Plant and Cell Physiology Advance Access originally published online on August 12, 2008
Plant and Cell Physiology 2008 49(9):1283-1293; doi:10.1093/pcp/pcn118
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Rapid Paper |
Inference of the japonica Rice Domestication Process from the Distribution of Six Functional Nucleotide Polymorphisms of Domestication-Related Genes in Various Landraces and Modern Cultivars
1 Plant Genome Research Unit, National Institute of Agrobiological Sciences, 2-1-2 Kannondai, Tsukuba, Ibaraki, 305-8602 Japan
2 QTL Genomics Research Center, National Institute of Agrobiological Sciences, 2-1-2 Kannondai, Tsukuba, Ibaraki, 305-8602 Japan
*Corresponding author: E-mail, tizawa{at}nias.affrc.go.jp. Fax, +81-29-838-7446
 |
Abstract |
|---|
 Top Abstract IntroductionResultsDiscussionMaterials and MethodsFundingAcknowledgmentsReferences |
|---|
Crop domestication can serve as a model of plant evolutionary processes. It involves a series of selection events from standing natural variation and newly occurring mutations and combinations of mutations as a result of natural crossings in populations during local adaptation and propagation of plant lines to other cultivation areas. Our earlier identification of three functional nucleotide polymorphisms (FNPs) of distinct genes involved in the rice domestication process led us to propose a model of the japonica rice domestication process. Here, we examined three more FNPs in two domestication-related genes involved in pigment synthesis during the development of seed pericarp color (Rc and Rd) in 91 landraces (and some modern cultivars) of japonica rice collected from throughout the area of distribution of rice. These polymorphisms were assigned by using genome-wide patterns of restriction fragment length polymorphisms (RFLPs) and the local origins of the landraces. The results led us to infer the process of japonica rice domestication in more detail and propose a more refined model of the japonica domestication process. In this model, the critical role of the Rc FNP at an early step of the domestication process was highlighted. Independent artificial selections of two defective Rd alleles were found, suggesting a role for Rd other than in pigment synthesis during rice domestication.
Keywords: Domestication - Evolution - Genome - RFLP - Rice (Oryza sativa)
Abbreviations: bHLH, basic helix–loop–helix; DFR, dihydroflavonol-4-reductase; FNP, funcional nucleotide polymorphism; ORF, open reading frame; RFLP, restriction fragment length polymorphism; SINE, short interspersed element
|
Introduction |
|---|
|
TopAbstractIntroductionResultsDiscussionMaterials and MethodsFundingAcknowledgmentsReferences |
|---|
As Darwin (1857
) noted, crop domestication resembles a rapid evolutionary process that results from artificial selection but that otherwise has all the characteristics of evolution by means of natural selection. With increasing availability of crop genome information, the accumulation of knowledge about domestication-related genes, and the identification of functional nucleotide polymorphisms (FNPs), the crop domestication process is being increasingly elucidated (Doebley et al. 2006). In maize, for example, extensive genome analysis to find genes with reduced natural variation among cultivars, landraces and their wild relatives has suggested that thousands of genes might have been subjected to selection during domestication (Gaut et al. 2000, Matsuoka et al. 2002, Yamasaki et al. 2005). In wheat, archeological analysis of plant remains has revealed that the domestication process took a few thousand years (Tanno and Willcox 2006, Dubcovsky and Dvorak 2007). In barley, natural variation among landraces from Europe to Asia has been examined to reveal how barley cultivation propagated throughout these regions, with some genome changes detected and two distinct domestication processes proposed for European and Asian barley (Morrell and Clegg 2007, Pourkheirandish and Komatsuda 2007, Saisho and Purugganan 2007). In addition, some domestication-related genes have been cloned, and key natural variations in these genes (i.e. FNPs) have been examined to elucidate the domestication process in several crops, including maize, wheat and barley (Wang et al. 2005, Doebley et al. 2006, Simons et al. 2006, Komatsuda et al. 2007), even though the domestication process itself remains largely unknown.
In rice (Oryza sativa), extensive genome analysis has revealed that the two major subspecies (japonica and indica) have relatively distinct genomes that may have coalesced into a common ancestor species >0.2–0.4 million years ago, indicating that at least two independent domestication processes occurred in rice (Ma and Bennetzen 2004, Vitte et al. 2004). Furthermore, the japonica subspecies may have a monophyletic origin with a group of wild relatives, Oryza rufipogon, whereas the origin of the indica subspecies is not yet clear in terms of the genome variation among the landraces of O. sativa and O. rufipogon accessions that have been identified from short interspersed element (pSINE) retroelement insertions (Cheng et al. 2003). The japonica subspecies comprises at least two subgroups, tropical japonica and temperate japonica (Garris et al. 2005, Caicedo et al. 2007). A recent genome analysis revealed that tropical japonica may exist as intermediate between O. rufipogon and temperate japonica in the phylogenetic tree, although archeological data do not yet support this hypothesis (Cheng et al. 2003).
