Patterns of Genetic Structure and Linkage Disequilibrium in a Large Collection of Pea Germplasm.

Pea (Pisum sativum, L.) is a major pulse crop used both for animal and human alimentation. Owing to its association with nitrogen-fixing bacteria, it is also a valuable component for low-input cropping systems. To evaluate the genetic diversity and the scale of linkage disequilibrium (LD) decay in pea, we genotyped a collection of 917 accessions, gathering elite cultivars, landraces, and wild relatives using an array of ∼13,000 single nucleotide polymorphisms (SNP). Genetic diversity is broadly distributed across three groups corresponding to wild/landraces peas, winter types, and spring types. At a finer subdivision level, genetic groups relate to local breeding programs and type usage. LD decreases steeply as genetic distance increases. When considering subsets of the data, LD values can be higher, even if the steep decay remains. We looked for genomic regions exhibiting high level of differentiation between wild/landraces, winter, and spring pea, respectively. Two regions on linkage groups 5 and 6 containing 33 SNPs exhibit stronger differentiation between winter and spring peas than would be expected under neutrality. Interestingly, QTL for resistance to cold acclimation and frost resistance have been identified previously in the same regions.

Genomic Prediction in Pea: Effect of Marker Density and Training Population Size and Composition on Prediction Accuracy.

Pea is an important food and feed crop and a valuable component of low-input farming systems. Improving resistance to biotic and abiotic stresses is a major breeding target to enhance yield potential and regularity. Genomic selection (GS) has lately emerged as a promising technique to increase the accuracy and gain of marker-based selection. It uses genome-wide molecular marker data to predict the breeding values of candidate lines to selection. A collection of 339 genetic resource accessions (CRB339) was subjected to high-density genotyping using the GenoPea 13.2K SNP Array. Genomic prediction accuracy was evaluated for thousand seed weight (TSW), the number of seeds per plant (NSeed), and the date of flowering (BegFlo). Mean cross-environment prediction accuracies reached 0.83 for TSW, 0.68 for NSeed, and 0.65 for BegFlo. For each trait, the statistical method, the marker density, and/or the training population size and composition used for prediction were varied to investigate their effects on prediction accuracy: the effect was large for the size and composition of the training population but limited for the statistical method and marker density. Maximizing the relatedness between individuals in the training and test sets, through the CDmean-based method, significantly improved prediction accuracies. A cross-population cross-validation experiment was further conducted using the CRB339 collection as a training population set and nine recombinant inbred lines populations as test set. Prediction quality was high with mean Q (2) of 0.44 for TSW and 0.59 for BegFlo. Results are discussed in the light of current efforts to develop GS strategies in pea.

Development of two major resources for pea genomics: the GenoPea 13.2K SNP Array and a high-density, high-resolution consensus genetic map.

Single nucleotide polymorphism (SNP) arrays represent important genotyping tools for innovative strategies in both basic research and applied breeding. Pea is an important food, feed and sustainable crop with a large (about 4.45 Gbp) but not yet available genome sequence. In the present study, 12 pea recombinant inbred line populations were genotyped using the newly developed GenoPea 13.2K SNP Array. Individual and consensus genetic maps were built providing insights into the structure and organization of the pea genome. Largely collinear genetic maps of 3918-8503 SNPs were obtained from all mapping populations, and only two of these exhibited putative chromosomal rearrangement signatures. Similar distortion patterns in different populations were noted. A total of 12 802 transcript-derived SNP markers placed on a 15 079-marker high-density, high-resolution consensus map allowed the identification of ohnologue-rich regions within the pea genome and the localization of local duplicates. Dense syntenic networks with sequenced legume genomes were further established, paving the way for the identification of the molecular bases of important agronomic traits segregating in the mapping populations. The information gained on the structure and organization of the genome from this research will undoubtedly contribute to the understanding of the evolution of the pea genome and to its assembly. The GenoPea 13.2K SNP Array and individual and consensus genetic maps are valuable genomic tools for plant scientists to strengthen pea as a model for genetics and physiology and enhance breeding.

Full-length de novo assembly of RNA-seq data in pea (Pisum sativum L.) provides a gene expression atlas and gives insights into root nodulation in this species.

