Skip to main content
NCBI home page
Plant Physiology logoLink to Plant Physiology
. 2016 Feb 19;170(4):1989–1998. doi: 10.1104/pp.15.01971

Integration of Experiments across Diverse Environments Identifies the Genetic Determinants of Variation in Sorghum bicolor Seed Element Composition1,[OPEN]

Nadia Shakoor 1,2,3,4,5, Greg Ziegler 1,2,3,4,5, Brian P Dilkes 1,2,3,4,5, Zachary Brenton 1,2,3,4,5, Richard Boyles 1,2,3,4,5, Erin L Connolly 1,2,3,4,5, Stephen Kresovich 1,2,3,4,5, Ivan Baxter 1,2,3,4,5,*
PMCID: PMC4825124  PMID: 26896393

Genome-wide association analysis identifies novel alleles controlling seed element accumulation.

Abstract

Seedling establishment and seed nutritional quality require the sequestration of sufficient element nutrients. The identification of genes and alleles that modify element content in the grains of cereals, including sorghum (Sorghum bicolor), is fundamental to developing breeding and selection methods aimed at increasing bioavailable element content and improving crop growth. We have developed a high-throughput work flow for the simultaneous measurement of multiple elements in sorghum seeds. We measured seed element levels in the genotyped Sorghum Association Panel, representing all major cultivated sorghum races from diverse geographic and climatic regions, and mapped alleles contributing to seed element variation across three environments by genome-wide association. We observed significant phenotypic and genetic correlation between several elements across multiple years and diverse environments. The power of combining high-precision measurements with genome-wide association was demonstrated by implementing rank transformation and a multilocus mixed model to map alleles controlling 20 element traits, identifying 255 loci affecting the sorghum seed ionome. Sequence similarity to genes characterized in previous studies identified likely causative genes for the accumulation of zinc, manganese, nickel, calcium, and cadmium in sorghum seeds. In addition to strong candidates for these five elements, we provide a list of candidate loci for several other elements. Our approach enabled the identification of single-nucleotide polymorphisms in strong linkage disequilibrium with causative polymorphisms that can be evaluated in targeted selection strategies for plant breeding and improvement.

Sorghum (Sorghum bicolor) is a globally cultivated source of food, feed, and fiber. Contrasting needs for elemental nutrient accumulation limit crop yield and quality for sorghum marketed to different sectors. The seed-bearing reproductive organs, or panicles, in sorghum represent up to 30% of the total dry matter yield (Amaducci et al., 2004). Plant-based diets, in which grains compose the major food source, require the accumulation of bioavailable essential elements in the plant seeds. Currently, iron (Fe) and zinc (Zn) deficiencies negatively affect the health of over two billion people worldwide (World Health Organization, 2002). Increased bioavailable elemental nutrient content in the edible portions of sorghum for human and animal nutrition could ameliorate this nutritional crisis (Graham et al., 1999; World Health Organization, 2002). Additional global health benefits could be achieved by increasing magnesium (Mg), selenium (Se), calcium (Ca), and copper (Cu; White and Broadley, 2005) while reducing the concentration of toxic elements, including arsenic (As) and cadmium (Cd; Ma et al., 2008).

Seed element accumulation results from interconnected biological processes, including element uptake by the roots, translocation and remobilization within the plant, and ultimately import, deposition, and assimilation/storage in the seeds. Element availability is further affected by the accumulation of metabolites in seeds (Vreugdenhil et al., 2004). High-throughput ionomic analysis, or concurrent measurement of multiple elements, allows for the quantitative and simultaneous measurement of an organism’s elemental composition, providing a snapshot of the functional state of an organism under different experimental conditions (Salt et al., 2008). Most studies of the plant ionome utilize inductively coupled plasma mass spectroscopy (ICP-MS). Briefly, inductively coupled plasma (ICP) functions to ionize the analyte into atoms, which are then detected by mass spectroscopy. Reference standards are used to identify and quantitate each element of interest in the sample. ICP-MS analysis can be accomplished in as little as 1 min per sample, which allows for high-throughput processing of thousands of samples (Salt et al., 2008). Previous studies have demonstrated that several elements, including Fe, manganese (Mn), Zn, cobalt (Co), and Cd, share mechanisms of accumulation (Yi and Guerinot, 1996; Vert et al., 2002; Connolly et al., 2003). Ionomic signatures derived from multiple elements also have been shown to better predict plant physiological status for some elements than the measure of the element’s concentration, including essential nutrients like Fe (Baxter et al., 2008). Holistically examining the ionome provides significant insights into the networks underlying ion homeostasis beyond single-element studies (Baxter and Dilkes, 2012).

There are over 45,000 catalogued lines of sorghum at the U.S. Department of Agriculture Germplasm Resource Information Network. This diverse collection of sorghum germplasm contains genetic variation with undiscovered impact on seed element composition (Das et al., 1997). Mapping quantitative trait loci for seed element concentration has been successful in a number of species, including Arabidopsis (Arabidopsis thaliana; Vreugdenhil et al., 2004; Waters and Grusak, 2008; Buescher et al., 2010), rice (Oryza sativa; Norton et al., 2010; Zhang et al., 2014), wheat (Triticum aestivum; Shi et al., 2008; Peleg et al., 2009), and maize (Zea mays; Simić et al., 2012; Baxter et al., 2013, 2014). Genome-wide association (GWA) mapping is well suited for uncovering the genetic basis for complex traits, including seed element accumulation. One of the key strengths of association mapping is that a priori knowledge is not necessary to identify new loci associated with the trait of interest. Furthermore, a GWA mapping population is composed of lines that have undergone numerous recombination events, allowing for a narrower mapping interval. Previous GWA studies in maize (Tian et al., 2011), rice (Huang et al., 2010), and sorghum (Morris et al., 2013) have been successful in identifying the genetic basis for various agronomic traits. Here, we analyzed the seed ionome from a community-generated association panel to identify potential loci underlying seed element accumulation in sorghum.

RESULTS

Phenotypic Diversity for Seed Element Concentrations in the Sorghum Association Panel

We grew 407 lines from the publicly available Sorghum Association Panel (SAP) selected for genotypic diversity and phenotypic variation (Casa et al., 2008; Supplemental Table S1). These lines were previously genotyped by sequencing (Morris et al., 2013). The SAP lines were grown in three experiments: Lubbock, Texas, in 2008 (SAP 2008); Puerto Vallarta, Mexico, in 2012 (SAP 2012); and two field replicates produced in Florence, South Carolina, in 2013 (SAP 2013-1 and SAP 2013-2). A total of 287 of the 407 SAP lines were present in all four grow outs.