Several domestication-related rice genes have been cloned, and FNPs have been identified for these genes. For instance, Waxy (Wx) encodes a granule-bound starch synthase that is required for amylose synthesis in rice grains (Hirano et al. 1998, Isshiki et al. 1998). The activity of Wx affects the texture and taste of cooked rice. There are two major Wx alleles (Wxa and Wxb). A single nucleotide polymorphism at the junction of the first exon and intron confers the difference between Wxa and Wxb due to a defect in the efficiency of splicing in Wxb (Hirano et al. 1998; Isshiki et al. 1998). However, this allele has been found in the majority of japonica rice on the basis of a haplotype analysis, suggesting that this allele was favored during the japonica rice domestication process (Olsen and Purugganan 2002). In order to simplify the story, Wxa is referred to as Wx, and Wxb is referred as wx in the latter part of this paper. Another gene, sh4, encodes a possible MYB3-type transcription factor; the defective allele reduces seed shattering and was thus selected during domestication (Li et al. 2006, Lin et al. 2007). The FNP for this gene was identified as an amino acid change in a conserved region in the sh4 open reading frame (ORF). Interestingly, this amino acid change has been identified in all tested landraces and cultivars in both japonica and indica. How this FNP was incorporated in all cultivated rice is unclear, since both subspecies have different histories of domestication. Therefore, several models to explain the distribution of this allele have been proposed (Sang and Ge 2007, Sweeney and McCouch 2007).
Recently, we cloned two additional domestication-related genes, qSH1, which is required for formation of the abscission layer during seed shattering, and qSW5, which controls seed grain width; we also identified their FNPs (Konishi et al. 2006, Shomura et al. 2008). The FNPs for Wx, qSH1 and qSW5 have been integrated into a genome variation map constructed on the basis of similarity in the restriction fragment length polymorphism (RFLP) pattern at 179 loci throughout the rice genome in hundreds of rice landraces and cultivars. This integrated map, combined with information on the local origins of rice landraces, elucidated the process of rice domestication and led us to infer a model of the japonica rice domestication process (Shomura et al. 2008). This model described several key events, such as the selection of newly occurring natural variation, selection of a combination of FNPs after natural crossings and propagation of various lines into various cultivation areas, following local adaptation with gradual genome fixation.
Recently, researchers have extensively analyzed another domestication-related gene (Rc), which encodes a basic helix–loop–helix (bHLH) transcription factor controlling several key enzymes that are active during pigment synthesis and the development of rice pericarp color (Sweeney et al. 2006, Sweeney et al. 2007). A major defective allele of Rc might have originated in the ancestor of japonica rice and subsequently introgressed into the majority of indica rice. Furthermore, two independent natural mutations in another gene related to pigment synthesis [Rd, which encodes a key enzyme, dihydroflavonol-4-reductase (DFR)] have also been reported among japonica rice (Furukawa et al. 2007). Therefore, we tried to integrate the FNP information for both Rc and Rd into the integrated map. In total, we integrated six FNPs of five domestication-related genes (i.e. Wx, qSH1, qSW5, Rc, Rd1 and Rd2) and genome-wide RFLP data at 179 loci, combined with local origin information for 91 japonica rice landraces including some modern cultivars, in the map. The creation of this integrated map led us to correct our previous model and propose a more refined model of japonica rice domestication, in which the critical role of Rc deficiency and strong selection for Rd after the rc mutation had been incorporated into rice landraces suggest a role for Rd deficiency unrelated to pigmentation.
|
Results |
|---|
|
TopAbstractIntroductionResultsDiscussionMaterials and MethodsFundingAcknowledgmentsReferences |
|---|
Recently, we proposed a model of japonica rice domestication that was based on an integrated map constructed from information on three FNPs for domestication-related genes of rice (Wx, qSH1 and qSW5), local origin information and RFLP variation patterns at 179 loci distributed throughout the rice genome (Shomura et al. 2008
). In this model, we found that some landraces still contain all the original alleles of the domestication-related genes at these three FNP positions, and we named these lines ‘heritage landraces’. These results, together with knowledge of the origins of landraces that possess the mutated FNPs, strongly suggested that the local origin of japonica rice was in Indonesia, the Philippines or a southern part of Indochina.
