* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Download The qSD12 Locus Controls Offspring Tissue-Imposed
Genomic imprinting wikipedia , lookup
Site-specific recombinase technology wikipedia , lookup
Dominance (genetics) wikipedia , lookup
Genetically modified crops wikipedia , lookup
Heritability of IQ wikipedia , lookup
Nutriepigenomics wikipedia , lookup
Hybrid (biology) wikipedia , lookup
Hardy–Weinberg principle wikipedia , lookup
Public health genomics wikipedia , lookup
Genetic engineering wikipedia , lookup
Genetic drift wikipedia , lookup
Genetically modified organism containment and escape wikipedia , lookup
Genome (book) wikipedia , lookup
Human genetic variation wikipedia , lookup
Designer baby wikipedia , lookup
History of genetic engineering wikipedia , lookup
Population genetics wikipedia , lookup
Copyright Ó 2008 by the Genetics Society of America DOI: 10.1534/genetics.108.092007 The qSD12 Locus Controls Offspring Tissue-Imposed Seed Dormancy in Rice Xing-You Gu,*,1 E. Brent Turnipseed* and Michael E. Foley† *Plant Science Department, South Dakota State University, Brookings, South Dakota 57007 and †Biosciences Research Laboratory United States Department of Agriculture, Agricultural Research Service, Fargo, North Dakota 58105 Manuscript received May 28, 2008 Accepted for publication June 12, 2008 ABSTRACT Seed component structures were grouped into maternal and offspring (embryo and endosperm) tissues to characterize a dormancy quantitative trait locus (QTL) for tissue-specific function using a marker-assisted genetic approach. The approach was devised to test if genotypic/allelic frequencies of a marker tightly linked to the QTL deviate from Mendelian expectations in germinated and nongerminated subpopulations derived from a segregation population of partially after-ripened seeds and was applied to the dormancy QTL qSD12 and qSD7-1 in a nearly isogenic background of rice. Experimental results unambiguously demonstrated that qSD12 functions in the offspring tissue(s) and suggested that qSD7-1 may control dormancy through the maternal tissues. These experiments also provide the first solid evidence that an offspring tissueimposed dormancy gene contributes to the segregation distortion in a mapping population developed from partially after-ripened seeds and, in part, to the germination heterogeneity of seeds from hybrid plants. Offspring and maternal tissue-imposed dormancy genes express in very early and late stages of the life cycle, respectively, and interact to provide the species with complementary adaptation strategies. The qSD12 locus was narrowed to the region of 600 kbp on a high-resolution map to facilitate cloning and marker-assisted selection of the major dormancy gene. S EEDS are dormant under the conditions conducive for germination because embryo emergence is blocked by its surrounding structures (i.e., coat-imposed dormancy) or repressed by inhibitory factors within the embryo (i.e., embryo dormancy) (Bewley and Black 1994). Wild or weedy species developed these dormancy mechanisms during evolution to enhance their survival under adverse natural or in human-disturbed environments by selection for an optimum time to germinate. In contrast, many crop species have lost part or all dormancy mechanisms due to the selections for rapid, uniform germination during domestication and breeding activities (Harlan et al. 1973). Lack of seed dormancy causes preharvest sprouting (PHS) in cereal crops (Gubler et al. 2005). The variation in this adaptive or domestication-related trait is controlled by multiple genetic factors ( Johnson 1935), which are known as quantitative trait loci (QTL). Many QTL for seed dormancy have been identified for crop and model plants since the early 1990s (Anderson et al. 1993; Ullrich et al. 1993; Lin et al. 1998; Alonso-Blanco et al. 2003). These QTL are yet to be related to coat-imposed or embryo dormancy to understand their underlying mechanisms. 1 Corresponding author: Plant Science Department, South Dakota State University, SNP248D, Box 2140C, Brookings, SD 57007. E-mail: [email protected] Genetics 179: 2263–2273 (August 2008) Structures surrounding the embryo include endosperm, pericarp, and testa. Grass seeds are also covered by a lemma and a palea (or hull). It has been difficult to distinguish effects of individual component tissues on seed dormancy. Researchers have resorted to excising embryos and removing the hull or pericarp/testa tissues from dormant seeds to infer coat-imposed or embryo dormancy in barley (Romagosa et al. 1999), wild oats (Simpson 1990; Foley 1992), rice (Takahashi 1963; Seshu and Dadlani 1991), and wheat (Morris et al. 1989; Lan et al. 2005). The detected effects using a somatic approach may be confounded with wounding, inhibitor leaching, and other effects. A genetic approach, which is based on the difference in germinability between reciprocal hybrid F1 seeds, was also used to infer dormancy types (Nair et al. 1965; Noll et al. 1982; Seshu and Sorrells 1986; Flintham 2000). Reciprocal difference detected by the genetic approach may be difficult to attribute to a single locus if its genetic background is unknown or when the parental lines differ in multiple loci controlling seed dormancy. Consequently, additional approaches are needed to characterize QTL for dormancy types on the basis of intact seeds in a known genetic background. Seed dormancy, as a key adaptive trait in natural populations, has not been related to segregation distortion, the phenomenon of selective advantage and frequently reported for experimental populations from 2264 X.-Y. Gu, E. B. Turnipseed and M. E. Foley TABLE 1 Genotypes for component tissues of seeds from plants heterozygous for a dormancy (D/d) locus tagged by a codominant marker (M/m) and hypothesis tests for the allelic frequencies FD and Fd Offspring tissues Maternal tissue (2n) genotypea DM/dm (NGS) Endosperm (3n) Embryo (2n) Hypothesis testing for embryo genotypes in subpopulationsc Genotype Frqb Genotype Frqb Germinated Nongerminated DM/DM/DM DM/DM/dm DM/dm/dm dm/dm/dm 0.25 0.25 0.25 0.25 DM/DM DM/dm 0.25 0.50 H0: FD ¼ Fd ¼ 0.50; or HA: FD , 0.50 and Fd . 0.50 H0: FD ¼ Fd ¼ 0.50; or HA: FD . 