Analysis of Magnaporthe grisea pathogenicity by insertion mutagenesis and hierarchical metabolomics

The project will utilize high throughput analytical biochemistry techniques to provide new information concerning the infection of cereals by an economically significant plant pathogenic fungus. Fungi cause the world's most serious and devastating crop diseases and this project will provide new information about the process of plant infection by these microbial agents of disease. We will investigate a fungus called Magnaporthe grisea which causes rice blast disease, a disease that destroys enough rice each year to feed 60 million people. We have produced mutants of the fungus that are impaired in plant infection, and these will be subject to metabolome analysis. Metabolomics techniques will allow us to resolve many of the low molecular weight compounds in the cells of the fungus (organic acids, amino acids, sugars, and polyols) and produce metabolite fingerprints for each mutant strain investigated. We will investigate a set of mutants affected in glycogen, lipid and trehalose metabolism that are required for development of turgor pressure in the specialised infection structures produced by the fungus. To break into a rice leaf, the fungus generates enormous pressure that allows the cuticle of the cereal leaf to be broken. We will also investigate a novel set of mutants that are affected in their ability to colonize living plant tissue, resulting in small, or mis-timed disease lesions. The techniques we propose to use allow the simultaneous global analysis of all the small molecules (several hundreds of metabolites) contained within a cell. This comprehensive analysis will ensure that all major areas of metabolism are studied which will allow us to develop new ideas of the precise differences exist between the normal, wild type fungus and each mutant. These complex data require specialised methods for data analysis to determine metabolic differences between strains. In the future, we expect that novel fungicides and disease-control strategies may be developed that disrupt the virulence-associated metabolic processes identified in this study.

The project aims to utilise metabolomics approaches to identify metabolic processes associated with pathogenicity in the fungus Magnaporthe grisea, a major disease of a range of cereals and grasses. The genome sequence of the fungus has been determined and tools are available for generating targeted gene replacement mutants, studying gene expression using genome microarrays, and carrying out detailed cell biological studies of plant infection (for review see Talbot, 2003). M. grisea is being subjected to intensive functional genomics analysis, including large-scale insertion mutagenesis projects. To date, mutants affecting pathogenicity are almost without exception impaired in ability to form infection structures (appressoria) and penetrate host epidermal cells. However, in new mutant screens carried out at Exeter and elsewhere, several new classes of mutant are emerging where the timing of lesion formation and subsequent lesion expansion is impaired. We hypothesize that the corresponding genes may make important contributions to plant tissue colonization and disease symptom formation by M. grisea. Molecular genetic analysis of early-phase infection mutants in M. grisea have largely been carried out ex planta by germination of spores on inert plastic surfaces, providing large synchronous populations of infection structures for biochemical analysis. In parallel, we have developed an accurate sampling system for in planta infection sites based on microscopy and GFP-tagging of the pathogen, and by using such sampling approaches have shown that metabolomic fingerprinting and supervised data analysis can detect reproducible major changes in metabolome during lesion development. We will carry out detailed metabolome phenotyping of M. grisea in order to understand the precise roles of genes involved in the development of fungal infection structures using already available mutants. We will also refine and carry out a screen for mutants affected in the timing, rate of growth and sporulation of disease lesions. Mutants representative of different phenotypic classes will be inoculated onto hosts in controlled environments and lesion material collected at several time points. Metabolome analysis will follow a hierarchical procedure initiated with high-throughput, low-resolution ESI-MS fingerprinting (LTQ linear ion trap) and GC-tof-MS fingerprinting (LECO Pegasus II). Discrimination of appropriate sample combinations will be determined by supervised data analysis. If there is evidence for metabolome differences (e.g. comparing mutants with an isogenic wild type strain at the same stage of infection) then a further subset of the same samples will be subjected to ESI-FTMS fingerprinting to generate high resolution peak tables. Corresponding GC-tof-MS chromatograms will also be processed to deconvolute and annotate peaks for data mining. Explanatory metabolite signals between different sample classes will be determined using machine learning procedures. In ESI-MS data, high ranked m/z signals will be examined in chromatograms of further LC-FTMS analyses to predict the mass of possible parent ions which will be further fragmented to obtain MS(n) spectral data. Metabolite mass tables and spectral libraries will be searched for matches with spectra representing discriminatory peaks, and predicted metabolites will be quantified against standards using targeted GC-tof-MS or LC-MS as appropriate. Metabolome differences centred on specific metabolites in the various mutants will be used to determine areas of metabolism that may impact on fungal pathogenicity in future experiments. Parallel genetic analysis of insertional mutant collections of M. grisea will focus on those in which metabolome differences are apparent during plant tissue invasion. Gene isolation by inverse PCR, complementation and validation by targeted gene replacement experiments will be used to define genes associated with disease lesion formation by M. grisea.