All small Tim proteins of the mitochondrial intermembrane space contain two conserved CX3C motifs, which form two intramolecular disulfide bonds essential for function, but only the cysteine-reduced, but not oxidized, proteins can be imported into mitochondria. We have shown that Tim10 can be oxidized by glutathione under cytosolic concentrations. However, it was unknown whether oxidative folding of other small Tims can occur under similar conditions and whether oxidative folding competes kinetically with mitochondrial import. In the present study, the effect of glutathione on the cysteine-redox state of Tim9 was investigated, and the standard redox potential of Tim9 was determined to be approx. −0.31 V at pH 7.4 and 25 °C with both the wild-type and Tim9F43W mutant proteins, using reverse-phase HPLC and fluorescence approaches. The results show that reduced Tim9 can be oxidized by glutathione under cytosolic concentrations. Next, we studied the rate of mitochondrial import and oxidative folding of Tim9 under identical conditions. The rate of import was approx. 3-fold slower than that of oxidative folding of Tim9, resulting in approx. 20% of the precursor protein being imported into an excess amount of mitochondria. A similar correlation between import and oxidative folding was obtained for Tim10. Therefore we conclude that oxidative folding and mitochondrial import are kinetically competitive processes. The efficiency of mitochondrial import of the small Tim proteins is controlled, at least partially in vitro, by the rate of oxidative folding, suggesting that a cofactor is required to stabilize the cysteine residues of the precursors from oxidation in vivo.

INTRODUCTION

Understanding the molecular mechanism of mitochondrial protein import is a fundamentally important issue in biochemistry and cell biology, because the mitochondrion is an essential organelle of the cell. Approx. 99% of the total mitochondrial proteins are synthesized in the cytosol and have to be imported into mitochondria for biogenesis of this organelle [13]. The small Tim proteins of the mitochondrial IMS (intermembrane space) play an essential role during the import of the mitochondrial inner- and outer-membrane proteins [4,5]. There are five homologous small Tim proteins in yeast, and the proteins are evolutionarily conserved throughout the eukaryotic kingdom. All members of the small Tim family have a molecular mass of approx. 10 kDa and contain a strictly conserved twin CX3C zinc-finger motif [6].

Tim9 and Tim10 are the two most important members of the small Tim family in yeast, which form a hexameric complex possessing a chaperone-like activity probably to prevent aggregation of the hydrophobic membrane proteins in the aqueous IMS. We have previously shown that both Tim9 and Tim10 are imported individually from the cytosol in a cysteine-reduced form, and the oxidized (disulfide-bonded) proteins are import-incompetent [7]. While keeping the protein in a reduced form is essential for its mitochondrial import, only the oxidized proteins can form the Tim9–Tim10 complex in the mitochondrial IMS [7]. The complex consists of three molecules of Tim9 and Tim10, with each subunit having two intramolecular disulfide bonds [79]. Interestingly, it has been shown that zinc can bind to the reduced but not to the oxidized protein, at a molar ratio of 1:1 in vitro, with a sub-nanomolar dissociation constant, suggesting a potential role for zinc in protecting the reduced protein from oxidation [10,11].

Detailed mitochondrial import studies have shown that import of the small Tim proteins does not require the TOM (translocase of the outer membrane) receptors of the outer membrane and it is ATP-independent [12]. It has been suggested that import may be a passive translocation process with the reduced proteins traversing the protein-conducting channel of the TOM complex freely [12,13]. On the other hand, recent studies have shown that import of the small Tim proteins and many other IMS proteins that could potentially form disulfide bonds is mediated by a newly identified Mia40 import machinery in a redox-sensitive manner [1417] (Mia40 is an essential protein of the mitochondrial IMS; synonyms YKL195W, FMP15 and Tim40). Mia40-substrate import intermediates linked by an intermolecular disulfide bond are observed during import of the substrate proteins. Import studies with cysteine mutants of Tim9 and Tim10, mutated systematically one by one, demonstrate that Mia40 specifically recognizes the first cysteine of the proteins during their import and thus acts as an import receptor in the mitochondrial IMS [18,19]. Taken together, all these results are consistent with only the reduced small Tim proteins being import-competent.

