Difference between revisions of "IC4R001-GWAS-2011-21829395"
(→Project Title) |
|||
| (33 intermediate revisions by 4 users not shown) | |||
| Line 1: | Line 1: | ||
== Project Title == | == Project Title == | ||
| − | ''' | + | ''' Genetic Architecture of Aluminum Tolerance in Rice( Oryza sativa ) Determined through Genome-Wide Association Analysis and QTL Mapping ''' |
==The Background of This Project== | ==The Background of This Project== | ||
| − | Rice | + | [[File:gwas-1.PNG|700px|thumb|right|'''Figure 1.''' '' GWA Analysis of Al Tolerance within and across Rice Subpopulations.'']] |
| + | * While rice (Oryza sativa) is significantly more Al tolerant than other cereals, no genes underlying Al tolerance in rice have been reported. Using genome-wide association(GWA) and bi-parental QTL mapping, we investigated the genetic architecture of Al tolerance in rice. Japonica varieties were twice as Al tolerant as indica and aus varieties. Overall, 57% of the phenotypic variation was correlated with subpopulation, consistent with observations that different genes and genomic regions were associated with Al tolerance in different subpopulations. Four regions identified by GWA co-localized with a priori candidate genes, and two highly significant regions co-localized with previously identified quantitative trait loci(QTL). Haplotype and sequence analysis around the candidate gene, Nrat1, identified a susceptible haplotype explaining 40% of the Al tolerance variation within the aus subpopulation and three non-synonymous mutations within Nrat1 that were predictive of Al sensitivity. Using Indica 6 Japonica mapping populations, we identified QTLs associated with transgressive variation where alleles from a susceptible indica or aus parent enhanced Al tolerance in a tolerant japonica background. This work demonstrates the importance of subpopulation in interpreting and manipulating complex traits in rice and provides a roadmap for breeders aiming to capture genetic value from phenotypically inferior lines. | ||
==Plant Culture & Treatment== | ==Plant Culture & Treatment== | ||
| − | + | * Plants were grown hydroponically in a growth chamber as described by Famoso et al. Al tolerance was determined based on relative root growth (RRG) after three days in Al (160 m M Al 3+ ) or control solution. The hydroponic solution used in this study was chemically designed and optimized for rice Al tolerance screening; for a detailed comparison of the phenotypic procedures employed in this work compared to previously published rice Al tolerance work see Famoso et al. (2010). To obtain uniform seedlings, 80 seeds were germinated and the 30 most uniform seedlings were visually selected and transferred to a control hydroponic solution for a 24 hour adjustment period. After the 24 hour adjustment period, root length was measured with a ruler and the 20 most uniform seedlings were selected and distributed to fresh control solution (0 uM Al 3+ ) or Al treatment solution (160 uM Al 3+ ). Plants were grown in their respective treatments for ,72 hours and the total root system growth was quantified using an imaging and root quantification system as described by Famoso et al.(2010). The mean total root growth was calculated for Al treated and control plants and RRG was calculated as mean growth (Al)/mean growth (control). The 373 genotypes screened for Al tolerance and used in the association analysis are part of a set of 400 O. sativa genotypes that have been genotyped with 44,000 SNPs as described by Zhao et al. | |
| − | * | ||
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
==Research Findings== | ==Research Findings== | ||
| − | * | + | * Two immortalized QTL mapping populations were analyzed for Al tolerance. One consisted of 134 recombinant inbred lines (RIL) derived from the cross IR64/Azucena , and the other was comprised of 78 backcross inbred lines (BIL) derived from the cross Nipponbare/Kasalath//Nipponbare. These populations were used to evaluate Al tolerance using three different indices of relative root growth (RRG), (1) longest root growth (LRG-RRG), (2) primary root growth (PGR-RRG) and total root growth (TRG-RRG) (see Materials and Methods for details). The phenotypic distribution was approximately normal for each population, no matter which root screening index was used. The QTL mapping populations allowed us to determine which of the three root evaluation methods would be most useful for evaluating the diversity panel as a whole.<br><br> |