Seed samples were taken from each replicate and weighed before analysis. A simple weight normalization and established methods to estimate weight from the element content were attempted (Lahner et al., 2003). However, both methods created artifacts, particularly in elements with concentrations near the level of detection for ICP-MS (Supplemental Fig. S1). To address this concern, we included weight as a cofactor in a linear model that included other sources of technical error and utilized the residuals of the model as the trait of interest for genetic mapping. The residuals from this transformation were used for all further analyses and outperformed any other method (data not shown).

We calculated broad-sense heritability for each trait to determine the proportion of the phenotypic variation explained by the genetic variation present in the SAP across the three environments (Table I). Heritability estimates ranged from 1% (sodium, Na) to 45% (Cu). We obtained moderate heritability (greater than 30%) for several elements, including Mg, phosphorus (P), sulfur (S), potassium (K), Ca, Mn, Fe, Co, Zn, strontium (Sr), and molybdenum (Mo). Low heritabilities were reported previously for seed accumulation of aluminum (Al) and As (Norton et al., 2010) as well as for Se, Na, Al, and rubidium (Rb) in a similarly designed study in maize seed kernels (Baxter et al., 2014). The relatively lower heritabilities for these elements, including boron (B), Cd, and Se, could be explained by environmental differences between the experiments, element accumulation near the limit of detection via ICP-MS, and the absence of genetic variation affecting the concentrations of these elements. Consistent with the hypothesis that the field environment was masking genetic variation, we calculated the heritability for two field replicates of the SAP in 2013 and found higher heritabilities for 12 elements (Table I).

Table I. Mean, sd, and broad-sense heritability of seed element concentrations from the SAP averaged across three environments.

Element concentration values are presented as mg kg−1, and broad sense heritability (H2) was calculated as described in “Materials and Methods.” Data represent averages of individual samples (n = 287) analyzed in four separate experiments.

Trait All SAP Replicates
 H2
Mean sd All SAP
Replicates 2013 Field
Replicates
B 12.8 5.88 0.02 0.05
Na 0.572 0.316 0.01 0.00
Mg 1580 246 0.42 0.38
Al 0.277 0.269 0.06 0.17
P 3350 623 0.38 0.35
S 890 127 0.36 0.33
K 3850 795 0.36 0.46
Ca 25.3 18.8 0.41 0.57
Mn 13.5 3.33 0.36 0.43
Fe 24.9 5.7 0.40 0.46
Co 0.00611 0.00448 0.32 0.28
Ni 0.18 0.187 0.22 0.10
Cu 2.54 1.21 0.45 0.47
Zn 19.9 5.71 0.35 0.38
As 0.0796 0.0361 0.08 0.19
Se 1.32 0.454 0.03 0.00
Rb 1.76 0.848 0.10 0.12
Sr 0.573 0.586 0.32 0.27
Mo 0.603 0.306 0.33 0.37
Cd 0.0683 0.0739 0.16 0.23
Open in a new tab

We detected significant effects of both genotype and environment on most of the elements (Fig. 1; Supplemental Table S2). The measured element concentrations of this study corroborate the broad range observed in the sorghum element literature (Mengesha, 1966; Neucere and Sumrell, 1980; Lestienne et al., 2005; Ragaee et al., 2006). Similar to a study carried out in wild emmer wheat (Triticum turgidum ssp. dicoccoides; Gomez-Becerra et al., 2010), grain Na and Ca showed large variations (5- and 4-fold, respectively). Compared with micronutrients, the remaining macronutrients (P, K, S, and Mg) measured in the study exhibited less phenotypic variation overall (Table I; Supplemental Table S3) and ranged between 1.6- and 1.8-fold across the SAP. Of the micronutrients, high variation was detected for Al and nickel (Ni; 8- and 6-fold, respectively). With the exception of these two elements, seed micronutrient concentration showed phenotypic variation ranging between 2.4- and 5.6-fold. High variation in Ni and Al may indicate strong environmental effects on seed Ni and Al concentrations or contamination during handling and analysis of the seeds, as suggested previously (Baxter et al., 2014). The element traits were well distributed across the sorghum subpopulations, with no specific subpopulations accumulating disproportionate levels of any element (Supplemental Fig. S2).

Figure 1.

Figure 1.

Open in a new tab

Box plots with median, minimum, and maximum values as well as interquartile ranges for the 20 elements in three SAP experimental populations. The raw concentration values for each of element were log transformed to obtain normally distributed phenotypes.

We used two different approaches to identify the shared regulation of element accumulation. Pairwise correlations were calculated and graphed (Fig. 2A; Supplemental Table S4), and principal component analysis (PCA) was carried out (Fig. 2B). Highly correlated element pairs in our data included Mg/P, Mg/Mn, P/S, and Mg/S. Divalent cations Ca/Sr and K/Rb are chemical analogs, and strong correlation was observed between these pairs of elements, consistent with previous reports in other species (Queen et al., 1963; Hutchin and Vaughan, 1968; Ozgen et al., 2011; Broadley and White, 2012). In the SAP, the first two principal components accounted for a large fraction of the phenotypic covariance (36%). PCA clustering of elements reflected known elemental relationships, including the covariation of Ca/Sr and K/Rb (Fig. 2). A cluster of the essential micronutrient transition metals, Fe, Zn, and Cu, is distinguishable, suggesting that their accumulation may be affected by a shared mechanism. Similarly, clustering of Mg and P is consistent with previous studies in wheat (Peleg et al., 2009). Seed P is stored predominantly as the Mg2+ salt of phytic acid (inositol hexaphosphate), which may explain the significant positive correlation of these elements (Maathuis, 2009; Marschner and Marschner, 2012).

Figure 2.

Figure 2.

Open in a new tab

A, Correlation network of seed element concentrations using rank average data calculated across replicates from SAPs. Green solid lines represent positive correlation values, and red dashed lines represent negative correlation values. The intensity and thickness of the lines indicate the degree of correlation. Element correlation values can be found in Supplemental Table S4. Correlation networks for SAP 2008, SAP 2012, and SAP 2013 can be found in Supplemental Figure S3. B, PCA applied to the rank average seed concentrations for 20 elements in the SAP lines across experiments. Each symbol represents a single element. PCA for SAP 2008, SAP 2012, and SAP 2013 can be found in Supplemental Figure S4. Outlined elements reflect clustering of known elemental relationships.