Glutathione, present in its reduced (GSH) and oxidized (GSSG) forms, is considered to be the major thiol–disulfide redox buffer of the cell [20]. Most of the glutathione in cells is usually found in the cytosol, which is the principal site of GSH biosynthesis, with the glutathione concentration in cells typically between 1 and 13 mM [2022]. The ratio of GSH to GSSG has been reported to range between 30:1 to 3000:1 in the cytosol. Our recent redox stability studies on Tim10 have shown that thermodynamically Tim10 can be oxidized by glutathione under cytosolic concentrations, and the presence of zinc ions can inhibit the oxidative folding of Tim10 kinetically [23]. Thus zinc binding is probably required to stabilize the precursor proteins in a reduced and import-competent form in the cytosol. However, oxidative folding studies with purified apoTim10 showed that the rate of the oxidation, under the cytosolic glutathione conditions, seems much slower than that of the protein import. Thus it is unclear whether import of the small Tim proteins can actually be inhibited by oxidative folding. So far, no study on the correlation between oxidative folding and mitochondrial import of the small Tim proteins has been reported; therefore, this is one of the main aims of the present study. In addition, all the small-Tim-protein-import studies so far have demonstrated that import of these proteins was typically approx. 10–20% of the total material or less [7,8,12,15], which is much less efficient than that of other mitochondrial subcompartment proteins, such as AAC (ADP–ATP carrier) and many matrix proteins [2426]. Why import of the IMS small Tim proteins is so inefficient is an important issue that needs to be addressed.

In the present study, we investigated the effect of glutathione on the redox states of recombinant yeast Saccharomyces cerevisiae Tim9 and the redox stability of the protein, using a thiol modification assay. The standard redox potential of Tim9 was determined to be −0.31 V at pH 7.4, showing that oxidized Tim9 is the thermodynamically stable form under cytosolic glutathione conditions. To determine whether oxidative folding will compete directly with mitochondrial import of the small Tim proteins kinetically, time courses of import and oxidative folding of Tim9 in the absence of Zn2+ were studied at identical conditions. The data analyses showed that the rate of oxidation was approx. 3-fold faster than that of mitochondrial import. Import was inhibited once the reduced Tim9 was oxidized, and only approx. 20% of the total precursor protein was imported into mitochondria added in excess. The same results were obtained with different preparations and with Nycodenz-gradient-purified mitochondria, showing that the fast oxidative folding was not due to the effect of a contamination in mitochondria. A similar correlation between import and oxidative folding for Tim10 was obtained. We conclude that oxidative folding is a strong competitor and can inhibit mitochondrial import of apoTim9 and apoTim10 kinetically. The efficiency of mitochondrial import of the small Tim proteins is controlled, at least partially, by the rate of oxidative folding.

EXPERIMENTAL

Materials

TCEP [tris-(2-carboxyethyl)phosphine] and AMS (4-acetamido-4′-maleimidylstilbene-2,2′-disulfonic acid) were obtained from Invitrogen Molecular Probes. EDTA was from BDH, and all other chemicals were obtained from Sigma at the highest grade.

Protein purification

Wild-type Tim9 and Tim9F43W mutant were purified from Escherichia coli using the same protocol for Tim10 as described previously [23,27]. Briefly, BL21-Codon plus cells (Stratagene) containing a plasmid expressing GST (glutathione transferase)–Tim9 were grown in LB (Luria–Bertani) medium containing ampicillin (0.1 mg/ml) at 30 °C until the D600 (attenuance at 600 nm) was approx. 0.4 and then the cells were induced for 3 h by IPTG (isopropyl β-D-thiogalactoside). The cells were harvested and lysed by sonication at 4 °C in buffer A (50 mM Tris/HCl, pH 7.4, 150 mM NaCl and 1 mM EDTA) plus 10 mM DTT (dithiothreitol). Most of the GST–Tim9 was isolated in inclusion bodies (the pellet), which was solubilized in 10 ml of buffer B (buffer A containing 8 M urea and 10 mM DTT) for 1 h at room temperature, and renatured for 2 h at 4 °C by 10-fold dilution with buffer A. The solubilized Tim9 was incubated overnight at 4 °C with approx. 2 ml of glutathione–Sepharose 4B beads equilibrated with buffer A. The beads were washed and incubated with buffer A containing 5 units/ml thrombin at 4 °C to release Tim9. Tim9 was further purified using an FPLC gel filtration column (Superdex 75; Amersham Biosciences) with buffer A at 0.5 ml/min.

Preparation of reduced Tim9

Oxidized Tim9 was typically incubated with 2 mM TCEP for approx. 1 h at 25 °C and pH 7.4, followed by gel filtration (Superdex 75) to remove TCEP. The reduced protein was always prepared freshly just before use.

AMS alkylation assay

The purified oxidized or reduced Tim9 (typically approx. 10 μM) in buffer A was incubated with TCEP, GSSG and GSH under various conditions according to the individual experiment, followed by addition of 5–20 mM thiol-specific reagent AMS and non-reducing SDS/PAGE sample buffer at 25 °C for approx. 15 min in the dark. Then samples were analysed by Tricine/SDS/16%-(w/v)-PAGE followed by CBB (Coomassie Brilliant Blue) staining. In this assay, the alkylation agent AMS, a 500 Da molecule, reacts specifically and efficiently with free thiols of the reduced protein, but not disulfide bonds of the oxidized protein. AMS modification increases the mass of the protein and thus decreases its mobility on SDS/PAGE. To avoid any potential oxidation of GSH, samples were prepared in freshly deoxygenated buffer A.

