Number of Candidate Effector Genes in Accessory Genomes Differentiates Pathogenic From Endophytic Fusarium oxysporum Strains

The fungus Fusarium oxysporum (Fo) is widely known for causing wilt disease in over 100 different plant species. Endophytic interactions of Fo with plants are much more common, and strains pathogenic on one plant species can even be beneficial endophytes on another species. However, endophytic and beneficial interactions have been much less investigated at the molecular level, and the genetic basis that underlies endophytic versus pathogenic behavior is unknown. To investigate this, 44 Fo strains from non-cultivated Australian soils, grass roots from Spain, and tomato stems from United States were characterized genotypically by whole genome sequencing, and phenotypically by examining their ability to symptomlessly colonize tomato plants and to confer resistance against Fusarium Wilt. Comparison of the genomes of the validated endophytic Fo strains with those of 102 pathogenic strains revealed that both groups have similar genomes sizes, with similar amount of accessory DNA. However, although endophytic strains can harbor homologs of known effector genes, they have typically fewer effector gene candidates and associated non-autonomous transposons (mimps) than pathogenic strains. A pathogenic ‘lifestyle’ is associated with extended effector gene catalogs and a set of “host specific” effectors. No candidate effector genes unique to endophytic strains isolated from the same plant species were found, implying little or no host-specific adaptation. As plant-beneficial interactions were observed to be common for the tested Fo isolates, the propensity for endophytism and the ability to confer biocontrol appears to be a predominant feature of this organism. These findings allow prediction of the lifestyle of a Fo strain based on its genome sequence as a potential pathogen or as a harmless or even beneficial endophyte by determining its effectorome and mimp number.