Radioimmunoassay of cortisone in serum, urine, and ... - CiteSeerX

Clinical Chemistry 43:8 1397–1407 (1997)

Endocrinology and Metabolism

Radioimmunoassay of cortisone in serum, urine, and saliva to assess the status of the cortisol– cortisone shuttle Gilles Morineau,1* Ahmed Boudi,2 Ahmed Barka,3 Micheline Gourmelen,4 Françoise Degeilh,5 Noah Hardy,1 Akram Al-Halnak,1 Hany Soliman,1 James P. Gosling,6 Rene´ Julien,1 Jean-Louis Brerault,1 Philippe Boudou,1 Philippe Aubert,3 Jean-Marie Villette,1 Andre´ Pruna,3 Herve´ Galons,2 and Jean Fiet1,7

We have developed a new assay for cortisone (E) in serum, saliva, and urine involving Celite® chromatography followed by RIA with 125I-labeled E and scintillation proximity assay. The chromatography step separates cortisol (F) from E, and in combination with their RIAs, permits assessment of the status of the F–E shuttle. We report the results of basal, postcorticotropin (ACTH), and postdexamethasone E and F concentrations and their circadian fluctuations in the serum, saliva, and urine of healthy volunteers. The serum and urine F/E ratios were increased in patients with ectopic ACTH secretion, whereas in adrenal adenoma and Cushing disease only the urinary ratio was increased. In chronic renal insufficiency this ratio was increased in serum (23.5 6 3.9) but diminished in saliva (0.38 6 0.11), and in apparent mineralocorticoid excess the ratios were high in serum (44.3 6 9.3) and urine (5.35 6 0.85) compared with those of healthy subjects (serum 9.8 6 3.5, urine 0.52 6 0.29, saliva 0.52 6 0.29).

1 Laboratoire de Biochimie Hormonale, Hoˆpital Saint-Louis, 75010 Paris, France. Laboratoire de 2 Chimie Organique and 7 Biochimie, Faculte´ de Pharmacie, 75006 Paris, France. 3 Service de Ne´phrologie, Hoˆpital Foch, 92151 Suresnes, France. 4 Laboratoire d’Explorations Endocriniennes, Hoˆpital Trousseau, 75012 Paris, France. 5 Service des Molećules Marqueés, CEA/Saclay, 91191 Gif-sur-Yvette, France. 6 Department of Biochemistry and National Diagnostics Centre, University College, Galway, Ireland. *Address correspondence to this author at: Laboratoire de Biochimie Hormonale, Hoˆpital Saint-Louis, 1 ave. Claude-Vellefaux, 75475 Paris ce´dex 10, France. Fax 1 33 1 42 49 42 80; e-mail [email protected]. Received December 19, 1996; revised March 13, 1997; accepted April 28, 1997.

scintillation proximity assay • apparent mineralocorticoid excess • 11b-hydroxysteroid dehydrogenase • Cushing syndrome

INDEXING TERMS:

In humans, cortisone (E) derives mainly from the oxidation of the 11-hydroxyl function of cortisol (F) into 11ketone by 11b-hydroxysteroid dehydrogenase (11-HSD2).8 11-HSD-2 is located in mineralocorticoid target tissues, principally kidney, colon [1], and parotid gland [2, 3]. The kidney is the major site of the conversion of F to E [4]. In vivo, 11-HSD-2 confers the mineralocorticoid receptor (MR) with specificity for circulating aldosterone in preference to far more abundant F and corticosterone (B). In vitro F and B bind to the MR with affinity equal to that of aldosterone [5], whereas in vivo, 11-HSD-2 transforms F and B (but not aldosterone) into their respective metabolites E and dehydrocorticosterone, which have much less affinity for the MR. Inversely, conversion of E to F takes place mainly in the liver [6], under the action of 11-HSD-1. This enzyme probably regulates access of glucocorticoids to their receptors [7]. In the adult, plasma F exceeds E 10-fold, whereas in the fetus, in which 11-HSD-2 is present in many tissues [3], the E concentration is threefold higher than that of F (11-HSD-2 is also abundant in the placenta [8, 9]). In the fetus, 11-HSD-2 is thought to prevent the deleterious

8 Nonstandard abbreviations: 11-HSD-2, 11b-hydroxysteroid dehydrogenase type 2; MR, mineralocorticoid receptor; E, cortisone; F, cortisol; B, corticosterone; AME-1, apparent mineralocorticoid excess type 1; THF, 5atetrahydrocortisol; aTHF, 5b-tetrahydrocortisol; THE, 5a-tetrahydrocortisone; ACTH, corticotropin; DXM, dexamethasone; CRI, chronic renal insufficiency; PGB, phosphate gelatin buffer; SPA, scintillation proximity assay; cortisone-3CMO, cortisone-3-(O-carboxymethyl) oxime; BSA, bovine serum albumin; and PRA, plasma renin activity.

1397

1398

Morineau et al.: RIA for cortisone

effects of high F concentrations and to help adrenal development [3]. Patients with mutations in the 11-HSD-2 gene [10 –14] have been reported to have a severe and sometimes fatal [15–17] hypertensive syndrome called apparent mineralocorticoid excess type 1 (AME-1) [18] and to have a low birth weight [19]. Thus, the F–E shuttle appears to be implicated in fetal development and arterial hypertension. Methods that aim to evaluate its status may be clinically relevant. Ulick et al. used gas chromatography to establish a diagnosis of AME on the basis of urinary assays of the tetrahydrometabolites of F (THF and aTHF) and E (THE) [18]. While useful, this method is lengthy and timeconsuming and the THF 1 aTHF/THE ratio obtained reflects not only the altered activities of 11-HSD-2, but also those of the 5a- and 5b-reductases encountered in AME-1 [20]. We describe a new, sensitive, specific, and simple RIA for E that, when used in combination with a F RIA, provides an accurate and unambiguous assessment of F/E equilibrium. After validating the new E RIA, we studied variations in basal and post-Cosyntropin {Synacthen® [a-1-24 corticotropin (ACTH)]} and dexamethasone (DXM) concentrations of E and F over 24 h in simultaneous samples of serum, saliva, and urine. We also evaluated the clinical applications of this new E assay in Cushing syndrome and in chronic renal insufficiency (CRI), as well as its diagnostic potential in AME.

Materials and Methods All investigations conformed with the ethical standards laid down by the Helsinki declaration (1975) as revised at Tokyo (1983).