In the present study, we observed seed grains of these heritage landraces and found that the colors of the rice seed pericarp were segregated among the existing heritage landraces (Fig. 1), suggesting that another FNP in a domestication-related gene involved in pigment synthesis had already been partly selected among these landraces. It has recently been reported that a 14 bp deletion in Rc, which encodes a bHLH transcription factor that controls pigment biosynthesis, was involved in the rice domestication process (Sweeney et al. 2006, Sweeney et al. 2007). Furthermore, natural mutations in Rd, another pigment synthesis gene that encodes a key enzyme, have been found in japonica rice cultivars (Furukawa et al. 2007). Therefore, we analyzed the FNPs in Rc and Rd so that we could incorporate this new information in our integrated map of japonica rice domestication. We amplified PCR fragments, including FNP regions of 91 rice landraces (including some modern cultivars) (Supplementary Table S1; hereafter we call them cultivars) and five accessions of O. rufipogon (W1943, W1948, W1953, W1958 and W1976), which were previously reported as accessions very close to japonica rice based on pSINE distribution in the rice genome (Cheng et al. 2003), and judged the genotype by using electrophoresis for the Rc FNP and by sequencing the two Rd FNPs (Fig. 2, Supplementary Table S1). Note that the two FNPs we detected in Rd (Rd1 and Rd2) would have behaved independently during the rice domestication process, since any recombination event between the FNPs was very unlikely because of the distance of around 250 bp between them. In total, we obtained information on the six FNPs for the five domestication-related genes for 91 cultivars of japonica rice (Supplementary Fig. S1) and the five accessions of O. rufipogon. First of all, the results indicated that all the six FNPs were the original types in the O. rufipogon accessions, except for qSW5 (Table 1). For the qSW5 gene, the other haplotypes in addition to the Kasalath type, but not the Nipponbare type were observed. Therefore, it is very likely that these six FNPs were not standing variations.
|
|
|
The number of possible FNP genotype combinations totals 48, which equals three-quarters of two raised to the sixth power because of the lack of recombination between the two Rd FNPs. However, in 91 cultivars, we found only 17 genotypes on the basis of combinations of the six FNPs (Table 2, Supplementary Fig. S2).
|
These results suggest that many FNP combinations have not existed in the japonica rice landraces and that only some of the FNP combinations were selected during rice domestication and remained after the selections. Therefore, it should be possible to infer the selection process in japonica rice domestication from these results.
With this new FNP information, we reduced the original 10 rice heritage landraces that retained all the original alleles of the domestication-related genes at the three FNP positions to four landraces with no mutations in any of the six FNPs including the three new ones. The origins of the remaining four landraces were Indonesia, the Philippines, Vietnam and Bhutan (Table 2, Fig. 3). Since only the RFLP genome pattern for the Bhutan landrace differed from those of the other three landraces, this landrace appears to be the product of relatively recent crossing or migration. In contrast, the genome patterns of the other three landraces were similar. Therefore, this result is consistent with our previous model (Shomura et al. 2008), in which the local origin of japonica rice was hypothesized to be an area around Indonesia, the Philippines or possibly Indochina (Fig. 3).
|
First, we focused on single-mutant cultivars among the 91 tested cultivars. When we searched for these cultivars among the 48 possible FNP combinations, we found cultivars with a single mutated FNP only for qSW5 and Rc (three and four cultivars, respectively; Table 2). We could not find any cultivars with a single mutation in the other three genes, strongly suggesting that plants with a mutation in either qSW5 or Rc were subjected to artificial selection at the beginning of the japonica rice domestication process. The local origins of the rc single-mutant cultivars were Indonesia and Malaysia, whereas the origins of the qsw5 single-mutant cultivars were Indonesia, the Philippines and China. The RFLP genome patterns for the four new heritage landraces and the single-mutant cultivars were similar, strongly suggesting that natural mutations in both qSW5 and Rc occurred around the local origins of these cultivars. The qsw5 single-mutant landrace of Chinese origin has a genome pattern close to that of the other three, suggesting recent migration to China. These results also support the hypothesis that Indonesia, the Philippines and possibly Indochina were the local origins of japonica rice (Table 2, Fig. 3).
We next focused on double-mutant cultivars, and we found three combinations of mutated FNPs. We found one double-mutant landrace for rc wx, two for qsw5 qsh1 and nine for qsw5 rc (Table 2). The limited number of single-mutant cultivars strongly suggested that the qsw5 rc double-mutant cultivars were produced by natural crossing between ancestor plants with qsw5 and rc single mutations. The origins of the qsw5 rc double-mutant cultivars were in Bangladesh, Sri Lanka, the Philippines, Myanmar and China (Table 2). These origins overlapped with those of the qsw5 or rc single-mutant cultivars, and the ancestors further propagated into Bangladesh and Sri Lanka (Fig. 3). In addition, the RFLP genome patterns of the double-mutant cultivars were similar to those of the single-mutant cultivars, with some exceptions of cultivars according to the distance between local origins. Therefore, it is very likely that the plants with the qsw5and rc single mutations were crossed to produce the ancestors of the qsw5 rc double-mutant cultivars and that the resulting plants spread into neighboring countries such as Bangladesh and Sri Lanka through Indochina to become the local qsw5 rc double-mutant cultivars (Fig. 3). The other double-mutant cultivars (rc wx and qsw5 qsh1) were rare, suggesting that they did not contribute much to the japonica rice domestication process (Supplementary Fig. S3). The local origins were Bhutan for rc wx and China for qsw5 qsh1 (Table 2). These RFLP patterns were also distinct from those of the single-mutant cultivars of qsw5 or rc, indicating that their genomes changed as a result of natural crossings between relatively distant early intermediates of japonica rice (Fig. 3). Therefore, a major event in japonica rice domestication might have been a natural crossing between qsw5 and rc single-mutant ancestor lines at their local origins. In addition, three qsw5 rc cultivars that originated in Bangladesh contain genome patterns that are clearly distinct from those of other japonica cultivars (Fig. 3, Supplementary Fig. S4). As an explanation for this difference, crossings between japonica-type and indica-type ancestor lines were involved in this genome diversity, leading to introgression of the qsw5 and rc mutations into the Bangladesh cultivars.
