Population Genomics Provide Insights into the Global Genetic Structure of Colletotrichum graminicola, the Causal Agent of Maize Anthracnose

Abstract

Understanding the genetic diversity and mechanisms underlying genetic variation in pathogen populations is crucial to the development of effective control strategies. We investigated the genetic diversity and reproductive biology of Colletotrichum graminicola isolates which infect maize by sequencing the genomes of 108 isolates collected from 14 countries using restriction site-associated DNA sequencing (RAD-seq) and whole-genome sequencing (WGS). Clustering analyses based on single-nucleotide polymorphisms revealed three genetic groups delimited by continental origin, compatible with short-dispersal of the pathogen and geographic subdivision. Intra- and intercontinental migration was observed between Europe and South America, likely associated with the movement of contaminated germplasm. Low clonality, evidence of genetic recombination, and high phenotypic diversity were detected. We show evidence that, although it is rare (possibly due to losses of sexual reproduction- and meiosis-associated genes) C. graminicola can undergo sexual recombination. Our results support the hypotheses that intra- and intercontinental pathogen migration and genetic recombination have great impacts on the C. graminicola population structure. IMPORTANCE Plant pathogens cause significant reductions in yield and crop quality and cause enormous economic losses worldwide. Reducing these losses provides an obvious strategy to increase food production without further degrading natural ecosystems; however, this requires knowledge of the biology and evolution of the pathogens in agroecosystems. We employed a population genomics approach to investigate the genetic diversity and reproductive biology of the maize anthracnose pathogen (Colletotrichum graminicola) in 14 countries. We found that the populations are correlated with their geographical origin and that migration between countries is ongoing, possibly caused by the movement of infected plant material. This result has direct implications for disease management because migration can cause the movement of more virulent and/or fungicide-resistant genotypes. We conclude that genetic recombination is frequent (in contrast to the traditional view of C. graminicola being mainly asexual), which strongly impacts control measures and breeding programs aimed at controlling this disease.

Keywords: fungal plant pathogen; migration; phylogeography; phytopathology; population genetics; population genomics; recombination.

FIG 1

Worldwide Colletotrichum graminicola sampling. Red circles indicate the number of isolates sampled. Artwork created with mapchart.net.

FIG 2

Population subdivision of Colletotrichum graminicola. (A) Principal-component analysis (PCA). Isolates are colored according to their continent of origin. (B) Minimum spanning network (MSN) showing the multilocus genotypes (MLGs) based on Nei’s genetic distance. The distance between nodes represents the genetic distance between MLGs. Each circle represents a distinct MLG, proportional in size to the number of isolates with the genotype.

FIG 3

Neighbor-net network showing relationships between Colletotrichum graminicola isolates. Red color indicates “migrant” isolates, i.e., isolates in the European clade. Isolates with pathogenic characterization are indicated by asterisks.

FIG 4

Linkage disequilibrium (LD) decay plots of three genetic populations of Colletotrichum graminicola (Brazilian, North American, and European clades). LD was calculated for all pairs of SNPs less than 100 kb apart from each other.

FIG 5

Pathogenic characterization of Colletotrichum graminicola isolates. (a) Graphic representation of 19 isolates in order of virulence (relative to the control strain M1.001) in ascending order. The isolates CBS252.59 and MAFF511343 were nonpathogenic to the maize cultivar evaluated and are not shown in the figure. Pathogenicity assay batch number is shown in parentheses. Isolates marked with an asterisk were significantly different (P < 0.001) from strain M1.001 based on a post hoc Tukey’s test using a Bonferroni correction. (b) Necrotic lesions on maize leaves at 4 days after inoculation with spore suspensions of the isolates.