PMCCPMCCPMCC

Search tips
Search criteria 

Advanced

 
Logo of annbotAboutAuthor GuidelinesEditorial BoardAnnals of Botany
 
Ann Bot. 2010 June; 105(7): 1221–1234.
Published online 2010 March 17. doi:  10.1093/aob/mcq050
PMCID: PMC2887070

Quantitative trait loci affecting seed mineral concentrations in Brassica napus grown with contrasting phosphorus supplies

Abstract

Background and Aims

Phosphorus (P) deficiency is one of the major limitations for crop production. A significant relationship exists between plant P uptake from soils and the accumulation of P and other mineral elements in seeds. The aims of this study were to identify and characterize genetic loci (QTLs) controlling the accumulation of mineral elements in seeds of Brassica napus grown with contrasting P availabilities.

Methods

A population of 124 recombinant inbred lines derived from a cross between P-inefficient ‘B104-2’ and P-efficient ‘Eyou Changjia’ was used for phenotypic investigation and QTL analysis. Two-year field trials were conducted with two P treatments. Concentrations of mineral elements (P, Ca, Mg, Fe, Zn, Cu and Mn) in seeds were determined and QTLs were identified by composite interval mapping.

Key Results

There was significant genetic variation in seed concentrations of all mineral elements, and correlations between most elements were observed. A total of 78 putative QTLs (45 under the normal-P conditions and 33 under the low-P conditions) were detected, covering 17 linkage groups and accounting for 7·84–18·1 % of the phenotypic variation. Sixteen of these QTLs were identified in at least two environments, and co-location of QTLs for different mineral traits was found on several linkage groups. By in silico mapping, 21 genes involved in ion homeostasis in Arabidopsis were mapped to the QTL intervals identified in B. napus.

Conclusions

The accumulation of mineral elements in seeds is controlled by multiple genes. Common physiological and molecular mechanisms could be involved in the accumulation of several mineral elements, and genes involved in these processes in B. napus are suggested. These results offer insights to the genetic basis of seed mineral accumulation across different P levels in B. napus.

Keywords: Brassica napus, phosphorus deficiency, seed mineral concentration, quantitative trait locus, in silico mapping, recombinant inbred line

INTRODUCTION

Mineral elements such as phosphorus (P), calcium (Ca), iron (Fe), magnesium (Mg), manganese (Mn) and zinc (Zn) perform a variety of functions in plant cells, and are essential for growth and development in plants as well as in animals and humans. The acquisition of mineral elements by higher plants is primarily from the soil, while humans obtain all essential elements mostly from higher plants (Grusak and Cakmak, 2005; White and Broadley, 2009). Unfortunately, it is estimated that over three billion people are currently malnourished because of lack of minerals, especially iron and zinc, in their diet (Welch and Graham, 2004; White and Broadley, 2009). Biofortification, which aims to increase mineral concentrations in edible crops, either agronomically or genetically using both conventional breeding and modern biotechnology, has been adopted by plant scientists to address this problem, and is considered to be the most promising and cost-effective approach to alleviate mineral malnutrition (White and Broadley, 2005, 2009). However, the accumulation of minerals in seeds is a complex phenomenon, which is most likely controlled by a number of genes. The movement of mineral elements from soils to seeds involves their mobilization from soils, uptake by roots, translocation to the shoot, redistribution within the plant and deposition in seeds (Grusak and DellaPenna, 1999; White and Broadley, 2009). The genetic basis of many of these processes is unknown.

P is an essential macronutrient for plants (Marschner, 1995). Due to its high adsorption capacity in soils (e.g. adsorbed by Al- and Fe-oxides in acid soils), mineralization (e.g. to calcium phosphates such as apatite) and formation of organic P complexes, P is one of the least available plant nutrients in the rhizosphere. Indeed, P deficiency is considered to be one of the major limitations for crop production worldwide and is probably the most widely studied macronutrient deficiency (Schachtman et al., 1998; Raghothama and Karthikeyan, 2005). Rapeseed (Brassica napus), which has high P requirement for its optimal seed yield and quality, is commonly used in animal feed or as an oil source for humans. Lack of available P in soils seriously limits the production of B. napus and genetic approaches to improve its P-use efficiency is considered as an efficient and promising way to maintain or increase yields whilst reducing P inputs.

Quantitative trait locus (QTL) analysis is a powerful genetic approach to dissect complex traits and to investigate natural genetic variation within a species (Paran and Zamir, 2003). In recent years, different genetic mapping populations have been developed to identify QTLs that contribute to mineral traits in various species. For example, in the model plant Arabidopsis thaliana, QTLs for phosphate accumulation in seed and leaf (Bentsink et al., 2003), shoot Ca concentration (White, 2005), shoot K concentration (Harada and Leigh, 2006), seed sulfate concentration (Loudet et al., 2007) and the accumulation of other mineral elements in shoots and seed (Payne et al., 2004; Vreugdenhil et al., 2004; Waters and Grusak, 2008a; Ghandilyan et al., 2009a, b) have been identified. Similarly, QTLs for the accumulation of mineral elements in seeds of other plant species such as rice (Garcia-Oliveira et al., 2009), wheat (Peleg et al., 2009) and Medicago truncatula (Sankaran et al., 2009) have been identified. Recently, QTLs have been identified in crop Brassicaceae that influence seed and leaf P (Zhao et al., 2008), shoot Ca and Mg (Broadley et al., 2008), shoot P (Hammond et al., 2009), shoot Na, Mg, Mn, Fe, Al, Sr, Zn, P (Wu et al., 2008), shoot B, Ca, Fe, Cu, Mg, P, Zn (Liu et al., 2009) and shoot K (White et al., 2010).

