- Email: [email protected]
QTL Analysis of Anoxic Tolerance at Seedling Stage in Rice
Rice Science, 2010, 17(3): 192–198 Copyright © 2010, China National Rice Research Institute. Published by Elsevier BV. All rights reserved DOI: 10.1016/S1672-6308(09)60017-2
QTL Analysis of Anoxic Tolerance at Seedling Stage in Rice WANG Yang1, 2 , GUO Yuan1, HONG De-lin1 (1State Key Laboratory of Crop Genetics and Germplasm Enhancement, Nanjing Agricultural University, Nanjing 210095, China; 2 Department of Agricultural Resource and Environment, Heilongjiang University, Harbin 150008, China)
Abstract：Coleoptile lengths of 7-day-old seedlings under anoxic stress and normal conditions were investigated in two permanently segregated populations and their parents in rice (Oryza sativa L.). Using anoxic response index, a ratio of coleoptile length under anoxic stress to coleoptile length under normal conditions, as an indicator of seedling anoxic tolerance (SAT), QTLs for SAT were detected. Two loci controlling SAT, designated as qSAT-2-R and qSAT-7-R, were detected in a recombinant inbred line (RIL) population (247 lines) derived from a cross between Xiushui 79 (japonica variety) and C Bao (japonica restorer line). qSAT-2-R, explaining 8.7% of the phenotype variation, was tightly linked with the SSR marker RM525. qSAT-7-R, explaining 9.8% of the phenotype variation, was tightly linked with the marker RM418. The positive alleles of the two loci came from C Bao. Six loci controlling SAT, designated as qSAT-2-B, qSAT-3-B, qSAT-5-B, qSAT-8-B, qSAT-9-B and qSAT-12-B, were detected in a backcross inbred line (BIL) population (98 lines) derived from a backcross of Nipponbare (japonica)/Kasalath (indica)//Nipponbare (japonica). The positive alleles of qSAT-2-B, qSAT-3-B and qSAT-9-B, which explained 16.2%, 11.4% and 9.5% of the phenotype variation, respectively, came from Nipponbare. Besides, the positive alleles of qSAT-5-B, qSAT-8-B and qSAT-12-B, which explained 7.3%, 5.8% and 14.0% of the phenotype variation, respectively, were from Kasalath. Key words: rice (Oryza sativa); seedling; anoxic tolerance; quantitative trait locus; recombinant inbred line; backcross inbred line
Direct seeding in rice (Oryza sativa L.) is a labor-saving and highly efficient cultivation pattern (Chen, 2003), and is now becoming more and more popular. Compared with the traditional transplanting cultivation pattern, it can increase number of panicles per unit area and 1000-grain weight of rice (Jing et al, 2008). In the rice production through direct seeding, one of the key factors affecting grain yield is the seedling establishment in paddy fields (Wu et al, 2006). In flooded or water-saturated soil, seedling rates are lowered due to anoxic stress (Fred et al, 1981). Breeding rice varieties with anoxic tolerance at the seedling stage would be helpful in increasing seedling rate in paddy fields of direct seeding. Rice seed can germinate under anoxic conditions, but root growth and leaf emergence are hindered. To get enough O2 for the growth of root and leaf, the coleoptile elongates faster and longer under anoxia than normal conditions (Yamauchi et al, 1993). Coleoptiles of rice varieties with anoxia tolerance had stronger elongation ability than those without anoxic tolerance Received: 1 December 2009; Accepted: 12 March 2010 Corresponding author: HONG De-lin ([email protected])
(Setter et al, 1994; Yamauchi and Biswas, 1997), and coleoptiles of japonica rice elongated faster and longer than those of indica rice under anoxia stress (Hou et al, 2004). We also found genetic variations in anoxic tolerance of seedlings among japonica rice varieties (Wang et al, 2009). These genetic variations can be utilized to develop varieties with anoxic resistance in rice breeding. Using the shoot length of seedlings germinated in water at a depth of 20 cm for 5 days in the dark as an indicator of anoxic germination (AG), Hou et al (2004) detected five QTLs located on chromosomes 1, 2, 5, 5 and 7 in a population of recombinant inbred lines (RIL). According to the direction of the additive effects, the positive alleles at the loci of qAG-1, qAG-G2 and qAG-7 derived from Kinmaze, while at the loci of qAG-5a and qAG-5b, the DV85 alleles increased AG. However, the RIL population mentioned above is derived from a cross of Kinmaze (japonica)/DV85 (indica), and is not a breeding population. Studies on QTL mapping for seedling anoxic tolerance by using a breeding population of japonica rice have not been reported yet. To discover more genetic information for
WANG Yang, et al. QTL Analysis of Anoxic Tolerance at Seedling Stage in Rice
breeding rice varieties suitable for direct seeding, we conducted QTL mapping for seedling anoxic tolerance using a population of 247 RILs from the cross of Xiushui 79 (japonica variety) and C Bao (japonica restorer line), and a population of 98 lines derived from a backcross of Nipponbare (japonica)/Kasalath (indica)//Nipponbare.
