In vivo footprinting of an androgen-dependent enhancer reveals an accessory element integral to hormonal response.

A hormonally responsive enhancer that is specifically activated by androgens resides 2 kilobases upstream of the transcription start site of the mouse sex-limited protein (Slp) gene. We have previously shown that strong androgen induction in transfection requires a consensus hormone response element as well as several nonreceptor factor binding sites within this complex enhancer. To determine which accessory elements are required for androgen-dependent transcription, we have examined binding of nuclear proteins to the enhancer both in vitro and in vivo. In vitro footprinting assays demonstrated that multiple factors present in mouse liver and kidney nuclear extracts bound the enhancer, with tissue-specific but not sex-dependent differences in pattern. In contrast, examination of DNA sites occupied in liver chromatin identified a footprint (FPIV) that is well protected in males but sensitive to DNase I in females. FPIV was occupied in males in other sites of Slp expression, such as kidney, but not in tissues lacking expression, such as lung. FPIV protection was induced in females treated with androgen, abrogated in castrated males, and absent in immature mice, implying hormonal and developmental regulation of FPIV binding. Protection of the hormone response element, in contrast to FPIV, was not obvious but was discerned by analysis of densitometry data. Together with results from in vivo protein-DNA interactions determined for other steroid-dependent enhancers, this suggests that in some cases receptor may permit transcriptional activation by altering chromatin structure to allow access to other factors, which may not necessitate tight binding of receptor itself. This further emphasizes the crucial role of the nonreceptor factors in hormone response. The ubiquitous transcription factor Oct-1 forms complexes with an octamer motif present within FPIV by gel shift analysis with liver and kidney extracts, making Oct-1 an intriguing candidate for partnership in androgen regulation.