Samples. For the circadian study, in addition to 24-h urines, saliva and serum samples were collected at 8, 12, 16, and 20 h, and at 0 and 4 h the next day. Urine samples were also obtained at approximately the same times from 20 of the healthy subjects. Aliquots were stored at 220 °C until assay. We tested the sera and saliva of the CRI patients at 8 h, before they began dialysis. A neutral (flavorless) chewing-gum was chewed by both healthy subjects and patients, who had nothing by mouth for at least 1 h before obtaining 8-mL saliva samples from them. Reagents. Steroids were purchased from Sigma (SaintQuentin-Fallavier, France) and Steraloids (Wilton, NH), and tritiated F 2.40 TBq/mmol from Amersham (Les Ulis, France). Solvents were of analytical or HPLC grade (Merck, Nogent/Marne, France). Phosphate gelatin buffer (PGB) was prepared from 0.04 mol/L phosphate buffer, pH 7.4, 1 1 g/L gelatin. BCS scintillation liquid was purchased from Amersham and RPN140 antirabbit scintillation proximity assay reagent (SPA) from Amersham. 125

I-labeled ACTH IRMA kit. A Nichols Allegro HS-ACTH kit (ref. CA2194) purchased from Mallinckrodt-Diagnostica-France (Evry, France) was used to assay ACTH.

125

I-labeled aldosterone RIA kit. Aldosterone was assayed with DPC Coat-a-count® kit (TKAL20) obtained from SAPB Hoëchst Behring (Rueil-Malmaison, France). Angiotensin I, RIA 125I kit. RENCTK (P2721) bought from Sorin-Biomedica (Antony, France) was used for plasma renin activity assay. 125

I-labeled F RIA kit. An Incstar kit (ref. CA1549) obtained from Sorin-Biomedica was used for F determination, after extraction and Celite chromatography separation (see below).

Healthy volunteers. Fifty healthy Caucasian nonsmoking volunteers (27 women and 23 men) ranging in age from 21 to 46 years (mean 28) taking no medication or contraceptives and not eating licorice were included in the study to establish normal values. Thirty-one of them (17 women and 14 men) participated in a circadian study. Sixteen underwent dynamic testing with short-acting Synacthen (Novartis, Rueil-Malmaison, France) (0.25 mg intramuscularly), five with long-acting Synacthen (1 mg), and three were subjected to standard DXM suppression tests (3 mg 3 5 days).

Preparation of [1,2-3H]E. Prednisone (10 mg) in ethanol (3 mL) solution was reduced by tritium gas with tristriphenylphosphine rhodium chloride (Wilkinson catalyst) (20 mg) under constant stirring at ambient temperature and pressure for 18 h. Separation of prednisone and E was carried out by HPLC and a Zorbax C18 column with a mobile phase consisting of methanol:water (60:40 by vol) at a flow rate of 2 mL/min. The specific activity of tritiated E was found to be 1.04 TBq/mmol (28 Ci/mmol).

Patients. We also studied a selected population of patients in university teaching hospital endocrinology and nephrology services in the Paris area. Two of the patients had primary adrenal insufficiency (Addison disease), two had secondary adrenal insufficiency (hypopituitarism), 10 had adrenal adenoma, two had Cushing disease, five had ectopic ACTH secretion, 10 were hypertensive patients with CRI, and two were cases of 11b-OH steroid dehydrogenase deficiency (AME-1).

Preparation of the cortisone-3-(O-carboxymethyl) oxime hapten (cortisone-3-CMO). To a cooled (0 –5 °C) solution of E (0.180 g, 0.5 mmol) in 20 mL of methanol, anhydrous pyrrolidine (0.071 g, 1 mmol) in 20 mL of methanol was added. After stirring for 5 min, the same quantity of pyrrolidine was added, followed by O-(carboxymethyl) hydroxylamine hemihydrochloride (0.055 g, 0.5 mmol). The mixture was brought to and kept at 50 °C for 5 min, then cooled to room temperature. The methanol was

Clinical Chemistry 43, No. 8, 1997

evaporated under vacuum and the residue redissolved in 20 mL of distilled water. This aqueous solution was adjusted to pH 2 with HCl and extracted with methylene chloride (3 3 20 mL). The combined organic extracts were washed with water (1 3 10 mL), then dried over sodium sulfate and concentrated to dryness, yielding 0.190 g (86%) of cortisone-3-CMO. The product was a mixture of two geometric isomers: E (57%) and Z (43%) (Mp 5 86 –91 °C), 1H-NMR (CDC13, TMS, Bruker 270 MHz): 0.63 (s, 18-CH3); 1.28 (s, 19-CH3); 4.21 and 4.61 (dd, Jgem 5 19 Hz, J21H-OH 5 4 Hz, 21-CH2OH); 4.60 (s, CH2CO); 5.80 (s, 4-H, E-isomer); 6.50 (s, 4-H, Z-isomer). Preparation of the cortisone-3-CMO/bovine serum albumin (BSA) immunogen. BSA was coupled to cortisone-3-CMO with the mixed anhydride method described by Erlanger et al. [21]. Preparation of anti-E antibodies. Three New Zealand White male rabbits were immunized according to the method of Vaitukaitis et al. [22]. Preparation of 125I-labeled E. The coupling of histamine to cortisone-3-CMO, radioactive labeling with 125I by the chloramine-T method of Greenwood and Hunter, modified, and purification of radiolabeled E by HPLC were carried out as previously described [23]. The specific activity of iodinated E was found to be 25 TBq/mmol (450 mCi/mmol). Celite and columns. Celite obtained from Touzard et Matignon (Vitry/Seine, France) was washed in cyclohexane and heated for 16 to 18 h at 800 °C, then kept dry at 100 °C until use. Five mL x 5 mm (i.d.) Kimble pipettes obtained from STP (Paris, France) were siliconized and stoppered at the bottom with a glass bead. Measuring instruments. The following measuring instruments and equipment were used: a Wallac 1409 beta-ray counter (Pharmacia-LKB, Saint-Quentin-en-Yvelines, France), a model 400 HPLC solvent delivery system (Applied Biosystems, Foster City, CA), and a Zorbax ODS HPLC column, 5 m, 250 3 9.4 mm from Interchim (Montluçon, France). RIA of F and E. These assays were carried out after extraction and Celite chromatography [24], with slight changes. Two thousand cpm each of [3H]F and [3H]E in PGB (50 mL) were added to the serum samples (0.5 mL), as well as to the saliva (2 mL) and urine (0.5 mL) samples, both of which had been supplemented with 0.5 mL of steroid-free serum. Two milliliters of dichloromethane was added, the samples vortex-mixed for 60 s, and then centrifuged for 10 min at 2500g. The aqueous phase was decanted and the organic phase first evaporated, then redissolved in 1.5 mL of isooctane 1 20 mL/L dichloromethane. The extract obtained was introduced into the