Arabidopsis thaliana is an excellent model system with a small and relatively simple genome. The existence of a common ancestor for Brassica and Arabidopsis enables comparative genome mapping techniques. Recent research has indicated 24 conserved chromosomal blocks in A. thaliana and other Brassicaceae (Schranz et al., 2006). By in silico mapping between A. thaliana and B. napus, several genes involved in ion uptake and transport have been mapped to the QTL intervals in B. napus affecting shoot mineral concentrations (Liu et al., 2009). Such information can provide clues for researchers to identify genes related to mineral accumulation and to finally uncover genetic networks that control plant ion homeostasis. It is well known that the major storage form for P in seeds of plants is myo-inositol-1,2,3,4,5,6-hexakisphosphate (IP6, or phytic acid) which often chelates metal ions, such as Ca, Mn, Zn and Fe (Raboy, 1997). However, little is known about the relationship between the accumulation of P and other mineral elements, such as Ca, Mg, Fe, Mn, Cu and Zn, in seeds of B. napus.

To indentify genetic loci that control the accumulation of mineral elements in seeds of B. napus, 2-year field trials were conducted with two P treatments using an F10 recombinant inbred line (RIL) population. By employing WinQTL Cartographer 2·5, a total of 78 putative QTLs were detected on 17 linkage groups. Furthermore, 21 functional genes were mapped to the interval of the QTLs affecting seed mineral concentrations by in silico comparative genome mapping techniques. These results improve our genetic understanding of the accumulation of mineral elements in seeds and plant responses to P availability.

MATERIALS AND METHODS

Plant materials and genetic linkage map

A set of 124 F10 lines of a Brassica napus L. RIL population was developed by single-seed descent from a cross between cv. ‘B104-2’ (B) and cv. ‘Eyou Changjia’ (E). Genotype ‘Eyou Changjia’ is a P-efficient cultivar, which shows a higher P efficiency coefficient (defined as the quotient of shoot dry weight at the seedling stage, or seed yield at maturity, under low P availability and that under high P availability) than ‘B104-2’, a P-inefficient cultivar (Duan et al., 2009). The RIL population was previously genotyped with 553 molecular markers by Yang et al. (2010). Markers included 202 simple sequence repeat (SSR), 62 amplified fragment length polymorphism (AFLP), 234 sequence-related amplified polymorphism (SRAP) and 55 specific functional markers. Primer sequences for the SSR markers were obtained from various public sources (UK, http://www.brassica.bbsrc.ac.uk/BrassicaDB, prefixed by OL and Na; Australia, http://www.hornbill.cspp.latrobe.edu.au, prefixed by sA; Canada http://www.brassica.agr.gc.ca/index_e.shtml, prefixed by sR and sN; Japan, Suwabe et al., 2002, prefixed by BRMS; France, Piquemal et al., 2005, prefixed by BRAS, CB and MR; private communications prefixed by CNU, niab and HBr). The AFLP procedure was performed according to the protocol developed by Vos et al. (1995), and the SRAP according to Li and Quiros (2001). The specific functional markers were developed in the laboratory according to the method described by Li et al. (2006), and homologous sequences in B. napus of Arabidopsis thaliana functional genes related to P homeostasis were used for primer design. In general, the genetic map covered a total length of 1592·7 cM with an average distance of 2·88 cM between adjacent markers. The genetic distance of the 19 linkage groups varied from 13·4 to 148·2 cM. The order of SSR markers in each linkage group was in agreement with previous reports (Piquemal et al., 2005; Long et al., 2007). Alignment between the B. napus linkage groups and the A. thaliana genome was conducted using the comparative mapping approach described by Long et al. (2007). Twelve syntenic blocks and 17 insertion islands were identified between the A. thaliana genome and the B. napus linkage groups (M. Yang et al., unpubl. res.).

Field trial

The phenotypes of the RIL population were tested by 2-year field trials in a paddy soil with 12·5 mg kg−1 of Olsen-P located in Daye, Hubei Province, China. The field trials were conducted during 2006–2007 and 2007–2008, respectively, with two P treatments of 30 kg P2O5 ha−1 (low-P) and 90 kg P2O5 ha−1 (normal-P). Each treatment was performed in triplicate. Each replicate consisted of 124 lines as well as two parents, and each genotype was planted in two rows with nine plants per row. The application amount of N, K and B fertilizers was calculated according to the following nutrient rates: 180 kg N ha−1, 83 kg K ha−1 and 15 kg H3BO3 ha−1.

Mineral concentration analysis in seeds

After harvesting in May (2007 and 2008), all seed samples were weighed after drying at 60 °C to a constant mass. About 500 mg of each sample was weighed and then transferred to a digestion vessel. To each sample, 8 mL concentrated nitric acid and 2 mL of perchloric acid were added. This mixture was kept over night and then digested at 140 °C for 2 h and 220 °C for 4 h. After cooling, digested samples were diluted to 50 mL using deionized water and finally transferred to 50-mL polyethylene flasks. Concentrations of P, Ca, Mg, Cu, Fe, Mn and Zn were determined by inductively coupled plasma-optical emission spectroscopy (VISTA-MPX; Varian, Inc., USA).

Statistical analysis

Seed mineral concentrations were calculated for all lines, and statistical analysis for all traits was conducted using the ‘PROC GLM’ procedure of SAS 8·1 (SAS Institute, Cary, NC, USA). Histograms and normality tests (Pearson chi-square test) were used to describe the variation of the phenotypic traits, and Pearson's correlation analysis was performed to examine their phenotypic association. The broad-sense heritability estimate (h2) was calculated for each trait across different P environments according to the approach described by Shi et al. (2009).