We further focused on triple-mutant cultivars and found five combinations. We found 15 triple-mutant cultivars for qsw5 rc wx, five for rc wx rd2, three for qsw5 rc rd1, two for qsw5 rc rd2, and one for qsw5 rc qsh1 (Table 2). Of these, four of five FNP combinations contained the qsw5 rc double mutation (Table 2). Therefore, it is possible that those triple-mutant cultivars resulted from a third mutation in some qsw5 rc double-mutant ancestor lines. It is very likely that the qsw5 rc wx triple-mutant cultivars were produced by selection of the wx mutation occurring in qsw5 rc double-mutant ancestor lines, since most of these triple-mutant cultivars originated in Indochina and the RFLP genome patterns gradually diverged from those of the qsw5 rc double-mutant cultivars with the same or adjacent local origins (Fig. 3). This finding for the wx mutation is not consistent with our previous model (Shomura et al. 2008). We will discuss this difference later. When we considered FNP variations for three genes (Rc, qSW5 and Wx), the RFLP variations of rice cultivars with local origins in southern Asia, including the Philippines, Indonesia, Indochina, Bangladesh and Sri Lanka, could be explained well (Fig. 3). In other words, the selected FNPs in qSH1 and Rd were rarely found in japonica cultivars with these local origins (Table 2). Of the other triple-mutant cultivars, rc wx rd2 (five cultivars), qsw5 rc qsh1 (one), qsw5 rc rd1 (three) and qsw5 rc rd2 (two) were less popular than the qsw5 rc wx triple-mutant cultivars (15). Furthermore, genome RFLP patterns of four minor triple-mutant cultivars were diverse among the same triple-mutant cultivars (Supplementary Fig. S5). Therefore, it may be difficult to derive a concrete model of whether the third mutation occurred in rc qsw5 double-mutant cultivars from only the genome patterns. We found that it is possible that the rd1 mutation occurred in qsw5 rc double-mutant cultivars (or qsw5 rc wx triple-mutant cultivars) that originated in the Philippines and was transferred into temperate japonica through Japanese upland rice cultivars, since the rd1 mutation was observed in only two cultivars that originated in the Philippines, several cultivars of Japanese upland rice and many temperate japonica cultivars found only in Japan (Supplementary Fig. S6). If this hypothesis is not correct, then we cannot explain the loss of rd1 in cultivars that originated in Indochina and China. Note that Japanese upland rice is believed to have originated from Southern countries, but not from China and Korea.
We further focused on tetrad mutant cultivars and found four combinations of the six FNPs among these cultivars. We found 10 tetrad mutant cultivars for qsw5 rc wx rd2, seven for qsw5 rc wx rd1, four for qsw5 rc rd2 qsh1 and one for qsw5 rc wx qsh1. Of these, three of the four combinations contained the qsw5 rc wx triple mutation (Table 2). Therefore, it is possible that these tetrad mutant cultivars were produced by a fourth mutation in some qsw5 rc wx triple-mutant ancestor lines. It is very likely that the qsw5 rc wx rd2 tetrad mutant cultivars were produced by selection of plants with the rd2 mutation from the qsw5 rc wx triple-mutant ancestor lines, since the RFLP genome patterns gradually diverged from the qsw5 rc wx mutant cultivars with neighboring local origins (Supplementary Fig. S7).
