We describe the develop- ment of a ... RIA procedure for 18- hydroxycortisol suitable for direct assay of urine ..... the development of the RIA was straightfor- ward ...
CLIN.
CHEM.
31/6,
849-852
(1985)
A Radioimmunoassay
for 18-Hydroxycortisol
John E. T. Corrle,1 Christopher
R. W. Edwards,2
and Peter S. Budd1
Increased excretion of 18-hydroxycortisol has been proposed as a specific biochemical marker for differential diagnosis of primary aldosteronism. We describe the development of a direct AlA with an 1251 label that permits measurement of the steroid in 0.5 L of urine or 25 pt of plasma. For control subjects, the mean concentrations of 1 8-hydroxycortisol in urine and plasma were 310 (SD 178) nmol/24 h (n = 32) and 2.27 (SD 0.80) nmol/L (n = 37), respectively; patients
with Conn’s
adenoma
or glucocorticoid-remediable
aldosteronism had values for urine in the range 1084 to 6534 nmol/24 h and concentrations in plasma ranging from 6.49 to 31.20 nmol/L. Patients with idiopathic zona glomerulosa hyperplasia had values for urine and plasma ranging from 353 to 734 nmol/24 h and from 0.26 to 6.60 nmol/L, respectively. Concentrations of 18-hydroxycortisol in urine clearly discriminate patients with idiopathic hyperplasia from those with other forms of primary aldosteronism, but further work is required to assess the diagnostic accuracy of determinations in plasma. AddItIonal Keyphrases: reference interval differentiating idiopathic zona glomerulosa hyperplasia from other forms of primary akiosteronism liquid chromatography compared steroids
Measurement of urinary 18-hydroxycortisol (I, Figure 1) may be definitive for the differential diagnosis of primary aldosteronism, because patients with a proven Conn’s adonoma or glucocorticoid-remediable aldosteronism reportedly excrete more of the steroid than do normal subjects or patients with idiopathic zona glomerulosa hyperplasia (1,2). The ability to make such a distinction would be useful because current diagnostic techniques are not entirely specific (3) and, although evaluation of 18-hydroxycorticosterone in plasma has been proposed as an improved discriminator (4), its measurement, usually involving chemical modification before REA (5), is technically demanding.
The original
measurements of 18-hydroxycortisol in urine chromatography-mass spectrometry, but such equipment is not widely available. In our attempts to establish an assay for studying the physiology and pathophysiology of this new steroid and assessing its diagnostic utility, we first tried a “high-pressure” liquid chromatographic (HPLC) method. However, its low sample throughput, the requirement for solvent extraction of samples, and its susceptibility to analytical interference from drugs made it unacceptable for routine use. Subsequently we have established a sensitive and specific RIA procedure for 18hydroxycortisol suitable for direct assay of urine and plasma samples. were
by gas
Materials
and Methods
Materials 18-Hydroxycortisol, 18-hydroxycortisone, and 18-oxocortiaol were generous gifts from Professor C. Gomez-Sanchez, VA Medical Center, University of South Florida, Tampa, FL. Other steroids for cross-reaction studies were from Sigma Ltd., London, U.K. Na’I (cat. no. IMS.30) was from Amershain Jnternational, Amersham, Bucks., U.K. Donkey anti-rabbit antiserum and normal rabbit serum were from the Scottish Antibody Production Unit, Carluke, Lanarkshire, U.K. New Zealand White rabbits were from Ranch Rabbits Ltd., Croydon, Surrey, U.K. Assay diluent, pH 4, was prepared by dissolving 10.5 g of citric acid monohydrate, 8.4 g of anhydrous Na2HPO4, and 1 g of gelatin in 1 L of distilled water. Polystyrene assay tubes (cat. no.123) were from Luckham Ltd., Burgess Hill, U.K. 18-Hydroxycortisol standards were prepared in assay diluent in doubling concentrations over the range 0.025813.20 nmol/L (1.95-1000 pg per assay tube) and stored in aliquots (one set per assay) at -20 #{176}C.
Development
of the Radioimmunoassay
of 3-O-(carboxymethyl)oximino-18-hydroxycor7.24 mg of 18-hydroxycortisol and 5.15 mg of carboxymethoxylamine hemihydrochioride in a halfsaturated methanolic solution of sodium hydrogen carbonate (1.07 mL) and left this at room temperature for 25 mm (6). After evaporating the solvent under reduced pressure, we dissolved the residue in 3.6 mL of water and acidified this to pH 4 with 0.1 mol/L hydrochloric acid. We extracted the aqueous mixture with four 3.6-mL portions of ethyl acetate and washed the combined extracts once with satuSynthesis
0H
tisol.
