Comparative genomics and prediction of genome features in Fusarium oxysporum f. sp. vasinfectum isolates to aid in the identification of virulence factors specific for cotton

Abstract

Fusarium oxysporum f. sp vasinfectum (Fov), the causal agent of Fusarium wilt of cotton (Gossypiym species), is causing a devastating decline in worldwide cotton production. Although many factors associated with pathogenicity of F. oxysporum species have been revealed, individual members in the F. oxysporum species complex (FOSC) display host specificity. This indicates that characterized virulence factors may not function similarly in other isolates virulent on different plant hosts. Thus, individual Fusarium isolates may need to be evaluated independently on each host plant. The genomes of five Fov isolates, including a representative of the highly virulent race 4 genotype, were sequenced at a high coverage to obtain reference quality assemblies. Various bioinformatic approaches were employed to identify genomic features of Fov. In Chapter 2, the core (CC) and accessory (AC) components of the genome of the five Fov isolates were identified. Scaffolds indicated as part of the AC were searched for syntenic regions between the five Fov genomes. Except for the Fov race 4 (89-1A) genome, AC regions of the remaining four Fov genomes shared syntenic loci. Homologs in Verticillium species that share the same cotton host with Fov that were confirmed to be involved in virulence were used to narrow down a key region in the ACs for pathogenicity on cotton. As a result, about 360 kb of the genomes of TF1, LA3B, and 14-004 was inferred to be important for conferring pathogenicity on cotton. In addition, the most virulent strain, Fov 89-1A had unique genomic features; a larger genome size and proportion of repetitive elements than the other four Fov genotypes included in the study. In Chapter 3, comparative genomics was applied to identify potential secreted protein sets from six Fov genomes (cotton) and ten publicly available F. oxysporum genomes to uncover putative secreted proteins which could be contributing to virulence (fungal effectors) on cotton. Among the Fov isolates, the highly virulent race 4 is of particular interest since no highly resistant cotton cultivars are commercially available and it is able to infect plants without the presence of nematodes. To narrow down unique putative effector gene specific for Fov, in silico analysis was universally applied to identify proteins containing a signal peptide and exclude those with transmembrane domains. After filtering all the secretomes to only proteins less than 300 amino acids in length, the refined secretome of each individual FOSC was analyzed using BLASTP against of the SSP set of 687 proteins from Fov race 4 (89-1A) to cluster SSPs. A total of 78 SSPs were shared between at least three isolates from cotton but were absent in all the other FOSC genomes not pathogenic on cotton. Four mutants of SSP genes, F02527, F02627, F07669/07679 and F11624 were chosen for gene deletion to evaluate their contribution to fungal pathogenicity on cotton. Three pathogenicity assays, stem, hydroponic, and pot, were conducted with the Pima cultivar DP744 to assess whether these SSPs had any influence on the virulence to cotton. SSP F11624 was confirmed to serve as a virulence factor on cotton and the gene expression was up-regulated during infection. In Chapter 4, in silico analysis was performed to identify secondary metabolite (SM) clusters to elucidate their biological significance within the FOSC. Twenty-one genomes of members of the FOSC including six Fov genomes were queried to identify putative secondary metabolite biosynthetic gene clusters. Core and accessory regions of all the FOSC genomes were defined using the well-established Fol 4287 genome as a reference for chromosomal structure. Altogether, comparative analysis of the SM cluster types revealed 28 clusters in the core regions are shared in all 21 FOSC genomes. Furthermore, to understand whether specific clusters exist in the Fov isolates including race 7 (NRRL 25433), the type of cluster, uniqueness, and location in sub-telomeric regions were manually curated. The putative siderophore cluster in the middle of chromosome 11 (CHR11) harboring the core gene encoded by NRPS6 was observed in all Fov and Fol 4287 isolates and has high conservation. Additional putative siderophore clusters encoding a homolog of NRPS1 on CHR12 were found in both Fov 14-004 and LA3B isolates. The Fov 14-004 genome contains many unique SM clusters not observed in the other Fov genomes, and four Fov isolates, TF1, 89-1A, 25433, and LA3B had one to three SM clusters encoded in the accessory region. In conclusion, the Fov genomes provide a necessary resource for comparative genomic analyses to identify genes and genomic features that are involved in pathogenicity on cotton and may ultimately be used to develop improved management strategies. Elucidating the function of many putative SSPs could be valuable in identifying novel and isolate-specific effector genes in Fov and more importantly could facilitate identification of corresponding resistance gene(s) in cotton for use in breeding programs. Such fundamental studies are necessary for developing novel methods to control the threat caused by Fusarium and SMs for agricultural biosecurity.