Rice Science, 2014, 21(1): 20?28 Copyright ? 2014, China National Rice Research Institute Published by Elsevier BV. All rights reserved DOI: 10.1016/S1672-6308(13)60165-1
Seed Length Controlled by Same Locus in Four Different AA Genome Species of Genus Oryza
1, #1, #1111, 21ZHANG Yu, LI Jing, ZHOU Jia-wu, XU Peng, DENG Xian-neng, YANG Fei, DENG Wei,
11HU Feng-yi, TAO Da-yun 12(Institute of Food and Crops Research, Yunnan Academy of Agricultural Sciences, Kunming 650205, China; Life-Sciences School, #Yunnan University, Kunming 650091, China; These authors contributed equally to this study)
Abstract: To broaden the genetic basis and overcome the yield plateau in Asian cultivated rice, the
exploitation and utilization of favorable alleles from rice species with the AA genome has become
important and urgent in modern breeding programs. Four different interspecific populations were used to
detect quantitative trait locus (QTL) for seed length, including a BCF population derived from Oryza 42
glumaepatula crossed with Dianjingyou 1 (a japonica cultivar), a BCF population derived from O. nivara 42
crossed with Dianjingyou 1, a BCF population derived from a cross between O. longistaminata and 71
RD23 (an indica cultivar), and a BCFpopulation derived from a cross between O. glaberrima and 81
Dianjingyou 1. The QTLs for seed length in four different populations were termed as SL-3a, SL-3b,
SL-3c and SL-3d, respectively. They had good collinearity and accounted for 49% to 60% of the
phenotypic variations. Sequencing data indicated that four QTLs were different alleles of GS3 which were
responsible for the seed length variation between O. sativa and its four AA genome relatives. These
results will be valuable for confirming the evolution of GS3 and also be helpful for rice breeding.
Key words: seed length; O. glumaepatula; O. nivara; O. longistaminata; O. glaberrima; quantitative trait locus
Rice is one of the most important food crops for half indicates that 40% of alleles of wild rice is lost during of the world’s population. In the past half century, two the domestication from wild rice to cultivated rice major breakthroughs in rice breeding were made (Sun et al, 2002), and only 10% to 20% of genetic through the use of semidwarf gene and heterosis diversities in wild species are retained in two subspecies (Patnaik et al, 1990; Spielmeyer et al, 2002). But in of cultivated rice (Zhu et al, 2007). Thus, the exploitation recent years, rice yields have not significantly increased, and utilization of favorable alleles of wild rice and mainly due to the narrow genetic basis of parental related species may overcome yield bottleneck of the materials (Tanksley and McCouch, 1997). The genus, Asian cultivated rice (Xiao et al, 1998).
Seed length is one of the major determination factors Oryza, is comprised of two cultivated rice species and
22 wild rice species, and all species of the genus Oryza of rice yield (Xing and Zhang, 2010). Seed length not are classified into six diploid genome types (AA, BB, only determines seed appearance, but also affects CC, EE, FF and GG) and four tetraploid genome types milling, cooking and eating qualities of rice (Fan et al,
(BBCC, CCDD, HHJJ and HHKK) (Vaughan, 1994; 2006), moreover, seed length is also important in the Khush, 1997). The species with AA genome are evolution of cereal crops because long grains tend to classified into two cultivated species (O. sativa and O. be selected during the early domestication process, as glaberrima), and six wild rice species (O. nivara, O. evidenced by the fact that most cultivated species have rufipogon, O. barthii, O. glumaepatula, O. longistaminata longer grains than their wild relatives (Li et al, 2004). and O. meridionalis). Since they share the same AA Elucidating the molecular and genetic basis of seed genome, O. glaberrima and the six wild rice species are length will lay the foundation for improving rice yield the most accessible genetic resources for the improvement and quality, and contribute to the dissection of evolution of cultivated rice (Ren et al, 2003). Previous reports mechanism underlying grain size in the genus Oryza.
Quantitative trait locus (QTL) mapping has been
proven to be an effective approach to reveal the
genetic basis of phenotypic evolution and domestication Received: 17 May 2013; Accepted: 13 October 2013
(Tanksley and McCouch, 1997; Barton and Keightley, Corresponding author: TAO Da-yun (email@example.com)
21 ZHANG Yu, et al. Seed Length Controlled by Same Locus in Four AA Genome Species of Oryza
Four populations were used to identify the QTL for 2002). Seed size is not only a typical quantitative trait,
seed length. The first population was raised as follows: but also complex in its genetic basis (Gupta et al,
2006). Many QTLs associated with grain or fruit sizes O. glumaepatula as the donor parent, was crossed with were identified in the last decade and some of them was successively Dianjingyou 1 as the recurrent parent. F1
have been isolated by QTL-mapping. Grain size 3 backcrossed with Dianjingyou 1 and then self-fertilized. (GS3), which is a major QTL controlling seed length Meanwhile, phenotypic selection (short grain) was and weight in the modern subpopulations of O. sativa, done until producing the BCF as an introgression 37
encodes a putative transmembrane protein with a line (IL). BCF was produced through BCF IL crossing 4137
phosphatidyl-ethanolamine-binding protein (PEBP)- with Dianjingyou 1. The BCF individuals were used 41
like domain, a putative tumor necrosis factor receptor/ to raise the secondary population by self-pollination, nerve growth factor receptor (TNFR/NGFR) domain and then BCF individuals were raised for further 42
and a von Willebrand factor type C (VWFC) domain analysis. Similarly, other populations were developed (Fan et al, 2006, 2009; Takano-Kai et al, 2009; Mao et al, using the same strategy. Finally, a total of 300 BCF 42
2010). Grain weight 2 (GW2), a QTL for seed width individuals derived from the cross between O.