1399

top of a 5 mL x 5 mm (i.d.) pipette packed with 0.75 g of Celite/ethylene glycol (1 g/0.5 mL) to a depth of 6 cm and eluted under positive air pressure. The column was flushed first with 5 mL of 35:65, then with 5 mL of 50:50 dichloromethane:isooctane. One 5-mL faction of 80:20 dichloromethane:isooctane (E) and one 5-mL fraction of 60:40 ethyl acetate:isooctane (F) were collected. The eluates were evaporated in an air current and the eluted steroids redissolved in 1 mL of PGB. One-hundred microliters were counted in 4 mL of scintillation liquid for 300 s to estimate recovery. F RIA was carried out in duplicate in various aliquot volumes: 300 mL for saliva, 100 mL for urine, and 30 mL for serum, to each of which 10 mL of zero calibrator was added, with the Incstar kit. E RIA was also carried out in duplicate on 100 mL (saliva and serum) and 40 mL (urine). Calibrators consisted of duplicates containing 0, 10, 20, 50, 100, 200, 500, 1000, and 2000 pg of steroid/tube, in 100 mL of PGB solution. To these samples were added 100 mL of 125I-labeled E (10 000 cpm), 100 mL of rabbit anti-E antibody, and 100 mL of antirabbit SPA reagent [25]. The samples were vortex-mixed, then incubated at room temperature for 15 to 20 h under rotating agitation (2000 rpm) before being counted for 60 s with a b-ray counter. Calibration curves were plotted by using RIAcalc Sofware (Pharmacia LKB, Bromma, Sweden) and smooth-spline function fitting. We also performed this same E assay using [3H]E to compare the sensitivity obtained with the tritiated and the iodinated tracers. Direct RIAs of F and E (without extraction and chromatography) were also carried out in serum, saliva, and urine and the results of these analyses compared with those obtained after extraction and Celite chromatography. Statistics. The results of the F and E concentrations obtained by the two methods were compared with the Wilcoxon test. Comparisons between group mean values were carried out with Student’s t-test or the Mann– Whitney U-test. Variations in F and E and in the F/E ratio over a 24-h period, as well as before and after Synacthen or DXM administration, were evaluated by ANOVA for repeated measurements. P values .0.05 were considered nonsignificant.

Results characteristics of the anti-e antibody Antibody titers. Immunoserum dilutions that bound 50% of the iodinated tracer increased from 0.2 3 1022 during the third month to 1 3 1026 during the 14th month and remained between 1026 and 0.2 3 1026 for the following year in all three immunized rabbits. To date, we have only used immunoserum #3101 in a final PGB dilution of 1:160 000. Affinity constant. The resulting affinity constant, calculated according to Scatchard, was K 5 2.33 3 109 L mol21.

1400


Specificity (Table 1). Anti-E antiserum cross-reactions corresponding to 50% displacement of the iodinated tracer were determined according to Abraham [26] and with SPA. Antiserum #3101 was much more highly specific for E than for F, with only 0.42% cross-reactivity. Most of the interference was from the following compounds: prednisone (52%), 5a-THE (9.75%), 6b-OH-cortisone (0.71%), 11-deoxycortisol (0.70%), 5b-THE (0.53%), and prednisolone (0.46%).

assay reliability Separation of F and E by Celite chromatography. Less than 3% of each steroid eluate was found in the other fraction, resulting in negligible errors (,1% in the worst case, which was a serum E assay in which the F concentration was 10- to 20-fold that of E). In a preliminary study, we compared the direct F and E assays with the new assay, which includes extraction and Celite chromatography steps. We obtained significantly higher concentrations of F and E with the direct assays (Wilcoxon test, P ,0.001). For serum (n 5 28), saliva (n 5 20), and urine (n 5 39), they were respectively 1.20-, 1.6-, and 1.5-fold higher for F and 1.2-, 1.4-, and 1.8-fold higher for E. Therefore we decided to carry out the assays after extraction and Celite chromatography steps.

Table 1. Cross-reactivities (CR) of anti-E antiserum #3101 and anti-F antiserum of the Incstar F kit.

E 5a-THE 5b-THE 6b-OH-cortisone F 5a-DHF 5b-DHF 20a-DHF 20b-DHF 5a-THF 5b-THF 6b-OH-cortisol 11-deoxycortisol 21-deoxycortisol B 11-deoxycorticosterone Aldosterone Progesterone 17-OH-progesterone Prednisone Prednisolone Methylprednisolone DXM

Anti-E antiserum, CR %

Anti-F antiserum (Insctar kit), CR %

100 9.75 0.53 0.71 0.42 0.07 0.07 ,0.01 0.03 0.02 ,0.01 ,0.01 0.70 0.01 0.04 0.05 ,0.01 ,0.01 0.05 52.00 0.46 0.07 ,0.01

0.39 0.17 0.02 ,0.10 100 9.00 2.40 0.21 0.42 2.70 0.73 3.30 5.75 21.20 0.60 0.18 ,0.01 ,0.10 ,0.10 0.45 69.70 36.00 0.07

CR of antiserum #3101 was ,0.001% for pregnenolone, 17-OH-pregnenolone, DHEA, and DHEA-S.

Recovery of tritiated markers. The recovery rates after extraction and chromatography for F were (mean 6 SD, n 5 28): 70.1 6 9.1, 64.5 6 4.2, and 65.8 6 9.3 respectively for serum, urine, and saliva; for E they were 75.7 6 10.5, 68.4 6 6.5, and 63.2 6 7.2. Blank values. Blank assay values were found to be ,30 pg/tube for F and ,4 pg/tube for E. Accuracy. Two serum, two urine, and two saliva samples were supplemented with known quantities of F and E to the following concentrations (nmol/L): S1: F 5 310, E 5 72; S2: F 5 560, E 5 165; U1: F 5 22, E 5 72; U2: F 5 300, E 5 360; Sa1: F 5 10, E 5 20; Sa2: F 5 40, E 5 80, and assayed five times. The recovery rates were respectively (mean 6 SD): S1: F 5 91 6 6.4, E 5 86 6 6.5%; S2: F 5 93 6 6.3, E 5 97 6 5.7%; U1: F 5 110 6 7.1, E 5 104 6 5.6%; U2: F 5 93 6 6.1, E 5 95 6 6.1%; Sa1: F 5 98 6 6.2, E 5 99 6 7.3%; Sa2: F 5 101 6 7.5, E 5 94 6 5.9%. One serum, one urine, and one saliva with high concentrations of F and E (S3: F 5 1186, E 5 266; U3: F 5 728, E 5 794; Sa3: F 5 43, E 5 78 nmol/L) were progressively diluted (2- to 10-fold) with steroid-free serum and water (respectively) and assayed in triplicate. Mean results ranged from 89% to 112% of the expected values in serum, 92% to 109% in urine, and 94% to 108% in saliva. Reproducibility. Intra- (n 5 10) and interassay reproducibility (n 5 20) were respectively: S1: 6.6%, 10.3%; S2: 7.8%, 10.4%; U1: 6.2%, 11.8%; U2: 4.7%, 7.9%; Sa1: 8.2%, 10.8; Sa2: 6.4%, 8.8% for F and S1: 8.5%, 11.6%; S2: 7.7%, 11%; U1: 5.7%, 11%; U2: 7.0%, 10.2%; Sa1: 5.9%, 8.1%; Sa2: 5.3%, 6.7% for E. Sensitivity. For Incstar F RIA, the least detectable dose (zero 1 3SD) was 38 pg/tube, corresponding to the usual lower limit of F detection (loss-corrected) of 55 pg/mL (0.15 nmol/L) in saliva, 2.5 ng/mL (7 nmol/L) in serum, and 0.7 ng/mL (2 nmol/L) in urine. For E RIA, the least detectable dose was 4.5 pg/tube, corresponding to a pratical E detection limit of 21 pg/mL (0.06 nmol/L) in saliva, 90 pg/mL (0.25 nmol/L) in serum, and 210 pg/mL (0.58 nmol/L) in urine. When the sensitivity of the 125Ilabeled E RIA was compared with the same RIA carried out with [3H]E, it was found to be lower (4.5 6 1.6 vs 10 6 2.3 pg/tube, P ,0.001, n 5 6).

