Article pubs.acs.org/JAFC
Mapping of Quantitative Trait Loci for Contents of Macro- and Microelements in Milled Rice (Oryza sativa L.) Yong-Hong Yu, Ya-Fang Shao, Jie Liu, Ye-Yang Fan, Cheng-Xiao Sun, Zhao-Yun Cao, and Jie-Yun Zhuang* China National Rice Research Institute, Hangzhou, Zhejiang 310006, People’s Republic of China ABSTRACT: Macro- and microelement contents are important traits for nutritional quality in rice. In this study, quantitative trait loci (QTLs) for the contents of seven mineral elements in milled rice were detected using recombinant inbred lines (RILs) of the indica rice cross Zhenshan 97/Milyang 46, followed by the validation and fine mapping of a QTL region on the short arm of chromosome 6. A total of 20 QTLs distributed on chromosomes 1, 3, 5, 6, 10, and 11 were detected in the RIL population. Co-localizations of QTLs for multiple traits were observed, of which the qP3/qMg3/qZn3 region was shown to have the largest effects for the contents of phosphorus, magnesium, and zinc, and the qK6.1/qCa6/qZn6/qMn6/qCu6 region was found to be responsible for five of the seven traits. Using near isogenic lines having sequential segregating region, the target QTL on chromosome 6 was delimitated to a 29.9 kb region flanked by RM19410 and Si2944. This QTL showed major effects for all seven traits, with the enhancing alleles derived from the male parent Milyang 46. KEYWORDS: rice, quantitative trait locus, recombinant inbred line, residual heterozygous line, macroelement, microelement
■
contents in milled and brown rice,7,8 and F2 plants of the indica rice cross PAU201/Palman 579 and double haploid lines of the inter-subspecific crosses IR64/Azucena and Chunjiang 06/TN1 were used to detect QTLs for mineral contents in brown rice.9−11 Validation and fine mapping of these QTLs would greatly facilitate their utilization in rice breeding. It has been known that the contents of most mineral elements, such as P, K, Mg, Ca, Fe, and Mn, are not evenly distributed in rice grain.12,13 Because rice is mostly consumed after milling, it is more important to determine QTLs for mineral contents in milled rice than in brown rice. Therefore, this study was conducted to determine QTLs for the contents of seven mineral elements in milled rice using RILs of an indica rice cross and to fine map a QTL that was responsible for multiple elements and detected in various studies. Following the detection of 20 QTLs in the RIL population, a QTL region on the short arm of chromosome 6 was selected for validation and fine mapping using near isogenic lines (NILs) derived from the same cross. This QTL was finally delimitated to a 29.9 kb region, showing major effects for all seven traits analyzed.
INTRODUCTION Rice (Oryza sativa L.), one of the most important staple crops, feeds almost half of the world’s population. It is also an important source of macro- and microelements for people getting energy solely from rice. Currently, mineral malnutrition is considered to be a serious global challenge for humans.1,2 Microelements, such as manganese (Mn), zinc (Zn), iron (Fe), and copper (Cu), can be found in many enzymes that are of great importance for maintaining normal metabolic functions. Macroelements, such as phosphorus (P), potassium (K), calcium (Ca), and magnesium (Mg), which can be found in every cell, are very important for proper fluid balance, blood pressure regulation, nerve transmission, and immune system health.3 Understanding of the genetic basis underlying natural variations of the macro- and microelement contents in rice grains will greatly facilitate the improvement of mineral nutrition in rice breeding. In the past decade, a large number of quantitative trait loci (QTLs) for mineral contents in rice grain has been detected in various populations derived from interspecific, inter-subspecific, or intra-subspecific crosses.4−11 Using recombinant inbred lines (RILs) and introgression lines (ILs) derived from the intersubspecific cross Lemont/Teqing, 134 QTLs were detected for the contents of 16 elements in brown rice, which were distributed on 39 genomic regions of the 12 rice choromosomes.4 Using RILs of the indica rice cross Zhenshan 97/ Minghui 63, Lu et al.5 found that the contents of Cu, Ca, Zn, Mn, and Fe in milled rice were controlled by 10 main-effect QTLs and 28 digenic interactions involving 46 epistatic QTLs. When ILs constructed from a cross between a rice cultivar and a wild rice (Oryza rufipogon) were used, 31 QTLs for the contents of Fe, Zn, Mn, Cu, Ca, Mg, P, and K in brown rice were identified, with the enhancing alleles derived from the wild rice for 26 QTLs.6 In other studies, RILs of the inter-subspecific cross Bala/Azucena were used to identify QTLs for mineral © 2015 American Chemical Society
