in water (dashed line) on a column packed with Apex ODS. Chromatography was performed at 1 ml/min and 45% The steroids injected were (1) aldosterone, (2) cortisone, (3) cortisol, (4) 1l&hydroxyandrostenedione, (5) corticosterone, (6) 1 Ideoxycortisol, (7) 16+hydroxypregnenalone, (8) androstendione, (9) 1 Ideoxycorticosterone, (10) 17a-hydroxy-20adihydroprogesterone, (11) testosterone, ( 12) 17a-hydroxyprogesterone, ( 13) 17a-hydroxypregnenolone, ( 14) dehydroepiandrosterone, ( 15) 20adihydroprogesterone, ( 16) progesterone, and (17) pregnenolone. Detection is by uv absorbance at 280 nm and attenuation is 0.02 absorbance full scale (solid line).
hydroepiandrosterone from 17a-hydroxyprogesterone and 17a-hydroxypregnenolone, which coeluted, and did not resolve 17~ hydroxy-20adihydroprogesterone from 17~ hydroxyprogesterone or 20a-dihydroprogesterone from progesterone. There are four pairs of steroids included in this list that coelute or only partially separate with the propanol-water system; however, all of these pairs separated with the methanol-water gradient.
CHROMATOGRAPHIC
SEPARATION TABLE
OF ADRENAL
377
STEROIDS
1
COMPARISONOFSTEROIDRETENTIONTIMESWITH I-PROPANOL-WATER OR METHANOL-WATER SOLVENT SYSTEMS Retention time (min) Trivial name
Steroid 11@,21-Dihydroxy- 18-al4pregnene3,20-dione 1l&l 8,2 1-Trihydroxy4pregnene3,20dione 17cy,2I-Dihydroxy-4-pregnene 3,11,20 trione 4-Androstene-3,ll ,I7-trione I l&17&,2 l-Trihydroxy-4-pregnene3,2Odione 2 I-Hydroxy4pregnene-3,11,20trione 1 la-Hydroxy-4-androstene-3,17dione 18,2 I-Dihydroxy4pregnene-3,20dione 3fl,l6a-Dihydroxyandrost-5-en- 17one 1l&21-Dihydroxy-4-pregnene-3,20dione 17n,2 1-Dihydroxy&pregnene-3,20dione 3&16n-Dihydroxy-5-pregnene-3-one 4-Androstene-3,17dione 21-Hydroxy-4-pregnene-3,2Odione 17a,20a-Dihydroxy4pregnene-3-one 17a-Hydroxy4androstene-3-one 31%17a,20a-Trihydroxy-S-pregnen20-one 17a-Hydrnxy4pregnene-3,2Odione 17a-2Ofi-Dihydroxy-4-pregnene-3-one 30-17~Dihydroxy-5-pregnen-20-one 3&Hydroxyandrost-S-en-17-one 20a-Hydroxy-4-pregnene-3-one 4-Pregnene-3,20dione 20&Hydroxy4pregnene-3-one 3&Hydroxy-5-pregnene-3-one
1Propanolwater
Methanolwater 11.8
Aldosterone
9.8
18-Hydroxycorticosterone
11.7
Cortisone
12.3
15.2
Adrenosterone Cortisol
14.0 14.1
16.3 17.5
1 l-Dehydrocorticosterone
15.4
1 l&Hydroxyandrostenedione
17.9
18-Hydroxy- 11deoxycorticosterone
19.9
16cY-Hydroxydehydroepiandrosterone
20.3
21.0
Corticosterone
20.4
25.5
1 I -Deoxycortisol
23.8
28.1
t6a-Hydroxypregnenolone Androstenedione 1I-Deoxycorticosterone 17a-Hydroxy-20&dihydroprogesterone Testosterone 17a-Hydroxy-20~ dihydropregnenolone 17~~Hydroxyprogesterone 17a-Hydroxy-20& dihydroprogesterone 17a-Hydroxypregnenolone Dehydrcepiandrosterone 20a-Dihydroprogesterone Progesterone 20&Dihydroprogesterone Pregnenolone
26.9 29.4 30.2 31.1
31.3 33.4 36.0
31.9 32.1
33.8 39.1
32.6 33.0
36.1
34.2 34.7 37.5 38.5 40.2 40.3
36.1 35.7 41.2 41.4 44.5 43.5
21.5
Note. IUPAC systemic and trivial names of steroid standards that were chromatographed are shown with the retention times on gradient systems of I-propanol-water or methanol-water. Both solvent systems used an Apex ODS column with other conditions as under Materials and Methods.
