A comprehensive map of preferentially located motifs reveals distinct proximal cis-regulatory elements in plants

Abstract

The identification of cis-regulatory elements controlling gene expression is an arduous challenge that is being actively explored to discover the key genetic factors responsible for traits of agronomic interest. In this regard, in silico predictive methods have the advantage of being of high-resolution and independent of experimental conditions. Here, we expanded the preferentially located motif (PLM) detection method to comprehensively analyze the gene-proximal regions of Arabidopsis thaliana and Zea mays. We identified three groups of PLMs in each region for each species and highlighted conserved PLMs in both species, particularly in the 3′-proximal region. Moreover, we showed that PLMs occurred at both transcription factor and microRNA binding sites in each group with specific positional distributions for each region. Although the majority of PLMs were unassigned PLMs, it is noteworthy that some were supported by MNase-defined cistrome occupancy analysis. Enrichment analysis also revealed that unassigned PLMs provide functional predictions distinct from those inferred by the other PLMs. Using PLMs that occured at transcription factor binding sites, we inferred the regulatory network of a poorly characterized Z. mays-specific gene family, paving the way for further characterization of its individual members.