Cyclic, nonequilibrium models of glucocorticoid antagonism: role of activation, nuclear binding and receptor recycling.

Quantitative models that have been proposed to date to explain mechanisms of glucocorticoid antagonism have generally been of the equilibrium type, involving hypothetical allosteric equilibria between active and inactive states of the receptor or the steroid-receptor complex. We describe here the agonist-antagonist relationships predicted by a nonequilibrium cyclic model that we have recently devised to account for the kinetic behavior of glucocorticoid-receptor complexes in intact rat thymus cells. This model simulates quantitatively most kinetic and steady state results that have been obtained so far. It postulates the existence of only well-established receptor species, and its kinetic parameters can in principle be determined by receptor measurements with intact cells. To calculate the steady state agonist-antagonist properties it is assumed that biological activity is proportional to the total amount of nuclear-bound complex, whether formed by agonist or antagonist. The agonist activity of a steroid is determined by the steady state ratio of nuclear-bound to total complexes it forms. This ratio varies from 0 for a pure antagonist to 1 for a pure agonist. It turns out to be independent of agonist and antagonist concentrations, and a function only of the rate constants for the reactions of the complexes formed by a steroid. Analysis of the dependence of the ratio on each rate constant shows quantitatively how each reaction in the cyclic model--activation of the nonactivated complex, nuclear binding of the activated complexes, and dissociation and recycling of activated and nuclear-bound complexes--affects antagonist properties. Steady state interactions of agonists with antagonists are found to be determined by equations that are identical to those for competition in simple equilibrium systems. Predicted dose-response relations agree qualitatively with experimentally observed relations. They are similar to those predicted by two-state allosteric models, although the cyclic model has no allosteric mechanisms and is based on quite different assumptions. Present limitations of the model arise particularly from lack of information about the mechanisms by which nuclear-bound complexes generate biological activity; for lack of such information the model includes no steps to account for substances that have low agonist activity despite forming nuclear-bound complexes.