0013-7227/90/1263-1410$02.00/0 Endocrinology Copyright © 1990 by The Endocrine Society
Vol. 126, No. 3 Printed in U.S.A.
Mineralocorticoid and Renal Receptor Binding Activity of 21-Deoxyaldosterone HIDEO KOSHIDA, ISAMU MIYAMORI, RYUICHIRO SOMA, TAKAO MATSUBARA, MASATOSHI IKEDA, RYOYU TAKEDA, SHINICHI NAKAMURA, FUMIYUKI KIUCHI, AND YOSHISUKE TSUDA Second Department of Internal Medicine, and Department of Microbiology (S.N.), School of Medicine, and Faculty of Pharmaceutical Sciences (F.K., Y. TJ, Kanazawa University, Kanazawa 920, Japan
ABSTRACT. Since several aldosterone metabolites are known to be active, we have assessed the mineralocorticoid biological and renal receptor binding activities of the aldosterone metabolites, 21-deoxyaldosterone (21-deoxy-Aldo), 21-deoxytetrahydroaldosterone (21-deoxy-THAldo), and 3a, 5/3-tetrahydroaldosterone (THAldo). We synthesized these steroids by bioreduction of aldosterone with intestinal bacteria. Mineralocorticoid agonist activity of 21-deoxy-Aldo, 21-deoxy-THAldo and THAldo, determined by bioassay using adrenalectomized rats, was 1-5%, less than 0.01%, and 0.1-0.5% that of aldosterone, respectively. 21-De.oxy-Aldo showed no antagonist activity. The relative affinity in competing with [3H] aldosterone for binding to mineralocorticoid receptors in adrenalectomized rat kidney cytosols
P
was 94%, less than 0.01%, and less than 0.01% that of aldosterone. The relative binding affinity for rat renal glucocorticoid receptors was 23%, less than 0.01%, and less than 0.01% that of dexamethasone, and for corticosteroid-binding globulin 17%, less than 0.01%, and less than 0.01% that of cortisol. These results show that the naturally occurring steroid, 21-deoxy-Aldo, possesses mineralocorticoid agonist activity which is equivalent to that of 11-deoxycorticosterone, and has substantial affinity for rat renal mineralocorticoid and glucocorticoid receptors. The results also implicate the pathophysiological role of 21-deoxyAldo as a potential mineralocorticoid in 21-hydroxylase deficiency, where urinary excretion of this steroid is invariably elevated. (Endocrinology 126: 1410-1415,1990)
The only studies on the mineralocorticoid activity and the affinity for mineralocorticoid receptors of 21-deoxyAldo are those by Hesse and Pechet (8) and Wynne et al. (9), in which only results of 21-deoxy-Aldo derivatives and 17-iso-21-deoxy-Aldo were described. In the present study we undertook to synthesize 21-deoxy-Aldo and 21deoxytetrahydroaldosterone (21-deoxy-THAldo) from aldosterone using intestinal bacteria, and to determine the mineralocorticoid activity and binding affinity of these steroids for rat renal mineralocorticoid and glucocorticoid receptors. 21-Deoxy-Aldo was found to have mineralocorticoid activity approximately equivalent to that of 11-deoxycorticosterone, and to have a binding affinity for mineralocorticoid receptors similar to that of aldosterone.
REVIOUS studies have demonstrated the presence of 21-deoxy metabolites of aldosterone in urine (1). Since mammalian enzymes are, however, not capable of reducing 21-hydroxy corticoids to 21-deoxy derivatives, the 21-deoxy metabolites in urine are considered to be formed by intestinal bacteria (2, 3). Indeed, our previous study (4) has shown that intestinal flora participate in the metabolism of aldosterone in the enterohepatic circulation of both normal subjects and patients with cirrhosis of the liver. In contrast, 21-deoxyaldosterone (21deoxy-Aldo) can be 21-hydroxylated to aldosterone by adrenal tissue from cows (5), bullfrogs, and humans (6). Lewicka et al. (7) have demonstrated that urinary excretion of 21-deoxy-Aldo is invariably elevated in 21-hydroxylase deficient patients of congenital adrenal hyperplasia, and assumed that 21-hydroxylase deficiency in the zona glomerulosa of the adrenal gland resulted in the excessive release of 21-deoxy-Aldo into circulation.
Steroids
Received July 12, 1989. Address requests for reprints to: Hideo Koshida, M.D., Second Department of Internal Medicine, School of Medicine, Kanazawa University, Takara-Machi 13-1, Kanazawa 920, Japan. 1 This work was supported in part by a research grant from the Intractable Diseases Division, Public Health Bureau, Ministry of Health and Welfare, Japan, and by a research grant from the Ministry of Education, Japan (A 63440091).
Aldosterone, 3a, 5/?-tetrahydroaldosterone (THAldo), dexamethasone, cortisol, and 11-deoxycorticosterone were purchased from Sigma (St. Louis, MO). D-[1,2-3H]-Aldosterone (46.9 Ci/mmol), [6,7-3H]dexamethasone (41.9 Ci/mmol), [1,23 H]cortisol (45.4 Ci/mmol), and ZK 91587 (7a-methoxycarbonyl-15/3, 16j8-methylene-3-oxo-17a;-pregn-4-ene-21,17-carbolactone) were obtained from New England Nuclear (Boston,
Materials and Methods
1410
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 19 November 2015. at 19:16 For personal use only. No other uses without permission. . All rights reserved.
21-DEOXYALDOSTERONE MA). RU 28362 [ll/3,17|3-dihydroxy-6-methyl-17a (1-propynyl) androsta-l,4,6-trien-3-one] was kindly provided by Roussel-Uclaf (Romainville, France). Preparation of tritiated and unlabeled 21-deoxy-Aldo and unlabeled 21-deoxy-THAldo Unlabeled 21-deoxy-Aldo and 21-deoxy-THAldo were biosynthesized from unlabeled aldosterone as previously described by Bokkenheuser et al. (3) with a minor modification. Briefly, 1 liter media (Brain Heart Infusion Broth; Baltimore Biological Laboratories, Cockeysville, MD) supplemented with cysteine HC1, NaHCO3, and aqueous resazurin was seeded with 5 ml Eubacterium lentum (VPI 1947-1, kindly supplied by Dr. R. Tanaka, Yakult Institute for Microbiological Research, Tokyo, Japan) and 2 ml 24-h-old culture of Escherichia coli (K-12) for the synthesis of 21-deoxy-Aldo, and with 5 ml E. lentum (VPI 1947-1), 5 ml Clostridiumparaputrificum (ATCC 17864), and 2 ml E. coli (K-12) for the synthesis of 21-deoxy-THAldo. The media were incubated at 37 C for 7 days. At the end of incubation, the media were mixed with 5 vol methanol, filtered, and evaporated under vacuum. Residues were dissolved in 100 ml water and extracted twice with 3 vol dichloromethane. The extracts were dehydrated over sodium sulfate and evaporated under vacuum. Two hundred microcuries of [3H]aldosterone was incubated in 60 ml media for the synthesis of [3H] 21deoxy-Aldo as described above. Dried residues were then subjected to chromatography to isolate unlabeled and labeled aldosterone metabolites. Preparative TLC was performed on silica gel (1 mm thickness; Silica Gel 60 F254, E. Merck, Darmstadt, FRG) using solvent systems of chloroform/rc-butanol/water (80:20:1) (vol/vol), and benzene/acetone (1:1) (vol/vol) (developed twice). The individual metabolites were located under UV light at 254 nm and by Spraying an 8% phosphomolybdate ethanol solution on the developed plates for unlabeled steroids, and by employing a TLC Scanner (TLC-500, Aloka, Tokyo, Japan) for [3H]21deoxy-Aldo. Eluates were evaporated under vacuum and applied to HPLC. The HPLC system consisted of a model LC-5A pump (Shimadzu, Kyoto, Japan), a spectrophotometric detector SPD1 (Shimadzu), and a FRAC-100 fraction collector (Pharmacia, Uppsala, Sweden). A Shim-pak PREP ODS column (20 x 250 mm; 15 /an, Shimadzu) was used for separation with 55% aqueous methanol as a mobile phase at 40 C at a constant flow rate of 7.5 ml/min. Corresponding fractions for each metabolite were determined by a spectrophotometric detector for unlabeled 21-deoxy-Aldo, and by a liquid scintillation counter model 700 (Aloka) and Aquasol-2 (New England Nuclear) as a scintillation cocktail solution for [3H]21-deoxy-Aldo, and by adding phosphomolybdate for unlabeled 21-deoxy-THAldo. After separation of unlabeled 21-deoxy-Aldo on this HPLC method, there was an additional small UV-absorbent peak (retention time; 27.5 min; abbreviated as K-l) which was more polar than that of 21-deoxy-Aldo (retention time; 33.2 min). The purity and identification of HPLC-separated unlabeled 21-deoxy-Aldo and 21-deoxy-THAldo were confirmed by gas chromatography (GC), gas chromatography-mass spectrometry (GC-MS) and proton nuclear magnetic resonance (XH-NMR) spectra. Methyloxime-trimethylsilyl ether derivatives of these
