Reconstitution and properties of the recombinant glyceraldehyde-3-phosphate dehydrogenase/CP12/phosphoribulokinase supramolecular complex of Arabidopsis

Abstract

Calvin cycle enzymes glyceraldehyde-3-phosphate dehydrogenase (GAPDH) and phosphoribulokinase (PRK) form together with the regulatory peptide CP12 a supramolecular complex in Arabidopsis (Arabidopsis thaliana) that could be reconstituted in vitro using purified recombinant proteins. Both enzyme activities were strongly influenced by complex formation, providing an effective means for regulation of the Calvin cycle in vivo. PRK and CP12, but not GapA (A(4) isoform of GAPDH), are redox-sensitive proteins. PRK was reversibly inhibited by oxidation. CP12 has no enzymatic activity, but it changed conformation depending on redox conditions. GapA, a bispecific NAD(P)-dependent dehydrogenase, specifically formed a binary complex with oxidized CP12 when bound to NAD. PRK did not interact with either GapA or CP12 singly, but oxidized PRK could form with GapA/CP12 a stable ternary complex of about 640 kD (GapA/CP12/PRK). Exchanging NADP for NAD, reducing CP12, or reducing PRK were all conditions that prevented formation of the complex. Although GapA activity was little affected by CP12 alone, the NADPH-dependent activity of GapA embedded in the GapA/CP12/PRK complex was 80% inhibited in respect to the free enzyme. The NADH activity was unaffected. Upon binding to GapA/CP12, the activity of oxidized PRK dropped from 25% down to 2% the activity of the free reduced enzyme. The supramolecular complex was dissociated by reduced thioredoxins, NADP, 1,3-bisphosphoglycerate (BPGA), or ATP. The activity of GapA was only partially recovered after complex dissociation by thioredoxins, NADP, or ATP, and full GapA activation required BPGA. NADP, ATP, or BPGA partially activated PRK, but full recovery of PRK activity required thioredoxins. The reversible formation of the GapA/CP12/PRK supramolecular complex provides novel possibilities to finely regulate GapA ("non-regulatory" GAPDH isozyme) and PRK (thioredoxin sensitive) in a coordinated manner.

Figure 1.

Multiple alignment of Arabidopsis CP12s (CP12-1, At2g47400; CP12-2, At3g62410; CP12-3, At1g76650; according to Marri et al. [2005]). The alignment was performed by ClustalW. The transit peptide as predicted by ChloroP is shown on a gray field. Conserved Cys are on a black field. The sequence of recombinant CP12-2 retained at the N terminus four residues from the His tag. Therefore, the N terminus of recombinant CP12-2 started with GSHM (not shown), followed by AAPEGG….

Figure 2.

SDS-PAGE of GapA, PRK, and CP12 of Arabidopsis expressed in E. coli and purified to homogeneity. Gels were stained with Coomassie Brilliant Blue R-250. A, GapA and PRK were treated with sample buffer including reductant (5 mm DTT) and loaded on a 12.5% polyacrylamide gel. B, Reduced (rd) and oxidized (ox) CP12 samples were obtained by incubation for 2 h at 25°C with equimolar concentrations of prereduced or preoxidized thioredoxin, respectively (see "Materials and Methods"). Samples were then boiled for 3 min in sample buffer with no reductants, and the proteins were separated on 15% polyacrylamide gels. Thioredoxin migrated below the 14.4-kD marker. The apparent molecular mass of CP12 (arrowheads) shifted from 20 to 16 kD depending on redox conditions.

Figure 3.

