Genome-wide association studies reveal the complex genetic architecture of DMI fungicide resistance in Cercospora beticola

Cercospora leaf spot is the most important disease of sugar beet worldwide. The disease is caused by the fungus Cercospora beticola and is managed principally by timely application of fungicides including those of the sterol demethylation inhibitor (DMI) class. However, reliance on DMIs has caused an increase in resistance to this class of fungicides in multiple C. beticola populations. To better understand the genetic and evolutionary basis for resistance in C. beticola, a genome-wide association study (GWAS) and selective sweep analysis were conducted for the first time in this fungal plant pathogen. We performed whole genome resequencing of 190 C. beticola isolates predominantly from North Dakota and Minnesota that were phenotyped for sensitivity to tetraconazole, the most widely used DMI fungicide in this region. GWAS identified mutations in genes associated with DMI fungicide resistance including a Regulator of G-protein Signaling (RGS) protein, an ATP-binding cassette (ABC) pleiotropic drug resistance transporter, a dual-specificity tyrosine phosphorylation-regulated kinase (DYRK), and a gene annotated as a hypothetical protein. A SNP upstream of CbCYP51, the gene encoding the target of DMI fungicides, was also identified via GWAS. Haplotype analysis of CbCYP51 identified a synonymous mutation (E170) in high linkage disequilibrium with the upstream SNP, and multiple non-synonymous mutations (L144F, I387M and Y464S) associated with DMI resistance. Additionally, a putative codon bias effect for the L144F substitution was identified that generated different resistance potentials. We also identified a CbCYP51 paralog in C. beticola, CbCYP51-like, with high protein homology to CYP51C found uniquely in Fusarium species but CbCYP51-like does not appear to influence DMI sensitivity. Genome-wide scans of selection showed that several of the GWAS mutations for fungicide resistance resided in regions that have recently undergone a selective sweep. Using radial plate growth on selected media as a fitness proxy, we did not find a trade-off associated with DMI fungicide resistance suggesting that resistance mutations can persist in C. beticola populations. Taken together, we show that population genomic data from a crop pathogen can allow the identification of mutations conferring fungicide resistance and inform about their origins in the pathogen population.