Cytological Studies Of Hybrid Pollen Sterility And Fine Mapping Of Two Genes Causing Hybrid Pollen Sterility In Rice (Oryza Sativa L.)

Rice (Oryza sativa L.) is not only one of the most important food crops in the world, but also a model plant for studying the developmental biology of monocots, and more than half of the world’s population depends on it as main source of nutrition. Recently, due to the wide application of improved varieties and breeder’s preference for parents, the gene resources of rice were becoming more and more single and genetic basis was becoming more and more narrow, which seriously affected increase of rice yield. In order to enrich the genetic basis of rice and break the bottleneck in rice breeding, at present, the most effective way is to improve rice materials by introgression of valuable genes from distant species. Because weedy rice and Oryza glaberrima Steud. possed so many valuable genes, such as disease and insect resistance, saline tolerance, drought resistance and high temperature resistance. If we can introduce these valuable genes into Oryza sativa L., it will certainly bring a new leap for rice breeding. However, the reproductive isolation caused extreme sterility of F1 hybrids between Oryza sativa L. and weedy rice, or Oryza sativa L. and Oryza glaberrima Steud., which greatly limited transfer of favorable genes from weedy rice and Oryza glaberrima Steud. to the Oryza sativa L and the use of distant heterosis. In order to overcome interspecific hybrid sterility and take full use of the strong distant heterosis, it is necessary to more widely evaluate cytological mechanisms of hybrid sterility and find more hybrid sterility gene between Oryza sativa L. and weedy rice, or Oryza sativa L. and Oryza glaberrima Steud..In this study, we further investigate cytological mechanism of pollen abortion in F1 hybrid between the japonica wide compatibility rice cultivar 02428 and a weedy rice accession from Yunnan province. Genetic mapping in a BC1F1 population (02428//Yunnan weedy rice (YWR)/02428) showed that a major QTL for hybrid pollen sterility iqPS-1) was present on chromosome 1, which was also fine-mapped. Simultaneously, in order to explore the possible origins of YWR, a phylogenetic analysis of YWR, cultivated rice and wild rice based on microsatellite genotyping was carried out. Moreover, we developed an NIL at the locus S37 via repeated backcrossing and molecular marker-assisted selection (MAS), where the japonica variety Dianjingyou 1 was used as the receptor parent and O. glaberrima Steud. variety IRGC102295 as the donor parent. An F1 pollen semi-sterility locus, S37, was identified on rice chromosome 12 between NIL and Dianjingyou 1. We further elucidated cytological mechanism of pollen abortion in F1 hybrid between NIL and Dianjingyou 1, and also described the fine mapping and candidate-gene screening of S37. This study established a solid foundation for better understanding hybrid sterility and finally utilizing strong heterosis between indica and japonica subspecies.The main results were as follows:1. The pollen stain ability and in vitro germination tests revealed that both YWR and 02428 pollen fertility was normal, whereas their F1 hybrid showed clear pollen sterility. The type of pollen abortion contained typical abortion, spherical abortion, and stained abortion, wherein stained abortion was observed in most of aborted pollens. The two reciprocal F1, hybrids showed a similar level of pollen fertility to one another (P<0.05). Cytological studies have shown that pollen abortion in the F1 hybrid occured at the early bicellular pollen stage and probably occurred because the failure in the first mitosis prevented the formation of a functional reproductive nucleus. No abnormality in the development of tapetum or other anther walls was apparent. Aniline blue staining revealed that many pollen grains adhered to the stigmas and were able to germinate in the parental plants, but in the F1 hybrid, adherence and germination were restricted, and no pollen tubes were able to penetrate the style. Thus, the reduced spikelet fertility of F1 hybrids was the cumulative result of pollen abortion and poor stigma adherence of any remaining viable pollen, but the embryo sac of F1 hybrids appeared to be uncompromised.2. Of 805 SSR primers tested,313 were informative in the YWR×02428 population, and an outline linkage map based on 02428//YWR/02428 BC1F1 population was constructed from 133 of these. A QTL analysis performed using this map indicated the likely existence of two hybrid pollen sterility QTL, one (qPS-1) on chromosome 1 and the other (qPS-8) on chromosome 8. The location of the former was close to RM5, and this QTL accounted for～23% of the phenotypic variation for hybrid pollen sterility. qPS-8 accounted for～12% of the phenotypic variation and was linked to RM210. No interaction was detected between the two loci, which affected pollen sterility independently with additive effect, and qPS-1 was a major and stable hybrid pollen sterility loci.3. A total of 795 extreme individuals containing 403 plants from 02428//YWR/02428 population and 392 plants from KN/N22//KN population were genotyped using the flanking markers RM493 and RM5, respectively. As the results, qPS-1 was fine mapped between LI1 and LI14-1, about 110-kb in length on a single PAC clone (P0013G02). Gene prediction analysis of the 110-kb region showed that there were 27 putative open reading frames (ORFs) in this region, of which ORF9 and ORF10 encoded putative SaF+ and putative SaM, respectively. The sequencing results revealed that the qPS-1 locus is actually allelic with Sa.4. Dular and IR36 were assumed to carry the sterility-neutral allele, Sa", at Sa locus. SNPs in the two subgenes were examined in 23 wild species,9 weedy strains, and in 82 cultivars. The results indicated that a single substitution of "T" or "C" and "T"or "G" at gene of SaM and SaF, respectively, arose in wild rice, weedy rice, and indica cultivars, while only "C" and "T" was found at SaF and SaM, respectively, in japonica cultivars. The obtained dendrogram showed that the weedy rice was classified into indica or japonica type in cultivar (O. sativa L.), and the YWR was distributed among the wild rice and indica type, being separated from the japonica type. Accordingly, we extrapolated that YWR most probably originated from hybridization between Oryza sativa indica cultivars and Oryza rufipogon.5. We developed an NIL via repeated backcrossing and molecular marker-assisted selection (MAS), where the japonica variety Dianjingyou 1 was used as the receptor parent and O. glaberrima Steud. variety IRGC102295 as the donor parent. F1 pollen fertility of NIL/DJY1 exhibited typical semi-sterility, and the type of pollen abortion exhibited stained abortive. At the same time, we examined the DJY1 and F1 pollens using scanning electron microscopy and transmission electron microscopy and found the stained abortive pollens in F1 hybrid have small volume and most of them are shrunken, which have no accumulations of starch granules. Cytological studies have shown that pollen abortion in the F1 hybrid occured at the mature pollen stage and probably occurred because the failure in the second mitosis prevented the formation of a functional trinuclear pollen.6. According to pollen fertility, we constructed a linkage map covering the S37 region by 743 plants randomly selected from F2 progenies. The results indicated that the S37 locus was located to 2.2cM in the interval between NJ5 and G4. Finally, S37 was mapped to the region between HP 14 and G21 using a large F2 population contained 18014 plants, with physical size of about 73 kb. Thirteen open reading frames can be predicted by a sequence analysis of this fragment. Quantitative analysis of anther showed that LOC_Os12g02800 was the most likely candidate genes, which codes a cysteine-rich family protein precursor. Further complementation experiments and functional studies are ongoing.