Genetic characterization of a core collection of flax (Linum usitatissimum L.) suitable for association mapping studies and evidence of divergent selection between fiber and linseed types

Abstract

Background: Flax is valued for its fiber, seed oil and nutraceuticals. Recently, the fiber industry has invested in the development of products made from linseed stems, making it a dual purpose crop. Simultaneous targeting of genomic regions controlling stem fiber and seed quality traits could enable the development of dual purpose cultivars. However, the genetic diversity, population structure and linkage disequilibrium (LD) patterns necessary for association mapping (AM) have not yet been assessed in flax because genomic resources have only recently been developed. We characterized 407 globally distributed flax accessions using 448 microsatellite markers. The data was analyzed to assess the suitability of this core collection for AM. Genomic scans to identify candidate genes selected during the divergent breeding process of fiber flax and linseed were conducted using the whole genome shotgun sequence of flax.

Results: Combined genetic structure analysis assigned all accessions to two major groups with six sub-groups. Population differentiation was weak between the major groups (F(ST) = 0.094) and for most of the pairwise comparisons among sub-groups. The molecular coancestry analysis indicated weak relatedness (mean = 0.287) for most individual pairs. Abundant genetic diversity was observed in the total panel (5.32 alleles per locus), and some sub-groups showed a high proportion of private alleles. The average genome-wide LD (r²) was 0.036, with a relatively fast decay of 1.5 cM. Genomic scans between fiber flax and linseed identified candidate genes involved in cell-wall biogenesis/modification, xylem identity and fatty acid biosynthesis congruent with genes previously identified in flax and other plant species.

Conclusions: Based on the abundant genetic diversity, weak population structure and relatedness and relatively fast LD decay, we concluded that this core collection is suitable for AM studies targeting multiple agronomic and quality traits aiming at the improvement of flax as a true dual purpose crop. Our genomic scans provide the first insights into candidate regions affected by divergent selection in flax. In combination with AM, genomic scans have the ability to increase the power to detect loci influencing complex traits.

Figure 1

Genetic relationships and population structure of the 407 flax accessions of the core collection. (a) Phylogenetic tree created using the Neighbor-joining (NJ) algorithm [62] and information from 414 neutral SSRs. Colored clusters represent the sub-groups within major groups. The scale bar indicates the Nei [62] minimum genetic distance. (b) Bayesian clustering (STRUCTURE K = 2). Sub-groups within groups are distributed according to the clustering obtained by the NJ analysis. Accessions with a membership coefficient Q < 0.7 were classified as admixture Group 2. (c) Average log-likelihood values (mean lnP(D) ± SD for 10 iterations) and ad-hoc statistic ΔK[70] for K values ranging from 1 to 12.

Figure 2

Distribution of pairwise molecular coancestry estimates and linkage disequilibrium decay. (a) Global pairwise molecular coancestry estimates of the 407 flax accessions of the core collection. Only kinship values ranging from 0 to 0.5 are shown. (b) Scatter plot of LD decay (r²) against the genetic distances (cM) for pairs of linked SSRs across the 15 linkage groups. The inner panel shows a detailed view of LD decay for markers located within 5 cM. The decay curves were plotted according to Breseghello and Sorells [75]. The blue line represents the threshold level of significance (r² = 0.1). The red line represents the average genome-wide LD of linked markers. (c) Pairwise molecular coancestry estimates [72] within each of the six sub-groups. The diagonal values correspond to the intra sub-group molecular coancestry. (d) Average genome-wide LD decay curves for linked markers within each of the six sub-groups.