■
MATERIALS AND METHODS
Plant Materials. A RIL population and three sets of NILs developed from an indica rice cross between Zhenshan 97 (ZS97) and Milyang 46 (MY46) were used, in which ZS97 and MY46 are maintainer and restorer lines of the commercial three-line hybrid rice Shanyou 10, respectively. The three NIL sets were derived from a residual heterozygote (RH) carrying a 7.3 Mb heterozygous segment on the short arm of chromosome 6, which was previously selected from a F7 population of ZS97/MY46 using 208 polymorphic simple sequence repeat (SSR) markers.14 Selfing progenies of the RH plant Received: Revised: Accepted: Published: 7813
June 12, 2015 July 30, 2015 August 23, 2015 August 24, 2015 DOI: 10.1021/acs.jafc.5b02882 J. Agric. Food Chem. 2015, 63, 7813−7818
Article
Journal of Agricultural and Food Chemistry
Figure 1. Genotypes of three sets of NILs, TF6-2, TF6-15, and TF6-17, in the interval RM19350−RM225 on the short arm of rice chromosome 6. The position of qZn6 for the contents of seven elements in milled rice is indicated.
Table 1. Phenotypic Performance of the Contents of Seven Mineral Elements in Milled Rice of the ZS97/MY46 RIL Population parental mean traita
mean
SD
CV
range
kurtosis
skewness
ZS97
MY46
P K Mg Ca Zn Mn Cu
1715.0 1037.1 349.3 74.4 22.6 15.8 3.7
319.5 208.9 96.8 15.8 4.6 3.7 0.9
0.186 0.201 0.277 0.212 0.204 0.234 0.243
828.8−3250.0 270.3−1622.0 81.1−633.5 40.7−123.3 13.6−44.4 7.7−25.9 1.9−7.2
3.32 0.55 0.16 0.46 3.18 0.02 1.04
1.00 −0.26 0.15 0.77 1.22 0.62 0.80
2632.0 1262.0 423.1 152.8 18.1 9.6 3.9
2390.0 1130.0 305.1 124.3 20.7 12.0 5.5
a
The contents of the mineral elelments are presented as milligrams per kilogram in the milled rice. P, phosphorus; K, potassium; Mg, magnesium; Ca, calcium; Zn, zinc; Mn, manganese; and Cu, copper.
were selected for three generations based on genotyping of polymorphic markers located in the target QTL region. Three new RHs with sequential heterozygous segments, which jointly covered the entire QTL region, were identified. The three plants were each selfed to produce a NIL-F2 population, from which 20 non-recombinant homozygous plants, including 10 maternal and 10 paternal homozygotes, were selected. Selfing seeds of these plants resulted in the development of three NIL sets, which were segregated in the intervals RM4923−RM19410, RM19410−Si2950, and Si2944− RM204 and named TF6-2, TF6-15, and TF6-17, respectively (Figure 1). Field Experiments. The rice lines were planted in the paddy field at the China National Rice Research Institute, Hangzhou, Zhejiang, China. For each line, 12 plants were grown in one row with spacing of 16.7 cm between plants and 26.7 cm between rows. While 243 RILs were tested without replication, the NILs were planted in two replications with a randomized complete block design. Field management followed the normal agricultural practice. At maturity, seeds from five middle plants of each row were harvested. Phenotype Determination. The grains were dried to a moisture content of about 13% and stored at room temperature for 3 months. They were then dehulled on a Satake Rice Machine (Satake Co., Tokyo, Japan). After polishing on a desktop rice miller (NSART, Sangyong Machinery Industry Co., Jungu, South Korea), the milled rice was ground into flour using a Cyclotec 1093 Sample Mill (Foss Tecator, Sweden) and passed through a 100-mesh sieve. The rice flour was stored at 4 °C. Extraction of macro- and microelements was conducted by a digestion method. In briefly, about 0.50 g of rice flour and 5 mL of ultrapure nitric acid were added to a Teflon digestion vessel and then put on the microwave digestion instrument (CEM Mars, Matthews, NC). The digestion procedure was set as the followings: raising from room temperature to 120 °C, 0−5 min; holding at 120 °C, 6−7 min; raising from 120 to 180 °C, 8−17 min; and holding 180 °C, 18−32 min. After cooling to room temperature, the solution was transferred to a 50 mL volumetric flask and diluted to 50 mL with deionized water. The contents of P, K, Ca, Mg, Mn, Zn, and Cu were determined