1411
steroids were prepared for GC and GC-MS analysis according to the previously described method (10). GC was performed on an MS-GCG06 instrument (JEOL, Tokyo, Japan) with a flame ionization detector, using a DB-17 column (15 m x 0.53 mm) (J & W Scientific Inc., Cordova, CA) under the following conditions; column temperature: 200-250 C, 10 C/min; injection temperature: 260 C; carrier gas/He, flow rate: 20 ml/min. Mass spectra and GC-MS were determined on a JEOL D-300 machine. For GC-MS, electron impact ionization (70 electron V) was employed and spectra were recorded by automatic repetitive scanning over the mass range of 0-700 atomic mass units. The derivatives showed the following prominent peaks; 21-deoxy-Aldo mass per unit charge (m/z): 474(M+), 459(M+-15), 443(M+-31), 384(M+-90), 371(M+-103); 21deoxy-THAldo m/z: 521(M+), 506(M+-15), 490(M+-31), 431(M+—90). The methyloxime-trimethylsilyl derivatives of K1 showed mass spectra similar to that of 21-deoxy-Aldo. X H-NMR spectra were measured on a JEOL JNM GX-400 spectrometer (400 MHZ) in CDC13 with tetramethylsilane as an internal standard. 21-Deoxy-Aldo and K-l showed the following signals; 21-deoxy-Aldo 5 ppm; 1.28 (3H, s, 19-Me), 2.22 (3H, s, 21-Me), 4.59 (1H, d, J = 5.9 Hz, H n ), 5.04 (1H, s, H,8), 5.73 (1H, s, H4); K-l 8 ppm: 1.26 (3H, s, 19-Me), 2.18 (3H, s, 21-Me), 4.56 (1H, d, J = 6.5 Hz, H n ), 4.96 (1H, d, J = 4.5 Hz, H18), 5.72 (1H, brs, H4). Chemical shifts of the signals for K-l were in good agreement with those reported for 17a-21-deoxyAldo (9), indicating that K-l has a 17a-configuration. 17(3-Orientation of the methyl ketone side chain in 21-deoxyAldo prepared by us was indicated by its optical rotation value ([a]D + 199°, c = 0.22, CHC13; literature (8): [a]D + 195°). This was further confirmed by base catalyzed isomerization of 21deoxy-Aldo. 21-Deoxy-Aldo was dissolved in a mixture of methanol and 1 N NaOH (10:1) and stored at room temperature for 3 h. HPLC analysis of this mixture showed that 90% of 21deoxy-Aldo changed to K-l. This indicates that 21-deoxy-Aldo has an unstable 17/3-configuration and that K-l is a more stable 17a-isomer. From these analyses we confirmed that the structures of the aldosterone metabolites prepared by us were 17/?-21-deoxy-Aldo (11,18-hemiacetal form) and 21-deoxy-THAldo, and we used these steroids in the following experiments. Mineralocorticoid bioassay
A modification of the procedure of Morris and co-workers (11) was used. Male Wistar rats weighing 150-200 g were bilaterally adrenalectomized under ether anesthesia and received 0.9% NaCl and regular rat chow ad libitum for 5-7 days. Rats were given no food overnight and the drinking fluid was removed 2 h before the assay. At the beginning of the experiment, rats were made to urinate with a whiff of ether and tail pinching, and injected ip with 3 ml 0.9% NaCl/100 g BW and sc with various doses of steroids/100 g BW dissolved in 0.2 ml 0.9% NaCl-2% ethanol. All steroids were freshly dissolved in ethanol on the same day of the assay and diluted to final concentrations within 2 h of the injections. In addition, 1 /xg aldosterone with 0,10, 30, and 100 ng 21-deoxy-Aldo/100 g BW was injected into rats to examine the possible antagonist activity of 21-deoxy-Aldo. Control rats received vehicle alone. Rats
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 19 November 2015. at 19:16 For personal use only. No other uses without permission. . All rights reserved.
21-DEOXYALDOSTERONE
1412
were induced to urinate with ether and tail pinching at 1 and 4 h after injection. Urine samples at 4 h were analyzed for Na and K by flame photometry.
Endo • 1990 Voll26«No3
concentrations required to reduce [3H] labeled steroid binding to half.
Results
Mineralocorticoid and glucocorticoid RRA
Mineralocorticoid RRA were performed on renal cytosol from adrenalectomized rats as described previously in detail (12). The relative abilities of various steroids to displace [3H]aldosterone were studied in the presence of 0.1 nM RU 28362, a highly specific glucocorticoid agonist (13), to block binding to glucocorticoid receptors. Preparation of kidney cytosol and procedure for the glucocorticoid RRA were identical to the mineralocorticoid RRA, except that the labeled tracer was [3H]dexamethasone at a concentration of 5 nM, a nonspecific binding was determined in parallel incubations containing a 100-fold excess of unlabeled dexamethasone. Competition by various steroids for [3H]dexamethasone binding to rat renal glucocorticoid receptors was studied in the presence of 0.1 ^M ZK 91587, a highly specific mineralocorticoid, which hardly interacts with glucocorticoid receptors (unpublished observation). Scatchard analyses of [3H]aldosterone and [3H]21deoxy-Aldo binding were performed by incubating various concentrations of [3H]aldosterone (10~10 to 10~8 M) or [3H]21-deoxy-Aldo (10"10to 10~8 M) with renal cytosol. Nonspecific binding was determined by parallel incubation with 1000-fold unlabeled aldosterone or 21-deoxyAldo.
Figure 1 shows the effects of aldosterone metabolites on urinary Na/K ratio in the adrenalectomized rat bioassay. The urinary Na/K ratio was significantly reduced at 30 Mg (P < 0.05) and 100 jug (P < 0.01) of 21-deoxyAldo/100 g BW in comparison to that of rats given vehicle alone. Although the urinary Na/K ratio was significantly reduced at 10 ng (P < 0.05) and 30 ng (P < 0.01) of THAldo, 21-deoxy-THAldo did not, even at a dose of 100 /xg. The mineralocorticoid activity of 21deoxy-Aldo, 21-deoxy-THAldo, and THAldo judged from urinary Na/K ratio was 1-5%, less than 0.01%, and 0.10.5% that of aldosterone, respectively. When 0, 10, 30, and 100 ng 21-deoxy-Aldo were injected with 1 ixg aldosterone, no antagonist activity was observed (Table 1). The mineralocorticoid activity of 21-deoxy-Aldo was due to both antinatriuretic (Fig. 2) and kaliuretic effects (Fig. 3). A Scatchard analysis of [3H]aldosterone binding to renal cytosol from adrenalectomized rats showed that the apparent maximum binding (Bmax) of aldosterone to the higher affinity sites was 21 ± 5 fmol/mg cytosol protein and the dissociation constant (Kd) was (1.8 ± 0 O • A • X
0_J
Saline Aldo 21-deoxy-Aldo THAldo 21-deoxy-THAldo 11-Deoxycorticosterone
Binding to corticosteroid-binding globulin (CBG)
The assay was carried out according to the procedure of Murphy (14). In brief, 0.9 ml 5% human plasma diluted with distilled water, containing 8 x 104 dpm [3H]cortisol was incubated with 0.1 ml 2% ethanol solution of various concentrations of steroids examined for 10 min at 10 C. Our preliminary experiment confirmed that 10 min is sufficient to achieve equilibrium. After incubation, 1 ml 0.5% charcoal (Norit A, Sigma, St. Louis, MO) and 0.05% Dextran T-70 (Pharmacia, Piscataway, NJ) in 5 nM barbital-HCl buffer (pH 7.4) was added to separate bound from free cortisol. After vigorous shaking and then standing for 10 min, the charcoal was sedimented by centrifugation for 15 min at 1000 x g. One milliliter of the supernatant was counted in 10 ml Aquasol-2.