normal results in healthy subjects

Normal values and circadian fluctuations. Usual values (n 5 50) (Table 2, upper part): Whereas in our healthy adult subjects the serum E concentration was found to be around one-tenth the F concentration, its concentrations in saliva and urine were twice those of F. At 0800, there was a positive correlation between F/E and F [F/E 5 0.057F (nmol/L) 1 0.248; r 5 0.610; P 5 0.003] and a negative correlation between F/E and E [F/E 5 20.437E (nmol/L) 1 18.239; r 5 0.688; P 5 0.0001], but no

7.2

182 124

6.9 8

33 36

215 193 146

6.3 4 2.8

11.4 –23.3

265 –568

62 69

569 358

18.4 6 3.6

39.5

282

387 6 98

39.0–127

786–9865

51 39

127 85.4 39 79.7 6 34.0

4389 3925

58.4–184

163 135 125 6 50.3

61 55

4.4–13.9

83 88

16 19

,2

,2 ,2

2.8 4.7

28–75

45.6 6 13

t11h

14–60

51.4 6 16.7

t0

E, nmol/L

9865 1574 786 2959 6 3874

411 372

568–1377

1258 1068

406–1040

138–460

1377 1145 973 6 266

705.6 6 144.1

t11h

344 6 81

t0

a

1 h after ACTH 0.25 mg. b First three letters of patients’ family name. c Under effective treatment with DXM and Nifedipine.

a

Usual values Mean 6 SD t0 (n 5 50); t 1 1 h (n 5 16) Range Primary adrenal insufficiency MAIb FAI Secondary adrenal insufficiency ADE MER Adrenal adenoma TUR ORT Mean 6 SD (n 5 10) Range Cushing disease MON LAN Ectopic ACTH secretion BRE MOR IMO Mean 6 SD (n 5 5) Range Cushing disease under OP’DDD MON Ectopic ACTH under OP’DDD BRE MOR Chronic renal insufficiency Mean 6 SD (n 5 10) Range 11-HSD deficiency GON J-C GON A GON Ac

F, nmol/L

Serum

34 47 52

16.8 –28.3

23.5 6 3.9

9.2 5.2

7.1

16.4–77.9

77.9 18.4 20.2 30.5 6 26.6

6.4 7.6

8.5 8.5 9.5 6 3.1

12 7.6

5.1–16.7

9.8 6 3.5

t0

F/E

17.3 31.8

15.2 12

11.4 6.5

10.2–28.1

16.4 6 5.0

t11h

6.9

8.7

145 34 27 84 6 62

21 16

2.4 1.8

1.1 1.5

49 28

4.4–18

t0

8 8.2 2.4

4.1 –18.5

9.9 6 5.8

3.5–17

9.3 6 4.0

ACTH, pmol/L

7.2 8.6 2.2

12.5 –39.9

25.1 6 10.4

5.5–35

17.9 6 7.6

F, nmol/L

Saliva

1.11 0.95 1.10

0.18 –0.53

0.38 6 0.11

0.24–1.11

0.50 6 0.19

E, nmol/L

F/E

223 177 54

114

181436 3568 1903

915 543

136 –2309

2309 1241 1199 6 801

19 11

19–205

96.3 6 43.6

Table 2. Variations in serum, saliva, and urine concentrations of F and E (and in their ratios).

44 34 8.5

206

2662 490 607

821 552

116 –1705

1705 775 717 6 537

67 35

47–417

130 6 104

F, nmol/24 h

Urine

4.55 5.26 6.25

0.56

68.2 7.14 3.1

1.11 0.98

1.11 –2.02

1.35 1.61 1.55 6 0.39

0.29 0.31

0.12–0.91

0.52 6 0.29

E, nmol/24 h

F/E


1401

1402


significant correlation between E and F (r 5 0.063; P 5 0.0921). Circadian variations (Figs. 1 and 2, Table 3) in the serum E concentrations among our healthy subjects were seen to parallel those classically described for F; however, their individual E concentrations were practically identical at 0800 and noon. Circadian variations were also found in their saliva and urine. The highest E concentrations were observed in these subjects’ serum and saliva at 0800 and the lowest at midnight. In urine this occured at noon and at 0400. The F/E ratio fluctuated significantly over the course of the day (Fig. 2 and Table 3) (ANOVA P ,0.001 for serum and saliva, P 5 0.045 for urine). In serum and saliva, the highest F/E ratio was found at 0800 and in urine at 1600 (P ,0.001). This ratio was significantly lower for urine during the day (0800 –1600) than at night (2000 – 0400) (P 5 0.0042). Dynamic tests (Table 4). Synacthen stimulation tests: While a quite classically significant increase in the serum F concentration was observed in our healthy subjects 1 h after injection of short-acting Synacthen, a significant decrease in E was found. In contrast, serum E was found to be increased 4 h after administration of long-acting Synacthen. However, both these stimulation tests resulted in a significant increase in the F/E ratio (P ,0.001). DXM suppression test: Serum F and E significantly decreased (P ,0.001) in our healthy subjects during the DXM suppression test. However, the decrease in F was greater than that of E, resulting in a nonsignificant decrease in the F/E ratio.

pathological results in patients Adrenal diseases (Table 2). Primary adrenal insufficiency: The two Addisonians had nearly indetectable serum F ('12 nmol/L) and E (,2 nmol/L) that failed to respond to short-acting Synacthen infusion. Secondary adrenal insufficiency: The two hypopituitary patients had depressed serum F and E concentrations that responded to short-acting Synacthen stimulation 1 h later. The responses of these two steroids to 1-24 ACTH were parallel, resulting in a constant F/E ratio. Cushing syndrome: (a) Adrenal adenoma (10 cases). Both serum F and E concentrations were high and failed to increase after ACTH in the two patients who underwent the short-acting Synacthen test. Nonetheless in these two patients, as well as in the eight others who did not undergo the Synacthen test, serum F/E ratios were normal. However, all adrenal adenoma patients were found to have abnormally high urinary F/E ratios (in the 1.1 to 2.0 range). (b) Cushing disease (two cases). The explosive increase in serum F following short-acting Synacthen was accompanied by a smaller increase in serum E, and the increase in their serum F/E ratio was similar to that observed in healthy subjects (2.78 6 0.91). However, their urinary F/E ratios were higher than those found in healthy subjects. (c) Ectopic ACTH secretion (five cases).