MATERIALS AND METHODS Rice materials A population of 247 recombinant inbred lines (RILs) from a cross of Xiushui 79 (japonica variety) and C Bao (japonica restorer line) and their parents were used. In 2008, the population of F10:11 generation was obtained. Another population of 98 backcross inbred lines (BILs) from a backcross of Nipponbare (japonica)/Kasalath (indica)//Nipponbare and their parents were also used. The BIL population of BC1F13 was obtained in 2007. Both populations were developed by the single-seed descent method. Field planting In 2007, seeds of the BIL population and their parents were sown in a rice seedling bed in May, and the seedlings were transplanted in June at Jiangpu Experimental Station, Nanjing Agricultural University, China. In 2008, the experiment was conducted in the same way for the seeds of the RIL population and their parents. Each material was transplanted to one plot in a paddy field. Twelve hills per plot were transplanted at a density of 16.7 cm×20.0 cm, with one seedling per hill. Field managements followed traditional practice. Three thousand seeds were harvested from middle five hills of each plot at 45–50 days after heading (harvesting time was determined according to the panicle size). Seed dormancy was broken by keeping the seeds at 50ºC in a drying oven for five days. Measurement of coleoptile length of seedlings germinated under anoxic conditions Two hundred seeds of each material were surface sterilized by immersion in 0.6% sodium hypochlorite solution for 15 min and rinsed with tap water for three
193
times. Then the seeds were soaked in water for 48 h at 25ºC. After that, 20 seeds per line which absorbed enough water were put into a plastic culture tray (modified from seedling trays) with 150 compartments (1.9 cm × 1.9 cm × 3.8 cm) and a drain hole (0.6 cm diameter) under each compartment. Four seeds were placed at the bottom of each compartment and then germinated in the depth of 5 cm water. The plastic culture tray was put on the surface of vermiculite with a depth of 5 cm water in a plastic transfer box (37 cm × 25 cm × 9 cm). The distance between seeds and water surface was 5 cm. The length of the coleoptile of 7-day-old seedlings was measured from the base of coleoptile to the top and accurate to 1 mm. The mean of the coleoptile length of 10 seedlings was calculated in each line. The germination experiment was conducted in a GXZ-type incubator under the photoperiod of 16 h darkness at 20°C and 8 h light at 30°C with two replications. Measurement of coleoptile length of seedlings germinated under normal conditions Twenty seeds per line which absorbed enough water were put on the surface of two layers of filter paper covering the slope board at 70-degree tilt position in a germination box. Water in the germination box was kept 3 cm high in order to make paper moist without soaking seeds. The other germination conditions and the measurement of coleoptile length of 7-day-old seedlings were the same as those under the anoxic conditions. Calculation of seedling anoxic response index The ratio of coleoptile length under anoxic conditions over that under normal conditions was calculated as the anoxic response index of seedlings, and used as an indicator of seedling anoxic tolerance. Data analysis A genetic linkage map of SSR markers for the RIL population has been constructed by our laboratory (Guo et al, 2010). The linkage map spans 744.6 cM of rice genome with 74 SSR markers from all 17 individual linkage groups. The molecular data of the BIL population were obtained from Dr. Yano at the National Institute of
Rice Science, Vol. 17, No. 3, 2010
194
Agrobiological Resources, Japan. The 245 RFLP markers are distributed evenly on 12 chromosomes, covering 1179.9 cM of rice genome with an average distance of 4.8 cM between two markers. The mean of the coleoptile length of each line over two replications was used as a unit for the following statistical analysis. QTL detection was carried out by the Composite Interval Mapping (CIM) method in the Win QTL Cartographer 2.5 software (Wang et al, 2006). The LOD scores were calculated every 2 cM on 12 chromosomes, with the threshold of 2.5. A putative QTL was determined between the markers when the LOD score was larger than the threshold. QTL nomenclature followed the propositions of the Rice Genetics Cooperative (McCouch et al, 1997). In order to distinguish QTLs detected in two different populations, ‘R’ was added to the name of QTLs detected in the RIL population, and ‘B’ was added to the name of QTLs detected in the BIL population. The statistical analyses were executed with the Excel software according to the Method of Experimental Statistics (Gai, 2000).
RESULTS Difference between two parents and variations in two populations for seedling anoxic response index In the RIL population, the mean of the seedling anoxic response index of C Bao (9.30±0.20) was significantly higher than that of Xiushui 79 (7.93±0.04) (t=9.63). The seedling anoxic response index of the
Fig. 1. Frequency distributions of anoxic response index in rice seedlings of Xiushui 79/C Bao RIL population (A) and Nipponbare/ Kasalath//Nipponbare BIL population (B). Anoxic response index is the ratio of coleoptile length of seedlings with seed germinated in water at a depth of 5 cm to coleoptile length under normal germination conditions.
RIL population averaged 9.13±1.89 with a range of 4.21–14.41, which showed approximately normal distribution and transgressive segregation with a coefficient of variation of 20.7% (Fig. 1-A). In the BIL population, the mean of the seedling anoxic response index of Nipponbare (4.58±0.45) was significantly higher than that of Kasalath (2.33±0.07) (t=7.00). The seedling anoxic response index of the BIL population averaged 3.89±0.74 with a range of 2.36–5.56, which showed approximately normal distribution and transgressive segregation with a coefficient of variation of 19.1% (Fig. 1-B). QTLs for seedling anoxic tolerance detected in the two populations In the RIL population, two QTLs conditioning seedling anoxic tolerance were detected on chromosomes 2 and 7 (Table 1 and Fig. 2). qSAT-2-R explained 8.7%
Table 1. QTLs identified for seedling anoxic tolerance (SAT) in the RIL and BIL populations. Population and QTL
Marker interval a
Distance (cM) b
LOD
Variance explained (%)
Additive effect c
RIL population qSAT-2-R RM525–RM2127 0.1 3.27 8.7 -0.77 qSAT-7-R RM418–RM11 2.0 3.36 9.8 -0.87 BIL population qSAT-2-B R3393–C747 1.0 5.34 16.2 +0.30 qSAT-3-B C1488–C63 2.0 3.82 11.4 +0.30 qSAT-5-B R830–R3166 0.0 2.60 7.3 -0.15 qSAT-8-B R1813–C1121 0.3 2.82 5.8 -0.20 qSAT-9-B R2272–R2638 0.0 3.63 9.5 +0.30 qSAT-12-B C1336–R642 3.5 2.99 14.0 -0.30 a Bold letters indicate the nearest marker. b Distance from the nearest marker to putative QTL. c ‘+’ means that positive alleles come from Xiushui 79 in the RIL population or from Nipponbare in the BIL population; ‘-’ means that positive alleles come from C Bao in the RIL population or from Kasalath in the BIL population.