2.0-
1.0-
Statistical analysis Statistical analyses of differences in urinary electrolytes within the same steroid group were performed by KruskalWallis test with repeated measures followed by Scheffe's multiple range test. The relative bioassay potencies of various steroids were calculated as previously described (15). Scatchard plots were resolved by graphic analysis (16). The relative potencies of various steroids for mineralocorticoid, glucocorticoid receptors, and CBG were estimated from a comparison of the
0
0.01
0.1
10
100
Steroid Ug/100g)
FIG. 1. Effects of various steroids on urinary Na/K ratio in male adrenalectomized rats. Values are mean ± SE; n = 4-10 for each dosage. , saline; O, aldosterone (Aldo); • , 21-deoxy-Aldo; A, THAldo; • , 21deoxy-THAldo; x, 11-deoxycorticosterone. ', P < 0.05; *', P < 0.01 vs. saline.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 19 November 2015. at 19:16 For personal use only. No other uses without permission. . All rights reserved.
21-DEOXYALDOSTERONE
1413
TABLE 1. Effects of 21-deoxyaldosterone (21-deoxy-Aldo) on the urinary electrolyte excretion in adrenalectomized rats in the presence of aldosterone (Aldo) Aldo (ng/100 g) 21-deoxy-Aldo (/ig/100 g) Urinary Na/K ratio Urinary Na excretion (jtmol/3 h) Urinary K excretion (/umol/3 h)
0.78 ± 0.08 138 ± 10 180 ± 10
1 10
1 30
1 100
0.80 ± 0.10 142 ± 15 175 ± 8
0.75 ± 0.12 136 ± 13 186 ± 15
0.77 ± 0.06 135 ± 9 175 ± 8
Value are the mean ± SE; n = 5-8 for each dosage. ® O • A • x
Saline Aldo 21 -deoxy-Aldo THAIdo 21-deoxy-THAIdo 11 -Deoxycorticosterone
25OH
200-
S 200-
150-
150-
100
100 0
Saline Aldo 21 -deoxy-Aldo. THAIdo 21-deoxy-THAIdo 11-Deoxycorticosterone
0
0.01
0.1
1 Steroid (^g/100g)
10
100
FIG. 2. Effects of various steroids on urinary Na excretion in male adrenalectomized rats. Values are mean ± SE; n = 4-10 for each dosage. ®, saline; O, aldosterone (Aldo); • , 21-deoxy-Aldo; A, THAIdo; • , 21deoxy-THAldo; X, 11-deoxycorticosterone. ', P < 0.05; ", P < 0.01 us. saline.
0.3) x 10"9 M. The Kd for lower affinity sites was (3.5 ± 0.6) x 10~8 M (mean ± SE, n = 5). The Scatchard plot of [3H] 21-deoxy-Aldo was also curvilinear. Figure 4 shows a representative Scatchard plot. The B max and Kd of 21deoxy-Aldo for the higher affinity sites was 12 ± 3 fmol/ mg cytosol protein and (1.1 ± 0.2) x 10"9 M. The Kd for lower affinity sites was (4.8 ± 0.7) X 10~8 M (mean ± SE, n = 5), values very similar to those for aldosterone. Competition by 21-deoxy-Aldo, 21-deoxy-THAldo, and THAIdo for [3H] aldosterone binding to rat renal mineralocorticoid receptors revealed that the potencies relative to aldosterone were 94%, less than 0.01%, and less than 0.01%, respectively, as shown in Fig. 5.
0.01
0.1
1 Steroid (^g/100g)
10
100
FIG. 3. Effects of various steroids on urinary K excretion in male adrenalectomized rats. Values are mean ± SE; n = 4-10 for each dosage. ®, saline; O, aldosterone (Aldo); • , 21-deoxy-Aldo; A, THAIdo; • , 21deoxy-THAldo; x, 11-deoxycorticosterone.', P < 0.05; ", P < 0.01 us. saline.
Competition by aldosterone metabolites for [3H]dexamethasone binding to rat renal glucocorticoid receptors is shown in Fig. 6. The relative binding potencies of 21deoxy-Aldo, 21-deoxy-THAldo, and THAIdo were 23%, less than 0.01%, and less than 0.01% that of dexamethasone, respectively. The relative binding potencies of 21-deoxy-Aldo, 21deoxy-THAldo, and THAIdo for CBG were 17%, less than 0.01%, and less than 0.01% that of cortisol, respectively (Fig. 7). Discussion In the present study, we undertook to synthesize 21deoxy-Aldo from aldosterone using intestinal bacteria.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 19 November 2015. at 19:16 For personal use only. No other uses without permission. . All rights reserved.
Endo • 1990 Vol 126 • No 3
21-DEOXYALDOSTERONE
1414 X1O 2
THAldo 100
2.0
KdI = 1.0iX10-9M KdI=4.59xiO"8M
1.5
75
1.0
50 Dex 25
0.5
10 3
H-21-deoxy-Aldo
20
30
0—T-
40
10"
Bound (fmol/mg protein)
10"
10"
10"
10"
Molar Concentration of Competitor
FIG. 4. Scatchard analysis of 3H-21-deoxy-Aldo binding to adrenalectomized rat renal cytosol. Cytosol was incubated with 10~10-10~8 M [3H] 21-deoxy-Aldo at 4 C for 4 h. Values shown were corrected for nonspecific binding by subtracting the binding obtained after incubation with 1000-fold unlabeled 21-deoxy-Aldo. Lines were calculated using graphic analysis.
FIG. 6. Competition of unlabeled dexamethasone (Dex; • ) , aldosterone (Aldo; O), 21-deoxy-Aldo; • , THAldo; A and 21-deoxy-THAldo; • for [3H]dexamethasone (5 nM) binding to adrenalectomized rat renal cytosol in the presence of 0.1 nM ZK 91587. Values are expressed as mean (n = 3-5), corrected for nonspecific binding in the presence of 100-fold unlabeled dexamethasone.
• THAldo
100-
THAldo
100
75-
a
75
Aldo
50-
50
25-
25
21-deoxy-Aldo
10"
10-
9
10" 8
10- 7
10- 6
Molar Concentration of Competitor
FIG. 5. Competition of unlabeled aldosterone (Aldo; O), 21-deoxyaldosterone (21-deoxy-Aldo; • ) , THAldo; A, and 21-deoxy-THAldo; • for [3H]aldosterone (2 nM) binding to adrenalectomized rat renal cytosol in the presence of 0.1 /zM RU 28362. Values are expressed as mean (n = 3-5), corrected for nonspecific binding in the presence of 100-fold unlabeled aldosterone.
