Oxidation of Arabidopsis thaliana COX19 Using the Combined Action of ERV1 and Glutathione

Abstract

Protein import and oxidative folding within the intermembrane space (IMS) of mitochondria relies on the MIA40-ERV1 couple. The MIA40 oxidoreductase usually performs substrate recognition and oxidation and is then regenerated by the FAD-dependent oxidase ERV1. In most eukaryotes, both proteins are essential; however, MIA40 is dispensable in Arabidopsis thaliana. Previous complementation experiments have studied yeast mia40 mutants expressing a redox inactive, but import-competent versions of yeast Mia40 using A. thaliana ERV1 (AtERV1) suggest that AtERV1 catalyzes the oxidation of MIA40 substrates. We assessed the ability of both yeast and Arabidopsis MIA40 and ERV1 recombinant proteins to oxidize the apo-cytochrome reductase CCMH and the cytochrome c oxidase assembly protein COX19, a typical MIA40 substrate, in the presence or absence of glutathione, using in vitro cysteine alkylation and cytochrome c reduction assays. The presence of glutathione used at a physiological concentration and redox potential was sufficient to support the oxidation of COX19 by AtERV1, providing a likely explanation for why MIA40 is not essential for the import and oxidative folding of IMS-located proteins in Arabidopsis. The results point to fundamental biochemical differences between Arabidopsis and yeast ERV1 in catalyzing protein oxidation.

Keywords: ERV1; MIA40; glutathione; mitochondrial intermembrane space; oxidative folding.

Figure 1

In vitro oxidation of Arabidopsis COX19 protein using the ERV1–MIA40 couple from Arabidopsis thaliana or Saccharomyces cerevisiae. Capacity of ERV1 alone or in the presence of MIA40 to oxidize Arabidopsis COX19 using A. thaliana (A,B) or S. cerevisiae (C,D) proteins. In (A,C), the electron transfer from reduced COX19 (40 µM, grey line) was followed by recording the reduction of 20 µM cytochrome c at 550 nm over time in the presence of 4 µM ERV1 alone (red line) or with 4 µM MIA40 (blue line). A reference was made using 40 µM DTT as an electron donor (black line). In (B,D), COX19 (40 µM) oxidation was followed over time using cysteine alkylation using mmPEG₂₄ and SDS-PAGE separation after incubation with 4 µM ERV1 alone or in the presence of 4 µM MIA40.

Figure 2

In vitro oxidation of Arabidopsis COX19 protein using the ERV1–MIA40 couple from Arabidopsis thaliana or Saccharomyces cerevisiae in the presence of glutathione. The oxidation of AtCOX19 (40 µM) was followed over time using a cysteine alkylation assay (mmPEG₂₄ alkylation and SDS-PAGE separation) in the presence of a total glutathione concentration of 8 mM, adjusted to a midpoint redox potential for the GSH/GSSG couple of—274 mV at pH 7.4 (A). The capacity of Arabidopsis or of yeast ERV1 alone, or in presence of MIA40, each at a 4 µM concentration, was tested in the same conditions (B,C). The amount of fully oxidized AtCOX19 relative to the total AtCOX19 amounts (represented by a sum of the intensity of the three bands corresponding to the various redox states of AtCOX19) was quantified in each condition. Mean values ± SD of at least three replicates are shown.

Figure 3

In vitro oxidation of Arabidopsis CCMH using the ERV1–MIA40 couple from Arabidopsis thaliana or Saccharomyces cerevisiae. Capacity of A. thaliana (A,B) or S. cerevisiae (C,D) ERV1–MIA40 couples to oxidize Arabidopsis CCMH using cysteine alkylation with mPEG2000 or mmPEG24 (as indicated) and SDS-PAGE separation or cytochrome c reduction. In all of these assays, 40 µM CCMH was incubated with 4 µM ERV1 alone or in presence of 4 µM MIA40. In (E,F), the same tests were performed using a hybrid system comprising S. cerevisiae Erv1 and A. thaliana MIA40.

Figure 4

Redox titration of the catalytic disulfides in AtERV1 and AtMIA40. The titrations of the dithiol-disulfide couples of AtERV1 SEQKSS (A), AtERV1 SERS (B), and AtMIA40 (C) were carried out using mixtures of reduced and oxidized DTT for a total DTT concentration of 2 mM for 2 h at pH 7.0. Free thiol groups were labeled using mBBr and the resulting fluorescence emission was expressed as % of maximal fluorescence and fitted to the redox potential of the solution. The obtained E_m value is the mean ± SD of three replicates. (D) Summary of the midpoint redox potentials measured for the disulfides of MIA40 substrates, MIA40 CPC active motif, and ERV1 from S. cerevisiae (Sc) and A. thaliana (At). For ERV1, we have distinguished the shuttle (shut) and internal (int) disulfides and added the redox potential of the FAD/FADH₂ couple when known. Different values, labeled a or b, have been obtained for ScERV1 and are based, respectively, on references [57] or [58].

Figure 5

Proposed model of GSH- and AtERV1-mediated oxidation of Arabidopsis COX19 in the absence of MIA40. In this scheme, reduced glutathione (GSH) reacts with the shuttle disulfide of ERV1 (Cx₄C) (rather than the catalytic disulfide, which is not in contact with the substrates), leading to the formation of a glutathione adduct on one of these cysteines. The glutathione molecule would then be transferred to one of the cysteine residues, forming the more internal disulfide in the substrates, to promote disulfide bond formation and the release of GSH. The second disulfide bridge of the substrate could be formed using the same mechanism (ERV1 + GSH) or, eventually, by the action of GSSG or of O₂ in conjunction with metal traces. The shuttle disulfide of ERV1 is subsequently regenerated by transferring electrons to the catalytic Cx₂C motif, FAD and oxygen (O₂) or cytochrome c (cyt c).