The propanol-water gradient was compared to the methanol-water gradient on an Apex ODS column for HPLC analysis of
steroid production by human adrenocortical cell cultures. Because of the lower sensitivity for uv-absorbance detection of the 5-ene-3&
378
CAPP AND SIMONIAN
01 steroids relative to the 4-ene-3-one steroids, the production of both types of steroids from cell cultures was determined by use of a radiolabeled 5-ene-3/3-ol precursor. Sevenday-old cell cultures were treated for 48 h without or with 10 nM ACTH and subsequently incubated for 8 h with 5 PM [3H]pregnenolone in the absence of ACTH. Incubation with exogenous [3H]pregnenolone for this length of time results in substantial conversion to both unconjugated steroid products (cortisol, 1 lp-hydroxyandrostenedione, and corticosterone) and/or steroid intermediates (e.g. 17a-hydroxypregnenolone, dehydroepiandrosterone, and 17a-hydroxyprogesterone), whereas longer incubation times (> 12 h) resulted in nearly complete conversion to the unconjugated 4-ene-3-one steroid products and sulfated 5-ene-3P-ol steroid products (10,14). As shown in Fig. 2, the steroid production pattern for ACTHtreated cultures was the same with both
solvent systems except that the propanolwater system resolved 17a-hydroxyprogesterone, dehydroepiandrosterone, and 17a-hydroxypregnenolone. These three steroids eluted as a single peak with the methanolwater gradient system. The total steroid production for duplicate ACTH-treated cultures was 147 and 169 pmol/5 X lo4 cells as determined by the propanol-water system and 142 and 162 pmol/5 X lo4 cells by the methanol-water gradient system, respectively. Without ACTH treatment, duplicate control cultures produced qualitatively fewer and quantitatively less steroids from [3H]pregnenolone as determined by HPLC analysis. With the propanol-water gradient, 79 and 84 pmol/ 5 X lo4 cells total for three steroids, 17ahydroxyprogesterone ( 17.4%), 17Lu-hydroxypregnenolone (44.7%) and dehydroepiandrosterone (37.9%) was detected in duplicate control cultures, whereas analysis of the same cultures with the methanol-water gradient detected only a single peak of 67 and 80 pmol/5 X lo4 cells, respectively, with the retention time common to these three steroids. This qualitative and quantitative stimulation of the steroidogenic capacity by ACTH agrees with published reports for these cell cultures ( 10,14). As measured by uv absorbance, the 4-ene3-one steroids formed from unlabeled exogenous pregnenolone by cell cultures also has been determined by HPLC separation with the propanol-water and methanol-water graFIG. 2. HPLC of radiolabeled steroids metabolized dients. In the absence of ACTH, 2-day-old from [‘H]pregnenolone by lo-day-old cultures of ACTHcultures of human adrenocortical cells were treated human fetal adrenal cells. Cell cultures maintained incubated subsequently for 24 h with 5 PM in serum-supplemented medium without ACTH for 7 days were then treated with 10 nM ACTH for 2 days in pregnenolone, and the 4-ene-3-one steroid serum-free medium. Cultures were subsequently incu- content of the medium was analyzed. As bated for 8 h in this medium with 5pM [3H]pregnenolone determined with both gradient solvent sys(1 .O Ci/mmol) in the absence of ACTH. Equal ahquots of a methylene chloride extract of the medium were tems, early cultures from two different adrenal analyzed by the I-propanol-water gradient (A) and by preparations produced primarily steroid end the methanol-water gradient (B). The retention times of products (Table 2) which indicates high basal steroid standards are indicated above the peaks of radio- activity of the steroidogenic enzymes for activity: (I) cortisol, (2) I I@-hydroxyandrostenedione, pregnenolone metabolism as published for (3) corticosterone, (4) deoxycortisol, (5) androstenedione, similar aged cultures (IO). For each cell cul(6) 17~hydroxyprogesterone, (7) 17n-hydroxypregnenture preparation, the total and major steroid alone, (8) dehydroepiandrosterone, and (9) pregnenolone.
CHROMATOGRAPHIC
SEPARATION
OF ADRENAL
379
STEROIDS
TABLE 2 4-ENE-~-ONE STEROID PRODUCTION FROM PREGNENOLONEBY ADRENOCORTICAL CELL CULTURES Production (pmol/5 X 1O4 cells) Preparation 1 Propanolwater
Steroid Cortisol I 1fi-Hydroxyandrostenedione Corticosterone Deoxycortisol Androstenedione Total
Preparation 2
Methanolwater
Propanolwater
Methanolwater
193 51 26 5.6 9.5
198 58 32 4.0 9.4
143 53 40 5.0 11.2
148 56 41 5.4 15.4
285.1
301.4
252.2
265.8
Note. Two-day-old cultures of human fetal adrenal cells were incubated in serum-free medium with 5 FM pregnenolone for 24 h. Equal aliquots of a methylene chloride extract of the medium were analyzed by the Ipropanol-water and methanol-water gradients. The 4-ene-3-one steroids were detected and quantitated by absorbance at 246 nm. The production results are for two different cell culture preparations (Preparations I and 2).
