Genetic Analysis Of A Novel Aegilops Tauschii-derived Gene Conferring Race-specific Resistance To Blumeria Graminis DC F.sp.tritici

Wheat powdery mildew disease caused by Blumeria graminis DC f.sp.tritici is one of themost devastating diseases prevalent in the wheat production area worldwide.Development of resistant cultivars is the most cost efficient and environmental safe wayto prevent its epidemic. Mining effective resistance genes and resources, however, is theprerequisite to developing the resistant cultivars. Genetic dissection of the genes ofinterest provides a foundation for marker-assisted selection, subsequent gene isolationand for understanding of plant disease resistance mechanism.Aegilops tauschii (syn. Aegilops squarrosa, 2n=2x=14, DD) is the diploid donor of Dgenome of allohexaploid bread wheat (Triticum aestivum,2n=6x=42,AABBDD). Itcarries a wide variety of superior genes which can be transferred to bread wheat withoutmuch effort because the two species share D genome. M53 (YAV2/TEZ//Aegilopssquarrosa249) is the synthetic hexaploid wheat of Aegilops tauschii and tetraploidTriticum durum(2n=4x=28,AABB). It was introduced from CYMMIT and exhibitedeverlasting immunity to the predominant isolates of Blumeria graminis DC f.sp.tritici(Bgt) in Beijing and Yangzhou wheat fields of China. In this paper, we explored theorigin and the inheritance mode of the gene of interest carried by M53 and the differences of disease response patterns of the target gene from known powdery mildewresistant genes (Pm) through inoculation of an array of isolates of Bgt. Chromosomelocation and integrated genetic mapping of the gene of interest were done employingAFLP, SSR and a set of aneuploids with Chinese Spring (CS) genetic backgrounds.Considering the cloned R genes have functional and structural similarities in proteinsequences irrespective of host type and pest type, we designed both degenerate andspecific PCR primers based on the conserved motifs and isolated the resistant geneanalogs (RGAs) from genomic DNA of M53. The following are the major points:Under infection of isolate No. 15 of Bgt, which is prevalent in wheat fields of Beijing area,both of F₁ reciprocal crosses of M53/Wan7107 demonstrated resistant phenotypes.indicating the gene resides in the nucleus in a dominant form. Further investigation of F₂individuals showed that the gene of interest segregated in a Mendelian factor waysuggested by significant 3:1 ratio of resistant/susceptible at Chi-square level. Randomlysampled F₂ individuals were advanced to F₃ generation. The disease response pattern ofF₃ individuals were in concert with the expectation, indicating the reliable scoring of F₂phenotypes. In order to make clear of the origin of the gene of interest, the two parents ofM53, YAV2/TEZ and Aegilops squarrosa249, were inoculated with the same isolate ofBgt. Consequently, YAV2/TEZ exhibits considerable susceptibility and Aegilopssquarrosa249 showed immunity to the tested isolate. These results figured out the factthat the gene is originated from Aegilops squarrosa249, that is, from D genome. Thegene of interest is temporarily named Pm-M53.To date a total of three Pm genes were assigned to D genome of wheat and its wildrelatives. They are Pm2, Pm19 and Pm24. All of these three Pm genes respondeddifferently from Pm-M53 to isolate No. 15 infection either at seedling stage or at adultstage. Pm-M53 was immune at both stages, whereas Pm19 and Pm24 showedsusceptibility at both stages. and Pm2 was resistant at seedling stage and susceptible atadult stage. Further infection with 14 additional isolates of Bgt differentiated Pm-M53from the other three Pm genes. The facts above showed that Pm-M53 is a race-specificand single dominant R gene, and seemed to be different from known Pm genes. Molecular markers are now becoming indispensable tools for marker-assisted selectionbreeding (MAS), genetic mapping and map-based cloning of agronomically importantgenes. AFLP marker technique is very powerful and popular