and weight in rice, encodes a RING-type E3 ubiquitin glumaepatula and Dianjingyou 1 (2009H3E669), 198 ligase (Matsuoka and Ashikari, 2007; Song et al, 2007). BCF individuals derived from the cross between O. 42
A QTL for seed width, qSW5, encodes an unknown nivara and Dianjingyou 1 (2009H3E705), 117 BCF 71
protein and regulates cell division during seed individuals derived from the cross between O.
development (Shomura et al, 2008; Weng et al, 2008). longistaminata and RD23 (2005H2E482), and 140
Recently, it was reported that GW8 was synonymous BCF individuals derived from the cross between O. 81with OsSPL16, which controlls grain size, shape and glaberrima and Dianjingyou 1 (2004H3E242) were quality by regulating cell proliferation (Wang et al, 2012). subjected to phenotypic and genotypic survey, respectively. And qGL3, which encodes a putative protein phosphatase All the plant materials were grown in the paddy field with Kelch-like repeat domain (OsPPKL1), controlls at the Winter Breeding Station located in Sanya, Hainan grain length and grain yield (Zhang et al, 2012). In Province, China. 2004H3E242, 2009H3E669 and addition, some genes controlling fruit size have been 2009H3E705 were grown in 2004, 2009 and 2009, identified in tomato. For example, fw2.2, the first gene respectively, in Winter Crop Season (November to known to control fruit size, encodes an unknown April), whereas 2005H2E482 was grown in 2005 in functional protein regulating cell cycle during fruit Late Crop Season (July to October). The plant density development (Frary et al, 2000). SUN, one of the and field management followed essentially the ordinary major genes controlling the elongation of fruit shape in agricultural practices.
tomato, encodes a member of the IQ67 domain-containing Phenotypic evaluation family (Xiao et al, 2008). These results have elucidated
the molecular mechanism of grain or fruit development. Seeds were sampled from primary panicle and stored In order to identify the QTLs for seed length in four at room temperature for at least three months before AA genome relatives, four different interspecific testing. A total of 20 unmilled seeds were used to populations were developed: an O. glumaepatula × measure seed length from each plant of the six parent Dianjingyou 1 (a japonica cultivar) cross, an O. nivara lines and the four populations. Seed length was × Dianjingyou 1 cross, an O. longistaminata × RD23 evaluated after removal of the awn. Seed photographs (an indica cultivar) cross and an O. glaberrima × per line were taken using a stereomicroscope (Nikon Dianjingyou 1 cross. QTL detection, collinearity analysis SMZ1000, Japan), and then seed length was measured and gene sequencing were performed to confirm as the distance from the opposing tips of a seed by QTLs for seed length in four AA genome relatives, software Image J. The average length of 20 seeds was which will be benefit for better understanding the used as the phenotypic value. domestication of GS3 and the application of this gene
DNA extraction and PCR for rice breeding program.
DNA extraction was performed as previously MATERIALS AND METHODS described (Edwards et al, 1991). The required markers
for identifying seed length locus were achieved by Rice materials using previously published simple sequence repeat
22 Rice Science, Vol. 21, No. 1, 2014
(SSR) markers in rice (McCouch et al, 2002). For to 8.4 mm as the boundary, respectively (Fig. 1). In mapping, PCR was performed as follows: a total the 2009H3E669 population, peak values were found volume of 10 μL containing 10 ng of template DNA, 1 × at 8.5 and 9.5 mm of seed length, respectively. Similarly, buffer, 0.5 μmol/L of each primer, 50 μmol/L of peak values were observed at 7.9 and 8.9 mm in the dNTPs and 0.5 U of Taq polymerase. A total of 30 2009H3E705 population, 9.1 and 9.9 mm in the cycles were carried out, with an initial 4 min at 94 ?C, 2005H2E482 population, 8.1 and 8.7 mm in the followed by cycles of 30 s at 94 ?C, 30 s at 55 ?C and 2004H3E242 population. Thus, those results suggested 30 s at 72 ?C, and finally 5 min at 72 ?C. PCR products that seed length is controlled by a major QTL in the were separated on 8% non-denaturing polyacrylamide four populations, respectively.
