Hitzelberger, John Christian (2019-08). Development and Characterization of Gossypium hirsutum L. x G. mustelinum Miers ex Watts Chromosome Segment Substitution Line Population. Master's Thesis.
Thesis
Genetic improvements of Upland cotton (Gossypium hirsutum L.) must be continuous if the crop is to remain biologically and economically viable. The notoriously low genetic diversity found among Upland cottons constrains opportunities for improvements by breeding based on conventional elite-by-elite crosses. Fortunately, each related wild AD-genome species harbors about 80,000 genes and so introgression of a wild species genome could significantly increase the breadth of variation available among Upland cottons. Most genes from a non-domesticated donor species would expectedly be agriculturally neutral or deleterious, thus, for multi-genic traits, it would expectedly be virtually impossible to discern the presence of a beneficial allele or gene in a donor genome until that DNA variant is selectively integrated into an Upland cotton genetic background. In this study, sub-chromosomal segments of the G. mustelinum genome were concomitantly introgressed and separated by developing chromosome segment substitution lines (CSSLs). Modified backcross-inbreeding and marker-based selections enabled the creation of a panel of CSSLs, each containing one to several small sub-chromosomal introgressed alien segments but otherwise isogenic to the G. hirsutum recurrent inbred line parent and to each other. Genotyping was based on single-nucleotide polymorphism markers (SNPs). Segmenttargeted genotyping was based on PACE or KASP assays for small sets of two to several interspaced SNPs per segment, drawn from a recently developed genome-spanning panel of ~260 such assays. In contrast, genome-wide high-density genotyping was based on the Illumina Cotton63KSNP array for 15,000+ SNPs. Coverage and pedigree-based tracking of specific segments at BC4 and BC5 generations was enabled by prior CottonSNP63K-based genotyping of 18 BC2F1s. In 2017, 410 BC4F1s were backcrossed to G. hirsutum and selected, of which 92 were genotyped. In 2018, 378 of 933 BC5F1s were genotyped, self-pollinated and selected. Based on targeted analyses with spaced SNPs completed before January 2018, the CSSL panel comprises 65 BC5F1 plants that collectively contain approximately 50% of the G. mustelinum genome in a heterozygous state; these descend from 18 different BC2F1s. In subsequent research, each heterozygous donor segment must be recovered in homozygous form, and additional CSSLs with complementary genome coverage (~50%) must be identified to attain 100% of donor genome coverage. Available germplasm resources include BC5S1 seed from 378 BC5F1 plants and 77 BC4:5F1 families. To facilitate the follow-through efforts, I created well-organized computer spreadsheets that integrate relevant pedigree and SNP data; these help identify which segments to target, which pedigree to use for a given segment, and which SNPs to genotype for selection of heterozygotes and homozygotes. Recovery of homozygotes and complementary segments will be facilitated by the availability of plants and/or seed at BC5S1, BC5F1, and BC4F1 generations. To begin gauging if donor genes affect fiber quality traits, BC5F1 fiber samples were harvested on a single-plant basis within BC2-derived families and characterized using High Volume Instrument (HVI) analysis. ANOVA of HVI data showed that differences were significant among families (?=0.05) for micronaire (p = 0.0342), upper half mean length (p = 0.0004), elongation (p = 0.0253), and strength (p = 0.0224). If substantiated, the results would reflect dominant or co-dominant effects, but insufficient experimental replication precludes conclusiveness at this time. More authoritative deductions about dominant, co-dominant and recessive genotypic effects will be possible once homozygous BC5Sn CSSL lines are established, as these will be amenable to seed increases, use of larger experimental units, replication and multiple environments.
