Combinatorial complexity in a transcriptionally centered signaling hub in Arabidopsis

A subfamily of four Phytochrome (phy)-Interacting bHLH transcription Factors (PIFs) collectively promote skotomorphogenic development in dark-grown seedlings. This activity is reversed upon exposure to light, by photoactivated phy molecules that induce degradation of the PIFs, thereby triggering the transcriptional changes that drive a transition to photomorphogenesis. The PIFs function both redundantly and partially differentially at the morphogenic level in this process. To identify the direct targets of PIF transcriptional regulation genome-wide, we analyzed the DNA-binding sites for all four PIFs by ChIP-seq analysis, and defined the genes transcriptionally regulated by each PIF, using RNA-seq analysis of pif mutants. Despite the absence of detectable differences in DNA-binding-motif recognition between the PIFs, the data show a spectrum of regulatory patterns, ranging from single PIF dominance to equal contributions by all four. Similarly, a broad array of promoter architectures was found, ranging from single PIF-binding sites, containing single sequence motifs, through multiple PIF-binding sites, each containing one or more motifs, with each site occupied preferentially by one to multiple PIFs. Quantitative analysis of the promoter occupancy and expression level induced by each PIF revealed an intriguing pattern. Although there is no robust correlation broadly across the target-gene population, examination of individual genes that are shared targets of multiple PIFs shows a gradation in correlation from strongly positive, through uncorrelated, to negative. This finding suggests a dual-layered mechanism of transcriptional regulation, comprising both a continuum of binding-site occupancy by each PIF and a superimposed layer of local regulation that acts differentially on each PIF, to modulate its intrinsic transcriptional activation capacity at each site, in a quantitative pattern that varies between the individual PIFs from gene to gene. These findings provide a framework for probing the mechanisms by which transcription factors with overlapping direct-target genes integrate and selectively transduce signals to their target networks.