gels and detected using the silver staining method. Molecular mapping of QTLs for seed length QTL analysis To map the QTL controlling seed length in the For QTL analysis, linkage maps of the four different 2009H3E669 population, 443 SSR markers on 12 rice populations were constructed using MAPMAKER 3.0 chromosomes were used for polymorphic analysis with the minimum LOD score of 3.0 (Lincoln and between the IL and the recurrent parent Dianjingyou 1, Lander, 1992). The QTL responsible for seed length and 15 polymorphic markers on chromosome 3 and 7 was identified by the interval mapping analysis using on chromosome 6 were selected for further analysis. the QTLCARTOGRAPHER software package (Basten Seed lengths of 300 individuals of this population were et al, 1998). The LOD threshold significance level was evaluated, and their genotypes were detected using 22 determined from 1 000 permutation tests, as implemented polymorphic markers. Linkage analysis indicated that by the QTL Cartographer (Churchill and Doerge, 1994). the QTL for seed length in this population was
associated with SSR marker RM15281 on chromosome Gene sequencing 3. Thus, this QTL for seed length, named SL-3a, was
For sequencing, PCR was carried out as follows: a mapped to the region between RMw315 and RM7395 total volume of 25 μL containing 100 ng of template on chromosome 3 (Fig. 2-A). In this population, the DNA, 1 × buffer, 0.5 μmol/L of each primer, 50 μmol/L average seed length of the homozygous O. glumaepatula
of dNTPs and 0.5 U of PrimerSTAR HS DNA class (8.2 mm) was significantly different from that of Polymerase (Takara, Dalian, China). A total of 30 the heterozygous (8.4 mm) (P < 0.01), and the average cycles were performed, with an initial 2 min at 98 ?C, of the heterozygous class was also significantly followed by cycles of 15 s at 98 ?C, 6 s at 55 ?C and 1 different from that of the Dianjingyou 1 homozygous min at 72 ?C, and finally 5 min at 72 ?C. Each fragment class (9.22 mm) (P < 0.0001), which indicated that the of the GS3 genomic sequence was amplified three O. glumaepatula allele decreased seed length under times independently and PCR products were sequenced the Dianjingyou 1 background, and seed length of the using both forward and reverse primers on an ABI heterozygous class was skewed toward the small-value 3730xl DNA Analyzer (ABI Company, USA). Sequences
were aligned using the MegAlign program (DNAStar Table 1. Primers used for GS3 sequencing. Lasergene 7.1). Heterozygous sites were identified by
Primer name Primer sequence (5′–3′) visual inspection of Sequencing Chromatograms for
double peaks. Genomic sequence of the GS3 gene GS3A-F GAAAAAAGGTGAAGGACGAGGGG
GS3A-R AGAAGAACGAAGATAAAAGGAGTGG included 540 bp upstream and 338 bp downstream of GS3B-F CTCTCATTGTAAGTTTTCCCCAGC this gene. A list of all primers for sequencing is shown GS3B-R GCTTATAGCTCGAAGACCTGTGAAG in Table 1. GS3C-F GCTACTATTTTAGACTTCACAGGTC
GS3C-R GTAAGAGTAAAGACGAGAAGAAATG GS3D-F CTGAAAAGTTGACAGGCTAAACAC RESULTS GS3D-R CTGAAGAACACAGGGAATAGGGAT
GS3E-F GAATCCCTATTCCCTGTGTTCTTC GS3E-R CGATGAACTGCTTGTAAACAGAGA Genetic analysis of seed length in the four GS3F-F TCTCTGTTTACAAGCAGTTCATCG interspecific populations GS3F-R AATGTGATTTCACCTGGTTCTTGT GS3G-F ATTACAAGAACCAGGTGAAATCAC In the four populations, all the seed lengths showed a GS3G-R AAACTAACAAATGTGGCAGACAAG pattern of continuous and bimodal distribution, with GS3H-F TTGTCTGCCACATTTGTTAGTTTG GS3H-R GCTCTAAGGAGTATGAACCAAAGG 9.0 to 9.2 mm, 8.4 to 8.6 mm, 9.4 to 9.6 mm and 8.2
23 ZHANG Yu, et al. Seed Length Controlled by Same Locus in Four AA Genome Species of Oryza
Fig. 1. Frequency distribution of seed length in the four interspecific cross populations.
F individuals derived from the cross between O. glaberrima and Dianjingyou 1; 2005H2E482, BCF individuals derived 2004H3E242, BC8171from the cross between O. longistaminata and RD23; 2009H3E669, BCF individuals derived from the cross between O. glumaepatula and 42Dianjingyou 1; 2009H3E705, BCF individuals derived from the cross between O. nivara and Dianjingyou 1. 42IL, Introgression line; DP, Donor parent; DJY1, Dianjingyou 1.
parent (Table 2). The implication of this analysis was 2009H3E705 population, a total of 443 SSR markers that SL-3a acted on seed length in negatively partial distributed throughout the whole genome were dominance mode. In addition, the additive effect of employed for the polymorphic analysis between the the O. glumaepatula allele was -0.49 mm in the ILs BCF and the recurrent parent Dianjingyou 1, and 37
2009H3E669 population and SL-3a explained 60% of among which 16 markers were polymorphic and the total phenotypic variation (Table 2). distributed on chromosomes 1, 2, 3, 5, 6, 10 and 12 To locate the QTL for seed length in the respectively. Then, seed lengths of 198 individuals in
the population were evaluated, and genotypes were
surveyed using the 16 polymorphic markers. QTL
analysis indicated that a QTL for seed length tightly
linked to RM15281, named SL-3b, was mapped to the
region between RMw315 and RM411 on chromosome
3 (Fig. 2-B). In this population, there were significant
differences in the average seed lengths among each of
the genotypic classes (P < 0.01). The homozygous O.
nivara class had an average seed length of 7.8 mm, the
heterozygous class of 7.9 mm and the Dianjingyou 1
homozygous class of 8.6 mm. The seed length of the
heterozygous class was also skewed toward the
small-value homozygous class (Table 2). Those results
suggested that the O. nivara allele also contributed to
the decrease of seed length in partial dominant fashion. Fig. 2. Co-linear analysis among SL-3a, SL-3b, SL-3c and SL-3d on The additive effect of the O. nivara allele was -0.53 chromosome 3 from the four different populations. mm in this population and SL-3b explained 54% of the Black squares indicate the QTL regions of SL-3a, SL-3b, SL-3c total phenotypic variation (Table 2). and SL-3d, respectively.