GWA Mapping of Seed Element Traits

To dissect the genetic basis of natural variation for seed element concentration in sorghum seed, GWA mapping was performed using both an optimal model obtained from the multilocus mixed model (MLMM) algorithm and a compressed mixed linear model (CMLM) that accounts for population structure. For the MLMM analysis, we considered several models to account for population structure as well as two different models to determine the number of cofactors to add into the analyses (Supplemental Fig. S5; see “Materials and Methods”). We chose a kinship model to account for population structure and the most conservative multiple-Bonferroni (mBonf) criterion model for selecting cofactors. We also used a conservative Bonferroni-corrected threshold (P < 0.05) for CMLM. These two approaches identified overlapping single-nucleotide polymorphisms (SNPs) significantly associated with seed element concentration (Supplemental Tables S5 and S6). Compared with traditional single-locus approaches (e.g. CMLM), MLMM uses multiple loci in the model, which contribute to a higher detection power and a lower potential for false discoveries (Segura et al., 2012). MLMM also identified additional associations of interest. Significant SNPs identified with the MLMM approach were prioritized for further analyses (Supplemental Table S5).

In an effort to comprehensively identify significant SNPs associated with element concentration, we created several data sets for GWA analysis. After averaging the two SAP 2013 grow outs, each location was treated as an individual experiment. To link SAP experiments across environments, we ranked the individual lines of each experiment by element concentration and derived a robust statistic describing element accumulation for GWA using the average of ranks across the three SAP environments. By utilizing rank order, we eliminated skewness and large variations in element concentration due to environmental differences (Conover and Iman, 1981). GWA scans across individual experiments identified 270, 228, and 207 significant SNPs for all 20 element traits in the SAP 2008, SAP 2012, and SAP 2013 panels, respectively. In total, we identified 255 significant loci in the ranked data set for the 20 element traits (Supplemental Table S5). The number of significant SNPs per element trait ranged from two (B) to 33 (Ca) and roughly reflected their heritabilities in this data set (Table I; Supplemental Table S7).

We identified several SNPs common to multiple environments (Supplemental Table S8). For example, GWA for Ca concentration in all three of our SAP experiments identified significant SNPs within 5 kb of locus Sobic.001G094200 on chromosome 1. Sobic.001G094200 is a putative Ca homeostasis regulator (CHoR1; Zhang et al., 2012). We also identified several significant SNPs that colocalized for multiple element traits (Fig. 3; Supplemental Table S9). Several of these SNPs were detected as significantly associated with multiple elements that are known to be coordinately regulated (Yi and Guerinot, 1996; Vert et al., 2002; Connolly et al., 2003; Lahner et al., 2003) and implicate candidate genes involved in the regulation of multiple elements. For example, a SNP on chromosome 1 (S1_18898717) was a significant peak in both Mg and Mn GWA analysis (Fig. 3). This SNP peak is in linkage disequilibrium (LD) with the Arabidopsis homolog of AT3G15480. AT3G15480 is a protein of unknown function; however, transfer-DNA (T-DNA) knockout lines display mutant phenotypes in both Mg and Mn accumulation (www.ionomicshub.org; SALK_129213, Tray 449). T-DNA knockout lines in Arabidopsis also validated the significant peak for Co accumulation in this study (S2_8464347). This SNP is linked to the homolog of AT5G63790, a NAC domain-containing protein that imparts a significantly decreased Co phenotype in the T-DNA knockout line (www.ionomicshub.org; SALK_030702, Tray 1137).

Figure 3.

Figure 3.

Open in a new tab

Heat map displaying the log P values of shared significant SNPs across 20 elements in the rank average data set. Significance values below 2 are white, and the ranges from 2.01 to 9.01 are shown in green (light to dark). Outlined in red are biologically relevant SNPs that colocalized for multiple elements.

We focused our interpretation efforts on the GWA results from the SAP rank average data set, as these are the most likely to provide the tools to manipulate seed element concentration in multiple environments. The GWA results for each element trait obtained at the optimal step of the MLMM were compiled. The data for Cd using the SAP rank average data set is presented in Figure 4 as an example of the analysis procedure. GWA with the optimal MLMM (mBonf) across multiple environments identified one significant SNP (S2_8883293) associated with Cd levels. (Fig. 4A). The distribution of expected versus observed P values, quantile-quantile (QQ) plots (Fig. 4B; Supplemental Fig. S5), suggests that population structure was well controlled and false positive association signals were minimized using the kinship matrix plus cofactors. Furthermore, this SNP explained 18% of the phenotypic variation in Cd (Fig. 4C), and the allelic effects of each genotype were estimated (Fig. 4D). This major-effect locus on chromosome 2 is in LD with a homolog of a well-characterized Cd transporter in plants, HEAVY METAL ATPASE2 (HMA2).

Figure 4.

Figure 4.

Open in a new tab

A, Manhattan plot displaying Cd GWA study results [−log10(P)] for the 10 sorghum chromosomes (x axis) and associated P values for each marker (y axis). The red lines indicate a Bonferroni-corrected threshold of 0.05. B, QQ plot of observed P values against the expected P values from the GWA analysis for Cd element concentration. The MLMM includes cofactors that reduce the inflation of P values (green line). The null model, which does not consider significant cofactors, indicates the presence of P value inflation (blue line). The red line indicates the expected P value distribution under the null hypothesis. C, Evolution of genetic variance at each step of the MLMM (blue, genetic variance explained; green, total genetic variance; red, error). The yellow line indicates the variance with the inclusion of S2_8883293. The orange line indicates the optimal model selected by the mBonf criterion. D, Allelic effect for the significant SNP marker on chromosome 2. The y-axis indicates the rank average values for Cd. The x-axis depicts the genotypes at marker S2_8883293. AA and TT refers to the homozygous alleles present at marker S2_8883293.

DISCUSSION

Ionome Profiling for Improved Sorghum Seed Quality

Increasing the concentrations of elements essential for human and animal nutrition (e.g. Fe and Zn) while simultaneously minimizing and increasing tolerance to antinutrients and toxic elements (e.g. As, Cd, and Al) is a significant goal of fundamental research directed toward global crop improvement (Schroeder et al., 2013). Element homeostasis in plants is affected by genotype, environment, soil properties, and nutrient interactions (Gregorio et al., 2000).

While determining strategies to enhance or reduce element content for food or fuel, several components of seed element traits must be considered. These include the heritability of the various element traits, genotype-environment interactions, and the availability of high-throughput element content screening tools (Ortiz-Monasterio et al., 2007). Differences in seed organic composition also can have large effects on the element composition of seeds, as different seed compartments will contain elements in different proportions. Variation in seed composition, together with variation in sorghum seed size, violates the assumption of a uniform elemental concentration inherent in simple weight normalizations. Our data were not well modeled by a simple weight normalization (Supplemental Fig. S1), and we subsequently employed a rank transformation of the phenotypic data and linear model in the analysis (Ayana and Bekele, 2000; Baxter et al., 2014).

