A. Lambert

THE ROLE OF MODERN BIOASSAYS IN CLINICAL ENDOCRINOLOGY

Bioassays measure a functional effect and relate to a hormone’s activity. Immunoassays measure antigenic determinant(s) and relate to a hormone’s concentration. However, whatever assay is used the information gained must be clinically or physiologically relevant and this is the theme of this review. It will not attempt to give a comprehensive overview of the history and practice of bioassays. The interested reader is directed to excellent general reviews on bioassays in pharmacology and endocrinology (Daly, 1978; Bangham, 1983a), to the problems of biological standardization in horrnonc assays (Bangham, 1979, 1983b), and to descriptions of the biometric principles and validation of assays (Gaddum, 1953; Ekins, 1976; Finney, 1978; see also British Pharmacopoeia, 1980, the statistical appendix, BP 11, Appendix XIV). We are restricting our remarks to the polypeptide and protein hormones (and molecules which mimic or modulate their actions) and our purpose is to bring to the attention of endocrinologists the continuing and expanding relevance of bioassays to their discipline. ‘No one would do bioassays if he could help it’. This quote from Gaddum (1953) reflects in many ways the feelings of endocrinologists with traditional, mainly in vivo, bioassay techniques. Daly (1978) wrote ‘the idea of a bioassay involving, for instance, the meticulous dissection of the uteri from dozens of pre-adolescent mice, may seen merely quaint’. Despite this, classicial bioassays continue to be fundamentally important in pharmacology and endocrinology. Such bioassays define the biological function of most hormones, and are still the chosen technique in estimating the potency of hormone preparations under consideration as candidates for use as International Reference Preparations (WHO Expert Committee on Biological Standardization, 1982). However, to the clinical endocrinologist bioassays may seem an irrelevance and with the success and universal application of immunoassays, largely because of their sensitivity, precision,

Cortisol production by dispersed guinea-pig adrenal cells; a specific, sensitive and reproducible response to acth and it's fragments

Naturally occurring steroids and peptide hormones, tested at supraphysiological concentrations, were without effect on basal and human (h) 1-39 ACTH (NIBSC code 74/555, 25 ng/l (5.5 X 10(-12) mol/l] stimulated cortisol production. Further, low concentrations of angiotensin II, N-pro-opiocortin (N terminal fragment 1-76) and gamma-MSH all of which have been reported to synergise with ACTH with regard to cortisol production, were without significant effect alone or in combination with ACTH over the range 2.2 X 10(-13) to 5.5 X 10(-12) mol/l. The activity of h 1-39 was compared with that of the ACTH related peptides 1-24, 1-18, 1-17, 1-16, 1-13-NH2 (alpha MSH), 1-10 and 4-10. The dose responses were parallel and the same maximal cortisol output was observed with all the peptides except the 1-10 fragment. Half maximal stimulation occurred at 3.1 X 10(-12) (1-24), 4.4 X 10(-12) (h 1-39), 1.5 X 10(-11) (1-39), 3.3 X 10(-10) (1-18), 5 X 10(-9) (1-13-NH2), 8 X 10(-9) (1-17), 2 X 10(-7) (1-16) and 1 X 10(-5) (4-10) mol/l respectively. Interference by the above ACTH-derived peptides in cortisol secretion by the cells in response to 5.5 X 10(-12) mol/l h 1-39 ACTH was minimal over the range 5.2 X 10(-12)-2.2 X 10(-6) mol/l. The sensitivity of the adrenal cells to h 1-39 ACTH was such that 2 ng/l (4.4 X 10(-13) mol/l) provoked cortisol secretion over the control (P less than 0.05, n = 17). The coefficient of variation within assay for each dose on the full standard curve (2.2 X 10(-13)-1.1 X 10(-10) mol/l) was less than 10% (n = 6). Half maximal stimulation was given by 14.5 ng/l (3.2 X 10(-12) mol/l). Between control and 1.1 X 10(-10) mol/l ACTH there was a 32 +/- 8 (mean +/- SD, n = 9) fold change in cortisol production.