24 Rice Science, Vol. 21, No. 1, 2014
Table 2. Detection of QTLs for seed length in the four populations.
Seed length (mm) Population Population Additive a2bQTL Donor parent Flanking marker LOD R (%) name type effect (mm) DD DR RR
2009H3E669 SL-3a O. glumaepatula BCF RM7642–RM7395 8.2 ? 0.4 8.4 ? 0.4 9.2 ? 0.5 45.5 60 -0.49 422009H3E705 SL-3b O. nivara BCF RMw315–RM411 7.8 ? 0.4 7.9 ? 0.3 8.6 ? 0.4 26.2 54 -0.53 422005H2E482 SL-3c O. longistaminata BCF RM1164–RMw487 9.1 ? 0.3 9.7 ? 0.3 16.8 49 712004H3E242 SL-3d O. glaberrima BCF RMw315–RMw329 7.9 ? 0.3 8.6 ? 0.3 22.3 57 81
DD, Donor parent homozygote; DR, Donor/recurrent parent heterozygote; RR, Recurrent parent homozygote. a, O. glumaepatula (IRGC 100184) introduced from International Rice Research Institute (IRRI); O. nivara (IRGC 100195) introduced from IRRI; O. longistaminata (an unnamed accession) with long and strong rhizomes, originally collected from Niger and kindly provided by Hiroshi Hyakutake, Institute of Physical and Chemical Research, Saitama, Japan. Now preserved at Yunnan Academy of Agricultural Sciences (YAAS), bKunming, Yunnan, China; O. glaberrima (IRGC 101901) introduced from IRRI; , Phenotype variation explained by the QTL.
Similarly, 19 heterozygous markers were used to The QTLs identified from the four populations were investigate the genotype of the 2005H2E482 population. mapped to the same chromosomal region, where the A QTL, named SL-3c, resided in the region between known GS3 gene located. In order to identify if those RM1164 and RM487 (Fig. 2-C). The average of seed QTLs harbor GS3 alleles and look for the possible length of the heterozygous class was significantly different relationships between the GS3 sequence variation and from that of the homozygous class (P < 0.0001), seed length, the sequencing analysis of GS3 in the four
suggesting that the allele from O. longistaminata donor parents and two recurrent parents were performed. decreased seed length under the RD23 background A total of 125 single nucleotide polymorphisms (SNPs), and SL-3c explained 49% of the total phenotypic 17 InDels and 11 SSRs in the 6.2 kb of aligned variation (Table 2). Meanwhile, 22 heterozygous sequenced DNA were identified (Table 3). Of these,
the C-A mutation in the second exon of GS3 (SF28) markers were employed to investigate the genotype of the
2004H3E242 population, and a QTL for seed length, was discovered, and the A allele conferring the long named SL-3d, fell in the interval between RMw315 and grain was observed in Dianjingyou 1 and RD23, RMw329 (Fig. 2-D). There were significant differences whereas the C allele conferring the short grain was
detected in O. glumaepatula, O. nivara, O. longistaminata in the average seed lengths between the heterozygous
and the homozygous classes (P < 0.0001), indicating and O. glaberrima (Fig. 3), which was the same as the that the O. glaberrima allele also contributed to the results reported by Takano-Kai et al (2009). It decrease of seed length under the Dianjingyou 1 suggested that a common GS3 allele was identified
background and SL-3d explained 57% of the seed from four different donors, and the GS3 could be
length variation. Thus, SL-3a, SL-3b, SL-3c and SL-3d responsible for the seed length differences between O.
were all mapped to the region in the centromere sativa and its four AA genome relatives. In addition, region of rice chromosome 3, where the known GS3 we found one InDel polymorphism (named I1446) and gene located (Fan et al, 2006; Takano-Kai et al, 2009). eight SNPs polymorphism (named S316, S1206, S2555,
S3809, S3938, S4036, S4086 and S4102, respectively) Collinearity analysis of seed length QTLs within GS3 between the donor and recurrent parents In the four different populations, the QTLs for seed (Fig. 3). Thus, those novel polymorphic loci could be length were all located in the pericentromeric region potential factors generating the seed length differences of chromosome 3. SL-3a was tightly linked to SSR between O. sativa and its four AA genome relatives. markers RMw315 and RM15281, as did SL-3b (Fig.
2); both SL-3b and SL-3c resided in the region Table 3. Numbers of nucleotide polymorphism within GS3 based on between RMw315 and RMw330, and SL-3d was also sequencing of O. sativa and its four AA genome relatives. mapped to the region between RMw315 and RMw329 Region No. of SNPs No. of InDels No. of SSRs with the peak closest to RMw330. These results Coding region 23 3 0 indicated that the four QTLs existed collinearity (Fig. In frame 17 3 0 2) and suggested the occurrence of an orthologous Frame shift 0 0 0
Amino acid substitution 6 0 0 seed length QTL at this locus that might be associated Introns 87 10 11 with domestication in the genus Oryza. 5′-UTR 5 2 0
3′-UTR 10 2 0 The same locus GS3 is associated with seed length Total 125 17 11 variation in four interspecific populations SNP, Single nucleotide polymorphism; SSR, Simple sequence repeat.
