Skip to main content
NCBI home page
Molecules and Cells logoLink to Molecules and Cells
. 2011 Apr 30;31(4):385–392. doi: 10.1007/s10059-011-0049-z

Identification of QTLs for Seed Germination Capability after Various Storage Periods Using Two RIL Populations in Rice

Wenzhu Jiang 1, Joohyun Lee 2, Yong-Mei Jin 3, Yongli Qiao 2, Rihua Piao 2, Sun Mi Jang 2, Mi-Ok Woo 2, Soon-Wook Kwon 4, Xianhu Liu 5, Hong-Yu Pan 1, Xinglin Du 1,*, Hee-Jong Koh 2,*
PMCID: PMC3933968  PMID: 21399992

Abstract

Seed germination capability of rice is one of the important traits in the production and storage of seeds. Quantitative trait loci (QTL) associated with seed germination capability in various storage periods was identified using two sets of recombinant inbred lines (RILs) which derived from crosses between Milyang 23 and Tong 88-7 (MT-RILs) and between Dasanbyeo and TR22183 (DT-RILs). A total of five and three main additive effects (QTLs) associated with seed germination capability were identified in MT-RILs and DT-RILs, respectively. Among them, six QTLs were identified repeatedly in various seed storage periods designated as qMT-SGC5.1, qMT-SGC7.2, and qMT-SGC9.1 on chromosomes 5, 7, and 9 in MT-RILs, and qDT-SGC2.1, qDTSGC3.1, and qDT-SGC9.1 on chromosomes 2, 3, and 9 in DT-RILs, respectively. The QTL on chromosome 9 was identified in both RIL populations under all three storage periods, explaining up to 40% of the phenotypic variation. Eight and eighteen pairs additive × additive epistatic effect (epistatic QTL) were identified in MT-RILs and DT-RILs, respectively. In addition, several near isogenic lines (NILs) were developed to confirm six repeatable QTL effects using controlled deterioration test (CDT). The identified QTLs will be further studied to elucidate the mechanisms controlling seed germination capability, which have important implications for long-term seed storage.

Keywords: near isogenic line, quantitative trait loci, rice, RIL, seed germination capability

INTRODUCTION

Seed germination capability during storage can be defined as the maximum time period that pure seeds retain germination viability when stored under ideal ambient conditions. It varies among the species due to the occurrence of natural variability and is usually regarded to be related with seed longevity or seed storability traits (Bentsink et al., 2000; Ellis and Roberts, 1981; Singh and Ram, 1986). Seed germination capability is an important index for determining the regeneration cycle of germplasm conserved in genebanks (Miura et al., 2002). Without proper storage facilities, rapid seed deterioration in humid tropical climates can be a serious problem for rice production (Siddique et al., 1988).

Although storage conditions are important for seed germination capability, genetic factors also largely affect seed germination capability (Bewley and Black, 1994; Clerkx et al., 2004; Miura et al., 2002). To date, some quantitative genetic studies have been conducted to detect QTLs for analysis of seed germination capability (Miura et al., 2002; Sasaki et al., 2005; Xue et al., 2008; Zeng et al., 2006). Among these QTL analyses, the QTL on chromosome 9 was consistently identified in three independent studies (Miura et al., 2002; Sasaki et al., 2005; Xue et al., 2008), and subsequently, the effect of this QTL was confirmed with chromosome segment substitution lines (CSSLs) by Shigemune et al. (2008).

The genetics and molecular basis of seed germination capability have not yet been clearly elucidated, mainly due to the difficulty in accurate phenotyping. Since it takes a long period of time for seeds to age naturally, an alternative method for evaluating seed germination capability was suggested based on the fact that moisture content and storage temperatures are the most important factors affecting the germination capability of stored seeds (Ellis et al., 1982; Roberts, 1972). The combination of temperature, humidity, time, and controlled deterioration test (CDT) are considered a reliable method to measure seed germination capability and seed longevity (Padma and Reddy, 2000; Powell and Matthews, 1984; Xue et al., 2008; Zeng et al., 2002; 2006). This treatment is presumed to mimic natural aging (Delouche and Baskin, 1973), while allowing considerably accelerates the seed deteriorations, which is convenient for quickly assessing the seed germination capability. Therefore, seed companies largely rely on this treatment as a prognosis for seed germination capability in stored seeds (Rajjou and Debeaujon, 2008). The procedure in rice is simple: seeds are placed in a chamber controlled at a temperature of 40℃ and 95% relative humidity (RH%) for 10-14 days, and then germination rate is scored.

In the present study, we conducted QTL analysis for seed germination capability in various seed storage periods under natural conditions using two RIL populations, both derived from the combination of indica and japonica varieties (Milyang 23 × Tong 88-7 and Dasanbyeo × TR22183). We also attempted to confirm the effects of the QTLs using two sets of NILs that separately developed from each RIL population through backcrossing and marker-assisted selection (MAS) using a CDT.

