Analysis of the MLO protein family and characterization of a mlo-virulent Bgh isolate

The Mildew Locus O (MLO) protein family is best-known for its prominent role in the plant-powdery mildew interaction. Loss-of-function mutation of barley Mlo, respective wheat TaMlo1 orthologues, Arabidopsis AtMLO2, AtMLO6 and AtMLO12, and of tomato SlMlo1 (to name but a few) results in durable broad-spectrum resistance to powdery mildew across these diverse plant species. However, the MLO protein family is more complex. In Arabidopsis, for instance, 15 MLO proteins are present, and in addition to the three above-mentioned MLOs, AtMLO7 is required for proper pollen tube perception of the female gametophyte, and AtMLO4 plus AtMLO11 are involved in root thigmomorphogenesis and respective loss-of-function mutations lead to an aberrant root curling phenotype upon a tactile stimulus.I further investigated the diversity of the MLO family across eukaryotic species. First, I performed comprehensive phylogenetic analysis to resolve the relationship of the MLO subfamilies (clades) and to trace the origin of the MLO protein. By this work, I found that the MLO protein family likely originated in a single-celled eukaryotic organism that was the ancestor of both the plant kingdom and oomycetes, and diversified into seven MLO clades in land plants. Next, I addressed the question whether the MLO clade diversity reflects functional diversity. To this end, I explored the yet uncharacterized MLO clades II and VI in Arabidopsis based on available expression data and co-expression networks, i.e. AtMLO1 and AtMLO15 (clade II) for root hair development and AtMLO3 (clade VI, closely related to AtMLO2) for biotic stress response to various pathogens of fungal, oomycete and bacterial origin. However, I could not identify aberrant phenotypes in respective T-DNA insertion mutants. Furthermore, I characterized clade III (AtMLO5, AtMLO7, AtMLO8, AtMLO9, and AtMLO10) single mutants and various mutant combinations thereof for fertility and analyzed mlo7 and mlo8 mutants for abiotic and biotic stress responses. Here, I found that only mlo7 mutants are clearly affected in fertility (seed number) by 10-20%, while mlo8 mutants might be somewhat more susceptible to the adapted pathogen Hyaloperonospora arabidopsidis (Hpa). Taken together, the data suggests that the MLO clades are functionally distinct, with potential further specialization within some of the clades.Additionally, I investigated the role of further plant MLO proteins in the interaction with powdery mildews. Here, I used Arabidopsis mlo2 mlo6 mlo12 triple mutant plants that heterologously express one of four selected rose (Rosa hybrida) MLOs (RhMLO1-RhMLO4) and found that RhMLO1 and potentially RhMLO3 and RhMLO4 partially restore susceptibility to the powdery mildew fungus Golovinomyces orontii. A TALEN- and Cas9-based proof-of-principle approach to introduce powdery mildew resistance by targeted gene knock-out of MLO genes in barley and Arabidopsis was not successful.Finally, I generated and characterized a barley powdery mildew (Blumeria graminis f.sp. hordei, Bgh) isolate, termed Super-virulent K1 (SK1), that is partially virulent on mlo mutant plants. I quantified the extent of this limited virulence in comparison with the mlo-avirulent parent isolate Bgh K1 and demonstrated that Bgh SK1 shows about 20-30% penetration success on barley mlo mutants and successfully completes its life cycle on this genotype, as evidenced by the formation of macroscopically visible colonies. A clear difference between the two fungal isolates on the non-hosts wheat and Arabidopsis was not found. Interestingly, Bgh SK1 is hyper-virulent on mlo-5 ror1 and mlo-5 ror2 double mutants (ror1 and ror2 are suppressors of mlo resistance). On these double mutants penetration rates reach levels like on the wild type Mlo genotype cultivar Ingrid, while Bgh K1 only shows 10-20% entry rate (ΔK1/SK145%). I performed whole transcript shotgut sequencing (RNAseq) analysis of both isolates on both Ingrid and mlo as host plants at four time points (6, 12, 18, and 24 hours post inoculation (hpi)) to uncover the molecular basis for this phenotype. In preliminary analysis of the data I uncovered >600 genes being at least 2-fold upregulated in Bgh SK1 on barley mlo at 6 hpi compared to wild type as the host. The data suggests that Bgh SK1 overcomes mlo resistance by breaking a pre-penetration resistance pathway distinct from but potentially related to pathways involving Ror1 or Ror2, coinciding with significant transcriptomic adaptations at early infection stages to the new host barley mlo