Although Wynne et al. (9) attempted to synthesize this steroid by a chemical procedure using 11/3-hydroxyprogesterone-3,20-bisethylene acetal, they obtained only 17a-21-deoxy-Aldo or derivatives of 17/3-21-deoxy-Aldo, and failed to get 17jS-21-deoxy-Aldo. They supposed that the unintentional conversions were due to the use of a silica gel column and the prolonged exposure of the steroid to methanol during the separation procedure. Thus we chose the biosynthetic procedure using intestinal bacteria, avoided these separation methods, and confirmed the chemical structure of the isolated steroid
10"
10"
10"
10"
10"
Molar Concentration of Competitor
FIG. 7. Competition of unlabeled cortisol (A)4, aldosterone (Aldo; O), 21-deoxyaldosterone (21-deoxy-Aldo; • ) , THAldo; A, and 21-deoxyTHAldo; • for [3H]cortisol binding (0.8 nM) to diluted human plasma. Values are expressed as mean (n = 3-5).
to be 170-21-deoxy-Aldo by GC-MS and 'H-NMR studies. Although Hesse et al. (8) reported that 21-deoxy-Aldo exerted very little influence on the transport of sodium, they did not present precise data. In the report of Wynne et al. (9), 18- and 20-methyl ethers of 21-deoxy-Aldo exhibited about l/100th of the mineralocorticoid activity of aldosterone, and 20-methyl ether of 21-deoxy-Aldo did not show any mineralocorticoid antagonism in the presence of aldosterone. Our present study showed that 21deoxy-Aldo possessed mineralocorticoid agonist activity which was almost equal to 11-deoxycorticosterone, but
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 19 November 2015. at 19:16 For personal use only. No other uses without permission. . All rights reserved.
21-DEOXYALDOSTERONE
did not show antagonism, in the adrenalectomized rat bioassay. From the viewpoint of structure-activity relationships, it is generally assumed that the C-21 hydroxy group of naturally occurring adrenal mineralocorticoids is essential for their effect on electrolyte balance (17). Elimination of the C-21 hydroxy group from aldosterone diminished but did not abolish mineralocorticoid activity, and did not change the affinity for renal mineralocorticoid receptors. The discrepancy between biological activity and receptor affinity may reflect the half-life, uptake, binding to plasma proteins, distribution and metabolism of this compound (18). Higher binding to plasma proteins than that of aldosterone in the present study could contribute at least in part to the diminished bioactivity seen in vivo. The undiminished receptor affinity may be in part due to the presence of the 11,18-hemiacetal bridge of 21-deoxy-Aldo, which provides protection from inactivation by 110-hydroxysteroid dehydrogenase (19). The urinary excretion of 21-deoxy-Aldo in normal subjects and in patients with liver cirrhosis has been reported (4). The wide range of results among these subjects may be attributable to the quantities and/or enzyme activities of intestinal flora, which metabolize aldosterone to 21-deoxy-Aldo or 21-deoxy-THAldo. Previously, Lewicka et al. (7) have suggested that because 21-deoxy-Aldo could be converted to aldosterone by adrenal tissue from cows (5), bullfrogs, and humans (6), the zona glomerulosa of the adrenal gland released excessive 21-deoxy-Aldo in cases of 21-hydroxylase deficiency. They reported that the urinary excretion of 21deoxy-Aldo in 21-hydroxylase deficiency, including salt losing and simple virilizing forms, ranged from 1.8 ng/ day to 508.2 /ug/day. Similarly vast ranges for plasma 21deoxycortisol and 21-deoxycorticosterone in 21-hydroxylase deficiency have also been reported (20). The mineralocorticoid activity of 21-deoxy-Aldo is approximately 1-5% that of aldosterone; in theory, at least in some cases of 21-hydroxylase deficiency with extremely elevated urinary excretion of 21-deoxy-Aldo, it may contribute substantially to the sum total of mineralocorticoid activity. While the mechanism of salt wasting in patients with 21-hydroxylase deficiency remains controversial, it has been postulated that salt losers are unable to produce sufficient quantities of aldosterone, or alternatively produce elevated levels of progesterone and 17hydroxyprogesterone which are mineralocorticoid antagonists (21, 22). References 1. Kelly WG, Bandi L, Lieberman S 1963 Isolation and characterization of human urinary metabolites of aldosterone. 5. dihydroaldos-
1415
terone and 21-deoxytetrahydroaldosterone. Biochemistry 2:1249 2. Eriksson H, Gustafsson JA, Sjovall J 1969 Steroids in germfree and conventional rats. 21-dehydroxylation by intestinal microorganisms. Eur J Biochem 9:550 3. Bokkenheuser VD, Winter J, Honour JW, Shackleton CHL 1979 Reduction of aldosterone by anaerobic bacteria: origin of urinary 21-deoxy metabolites in man. J Steroid Biochem 11:1145 4. Miyamori I, Koshida H, Matsubara T, Ikeda M, Takeda Y, Takeda R, Vecsei P 1988 Role of intestinal bacteria in the metabolism of aldosterone in man. Horm Res 29:147 5. Wettstein A 1961 Biosynthese des hormones steroides. Experientia 17:329 6. Ulick S Normal and alternate pathways in aldosterone biosynthesis. Fourth International Congress of Endocrinology, Washington DC, 1972, p 761 7. Lewicka S, Winter J, Bige K, Bokkenheuser V, Vecsei P, Abdelhamid S, Heinrich U 1987 21-Deoxyaldosterone excretion in patients with primary aldosteronism and 21-hydroxylase deficiency. J Clin Endocrinol Metab 64:771 8. Hesse RH, Pechet MM 1965 Substitution at unactivated carbon. The synthesis of 18- and 19-substituted derivatives of 11/3-hydroxyprogesterone. J Org Chem 30:1723 9. Wynne KN, Rae ID, O'Keefe DF, Adam WR, Pearce P, Stockigt JR, Funder JW 1981 Mineralocorticoid activity of 21-deoxyaldosterone derivatives: structure-function studies. J Steroid Biochem 14:1041 10. Honour JW, Shackleton CHL 1977 Mass spectrometric analysis for tetrahydroaldosterone. J Steroid Biochem 8:299 11. Morris DJ, Kenyon CJ, Latif SA, McDermott M, Goodfriend TL 1983 The possible biological role of aldosterone metabolites. Hypertension [Suppl l]5:l-35 12. Takeda R, Miyamori I, Soma R, Matsubara T, Ikeda M 1987 Glycyrrhizic acid and its hydrolysate as mineralocorticoid agonist. J Steroid Biochem 27:845 13. Teutsch G, Costerousse G 1981 17a-alkynyl-ll/3, 17-dihydroxyandrostane derivatives: a new class of potent glucocorticoids. Steroids 38:651 14. Murphy BEP 1967 Some studies of the protein-binding of steroids and their application to the routine micro and ultramicro measurement of various steroids in body fluids by competitive proteinbinding radioassay. J Clin Endocrinol Metab 27:973 15. Colquhoun D 1971 Lectures on biostatistics. Oxford University Press, London 16. Chamness GC, McGuire WL 1975 Scatchard plots: common errors in correction and interpretation. Steroids 26:538 17. Hall CE, Gomez-Sanchez CE, Hungerford S, Gomez-Sanchez EP 1981 Mineralocorticoid and hypertensive effects of 19-nor-progesterone. Endocrinology 109:1168 18. Morris DJ, Brem AS, Saccoccio NA, Pacholski M, Harnik M 1986 Mineralocorticoid activity of 19-hydroxyaldosterone, 19-nor-aldosterone, and 3/3-hydroxy-As-aldosterone: relative potencies measured in two bioassay systems. Endocrinology 118:2505 19. Funder JW, Pearce PT, Smith R, Smith AI1988 Mineralocorticoid action: target tissue specificity is enzyme, not receptor, mediated. Science 242:583 20. Gueux B, Fiet J, Galons H, Bonete R, Villette JM, Vexiau P, Pham-Huu-Trung MT, Raux-Eurin MC, Gourmelen M, Brerault JL, Julien R, Dreux C 1987 The measurement of ll/3-hydroxy-4pregnene-3,20-dione (21-deoxycorticosterone) by radioimmunoassay in human plasma. J Steroid Biochem 26:145 21. Stoner E, Dimartino-Nardi J, Kuhnle U, Levine LS, Oberfield SE, New MI 1986 Is salt-wasting in congenital adrenal hyperplasia due to the same gene as the fasciculata defect? Clin Endocrinol (Oxf) 24:9 22. White PC, New MI, Dupont B 1987 Congenital adrenal hyperplasia (first of two parts). N Engl J Med 316:1519
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 19 November 2015. at 19:16 For personal use only. No other uses without permission. . All rights reserved.