On the assessment of the in vitro biopotency and site(s) of action of drugs affecting adrenal steroidogenesis

Dispersed guinea-pig adrenal cells have been employed in the in vitro estimation of the biological potency and sites of action of drugs acting against the adrenal. The effect of 12 drugs on cortisol secretion from cells stimulated with adrenocorticotrophin (ACTH, 50 ng/L, a 95% saturating dose) has been tested. All the drugs depressed cortisol output in a dose-related fashion. The concentration of drug which inhibited secretion by 50% was (μmol/L, mean±SEM): etomidate 0·1±0·002; epostane 0·44±0·02: 17-ketotrilostane 0·55±0·04: trilostane 1·3±0·1: metyrapone 3·5±0·6: cyproterone acetate 4·6±0·2: megestrol acetate 11±2: danazol 22±2: aminoglutethimide 41±5: stanozolol 50±4: thiopentone 160±18: propofol 170±18. The sites of the anti-steroidogenic effect of seven of these drugs have also been established using a method based upon the sequential stimulation by the exogenous precursor steroids of the various steps leading to the biosynthesis of cortisol by adrenal cells. Propofol acts between ACTH binding and pregnenolone production, trilostane, megestrol acetate and cyproterone acetate are 3β-hydroxysteroid dehydrogenase inhibitors whereas metyrapone, etomidate and thiopentone act at 11β-hydroxylase.

ON THE SITE OF ACTION OF THE ANTI‐ADRENAL STEROIDOGENIC EFFECT OF ETOMIDATE AND MEGESTROL ACETATE

The sites of action of the anti‐steroidogenic action of etomidate and megestrol acetate have been established with a novel in vitro approach based upon the inhibition of cortisol (Co) secretion by dispersed guinea‐pig adrenal cells. The cells were challenged with the Co precursor steroids (all at 10−5 mol/l) pregnenolone (Pe), 17‐hydroxy‐pregnenolone (17‐Pe), progesterone (Po), 17‐hydroxyprogesterone (17‐Po), 21‐deoxycortisol (21‐DOC) and 11‐deoxycortisol (11‐DOC), or 1‐24 adrenocorticotrophin (ACTH, 100 ng/l) in the presence or absence of either etomidate, megestrol acetate, metyrapone or trilostane (all at 5 ± 10−5 mol/l). In the absence of drug, the steroid precursors or ACTH provoked a cortisol secretion of > 14 times that secreted by cells incubated in their absence. ACTH‐stimulated Co secretion was inhibited by > 85% by all the drugs employed. In the presence of trilostane and megestrol acetate, Co secretion provoked by the Δ4 3‐keto steroids (Po, 17‐Po, 21‐DOC and 11‐DOC) was similar to the controls. However, with the Δ5, 3β‐hydroxy steroids, 17‐Pe and Pe, Co secretion was inhibited by > 57% in the presence of these drugs. In contrast, etomidate and metyrapone inhibited Co secretion by >60% when 11‐deoxycortisol was employed, indicative of a block at 11 β‐hydroxylase, the final step in the cortisol biosynthetic pathway. Similar results were seen with Pe, 17‐Pe, Po and 17‐Po, all of which are converted to cortisol via a biosynthetic route which includes catalysis by 11 β‐hydroxylase. In contrast the transformation of 21‐deoxycortisol to Co, a step which proceeds directly via 21‐hydroxylase, was unaffected by the presence of any of the drugs. The site of action for etomidate was confirmed by a second method. Adrenal cells were incubated for 90 min with ACTH (1–24, 100 ng/l) and 14C‐pregnenolone in the presence and absence of either etomidate or trilostane. 11‐DOC and Co were extracted from the incubates and assayed by thin‐layer chromotography. The ratio of 11‐DOC to Co was 0.20, 0.27 and 1.64 for the control, trilostane and etomidate incubations respectively. We conclude that trilostane and megestrol acetate are 3a‐hydroxysteroid dehydrogenase, A5‐A4 isomerase inhibitors, whereas metyrapone and etomidate act at the level of 1 la‐hydroxylase.