0.50 and Fd , 0.50 dm/dm 0.25 a No genotypic segregation (NGS) for the tissues includes hull, pericarp, and testa; the letter n indicates the gametic number of chromosomes; the D and M (or d and m) alleles are linked in coupling. b Expected genotypic frequencies for a population of fully after-ripened seeds. c The subpopulations are obtained by germinating partially after-ripened seeds; allelic frequencies FD and Fd are estimated on the basis of marker genotypes for the embryos; and H0 or HA are null or alternative hypotheses. distant crosses (Lyttle 1991; Taylor and Ingvarsson 2003). Segregation distortion refers to deviations of allelic and genotypic frequencies from Mendelian expectations in an experimental population. Biological mechanisms for distorted segregations are explained as preferential fertilization of male or female gametes or zygotic difference in viability resulting from gene or chromosomal mutations (Oka 1988; Lyttle 1991; Taylor and Ingvarsson 2003). Segregation distortion loci (SDL) were reported for all 12 chromosomes in rice (Xu et al. 1997). Most of the SDL are related to gameophyte or zygote sterility genes contributing to interspecific or subspecific differentiations and the biological bases for remaining SDL are uncertain (Oka 1988; Xu et al. 1997). We observed segregation distortions in our research on seed dormancy, which was characterized by reduced germination rate. This preliminary observation was made in a nearly isogenic background in which there is no detectable reproductive barrier or viability problem. Therefore, the observed segregation distortion may have something to do with the dormancy QTL or partial dormancy release. Rice (Oryza sativa) is a model species for seed dormancy research. More than 30 QTL associated with seed dormancy or PHShave been reported for cultivated (Lin et al. 1998; Dong et al. 2003; Guo et al. 2004; Wan et al. 2005, 2006), wild (Cai and Morishima 2000; Thomson et al. 2003; Lee et al. 2005; Li et al. 2006), and weedy (Gu et al. 2004) rice, but none of them has been associated with seed component tissues. We introduced a set of dormancy QTL alleles, including those at the qSD12 and qSD7-1 loci, from weedy into cultivated rice (Gu et al. 2006a). The qSD12 locus might have played a major role in rice domestication and obviously harbors a key dormancy gene that likely imparts resistance to PHS, because it explained .50% of the phenotypic variance in the nearly isogenic background and its effect on seed dormancy was enhanced by relatively high temperatures during seed development (Gu et al. 2006a,b). The objectives of this research were (1) to develop a marker-assisted genetic approach to characterize a QTL for maternal or offspring tissue-imposed seed dormancy, (2) to determine the relationship of the qSD12 dormancy QTL with segregation distortion, and (3) to develop a high-resolution map for the qSD12 region to facilitate cloning, molecular characterization, and marker-assisted selection of the major dormancy gene. MATERIALS AND METHODS Genetic assumptions: Seed structures are grouped into maternal (2n) and offspring tissues, with the latter consisting of triploid (3n) endosperm and diploid (2n) embryo (Table 1). It is assumed that in a diploid species (1) a population of seeds is harvested from plants heterozygous for the same selected dormancy QTL region(s) and grown under the same conditions, and accordingly these seeds are identical for their maternal tissue genotype and macro (between-plants) environmental variation in germination; (2) the dormancy locus (D/d) completely or tightly links to a codominant marker (M/m), with the D (M) and d (m) alleles derived from the dormant and nondormant parents, respectively; (3) in the QTL region there is no gametophytic/sporophytic sterility gene or other gene resulting in preferential fertilization and zygotic selection; and (4) the seeds are partially after-ripened (i.e., stored in warm, dry conditions for given days) so that they can be divided into germinated and nongerminated subpopulations by a germination experiment. In the above population, there is no segregation for the maternal genotype and there is an association between endosperm and embryo genotypes due to the double fertilization. Genotypic identification with the marker is much easier for embryo than for endosperm tissues because the embryonic genotype can be identified with the seedling tissue. Therefore, hypothesis testing can be conducted on the basis of segregation ratios or allelic frequencies for embryonic genotypes in the subpopulations (Table 1). There are three possibilities for a dormancy gene to express in (1) only the maternal tissue(s), (2) only the offspring tissue(s), and (3) both the maternal and offspring tissues. If the QTL associates with dormancy imposed by the maternal tissues (i.e., possibility 1), the segregation ratio (DD:Dd:dd ¼ 1:2:1) or allelic (FD ¼ Fd ¼ Offspring Tissue-Imposed Seed Dormancy 0.50) frequencies of the embryos follow the Mendelian expectations in either of the subpopulations. If the QTL controls dormancy imposed by the offspring tissues (i.e., possibility 2), the segregation ratio or allelic frequencies would deviate from the expectations. The deviation varies in magnitude with degree of dormancy or germination rates (the lower the germination, the higher the magnitude of deviation) and in the parental direction between germinated (Fd . FD) and nongerminated (FD . Fd) subpopulations (Table 1). If the QTL controls dormancy through the maternal and offspring tissues (i.e., possibility 3), the deviations due to the offspring tissue would be the same as those for possibility 2, and the maternal effect on dormancy cannot be detected because of no genotypic segregation for the covering tissues. Therefore, accepting the null (H0) and rejecting the alternative (HA) hypotheses imply the QTL may control dormancy through the maternal tissue(s), and rejecting the H0 and accepting the HA suggest involvement of the QTL in offspring tissue-imposed dormancy. A partially after-ripening treatment may be used to manipulate the germination rate to better test the hypotheses (Table 1). Tests for both germinated and nongerminated subpopulations are necessary to verify if other distortion factors, such as gamete preferential fertilization genes, are present in the dormancy QTL region. A deviation due to a distorter unrelated to germination would favor the same parent in both the germinated and nongerminated subpopulations. Plant materials and cultivation: This research started with selection for the plant heterozygous for the qSD12 and qSD7-1 regions from the advanced backcross (BC4F3) population (Gu et al. 2006a). The donor parent of the dormancy alleles at the QTL is the weedy rice (O. sativa) accession SS18-2, originated from Thailand and the recipient parent of the alleles is the nondormant indica line EM93-1. The qSD12 and qSD7-1 QTL peak positions are closest (,1 cM) to the markers RM270 and RM5672, respectively, on chromosomes 12 and 7, respectively (Gu et al. 2004, 2005). The genetic background of the selected dihybrid is identical to EM93-1, which was checked with 150 markers that are relatively evenly distributed along the framework genetic map (Gu et al. 2004, 2006a). A preliminary experiment was conducted with a sample of seeds harvested from the dihybrid plant. After 21 days of after-ripening (DAR) at room temperature (25°), the seeds were germinated to develop a segregation population for preliminary research. Follow-up experiments focused on the qSD12 region. The initial plant containing one copy of the weedy rice-derived segment from RM3331 through RM235 (Figure 2B) was selected from the BC4F4 population (Gu et al. 2006a). This monohybrid plant was multiplied by a ratooning technique to increase seeds. Seeds were harvested at 40 days after flowering and air-dried in a greenhouse for 3 days. Dried seeds were partially after-ripened at room temperature for 10, 20, and 30 days prior to germination. Germinated seeds from each germination experiment were collected as a germinated subpopulation and the seedlings genotyped with RM3331, RM270, and RM235. Segregation ratios of the markers from each subpopulation were used to test the hypotheses (Table 1). Recombinants between the markers were transplanted into 13-cm pots (one plant per pot) filled with a mixture of clay and greenhouse medium to harvest seeds. To narrow qSD12 and estimate its gene effects on germination, three recombinants (i.e., nos. 1–3 in Figure 2C), which are different in length from the SS18-2-derived segments containing the qSD12 peak position, were determined for seed dormancy by progeny testing. Fully after-ripened seeds from each recombinant were germinated to develop a progeny population. Individual seedlings derived from recombinant nos. 1 and 3 were genotyped with selected markers on the introgression segments (Figure 2C). Plant cultivation and seed harvest and drying were conducted as described above. Dried seeds were stored at 20° to maintain the dormancy status. 