WANG Yang, et al. QTL Analysis of Anoxic Tolerance at Seedling Stage in Rice
of the phenotypic variance and its positive allele came from C Bao. The amplification product of SSR marker RM525, closely linked to qSAT-2-R, showed a band of 140 bp using the total DNA of C Bao as template. qSAT-7-R explained 9.8% of the phenotypic variance and its positive allele came from C Bao. The amplification product of SSR marker RM418, closely linked to qSAT-7-R, showed a band of 250 bp using the total DNA of C Bao as template. In the BIL population, six QTLs governing seedling anoxic tolerance were detected on chromosomes 2, 3, 5, 8, 9 and 12 (Table 1 and Fig. 3). These QTLs explained 5.8% to 16.2% of the phenotypic variances. Nipponbare carried positive alleles on the loci of qSAT-2-B, qSAT-3-B and qSAT-9-B, and Kasalath possessed positive alleles on the loci of qSAT-5-B, qSAT-8-B and qSAT-12-B.
195
Fig. 2. Chromosomal location of QTLs for seedling anoxic tolerance and coleoptile length under normal germination conditions in the RIL population.
Fig. 3. Chromosomal location of QTLs for seedling anoxic tolerance and coleoptile length under normal germination conditions in the BIL population.
Rice Science, Vol. 17, No. 3, 2010
196
Table 2. QTLs identified for coleoptile length under normal germination conditions. Population and QTL
Marker interval a
Distance (cM) b
LOD
Variance explained (%)
Additive effect c
RIL population qCL-2-R RM525–RM2127 13.0 2.50 5.2 +0.14 BIL population qCL-1-B C1370–C122 2.0 4.12 10.3 +0.50 qCL-2-B C1221–G275 0.9 2.52 4.0 -0.30 qCL-8-B C1121–R902 0.0 3.18 8.1 -0.50 qCL-11-B C1172–S2260 1.2 6.39 9.7 -0.30 a Bold letters indicate the nearest marker. b Distance from the nearest marker to putative QTL. c ‘+’ means that positive alleles come from Xiushui 79 in the RIL population or from Nipponbare in the BIL population; ‘-’ means that positive alleles come from C Bao in the RIL population or from Kasalath in the BIL population.
Difference between two parents and variations in the two populations for coleoptile length of seedlings germinated under normal conditions In the RIL population, the average coleoptile length of C Bao (5.5±0.1 mm) was significantly higher than that of Xiushui 79 (4.9±0.1 mm) (t=8.40). The average coleoptile length of the RIL population was 5.1±0.6 mm with a range of 3.9–7.1 mm, which indicated an approximately normal distribution and transgressive segregation. The coefficient of variation of the RIL population was 12.5% for the trait (Fig. 4-A). In the BIL population, the average coleoptile length of Kasalath (11.0±1.0 mm) was significantly higher than that of Nipponbare (5.5±0.3 mm). The average coleoptile length of the BIL population was 6.4±1.3 mm with a range of 3.0–10.1 mm, which showed an approximately normal distribution and transgressive segregation. The coefficient of variation of the BIL population was 20.2% for the trait (Fig. 4-B).
Fig. 4. Frequency distributions of coleoptile length in Xiushui 79/C Bao RIL population (A) and Nipponbare/Kasalath//Nipponbare BIL population (B).
QTLs for seedling coleoptile length detected in the two populations under normal conditions In the RIL population, only one QTL (qCL-2-R) conditioning coleoptile length was detected. The QTL located at RM525–RM2127 interval on chromosome 2. qCL-2-R explained 5.2% of the phenotypic variance (Table 2 and Fig. 2) and its positive allele came from Xiushui 79. In the BIL population, four QTLs controlling coleoptile length were detected and located on chromosomes 1, 2, 8, and 11. These QTLs explained 4.0% to 10.3% of the phenotypic variances (Table 2 and Fig. 3). Positive allele of qCL-1-B was derived from Nipponbare.
