Magnaporthe oryzae is the causal agent of rice blast disease, a devastating problem worldwide. This fungus has caused breakdown of resistance conferred by newly developed commercial cultivars. To address how the rice blast fungus adapts itself to new resistance genes so quickly, we examined chromosomal locations of AVR-Pita, a subtelomeric gene family corresponding to the Pita resistance gene, in various isolates of M. oryzae (including wheat and millet pathogens) and its related species. We found that AVR-Pita (AVR-Pita1 and AVR-Pita2) is highly variable in its genome location, occurring in chromosomes 1, 3, 4, 5, 6, 7, and supernumerary chromosomes, particularly in rice-infecting isolates. When expressed in M. oryzae, most of the AVR-Pita homologs could elicit Pita-mediated resistance, even those from non-rice isolates. AVR-Pita was flanked by a retrotransposon, which presumably contributed to its multiple translocation across the genome. On the other hand, family member AVR-Pita3, which lacks avirulence activity, was stably located on chromosome 7 in a vast majority of isolates. These results suggest that the diversification in genome location of AVR-Pita in the rice isolates is a consequence of recognition by Pita in rice. We propose a model that the multiple translocation of AVR-Pita may be associated with its frequent loss and recovery mediated by its transfer among individuals in asexual populations. This model implies that the high mobility of AVR-Pita is a key mechanism accounting for the rapid adaptation toward Pita. Dynamic adaptation of some fungal plant pathogens may be achieved by deletion and recovery of avirulence genes using a population as a unit of adaptation.

We have developed a pHANNIBAL-like silencing vector, pSilent-1, for ascomycete fungi, which carries a hygromycin resistance cassette and a transcriptional unit for hairpin RNA expression with a spacer of a cutinase gene intron from the rice blast fungus Magnaporthe oryzae. In M. oryzae, a silencing vector with the cutinase intron spacer (147 bp) showed a higher efficiency in silencing of the eGFP gene than did those with a spacer of a GUS gene fragment or a longer intron (850 bp) of a chitin binding protein gene. Application of pSilent-1 to two M. oryzae endogenous genes, MPG1 and polyketide synthase-like gene, resulted in various degrees of silencing of the genes in 70-90% of the resulting transformants. RNA silencing was also induced by a pSilent-1-based vector in Colletotrichum lagenarium at a slightly lower efficiency than in M. oryzae, indicating that this silencing vector should provide a useful reverse genetic tool in ascomycete fungi.

SummaryWe developed an RNA-silencing vector, pSilent-Dual1 (pSD1), with a convergent dual promoter system that provides a high-throughput platform for functional genomics research in filamentous fungi. In the pSD1 system, the target gene was designed to be transcribed as a chimeric RNA with enhanced green fluorescent protein (eGFP) RNA. This enabled us to efficiently screen the resulting transformants using GFP fluorescence as an indicator of gene silencing. A model study with the eGFP gene showed that pSD1-based vectors induced gene silencing via the RNAi pathway with slightly lower efficiency than did hairpin eGFP RNA-expressing vectors. To demonstrate the applicability of the pSD1 system for elucidating gene function in the rice-blast fungus Magnaporthe oryzae, 37 calcium signalling-related genes that include almost all known calcium-signalling proteins in the genome were targeted for gene silencing by the vector. Phenotypic analyses of the silenced transformants showed that at least 26, 35 and 15 of the 37 genes examined were involved in hyphal growth, sporulation and pathogenicity, respectively, in M. oryzae. These included several novel findings such as that Pmc1-, Spf1-and Neo1-like Ca 2+ pumps, calreticulin and calpactin heavy chain were essential for fungal pathogenicity.

Eighty-five Pyricularia isolates were collected from 29 host species of Gramineae, Bambusideae and Zingiberaceae plants sampled in Brazil, Uganda, Ivory Coast, India, Nepal, China, Indonesia and Japan. These isolates were compared on the basis of pathogenicity, mating ability and restriction fragment length polymorphisms with single-copy DNA probes. Based on the pathogenicity to eight differential gramineous plants, these isolates were classified into seven pathotypes : finger millet type, foxtail millet type, common millet type, rice type, crabgrass type, Italian ryegrass/ weeping lovegrass type, and non-cereal/grass type. Genetic variation among these isolates was assessed by RFLP analysis with two restriction enzymes and nine single-copy DNA probes isolated from a finger millet strain. An UPGMA dendrogram based on the RFLPs revealed that the 86 isolates could be classified into seven major groups. Isolates from cereal crops (finger millet, foxtail millet, common millet, wheat and rice) and a grass, Brachiaria plantaginea, were clustered into a single group. They were further divided into six subgroups corresponding to the pathotypes. Among cereal crop isolates only an isolate from pearl millet was located into a different group. The remaining isolates were clustered into five groups designated as the crabgrass group, the buffelgrass and jungle rice group, the rice cutgrass, knotroot bristlegrass and Setaria tomentosa group, the bamboo and bamboo grass group and the Zingiber mioga group. The isolates from cereal crops were generally capable of mating with finger millet strains and constituted a closed mating compatibility group. These results suggested that the isolates from cereal crops form a single group with a common ancestor although they are pathogenic to taxonomically diverse plants. A combined analysis of the pathogenicity and genetic similarity suggested that the transmission of M. grisea isolates occurs in natural agroecosystems between finger millet and Eleusine africana, goosegrass or Bambusa arundinacea, between foxtail millet and green bristlegrass, and between rice and tall fescue, Italian ryegrass, sweet vernalgrass, reed canarygrass or Oryza longistaminata.

Understanding the molecular basis of plant responses to pathogen-associated molecular patterns (PAMPs) is an active area of research in the field of plant-microbe interactions. A growing number of plant genes involved in various steps of PAMP-triggered immunity (PTI) pathways and microbial factors involved in the elicitation or suppression of PTI have been identified. These studies have largely relied on Arabidopsis thaliana and, therefore, most of the PTI assays have been developed and optimized for that model plant system. Although PTI is a conserved feature among plants, the response spectra vary across different species. Thus, there is a need for robust PTI assays in other pathosystems, such as those involving Solanaceae plant-pathogen interactions, which include many economically important plants and their diseases. We have optimized molecular, cellular, and whole-plant methods to measure PTI responses in two widely studied solanaceous species, tomato (Solanum lycopersicum) and Nicotiana benthamiana. Here, we provide detailed protocols for measuring various PTI-associated phenotypes, including bacterial populations after pretreatment of leaves with PAMPs, induction of reporter genes, callose deposition, activation of mitogen-activated protein kinases, and a luciferase-based reporter system. These methods will facilitate limited genetic screens and detailed characterization of potential PTI-related genes in model and economically important Solanaceae spp.-pathogen interactions.