Our previous model proposed that the qsh1 mutation occurred in a qsw5 Wx background of Chinese origin (Shomura et al. 2008). Our new genotype data instead suggest that it is likely that the qsh1 mutation occurred in a qsw5 rc Wx rd2 background. This is consistent with the previous model, but adds more data on the ancestor. Here, we found that four qsh1 cultivars of Chinese origin were from the qsw5 rc Wx rd2 qsh1 background whereas one landrace of Chinese origin was from the qsw5 rc Wx Rd2 qsh1 background (Supplementary Fig. S8). Earlier in the Results section, we inferred that the rd2 mutation occurred in a qsw5 rc wx background, but the new data suggest that the qsh1 mutation might have occurred in the qsw5 rc Wx rd2 background. We cannot currently explain how an ancestor with the qsw5 rc Wx rd2 background was produced, although it would be possible to produce such background ancestor lines by natural crossings in China based on the genotypes of the analyzed cultivars.
All of the cultivars of temperate japonica of Japanese origin contain the qsw5 rc wx genotype (Fig. 4). On the other hand, the qsh1 and rd1/rd2 mutations are segregating in these cultivars (Fig. 4). Many of the qsw5 rc wx genotypes contain either rd1 or rd2 mutations, although they have distinct origins. Genome-wide RFLP analysis did not support a refined population structure associated with the rd1 or rd2 mutation. Therefore, any defect in the Rd gene might have been favored in temperate japonica cultivars in Japan.
|
|
Discussion |
|---|
|
TopAbstractIntroductionResultsDiscussionMaterials and MethodsFundingAcknowledgmentsReferences |
|---|
Local origin of japonica rice
Domestication of a crop was thought to be a kind of rapid change and was considered as an event occurring in the wild species. One now also considers it as a process. Several pieces of genetic data including our data supported this explanation (Sweeney et al. 2007
, Shomura et al. 2008). Tanno and Willcox (2006) proposed a model in which the domestication of wheat had taken >3,000 years. On the other hand, there might be a critical event at the beginning of crop domestication during the relatively long selection term as we described here. To elucidate not only the event at the beginning, but also the whole process of rice domestication, all the FNP changes identified in domestication-related genes should be examined as in depth as possible. To this end, we recently demonstrated that the local origin of japonica rice during domestication would have been in Indonesia, the Philippines and possibly southern Indochina, as determined from genome-wide RFLP data and the FNP distributions for three genes (qSW5, Wx and qSH1; Shomura et al. 2008). Here, we integrated information on three more FNPs in two domestication-related genes (Rc and Rd), both of which are involved in the development of color pigmentation in the rice seed pericarp, which is a domestication trait. With the combined information on the six FNPs, we again clearly demonstrated this local origin. An important point is that in this latest study we not only examined the local origins of heritage landraces that possessed all the original and functional FNPs, but we also extended the analysis to the local distribution of single-mutant cultivars (here, of qsw5 and rc). Analyses of archeological data have so far suggested that China's Yangtze River could have been the origin of japonica rice during domestication (Vaughan et al. 2008). In addition, the local origin of tropical japonica has long been an enigma. Our data demonstrate that many cultivars of tropical japonica rice contain many of the original alleles of domestication-related genes that existed before artificial selection, and some tropical japonica rice facilitate the same selected FNPs with temperate japonica. Therefore, we hypothesize that temperate and tropical japonica rice share a local origin. Although many tropical japonica rice lines are currently grown as upland rice, it is noteworthy that our data do not prove that the ancestor of japonica rice was cultivated as upland rice. Since all tested japonica-like O. rufipogon accessions did not carry any ‘selected FNPs’ (Table 1), the FNPs might have occurred as mutations in ancient early-intermediates of rice. Although it is likely that there would be partial sharing of chromosomal segments with the wild ancestral populations (and/or recent gene flow) among two major subspecies of rice (Gao and Innan 2008), according to our data this sharing might be events before the selection processes of these FNPs.
Selection of the rd mutation in temperate japonica rice in Japan
Our process suggests that the rd mutation occurred twice independently during japonica rice domestication. The qsh1 mutation might have occurred in China and was associated with the rd2 mutation. This is suggested by the genetic linkage between qsh1 and rd2, which are both located on chromosome 1 and are 
2.0 Mbp (10 cM) distant. On the other hand, the rd1 mutation may have occurred in the Philippines and was subsequently transferred into the ancestors of Japanese upland rice. Since some cultivars have both qsh1 and rd1 mutations in temperate japonica cultivars grown in Japan, some natural crossings must have produced these cultivars. Most temperate japonica cultivars in Japan (29 out of 31) contained either rd1 or rd2. Of these, eight cultivars contained the qsh1 rd1 double mutation that required recombination between qsh1 and rd loci during natural crossings. This clearly indicates that the defective Rd alleles have been repeatedly selected during the development of japonica rice cultivars in Japan. Since those cultivars already contained the rc mutation, an agronomic trait other than seed pericarp color would have been selected to explain this. Rd encodes a key enzyme for phenylpropanoid products and could thus be involved in some form of plant disease resistance (Furukawa et al. 2007). Note that the sd1 gene, which is related to plant stature and was selected during the ‘Green Revolution’, is also located in this region of chromosome 1 (Nagano et al. 2005). Therefore, selection for another gene rather than for the defective rd mutation might have been responsible for the rd qsh1 combination, although it is not easy to explain how both the rd1 and rd2 mutations were selected during domestication.