| − | * | + | * To identify Al tolerance loci based on genome-wide association(GWA) mapping, we used an existing genotypic dataset consisting of 36,901 SNPs, and the total root growth (TRG-RRG) Al tolerance phenotype generated on 373 O. sativa accessions over the course of this study. GWA mapping was conducted, using SNPs with a MAF.0.05, across all 373 genotypes as well as independently within the indica, aus, temperate japonica, and tropical japonica subpopulations '''(Figure 1)'''. The Efficient Mixed-Model Association (EMMA) model was used in each analysis (both within and across subpopulations) to correct for confounding effects due to subpopulation structure and relatedness between individuals. As the subpopulation structure was highly correlated with Al tolerance, it was observed that analyzing all samples (373) together with the EMMA model resulted in an overcorrection (causing type 2 error) and a corresponding reduction in SNP significance. To address this problem, a PCA approach was also employed when analyzing all (373) samples together. However, the PCA approach resulted in a slight under-correction for population structure, demonstrating that results from each GWA method has limitations when used across all germplasm in this highly structured diversity panel. |
| − | + | * We chose to further investigate the variation in and around the Nrat1 gene on chromosome 2 because multiple independent lines of evidence supported the existence of a gene(s) in this region responsible for a significant portion of the variation for Al tolerance in rice. Evidence included a strong GWA peak in the aus subpopulation, a previously reported QTL, and the localization of the Nrat1 Al transporter gene. Using the 44 K SNP data, LD in this region was calculated to be ,150 kb in the aus subpopulation and 11 distinct haplotypes were observed in the entire diversity panel across a 139 kb region around the Nrat1 gene(1.536 Mb–1.675 Mb on chr. 2) (Figure 2). Haplotype 1 (Hap.1), which was unique to the aus subpopulation, was found in 8 Al sensitive aus accessions and one Al sensitive aus/indica admixed line. These 9 genotypes were among the least Al tolerant (7 th percentile, mean RRG=0.16) of the 373 accessions screened. Haplotype 1 explained 40% of the phenotypic variation for Al tolerance within the aus subpopulation. In addition, four aus accessions that were highly or moderately Al tolerant were found to contain a tropical japonica introgression across this region (described in the section on Introgression analysis below). | |
| − | * We | + | [[File:IC4R001-GWAS-2011-21829395-2.PNG|700px|thumb|right|'''Figure 2.''' '' Haplotype analysis of the Nrat1 gene region.'']] |
| − | |||
| − | [[File: | ||
| − | |||
| − | + | == Labs working on this Project == | |
| + | * Department of Plant Breeding and Genetics, Cornell University, Ithaca, New York, United States of America | ||
| + | * Department of Biological Statistics and Computational Biology, Cornell University, Ithaca, New York, United States of America | ||
| + | * Robert W. Holley Center for Agriculture and Health, Agricultural Research Service, US Department of Agriculture, Cornell University, Ithaca, New York, United States of America | ||
| − | == | + | ==Corresponding Author== |
| − | * | + | * ''' Susan R. McCouch'''(srm4@cornell.edu) |
| − | |||
Latest revision as of 09:48, 21 June 2016
Contents
Project Title
Genetic Architecture of Aluminum Tolerance in Rice( Oryza sativa ) Determined through Genome-Wide Association Analysis and QTL Mapping
The Background of This Project
- While rice (Oryza sativa) is significantly more Al tolerant than other cereals, no genes underlying Al tolerance in rice have been reported. Using genome-wide association(GWA) and bi-parental QTL mapping, we investigated the genetic architecture of Al tolerance in rice. Japonica varieties were twice as Al tolerant as indica and aus varieties. Overall, 57% of the phenotypic variation was correlated with subpopulation, consistent with observations that different genes and genomic regions were associated with Al tolerance in different subpopulations. Four regions identified by GWA co-localized with a priori candidate genes, and two highly significant regions co-localized with previously identified quantitative trait loci(QTL). Haplotype and sequence analysis around the candidate gene, Nrat1, identified a susceptible haplotype explaining 40% of the Al tolerance variation within the aus subpopulation and three non-synonymous mutations within Nrat1 that were predictive of Al sensitivity. Using Indica 6 Japonica mapping populations, we identified QTLs associated with transgressive variation where alleles from a susceptible indica or aus parent enhanced Al tolerance in a tolerant japonica background. This work demonstrates the importance of subpopulation in interpreting and manipulating complex traits in rice and provides a roadmap for breeders aiming to capture genetic value from phenotypically inferior lines.