A simple in vitro approach to the estimation of the biopotency of drugs affecting adrenal steroidogenesis

Dispersed guinea-pig adrenal cells can be maximally stimulated to secrete cortisol by adrenocorticotrophin (ACTH greater than 50 ng/l). Further, this stimulation appears to be specific to ACTH alone, with other naturally occurring chemicals (e.g. steroids, protein hormones) at supra-physiological concentrations being without effect on cortisol production. The effect of drugs of differing structure and therapeutic function (aminogluthethimide, metyrapone, trilostane, 17-ketotrilostane, danazol, epostane, megestrol acetate, stanozolol and etomidate) on ACTH-stimulated (50 ng/l) cortisol production has been tested in this system. All the drugs depressed steroid output in a similar dose-related fashion. The concentration of drug which inhibited cortisol output by 50% was (mumol/l, mean +/- SEM): etomidate 0.097 +/- 0.002: epostane 0.44 +/- 0.02: 17-ketotrilostane 0.55 +/- 0.04: trilostane 1.3 +/- 0.1: metyrapone 3.5 +/- 0.6: megestrol acetate, 11 +/- 2: danazol 22 +/- 2: aminogluthethimide 41 +/- 5: stanozolol 50 +/- 4. Thus, etomidate, an anaesthetic, is more potent than the established anti-steroidogenic drugs metyrapone, aminogluthethimide and trilostane. Further, direct anti-steroidogenic effects have been demonstrated for megestrol acetate and stanozolol for the first time. We conclude that this technique offers a promising new approach to the assessment of biological potency of drugs affecting endocrine tissues.

Preliminary experiences with a bioassay for adrenocorticotrophin (ACTH) in unextracted human plasma using dispersed guinea-pig adrenal cells.

A simple, accurate and precise dispersed cell bioassay suitable for the assay of ACTH levels >50 ng/1 in unextracted human plasma is described. Isolated adrenal cells were prepared by tryptic digestion of the guinea‐pig adrenal gland and cortisol production by these cells was controlled specifically by ACTH. The cortisol response to 2 ng/1 of human pituitary ACTH (1–39) was always significantly (P < 0·05, n= 17) different from the control. ACTH concentrations between 1 and 20 ng/1 gave a 14 ± 3 fold (n= 17) increment in cortisol response. Dilutions of test plasmas were parallel with standard ACTH (1–39) when the plasma concentrations were maintained throughout at 4% (1 in 25 dilution) by the addition of dexamethasone‐suppressed plasma (‘ACTH free’). Using a 25 fold plasma dilution, the limit of detection of ACTH in unextracted plasma was 50 ng/1. Recovery of ACTH (1–39) spiked into dexamethasone suppressed plasma at concentrations of 250 and 100 ng/1 was 97·10% (n=17) and 114 ± 15% (n= 17) respectively. The within assay coefficient of variation (CV) of both quality control and patients samples never exceeded 9% while between assay variation was 13%. The sample throughput was 30 plasma samples/week/technician.

On the site of action of the anti-adrenal steroidogenic effect of cyproterone acetate