MATERIALS AND METHODS

Plant materials

Two sets of RILs were used. One set of RILs (F6, MT-RILs) consisted of 166 lines derived from the cross between Milyang 23 and Tong 88-7 by single seed descent (SSD) (Jiang et al., 2010). The other set of RILs (F8, DT-RILs) consisted of 166 lines derived from the cross between Dasanbyeo and TR22183 by SSD (Cho et al., 2007). Milyang 23 and Dasanbyeo are Korean tongil-type rice varieties derived from an indica × japonica cross and similar to indica in its genetic make-up (Chung and Heu, 1991; Kown et al., 2001), whereas Tong 88-7 and TR22183 are temperate japonica varieties, both originating from northeast China.

To confirm QTL effects of seed germination capability, three RILs harboring one or more target QTL regions were separately selected in each RIL population for developing NILs. These RILs were backcrossed with their tongil-type parents (Milyang 23 or Dasanbyeo) twice to produce two sets of BC2F1 populations. A total of 123 markers for MT-NILs (developed from MTRILs) and 132 markers for DT-NILs (developed from DT-RILs) were used to background and foreground selection. After a marker selection, several BC2F1 plants which contained japonica segments harboring target QTLs (alleles from Tong 88-7 or TR22183), and a small portion of other segments originating from japonica varieties, were selected and backcrossed again to produce BC3F1 populations. In this generation, 24 markers for MT-NILs and 25 markers for DT-NILs were used in selecting NILs. Finally, 10-18 BC3F1 plants of each introgression type were selected and the seed germination capabilities were evaluated using the BC3F2 seeds with three replications per plant.

Seed production and field experiments

Two sets of RILs were seeded on plastic-tunnel seed beds, and the seedlings were transplanted at a planting density of 30 × 15 cm in the experiment field of Seoul National University, Suwon, Korea (127°36’E, 37° 51’N). Fertilizer was applied at the rate of 110-80-80 kg N-P-K ha-1. Field management and chemical input for disease and pest control followed the conventional methods of the experimental farm. The seeds were harvested 45 days after the heading date, from the beginning of September to the beginning of October, and they were dried in a wellventilated glass house to a constant weight. Well-dried seeds were put in envelopes and then stored under natural room conditions at temperatures from > 10℃ in January to < 35℃ in August and relative humidity (RH%) from 50% in February to 85% in July (monthly average data from year 2004 to 2006).

Evaluation of seed germination capability

Bulked seeds from five plants of each line were harvested in 2003, 2004, and 2005 and stored for 3, 2, and 1 years, respectively. The seeds harvested in 2006 served as a control. Seed germination capability was evaluated in March 2007. Seed germination tests for each storage period were replicated three times with 100 seeds per replication. Seeds were placed on a whatman filter paper moistened with distilled water in a 6-cm Petri-dish and incubated at 30℃ and 100% relative humidity in the dark for 7 days. Germination was evaluated visually by protrusion of the radicle from the hull by 2 mm. Seed germination capability (%) was generated by SGR/NSGR × 100, where the SGR was the germination rate of stored seeds, and the NSGR was the germination rate of the non-stored seeds. The same formula applied to all storage periods. Seed germination capability was transformed by sin-1 (x)-0.5 for statistical analysis (Gu et al., 2004).

For evaluating seed germination capability on the developed NILs, the CDT was used. Seeds of NILs and their parents were harvested in autumn in 2008 and well-dried in a greenhouse. To break seed dormancy, seeds were kept in an oven at 30℃ for 1 day, followed by 45℃ for 3 days, then followed by 30℃ for 1 day. For prompt aging, the seeds were put in a Constant Temperature and Humidity Chamber (Hanbaek Sci. Co., Ltd., Korea) at 40℃ and > 90% RH for 2 weeks. After germination test, the seed germination capability was calculated by SGR/ NSGR × 100, where the SGR was the seed germination rate for CDT, and the NSGR was the germination rate of the non-CDT seeds.

Data analysis

Following classical quantitative genetics theory (Falconer and Mackay, 1996), the phenotypic value of a RIL (yijk) is described by the genetic model:

graphic file with name ksmcb-31-385-e100.jpg

Where yijk is the kth observation for the jth storage period of the ith RIL; μ is the mean of germination capability over all lines and all storage periods, αi and βj are the main genotype, storage period effects, respectively; (αβ)ij is the genotype × storage period interaction, and єijk is the error term, including random error and residual effect. ANOVA for each RIL population over storage periods was performed using the SAS PROC MIXED procedure (SAS Institute, 1999). Variance components for estimating broad sense heritability (h2) were estimated by a model, wherein all factors were considered randomly using the restricted maximum likelihood (REML) option of the SAS PROC VARCOMP procedure (SAS Institute, 1999). The h2 of seed germination capability over storage periods was estimated as:

graphic file with name ksmcb-31-385-e101.jpg

where

graphic file with name ksmcb-31-385-e102.jpg

are the genotype, genotype × storage period, and residual variance components, respectively; j is the storage period; and k is the replicates (Hill et al., 1998).

