Tamoxifen — a selective oestrogen receptor modulator (SERM) that regulates oestrogen receptor (ER)-induced gene expression — has a split personality: it inhibits gene expression, and hence cell growth, in the breast, but promotes it in the endometrium. Treating breast cancer patients with tamoxifen might therefore promote carcinogenesis in the uterus. So what determines these tissue-specific effects? Yongfeng Shang et al., reporting in Science, show that tissue-specific co-regulator expression levels determine tamoxifen's effects.

The ER regulates gene transcription both directly, by binding to an oestrogen-responsive element (ERE) in gene promoters, and indirectly, by binding through other transcription factors. How do SERMs modulate these different promoter types? Tamoxifen and raloxifene — a SERM that acts as an oestrogen antagonist in both breast and endometrium — were administered to MCF-7 breast carcinoma cells and Ishikawa endometrial carcinoma cells, to investigate their effect on gene expression. Tamoxifen, but not raloxifene, stimulated transcription of the indirect ER target genes c-MYC and IGF1 , but not the direct ER target genes CTSD and EBAG7 , in the Ishikawa cells, but not in the MCF-7 cells. The promoter type is therefore an important determinant of tamoxifen activity.

But what mediates this differential activity? Chromatin immunoprecipitation experiments showed that tamoxifen recruited transcriptional repressors to both promoter types in MCF-7 cells, and to the direct promoters in Ishikawa cells. However, tamoxifen recruited transcriptional activators to indirect promoters in Ishikawa cells. Might different expression levels of these transcriptional regulators explain the tissue-specific effects of tamoxifen?

Levels of the transcriptional activator NCOA1 (also known as SRC1) were lower in mammary cells than in endometrial cells. Overexpression of SRC1 in MCF-7 cells stimulated tamoxifen-induced transcription. By contrast, its silencing in Ishikawa cells by RNA interference abolished tamoxifen-induced transcription and the associated cell-cycle progression.

So, tissue-specific differences in SRC1 levels determine whether tamoxifen is pro- or anti-oestrogenic on some promoter types. Let's hope that this mechanistic knowledge will lead to better SERMs without carcinogenic side effects.