A disulfide bond in the TIM23 complex is crucial for voltage gating and mitochondrial protein import

Abstract

Tim17 is a central, membrane-embedded subunit of the mitochondrial protein import machinery. In this study, we show that Tim17 contains a pair of highly conserved cysteine residues that form a structural disulfide bond exposed to the intermembrane space (IMS). This disulfide bond is critical for efficient protein translocation through the TIM23 complex and for dynamic gating of its preprotein-conducting channel. The disulfide bond in Tim17 is formed during insertion of the protein into the inner membrane. Whereas the import of Tim17 depends on the binding to the IMS protein Mia40, the oxidoreductase activity of Mia40 is surprisingly dispensable for Tim17 oxidation. Our observations suggest that Tim17 can be directly oxidized by the sulfhydryl oxidase Erv1. Thus, import and oxidation of Tim17 are mediated by the mitochondrial disulfide relay, though the mechanism by which the disulfide bond in Tim17 is formed differs considerably from that of soluble IMS proteins.

Figure 1.

Tim17 contains a structural disulfide bond. (A) Schematic representation of the Tim17 sequence. Invariant residues (see Fig. S1) are indicated above and cysteine residues are indicated below the representation. TM, transmembrane segments. The inset shows the topology of Tim17 and the position of the two conserved cysteine residues. (B) Mitochondrial proteins were TCA precipitated and incubated with the reductant TCEP and/or the thiol-modifying compound mmPEG₂₄. The samples were resolved by SDS-PAGE and used for Western blotting with antibodies against Tim17, Cmc1 (a protein with four oxidized cysteines), or Mdj1 (a protein with 10 reduced cysteines). (C) Mitochondria were isolated from Tim17 mutants in which the indicated cysteines were replaced by alanine residues and used for an mmPEG₂₄ shift experiment as described for B. WT, wild type; AC, Tim17AC; CA, Tim17CA; AA, Tim17AA. (D) To assess the stability of the disulfide bond in Tim17, wild-type mitochondria were pretreated with the indicated concentrations of DTT for 15 min at 28°C. Mitochondrial proteins were TCA precipitated before the redox states of the cysteines in Tim17 were analyzed as described for B. (E) Yeast cells were grown in the presence or absence of molecular oxygen. Cellular proteins were TCA precipitated and the redox state of the cysteines in Tim17 was analyzed. (F) Strains were generated that overexpress Cyb2-dihydrofolate reductase (DHFR) and Cyb2Δ-DHFR from a GAL10 promoter. Mitochondria were isolated from these strains as well as from controls. The redox state of Tim17 was analyzed as described for B. (G) Topology of Tim17 indicating the position of the disulfide bond. IM, inner membrane.

Figure 2.

Tim17 import depends on Mia40. (A and B) ³⁵S-labeled Tim17 or Tim17AA was incubated with wild-type mitochondria at 25°C for the times indicated. Nonimported protein was removed by treatment with proteinase K (PK). Samples were split into two fractions. One was directly subjected to SDS-PAGE and autoradiography (Total), from the other mitochondria were incubated with 0.1 M sodium carbonate for 30 min on ice before carbonate-resistant proteins were pelleted by ultracentrifugation at 100,000 g for 30 min. Western blot signals of the integral membrane protein Oxa1 and the nonmembrane protein Sod1 are shown for control. Quantified signals are shown in B. (C) Model for the import and membrane insertion of Tim17. (D and E) Tim17 or Cmc1, respectively, were incubated with mitochondria from Mia40-containing (Mia40↑) or Mia40-depleted (Mia40↓) cells in the presence or absence of membrane potential (Δψ). The soluble IMS protein Cmc1 is an established Mia40 substrate. (F–H) Import experiments of Tim17 and Tim17AA with wild-type (WT) and mia40-4 mitochondria. (I and J) GAL-Mia40 cells expressing either Mia40 or an SPS mutant or an F315,318E mutant of Mia40 lacking the redox-active cysteines were cultured on glucose for 72 h. Mitochondria were isolated and used for import experiments with radiolabeled Tim17 or Cmc1, respectively.

Figure 3.

Tim17 directly binds to Mia40. (A) Radiolabeled Tim17, Tim17AC, and Tim17AA were imported into wild-type mitochondria for 2 min. The cross-linker DSP was added to the samples for 5 min. Cross-linking was blocked by addition of 100 mM glycine. Mitochondria were reisolated, lysed with 1% SDS, and subjected to immunoprecipitation with antibodies against Mia40 or with preimmune serum (pi). Samples were dissolved in sample buffer containing or lacking 10 mM DTT and analyzed by SDS-PAGE. 10% of the extract used per immunoprecipitation sample was loaded for control (T, total). Arrowheads depict the signals of the Tim17 protein that was pulled down together with Mia40. (B) Tim17 and Tim17AA were imported into Mia40↑ and Mia40↓ mitochondria for 30 min. Nonimported material was removed by protease treatment. Subsequently, the redox state of the imported Tim17 proteins was determined as described for Fig. 1. To show the maximal shift, samples were boiled in SDS and TCEP before incubation with mmPEG₂₄ (lanes 2 and 6). The percent values show the signal intensity of the Mia40-depleted mitochondria relative to those containing Mia40. (C) Tim17 was imported into Mia40↑ and Mia40↓ mitochondria for the indicated time periods. Samples were protease treated and TCA precipitated, and the redox state of the imported Tim17 was analyzed by mmPEG₂₄ treatment as described for Fig. 1 B.