PCR amplification of markers and linkage map construction

The DNA of each RIL was extracted from leaves collected at the maximum tillering stage, according to the method of Jiang et al. (2008). The new genetic map in MT-RILs was constructed. A total of 363 SSR and 525 STS markers were tested for polymorphisms, which found 196 of SSR (54.0%) and 321 of STS (61.1%) polymorphic markers between Milyang 23 and Tong 88-7. An integrated genetic linkage map consisting of 119 SSR and 97 STS markers was constructed, which covered a total length of 1,666.2 cM, with an average distance of 7.64 cM between adjacent markers. Molecular markers for the DT-RILs were essentially the same as those described earlier (Cho et al., 2007). This map contained 216 markers consisting of 113 SSR and 103 STS markers. The total map length was 1409.0 cM, with average distance of 6.5 cM between adjacent markers. The molecular linkage map was constructed using “Mapmaker/ EXP 3.0” software, and the Kosambi function was used for calculating map distances. A log of the odds (LOD) score of 3.0 was used as the threshold for declaring linkage (Kosambi, 1944; Lincoln et al., 1992).

QTL analysis

The chromosomal locations of main additive effects (QTLs) and additive × additive epistatic effects (epistatic QTLs) were per-formed separately in each storage period and RIL population by composite interval mapping using the mixed linear model ap-proach and QTLMapper 1.0 (Wang et al., 1999). The value of likelihood ratio (LR) corresponding to P ≤ 0.005 (equivalent to LOD = 2.79 for df = 3) was used in this QTL analysis. To deter-mine the empirical significance threshold for declaring a QTL, 10,000 permutations were performed to calculate the thresh-olds of LOD for seed germination capability of each storage period at P = 0.05 using the software Qgene 3.06 for Macintosh (Kim et al., 2004; Nelson, 1997). The threshold of LOD score (P = 0.05) by permutation test was ranged from 2.96 to 3.02 in various storage periods of both populations. For epistatic QTLs, the LR value corresponding to P ≤ 0.001 was used as the threshold for claiming the presence of putative epistatic QTLs. The description of epistatic QTLs was followed by Yang et al. (2010). The proportion of total phenotypic variance explained (PVE) was estimated by the sum of phenotypic effects (R2) of each QTL or epistatic QTL for each storage period. Nomencla-ture for the QTL was modified from that described by McCouch et al. (1997).

RESULTS

Variation of seed germination capability

The seed germination capability of the two indica varieties, Milyang 23 and Dasanbyeo, maintained high seed germination capability (above 80%) in all storage periods. However, the two japonica varieties, Tong 88-7 and TR22183, was reduced as the storage period increased, extremely reduced at the 3-year storage period (below 20%) (Fig. 1). The means of seed germi-nation capability were evidently reduced in 3-year storage peri-od than the other two storage periods in both RIL populations, indicating that germination capability was largely affected by storage period (Fig. 1). The h2 computed across storage peri-ods were 87.2% in the MT-RILs and 82.5% in the DT-RILs, showing high amounts of genetic effect in seed germination capability.

Fig. 1. Frequency distributions of seed germina-tion capability in the two RIL populations. Arrow-heads showed mean values for parents and RILs. M, T, D, and TR on arrowheads indicated Milyang 23, Tong 88-7, Dasanbyeo, and TR22183, respectively. Bars under arrowheads indicated standard deviation (10 replications) for seed germination capability in respective parents.

Fig. 1.

Open in a new tab

Main additive effect (QTL)

In the MT-RILs, a total of five QTLs for all seed storage periods were identified on chromosomes 1, 5, 7 (2 regions), and 9 (Table 1 and Fig. 2). Two QTLs, qMT-SGC7.2 (S07080-RM234) and qMT-SGC9.1 (S09040-S09045), were consistently identi-fied in all seed storage periods. One QTL, qMT-SGC5.1 (S05095-RM480), was identified in 2-year and 3-year seed storage peri-ods, whereas both qMT-SGC1.1 and qMT-SGC7.1 were identi-fied only in one seed storage period (3-year and 1-year storage period, respectively). The phenotypic variance explained by individual QTL ranged from 8.5% to 17.5%. Cumulative effects of these QTLs in each storage period ranged from 32.4% (2-year storage) to 47.5% (3-year storage) of the PVE. Milyang 23 alleles at four loci (qMT-SGC5.1, qMT-SGC7.1, qMT-SGC7.2, and qMT-SGC9.1) seemed to increase seed germination capa-bility, but the allele at qMT-SGC1.1 had the opposite effect.

Table 1.

Additive main effect (QTL) characteristics for seed germination capability in two RIL populations

