Study On Gene Interaction Involving Hybrid Sterility And Wide Compatibility In Rice

Rice is the staple food for about half of the world population, and also the model plant of Gramineae research. With the increase of world population and the decrease of arable land, yield per unit area of rice should be elevated, to meet the consuming needs. Asian cultivated rice (Oryza sativa L.) comprises two subspecies, indica and japonica rice, and inter-subspecific hybrids are usually sterile. Yield of hybrid rice in our country is presently stepping into the plateau phase, and inter-subspecific heterosis utilization is one way to significantly increase the yield of hybrid rice. Though inter-subspecific hybrid rices have stronger heterosis than intra-subspecific ones, sterility accompanying inter-subspecific hybrid limits effective utilization of its strong heterosis. Wide compatibility varieties (WCVs) can produce fertile hybrids when crossed with both indica and. japonica rice, which bring hope for utilization of inter-subspecific heterosis.The S5and f5loci, which affect fertility of inter-subspecific hybrid significantly, were identified to control hybrid sterility (HS) and wide compatibility of female and male gametes respectively. To dissect the genetic mechanism of S5and f5thoroughly, we developed comprehensive near isogenic lines (NILs) of these two loci:WCVs as the locus donor, indica and japonica as the recipients; indica as the locus donor, WCVs and japonica as the recipients; japonica as the locus donor, indica and WCVs as the recipients. Genetic analysis of NILs showed that segregation of indica/japonica allele at S5(under japonica/japonica or indica/japonica genomic background) and f5(under japonica/japonica genomic background) didn’t fit Mendel ratio, demonstrating significant segregation distortion (SD). S5i/S5j or f5i/f5j hybrid preferentially transmitted indica-type gametes to the progeny, accompanying the abortion of japonica-type gametes. WCVs showed no SD when crossed with both indica and japonica rice. Hybrid sterility caused by S5and f5resulted from the death of japonica gametes, leading to SD.Early studies ascribed control of S5on HS to single locus. Our group previously mapped the S5locus to a region of40kb in sixth chromosome, with five predicted genes (ORF1～ORF5). Transgenic plants with ORF5+from indica rice Nanjing11transformed into Balilla, a japonica variety, showed complete spikelet sterility, due to embryo-sac abortion. Since transformants of ORF5mimicked phenotype of indica/japonica hybrids, our group previously took ORF5as the causative gene of HS at S5locus. In our present research, heterozygous NILs with S5fragment from Nanjing11introgressed into Balilla showed spikelet semi-sterility. If only ORF5participates in fertility control at S5region, introgression of both ORF5and S5into Balilla should result in same fertility phenotype. The fertility discrepancy between ORF5+transformants and S5NILs suggested that there are still other genes at S5region, in addition to ORF5, participate in fertility control. Results from animal showed that gain-of-function hybrid SD is controlled by at least two closely linked genes, which bring new thread to our study. Since S5demonstrated single-locus genetic behavior, HS at S5is controlled either by single gene or by several closely linked genes.We compared genomic sequences of ORFs1-5from indica, japonica and WCVs, and detected functional polymorphisms in ORFs3-5. To probe relationship among ORF3, ORF4and ORF5, we transformed them into the NILs and Balilla respectively. Genetic analyses of the progenies from crosses of various combinations among the NILs and transgenic lines uncovered a killer-protector system regulating HS and SD at S5. This system consists of a killer, a partner and a protector encoded by three tightly linked genes (ORFs3-5).ORF4and ORF5act sporophytically, whereas ORF3act gametophytically. In indica/japonica hybrid, the killer (ORF5+) and partner (ORF4+) work together to kill female gametes not carrying the functional protector (usually japonica gametes), resulting in preferential transmission of gametes (usually indica gametes) carrying the functional protector (ORF3+), causing segregation distortion in the progeny. WCVs have no killing effect and can protect itself, so WCVs produce fertile hybrid when hybridized with both indica and japonica rice.Microarray data showed that the joint action of ORF5+and ORF4+causes endoplasmic reticulum (ER) stress, and ORF3+can prevent ER stress and produce normal gametes. But ORF3-cannot prevent ER stress, resulting in premature programmed cell death and eventually leads to embryo-sac abortion.Genetic diversity analysis of S5showed that ORF3+/ORF4+/ORF5+is the ancestral type and also the haplotype with the highest ratio in cultivated rice and common wild rice. Typical japonica haplotype (ORF3-/ORF4+/ORF5-) was found in extant common wild rice with the highest latitude from Dongxiang County of China, which means that Chinese common wild rices are probably the founder population for domestication of japonica rice.Based on published results at present, we proposed the simplest SD model for hybrid sterility and wide compatibility of rice. In this model, indica rice carries toxin and antitoxin genes;japonica rice carries non-toxin and non-antitoxin genes; while WCV carries non-toxin and antitoxin genes.The influence of genomic background on SD Can not be ignored. S5i/S5j hybrids showed SD under japonica/japonica or indica/japonica genomic background, but no SD under indica/indica background. Occurrence of SD at S5probably needs an enhancer in the japonica background. f5i/f5j hybrids showed SD under japonicaljaponica genomic background, but no SD under indicalindica or indicaljaponica background. Disappearance of SD at f5probably resulted from a repressor in the indica background. In short, for SD research, please keep your eyes on not only S5or f5locus but also the genomic background.Former researchers suggested that wide compatibility at S5locus of rice was controlled by single gene (wide compatibility gene, WCG), but our results showed that this so-called WCG is in fact a gene team composed of three genes. Previously published SD systems all regulate male gametes; S5is the firstly cloned SD system that regulates female gametes. Toxin in other species comprised either single gene (Drosophila melanogaster) or several additive genes (Mus musculus); whereas toxin comprising two nonadditive genes like that at S5has not been found in other species.Our results characterized molecular mechanism of hybrid sterility and wide compatibility in rice, which will provide reference to breeding of inter-subspecific hybrid rice. These results would also enhance our understanding of the origin and nature of indica and japonica rice.