Rapid turnover of effectors in grass powdery mildew (Blumeria graminis)

Grass powdery mildew (Blumeria graminis, Ascomycota) is a major pathogen of cereal crops and has become a model organism for obligate biotrophic fungal pathogens of plants. The sequenced genomes of two formae speciales (ff.spp.), B.g. hordei and B.g. tritici (pathogens of barley and wheat), were found to be enriched in candidate effector genes (CEGs). Similar to other filamentous pathogens, CEGs in B. graminis are under positive selection. Additionally, effectors are more likely to have presence-absence polymorphisms than other genes among different strains.

Here we identified effectors in the genomes of three additional host-specific lineages of B. graminis (B.g. poae, B.g. avenae and B.g. infecting Lolium) which diverged between 24 and 5 million years ago (Mya). We found that most CEGs in B. graminis are clustered in families and that most families are present in both reference genomes (B.g. hordei and B.g. tritici) and in the genomes of all three newly annotated lineages. We identified conserved protein domains including a novel lipid binding domain. The phylogenetic analysis showed that frequent gene duplications and losses shaped the diversity of the effector repertoires of the different lineages through their evolutionary history. We observed several lineage-specific expansions where large clades of CEGs originated in only one lineage from a single gene through repeated gene duplications. When we applied a birth-death model we found that the turnover rate (the rate at which genes are deleted and duplicated) of CEG families is much higher than for non-CEG families. The analysis of genomic context revealed that the immediate surroundings of CEGs are enriched in transposable elements (TE) which could play a role in the duplication and deletion of CEGs.

The CEG repertoires of related pathogens diverged dramatically in short evolutionary times because of rapid turnover and of positive selection fixing non-synonymous mutations. While signatures of positive selection on effector sequences are the expected outcome of the evolutionary "arms race" between pathogen and plant immune system, it is more difficult to infer the mechanisms and evolutionary forces that maintained an extreme turnover rate in CEG families of B. graminis for several millions of years.