Fluorescence measurements

Fluorescence spectra were acquired using a Cary Eclipse fluorescence spectrophotometer. A fluorescence cell of 10 mm×2 mm was used. Fluorescence spectra at λexcitation=280 nm and λemission=290–500 nm were recorded. A slit width of 5 nm was used for both excitation and emission.

RP-HPLC (reverse-phase HPLC)

RP-HPLC (the Pharmacia Biotech SMART system) was used to separate the reduced and oxidized forms of Tim9. All samples were acid-quenched with 0.1% TFA (trifluoroacetic acid), both to prepare the samples for RP-HPLC and to prevent any further disulfide exchange, prior to loading on to an Ace 5 C4-300 reverse-phase column. Buffer 1 was 0.1% TFA in water, and buffer 2 was 0.1% TFA in acetonitrile. The column was equilibrated with 10% buffer 2 prior to sample injection. Separation was performed by running a linear gradient from 10–60% buffer 2, with a flow rate of 200 μl/min. The injection volume was 100 μl.

Determination of redox potential

Oxidized Tim9 in buffer A plus 1 mM EDTA was incubated with DTT and OxDTT (oxidized form of DTT) at various ratios: 0.5 mM OxDTT plus 0.5 μM–20 mM DTT, at 25 °C for 2 h to overnight in the cases where reduction was slow (low DTT concentration). For the wild-type Tim9, reactions were acidified by the addition of 0.1% TFA and subjected to RP-HPLC. The percentage of Tim9 in the oxidized form was calculated as area under the curve of absorbance at 215 nm. For the Tim9F43W, fluorescence spectra before and after incubation with DTT/OxDTT were measured. Fluorescence intensity changes at 350±5 nm were used for further data analyses. The equilibrium constant Keq for redox reaction 1:

 
formula

was calculated according to eqn (1):

 
formula
(1)

Y is the measured fluorescence intensity of mutant Tim9 at various ratios of DTT/OxDTT; YR and YO are the calculated fluorescence intensity of reduced (R) and oxidized (O) states respectively. In the case of wild-type proteins, Y is the fractions of the oxidized protein at various DTT/OxDTT concentrations, and YO and YR are set at 100% and 0 respectively.

Then, the standard redox potential was calculated according to the Nernst equation (eqn 2) at 25 °C with the number of electrons n=4 (eqn 2):

 
formula
(2)

The standard redox potential of DTT (E0, DTT/OxDTT), −330 mV, at 25 °C and pH 7.4 was used, based on the equilibrium constant for the reaction GSSG+DTT=2GSH+OxDTT, which is approx. 200 M [28,29].

Mitochondria isolation

The wild-type yeast strain D273-10B (MATα) was grown at 30 °C, and the T9ts strain was grown at 24 °C for 30 h, followed by a further 8 h growth at 37 °C. Both strains were grown in media containing 2.2% (v/v) lactic acid and 0.3% (w/v) yeast extract, with the pH adjusted to 5.5, until the D600 was ∼1.5–2.0. Then, mitochondria were isolated as previously described [30].

Mitochondria import

35S-labelled proteins were synthesized using the TNT SP6 coupled transcription/translation kit (Promega). Typically, 5 μl of lysate was added to 150 μl of import reaction mixture containing mitochondria at 0.5 mg/ml in sorbitol-based import buffer, consisting of 0.6 M sorbitol, 50 mM KCl, 0.75 mg/ml L-methionine, 1 mg/ml fatty acid-free BSA, 50 mM Hepes/KOH (pH 7.4) and 2 mM EGTA. Import was performed at 25 °C for the times indicated and stopped by decanting an aliquot into a test tube containing 5 vol. of ice-cold 0.6 M sorbitol plus 50 μg/ml trypsin to stop import and to digest any surface-bound un-imported materials for 10 min. The mitochondria were re-isolated followed by resuspension in gel sample buffer for Tris/Tricine SDS/PAGE analysis and visualized by autoradiography. For analysis of the redox state of un-imported proteins, aliquots were removed from the import reaction and mitochondria were removed by centrifugation. The supernatant was added to sample buffer containing 10 mM AMS for thiol modification. Then, samples were subjected to non-reducing Tris/Tricine SDS/PAGE separation and visualized by autoradiography. The result was quantified by two-dimensional densitometry using Aida Image Analyser V.4.00, and the data were further analysed using a single or a double exponential function: y=y0+A1×exp(−k1×x)+A2×exp(−k2×x).