(I)
in Plasma and Urine
Ill)
R=OH
(ill)
B
(IV) R
-
NH-Bovine
serum
albumin
NH
Fig. 1. Structure of 18-hydroxycortisol (I), its 3-carboxymethyloxime derivative (II), and conjugates with bovine serum albumin (Ill) and radlolodinated histamine (IV) 1 MRC Clinical and Population Cytogenetics Unit, Western Genera! Hospital, Edinburgh, EH4 2XU, U.K. 2Department of Medicine, Western General Hospital, Edinburgh EH4 2XU, U.K. Received January 22, 1985; accepted March 13, 1985.
We dissolved
rated NaC1 solution, dehydrated them over anhydrous sodium sullhte, and evaporated the solvent under reduced pressure. We dissolved the residual solid in 5 mL of methanol and diluted an aliquot of this 100-fold for quantification by ultraviolet spectroscopy at 251 nm. On the basis of e = 23100 for 3.O-(carboxymethyl)oximinocortisol, the calculated yield of the oxune was 5.50 mg (63.8%). Analysis by thinlayer chromatography on silica gel with a solvent of benzene/ethanol/acetic acid, 75/25/0.5 by vol, showed only a single component, which was more polar than 18-hydroxycortisol or 3-O-(carboxymethyl)oximmnocortisol. The NMR spectrum (methanol-d4) showed 5 6.36 and 5.64 (total 1H, CLINICAL
CHEMISTRY,
Vol. 31, No. 6, 1985
849
both a, H-4; Z.E 2:3), 4.53 and 4.49 (total 2H, both a, =NOCH2-), 4.32 and 3.74 (211, d/d, H-18, J = 10Hz), 4.34 (1H, s, H-il), 3.73 and 3.58 (2H, did, H-21, J = 12Hz), 1.32 (3H, s,H-19). Because there was no evidence of conf.aniination by starting material or 3,20-dioxime, we used this material after evaporation of the methanol without further purification. Synthesis of immunogen. We dissolved 5.5 mg of the oxime in 0.1 mL of dimethylformamide and coupled it to 20 mg of bovine serum albumin by the modified Erlanger procedure previously described (7). Analysis of the conjugate showed a coupling ratio of 13.5 18-hydroxycortisol residues per protein molecule. The steroid-protein conjugate was lyophilized and stored at -20 #{176}C. Antisera to the conjugate were raised in New Zealand White rabbits. Each animal was injected by the multiple intradermal technique (8) with 100 ig of immunogen in 1 mL of an equivolume emulsion of saline and Freund’s complete adjuvant. Subsequent booster injections were made every six weeks by combined intramuscular and subcutaneous injection of the same dose in saline/Freund’s incomplete adjuvant. The animals were bled 10 days after each booster injection. lodinated radiolzgand was prepared by activating the oxime in dimethylformamide solution and subsequent coupling to radioiodinated histamine as previously described (9). The radioligand was recovered by extracting the reaction mixture three times with 0.25-mL portions of ethyl acetate and the combined extract was chromatographed on a plastic-backed silica gel plate (20 x 20 cm x 0.1 mm) in toluene/ethanollacetic acid (70/30/1, by vol). The major radioactive band, R 0.18-0.24, was located by autoradiography (Kodak X-omat film), cut from the plate, and eluted with 6 mL of ethanol containing 60 mg of triethylamine. The recovered tracer solution was stored at -20 #{176}C. After determining what dilution of freshly prepared tracer in assay buffer would yield 2 x i0 counts/mm per milliliter, we used the same dilution factor throughout the life of the tracer. Assay procedure. Dispense into assay tubes 200 iL of standard solutions, urine specimens diluted 400-fold in assay diluent, and plasma specimens diluted eightfold in assay diluent. (For all dilutions and dispensings we used Hamilton Microlab M apparatus.) To each tube add 50 L of radioligand (10 000 counts/mm with freshly prepared tracer) and 50 L of 7500-fold diluted antiserum. Mix, then incubate at room temperature for 4 h. Next add 50 uL of 10fold diluted donkey anti-rabbit serum and 50 iL of 150-fold diluted normal rabbit serum containing 1 mmol of EDTA per liter, incubate at 4#{176}C overnight, then dilute with 1.5 mL of assay buffer and centrifuge (4#{176}C, 2000 x g, 0.5 h). Decant the supernates and count the radioactivity of the remaining (antibody-bound)
fraction.
(We
used
a Wallac
CLINICAL
CHEMISTRY,
Vol. 31, No. 6, 1985
of initial
dilutions among samples, as picograms of 18-hydroxycortisol
are expressed tube.
all results per assay
Results Assay
Development
Antisera. Every animal immunized showed at least a weak antibody response, and three of the six had antibody titers exceeding 1:3500 after the first booster injection. Antiserum R 22C/3 had the highest titer after a second booster, and was used in further study. Figure 2 illustrates a typical standard curve. Ratholgand. The iodinated radioligand showed good stability on storage as judged by binding to antibody in the absence of added 18-hydroxycortisol: 48.1% (SD 2.2%) over 150 days (2.5 half-lives for ‘9). The nonspecific binding was