25 ZHANG Yu, et al. Seed Length Controlled by Same Locus in Four AA Genome Species of Oryza
Table 4. Numbers of nucleotide polymorphism among O. sativa and its four AA genome relatives, respectively.
Region DJY1 RD23 O. glum O. nivara O. long O. glab
Coding region 0 0 4 3 19 2
In frame 0 0 1 1 15 2 Frame shift 0 0 0 0 0 0
Amino acid substitution 0 0 3 2 4 0
Intron and UTR 1 2 29 27 84 34
Total 1 2 33 30 103 36 Fig. 3. Functional variation of GS3 between O. sativa and its four DJY1, Dianjingyou 1; O. glum, O. glumaepatula; O. long, O. AA genome relatives. longistaminata; O. glab, O. glaberrima. Red columns represent the alleles of O. sativa, yellow squares
represent the variants of the corresponding alleles. Green columns show the heterozygous sites and the former nucleotide exhibits the Grain size plays an important role in the domestication
higher peak than the latter one by visual inspection of chromatograms. of cereal crops. The grains of wild relatives are Position indicates that the nucleotide sequences of the GS3 gene are usually small and round in shape for high fecundity compared with those of Dianjingyou 1. and easy dispersal by vectors. Rice domestication is
accompanied by increased size variation (Morishima Interestingly, S316 in the first intron was heterozygous et al, 1992). QTLs responsible for grain size variation in O. glumaepatula, and heterozygous sites were also were detected in the same chromosome region where observed at S1206, S2555, S3809 and S3938 in O. GS3 located in some subspecies crosses and in nivara, S4036 and S4086 in the last intron of GS3 in interspecies crosses between O. sativa, O. glaberrima O. longistaminata (Fig. 3). This might be related to and O. rufipogon (Li et al, 1997; Redoña and Mackill, the higher natural cross-pollination in wild rice 1998; Xiao et al, 1998; Kubo et al, 2001; Moncada et al, (Ghesquiere, 1985; Caicedo et al, 2007). 2001; Xing et al, 2002; Thomson et al, 2003; Aluko et al,
2004; Li et al, 2004), but has not been reported from O. Nucleotide polymorphism in the GS3 gene of O.
glumaepatula, O. nivara and O. longistaminata. sativa and its related species
In this study, four major QTLs for seed length were A total of 153 polymorphic loci were found in the 6.2 mapped to the pericentromeric region on chromosome kb of aligned sequenced DNA. Of those changes, only 3 in four different interspecific populations, including three polymorphic loci (0.02% of the total an O. glumaepatula × Dianjingyou 1 cross, an O. polymorphism) were observed in O. sativa, but 33 in nivara × Dianjingyou 1 cross, an O. longistaminata × O. glumaepatula (21.43%), 30 in O. nivara (19.48%), RD23 cross and an O. glaberrima × Dianjingyou 1 103 in O. longistaminata (66.88%) and 36 in O. cross. Compared with the recurrent parents exhibiting glaberrima (23.38%) (Table 4). It is suggested that the long seeds, the four donor parents all confer short GS3 gene in the related relatives of Asian cultivated seeds. The total phenotypic variations were explained rice had a higher level of nucleotide polymorphism from 49% to 60% (Table 2). Furthermore, the sequencing than that in O. sativa. In addition, amino acid analysis identified allelic variants of GS3 that may be substitutions were found in the wild rice. Of those responsible for the seed length differences between O. substitutions, one amino acid substitution (R53G) was sativa and its four AA genome relatives (Fig. 3). found in the OSR (organ size regulation) domain in O. These results contributed to understanding grain size glumaepatula, and a serine was substituted for tyrosine evolution in the genus Oryza. at the amino acid position 145 in the TNFR/NGFR
‘Wild’ alleles are functionally the same domain in O. nivara (data not shown). No amino acid
substitution was detected in the TM (transmembrane Asian cultivated rice (O. sativa) is thought to have domain) and VWFC domain (data not shown), indicating been domesticated from divergent populations of wild that both domains were conserved in O. sativa and its rice. During domestication, rice has undergone four related species. significant phenotypic changes in seed size, color,
shattering, seed dormancy and tillering (Kovach et al, DISCUSSION 2007). In this study, the two donor parents (O.
glumaepatula and O. nivara) all confer short seed A common locus is associated with seed length in length, and the estimated additive effects in the four different AA genome species populations 2009H3E669 and 2009H3E705 were
26 Rice Science, Vol. 21, No. 1, 2014
-0.49 and -0.53, respectively (Table 2). In addition, the introns of GS3 were discovered between O. sativa and
total phenotypic variations explained by the GS3 its four AA genome relatives (Fig. 3), suggesting that alleles were quite similar (54% and 57%, respectively) those nine polymorphic loci are also under artificial (Table 2). Furthermore, the sequencing data indicated selection. Secondly, a total of 135 nucleotide polymorphic that the A allele conferring long seed length was loci were found in the wild rice, including O. nivara,
observed in Dianjingyou 1, whereas the C allele O. glumaepatula and O. longistaminata, suggesting
conferring short seed length was detected in O. variations of these loci should occur earlier than the glumaepatula, O. nivara and O. glaberrima. Similarly, C-A mutation. Thirdly, although two polymorphism the accession of O. longistaminata with C allele also loci of GS3 in the third and the last exons, except for
C-A mutation, were found between the Asian conferred short grain in RD23 background (Fig. 3).