RESULTS AND DISCUSSION

Effects of glutathione on the cysteine-redox state of Tim9

Glutathione (GSH/GSSG) is considered the major thiol-disulfide redox buffer of the cytosol and has a standard redox potential of approx. −0.26 V at pH 7.4 [20]. Therefore it is important to know the relative redox stability of a protein to glutathione, in order to ascertain whether disulfide bonds could form within the cytosol. To address this issue, the effects of GSH and GSSG on the redox state of purified Tim9 were investigated. First, oxidized Tim9, containing two native intramolecular disulfide bonds, was incubated with GSH at various concentrations for 2 h. Then, excess amount of thiol-specific reagent AMS was added to react covalently with reduced cysteine thiol groups of the protein. Under these conditions, reduced but not oxidized Tim9 was modified, with an increased molecular mass of approx. 0.5 kDa per thiol. The redox state was examined using non-reducing SDS/PAGE (Figure 1a). A clear difference in mobility of approx. 2 kDa between the oxidized (unmodified) and the TCEP-reduced Tim9 (Figure 1, lanes 1 and 2) was shown, which is consistent with our previous observation of reductive unfolding in the presence of DTT. More importantly, the result showed that oxidized Tim9 was stable against glutathione reduction; no reduced Tim9 was observed even in the presence of 10 mM GSH. Next, reduced Tim9 was incubated with GSH/GSSG at various conditions, followed by the AMS assay as described above (Figure 1b). The results showed that reduced Tim9 can be oxidized by GSSG under physiological glutathione redox conditions, even in the presence of as little as 2 μM GSSG and 2500-fold of GSH (Figure 1). Taken together, these results show that oxidized Tim9 is the thermodynamically stable form under the cytosolic glutathione conditions. The newly synthesized reduced protein can be oxidized in the cytosol in the absence of a cofactor to stabilize it.

Effect of glutathione on the redox state of Tim9 measured by AMS assays

Figure 1
Effect of glutathione on the redox state of Tim9 measured by AMS assays

CBB-stained non-reducing Tricine SDS/PAGE of (a) untreated oxidized Tim9 (lane 1) and AMS-treated Tim9 (lanes 2–7). The protein was pre-incubated with 1 mM TCEP (lane 2) or 0–10 mM GSH (lanes 3–7) respectively at 25 °C for 2 h prior to AMS treatment. The reduced (R) and oxidized (O) states are indicated. (b) Reduced Tim9 was pre-incubated with 5 mM GSH plus 2–1000 μM GSSG (lanes 2–10), followed by the AMS assay. Lane 1 is the AMS-treated starting material of reduced Tim9. The reduced (R) and oxidized (O) states are indicated. 2.5 k, 2500.

Figure 1
Effect of glutathione on the redox state of Tim9 measured by AMS assays

CBB-stained non-reducing Tricine SDS/PAGE of (a) untreated oxidized Tim9 (lane 1) and AMS-treated Tim9 (lanes 2–7). The protein was pre-incubated with 1 mM TCEP (lane 2) or 0–10 mM GSH (lanes 3–7) respectively at 25 °C for 2 h prior to AMS treatment. The reduced (R) and oxidized (O) states are indicated. (b) Reduced Tim9 was pre-incubated with 5 mM GSH plus 2–1000 μM GSSG (lanes 2–10), followed by the AMS assay. Lane 1 is the AMS-treated starting material of reduced Tim9. The reduced (R) and oxidized (O) states are indicated. 2.5 k, 2500.

Determination of the standard redox potential of Tim9

The above studies show that Tim9 is a stronger reductant than glutathione and probably has a standard redox potential below −0.3 V. Next, in the present study, two alternative methods were developed to determine the standard redox potential of Tim9. We chose a stronger disulfide-bond reducing agent, DTT, and its oxidized form (OxDTT) (E0,DTT/OxDTT=−0.33 V) [23,28] as the redox buffer, coupled with the use of RP-HPLC or fluorescence measurements in determining the redox potential of Tim9.

First, the oxidized wild-type Tim9 was incubated with various ratios of DTT/OxDTT in nitrogen-saturated buffer A (50 mM Tris, 150 mM NaCl and 1 mM EDTA, pH 7.4). Then, the mixtures of the oxidized and reduced proteins were acidified and separated by RP-HPLC (Figure 2a and the Experimental section). The relative amounts of the oxidized and reduced proteins were calculated based on the area of the corresponding peaks at approx. 25.8 ml for the oxidized form (O) and at 30.5 ml for the reduced form (R) respectively. There were also a few very small peaks that were eluted between the two major ones, possibly due to impurity and/or presence of single-disulfide-bond species, which were not included in our data analysis. The fractions of oxidized Tim9 were plotted against ratios of DTT/OxDTT as shown in Figure 2(b). The redox equilibrium constant (Keq) for Reaction 1 (see above) was determined based on eqn (1) (see the Experimental section) to be 90±30. Then, the standard redox potential of Tim9 (E0,Tim9) was calculated according to the Nernst equation (eqn 2 in the Experimental section) to be approx. −0.3 V. A potential problem for the E0,Tim9 determination is that the acid and organic solvents used for the protein separation by RP-HPLC may cause loss of some reduced and oxidized Tim9 to a different extent. Thus an alternative approach was required to confirm the redox potential of Tim9.