Vol. 126, No. 3 Printed in U.S.A.
Mineralocorticoid and Renal Receptor Binding Activity of 21-Deoxyaldosterone HIDEO KOSHIDA, ISAMU MIYAMORI, RYUICHIRO SOMA, TAKAO MATSUBARA, MASATOSHI IKEDA, RYOYU TAKEDA, SHINICHI NAKAMURA, FUMIYUKI KIUCHI, AND YOSHISUKE TSUDA Second Department of Internal Medicine, and Department of Microbiology (S.N.), School of Medicine, and Faculty of Pharmaceutical Sciences (F.K., Y. TJ, Kanazawa University, Kanazawa 920, Japan
ABSTRACT. Since several aldosterone metabolites are known to be active, we have assessed the mineralocorticoid biological and renal receptor binding activities of the aldosterone metabolites, 21-deoxyaldosterone (21-deoxy-Aldo), 21-deoxytetrahydroaldosterone (21-deoxy-THAldo), and 3a, 5/3-tetrahydroaldosterone (THAldo). We synthesized these steroids by bioreduction of aldosterone with intestinal bacteria. Mineralocorticoid agonist activity of 21-deoxy-Aldo, 21-deoxy-THAldo and THAldo, determined by bioassay using adrenalectomized rats, was 1-5%, less than 0.01%, and 0.1-0.5% that of aldosterone, respectively. 21-De.oxy-Aldo showed no antagonist activity. The relative affinity in competing with [3H] aldosterone for binding to mineralocorticoid receptors in adrenalectomized rat kidney cytosols
P
was 94%, less than 0.01%, and less than 0.01% that of aldosterone. The relative binding affinity for rat renal glucocorticoid receptors was 23%, less than 0.01%, and less than 0.01% that of dexamethasone, and for corticosteroid-binding globulin 17%, less than 0.01%, and less than 0.01% that of cortisol. These results show that the naturally occurring steroid, 21-deoxy-Aldo, possesses mineralocorticoid agonist activity which is equivalent to that of 11-deoxycorticosterone, and has substantial affinity for rat renal mineralocorticoid and glucocorticoid receptors. The results also implicate the pathophysiological role of 21-deoxyAldo as a potential mineralocorticoid in 21-hydroxylase deficiency, where urinary excretion of this steroid is invariably elevated. (Endocrinology 126: 1410-1415,1990)
The only studies on the mineralocorticoid activity and the affinity for mineralocorticoid receptors of 21-deoxyAldo are those by Hesse and Pechet (8) and Wynne et al. (9), in which only results of 21-deoxy-Aldo derivatives and 17-iso-21-deoxy-Aldo were described. In the present study we undertook to synthesize 21-deoxy-Aldo and 21deoxytetrahydroaldosterone (21-deoxy-THAldo) from aldosterone using intestinal bacteria, and to determine the mineralocorticoid activity and binding affinity of these steroids for rat renal mineralocorticoid and glucocorticoid receptors. 21-Deoxy-Aldo was found to have mineralocorticoid activity approximately equivalent to that of 11-deoxycorticosterone, and to have a binding affinity for mineralocorticoid receptors similar to that of aldosterone.
REVIOUS studies have demonstrated the presence of 21-deoxy metabolites of aldosterone in urine (1). Since mammalian enzymes are, however, not capable of reducing 21-hydroxy corticoids to 21-deoxy derivatives, the 21-deoxy metabolites in urine are considered to be formed by intestinal bacteria (2, 3). Indeed, our previous study (4) has shown that intestinal flora participate in the metabolism of aldosterone in the enterohepatic circulation of both normal subjects and patients with cirrhosis of the liver. In contrast, 21-deoxyaldosterone (21deoxy-Aldo) can be 21-hydroxylated to aldosterone by adrenal tissue from cows (5), bullfrogs, and humans (6). Lewicka et al. (7) have demonstrated that urinary excretion of 21-deoxy-Aldo is invariably elevated in 21-hydroxylase deficient patients of congenital adrenal hyperplasia, and assumed that 21-hydroxylase deficiency in the zona glomerulosa of the adrenal gland resulted in the excessive release of 21-deoxy-Aldo into circulation.
Steroids
Received July 12, 1989. Address requests for reprints to: Hideo Koshida, M.D., Second Department of Internal Medicine, School of Medicine, Kanazawa University, Takara-Machi 13-1, Kanazawa 920, Japan. 1 This work was supported in part by a research grant from the Intractable Diseases Division, Public Health Bureau, Ministry of Health and Welfare, Japan, and by a research grant from the Ministry of Education, Japan (A 63440091).
Aldosterone, 3a, 5/?-tetrahydroaldosterone (THAldo), dexamethasone, cortisol, and 11-deoxycorticosterone were purchased from Sigma (St. Louis, MO). D-[1,2-3H]-Aldosterone (46.9 Ci/mmol), [6,7-3H]dexamethasone (41.9 Ci/mmol), [1,23 H]cortisol (45.4 Ci/mmol), and ZK 91587 (7a-methoxycarbonyl-15/3, 16j8-methylene-3-oxo-17a;-pregn-4-ene-21,17-carbolactone) were obtained from New England Nuclear (Boston,
Materials and Methods
1410
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 19 November 2015. at 19:16 For personal use only. No other uses without permission. . All rights reserved.
21-DEOXYALDOSTERONE MA). RU 28362 [ll/3,17|3-dihydroxy-6-methyl-17a (1-propynyl) androsta-l,4,6-trien-3-one] was kindly provided by Roussel-Uclaf (Romainville, France). Preparation of tritiated and unlabeled 21-deoxy-Aldo and unlabeled 21-deoxy-THAldo Unlabeled 21-deoxy-Aldo and 21-deoxy-THAldo were biosynthesized from unlabeled aldosterone as previously described by Bokkenheuser et al. (3) with a minor modification. Briefly, 1 liter media (Brain Heart Infusion Broth; Baltimore Biological Laboratories, Cockeysville, MD) supplemented with cysteine HC1, NaHCO3, and aqueous resazurin was seeded with 5 ml Eubacterium lentum (VPI 1947-1, kindly supplied by Dr. R. Tanaka, Yakult Institute for Microbiological Research, Tokyo, Japan) and 2 ml 24-h-old culture of Escherichia coli (K-12) for the synthesis of 21-deoxy-Aldo, and with 5 ml E. lentum (VPI 1947-1), 5 ml Clostridiumparaputrificum (ATCC 17864), and 2 ml E. coli (K-12) for the synthesis of 21-deoxy-THAldo. The media were incubated at 37 C for 7 days. At the end of incubation, the media were mixed with 5 vol methanol, filtered, and evaporated under vacuum. Residues were dissolved in 100 ml water and extracted twice with 3 vol dichloromethane. The extracts were dehydrated over sodium sulfate and evaporated under vacuum. Two hundred microcuries of [3H]aldosterone was incubated in 60 ml media for the synthesis of [3H] 21deoxy-Aldo as described above. Dried residues were then subjected to chromatography to isolate unlabeled and labeled aldosterone metabolites. Preparative TLC was performed on silica gel (1 mm thickness; Silica Gel 60 F254, E. Merck, Darmstadt, FRG) using solvent systems of chloroform/rc-butanol/water (80:20:1) (vol/vol), and benzene/acetone (1:1) (vol/vol) (developed twice). The individual metabolites were located under UV light at 254 nm and by Spraying an 8% phosphomolybdate ethanol solution on the developed plates for unlabeled steroids, and by employing a TLC Scanner (TLC-500, Aloka, Tokyo, Japan) for [3H]21deoxy-Aldo. Eluates were evaporated under vacuum and applied to HPLC. The HPLC system consisted of a model LC-5A pump (Shimadzu, Kyoto, Japan), a spectrophotometric detector SPD1 (Shimadzu), and a FRAC-100 fraction collector (Pharmacia, Uppsala, Sweden). A Shim-pak PREP ODS column (20 x 250 mm; 15 /an, Shimadzu) was used for separation with 55% aqueous methanol as a mobile phase at 40 C at a constant flow rate of 7.5 ml/min. Corresponding fractions for each metabolite were determined by a spectrophotometric detector for unlabeled 21-deoxy-Aldo, and by a liquid scintillation counter model 700 (Aloka) and Aquasol-2 (New England Nuclear) as a scintillation cocktail solution for [3H]21-deoxy-Aldo, and by adding phosphomolybdate for unlabeled 21-deoxy-THAldo. After separation of unlabeled 21-deoxy-Aldo on this HPLC method, there was an additional small UV-absorbent peak (retention time; 27.5 min; abbreviated as K-l) which was more polar than that of 21-deoxy-Aldo (retention time; 33.2 min). The purity and identification of HPLC-separated unlabeled 21-deoxy-Aldo and 21-deoxy-THAldo were confirmed by gas chromatography (GC), gas chromatography-mass spectrometry (GC-MS) and proton nuclear magnetic resonance (XH-NMR) spectra. Methyloxime-trimethylsilyl ether derivatives of these
