Restricted Two-Stage Multi-Locus Genome-Wide Association Analysis and Candidate Gene Prediction of Boll Opening Rate in Upland Cotton

Boll opening rate (BOR) is one of the most important indicators reflecting the early maturing trait of upland cotton (Gossypium hirsutum L.). The genome-wide association study (GWAS) was applied to dissect the QTL (quantitative trait locus) and its genetic effect for providing a theoretical basis for molecular breeding of early maturing traits in upland cotton.

The natural population composed of 315 different upland cotton varieties (lines) were used to identify the BOR under three environments. Simultaneously, a total of 9 244 SNP linkage disequilibrium block (SNPLDB) markers with multiple alleles were constructed. Then, the restricted two-stage multi-locus GWAS (RTM-GWAS) was utilized to detect SNPLDB loci significantly associated with BOR, estimate its phenotypic effect value, establish QTL-Allele matrix for significantly associated loci in the population, and further detected the stable major SNPLDB loci and elite haplotypes. Finally, according to the gene expression levels of the two transcriptome data, candidate genes that may be related to the target trait were mined within the 1 Mb genome range of the flanking sequence of the significant SNPLDB loci.

The variation of BOR was ranged from 37.78% to 100.00% and the broad-sense heritability was 67.03% in the natural population under three environments. The multi-environment variance analysis revealed that the BOR was significantly different among genotype, environment and genotype × environment interaction (P<0.001). A total of 52 SNPLDB loci significantly associated with BOR were detected through the RTM-GWAS procedure, containing 179 alleles or haplotypes, among them, the effect values of 90 increasing alleles or haplotypes ranged from 0.014 to 19.43, and the effect values of 89 decreasing alleles or haplotypes ranged from -21.49 to -0.039. Among the significant SNPLDB loci mentioned above, 6 SNPLDB loci were detected simultaneously in both multi-environment and single environment, which were considered as stable SNPLDB loci significantly associated with BOR. Through the significance analysis of phenotypic traits corresponding to different allelic variations of the above six stable SNPLDB loci, the four favorable alleles were identified as LDB_16_37952328(TT), LDB_5_96395565(AA), LDB_16_49503485(TT), and LDB_4_81118668(TT). Besides, further analysis showed that there were significant differences in the frequency distribution of favorable alleles among varieties (lines) in four different ecological regions. Additionally, a total of 178 genes were annotated and 23 potential candidate genes were predicted in the adjacent regions of 4 stable major SNPLDB loci by transcriptome data analysis.

A total of 52 SNPLDB loci significantly associated with BOR were identified, of which 4 loci were stable major SNPLDB loci. Furthermore, it was predicted that 23 genes might be related to the BOR of upland cotton. These SNPLDBs loci and candidate genes will provide a theoretical basis for marker-assisted breeding of early maturity in upland cotton.