| Higher quality natural fibers are highly desirable for the current textile market to meet demands based on the increasing standard of living and advanced textile techniques.While,the sustained reduction of available arable land area further challenges the fiber yield of cotton.Therefore,it is an essential goal to develop novel cultivars with both more excellent fiber quality and higher yield in current cotton breeding.However,the simultaneous improvement of cotton fiber quality and yield-related traits（lint percentage）is seriously restricted by the significant negative correlations between them.As a result,it is very necessary to analyze the genetic basis of interactive development and phenotypic negative correlation between related traits of cotton fiber,and to mine favorable loci which can effectively achieve simultaneous improvement of cotton fiber quality and yield-related traits.Which are of important scientific interest to break through the bottleneck of current breeding.In this study,a series of related studies were carried out by using the chromosome segment substitution lines（CSSLs,or introgression lines）,which were generated by hybridization and high generation backcrossing between Gossypium hirsutum（Gh）,with high-yield and wide-adaptability,and Gossypium barbadense（Gb）with excellent fiber quality.Two sets of CSSLs with the different Gh genetic backgrounds,including 408 BC5F5and 332 BC4F5,were generated by introgressing the chromosome segments from Hai1（Gb）into CCRI36（Gh）and CCRI45（Gh）,respectively,and evaluated in multiple environments by our team.In this study,after the phenotypic comprehensive evaluation of CSSL,we improved genotyping strategies and carried out related mining of simultaneous improvement introgression segments based on population genetics using the simplified genome sequencing data of CSSLs populations and the deep resequencing data of parents.Furthermore,a new analysis strategy was proposed based on mixed pool sequencing and transcriptome sequencing,and was used to carry out the related efficient identification of important genes with simultaneous improvement of multiple traits.The main results are as follows:1)Phenotypic identification of CSSLs populations.Phenotypic identification of CCRI36 and CCRI45 CSSLs was carried out in 6 and 8 environments,respectively,including fiber quality related traits（fiber length（FL）,fiber strength（FS）and micronaire（FM））and yield related traits（lint percentage（LP）,boll weight（BW）,seed index（SI）and lint index（LI））.The normal distribution of phenotypic values was observed from each trait in each environment,and series of high heritability（0.85-0.93）was observed for each trait.The trait pairs（FS and FL,FM and LP）showed significant medium or high positive correlations in most of the environments;the trait pairs（FL/FS and FM,FL/FS and LP）showed significant negative correlations in most of the environments;the trait pairs（FL/FS/FM and BW）showed no or weak positive or negative correlations in most of the environments.The averages of fiber quality related traits were better than those of their respective recurrent parents,and the averages of fiber yield related traits were similar to those of their respective recurrent parents.However,both the coefficient of variations and the averages of transgressive rate over the recurrent parent of each trait in CCRI45 genetic background were generally higher than those in CCRI36 genetic background.The results showed that the fiber quality related traits of Gh were significantly improved by the abundant genetic variations,which were produced by the interspecific introgression of Gb into Gh.In addition,the fiber quality and yield were simultaneously improved in some introgression lines,but the improvement effects were more significant in the genetic background of CCRI45.Twelve excellent lines with excellent FL/FS and high LP were screened as valuable resources from the CCRI45 population,which were more stable than those 3 from CCRI36 population.2)Improving of genotyping strategy and genotyping of CSSLs populations.The 740CSSLs of two populations were sequenced by simplified sequencing（SLAF-seq）with an average sequence depth of about 1×,and the genomes of three parents were re-sequenced with a sequence depth of more than 45×.First,45,343 and 46,181 high-quality SNP were detected in CCRI36 and CCRI45 populations,respectively,based on the routine SNP genotyping strategy.However,the results showed that the distribution of SNP in the whole genome was extremely uneven and many large-gaps were identified on most chromosomes.Then the kmer genotyping strategy was improved and used to checked kmer variants in 2sets of CSSLs,which considering the characteristics of the parental genetic population and the genomic characteristics of the introgression lines.As a result,68,912 and 83,352introgressed kmers with high quality were obtained in CCRI36 and CCRI45 populations,respectively,of which 29,390 were shared in the two populations.All introgression kmers were completely anchored to the reference genome by step-by-step mount to determine their physical positions,which improved the previous kmer genotyping strategy.The superiority of kmer variants were highlighted through the comparisons of