productions measured with each solvent gradient system agreed within 12%. The use of the propanol-water gradient with other octadecylsilyl packings did not achieve the separation of both 5-ene-3&ol and 4-ene-3-one adrenal steroids. With Spherisorb SS-ODS packing, which is a lowcoverage, spherical packing (8), the steroids chromatographed into three groups: low, medium, and high polarity. There was particularly poor resolution of the polar steroids, 17cr-hydroxypregnenolone and dehydroepiandrosterone were poorly separated, and deoxycorticosterone and 17cr-hydroxyprogesterone coeluted. With LiChrosorb RP 18, which is a high-coverage, irregular packing (8), the overall elution pattern was similar to that with Apex ODS; however, androstenedione and deoxycorticosterone separated poorly and 17a-hydroxyprogesterone and 17a-hydroxypregnenolone coeluted. With neither the Spherisorb SS-ODS nor the LiChrosorb RP 18 were 17a-hydroxy-20a-dihydroprogesterone and 20a-dihydroprogesterone separated from the other steroids by the propanol-water system. When these two
packings were used with the methanol-water gradient, the respective steroid elution patterns were similar to that with the propanolwater gradient except that the peaks were generally narrower and more symmetrical with propanol as the secondary solvent. As a result, the methanol-water gradient demonstrated poor resolution of dehydroepiandrosterone, 17a-hydroxypregnenolone, and 17ahydroxyprogesterone as compared to the propanol-water gradient with a given packing. With all three packings, the acetonitrilewater system separated cortisone from cortisol poorly and did not resolve the 5-ene-3@-ol and 4-ene-3-one adrenal steroids. However, with Apex ODS packing, this solvent system resolved 17a-hydroxy-20a-dihydroprogesterone and 20a-dihydroprogesterone from the other steroids, a useful feature achieved also by propanol-water. DISCUSSION
While propanol has been used in RPHPLC of proteins because of its high solvent strength and relatively good solubility for
380
CAPP AND SIMONIAN
proteins (15), it has not been much used in reversed-phase partition chromatography, where it exhibits a unique selectivity for steroids, differing even from other primary alcohols. The similarities in elution patterns of the steroids with propanol and methanol on Apex ODS, LiChrosorb RPl8, and Spherisorb SS-ODS packings suggest that within a class of solvent (e.g., primary alcohols) a given packing will maintain its general retention characteristics but that by altering the solvent within that class fine tuning of separations may be achieved. Steroid selectivity differences among different reversed-phase packings has not been attributed to particle shape but instead to differences in accessible silanol groups on the packing (8). Among the three packings used in this study, Apex ODS has the least and Spherisorb SS-ODS the greatest number of free silanols (8,16). Thus the simultaneous separation of all the major unconjugated adrenal steroids by the propanol-water gradient with the Apex ODS packing is presumably the result of the appropriate combination of solvent and packing selectivities. The propanol-water and methanol-water gradient systems were compared and the quantitative results agreed for the HPLC analysis of steroidogenesis by human adrenocortical cell cultures. The advantage of the propanol-water gradient was demonstrated by the simultaneous separation of 5-ene-3/301 steroid intermediates and 4-ene-3-one steroid intermediates and end products metabolized from radiolabeled pregnenolone. For cell culture samples of ~5 X lo6 cells, detection of 5-ene-3/3-ol steroid production requires a radiolabeled precursor as used in this study or a specific radioimmunoassay of eluant fractions (8). However, detection of l-2 ng/ peak of 4-ene-3-one steroids by absorbance at 246 nm after HPLC separation allowed measurement of the production of these steroids from an unlabeled exogenous precursor in cultures of >30,000 cells. The use of the
propanol-water gradient for HPLC analysis would not be limited to steroid production studies with in vitro culture systems. Just as the methanol-water gradient system has been used to analyze steroids in solid adrenal tissue after appropriate preparation and extraction (8), propanol-water also could he used for analysis of these samples. As with any chromatographic system used to analyze biological systems, the ability to extract the solute of interest is the primary limitation for the analysis. The adrenal steroids normally produced by most animal species and humans can be separated in a single chromatographic analysis using the propanol-water gradient and a fully covered octadecylsilyl packing. However, given the uncertainty of any biological sample contents, confirmation of steroid identity by analysis with two solvent systems with different selectivities is routinely performed in this laboratory ( 10,12,13). An increasing number of multisolvent optimization programs are being developed for both isocratic and gradient elutions, enabling the chromatographer to design chromatographic systems giving maximal resolution of a complex mixture of compounds ( 17). However, these solvent systems require chromatographic systems with ternary or quaternary solvent capability and considerable amount of computer time. In laboratories with small sample throughput, an optimization can take longer to run than the samples for a complete project. REFERENCES 1. Vinson, G. P., and Whitehouse, J. B. (1969) Acta Endocrinol. 61, 695-708. 2. O’Hare, M. J., and Neville, A. M. (1973) J. Endocrinol.
58, 447-462.
3. Schoneshofer, M., Skobolo, R., and Dulce, H. J. (1980) J. Chromatogr. 222, 478-48 1. 4. Cavina, G., Moretti, G., Alimenti, R., and Gallinella, B. (1979) J. Chromatogr. 175, 125-140. 5. O’Hare, M. J., Nice, E. C., Magee-Brown, R., and Bullman, H. (1976) J. Chromatogr. 125, 357367.
CHROMATOGRAPHIC
SEPARATION
6. Schoneshofer, M., Fenner, A., and Duke, H. J. (1980) J. Steroid Biochem. 14, 377-386. 7. Schoneshofer, M., Maxeiner, J., and Fenner, A. (1981) J. Chromatogr. 224, 229-237. 8. O’Hare, M. J., and Nice, E. C. (1981) in Steroid Analysis by HPLC Recent Applications (Kautsky, M. P., ed.), pp. 217-322, Dekker, New York/ BaseI. 9. Huang, F., Ke, F., Hwang, J., and Lo, T. (1983) Arch. Biochem. Biophys. 225, 5 12-5 17. 10. Simonian, M. H., and Capp, M. W. (1984) J. Clin. Endocrinol. Metab. 59, 643-65 1. 11. Simonian, M. H., and Gill, G. N. (198 1) Endocrinology 108, 1769-1779.
