Comparison of glucocorticoid-receptor binding kinetics with predictions from a biomolecular model.

Previous studies of the kinetics of steroid binding to receptors have shown a lack of agreement between the dissociation constant obtained at equilibrium by Scatchard analysis (Scatchard, G. (1949) Ann. N. Y . Acad Sci. 51, 660-672) and that calculated using the rate constants of dissociation and association (kd/k,) (Pratt, W. B., Kaine, J. L., and Pratt, D. V. (1975) J. Biol. Chem. 250, 4584-4591; Hansen, P. E., Johnson, A., Schrader, W. T., and O’Malley, B. W. (1976) J. Steroid Biochem. 7, 723-732; Aranyi, P. (1979) Biochim. Biophys. Acta 584, 529-537). It was also found that the rate of glucocorticoid receptor formation did not increase linearly as expected when the free steroid concentration was increased (Pratt, W. B., Kaine, J. L., and Pratt, D. V. (1975) J. Biol. Chem. 250,4584-4591). These experimental observations have led others to conclude that the steroid-receptor-binding process was not a simple bimolecular reaction but involved multiple steps. This paper reports similar experimental observations for glucocorticoid binding to the AtT-20 cell cytosol receptor. A plot of the association rate versus free steroid concentration was nonlinear at higher concentrations of triamcinolone acetonide, dexamethasone, or corticosterone. The dissociation constants (Kd) for these steroids were determined by Scatchard analysis of binding at apparent equilibrium. In every instance these dissociation constants were larger than the ratios of the dissociation and association rate constants (kd/ ha,). The differences were greatest for triamcinolone acetonide (80-fold) and smallest for corticosterone (7fold). Since these observations agreed with those of others it might appear that glucocorticoids also bind to the AtT-20 cell receptor by a complex mechanism. However, we next used computer simulations to see if these data could be explained by a simple bimolecular reaction. We found that the simulated curve of the association rate versus free steroid concentration was very similar to the experimental curve and that they both deviated from linearity at identical points. We also measured the rate of inactivation of the unbound receptor and included that term in a computer simulation of the Scatchard plot. This led to a considerable increase (23-fold) in the estimation of the dissociation constant and produced simulated data very similar to our experimental findings. The agreement of our experimental observations