1411
steroids were prepared for GC and GC-MS analysis according to the previously described method (10). GC was performed on an MS-GCG06 instrument (JEOL, Tokyo, Japan) with a flame ionization detector, using a DB-17 column (15 m x 0.53 mm) (J & W Scientific Inc., Cordova, CA) under the following conditions; column temperature: 200-250 C, 10 C/min; injection temperature: 260 C; carrier gas/He, flow rate: 20 ml/min. Mass spectra and GC-MS were determined on a JEOL D-300 machine. For GC-MS, electron impact ionization (70 electron V) was employed and spectra were recorded by automatic repetitive scanning over the mass range of 0-700 atomic mass units. The derivatives showed the following prominent peaks; 21-deoxy-Aldo mass per unit charge (m/z): 474(M+), 459(M+-15), 443(M+-31), 384(M+-90), 371(M+-103); 21deoxy-THAldo m/z: 521(M+), 506(M+-15), 490(M+-31), 431(M+—90). The methyloxime-trimethylsilyl derivatives of K1 showed mass spectra similar to that of 21-deoxy-Aldo. X H-NMR spectra were measured on a JEOL JNM GX-400 spectrometer (400 MHZ) in CDC13 with tetramethylsilane as an internal standard. 21-Deoxy-Aldo and K-l showed the following signals; 21-deoxy-Aldo 5 ppm; 1.28 (3H, s, 19-Me), 2.22 (3H, s, 21-Me), 4.59 (1H, d, J = 5.9 Hz, H n ), 5.04 (1H, s, H,8), 5.73 (1H, s, H4); K-l 8 ppm: 1.26 (3H, s, 19-Me), 2.18 (3H, s, 21-Me), 4.56 (1H, d, J = 6.5 Hz, H n ), 4.96 (1H, d, J = 4.5 Hz, H18), 5.72 (1H, brs, H4). Chemical shifts of the signals for K-l were in good agreement with those reported for 17a-21-deoxyAldo (9), indicating that K-l has a 17a-configuration. 17(3-Orientation of the methyl ketone side chain in 21-deoxyAldo prepared by us was indicated by its optical rotation value ([a]D + 199°, c = 0.22, CHC13; literature (8): [a]D + 195°). This was further confirmed by base catalyzed isomerization of 21deoxy-Aldo. 21-Deoxy-Aldo was dissolved in a mixture of methanol and 1 N NaOH (10:1) and stored at room temperature for 3 h. HPLC analysis of this mixture showed that 90% of 21deoxy-Aldo changed to K-l. This indicates that 21-deoxy-Aldo has an unstable 17/3-configuration and that K-l is a more stable 17a-isomer. From these analyses we confirmed that the structures of the aldosterone metabolites prepared by us were 17/?-21-deoxy-Aldo (11,18-hemiacetal form) and 21-deoxy-THAldo, and we used these steroids in the following experiments. Mineralocorticoid bioassay
A modification of the procedure of Morris and co-workers (11) was used. Male Wistar rats weighing 150-200 g were bilaterally adrenalectomized under ether anesthesia and received 0.9% NaCl and regular rat chow ad libitum for 5-7 days. Rats were given no food overnight and the drinking fluid was removed 2 h before the assay. At the beginning of the experiment, rats were made to urinate with a whiff of ether and tail pinching, and injected ip with 3 ml 0.9% NaCl/100 g BW and sc with various doses of steroids/100 g BW dissolved in 0.2 ml 0.9% NaCl-2% ethanol. All steroids were freshly dissolved in ethanol on the same day of the assay and diluted to final concentrations within 2 h of the injections. In addition, 1 /xg aldosterone with 0,10, 30, and 100 ng 21-deoxy-Aldo/100 g BW was injected into rats to examine the possible antagonist activity of 21-deoxy-Aldo. Control rats received vehicle alone. Rats
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 19 November 2015. at 19:16 For personal use only. No other uses without permission. . All rights reserved.
21-DEOXYALDOSTERONE
1412
were induced to urinate with ether and tail pinching at 1 and 4 h after injection. Urine samples at 4 h were analyzed for Na and K by flame photometry.
Endo • 1990 Voll26«No3
concentrations required to reduce [3H] labeled steroid binding to half.
Results
Mineralocorticoid and glucocorticoid RRA
Mineralocorticoid RRA were performed on renal cytosol from adrenalectomized rats as described previously in detail (12). The relative abilities of various steroids to displace [3H]aldosterone were studied in the presence of 0.1 nM RU 28362, a highly specific glucocorticoid agonist (13), to block binding to glucocorticoid receptors. Preparation of kidney cytosol and procedure for the glucocorticoid RRA were identical to the mineralocorticoid RRA, except that the labeled tracer was [3H]dexamethasone at a concentration of 5 nM, a nonspecific binding was determined in parallel incubations containing a 100-fold excess of unlabeled dexamethasone. Competition by various steroids for [3H]dexamethasone binding to rat renal glucocorticoid receptors was studied in the presence of 0.1 ^M ZK 91587, a highly specific mineralocorticoid, which hardly interacts with glucocorticoid receptors (unpublished observation). Scatchard analyses of [3H]aldosterone and [3H]21deoxy-Aldo binding were performed by incubating various concentrations of [3H]aldosterone (10~10 to 10~8 M) or [3H]21-deoxy-Aldo (10"10to 10~8 M) with renal cytosol. Nonspecific binding was determined by parallel incubation with 1000-fold unlabeled aldosterone or 21-deoxyAldo.
Figure 1 shows the effects of aldosterone metabolites on urinary Na/K ratio in the adrenalectomized rat bioassay. The urinary Na/K ratio was significantly reduced at 30 Mg (P < 0.05) and 100 jug (P < 0.01) of 21-deoxyAldo/100 g BW in comparison to that of rats given vehicle alone. Although the urinary Na/K ratio was significantly reduced at 10 ng (P < 0.05) and 30 ng (P < 0.01) of THAldo, 21-deoxy-THAldo did not, even at a dose of 100 /xg. The mineralocorticoid activity of 21deoxy-Aldo, 21-deoxy-THAldo, and THAldo judged from urinary Na/K ratio was 1-5%, less than 0.01%, and 0.10.5% that of aldosterone, respectively. When 0, 10, 30, and 100 ng 21-deoxy-Aldo were injected with 1 ixg aldosterone, no antagonist activity was observed (Table 1). The mineralocorticoid activity of 21-deoxy-Aldo was due to both antinatriuretic (Fig. 2) and kaliuretic effects (Fig. 3). A Scatchard analysis of [3H]aldosterone binding to renal cytosol from adrenalectomized rats showed that the apparent maximum binding (Bmax) of aldosterone to the higher affinity sites was 21 ± 5 fmol/mg cytosol protein and the dissociation constant (Kd) was (1.8 ± 0 O • A • X
0_J
Saline Aldo 21-deoxy-Aldo THAldo 21-deoxy-THAldo 11-Deoxycorticosterone
Binding to corticosteroid-binding globulin (CBG)
The assay was carried out according to the procedure of Murphy (14). In brief, 0.9 ml 5% human plasma diluted with distilled water, containing 8 x 104 dpm [3H]cortisol was incubated with 0.1 ml 2% ethanol solution of various concentrations of steroids examined for 10 min at 10 C. Our preliminary experiment confirmed that 10 min is sufficient to achieve equilibrium. After incubation, 1 ml 0.5% charcoal (Norit A, Sigma, St. Louis, MO) and 0.05% Dextran T-70 (Pharmacia, Piscataway, NJ) in 5 nM barbital-HCl buffer (pH 7.4) was added to separate bound from free cortisol. After vigorous shaking and then standing for 10 min, the charcoal was sedimented by centrifugation for 15 min at 1000 x g. One milliliter of the supernatant was counted in 10 ml Aquasol-2.