OF ADRENAL
STEROIDS
381
12. Simonian, M. H., White, M. L., and Gill, G. N. (1982) Endocrinology 111, 919-927. 13. Simonian, M. H., Homsby, P. J., Ill, C. R., O’Hare, M. J., and Gill, G. N. (1979) Endocrinology 105, 99-108. 14. Homsby, P. J., and Aldem, K. A. (1984) J. Clin. Endocrinol. Metab. 58, 12l-l 27. 15. O’Hare, M. J., Capp, M. W., Nice, E. C., Cooke, N. H. C., and Archer, B. G. (1982) Anal. Biochem. 126, 17-28. 16. Simonian, M. H., and Capp, M. W. (1984) J. Chromatogr. 287, 97-104. 17. D’Agostino, G., Mitchell, F., Castagnetta, L., and O’Hare, M. J. (1984) J. Chromatogr. 305, 13-26.
BIOCHEMISTRY
Separation
147, 374-381 (1985)
of the Major Adrenal Steroids by Reversed-Phase High-Performance Liquid Chromatography’ MICHAEL
W. CAPP AND MICHAEL
H. SIMONIAN~
Department o/Physiology and Biophysics, The University oflowa, Iowa City, Iowa 52242 Received June 20. 1984 The major adrenal steroids were separated by multistep gradient elution with a reversedphase high-performance liquid chromatography system, employing water and I-propanol as solvents. With this solvent system, a wide range of 2 I 5-ene-3@-ol and 4-ene-3-one steroids can be resolved in a single chromatogram, which was not possible with previously published gradient solvent systems. In particular. intermediate steroids of the biosynthetic pathway, 17ahydroxyprogesterone, 17a-hydroxypregnenolone, and dehydroepiandrosterone, were separated with baseline or sufficient resolution to allow accurate quantitation. Using the l-propanolwater gradient, the separations of 5-ene and 4-ene steroids were compared on different octadecylsilyl packings. Optimum resolution was obtained with a fully covered, spherical particle. The I-propanol-water gradient was compared to a previously published methanolwater gradient in the analysis of steroidogenesis by adrenocortical cell cultures. HPLC analysis of the steroid production was quantitatively the same with both gradient solvent systems. However, qualitatively, the methanol-water gradient system did not resolve the abovementioned intermediate steroids. Q 1985 Academic press. IX. KEY WORDS: 5-ene steroids; 4-ene steroids; adrenal; HPLC; binary gradient.
The major adrenal steroids have been separated by gas chromatography (1) and thinlayer chromatography (2). In the last decade, high-performance liquid chromatography has eclipsed these methods with both normalphase (3,4) and reversed-phase chromatography (5) being used with isocratic elution and gradient elution, which has enabled a wider range of steroids to be separated efficiently. Schoneshoffer and co-workers (6,7) used normal-phase gradient systems to separate a range of circulating steroids in human serum. O’Hare and co-workers (5,8) published a number of binary-solvent reversed-phase gradient systems utilizing solvent selectivity to obtain separation of a complex mixture of adrenal and testicular steroids from cell cultures and intact tissue. Huang et al. (9) ’ This work was supported by NIH research Grant HD15882 from the Institute of Child Health and Human Development. 2 To whom correspondence should be addressed. 0003-2697185 $3.00 Copyright 0 1985 by Academic Press. Inc. All rights of reproductmn in any form reserved.
374
used a multiple-solvent reversed-phase gradient system to separate 15 assorted steroids found in carp. None of these systems has allowed a complete separation of all of the major human adrenal steroids, including 5ene and 4-ene steroids, without using a second system with different selectivity to resolve pairs of steroids which coelute on the first system. This study demonstrates a binarysolvent gradient system, using I-propanol in water, which separates the major adrenal steroids in a single HPLC analysis. MATERIALS
AND
METHODS
HPLC methods. HPLC was performed on a Beckman 324 chromatograph (Beckman Instruments, Berkeley, Calif) equipped with an LDC Spectromonitor III variable-wavelength uv detector (Laboratory Data Control, Riviera Beach, Fla.). All chromatography was performed at 45°C and at flow rates of 1 ml/ min. Columns (150 X 4.6 mm) were labo-
CHROMATOGRAPHIC
SEPARATION
OF ADRENAL
STEROIDS
375
ratory packed with 5 pm Apex ODS (Jones 6 X lo4 cells/well) that were coated with 2 Chromatography), LiChrosorb RP 18 (E. pg/cm2 of human fibronectin. For experiMerck), or Spherisorb SS-ODS (Phase Sep) ments with unlabeled pregnenolone, cells were allowed 24 h for attachment and cultures using a Shandon slurry-packing instrument were changed to a serum-free, defined meall from Jones Chromatography (Columbus, dium described previously (10). After 24 h Ohio). Solvents were reagent-grade methylene chloride and HPLC-grade methanol, aceto- the medium was replenished, now containing 5j~M pregnenolone, and the cultures were nitrile (Fisher Scientific), and 1-propanol incubated for 24 h. At the end of the incu(Burdick and Jackson, Muskegon, Mich.). bation with exogenous steroid, the medium Steroid standards were from Sigma Chemical Company (St. Louis, MO.) with the excep- was collected and stored at -20°C and the cell number was quantitated with a Coulter tion of I7a-hydroxy-20a-dihydroprogesterone counter as published (12). and 17a-hydroxy-20a-dihydropregnenolone which were from Steraloids, inc. (Wilton, N. For experiments with radiolabeled pregH.). All steroid standards were dissolved in nenolone, cells were allowed to attach for absolute ethanol and 10 ~1, containing 100 48 h and cultures were changed to Ham’s Fng of each 4-ene-3-one steroid and 10 pg of 12 medium supplemented with 10% horse each 5-ene-3/3-ol steroid, was injected into serum and 50 rig/ml fibroblast growth