Determination of the standard redox potential of Tim9

Figure 2
Determination of the standard redox potential of Tim9

(a) RP-HPLC profiles of the wild-type Tim9 after incubation with 0.5 mM OxDTT and various DTT concentrations, from 0.5 μM to 20 mM, for 16 h at 25 °C (see the Experimental section). (b) The fractions of the oxidized Tim9 were calculated based on the areas of the oxidized and reduced proteins shown in (a) and plotted against the ratio of DTT/OxDTT. The data were analysed as described in the Experimental section. The equilibrium constant (Keq) of Tim9 reduction and the standard redox potential of Tim9, at 25 °C and pH 7.4, were determined to be 90±30 and −0.3 V respectively. (c) The differential fluorescence spectra of Tim9F43W, after and before the protein was incubated with the various DTT/OxDTT buffers as described above. (d) Fluorescence intensity change at 350 nm was plotted against the ratio DTT/OxDTT, and the data were analysed as described in the Experimental section. The equilibrium constant and the standard redox potential for Tim9F43W, at 25 °C and pH 7.4, were determined to be 33±6 and −0.31 V respectively.

Figure 2
Determination of the standard redox potential of Tim9

(a) RP-HPLC profiles of the wild-type Tim9 after incubation with 0.5 mM OxDTT and various DTT concentrations, from 0.5 μM to 20 mM, for 16 h at 25 °C (see the Experimental section). (b) The fractions of the oxidized Tim9 were calculated based on the areas of the oxidized and reduced proteins shown in (a) and plotted against the ratio of DTT/OxDTT. The data were analysed as described in the Experimental section. The equilibrium constant (Keq) of Tim9 reduction and the standard redox potential of Tim9, at 25 °C and pH 7.4, were determined to be 90±30 and −0.3 V respectively. (c) The differential fluorescence spectra of Tim9F43W, after and before the protein was incubated with the various DTT/OxDTT buffers as described above. (d) Fluorescence intensity change at 350 nm was plotted against the ratio DTT/OxDTT, and the data were analysed as described in the Experimental section. The equilibrium constant and the standard redox potential for Tim9F43W, at 25 °C and pH 7.4, were determined to be 33±6 and −0.31 V respectively.

Protein fluorescence is a sensitive, convenient and well-defined technique and can be used to study protein folding at native conditions; however, there is no tryptophan residue in Tim9, and fluorescence of the single tyrosine of Tim9 is negligible. We have shown previously that cleavage of Tim9 disulfide bonds is accompanied by protein unfolding. Therefore, in order to measure the standard redox potential and study the oxidative folding using the fluorescence technique, we made a tryptophan mutant by mutating the Phe43 at the loop between the two CX3C motifs. As expected, the Tim9F43W mutant showed the same characteristic as the wild-type protein in terms of far-UV CD spectra, complex formation with Tim10 and effects of glutathione on the redox state as studied by the AMS assay (results not shown). Importantly, Tim9F43W showed a significant difference in fluorescence intensities between the oxidized and reduced forms, with the intensity of the oxidized form being approximately three times higher than the reduced form (Figure 2c). Next, fluorescence spectra of the mutant before and after incubation with various DTT/OxDTT redox buffers for 16 h were measured, and the standard redox potential of Tim9F43W was calculated based on the tryptophan fluorescence intensity change at 350±5 nm (Figures 2c and 2d). The data were analysed using eqn (1) (see the Experimental section for details) and the Keq was calculated to be 33±6. The standard redox potential obtained for Tim9F43W was −0.31 V, which is highly similar to that for wild-type Tim9. Thus the results for both the wild-type and the mutant showed that Tim9 is a much stronger reductant than glutathione. Based on the Keq of 33 and the standard redox potential for GSH/GSSG and DTT/OxDTT of −0.26 and −0.33 V respectively at 25 °C and pH 7.4, the equilibrium constant for the reaction:

 
formula

was calculated to be 1.7×103 M2. This confirms that oxidized Tim9 is the thermodynamically stable and dominant form under cytosolic glutathione conditions. The standard redox potential of Tim9 is similar to that of Tim10 (−0.32 V) as we reported previously [23], suggesting that the small Tim proteins have very similar redox stability and may share the same oxidative folding mechanism. The fact that the standard redox potentials of these small Tim proteins are lower than that of PDI (protein disulfide-isomerase) (−0.18 V) [31] is thus consistent with the fact that PDI can catalyse the oxidative folding of these proteins in vitro [23]. Similarly, other thiol oxidoreductases, such as thioredoxin (−0.27 V) [32] and glutaredoxin (−0.2 to −0.23 V) [33], may be able to catalyse the oxidative folding of the small Tim proteins as well.

