Differential loss of effector genes in three recently expanded pandemic clonal lineages of the rice blast fungus

Abstract

Background: Understanding the mechanisms and timescales of plant pathogen outbreaks requires a detailed genome-scale analysis of their population history. The fungus Magnaporthe (Syn. Pyricularia) oryzae-the causal agent of blast disease of cereals- is among the most destructive plant pathogens to world agriculture and a major threat to the production of rice, wheat, and other cereals. Although M. oryzae is a multihost pathogen that infects more than 50 species of cereals and grasses, all rice-infecting isolates belong to a single genetically defined lineage. Here, we combined the two largest genomic datasets to reconstruct the genetic history of the rice-infecting lineage of M. oryzae based on 131 isolates from 21 countries.

Results: The global population of the rice blast fungus consists mainly of three well-defined genetic groups and a diverse set of individuals. Multiple population genetic tests revealed that the rice-infecting lineage of the blast fungus probably originated from a recombining diverse group in Southeast Asia followed by three independent clonal expansions that took place over the last ~ 200 years. Patterns of allele sharing identified a subpopulation from the recombining diverse group that introgressed with one of the clonal lineages before its global expansion. Remarkably, the four genetic lineages of the rice blast fungus vary in the number and patterns of presence and absence of candidate effector genes. These genes encode secreted proteins that modulate plant defense and allow pathogen colonization. In particular, clonal lineages carry a reduced repertoire of effector genes compared with the diverse group, and specific combinations of presence and absence of effector genes define each of the pandemic clonal lineages.

Conclusions: Our analyses reconstruct the genetic history of the rice-infecting lineage of M. oryzae revealing three clonal lineages associated with rice blast pandemics. Each of these lineages displays a specific pattern of presence and absence of effector genes that may have shaped their adaptation to the rice host and their evolutionary history.

Keywords: Cereals; Effectors; Fungi; Genomes; Infectious diseases; Pandemics; Pathogens; Plants; Population history; Rice.

Fig. 1.

Genetic clustering of Magnaporthe oryzae reveals three defined groups and a diverse set of individuals. The pairwise relatedness between M. oryzae samples (X and Y) was estimated using f3-outgroup statistics of the form f3(X, Y; outgroup), which measures the amount of shared genetic history (genetic drift) between X and Y after the divergence from an outgroup (M. oryzae strain from Setaria). The hierarchical clustering is based on f3-scores resulting from f3-outgroup statistic calculations. Darker colors indicate more shared drift

Fig. 2.

Geographic location of Magnaporthe oryzae isolates shows global distribution of defined genetic groups (II–III) and a preferential Southeast Asian location for the diverse group (I). a Dendrogram showing the hierarchical clustering based on pairwise f3 values (same as Fig. 1). Prefixes of the isolate names correspond to the database source: G = Gladieux et al., 2018 [31]; Z = Zhong et al., 2018 [32]. b Country of origin for M. oryzae isolates. The overall size of the boxes represents the total number of samples within each genetic group. The size of each internal box is proportional to the number of isolates per country. Countries are represented as three-letter codes (ISO 3166-1 alpha-3): BDI = Burundi, BRA = Brazil, CHN = China, CIV = Côte d’Ivoire, COL = Colombia, ESP = Spain, GHA = Ghana, HUN = Hungary, IND = India, JPN = Japan, LAO = Lao People’s Democratic Republic, MAR = Morocco, MDG = Madagascar, MLI = Mali, NPL = Nepal, PHL = Philippines, PRT = Portugal, SUR = Suriname, THA = Thailand, TWN = Taiwan, Province of China, USA = United States of America. c Geographical origin of samples used in this study. A random jitter was used on the coordinates of geographical-close samples for better visualization

Fig. 3.

Magnaporthe oryzae population structure is driven by recombination and global clonal expansions. a Phylogenetic network showing the three well-defined groups (green, blue, and red) and the diverse set of individuals (orange) from Fig. 1. b Within-population comparisons of nucleotide diversity measured as π. c Recombination proxy calculated by dividing the number of violations of the four-gamete test by the total number of SNPs. d Genetic distances between groups measured as fixation indices (F_st). The box colors depict the pairwise comparisons between groups. e Tajima’s D

Fig. 4.

Clonal expansions of Magnaporthe oryzae took place in the last 200 years. Bayesian tip calibrated phylogenetic tree using individuals belonging to clonal lineages. Average, and HPD 95% confidence intervals are shown in calendar years. The Bayesian posterior probability is shown in red for nodes leading to the clonal lineage expansions

Fig. 5.

Patterns of allele frequency sharing identify introgression between a Chinese Magnaporthe oryzae subpopulation and clonal lineage II. D-statistics using three different phylogenetic configurations (depicted as colored inset trees). aD (outgroup, orange; blue, red). bD (outgroup, orange; green, red). cD (outgroup, orange; green, blue). In all cases, a M. oryzae strain from wheat was used as an outgroup and a fixed individual was selected as representative from each clonal lineage (blue, orange, red). Points represent D-statistic tests for each of the 22 individuals assigned to the diverse clade (orange), and lines depict 95% confidence intervals. Purple dots in b and c correspond to Chinese individuals CH1016 and HB-LTH18, which are the closest individuals to the green clonal lineage

Fig. 6.

Rice blast genetic lineages vary in the number and patterns of presence and absence of candidate effector genes. a Clonal lineages carry a reduced repertoire of effector genes compared with the diverse group I. b The box-and-whisker plots show the distribution of effector number per isolate for each genetic group. Asterisks represent a p value < 0.01 for a one-tailed Wilcoxon non-parametric test. c The dendrogram shows the clustering based on f3-outgroup statistic (as in Fig. 1). Light and dark colors on the rows represent absence and presence of effectors, respectively. Rows were grouped using a hierarchical clustering algorithm. Labels in green and blue font denote effectors missing in clonal groups II and III, respectively