DISCUSSION In both of the two populations, QTL controlling seedling anoxic tolerance was detected on chromosome 2. In the RIL population, the SSR marker tightly linked to qSAT-2-R was RM525, which is located at the position of 118.1 cM from the end of the short arm of chromosome 2 (McCouch et al, 2002). In the BIL population, the RFLP marker tightly linked to qSAT-2-B was C747, corresponding to SSR marker RM1367 which is located at the position of 110.9 cM from the end of the short arm of chromosome 2 (McCouch et al, 2002). Physical distance between qSAT-2-R and qSAT-2-B was 1230 kb according to the physical map (http://rgp. dna.affrc.go.jp/E/IRGSP/download.html). It suggests that qSAT-2-B and qSAT-2-R might be the same QTL controlling seedling anoxic tolerance. Under normal germination conditions, the difference of coleoptile length between Xiushui 79 and C Bao
WANG Yang, et al. QTL Analysis of Anoxic Tolerance at Seedling Stage in Rice
was only 1 mm and the coefficient of variation was 12.5% in the RIL population. However, the anoxic response index of seedling in the two parents was significantly different. The average coleoptile length of seedlings in the RIL population (247 lines) under the anoxic stress was nine times as that under the normal conditions. In addition, the coefficient of variation of seedling anoxic response index also increased. These results indicate that under the anoxic stress, the differences of the elongation ability of seedling coleoptile became larger among the population lines and the two parents. Furthermore, as the difference of seedling coleoptile length under the normal conditions was small among the population lines and between parents, the anoxic response index was more suitable to reflect the anoxic tolerance among lines. Therefore, QTLs controlling seedling anoxic tolerance detected in the RIL population derived from japonica-japonica cross were more reliable than those detected in the BIL population derived from indica-japonica cross. The RIL population had smaller difference in seedling coleoptile length among lines than the BIL population under the normal conditions, thus the variation of the anoxic response index directly reflected the variation of seedling coleoptile length under the anoxic conditions. In the BIL population, five QTLs for seedling anoxic tolerance did not overlap with QTLs for seedling coleoptile length germinated under the normal conditions except the locus qSAT-8-B. qSAT-8-B and qCL-8-B were both located between R1813 and C1121 on chromosome 8, with a genetic distance of 1.0 cM. Genes controlling coleoptile elongation under anoxic stress were different from those under normal conditions. Under anoxic stress, anaerobic carbohydrate catabolism takes the place of aerobic carbohydrate catabolism, and sugars appear to play a signaling role under anoxia, with several genes such as RAMY3D, PDC, ADH1 and ADH2 indirectly up-regulated by anoxiadriven sugar starvation (Gibbs et al, 2000). In the anoxic rice coleoptile, six among 34 cell expansion genes were up-regulated (Rasika et al, 2007), especially EXPA7 and EXPB12. Furthermore, the expressions of some genes encoding functional proteins altered under anoxia. For example, rice HSP20 (Os02g52150) showed a higher expression, while some dehydrogenase genes were down-regulated under anoxia. Therefore, genes
197
controlling coleoptile elongation under anoxic stress were different from those in air.
ACKNOWLEDGEMENT This work was supported by the Program of Introducing Talents of Discipline to Universities, China (Grant No. B08023).
REFERENCES Chen J. 2003. Evolution and development of rice planting pattern. J Shenyang Agric Univ, 34(5): 389–393. (in Chinese with English abstract) Fred T T, Chen C C, Garry N M. 1981. Morphological development of rice seedling in water at controlled oxygen levels. Agron J, 73: 556–560. Gai J Y. 2000. Methods of Experimental Statistics. Beijing: China Agriculture Press: 35–125. (in Chinese) Gibbs J, Morrell S, Valdez A, Setter E S, Greenway H. 2000. Regulation of alcoholic fermentation in coleoptiles rice cultivars differing in tolerances to anoxia. J Exp Bot, 51: 785–796. Guo Y, Chen B S, Hong D L. 2010. Construction of SSR linkage map and analysis of QTL for rolled leaf in japonica rice. Rice Sci, 17(1): 28–34. Hou M Y, Jiang L, Wang C M, Wan J M. 2004. Quantitative trait loci and epistatic analysis for seed anoxia germinability in rice (Oryza sativa). Chin J Rice Sci, 18(6): 483–488. (in Chinese with English abstract) Jing D D, Diao L P, Qian H F, Sheng S L, Lin T Z, Hu C M. 2008. The comparison between rice direct sowing and transplanting and the strategy of breeding varieties in rice. Acta Agric Jiangxi, 20(7): 17–20. (in Chinese with English abstract) McCouch S R, Cho Y G, Yano M. 1997. Report on QTL nomenclature. Rice Genet Newsl, 14: 11–13. McCouch S R, Teytelman L, Xu Y. 2002. Development and mapping of 2240 new SSR markers for rice (Oryza sativa L.). DNA Res, 9: 199–207. Rasika L K, Leonardo M, Elena L, Silvia G, Francesco L, Giacomo N, Ottavio B, Federico V, Amedeo A, Pierdomenico P. 2007. Transcript profiling of the anoxic rice coleoptile. Plant Physiol, 51: 785–796. Setter E S, Ella E S, Valdez A P. 1994. Relationship between coleoptile elongation and alcoholic fermentation in rice exposed to anoxia: II. Cultivar differences. Ann Bot, 74: 273– 279. Wang S C, Basten C J, Zeng Z B. 2006. Windows QTL Cartographer Version 2.5. [2006-09-15]. http://statgen.ncsu. edu/qtlcart/WinQTLCart.htm Wang Y, Wang Y Y, Hong D L. 2009. Genetic variation of seed vigor and tolerance to anoxia among rice (Oryza sativa L.) accessions in Tai lake region. J Nanjing Agric Univ, 32(3): 1–7.
Rice Science, Vol. 17, No. 3, 2010
198 (in Chinese with English abstract) Wu W G, Chen Y, Qian Y F, Wang X J, Wu Y M. 2006. The current
(Oryza sativa L.) germplasm suitable for direct sowing under flooded soil surface. Euphytica, 67: 177–184.
status and progress of the research on direct seeding rice. Rev
Yamauchi M, Biswas J K. 1997. Rice cultivar difference in
China Agric Sci & Technol, 8(4): 32–36. (in Chinese with English
seedling establishment in flooded soil. Plant & Soil, 189: 145–
abstract)
153.