QTL analysis and association with functional genes

QTL analysis was performed with WinQTL Cartographer 2·5 (Wang et al., 2006). The composite interval mapping (CIM) model was used to scan the genetic map and estimate the likelihood of a QTL and its corresponding effect at every 2 cM. The genome wide LOD score threshold for QTL detection was determined for each trait by performing a 1000-permutation test at a significance level of P = 0·05 (Churchill and Doerge, 1994). The estimated additive effect and the percentage of phenotypic variation explained by each putative QTL were obtained by the software in the CIM model. Then QTLs for the same trait, which were detected in both years and mapped to the same region of a chromosome, were integrated into one QTL by QTL meta-analysis following the method reported by Shi et al. (2009). All QTLs identified were named using italic trait abbreviation, followed by ‘NP’ or ‘LP’ (abbreviation of normal-P or low-P, indicating the QTL was detected under normal-P or low-P conditions), and then suffixed with the corresponding linkage group number (A1–A10 or C1–C9) and the serial letter (a, b, c … ) in the same linkage group. For example, CaNP-A7a indicates a QTL for Ca concentration detected in the NP treatment at the first locus of chromosome A7. After comparative mapping between the B. napus linkage groups and the A. thaliana genome, genes affecting mineral concentrations in each syntenic block of A. thaliana were then aligned with the linkage map according to the position of close anchor marker(s) in the same syntenic block. If the position of an aligned gene(s) was located in the confidence interval of a QTL, the orthologous gene(s) was considered to be associated with the target QTL.

RESULTS

Phenotypic diversity for seed mineral concentrations

The mean values, ranges and heritability estimates of seven seed mineral concentrations of the two parental lines and the 124-line RIL population under four environmental conditions are presented in Table 1. Analysis of variance (ANOVA) showed significant (P < 0·05) genetic variation for all the minerals analysed as well as for environmental effects (data not shown). All the mineral traits showed large variation among RILs at both P supplies, but wider ranges were observed under low-P conditions than that under normal-P conditions, except for P in 2007 and Cu in both years. Among the elements studied, P was the most abundant element ranging from 3508 µg g−1 to 12397 µg g−1 (Table 1). Among the micro-nutrients, Mn was most abundant ranging from 41·79 µg g−1 to 103·66 µg g−1. For the two parental lines, higher concentrations were observed for all mineral elements except Cu under normal-P conditions than under low-P conditions. The P-efficient genotype ‘Eyou Changjia’ accumulated more P in seeds in low-P environments than the P-inefficient genotype ‘B104-2’. Seed concentrations of nearly all minerals assayed in the RIL population fell into relatively normal distributions under both normal-P (Fig. 1) and low-P supply (Fig. 2). Considerable transgressive segregations for seed mineral concentrations were generally observed in both directions, suggesting multiple genes act on these traits. However, there were some exceptions, such as seed P concentrations in plants with normal-P supply and seed Ca concentrations in plants with low-P supply, for which the transgression was solely towards higher concentrations. To estimate the proportion of phenotypic variation in the population that can be attributed to genetic variation, the h2 values for all traits were calculated. These varied from 0·29 for Mg to 0·70 for P (Table 1).

Table 1.
Mean values, ranges and heritability (h2) estimates of seven seed mineral concentrations (μg g1) in 124 recombinant inbred lines and two parental lines of Brassica napus under four P conditions
Fig. 1.
Frequency distributions of seed mineral concentrations in a Brassica napus ‘B104-2’ × ‘Eyou Changjia’ RIL population grown under normal-phosphorus conditions in 2-year field trials. B and E indicate the position ...
Fig. 2.
Frequency distributions of seed mineral concentrations in a Brassica napus ‘B104-2’ × ‘Eyou Changjia’ RIL population grown under low-phosphorus conditions in 2-year field trials. B and E indicate the position of ...

Correlation analysis showed that positive correlations existed among mineral traits (Table 2). Macro-nutrients (P, Ca and Mg) showed relatively high and significant correlations with each other in all environments with the highest value (0·86) between P and Mg in 2007. Among the micro-nutrients (Zn, Cu, Fe and Mn), the correlations between Zn and Cu were significant under both P levels in both years, but correlations between other elements were often either non-significant or inconsistent between years and/or P treatments. Significant positive correlations were observed between some macro-nutrients and micro-nutrients, such as P and Zn, and Ca and Zn, and weak positive correlations were observed in others, especially under low-P conditions. Correlations between seed mineral traits and seed yield were weak, or even negative, except between Mg and seed yield in 2007 under normal-P conditions (Table 2).

Table 2.
Correlations among the concentrations of seed minerals and seed yield (SY) in 124 RI lines of Brassica napus grown under low-phosphorus (below diagonal) and normal-phosphorus (above diagonal) conditions

General characteristics of the QTLs detected

A total of 78 putative QTLs, including 45 under normal-P conditions and 33 under low-P conditions, were detected in the 2-year field trials (Table 3). These QTLs were mainly located on 17 linkage groups, explaining 7·84–18·1 % of the phenotypic variation. About 65 % of these QTLs were co-located with at least one other QTL. Alleles from the P-efficient cultivar ‘Eyou Changjia’ contributed to increase seed mineral concentrations at 44 QTLs (57 %), while alleles from the P-inefficient cultivar ‘B104-2’ contributed to increase seed mineral concentrations at 34 QTLs (43 %). Chromosome C6 had the largest number of QTLs (a total of 11 QTLs), including QTLs for all nutrients except Mn, whereas the smallest number of QTLs was observed on C5 and C9 (one QTL each). For each mineral trait, at least two QTLs were located on different chromosomes. Sixteen QTLs were identified in at least two environments, but the rest were detected in only one environment. The co-location of QTLs occurred on most chromosomes, except A8, C5 and C9 (Fig. 3).

Fig. 3.Fig. 3.Fig. 3.
Chromosomal locations of putative QTLs for seed mineral concentrations in a Brassica napus ‘B104-2’ × ‘Eyou Changjia’ RIL population and genes aligned with the pseudo chromosomes of Arabidopsis thaliana. The length ...
Table 3.
Significant QTLs for seed mineral concentrations in a Brassica napus ‘B104-2’ × ‘Eyou Changjia’ RIL population grown under two phosphorus conditions

QTLs detected for each mineral across different P levels

Phosphorus

In total, 12 putative QTLs explaining 7·84–15·94 % of the phenotypic variation for seed P concentration were detected. These were located on ten different chromosomes and had LOD scores ranging from 3·27 to 5·30. Seven QTLs were detected under normal-P conditions, and five were detected under low-P conditions. Three QTLs (PLP-A1, PLP-C3b and PNP-C6a), locating on A1, C3 and C6, respectively, were identified in both years and accounted for a high percentage of phenotypic variation (15·94 %). However, no QTLs were common to both P conditions. Favourable alleles for increasing seed P concentration were contributed by the P-efficient cultivar ‘Eyou Changjia’ at seven loci and from the P-inefficient cultivar ‘B104-2’ at five loci (Table 3).