2265 Because recombinant no. 3 has the shortest segment from the weedy rice parent (Figure 2C), it was used to develop a high-resolution map for the narrowed qSD12 region. In addition, the two recombinant-derived homozygotes (refer to the DD and dd genotypes in Figure 3B) were selected as the isogenic lines for the dormancy (ILSD12DD ) and nondormancy (ILSD12dd ) alleles, respectively. Furthermore, seeds from the recombinant-derived heterozygous plants (refer to the Dd genotype in Figure 3B) were bulked to develop a large segregation population to test the hypotheses (Table 1). Identification of seed dormancy and development of germinated and nongerminated subpopulations: Seed dormancy was quantified by germination percentage. Seed samples were maintained at 25° for the given periods of time for partial after-ripening. Fifty seeds were placed in a 9-cm petri dish lined with Whatman no. 1 filter paper, wetted with 10-ml deionized water, and then incubated at 30° and 100% relative humidity in the dark. Germination was evaluated visually by protrusion of the radicle or coleoptile from the hull by $3 mm at day 7 or daily for a period of 10 days, depending on the purpose of the experiment. Three replications were used for a germination experiment. To prepare segregation populations for hypothesis testing, germinated seeds were transferred to large containers lined with three layers of moist germination paper and maintained at room temperature in the light for 10 days. Seedlings collected from one germination experiment were grouped as a germinated subpopulation. The seeds not germinated after 7 days of incubation were collected and air-dried in a greenhouse for .60 days to break dormancy. These dried seeds were sterilized with 50% bleach for 15 min, rinsed with running water for 30 min, and then incubated at 30° and 100% relative humidity in the light to promote germination (about 90%). Germinated seeds from the second incubation were transferred to the same conditions as the above germinated subpopulation and the seedlings grouped as a nongerminated subpopulation. Marker development and genotypic identification: To delimit the initial introgression segment, 34 rice microsatellite (RM) markers between RM309 and RM17 on chromosome 12 (McCouch et al. 2002) were screened for polymorphism between EM93-1 and SS18-2. Polymorphic markers were used to genotype the initial plant and then aligned against the Nipponbare (a japonica variety of rice) genome sequence (International Rice Genome Sequencing Project 2005) to physically delimit the introgression segments. RM270 was the marker nearest to the qSD12 QTL peak position (Gu et al. 2004) and its primers were designed on the basis of the cDNA sequence from an indica variety of rice (Temnykh et al. 2001). Unfortunately, RM270 was unable to be positioned on the japonica genome ( J. Ni, personal communication). Therefore, polymorphic markers on the introgression segment were used to genotype available DNA samples from the population of 204 BC1 (EM93-1//EM93-1/SS18-2) plants (Gu et al. 2005) and RM270 was positioned on the basis of its genetic distances to the flanking markers estimated using MAPMAKER/EXP 3.0 (Lincoln et al. 1992). Forty-five additional RM markers were developed to construct a high-resolution map for the narrowed qSD12 interval between RM5479 and RM6123. The primers were designed to target the same simple sequence repeats as those for the rice RM markers on the Gramene Oryza sativa japonica MapView (http://www.gramene.org) using the Primer3 software (Rozen and Skaletsky 2000). The new primers were used to amplify the same RM markers because the estimated sizes of PCR products (231 6 43 bp) fit our electrophoresis system better than those (255 6 138 bp) amplified using the recommended primers (http://www.gramene.org). Markers polymorphic be- 2266 X.-Y. Gu, E. B. Turnipseed and M. E. Foley TABLE 2 Information about new polymorphic markers Marker name RM28595 RM28603 RM28607 RM28608 RM28621 RM28638 RM28642 RM28643 RM28645 RM28651 RM28652 RM28656 RM28659 RM28661 RM28662 RM28664 RM28665 RM28682 RM28683 Position (bp) Size (bp) Forward primer Reverse primer 24557835 24683314 24766244 24767480 24942198 25069797 25121458 25124840 25167389 25243158 25276121 25352787 25392037 25430188 25435457 25442822 25455125 25737201 25747576 240 207 250 248 249 288 237 190 231 250 219 363 220 169 195 243 220 192 183 tacaacgcaccccatctgta cattggactcacctggaagg agaactagagagatgagaaaggaaag tgtgtattagtttccatatggtcttca cacagtcgaattgcaaagga ccatcctgcctctagcatgt agctgctgagaacacaatcg tgggggttgtactcttctcc gggcgccagtattagtgttg cttcctggcttcaactctgg aacaacattctctgcaattttcc gcacacagtggaagtacgtttg aaacgcatgcagaagaacct acagtgactttggccgtgtt ttggagctgattttggagtttt gccttagcttctccctgctt tgcagatggtgaggaagttg tctccgagagggtacgtgtc tgcgtggtgaattgtgattt gcccaatcatcttgcatctt caccaatctccgccattact ggcagctcaaccctttcatag actacaatatggggcggatg gccaaaaggtcagggttaca ctgaagagctgcgagaatcc gtacctctccacccatcgac ccgatgttgagacaaggtga acgcagcatgtaggagaggt cggaactgccgtttattcat tctcaattgcactccatcca tccgattataccattgtattcgtt ccatcgaaagatgtgtggaa cgcgtgttgtatggtttcac gttttaaggcccccatcatt tgggaagcagaagagtttttg ctcaaggacgtgtggaacg tctcctctgcatcacaatcaa aagtccgtctccgtcttcg Markers were designed to target the same simple sequence repeats (SSR) as those for the rice microsatellite (RM) markers listed on the Gramene Oryza sativa japonica MapView at http://www.gramene.org/Oryza_sativa_ japonica/mapview?chr¼12. tween EM93-1 and SS18-2 were screened (Table 2) and used to saturate the target region. Genomic DNA was prepared from a leaf segment from the individual seedlings at 3 weeks. DNA extraction, PCR conditions, and PCR product display were conducted using the methods of Gu et al. (2004). Genetic and statistic analyses: Germination percentage (y) for a sample was transformed by sin1(y)0.5 and the transformed values were averaged over three replications for genetic analysis. For the population segregating for both qSD12 and qSD7-1 regions, the Cockerham’s model (Kao and Zeng 2002) was used to partition the genetic effect and variance of the two loci into the additive (a), dominance (d), and epistatic components. The two QTL were represented by their tightly linked markers RM270 and RM5672, respectively, in the linear model. Regression analysis was performed using the SAS REG procedure with a stepwise selection set at a significant level of 5% (SAS Institute 1999). For