by inductively coupled plasma−atomic emission spectrometry (ICP− AES, Thermo Elemental, Franklin, MA) at the wavelengths of 213.6, 766.5, 183.8, 280.3, 257.6, 213.9, and 324.8 nm, respectively. The contents of elements were calibrated using the certified standards in rice flour samples (GBW10010, CRM Rice) obtained from the National Institute Center of Standards. Results were presented as milligrams per kilogram of rice flour. All of the determinations were conducted in duplication. Marker Genotyping. A total of 16 DNA markers as shown in Figure 1 were used in the NIL development. Total DNA was extracted following the method of Zheng et al.15 Polymerase chain reaction (PCR) amplification was performed according to Chen et al.,16 and the products were visualized on 6% non-denaturing polyacrylamide gels using silver staining. All of the SSR markers were selected from the Gramene database (www.gramene.org). InDel markers Si2944 and Si2950 were designed according to the difference between ZS97 and MY46 detected by whole-genome re-sequencing. Primers used were 5′-AAACCACAAGATTAGGCTCTAAGT-3′ for Si2944-F and 5′AAGTGAGGGAAACCACATTCTAC-3′ for Si2944-R and 5′-AGGCACTCACAACCATAAC-3′ for Si2950-F and 5′-TAGTACATGACAATGGAGCTATC-3′ for Si2950-R. QTL Mapping. A linkage map has been constructed for the ZS97/ MY46 RIL population, consisting of 256 DNA markers and spanning 1814.7 cM.17,18 QTL analysis was performed with the Composite Interval Mapping (CIM) and Multiple Interval Mapping (MIM) of the Windows QTL Cartographer 2.5.19 Candidate loci were selected with CIM using a threshold of log of odds (LOD) > 2.5 and tested with MIM using the Bayesian Information Criterion (BIC) default value c(n) = ln(n). Those showing significant effects were claimed as putative QTLs and designated as proposed by McCouch and the Committee on Gene Symbolization, Nomenclature and Linkage (CGSNL).20 For each of the three NIL sets, two-way analysis of variance (ANOVA) was conducted to test phenotypic differences between the two genotypic groups using SAS procedure general linear model (GLM) as described by Dai et al.21 Given the detection of significant difference (p < 0.05), the same model was applied to estimate the 7814
DOI: 10.1021/acs.jafc.5b02882 J. Agric. Food Chem. 2015, 63, 7813−7818
Article
Journal of Agricultural and Food Chemistry
that Mn was not significantly correlated with Zn and Cu.10 It seems that positively significant correlations between the contents of different elements in rice grain are more achievable in populations that are less diverse. QTLs Detected in the RIL Population. A total of 20 QTLs distributed on six chromosomes were detected for the seven traits (Table 3 and Figure 2). The proportion of phenotypic variance explained by a single QTL (R2) ranged from 4.2 to 11.8%. For the contents of macroelements P, K, Mg, and Ca, the numbers of QTLs detected were 2, 3, 1, and 3, respectively. All of these QTLs, except qCa5, had the enhancing allele derived from the male parent MY46. Of the two QTLs for phosphorus content, qP3 explained 11.8% of the phenotypic variance with an additive effect of 114.78 mg/kg, which were much larger than the values of 4.3% and 67.11 mg/kg estimated for qP1, respectively. It is also noted that the only QTL detected for magnesium content, qMg3, had the same location and allelic direction as qP3. The three QTLs for potassium content showed little difference in the magnitude, along with the three QTLs for calcium content. While the R2 of qK1, qK6.1, and qK6.2 ranged from 4.5 to 6.7% with the additive effects ranging from 45.19 to 55.06 mg/kg, the R2 of qCa5, qCa6 and qCa10 ranged from 4.3 to 6.6% with the additive effects ranging from 3.33 to 4.27 mg/kg. For the microelements Mn, Zn, and Cu, the numbers of QTLs detected were 4, 5, and 2, respectively. The QTL having the largest effect for manganese content, qMn6, was located at a region close to marker RM190 for the Wx gene on the short arm of chromosome 6, explaining 10.3% of the phenotypic variance, with the MY46 allele increasing the Mn content by 1.20 mg/kg. Two other QTLs, qMn1.1 