Oxidative folding competes with mitochondria import of Tim9 and Tim10

The above study shows that newly synthesized reduced Tim9 can be oxidized under physiological glutathione concentrations. Therefore we tested whether both oxidized and reduced Tim9 can be imported into mitochondria isolated from a temperature-sensitive yeast strain, tim9ts, in which both Tim9 and Tim10 are not detectable [34]. After incubation of the oxidized or reduced Tim9 with the mitochondria for 30 min, import was stopped and the mitochondria were treated with proteinase to remove any surface-bound Tim9. Then, imported Tim9 was analysed by SDS/PAGE followed by Western blotting. As shown in Figure 3(a), the reduced Tim9 can be imported into mitochondria, but the oxidized protein cannot. The result was the same as that demonstrated previously for Tim10 [7].

Mitochondrial import and oxidative folding of purified (a) and 35S-radiolabelled (b–e) Tim9

Figure 3
Mitochondrial import and oxidative folding of purified (a) and 35S-radiolabelled (b–e) Tim9

(a) Western blotting of Tim9 imported using reduced or oxidized recombinant protein and mitochondria isolated from tim9ts yeast. (b) Time course of import of [35S]Tim9 into isolated mitochondria from yeast, analysed by reducing SDS/PAGE. (c) AMS assay of the un-imported 35S-Tim9 during the time course experiment shown in (b). The reduced (R), partially oxidized intermediate (I) and fully oxidized (O) states are indicated. (d) The quantified level of import as shown in (b) was plotted against time, and the data were analysed with a single exponential function (continuous line) giving a rate constant of 0.07±0.02 min−1. (e) The percentage of un-imported Tim9 as reduced (R), intermediate (I), fully oxidized (O) or intermediate plus fully oxidized (I+O) was plotted against time and analysed with a single- or double-exponential function. The calculated rate constant for disappearance of the reduced Tim9 was 0.2±0.06 min−1, and formation of the fully oxidized Tim9 was 0.025±0.005 min−1. The error bars shown in (d, e) represent S.E.M. for three sets of repeat experiments.

Figure 3
Mitochondrial import and oxidative folding of purified (a) and 35S-radiolabelled (b–e) Tim9

(a) Western blotting of Tim9 imported using reduced or oxidized recombinant protein and mitochondria isolated from tim9ts yeast. (b) Time course of import of [35S]Tim9 into isolated mitochondria from yeast, analysed by reducing SDS/PAGE. (c) AMS assay of the un-imported 35S-Tim9 during the time course experiment shown in (b). The reduced (R), partially oxidized intermediate (I) and fully oxidized (O) states are indicated. (d) The quantified level of import as shown in (b) was plotted against time, and the data were analysed with a single exponential function (continuous line) giving a rate constant of 0.07±0.02 min−1. (e) The percentage of un-imported Tim9 as reduced (R), intermediate (I), fully oxidized (O) or intermediate plus fully oxidized (I+O) was plotted against time and analysed with a single- or double-exponential function. The calculated rate constant for disappearance of the reduced Tim9 was 0.2±0.06 min−1, and formation of the fully oxidized Tim9 was 0.025±0.005 min−1. The error bars shown in (d, e) represent S.E.M. for three sets of repeat experiments.

Next, we asked whether the folding and import processes are kinetically controlled. In other words, if oxidative folding is a kinetically unfavoured slow process compared with the rate of protein import, it will have little effect on the mitochondrial import. On the other hand, protein import will be inhibited by fast folding.

To understand the correlation between oxidative folding and mitochondria import in a biologically relevant condition, a cell-free import system was used. It allows us to study the rate of mitochondrial import and oxidative folding of Tim9 under identical conditions. Radioactive labelled Tim9 ([35S]Tim9) synthesized in the rabbit reticulocyte lysate was incubated with mitochondria isolated from yeast in a standard import buffer. After various times of incubation, import was stopped and mitochondria were separated from the un-imported material. While the isolated mitochondria were treated with trypsin to remove any surface-bound materials, the un-imported materials were treated immediately with AMS for the protein redox-state analysis. Then, both imported (Figure 3b) and un-imported (Figure 3c) Tim9 were analysed using SDS/PAGE and autoradiography. As expected, the level of import increased with time and reached a plateau by approx. 30 min (Figure 3b). However, only approx. 20% of the total material was imported into mitochondria. Nearly all of the un-imported Tim9 was fully (O) or partially (I) oxidized by 30 min (Figure 3c), showing that import was accompanied by protein oxidation. The data were quantified and further analysed using a single exponential function for the mitochondrial import (Figure 3d), as well as the reduced and fully oxidized un-imported Tim9 (Figure 3e). The analyses show that the level of reduced protein decreased with a rate constant of 0.2±0.06 min−1 and the fully oxidized Tim9 formed at a rate of 0.025±0.005 min−1, whereas the rate of import was 0.07±0.02 min−1. Thus the import was approximately three times slower than the first step of the oxidative folding, and approximately three times faster than the formation of fully oxidized protein. For the intermediate (I), the data were not clear enough to give an independent kinetic parameter, but it was well described by the double exponential function using the two rate constants obtained from the reduced and fully oxidized proteins (Figure 4e). Thus the fully oxidized Tim9 was formed after formation of a single-disulfide-bond intermediate, and its formation was approx. 10-fold slower than that of the partially oxidized form. The increase in partially and fully oxidized proteins (I+O), which fitted the same rate constant as the reduced Tim9, was shown as well (Figure 3e).