Limitation for modeling the domestication process
We previously proposed a model on the basis of three FNPs (qSW5, Wx and qSH1) in which multiple natural crossings were hypothesized to explain the observed genome-wide RFLP variations. In this work, we proposed a new model in which the wx mutation was considered to be a third mutation that occurred in qsw5 rc double-mutant ancestor landraces (Fig. 5). Therefore, by incorporating information on the rc FNP, one of the natural crossings between qsw5 and wx lines in our previous model can be explained by a third mutation that occurred after a natural crossing between rc and qsw5 lines. This shows how the model can be refined as we obtain more data. This example of using data on rc and wx improves the previous model greatly, since analysis of the additional data can rule out some of the formerly possible selections and retain only the more likely selections.
|
In contrast, by considering the origin of the qsh1 mutation, we hypothesized a second mutation in a qsw5 Wx line in China in the previous model. Here, we hypothesize that the qsh1 mutation occurred in an qsw5 rc Wx rd2 line, which is consistent with the previous one. However, we also hypothesize that the rd2 mutation was a fourth mutation in an qsw5 rc wx line. Therefore, we must hypothesize a natural crossing between qsw5 rc Wx Rd2 and qsw5 rc wx rd2, although existing cultivars do not support the production of an qsw5 rc Wx rd2 mutant line before the qsh1 mutation occurred (Fig. 5). Note that there are two cultivars that contain the qsw5 rc Wx rd2 genotype, but their local origins and genome-wide RFLP patterns did not fit well with the hypothesized ancestors. In this case, the new data did not support one clear possible model more strongly than others.
As we have described here, to propose a new model in which the most likely process should be proposed to explain the data, we facilitated our modeling of the rice domestication process by using six FNPs from 91 japonica cultivars. Accordingly, the new information improved the model, but more information on other cultivars is needed to let us derive a completely consistent model of rice domestication. Although we have recently introduced such modeling for crop domestication as a form of plant evolution (Shomura et al. 2008), balancing the amount of genotype data with the sample size (here, the number of cultivars) is an important factor in developing a good model. Here, we were able to propose a more refined model for japonica rice domestication by using data for six FNPs, unlike the previous model, which was based on only three FNPs. However, it is noteworthy that the use of only three FNPs in our previous model for the same 91 cultivars led us again to produce a likely and reasonable model, which is the same as the previous one (Supplementary Fig. S9, Supplementary Table S3). The difference between these models arises mainly from whether their genome divergences are hypothesized to be derived from natural crossings between related ancestor lines or gradual fixation of segregating loci in local populations. The cause of these divergences could be local adaptation due to climate change, differences in cultivation style or taste preference, and other factors. Since, in some cases, it is difficult to distinguish between crossings of relatives and genome-wide fixation of loci from information on a limited number of existing cultivars, what we can do is to produce the most likely model by using only the currently available experimental data. Therefore, the reliability of the model developed by this approach should be tested again with additional phylogenetic, archeological, paleogeographic and anthropological evidence. In addition, the validity of our rice collection to investigate this type of analysis should be re-examined in the future. However, selection of landraces and accessions based on genome diversity may just over-represent the minor genome compositions in terms of domestication history, which may not be useful for this type of analysis. Therefore, other natural variations for agronomic traits such as plant size, flowering time and seed size, and grain number in a panicle, would be better to select rice lines and make a representative collection for this purpose. Broad and even distribution of local origins throughout Asia is another key point of this collection. Here, we originally utilized 332 lines which were subjected to the genome-wide RFLP analysis for these analyses (Kojima et al. 2005, Shomura et al. 2008) and focused on 91 japonica cultivars. These lines were selected based on coverage of a large range of natural variations of agronomic traits in these rice cultivars with widely distributed local origins. Therefore, we believe that our collection is a representative collection of rice landraces and cultivars. In addition, it would be important to examine the same FNPs for the different collections such as those described previously (Garris et al. 2005, Caicedo et al. 2007, Sweeney et al. 2007).
We demonstrated here that the integration of FNP data for domestication-related genes in various cultivars with genome-wide and neutral RFLP variations and information on the biogeographic origins of the landraces revealed several key events during the process of japonica rice domestication. This example shows how we can address the evolution of a plant species by considering current DNA variation as a kind of historical record. This approach should provide many insights into the molecular process of plant evolution.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionResultsDiscussionMaterials and MethodsFundingAcknowledgmentsReferences |
|---|
Plant materials
We used genome DNA from the 91 japonica cultivars (Kojima et al. 2005
) in the NIAS gene bank (see Supplementary Table S1).