the population segregating only for the qSD12 region, the genic effects (a and d) and their standard errors were estimated following the single-locus model (Gu et al. 2006a). Significance of the estimates was determined by a t-test. Genotypic frequencies FDD, FDd, and Fdd were estimated as proportions of the numbers of DD, Dd, and dd seedlings, respectively, to the total number (N) of the subpopulation. Fitness of an observed segregation ratio was determined by chisquare testing against the 1:2:1 monogenic expectation. The allelic frequency FD or Fd was estimated as FDD 1 12 FDd or Fdd 1 1 0.5 2 FDd and the standard error calculated as (FD 3 Fd/N) . A binomial test was used to determine fitness of the observed allelic frequencies to the 0.50 expectation. A slight but statistically not significant segregation distortion was observed with markers near the qSD12 region in the BC1 population (Gu et al. 2004). The multipoint mapping method (Vogl and Xu 2000) was applied to available marker genotype data to map the SDL. The likelihood distribution generated by multipoint mapping was compared with the distribution of test statistics computed by simple interval mapping (Tinker and Mather 1995) for the same chromosomal region to determine the positional relation between the SDL and qSD12. RESULTS Difference in segregation ratio between the qSD12 and qSD7-1 regions: About 70% of seeds from the dihybrid plant were germinated after 21 DAR. A population of 177 seedlings from the germinated seeds was genotyped for the RM270 and RM5672 loci, which mark the qSD12 and qSD7-1 QTL,respectively. Segregation distortion was detected for RM270, with the allelic frequencies FD or Fd being significantly lower or higher than the expected 0.50; whereas, the segregation ratio for three RM5672 genotypes fits the 1:2:1 expectation in the germinated subpopulation (Figure 1A). These results imply that there may be a genetic factor in the qSD12, but not in the qSD7-1 region, that reduced the frequency for the allele from the dormant parent. This factor could be the dormancy gene according to the hypothesis (Table 1). Plants in the dihybrid-derived population varied in their degree of seed dormancy, as determined by the distribution of germination for 21-days after-ripened seeds. Genotypic differences in germination are much larger for qSD12 than for qSD7-1 (Figure 1B). Both loci consisted of gene additive and dominance effects, and qSD12 and qSD7-1 accounted for 82 and 8% of phenotypic variance, respectively (Table 3). Because qSD12 and qSD7-1 both had significant effects on Offspring Tissue-Imposed Seed Dormancy 2267 TABLE 3 2 Contributions (R ) of qSD12 and qSD7-1 component gene effects to the phenotypic variation in germination of seeds after-ripened for 21 days Componenta m a1 d1 a2 d2 Effectb 0.60 0.42 0.15 0.11 0.10 6 6 6 6 6 0.012 0.018 0.024 0.017 0.023 Total t-value Probability R 2 (%) 49.6 22.9 6.2 6.4 4.4 ,0.0001 ,0.0001 ,0.0001 ,0.0001 ,0.0001 NA 78 4 6 2 90 a m, mean of the model; a1 (d1) and a2 (d2), gene additive (dominance) effects for qSD12 and qSD7-1, respectively, in the population segregating only for the two QTL-containing regions (Figure 1). b Arc-sin square root transformed germination (mean 6 standard deviation). The minus sign indicates that the weedy rice-derived alleles reduce germination or enhance seed dormancy. Figure 1.—Genotypic frequencies (A) and differences in germination (B) in a population segregating only for the qSD12 and qSD7-1 QTL regions. Germination was evaluated for individual plants with their seeds partially after-ripened for 21 days. RM270 and RM5672 are markers tightly linked to the loci qSD12 on chromosome 12 and qSD7-1 on chromosome 7, respectively. Italic letters DD, Dd, and dd indicate three genotypes of plants for qSD12 or qSD7-1 in the population, with D and d representing the alleles from the dormant and nondormant parents, respectively. N is the population size, P is the x2-test probability for fitness of an observed dd:Dd:DD ratio to the 1:2:1 expectation, and FD and Fd are D and d allelic frequencies. germination but differed in segregation distortion, we suspected that these two loci may be involved in the genetic control of offspring and maternal tissue-imposed seed dormancy, respectively. An overwhelming majority (78%) of the genetic variance in the above population was contributed by the additive effect of qSD12 (Table 3). Because of its association with segregation distortion and major effect, we focused on the qSD12 QTL. Fine mapping and genic effects of qSD12: Fourteen of the 34 markers were polymorphic between EM93-1 and SS18-2 (Figure 2A) and located on the introgression segment in the initial plant (Figure 2B). The introgression segment ranges from 2.71 to 5.50 Mbp according to the physical positions of markers flanking the segment ends. RM270, which marked the qSD12 peak in previous experiments (Gu et al. 2004, 2005), was located between RM5479 and RM6265 (Figure 2A), with distances to the two markers being 5 and 0.5 cM, respectively, on the basis of the BC1 population (Gu et al. 2005). About 20 recombinants between the RM3331, RM270, and RM235 loci were identified from germinated subpopulations derived from the initial plant (Figure 2B). These recombinants were genotyped for the 14 polymorphic markers to determine the length of retained introgression segments. None of the recombinants was for the RM270–RM3739 or RM270–RM3739 regions. One recombinant (no. 3) for the RM5479– RM6123 region was most likely derived from a twocrossover event (Figure 2C). Three recombinants (nos. 1, 2, and 3) with different lengths of the introgression segments containing the qSD12 peak position (Figure 2C) were confirmed for seed dormancy by progeny testing. Both recombinants nos. 1 and 3 are heterozygous for the introgression segment (Figure 2C) and their progeny lines segregated for seed dormancy (Figure 3, A and B). The qSD12 locus had no significant dominance effect but displayed predominant additive effects, which reduced the germination rate by 22 and 36% in the recombinants nos. 1- and 3-derived populations, respectively (Figure 3, A and B). Recombinant no. 2 is homozygous for the qSD12 dormancy allele (Figure 2C), and its progeny line had a much longer duration of dormancy than the parental line EM93-1, which was evaluated with the seeds from both lines grown under the same (winter) conditions in a greenhouse (solid lines in Figure 3C). It took 4 and 1 weeks of afterripening for the recombinant no. 2-derived line and EM93-1 to reach 90% germination, respectively. The above progeny tests confirmed the dormancy gene was retained in all three recombinants, indicating the location of qSD12 on an overlapped interval of ,1.69 Mbp between RM5479 and RM6123 (Figure 2C). Seeds from the recombinant no. 3-derived homozygotes DD and dd (Figure 3B) were selected as the putative isogenic lines ILSD12DD and ILSD12dd, respectively. One more generation of progeny testing conducted in the summer season confirmed that ILSD12DD and ILSD12dd are homozygous for the wild and mutant alleles, respectively, at the dormancy locus, as reflected by their difference in the germination profile (dotted 2268 X.-Y. Gu, E. B. Turnipseed and M. E. Foley Figure 2.