Deprivation of estrogen causes breast tumors in women to adapt and develop enhanced sensitivity to this steroid. Accordingly, women relapsing after treatment with oophorectomy, which substantially lowers estradiol for a prolonged period, respond secondarily to aromatase inhibitors with tumor regression. We have utilized in vitro and in vivo model systems to examine the biologic processes whereby long-term estradiol deprivation (LTED) causes cells to adapt and develop hypersensitivity to estradiol. Several mechanisms are associated with this response, including up-regulation of estrogen receptor-a (ERa) and the MAP kinase, phosphoinositol 3 kinase (PI3-K) and mammalian target of rapamycin (mTOR) growth factor pathways. ERa is four-to tenfold up-regulated and co-opts a classical growth factor pathway using Shc, Grb-2 and Sos. This induces rapid non-genomic effects which are enhanced in LTED cells. The molecules involved in the non-genomic signaling process have been identified. Estradiol binds to cell membrane-associated ERa, which physically associates with the adaptor protein Shc, and induces its phosphorylation. In turn, Shc binds Grb-2 and Sos, which result in the rapid activation of MAP kinase. These non-genomic effects of estradiol produce biologic effects as evidenced by Elk-1 activation and by morphologic changes in cell membranes. Additional effects include activation of the PI3-K and mTOR pathways through estradiolinduced binding of ERa to the IGF-I and epidermal growth factor receptors. A major question is how ERa locates in the plasma membrane since it does not contain an inherent membrane localization signal. We have provided evidence that the IGF-I receptor serves as an anchor for ERa in the plasma membrane. Estradiol causes phosphorylation of the adaptor protein, Shc and the IGF-I receptor itself. Shc, after binding to ERa, serves as the 'bus' which carries ERa to Shc-binding sites on the activated IGF-I receptors. Use of small inhibitor (si) RNA methodology to knockdown Shc allows the conclusion that Shc is needed for ERa to localize in the plasma membrane. In order to abrogate growth factorinduced hypersensitivity, we have utilized a drug, farnesylthiosalicylic acid, which blocks the binding of GTP-Ras to its membrane acceptor protein, galectin 1, and reduces the activation of MAP kinase. We have also shown that this drug is a potent inhibitor of mTOR as an additional mechanism of inhibition of cell proliferation. The concept of 'adaptive hypersensitivity' and the mechanisms responsible for this phenomenon have important clinical implications. The efficacy of aromatase inhibitors in patients relapsing on tamoxifen could be explained by this mechanism and inhibitors of growth factor pathways should reverse the hypersensitivity phenomenon and result in prolongation of the efficacy of hormonal therapy for breast cancer.

Clinical observations suggest that human breast tumors can adapt to endocrine therapy by developing hypersensitivity to estradiol (E 2 ). To understand the mechanisms responsible, we examined estrogenic stimulation of cell proliferation in a model system and provided in vitro and in vivo evidence that long-term E 2 deprivation (LTED) causes 'adaptive hypersensitivity'. The enhanced responses to E 2 do not involve mechanisms acting at the level of transcription of estrogen-regulated genes. We found no evidence of hypersensitivity when examining the effects of E 2 on regulation of c-myc, pS2, progesterone receptor, several estrogen receptor (ER) reporter genes, or c-myb in hypersensitive cells. Estrogen deprivation of breast cells long-term does up-regulate both the MAP kinase and phosphatidyl-inositol 3-kinase pathways. As a potential explanation for up-regulation of these signaling pathways, we found that ERα is 4-to 10-fold up-regulated and co-opts a classic growth factor pathway using Shc, Grb-2 and Sos. This induces rapid non-genomic effects which are enhanced in LTED cells. E 2 binds to cell membrane-associated ERα, physically associates with the adapter protein SHC, and induces its phosphorylation. In turn, Shc binds Grb-2 and Sos, which results in the rapid activation of MAP kinase. These non-genomic effects of E 2 produce biological effects as evidenced by Elk activation and by morphological changes in cell membranes. Further proof of the non-genomic effects of E 2 involved use of cells which selectively expressed ERα in the nucleus, cytosol and cell membrane. We created these COS-1 'designer cells' by transfecting ERα lacking a nuclear localization signal and containing a membrane localizing signal.The concept of 'adaptive hypersensitivity' and the mechanisms responsible for this phenomenon have important clinical implications. Adaptive hypersensitivity would explain the superiority of aromatase inhibitors over the selective ER modulators (SERMs) for treatment of breast cancer. The development of highly potent third-generation aromatase inhibitors allows reduction of breast tissue E 2 to very low levels and circumvents the enhanced sensitivity of these cells to the proliferative effects of E 2 . Clinical trials in the adjuvant, neoadjuvant and advanced disease settings demonstrate the greater clinical efficacy of the aromatase inhibitors over the SERMs. More recent observations indicate that the aromatase inhibitors are superior for the prevention of breast cancer as well. These observations may be explained by the hypothesis that estrogens induce breast cancer both by stimulating cell proliferation and by their metabolism to genotoxic products. The SERMs block ER-mediated proliferation only, whereas the aromatase inhibitors exert dual effects on proliferation and genotoxic metabolite formation.