Contrasting levels of genetic predictability in the evolution of resistance to major classes of fungicides.

The evolution of resistance has been seen across all major classes of xenobiotics, including antimicrobial drugs and agricultural pesticides. This repeated emergence of resistance is a case of phenotypic parallel evolution, but often the parallelism extends to the molecular level too, with multiple species gaining the same mutation in response to the same chemical treatment. We review the degree of repeatability in target-site resistance mutations affecting different classes of site-specific agricultural fungicides used in crop protection, comparing the extent to which resistance in different pathogen species has evolved via the same or different mutations. For all major fungicide target sites, substantial levels of molecular parallel evolution can be seen, with at least one mutation recurring in over 50% of species. Target-site mutations appear to be most repeatable in cytochrome b, target site of quinone-outside inhibitor fungicides, and least predictable for CYP51, target site of the azoles. Intermediate levels of repeatability are seen for the MBC target site β-tubulin, and the SDHI target site succinate dehydrogenase. Repeatability may be lower where there are selective trade-offs between resistance and pleiotropic fitness penalties, or differing levels of cross-resistance across members of a fungicide class; or where single mutations confer only partial resistance, and epistatic interactions between multiple mutations result in a rugged fitness landscape. This affects the predictive power of in vitro mutation studies, and has practical implications for resistance monitoring strategies and diagnostic methods.