—Partial physical maps (A and D) for the qSD12 QTL region and graphic representation of genotypes for selected recombinants (B and C) and isogenic line (E). Rice microsatellite (RM) markers are placed according to their positions on the Nipponbare genome sequence for chromosome 12 excluding RM270 (dotted line), which is placed on the basis of its linkage to the flanking markers as its physical position was not determined. The arrow indicates the putative peak position of qSD12. A solid horizontal bar depicts an introgression segment from the weedy rice accession SS18-2 in the genetic background of the cultivated rice line EM93-1 (open bar), and a shaded bar indicates that the segment was not determined for its origin from SS18-2 or EM93-1; the letters DD and Dd indicate the individuals homozygous and heterozygous, respectively, for the SS18-2-derived segment. lines in Figure 3C). Comparing the germination data collected from the aforementioned winter and summer seasons, ILSD12dd was similar to EM93-1, while ILSD12DD displayed longer dormancy duration than the recombi- nant no. 2-derived line (Figure 3C). The difference between the recombinant no. 2-derived line and ILSD12DD could be attributed to environmental variation, because the effect of qSD12 on dormancy was enhanced by Figure 3.—Germination profiles of the qSD12 (D/d) genotypes. (A) Distribution of the recombinant no. 1-derived population (Figure 2C) for germination of seeds after-ripened for 7 days. Number of plants is indicated in parentheses after the genotypes DD, Dd, and dd. (B) Cumulative germination of seeds afterripened for 10 days from recombinant no. 3-derived plants (Figure 2C). (C) Duration of seed dormancy for the lines homozygous for qSD12 dormancy (solid circle) and nondormancy (open circle) alleles. The recipient parent EM93-1 was used as the nondormant control for recombinant no. 2 derivatives. ILSD12DD and ILSD12dd are isogenic lines selected from recombinants no. 3-derived DD and dd genotypes (Figure 3B), respectively. An asterisk (*) and the letters ‘‘ns’’ indicate the gene additive (a) and dominance (d) effects were significant at a probability level of P , 0.01 and not significant, respectively. Solid and dotted lines indicate the seeds harvested in winter and summer seasons, respectively. Bars indicate means and standard deviations of 10 plants. Offspring Tissue-Imposed Seed Dormancy 2269 TABLE 4 Observed genotypic frequencies for the markers linked to the qSD12 locus in germinated subpopulations DAR (% germination rate)a Number of seedlings for genotypec Fitness test for dd:Dd:DD ¼ 1:2:1 Allelic frequencyd Markerb dd Dd DD Total x2 value Prob. 10 (35) RM3331 RM270 RM235 91 100 92 111 106 110 22 18 21 224 224 223 42.5 60.7 45.3 0.0000 0.0000 0.0000 FD ¼ 0.32 Fd ¼ 0.68 20 (79) RM3331 RM270 RM235 52 58 51 101 99 103 40 36 39 193 193 193 1.9 5.1 2.4 0.3844 0.0763 0.3061 FD ¼ 0.44 Fd ¼ 0.56 30 (92) RM3331 RM270 RM235 58 60 51 122 124 126 52 48 55 232 232 232 0.9 2.3 1.9 0.6278 0.3096 0.3941 FD ¼ 0.47 Fd ¼ 0.53 a Days of after-ripening (DAR) prior to the germination experiments. RM270 is nearest to the qSD12 peak position, and RM3331 and RM235 are 1.3 Mbp away from the peak (Figure 2A). c D and d indicate the alleles derived from the dormant and nondormant parents, respectively. d FD or Fd are D or d allelic frequencies for RM270. b relatively high temperatures during the seed development in the previous research (Gu et al. 2006b). Of the 45 newly designed markers, 21 and 11 are polymorphic between EM93-1 and SS18-2 and between ILSD12dd and ILSD12DD, respectively (Figure 2, D and E). With these new markers, the introgression segment retained in ILSD12DD was located between RM28595 and RM28656 and its physical length ranges from 593 to 795 kbp. Association of qSD12 with offspring tissue-imposed seed dormancy: The 10-, 20-, and 30-DAR treatments resulted in 35, 79, and 92% germination (Table 4), respectively, for seeds sampled from the initial plant (Figure 2B). The germinated seeds were genotyped with the relatively loosely linked markers RM3331, RM270, and RM235 (Figure 2A). This experimental design provided a set of subpopulations to examine relationships of segregation deviation with degrees of dormancy and linkage strengths between markers and dormancy QTL. As expected, segregation distortion occurred with linked markers in the three subpopulations (Table 4). The most significant deviation occurred at 10 DAR, when the frequency of the nondormant parent-derived allele (0.68) was twice as much as that of the nondormant parent-derived allele (0.32) at RM270 in the germinated subpopulation. The derivations diminished with increased (79 to 92%) germination following longer periods of after-ripening. The degree of deviation was greater for RM270 than for RM3331 or RM235, as quantified by the chi-square values (Table 4). This quantitative variation with the linked markers coincides with the distribution of test statistics (TS) for the QTL mapping where RM270 had higher TS value than the remaining markers (Gu et al. 2004, 2005). The above results were verified with separate experiments. The first verification was made with a small sam- ple of 60 seedlings from the nongerminated seeds collected from the experiments reported in Table 4. The observed segregation ratio (dd:Dd:DD ¼ 8:19:33) strongly skewed to the dormant homozygote, with FD and Fd equal to 0.71 and 0.29, respectively. The second verification was made with a large population segregating for the greatly narrowed qSD12 region (,800 kbp, refer to Figure 2E). These seeds were harvested from the recombinant no. 3-derived heterozygotes (Dd in Figure 3B) and divided into germinated and nongerminated subpopulations by a 21-DAR treatment. Similar to the above observations, FD and Fd were significantly lower and higher than 0.50, respectively, in the germinated subpopulation of 860 seedlings or higher and lower than 0.50, respectively, in the nongerminated subpopulation of 2738 seedlings (Figure 3). We also noted that a distorted segregation was not present in the recombinant no. 1-derived population (Figure 3A), which was developed from fully after-ripened seeds. The above results strongly indicate that in the qSD12 linkage region there is no ‘‘segregation distorter’’ other than the dormancy gene. The last verification was conducted with seeds also from a recombinant no. 3-derived heterozygote, but the germination was conducted with seeds immediately removed from the freezer. After 10 days of incubation, the sample had a cumulative germination rate of 15%, yielding 139 seedlings for marker genotyping. In this germinated subpopulation, the allelic frequencies dramatically skewed to the allele from the nondormant parent (Fd ¼ 0.81). More interestingly, the nondormancy allele preferentially distributed in the early (2–5 days of incubation) germinated seeds and the dormancy allele was detected only in the seeds that required $5 days of incubation (Figure 4). In summary, the dormancy effect of qSD12 on its allelic segregation could be 2270 X.-Y. Gu, E. B. Turnipseed and M. E. Foley Figure 4.