Our results demonstrate that the environmental effects on the range and means of element concentrations are largely element specific. In general, seed element concentrations did not exhibit large variation due to environmental effects. This contributed to high heritabilities for several elements and homeostasis of individual element concentrations across very diverse environments (Fig. 1; Table I). The high heritabilities for these traits demonstrate the feasibility of breeding strategies for the improvement of sorghum for seed element accumulation. Furthermore, due to the known genetic contributors to trait covariation, selection strategies can include alteration of multiple traits, enhancing phenotypic correlations between traits or counter selection for undesirable traits (e.g. As and Cd accumulation). The high heritability and the relationships we report between important elements, including Fe and Zn, are encouraging for the development of breeding schema for improved sorghum elemental profiles.

Trait correlations and covariation were used to uncover genetic associations for multiple elements. The observed correlations of several elements indicate that changes in one or more elements can simultaneously affect the concentration of other elements present in the seed (Fig. 2A). However, the individual effects of particular alleles can deviate from this pattern. Even without more complicated analyses, we detected colocalized effects on several element traits (Supplemental Tables S4 and S8). For example, several significant SNPs colocalized for the strongly correlated element pairs Ca and Sr (r = 0.79) as well as Mg and P (r = 0.71). Shared SNPs and colocalization of significant loci across multiple element traits suggest the possibility of tightly linked genes or genes with pleiotropic effects and have been documented in recent GWA studies, including experiments in tomato (Solanum lycopersicum; Sauvage et al., 2014) and rice (Zhao et al., 2011). In our analysis, we applied a conservative threshold in our MLMM implementation and identified SNPs from the most complex model, in which the P values of cofactors were below a defined threshold of 0.05. We implemented stringent parameters to eliminate false positives but also risked the elimination of true positives. To identify additional candidate SNPs for further investigation, these stringent parameters can be relaxed to include association signals below the threshold.

Candidate Genes

One of the primary goals of this study was to utilize GWA analyses to identify candidate genes and novel loci implicated in the genetic regulation of sorghum seed element traits. We identified numerous significant SNPs for all 20 element traits that currently do not associate with known elemental accumulation genes. Although it is likely that a small fraction of these SNPs are false positives, many more may be novel associations with as-yet undiscovered causal genes and merit further investigation. We did, however, identify several significant SNPs that fall directly within a characterized candidate gene or are in close proximity, or LD, with putative candidates.

Zn

Zn deficiency is a critical challenge for food crop production that results in decreased yields and nutritional quality. Zn-enriched seeds result in better seedling vigor and higher stress tolerance on Zn-deficient soils (Cakmak, 2008). Here, we identify a strong candidate for genetic improvement of Zn concentration in the sorghum seed, Sobic.007G064900, a homolog of Arabidopsis ZIP5, a Zn transporter precursor (AT1G05300; Table II). AT1G05300 is a member of the ZIP family of metal transporter genes, and overexpression lines of this gene display increased Zn accumulation in Arabidopsis (www.ionomicshub.org; 35SZip5_2 _Tray 700).

Table II. Detailed information for selected significant associations detected within the 20 element traits analyzed using the MLMM.

Phenotype SNP Locus Name Chromosome P Sorghum bicolor Annotation Arabidopsis Homolog
Cd S2_8883293 Sobic.002G083000 2 9.67E-10 Cation-transporting ATPase AtHMA2; HEAVY METAL ATPASE2
Mo S3_64823106 Sobic.003G320600 3 1.65E-08 Membrane protein-like Sulfite exporter TauE/SafE familyprotein
Ni S6_53175238 Sobic.006G164300 6 1.84E-07 IRON TRANSPORT PROTEIN2 AtYSL3; YELLOW STRIPE LIKE3
Mg S1_64935466 Sobic.001G443900 1 3.20E-07 Peptide transporter PTR2, putative, expressed AtPTR2; PEPTIDE TRANSPORTER2
Fe S1_19766414 Sobic.001G213400 1 4.92E-07 Homeodomain domain-containing protein, expressed Metal-dependent phosphohydrolase
K S6_45971634 Sobic.006G082200 6 4.25E-06 OSIGBa0160I14.4 protein MGT4, MRS2-3; MAGNESIUM TRANSPORTER4
B S4_52068874 Sobic.004G174600 4 5.45E-06 Putative multidrug resistanceprotein ABC transporter family protein
Rb S8_6186108 Sobic.008G058700 8 7.06E-06 Putative uncharacterized
protein OSJNBa0065C11.1 ZIP metal ion transporter family
P S1_64935466 Sobic.001G443900 1 7.14E-06 Peptide transporter PTR2, putative, expressed AtPTR2; PEPTIDE TRANSPORTER2
Zn S7_6880986 Sobic.007G064900 7 8.11E-06 Zn transporter ZIP5; ZN TRANSPORTER5 precursor
Sr S4_5986126 Sobic.004G073500 4 8.96E-06 Putative multidrug resistance
protein ABC transporter family protein
Mn S3_66960028 Sobic.003G349200 3 6.37E-05 Cation efflux family protein MTP11; ABC transporter family protein
Open in a new tab

Mn

Associated with amino acid, lipid, and carbohydrate metabolism, Mn is one of the essential elements critical to human and animal nutritional requirements (Aschner and Aschner, 2005). We identified significant GWAs in the putative sorghum homolog for a member of the metal transporter-encoding cation diffusion facilitator gene family, MTP11 (Sobic.003G349200; Table II). The Arabidopsis homolog, AtMTP11, has been reported to confer Mn tolerance and transport Mn2+ via a proton-antiport mechanism in Saccharomyces cerevisiae (Delhaize et al., 2007).

Cd

The seeds are a major source of essential nutrients but also can be a source of toxic heavy metals, including Cd. Contamination of groundwater and subsequent uptake and absorption by the plant can result in high levels of Cd contamination in the seed (Arao and Ae, 2003). GWA analysis identified significant SNPs associated with a paralogous set of cation-transporting ATPases (Fig. 4), Sobic.002G083000 and Sobic.002G083100. These are sorghum homologs of Arabidopsis HMA genes in the heavy metal-transporting subfamily of the P-type ATPases. AtHMA3 participates in the vacuolar storage of Cd in Arabidopsis, and a recent study revealed that HMA3 is a major-effect locus controlling natural variation in leaf Cd (Morel et al., 2009; Chao et al., 2012). These SNP alleles could be used immediately to potentially produce sorghum seed with lowered Cd2+ accumulation.