Cyproterone acetate (CA) inhibited 1-24 ACTH (50 ng/l)-stimulated cortisol production by dispersed guinea-pig adrenal cells in a dose-related manner. Inhibition occurred over the range 10(-6) to 5 X 10(-5) moles/l. The concentration of drug which induced 50% inhibition was 4.6 X 10(-6) moles/l. The sites of action of this anti-steroidogenic effect have been established. Dispersed adrenal cells were challenged with the cortisol precursor steroids (all at 10(-5) moles/l) pregnenolone (Pe), 17 alpha hydroxypregnenolone (17-OH Pe), 17 alpha-hydroxyprogesterone (17-Po), and 11 deoxycortisol or 1-24 ACTH (100 ng/1) in the absence and presence of increasing concentrations of CA (10(-5) to 10(-4) moles/l). In the absence of drug, the steroid precursors or ACTH provoked cortisol secretion greater than 10-fold that secreted by cells incubated in their absence. ACTH-stimulated cortisol secretion was inhibited greater than 68% at concentrations of CA greater than 10(-5) moles/l. CA (10(-5) moles/l) had no significant effect on steroid-stimulated cortisol production when the delta 4, 3 ketosteroids 17-OH Po and 11-deoxycortisol were used, but depressed secretion by greater than 61% (P less than 0.001) when the delta 5 3 beta hydroxysteroids (Pe, 17-OH Pe) were employed. Increasing CA concentrations to 10(-4) moles/l had little effect on cortisol secretion provoked by 11-deoxycortisol, but significantly (P less than 0.05) depressed cortisol secretion stimulated by 17-OH Po. These results suggest that the major site of action of CA is delta 5, 3 beta hydroxysteroid dehydrogenase (3 beta-HSD) with a secondary effect on 21-hydroxylase activity. To confirm these findings cortisol secretion provoked by 17, 21, dihydroxypregnenolone (17, 21 diOH Pe) and 21 deoxycortisol (21-DOC) was measured in the absence and presence of increasing concentrations of CA. At the lowest concentration of CA (5 X 10(-6) moles/l), cortisol secretion provoked by 17, 21 diOH Pe was inhibited by 28% (P less than 0.01) whereas secretion provoked by 21-DOC was not significantly affected. At the highest concentration of CA (10(-4) moles/l), the relative inhibition was 80% for 17, 21 diOH Pe and 38% for 21-DOC. We conclude that cyproterone acetate inhibits adrenal steroidogenesis at both 3 beta-HSD and 21-hydroxylase, the degree of inhibition being more pronounced at 3 beta-HSD.

The effect of some antiprostatic steroids upon cortisol production by guinea-pig adrenal cells stimulated by ACTH

The antiprostatic steroids 6-methylene-4-pregnene-3,20-dione (6-MP) (I), 17-alpha-acetoxy-6, 16-dimethylene-4-pregnene-3,20-dione (II), and melengestrol acetate (MGA) (III) were incubated with guinea-pig adrenal cells, both alone and maximally stimulated with ACTH. Cortisol output was then measured by RIA. Increased cortisol-like secretion was obtained with 6-MP in the absence of ACTH. In the presence of ACTH, cortisol-like steroid secretion was the sum of that seen with ACTH and 6-MP alone. It follows that 6-MP stimulates in vitro a cortisol-like steroid cross reacting with the cortisol antibody by a mechanism that by-passes ACTH. Steroid (II) weakly inhibited cortisol output. MGA, in contrast, proved to be a strong inhibitor of cortisol output (ID50 of 2.3 mumol/l). Its site of action was established by adding it to adrenal cells incubated with precursor steroids on the cortisol pathway. Conversion of 3 beta-hydroxysteroids to cortisol was inhibited whereas conversion of 3-keto steroids was not affected. It follows that MGA inhibits 3 beta-hydroxysteroid dehydrogenase.

ACTH Adrenal Cell Bioassay: Improved Sensitivity (12 ng/L) Achieved by Immunoextraction of Acth from Human Plasma by a Monoclonal Antibody

We have previously reported a bioassay for human plasma ACTH based upon trypsin dispersed guinea-pig adrenal cells which was sensitive to 100 ng/L ACTH in unextracted human plasma when measured against human pituitary ACTH (1–39) standard 74/555. We now present a bioassay of increased sensitivity (12 ng/L) which incorporates three major changes. The trypsin/trypsin inhibitor step in the cell dispersion protocol has been replaced with collagenase, donor calf serum (3%) has been incorporated into the standard curve and ACTH has been extracted from human plasma and dilutions of standard hormone by a sephacryl bound monoclonal antibody (2A3) directed towards the 25–39 sequence. The extracted standard curve has a detection limit of 6 ng/L and the cells can tolerate up to 50% plasma equivalent concentration. Thus, the improved assay has a detection limit of 12 ng/L ACTH in plasma. The assay can now measure bioactive plasma ACTH levels reliably in the normal range.