Genome-wide association study of frost tolerance in Vicia faba reveals syntenic loci in cool-season legumes and highlights relevant candidate genes

Cool-season grain legumes are mostly grown over spring and summer due to poor frost tolerance. However, fall-sown varieties often provide higher yields, earlier harvests and avoid late-season drought and heat. Understanding the genetic determinism and molecular basis of frost tolerance is therefore crucial for developing high-performing winter varieties. This study aimed to (1) investigate the genetic architecture of frost tolerance in Vicia faba L. using 247 accessions phenotyped under four field environments, and (2) explore the conservation of frost tolerance loci in cool-season legumes using the OrthoLegKB translational research database. A genome-wide association study identified nineteen V. faba genomic regions with a high density of markers significantly associated with frost tolerance, on all chromosomes. Mapping of frost tolerance QTL from V. faba and related species obtained from the literature onto their respective reference genomes and their integration into OrthoLegKB revealed synteny of major QTL across V. faba, Pisum sativum, and/or Medicago truncatula, particularly near clusters of CBF/DREB1 genes. Frost tolerance QTL at the P. sativum Le locus, which controls internode length, were also syntenic with a frost tolerance QTL in V. faba. Synteny between frost tolerance QTL and those controlling phenology and physiology was found at other loci, suggesting pleiotropy. Finally, expression data from P. sativum and C. arietinum accessions grown under low temperature were considered as information source to highlight potential candidate genes underlying the conserved QTL. Overall, these results provide a valuable resource for understanding and improving frost tolerance in V. faba and other cool-season legumes, including orphan crops by knowledge transfer. The use of OrthoLegKB to explore the genetic and molecular determinism of target traits across species is worth generalising.