Ni

Ni is an essential nutrient required for plant growth. However, similar to Cd, high Ni concentrations in the soil can be toxic to the plant, resulting in reduced biomass and crop yield. The most significant SNP for Ni concentration in the SAP 2008 environment (and present in SAP 2012 and the ranked data set) was S6_53175238. This SNP is in LD with the candidate gene Sobic.006G164300, a homolog of the YELLOW STRIPE-LIKE3 (YSL) family of proteins (Table II). Originally identified in maize, the YSL proteins are a subfamily of transporters involved in metal chelate uptake, metal homeostasis, and long-distance transport (Curie et al., 2009). YSL3 is demonstrated to transport metals bound to nicotianamine (Curie et al., 2001), and in the metal hyperaccumulator Thlaspi caerulescens, YSL3 functions as an Ni-nicotianamine influx transporter (Gendre et al., 2007).

CONCLUSION

In this study, we utilized GWA mapping and rank transformation of phenotypic data to scale genotype-environment interactions and identify a number of genetic loci and candidate gene associations for immediate study and application to breeding strategies. The use of a multielement, or ionomic, approach in the analysis allowed for the identification of SNPs that confer multiple advantageous traits that can be selected for in breeding programs. We identified colocalization of significant SNPs for different elements, indicating potential coregulation through physiological processes of elemental uptake, transport, traffic, and sequestration. Our results suggest that combining elemental profiling with GWA approaches can be useful for understanding the genetic architecture underlying elemental accumulation and for improving the nutritional content of sorghum. The data and analysis scripts used for this article can be found at www.ionomicshub.org.

MATERIALS AND METHODS

Plant Material

The SAP has been described previously (Casa et al., 2008). Sorghum (Sorghum bicolor) seeds harvested from 407 lines that constitute the SAP were utilized for this study. SAP 2008 seeds were obtained from the Germplasm Resources Information Network and were produced in Lubbock, Texas, by the U.S. Department of Agriculture-Agricultural Research Service Cropping Systems Research Laboratory in 2008 and 2009. SAP 2012 seeds were produced in Puerto Vallarta, Mexico, in 2012. SAP 2013 seeds were produced in Florence, South Carolina, in 2013.

Phenotypic Elemental Analysis

Four seeds per replicate were weighed from each individual, and a minimum of two replicates from each line of the SAP 2008 and SAP 2013 panels were analyzed by ICP-MS. Each sample was digested with 2.5 mL of concentrated nitric acid at 95°C for 3 h. Elemental analysis was performed with ICP-MS for B, Na, Mg, Al, P, S, K, Ca, Mn, Fe, Co, Ni, Cu, Zn, As, Se, Rb, Sr, Mo, and Cd following established protocols (Baxter et al., 2010). A reference sample derived from a pool of sorghum seed samples was generated and run after every ninth sample to correct for ICP-MS run-to-run variation and within-run drift.

Data Processing and Analysis

Phenotype data were generated for 407 SAP lines. The genotyped by sequencing SNP markers for the SAP lines used in this study have been described previously (Morris et al., 2013). After removing SNPs with more than 20% missing data and minor allele frequencies below 0.05, genotype data for 78,012 SNPs remained. ANOVA was performed to calculate broad sense heritability using the lmer function in the lme4 package with the experimental replicates of the SAP using previously described methods (Van Poecke et al., 2007; Bates et al., 2014). To ensure normality in the data distribution of the phenotype, the Box-Cox procedure was carried out on the phenotype scores to identify the best transformation method (Box and Cox, 1964). The boxcox function in the MASS package was utilized in R to carry out the transformations (R Development Core Team, 2014; Ripley et al., 2015). To address potential confounding factors in the GWA analysis, specifically ICP run-to-run variation and the weight correction calculation, we used linear regression to compute residuals adjusted for ICP run and sample weight. These residuals were used to test for association with qualifying SNPs in the GWA analysis.

GWA

GWA was executed in R with GAPIT using CMLM (Zhang et al., 2010; Lipka et al., 2012). Significant associations were determined by estimates of false discovery rate (P = 0.05; Benjamini and Hochberg, 1995). The CMLM uses a VanRaden kinship matrix and the first three principal components as covariates to account for population structure. MLMM is based on EMMA (Kang et al., 2008) and relies on the iterative use of a simple K, or Q+K, mixed-model algorithm. The kinship term, K, provides a fine-grained estimate of familial relatedness between lines. In addition, GWA models often include a more granular measurement of population membership for each line, Q. To determine the necessity of using the more complex Q+K model to control for spurious allele associations, we analyzed QQ plots generated from MLMM GWA using a simple K model plus cofactors (Supplemental Fig. S5) and phenotypic distributions across known sorghum subpopulations (Supplemental Fig. S2). Phenotypic distributions across subpopulations were similar, suggesting that population structure does not play a strong role in elemental accumulation. The QQ plots show that, after the addition of major-effect loci to the model as cofactors, the P value distribution does not deviate drastically from the expected uniform distribution. These results indicate that the mixed model containing only the kinship matrix, K, plus cofactors is sufficient to control for spurious allele associations due to population structure and cryptic relatedness.

At each step of the MLMM, the phenotypic variance is divided into genetic, random, and explained variance. The most significant marker is included as a cofactor, and the variance components of the model are recalculated. With each successive iteration, the remaining genetic variance approaches zero, and an optimal model including cofactors that explains the genetic fraction of the phenotypic variance is determined. The MLMM method selects two models using stop criteria determined by two test statistics termed the mBonf and the extended Bayesian information criterion. The mBonf criterion selects a model wherein all cofactors have a P value below a Bonferroni-corrected threshold (Segura et al., 2012), and in our experiments, this was the more stringent of the two model selection criteria (i.e. it favored less complex models) and was used for all reported GWA analyses. We utilized a genome-wide significance threshold of P < 0.05 for the Bonferroni correction. A kinship matrix was constructed to correct for population structure and cryptic relatedness (Supplemental Table S10). The kinship matrix was estimated from all of the SNPs in the data set using the VanRaden method (VanRaden, 2008) in GAPIT (Lipka et al., 2012). Kinship was included as a random effect in the MLMM. In addition, the genetic variance partition described above provides an estimate of heritability, termed pseudoheritability (Kang et al., 2010; Segura et al., 2012), explained by the model at each step. The missing heritability can be calculated from the model at the optimal step (mBonf). The percent variance explained by the model is the difference between the genetic variance at step zero and the optimal step (Supplemental Table S7).

The data sets supporting the results of this article are available through the Purdue Ionomics Information Management System at http://www.ionomicshub.org/home/PiiMS/fileDownload?file=41.

