Marlene Castro, Salim Charaniya, Siguang Sui, and Wei-Shou Hu. Department of Chemical Engineering and Materials Science, University of Minnesota, 151 Amundson Hall, 421 Washington Ave SE, Minneapolis, MN 55455
Biosynthesis of secondary metabolites in Streptomycetes is regulated by a multilayer, dynamic network, interconnected to morphological differentiation. In such a complex regulatory network, many smaller overlapping networks exist. However, there is little information on the regulatory elements integrating such networks. In order to uncover potential regulatory interactions, we combined large scale gene expression data with genome and functional information. More than 500 transcriptome profiles of Streptomyces coelicolor, available from in-house and public data, were compiled. The multiplatform dataset probes the temporal dynamics of wild-type and regulatory mutant strains, as well as growth under different media and stress conditions. By combining temporal transcriptome profiles with other similarity features, such as intergenic distance and synteny, a whole-genome operon map was predicted. This operon prediction was used to validate clusters of correlated genes. Sets of correlated genes were used to predict transcriptional networks. Some of the resulting networks represent functional modules, integrated by functionally coherent genes, such as those for fatty acid synthesis. The upstream regions of putative regulons were tested for the occurrence of overrepresented sequence motifs. Such an integrated approach, combining gene expression with sequence and function data, will help in unveiling regulatory network interactions.