Calcium

For seed Ca concentration, a total of 14 significant QTLs explaining 8·44–18·10 % of the phenotypic variation were identified. These were located on nine chromosomes and had LOD scores ranging from 3·19 to 6·35 (Table 3). Two QTLs, CaNP-C6a and CaLP-C2b, were detected in both years. One QTL (CaLP-A1) explaining the highest phenotypic variation (18·10 %) was only detected under low-P conditions in 2007. Several QTLs on different chromosomes (A1, A4, A5, A7, A9, C2 and C6) co-located with the QTLs controlling seed concentrations of other mineral elements. Favourable alleles for increasing seed Ca concentration were contributed by ‘Eyou Changjia’ at ten loci and by ‘B104-2’ at four loci.

Magnesium

A total of 12 significant QTLs, scattered across nine chromosomes with LOD scores ranging from 3·28 to 6·23, were identified for seed Mg concentration. These QTLs accounted for a range of 9·11–16·49 % of the total phenotypic variation. Six of the QTLs were detected under low-P conditions. Two QTLs, MgNP-C6b and MgLP-C8, were detected in both years. The QTL, MgNP-A7, located on A7 accounted for the largest proportion of phenotypic variation (16·49 %). Most QTLs for seed Mg concentration overlapped with QTLs affecting the concentrations of other mineral elements, especially P and Ca. A higher Mg concentration was conferred by the ‘Eyou Changjia’ allele at seven loci and by the ‘B104-2’ allele at five loci (Table 3).

Zinc

A total of ten significant QTLs, with LOD scores ranging from 3·76 to 4·84, distributed across eight chromosomes were associated with seed Zn concentration. Seven QTLs were detected under normal-P conditions and three QTLs were found under low-P conditions. One QTL, ZnLP-C3, detected in both years under low-P conditions, explained a high percentage of phenotypic variation (13·39 %), while another QTL, ZnNP-A10, was detected in both years under normal-P conditions. Six QTLs on five chromosomes (A10, C1, C3, C6 and C7) co-located with QTLs affecting seed concentrations of the other mineral elements. Higher seed Zn concentration was associated with the ‘Eyou Changjia’ alleles at six QTLs and by the 104-2 alleles at four QTLs. Interestingly, favourable alleles for increasing seed Zn concentration under low-P conditions were all contributed by ‘Eyou Changjia’.

Copper

Eleven putative QTLs explaining 7·84–16·82 % of the total phenotypic variation in seed Cu concentration were identified on five chromosomes (A6, A9, A10, C6 and C7) with LOD scores ranging from 3·35 to 6·50. One QTL, CuNP-A10a, was detected in both years under normal-P conditions, while another QTL, CuLP-C6, was detected in both years under low-P conditions. Three QTLs, CuNP-A10a, CuLP-A10 and CuNP-A10b, co-located on A10 under different P conditions, with QTL CuNP-A10b explaining the highest percentage of phenotypic variation (16·82 %). Similarly, QTLs CuNP-C7b and CuLP-C7 co-located on C7. Higher seed Cu concentration was conferred by ‘Eyou Changjia’ alleles at five loci and by ‘B104-2’ at six loci.

Iron

In total, nine significant QTLs, which were located on seven chromosomes with LOD scores ranging from 3·38 to 4·26, were associated with seed Fe concentration. These QTLs accounted for a range of 8·97–11·95 % of the total phenotypic variation. Two QTLs, FeNP-C7 and FeLP-A3 were detected in both years under normal- and low-P conditions, respectively. Six of the nine QTLs for Fe overlapped with QTLs affecting seed concentrations of other mineral elements (P, Mg, Zn and Cu). Favourable alleles for increasing seed Fe concentration were contributed by ‘Eyou Changjia’ at three loci and by ‘B104-2’ at six loci.

Manganese

A total of ten QTLs, with LOD scores ranging from 3·34 to 6·55, distributed across eight chromosomes were associated with seed Mn concentration, explaining between 9·06 % and 16·43 % of the total phenotypic variation. Two QTLs, MnNP-A10b and MnLP-C3, were detected in both years under normal and low-P conditions, respectively. The largest proportion of phenotypic variation (16·43 %) could be attributed to the QTL, MnLP-A6, which was observed under the low-P condition in 2007. Two QTLs (MnNP-A10a and MnLP-A10) detected under different P conditions co-located on A10 with the same allelic effects of ‘B104-2’. Higher seed Mn concentration was conferred by ‘Eyou Changjia’ alleles at six loci and by ‘B104-2’ alleles at four loci.

Association of QTLs with functional genes by in silico mapping

Twelve syntenic blocks and 17 insertion islands have been identified between the A. thaliana genome and B. napus linkage groups by in silico mapping (Fig. 3). In total, 21 orthologous genes, which are involved in ion homeostasis in A. thaliana (Table S1 in Supplementary data, available online), were mapped to the QTL intervals identified in B. napus (Fig. 3). Three PAP family members, PAP16, PAP17 and PAP28, which have acid phosphatase activity (Li et al., 2002), overlapped well with QTLs on A5 and C2. Three YSL family members, YSL1, YSL5 and YSL7, and one ZIP family member, ZIP1, which are all thought to be involved in metal homeostasis (Colangelo and Guerinot, 2006; Waters and Grusak, 2008b; White and Broadley, 2009), were located in the QTL intervals on four chromosomes (A1, A3, A5 and C6). The co-localized QTL intervals for Ca, Mg and P on A5 contained two CAX family genes, CAX2 and CAX9, and four other genes, MT3, MT2A, MRS2-3 and IPS1. One of the CAX family genes, CAX1, was also located in the QTL intervals for Ca and Mn on A4. Three P transporter genes, GPT1, PHT4;5 and PHT3;1, were mapped to the QTL intervals on C2 and C3. Another four genes, UGP, IRT1, NRAMP5 and AZF2, were mapped to the QTL intervals for Mn and P on A1 and A3, respectively.

