The Evolution of Secondary Metabolic Gene Clusters in Plants

Eukaryotic gene order is not random. In extreme cases, genes become physically linked on the genome, forming tight gene clusters responsible for the production of specialized metabolites. Recently characterized gene clusters in plants produce metabolites associated with stress tolerance, an ecologically important trait as well as a useful one for agriculturalists seeking to develop crops resistant to changing environmental conditions. Additionally, many clustered plant pathways code for small molecules with antimicrobial and anticancer properties. Despite the importance of these gene clusters, the bioinformatic tools that can lead to their discovery have not been developed for plants, and the mechanisms that make and maintain gene clusters in plant genomes are poorly characterized. Examination of 37 plant genomes for the presence of secondary metabolic gene clusters revealed thousands of genes putatively belonging to hundred of distinct gene clusters, with the average plant genome containing 44.6 metabolic gene clusters and 239.1 clustered genes. The number of metabolic gene clusters was variable across species, being high in many Angiosperms and low in algae.  GO terms overrepresented in clustered genes include those for biosynthesis of small molecules. To understand the evolution of gene content in plant gene clusters, we developed a phylogenetics pipeline to reconstruct the gene tree phylogenies for plant metabolic gene families. Preliminary results suggest that clustered genes are either members of large families consisting of largely unclustered genes or are members of small families consisting of only clustered genes present in one or two species. Ongoing analysis of the phylogenomic results will address the following questions: 1) Does gene family expansions correspond to an increase in secondary metabolism? and 2) Are clustered genes derived from duplication of existing protein families or from de novo origination and what is the relative contribution of the two processes in cluster formation?