Yamauchi M, Aguilar A M, Vaughan D A, Seshu D V. 1993. Rice
QTL Analysis of Anoxic Tolerance at Seedling Stage in Rice WANG Yang1, 2 , GUO Yuan1, HONG De-lin1 (1State Key Laboratory of Crop Genetics and Germplasm Enhancement, Nanjing Agricultural University, Nanjing 210095, China; 2 Department of Agricultural Resource and Environment, Heilongjiang University, Harbin 150008, China)
Abstract：Coleoptile lengths of 7-day-old seedlings under anoxic stress and normal conditions were investigated in two permanently segregated populations and their parents in rice (Oryza sativa L.). Using anoxic response index, a ratio of coleoptile length under anoxic stress to coleoptile length under normal conditions, as an indicator of seedling anoxic tolerance (SAT), QTLs for SAT were detected. Two loci controlling SAT, designated as qSAT-2-R and qSAT-7-R, were detected in a recombinant inbred line (RIL) population (247 lines) derived from a cross between Xiushui 79 (japonica variety) and C Bao (japonica restorer line). qSAT-2-R, explaining 8.7% of the phenotype variation, was tightly linked with the SSR marker RM525. qSAT-7-R, explaining 9.8% of the phenotype variation, was tightly linked with the marker RM418. The positive alleles of the two loci came from C Bao. Six loci controlling SAT, designated as qSAT-2-B, qSAT-3-B, qSAT-5-B, qSAT-8-B, qSAT-9-B and qSAT-12-B, were detected in a backcross inbred line (BIL) population (98 lines) derived from a backcross of Nipponbare (japonica)/Kasalath (indica)//Nipponbare (japonica). The positive alleles of qSAT-2-B, qSAT-3-B and qSAT-9-B, which explained 16.2%, 11.4% and 9.5% of the phenotype variation, respectively, came from Nipponbare. Besides, the positive alleles of qSAT-5-B, qSAT-8-B and qSAT-12-B, which explained 7.3%, 5.8% and 14.0% of the phenotype variation, respectively, were from Kasalath. Key words: rice (Oryza sativa); seedling; anoxic tolerance; quantitative trait locus; recombinant inbred line; backcross inbred line
Direct seeding in rice (Oryza sativa L.) is a labor-saving and highly efficient cultivation pattern (Chen, 2003), and is now becoming more and more popular. Compared with the traditional transplanting cultivation pattern, it can increase number of panicles per unit area and 1000-grain weight of rice (Jing et al, 2008). In the rice production through direct seeding, one of the key factors affecting grain yield is the seedling establishment in paddy fields (Wu et al, 2006). In flooded or water-saturated soil, seedling rates are lowered due to anoxic stress (Fred et al, 1981). Breeding rice varieties with anoxic tolerance at the seedling stage would be helpful in increasing seedling rate in paddy fields of direct seeding. Rice seed can germinate under anoxic conditions, but root growth and leaf emergence are hindered. To get enough O2 for the growth of root and leaf, the coleoptile elongates faster and longer under anoxia than normal conditions (Yamauchi et al, 1993). Coleoptiles of rice varieties with anoxia tolerance had stronger elongation ability than those without anoxic tolerance Received: 1 December 2009; Accepted: 12 March 2010 Corresponding author: HONG De-lin ([email protected])
(Setter et al, 1994; Yamauchi and Biswas, 1997), and coleoptiles of japonica rice elongated faster and longer than those of indica rice under anoxia stress (Hou et al, 2004). We also found genetic variations in anoxic tolerance of seedlings among japonica rice varieties (Wang et al, 2009). These genetic variations can be utilized to develop varieties with anoxic resistance in rice breeding. Using the shoot length of seedlings germinated in water at a depth of 20 cm for 5 days in the dark as an indicator of anoxic germination (AG), Hou et al (2004) detected five QTLs located on chromosomes 1, 2, 5, 5 and 7 in a population of recombinant inbred lines (RIL). According to the direction of the additive effects, the positive alleles at the loci of qAG-1, qAG-G2 and qAG-7 derived from Kinmaze, while at the loci of qAG-5a and qAG-5b, the DV85 alleles increased AG. However, the RIL population mentioned above is derived from a cross of Kinmaze (japonica)/DV85 (indica), and is not a breeding population. Studies on QTL mapping for seedling anoxic tolerance by using a breeding population of japonica rice have not been reported yet. To discover more genetic information for
WANG Yang, et al. QTL Analysis of Anoxic Tolerance at Seedling Stage in Rice
breeding rice varieties suitable for direct seeding, we conducted QTL mapping for seedling anoxic tolerance using a population of 247 RILs from the cross of Xiushui 79 (japonica variety) and C Bao (japonica restorer line), and a population of 98 lines derived from a backcross of Nipponbare (japonica)/Kasalath (indica)//Nipponbare.
