Sink/Source Balance of Leaves Influences Amino Acid Pools and Their Associated Metabolic Fluxes in Winter Oilseed Rape (Brassica napus L.)

Abstract

Nitrogen remobilization processes from source to sink tissues in plants are determinant for seed yield and their implementation results in a complete reorganization of the primary metabolism during sink/source transition. Here, we decided to characterize the impact of the sink/source balance on amino acid metabolism in the leaves of winter oilseed rape grown at the vegetative stage. We combined a quantitative metabolomics approach with an instationary ¹⁵N-labeling experiment by using [¹⁵N]_L-glycine as a metabolic probe on leaf ranks with a gradual increase in their source status. We showed that the acquisition of the source status by leaves was specifically accompanied by a decrease in asparagine, glutamine, proline and S-methyl-l-cysteine sulphoxide contents and an increase in valine and threonine contents. Dynamic analysis of ¹⁵N enrichment and concentration of amino acids revealed gradual changes in the dynamics of amino acid metabolism with respect to the sink/source status of leaf ranks. Notably, nitrogen assimilation into valine, threonine and proline were all decreased in source leaves compared to sink leaves. Overall, our results suggested a reduction in de novo amino acid biosynthesis during sink/source transition at the vegetative stage.

Keywords: Brassica napus; SMCSO; amino acid metabolism; isotopically non-stationary steady-state metabolic flux analysis; metabolomics; proline; senescence; sink/source balance; threonine; valine; winter oilseed rape.

Figure 1

Physiological indicators of the sink/source status of four leaf ranks (L15, L11, L7, L3) from oilseed rape plants grown at the vegetative stage. (A) Fresh weight (FW) per leaf area; (B) Dry weight (DW) per leaf area; (C) Ratio of fresh weight over dry weight; (D) Relative chlorophyll content; (E) Soluble protein content; (F) Total carbon content; (G) Total nitrogen content. Plants were grown for 60 days after sowing and developed 16 leaf ranks annotated from the bottom to the top, accordingly (L1 to L16) and the two oldest leaves had already fallen off (L1 and L2). Values are the mean ± SD of 3–5 independent biological replicates. Different letters indicate groups of mean values that are significantly different between the different leaf ranks (ANOVA-Tukey HSD, p-value < 0.05).

Figure 2

Principal Component Analysis of free amino acid contents from four leaf ranks (L15, L11, L7, L3) of oilseed rape plants grown at the vegetative stage. (A) Loading plot with Pearson correlation coefficients for each amino acid along the two major axis of the PCA; (B) Score plot of leaf groups along the two major axes of the PCA with 95% confidence ellipses; (C) Content of amino acids differentiating the leaf ranks. Values are the mean ± SD of 5 independent biological replicates. Different letters indicate groups of mean values that are significantly different between the different leaf ranks (ANOVA-Tukey HSD, p-value < 0.05). PCA analysis was performed on amino acids showing a normal distribution of their values (denoted in bold in the complete dataset available in Supplementary Table S1).

Figure 3

Time-course concentrations of the most abundant amino acids in leaf ranks with a different sink/source balance during an instationary labeling experiment with [¹⁵N]_L-glycine as the sole nitrogen source. Leaf discs from the four leaf ranks having a different sink/source balance were floated during 0, 30, 60 or 120 min in a buffer containing 10 mM [¹⁵N]_L-glycine, under continuous light and orbital shaking. Amino acid quantification was achieved with a UPLC-DAD system. Values are the mean ± SD of 3 independent biological replicates for each leaf group and each time-point. Complete dataset is available in Supplementary Table S2.

Figure 4

Time-course fractional ¹⁵N enrichment of the most abundant amino acids in leaf ranks having a different sink/source balance during an instationary labeling experiment with [¹⁵N]_L-glycine as sole nitrogen source. Leaf discs from the four leaf ranks having a different sink/source balance were floated during 0, 30, 60 or 120 min in a buffer containing 10 mM [¹⁵N]_L-glycine, under continuous light and orbital shaking. Analysis of nitrogen isotopologue distributions within amino acids was achieved with a HPLC-HRMS system. Values are the mean ± SD of 3 independent biological replicates for each leaf group and each kinetic. The complete dataset is available in Supplementary Table S3.

Figure 5

Estimation of metabolic fluxes for Thr, Val and Pro biosynthesis in leaf ranks with a different sink/source balance by using a ¹⁵N-INST-MFA approach. (A) Metabolic subsystems considered for the estimation of metabolic fluxes; (B) Comparison of estimated (flux-based) and observed fractional ¹⁵N enrichment for the considered amino acids; (C) Pro biosynthesis. Local estimation of metabolic fluxes (in nmoL.g⁻¹ DW.min⁻¹) was achieved using the ScalaFlux approach with the IsoSim v2software by minimizing the sum of squared weighted errors between the observed and simulated experimental data (amino acid concentrations and label propagations). Values are the mean ± SD of 3 independent biological replicates for each leaf group. Different letters indicate groups of mean values that are significantly different between the different leaf ranks (ANOVA-Tukey HSD, p-value < 0.05). For Vthr (ANOVA p-value = 0.068), additional student tests were performed between L15-L7 and L15-L3 (p-values shown within the graph). Complete dataset is available in Supplementary Tables S4 and S5.