factor, the chromatograph. A 50-min exponential which was replenished every 2 days. After 7 concave gradient (y = x2) of 40-100% methdays in culture, the cell cultures were incuanol in water or 20-100% acetonitrile in bated in the serum-free medium without and water was used as described by O’Hare and with 10 nM adrenocorticotropin (ACTH lNice (8). The multistep propanol-water gra- 24, synthetic) for 48 h with a complete dient consisted of 13, 15, 2 1, 28, 39, 39, and medium change at 24 h. After the 48 h 100% 1-propanol at time 0, 6, 22, 30, 38, treatment, cell cultures were incubated for 8 39, and 44 min, respectively, of the chro- h in the serum-free medium without ACTH3 matogram. The viscosity of propanol neces- and with 5j~M [3H]pregnenolone (sp act 1 sitated a dynamic mixer (Beckman InstruCi/mmol; Amersham, Arlington Heights, Ill.). ments) with a dead volume of 4-5 ml for At the end of the incubation, the medium generating the gradient with water. When was collected and stored and the cell number mixers with a lower dead volume were used, quantitated as described above. the mixing of propanol with water was inFor HPLC and analysis, each medium adequate and interfered with uv-absorbance sample (0.5 ml) was diluted with 1 ml water, detection, especially at high sensitivity. Steroid extracted with 3 X 6 ml methylene chloride, standards were detected by absorbance at and the combined extracts for each sample 280 nm with a sensitivity of 8-12 rig/peak were washed with 0.1 N NaOH. Equal alifor 4-ene-3-one steroids and 400-800 ng/ quots (9 ml) of each extract were evaporated peak for 5-ene-3P-ol steroids. With absor- and redissolved in 20 ~1 60% water and 40% bance at 246 nm, the 4-ene-3-one steroids of either propanol or methanol, depending are detected with an increased sensitivity of on the gradient system to be used for the 1-2 rig/peak. HPLC analysis as described above. Detection Cell culture methods. The dissection, prep- was by absorbance at 246 nm for culture aration, and treatment of definitive zone cell samples incubated with unlabeled pregnenocultures from human fetal adrenal glands lone or by liquid scintillation counting of 0.2 (13-l 8 weeks gestation) has been described (10,ll). Cell suspensions in Ham’s F- 12 me3 Abbreviations used: ACTH. corticotropin (adrenodium supplemented with 10% newborn bo- corticotropin, adrenocorticotropic hormone); RP, reversed vine serum were plated into 2-cm2 wells (3- phase.
376
CAPP AND SIMONIAN
or 1.0 ml fractions from chromatograms for [3H]pregnenolone-incubated samples (10). The 0.2-ml fractions were collected at the retention times corresponding to 17cY-hydroxyprogesterone, 17a-hydroxypregnenolone, and dehydroepiandrosterone. The identities of the radioactive peaks were determined by comparison to the retention times of authentic unlabeled 5-ene and 4-ene standards, which were detected by absorbance at 280 nm. The recovery of [3H]pregnenolone added to samples incubated with unlabeled pregnenolone was >98%. RESULTS
A gradient system of propanol in water was developed for separation of both 5-ene3&ol and 4-ene-3-one adrenal steroids. Using a fully covered, spherical packing, Apex ODS, this solvent system separated 17a-hydroxyprogesterone, 17+hydroxypregnenolone, and dehydroepiandrosterone sufficiently for quantitation while maintaining the complete separation of the major adrenal 4-ene-3-one steroids (Fig. 1). The dropping baseline with increasing propanol concentration shown in this figure was also observed in blank chromatograms without steroid standards. The shape of the earlier part of the gradient (O22 min) was critical to the separation of some of the later-eluting steroids, particularly the separation of 17a-hydroxypregnenolone and dehydroepiandrosterone. Either a shallower or a steeper early gradient reduced the resolution of these two steroids. In addition, this solvent system resolved 17cY-hydroxy20adihydroprogesterone and 20adihydroprogesterone from other steroids, which are produced by bovine adrenals (12,13). The retention times of 25 steroid standards on this solvent system are compared to those of the methanol-water system using an Apex ODS column (Table 1). A separation of 20.5 min between steroids was required for accurate quantitation of each steroid. Unlike the propanol-water system, the methanol-water system demonstrated poor resolution of de-
w
Retention
time
(mid
FIG. I. HPLC of S-ene and 4-ene steroids by a lpropanol-water gradient. A mixture of 17 steroid standards was separated with a gradient of I-propanol in water (dashed line) on a column packed with Apex ODS. Chromatography was performed at 1 ml/min and 45% The steroids injected were (1) aldosterone, (2) cortisone, (3) cortisol, (4) 1l&hydroxyandrostenedione, (5) corticosterone, (6) 1 Ideoxycortisol, (7) 16+hydroxypregnenalone, (8) androstendione, (9) 1 Ideoxycorticosterone, (10) 17a-hydroxy-20adihydroprogesterone, (11) testosterone, ( 12) 17a-hydroxyprogesterone, ( 13) 17a-hydroxypregnenolone, ( 14) dehydroepiandrosterone, ( 15) 20adihydroprogesterone, ( 16) progesterone, and (17) pregnenolone. Detection is by uv absorbance at 280 nm and attenuation is 0.02 absorbance full scale (solid line).
hydroepiandrosterone from 17a-hydroxyprogesterone and 17a-hydroxypregnenolone, which coeluted, and did not resolve 17~ hydroxy-20adihydroprogesterone from 17~ hydroxyprogesterone or 20a-dihydroprogesterone from progesterone. There are four pairs of steroids included in this list that coelute or only partially separate with the propanol-water system; however, all of these pairs separated with the methanol-water gradient.