Supplemental Data

The following supplemental materials are available.

Supplementary Material

Supplemental Data
supp_170_4_1989__index.html (2.4KB, html)

Acknowledgments

We thank Todd Mockler for his support of N.S. during the final stages of article writing as well as Kimberly Green and Janna Hutchinson for their contributions to sample preparation in the ionomics pipeline.

Glossary

Fe

iron

Zn

zinc

Mg

magnesium

Se

selenium

Ca

calcium

Cu

copper

As

arsenic

Cd

cadmium

Mo

molybdenum

ICP-MS

inductively coupled plasma mass spectroscopy

ICP

inductively coupled plasma

Mn

manganese

Co

cobalt

GWA

genome-wide association

SAP

Sorghum Association Panel

Na

sodium

Sr

strontium

Al

aluminum

Rb

rubidium

Ni

nickel

B

boron

P

phosphorus

S

sulfur

K

potassium

PCA

principal component analysis

MLMM

multilocus mixed model

CMLM

compressed mixed linear model

mBonf

multiple-Bonferroni

SNP

single-nucleotide polymorphism

T-DNA

transfer DNA

LD

linkage disequilibrium

QQ

quantile-quantile

Footnotes

1

This work was supported by the National Science Foundation (iHUB Visiting Scientist Program grant no. DBI 0953433 to David Salt and grant nos. EAGER 1450341 to I.B. and B.P.D., IOS 1126950 to I.B., IOS 0919739 to E.L.C.), the BMGF (grant no. OPP 1052924 to B.P.D.) and Chromatin, Inc..

[OPEN]

Articles can be viewed without a subscription.

