Genome-wide Association Study Of Rice Seeding Resistance To Blast

Rice is a vital and stable grain food, but rice blast disease is one of the most serious and recurrentproblems in rice-growing regions. Every year the rice capable of feeding60million people is destroyedby rice blast disease. Hence, cultivating rice varieties with highly efficient, durable resistance to blast isstill the most economically feasible and environmentally sound management approach in mostblast-prone rice ecosystems. Most previous studies cloned resistance genes were based on linkagemapping using bi-parental populations. Here, we extensively examined blast resistance in agenome-wide association study （GWAS） based on genotyping805,158SNPs variants across517diverse rice accessions. The main results were as follows:1. GWAS identified126associated loci at P<10-7in the compressed MLM. A total of51associatedSNPs were identified in indica panel. Only five associated loci were detected in japonica panel. AGWAS in the full panel was higher power with a total of72SNPs.2. Searching the flanking regions of the associated SNPs, we revealed18regions co-localized, withseven known cloned genes （Pia, Pik, Pi-1, Pik-h/Pi-54, Pik-m, Pik-p, Pi5/Pi3/Pi-i）, and eight QTLs （Pif,Pig, PiGD3, Pik-s, Pik-g（t）, Pitq5, Pitg6, Pi6）. In addition to the loci associated with the known, wealso found another18new loci referring to25previously unreported candidate genes, which describedas disease resistance protein and showed potential function for rice blast as R genes, separately17inindica panel, one in japonica panel and11in the full panel. Of25candidate genes，four simultaneouslyfound in indica and the full panel, and one simultaneously in japonica and the full panel.3. According to gene ontology （GO） analysis, the others encoding proteins involved53classifications.In the protein class, nucleic acid binding and hybrolase were the main. In the metabolic pathways, DNAreplication was major. In the biological process, metabolic process was the most. In the molecularfunction, catalytic activity and binding were the largest proportion.4. We chose four associated SNPs to identify the candidate genes, referring to Chr11₆526998,Chr12₁4696604, Chr12₁3690289and Chr12₁3032951. Through RT-PCR and expression profilesfrom microarray data, we observed:（1） Chr11₆526998was within the coding domain of the geneOs11g0225100, one of the rice Pia-blast resistance gene.（2） A signal, Chr12₁4696604, was detectedaround the QTL PiGD3. And, our study identified the gene Os12g0438300, a disease resistance proteincontaining NB-ARC and leucine-rich repeat domain, as the main candidate gene of PiGD3.（3） Forassociated SNP, Chr12₁3690289, the newly discovered R genes, Os12g0424700and Os12g0427000were presumed to be required for the full function of this locus.（4） In addition, for Chr12₁3032951,the region without R genes, Os12g0416300, Os12g0417100, and Os12g0417600were considered themost likely candidate genes for this region.5. Haplotype analysis around the candidate genes reflected that most of the SNPs were non-synonymousmutation. 6. Our results further confirmed that GWAS is a powerful complementary approach for dissecting thequantitative disease resistant genes to traditional QTL mapping.