Storage period QTLs Chromosome Flanking markers LOD Aa R2 (%) PVE (%)b
MT-RILs
   1-year qMT-SGC7.1 7 RM214-S07055 4.80 6.4 11.6 45.7
qMT-SGC7.2 7 S07080-RM234 8.21 7.9 17.5
qMT-SGC9.1 9 S09040-S09045 3.90 5.8 9.5
   2-year qMT-SGC5.1 5 S05095-RM480 3.15 5.9 7.9 32.4
qMT-SGC7.2 7 S07080-RM234 8.68 8.4 16.0
qMT-SGC9.1 9 S09040-S09045 3.17 5.9 8.5
   3-year qMT-SGC1.1 1 S01027-RM1 6.19 -8.4 10.7 47.5
qMT-SGC5.1 5 S05095-RM480 5.96 9.1 12.5
qMT-SGC7.2 7 S07080-RM234 6.76 9.1 12.5
qMT-SGC9.1 9 S09040-S09045 5.97 8.9 11.8
DT-RILs
   1-year qDT-SGC3.1 3 S03048-RM251 4.56 4.2 10.5 36.6
qDT-SGC9.1 9 S09040-S09049 9.16 6.6 26.1
   2-year qDT-SGC2.1 2 S02054-S02057 9.86 6.2 12.2 63.6
qDT-SGC3.1 3 S03048-RM251 7.25 6.0 11.3
qDT-SGC9.1 9 S09040-S09049 22.32 11.3 40.1
   3-year qDT-SGC2.1 2 S02054-S02057 7.50 6.0 12.4 48.4
qDT-SGC3.1 3 S03048-RM251 5.79 6.7 15.6
qDT-SGC9.1 9 S09040-S09049 10.29 7.7 20.5
Open in a new tab

aQTL effects of the maternal parents (Milyang 23 or Dasanbyeo); positive (+) means maternal genotype increasing seed germination capability and negative (-) means maternal genotype decreasing seed germination capability.

bTotal percentage of phenotypic variance explained (PVE) by all QTLs for each storage period

Fig. 2. Milyang 23/Tong 88-7 linkage map show-ing the genetic location of QTLs for seed ger-mination capability in various storage periods.

Fig. 2.

Open in a new tab

In the DT-RILs, a total of three QTLs, qDT-SGC2.1, qDT-SGC3.1 and qDT-SGC9.1, were identified on chromosomes 2, 3, and 9, respectively (Table 1 and Fig. 3). The two QTLs, qDT-SGC3.1 and qDT-SGC9.1, were consistently identified in all seed storage periods, and qDT-SGC2.1 was identified in 2-year and 3-year seed storage periods. The phenotypic variance explained by individual QTL ranged from 10.5% to 40.1%. Col-lectively, the PVE by these QTLs ranged from 32.4% (2-year storage) to 47.5% (3-year storage). The QTL, qDT-SGC9.1, had the largest phenotypic variance in all seed storage periods, accounting for 26.1% (1-year storage), 40.1% (2-year storage), and 20.5% (3-year storage) of the phenotypic effect. The PVE in each storage period ranged from 36.6% (1-year storage) to 63.6% (2-year storage). Dasanbyeo alleles increased seed germination capability in all QTLs.

Fig. 3. Dasanbyeo/TR22183 linkage map showing the genetic location of QTLs for seed germination ca-pability in various storage periods.

Fig. 3.

Open in a new tab

Additive × additive epistatic effect (epistatic QTL)

In MT-RILs, a total of eight epistatic QTL pairs were identified; three pairs in 1-year storage period, three pairs in 2-year stor-age period, and two pairs in 3-year storage period (Table 2). Among these loci, the phenotypic variance contributed by each epistatic QTL pair ranged from 3.5% to 9.7% and the PVE in the three storage periods ranged from 9.9% to 23.7%. These epistatic QTL pairs were partitioned into six interactions be-tween two loci with only epistatic effects (NN) and two interac-tions between a QTL with additive effect and a locus without significant additive effect (AN). In DT-RILs, six epistatic QTL pairs were separately identified in the three storage periods (Table 2). The phenotypic variance explained by each epistatic QTL pair ranged from 1.4% to 7.5% and the PVE in the three storage periods ranged from 11.9% to 34.2%. These epistatic QTL pairs were partitioned into fourteen NN interactions, two AN interactions, and three interactions between two QTLs with additive effects (AA). In the AA types, the interactions between two QTLs, qDT-SGC3.1 and qDT-SGC9.1, were identified in two storage periods with an increased epistatic effect originated from the parental digenic combination, and an epistatic QTL pair between qDT-SGC2.1 and qDT-SGC9.1 was identified in 2-year storage period with an increased epistatic effect arose from different parents.

Table 2.

Additive × additive epistatic interactions (epistatic QTLs) for seed germination capability identified in two RIL populations

