Fungi are diverse eukaryotic microorganisms with pivotal roles in ecosystems, but also notorious pathogens causing significant economic losses. Understanding the mechanisms underlying fungal pathogenicity is crucial for devising effective strategies to mitigate their negative impact. Fungi utilize various mechanisms to infect plants, including secreting effector proteins that promote virulence on susceptible hosts or trigger immune responses (i.e. avirulence) in resistant plants carrying resistance genes. To retain their virulence properties and abolish their avirulence ones, effectors often accumulate mutations in their coding sequence or are entirely deleted from the pathogen genome. Thus, knowing the mutability of effectors is crucial to selecting the right resistance genes when breeding for durable resistance. The tomato pathogen Cladosporium fulvum showcases skewed types of mutations in its effector genes to overcome resistance in tomato, which depending on the effector range from point mutations to complete gene deletions. This observation indicates that the type and frequency of mutations accumulating in effectors are to an extent driven by the genes’ genomic location and the propensity of these genomic regions to structural variations (SVs). However, the genome architecture of C. fulvum and its landscape of SVs were never investigated. Similarly, SVs are thought to be an important source of adaptation in the grape powdery mildew fungus Erysiphe necator, economically the most important foliar pathogen on this crop, in which gene duplications have been already associated with the development of resistance to fungicides. In this dissertation, I seek to address these gaps by obtaining high-quality chromosome-level genome assemblies and annotations for C. fulvum and E. necator, and investigating the extent to which their genomic architecture and SVs contribute to their evolution and pathogenicity. The genome of C. fulvum is organized into a variable number of 13 to 15 chromosomes, as two of them are dispensable for fungal growth and pathogenicity. The chromosomes of C. fulvum exhibit a peculiar ‘checkerboard’ pattern of gene-rich/repeat-poor regions, interspersed with gene-poor/repeat-rich regions. Comparisons with an additional five isolates of C. fulvum revealed that nearly all SVs corresponded to insertions or deletions in regions rich in transposable elements (TEs). Notably, three SVs that were likely induced by TEs effected the deletion of the effector genes Avr9, Avr5, and Avr4E, thereby mediating an escape of pathogen recognition by the cognate Cf-9, Cf-5 and Cf-4E resistance genes in tomato. In this dissertation we also investigated the landscape of alternative splicing (AS) events in C. fulvum genes during a complete infection cycle. The analysis showed that nearly 40% of the protein-coding genes in C. fulvum were AS at some stage during the infection process, suggesting that AS could have a role in finetuning infections of the host. Comparison of the location of the AS genes in the genome of C. fulvum revealed that AS genes are more abundant in repeat-rich core chromosomes, and exhibit significant longer 5’ intergenic regions richer in repetitive DNA compared to non-AS genes, indicating that the genome organization could have an effect on the occurrence of AS. Our studies on the grape powdery mildew pathogen E. necator showed that its genome is organized into 11 chromosomes which do not exhibit large-scale compartmentalization into gene-rich and repeat-rich regions. A total of 13.1% of the genes in E. necator were predicted to be duplicated and were particularly enriched for genes encoding candidate effectors. Comparative analysis among six isolates of E. necator revealed a total of 122 genes that varied in their copy numbers. One of these varied from 1 to 31 copies and encoded a putative secreted carboxylesterase (CE), which is a member of a novel family of CEs that is unique to powdery mildew fungi. Next to the nuclear genome, the organization of the mitochondrial genome of E. necator and that of other powdery mildew fungi were also analyzed. Comparative genomics among E. necator and three other species of powdery mildew fungi revealed a wide variation of mitochondrial genome sizes, ranging from 109.8 kb in B. graminis f. sp. tritici to 332.2 kb in G. cichoracearum, which has the largest mitochondrial genome of a fungal pathogen reported to date. Finally, the introns of the cytochrome b gene, which encodes for the target site of QoI fungicides, of powdery mildew fungi contained rare open reading frames encoding reverse transcriptases that were likely acquired horizontally. Collectively, the results presented in this dissertation reveal new evolutionary aspects of the genomes of C. fulvum and E. necator, and highlight the importance of genome organization and genomic structural variations in overcoming host resistance.