Genotyping of six FNPs in five domestication-related genes
For the Rc genotype, the presence of a key 14 bp deletion was detected by gel electrophoresis after PCR amplification of the corresponding region with the primers listed in Supplementary Table S2.
For the Rd and qSH1 genotypes, the PCR products were sequenced with the primers described in Supplementary Table S1 to identify the rd1 and rd2 genotypes.
For the qSW5 and Wx genotypes, the FNPs were detected by gel electrophoresis after PCR amplification of the corresponding region with the primers described in Supplementary Table S2.
Clustering of rice cultivars on the basis of RFLP patterns
The numbers of identical RFLP genomes among 179 loci distributed over all 12 chromosomes in a pair of rice cultivars were determined as the genome distance between the cultivars. We analyzed these pairwise genome distances by the pvclust clustering software in an R package for hierarchical clustering. The dendrogram obtained was arranged by eye to reduce inconsistency. The clustering was performed using 91 cultivars as described previously (Shomura et al. 2008).
Supplementary material
Supplementary material mentioned in the article is available to online subscribers at the journal website www.pcp.oxfordjournals.org.
|
Funding |
|---|
|
TopAbstractIntroductionResultsDiscussionMaterials and MethodsFundingAcknowledgmentsReferences |
|---|
The Integrated Research Project for Plants, Insects and Animals Using Genome Technology (grant No. GD2008) to T.I.; the Ministry of Agriculture, Forestry and Fisheries of Japan; Genomics for Agricultural Innovation project (grant No. QTL5001) to T.I.; Grants-in-Aid for Scientific Research Priority Areas ‘Genome Barrier in Plant Reproduction’ to T.I.
|
Acknowledgments |
|---|
|
TopAbstractIntroductionResultsDiscussionMaterials and MethodsFundingAcknowledgmentsReferences |
|---|
We thank Shomura of the Institute of the Society for Techno-Innovation of Agriculture, Forestry and Fisheries for performing the qSW5 and Wx genotyping, and Kojima of Toyama Agricultural Research Center for providing the genome-wide RFLP data.
|
References |
|---|
|
TopAbstractIntroductionResultsDiscussionMaterials and MethodsFundingAcknowledgmentsReferences |
|---|
Caicedo AL, Williamson SH, Hernandez RD, Boyko A, Fledel-Alon A, et al. Genome-wide patterns of nucleotide polymorphism in domesticated rice. PLoS Genet. (2007) 3:1745–1756.[Web of Science][Medline]
Cheng C, Motohashi R, Tsuchimoto S, Fukuta Y, Ohtsubo H, Ohtsubo E. Polyphyletic origin of cultivated rice: based on the interspersion pattern of SINEs. Mol. Biol. Evol. (2003) 20:67–75.
Darwin C. The Variations of Animals and Plants under Domestication. (1857) New York: D. Appleton.
Doebley JF, Gaut BS, Smith BD. The molecular genetics of crop domestication. Cell (2006) 127:1309–1321.[CrossRef][Web of Science][Medline]
Dubcovsky J, Dvorak J. Genome plasticity a key factor in the success of polyploid wheat under domestication. Science (2007) 316:1862–1866.
Furukawa T, Maekawa M, Oki T, Suda I, Iida S, Shimada H, Takamure I, Kadowaki K. The Rc and Rd genes are involved in proanthocyanidin synthesis in rice pericarp. Plant J. (2007) 49:91–102.[CrossRef][Web of Science][Medline]
Gao L-Z, Innan H. Nonindependent domestication of the two rice subspecies, Oryza sativa ssp. indica and ssp. japonica, demonstrated by multilocus microsatellites. Genetics (2008) 179:965–976.
Garris AJ, Tai TH, Coburn J, Kresovich S, McCouch S. Genetic structure and diversity in Oryza sativa L. Genetics (2005) 169:1631–1638.
Gaut BS, Le Thierry d’Ennequin M, Peek AS, Sawkins MC. Maize as a model for the evolution of plant nuclear genomes. Proc. Natl Acad. Sci. USA (2000) 97:7008–7015.
Hirano HY, Eiguchi M, Sano Y. A single base change altered the regulation of the Waxy gene at the posttranscriptional level during the domestication of rice. Mol. Biol. Evol. (1998) 15:978–987.[Abstract]
Isshiki M, Morino K, Nakajima M, Okagaki RJ, Wessler SR, Izawa T, Shimamoto K. A naturally occurring functional allele of the rice waxy locus has a GT to TT mutation at the 5' splice site of the first intron. Plant J. (1998) 15:133–138.[CrossRef][Web of Science][Medline]
Kojima Y, Ebana K, Fukuoka S, Nagamine T, Kawase M. Development of an RFLP-based rice diversity research set of germplasm. Breeding Sci. (2005) 55:431–440.[CrossRef]
Komatsuda T, Pourkheirandish M, He C, Azhaguvel P, Kanamori H, Perovic D, et al. Six-rowed barley originated from a mutation in a homeodomain-leucine zipper I-class homeobox gene. Proc. Natl Acad. Sci. USA (2007) 104:1424–1429.
