An understanding of the genetic determinism of frost tolerance is a prerequisite for the development of frost tolerant cultivars for cold northern areas. In legumes, it is not known to which extent vernalization requirement or photoperiod responsiveness are necessary for the development of frost tolerance. In pea (Pisum sativum L.) however, the flowering locus Hr is suspected to influence winter frost tolerance by delaying floral initiation until after the main winter freezing periods have passed. The objective of this study was to dissect the genetic determinism of frost tolerance in pea by QTL analysis and to assess the genetic linkage between winter frost tolerance and the Hr locus. A population of 164 recombinant inbred lines (RILs), derived from the cross Champagne x Terese was evaluated both in the greenhouse and in field conditions to characterize the photoperiod response from which the allele at the Hr locus was inferred. In addition, the population was also assessed for winter frost tolerance in 11 field conditions. Six QTL were detected, among which three were consistent among the different experimental conditions, confirming an oligogenic determinism of frost tolerance in pea. The Hr locus was found to be the peak marker for the highest explanatory QTL of this study. This result supports the hypothesis of the prominent part played by the photoperiod responsiveness in the determinism of frost tolerance for this species. The consistency of three QTL makes these positions interesting targets for marker-assisted selection.

To increase yield in pea (Pisum sativum L.), autumn sowing would be preferable. Hence, frost tolerance of pea became a major trait of interest for breeders. In order to better understand the cold acclimation in pea, Champagne a frost tolerant line and Terese, a frost sensitive line, and their recombinant inbred lines (RIL) were studied. RIL frost tolerance was evaluated by a frost damage scale under field as well as controlled conditions. A quantitative trait loci (QTL) approach was used to identify chromosomal regions linked to frost tolerance. The detected QTL explained from 6.5 to 46.5% of the phenotypic variance. Amongst them, those located on linkage groups 5 and 6 were consistent with over all experiments, in field as well as in controlled environments. In order to improve the understanding of the frost tolerance mechanisms, several cold acclimation key characters such as concentration of sugars, electrolyte leakage, osmotic pressure, and activity of RuBisCO were assessed. Some of these physiological QTL colocalised with QTL for frost damage, in particular two raffinose QTL on LG5 and LG6 and one RuBisCO activity QTL on LG6, explaining 8.8 to 27.0% of the phenotypic variance. In addition, protein quantitative loci were mapped; some of them colocalised with frost damage and physiological QTL on LG5 and LG6, explaining 16.0-43.6% of the phenotypic variance. Raffinose metabolism and RuBisCO activity and its effect on photosynthesis might play a major role in cold acclimation of pea.

Background Frost is a limiting abiotic stress for the winter pea crop ( Pisum sativum L.) and identifying the genetic determinants of frost tolerance is a major issue to breed varieties for cold northern areas. Quantitative trait loci (QTLs) have previously been detected from bi-parental mapping populations, giving an overview of the genome regions governing this trait. The recent development of high-throughput genotyping tools for pea brings the opportunity to undertake genetic association studies in order to capture a higher allelic diversity within large collections of genetic resources as well as to refine the localization of the causal polymorphisms thanks to the high marker density. In this study, a genome-wide association study (GWAS) was performed using a set of 365 pea accessions. Phenotyping was carried out by scoring frost damages in the field and in controlled conditions. The association mapping collection was also genotyped using an Illumina Infinium® BeadChip, which allowed to collect data for 11,366 single nucleotide polymorphism (SNP) markers. Results GWAS identified 62 SNPs significantly associated with frost tolerance and distributed over six of the seven pea linkage groups (LGs). These results confirmed 3 QTLs that were already mapped in multiple environments on LG III, V and VI with bi-parental populations. They also allowed to identify one locus, on LG II, which has not been detected yet and two loci, on LGs I and VII, which have formerly been detected in only one environment. Fifty candidate genes corresponding to annotated significant SNPs, or SNPs in strong linkage disequilibrium with the formers, were found to underlie the frost damage (FD)-related loci detected by GWAS. Additionally, the analyses allowed to define favorable haplotypes of markers for the FD-related loci and their corresponding accessions within the association mapping collection. Conclusions This study led to identify FD-related loci as well as corresponding favorable haplotypes of markers and representative pea accessions that might to be used in winter pea breeding programs. Among the candidate genes highlighted at the identified FD-related loci, the results also encourage further attention to the presence of C-repeat Binding Factors (CBF) as potential genetic determinants of the frost tolerance locus on LG VI.

Freezing is one of the most serious abiotic stress factors that affect cool-season legumes. It limits species geographic distribution and causes severe yield losses. Improving tolerance to freezing has long been a main concern for legume breeders. Medicago truncatula Gaertn. has been selected as a model species for legume biology. Various studies have shown significant macrosynteny between M. truncatula and agronomically important crop legumes. A major freezing tolerance quantitative trait locus (QTL), herein referred to as Mt-FTQTL6, was previously identified on M. truncatula chromosome 6. The physical location of this QTL was determined in this study and its corresponding chromosomal interval was enriched with additional markers. Markers were first developed using the draft sequence of M. truncatula euchromatin (release versions Mt3.0 and Mt3.5). Because Mt-FTQTL6 was found to coincide with an assembly gap, the Glycine max (L.) Merr. genome sequence was also used to generate markers. Five Mt-FTQTL6-linked markers were found to be common to a region on Pisum sativum L. linkage group VI harboring a QTL for freezing damage. A subset of markers was tested for transferability across 11 additional legume species. This study lays the groundwork for identifying the molecular basis of Mt-FTQTL6. Cross-legume markers will be useful in future efforts aiming to investigate the conservation of Mt-FTQTL6 in cool-season legumes and subsequently the existence of common mechanisms for response to freezing between M. truncatula and crop legumes