CHROMATOGRAPHIC
SEPARATION TABLE
OF ADRENAL
377
STEROIDS
1
COMPARISONOFSTEROIDRETENTIONTIMESWITH I-PROPANOL-WATER OR METHANOL-WATER SOLVENT SYSTEMS Retention time (min) Trivial name
Steroid 11@,21-Dihydroxy- 18-al4pregnene3,20-dione 1l&l 8,2 1-Trihydroxy4pregnene3,20dione 17cy,2I-Dihydroxy-4-pregnene 3,11,20 trione 4-Androstene-3,ll ,I7-trione I l&17&,2 l-Trihydroxy-4-pregnene3,2Odione 2 I-Hydroxy4pregnene-3,11,20trione 1 la-Hydroxy-4-androstene-3,17dione 18,2 I-Dihydroxy4pregnene-3,20dione 3fl,l6a-Dihydroxyandrost-5-en- 17one 1l&21-Dihydroxy-4-pregnene-3,20dione 17n,2 1-Dihydroxy&pregnene-3,20dione 3&16n-Dihydroxy-5-pregnene-3-one 4-Androstene-3,17dione 21-Hydroxy-4-pregnene-3,2Odione 17a,20a-Dihydroxy4pregnene-3-one 17a-Hydroxy4androstene-3-one 31%17a,20a-Trihydroxy-S-pregnen20-one 17a-Hydrnxy4pregnene-3,2Odione 17a-2Ofi-Dihydroxy-4-pregnene-3-one 30-17~Dihydroxy-5-pregnen-20-one 3&Hydroxyandrost-S-en-17-one 20a-Hydroxy-4-pregnene-3-one 4-Pregnene-3,20dione 20&Hydroxy4pregnene-3-one 3&Hydroxy-5-pregnene-3-one
1Propanolwater
Methanolwater 11.8
Aldosterone
9.8
18-Hydroxycorticosterone
11.7
Cortisone
12.3
15.2
Adrenosterone Cortisol
14.0 14.1
16.3 17.5
1 l-Dehydrocorticosterone
15.4
1 l&Hydroxyandrostenedione
17.9
18-Hydroxy- 11deoxycorticosterone
19.9
16cY-Hydroxydehydroepiandrosterone
20.3
21.0
Corticosterone
20.4
25.5
1 I -Deoxycortisol
23.8
28.1
t6a-Hydroxypregnenolone Androstenedione 1I-Deoxycorticosterone 17a-Hydroxy-20&dihydroprogesterone Testosterone 17a-Hydroxy-20~ dihydropregnenolone 17~~Hydroxyprogesterone 17a-Hydroxy-20& dihydroprogesterone 17a-Hydroxypregnenolone Dehydrcepiandrosterone 20a-Dihydroprogesterone Progesterone 20&Dihydroprogesterone Pregnenolone
26.9 29.4 30.2 31.1
31.3 33.4 36.0
31.9 32.1
33.8 39.1
32.6 33.0
36.1
34.2 34.7 37.5 38.5 40.2 40.3
36.1 35.7 41.2 41.4 44.5 43.5
21.5
Note. IUPAC systemic and trivial names of steroid standards that were chromatographed are shown with the retention times on gradient systems of I-propanol-water or methanol-water. Both solvent systems used an Apex ODS column with other conditions as under Materials and Methods.
