Steroid receptors, and the generation of closely coupled/biphasic dose-response curves

In most steroid-modulated systems, responses are closely coupled to receptor occupancy, with half-maximal responses at Kd, and near-maximal at greater than or equal to 10 Kd. Though glucocorticoids have been reported to increase, and progesterone to lower, levels of various milk proteins, clear differences in glucocorticoid dose-response profiles have been previously described. In mouse mammary gland explants in vitro, Ono and Oka showed that casein synthesis was stimulated by cortisol to plateau levels with progressively higher doses; in contrast, the dose-response curve for alpha-lactalbumin (alpha-LA) was biphasic, with stimulation at low doses and a return to baseline at higher cortisol levels. Since cortisol is both non-physiologic in the mouse, and has higher affinity for Type I (mineralocorticoid-like) than Type II (glucocorticoid) receptors, this might reflect an induction of alpha-LA synthesis via Type I receptor occupancy, and a suppression via Type II receptor occupancy. To examine this possibility we have used highly selective Type II receptor ligands, and in addition have defined the role of progestins as candidate inhibitors of alpha-LA synthesis. We have thus incubated mid-pregnant rat mammary gland explants in medium including insulin and prolactin, with increasing concentrations of highly specific Type II ligand (RU26988 or RU28362) and/or the selective progestin ORG 2058, and determined tissue and medium content of alpha-LA 48 h later. RU26988/RU28362 increased alpha-LA to a maximum of 3-6-fold basal at approximately 3 nM, well below the Kd (37 degrees C) for Type II receptors (8-28 nM); levels fell progressively with increasing concentration of ligand, returning to baseline by 30-300 nM. ORG 2058 lowered both basal and RU26988-stimulated levels, with the half-maximally effective dose being approximately 10(-10) M, again well below Kd for receptor binding. To accommodate the observed biphasic dose-response curve for highly specific glucocorticoids, and the shift-to-the-left for both classes of effector, we have proposed a model of a single "turn-on" nuclear acceptor site for glucocorticoid receptors (glucocorticoid regulatory element: GRE) and multiple, pre-emptive "turn-off" GREs, occupancy of any one of which will abrogate transcription; for progestins, we propose that all the GRE are alternate "turn-off" sites. Consistent with such a hypothesis, Qasba and Safaya have shown that the alpha-LA genome contains six core consensus sequences (TGT TCCT) which are putative GRE, and von der Ahe et al. that both progesterone receptors and glucocorticoid receptors may bind to GRE sites.