Import and AMS assay of the un-imported [35S]Tim9

Figure 4
Import and AMS assay of the un-imported [35S]Tim9

Time course of oxidation of [35S]Tim9 during import into mitochondria isolated (a) without or (b) with further purification by Nycodenz gradient. (c) Time course of [35S]Tim9 mitochondrial import in the absence (control) or presence of 1 mM TCEP (‘+TCEP’), analysed by reducing SDS/PAGE and autoradiography. (d) AMS assay of the un-imported material from the time courses shown in (c). An observed degradation band of [35S]Tim9 seen in the presence of TCEP is indicated by ‘*’. (e) The quantified level of import for the control (grey) and in the presence of TCEP (black) in (c) plotted against time. Error bars represent the S.E.M. for three independent experiments. (f) Time course of [35S]Tim9 import into mitochondria in the presence of 10 mM GSH; samples were analysed by SDS/PAGE and autoradiography. (g) AMS assay of the un-imported materials in (f). The reduced, intermediate and oxidized states in the AMS assays are indicated by R, I and O respectively.

Figure 4
Import and AMS assay of the un-imported [35S]Tim9

Time course of oxidation of [35S]Tim9 during import into mitochondria isolated (a) without or (b) with further purification by Nycodenz gradient. (c) Time course of [35S]Tim9 mitochondrial import in the absence (control) or presence of 1 mM TCEP (‘+TCEP’), analysed by reducing SDS/PAGE and autoradiography. (d) AMS assay of the un-imported material from the time courses shown in (c). An observed degradation band of [35S]Tim9 seen in the presence of TCEP is indicated by ‘*’. (e) The quantified level of import for the control (grey) and in the presence of TCEP (black) in (c) plotted against time. Error bars represent the S.E.M. for three independent experiments. (f) Time course of [35S]Tim9 import into mitochondria in the presence of 10 mM GSH; samples were analysed by SDS/PAGE and autoradiography. (g) AMS assay of the un-imported materials in (f). The reduced, intermediate and oxidized states in the AMS assays are indicated by R, I and O respectively.

Clearly, the import was inhibited as soon as the fully reduced Tim9 was gone, although the partially oxidized Tim9 was still populated. Together with the result that only approx. 20% of the total protein was imported into the excess amount of mitochondria used in the experiment, our study suggests that the partially oxidized Tim9 intermediate(s) may be largely importincompetent as well. If the partially oxidized Tim9 were import-competent, a higher level of import would be expected as the fully oxidized Tim9 was formed approx. three times slower than Tim9 import; statistically, up to 75% of the total precursor protein would be expected.

We have shown previously that oxidative folding of the small Tim proteins can be accelerated by PDI in vitro [23]. There are many PDI-like oxidoreductases in the ER (endoplasmic reticulum) responsible for disulfide bond formation within this organelle [3538]. Thus one potential reason for the fast oxidative folding during the mitochondrial import experiments is the presence of contaminants from the ER and/or other organelles in the isolated mitochondria, which can accelerate the oxidation of these proteins. To address this issue, we carried out the import experiment using different preparations of mitochondria and mitochondria isolated with or without further purification by Nycodenz gradient, and very similar results were obtained (Figure 4). This confirmed that the observed oxidative folding of the un-imported protein was not due to contamination of other cellular components. On the other hand, the presence of the disulfide-reducing agent, TCEP, maintains Tim9 in the reduced form and enhances the import efficiency, with approx. 2-fold increase in import level at 60 min (Figures 4c–4e). After approx. 30 min, the reduced protein was not stable and degraded (Figure 4d). Furthermore, the presence of 10 mM GSH does not have an obvious effect on the rates of both import and oxidation of Tim9 (Figures 4f and 4g), suggesting that the accelerator for Tim9 oxidation is independent of GSH. The factor(s) that is responsible for the fast folding is currently unknown, and it would be interesting to test, in future, whether the reactive oxygen species generated by mitochondria are responsible.

