Molecular Evidence Of Directional Selection Of Magnaporthe Grisea And Molecular Breeding Of The Near-isogenic Pyramid Lines With Major Blast Resistance Genes

Rice blast (Magnaporthe grisea) is a destructive disease of rice worldwide. It is very difficult to control the disease because of the diversity and variability of the pathogen. Although resistant variety appears to be the most effective and promising method for control of blast, a variety with a single R-gene often breaks down in 3-5 years after its release and a variety with multiple R-genes is bred difficultly due to the mask between R-genes so that it is important to understand how the pathogen evolves in virulence molecularly and to breed resistant variety by marker-aided selection.( MAS) The dissertation reviews the advances on the pathogen and resistance breeding and focuses on the directional selection, evolution and competition fitness of the pathogen and breeding a set of near isogenic resistance pyramid lines, using CO39 near isogenic (NILs) and pyramid lines (PLs) and molecular biotechnologies. 1. Inoculating 4 isolates representing lineages 4, 7, 17 and 44 by individual and pairing the isolates, 306,531 lesions appeared on CO39 NILs and PLs in 13 inoculation cycles, where the inoculia successively came from the lesions produced in the previous cycle and 9 isolates adapted to major genes. The mutation frequencies largely varied in adaptation to the different R-genes and among the isolates. Among the 9 isolates, 8 is AV-1⁺ for adaptation of Pi-1 with a frequency of 1.04x10^-4, 1 is AV-4a⁺ for Pi-4a with 1.23x10^-7 and no isolate is AV-2+ for Pi-2; the mutation frequencies of P06-6, 9239-4, 101-7-2-1-1 and 9248-6 are 0, 2.5x10^-5, 9.09x10^-5 and 2.82x10^-5, respectively. 2. The R-genes, Pi-k and Pi-1 were proved to be the same gene according to the pathotypic analyses of the isolates adapted and collected from the blast nursery. 3. The DNA fingerprints showed the adapted isolates originated from the parental isolates based on Pot2-PCR, PMC1-PCR, MGR586-RFLP, RAPD and AFLP, excluding the immigration from other sources. The Pot2-PCR based fingerprints of the 6 adapted isolates were identical to their own parental, 3 isolates lost 1, 2 and 3 bands compared with the parental, respectively. The PMC1-PCR based fingerprints of the 9 adapted isolates were the same as their own parental. The MGR586-RFLP based fingerprints of only one adapted isolate showed one band fewer than its parental. The profiles based on RAPD and AFLP with 9 primers and 29 pairs of primers, respectively, indicated the 9 adapted isolates were identical to their own parental. Three RAPD primers could distinguish isolate LC1(4)57-4 from its parental, 101-7-2-1-1, and other isolates. HC0(3)PKT lost 1-3 bands based on the amplification with 30 pairs of AFLP primers. RC1(3+4), ALC(5)141 and LC1(5)57-4 deriving from 9248-6 lost one band based on the AFLP primer pair, 92S08/92G08. 4. The Pot2-PCR based fingerprints showed that lineage closely related to the cultivar with given R-genes. The fingerprints and pathotypic analysis demonstrated that B97414 defeating the pyramids with Pi-1 and Pi-2 may mutate from the isolate infecting the line with Pi-2, but not Pi-1 to the one with AV-1+ mutation. 5. The isolate often changes its virulence with other traits. When they were inoculated together onto a susceptible line CO39, the isolates had different competition fitness, judging from the Pot2-PCR based markers. The adapted isolates with the mutation of high sporulation capacity prevailed over those with low sporulation capacity though it is necessary to mutate from avirulence to virulence. 6. By inoculation of isolates and MAS, 3 groups of PLs including 13 PLs with R-gene combinations: Pi-1,3, Pi-2,3 and Pi-1,2,3 were bred and identified in the blast nursery, using the cross C104PKT (Pi-3)/A57-119 (Pi-1,2). The PLs with Pi-3 were highly resistant in the field.