Clinically important chemopreventive agents of the antiestrogen and retinoid families have been shown to upregulate expression of the growth inhibitory transforming growth factor-beta (TGF-beta) family in vitro. This has led to the hypothesis that the chemopreventive action of these agents may be mediated at least in part by the TGF-betas, and, if so, that TGF-betas might be useful biomarkers of chemopreventive efficacy. We have tested this hypothesis in a standard rat model of mammary tumorigenesis, in which Sprague-Dawley rats were initiated with the carcinogen N-methyl-nitrosourea, and were then either left untreated or were treated with the antiestrogen, tamoxifen, or the retinoids, 4-N- (hydroxy)phenyl retinamide or 9-cis retinoic acid, alone or in combination. No changes were observed in the immunohistochemical expression of TGF-betas-1, 2 or 3, the type I or type II TGF-beta receptors, or the latent TGF-beta binding protein, in the mammary glands of treated compared with untreated rats. This suggests that the observed chemopreventive efficacy of antiestrogens or retinoids in this rat model is independent of effects on the TGF-beta system, and further that the TGF-betas may not be useful biomarkers in clinical breast cancer chemoprevention trials using these agents. To address the mechanism underlying the upregulation of TGF-beta1 by tamoxifen that had been observed in vitro, we mapped the start sites of all three rat TGF-beta1 transcripts for the first time, and used deletion analysis to demonstrate that the 51 untranslated regions contain multiple cis-regulatory elements that affect translational efficiency. An understanding of the mechanisms whereby steroids and related compounds regulate the production and activity of the TGF-beta family of growth inhibitors may allow the rational design of more potent pharmacological agents for use in chemoprevention or chemotherapy of cancer. Following the recommendation of the 1996 Site Visit report, this project was terminated in April 1997.
