Development Of Chromosome Segment Introgression Lines From G. Arboreum Into G. Hirsutum And Fine-mapping Of A Fiber Strength QTL From G. Hirsutum Race Yucatanense

As a major natural textile fiber crop,cotton(Gossypium spp.)occupies an important economic status in the world.G.hirsutum(2n=4x=52,AADD)(Gh)is widely cultivated due to its high yield properties,and it accounts for more than 95%of worldwide cotton production.The genetic diversity of G.hirsutum was greatly reduced due to the excessive use of limited cotton resources for artificial breeding.To continue to broaden the genetic basis of G.hirsutum,it is necessary to transfer elite genes of diploid cotton into G.hirsutum through breaking the interspecific isolation.G.arboreum(2n=2x=26,A2)(Ga),the putative progenitor of At subgenome of G.hirsutum,is an important source for broadening the genetic diversity of Gh due to its strong boll setting,high single fiber strength and resistance to biotic/abiotic stress.Introgressing these elite traits into Gh will be of great significance to the genetic improvement.In previous study,we overcame the cross-incompatibility between Ga and Gh through an improved embryo rescue technique,and successfully obtained synthetic amphiploid(2n=78,AADDA2A2),providing the foundation for genome-wide introgression from Ga into Gh.In present study,G.hirsutum,TM-1,was used as the recurrent parent to continuous backcross with the amphiploid,and develop the first Ga chromosome segment introgression lines(ILs)in the Gh background.In addition,a single segment introgression line(SSIL),IL-D2-2,deriving from TM-1×TX-1046,fiber strength(FS)of which was significantly improved,was selected to construct a secondary separation population to carry out study fine-mapping of QTL for FS and candidate gene analysis.The main results were as follows:1.Development of introgression lines and introgression analysisBased on the high-quality reference genomes,401 pairs of primers were developed and synthesized,whose products showed clear polymorphism between parents,which was used to construct the physical map of the current study and for molecular identification.In this study,289 Ga introgression lines in Gh background were developed.The introgression analysis showed the total coverage length of the introgressed segments was 1116.29 Mb,covering 78.48%of the At-subgenome in Gh background.The highest coverage rate was on All(99.71%)and the lowest on A01(39.83%).The coverage rates on A06-A13 were all>80%.The introgression segments from Ga in each IL and on each chromosome are shown in Fig.5 and Table 3.Among 289 ILs,there were 50 ILs contained a single introgressed segment.Total length of all introgressed segments was 11181.49 Mb.The total length of introgressed segments on each chromosome was different,the shortest was 88 Mb on Chr.A02 while the largest was 3531.37 Mb on Chr.A12.The length of each segment varied from 0.23 Mb(Chr.A03)to 94.96 Mb(Chr.A06),with an average length of 9.74 Mb.We found that reciprocal translocations involving A01,A02,A03,A04 and A05 strongly hindered introgression and inversions between chromosomes easily led to large-segment introgressions.It is the first report that chromosomal structural differentiation has huge effects on genome-wide introgression in cotton.2.QTL mapping of yield-related and fiber quality traitsBased on 11 traits in three environments,there were 47 co-QTLs detected through QTL mapping analysis(RSTEP-LRT)and association analysis(MLM).A total of 47 co-QTLs were detected in this study,of which 32 were related to yield and 15 were related to fiber quality.There were five co-QTLs for BW,among which two markers,INTR0176 and INTR0563,were associated with BW detected in three environments by two methods,which were stable and reliable QTL.There are seven co-QTLs associated with the SI,of them the marker INTR0373 was detected by both methods.Only two co-QTLs were related to LP,of them marker INTR0091 was detected in the E3 environment by two methods,while INTRO 180 was detected in the two environments by RSTEP-LRT.There were six co-QTLs related to PH,FBN and BN,respectively.Of them,INTRO 176 was stable and reliable,which could be detected to be related with PH in three environments by two methods;INTRO 177 was only related with PH by MLM method and could be detected in all three environments;and INTR0088 was related to BN detected in two environments by MLM