—Frequency distributions of the qSD12 dormancy (D) and nondormancy (d) alleles in germinated and nongerminated subpopulations of partially after-ripened seeds. The seeds were bulked from heterozygous (Dd) plants in the recombinant no. 3-derived population (Figure 3B) and genotyped with the marker RM28621 (Figure 2D). The letter N indicates the number of genotyped seedlings and an asterisk (*) indicates a frequency significantly (P , 0.0001) deviated from the Mendelian expectation (horizontal dashed line). detected with both germinated and nongerminated subpopulations of partially after-ripened seeds and with the seeds germinated after different days of incubation. DISCUSSION The marker-assisted genetic approach: Seed dormancy is imposed by the maternal or offspring tissues, or a combination of both. We demonstrated that the marker-assisted approach (Table 1) could be used to determine if a QTL is involved in the genetic control of seed dormancy imposed by the maternal or offspring tissues. This approach, which combines marker genotyping with after-ripening and germination techniques and is based on segregation populations of a reasonable size, could be a component of experiments for QTL allele introgression, fine mapping, and map-based cloning (Han et al. 1999; Bentsink et al. 2006; Gu et al. 2006a). This marker-assisted approach is reliable and valuable, because it is based on Mendel’s first law and can be used to directly characterize a dormancy gene/ QTL for tissue-specific expression at a phenotypic level. Among the factors affecting the efficacy of this approach could be the phenotypic effect and genetic background of the dormancy QTL and its linkage strength with selected markers. Although this marker-assisted approach was devised for research on seed dormancy in diploid rice, it can be extended to polyploidy species that follow a diploid segregation (e.g., wheat) or modified for research on other seed traits (e.g., seed desiccation tolerance, longevity, and vigor). Because the allelic frequencies in this research were estimated on the basis of seedling (or embryo) genotypes, a statistically significant effect could arise from the same gene Figure 5.—Frequency distribution of the qSD12 (D/d) genotypes for time of incubation required for onset of germination. Seeds were taken from a recombinant no. 3-derived heterozygous (Dd) plant (Figure 3B) and germinated without an after-ripening treatment. The letter N indicates the number of total seedlings genotyped for the marker RM28621 (Figure 2D), and the asterisk (*) indicates the frequency of dormancy (FD) or nondormancy (Fd) allele significantly (P , 0.0001) deviated from the expected 0.50. expressed in the embryo, endosperm, or both tissues due to the double fertilization. Therefore, other methods, such as the triploid genetic models for generation mean or variance analysis (Mo 1988), could be applied to the populations developed from nearly isogenic lines to distinguish endosperm from embryo effects of the dormancy QTL. A series of observations (Figures 1 and 4–6) were made to test the hypotheses (Table 1) and conclude that qSD12 is involved in offspring tissue-imposed seed dormancy. The qSD7-1 underlying gene appeared to control seed dormancy through the maternal tissue(s) because a distorted segregation was not detected for embryo genotypes of the linked marker when the QTL effect was significant in the preliminary observation (Figure 1 and Table 3). We are conducting additional experiments to verify the preliminary observation on the basis of qSD7-1 isogenic lines (X. Gu and M. Foley, unpublished data) and will test the remaining dormancy QTL isolated from weedy rice using the same approach. Seed dormancy gene-related segregation distortion: Allelic differentiation at the QTL associated with offspring tissue-imposed dormancy may cause segregation distortion in an experimental population. The dormancy gene-related distortion differs from the segregation distortion due to meiotic drive or zygotic selections (Lyttle 1991; Taylor and Ingvarsson 2003), which can be exemplified by the rice qSD12 and ga-13 loci on the long arm of chromosome 12. The ga-13 locus was identified as a gametophyte gene by distorted segregation for the linked isozyme gene Acp1 in the indica/ japonica cross (Rha et al. 1994). The Acp1 locus is located at 67 cM on the classical genetic map (http://www.shigen. nig.ac.jp/rice/oryzabase/genes/symbolDetailAction.do? Offspring Tissue-Imposed Seed Dormancy Figure 6.—Distributions of likelihood for a segregation distortion locus (SDL) (A) and test statistics for the qSD12 dormancy QTL (B). Original data were obtained from the BC1 population (Gu et al. 2004). Only part of chromosome 12 is shown and open circles on the horizontal axes indicate marker positions. Seed dormancy was measured by germination of seeds at 1, 11, and 21 days after-ripening (DAR). Arrows indicate the SDL and qSD12 peak positions detected using multipoint mapping (Vogl and Xu 2000) and simple interval mapping (Tinker and Mather 1995) strategies, respectively. The observed segregation ratio for the RM270 marker and its fitness test are shown because it is nearest the peaks, and the letters D and d represent the alleles from the dormant and nondormant parents, respectively. mutantGeneld¼966) and qSD12 is estimated at 97 cM on the basis of its tightly linked marker RM6265 (Figure 2D) based on the molecular genetic map (McCouch et al. 2002). The ga-13-caused distortion favored the indica allele, the distortion pattern did not change across the three environments, and the distortion is due to the genotypic difference of male gametes in competition for fertilization (Rha et al. 1994). In sharp contrast, the qSD12-related segregation distortion favored the alleles from the nondormant parent in the germinated populations and the dormant parent in the nongerminated subpopulations across generations (Table 4, Figures 1A, 4, and 5) and did not occur in the populations developed from fully after-ripened seeds (Table 4, Figure 3A). In addition, the distortion is due to the gene expressed in offspring tissue(s) and functions in a bilateral way during the period from seed maturation to germination to adapt to environmental variations. Similar to the segregation distortions due to other factors, a dormancy gene-related distortion also has a selective advantage in natural populations and a negative impact on linkage analysis. Dormancy distributes 2271 germination over a period of time with nondormant genotypes germinating earlier than the dormant ones in the soil seed bank. This germination distribution pattern coincides with the allelic distribution for the gene expressed in the offspring tissues, in that the nondormancy (mutant) allele has a higher frequency in the early germinating subpopulation and the dormancy (wild) allele has a higher frequency in the later germinating subpopulation due to the segregation distortion (Figure 4). Natural selection may favor less dormant seeds that can germinate and complete their life cycle in the next season, but favor more dormant seeds when environmental conditions are adverse. Apparently, the wild allele, together with its flanking chromosomal segments, would have a greater chance than the mutant allele to be present in the next generation in adverse environments. An