Bioinformatically informed design of synthetic mammalian promoters

Abstract

Synthetic promoters have been developed in a number of different organisms and are capable of mediating specific and enhanced levels of gene expression. Typically, cis-regulatory regions from a few genes are randomly combined to generate a synthetic promoter library, and the sequences with the highest activity are selected for in target cell lines. Here we describe a novel approach that can be employed in the construction of synthetic promoters. Specifically, we use gene expression profiles obtained from microarray datasets to select the cis-regulatory elements that comprise the synthetic promoter library. By adopting this approach, we were able to construct several promoters that could specifically mediate gene expression in colorectal cancer cells. We develop a new selection criteria based on the observed transcriptome of target cells, the frequency that identified cis-regulatory sequences occur in identified gene modules, and the length of identified cis-regulatory regions. Our method allows for the generation of synthetic promoter libraries with increased level of specificity and facilitates the selection of promoters that are highly active only under predefined gene expression profiles.