Plant natural promoters are always very long and contain many different promoter motifs, providing complex expression patterns, while synthetic promoters can be constructed to be very short in sequence and very strong in promoter strength. Bioinformatics-assisted de novo promoter motif discovery searches for statistically overrepresented motifs without the inclusion of biological information, leading to limited prediction efficiency. To overcome this limitation, we have developed a novel ensemble approach by mapping the motifs detected by a set of selected bioinformatics tools back to the promoter sequences and looking for overlapping motif regions among the detected motifs. Using this approach, we searched and identified novel constitutive promoter motifs from the soybean genome. Seven user-friendly bioinformatics tools, including BioProspector, CisGenome, HOMER, MEME, MotifSuite, RSAT Plants, and Weeder were employed for the de novo discovery of constitutive motifs among 11 published soybean constitutive promoters. A total of 62 promoter motifs were detected among the 11 soybean constitutive promoters by at least four of the seven bioinformatics tools. A tetramer (4×) of each promoter motif was cloned in front of the minimal 35S promoter driving GUS reporter gene expression, and used for tobacco leaf agroinfiltration and stable Arabidopsis transformation. Quantitative GUS activity assays following transient tobacco leaf agroinfiltration identified 26 of the 62 promoter motifs that drove GUS expression significantly higher than the basal level conferred by the minimal 35S promoter. Histochemical GUS analysis of stable transgenic Arabidopsis seedlings found that 16 of the 26 promoter motifs were 19 ~ 60 bp in length and exhibited constitutive expression with variable promoter strength, and 7 of the 26 promoter motifs showed strong constitutive expression which was comparable to (slightly weaker than) the 35S promoter. Thus, these novel constitutive motifs can be used to drive constitutive gene expression in dicot species.
