The radioimmunoassay of aldosterone in serum and urine: Theoretical and practical aspects

Abstract

Simple methods for the radioimmunoassay of aldosterone in serum and urine are described. The methods rely on an aldosterone antiserum raised in rabbits using aldosterone-3-oxime coupled to bovine serum albumin as antigen. Aldosterone is extracted from serum using a simple single stage solvent extraction procedure followed by a single paper chromatographic step. The isolated hormone is eluted from the chromatogram with buffer and subsequently radioimmunoassayed by traditional methods. Free hormone is sequestrated from the incubation mixture by direct extraction into toluene scintillator which is subsequently placed in liquid scintillation counting vials for radioassay. This approach yields blank values below the assay detection limit, that is, below roughly 1 pg/incubation tube. The technique will be used to illustrate the theoretical principles relating to optimal assay design. Our present procedure is distinguished from that adopted in many radioimmunoassays by a separation step which entails the counting of the “free” labelled hormone. Optimal reaction mixtures yielding maximum precision differ in composition depending on whether: (a) free, (b) bound, (c) both free and bound fractions are counted. Assuming that no experimental errors are made in the estimate of the response metameter, and that only counting errors govern the precision of measurement of the “dose”, it may be shown that, in case (a), optimal sensitivity is achieved if concentrations of labelled hormone and of antibody given by 9/K and 8/K are selected (where K represents the equilibrium constant of the predominant antibody binding site in the system). In this case, 33 per cent of the label will be free at the zero point on the response curve. Conversely in case (b), optimal concentrations are 2.25/K and 1.25/K respectively, and 33 per cent of the tracer is bound. If both fractions are counted, optimal concentrations are 4/K and 3/K, and 50 per cent of the total activity is bound at the zero point. In practice, experimental errors in addition to counting errors are always incurred and the resulting optimisation equations are not normally susceptible to analytic solution. Computer techniques are described enabling a choice of optimal assay mixtures for maximal precision of measurement of any desired hormone concentration to be made and their application to the aldosterone method illustrated. The theoretical aspects of cross-reaction between competing steroids are discussed, and experimental observations in the aldosterone assay system described. In general, lack of specificity becomes more evident as the fraction of labelled hormone bound to antibody increases. Though this finding conflicts with the requirements in the present assay system for the free fraction to be small (for highest assay precision), experimental data indicates that our procedures exclude steroidal non-specific effects, and that assay estimates in blood and urine are accurate. Our results obtained in a few representative physiological studies will be briefly discussed.