| 其他摘要 | During the past several tens of thousands of years, human civilizations began and thrived with the domestication of plants and animals. Artificial selection is the internal driver that triggered species domestication. Under human selection, wild species were finally changed into domesticated species. The first domesticated plants or animals were called landraces, which have relatively poor agronomic characteristics, for example low yield and poor eating quality. As the expansion of human population and human needs, breeders began to improve landraces, which finally gave rise to many elite varieties with elite agronomic performances. The improvement process constitutes the second stage of artificial selection. These elite varieties probably contain some elite alleles that account for the elite traits. Traditionally, people use QTL mapping to identify those elite varieties. This approach has been proved to be useful and succeeded in identifying a series of important agronomic genes, such as Gn1a controlling seed number, GS5 and GW3 influencing seed size, and DEP1 conferring the erect panicle trait, etc. However, QTL mapping is labor and time consuming, taking several years to construct the segregating population. In the recent years, more and more people take advantage of association analysis to identify agronomic genes, but this approach has little power in identifying rare elite alleles. In this dissertation, we proposed a new approach for guiding mining rare elite alleles in elite varieties, which we named ETAS analysis. Using this approach, we identified an important ETAS in the famous upland rice variety, IRAT104. Functional analysis shows that this ETAS allele increases the ABA level and enhances the root development in upland rice. This adaptive phenotype might probably be the reason why it was selected for upland agriculture. One of the main purposes of artificial selection is to enhance crops’ adaption to specific cultural environments. So far, breeders have been able to breed many upland rice varieties that adapt to dry land cultivation. However, it remains an open question that whether upland or irrigated rice emerge first in the evolution. Moreover, the genetic mechanisms underlying dry land adaption of upland rice have been hardly investigated. In second part of this study, we sequenced the genomes of 84 upland rice accessions and 82 irrigated rice accessions sampled around the world. By incorporating the genomic data of wild rice sequenced previously by our lab, we revealed the whole genome phylogenetic relationship between upland, irrigated and wild rice accessions. Our results show that upland rice probably came from improvement using irrigated japonica as basic materials. Furthermore, we did population genetics analysis by comparing the upland and irrigated rice genomes thereby to identify some ecotype differentiated regions (EDRs). Within these EDRs, we identified 154 ecotype differentiated genes (EDGs), many of which could be linked to the phenotypic differentiation between upland and irrigated rice. Among these EDGs, we were very interested in one gene, Os09g0410500, annotated as “similar to teosinte branched one (tb1)”. In our study, we named Os09g0410500 as Ostb2 so as to differentiate it from another homolog, Ostb1, on rice chromosome03. Tiller branching is an important trait that has been shaped radically during artificial selection of gramineous crops. The reduction of branches seems to be the convergent direction of artificial selection in many crops. For example, the tillering ability of maize, sorghum, pearl millet and foxtail millet all decrease after domestication. Fewer tillers could enhance the yield of single panicle and simultaneously permit close planting, therefore increase production per acre. During the domestication of maize, sorghum, pearl millet and foxtail millet, tb1 locus all contributes to their tiller decrease. In our study, we focus the tb1 homolog in rice, Ostb2, and study it |
修改评论