The Extra-Pathway Interactome of the TCA Cycle: Expected and Unexpected Metabolic Interactions

Abstract

The plant tricarboxylic acid (TCA) cycle provides essential precursors for respiration, amino acid biosynthesis, and general nitrogen metabolism; moreover, it is closely involved in biotic stress responses and cellular redox homeostasis. To further understand the in vivo function of the TCA cycle enzymes, we combined affinity purification with proteomics to generate a comprehensive extra-pathway protein-protein interaction network of the plant TCA cycle. We identified 125 extra-pathway interactions in Arabidopsis (Arabidopsis thaliana) mostly related to the mitochondrial electron transport complex/ATP synthesis and amino acid metabolism but also to proteins associated with redox stress. We chose three high-scoring and two low-scoring interactions for complementary bimolecular fluorescence complementation and yeast two-hybrid assays, which highlighted the reliability of our approach, supported the intimate involvement of TCA cycle enzymes within many biological processes, and reflected metabolic changes reported previously for the corresponding mutant lines. To analyze the function of a subset of these interactions, we selected two mutants of mitochondrial glutaredoxin S15 and Amidase, which have not yet been analyzed with respect to their TCA cycle function, and performed metabolite profiling and flux analysis. Consistent with their interactions identified in this study, TCA cycle metabolites and the relative TCA flux of the two mutants were altered significantly.

Figure 1.

Work flow of the on-bead trypsin/LysC digestion method for AP-MS analysis of protein-protein interaction. The bait was expressed in plant cell culture and checked by confocal microscopy. Following affinity purification using a GFP-binding protein (GFP-Trap; ChromoTek), the protein complexes were digested on beads followed by label-free liquid chromatography (LC)-MS/MS quantification. The intensities of proteomics were analyzed by CRAPome to get the FC score.

Figure 2.

Graphical representation of the protein-protein interaction network of Arabidopsis TCA cycle enzymes. Node color represents the enzyme subunits and isoforms. A, Overview of all detected 257 interactions, including 132 interactions between the enzymes of the TCA cycle and 125 novel interactions between subunits of enzymes and other pathway enzymes or proteins. B, Classification of all 37 preys with molecular function. Ten groups of preys were detected by affinity purification, with large ratios of 27% amino acid metabolism, 19% stress, 13% mitochondrial electron transport/ATP synthesis, and 8% redox.

Figure 3.

Confirmation of selected protein-protein interactions. A, List of the three high-FC-scoring protein interactions and the two low-scoring interactions that were further tested by Y2H and BiFC. B, Y2H assay to confirm the three high-FC-scoring protein interactions and the two low-scoring interactions. PDC1a-1/ODC1-1 was used as the negative control. The interaction was performed in synthetic dextrose medium with 10 mm 3-aminotriazole and without Leu, Trp, and His. C, The three high-FC-scoring protein interactions and the two low-scoring interactions were tested further by BiFC with transient expression of tagged proteins in Arabidopsis mesophyll protoplasts. Images from left to right show the BiFC signal, fluorescence from MitoTracker Orange staining, autofluorescence, bright-field images, and merged images. Numbered rows are as follows: 1, IDH1-SCYCE/GRXS15-VYNE; 2, mtLPD2-SCYCE/Amidase-VYNE; 3, SDH8-SCYCE/Amidase-VYNE; 4, SDH2-1-SCYCE/EMB1467-VYNE; 5, SDH5-SCYCE/TS-VYNE; 6, PDC1a-1-SCYCE/ODC1-1-VYNE, shown as a representative negative control, which was not detected by affinity purification.

Figure 4.

Metabolic content was analyzed using GC-MS. A, Two mutants (amidase and GRX4), as well as the wild type (WT), were sown on soil and grown for 35 d in short-day conditions (8 h of light/16 h of dark). Metabolic content was analyzed using GC-MS (n = 5). Log² values of the relative metabolic content are presented. Significant differences compared with the wild type following Student’s t test are denoted by asterisks (*, P < 0.05 and **, P < 0.01). B, PCA of the metabolite data.

Figure 5.

¹⁴CO₂ emission in the GRXS15 and amidase mutants and the wild type (WT). Emission of ¹⁴CO₂ from the C1 and C3:4 positions of Glc in leaf discs of mutant and wild-type leaves is shown after 35 d of growth under short-day conditions. A, Evolution of ¹⁴CO₂ following incubation in C1-labeled Glc. B, Evolution of ¹⁴CO₂ following incubation in C3:4-labeled Glc. C, Ratio of ¹⁴CO₂ emission from C1 to C3:4 positions, indicating the relative activities of glycolysis and the TCA cycle. Values are means ± sd of determinations on three to four independent samples. Significant differences compared with the wild type following Student’s t test are denoted by asterisks (*, P < 0.05 and **, P < 0.01).