Small seed conferred by the ‘wild’ alleles of GS3 is cultivated rice and the African cultivated rice, there often favored under nature selection. It suggested that was no difference in the GS3 protein sequence (Table the ‘wild’ alleles were functionally the same in seed 4), indicating that GS3 protein was conserved in two length. Previous studies have reported that ‘wild’ cultivated rice species. Interestingly, amino acid alleles in wild rice relatives produce the same original substitutions within GS3 were found in wild rice. One phenotype, compared with ‘domesticated’ alleles in amino acid substitution (R53G) was found in the OSR cultivated rice, such as sh4 and PROG1. A nucleotide domain in O. glumaepatula, and a serine was
substitution of G for T within sh4 was responsible for substituted for tyrosine at the amino acid position 145 the reduction of grain shattering. The G alleles in the TNFR domain in O. nivara (data not shown),
conferring grain non-shattering were detected in O. which were not observed in cultivated rice. It suggested sativa, while the T alleles exhibiting grain shattering that two loci might be under selection in rice. The were observed in wild rice (Li et al, 2006). The plant-specific domain of OSR is both necessary and replacement of A by T within PROG1 is responsible sufficient for GS3 to function as a negative regulator for compact and fewer tillers. Wild rice with the T of grain length, and loss of function of this domain allele shows a prostrate growth habit with many tillers, would produce long grains (Mao et al, 2010). However, while cultivated rice with the A allele exhibits an erect O. glumaepatula still exhibits short grains, suggesting growth and fewer tillers (Jin et al, 2008). that this substitution (R53G) has little effect on the
grain length relative to that of the C-allele of GS3. The Evolution of GS3 C-terminal TNFR/NGFR domain exhibits an inhibitory The higher survival of wild rice may result from small effect on the OSR function. Loss of function of those seeds, which is associated with more rapid maturity domains would result in very short grains (Mao et al, and high fecundity, feasible to dispersal by abiotic 2010), which is consistent with the short grain forces and biotic vectors. Compared with their wild exhibiting in O. nivara. These results would be helpful relatives, cultivated rice usually shows longer grains for better understanding of the evolution of GS3 and
in shape during domestication (Purugganan and Fuller, be consistent with the opinion that domestication is a 2009). A nucleotide substitution of C for A could be dynamic and cumulative evolutionary process that viewed as a turning point during the process of occurs over times (Gepts, 2010).
domestication for grain length, and A allele for long ‘Wild’ alleles of GS3 for cultivated rice improvement grain is associated with strong artificial selection in
tropical japonica (Takano-Kai et al, 2009). Besides the Recent years, narrow genetic basis has been the C-A mutation at locus SF28, two polymorphic loci, limiting factor of yield in cultivated rice (Tanksley and RGS1 and RGS2 in the last intron and the last exon of McCouch, 1997). Wild rice retains abundant genetic GS3, are also associated with grain length, and they diversity under long-term natural selection and is an should occur earlier variation comparing to the C-A excellent gene pool for Asian cultivated rice mutation because RGS1 and RGS2 had differentiation improvement (Xiao et al, 1998). In this study, a total in O. rufipogon (Wang et al, 2011). The present study of 150 polymorphic loci in GS3 derived from related
provided evidences supporting the dynamic and species of Asian cultivated rice were identified (Table cumulative domestication process of GS3 based on the 3). It is helpful to improve seed size and broaden sequence of GS3 in O. sativa and its four AA genome genetic basis of Asian cultivated rice by introgressing relatives. Firstly, nine novel polymorphic loci in the the GS3 alleles from O. glaberrima, O. glumaepatula,
27 ZHANG Yu, et al. Seed Length Controlled by Same Locus in Four AA Genome Species of Oryza
and minor QTL for grain width and thickness in rice, encodes a O. nivara and O. longistaminata into O. sativa.
putative transmembrane protein. Theor Appl Genet, 112(6): Moreover, one InDel and eight SNPs as the candidate 1164–1171. functional markers were found between O. sativa and Fan C C, Yu S B, Wang C R, Xing Y Z. 2009. A causal C-A their related species, including I1446, S316, S1206, mutation in the second exon of GS3 highly associated with rice S2555, S3809, S3938, S4036, S4086 and S4102 (Fig. grain length and validated as a functional marker. Theor Appl 3), which were significantly correlated with seed Genet, 118(3): 465–472. length. Hence, those markers could directly differentiate Frary A, Nesbitt T C, Grandillo S, Knaap E, Cong B, Liu J, Meller alleles conferring long and short grains. Especially, it J, Elber R, Alpert K B, Tanksley S D. 2000. fw2.2: A quantitative is difficult for breeders to select desirable traits such trait locus key to the evolution of tomato fruit size. Science, 289: as grain length using conventional approach, thus, it 85–88.
would be particularly helpful for direct selection with Gepts P. 2010. Crop domestication as a long-term selection
appropriate seed length in early generations of breeding experiment. Plant Breed Rev, 24: 1–44.
