We present the structure-based optimization of a series of estro
gen receptor-
![](/ima<font color=)
ges/
gifchars/beta2.
gif" BORDER=0 ALIGN="middle"> (ER
![](/ima<font color=)
ges/
gifchars/beta2.
gif" BORDER=0 ALIGN="middle">) selectiveli
gands. X-ray cocrystal structures of these li
gands complexed to both ER
![](/ima<font color=)
ges/
gifchars/alpha.
gif" BORDER=0> and ER
![](/ima<font color=)
ges/
gifchars/beta2.
gif" BORDER=0 ALIGN="middle"> are described. Wealso discuss how molecular modelin
g was used to take advanta
ge of subtle differences between the twobindin
g cavities in order to optimize selectivity for ER
![](/ima<font color=)
ges/
gifchars/beta2.
gif" BORDER=0 ALIGN="middle"> over ER
![](/ima<font color=)
ges/
gifchars/alpha.
gif" BORDER=0>. Quantum chemical calculations are utilizedto
gain insi
ght into the mechanism of selectivity enhancement. Despite only two relatively conservativeresidue substitutions in the li
gand bindin
g pocket, the most selective compounds have
greater than 100-fold selectivity for ER
![](/ima<font color=)
ges/
gifchars/beta2.
gif" BORDER=0 ALIGN="middle"> relative to ER
![](/ima<font color=)
ges/
gifchars/alpha.
gif" BORDER=0> when measured usin
g a competitive radioli
gand bindin
g assay.