experimental population is generally developed from germinated seeds. Segregation distortion in a population due to offspring tissue-imposed dormancy genes might occur if the seeds had not been fully after-ripened. This distortion is more likely in populations from crosses between cultivated and wild accessions, as the latter usually have strong seed dormancy and biases the estimation for parameters such as genetic distance and effects. To avoid the negative impact on map construction and QTL analysis, germination evaluation under controlled conditions is necessary to determine the degree of dormancy for seeds used to develop the mapping population. On the other hand, marker genotyping data may be employed to assist with identification of a dormancy QTL. The available genotyping data for the markers on the long arm of chromosome 12 from the BC1 population (Gu et al. 2004) were analyzed with the multipoint mapping method (Vogl and Xu 2000) to prove the above hypothesis. The modulated peak of the segregation distortion locus (Figure 6A) was identical to the qSD12 peak (Figure 6B), which was computed using simple interval mapping software (Tinker and Mather 1995). We noted that the observed segregation ratio for the two RM270 (on the peak positions) genotypes (dd:Dd ¼ 84:72) skewed in favor of the nondormant parent (dd), but this observation is not different from the 1:1 expectation (Figure 6A). This was because the mapping population was developed with hybrid seeds that received different after-ripening or dormancy-breaking treatments, such as hull removal, warm drying, and soaking overnight with running water, before germination. Implications of an offspring tissue-imposed dormancy gene: The qSD12 QTL is the first naturally occurring locus that has been physically finely mapped and unambiguously determined to regulate offspring tissue-imposed seed dormancy. Discovery of the character of qSD12 provides an additional insight into the genetic system controlling the adaptive trait. A population of seeds is characterized by heterogeneity of germination, which refers to the within-sample variation in 2272 X.-Y. Gu, E. B. Turnipseed and M. E. Foley onset of germination or termination of dormancy (Bewley and Black 1994). This heterogeneity has been a key characteristic confounding research on seed dormancy (Bewley 1997) and little is known about its genetic basis (Squire et al. 1997). The qSD12 data (Figures 4 and 5) clearly demonstrated that genotypic differences of the QTL alone contribute, in part, to the germination heterogeneity in the population of seeds from the hybrid plant. Outcrossing species are characterized by heterozygosity, resulting in segregation at loci including dormancy genes in the next generation of seed population. Therefore, germination heterogeneity in a sample of seeds from the same plant or panicle cannot be simply attributed to environmental variation when the plant genotype is unknown. A multigenic system, such as the weedy rice parent, most likely consists of both offspring and maternal tissue-imposed dormancy genes (Figure 1), which function in early and late stages of the life cycle, respectively. In other words, the genes expressed in offspring tissues have an immediate dormancy effect on the seeds; while the genes expressed in maternal tissues have a dormancy effect on the seeds belonging to the following generation. In field conditions, outcrossing or gene flow occurs between crops and their conspecific weeds to produce hybrid seeds and fertile hybrid plants, such as the case between rice and weedy (or red) rice (Langevin et al. 1990; Gealy et al. 2003). It is expected that hybrid seeds on the weedy plant are more dormant than the reciprocal hybrid seeds on the crop plant, because of the differences in both maternal tissue and endosperm tissue genotypes between the hybrids. However, hybrid seeds on the crop plants may also survive the soil seed bank to be potential ‘‘volunteer weeds’’ in the following season. This potential exists because offspring tissue-imposed dormancy genes confer a certain degree of dormancy to the seeds even in the absence of maternal tissue-imposed dormancy genes such as qSD7-1. Similar to the red pericarp color gene in cereal crops that expresses not in the hybrid F1 seed but in the pericarp tissue of the F2 seeds, a maternal tissueimposed dormancy gene in the F1 generation would enhance overall dormancy of the F2 seeds. Therefore, dormancy genes functioning in maternal and offspring tissues provide the species with complementary adaptation strategies. Genetic analysis for dormancy has been conducted on the basis of the seed maternal (plant) genotypes almost without exception. A QTL analysis based on a nonpermanent segregation population (e.g., BC1 or F2) may bias estimates of genetic effects for the loci associated with seed dormancy imposed by offspring tissues, such as qSD12. This is because the phenotypic value for a segregate is the mean germination of its progeny seeds, which ignores the within-sample genotypic variation resulting from gene dominance and epistatic effects. Therefore, caution should be used to interpret the estimates for dormancy QTL when the gene nonadditive effects are significant. The qSD12 locus was narrowed to an 600-kbp region (Figure 2E) suitable for map-based cloning of the underlying gene. Molecular characterization of the major dormancy gene will unravel an underlying mechanism regulating germination in the offspring tissue(s). We thank C. Vogl for help in data analysis, and Y. Wang, B. Carsrud, S. Franklin, C. Doddivenaka, and K. Hagedorn for technical supports. Funding for this research was supported by grants from United States Department of Agriculture Cooperative State Research, Education, and Extension Service (SD00074-G), South Dakota Agriculture Extension Station (SD00H171-06IHG), and National Science Foundation (0641376). LITERATURE CITED Alonso-Blanco, C., L. Bentsink, C. J. Hanhart, H. B. E. Vries and M. Koornneef, 2003 Analysis of natural allelic variation at seed dormancy loci of Arabidopsis thaliana. Genetics 164: 711–729. Anderson, J. A., M. E. Sorrells and S. D. Tanksley, 1993 RFLP analysis of genomic regions associated with resistance to pre-harvest sprouting in wheat. Crop Sci. 33: 453–459. Bentsink, L., J. Jowett, C. J. Hanhart and M. Koornneef, 2006 Cloning of DOG1, a quantitative trait locus controlling seed dormancy in Arabidopsis. Proc. Natl. Acad. Sci. USA 103: 17042–17047. Bewley, J. D., 1997 Seed germination and dormancy. Plant Cell 9: 1055–1066. Bewley, J. D., and M. Black, 1994 Seeds: Physiology of Development and Germination. Plenum Press, New York. Cai, H. W., and H. Morishima, 2000 Genomic regions affecting seed shattering and seed dormancy in rice. Theor. Appl. Genet. 100: 840–846. Dong, Y., E. Tsozuki, H. Kamiunten, H. Terao, D. Lin et al., 2003 Identification of quantitative trait loci associated with preharvest sprouting resistance in rice (Oryza sativa L.). Field Crops Res. 81: 133–139. Flintham, J. E., 2000 Different genetic components control coatimposed and embryo-imposed dormancy in wheat. Seed Sci. Res. 10: 43–50. Foley, M. E., 1992 Effect of soluble sugars and gibberellic acid in breaking dormancy of excised wild oat (Avena fatua) embryos. Weed Sci. 40: 208–214. Gealy, D. R., D. H. Mitten and J. N. Rutger, 2003 Gene flow between red rice (Oryza sativa) and herbicide-resistant rice (O. sativa): implications for weed management. Weed Tech. 17: 627– 645. Gu, X.