because of no requirementof sequence information of organisms and being accessible to many loci at one-time PCRamplification. When combined with bulked segregant analysis (BSA), it is quite effectiveto search for closely-linked markers and to saturate mapping of the target region. SSRmarkers also have distinct advantages of widely and evenly distributions and knownpositions on wheat chromosomes and of displaying co-dominant band patterns which candifferentiate homozygote from heterozygote. Therefore, they have been universally usedfor genetic mapping and gene localization. Application of these powerful molecularmarkers facilitates identification and characterization of wheat Pm genes.In this study, F₂ segregation population from cross of M53/Wan7107 was investigated tomap Pm-M53 by combining employment of AFLP, SSR and BSA. Each of 256 AFLPprimer pairs can produce over thirty bands both in M53 and Wan7107 genomic DNA,and polymorphic loci ranged from one to eleven between M53 and Wan7107, but only asmall fraction of them, a total of 16, were found to repeat well between the bulkedresistant genomic DNA pool and the bulked susceptible genomic DNA pool. Usinglinkage analysis software JoinMap version 3.0 and Kosambi's function at LODthreshold of 3.0, only two AFLP polymorphic fragments, Apm109 and Apm161, werefound to be closely linked to Pm-M53 with genetic distances of 1.0 and 3.0 cM,respectively. These two fragments flanked the target gene. Apm109 repeat well inresistant F₃ individuals. Amongst of 70 SSR primers sampled from D genome, only fiveprimers, Xwmc289, Xgwm583, Xgwm292, Xgwm191 and Xwme161, were able toproduce required polymorphic loci between bulked resistant genomic DNA pool andbulked susceptible genomic DNA pool. Interestingly, all of these five markers werelocated to the chromosome 5D with Xgwm191 on the short ann and the remaining fouron the long ann. These five polymorphic primers were subsequently used to analyze F₂mapping population, as a result, three markers, Xwmc289, Xgwm583 and Xgwm292,were linked to and flanked Pm-M53 with genetic distances of 20.0, 33.0 and 24.0 cM,respectively, indicating that Pm-M53 was probably located on the long arm of chromosome 5D. Unexpectedly, when the band of interest of marker Xwmc289 fromM53 and its 'allele band' from Wan7107 was respectively excised, cloned and sequenced,each band was found to be involved with two distinct types of nucleotide sequences ofsame size. That is, one band, two loci. The size of co-migration band from M53 was 159bp and its 'allele band' from Wan7107 was 161 bp. Two-nucleotide motifs of CT and GAreside in each co-migration band, whereas each type of sequence from M53 has acountpart in Wan7107 with polymorphisms of a C/G substitution and a GA-unitdeletion/insertion.Integrated genetic map encompassing Pm-M53 was constructed with two screened AFLPmarkers and three SSR markers. Markers Apm161, Xgwm583 and Xwmc289 wereplaced at one side, and Apm109 and Xgwm292 at the other side of Pm-M53. Thearrangement of the screened markers and Pm-M53 on 5DL, from proximal to distal,follows the sequence of Xgwm583, Xwmc289, Apm161, Pm-M53, Apm109 andXgwm292. The sequence of the three SSR markers on our genetic map was consistentwith the reference map. The fact that flanking SSR and AFLP markers covered the targetregion further indicated that the target gene was located on 5DL. This assignation ofPm-M53 to 5DL was also enhanced by chromosome location using a set of CSaneuploids. Closely linked AFLP marker Apm109 was present both innullisomic-tetrasomic line CSN5BT5D and ditelosomic line CSDT5DL, but absent innullisomic-tetrasomic line CSN5DT5B and ditelosomic line CSDT5DS.Cloned R genes share a strong functional and structural similarity in protein sequence.One or more conserved domains and motifs are present in cloned R genes irrespective ofhost type and pest type. These domains are typically categorized into four types,NBS-LRR, extracellular LRR, receptor kinase and cytosomic protein kinase. Each