Fig. 1. Circadian variations in the mean 6 SEM F and E concentrations in serum, saliva, and urine. Significant variations for each curve (ANOVA, P ,0.001).

1403


both serum and urine (P ,0.001). (d) OP’DDD therapy. The return of serum F to normal concentrations was accompanied by a return to normal E concentrations and to a normal F/E ratio. CRI patients (10 patients) (Table 2). All 10 patients had been on hemodialysis (4 h, 3 times a week) for more than a year and all were hypertensive (mean 6 SD, 141.7 6 10.9/ 86.2 6 9.5 mmHg). Hypertension was not the cause of their nephropathy, but occurred secondarily; all had normal liver function. Their serum F was slightly reduced (386 vs 475 nmol/L for the controls, P 5 0.04), but E was dramatically reduced (18.4 vs 51.4 nmol/L, P 5 0.0001), and the F/E ratio frankly increased compared with controls (23.5 vs 9.8, P 5 0.0001), but less increased than in 11-HSD-2 deficiency (23.5 vs 40.5, P 5 0.03). Saliva F was normal (9.9 vs 9.3 for the controls) and E slightly but not significantly increased (25.1 vs 17.9, P 5 0.054). Nevertheless, the F/E ratio was frankly reduced (0.38 vs 0.50, P 5 0.0075). 11b-OH-Steroid dehydrogenase-2 deficiency (two cases). Two young brothers (A and JC, respectively 21⁄2 and 6 years of age) who had high blood pressure, hypokalemia, undetectable plasma aldosterone (,14 pmol/L), and undetectable plasma renin activity (PRA ,0.05 nmol/L s) [27] presented low normal concentrations of serum F associated with very low concentrations of E and an F/E ratio three to five times the mean normal ratio. Urinary free F was normal and free urinary E depressed below the lower limit of the normal range with, as in plasma, an increased

Fig. 2. Circadian variations in the mean 6 SEM F/E ratio in serum, saliva, and urine. Significant variations for each curve (ANOVA, P ,0.001 for serum and saliva, P 5 0.045 for urine).

Excessively high serum F concentrations were associated in these patients with a much lower excess of serum E, and consequently with a greatly increased F/E ratio in

Table 3. Circadian variations in serum, saliva, and urinary F and E (and their ratios) (mean 6 SD). Serum (n 5 31)

Saliva (n 5 31)

Urine (n 5 20)

Time of day

F, nmol/L

E, nmol/L

F/E

F, nmol/L

E, nmol/L

F/E

F, nmol/L

E, nmol/L

F/E

0400 0800 Noon 1600 2000 Midnight

171 6 90.2 477 6 90.5 339 6 117.5 287 6 108.4 198 6 85.5 152 6 96.6

29.2 6 15.7 51.3 6 16 51 6 17.4 37.2 6 15.2 34.2 6 16.7 24 6 14.3

7.2 6 3.0 10.1 6 3.8 7.0 6 2.6 8.4 6 3.7 6.4 6 2.1 8.2 6 3.2

2.8 6 1.1 9.1 6 4.1 5.0 6 3.0 3.6 6 2.2 2.5 6 1.1 2.8 6 1.4

6.1 6 5.0 18.2 6 7.4 12.7 6 6.6 9.4 6 4.4 5.8 6 5.2 5.0 6 4.1

0.32 6 0.12 0.48 6 0.18 0.33 6 0.11 0.35 6 0.14 0.29 6 0.14 0.40 6 0.16

19.3 6 14.3 80 6 61.5 104 6 93.2 74.2 6 42.2 38.6 6 36.4 37.5 6 26.5

58.6 6 31.4 137 6 59.9 165 6 109.8 118 6 61 71.7 6 29.8 65.4 6 28.6

0.33 6 0.15 0.50 6 0.23 0.46 6 0.36 0.59 6 0.21 0.37 6 0.17 0.44 6 0.25

Table 4. Effects of ACTH stimulation and DXM suppression on serum F and E concentrations (nmol/L) in healthy subjects (mean 6 SD). ACTH stimulation (n 5 16) 0.25 mg E

t0 52.1 6 22.4 t 1 1 h 45.6 6 13 t14h D0 D5

DXM suppression (n 5 5), 3 mg 3 5 days

1 mg

F

F/E

E

319.8 6 152.1 705.6 6 144.1

6.4 6 2.6 16.4 6 5.0

41 6 7.8 62.9 6 12.3

F

278 6 38.6 1625 6 210

F/E

E

F

F/E

7.0 6 1.7 26.4 6 4.9 44.8 6 9.7 495 6 52.5 11.7 6 4.0 2.7 6 0.5 21.8 6 6.5 9.0 6 5.0

D0, Day of test, before DXM administration (baseline value); D5, morning of the 6th day of test, after 5 days of DXM administration.

1404


F/E ratio. The two brothers were successfully treated with DXM alone (1 mg 3 10 days), resulting in transient normalization of aldosterone and PRA and correction of arterial hypertension. JC has since ceased to follow his therapeutic regimen, but A remains under combined DXM (0.5 mg/day) 1 Nifedipine (20 mg 3 2/day) treatment. Since he has been under DXM treatment, A’s serum and urine F and E concentrations have decreased. However, in spite of the clinical effectiveness of his therapy, the F/E ratios in these two media remain unchanged, which accords with the absence of decrease in the ratio of the hydrogenated metabolites of F and E (THF 1 aTHF/THE) previously described [27]. Moreover, A’s PRA and aldosterone have stabilized at low concentrations [upright PRA 0.048 nmol/L (normal range 0.53– 1.79), aldosterone ,28 pmol/L (normal range 196 – 826)] (unpublished data). In addition, A’s saliva F/E ratio before and during DXM treatment, as well as JC’s before DXM treatment, were at the upper limit of normal values.