MATERIALS AND METHODS Rice materials A population of 247 recombinant inbred lines (RILs) from a cross of Xiushui 79 (japonica variety) and C Bao (japonica restorer line) and their parents were used. In 2008, the population of F10:11 generation was obtained. Another population of 98 backcross inbred lines (BILs) from a backcross of Nipponbare (japonica)/Kasalath (indica)//Nipponbare and their parents were also used. The BIL population of BC1F13 was obtained in 2007. Both populations were developed by the single-seed descent method. Field planting In 2007, seeds of the BIL population and their parents were sown in a rice seedling bed in May, and the seedlings were transplanted in June at Jiangpu Experimental Station, Nanjing Agricultural University, China. In 2008, the experiment was conducted in the same way for the seeds of the RIL population and their parents. Each material was transplanted to one plot in a paddy field. Twelve hills per plot were transplanted at a density of 16.7 cm×20.0 cm, with one seedling per hill. Field managements followed traditional practice. Three thousand seeds were harvested from middle five hills of each plot at 45–50 days after heading (harvesting time was determined according to the panicle size). Seed dormancy was broken by keeping the seeds at 50ºC in a drying oven for five days. Measurement of coleoptile length of seedlings germinated under anoxic conditions Two hundred seeds of each material were surface sterilized by immersion in 0.6% sodium hypochlorite solution for 15 min and rinsed with tap water for three
193
times. Then the seeds were soaked in water for 48 h at 25ºC. After that, 20 seeds per line which absorbed enough water were put into a plastic culture tray (modified from seedling trays) with 150 compartments (1.9 cm × 1.9 cm × 3.8 cm) and a drain hole (0.6 cm diameter) under each compartment. Four seeds were placed at the bottom of each compartment and then germinated in the depth of 5 cm water. The plastic culture tray was put on the surface of vermiculite with a depth of 5 cm water in a plastic transfer box (37 cm × 25 cm × 9 cm). The distance between seeds and water surface was 5 cm. The length of the coleoptile of 7-day-old seedlings was measured from the base of coleoptile to the top and accurate to 1 mm. The mean of the coleoptile length of 10 seedlings was calculated in each line. The germination experiment was conducted in a GXZ-type incubator under the photoperiod of 16 h darkness at 20°C and 8 h light at 30°C with two replications. Measurement of coleoptile length of seedlings germinated under normal conditions Twenty seeds per line which absorbed enough water were put on the surface of two layers of filter paper covering the slope board at 70-degree tilt position in a germination box. Water in the germination box was kept 3 cm high in order to make paper moist without soaking seeds. The other germination conditions and the measurement of coleoptile length of 7-day-old seedlings were the same as those under the anoxic conditions. Calculation of seedling anoxic response index The ratio of coleoptile length under anoxic conditions over that under normal conditions was calculated as the anoxic response index of seedlings, and used as an indicator of seedling anoxic tolerance. Data analysis A genetic linkage map of SSR markers for the RIL population has been constructed by our laboratory (Guo et al, 2010). The linkage map spans 744.6 cM of rice genome with 74 SSR markers from all 17 individual linkage groups. The molecular data of the BIL population were obtained from Dr. Yano at the National Institute of
Rice Science, Vol. 17, No. 3, 2010
194
Agrobiological Resources, Japan. The 245 RFLP markers are distributed evenly on 12 chromosomes, covering 1179.9 cM of rice genome with an average distance of 4.8 cM between two markers. The mean of the coleoptile length of each line over two replications was used as a unit for the following statistical analysis. QTL detection was carried out by the Composite Interval Mapping (CIM) method in the Win QTL Cartographer 2.5 software (Wang et al, 2006). The LOD scores were calculated every 2 cM on 12 chromosomes, with the threshold of 2.5. A putative QTL was determined between the markers when the LOD score was larger than the threshold. QTL nomenclature followed the propositions of the Rice Genetics Cooperative (McCouch et al, 1997). In order to distinguish QTLs detected in two different populations, ‘R’ was added to the name of QTLs detected in the RIL population, and ‘B’ was added to the name of QTLs detected in the BIL population. The statistical analyses were executed with the Excel software according to the Method of Experimental Statistics (Gai, 2000).
RESULTS Difference between two parents and variations in two populations for seedling anoxic response index In the RIL population, the mean of the seedling anoxic response index of C Bao (9.30±0.20) was significantly higher than that of Xiushui 79 (7.93±0.04) (t=9.63). The seedling anoxic response index of the
Fig. 1. Frequency distributions of anoxic response index in rice seedlings of Xiushui 79/C Bao RIL population (A) and Nipponbare/ Kasalath//Nipponbare BIL population (B). Anoxic response index is the ratio of coleoptile length of seedlings with seed germinated in water at a depth of 5 cm to coleoptile length under normal germination conditions.
RIL population averaged 9.13±1.89 with a range of 4.21–14.41, which showed approximately normal distribution and transgressive segregation with a coefficient of variation of 20.7% (Fig. 1-A). In the BIL population, the mean of the seedling anoxic response index of Nipponbare (4.58±0.45) was significantly higher than that of Kasalath (2.33±0.07) (t=7.00). The seedling anoxic response index of the BIL population averaged 3.89±0.74 with a range of 2.36–5.56, which showed approximately normal distribution and transgressive segregation with a coefficient of variation of 19.1% (Fig. 1-B). QTLs for seedling anoxic tolerance detected in the two populations In the RIL population, two QTLs conditioning seedling anoxic tolerance were detected on chromosomes 2 and 7 (Table 1 and Fig. 2). qSAT-2-R explained 8.7%
Table 1. QTLs identified for seedling anoxic tolerance (SAT) in the RIL and BIL populations. Population and QTL
Marker interval a
Distance (cM) b
LOD
Variance explained (%)
Additive effect c
RIL population qSAT-2-R RM525–RM2127 0.1 3.27 8.7 -0.77 qSAT-7-R RM418–RM11 2.0 3.36 9.8 -0.87 BIL population qSAT-2-B R3393–C747 1.0 5.34 16.2 +0.30 qSAT-3-B C1488–C63 2.0 3.82 11.4 +0.30 qSAT-5-B R830–R3166 0.0 2.60 7.3 -0.15 qSAT-8-B R1813–C1121 0.3 2.82 5.8 -0.20 qSAT-9-B R2272–R2638 0.0 3.63 9.5 +0.30 qSAT-12-B C1336–R642 3.5 2.99 14.0 -0.30 a Bold letters indicate the nearest marker. b Distance from the nearest marker to putative QTL. c ‘+’ means that positive alleles come from Xiushui 79 in the RIL population or from Nipponbare in the BIL population; ‘-’ means that positive alleles come from C Bao in the RIL population or from Kasalath in the BIL population.
