| Cotton (Gossypium spp.) is the most important natural fiber crop worldwide, providing leading raw material for textile industry. Upland cotton (Gossypium hirsutum L.) accounts for over95%of world’s cotton fiber production. China is the world’s largest consumer of cotton fiber and textile industry plays a critical role in our national economy, working as a core export industry and creating numerous jobs. At present, the fiber quality traits of major commercial cotton cultivars generally can meet the requirements of textile industry in China, but the fiber length is limited to middle level, fiber strength is low and fiber fineness is poor. High yield and excellent fiber quality cultivars are in shortage and each year, a large amount of high quality cotton fiber has to be imported abroad. Thus, breeding high yield cotton cultivars with excellent fiber quality traits is very urgent for cotton breeders at present.Cotton fiber quality traits are controlled by multiple quantitative genes of minor effects. Traditional breeding strategy based on hybridization between existing elite cultivars with selection for high lint yielding recombinants has contributed to improvements in fiber quality. This method is time-consuming and of low efficiency and it has limited potential to improve both lint yield and fiber quality, because of the negative correlation between these traits and their quantitative inheritance. Besides, there are also complex negative correlations among cotton fiber quality traits. Thus, it is very difficult to balance the yield and fiber quality traits in breeding projects.Marker assisted selection (MAS) provides a quick and accurate selection method for breeders, aiming at favorable gene pyramiding and breaking the unfavorable linkages, which will increase the efficiency of breeding of high yield and high fiber quality cotton cultivars. MAS must be based on conditions:(1) elucidating the molecular mechanism of fiber yield and quality traits;(2) mapping the QTL controlling fiber yield and quality and cloning the favorable alleles;(3) developing specific molecular markers closely linked to favorable alleles, using as selection marker.In this study, two well konwn elite upland cotton cultivar\line Yumian1and7235in China, which are featured with high fiber quality traits, were crossed to establish a recombinant lines (RIL) population. Based on this population, QTL controlling fiber quality traits were analyzed to mine the favorable alleles on the genomes of two parents. Stable and high effect QTL are expected, which could be further used to study the molecular mechanism of fiber quality traits and applied to MAS breeding projects. Based on the QTL identified in RIL population, a secondary F2mapping population was established to further map the fiber quality QTL on Chr.10, targeting at narrowing the QTL confidence interval, which laid foundation for map-based cloning of these QTL in the future. The main results are as follows.1. PGML primer pairsG. raimondii Ulbr. is diploid cotton species of D5genome and is considered as the ancestor of Dt subgenome of allotetraploid cotton species. A total of5,000SSR primer pairs were designed according to G. raimondii BAC-ends sequences from Professor Andrew H. Paterson in the University of Georgia, and named as PGML. These primer pairs were used to increase the preciseness of genetic map and QTL mapping with RIL population. Analysis showed that pentanucleotides (29.26%) were the most common SSR motif type, followed by dinucleotides (18.54%), tetranucleotides (17.56%), trinucleotide (14.22%), mononucleotides (10.24%), hexanucleotides (9.94%), and others (0.22%). In total,299PGML primer pairs revealed polymorphism between parents, with polymorphism rate around6.0%and produced326polymorphic loci in RIL population. PGML loci were widely mapped on26chromosomes, indicating that these primer pairs were suitable for genetic map construction and QTL mapping of upland cotton.2. The genetic map based on RIL populationAmong25,313SSR markers screened,1,468showed clear polymorphism between the parents Yumian1and7235, with the polymorphism ratio of5.8%. These polymorphic primer pairs produced1,582polymorphic loci in RIL population. Genetic map construction was based on the1,582loci. In total,1,540loci were mapped on26chromosomes, covering2842.06cM, with average distance of1.85cM between adjacent markers. This genetic map accounted for528loci (1,333.8cM) with2.53cM interval distance in the At subgenome and1,012loci (1,508.3cM) with1.49cM interval distance in the Dt subgenome, respectively.Loci were not evenly distributed on chromosomes, ranging from a high of447loci on Chr.24, to a low of17loci on Chr.04. Chr.26had the longest recombination length (172.3cM) while Chr.24having the shortest (56.5cM). Fifty gaps (interval distance of>10cM) were identified on23chromosomes, and the largest gap on Chr.15(44.5cM). The average marker distance varied on different chromosomes, with the largest on Chr.03(5.96cM) and the smallest on Chr.24(0.13cM). 