Another potential source of variable affinities is inherent
Another potential source of variable affinities is inherent in the method of RBA determination. While several reports simply use relative IC50 values as RBAs , , , , , these are only an approximate measure of relative affinity since IC50 varies due to experimental and biological parameters, such as the concentration of radiolabeled steroid and the concentration of the receptor being investigated. Fig. 1A illustrates how the IC50 changes in a competition binding assay as a function of receptor concentration. Interestingly, if the Hill slope is not fixed during plotting, it also increases as receptor concentration increases, suggesting that caution should be exerted when interpreting changes in Hill slope in competitive binding studies when receptor concentration is much greater than the true K. A more accurate measure of affinity of progestogens for a SR that circumvents these problems can be obtained by saturation binding to obtain an equilibrium dissociation constant (K). These K values are likely to be more accurate than RBAs determined by competitive binding, provided that the K is greater than the total receptor concentration in the assay and that other sources of technical and samples source error are not present. However, only a few reports use saturation binding to measure RBAs of progestogens. For example, K values of 10.9nM and 4.42nM for the PR have been determined for drospirenone (DRSP), and R5020, respectively , while a K value of 1.7nM was determined for medroxyprogesterone acetate (MPA) for the AR . Sometimes RBAs are calculated from K values. For example, K values were determined for NOMAC and Org2058 for the PR by saturation binding in rat uterus (K=5nM and 0.6nM, respectively) and human T47D breast cancer GSK1904529A australia (K=4nM and 3nM, respectively). RBAs were then calculated relative to Prog set with a RBA of 100%, such that NOMAC and Org2058 displayed RBAs of 67% and 692%, respectively for the PR in rat uterus, and 192% and 212%, respectively for the PR in human T47D cells (Table 1 and Supplementary Table 1). The differences in the K values for Org2058 and the RBAs for both ligands relative to Prog most likely reflect different off-target and/or metabolism and/or species effects in the two systems.As an alternative to saturation binding, homologous or heterologous competition binding displacement assays can be used to determine accurate K or K values using the Cheng–Prussof equation (Supplementary Table 1), provided the concentration of radiolabeled ligand is less than the IC50 , . The K is the equilibrium dissociation constant of the unlabeled competitor or inhibitor, and is a true constant that does not vary with receptor concentration in the assay, provided certain experimental restrictions are adhered to. Using this method, similar K values were obtained by two groups for Prog and MPA binding to the GR (K: 95.2nM and 215nM and K: 3.7nM and 10.8nM, respectively) , . RBAs for the GR can also be calculated from K values (e.g. relative to dexamethasone set at 100%, RBAs for Prog: 0.84%, and 2%, RBAs for MPA: 21.6% and 39%) (Table 1 and Supplementary Table 1). However, when comparing K values and RBAs calculated from K values obtained from different groups, large discrepancies are still often found. For example, for binding to the AR, a two-fold difference in the RBA of MPA has been reported, while a five- to nine-fold difference in the K values has been reported (151%, K=19.4nM ; 75%, K=1.7–3.6nM , RBAs relative to DHT set at 100%), (Table 1 and Supplementary Table 1). Possible reasons for these discrepancies may be failure to establish equilibrium and/or using a concentration of radiolabeled ligand greater than the IC50. Very different RBAs are likely to be obtained when using K values obtained by the Cheng–Prussof method compared to simply using IC50 values. For example the values obtained by Africander et al., for MPA for the MR (aldosterone=100%) (0.08%; K=197nM) and NET-A (0.07%; K=229nM) are 40-fold different to previously reported values of 3.1% and 2.7%, respectively (Table 1 and Supplementary Table 1) , .