DISCUSSION

Concentrations of seed mineral nutrients are a key factor for vigorous germination and successful seedling establishment (Welch, 1999). Mineral contents of plant seeds depend on both environmental and genetic factors, and a number of researchers have identified QTLs affecting seed mineral concentrations in different plant species in various environments. In this research, QTLs affecting the concentrations of mineral elements in seeds of Brassica napus have been identified under two P conditions in an effort to reveal the genetic complexity of seed mineral accumulation.

Genetic variation is essential for achieving significant increases in seed mineral concentrations. Recent studies have shown significant genetic variation in seed mineral concentrations in Brassica species (Wu et al., 2007; Broadley et al., 2008; Liu et al., 2009). In the research presented here, a 1·5- to 3·4-fold variation was observed in the concentrations of all mineral elements studied (P, Ca, Mg, Fe, Mn, Cu and Zn) in seeds of a B. napus ‘B104-2’ × ‘Eyou Changjia’ RIL population grown at two P supplies, even though the two parental lines were not very different for most traits except for their seed P concentrations. Wider ranges of phenotypic variation in the seed concentrations of most mineral elements studied were observed in low-P environments than in high P environments, except seed P concentration in 2007 and seed Cu concentrations in both years. Estimates of heritability for these mineral traits varied greatly (Table 1). The wide phenotypic variation and the low level of heritability indicate the influence of environmental factors on quantitative traits and/or significant G × E interactions, since a number of epistatic interaction pairs for shoot mineral concentrations were identified by whole genome analysis of QTL and epistatic interactions in B. napus (Liu et al., 2009). The frequency distributions in the RIL population of nearly all mineral elements assayed in plants grown under both P conditions were relatively normal, and considerable transgressive segregation was observed (Figs 1 and and2).2). This suggests that these are quantitatively inherited traits controlled by multiple genes. Hence, this population can be used to identify QTLs affecting concentrations of mineral elements in seeds and the genes underpinning these traits.

Correlations among different seed mineral traits were observed across P levels. Seed concentrations of macro-nutrients (P, Ca and Mg) were significantly positively correlated, as were seed concentrations of some micro-nutrients, such as Zn and Cu, suggesting that similar transport and chelation processes affected the accumulation of both elements in seeds (Table 2). Most mineral traits were correlated under normal-P conditions, and better correlations were observed under normal-P conditions than under low-P conditions. Similar results were reported by Liu et al. (2009) for the accumulation of mineral elements in shoots of B. napus. This may point to common molecular mechanisms controlling the uptake and transport of several mineral elements in plants under normal-P supply. However, when plants lack P, different adaptive strategies are required for the plants to survive. High yield capacity is a major requirement for any crop cultivar. Despite positive correlations between concentrations of many mineral elements in seeds, very weak or even negative correlations between seed mineral concentrations and seed yield were observed, except between seed Mg concentration and seed yield in 2007 under normal-P conditions. This result suggests that seed yield may have exerted a ‘dilution effect’ on seed mineral concentrations. Moreover, it may indicate that QTLs for seed mineral concentrations are not directly helpful in breeding crops with enhanced seed yield under reduced input agriculture. However, it is possible that QTLs affecting leaf mineral concentrations have more effect on seed yield than QTLs for seed mineral concentrations.

Movement of mineral elements from soil through the plant to the seeds involves a series of transport processes that are tightly regulated by element-specific and element-non-specific transcriptional and post-transcriptional controls (Dunlop and Phung, 2002; Amtmann et al., 2006; Schachtman and Shin, 2007). However, not much is known about the genetic control and molecular-physiological mechanisms that contribute to the high accumulation of mineral elements in the seeds of different crop species, especially under abiotic stresses such as drought and mineral toxicity or deficiency. With the development of molecular marker technology, QTL analysis has been widely used as a powerful tool to dissect complex agronomic traits. A number of QTLs associated with seed mineral concentrations under abiotic stress have been identified in plants such as A. thaliana (Ghandilyan et al., 2009a) and wheat (Peleg et al., 2009). In the present study, a previously constructed linkage map was employed to detect QTLs affecting seed mineral concentrations in B. napus grown at two P supplies. A total of 78 putative QTLs were identified covering 17 linkage groups (Table 3 and Fig. 3). For each trait assayed, at least two QTLs were identified in different chromosomal regions contributing to small percentages of the phenotypic variation under different P conditions. This suggests that the accumulation of mineral elements in seeds of B. napus is a complex trait involving several genes. Although 16 QTLs were identified in both years, the majority were observed in only one year. This has also been observed for QTLs affecting mineral traits in other plant species such as A. thaliana (Waters and Grusak, 2008a) and Medicago truncatula (Sankaran et al., 2009). Many factors such as genetic variation in individuals, environmental variation, population size, number of markers and experimental error can influence the detection of QTLs in a segregating population (Collard et al., 2005). In this research, most of the single-occurrence QTLs have small effects as well as low, although significant, LOD scores. Thus, the detection of these QTLs may vary according to environmental conditions. Moreover, because of the limited molecular markers and their unbalanced distribution, the genetic map used here was only 1592·72 cM and individual linkage groups ranging from 13·42 to 148·16 cM. Thus, many QTLs could remain undetected due to the relatively small linkage map which did not cover the whole genome of B. napus.