and qMn1.2, were linked in repulsion on the short arm of chromosome 1, having slightly lower effects compared to qMn6. The remaining QTL for this trait, qMn11, had a R2 of 4.2%, with the MY46 allele decreasing the Mn content by 0.77 mg/kg. All five QTLs for zinc content, except qZn11, had the enhancing allele derived from MY46. The QTL having the largest effect, qZn3, explained 11.7% of the phenotypic variance, with an additive effect of 1.59 mg/kg. The four other QTLs had R2 ranging from 4.4 to 6.5% and additive effects ranging from 0.98 to 1.18 mg/kg. The two QTLs for copper content both had the enhancing allele derived from MY46. While qCu6 located near RM190 had a R2 of 11.2% and an additive effect of 0.31 mg/kg, the values estimated for qCu10 were 7.4% and 0.25 mg/kg, respectively. Two of the QTL regions described above were shown to be especially important. One was the RZ142−RZ613 interval on the long arm of chromosome 3, in which three QTLs were located, including qMg3, the only QTL detected for magnesium, and qP3 and qZn3, the two QTLs having the largest effects for the contents of phosphorus and zinc in the ZS97/MY46 RIL population, respectively. The other was a region near the marker RM190 on the short arm of chromosome 6, in which five QTLs were located, including qMn6 and qCu6 having the largest effects for the contents of manganese and copper and qK6.1, qCa6, and qZn6 that were associated with the contents of potassium, calcium, and zinc, respectively. On the basis of the physical position of DNA markers linked to QTLs for mineral contents in rice grain, 12 of the QTLs detected in this study were coincident with those reported previously (Table 3). It is noteworthy that three members of
additive effect of the QTL and the proportion of phenotypic variance explained by the QTL.
■
RESULTS AND DISCUSSION Phenotypic Performance of the RIL Population. Descriptive statistics of the mineral contents in milled rice of the ZS97/MY46 RIL population are presented in Table 1. Large variations with strong transgressive segregation were observed for all of the traits. While the contents in the female parent ZS97 were higher for macroelements (P, K, Mg, and Ca) and lower for microelements (Zn, Mn, and Cu) than in the male parent MY46, the parental differences were small compared to the variations among the RILs. Although the contents of macroelements were higher than that of microelements, the coefficients of variation were similar among the seven elements. In comparison to the mineral contents in brown rice determined for 304 indica landraces and 28 indica improved cultivars,22 the contents in milled rice measured in our study were much smaller for P, K, Mg, Ca, and Cu but comparable for Zn. Consistent content of Zn in brown and milled rice was also found when more reports were compared,6,9,11,23,24 along with the higher contents of P, K, Mg, Ca, and Cu in brown rice.6,11,23 These results are in accordance with the understanding that most of the mineral elements are concentrated in the outer part of the rice grain but Zn has an relatively even distribution.12,13,25,26 For the remaining element tested in our study, Mn, no consensus could be achieved because the content in brown rice had a large variation ranging from 7.3 to 35.0 mg/kg.6,8,11,22,23 Correlation coefficients between the contents of P, K, Ca, Mg, Mn, Zn, and Cu of the 243 RILs were all highly significant (p < 0.01), indicating the feasibility of improving these elements simultaneously in rice grain. The coefficients between the three elements having the largest contents, P, K, and Mg, were larger than those between the microelements. While the largest coefficient of 0.851 was observed between K and Mg, the lowest value of 0.213 was found between K and Cu (Table 2). In a previous QTL mapping for mineral contents using an Table 2. Coeffcients of Correlation between the Seven Traits in the RIL Population traita
K
Mg
Ca
Zn
Mn
Cu
P K Mg Ca Zn Mn
0.621b
0.750b 0.851b
0.515b 0.320b 0.434b
0.608b 0.268b 0.353b 0.596b
0.495b 0.415b 0.558b 0.689b 0.448b
0.354b 0.213b 0.258b 0.432b 0.580b 0.415b
a
The contents of the mineral elelments are presented as milligrams per kilogram in the milled rice. P, phosphorus; K, potassium; Mg, magnesium; Ca, calcium; Zn, zinc; Mn, manganese; and Cu, copper. b p < 0.01.