Discussion

0400, slightly later than in serum and saliva (respectively at 0800 and midnight), reflecting the additional time it takes for urine to be produced and stored. E concentrations in saliva and urine are higher than those of F because of the oxidase activity of 11-HSD-2 (present in parotid gland and kidney) [2, 3]. The F/E fluctuation in serum is slightly attenuated compared with those encountered in saliva and urine, probably because serum reflects the antagonistic oxidation of F to E (by NAD-dependent 11-HSD-2 in kidney) and the reduction of E to F (by NADP-dependent 11-HSD-1 isoenzyme in liver).

cushing syndrome Only our ectopic ACTH patients had increased serum F/E ratios, not those with Cushing disease or the adrenal adenoma patients, as previously found by Walker et al. [6]. In contrast, we found that all Cushing syndrome patients had increased urinary F/E ratios, which corroborates previous results for the THF1aTHF/THE ratio found by Ulick et al. [37] and more recently by Stewart et al. [38].

e ria Existing E assay methods include HPLC [28], double isotope derivative assay [29, 30], protein binding assay [29, 31], and RIA with tritiated E [4, 32–34] following a separation step with paper [29 –31, 33], thin-layer [29, 32], reversed-phase [4, 28], or Sephadex LH 20 [32] chromatography. Our method includes a Celite chromatography purification step followed by a sensitive (12.5 fmol/tube), specific, and simple E RIA involving 125I-labeled E, which, after g emission, emits Auger electrons, well-suited for SPA. The use of SPA [35] has greatly improved RIA in general. This new E RIA 1 SPA assay is carried out in a homogeneous (or pseudohomogenous) phase and does not require time-consuming and laborious methods for separating the bound and free fractions, such as the use of charcoal– dextran or precipitation with a second antibody. The imperfect separation by Celite chromatography of prednisone from E and of prednisolone and methylprednisolone from F requires 24-h abstention from these therapeutic agents in patients who are taking them, to obtain true E and F results (data not shown).

physiological variations in f and e concentrations The usual values of F and E that we found accord well with previously published results for serum [4, 31, 32], saliva [29, 33], and urine [28, 36]. E varies in these three media over the course of the day, with timing similar to those for F and ACTH. One previous study [6] found no significant variation in the serum E concentration (probably because it included only a small number of healthy subjects, n 5 7), but other authors have reported cyclic variations of E in serum [32] and saliva [29], which agrees with our results. In our fractionated urine samples, maximum and minimum E concentrations occur respectively at noon and

role of acth At present, the interaction between ACTH and 11-HSD-2 is still in dispute, some authors arguing in favor of an inhibitory effect of ACTH on 11-HSD-2 activity, others against. As Walker et al. [6] have previously shown, although they did not find any significant variations in plasma E, we observed that circadian fluctuations in E and F in both serum and saliva displayed an increased F/E ratio at 0800, when ACTH is maximum. Katz and Shannon [29] reported that 2 h after administration of 40 IU of ACTH there was an increase in the F/E ratio in parotid fluid, as we found in serum after short- and long-acting Synacthen stimulation. Kornel [39] also demonstrated that ACTH prolongs the half-life of F and shortens the half-life of E and found an increased F/E ratio in urine [40]. All this suggests a possible inhibitory effect of ACTH on 11-HSD-2, as first proposed by Walker et al. [6]. In contrast, Ulick et al. [37] explained the high urinary F/E ratio encountered in Cushing syndromes (that we also observed) as being due to an overload of the capacity of 11-HSD-2 to oxidize F to E in these hypercortisolic patients. Such overload has been demonstrated in vitro by Diederich et al. [41]. Our observation of the absence of increase in the F/E ratio after ACTH in patients with secondary adrenal insufficiency also argues against the inhibitory effect of ACTH on 11-HSD-2 activity, but in hypocortisolic patients. A comprehensive explanation would also take into account the retroinhibition of 11-HSD-2 activity by increased serum E that Rusvai and Fejs-Toth [42] and Stewart et al. [3] have demonstrated in vitro. At this time, no single hypothesis can explain all the observations reported. Some combination of the three phenomena might explain (for example) the data we

1405


observed in the sera of the healthy subjects. Progressive 11-HSD-2 overload could account for the positive correlation between the F/E ratio and the F concentration, whereas retroinhibition of 11-HSD-2 by E might play a role in the negative correlation between F/E and E. Finally, ACTH inhibition of 11-HSD-2 could explain these two simultaneous correlations.

cri patients on hemodialysis Whitworth et al. [4] found plasma creatinine and E to be inversely correlated in renal disease, due to progressive destruction of kidney tissue (and 11-HSD-2). Because some functional 11-HSD-2 activity did remain in our CRI patients, their serum E concentrations were less decreased than in AME. In saliva, the increase in E and the low F/E ratio suggest stimulation of parotid 11-HSD-2 in response to the impaired renal 11-HSD-2 activity encountered in CRI patients. In our study, all the CRI patients were hypertensive and had excessive serum F/E ratios. Vierhapper et al. [43], studying 22 CRI patients not on hemodialysis, found that the ratio of the urinary metabolites (THF/THE) increased from control subjects to normotensive CRI patients to hypertensive CRI patients. This matter deserves further investigation to determine whether 11-HSD-2 deficiency plays a causal role in the onset of hypertension among CRI patients.

ame-1 We used our E assay to study the blood and urine of two young brothers in whom a diagnosis of 11b-HSD-2 deficiency (AME-1) had been made by capillary chromatography assay of the hydrogenated urinary metabolites of F (THF, aTHF) and of E (THE). According to Ulick et al.’s criteria [18], the ratio of the urinary metabolites (r 5 THF 1 aTHF/THE) was very high in both brothers (for JC, r 5 42; for A, r 5 21; normal range 0.58 –2.65) [27]. It is therefore not surprising that their urinary F/E ratios were .4, whereas this ratio is normally ,1 (normal range 0.12– 0.91). The serum F/E ratio in these two patients also was higher than normal. In addition, during effective DXM treatment of A, the unmodified F/E ratios found in both his plasma and urine corroborated our previously published results on the urine THF/THE ratio [27]. In spite of decreased serum and urine F and E concentrations and low PRA and aldosterone concentrations under longterm DXM therapy, the increased F/E ratios we found suggest that the effectiveness of DXM treatment is potentially mediated partly by the formation of a DXM-MR complex [44]. Indeed, the transcriptional efficiency of this complex is 100-fold lower than those of F-MR and aldosterone-MR [45, 46]. The fact that salivary F and E concentrations and the F/E ratio were normal in these two AME-1 patients demonstrates that saliva is not appropriate for carrying out AME-1 diagnosis.

