Whole-genome Resequencing Of Gossypium Barbadense Accessions In Xinjiang Reveals Genetic Variation And Genes Associated With Fiber Traits

Sea-island cotton(Gossypium barbadense L.)is one of the most widely cultivated cotton in the world.It is well-known for the outstanding performance in fiber quality and disease resistance.Recently,genome sequencing has gradually developed into a new era with the characteristics of high-throughput,high-precision,and low-cost.The high-quality genome of G.barbadense had been assemblied and sequenced based on the advanced technology,which have made the fine mapping of G.barbadense based on whole genome resequencing come into reality.All these achievements porovide impetus for the utilization of G.barbadense genetic resources.Currently,Xinjiang is the only production region of sea-island cotton in China.Using the molecular markers to explore the genetic variation characteristics of independent G.barbadense breeding lines and identify the valuable genes for genetic improvement will accelerate the breeding process of excellent cotton cultibars.In the present study,a diversity panel consisting of 240 accessions of G.barbadense were collected to detect the genotype variations by resequencing.The ultradense genetic variation map with coverage of the entire G.barbadense genome was used to dissect the characteristics like population structure,linkage disequilibrium(LD)decay rate of the population,and genetic diversity of each group.Phenotypic variations of 12 important traits were evaluated across four agroecological environments and a GWAS was performed based on these data.It will provide important resources for the molecular biology research and genetic improvement of G.barbadense.The main contents of this work are as follows:1.220 independent breeding lines in Xinjiang,China and 20 accessions from other cotton production areas were resequenced.Totally,6.34 Tb original data were generated.After sequence alignment and population variation detection,3632231 highquality SNPs(Single nucleotide polymorphism)and 221354 In Dels(Insertion-deletion)were screened out.2.To quantify the population structure of these cultivars and breeding lines,the phylogenetic,principal component analysis(PCA),and population structure analyses were performed.It was suggested that the population could be clustered into five groups,the cultivars and breeding lines from Xinjiang consisted of four subgroups,and those outside of Xinjiang were clustered together with a certain genetic distance from Xinjiang accessions.It illustrated that Xinjiang sea-island cotton has gradually formed into a unique type.3.The phenotypic variation among different subgroups in this populaiton was abundant,especially for the fiber-quality related traits.The population genetics study revealed that the nucleotide diversity(π)was similar among different subgroups.The genetic diversity of foreign group was richer than that of Xinjiang independent cultivars and breeding lines.The LD decay throughout the G.barbadense genome was much higher than G.hirsutum and G.arboreum.The magnitude of LD decay different drastically between the At sub-genome and Dt sub-genome of G.barbadense with higher LD decay distance in the At.4.A total of 12 traits were evaluated across two locations for different two cropping seasons in 2016 and 2018.Phenotypic correlations and broad sense heritability(BSH)were calculated and the phenotypic variation were estimated.It showed that several tarit-pairs were positively correlated with each other in a significant level(p<0.01),including boll weight and seedcotton yield,boll weight and lint yield,plant height and node of first fruiting branch,plant height and boll weight,noll number per plant and cotton yield.On the contrary,node of first fruiting branch was negatively correlate with seedcotton yield,lint yield,lint percentage,and boll number per plant.Overall,the correlation among quality-related traits is stronger than that of yield component traits.The results of BSH showed that the lint percentage was the trait with the highest heritability(0.57)and the boll number of per plant had the lowest heritability with the value of 0.17.The heritability level of fiber quality-related traits is higher than that of yield-related traits,which illustrated that fiber yield-related traits are more susceptible to environmental factors.5.The best linear unbiased prediction(BLUP)values of 12 traits for the two locations over two years,as well as each location over two years,were used as phenotypic values for GWAS.Based on the P-value thresholds for significant loci identification(p<0.05/n),168 significant SNPs were screened.Based on the P-value thresholds for suggestive loci(p<1/n),2645 SNP signals were identified.These identified SNPs are unevenly distributed in the genome,with the largest number of SNPs on chromosome D11,followed by chromosome A07.More SNPs were found in Dt subgenome than that in the At subgenome.Fiber strength was the trait with the most associated signals,which was followed by the lint percentage.While,no significant associated signals were identified for seedcotton yield and the node of first fruit branch.Several reasons like the complexity of traits and the environmental factors may account for it.6.Pairwise linkage disequilibrium(LD)correlation,functional annocation and RNA-seq were used to narrow down the number of candicated genes.Then the expression analysis of candidate genes was carried out in G.barbadense accessions with different phenotypic characteristics using q PCR.Finally,three promising genes related to fiber strength were identified,including one gene predicted to encode a casein kinase I isoform delta-like protein(GB＿D11G3437)and two genes involving in microtubule cytoskeleton organization and microtubule-based process(GB＿D11G3460and GB＿D11G3471).One gene encoding a receptor kinase regulated by brassinosteroids was found to be associated with lint percentage(GB＿A07G1034).In summary,a population of sea-island cotton was genotyped using resequencing,and the characteristics like population structure,linkage disequilibrium,and genetic diversity were further explored.Moreover,12 important traits were identified to perform the GWAS.Among these traits,fiber strength and lint percentage were selected to be emphasized.Finally,two promising genes related to fiber strength and one candidate gene related to lint percentage were found out.It will provide important theoretical basis and target genes for genetic improvement of sea-island cotton.