UC Berkeley

Powdery mildews are obligate biotrophic fungi that rely entirely on their host to complete the life cycle. They establish metabolic sinks at infection sites and rewire plant metabolism for nutrient acquisition. These pathogens have lost many primary and secondary metabolism genes but possess an extensive effectorome to manipulate host function. This work uses the model system Golovinomyces orontii and Arabidopsis thaliana to explore the interactions between host and pathogen, focusing on lipid metabolism, plastoglobule function, and genomic characteristics.First, the mechanism for host response to fungal lipid demand was investigated. Previous lipid profiling revealed that powdery mildew induces a ~3 fold increase in triacylglycerols (TAGs) in host infected tissues. Arabidopsis DIACYLGLYCEROL ACYLTRANSFERASE 3 (DGAT3) was then identified as a critical player responsible for TAG accumulation. I determined that DGAT3 is chloroplast-localized and that concurrent chloroplast degradation at the infection site fuels TAG synthesis via DGAT3. Knocking down or knocking out host DGAT3 reduces fungal spore production. These findings support a model in which DGAT3-mediated TAG accumulation in chloroplasts promotes fungal reproduction at the expense of thylakoid membrane lipids. In Chapter 3, the role of plastoglobules in the powdery mildew-host interaction is explored. Initially, I demonstrated that DGAT3 localizes within chloroplasts to plastoglobules (PGs), the lipid-containing compartments attached to chloroplast thylakoid membranes. PGs are often induced by environmental stress yet their functions in pathogen defense remain unclear. Upon powdery mildew infection, DGAT3 proteins and PG formation were found to be induced. Furthermore, genetic analysis identified key PG proteins, including Fibrillin (FBN)1a, FBN1b, and FBN2, that limit powdery mildew proliferation. Additional evidence of effector-host protein interactions also supports the involvement of PGs during powdery mildew infection. A powdery mildew effector was identified as a contributor to virulence and this effector specifically interacts with the PG protein FBN2. The interactions were validated using yeast-two-hybrid, co-immunoprecipitation and split-luciferase complementation assays. These results underscore PGs as essential but underappreciated hubs in plant-pathogen interactions. The final two chapters apply bioinformatic tools to analyze the powdery mildew genome. Powdery mildew genomes can contain up to 85% transposable elements (TEs), making the accurate annotation of genomes challenging. In a collaborative project, I co-developed TEtrimmer, a python-based tool to automate TE curation through multiple sequence alignment clustering, boundary definition, and classification. TEtrimmer offers robust improvements over existing methods and can be applied to other genomes including non-model organisms. Finally, to understand the genomic features of powdery mildews, I utilized the genome assembly and transcriptome data acquired through DOE JGI Community Science Project #1657 in which PacBio long-read sequencing data coupled with HiC proximity data to obtain a near-chromosome-level G. orontii MGH1 genome assembly of 211 Mbp from 123 scaffolds, with 20 large scaffolds covering 96.6% of the genome. The genome size and gene content were approximately double that of other powdery mildew genomes. Genomic analysis, including synteny mapping, k-mer histograms, and RNA sequencing, revealed that G. orontii MGH1 underwent a recent whole-genome duplication (WGD). To rule out contamination by a closely related strain, G. orontii MGH1’s mitochondrial genome and ITS region sequences, nuclear DNA content, and gene sequencing from single colony isolations were performed. All results confirmed the absence of contamination. Synonymous substitution rate (dS) values suggest that the WGD is very recent, yet it has already led to divergent expression in 72% of duplicated genes. Notably, a small set of gene pairs showed signs of positive selection, with candidate secreted effector proteins (CSEPs) under significantly higher selective pressure and a rapid loss of duplicate copies. Following WGD, selective gene loss has left a unique set of 828 single-copy genes, 6.5% of total, that are likely essential for pathogen survival and represent dosage-dependent critical enzymes. These include genes in functional categories previously identified for genes retained as single-copy following WGD such as replication, transcription, and translation; and cell morphology and cell cycle regulation. In addition, single copy genes include a number of metabolic enzymes whose regulation is critical to a given biosynthetic pathway or to coordination among biosynthetic pathways. Interrogation of these single-copy genes using spray-induced gene silencing found 9 out of 10 to significantly impact powdery mildew proliferation on Arabidopsis. This work provides insights into the genomic adaptations of powdery mildews and shed light on the genetic mechanisms that underpin their success in host colonization.