Despite the sources of error discussed above several

Despite the sources of error discussed above, several trends have emerged and much valuable data are available regarding binding affinities of progestogens for different steroid receptors. All the progestogens bind to the PR with a high affinity, typically in the nanomolar range. For example the synthetic agonist R5020 was found to bind to the PR expressed in human or calf uterine cytosol with K values of 4.42nM and 5.6nM, respectively [16], [27]. In contrast, progestogens do not bind to the ER, as expected, due to the low homology between the ER ligand binding domain and the PR, although there are reports that some progestogen metabolites do bind the ER [2]. Consistent with the structural similarities between some progestogens and testosterone, several progestogens bind with relatively high affinity to the AR (Table 1), although the reported affinities vary greatly, most likely due to many of the factors described above. For example, some of the older generation progestogens such as MPA, norethisterone (NET) and levonorgestrel bind the AR with high affinity relative to metribolone [27], DHT [11], [30] and football [18], [25]. However other researchers report that Prog, MPA, norethisterone-actetate (NET-A), cyproterone acetate and DHT all have similar and relatively high affinities for the AR [17], [23], [24] (Table 1). In contrast, DRSP, dienogest and TMG exhibit low RBAs [1], [6], [25], [31], while nestorone does not bind to the AR [12], [32]. It is not surprising that several progestogens structurally similar to Prog, bind to the MR, given the high affinity of Prog for the MR (K for Prog 1.69nM [23]). Indeed, TMG [25], [33] and DRSP (Table 1) [26], [27], [31], the latter derived from the MR antagonist spironolactone, both bind the MR with high affinity. These progestogens were specifically developed for their anti-mineralocorticoid properties for contraceptive usage [34] and for their predicted beneficial effects on blood pressure and cardiovascular function [26], [35], [36]. In contrast, many progestogens such as chlormadinone acetate, dienogest and NOMAC exhibit undetectable binding to the MR [1], [37], while MPA (K=197nM) and NET-A (K=229nM) bind weakly to the MR [2], [23]. Unlike for the PR, AR and MR, few progestogens bind to the GR with affinities in the significant pharmacological range. Of note, MPA has a high affinity in the nanomolar range for the GR (Table 1) [22], [25], [37], [38], with a K value of 10.8±1.1nM [22], significantly higher than the endogenous glucocorticoid cortisol in humans [38]. Interestingly gestodene also binds the GR with a relatively high affinity, while progestogens such as NET, levonorgstrel, dienogest, TMG and DRSP, like Prog, bind the GR with low relative affinity [1], [6], [22], [25], [26], [38], [39].
Potency, efficacy and biocharacter Dose response analysis (Fig. 2) can be performed on cells in culture, in tissue, in animals or even in patients. Many different responses can be measured, including mRNA levels, reporter gene activity, gene product (e.g. enzyme) activity or even physiological functions such as inhibition of ovulation. As in the case of steroid ligand binding curves, dose response curves ex vivo generally follow a sigmoidal shape with a Hill slope value of 1 (Fig. 2). Dose response curves with the same Hill slope are considered to be parallel, while those that differ are non-parallel. In the absence of dose response analysis, relative responses obtained by single doses of progestogens will vary depending on which part of the dose response curve is represented for each progestogen. For example, if the dose for all progestogens is already at the maximal response, then the progestogens could be incorrectly reported to have the same ‘potency’. Furthermore, if the data from incomplete curves represents different parts of the curve, or off-target competing receptors are present, plots will be non-parallel. Taken together, this will result in misleading estimates of relative potency, efficacy and biocharacter.