The F–E shuttle has become a subject of investigation in the study of hypertension, both clinically and experimentally. Numerous factors may potentially interfere with the interpretation of results: sample timing; licorice ingestion by the subject [47]; and liver [48], thyroid [49], renal, and adrenal status. These must all be carefully taken into account. The E RIA we describe is simple, selective, sensitive, and well suited for E assay in serum, saliva, and urine for diagnostic purposes, as well as for experimental and therapeutic applications. The easy-to-carry out parallel RIAs of E and F are an alternative or complement to gas chromatography assay of the urinary metabolites of these two steroids. Using these parallel RIAs, we obtained abundant physiological and pathological data concerning the F–E shuttle. We demonstrated that the circadian variation in the F/E ratio is similar to those of F and E themselves, suggesting that the activity of 11-HSD-2 varies over the course of 24 h. We showed that the urinary F/E ratio is increased in Cushing syndrome patients, but that only patients with ectopic ACTH secretion had an increased F/E ratio in serum. We found the serum F/E ratio to be increased in anuric CRI patients under hemodialysis, which we attribute to the destruction of renal 11-HSD-2. In contrast, we found the salivary F/E ratio to be decreased in these patients, which could be due to a compensatory increase in parotid 11-HSD-2 activity. In addition, in the two cases of AME we studied, the serum and urinary F/E ratios were found to be highly increased. These data suggest that parallel serum and urinary RIA of F and E may be a useful indicator in the detection of AME in patients who consult for hypertension associated with hypoaldosteronemia and hyporeninemia. Although the data we present appear to be of interest for the study of AME as well as for the differential diagnosis of Cushing syndrome, the small number of patients included did not enable us to determine F/E ratio cutoff values usable as diagnostic criteria. To do so would require additional study and larger numbers of subjects.

We are indebted to D. Cuvillier, F. Quilin, C. Ouabdelkader, V. Vimont, and to the nurses of the nephrology unit of Hoˆpital Foch (Suresnes, France) for providing samples and immunosera, as well as to C. Courtois for typing the manuscript. We thank Professor Gilbert Schaison for providing us with some of the Cushing syndrome samples.

References 1. Agarwal AK, Mune T, Monder C, White PC. NAD1 dependent isoform of 11b-hydroxysteroid dehydrogenase cloning and characterization of cDNA from sheep kidney. J Biol Chem 1994;269: 25959 – 62.

1406


2. Roland BL, Funder JW. Localization of 11b hydroxysteroid dehydrogenase type 2 in rat tissues: in situ studies. Endocrinology 1996;137:1123– 8. 3. Stewart PM, Whorwood CB, Mason JI. Type 2 11b hydroxysteroid dehydrogenase in foetal and adult life. J Steroid Biochem Mol Biol 1995;55:465–71. 4. Whitworth JA, Stewart PM, Burt D, Atherden SM, Edwards CRW. The kidney is the major site of cortisone production in man. Clin Endocrinol (Oxf) 1989;31:355– 61. 5. Krozowski ZS, Funder JW. Renal mineralocorticoid receptors and hippocampal corticosterone-binding species have identical intrinsic steroid specificity. Proc Natl Acad Sci U S A 1983;80:6056 – 60. 6. Walker BR, Campbell JC, Fraser R, Stewart PM, Edwards CRW. Mineralocorticoid excess and inhibition of 11b-hydroxysteroid dehydrogenase in patients with ectopic ACTH syndrome. Clin Endocrinol (Oxf) 1992;37:483–92. 7. Whorwood GB, Franklyn JA, Sheppard MC, Stewart PM. Tissue localization of 11b-hydroxysteroid dehydrogenase and its relationship to the glucocorticoid receptor. J Steroid Biochem Mol Biol 1992;41:21– 8. 8. Beitins IZ, Bayard F, Ances IG, Kowarski A, Migeon CJ. The metabolic clearance rate, blood production, interconversion and transplacental passage of cortisol and cortisone in pregnancy near term. Pediatr Res 1973;7:509 –19. 9. Nahoul K, Daffos F, Forestier F, Chartier M, Scholler R. Plasma corticosteroid patterns in the fetus. J Steroid Biochem 1988;29: 635– 640. 10. Mune T, Rogerson FM, Nikkila¨ H, Agarwal AK, White PC. Human hypertension caused by mutations in the kidney isoenzyme of 11b hydroxysteroid dehydrogenase. Nature Genet 1995;10:394 –9. 11. Wilson RC, Krozowski ZS, Li K, Obeyesekere VR, Razzaghy-Azar M, Harbison MD, et al. A mutation in the HSD11b2 gene in a family with apparent mineralocorticoid excess. J Clin Endocrinol Metab 1995;80:2263– 6. 12. Wilson RC, Harbison MD, Krozowski ZS, Funder JW, Shackleton CHL, Hanauske-Abel HM, et al. Several homozygous mutations in the gene for 11b-hydroxysteroid dehydrogenase type 2 in patients with apparent mineralocorticoid excess. J Clin Endocrinol Metab 1995;80:3145–50. 13. Obeyesekere VR, Ferrari P, Andrews RK, Wilson RC, New MI. The R337C mutation generates a high 11b-hydroxysteroid dehydrogenase type II enzyme in a family with apparent mineralocorticoid excess. J Clin Endocrinol Metab 1995;80:3381–3. 14. Stewart PM, Krozowski ZS, Gupta A, Milford DV, Howie AJ, Sheppard MC, Whorwood CB. Hypertension in the syndrome of apparent mineralocortocoid excess due to the mutation of the 11b-hydroxysteroid dehydrogenase type 2 gene. Lancet 1996; 347:88 –91. 15. Milford DV, Shackleton CHL, Stewart PM. Mineralocorticoid hypertension and congenital deficiency of 11b-hydroxysteroid dehydrogenase in a family with the syndrome of apparent mineralocorticoid excess. Clin Endocrinol 1995;43:241– 6. 16. New MI. Apparent mineralocorticoid excess: a personal history. Steroids 1994;59:66 – 8. 17. Shackleton CHL, Rodriguez J, Arteaga E, Lopez JM, Winter JSD. Congenital 11b-hydroxysteroid dehydrogenase deficiency associated with juvenile hypertension: corticosteroid metabolite profile of four patients and their families. Clin Endocrinol (Oxf) 1985;22: 701–12. 18. Ulick S, Levine LS, Gunczler P, Zanconato G, Ramirez LC, Rauh W, et al. A syndrome of apparent mineralocorticoid excess associated with defects in the peripheral metabolism of cortisol. J Clin Endocrinol Metab 1979;49:757– 64. 19. Kitanaka S, Tanae A, Hibi I. Apparent mineralocorticoid excess

20.

21.

22.

23.

24.

25. 26. 27.

28.

29. 30.

31.

32.

33.

34.

35.

36. 37.