WANG Yang, et al. QTL Analysis of Anoxic Tolerance at Seedling Stage in Rice
of the phenotypic variance and its positive allele came from C Bao. The amplification product of SSR marker RM525, closely linked to qSAT-2-R, showed a band of 140 bp using the total DNA of C Bao as template. qSAT-7-R explained 9.8% of the phenotypic variance and its positive allele came from C Bao. The amplification product of SSR marker RM418, closely linked to qSAT-7-R, showed a band of 250 bp using the total DNA of C Bao as template. In the BIL population, six QTLs governing seedling anoxic tolerance were detected on chromosomes 2, 3, 5, 8, 9 and 12 (Table 1 and Fig. 3). These QTLs explained 5.8% to 16.2% of the phenotypic variances. Nipponbare carried positive alleles on the loci of qSAT-2-B, qSAT-3-B and qSAT-9-B, and Kasalath possessed positive alleles on the loci of qSAT-5-B, qSAT-8-B and qSAT-12-B.
195
Fig. 2. Chromosomal location of QTLs for seedling anoxic tolerance and coleoptile length under normal germination conditions in the RIL population.
Fig. 3. Chromosomal location of QTLs for seedling anoxic tolerance and coleoptile length under normal germination conditions in the BIL population.
Rice Science, Vol. 17, No. 3, 2010
196
Table 2. QTLs identified for coleoptile length under normal germination conditions. Population and QTL
Marker interval a
Distance (cM) b
LOD
Variance explained (%)
Additive effect c
RIL population qCL-2-R RM525–RM2127 13.0 2.50 5.2 +0.14 BIL population qCL-1-B C1370–C122 2.0 4.12 10.3 +0.50 qCL-2-B C1221–G275 0.9 2.52 4.0 -0.30 qCL-8-B C1121–R902 0.0 3.18 8.1 -0.50 qCL-11-B C1172–S2260 1.2 6.39 9.7 -0.30 a Bold letters indicate the nearest marker. b Distance from the nearest marker to putative QTL. c ‘+’ means that positive alleles come from Xiushui 79 in the RIL population or from Nipponbare in the BIL population; ‘-’ means that positive alleles come from C Bao in the RIL population or from Kasalath in the BIL population.
Difference between two parents and variations in the two populations for coleoptile length of seedlings germinated under normal conditions In the RIL population, the average coleoptile length of C Bao (5.5±0.1 mm) was significantly higher than that of Xiushui 79 (4.9±0.1 mm) (t=8.40). The average coleoptile length of the RIL population was 5.1±0.6 mm with a range of 3.9–7.1 mm, which indicated an approximately normal distribution and transgressive segregation. The coefficient of variation of the RIL population was 12.5% for the trait (Fig. 4-A). In the BIL population, the average coleoptile length of Kasalath (11.0±1.0 mm) was significantly higher than that of Nipponbare (5.5±0.3 mm). The average coleoptile length of the BIL population was 6.4±1.3 mm with a range of 3.0–10.1 mm, which showed an approximately normal distribution and transgressive segregation. The coefficient of variation of the BIL population was 20.2% for the trait (Fig. 4-B).
Fig. 4. Frequency distributions of coleoptile length in Xiushui 79/C Bao RIL population (A) and Nipponbare/Kasalath//Nipponbare BIL population (B).
QTLs for seedling coleoptile length detected in the two populations under normal conditions In the RIL population, only one QTL (qCL-2-R) conditioning coleoptile length was detected. The QTL located at RM525–RM2127 interval on chromosome 2. qCL-2-R explained 5.2% of the phenotypic variance (Table 2 and Fig. 2) and its positive allele came from Xiushui 79. In the BIL population, four QTLs controlling coleoptile length were detected and located on chromosomes 1, 2, 8, and 11. These QTLs explained 4.0% to 10.3% of the phenotypic variances (Table 2 and Fig. 3). Positive allele of qCL-1-B was derived from Nipponbare.
