In silico prediction and characterization of secondary metabolite biosynthetic gene clusters in the wheat pathogen Zymoseptoria tritici

BACKGROUND: Fungal pathogens of plants produce diverse repertoires of secondary metabolites, which have functions ranging from iron acquisition, defense against immune perturbation, to toxic assaults on the host. The wheat pathogen Zymoseptoria tritici causes Septoria tritici blotch, a foliar disease which is a significant threat to global food security. Currently, there is limited knowledge of the secondary metabolite arsenal produced by Z. tritici, which significantly restricts mechanistic understanding of infection. In this study, we analyzed the genome of Z. tritici isolate IP0323 to identify putative secondary metabolite biosynthetic gene clusters, and used comparative genomics to predict their encoded products. RESULTS: We identified 32 putative secondary metabolite clusters. These were physically enriched at subtelomeric regions, which may facilitate diversification of cognate products by rapid gene rearrangement or mutations. Comparative genomics revealed a four gene cluster with significant similarity to the ferrichrome-A biosynthetic locus of the maize pathogen Ustilago maydis, suggesting this siderophore is deployed by Z. tritici to acquire iron. The Z. tritici genome also contains several isoprenoid biosynthetic gene clusters, including one with high similarity to a carotenoid/opsin producing locus in several fungi. Furthermore, we identify putative phytotoxin biosynthetic clusters, suggesting Z. tritici can produce an epipolythiodioxopiperazine, and a polyketide and non-ribosomal peptide with predicted structural similarities to fumonisin and the Alternaria alternata AM-toxin, respectively. Interrogation of an existing transcriptional dataset suggests stage specific deployment of numerous predicted loci during infection, indicating an important role of these secondary metabolites in Z. tritici disease. CONCLUSIONS: We were able to assign putative biosynthetic products to numerous clusters based on conservation amongst other fungi. However, analysis of the majority of secondary metabolite loci did not enable prediction of a cluster product, and consequently the capacity of these loci to play as yet undetermined roles in disease or other stages of the Z. tritici lifecycle is significant. These data will drive future experimentation for determining the role of these clusters and cognate secondary metabolite products in Z. tritici virulence, and may lead to discovery of novel bioactive molecules. ELECTRONIC SUPPLEMENTARY MATERIAL: The online version of this article (doi:10.1186/s12864-017-3969-y) contains supplementary material, which is available to authorized users.