One of the main objectives of this research was to detect QTLs affecting seed concentrations of mineral elements in B. napus and their response to P supply. A total of 33 significant QTLs were identified under low-P conditions (Table 3). Most of these QTLs differed from the QTLs identified for the same mineral elements under normal-P conditions. Similar observations were made by Liu et al. (2009) for QTLs affecting shoot mineral concentrations in B. napus grown with different boron availabilities. Considering the large number of solute-specific and non-specific transport proteins and chelates in plants (Maser et al., 2001), it is not surprising that different molecular mechanisms might be expressed under different P conditions, especially since P supply can alter the bio-availabilities of mineral elements in the soil. Analysis of QTLs affecting mineral accumulation in A. thaliana has also indicated that there is a strong dependence of QTLs on environment (Waters and Grusak, 2008a; Ghandilyan et al., 2009a, b). In addition, the fusion of B. oleracea (CC genome) and B. rapa (AA genome) progenitors has resulted in the genome of B. napus (AACC) being more than six times the size of the A. thaliana genome (Johnston et al., 2005) and it is suggested that B. napus might have two to eight orthologous genes corresponding to each single-copy gene in A. thaliana (Cavell et al., 1998). Some members of these gene families might have become specialized during the process of evolution to function optimally under different growth conditions, such as mineral deficiency. Thus, different family members distributing randomly in different chromosomal regions might account for differences in the QTLs identified under contrasting environmental conditions.

Co-localization of QTLs affecting different traits suggests either a single pleiotropic locus is involved in controlling of multiple traits or that several separate loci affecting independent traits are in close proximity. In this research, many QTLs (51 out of 78) co-localized with at least one other QTL. In particular, QTLs affecting seed P concentration co-located with QTLs affecting seed concentrations of other mineral elements. This is consistent with correlations between seed concentrations of P and those of other mineral elements in plants grown under different P conditions, suggesting a strong genetic association between mechanisms affecting seed P concentrations and concentrations of other mineral elements. Cations in seeds are often stored in a complex with phytic acid (1,2,3,4,5,6 hexakisphosphate) which is well known as the main storage compound for phosphorus in seeds (Raboy, 1997). The association of QTLs affecting seed P concentration with QTLs affecting the concentrations of other mineral elements in seeds could indicate a physiological process affecting several mineral elements similarly or physiological coupling in sink or source tissues. Co-localization of QTLs involved in the accumulation of different mineral elements, such as Ca/Mg and Fe/Mn/Cu/Zn, were also detected in this mapping population, suggesting a possible relationship at the molecular level among these traits. Similar phenomena have also been reported for QTLs affecting mineral traits in other plant species (Vreugdenhil et al., 2004; Broadley et al., 2008; Waters and Grusak, 2008a; Garcia-Oliveira et al., 2009; Ghandilyan et al., 2009a, b; Liu et al., 2009; Peleg et al., 2009; Sankaran et al., 2009). Co-localization of QTLs was not always consistent with the correlations observed between mineral traits. For instance, although seed concentrations of most mineral elements were correlated under normal-P supply, the QTLs identified for these traits were not always co-located. Similar results have been reported by Ghandilyan et al. (2009b). Correlations among seed concentrations of mineral elements reflect the association of total phenotypic variation which may be due to various genetic effects. However, because of the quantitative nature of seed mineral traits, weak statistical significance of QTLs is prevalent and co-location of QTLs may be obscured. QTLs for seed concentrations of several mineral elements, including P, Ca, Mg, Cu and Zn, overlapped on the top of C6, and two of these QTLs, MgNP-C6b and PNP-C6a, were detected under normal-P conditions in both years (Fig. 3), suggesting it was a hotspot locus for mineral traits under normal-P conditions. However, under low-P conditions most of the QTLs explaining a small percentage of the phenotypic variation were detected in only one year and were located on different chromosomal regions, suggesting that it maybe difficult to set up an efficient breeding programme for increasing seed mineral concentrations in rapeseed grown in a low-P environment.

QTLs should be robust enough for further fine mapping or a breeding scheme. Hence, it is important to compare the present results with those from other research groups. Due to its amphidiploid nature, B. napus (AACC genome) contains the diploid CC genome of B. oleracea and AA genome of B. rapa and the QTLs identified in this research can be related to those found by other research groups studying other Brassicaceae. Twenty-three QTLs controlling different aspects of mineral accumulation were found on the same chromosomes of different Brassica species (Table 3). However, because no common anchoring markers were used in these studies, it is difficult to determine whether these QTLs were located on the same chromosomal regions. Hence, further comparative QTL research should be conducted to confirm these QTLs using common molecular markers.

The existence of a common ancestor for Brassica and Arabidopsis makes comparative genome mapping possible. Twenty-four conserved genomic blocks have been identified in different Brassicaceae species by comparing the sequences of molecular markers with the annotated genome sequence of A. thaliana (Schranz et al., 2006). By in silico mapping, 21 genes involved in ion homeostasis in A. thaliana (Table S1 in Supplementary data, available online) were mapped to the QTL intervals of B. napus, including three PAP family members, three YSL family members, one ZIP family member, three P transporter genes, three CAX family genes and several other genes (Fig. 3). Our hypothesis was that QTLs were characterized by orthologous genes in both A. thaliana and B. napus. The approach of associating QTLs with functional genes in Arabidopsis by in silico mapping may provide more information for QTL analysis in B. napus. However, further fine mapping and analysis of near isogenic lines or mutants will be needed to confirm the involvement of a candidate gene on the identified quantitative trait. Moreover, there were no orthologous genes associated with the hotspot locus on C6, offering the possibility of identifying unknown genes that affect mineral uptake and translocation to seeds in B. napus.

CONCLUSIONS

Several QTLs for seed mineral concentrations were identified in Brassica napus under two P conditions. In total, 78 putative QTLs were identified affecting seed concentrations of diverse mineral elements (P, Ca, Mg, Fe, Mn, Cu and Zn). For each mineral trait, at least two QTLs were identified on different chromosomes, indicating that these mineral traits are quantitatively inherited and controlled by multiple genes. Often QTLs were clustered within a linkage group, suggesting that common mechanisms might control the accumulation of several mineral elements in seeds. Frequently, different chromosomal regions affecting the same mineral trait were identified under normal-P and low-P conditions, indicating different molecular mechanisms were employed for seed mineral accumulation in these two environments. Several genes involved in ion homeostasis in A. thaliana were mapped to the QTL intervals affecting seed mineral concentrations in B. napus, but further studies need to be conducted to determine whether these genes are true candidates for further investigation. These results may offer insights to the genetic basis of mineral accumulation in seeds of B. napus grown under different P supplies and help to unravel the genetic complexity of mineral accumulation in seeds of B. napus. However, the low robustness of the QTLs identified could make it difficult to focus on a particular QTL for cloning, and also difficult to set up an efficient breeding programme for increasing seed mineral concentrations in rapeseed in low-P environments.