method.There were two co-QTLs respectively,being related to FL and FS.INTR0046 was related to FS detected by both methods,and detected in both environments of MLM method.There were four co-QTLs respectively,being related to MIC and FU,and three co-QTLs were related to FE.Among them,INTR0495 was detected to be related to MIC by two methods,but it could be detected by MLM method in both environments.INTR0101 and INTR0160 were related to MIC detected by RSTEP-LRT and MLM in two environments.Compared with other QTLs,co-QTLs have higher reliability,especially those detected in multiple environments by two methods,such as marker loci INTR0176 and INTR0563 for BW;and marker locus INTR0176 for PH.In addition,based on additive effect calculated by RSTEP-LRT method,we found there were 21 QTLs related to yield showing positive additive effect,most of them were related to PH(4/8),FBN(5/7),BN(5/8),suggesting that Ga has potential to improve Gh yield by increasing BN.In this study,we found five QTL clusters with 26 QTLs on four chromosomes.Of them,A07-cluster,A08-cluster-1,A08-cluster-2 and A12-cluster were mainly related to yield traits;A05-cluster was mainly related to fiber quality.The GO and KEGG analysis of candidate genes in QTL clusters were carried out.For QTL clusters related to yield traits,most of genes were significantly enriched in nitrogen compound transport,regulation process and metabolic process in Gh and Ga;In molecular function category,most genes were significantly related to proteases activity,playing an important regulatory role in cellular activity in Gh and Ga.For QTL clusters related to fiber quality,most of genes were significant related to phosphorylation pathway,including ATP binding and phosphatase activity in Gh and Ga.The results showed that the gene functions of the homologous genomic regions of Gh and Ga were mostly the same.However,there was also different function enrichment between Gh and Ga.For example,most genes of QTL clusters related to yield traits were enriched in generation of precursor metabolites and energy,while not found in Ga,suggesting that there were some differences on gene function between Gh and Ga.The functions of these different genes were most likely to affect yield or quality performance of Gh and Ga.3.Evaluation of excellent introgression linesBased on performance of yield and fiber quality in IL population,we identified 37 excellent ILs,of which 31 were related to yield and ten were related to fiber quality.Some ILs had great potential for improving yield and fiber quality of G.hirsutum.For instance,IL_NB087 was significantly higher than TM-1 for BW and SI,which showed the characteristics of large boll,providing a basis for the genetic improvement of BW.SI of IL_NB250 was significantly lower than tnat of TM-1,but the LP and BN were significantly higher than that of that of TM-1,which had great potential to improve fiber yield.There were higher BN for IL_NB297 and IL_NB298 than TM-1,but BW and SI were significantly reduced,which showed the characteristics of small and more bolls,providing a basis for enhancing BN of upland cotton.FS and MIC of IL-NB032,and FL and FS of IL_NB034 were significantly improved,which provided a basis for improving fiber quality components.4.Fine-mapping QTL for fiber strength and candidate gene analysisA single segment introgression line,IL-D2-2,from the cross of(TM-1×TX-1046)reported in our previous studies,was found to have significantly improved FS compared with the recurrent parent TM-1.To fine-map the QTLs of the FS,we further crossed IL-D2-2 with its recurrent parent TM-1 to produce F2 and F3 populations.QTL analysis and substitution mapping showed qFS-Chr.D02 was anchored into a 550.66 kb-interval between two markers,INTR1027 and JESPR-231.This interval contained 67 genes,among which 27 genes related to cell-wall synthesis were selected to conduct qRT-PCR.The results revealed seven genes were expressed significantly differently during the fibre secondary-wallthickening stage(10-25 days post-anthesis),three being upregulated and four downregulated in IL-D2-2.Both GH_D02G2269(UDP-glucosyl transferase 84B1)and GH_D02G2289(unknown function(DUF869))with nonsynonymous SNPs in IL-D2-2 had significantly downregulated expression,suggesting they were candidates for qFS-Chr.D02.This research provides information about marker-assisted selection for cotton fibre strength improvement.