Storage period Chr/Ina Flanking markers Chr/In Flanking markers LOD Interactionb AAc R2 (aa) (%) PVEd
MT-RILs
   1-year 4/1 RM335-S04052 4/5 RM252-RM303 3.36 NN -4.9 6.2 23.7
4/9 S04097-RM317 6/6 RM50-S06033 3.79 NN 5.2 6.9
9/8 S09040-S09045 7/18 RM118-S07101 7.12 AN -6.5 10.6
   2-year 1/8 S01078-RM129 2/1 S02000-RM485 4.27 NN 7.3 9.7 21.5
4/4 RM119-RM252 8/6 S08038-RM72 3.29 NN -6.0 6.7
7/15 S07080-RM234 9/5 S09031-S09034 9.74 AN -5.3 5.2
   3-year 2/15 RM450-RM318 7/19 RM118-S07118 3.46 NN -5.7 4.5 9.9
5/10 RM430-RM440 8/10 RM44-RM331 2.97 NN 6.3 5.5
DT-RILs
   1-year 2/13 RM263-RM318 3/1 RM132-RM523 6.77 NN 4.7 6.9 34.2
3/16 S03048-RM251 9/6 S09040-S09049 5.78 AA 4.5 6.4
6/3 RM510-S06018 9/1 RM219-RM316 5.33 NN -4.1 5.4
8/13 S08107-RM477 9/13 S09073-S09075 5.12 NN 4.0 5.0
8/14 RM477-RM264 12/12 RM270-RM17 5.66 NN -4.0 5.0
10/1 RM184-RM467 10/6 S1001B-S10019 3.67 NN -4.2 5.5
   2-year 2/6 S02054-S02057 9/6 S09040-S09049 4.41 AA -4.3 2.1 11.9
3/16 S03048-RM251 7/9 S07099-S07103 10.50 AN -4.3 2.1
3/25 S03145-S03130 5/5 S05004-S05009 3.70 NN -3.2 1.2
4/2 S04060-S04058 5/16 RM413-RM249 3.85 NN 3.8 1.7
5/8 S05014B-S05029 9/17 RM215-RM201 2.89 NN 4.0 1.8
7/7 RM234-RM118 8/9 RM25-S08090 4.68 NN -5.2 3.1
   3-year 1/5 S01054-RM600 5/37 RM87-RM31 3.20 NN 4.5 4.1 32.8
2/4 RM341-RM475 5/34 RM173-RM538 5.47 NN -5.9 7.2
3/1 RM132-RM523 3/6 S03015A-S03015B 3.72 NN 4.3 3.8
3/16 S03048-RM251 9/6 S09040-S09049 12.70 AA 5.5 6.3
4/2 S04060-S04058 4/12 RM127-S04128 4.17 NN 4.3 3.8
4/7 S04097-RM349 5/17 RM249-RM289 5.01 NN -6.0 7.5
Open in a new tab

aChr/In and Chr/In represent the chromosome number interval of the tested points in the analysis.

bTypes of epistatic interactions. AA interactions between two QTL with additive effects, NA (AN) interactions between a QTL with additive effects and a locus without significant additive effects, NN interaction between two loci with epistatic effects only

cAA means the epistatic effects. Positive (+) means that parental digenic genotypes increasing seed germination capability, and negative (-) means that recombinant alleles from two parents increasing seed germination capability.

dTotal percentage of phenotypic variance explained (PVE) by all epistatic QTLs for each storage period

QTL validation for seed germination capability in NILs

Six NILs developed from the MT-RILs were evaluated for seed germination capability by CDT (Fig. 4). The MT_NIL1 and MT_NIL2 did not contain Tong 88-7 segments in QTL regions on chromosome 5, 7, and 9. As expected, the averages of seed germination capability in the MT_NIL1 and MT_NIL2 were 81.2% and 79.9%, respectively, which was similar to that of Milyang 23 (81.3%NIL3 and MT_NIL4 contained Tong 88-7 allele at the qMT-SGC7.2 and qMT-SGC5.1 on chromosome 7 and 5, respectively. In the MT_NIL3 and MT_ NIL4, seed germination capability was significantly reduced com- paring to that of Milyang 23, suggesting that the severe reduc-tion of germination capability in these NILs was caused by the effect of the Tong 88-7 alleles at qMT-SGC7.2 and qMT-SGC5.1.The MT_NIL5 and MT_NIL6 contained Tong 88-7 allele at qMT-SGC9.1 on chromosome 9, and their seed germi-nation capability was reduced the most severely among six MT-NILs.

Fig. 4. Genomic composition of NILs derived from backcrossing Milyang 23 to MT-RILs. Black bar: Tong 88-7 homozygote segment; white bar: Milyang 23 homozygote segment. *Seed germination ca-pability of NIL significantly different with Milyang 23 at P < 0.0001 based on the t-test. ns: not significant.

Fig. 4.

Open in a new tab

The same strategy was applied to DT-NIL for confirmation of three QTLs identified from the DT-RIL population. Among four DT-NILs, the DT_NIL2, DT_NIL3, and DT_NIL4 contained the TR22183 allele of qDT-SGC2.1, qDT-SGC3.1, and qDT-SGC9.1 on chromosome 2, 3, and 9, respectively (Fig. 5). Seed germination capability of the DT_NIL2, DT_NIL3, and DT_NIL4 was 61.0%, 53.3%, and 51.6%, respectively, which was signifi-cantly lower than that of recurrent parent Dasanbyeo.

Fig. 5. Genomic composition of NILs derived from backcrossing Dasanbyeo to DT-RILs. Black bar: TR22183 homozygote segment; white bar: Dasanbyeo homozygote segment; gray bar: heterozygote segment; broken line: chromo-some regions for the identified QTLs. *Seed germination capabil-ity of NIL significantly different with Dasanbyeo at P < 0.0001 based on the t-test. ns: not significant.

Fig. 5.

Open in a new tab

These results collectively demonstrate that the QTLs identi-fied in two sets of RILs reliably affected the seed germination capability of certain genotypes having different genetic back-grounds.

