Hormone-dependent tumors of the prostate and breast are the most commonly diagnosed tumors today. Therapy for metastatic disease directly targets the steroid hormone receptors. Therefore it is critical to understand the role of these receptors in cancer initiation and progression. In prostate cancer, although tumors initially respond to therapy, they eventually relapse, progressing to androgen-independence. However, the androgen receptor (AR) is still active as evidenced by expression of many of its target genes. Gain of function AR mutations have been identified in androgen-independent prostate cancer suggesting they are selected by treatment. To confirm selection of AR mutations and evaluate their prevalence, we examined a mouse prostate cancer model, thus avoiding the heterogeneity of human disease initiation and treatment.

Sequencing the Ar coding sequence from prostate tumors of mice undergoing different treatments identified numerous Ar mutations, each at low frequency. More abundant mutations localized in functional domains by treatment. Extending this analysis to patient metastases identified additional novel as well as previously reported mutations. Characterization of selected mutant ARs revealed diverse effects on receptor function including differential promoter activation, increased stability, and promiscuous ligand activation. Thus AR exploits multiple mechanisms to evade treatment.

Reduced function AR mutations in a few male breast cancer patients suggest that an active AR may be protective against breast cancer. Characterizing these mutants functionally revealed differential promoter activation as was seen with AR mutations in prostate cancer. To test further the role of AR in breast cancer, tumor development was assessed in AR-null XY mice. Surprisingly AR-null mice did not develop breast cancer, which proved to be due to the absence of mammary ductal development. Mammary gland transplants suggested that lack of ductal growth in AR-null mice is due to both systemic hormonal imbalances and loss of intrinsic AR effects on mammary gland proliferation.

AR activates a broad range of targets involved in cell proliferation and differentiation. These studies together suggest that mutant ARs may activate a subset of targets that influence disease in both prostate and breast. If identified, future therapies may be able to bypass AR and selectively inhibit these targets.