The cathepsin D (cath-D) gene, coding for a ubiquitous Iysosomal aspartyl protease, is overexpressed in aggressive human breast cancers, and its transcription is induced by estrogens in hormone-responsive breast cancer cells. We have determined the structure and function of the proximal 5' upstream region of the human cath-D gene from MCF7 cells. We show that the promoter has a compound structure with features of both housekeeping genes (high G+C content and potential transcription factor Spl sites) and regulated genes (TATAA sequence). By RNase protection assay, we show that transcription is initiated at five major transcription sites (TSSI to -V) spanning 52 base pairs. In hormoneresponsive breast cancer cells, estradiol increased by 6-to 10-fold the level of RNAs initiated at TSSI, which is located about 28 base pairs downstream from the TATA box. The specific regulation by estradiol of transcription starting at site I exclusively was confimed by primer extension. Moreover, the same estradiol effect was observed in the ZR75-1 cell line and in MDA-Ml231 estrogen-resistant breast cancer cells stably transfecte4 with the estrogen receptor. Site-directed mutagenesis in,#1c!Ited that the TATA box is essential for initiation ofcatfll') gene transcription at TSSI. In breast cancer biopsy samples, high levels of TATA-dependent transcription were correlated with overexpression of cath-D mRNA. We conclude that cath-D behaves, depending on the conditions, as a housekeeping gene with multiple start sites or as a hormoneregulated gene that can be controlled from its TATA box.Steroid repeptors increase the initiation of transcription of specific geneX lay interacting with the transcriptional machinery at thy promoter level (1). Cathepsin D (cath-D), a lysosomal gspartyl protease, is induced by estrogens in human breast cancer cell lines (2) and is produced in excess in cancer cells both in vitro and in vivo, where its concentration in the primary tumor is correlated with increased risk of metastasis (for review see ref.3).In eukaryotes, gene expression is controlled by both proximal and distal elements, generally located in the 5' upstream region of the gene (4). Many class II gene promoters contain a TATA box, which binds the transcription factor IID and defines the transcription initiation site; genes with these promoters have been called facultative or regulated genes. In contrast, promoters of housekeeping genes, such as those coding for lysosomal enzymes (5), lack a recognizable TATA box but contain multiple GC boxes acting as putative binding sites for the transcription factor Spl (6).We have previously shown, in human breast cancer cell lines, that estrogens stimulate transcription of the cath-D gene by means of estrogen-responsive sequences located in the proximal region of the promoter (7). Here we show, using RNase protection and primer extension, that cath-D gene transcription in breast cancer cells is initiated at multiple sites by a mixed promoter having characteristics of the promoters of both houseke...