38.

due to 11b-hydroxysteroid dehydrogenase deficiency: a possible cause of intrauterine growth retardation. Clin Endocrinol (Oxf) 1996;44:353–9. Ulick S, Ramirez LC, New MI. An abnormality in steroid reductive metabolism in a hypertensive syndrome. J Clin Endocrinol Metab 1977;44:799 – 802. Erlanger BF, Boreck F, Beisser SM, Liebermann S. Steroid protein conjugates: preparation and characterization of conjugates of bovine serum albumin with progesterone, deoxycorticosterone and estrone. J Biol Chem 1957;228:713–26. Vaitukaitis J, Robbins JB, Nieschlag E, Rossi G. A method for producing specific antisera with small doses of immunogen. J Clin Endocrinol Metab 1971;33:988 –91. Fiet J, Villette JM, Galons H, Boudou P, Burthier JM, Hardy N, et al. The application of a new highly sensitive radioimmunoassay for plasma 21 deoxycortisol to the detection of steroid-21-hydroxylase deficiency. Ann Clin Biochem 1994;31:56 – 64. Fiet J, Gourmel B, Villette JM, Brerault JL, Julien R, Cathelineau G, Dreux C. Simultaneous radioimmunoassay of androstenedione, dehydroepiandrosterone and 11b-hydroxyandrostenedione in plasma. Horm Res 1980;13:133– 49. Bosworth N, Towers P. Scintillation proximity assay. Nature 1989; 341:167– 8. Abraham GE. Radioimmunoassay of steroids in biological materials. Acta Endocrinol Suppl (Copenh) 1974;183:1– 42. Gourmelen M, Saint-Jacques I, Morineau G, Soliman H, Julien R, Fiet J. 11b-hydroxysteroid dehydrogenase deficit: a rare cause of arterial hypertension, diagnosis and therapeutic approach in two young brothers. Eur J Endocrinol 1996:135:238 – 44. Scho¨nesho¨fer M, Kage A, Weber B, Lenz I, Ko¨ttgen E. Determination of urinary free cortisol by “on-line” liquid chromatography. Clin Chem 1985;31:564 – 8. Katz FH, Shannon IL. Parotid fluid cortisol and cortisone. J Clin Invest 1969;48:848 –55. Hillman DA, Giroud CJP. Plasma cortisone and cortisol levels at birth and during the neonatal period. J Clin Endocrinol 1965;25: 243– 8. Few JD, Cashmore GC. Plasma cortisone in man: its determination, physiological variation and significance. Ann Clin Biochem 1980;17:227–32. Srivastava LS, Werk EE, Trasher K, Sholiton LS, Kozera R, Nolten W, Knowles HC. Plasma cortisone concentration as measured by radioimmunoassay. J Clin Endocrinol Metab 1973;36:937– 43. Meulenberg PMM, Ross HA, Swinkels LMJW, Benraad TJ. The effect of oral contraceptives on plasma free and salivary cortisol and cortisone. Clin Chim Acta 1987;165:379 – 85. Sippell WG, Bidlingmaier F, Becker H, Bru¨nig T, Do¨rr H, Hahn H, et al. Simultaneous radioimmunoassay of plasma aldosterone, corticosterone, 11-deoxycorticosterone, progesterone, 17-hydroxyprogesterone, 11-deoxycortisol, cortisol and cortisone. J Steroid Biochem 1978;9:63–74. Fiet J, Gosling JP, Soliman H, Galons H, Boudou Ph, Aubin Ph, et al. Hirsutism and acne in women: coordinated radioimmunoassay for eight relevant plasma steroids. Clin Chem 1994;40:2296 – 305. Minick MC. Cortisol and cortisone excretion from infancy to adult life. Metabolism 1966;15:359 – 63. Ulick S, Wang JZ, Blumenfeld JD, Dickering TG. Cortisol inactivation overload: a mechanism of mineralocorticoid hypertension in the ectopic adrenocorticotropin syndrome. J Clin Endocrinol Metab 1992;74:963–7. Stewart PM, Walker BR, Holder G, O’Halloran D, Shackleton CHL. 11b-hydroxysteroid deshydrogenase activity in Cushing’s syndrome: explaining the mineralocorticoid excess state of the


39.

40.

41.

42.

43.

44.

ectopic adrenocorticotropin syndrome. J Clin Endocrinol Metab 1995:80:3617–20. Kornel L. Corticosteroids in human blood V: extra adrenal effect of ACTH upon metabolism of cortisol. Steroidologia 1970;1:225– 45. Kornel L, Hill RS. Paper chromatographic pattern of endogeneous urinary corticosteroids in normal subjects. Metabolism 1961;10: 18 –26. Diederich S, Quinkler M, Miller K, Heilmann P, Scho¨nesho¨fer M, Oelkers W. Human kidney 11b-hydroxysteroid dehydrogenase: regulation by adrenocorticotropin? Eur J Endocrinol 1996;134: 301–7. Rusvai E, Fejs-Toth AN. A new isoform of 11b-hydroxysteroid deshydrogenase in aldosterone target cells. J Biol Chem 1993; 268:10717–20. Vierhapper H, Derfler K, Nowotny P, Hollenstein U, Waldhausl W. Impaired conversion of cortisol to cortisone in chronic renal insufficiency—a cause of hypertension or an epiphenomenon? Acta Endocrinol (Copenh) 1991;125:160 – 4. Arriza JL, Weinberger C, Cerelli G, Glaser TM, Handelin BL, Housman DE, Evans RM. Cloning of human mineralocorticoid

45.

46.

47.

48.

49.

1407

receptor complementary DNA: structural and functional kinship with the glucocorticoid receptor. Science 1987;237:268 –75. Lombes M, Kenouch S, Souque A, Farman N, Rafestin-Oblin ME. The mineralocorticoid receptor discriminates aldosterone from glucocorticoids independently of the 11b-hydroxysteroid deshydrogenase. Endocrinology 1994;135:834 – 40. Arriza JL, Simerly RB, Swanson LW, Evans RM. The neuronal mineralocorticoid receptor as a mediator of glucocorticoid response. Neuron 1988;1:887–900. Stewart PM, Burra P, Shackleton CHL, Sheppard MC, Elias E. 11b-hydroxysteroid deshydrogenase deficiency and glucocorticoid status in patients with alcoholic and non-alcoholic chronic liver disease. J Clin Endocrinol Metab 1993:76:748 –51. Zumoff B, Bradlow HL, Levin J, Fukushima DK. Influence of thyroid function on the in vivo cortisol 7 cortisone equilibrium in man. J Steroid Biochem 1983;18:437– 40. Stewart PM, Valentino R, Wallace AM, Burt D, Shackleton CHL, Edwards CRW. Mineralocorticoid activity of liquorice: 11-betahydroxysteroid dehydrogenase deficiency comes of age. Lancet 1987;ii:821– 4.