| XZM2 is a hybrid cotton variety with high yield heterosis. The parent CRI12 is an elite variety of high yield, good fiber quality and disease resistance and J8891 is a germplasm line with high yield potential. A recombinant inbred lines (RILs) population constructed from XZM2 and an immortalized F2 (IF2) population derived from crosses between the RILs were used to map QTL for agronomic and fiber quality traits; the genetic bases of heterosis were dissected by mapping of heterotic loci; and a transcriptome map of the IF2 population was constructed by cDNA-AFLP techniques, followed by QTL and candidate gene analyses for agronomic and fiber quality traits. The results are as follows:1. A genetic map mainly consisting of SSR markers was constructed using the RILs population, which was 941.2cM in recombined length, including 22 chromosomes and 1 linkage group, and covered 20.20% of cotton genome. The average interval between two markers was 6.67cM. Some of the markers were clustered on certain chromosomes.2. Based on data of RILs population in 8 environments in 3 years, QTL analysis was conducted using Composite Interval Mapping (CIM) procedure of Windows QTL cartographer 2.5. Totally 42 additively significant QTL for 7 yield and yield components were identified,18 and 24 of which were mapped on A and D subgenome respectively; Fifty nine QTL for 11 fiber quality traits were detected, with 22 distributed on A and 36 on D subgenome. Sixteen QTL for 7 morphological traits were mapped, with 1 on A and 15 on D subgenome. Some of these QTL were clustered on chromosome D2, D6, D9, A5 and A11, implying the existence of hot QTL region on the cotton genome.3. Utilizing the IF2 data in 6 environments in 3 years, QTL for 9 agronomic traits were analyzed by CIM and multiple interval mapping (MIM) of Cartographer 2.5 and multi-marker joint analysis methods respectively. Of the total 104 QTL detected by CIM, 37 can be detected in more than 2 environments,70 were also identified by MIM, and 43 were detected simultaneously by multi-marker joint analysis. Many of the QTL detected in different environments shared common makers. There were 28 relatively stable QTL concurrently detected by three methods, including 1 for seed-cotton yield; 1 for lint yield; 3 for bolls per plant; 5 for boll weight; 6 for lint percentage; 3 for seed index; 3 for lint index; 2 for plant height and 3 for fruit branch number respectively. About 26 QTL were at the same or close to the position of previously reported QTL of corresponding traits.Forty seven digenic epistatic QTL pairs were detected by MIM procedure. The number of QTL pairs involving in additive by additive (AA), additive by dominance (AD), dominance by additive (DA) and dominance by dominance (DD) interaction was 26,9,12 and 19 respectively, therefore the interaction modes between two loci were predominantly AA and DD, and some of the QTL pairs involved in more than two kinds of significant interaction. Some loci frequently interacted with other multiple loci, indicating there are hot regions of interaction in the cotton genome. Interaction between loci and environment for agronomic traits were analyzed using multi-marker joint analysis method. Totally 28 QE eptistatic QTL were identified including 10 on A and 18 on D subgenome. The results clearly demonstrated that the more susceptible to environment variations the traits, the more QE epistatic QTL were detected. Among these epistatic QTL,15 were also detected by CIM, suggesting that at least some of the QTL generally detected by CIM in previous researches were indeed QE epistatic QTL. Many digenic epistatic QTL pairs and QE epistatic QTL detected in this research implies the significance of epistasis in the constitution of heterosis in the hybrid cotton XZM2.4. QTL for mid-parent heterosis of the 9 agronomic traits were analyzed by multi-marker analysis method based on the data of IF2 in 4 environments. Totally 60 heterotic loci were detected, of which 19 were additive,31 were partial to complete dominant, and 10 were over-dominant. Forty three of these heterotic loci were found to be the QTL position of the corresponding traits, indicating these loci might indirectly affect the performance of heterosis by controlling the expression of traits. Sixteen pleiotropic QTL were found for heterosis of different traits, each simultaneously controlling heterosis of multiple traits. Seventy five QE epistatic heterotic QTL and 75 digenic epistatic QTL pairs were also detected by multi-marker analysis. That the heterosis variation explained by epistatic QTL was the largest among different gene action modes for all trait heterosis demonstrated that epistasis including digenic as well as QE interaction was the main genetic foundation of heterosis in XZM2. Additive, dominance and overdominance also played a role in heterosis.5. A transcriptome map of the IF2 was constructed via cDNA-AFLP techniques, using the top fully opened leaves during budding stage to extract RNA. The map was 2747.01cM in length, with 302 cDNA-AFLP markers distributed among 26 linkage groups. The average length of linkage groups was 95.27 cM, and the average interval between two markers was 8.23 cM. All markers distributed uniformly among linkage groups. Seventy six and 61 QTL were identified by CIM for 9 agronomic traits and 9 fiber quality traits respectively, using the transcriptome map and the 4-environment data. Some of the QTL were clustered on certain linkage groups. The majority of the QTL detected by this map were dominant and over-dormant.Most of the closely collocated or co-segregated candidate TDFs over 200bp in length were excised from the gels, successfully re-amplified and sequenced. Through homology searches, altogether 61 TDFs collocated with QTL of agronomic and fiber quality traits were detected with potential gene products or biological function. The putative gene products of these TDFs involved in controlling of transcription and translation, signal transduction, transport, cellulose synthesis, photosynthesis, metabolism of carbohydrate as well as lipid, and constitution components of cell structure. Several candidate genes for yield and yield heterosis were identified through correlation analysis between differential expression of TDFs and yield related traits. In the future, these functional candidate genes can be cloned by RACE or map based cloning, and the QTL associated TDFs can be converted into CAPS markers for the purpose of marker assisted selection (MAS). |
PDF Full Text Request