2.0-
1.0-
Statistical analysis Statistical analyses of differences in urinary electrolytes within the same steroid group were performed by KruskalWallis test with repeated measures followed by Scheffe's multiple range test. The relative bioassay potencies of various steroids were calculated as previously described (15). Scatchard plots were resolved by graphic analysis (16). The relative potencies of various steroids for mineralocorticoid, glucocorticoid receptors, and CBG were estimated from a comparison of the
0
0.01
0.1
10
100
Steroid Ug/100g)
FIG. 1. Effects of various steroids on urinary Na/K ratio in male adrenalectomized rats. Values are mean ± SE; n = 4-10 for each dosage. , saline; O, aldosterone (Aldo); • , 21-deoxy-Aldo; A, THAldo; • , 21deoxy-THAldo; x, 11-deoxycorticosterone. ', P < 0.05; *', P < 0.01 vs. saline.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 19 November 2015. at 19:16 For personal use only. No other uses without permission. . All rights reserved.
21-DEOXYALDOSTERONE
1413
TABLE 1. Effects of 21-deoxyaldosterone (21-deoxy-Aldo) on the urinary electrolyte excretion in adrenalectomized rats in the presence of aldosterone (Aldo) Aldo (ng/100 g) 21-deoxy-Aldo (/ig/100 g) Urinary Na/K ratio Urinary Na excretion (jtmol/3 h) Urinary K excretion (/umol/3 h)
0.78 ± 0.08 138 ± 10 180 ± 10
1 10
1 30
1 100
0.80 ± 0.10 142 ± 15 175 ± 8
0.75 ± 0.12 136 ± 13 186 ± 15
0.77 ± 0.06 135 ± 9 175 ± 8
Value are the mean ± SE; n = 5-8 for each dosage. ® O • A • x
Saline Aldo 21 -deoxy-Aldo THAIdo 21-deoxy-THAIdo 11 -Deoxycorticosterone
25OH
200-
S 200-
150-
150-
100
100 0
Saline Aldo 21 -deoxy-Aldo. THAIdo 21-deoxy-THAIdo 11-Deoxycorticosterone
0
0.01
0.1
1 Steroid (^g/100g)
10
100
FIG. 2. Effects of various steroids on urinary Na excretion in male adrenalectomized rats. Values are mean ± SE; n = 4-10 for each dosage. ®, saline; O, aldosterone (Aldo); • , 21-deoxy-Aldo; A, THAIdo; • , 21deoxy-THAldo; X, 11-deoxycorticosterone. ', P < 0.05; ", P < 0.01 us. saline.
0.3) x 10"9 M. The Kd for lower affinity sites was (3.5 ± 0.6) x 10~8 M (mean ± SE, n = 5). The Scatchard plot of [3H] 21-deoxy-Aldo was also curvilinear. Figure 4 shows a representative Scatchard plot. The B max and Kd of 21deoxy-Aldo for the higher affinity sites was 12 ± 3 fmol/ mg cytosol protein and (1.1 ± 0.2) x 10"9 M. The Kd for lower affinity sites was (4.8 ± 0.7) X 10~8 M (mean ± SE, n = 5), values very similar to those for aldosterone. Competition by 21-deoxy-Aldo, 21-deoxy-THAldo, and THAIdo for [3H] aldosterone binding to rat renal mineralocorticoid receptors revealed that the potencies relative to aldosterone were 94%, less than 0.01%, and less than 0.01%, respectively, as shown in Fig. 5.
0.01
0.1
1 Steroid (^g/100g)
10
100
FIG. 3. Effects of various steroids on urinary K excretion in male adrenalectomized rats. Values are mean ± SE; n = 4-10 for each dosage. ®, saline; O, aldosterone (Aldo); • , 21-deoxy-Aldo; A, THAIdo; • , 21deoxy-THAldo; x, 11-deoxycorticosterone.', P < 0.05; ", P < 0.01 us. saline.
Competition by aldosterone metabolites for [3H]dexamethasone binding to rat renal glucocorticoid receptors is shown in Fig. 6. The relative binding potencies of 21deoxy-Aldo, 21-deoxy-THAldo, and THAIdo were 23%, less than 0.01%, and less than 0.01% that of dexamethasone, respectively. The relative binding potencies of 21-deoxy-Aldo, 21deoxy-THAldo, and THAIdo for CBG were 17%, less than 0.01%, and less than 0.01% that of cortisol, respectively (Fig. 7). Discussion In the present study, we undertook to synthesize 21deoxy-Aldo from aldosterone using intestinal bacteria.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 19 November 2015. at 19:16 For personal use only. No other uses without permission. . All rights reserved.
Endo • 1990 Vol 126 • No 3
21-DEOXYALDOSTERONE
1414 X1O 2
THAldo 100
2.0
KdI = 1.0iX10-9M KdI=4.59xiO"8M
1.5
75
1.0
50 Dex 25
0.5
10 3
H-21-deoxy-Aldo
20
30
0—T-
40
10"
Bound (fmol/mg protein)
10"
10"
10"
10"
Molar Concentration of Competitor
FIG. 4. Scatchard analysis of 3H-21-deoxy-Aldo binding to adrenalectomized rat renal cytosol. Cytosol was incubated with 10~10-10~8 M [3H] 21-deoxy-Aldo at 4 C for 4 h. Values shown were corrected for nonspecific binding by subtracting the binding obtained after incubation with 1000-fold unlabeled 21-deoxy-Aldo. Lines were calculated using graphic analysis.
FIG. 6. Competition of unlabeled dexamethasone (Dex; • ) , aldosterone (Aldo; O), 21-deoxy-Aldo; • , THAldo; A and 21-deoxy-THAldo; • for [3H]dexamethasone (5 nM) binding to adrenalectomized rat renal cytosol in the presence of 0.1 nM ZK 91587. Values are expressed as mean (n = 3-5), corrected for nonspecific binding in the presence of 100-fold unlabeled dexamethasone.
• THAldo
100-
THAldo
100
75-
a
75
Aldo
50-
50
25-
25
21-deoxy-Aldo
10"
10-
9
10" 8
10- 7
10- 6
Molar Concentration of Competitor
FIG. 5. Competition of unlabeled aldosterone (Aldo; O), 21-deoxyaldosterone (21-deoxy-Aldo; • ) , THAldo; A, and 21-deoxy-THAldo; • for [3H]aldosterone (2 nM) binding to adrenalectomized rat renal cytosol in the presence of 0.1 /zM RU 28362. Values are expressed as mean (n = 3-5), corrected for nonspecific binding in the presence of 100-fold unlabeled aldosterone.