Figure 4.

Erv1 plays a direct role in the import and oxidation of Tim17. (A) Radiolabeled Tim17 was imported into wild-type (WT) and erv1-ts mitochondria for the times indicated. (B) The redox state of the Tim17 protein imported into erv1-ts mitochondria was analyzed as described for Fig. 3 C. (C) Radiolabeled Tim17, Tim17AC, and Tim17CA were imported into wild-type mitochondria for 2 min. The import was stopped by the addition of cold 1.2 M sorbitol, 20 mM Hepes, pH 7.4, plus 150 mM N-ethyl maleimide. Mitochondria were reisolated, lysed with 1% SDS in the absence or presence of 10 mM DTT and subjected to immunoprecipitation with antibody against Erv1 or with preimmune serum (pi). 10% of the extract used per immunoprecipitation sample was loaded for control (T, total). Arrowheads depict the signals of the Tim17 protein that was pulled down together with Erv1. (D and E) Coimmunoprecipitation experiments of Tim17 after incubation of mitochondria in the presence or absence of membrane potential as indicated and as described for C. (F) Mitochondria were pretreated with 0.5 µM valinomycin to dissipate the membrane potential (Δψ) and incubated with radiolabeled Tim17 for the times indicated. Mitochondria were reisolated, washed, and TCA precipitated. Then, the redox state of Tim17 was analyzed by incubation with mmPEG₂₄. Here, the pattern of the incubated Tim17 was identical to that of Tim17 that was boiled in SDS in the presence of TCEP, indicating that in the absence of membrane potential, all cysteine residues in Tim17 remained reduced. (G) Hypothetical model for the disulfide bond formation in Tim17.

Figure 5.

Mutants lacking the structural disulfide bond in Tim17 are temperature sensitive. (A) Plasmid shuffling to generate cysteine to alanine mutants in Tim17. Wild-type (WT) cells or Δtim17 cells containing the wild-type gene of TIM17 on URA3- or HIS3-containing plasmids or the TIM17AA allele on a HIS3-containing plasmid were dropped onto the indicated media. The inability to grow on uracil-deficient medium (SD-URA) and the growth on medium containing 5-fluoro orotic acid indicate the successful loss of the URA3-plasmid containing TIM17 by the shuffling strategy. (B) The indicated strains were grown to log phase and diluted to OD 0.5 to 0.0005, from which 3 µl was spotted onto plates containing glucose or glycerol as carbon sources. (C) Δtim17 mutants expressing either Tim17 or Tim17AA were grown in microtiter plates at 30°C, and the OD was continually measured. (D) The indicated amounts of mitochondrial protein were applied to Western blotting using the indicated antibodies. (E) Radiolabeled Tim17, Tim17AA, and Tim23 were imported for the indicated times into wild-type mitochondria in the presence or absence of membrane potential. Nonimported protein was removed with proteinase K. Samples were solubilized with 1% digitonin and analyzed by BN-PAGE. The asterisk depicts the complex that is found with Tim17AA but barely with Tim17. (F) Isolated mitochondria from the indicated strains were solubilized in buffer containing 1% digitonin and analyzed by BN-PAGE and Western blotting using the indicated antibodies.

Figure 6.

The disulfide bond in Tim17 is critical for mitochondrial protein import. (A–C) The indicated radiolabeled preproteins were incubated with mitochondria at 25°C for the times indicated. Nonimported material was removed by protease treatment. Mitochondria were washed, reisolated, and subjected to SDS-PAGE and autoradiography. Signals of the mature imported proteins were quantified. The dashed lines show the intermediate forms of cytochrome b₂. (D and E) As in the above experiments, but the mitochondria were pretreated with import buffer at 37°C before the import reaction at 25°C. (F) The abundance of mitochondrial proteins in Tim17, Tim17CA, and Tim17AA were analyzed by SILAC-based quantitative mass spectrometry. The ratios of the values in the mutants relative to those in the wild type were normalized to the levels of carrier proteins (red dots).

Figure 7.

The disulfide bond in Tim17 is critical for the dynamic gating properties of the TIM23 channel. (A and B) Current–voltage plots (A) and sample current traces (B) are shown to illustrate the behavior and flickering rate of TIM23 channels recorded from wild type and Tim17AA mutant. Single channel containing patches where excised under voltage-clamp conditions, and the voltage was ramped from −40 to 40 mV (A) or held at −40 mV (B). The perfusion chamber and the patch microelectrode contained 150 mM KCl and 5 mM Hepes, pH 7.4. Current was recorded before (Control) and after sequential perfusion of the bath side with 20 µM of the presequence signal peptide yCoxIV_(1–13). Dotted lines in current traces (B) correspond to the closed state, and the total amplitude histograms (40–60-s duration) on the side define the channel’s substates registered and their corresponding probability as the percentage of time spent under each current level. Three different behaviors are reported in each case, corresponding to the wild-type channel (Tim17 WT) and the modified and frozen ones of the Tim17AA mutant. (C) Model for the role of the disulfide bond in Tim17 for the stabilization of the TIM23 complex. In the absence of the disulfide bond in Tim17, the gating of the TIM23 complex is severely compromised.