Design of cytotoxic steroids for prostate cancer.

Our object was to determine if the aromatic nucleus of estramustine (I) is optimal for binding affinity to prostate cytosolic proteins, and if C3 is the preferred position for the N-mustard carbamate moiety. To this end we have submitted 34 steroids for in vitro assay of binding affinity to total prostate cytosolic proteins. Our structures included aromatic and hydroaromatic steroids containing N-mustard carbamate and other substituents at C3, C6, C11, C16, C17, C20, and C21. Our results show that binding affinity to prostate proteins is optimally present in C3-nitrogen mustard carbamates attached directly to a totally planar aromatic ring as in (IV). Partial deviation from total planarity as in enol-carbamates (V) leads to some loss of binding affinity, which largely disappears in hydroaromatic structures (VI). Thus, our data lead to the Ring A aromatic structure (X) as a basis for the design of steroidal N-mustard carbamates with prostate selectivity. Preliminary in vivo studies using the Dunning R3327AT prostatic adenocarcinoma implanted in the Copenhagen rat generally support our in vitro data.