Résumé
In spite of their importance in the resistance durability phenomena, molecular mechanisms for the recognition of fungal pathogens by their host plant remain poorly understood. To date, Twelve fungal avirulence genes have been cloned , including 4 from Cladosporium fulvum and 5 from Magnaporthe grisea. The peptides encoded by these genes share no structural homology, except for their size which is generally (but not always) small. Thus, studying more avirulence genes is desirable. The genome sequence of Magnaporthe grisea was released recently [Magnaporthe Sequencing Project. Ralph Dean, Fungal Genomics Laboratory at North Carolina State University (http://www.fungalgenomics.ncsu.edu), and Whitehead Institute/MIT Center for Genome Research (http://www-genome.wi.mit.edu)]. This provides a useful tool to rapidly develop microsatellite molecular markers. Among the 1,436 microsatellite sequences identified in the genome (that were composed of di, tri and tetra nucleotides of at least 18 nucleotides), we chose 341 sequences, distributed throughout the genome, to define primers pairs allowing for their amplification by PCR. Eight percent of these primers did not amplify any product. Among the others, 43% revealed polymorphism between the parents of a reference cross (Guy 11 x 2539) as analysed on 3% agarose gels. One hundred-thirty four microsatellite markers could be added to the existing reference map. These markers are distributed over the six chromosomes at an average distance of 10 cM. The descendants of four crosses between different M grisea strains were analyzed by pathology testing and 9 avirulence genes were identified. The reference microsatellite map was used to choose markers that allowed us to map them in the new crosses or to increase the density of markers in areas surrounding the avirulence genes. In most cases, when comparing the location of markers in different crosses, the order of markers and distances between markers were conserved. Some evidence for rearrangements were, however, found. The use of a reference microsatellite map turned out to be very rapid and very efficient for the construction of framework genetic maps and rough mapping of a gene. For example, for a new cross for which no information was known, we could construct in just 3 weeks, a framework map using 94 progeny and 52 markers covering the 6 chromosomes. Since the genome sequence is available, increasing the number of markers in one particular area should become relatively easy. (Texte intégral)