Supplementary Material

[Supplementary Data]

ACKNOWLEDGEMENTS

This work was supported by National Basic Research and Development Program (2005CB120905), and National 863 High Technology Program (2006AA10A112), China. We thank Professor Philip J. White (SCRI, UK) for his help in revising the manuscript.

LITERATURE CITED

  • Amtmann A, Hammond JP, Armengaud P, White PJ. Nutrient sensing and signalling in plants: potassium and phosphorus. Advances in Botanical Research. 2006;43:209–257.
  • Bentsink L, Yuan K, Koornneef M, Vreugdenhil D. The genetics of phytate and phosphate accumulation in seeds and leaves of Arabidopsis thaliana, using natural variation. Theoretical and Applied Genetics. 2003;106:1234–1243. [PubMed]
  • Broadley MR, Hammond JP, King GJ, et al. Shoot calcium and magnesium concentrations differ between subtaxa, are highly heritable, and associate with potentially pleiotropic loci in Brassica oleracea. Plant Physiology. 2008;146:1707–1720. [PubMed]
  • Colangelo EP, Guerinot ML. Put the metal to the petal: metal uptake and transport throughout plants. Current Opinion in Plant Biology. 2006;9:322–330. [PubMed]
  • Cavell AC, Lydiate DJ, Parkin IAP, Dean C, Trick M. Collinearity between a 30-centimorgan segment of Arabidopsis thaliana chromosome 4 and duplicated regions within the Brassica napus genome. Genome. 1998;41:62–69. [PubMed]
  • Churchill GA, Doerge RW. Empirical threshold values for quantitative trait mapping. Genetics. 1994;138:963–971. [PubMed]
  • Collard BCY, Jahufer MZZ, Brouwer JB, Pang ECK. An introduction to markers, quantitative trait loci (QTL) mapping and marker-assisted selection for crop improvement: the basic concepts. Euphytica. 2005;142:169–196.
  • Duan HY, Shi L, Ye XS, Wang YH, Xu FS. Identification of phosphorous efficient germplasm in oilseed rape. Journal of Plant Nutrition. 2009;32:1148–1163.
  • Dunlop J, Phung T. Transporter genes to enhance nutrient uptake: opportunities and challenges. Plant and Soil. 2002;245:115–122.
  • Garcia-Oliveira AL, Tan L, Fu Y, Sun C. Genetic identification of quantitative trait loci for contents of mineral nutrients in rice grain. Journal of Integrative Plant Biology. 2009;51:84–92. [PubMed]
  • Ghandilyan A, Barboza L, Tisné S, et al. Genetic analysis identifies quantitative trait loci controlling rosette mineral concentrations in Arabidopsis thaliana under drought. New Phytologist. 2009a;184:180–192. [PubMed]
  • Ghandilyan A, Ilk N, Hanhart C, et al. A strong effect of growth medium and organ type on the identification of QTLs for phytate and mineral concentrations in three Arabidopsis thaliana RIL populations. Journal of Experimental Botany. 2009b;60:1409–1425. [PubMed]
  • Grusak MA, Cakmak I. Methods to improve the crop-delivery of minerals to humans and livestock. In: Broadley MR, White PJ, editors. Plant nutritional genomics. Oxford: Blackwell; 2005. pp. 265–286.
  • Grusak MA, DellaPenna D. Improving the nutrient composition of plants to enhance human nutrition and health. Annual Review of Plant Physiology and Plant Molecular Biology. 1999;50:133–161. [PubMed]
  • Hammond JP, Broadley MR, White PJ, et al. Shoot yield drives phosphorus use efficiency in Brassica oleracea and correlates with root architecture traits. Journal of Experimental Botany. 2009;60:1953–1968. [PubMed]
  • Harada H, Leigh RA. Genetic mapping of natural variation in potassium concentrations in shoots of Arabidopsis thaliana. Journal of Experimental Botany. 2006;57:953–960. [PubMed]
  • Johnston JS, Pepper AE, Hall AE, et al. Evolution of genome size in Brassicaceae. Annals of Botany. 2005;95:229–235. [PMC free article] [PubMed]
  • Li D, Zhu H, Liu K, et al. Purple acid phosphatases of Arabidopsis thaliana. The Journal of Biological Chemistry. 2002;277:27772–27781. [PubMed]
  • Li G, Quiros CF. Sequence-related amplified polymorphism (SRAP), a new marker system based on a simple PCR reaction: its application to mapping and gene tagging in Brassica. Theoretical and Applied Genetics. 2001;103:455–461.
  • Li YY, Ma CZ, Fu TD, et al. Construction of a molecular functional map of rapeseed (Brassica napus L.) using differentially expressed genes between hybrid and its parents. Euphytica. 2006;152:25–39.
  • Liu J, Yang J, Li R, et al. Analysis of genetic factors that control shoot mineral concentrations in rapeseed (Brassica napus) in different boron environments. Plant and Soil. 2009;320:255–266.
  • Long Y, Shi J, Qiu D, et al. Flowering time quantitative trait loci analysis of oilseed Brassica in multiple environments and genomewide alignment with Arabidopsis. Genetics. 2007;177:2433–2444. [PubMed]
  • Loudet O, Saliba-Colombani V, Camilleri C, et al. Natural variation for sulfate content in Arabidopsis thaliana is highly controlled by APR2. Nature Genetics. 2007;39:896–900. [PubMed]
  • Marschner H. Mineral nutrition of higher plants. 2nd edn. London: Academic Press; 1995.
