Research in the May issue of Nature Structural Biology has shown that a small molecule known as THC antagonizes oestrogen receptor-β (ER-β) by means of a novel mechanism. THC blocks transcription without the usual physical obstruction of the positioning of helix 12 in the ER-β protein, which is necessary for receptor activation. Interestingly, whereas THC acts as an antagonist of ER-β, it is an agonist of ER-α. This study solved the crystal structures of THC bound to the ligand-binding domains of both ER-α and ER-β, in order to investigate the mechanisms of THC agonism and antagonism.

Both ER-α and ER-β mediate the physiological effects of both endogenous and synthetic oestrogens. These receptors are members of the nuclear-receptor superfamily of ligand-regulated transcription factors. The ligand-binding domains of ER-α and ER-β each have a transcriptional activation function, which is responsive to agonists, including oestrogen 17β-oestradiol (E2) and diethylstilbestrol. Antagonists, such as 4-hydroxytamoxifen and raloxifene, block the activation-function activity. Agonists stabilize a conformation of the receptor in which helix 12 lies across the opening of the binding pocket, thereby allowing the receptor to interact with the transcriptional co-activators that mediate ligand-dependent transcription of the receptor. Most ER antagonists have bulky side chains that cannot fit within the binding pocket, and so ER-β helix 12 cannot adopt the agonist-bound conformation, which occludes binding to transcriptional co-activators. But because THC lacks the bulky side chain that is typical of other ER antagonists, it must antagonize ER-β through a different mechanism.

Comparing the crystal structures of THC bound to either ER-β or ER-α shows that although THC binds to both chains in a similar manner overall, it fails to stabilize several of the binding-pocket interactions with ER-β in comparison to those with ER-α. The slight differences in sequence between the two ERs means that crucial residues in ER-β adopt a random-coil conformation instead of a helical one. This does not favour the agonist-bound conformation of helix 12, and actually stabilizes an inactive conformation of the helix.

This new type of antagonism is probably not unique to THC and ER-β. There are other examples of nuclear-receptor ligands that act as antagonists, even though they might be smaller than the endogenous agonists. Flutamide, a synthetic androgen-receptor antagonist, is similar in size to testosterone, and does not have a bulky side chain to act as an antagonist. Also, progesterone is smaller than aldosterone, but is a high-affinity antagonist of the mineralocorticoid receptor.

This insight into such non-classical antagonism of nuclear receptors highlights a new possible approach to designing antagonists, in which compounds could be tailored to selectively stabilize inactive conformations of certain nuclear receptors, and the active conformations of others.