the final kmer and SNP in many aspects,such as the number,genome distribution and the influence on the genetic relationship between samples.The introgression flow was identified by mapping kmers to the reference genomes of Gh and Gb,respectively.It was found that more than 90%of the introgressions occurred between the homologous chromosomes of Gh and Gb,and about 21%of the segments were inverted introgressions.The kmer-binmaps were constructed based on the genotypes of each introgression kmer,and 2277and 2320 recombinant bins were identified in CCRI36 and CCRI45 populations,respectively.The average length of the recombinant bins was about 1Mb,indicating that the exchanges and recombinations of chromosome segments in the CSSLs were sufficient.3)Mining of introgression segments for simultaneous improvement of fiber quality and yield.The association analysis was carried out using the genotypes of the kmer-binmaps and the phenotypic data of 7 traits.A total of 1849 QTL were detected in the two populations,of which 277,313,232,241,249,233 and 304 were associated with FL,FS,FM,LP,BW,SI and LI,respectively.There were 160 common QTL between 2 populations,and a total of 1267（68.52%）QTL could be verified by previous results.Furthermore,353introgressed segments with stable genetic effects were identified,of which 129 were beneficial for the improvement of fiber quality related traits,including 66（22.98%）for FL,64（21.99%）FS and 35（12.03%）for FM;and of which 48 were beneficial for the improvement of fiber yield related traits,including 31（10.65%）for LP and 18（6.19%）for BW.Thirty-nine introgression segments were detected with stable favorable additive effects for simultaneous improvement of 2 or more traits in Gh genetic background,of which 8 were beneficial to release/limit the negative correlation between FL/FS and LP,including 6 that could stably increase FL/FS and LP.In addition,the pyramiding effects and application potential of 3 pleiotropic segments were further validated in the segregating population,which was developed by the backcross of excellent line MBI7561and its recurrent parent CCRI45.Finally,3 additional genetic segments with the potential for stable simultaneous improvement of FL/FS and LP were identified by the comprehensive mining of the reported results in previous studies.4)A new strategy of"Comparison Analysis for Multiple Traits Graded Pool-Seq"（CAMTGS）was proposed for efficient mining of genetic resources for multiple traits,and used to identify key genes which regulate the simultaneous improvement of FS and LP.Four groups of double-traits graded pools were constructed focused on the FS and LP of CCRI45 population,since more stable excellent lines could be screened and more introgression segments favoring the simultaneous improvement of FL/FS and LP could be identified in the CCRI45 population.And 1845 candidate QTGs related to the comprehensive development of FS and LP were identified by CAMTGS analysis.These candidate QTGs could fully verify the stable genetic pleiotropy St-Clusters of FS and LP that identified in the QTL mapping of CCRI45 population,which showed the relatively high detection power on the identification of genetic resources for simultaneous improvement of FS and LP.Thus,the weighted co-expression network analysis（WGCNA）was carried out focus on the 1845 candidate QTGs,and the transcriptome sequencing data was from the recurrent parents and 4 excellent introgression lines（with excellent FS and high LP）at six fiber development stages（DPA:5,10,15,20,25,28）.It was identified that the 3 key development stages related to the simultaneous improvement of FS and LP were5DPA,15DPA and 25DPA.And 15,5 and 13 hubgenes,which played important roles in regulation,were identified in the corresponding stages,respectively.Of which 8,3 and 6genes were proved in previous studies,respectively.The results showed that the combination of CAMTGS and WGCNA had the advantages of high noise reduction efficiency and strong analytical power.The most important functional genes may include RING/U-box superfamily proteins,S-adenosyl-L-methionine-dependent methyltransferases superfamily proteins,Late embryogenesis abundant proteins（LEA）and cytochrome P450 family proteins.In this study,it was the first time to combine kmer genotyping strategy with bi-parental genetic population in cotton.The kmer genotyping strategy was improved,which inherits the advantages of capturing and characterizing more variation types and is suitable for the genetic analysis of bi-parental genetic populations with reference-based genomes,which provides a new analysis method for population genetics related studies.A new analysis strategy of"CAMTGS"was proposed for mining genetic resources of multi-traits interaction development,which provides an intensive and efficient study idea for mining simultaneous improvement genetic resources of multi-traits.It was also the first time to systematically identify introgression segments and genetic materials related to simultaneous improvement of cotton fiber quality and yield traits,which provides rich and novel reference resources for promoting design breeding based on multi-omics and realizing simultaneous improvement of cotton fiber quality and yield traits. |