DISCUSSION

We were able to identify five and three QTLs associated with seed germination capability in MT-RILs and DT-RILs in all stor-age periods, respectively (Table 1, Figs. 2 and 3). Among the QTLs identified in both RIL populations, one pair of QTLs was located at the same chromosome region on chromosome 9: qMT-SGC9.1 (S09040-S09045) for the MT-RILs and qDT-SGC9.1 (S09040-S09049) for the DT-RILs. This chromosomal region originated from indica varieties, was also identified for seed longevity or seed storability in the previous studies (Miura et al., 2002; Sasaki et al., 2005; Xue et al., 2008), suggesting that this QTL might be widespread among indica rice varieties. In the MT-RILs, one QTL, qMT-SGC7.2, on chromosome 7 was detected in all storage periods and possibly corresponds with RC7 detected by Sasaki et al. (2005). A novel QTL, qMT-SGC5.1, was repeatedly identified in 2-year and 3-year storage periods, implying that it was likely expressed in the severely deteriorated seeds caused by long-term seed storage. The qMT-SGC7.1 (RM214-S07055) on chromosome 7 is an inter-esting QTL because it was only detected in the 1-year storage period. Since 1-year storage period may not be regarded as long-term storage, qMT-SGC7.1 may represent QTL for seed dormancy rather than seed germination capability. It is difficult to clearly distinguish the phenotypes of seed dormancy from seed germination capability. Unexpectedly, the allele originating from Milyang 23 in the locus of qMT-SGC1.1 decreased seed germination capability, whereas all other alleles from Milyang 23 in the other loci of the QTLs increased seed germination capa-bility. It was revealed that seed germination capability exhibited different expression patterns between different rice varieties, which was consistent with the findings of previous studies (Kameswara and Jackson, 1996; 1997; Zeng et al., 2006). In DT-RILs, we identified two novel QTLs (qDT-SGC2.1 and qDT-SGC3.1) for seed germination capability, which increased the seed germination capability in Dasanbyeo alleles.

Developing NILs is an important fundamental step for QTL cloning and molecular marker-assisted selection. When a QTL is ‘Mendelized’, the candidate gene can be identified by high-resolution linkage mapping (Yamamoto et al., 2009). In rice, among the identified several QTLs for seed germination capa-bility on seven chromosomes, the QTL on chromosome 9 was identified consistently in various mapping populations (Miura et al., 2002; Sasaki et al., 2005; Shigemune et al., 2008; Xue et al., 2008). In the intermediate stage for cloning these QTLs, we developed NILs corresponding to QTLs, including qMT-SGC5.1, qMT-SGC7.2, qMT-SGC9.1 from MT-RILs and qDT-SGC2.1, qDT-SGC3.1, qDT-SGC9.1 from DT-RILs, by using molecular marker-assisted selection through foreground selection and background selection. After developing NILs using the CDT method, we confirmed that each genomic region on target QTLs controlled the seed germination capability (Figs. 4 and 5). This should allow the introgression of favorable alleles from high-germination capability varieties into high-yielding varieties, such as japonica hybrid rice, using marker-assisted backcross-ing.

Since we are also interested in the improvement of seed germination capability under the proper storage condition com-pared to that under the natural condition, the phenotype per-formance of seed germination capability was calculated by dividing SGR with NSGR, instead of using primary seed germi-nation rates even though this secondary data process may enlarge errors and the results may not be interpreted directly. In addition, we detected high correlations between seed germina-tion capability calculated with SGR/NSGR and primary seed germination rates in each seed storage period of both RIL populations (r > 0.993 in MT-RILs and r > 0.986 in DT-RILs). Moreover, the high h2 for seed germination capability in both RIL populations (86.9% in MT-RILs and 81.7% in DT-RILs) suggests that natural seed storage method can evaluate seed germination capability and reduce environmental errors. The indica alleles of six confirmed QTLs from the two RIL popula-tions showed that indica alleles employed increasing seed germination capability. Although these QTLs are the most im-portant determinants of the seed germination capability, epi-static interactions explained relatively high phenotypic variance in some storage periods, indicating that epistatic interactions were an important component on the genetic variance for seed germination capability. A series of NILs for a target trait provided the opportunity to analyze epistatic interactions among identi-fied genes by combining individual NILs (Doi et al., 2008; Rahman et al., 2009; Yano et al., 2001). In addition, sometimes the effects of QTLs are also subjected to environmental changes, which can lead to dramatic differences in the phenotypic effects of the QTLs. The results demonstrate that for the marker-assisted breeding programs for improving seed germination capability, attention should be paid to the direct effects of the minor QTLs, epistatic QTLs in addition to the most important process of pyramiding reliable QTLs (especially major QTLs).

The molecular identification of several reliable QTLs in the two sets of RILs, reported herein, may help to elucidate the poorly understood germination capability under different seed storage conditions. We expect that our results will provide the genetic information on map-based cloning and MAS that will be vital to the advancement of breeding strategies and breeding resources for increased seed germination capability.

Acknowledgments

This work was supported by a Korea Research Foundation (KRF) grant funded by the Korean government (MEST) (No. F00003).