In vitro reconstitution of binary (GapA/CP12) and ternary (GapA/CP12/PRK) complexes. A, Gel filtration (Superdex 200) of purified GapA, PRK (oxidized), and CP12 (either oxidized or reduced). Oxidized and reduced proteins were obtained by 3 h incubation at 25°C in the presence of 25 mm reduced or oxidized DTT. The four samples were individually loaded on the column and run under identical conditions. The absorption patterns at 280 nm were normalized and superimposed. Column equilibration buffer was 50 mm Tris-HCl, pH 7.5, 150 mm KCl, 1 mm EDTA; volume of loaded samples was 0.2 mL; and flow rate was 0.5 mL min⁻¹. Column calibration is reported at the bottom of the figure. Estimated molecular masses of samples were 120 kD (GapA), 110 kD (PRK, oxidized), 35 kD (CP12, reduced), and 29 kD (CP12, oxidized). B, Same type of chromatography as in A, except that the sample loaded on the column was a mixture of equimolar GapA and CP12 (oxidized) on subunit basis, incubated for 2 h at 4°C in the presence of 0.5 mm NADP before loading. Column equilibration buffer was as in A with the addition of 0.2 mm NADP. Insert, Western blots showing that anti-GAPDH polyclonal antibodies recognize GapA only in the peak corresponding to 120 kD, and anti-CP12 antibodies recognize CP12 only in the low M_r peak (29 kD). Stars indicate the column fractions (0.35 mL) which were concentrated and loaded on the gel for western blotting. C, Same experiments as in B except that NAD substituted NADP in both incubation and column buffers. The GapA peak shifted to the left to an apparent molecular mass of 150 kD. The elution pattern of B is superimposed for easy comparison. Insert, Western blots showing that anti-CP12 antibodies recognized a 16-kD peptide coeluting with GapA tetramers (36 kD in SDS-PAGE). D, Same experiment as in C except for the addition, in the incubation buffer, of equimolar PRK (oxidized) on subunit basis. Insert, Immunoblots showing that the peak at 640 kD contained GapA, CP12, and PRK.

Figure 4.

Activities of GapA and PRK as free enzymes, as enzymes embedded in complexes, and after complex treatment with several potential effectors. A, Activity expressed as percentage of the full activity of free tetrameric GapA (NADPH dependent). GapA/CP12 and GapA/CP12/PRK complexes were obtained under the same conditions as described in legends to Figures 3C and 3D, respectively. B, PRK activity was assayed and expressed as a percentage of fully reduced PRK. PRK oxidation was obtained by 3 h incubation at 25°C with 25 mm oxidized DTT. The ternary complex with GapA and CP12 was obtained as in Figure 3D. C, The GapA/CP12/PRK complex was reconstituted and chromatographed as in Figure 3D, re-equilibrated with 100 mm Tricine-NaOH, pH 7.9, in the absence of NAD, and incubated under different conditions (0.2 mm NAD; 0.5 mm ribulose-5-P; 2 mm ATP; 0.2 mm NADP; 43 μm BPGA; 5 mm reduced DTT plus 1 μg/mL thioredoxin). BPGA (43 μm) was produced at equilibrium by the reaction of phosphoglycerate kinase with 3 mm 3-phosphoglycerate and 2 mm ATP (initial concentrations). After 1 h incubation at 4°C with different effectors, the NADPH activity of GapA was assayed and expressed as percentage of the activity of GapA before complex reconstitution. An aliquot of the incubated sample was also loaded on a gel filtration column to check the aggregation state of the proteins. The stars indicate those conditions that did not lead to complex dissociation (NAD and ribulose-5-P; see also Fig. 5). D, Same conditions as in C except that PRK activity was assayed and expressed as percentage of the activity of the fully reduced enzyme. Ru5P, Ribulose-5-P.

Figure 5.

Effect of GapA and PRK ligands on GapA/CP12/PRK complex stability. The GapA/CP12/PRK complex was reconstituted and chromatographed as in Figure 3D, re-equilibrated with 100 mm Tricine-NaOH, pH 7.9, in the absence of NAD, and incubated under different conditions (A, 0.2 mm NAD; B, 0.5 mm ribulose-5-P; C, 2 mm ATP; D, 0.2 mm NADP; E, 43 μm BPGA; F, 5 mm reduced DTT plus 1 μg/mL thioredoxin) before loading on the gel filtration column (Superdex 200). The column equilibration buffer included different effectors as reported in the figure. The elution volumes of GapA/CP12/PRK (640 kD), GapA and PRK (110–120 kD), and CP12 (29–35 kD) are indicated. Under conditions of complex disruption (ATP, NADP, BPGA, or reductants), the CP12 peak was hardly detectable, partly due to the lower molar extinction coefficient in comparison with either GapA or PRK (see “Materials and Methods”).

Figure 6.

Schematic representation of the sequential formation of GapA/CP12 and GapA/CP12/PRK complexes. The percentage activity of GapA (NADPH dependent) and PRK in respect to free, fully activated enzymes is reported by numbers close to each enzyme form. The scheme indicates which enzyme forms participate in complex formation but does not represent the stoichiometry of complexes.