Abstracts |
Arbuscular mycorrhizal symbiosis (AMS), a widespread mutualistic association of land plants and fungi1, is predicted to have arisen once, early in the evolution of land plants2,3,4. Consistent with this notion, several genes required for AMS have been conserved throughout evolution5 and their symbiotic functions preserved, at least between monocot and dicot plants6,7. Despite its significance, knowledge of the plants' genetic programme for AMS is limited. To date, most genes required for AMS have been found through commonalities with the evolutionarily younger nitrogen-fixing Rhizobium legume symbiosis (RLS)8 or by reverse genetic analyses of differentially expressed candidate genes9. Large sequence-indexed insertion mutant collections and recent genome editing technologies have vastly increased the power of reverse genetics but selection of candidate genes, from the thousands of genes that change expression during AMS, remains an arbitrary process. Here, we describe a phylogenomics approach to identify genes whose evolutionary history predicts conservation for AMS and we demonstrate the accuracy of the predictions through reverse genetics analysis. Phylogenomics analysis of 50 plant genomes resulted in 138 genes from Medicago truncatula predicted to function in AMS. This includes 15 genes with known roles in AMS. Additionally, we demonstrate that mutants in six previously uncharacterized AMS-conserved genes are all impaired in AMS. Our results demonstrate that phylogenomics is an effective strategy to identify a set of evolutionarily conserved genes required for AMS. |