References

  • 1.Bentsink L., Alonso-Blanco C., Vreugdenhil D., Tesnier K., Groot S.P.C., Koornneef M. Genetic analysis of seed-soluble oligosaccharides in relation to seed storability of Ara-bidopsis. Plant Physiol. (2000);124:1595–1604. doi: 10.1104/pp.124.4.1595. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Bewley J.D., Black M. Seeds: physiology of develop-ment and germination. Plenum Press; New York, NY, USA: (1994). [Google Scholar]
  • 3.Cho Y.I., Jiang W., Chin J.H., Piao Z., Cho Y.G., McCouch S.R., Koh H.J. Identification of QTLs associated with physiological nitrogen use efficiency in rice. Mol. Cells. (2007);23:72–79. [PubMed] [Google Scholar]
  • 4.Chung G.S.,, Heu M.H. Improvement of tongil-type rice cultivars from indica/japonica hybridization in Korea. In Biotech-nology in Agriculture and Forestry 14-Rice, Y.P.S. Bajaj, ed. Springer; Berlin Heidelberg, New York: (1991). pp. 105–112. [Google Scholar]
  • 5.Clerkx E.J.M., Blankestijn-De Vries H., Ruys G.J., Groot S.P.C., Koornneef M. Genetic differences in seed longevity of various Arabidopsis mutants. Physiol. Plant. (2004);121:448–461. [Google Scholar]
  • 6.Doi K., Yasui H., Yoshimura A. Genetic variation in rice. Curr. Opin. Plant Biol. (2008);11:144–148. doi: 10.1016/j.pbi.2008.01.008. [DOI] [PubMed] [Google Scholar]
  • 7.Delouche J.C., Baskin C.C. Accelerated aging tech-niques for predicting the relative storability of seed lots. Seed Sci. Technol. (1973);1:427–452. [Google Scholar]
  • 8.Ellis R.H., Roberts E.H. The quantification of aging and survival in orthodox seeds. Seed Sci. Technol. (1981);9:373–409. [Google Scholar]
  • 9.Ellis R.H., Osei-Bonsu K., Roberts E.H. The influence of genotype, temperature and moisture on seed longevity in chickpea, cowpea and soybean. Ann. Bot. (1982);50:69–82. [Google Scholar]
  • 10.Falconer D.S., Mackay T.F.C. Introduction to quantita-tive genetics. Longmans Green; Harlow, Essex, UK: (1996). [Google Scholar]
  • 11.Gu X.Y., Kianian S.F., Foley M.E. Multiple loci and epistases control genetic variation for seed dormancy in weedy rice (Oryza sativa). Genetics. (2004);166:1503–1516. doi: 10.1534/genetics.166.3.1503. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Hill J., Becker H.C., Tigerstedt P.M.A. Quantitative and ecological aspects of plant breeding. Chapman & Hall; London, UK: (1998). [Google Scholar]
  • 13.Jiang W., Chu S.H., Piao R.H., Chin J.H., Jin Y.M., Lee J., Qiao Y., Han L., Piao Z., Koh H.J. Fine mapping and candidate gene analysis of hwh1 and hwh2, a set of comple-mentary genes controlling hybrid breakdown in rice. Theor. Appl. Genet. (2008);116:1117–1127. doi: 10.1007/s00122-008-0740-4. [DOI] [PubMed] [Google Scholar]
  • 14.Jiang W., Lee J., Chu S.H., Ham T.H., Woo M.O., Cho Y.I., Chin J.H., Han L.Z., Xuan Y., Yuan D., et al. Genotype × en-vironment interactions for chilling tolerance of rice recombinant inbred lines under different low temperature environments. Field Crops Res. (2010);117:226–236. [Google Scholar]
  • 15.Kameswara N., Jackson M.T. Seed longevity of envi-ronment and storage longevity of japonica rices (Oryza sativa L.). Seed Sci. Res. (1996);6:17–21. [Google Scholar]
  • 16.Kameswara N., Jackson M.T. Variation in seed lon-gevity of rice cultivars belonging to different isozyme groups. Gene Res. Crop Evol. (1997);44:159–164. [Google Scholar]
  • 17.Kim K.M., Kown Y.S., Lee J.J., Eun M.Y., Shon J.K. QTL mapping and molecular marker analysis for the resistance of rice to ozone. Mol. Cells. (2004);17:151–155. [PubMed] [Google Scholar]
  • 18.Kosambi D.D. The estimation of map distances from re-combination values. Ann. Eugn. (1944);12:172–175. [Google Scholar]
  • 19.Kwon Y.S., Kim K.M., Eun M.Y., Sohn J.K. Quantita-tive trait loci mapping associated with plant regeneration ability from seed derived calli in rice (Oryza sativa L.). Mol. Cells. (2001);11:64–67. [PubMed] [Google Scholar]
  • 20.Lincoln S.,, Daly M.,, Lander E.S. Constructing genetic maps with MAPMAKER/EXP 3.0. Whitehead Institute Technical Report. 2nd eds Whitehead Institute, Cambridge; Massachusetts, USA: (1992). [Google Scholar]