-Y., S. F. Kianian and M. E. Foley, 2004 Multiple loci and epistases control genetic variation for seed dormancy in weedy rice (Oryza sativa). Genetics 166: 1503–1516. Gu, X.-Y., S. F. Kianian, G. A. Hareland, B. L. Hoffer and M. E. Foley, 2005 Genetic analysis of adaptive syndromes interrelated with seed dormancy in weedy rice (Oryza sative L.). Theor. Appl. Genet. 110: 1108–1118. Gu, X.-Y., S. F. Kianian and M. E. Foley, 2006a Isolation of three dormancy QTLs from weedy rice as Mendelian factors. Heredity 96: 93–99. Gu, X.-Y., S. F. Kianian and M. E. Foley, 2006b Dormancy genes from weedy rice respond divergently to seed development environments. Genetics 172: 1199–1211. Gubler, F., A. A. Millar and J. V. Jacobsen, 2005 Dormancy release, ABA and pre-harvest sprouting. Curr. Opin. Plant Biol. 8: 183–187. Guo, L., L. Zhu, Y. Xu, D. Zeng, P. Wu et al., 2004 QTL analysis of seed dormancy in rice (Oryza sativa L.). Euphytica 140: 155–162. Han, F., S. E. Ullrich, J. A. Clancy and I. Romogosa, 1999 Inheritance and fine mapping of a major barley seed dormancy QTL. Plant Sci. 143: 113–118. Harlan, J. R., J. M. J. de Wet and E. G. Price, 1973 Comparative evolution of cereals. Evolution 27: 311–325. Offspring Tissue-Imposed Seed Dormancy International Rice Genome Sequencing Project, 2005 The map-based sequence of the rice genome. Nature 436: 793–800. Johnson, L. P. V., 1935 The inheritance of delayed germination in hybrids of Avena fatua and A. sativa. Can. J. Res. 13: 367–387. Kao, C.-H., and Z.-B. Zeng, 2002 Modeling epistasis of quantitative trait loci using Cockerham’s model. Genetics 160: 1243–1261. Langevin, S. A., K. Clay and J. B. Grace, 1990 The incidence and effects of hybridization between cultivated rice and its related weed red rice (Oryza sativa L.). Evolution 44: 1000–1008. Lan, X.-J., Y.-M. Wei, D. C. Liu, Z.-H. Yan and Y.-L. Zheng, 2005 Inheritance of seed dormancy in Tibetan semi-wild wheat accession Q1028. J. App. Genet. 46: 133–138. Lee, S.-J., C.-S. Oh, J.-P. Suh, S. R. McCouch and S.-N. Ahn, 2005 Identification of QTL for domestication-related and agronomic traits in an Oryza sativa 3 O. rufipogon BC1F7 population. Plant Breed. 124: 209–219. Li, C., A. Zhou and T. Sang, 2006 Genetic analysis of rice domestication syndrome with the wild annual species, Oryza nivara. New Phytol. 170: 185–193. Lin, S.-Y., T. Sasaki and M. Yano, 1998 Mapping quantitative trait loci controlling seed dormancy and heading date in rice, Oryza sativa L., using backcross inbred lines. Theor. Appl. Genet. 96: 997–1003. Lincoln, S., M. Daly and E. Lander, 1992 Constructing Genetic Maps with MAPMAKER/EXP 3.0, Ed. 3. Whitehead Institute, Cambridge, MA. Lyttle, T. W., 1991 Segregation distorters. Annu. Rev. Genet. 25: 511–557. McCouch, S. R., L. Teytelman, Y. Xu, K. B. Lobos, K. Clare et al., 2002 Development and mapping of 2240 new SSR markers for rice (Oryza sativa L.). DNA Res. 9: 199–207. Mo, H., 1988 Genetic expression for endosperm traits, pp. 478–487 in Proceedings of the Second International Conference on Quantitative Genetics. Sinauer Associates, Sunderland, MA. Morris, C. F., J. M. Moffatt, R. G. Sears and G. M. Paulsen, 1989 Seed dormancy and responses of caryopses, embryos, and calli to abscisic acid in wheat. Plant Physiol. 90: 643–647. Nair, V. G., B. W. X. Ponnaiya and V. S. Raman, 1965 Studies on seed dormancy in certain short-duration rice varieties. India J. Agric. Sci. 35: 234–246. Noll, J. S., P. L. Dyck and E. Czarnecki, 1982 Expression of RL4137 type of dormancy in F1 seeds of reciprocal crosses in common wheat. Can. J. Plant Sci. 2: 153–187. Oka, H. I., 1988 Origin of Cultivated Rice. Japan Scientific Society Press, Tokyo. Rha, E.-S., F.-S. Thseng and S.-T. Wu, 1994 Distorted segregation of Acp-1 for acid phosphase in a wide cross of rice. Rice Genet. Newsl. 11: 106. Romagosa, I., F. Han, J. A. Clancy and E. Ullrich, 1999 Individual locus effects on dormancy during seed development and after ripening in barley. Crop Sci. 39: 74–79. Rozen, S., and H. J. Skaletsky, 2000 Primer3 on the WWW for general users and for biologist programmers, pp. 365–386 in Bioinformatics Methods and Protocols: Methods in Molecular Biology, edited by S. Krawetz, and S. Misener. Humana Press, Totowa, NJ (http:// frodo.wi.mit.edu/cgi-bin/primer3/primer3_www.cgi). 2273 SAS Institute, 1999 Statview Reference Manual. SAS Institute, Cary, NC. Seshu, D. V., and M. E. Sorrells, 1986 Genetic studies on seed dormancy in rice, pp. 369–382 in Rice Genetics: Proceedings of the International Rice Genetics Symposium, edited by International Rice Research Institute. Manila, Philippines. Seshu, D. V., and M. Dadlani, 1991 Mechanism of seed dormancy in rice. Seed Sci. Res. 1: 187–194. Simpson, G. M., 1990 Seed Dormancy in Grasses. Cambridge University Press, Cambridge, UK. Squire, G. R., B. Marshall, G. Dunlop and G. Wright, 1997 Genetic basis of rate-temperature characteristics for germination in oilseed rape. J. Exp. Bot. 48: 869–875. Takahashi, N., 1963 Studies on the dormancy of wild rice seeds. Part 2. Role of seed coat, embryo and endosperm in dormant seeds, pp. 75–85 in The 241st Report of the Institute for Agricultural Research. Tohoku University, Sendai, Japan. Taylor, D. R., and P. K. Ingvarsson, 2003 Common features of segregation distortion in plants and animals. Genetica 117: 27–35. Temnykh, S., G. DeClerck, A. Lukashova, L. Lipovich, S. Cartinhour et al., 2001 Computational and experimental analysis of microsatellites in rice (Oryza sativa L.): frequency, length variation, transposon associations, and genetic marker potential. Genome Res. 11: 1441–1452. Thomson, M. J., T. H. Tai, A. M. McClung, X. H. Lai, M. E. Hinga et al., 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. Theor. Appl. Genet. 107: 479–493. Tinker, N. A., and D. E. Mather, 1995 MQTL: software for simplified composite interval mapping of QTL in multiple environments. J. Agr. Genomics 1: 2. Ullrich, S. E., P. M. Hayes, W. E. Dyer, T. K. Blake and J. A. Clancy, 1993 Quantitative trait locus analysis of seed dormancy in ‘Steptoe’ barley, pp. 136–145 in Pre-harvest Sprouting in Cereals, edited by M. K. Walker-Simmons and J. L. Ried. American Association of Cereal Chemists, St. Paul, MN. Vogl, C., and S. Xu, 2000 Multipoint mapping of viability and segregation distorting loci using molecular markers. Genetics 155: 1439–1447. Wan, J. M., L. Jiang, J. Y. Tang, C. M. Wang, M. Y. Hou et al., 2006 Genetic dissection of the seed dormancy trait in cultivated rice (Oryza sativa L.). Plant Sci. 170: 786–792. Wan, J. M., Y. J. Cao, C. M. Wang and H. Ikehashi, 2005 Quantitative trait loci associated with seed dormancy in rice. Crop Sci. 45: 712–716. Xu, Y., L. Zhu, J. Xiao, N. Huang and S. R. McCouch, 1997 Chromosomal regions associated with segregation distortion of molecular markers in F2, backcross, doubled haploid, and recombinant inbred populations in rice (Oryza sativa L.). Mol. Gen. Genet. 253: 535–545. Communicating editor: A. H. Paterson