The propanol-water gradient was compared to the methanol-water gradient on an Apex ODS column for HPLC analysis of
steroid production by human adrenocortical cell cultures. Because of the lower sensitivity for uv-absorbance detection of the 5-ene-3&
378
CAPP AND SIMONIAN
01 steroids relative to the 4-ene-3-one steroids, the production of both types of steroids from cell cultures was determined by use of a radiolabeled 5-ene-3/3-ol precursor. Sevenday-old cell cultures were treated for 48 h without or with 10 nM ACTH and subsequently incubated for 8 h with 5 PM [3H]pregnenolone in the absence of ACTH. Incubation with exogenous [3H]pregnenolone for this length of time results in substantial conversion to both unconjugated steroid products (cortisol, 1 lp-hydroxyandrostenedione, and corticosterone) and/or steroid intermediates (e.g. 17a-hydroxypregnenolone, dehydroepiandrosterone, and 17a-hydroxyprogesterone), whereas longer incubation times (> 12 h) resulted in nearly complete conversion to the unconjugated 4-ene-3-one steroid products and sulfated 5-ene-3P-ol steroid products (10,14). As shown in Fig. 2, the steroid production pattern for ACTHtreated cultures was the same with both
solvent systems except that the propanolwater system resolved 17a-hydroxyprogesterone, dehydroepiandrosterone, and 17a-hydroxypregnenolone. These three steroids eluted as a single peak with the methanolwater gradient system. The total steroid production for duplicate ACTH-treated cultures was 147 and 169 pmol/5 X lo4 cells as determined by the propanol-water system and 142 and 162 pmol/5 X lo4 cells by the methanol-water gradient system, respectively. Without ACTH treatment, duplicate control cultures produced qualitatively fewer and quantitatively less steroids from [3H]pregnenolone as determined by HPLC analysis. With the propanol-water gradient, 79 and 84 pmol/ 5 X lo4 cells total for three steroids, 17ahydroxyprogesterone ( 17.4%), 17Lu-hydroxypregnenolone (44.7%) and dehydroepiandrosterone (37.9%) was detected in duplicate control cultures, whereas analysis of the same cultures with the methanol-water gradient detected only a single peak of 67 and 80 pmol/5 X lo4 cells, respectively, with the retention time common to these three steroids. This qualitative and quantitative stimulation of the steroidogenic capacity by ACTH agrees with published reports for these cell cultures ( 10,14). As measured by uv absorbance, the 4-ene3-one steroids formed from unlabeled exogenous pregnenolone by cell cultures also has been determined by HPLC separation with the propanol-water and methanol-water graFIG. 2. HPLC of radiolabeled steroids metabolized dients. In the absence of ACTH, 2-day-old from [‘H]pregnenolone by lo-day-old cultures of ACTHcultures of human adrenocortical cells were treated human fetal adrenal cells. Cell cultures maintained incubated subsequently for 24 h with 5 PM in serum-supplemented medium without ACTH for 7 days were then treated with 10 nM ACTH for 2 days in pregnenolone, and the 4-ene-3-one steroid serum-free medium. Cultures were subsequently incu- content of the medium was analyzed. As bated for 8 h in this medium with 5pM [3H]pregnenolone determined with both gradient solvent sys(1 .O Ci/mmol) in the absence of ACTH. Equal ahquots of a methylene chloride extract of the medium were tems, early cultures from two different adrenal analyzed by the I-propanol-water gradient (A) and by preparations produced primarily steroid end the methanol-water gradient (B). The retention times of products (Table 2) which indicates high basal steroid standards are indicated above the peaks of radio- activity of the steroidogenic enzymes for activity: (I) cortisol, (2) I I@-hydroxyandrostenedione, pregnenolone metabolism as published for (3) corticosterone, (4) deoxycortisol, (5) androstenedione, similar aged cultures (IO). For each cell cul(6) 17~hydroxyprogesterone, (7) 17n-hydroxypregnenture preparation, the total and major steroid alone, (8) dehydroepiandrosterone, and (9) pregnenolone.
CHROMATOGRAPHIC
SEPARATION
OF ADRENAL
379
STEROIDS
TABLE 2 4-ENE-~-ONE STEROID PRODUCTION FROM PREGNENOLONEBY ADRENOCORTICAL CELL CULTURES Production (pmol/5 X 1O4 cells) Preparation 1 Propanolwater
Steroid Cortisol I 1fi-Hydroxyandrostenedione Corticosterone Deoxycortisol Androstenedione Total
Preparation 2
Methanolwater
Propanolwater
Methanolwater
193 51 26 5.6 9.5
198 58 32 4.0 9.4
143 53 40 5.0 11.2
148 56 41 5.4 15.4
285.1
301.4
252.2
265.8
Note. Two-day-old cultures of human fetal adrenal cells were incubated in serum-free medium with 5 FM pregnenolone for 24 h. Equal aliquots of a methylene chloride extract of the medium were analyzed by the Ipropanol-water and methanol-water gradients. The 4-ene-3-one steroids were detected and quantitated by absorbance at 246 nm. The production results are for two different cell culture preparations (Preparations I and 2).
productions measured with each solvent gradient system agreed within 12%. The use of the propanol-water gradient with other octadecylsilyl packings did not achieve the separation of both 5-ene-3&ol and 4-ene-3-one adrenal steroids. With Spherisorb SS-ODS packing, which is a lowcoverage, spherical packing (8), the steroids chromatographed into three groups: low, medium, and high polarity. There was particularly poor resolution of the polar steroids, 17cr-hydroxypregnenolone and dehydroepiandrosterone were poorly separated, and deoxycorticosterone and 17cr-hydroxyprogesterone coeluted. With LiChrosorb RP 18, which is a high-coverage, irregular packing (8), the overall elution pattern was similar to that with Apex ODS; however, androstenedione and deoxycorticosterone separated poorly and 17a-hydroxyprogesterone and 17a-hydroxypregnenolone coeluted. With neither the Spherisorb SS-ODS nor the LiChrosorb RP 18 were 17a-hydroxy-20a-dihydroprogesterone and 20a-dihydroprogesterone separated from the other steroids by the propanol-water system. When these two