  • 21.McCouch S.R., Cho Y.G., Yano M., Paul E., Kinoshita T. Report on QTL nomenclature. Rice Genet. Newslett. (1997);14:11–13. [Google Scholar]
  • 22.Miura K., Lin S.Y., Yano M., Nagamine T. Mapping quantitative trait loci controlling seed longevity in rice (Oryza sa-tiva L.). Theor. Appl. Genet. (2002);104:981–986. doi: 10.1007/s00122-002-0872-x. [DOI] [PubMed] [Google Scholar]
  • 23.Nelson J.C. QGENE, software for marker-based genome analysis and breeding. Mol. Breed. (1997);3:239–245. [Google Scholar]
  • 24.Padma V., Reddy B.M. Evaluation of rice genotypes for dormancy duration and seed storability under natural and ac-celerated ageing. Seed Res. (2000);28:158–165. [Google Scholar]
  • 25.Powell A., Matthews S. Application of the controlled deterioration vigour test to detect seed lots of Brussels sprouts with low potential for storage under commercial conditions. Seed Sci. Technol. (1984);12:649–657. [Google Scholar]
  • 26.Rahman M.L., Jiang W., Chu S.H., Qiao Y., Ham T.H., Woo M.O., Lee J.H., Khanam M.S., Chin J.H., Jeung J.U., et al. High-resolution mapping of two rice brown planthopper resistance genes, Bph20(t) and Bph21(t), originating from Oryza minuta. Theor. Appl. Genet. (2009);119:1237–1246. doi: 10.1007/s00122-009-1125-z. [DOI] [PubMed] [Google Scholar]
  • 27.Rajjou L., Debeaujon I. Seed longevity: survival and maintenance of high germination ability of dry seeds. C.R. Biol. (2008);331:796–805. doi: 10.1016/j.crvi.2008.07.021. [DOI] [PubMed] [Google Scholar]
  • 28.Roberts E.H. Loss of viability and crop yields. In Viability of Seeds, E.H. Roberts, ed. Chapman and Hall; London, UK: (1972). pp. 307–359. [Google Scholar]
  • 29.SAS Institute Inc. SAS/Stat User’s Guide, Version 8.2. SAS institute, Inc.; Cary, NC, USA: (1999). [Google Scholar]
  • 30.Sasaki K., Fukuta Y., Sato T. Mapping of quantitative trait loci controlling seed longevity of rice (Oryza sativa L.) after various periods of seed storage. Plant Breed. (2005);124:361–366. [Google Scholar]
  • 31.Shigemune A., Miura K., Sasahara H., Goto A., Yoshida T. Role of maternal tissues in qLG-9 control of seed lon-gevity in rice (Oryza sativa L.). Breed. Sci. (2008);58:1–5. [Google Scholar]
  • 32.Siddique S.B., Seshu D.V., Pardee W.D. Rice cultivar variability in tolerance for accelerated aging of seed. IRRI Res. Pap. Ser. (1988);131:2–7. [Google Scholar]
  • 33.Singh R.K., Ram H.H. Inheritance study of soybean seed storability using an accelerated aging test. Field Crops Res. (1986);13:89–98. [Google Scholar]
  • 34.Wang D.L., Zhu J., Li Z.K., Paterson A.H. Mapping QTLs with epistatic effects and QTL × environment interactions by mixed linear model approaches. Theor. Appl. Genet. (1999);99:1255–1264. [Google Scholar]
  • 35.Xue Y., Zhang S.Q., Yao Q.H., Peng R.H., Xiong A.S., Li X., Zhu W.M., Zhu Y.Y., Zha D.S. Identification of quantita-tive trait loci for seed storability in rice. Euphytica. (2008);164:739–744. [Google Scholar]
  • 36.Yamamoto T., Yonemaru J., Yano M. Towards the understanding of complex traits in rice: substantially of superfi-cially? DNA Res. (2009);16:141–154. doi: 10.1093/dnares/dsp006. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Yang X., Guo Y., Yan J., Zhang J., Song T., Rocheford T., Li J.S. Major and minor QTL and epistasis contribute to fat-ty acid composition and oil content in high-oil maize. Theor. Appl. Genet. (2010);120:665–678. doi: 10.1007/s00122-009-1184-1. [DOI] [PubMed] [Google Scholar]
  • 38.Yano M., Kojima S., Takahashi Y., Lin H.X., Sasaki T. Genetic control of flowering time in rice, a short-day plant. Plant Physiol. (2001);127:1425–1429. [PMC free article] [PubMed] [Google Scholar]
  • 39.Zeng D.L., Qian Q., Yasukumi K., Teng S., Hirosi F. Study on storability and morphological index in rice (Oryza sativa L.) under artificial aging. Acta Agron. Sin. (2002);28:551–554. [Google Scholar]
  • 40.Zeng D.L., Guo L.B., Xu Y.B., Yasukumi K., Zhu L.H., Qian Q. QTL analysis of seed storability in rice. Plant Breed. (2006);125:57–60. [Google Scholar]

Articles from Molecules and Cells are provided here courtesy of Korean Society for Molecular and Cellular Biology

RESOURCES