Fig. 1 General model of the receptor-induction mechanism of steroid hormone action. Modified from Feldman et al. [2].

Steroid receptors

The mode of steroid penetration into target cells has received little attention. In one system, however, uptake of estrogen by uterine cells appears to involve facilitated rather than simple diffusion [6].

The target cells responsive to a given steroid contain high affinity, saturable cytoplasmic binding proteins (mol. wt. ≅ 100,000 Daltons) and the uptake of the steroid precedes the appearance of the physiological effect. These and other findings, including (1) the stereospecificity of the high affinity binding proteins, (2) the close correlation between affinity of various steroids for the binding site and physiological potency, and (3) association of the steroid–protein complex with chromatin, lend credence to the inference that these high affinity binding proteins are receptors that mediate the biological response [1, 4].

The physical and biological characteristics of all of the steroid receptors, including that of 1,25 dihydrocholecalciferol (Vitamin D derivative) are remarkably homologous [1, 4]. The main features of these homologies will be illustrated for the most part by reference to studies with aldosterone, since this steroid has been the focus of my work for more than a decade [7].

The association of steroids with target cell nuclei was first revealed by autoradiography [7, 8]3H-Aldosterone was localized to the nuclear and perinuclear areas of toad bladder epithelium. In contrast, 3H-progesterone, an inactive steroid at low concentrations compared to aldosterone, was randomly distributed. Bogoroch (cited in Edelman [9]) explored the specificity of the observed binding in competition studies (Table 1). Excess estradiol-17β, which is inactive with respect to Na+ transport in the toad bladder, had no effect on the distribution of 3H-aldosterone but 9α-fluorocortisol, an active mineralocorticoid, significantly diminished both nuclear and cytoplasmic localization. These findings imply the existence of cytoplasmic and nuclear receptors, in vivo.

The receptors appear to reside in the cytoplasm pending availability of their respective steroids and then bind to chromatin sites after formation of the cytoplasmic complex. This “two-step” mechanism was independently proposed by Gorski et al. [10] and by Jensen et al. [11]. An important feature of this mechanism is the requirement for temperature activation of the complex for binding to the nucleus.

The cytoplasmic-nuclear transfer process and other features of the steroid–receptor system have been incorporated into the model shown in Fig. 2. The receptor is assumed to exist in an inactive and an active conformation in equilibrium, as formulated by Rubin and Changeux [12] for allosteric enzymes. This allosteric equilibrium model was applied to steroid receptors by Samuels and Tomkins [13] in order to account for the behaviour of steroids with mixed agonist and antagonist properties [suboptimal inducers—in their nomenclature]. The phenomenon of partial agonist-antagonist behaviour is illustrated in Table 2. In the isolated toad bladder system, 11-deoxycortisol elicited an increase in active Na+ transport, measured by the short-circuit current (SCC) technique, that was 30% of that caused by maximum doses of aldosterone and in combination with aldosterone inhibited the response by 71%. Thus total activity accounts for 100% occupancy of the putative receptors and can be rationalized as indicating that the affinity of 11-deoxycortisol for the two conformations is in the ratio of 30:70, active vs inactive. Primary agonists (e.g., aldosterone) would presumably have a high affinity for the active conformation and negligible affinity for the inactive conformation. The converse should hold for primary antagonists. Support for this model was obtained in studies with steroid antagonists. Baxter et al. [14] found that in hepatoma cells in tissue culture, progesterone inhibited induction of tyrosine amino-transferase by glucocorticoids and occupied cytoplasmic glucocorticoid receptor sites but failed to generate intra-nuclear complexes. Kaiser et al. [15, 16] reported that in rat thymocytes, the anti-glucocorticoid, cortexolone bound to the receptor but failed to transfer to nuclear binding sites. In our studies, the anti-mineralocorticoid, spirolactone-mineralocorticoid–receptor complexes (SC-26304) did not bind to nuclear acceptor sites in vivo (see Table 3), nor in reconstitution experiments with isolated nuclear or...