packings were used with the methanol-water gradient, the respective steroid elution patterns were similar to that with the propanolwater gradient except that the peaks were generally narrower and more symmetrical with propanol as the secondary solvent. As a result, the methanol-water gradient demonstrated poor resolution of dehydroepiandrosterone, 17a-hydroxypregnenolone, and 17ahydroxyprogesterone as compared to the propanol-water gradient with a given packing. With all three packings, the acetonitrilewater system separated cortisone from cortisol poorly and did not resolve the 5-ene-3@-ol and 4-ene-3-one adrenal steroids. However, with Apex ODS packing, this solvent system resolved 17a-hydroxy-20a-dihydroprogesterone and 20a-dihydroprogesterone from the other steroids, a useful feature achieved also by propanol-water. DISCUSSION
While propanol has been used in RPHPLC of proteins because of its high solvent strength and relatively good solubility for
380
CAPP AND SIMONIAN
proteins (15), it has not been much used in reversed-phase partition chromatography, where it exhibits a unique selectivity for steroids, differing even from other primary alcohols. The similarities in elution patterns of the steroids with propanol and methanol on Apex ODS, LiChrosorb RPl8, and Spherisorb SS-ODS packings suggest that within a class of solvent (e.g., primary alcohols) a given packing will maintain its general retention characteristics but that by altering the solvent within that class fine tuning of separations may be achieved. Steroid selectivity differences among different reversed-phase packings has not been attributed to particle shape but instead to differences in accessible silanol groups on the packing (8). Among the three packings used in this study, Apex ODS has the least and Spherisorb SS-ODS the greatest number of free silanols (8,16). Thus the simultaneous separation of all the major unconjugated adrenal steroids by the propanol-water gradient with the Apex ODS packing is presumably the result of the appropriate combination of solvent and packing selectivities. The propanol-water and methanol-water gradient systems were compared and the quantitative results agreed for the HPLC analysis of steroidogenesis by human adrenocortical cell cultures. The advantage of the propanol-water gradient was demonstrated by the simultaneous separation of 5-ene-3/301 steroid intermediates and 4-ene-3-one steroid intermediates and end products metabolized from radiolabeled pregnenolone. For cell culture samples of ~5 X lo6 cells, detection of 5-ene-3/3-ol steroid production requires a radiolabeled precursor as used in this study or a specific radioimmunoassay of eluant fractions (8). However, detection of l-2 ng/ peak of 4-ene-3-one steroids by absorbance at 246 nm after HPLC separation allowed measurement of the production of these steroids from an unlabeled exogenous precursor in cultures of >30,000 cells. The use of the
propanol-water gradient for HPLC analysis would not be limited to steroid production studies with in vitro culture systems. Just as the methanol-water gradient system has been used to analyze steroids in solid adrenal tissue after appropriate preparation and extraction (8), propanol-water also could he used for analysis of these samples. As with any chromatographic system used to analyze biological systems, the ability to extract the solute of interest is the primary limitation for the analysis. The adrenal steroids normally produced by most animal species and humans can be separated in a single chromatographic analysis using the propanol-water gradient and a fully covered octadecylsilyl packing. However, given the uncertainty of any biological sample contents, confirmation of steroid identity by analysis with two solvent systems with different selectivities is routinely performed in this laboratory ( 10,12,13). An increasing number of multisolvent optimization programs are being developed for both isocratic and gradient elutions, enabling the chromatographer to design chromatographic systems giving maximal resolution of a complex mixture of compounds ( 17). However, these solvent systems require chromatographic systems with ternary or quaternary solvent capability and considerable amount of computer time. In laboratories with small sample throughput, an optimization can take longer to run than the samples for a complete project. REFERENCES 1. Vinson, G. P., and Whitehouse, J. B. (1969) Acta Endocrinol. 61, 695-708. 2. O’Hare, M. J., and Neville, A. M. (1973) J. Endocrinol.
58, 447-462.
3. Schoneshofer, M., Skobolo, R., and Dulce, H. J. (1980) J. Chromatogr. 222, 478-48 1. 4. Cavina, G., Moretti, G., Alimenti, R., and Gallinella, B. (1979) J. Chromatogr. 175, 125-140. 5. O’Hare, M. J., Nice, E. C., Magee-Brown, R., and Bullman, H. (1976) J. Chromatogr. 125, 357367.
CHROMATOGRAPHIC
SEPARATION
6. Schoneshofer, M., Fenner, A., and Duke, H. J. (1980) J. Steroid Biochem. 14, 377-386. 7. Schoneshofer, M., Maxeiner, J., and Fenner, A. (1981) J. Chromatogr. 224, 229-237. 8. O’Hare, M. J., and Nice, E. C. (1981) in Steroid Analysis by HPLC Recent Applications (Kautsky, M. P., ed.), pp. 217-322, Dekker, New York/ BaseI. 9. Huang, F., Ke, F., Hwang, J., and Lo, T. (1983) Arch. Biochem. Biophys. 225, 5 12-5 17. 10. Simonian, M. H., and Capp, M. W. (1984) J. Clin. Endocrinol. Metab. 59, 643-65 1. 11. Simonian, M. H., and Gill, G. N. (198 1) Endocrinology 108, 1769-1779.
OF ADRENAL
STEROIDS
381
12. Simonian, M. H., White, M. L., and Gill, G. N. (1982) Endocrinology 111, 919-927. 13. Simonian, M. H., Homsby, P. J., Ill, C. R., O’Hare, M. J., and Gill, G. N. (1979) Endocrinology 105, 99-108. 14. Homsby, P. J., and Aldem, K. A. (1984) J. Clin. Endocrinol. Metab. 58, 12l-l 27. 15. O’Hare, M. J., Capp, M. W., Nice, E. C., Cooke, N. H. C., and Archer, B. G. (1982) Anal. Biochem. 126, 17-28. 16. Simonian, M. H., and Capp, M. W. (1984) J. Chromatogr. 287, 97-104. 17. D’Agostino, G., Mitchell, F., Castagnetta, L., and O’Hare, M. J. (1984) J. Chromatogr. 305, 13-26.