Next-generation sequencing technologies allow an almost exhaustive survey of the transcriptome, even in species with no available genome sequence. To produce a Unigene set representing most of the expressed genes of pea, 20 cDNA libraries produced from various plant tissues harvested at various developmental stages from plants grown under contrasting nitrogen conditions were sequenced. Around one billion reads and 100 Gb of sequence were de novo assembled. Following several steps of redundancy reduction, 46 099 contigs with N50 length of 1667 nt were identified. These constitute the 'Caméor' Unigene set. The high depth of sequencing allowed identification of rare transcripts and detected expression for approximately 80% of contigs in each library. The Unigene set is now available online (http://bios.dijon.inra.fr/FATAL/cgi/pscam.cgi), allowing (i) searches for pea orthologs of candidate genes based on gene sequences from other species, or based on annotation, (ii) determination of transcript expression patterns using various metrics, (iii) identification of uncharacterized genes with interesting patterns of expression, and (iv) comparison of gene ontology pathways between tissues. This resource has allowed identification of the pea orthologs of major nodulation genes characterized in recent years in model species, as a major step towards deciphering unresolved pea nodulation phenotypes. In addition to a remarkable conservation of the early transcriptome nodulation apparatus between pea and Medicago truncatula, some specific features were highlighted. The resource provides a reference for the pea exome, and will facilitate transcriptome and proteome approaches as well as SNP discovery in pea.

Genetic diversity and trait genomic prediction in a pea diversity panel.

BACKGROUND: Pea (Pisum sativum L.), a major pulse crop grown for its protein-rich seeds, is an important component of agroecological cropping systems in diverse regions of the world. New breeding challenges imposed by global climate change and new regulations urge pea breeders to undertake more efficient methods of selection and better take advantage of the large genetic diversity present in the Pisum sativum genepool. Diversity studies conducted so far in pea used Simple Sequence Repeat (SSR) and Retrotransposon Based Insertion Polymorphism (RBIP) markers. Recently, SNP marker panels have been developed that will be useful for genetic diversity assessment and marker-assisted selection. RESULTS: A collection of diverse pea accessions, including landraces and cultivars of garden, field or fodder peas as well as wild peas was characterised at the molecular level using newly developed SNP markers, as well as SSR markers and RBIP markers. The three types of markers were used to describe the structure of the collection and revealed different pictures of the genetic diversity among the collection. SSR showed the fastest rate of evolution and RBIP the slowest rate of evolution, pointing to their contrasted mode of evolution. SNP markers were then used to predict phenotypes -the date of flowering (BegFlo), the number of seeds per plant (Nseed) and thousand seed weight (TSW)- that were recorded for the collection. Different statistical methods were tested including the LASSO (Least Absolute Shrinkage ans Selection Operator), PLS (Partial Least Squares), SPLS (Sparse Partial Least Squares), Bayes A, Bayes B and GBLUP (Genomic Best Linear Unbiased Prediction) methods and the structure of the collection was taken into account in the prediction. Despite a limited number of 331 markers used for prediction, TSW was reliably predicted. CONCLUSION: The development of marker assisted selection has not reached its full potential in pea until now. This paper shows that the high-throughput SNP arrays that are being developed will most probably allow for a more efficient selection in this species.

Unexpectedly low nitrogen acquisition and absence of root architecture adaptation to nitrate supply in a Medicago truncatula highly branched root mutant.

To complement N2 fixation through symbiosis, legumes can efficiently acquire soil mineral N through adapted root architecture. However, root architecture adaptation to mineral N availability has been little studied in legumes. Therefore, this study investigated the effect of nitrate availability on root architecture in Medicago truncatula and assessed the N-uptake potential of a new highly branched root mutant, TR185. The effects of varying nitrate supply on both root architecture and N uptake were characterized in the mutant and in the wild type. Surprisingly, the root architecture of the mutant was not modified by variation in nitrate supply. Moreover, despite its highly branched root architecture, TR185 had a permanently N-starved phenotype. A transcriptome analysis was performed to identify genes differentially expressed between the two genotypes. This analysis revealed differential responses related to the nitrate acquisition pathway and confirmed that N starvation occurred in TR185. Changes in amino acid content and expression of genes involved in the phenylpropanoid pathway were associated with differences in root architecture between the mutant and the wild type.

QTL analysis of frost damage in pea suggests different mechanisms involved in frost tolerance.