Ghesquiere A. 1985. Evolution of Oryza longistaminata. In: program by marker assisted selection.
Proceedings of the Third International Rice Genetics Symposium.
Manila, Philippines: International Rice Research Institutes: ACKNOWLEDGEMENTS
Gupta P K, Rustgi S, Kumar N. 2006. Genetic and molecular basis We kindly thank Dr. Kenneth L. MCNALLY of grain size and grain number and its relevance to grain (International Rice Research Institute, the Philippines) productivity in higher plants. Genome, 49: 565–571. for careful revision of our manuscript. This research Jin J, Huang W, Gao J P, Yang J, Shi M, Zhu M Z, Luo D, Lin H X. was funded by grants from the Ministry of Agriculture, 2008. Genetic control of rice plant architecture under China (Grant No. 2009ZX08009-107B), National domestication. Nat Genet, 40: 1365–1369. Natural Science Foundation of China (Grant Nos. Khush G S. 1997. Origin, dispersal, cultivation and variation of 31201196, 31160275, U1036605 and 31000704), Yunnan rice. Plant Mol Biol, 35(1/2): 25–34. Provincial National Science Foundation, China (Grant Kovach M J, Sweeney M T, McCouch S R. 2007. New insights into
Nos. 2011FB118 and 2010CI035), and Institute of the history of rice domestication. Trends Genet, 23(11): 578–587.
Food and Crop Research, Yunnan Academy of Kubo T, Takano-Kai N, Yoshimura A. 2001. RFLP mapping of
genes for long kernel and awn on chromosome 3 in rice. Rice Agricultural Sciences, China (Grant No. 2013LZS001). Genet Newsl, 18: 26–28.
Li C B, Zhou A L, Sang T. 2006. Rice domestication by reducing REFERENCES
shattering. Science, 311: 1936–1939.
Li J M, Thomson M, McCouch S R. 2004. Fine mapping of a Aluko G, Martinez C, Tohme J, Castano C, Bergman C, Oard J H. grain-weight quantitative trait locus in the pericentromeric 2004. QTL mapping of grain quality traits from the interspecific region of rice chromosome 3. Genetics, 168(4): 2187–2195. cross Oryza sativa × O. glaberrima. Theor Appl Genet, 109(3):
Li Z K, Pinson S R, Park W D, Paterson A H, Stansel J W. 1997. 630–639.
Epistasis for three grain yield components in rice (Oryza sativa Barton N H, Keightley P D. 2002. Multifactorial genetics:
L.). Genetics, 145(2): 453–465. Understanding quantitative genetic variation. Nat Rev Genet, 3:
Lincoln S D M, Lander E S. 1992. Constructing genetic maps with 11–21.
MAPMAKER/EXP3.0. 3rd edn. Technical Report. Combridge, Basten C J, Weir B S, Zeng Z B. 1998. QTL CARTOGRAPHER: A
MA: Whitehead Institute. reference manual and tutorial for QTL mapping. Raleigh:
Mao H L, Sun S Y, Yao J L, Wang C R, Yu S B, Xu C G, Li X H, Department of Statistics, North Carolina State University. Zhang Q F. 2010. Linking differential domain functions of the Caicedo A L, Williamson S H, Hernandez R D, Boyko A,
GS3 protein to natural variation of grain size in rice. Proc Natl Fledel-Alon A, York T L, Polato N R, Olsen K M, Nielsen R, Acad Sci USA, 107: 19579–19584. McCouch S R, Bustamante C D, Purugganan M D. 2007.
Matsuoka M, Ashikari M. 2007. A quantitative trait locus Genome-wide patterns of nucleotide polymorphism in
regulating rice grain width. Nat Genet, 39: 583–584. domesticated rice. PLoS Genet, 3(9): e163.
McCouch S R, Teytelman L, Xu Y B, Lobos K B, Clare K, Walton Churchill G A, Doerge R W. 1994. Empirical threshold values for M, Fu B Y, Maghirang R, Li Z K, Xing Y Z, Zhang Q F, Kono I, quantitative trait mapping. Genetics, 138(3): 963–971.
Yano M, Fjellstrom R, DeClerck G, Schneider D, Cartinhour S, Edwards K, Johnstone C, Thompson C. 1991. A simple and rapid Ware D, Stein L. 2002. Development and mapping of 2240 new method for the preparation of plant genomic DNA for PCR
SSR markers for rice (Oryza sativa L.). DNA Res, 9: 199–207. analysis. Nucl Acids Res, 19(6): 1349.
Moncada P, Martínez C P, Borrero J, Châtel M, Gauch H, Fan C C, Xing Y Z, Mao H L, Lu T T, Han B, Xu C G, Li X H, Guimaraes E P, Tohmé J, McCouch S R. 2001. Quantitative trait Zhang Q F. 2006. GS3, a major QTL for grain length and weight
loci for yield and yield components in an Oryza sativa × Oryza
28 Rice Science, Vol. 21, No. 1, 2014
rufipogon BCF population evaluated in an upland environment. Theor Appl Genet, 107(3): 479–493. 22
Theor Appl Genet, 102: 41–52. Vaughan D A. 1994. The Wild Relatives of Rice: A Genetic Morishima H, Sano Y, Oka H I. 1992. Evolutionary studies in Resources Handbook. Manila: The International Rice Research cultivated rice and its wild relatives. Oxf Surv Evol Biol, 8: Institute, Philippines.
