The host-selective toxins of Alternaria show a pattern of disjunct taxonomic distribution similar to the Cochliobolus host-selective toxins, i.e., production of a particular HST is typically restricted to specific strains (pathovars) or species. Compared to other groups of fungi, these two genera appear to have a particularly dynamic capacity to acquire new secondary metabolite potential, which they have successfully exploited to colonize new plant pathogenic niches. The

mechanistic basis of the generation of the extraordinary metabolic diversity in Cochliobolus and Alternaria, and more Epoxomicin generally in the filamentous fungi, is not clear. The most plausible explanations are horizontal gene transfer and/or gene duplication followed by rapid divergence and rapid loss. Horizontal gene transfer has become increasingly accepted as an explanation for many examples of disjunct distribution of secondary metabolites and their genes. Clustering of pathway genes, a common observation, would facilitate horizontal transfer, and trans-species hyphal fusion provides a mechanism of DNA transfer [32–38]. Horizontal transfer is neither supported nor Selleck Caspase Inhibitor VI refuted by the example of Mdivi1 supplier HC-toxin described in this paper, because the two genera are so closely related. It is equally plausible that

a common ancestor of Alternaria and Cochliobolus produced HC-toxin, and this trait was lost from most of the species in the two genera. It is now possible to correlate genes and metabolites for three cyclic tetrapeptides of the HC-toxin family in three fungal species. A. jesenskae and C. carbonum both make HC-toxin, and their orthologous NRPS genes are 82% identical. F. incarnatum makes a chemically related molecule,

apicidin, Epothilone B (EPO906, Patupilone) and its cognate NRPS (APS1) is 44% identical to HTS1. The known genes in common among the three pathways are HTS1, TOXA, TOXC, TOXD, TOXF, and TOXE. Apicidin does not contain any D amino acids besides D-proline (or D-pipecolic acid), whose production from L-proline is presumably catalyzed by the epimerase domain of APS1, and therefore an alanine racemase (TOXG) is not needed for its biosynthesis [14]. The TOX2 cluster of C. carbonum contains a gene for a fatty acid synthase beta subunit (TOXC) and one for the alpha subunit (TOXH). The apicidin cluster does not contain a beta subunit gene. Either apicidin biosynthesis uses the housekeeping beta subunit, or, more likely, the gene for the dedicated beta subunit is elsewhere in the genome. The family of cyclic peptides related to HC-toxin has seven members (from seven fungi in the Sordariomycetes and Dothideomycetes) [5]. The biosynthetic genes for the other members have not yet been characterized.