Global gene expression profiling of progesterone receptor modulators in T47D cells provides a new classification system

Abstract

Progesterone receptor modulators (PRMs) play an important role in women's health. They are widely used in oral contraception or hormone therapy, and provide an attractive treatment approach for gynecological disorders such as uterine leiomyomas, endometriosis or breast cancer. Due to the broad range of activities, various studies were conducted to assess progesterone receptor antagonists (PAs) and selective progesterone receptor modulators (SPRMs) with respect to progesterone receptor (PR) agonistic and antagonistic activities in vivo. These properties are not always adequately reflected in classical in vitro models, especially differences in the agonistic potential of SPRMs, such as asoprisnil, J1042, and J912, and mixed antagonists, such as mifepristone, are not sufficiently substantiated. The effects of PRMs upon gene expression in progesterone target tissues such as breast epithelium and uterus are poorly understood. This study compares the properties of PR ligands using mammalian two-hybrid assays and gene expression profiling. The protein–protein interaction analyses in HeLa cells provide for specific ligand-induced PR conformations, whereas Affymetrix GeneChip HG-U133Plus2.0 analyses in T47D breast cancer cells indicate the transcriptional activity on the level of target genes. The analyses comprise the pure agonist R5020, the non-steroidal PR modulator PRA-910, SPRMs (J1042, asoprisnil, J912), the mixed antagonist mifepristone, classical antagonists (onapristone, ZK 137316) and the pure antagonist lonaprisan to consider all types of ligands described before. Marginal differences were identified in coactivator interaction profiles at all, but significant differences between SPRMs and PR antagonists (PAs) were observed in recruiting the LXXLL-motif containing peptide (LX-H10), very similar to in vivo activities in endometrial transformation in the rabbit (McPhail test). Global gene expression profiles demonstrated progesterone-independent effects for all PR modulators examined and emphasised similarities of asoprisnil and J1042 compared to J912 and all types of PR antagonists. In summary, the data support the popular concept of PR modulator classification in agonists, selective progesterone receptor modulators, mixed and pure antagonists. It further refines previous classification models and accentuates unique effects for each PR modulator.