摘要
Understanding how plants respond to drought can benefit drought resistance （DR） breeding. Using a non-destructive phenotyping facility, 51 image-based traits （i-traits） for 507 rice accessions were extracted. These i-traits can be used to monitor drought responses and evaluate DR. High heritability and large variation of these traits was observed under drought stress in the natural population. A genome-wide as- sociation study （GWAS） of i-traits and traditional DR traits identified 470 association loci, some containing known DR-related genes. Of these 470 loci, 443 loci （94%） were identified using i-traits, 437 loci （93%） co- localized with previously reported DR-related quantitative trait loci, and 313 loci （66.6%） were reproducibly identified by GWAS in different years. Association networks, established based on GWAS results, revealed hub i-traits and hub loci. This demonstrates the feasibility and necessity of dissecting the complex DR trait into heritable and simple i-traits. As proof of principle, we illustrated the power of this integrated approach to identify previously unreported DR-related genes. OsPP15 was associated with a hub i-trait, and its role in DR was confirmed by genetic transformation experiments. Furthermore, i-traits can be used for DR linkage analyses, and 69 i-trait locus associations were identified by both GWAS and linkage analysis of a recom- binant inbred line population. Finally, we confirmed the relevance of i-traits to DR in the field. Our study pro- vides a promising novel approach for the genetic dissection and discovery of causal genes for DR.
Understanding how plants respond to drought can benefit drought resistance （DR） breeding. Using a non-destructive phenotyping facility, 51 image-based traits （i-traits） for 507 rice accessions were extracted. These i-traits can be used to monitor drought responses and evaluate DR. High heritability and large variation of these traits was observed under drought stress in the natural population. A genome-wide as- sociation study （GWAS） of i-traits and traditional DR traits identified 470 association loci, some containing known DR-related genes. Of these 470 loci, 443 loci （94%） were identified using i-traits, 437 loci （93%） co- localized with previously reported DR-related quantitative trait loci, and 313 loci （66.6%） were reproducibly identified by GWAS in different years. Association networks, established based on GWAS results, revealed hub i-traits and hub loci. This demonstrates the feasibility and necessity of dissecting the complex DR trait into heritable and simple i-traits. As proof of principle, we illustrated the power of this integrated approach to identify previously unreported DR-related genes. OsPP15 was associated with a hub i-trait, and its role in DR was confirmed by genetic transformation experiments. Furthermore, i-traits can be used for DR linkage analyses, and 69 i-trait locus associations were identified by both GWAS and linkage analysis of a recom- binant inbred line population. Finally, we confirmed the relevance of i-traits to DR in the field. Our study pro- vides a promising novel approach for the genetic dissection and discovery of causal genes for DR.
基金
We thank the National Key Research and Development Program of China （2016YFD0100600, 2016YFD0100101-18）, the National Program on High Technology Development （2014AA10A600）, the National Natural Science Foundation of China （31770397）, and the Fundamental Research Funds for the Central Universities （2662015PY126, 2662016PY0092, 662017PY058）. We also thank Prof. John Doonan of Aberystwyth University for the English language improvement.