interspecific cross,6 all seven elements analyzed in our study were included and tested for 2 years. Of the 42 correlations between the seven elements, 24 were positively significant and the others were not significant. While correlations between the macroelements were all significant, Zn was not significantly correlated with other elements in 1 or 2 years and Cu was only significantly correlated with one element in 1 year. In another study using an inter-subspecific cross, positively significant correlations were found between P, Mn, Zn, and Cu, except 7815
DOI: 10.1021/acs.jafc.5b02882 J. Agric. Food Chem. 2015, 63, 7813−7818
Article
Journal of Agricultural and Food Chemistry Table 3. QTLs Detected for the Contents of Seven Elements in Milled Rice element P K
Mg Ca
Mn
Zn
Cu
QTL
interval
LOD
Aa
R2 (%)b
previous reportc
qP1 qP3 qK1 qK6.1 qK6.2 qMg3 qCa5 qCa6 qCa10 qMn1.1 qMn1.2 qMn6 qMn11 qZn1 qZn3 qZn5 qZn6 qZn11 qCu6 qCu10
RZ538−RG381 RZ142−RZ613 RG472−RZ543 RZ516−R1962 RZ140−RZ405 RZ142−RZ613 C246−RM274 RM204−RM225 RM171−RM1108 RZ543−RG313 RM1195−RM5359 RZ516−R1962 RG167−RM287 RM315−RZ538 RZ142−RZ613 RG119−RG346 RM204−RM225 RM7557−RZ816 RM190−RZ516 RM1859−RM6704
2.60 6.88 3.39 3.55 2.82 4.61 3.40 2.74 4.11 4.00 4.85 5.66 2.51 2.64 5.98 3.04 4.18 2.62 5.63 2.78
67.11 114.78 52.14 55.06 45.19 27.60 −4.27 3.33 4.13 1.11 −1.16 1.20 −0.77 1.02 1.59 1.05 1.18 −0.98 0.31 0.25
4.3 11.8 6.1 6.7 4.5 7.8 6.5 4.3 6.6 8.2 9.0 10.3 4.2 4.8 11.7 5.0 6.5 4.4 11.2 7.4
4, 6, and 7 4 11
4 and 6 5 and 10 7 8 11 4 4, 7, and 8 5 6
a
Additive effect of replacing a maternal allele by a paternal allele. bProportion of phenotypic variance explained by the given QTL. cPrevious reports in which a QTL for the content of the same elelment in brown or milled rice was detected in a simliar location.
Figure 2. Genomic distribution of QTLs for mineral element contents in milled rice detected in the RIL population of ZS97/MY46.
the QTL cluster qK6.1/qCa6/qMn6/qZn6/qCu6 have been previously detected, of which qZn6 was detected for the Zn content in milled or brown rice in three studies4,7,8 and qK6.1 and qCu6 were detected for the contents of K and Cu in brown rice as reported by Du et al.11 and Garcia-Oliveira et al.,6 respectively. Because this region was found to be associated
with multiple mineral elements in a few independent studies, it was selected for validation and fine mapping. QTL Fine Mapping Using NILs. Three sets of NILs with sequential segregating regions jointly covering the entire region flanked by RM4923 and RM204 (Figure 1) were used for QTL fine mapping. For NIL sets TF6-2, TF6-15, and TF6-17, which consisted of two homozygous genotypic groups differing in the 7816
DOI: 10.1021/acs.jafc.5b02882 J. Agric. Food Chem. 2015, 63, 7813−7818
Article
Journal of Agricultural and Food Chemistry Table 4. QTL Effects Detected in the Three Sets of NILs phenotype (mean ± SD) NIL name
segregating region
element
TF6-2 TF6-15
RM4923−RM19410 RM19410−Si2950
TF6-17
Si2944−RM204
Zn P K Mg Ca Mn Zn Cu P K Mg Mn
ZS97a 18.0 1370.5 942.1 305.7 59.3 20.4 20.1 4.3 1567.1 977.8 377.0 23.8
± ± ± ± ± ± ± ± ± ± ± ±
MY46b
1.3 108.3 90.0 51.3 4.1 1.4 0.7 0.4 105.9 82.1 50.2 1.4
18.7 1883.9 1286.0 517.9 84.9 29.4 25.6 5.9 1486.7 926.6 344.9 22.7
± ± ± ± ± ± ± ± ± ± ± ±
1.7 160.5 134.3 64.9 10.9 2.3 1.3 0.4 138.5 106. 8 51.7 1.6
P < < < < < <