135–184. Wang C R, Chen S, Yu S B. 2011. Functional markers developed Purugganan M D, Fuller D Q. 2009. The nature of selection during from multiple loci in GS3 for fine marker-assisted selection of plant domestication. Nature, 457: 843–848. grain length in rice. Theor Appl Genet, 122(5): 905–913.
Patnaik R N, Pande K, Ratho S N, Jachuck P J. 1990. Heterosis in Wang S K, Wu K, Yuan Q B, Liu X Y, Liu Z B, Lin X Y, Zeng R Z, rice hybrids. Euphytica, 49: 243–247. Zhu H T, Dong G J, Qian Q, Zhang G Q, Fu X D. 2012. Control Redoña E D, Mackill D J. 1998. Quantitative trait locus analysis of grain size, shape and quality by OsSPL16 in rice. Nat Genet,
for rice panicle and grain characteristics. Theor Appl Genet, 96: 44: 950–954.
957–963. Weng J F, Gu S H, Wan X Y, Gao H, Guo T, Su N, Lei C L, Zhang Ren F G, Lu B R, Li S Q, Huang J Y, Zhu Y G. 2003. A comparative X, Cheng Z J, Guo X P, Wang J L, Jiang L, Zhai H Q, Wan J M. study of genetic relationships among the AA-genome Oryza 2008. Isolation and initial characterization of GW5, a major QTL
species using RAPD and SSR markers. Theor Appl Genet, 108: associated with rice grain width and weight. Cell Res, 18:
Shomura A, Izawa T, Ebana K, Ebitani T, Kanegae H, Konishi S, Xiao H, Jiang N, Schaffner E, Stockinger E J, van der Knaap E. Yano M. 2008. Deletion in a gene associated with grain size 2008. A retrotransposon-mediated gene duplication underlies increased yields during rice domestication. Nat Genet, 40: morphological variation of tomato fruit. Science, 319:
Song X J, Huang W, Shi M, Zhu M Z, Lin H X. 2007. A QTL for Xiao J H, Li J M, Grandillo S, Ahn S N, Yuan L P, Tanksley S D, rice grain width and weight encodes a previously unknown McCouch S R. 1998. Identification of trait-improving quantitative RING-type E3 ubiquitin ligase. Nat Genet, 39: 623–630. trait loci alleles from a wild rice relative, Oryza rufipogon.
Spielmeyer W, Ellis M H, Chandler P M. 2002. Semidwarf (sd-1), Genetics, 150(2): 899–909.
‘green revolution’ rice, contains a defective gibberellin Xing Y Z, Zhang Q F. 2010. Genetic and molecular bases of rice 20-oxidase gene. Proc Natl Acad Sci USA, 99(13): 9043–9048. yield. Annu Rev Plant Biol, 61: 421–442.
Sun C, Wang X, Yoshimura A, Doi K. 2002. Genetic differentiation Xing Y Z, Tan Y F, Hua J P, Sun X L, Xu C G, Zhang Q F. 2002. for nuclear, mitochondrial and chloroplast genomes in common Characterization of the main effects, epistatic effects and their wild rice (Oryza rufipogon Griff.) and cultivated rice (Oryza environmental interactions of QTLs on the genetic basis of yield sativa L.). Theor Appl Genet, 104(8): 1335–1345. traits in rice. Theor Appl Genet, 105(2/3): 248–257.
Takano-Kai N, Jiang H, Kubo T, Sweeney M, Matsumoto T, Zhang X J, Wang J F, Huang J, Lan H X, Wang C L, Yin C F, Wu Y Kanamori H, Padhukasahasram B, Bustamante C, Yoshimura A, Y, Tang H J, Qian Q, Li J Y, Zhang H S. 2012. Rare allele of
OsPPKL1 associated with grain length causes extra-large grain Doi K, McCouch S. 2009. Evolutionary history of GS3, a gene
and a significant yield increase in rice. Proc Natl Acad Sci USA, conferring grain length in rice. Genetics, 182: 1323–1334.
109(52): 21534–21539. Tanksley S D, McCouch S R. 1997. Seed banks and molecular
Zhu Q, Zheng X, Luo J, Gaut B S, Ge S. 2007. Multilocus analysis maps: Unlocking genetic potential from the wild. Science, 277:
of nucleotide variation of Oryza sativa and its wild relatives: 1063–1066.
Severe bottleneck during domestication of rice. Mol Biol Evol, Thomson M J, Tai T H, McClung A M, Lai X H, Hinga M E, Lobos
24(3): 875–888. K B, Xu Y, Martinez C P, McCouch S R. 2003. Mapping
quantitative trait loci for yield, yield components and
morphological traits in an advanced backcross population
between Oryza rufipogon and the Oryza sativa cultivar Jefferson.
The corresponding authors of the published article ‘Photoperiodic Control of Flowering and Flower Development in Rice’ on Rice Science (2013, Volume 20, Issue 2, Pages 79–87) should be addressed to both GAO Yong-ming
(firstname.lastname@example.org) and SHI Ying-yao (email@example.com).