DISCUSSION In both of the two populations, QTL controlling seedling anoxic tolerance was detected on chromosome 2. In the RIL population, the SSR marker tightly linked to qSAT-2-R was RM525, which is located at the position of 118.1 cM from the end of the short arm of chromosome 2 (McCouch et al, 2002). In the BIL population, the RFLP marker tightly linked to qSAT-2-B was C747, corresponding to SSR marker RM1367 which is located at the position of 110.9 cM from the end of the short arm of chromosome 2 (McCouch et al, 2002). Physical distance between qSAT-2-R and qSAT-2-B was 1230 kb according to the physical map (http://rgp. dna.affrc.go.jp/E/IRGSP/download.html). It suggests that qSAT-2-B and qSAT-2-R might be the same QTL controlling seedling anoxic tolerance. Under normal germination conditions, the difference of coleoptile length between Xiushui 79 and C Bao
WANG Yang, et al. QTL Analysis of Anoxic Tolerance at Seedling Stage in Rice
was only 1 mm and the coefficient of variation was 12.5% in the RIL population. However, the anoxic response index of seedling in the two parents was significantly different. The average coleoptile length of seedlings in the RIL population (247 lines) under the anoxic stress was nine times as that under the normal conditions. In addition, the coefficient of variation of seedling anoxic response index also increased. These results indicate that under the anoxic stress, the differences of the elongation ability of seedling coleoptile became larger among the population lines and the two parents. Furthermore, as the difference of seedling coleoptile length under the normal conditions was small among the population lines and between parents, the anoxic response index was more suitable to reflect the anoxic tolerance among lines. Therefore, QTLs controlling seedling anoxic tolerance detected in the RIL population derived from japonica-japonica cross were more reliable than those detected in the BIL population derived from indica-japonica cross. The RIL population had smaller difference in seedling coleoptile length among lines than the BIL population under the normal conditions, thus the variation of the anoxic response index directly reflected the variation of seedling coleoptile length under the anoxic conditions. In the BIL population, five QTLs for seedling anoxic tolerance did not overlap with QTLs for seedling coleoptile length germinated under the normal conditions except the locus qSAT-8-B. qSAT-8-B and qCL-8-B were both located between R1813 and C1121 on chromosome 8, with a genetic distance of 1.0 cM. Genes controlling coleoptile elongation under anoxic stress were different from those under normal conditions. Under anoxic stress, anaerobic carbohydrate catabolism takes the place of aerobic carbohydrate catabolism, and sugars appear to play a signaling role under anoxia, with several genes such as RAMY3D, PDC, ADH1 and ADH2 indirectly up-regulated by anoxiadriven sugar starvation (Gibbs et al, 2000). In the anoxic rice coleoptile, six among 34 cell expansion genes were up-regulated (Rasika et al, 2007), especially EXPA7 and EXPB12. Furthermore, the expressions of some genes encoding functional proteins altered under anoxia. For example, rice HSP20 (Os02g52150) showed a higher expression, while some dehydrogenase genes were down-regulated under anoxia. Therefore, genes
197
controlling coleoptile elongation under anoxic stress were different from those in air.
ACKNOWLEDGEMENT This work was supported by the Program of Introducing Talents of Discipline to Universities, China (Grant No. B08023).
REFERENCES Chen J. 2003. Evolution and development of rice planting pattern. J Shenyang Agric Univ, 34(5): 389–393. (in Chinese with English abstract) Fred T T, Chen C C, Garry N M. 1981. Morphological development of rice seedling in water at controlled oxygen levels. Agron J, 73: 556–560. Gai J Y. 2000. Methods of Experimental Statistics. Beijing: China Agriculture Press: 35–125. (in Chinese) Gibbs J, Morrell S, Valdez A, Setter E S, Greenway H. 2000. Regulation of alcoholic fermentation in coleoptiles rice cultivars differing in tolerances to anoxia. J Exp Bot, 51: 785–796. Guo Y, Chen B S, Hong D L. 2010. Construction of SSR linkage map and analysis of QTL for rolled leaf in japonica rice. Rice Sci, 17(1): 28–34. Hou M Y, Jiang L, Wang C M, Wan J M. 2004. Quantitative trait loci and epistatic analysis for seed anoxia germinability in rice (Oryza sativa). Chin J Rice Sci, 18(6): 483–488. (in Chinese with English abstract) Jing D D, Diao L P, Qian H F, Sheng S L, Lin T Z, Hu C M. 2008. The comparison between rice direct sowing and transplanting and the strategy of breeding varieties in rice. Acta Agric Jiangxi, 20(7): 17–20. (in Chinese with English abstract) McCouch S R, Cho Y G, Yano M. 1997. Report on QTL nomenclature. Rice Genet Newsl, 14: 11–13. McCouch S R, Teytelman L, Xu Y. 2002. Development and mapping of 2240 new SSR markers for rice (Oryza sativa L.). DNA Res, 9: 199–207. Rasika L K, Leonardo M, Elena L, Silvia G, Francesco L, Giacomo N, Ottavio B, Federico V, Amedeo A, Pierdomenico P. 2007. Transcript profiling of the anoxic rice coleoptile. Plant Physiol, 51: 785–796. Setter E S, Ella E S, Valdez A P. 1994. Relationship between coleoptile elongation and alcoholic fermentation in rice exposed to anoxia: II. Cultivar differences. Ann Bot, 74: 273– 279. Wang S C, Basten C J, Zeng Z B. 2006. Windows QTL Cartographer Version 2.5. [2006-09-15]. http://statgen.ncsu. edu/qtlcart/WinQTLCart.htm Wang Y, Wang Y Y, Hong D L. 2009. Genetic variation of seed vigor and tolerance to anoxia among rice (Oryza sativa L.) accessions in Tai lake region. J Nanjing Agric Univ, 32(3): 1–7.
Rice Science, Vol. 17, No. 3, 2010
198 (in Chinese with English abstract) Wu W G, Chen Y, Qian Y F, Wang X J, Wu Y M. 2006. The current
(Oryza sativa L.) germplasm suitable for direct sowing under flooded soil surface. Euphytica, 67: 177–184.
status and progress of the research on direct seeding rice. Rev
Yamauchi M, Biswas J K. 1997. Rice cultivar difference in
China Agric Sci & Technol, 8(4): 32–36. (in Chinese with English
seedling establishment in flooded soil. Plant & Soil, 189: 145–
abstract)
153.
Yamauchi M, Aguilar A M, Vaughan D A, Seshu D V. 1993. Rice