Konishi S, Izawa T, Lin SY, Ebana K, Fukuta Y, Sasaki T, Yano M. An SNP caused loss of seed shattering during rice domestication. Science (2006) 312:1392–1396.
Li C, Zhou A, Sang T. Rice domestication by reducing shattering. Science (2006) 311:1936–1939.
Lin Z, Griffith ME, Li X, Zhu Z, Tan L, Fu Y, Zhang W, Wang X, Xie D, Sun C. Origin of seed shattering in rice (Oryza sativa L.). Planta (2007) 226:11–20.[CrossRef][Web of Science][Medline]
Ma J, Bennetzen JL. Rapid recent growth and divergence of rice nuclear genomes. Proc. Natl Acad. Sci. USA (2004) 101:12404–12410.
Matsuoka Y, Vigouroux Y, Goodman MM, Sanchez GJ, Buckler E, Doebley J. A single domestication for maize shown by multilocus microsatellite genotyping. Proc. Natl Acad. Sci. USA (2002) 99:6080–6084.
Morrell PL, Clegg MT. Genetic evidence for a second domestication of barley (Hordeum vulgare) east of the Fertile Crescent. Proc. Natl Acad. Sci. USA (2007) 104:3289–3294.
Nagano H, Onishi K, Ogasawara M, Horiuchi Y, Sano Y. Genealogy of the ‘Green Revolution’ gene in rice. Genes Genet. Syst. (2005) 80:351–356.[CrossRef][Web of Science][Medline]
Olsen KM, Purugganan MD. Molecular evidence on the origin and evolution of glutinous rice. Genetics (2002) 162:941–950.
Pourkheirandish M, Komatsuda T. The importance of barley genetics and domestication in a global perspective. Ann. Bot. (2007) 100:999–1008.
Saisho D, Purugganan MD. Molecular phylogeography of domesticated barley traces expansion of agriculture in the Old World. Genetics (2007) 177:1765–1776.
Sang T, Ge S. Genetics and phylogenetics of rice domestication. Curr. Opin. Genet. Dev. (2007) 17:533–538.[CrossRef][Web of Science][Medline]
Shomura A, Izawa T, Ebana K, Ebitani T, Kanegae H, Konishi S, Yano M. Deletion of a gene associated with grain size increased yields during rice domestication. Nat. Genet. (2008) 40:1023–1028.[CrossRef][Web of Science][Medline]
Simons KJ, Fellers JP, Trick HN, Zhang Z, Tai YS, Gill BS, Faris JD. Molecular characterization of the major wheat domestication gene Q. Genetics (2006) 172:547–555.
Sweeney M, McCouch S. The complex history of the domestication of rice. Ann. Bot. (2007) 100:951–957.
Sweeney MT, Thomson MJ, Cho YG, Park YJ, Williamson SH, Bustamante CD, McCouch SR. Global dissemination of a single mutation conferring white pericarp in rice. PLoS Genet. (2007) 3:1418–1424.[Web of Science]
Sweeney MT, Thomson MJ, Pfeil BE, McCouch S. Caught red-handed: Rc encodes a basic helix–loop–helix protein conditioning red pericarp in rice. Plant Cell (2006) 18:283–294.
Tanno K, Willcox G. How fast was wild wheat domesticated? Science (2006) 311:1886.
Vaughan DA, Lu BR, Tomooka N. The evolving story of rice evolution. Plant Sci. (2008) 174:394–408.
Vitte C, Ishii T, Lamy F, Brar D, Panaud O. Genomic paleontology provides evidence for two distinct origins of Asian rice (Oryza sativa L.). Mol. Genet. Genomics (2004) 272:504–511.[CrossRef][Web of Science][Medline]
Wang H, Nussbaum-Wagler T, Li B, Zhao Q, Vigouroux Y, Faller M, Bomblies K, Lukens L, Doebley JF. The origin of the naked grains of maize. Nature (2005) 436:714–719.[CrossRef][Web of Science][Medline]
Yamasaki M, Tenaillon MI, Bi IV, Schroeder SG, Sanchez-Villeda H, Doebley JF, Gaut BS, McMullen MD. A large-scale screen for artificial selection in maize identifies candidate agronomic loci for domestication and crop improvement. Plant Cell (2005) 17:2859–2872.
(Received July 17, 2008; Accepted August 8, 2008)
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  T. Yamamoto, J. Yonemaru, and M. Yano Towards the Understanding of Complex Traits in Rice: Substantially or Superficially? DNA Res, June 1, 2009; 16(3): 141 - 154. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||