ofdomains has one or more highly conserved motifs. Therefore, PCR amplification wereoften utilized to search for candidate R genes or defense response related genes, usingdegenerate or specific primers which were designed based on amino acid or nucleotidesequence of conserved motifs. In this study, both degenerate and specific primers weredesigned based on the conserved amino acid or nucleotide sequences of NBS-LRR likeprotein from Aegilops tauschii and cloned Pm genes from wheat and barley. For M53. Wan7107, bulked resistant genomic DNA pool and bulked susceptible genomic DNApool, both types of primers can produce over eight loci. Only two specific primers,RGAP-2 and RGAP-6, amplified required polymorphic bands, present in M53 andbulked resistant genomic DNA pool, but absent in W7107 and bulked susceptiblegenomic DNA pool. None of degenerate primers could produce any polymorphic bandsbetween them. The polymorphie bands were then excised, cloned and sequenced. Thesizes of RGAP-2 and RGAP-6 fragments are 331bp and 259bp, respectively.Interestingly, when searching for similarities in GenBank non-redundant nucleotidedatabase using NCBI BLASTN function, the top three or four hits were all originatedfrom Aegilops tauschii and related to disease resistance. RGAP-2 fragment shared ahigher homology with top hits. Moreover, RGAP-2 fragment could be translated intoprotein sequence without interruption, whereas RGAP-6 fragment could not, resulting inthe prediction that RGAP-2 fragment may be a partial sequence of a functionallyunknown gene, while RGAP-6 fragment may be an intron or a regulatory sequence.When searching for similarities in non-redundant protein database of GenBank usingNCBI BLASTX function, all of the hits were related to plant disease resistance withobvious NB or kinase domains and motifs. The top four hits share a strong homologywith LZ-NBS-LRR type of protein sequences originated from Aegilops tauschii.RGAP-2 fragment shared with top 10 hits several conserved amino acid motifs likeI(V/L)IDD(K/I)WDK, PLS, KIL, KCGGVPLAII and N(N/E)CGSR(I/V)I(T/A)TTRI,amongst of which, I(V/L)IDD(K/I)WDK, N(N/E)CGSR(I/V)I(T/A)TTRI andKCGGVPLAII were correspondingly similar to kinase2, kinase3a and conservedhydrophobic GLPLAII motifs of cloned R genes. Further 'domain search' with NCBIRPSBLAST function in GenBank showed that RGAP-2 fragment possesses a NB-ARCdomain which shared by cloned R genes of Arobidopsis thaliana-derived RPS-2, RPM1,RPP8 and RPP13, and tomato-derived Prf, I2C-1 and I2C-2, and flax-derived L6, andprogrammed cell death related proteins from human and fruit fly. These results showedthat RGAP-2 fragment was probably originated from Aegilops tauschii, and may be apartial sequence of a functional R gene. But relative low score of identity and high Evalue suggested that RGAP-2 and RGAP-6 were distinct sequences from known Aegilops tauschii-derived sequences in GenBank.In this paper, the following issues were discussed: The relationship of Pm-M53 withknown Pm genes; scoring system of powdery mildew inducible phenotype; conversion ofAFLP markers into locus-specific markers; the possible relationship of RGAP-2fragment with the cloned R genes in cereals, and with the cloned Pm genes in plant, andwith powdery mildew inducible defense response genes or candidate genes in wheat andbarley, respectively. Additionally, the advances of relevant issues were mini-reviewed inthe beginning pages of this paper: The biological characteristics of Blumeria graminisDC f.sp.tritici; identification and characterization of known Pm genes and QTLs;exploitation of closely linked or genie markers of known Pm genes; the inheritancemodes of powdery mildew resistance genes or QTLs; molecular breeding strategies ofdevelopment of powdery mildew resistance cultivars; mechanism of plant-pathogeninteraction and R gene cloning; strategies of engineering enhanced plant resistance.Finally, the disadvantages and disadvantages, progress and optimization of AFLPsystems were thoroughly reviewed.