Starting from the human estrogen receptor α, the authors employed directed evolution: they changed the residues in proximity of the ligand by mutagenesis, screened the resulting mutants, and selected the best receptor mutants for the next round. After the third round of directed evolution, they came up with an optimized mutant that bound to CV3320 with an EC₅₀ of 55 nM, while the affinity to 17β-estradiol was reduced by a factor of 10 (4 nM).

While the authors admit that the selectivity over 17β-estradiol could still be improved, it still is a nice piece of work that demonstrates the power of directed evolution. This way, a protein receptor for a substrate that does not occur in nature can be made. Such a receptor can be used to make biological switches.

An overlay of the 3D structures of estrone (1, yellow) and 3 (white, R = 2-Cl, R’ = Me) shows a considerable offset between the corresponding carbonyl groups (O – O distance of 1.67 Å). I also wonder about the rotational flexibility of the linker – the “text-book”-property of steroids is their rigidity, but the new analogues can rotate about the triple bond.

The biological properties of these new compounds have not yet been measured. I wonder if they will be comparable to estrogens. If this radical structural simplification still yields bioactive compounds, this will be a remarkable achievement. On the other hand, flexibility always has an entropic cost when the molecules bind to their target, so I don’t expect the activities to be too high. Also, the new class of compounds is probably less selective than the original estrogens. But this is all speculation as long as the biology hasn’t been done.