Molecular Mechanisms Underlying Flax (Linum usitatissimum L.) Tolerance to Cadmium: A Case Study of Proteome and Metabolome of Four Different Flax Genotypes

Abstract

Cadmium is one of the most toxic heavy metal pollutants, and its accumulation in the soil is harmful to agriculture. Plants have a higher cadmium tolerance than animals, and some species can be used for phytoremediation. Flax (Linum usitatissimum L.) can accumulate high amounts of cadmium, but the molecular mechanism behind its tolerance is unknown. Here, we employed four genotypes representing two fiber cultivars, an oilseed breeding line, and a transgenic line overexpressing the metallothionein domain for improved cadmium tolerance. We analyzed the proteome of suspensions and the proteome and metabolome of seedling roots in response to cadmium. We identified more than 1400 differentially abundant proteins representing putative mechanisms in cadmium tolerance, including metal-binding proteins and transporters, enzymes of flavonoid, jasmonate, polyamine, glutathione metabolism, and HSP70 proteins. Our data indicated the role of the phytohormone cytokinin in the observed responses. The metabolome profiling found that pipecolinic acid could be a part of the cadmium accumulation mechanism, and the observed accumulation of putrescine, coumaric acid, cinnamic acid, and coutaric acid confirmed the role of polyamines and flavonoids in tolerance to cadmium. In conclusion, our data provide new insight into cadmium tolerance and prospective targets for improving cadmium tolerance in other plants.

Keywords: Cd2+; HSP70; heavy metals; phenolic compounds; pipecolinic acid; polyamines; proteome; toxicity.

Figure 1

Cadmium tolerance in four genotypes employed in the study. (a) Relative cell viability of suspension cultures, (b) root length, and (c) cadmium content in shoots of six-day-old flax seedlings. The plots represent the means and standard deviation of at least five (a,b) and three (c) biological replicates. Different letters indicate significant differences (Kruskal-Wallis test, p < 0.05); p-values represent results of pair-wised comparisons and Student’s t-test.

Figure 2

The proteome profile of the flax suspension cultures did not show the expected separation of the mock and cadmium treated cells. (a) Results of two-way ANOVA analysis visualized in a Venn diagram and (b) PCA based on profile of all 969 differentially abundant proteins. For details, see Supplementary Tables S1 and S2.

Figure 3

Cadmium-responsive proteins identified in flax suspension cultures. (a,b) Genotype-specific effect on protein abundance combined with (a) the cadmium-induced accumulation and (b) a decrease in abundance. (c,d) Genotype-independent effects of cadmium on protein accumulation (c) and decrease in abundance (d). Heat maps represent the mean relative abundances of five biological replicates; letters represent the results of Kruskal-Wallis and Dunn’s test (p < 0.05); blue, significant effects of genotype; red, significant effects of cadmium; asterisks represent significant interaction between genotype and cadmium concentration.

Figure 4

Cadmium response in root tissue of four different genotypes. Comparison of mock-treated roots and response to 100 µM Cd²⁺. (a) Results of the two-way ANOVA analysis visualized in a Venn diagram, (b) PCA based on the profile of 1435 differentially abundant proteins (p < 0.05, at least 1.5-fold change), (c) visualization of functional categories in the ProteoMap, and (d) significant differences visualized on a heat map. The ProteoMap corresponds to the estimated content in the mock-treated AGT plants. Asterisks indicate a significant interaction of genotype and cadmium concentration on protein abundances. The letters represent significant differences (p < 0.05, ANOVA, Tukey’s HSD). For details, see Supplementary Table S3.

Figure 5

Metabolome analysis of flax root in response to cadmium. (a) Results of two-way ANOVA analysis visualized in a Venn diagram and the corresponding heat map visualization of subsets of (b) 24 metabolites whose abundances were affected by cadmium and genotype with significant interaction of these two factors, (c) 44 metabolites that showed genotype- and cadmium- specific responses and no interaction of these two factors, and (d) five metabolites that showed only a genotype-independent response to cadmium. Different letters indicate significant differences (Kruskal-Wallis and Dunn’s test, p < 0.05). For simplicity, the tests and normalization in the heat maps are limited to the genotype level. For cross-genotype comparisons, see the Supplementary Table S6.

Figure 6

Identification of proteins and metabolites that show significant correlation with cadmium effects on the molecular composition of roots. Orthogonal partial least squares discriminant analysis (a) followed by VIP (variable importance in projection); (b) Identified proteins and metabolites with significant correlation (absolute threshold 0.75) are listed. For details, see Supplementary Tables S4 and S5.

Figure 7

Integrative analysis of proteomics and metabolomics data. (a) Visualization of metabolic pathways significantly impacted by cadmium treatment and (b) significant intra-omics correlations. The pathway impact analysis was visualized by MetaboAnalyst. Omics data interaction was evaluated by OmicsAnalyst using Pearson’s correlation threshold of 0.75 for intra-omics interactions. For details, see Supplementary Tables S4 and S5.