Comparative gene expression profiling reveals partially overlapping but distinct genomic actions of different antiestrogens in human breast cancer cells.

Antiestrogens used for breast cancer (BC) treatment differ among each other for the ability to affect estrogen receptor (ER) activity and thereby inhibit hormone-responsive cell functions and viability. We used high-density cDNA microarrays for a comprehensive definition of the gene pathways affected by 17beta-estradiol (E2), ICI 182,780 (ICI), 4OH-tamoxifen (Tamoxifen), and raloxifene (RAL) in ER-positive ZR-75.1 cells, a suitable model to investigate estrogen and antiestrogen actions in hormone-responsive BC. The expression of 601 genes was significantly affected by E2 in these cells; in silico analysis reveals that 86 among them include one or more potential ER binding site within or near the promoter and that the binding site signatures for E2F-1, NF-Y, and NRF-1 transcription factors are significantly enriched in the promoters of genes induced by estrogen treatment, while those for CAC-binding protein and LF-A1 in those repressed by the hormone, pointing to novel transcriptional effectors of secondary responses to estrogen in BC cells. Interestingly, expression of 176 E2-regulated mRNAs was unaffected by any of the antiestrogens tested, despite the fact that under the same conditions the transcriptional and cell cycle stimulatory activities of ER were inhibited. On the other hand, of 373 antiestrogen-responsive genes identified here, 52 were unresponsive to estrogen and 25% responded specifically to only one of the compounds tested, revealing non-overlapping and clearly distinguishable effects of the different antiestrogens in BC cells. As some of these differences reflect specificities of the mechanism of action of the antiestrogens tested, we propose to exploit this gene set for characterization of novel hormonal antagonists and selective estrogen receptor modulators (SERMs) and as a tool for testing new associations of antiestrogens, more effective against BC.

Chp-1 and melusin, two CHORD containing proteins in vertebrates.

Melusin is a muscle specific protein required for heart hypertrophy in response to mechanical overload. Here we describe a protein 63% homologous to melusin, named chp-1, expressed in all tissues tested, including muscles, and highly conserved from invertebrates to human. Both proteins are characterized in their N-terminal half by a tandemly repeated zinc binding 60 amino acid domain with a motif of uniquely spaced cysteine and histidine residues. These motives are highly conserved from plants to mammals and have been recently named CHORD (for cysteine and histidine rich domain) domains. At the C-terminal end melusin contains a calcium binding stretch of 30 acidic amino acid residues which is absent in chp-1. While invertebrate genome contains only one gene coding for a chp-1 homolog, two genes coding for CHORD containing proteins (chp-1 and melusin) are present in vertebrates. Sequence analysis suggests that the muscle specific CHORD containing protein melusin originated by a gene duplication event during early chordate evolution.