Regulatory plasticity within a complex cytokine-sensing mammary enhancer during lactation

Enhancers are transcription factor platforms that synergize with promoters to activate gene expression up to several-thousand-fold. While genome-wide structural studies are used to predict enhancers, the in vivo significance is less clear. Specifically, the biological importance of individual transcription factors within enhancer complexes remains to be understood. Here we investigate the structural and biological importance of individual transcription factor binding sites and redundancy among transcription components within a complex enhancer in vivo. The Csn1s2b gene is expressed exclusively in mammary tissue and activated several thousand-fold during pregnancy and lactation. Using ChIP-seq we identified a complex lactation-specific candidate enhancer that binds multiple transcription factors and coincides with activating histone marks. Using experimental mouse genetics, we determined that deletion of canonical binding motifs for the transcription factors NFIB and STAT5, individually and combined, had a limited biological impact. Loss of these sites led to a shift of transcription factor binding to juxtaposed sites, suggesting exceptional plasticity that does not require direct protein-DNA interactions. Additional deletions revealed the critical importance of a non-canonical STAT5 binding site for enhancer activity. Our data also suggest that enhancer RNAs are not required for the activity of this specific enhancer. While ChIP-seq experiments predicted an additional candidate intronic enhancer, its deletion did not adversely affect gene expression, emphasizing the limited biological information provided by structural data. Our study provides comprehensive insight into the anatomy and biology of a composite mammary enhancer that activates its target gene several hundred-fold during lactation.