References

  1. Amaducci S, Monti A, Venturi G (2004) Non-structural carbohydrates and fibre components in sweet and fibre sorghum as affected by low and normal input techniques. Ind Crops Prod 20: 111–118 [Google Scholar]
  2. Arao T, Ae N (2003) Genotypic variations in cadmium levels of rice grain. Soil Sci Plant Nutr 49: 473–479 [Google Scholar]
  3. Aschner JL, Aschner M (2005) Nutritional aspects of manganese homeostasis. Mol Aspects Med 26: 353–362 [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Ayana A, Bekele E (2000) Geographical patterns of morphological variation in sorghum (Sorghum bicolor (L.) Moench) germplasm from Ethiopia and Eritrea: quantitative characters. Euphytica 115: 91–104 [Google Scholar]
  5. Bates D, Mächler M, Bolker B, Walker S (2014) Fitting linear mixed-effects models using lme4. arXiv 1406.5823 (December 1, 2015)
  6. Baxter I, Brazelton JN, Yu D, Huang YS, Lahner B, Yakubova E, Li Y, Bergelson J, Borevitz JO, Nordborg M, et al. (2010) A coastal cline in sodium accumulation in Arabidopsis thaliana is driven by natural variation of the sodium transporter AtHKT1;1. PLoS Genet 6: e1001193. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Baxter I, Dilkes BP (2012) Elemental profiles reflect plant adaptations to the environment. Science 336: 1661–1663 [DOI] [PubMed] [Google Scholar]
  8. Baxter IR, Gustin JL, Settles AM, Hoekenga OA (2013) Ionomic characterization of maize kernels in the intermated B73 × Mo17 population. Crop Sci 53: 208–220 [Google Scholar]
  9. Baxter IR, Vitek O, Lahner B, Muthukumar B, Borghi M, Morrissey J, Guerinot ML, Salt DE (2008) The leaf ionome as a multivariable system to detect a plant’s physiological status. Proc Natl Acad Sci USA 105: 12081–12086 [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Baxter IR, Ziegler G, Lahner B, Mickelbart MV, Foley R, Danku J, Armstrong P, Salt DE, Hoekenga OA (2014) Single-kernel ionomic profiles are highly heritable indicators of genetic and environmental influences on elemental accumulation in maize grain (Zea mays). PLoS ONE 9: e87628. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Benjamini Y, Hochberg Y (1995) Controlling the false discovery rate: a practical and powerful approach to multiple testing. J R Stat Soc B 57: 289–300 [Google Scholar]
  12. Box GE, Cox DR (1964) An analysis of transformations. J R Stat Soc B 26: 211–252 [Google Scholar]
  13. Broadley MR, White PJ (2012) Some elements are more equal than others: soil-to-plant transfer of radiocaesium and radiostrontium, revisited. Plant Soil 355: 23–27 [Google Scholar]
  14. Buescher E, Achberger T, Amusan I, Giannini A, Ochsenfeld C, Rus A, Lahner B, Hoekenga O, Yakubova E, Harper JF, et al. (2010) Natural genetic variation in selected populations of Arabidopsis thaliana is associated with ionomic differences. PLoS ONE 5: e11081. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Cakmak I. (2008) Enrichment of cereal grains with zinc: agronomic or genetic biofortification? Plant Soil 302: 1–17 [Google Scholar]
  16. Casa AM, Pressoir G, Brown PJ, Mitchell SE, Rooney WL, Tuinstra MR, Franks CD, Kresovich S (2008) Community resources and strategies for association mapping in sorghum. Crop Sci 48: 30–40 [Google Scholar]
  17. Chao DY, Silva A, Baxter I, Huang YS, Nordborg M, Danku J, Lahner B, Yakubova E, Salt DE (2012) Genome-wide association studies identify heavy metal ATPase3 as the primary determinant of natural variation in leaf cadmium in Arabidopsis thaliana. PLoS Genet 8: e1002923. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Connolly EL, Campbell NH, Grotz N, Prichard CL, Guerinot ML (2003) Overexpression of the FRO2 ferric chelate reductase confers tolerance to growth on low iron and uncovers posttranscriptional control. Plant Physiol 133: 1102–1110 [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Conover WJ, Iman RL (1981) Rank transformations as a bridge between parametric and nonparametric statistics. Am Stat 35: 124–129 [Google Scholar]
  20. Curie C, Cassin G, Couch D, Divol F, Higuchi K, Le Jean M, Misson J, Schikora A, Czernic P, Mari S (2009) Metal movement within the plant: contribution of nicotianamine and yellow stripe 1-like transporters. Ann Bot (Lond) 103: 1–11 [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Curie C, Panaviene Z, Loulergue C, Dellaporta SL, Briat JF, Walker EL (2001) Maize yellow stripe1 encodes a membrane protein directly involved in Fe(III) uptake. Nature 409: 346–349 [DOI] [PubMed] [Google Scholar]
  22. Das P, Samantaray S, Rout GR (1997) Studies on cadmium toxicity in plants: a review. Environ Pollut 98: 29–36 [DOI] [PubMed] [Google Scholar]
  23. Delhaize E, Gruber BD, Pittman JK, White RG, Leung H, Miao Y, Jiang L, Ryan PR, Richardson AE (2007) A role for the AtMTP11 gene of Arabidopsis in manganese transport and tolerance. Plant J 51: 198–210 [DOI] [PubMed] [Google Scholar]
  24. Gendre D, Czernic P, Conéjéro G, Pianelli K, Briat JF, Lebrun M, Mari S (2007) TcYSL3, a member of the YSL gene family from the hyper-accumulator Thlaspi caerulescens, encodes a nicotianamine-Ni/Fe transporter. Plant J 49: 1–15 [DOI] [PubMed] [Google Scholar]
  25. Gomez-Becerra HF, Yazici A, Ozturk L, Budak H, Peleg Z, Morgounov A, Fahima T, Saranga Y, Cakmak I (2010) Genetic variation and environmental stability of grain mineral nutrient concentrations in Triticum dicoccoides under five environments. Euphytica 171: 39–52 [Google Scholar]
  26. Graham R, Senadhira D, Beebe S, Iglesias C, Monasterio I (1999) Breeding for micronutrient density in edible portions of staple food crops: conventional approaches. Field Crops Res 60: 57–80 [Google Scholar]
  27. Gregorio GB, Senadhira D, Htut H, Graham RD (2000) Breeding for trace mineral density in rice. Food Nutr Bull 21: 382–386 [Google Scholar]
  28. Huang X, Wei X, Sang T, Zhao Q, Feng Q, Zhao Y, Li C, Zhu C, Lu T, Zhang Z, et al. (2010) Genome-wide association studies of 14 agronomic traits in rice landraces. Nat Genet 42: 961–967 [DOI] [PubMed] [Google Scholar]
  29. Hutchin ME, Vaughan BE (1968) Relation between simultaneous Ca and Sr transport rates in isolated segments of vetch, barley, and pine roots. Plant Physiol 43: 1913–1918 [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Kang HM, Sul JH, Service SK, Zaitlen NA, Kong SY, Freimer NB, Sabatti C, Eskin E (2010) Variance component model to account for sample structure in genome-wide association studies. Nat Genet 42: 348–354 [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Kang HM, Zaitlen NA, Wade CM, Kirby A, Heckerman D, Daly MJ, Eskin E (2008) Efficient control of population structure in model organism association mapping. Genetics 178: 1709–1723 [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Lahner B, Gong J, Mahmoudian M, Smith EL, Abid KB, Rogers EE, Guerinot ML, Harper JF, Ward JM, McIntyre L, et al. (2003) Genomic scale profiling of nutrient and trace elements in Arabidopsis thaliana. Nat Biotechnol 21: 1215–1221 [DOI] [PubMed] [Google Scholar]
  33. Lestienne I, Icard-Vernière C, Mouquet C, Picq C, Trèche S (2005) Effects of soaking whole cereal and legume seeds on iron, zinc and phytate contents. Food Chem 89: 421–425 [Google Scholar]
  34. Lipka AE, Tian F, Wang Q, Peiffer J, Li M, Bradbury PJ, Gore MA, Buckler ES, Zhang Z (2012) GAPIT: genome association and prediction integrated tool. Bioinformatics 28: 2397–2399 [DOI] [PubMed] [Google Scholar]
  35. Ma JF, Yamaji N, Mitani N, Xu XY, Su YH, McGrath SP, Zhao FJ (2008) Transporters of arsenite in rice and their role in arsenic accumulation in rice grain. Proc Natl Acad Sci USA 105: 9931–9935 [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Maathuis FJ. (2009) Physiological functions of mineral macronutrients. Curr Opin Plant Biol 12: 250–258 [DOI] [PubMed] [Google Scholar]