KEY MESSAGE: Avoidance mechanisms and intrinsic resistance are complementary strategies to improve winter frost tolerance and yield potential in field pea. The development of the winter pea crop represents a major challenge to expand plant protein production in temperate areas. Breeding winter cultivars requires the combination of freezing tolerance as well as high seed productivity and quality. In this context, we investigated the genetic determinism of winter frost tolerance and assessed its genetic relationship with yield and developmental traits. Using a newly identified source of frost resistance, we developed a population of recombinant inbred lines and evaluated it in six environments in Dijon and Clermont-Ferrand between 2005 and 2010. We developed a genetic map comprising 679 markers distributed over seven linkage groups and covering 947.1 cM. One hundred sixty-one quantitative trait loci (QTL) explaining 9-71 % of the phenotypic variation were detected across the six environments for all traits measured. Two clusters of QTL mapped on the linkage groups III and one cluster on LGVI reveal the genetic links between phenology, morphology, yield-related traits and frost tolerance in winter pea. QTL clusters on LGIII highlighted major developmental gene loci (Hr and Le) and the QTL cluster on LGVI explained up to 71 % of the winter frost damage variation. This suggests that a specific architecture and flowering ideotype defines frost tolerance in winter pea. However, two consistent frost tolerance QTL on LGV were independent of phenology and morphology traits, showing that different protective mechanisms are involved in frost tolerance. Finally, these results suggest that frost tolerance can be bred independently to seed productivity and quality.

A PQL (protein quantity loci) analysis of mature pea seed proteins identifies loci determining seed protein composition.

Legume seeds are a major source of dietary proteins for humans and animals. Deciphering the genetic control of their accumulation is thus of primary significance towards their improvement. At first, we analysed the genetic variability of the pea seed proteome of three genotypes over 3 years of cultivation. This revealed that seed protein composition variability was under predominant genetic control, with as much as 60% of the spots varying quantitatively among the three genotypes. Then, by combining proteomic and quantitative trait loci (QTL) mapping approaches, we uncovered the genetic architecture of seed proteome variability. Protein quantity loci (PQL) were searched for 525 spots detected on 2-D gels obtained for 157 recombinant inbred lines. Most protein quantity loci mapped in clusters, suggesting that the accumulation of the major storage protein families was under the control of a limited number of loci. While convicilin accumulation was mainly under the control of cis-regulatory regions, vicilins and legumins were controlled by both cis- and trans-regulatory regions. Some loci controlled both seed protein composition and protein content and a locus on LGIIa appears to be a major regulator of protein composition and of protein in vitro digestibility.

Highly-multiplexed SNP genotyping for genetic mapping and germplasm diversity studies in pea.

BACKGROUND: Single Nucleotide Polymorphisms (SNPs) can be used as genetic markers for applications such as genetic diversity studies or genetic mapping. New technologies now allow genotyping hundreds to thousands of SNPs in a single reaction.In order to evaluate the potential of these technologies in pea, we selected a custom 384-SNP set using SNPs discovered in Pisum through the resequencing of gene fragments in different genotypes and by compiling genomic sequence data present in databases. We then designed an Illumina GoldenGate assay to genotype both a Pisum germplasm collection and a genetic mapping population with the SNP set. RESULTS: We obtained clear allelic data for more than 92% of the SNPs (356 out of 384). Interestingly, the technique was successful for all the genotypes present in the germplasm collection, including those from species or subspecies different from the P. sativum ssp sativum used to generate sequences. By genotyping the mapping population with the SNP set, we obtained a genetic map and map positions for 37 new gene markers. CONCLUSION: Our results show that the Illumina GoldenGate assay can be used successfully for high-throughput SNP genotyping of diverse germplasm in pea. This genotyping approach will simplify genotyping procedures for association mapping or diversity studies purposes and open new perspectives in legume genomics.

Dissecting the proteome of pea mature seeds reveals the phenotypic plasticity of seed protein composition.

Pea (Pisum sativum L.) is the most cultivated European pulse crop and the pea seeds mainly serve as a protein source for monogastric animals. Because the seed protein composition impacts on seed nutritional value, we aimed at identifying the determinants of its variability. This paper presents the first pea mature seed proteome reference map, which includes 156 identified proteins (http://www.inra.fr/legumbase/peaseedmap/). This map provides a fine dissection of the pea seed storage protein composition revealing a large diversity of storage proteins resulting both from gene diversity and post-translational processing. It gives new insights into the pea storage protein processing (especially 7S globulins) as a possible adaptation towards progressive mobilization of the proteins during germination. The nonstorage seed proteome revealed the presence of proteins involved in seed defense together with proteins preparing germination. The plasticity of the seed proteome was revealed for seeds produced in three successive years of cultivation, and 30% of the spots were affected by environmental variations. This work pinpoints seed proteins most affected by environment, highlighting new targets to stabilize storage protein composition that should be further analyzed.