3. SSR loci segregation distortionAmong the1,582polymorphic SSR loci,662(41.8%) showed segregation distortion (P <0.05) with629(95.0%) favoring the Yumian1alleles and33(5.0%) favoring the7235alleles. Far more distorted loci were located on the Dt subgenome than on the At subgenome (582versus80), and Chr.24was heavily concentrated with distorted loci, with447loci showing significant distortion toward Yumian1, forming a large segregation distortion region (SDR). Beyond Chr.24,131(60.9%) of the remaining215distorted loci were clustered into16SDRs.The distorted loci in a given SDR skewed toward the same allele, showing strong "hitchhiking"effect.4. Fiber quality traits performanceFiber quality traits data across four years showed that the two parents had good fiber quality traits on the whole. Comparison between parents indicated that fiber length and micronaire of7235was superior to Yumian1, while for other traits, there was no significant difference. For RIL population, all traits were observed to segregate continuously. Skewness and kurtosis values showed that all traits were approximately in normal distribution. Transgressive segregations were also observed for all traits. Variance analysis based on four-year fiber quality traits data showed that fiber quality traits presented significant environmental and genetic effects (P<0.01). Correlations among fiber quality traits in different year indicated that fiber length, strength and uniformity were significantly positively correlated to each other; Fiber micronaire was generally negatively correlated to fiber strength and uniformity, while had inconsistent correlation with fiber length across years. Surprisingly, fiber elongation had significant converse correlations with all other traits across years.5. QTL for fiber quality traits based on RIL populationA total of61QTL, including18significant QTL and43suggestive QTL, were identified for five fiber quality traits with a range of11-14QTL per trait. The percentage of phenotypic variance explained by each QTL for each trait ranged from5.0%to28.1%. Among these QTL,25were detected using both combined analysis and single environment analysis and seven QTL were detected only by using combined analysis. Eight QTL (qFL10.1, qFL24.1, qFM23.1, qF24.1,qFE10.1, qFE24.1, qFS10.1and qFS24.1) could be detected in at least three years and these eight QTL were of high value for molecular mechanism research and application, including map-based QTL cloning and MAS. Different QTL were clustered on Chr.05, Chr.08, Chr.10, Chr.18 and Chr.24. For each trait, favorable alleles were conferred by both parents. Yumian1and7235conferred favorable alleles for17and42QTL respectively, except qFE10.1and qFE24.1which were conferred by different parents in different years.6. Abnormal phenomenon on Chr.24There are several abnormal phenomena on Chr.24in this study:(1) it is rich in polymorphic loci, containing447loci and all loci show significant segregation distortion toward Yumian1alleles;(2) it is a QTL "hotspot", mapping4QTL, including3very stable QTL;(3) it is the shortest chromosome in terms of recombination length, resulting from the reduced recombination frequency as revealed by the SSR loci, which lead to that QTL intervals on this chromosome have been enlarged to cover almost the whole chromosome. These abnormal phenomena further indicate that Chr.24of parent7235contains introgression segments from G. anomalum or G. barbadense as reported before.7. Mapping fiber quality QTL on Chr.10with secondary populationRIL-179, of which QTL interval on Chr.10was conferred by7235, was crossed with Yumian1to establish a F2population. This population comprised of1,038plants and was used as secondary population to further map QTL on Chr.10identified in RIL population. A total of1,663SSR primer pairs were developed according to the homologous sequence of QTL interval in G. raimondii genome and they were used to detect polymorphic SSR loci in QTL interval. In total,30primer pairs produced clear polymorphic loci in F2population. The local genetic map of QTL interval comprised of30SSR loci, covering34.6cM with adjacent loci distance1.57cM. The loci order on genetic map and physical map was highly consistent. Based on this local genetic map and fiber quality trait data of F2population, qFL10.1, qFS10.1, qFE10.1and qFM10.1were mapped to interval between SWU4439and SWU4772, with LOD value17.3-26.7and variance explained by11.6-17.3%. The qFM10.1was a new QTL that was not detected in RIL population. QTL1-LOD confidence intervals of these four QTL were highly overlapped and corresponded to a physical interval of1.2Mb on G. raimondii genome. This physical interval contains117putative genes according to the annotation information. This mapping work laid the ground work for map-based cloning of these QTL in the future. |
PDF Full Text Request