Plant Culture & Treatment
- Plants were grown hydroponically in a growth chamber as described by Famoso et al. Al tolerance was determined based on relative root growth (RRG) after three days in Al (160 m M Al 3+ ) or control solution. The hydroponic solution used in this study was chemically designed and optimized for rice Al tolerance screening; for a detailed comparison of the phenotypic procedures employed in this work compared to previously published rice Al tolerance work see Famoso et al. (2010). To obtain uniform seedlings, 80 seeds were germinated and the 30 most uniform seedlings were visually selected and transferred to a control hydroponic solution for a 24 hour adjustment period. After the 24 hour adjustment period, root length was measured with a ruler and the 20 most uniform seedlings were selected and distributed to fresh control solution (0 uM Al 3+ ) or Al treatment solution (160 uM Al 3+ ). Plants were grown in their respective treatments for ,72 hours and the total root system growth was quantified using an imaging and root quantification system as described by Famoso et al.(2010). The mean total root growth was calculated for Al treated and control plants and RRG was calculated as mean growth (Al)/mean growth (control). The 373 genotypes screened for Al tolerance and used in the association analysis are part of a set of 400 O. sativa genotypes that have been genotyped with 44,000 SNPs as described by Zhao et al.
Research Findings
- Two immortalized QTL mapping populations were analyzed for Al tolerance. One consisted of 134 recombinant inbred lines (RIL) derived from the cross IR64/Azucena , and the other was comprised of 78 backcross inbred lines (BIL) derived from the cross Nipponbare/Kasalath//Nipponbare. These populations were used to evaluate Al tolerance using three different indices of relative root growth (RRG), (1) longest root growth (LRG-RRG), (2) primary root growth (PGR-RRG) and total root growth (TRG-RRG) (see Materials and Methods for details). The phenotypic distribution was approximately normal for each population, no matter which root screening index was used. The QTL mapping populations allowed us to determine which of the three root evaluation methods would be most useful for evaluating the diversity panel as a whole.
- To identify Al tolerance loci based on genome-wide association(GWA) mapping, we used an existing genotypic dataset consisting of 36,901 SNPs, and the total root growth (TRG-RRG) Al tolerance phenotype generated on 373 O. sativa accessions over the course of this study. GWA mapping was conducted, using SNPs with a MAF.0.05, across all 373 genotypes as well as independently within the indica, aus, temperate japonica, and tropical japonica subpopulations (Figure 1). The Efficient Mixed-Model Association (EMMA) model was used in each analysis (both within and across subpopulations) to correct for confounding effects due to subpopulation structure and relatedness between individuals. As the subpopulation structure was highly correlated with Al tolerance, it was observed that analyzing all samples (373) together with the EMMA model resulted in an overcorrection (causing type 2 error) and a corresponding reduction in SNP significance. To address this problem, a PCA approach was also employed when analyzing all (373) samples together. However, the PCA approach resulted in a slight under-correction for population structure, demonstrating that results from each GWA method has limitations when used across all germplasm in this highly structured diversity panel.
- We chose to further investigate the variation in and around the Nrat1 gene on chromosome 2 because multiple independent lines of evidence supported the existence of a gene(s) in this region responsible for a significant portion of the variation for Al tolerance in rice. Evidence included a strong GWA peak in the aus subpopulation, a previously reported QTL, and the localization of the Nrat1 Al transporter gene. Using the 44 K SNP data, LD in this region was calculated to be ,150 kb in the aus subpopulation and 11 distinct haplotypes were observed in the entire diversity panel across a 139 kb region around the Nrat1 gene(1.536 Mb–1.675 Mb on chr. 2) (Figure 2). Haplotype 1 (Hap.1), which was unique to the aus subpopulation, was found in 8 Al sensitive aus accessions and one Al sensitive aus/indica admixed line. These 9 genotypes were among the least Al tolerant (7 th percentile, mean RRG=0.16) of the 373 accessions screened. Haplotype 1 explained 40% of the phenotypic variation for Al tolerance within the aus subpopulation. In addition, four aus accessions that were highly or moderately Al tolerant were found to contain a tropical japonica introgression across this region (described in the section on Introgression analysis below).
Labs working on this Project
- Department of Plant Breeding and Genetics, Cornell University, Ithaca, New York, United States of America
- Department of Biological Statistics and Computational Biology, Cornell University, Ithaca, New York, United States of America
- Robert W. Holley Center for Agriculture and Health, Agricultural Research Service, US Department of Agriculture, Cornell University, Ithaca, New York, United States of America
Corresponding Author
- Susan R. McCouch(srm4@cornell.edu)
