Genome- and transcriptome-wide association studies reveal the genetic basis and the breeding history of seed glucosinolate content in Brassica napus

Abstract

A high content of seed glucosinolates and their degradation products imposes anti-nutritional effects on livestock; therefore, persistent efforts are made to reduce the seed GSL content to increase the commercial value of rapeseed meal. Here, we dissected the genetic structure of SGC by genome-wide association studies (GWAS) combined with transcriptome-wide association studies (TWAS). Fifteen reliable quantitative trait loci (QTLs) were identified to be associated with the reduced SGC in modern B. napus cultivars by GWAS. Analysis of the selection strength and haplotypes at these QTLs revealed that low SGC was predominantly generated by the co-selection of qGSL.A02.2, qGSL.C02.1, qGSL.A09.2, and qGSL.C09.1. Integration of the results from TWAS, comprehensive bioinformatics, and POCKET algorithm analyses indicated that BnaC02.GTR2 (BnaC02g42260D) is a candidate gene underlying qGSL.C02.1. Using CRISPR/Cas9-derived Bna.gtr2s knockout mutants, we experimentally verified that both BnaC02.GTR2 and its three paralogs positively regulate seed GSL accumulation but negatively regulated vegetative tissue GSL contents. In addition, we observed smaller seeds with higher seed oil content in these Bna.gtr2 mutants. Furthermore, both RNA-seq and correlation analyses suggested that Bna.GTR2s might play a comprehensive role in seed development, such as amino acid accumulation, GSL synthesis, sugar assimilation, and oil accumulation. This study unravels the breeding selection history of low-SGC improvement and provides new insights into the molecular function of Bna.GTR2s in both seed GSL accumulation and seed development in B. napus.