  • Maser P, Thomine S, Schroeder JI, et al. Phylogenetic relationships within cation transporter families of Arabidopsis. Plant Physiology. 2001;126:1646–1667. [PubMed]
  • Paran I, Zamir D. Quantitative traits in plants: beyond the QTL. Trends in Genetics. 2003;19:303–306. [PubMed]
  • Payne KA, Bowen HC, Hammond JP, et al. Natural genetic variation in caesium (Cs) accumulation by Arabidopsis thaliana. New Phytologist. 2004;162:535–548.
  • Peleg Z, Cakmak I, Ozturk L, et al. Quantitative trait loci conferring grain mineral nutrient concentrations in durum wheat×wild emmer wheat RIL population. Theoretical and Applied Genetics. 2009;119:353–369. [PubMed]
  • Piquemal J, Cinquin E, Couton F, et al. Construction of an oilseed rape (Brassica napus L.) genetic map with SSR markers. Theoretical and Applied Genetics. 2005;111:1514–1523. [PubMed]
  • Raboy V. Accumulation and storage of phosphate and minerals. In: Larkins BA, Vasil IK, editors. Cellular and molecular biology of plant seed development. Dordrecht: Kluwer; 1997. pp. 441–477.
  • Raghothama K, Karthikeyan A. Phosphate acquisition. Plant and Soil. 2005;274:37–49.
  • Sankaran R, Huguet T, Grusak M. Identification of QTL affecting seed mineral concentrations and content in the model legume Medicago truncatula. Theoretical and Applied Genetics. 2009;119:241–253. [PubMed]
  • Schachtman DP, Shin R. Nutrient sensing and signaling: NPKS. Annual Review of Plant Biology. 2007;58:47–69. [PubMed]
  • Schachtman DP, Reid RJ, Ayling SM. Phosphorus uptake by plants: from soil to cell. Plant Physiology. 1998;116:447–453. [PubMed]
  • Schranz ME, Lysak MA, Mitchell-Olds T. The ABC's of comparative genomics in the Brassicaceae: building blocks of crucifer genomes. Trends in Plant Science. 2006;11:535–542. [PubMed]
  • Shi JQ, Li RY, Qiu D, et al. Unraveling the complex trait of crop yield with QTL mapping in Brassica napus. Genetics. 2009;182:851–861. [PubMed]
  • Suwabe K, Iketani H, Nunome T, Kage T, Hirai M. Isolation and characterization of microsatellites in Brassica rapa L. Theoretical and Applied Genetics. 2002;104:1092–1098. [PubMed]
  • Vos P, Hogers R, Bleeker M, et al. AFLP: a new technique for DNA-fingerprinting. Nucleic Acids Research. 1995;23:4407–4414. [PMC free article] [PubMed]
  • Vreugdenhil D, Aarts MGM, Koornneef M, Nelissen H, Ernst WHO. Natural variation and QTL analysis for cationic mineral content in seeds of Arabidopsis thaliana. Plant, Cell & Environment. 2004;27:828–839.
  • Wang SC, Bastern J, Zeng ZB. Raleigh, NC, USA: Department of Statistics, North Carolina State University; 2006. Windows QTL Cartographer 2·5.
  • Waters BM, Grusak MA. Quantitative trait locus mapping for seed mineral concentrations in two Arabidopsis thaliana recombinant inbred populations. New Phytologist. 2008a;179:1033–1047. [PubMed]
  • Waters BM, Grusak MA. Whole-plant mineral partitioning throughout the life cycle in Arabidopsis thaliana ecotypes Columbia, Landsberg erecta, Cape Verde Islands, and the mutant line ysl1ysl3. New Phytologist. 2008b;177:389–405. [PubMed]
  • Welch RM. Importance of seed mineral nutrient reserves in crop growth and development. In: Rengel Z, editor. Mineral nutrition of crops: fundamental mechanisms and implications. New York, NY: Food Products Press; 1999. pp. 205–226.
  • Welch RM, Graham RD. Breeding for micronutrients in staple food crops from a human nutrition perspective. Journal of Experimental Botany. 2004;55:353–364. [PubMed]
  • White PJ. Calcium. In: Broadley MR, White PJ, editors. Plant nutritional genomics. Oxford: Blackwell; 2005. pp. 66–86.
  • White PJ, Broadley MR. Biofortifying crops with essential mineral elements. Trends in Plant Science. 2005;10:586–593. [PubMed]
  • White PJ, Broadley MR. Biofortification of crops with seven mineral elements often lacking in human diets – iron, zinc, copper, calcium, magnesium, selenium and iodine. New Phytologist. 2009;182:49–84. [PubMed]
  • White PJ, Hammond JP, King GJ, et al. Genetic analysis of potassium use efficiency in Brassica oleracea. Annals of Botany. 2010;105:1199–1210. [PMC free article] [PubMed]
  • Wu J, Schat H, Sun R, Koornneef M, Wang X, Aarts M. Characterization of natural variation for zinc, iron and manganese accumulation and zinc exposure response in Brassica rapa L. Plant and Soil. 2007;291:167–180.
  • Wu J, Yuan YX, Zhang XW, et al. Mapping QTLs for mineral accumulation and shoot dry biomass under different Zn nutritional conditions in Chinese cabbage (Brassica rapa L. ssp. pekinensis) Plant and Soil. 2008;310:25–40.
  • Yang M, Ding G, Shi L, Feng J, Xu F, Meng J. Quantitative trait loci for root morphology in response to low phosphorus stress in Brassica napus. Theoretical and Applied Genetics. 2010 in press. doi: 10.1007/s00122-010-1301-1. [PubMed]
  • Zhao J, Jamar DL, Lou P, et al. Quantitative trait loci analysis of phytate and phosphate concentrations in seeds and leaves of Brassica rapa. Plant, Cell & Environment. 2008;31:887–900. [PubMed]

Articles from Annals of Botany are provided here courtesy of Oxford University Press