Although Wynne et al. (9) attempted to synthesize this steroid by a chemical procedure using 11/3-hydroxyprogesterone-3,20-bisethylene acetal, they obtained only 17a-21-deoxy-Aldo or derivatives of 17/3-21-deoxy-Aldo, and failed to get 17jS-21-deoxy-Aldo. They supposed that the unintentional conversions were due to the use of a silica gel column and the prolonged exposure of the steroid to methanol during the separation procedure. Thus we chose the biosynthetic procedure using intestinal bacteria, avoided these separation methods, and confirmed the chemical structure of the isolated steroid
10"
10"
10"
10"
10"
Molar Concentration of Competitor
FIG. 7. Competition of unlabeled cortisol (A)4, aldosterone (Aldo; O), 21-deoxyaldosterone (21-deoxy-Aldo; • ) , THAldo; A, and 21-deoxyTHAldo; • for [3H]cortisol binding (0.8 nM) to diluted human plasma. Values are expressed as mean (n = 3-5).
to be 170-21-deoxy-Aldo by GC-MS and 'H-NMR studies. Although Hesse et al. (8) reported that 21-deoxy-Aldo exerted very little influence on the transport of sodium, they did not present precise data. In the report of Wynne et al. (9), 18- and 20-methyl ethers of 21-deoxy-Aldo exhibited about l/100th of the mineralocorticoid activity of aldosterone, and 20-methyl ether of 21-deoxy-Aldo did not show any mineralocorticoid antagonism in the presence of aldosterone. Our present study showed that 21deoxy-Aldo possessed mineralocorticoid agonist activity which was almost equal to 11-deoxycorticosterone, but
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 19 November 2015. at 19:16 For personal use only. No other uses without permission. . All rights reserved.
21-DEOXYALDOSTERONE
did not show antagonism, in the adrenalectomized rat bioassay. From the viewpoint of structure-activity relationships, it is generally assumed that the C-21 hydroxy group of naturally occurring adrenal mineralocorticoids is essential for their effect on electrolyte balance (17). Elimination of the C-21 hydroxy group from aldosterone diminished but did not abolish mineralocorticoid activity, and did not change the affinity for renal mineralocorticoid receptors. The discrepancy between biological activity and receptor affinity may reflect the half-life, uptake, binding to plasma proteins, distribution and metabolism of this compound (18). Higher binding to plasma proteins than that of aldosterone in the present study could contribute at least in part to the diminished bioactivity seen in vivo. The undiminished receptor affinity may be in part due to the presence of the 11,18-hemiacetal bridge of 21-deoxy-Aldo, which provides protection from inactivation by 110-hydroxysteroid dehydrogenase (19). The urinary excretion of 21-deoxy-Aldo in normal subjects and in patients with liver cirrhosis has been reported (4). The wide range of results among these subjects may be attributable to the quantities and/or enzyme activities of intestinal flora, which metabolize aldosterone to 21-deoxy-Aldo or 21-deoxy-THAldo. Previously, Lewicka et al. (7) have suggested that because 21-deoxy-Aldo could be converted to aldosterone by adrenal tissue from cows (5), bullfrogs, and humans (6), the zona glomerulosa of the adrenal gland released excessive 21-deoxy-Aldo in cases of 21-hydroxylase deficiency. They reported that the urinary excretion of 21deoxy-Aldo in 21-hydroxylase deficiency, including salt losing and simple virilizing forms, ranged from 1.8 ng/ day to 508.2 /ug/day. Similarly vast ranges for plasma 21deoxycortisol and 21-deoxycorticosterone in 21-hydroxylase deficiency have also been reported (20). The mineralocorticoid activity of 21-deoxy-Aldo is approximately 1-5% that of aldosterone; in theory, at least in some cases of 21-hydroxylase deficiency with extremely elevated urinary excretion of 21-deoxy-Aldo, it may contribute substantially to the sum total of mineralocorticoid activity. While the mechanism of salt wasting in patients with 21-hydroxylase deficiency remains controversial, it has been postulated that salt losers are unable to produce sufficient quantities of aldosterone, or alternatively produce elevated levels of progesterone and 17hydroxyprogesterone which are mineralocorticoid antagonists (21, 22). References 1. Kelly WG, Bandi L, Lieberman S 1963 Isolation and characterization of human urinary metabolites of aldosterone. 5. dihydroaldos-
1415
terone and 21-deoxytetrahydroaldosterone. Biochemistry 2:1249 2. Eriksson H, Gustafsson JA, Sjovall J 1969 Steroids in germfree and conventional rats. 21-dehydroxylation by intestinal microorganisms. Eur J Biochem 9:550 3. Bokkenheuser VD, Winter J, Honour JW, Shackleton CHL 1979 Reduction of aldosterone by anaerobic bacteria: origin of urinary 21-deoxy metabolites in man. J Steroid Biochem 11:1145 4. Miyamori I, Koshida H, Matsubara T, Ikeda M, Takeda Y, Takeda R, Vecsei P 1988 Role of intestinal bacteria in the metabolism of aldosterone in man. Horm Res 29:147 5. Wettstein A 1961 Biosynthese des hormones steroides. Experientia 17:329 6. Ulick S Normal and alternate pathways in aldosterone biosynthesis. Fourth International Congress of Endocrinology, Washington DC, 1972, p 761 7. Lewicka S, Winter J, Bige K, Bokkenheuser V, Vecsei P, Abdelhamid S, Heinrich U 1987 21-Deoxyaldosterone excretion in patients with primary aldosteronism and 21-hydroxylase deficiency. J Clin Endocrinol Metab 64:771 8. Hesse RH, Pechet MM 1965 Substitution at unactivated carbon. The synthesis of 18- and 19-substituted derivatives of 11/3-hydroxyprogesterone. J Org Chem 30:1723 9. Wynne KN, Rae ID, O'Keefe DF, Adam WR, Pearce P, Stockigt JR, Funder JW 1981 Mineralocorticoid activity of 21-deoxyaldosterone derivatives: structure-function studies. J Steroid Biochem 14:1041 10. Honour JW, Shackleton CHL 1977 Mass spectrometric analysis for tetrahydroaldosterone. J Steroid Biochem 8:299 11. Morris DJ, Kenyon CJ, Latif SA, McDermott M, Goodfriend TL 1983 The possible biological role of aldosterone metabolites. Hypertension [Suppl l]5:l-35 12. Takeda R, Miyamori I, Soma R, Matsubara T, Ikeda M 1987 Glycyrrhizic acid and its hydrolysate as mineralocorticoid agonist. J Steroid Biochem 27:845 13. Teutsch G, Costerousse G 1981 17a-alkynyl-ll/3, 17-dihydroxyandrostane derivatives: a new class of potent glucocorticoids. Steroids 38:651 14. Murphy BEP 1967 Some studies of the protein-binding of steroids and their application to the routine micro and ultramicro measurement of various steroids in body fluids by competitive proteinbinding radioassay. J Clin Endocrinol Metab 27:973 15. Colquhoun D 1971 Lectures on biostatistics. Oxford University Press, London 16. Chamness GC, McGuire WL 1975 Scatchard plots: common errors in correction and interpretation. Steroids 26:538 17. Hall CE, Gomez-Sanchez CE, Hungerford S, Gomez-Sanchez EP 1981 Mineralocorticoid and hypertensive effects of 19-nor-progesterone. Endocrinology 109:1168 18. Morris DJ, Brem AS, Saccoccio NA, Pacholski M, Harnik M 1986 Mineralocorticoid activity of 19-hydroxyaldosterone, 19-nor-aldosterone, and 3/3-hydroxy-As-aldosterone: relative potencies measured in two bioassay systems. Endocrinology 118:2505 19. Funder JW, Pearce PT, Smith R, Smith AI1988 Mineralocorticoid action: target tissue specificity is enzyme, not receptor, mediated. Science 242:583 20. Gueux B, Fiet J, Galons H, Bonete R, Villette JM, Vexiau P, Pham-Huu-Trung MT, Raux-Eurin MC, Gourmelen M, Brerault JL, Julien R, Dreux C 1987 The measurement of ll/3-hydroxy-4pregnene-3,20-dione (21-deoxycorticosterone) by radioimmunoassay in human plasma. J Steroid Biochem 26:145 21. Stoner E, Dimartino-Nardi J, Kuhnle U, Levine LS, Oberfield SE, New MI 1986 Is salt-wasting in congenital adrenal hyperplasia due to the same gene as the fasciculata defect? Clin Endocrinol (Oxf) 24:9 22. White PC, New MI, Dupont B 1987 Congenital adrenal hyperplasia (first of two parts). N Engl J Med 316:1519
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 19 November 2015. at 19:16 For personal use only. No other uses without permission. . All rights reserved.