  37. Marschner H, Marschner P (2012) Marschner’s Mineral Nutrition of Higher Plants, Vol 89 Academic Press, San Diego, CA [Google Scholar]
  38. Mengesha MH. (1966) Chemical composition of teff (Eragrostis tef) compared with that of wheat, barley and grain sorghum. Econ Bot 20: 268–273 [Google Scholar]
  39. Morel M, Crouzet J, Gravot A, Auroy P, Leonhardt N, Vavasseur A, Richaud P (2009) AtHMA3, a P1B-ATPase allowing Cd/Zn/Co/Pb vacuolar storage in Arabidopsis. Plant Physiol 149: 894–904 [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Morris GP, Ramu P, Deshpande SP, Hash CT, Shah T, Upadhyaya HD, Riera-Lizarazu O, Brown PJ, Acharya CB, Mitchell SE, et al. (2013) Population genomic and genome-wide association studies of agroclimatic traits in sorghum. Proc Natl Acad Sci USA 110: 453–458 [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Neucere NJ, Sumrell G (1980) Chemical composition of different varieties of grain sorghum. J Agric Food Chem 28: 19–21 [DOI] [PubMed] [Google Scholar]
  42. Norton GJ, Deacon CM, Xiong L, Huang S, Meharg AA, Price AH (2010) Genetic mapping of the rice ionome in leaves and grain: identification of QTLs for 17 elements including arsenic, cadmium, iron and selenium. Plant Soil 329: 139–153 [Google Scholar]
  43. Ortiz-Monasterio J, Palacios-Rojas N, Meng E, Pixley K, Trethowan R, Pena R (2007) Enhancing the mineral and vitamin content of wheat and maize through plant breeding. J Cereal Sci 46: 293–307 [Google Scholar]
  44. Ozgen S, Busse JS, Palta JP (2011) Influence of root zone calcium on shoot tip necrosis and apical dominance of potato shoot: simulation of this disorder by ethylene glycol tetra acetic acid and prevention by strontium. HortScience 46: 1358–1362 [Google Scholar]
  45. Peleg Z, Cakmak I, Ozturk L, Yazici A, Jun Y, Budak H, Korol AB, Fahima T, Saranga Y (2009) Quantitative trait loci conferring grain mineral nutrient concentrations in durum wheat × wild emmer wheat RIL population. Theor Appl Genet 119: 353–369 [DOI] [PubMed] [Google Scholar]
  46. Queen WH, Fleming HW, O’Kelley JC (1963) Effects on Zea mays seedlings of a strontium replacement for calcium in nutrient media. Plant Physiol 38: 410–413 [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. R Development Core Team (2014) R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna. http://cran.r-project. org (December 1, 2015)
  48. Ragaee S, Abdel-Aal ESM, Noaman M (2006) Antioxidant activity and nutrient composition of selected cereals for food use. Food Chem 98: 32–38 [Google Scholar]
  49. Ripley B, Venables B, Bates DM, Hornik K, Gebhardt A, Firth D, Ripley MB (2015) Package ‘MASS’. In Retrieved from CRAN: http://cran.r-project.org/web/packages/MASS/MASS.pdf
  50. Salt DE, Baxter I, Lahner B (2008) Ionomics and the study of the plant ionome. Annu Rev Plant Biol 59: 709–733 [DOI] [PubMed] [Google Scholar]
  51. Sauvage C, Segura V, Bauchet G, Stevens R, Do PT, Nikoloski Z, Fernie AR, Causse M (2014) Genome-wide association in tomato reveals 44 candidate loci for fruit metabolic traits. Plant Physiol 165: 1120–1132 [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Schroeder JI, Delhaize E, Frommer WB, Guerinot ML, Harrison MJ, Herrera-Estrella L, Horie T, Kochian LV, Munns R, Nishizawa NK, et al. (2013) Using membrane transporters to improve crops for sustainable food production. Nature 497: 60–66 [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Segura V, Vilhjálmsson BJ, Platt A, Korte A, Seren Ü, Long Q, Nordborg M (2012) An efficient multi-locus mixed-model approach for genome-wide association studies in structured populations. Nat Genet 44: 825–830 [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. Shi R, Li H, Tong Y, Jing R, Zhang F, Zou C (2008) Identification of quantitative trait locus of zinc and phosphorus density in wheat (Triticum aestivum L.) grain. Plant Soil 306: 95–104 [Google Scholar]
  55. Simić D, Mladenović Drinić S, Zdunić Z, Jambrović A, Ledenčan T, Brkić J, Brkić A, Brkić I (2012) Quantitative trait loci for biofortification traits in maize grain. J Hered 103: 47–54 [DOI] [PubMed] [Google Scholar]
  56. Tian F, Bradbury PJ, Brown PJ, Hung H, Sun Q, Flint-Garcia S, Rocheford TR, McMullen MD, Holland JB, Buckler ES (2011) Genome-wide association study of leaf architecture in the maize nested association mapping population. Nat Genet 43: 159–162 [DOI] [PubMed] [Google Scholar]
  57. Van Poecke RM, Sato M, Lenarz-Wyatt L, Weisberg S, Katagiri F (2007) Natural variation in RPS2-mediated resistance among Arabidopsis accessions: correlation between gene expression profiles and phenotypic responses. Plant Cell 19: 4046–4060 [DOI] [PMC free article] [PubMed] [Google Scholar]
  58. VanRaden PM. (2008) Efficient methods to compute genomic predictions. J Dairy Sci 91: 4414–4423 [DOI] [PubMed] [Google Scholar]
  59. Vert G, Grotz N, Dédaldéchamp F, Gaymard F, Guerinot ML, Briat JF, Curie C (2002) IRT1, an Arabidopsis transporter essential for iron uptake from the soil and for plant growth. Plant Cell 14: 1223–1233 [DOI] [PMC free article] [PubMed] [Google Scholar]
  60. Vreugdenhil D, Aarts M, Koornneef M, Nelissen H, Ernst W (2004) Natural variation and QTL analysis for cationic mineral content in seeds of Arabidopsis thaliana. Plant Cell Environ 27: 828–839 [Google Scholar]
  61. Waters BM, Grusak MA (2008) Quantitative trait locus mapping for seed mineral concentrations in two Arabidopsis thaliana recombinant inbred populations. New Phytol 179: 1033–1047 [DOI] [PubMed] [Google Scholar]
  62. White PJ, Broadley MR (2005) Biofortifying crops with essential mineral elements. Trends Plant Sci 10: 586–593 [DOI] [PubMed] [Google Scholar]
  63. World Health Organization (2002) The World Health Report 2002: Reducing Risks, Promoting Healthy Life. World Health Organization, Geneva, Switzerland [Google Scholar]
  64. Yi Y, Guerinot ML (1996) Genetic evidence that induction of root Fe(III) chelate reductase activity is necessary for iron uptake under iron deficiency. Plant J 10: 835–844 [DOI] [PubMed] [Google Scholar]
  65. Zhang M, Pinson SR, Tarpley L, Huang XY, Lahner B, Yakubova E, Baxter I, Guerinot ML, Salt DE (2014) Mapping and validation of quantitative trait loci associated with concentrations of 16 elements in unmilled rice grain. Theor Appl Genet 127: 137–165 [DOI] [PMC free article] [PubMed] [Google Scholar]
  66. Zhang T, Zhao X, Wang W, Pan Y, Huang L, Liu X, Zong Y, Zhu L, Yang D, Fu B (2012) Comparative transcriptome profiling of chilling stress responsiveness in two contrasting rice genotypes. PLoS ONE 7: e43274. [DOI] [PMC free article] [PubMed] [Google Scholar]
  67. Zhang Z, Ersoz E, Lai CQ, Todhunter RJ, Tiwari HK, Gore MA, Bradbury PJ, Yu J, Arnett DK, Ordovas JM, et al. (2010) Mixed linear model approach adapted for genome-wide association studies. Nat Genet 42: 355–360 [DOI] [PMC free article] [PubMed] [Google Scholar]
  68. Zhao K, Tung CW, Eizenga GC, Wright MH, Ali ML, Price AH, Norton GJ, Islam MR, Reynolds A, Mezey J, et al. (2011) Genome-wide association mapping reveals a rich genetic architecture of complex traits in Oryza sativa. Nat Commun 2: 467. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplemental Data
supp_170_4_1989__index.html (2.4KB, html)
supp_pp.15.01971_01971suppfigstbls.pdf (9.4MB, pdf)
supp_pp.15.01971_PP2015-01971DR1_Supplemental_table_1.xlsx (44.4KB, xlsx)
supp_pp.15.01971_PP2015-01971DR1_Supplemental_table_5.xlsx (776.2KB, xlsx)
supp_pp.15.01971_PP2015-01971DR1_Supplemental_table_6.xlsx (59.4KB, xlsx)
supp_pp.15.01971_PP2015-01971DR1_Supplemental_table_7.xlsx (12.2KB, xlsx)
supp_pp.15.01971_PP2015-01971DR1_Supplemental_table_8.xlsx (44.1KB, xlsx)
supp_pp.15.01971_PP2015-01971DR1_Supplemental_table_9.xlsx (71.5KB, xlsx)
supp_pp.15.01971_PP2015-01971DR1_Supplemental_table_10.csv (1.6MB, csv)

Articles from Plant Physiology are provided here courtesy of Oxford University Press

RESOURCES