Abstract

Gilia was the model species in this research to study the genetic regulation of flower color. I generated an F2 population from a cross of two closely related leafy-stemmed Gilia: G. capitata with blue and G. yorkii with white flowers, to obtain the fertile F1, which was selfed. The F2 population showed a segregation ratio between pigmented versus white flowers 3:1, suggesting a single locus with a major effect on the regulation of flower color. Then, the G. yorkii genome was assembled and annotated to become the reference genome to genotype-by-sequencing (GBS) 189 F2 individuals. The Quantitative Trait Loci (QTL) analysis showed a strong linkage at chromosome one (Gy1) and a weak linkage with a minor effect at chromosome four (Gy4). Underneath the Gy1 QTL for flower color, tandem duplication of the Flavonoid 3'5' hydroxylase (F3'5'H) was detected, a gene well known involved in the accumulation of blue anthocyanins. However, this QTL comprises 92 genes that could not be ruled out. I followed a Fine mapping approach to detect the causative gene in the QTL for flower color. Results suggested that the F3'5'H is a good candidate gene for flower color. Also, G. capitata was assembled and sequenced, showing an F3'5'H tandem duplication. The only difference was that G. yorkii has two F3'5'H orthologs denoted GY000172 and GY000173, and G. capitata has four F3'5'H orthologs denoted GcapC00000000179, GcapC00000000180, GcapC00000000181, and GcapC00000000182. Gene expression and sequence analysis ruled out a change by F3'5'H copy number but they cannot rule out coding changes, cis-regulatory effects, or both in flower color regulation. The results showed that GY000173 and GcapC00000000181 are similar in their genomic sequence but GY000173 has an insertion of a transposable element, which may produce a splicing that leads to a possible gene disruption, making it different from the ortholog GcapC00000000181 which is functional. Gene expression data showed that GY000172 is not expressed and GY000173 is only weakly expressed in petals. In contrast, three F3'5'H orthologs are not expressed, and only GcapC00000000181 is strongly expressed in the G. capitata petals. Furthermore, the expression of GcapC00000000181 is ninefold compared to GY000173. In addition, the hybrid F1 , which has light blue petals, showed a strong allelic imbalance among GcapC00000000181 alleles in three Single Nucleotide Polymorphism (SNP)s, since the G. yorkii alleles are silenced in the transcripts of the F1 , which is consistent with cis-regulation of the expression differences. Finally, the chemical analysis confirmed that the F3'5'H is necessary for delphinidin production and is likely the causative gene for flower color in Gilia.

Degree

PhD

College and Department

Life Sciences; Biology

Rights

https://lib.byu.edu/about/copyright/