Next, to find out whether oxidative folding is a strong competitor for import of the small Tim proteins in general, we performed the same experiment with [35S]Tim10. As shown in Figure 5, Tim10 was imported into mitochondria with a similar time course to that of Tim9. Data analysis showed that the rate constant for Tim10 import was approx. 0.13 min−1, slightly faster than that of Tim9, whereas the overall level of import was approx. 11% and thus less efficient than that of Tim9 (approx. 20%). Unfortunately, the oxidative folding of the un-imported Tim10 was poorly resolved, and the reason is unclear. Although we were not able to get a quantitative measurement of the rate of Tim10 oxidation, the result of the AMS assay did suggest that the un-imported Tim10 was rapidly oxidized, with the reduced form disappearing in less than 15 min, at which point the import was inhibited. Thus the correlation between import and oxidation of Tim10 is consistent with that of Tim9 in general, demonstrating that oxidative folding can compete with and inhibit mitochondrial import of the small Tim proteins kinetically.

Mitochondrial import and oxidative folding of 35S-radiolabelled Tim10

Figure 5
Mitochondrial import and oxidative folding of 35S-radiolabelled Tim10

(a) Time course of import of [35S]Tim10 into isolated mitochondria from yeast, analysed by reducing SDS/PAGE. (b) AMS assay of the un-imported [35S]Tim10 during the time course experiment shown in (a). (c) The quantified level of import as shown in (a) was plotted against time, and the data were analysed with a single exponential function (continuous line) giving a rate constant of 0.13±0.02 min−1. The error bars represent the S.E.M. for three sets of repeat experiments.

Figure 5
Mitochondrial import and oxidative folding of 35S-radiolabelled Tim10

(a) Time course of import of [35S]Tim10 into isolated mitochondria from yeast, analysed by reducing SDS/PAGE. (b) AMS assay of the un-imported [35S]Tim10 during the time course experiment shown in (a). (c) The quantified level of import as shown in (a) was plotted against time, and the data were analysed with a single exponential function (continuous line) giving a rate constant of 0.13±0.02 min−1. The error bars represent the S.E.M. for three sets of repeat experiments.

It is noteworthy that import of the small Tim proteins is typically approx. 10–20% as shown previously by us and other research groups, based on the use of radioactive precursor and excess amount of mitochondria. In the present study, an explanation for the inefficient import of the small Tim proteins was provided. We demonstrated that oxidative folding is a significant factor limiting the mitochondrial import of small Tim proteins. Consistent with this hypothesis, in the presence of TCEP, the import level of Tim9 was increased approx. 2-fold. Although other factors may also play a role, the efficiency of the import of the small Tim proteins is at least partially controlled by the rate of oxidative folding. This result suggests that a cofactor is present in vivo to stabilize the cysteine from oxidation. Since zinc binding can stabilize the reduced Tim10 [23] and reduced Tim9 (B. Morgan and H. Lu, unpublished work) from oxidation in vitro, zinc ions as well as other components may play a role during import of the proteins, which is under investigation.

In summary, the present study has demonstrated that the newly synthesized reduced Tim9 in the absence of a cofactor is not stable and can be oxidized thermodynamically under the cytosolic glutathione conditions. The standard redox potential of Tim9 was determined, using two alternative methods developed in the present study, to be approx. −0.31 V at pH 7.4, and the equilibrium constant for the reaction between reduced Tim9 and GSSG was determined to be 1.7×103 M2. These results confirm that Tim9 is a stronger reductant than glutathione. Most importantly, we showed that for both Tim9 and Tim10, oxidative folding is a major competitive process that can inhibit the import of these proteins kinetically. Our results also suggest that both partially and fully oxidized proteins are incompetent for mitochondrial import, and cofactors are required to stabilize the precursors of the small Tim proteins in the reduced form in vivo.

We are grateful to Professor Neil Bulleid and Dr Martin Pool (both of the Faculty of Life Sciences, University of Manchester, Manchester, U.K.) for helpful comments on this paper, to Professor Kostas Tokatlidis (Institute of Molecular Biology and Biotechnology, Foundation of Research and Technology Hellas, Heraklion, Crete, Greece) for protein plasmids and to Professor Carla Koehler (Department of Chemistry and Biochemistry, University of California at Los Angeles, Los Angeles, CA, U.S.A.) for the yeast strain tim9ts. H. L. is supported by a Royal Society University Research Fellowship (2003). Research in H. L.'s laboratory is financially supported by The Royal Society and BBSRC (Biotechnology and Biological Sciences Research Council grants BB/C514323/1 and BBS/S/A/2004/10901). B. M. is sponsored by a BBSRC Committee Ph.D. Studentship (BBS/S/A/2004/10901).

Abbreviations

     
  • AMS

    4-acetamido-4′-maleimidylstilbene-2,2′-disulfonic acid

  •  
  • CBB

    Coomassie Brilliant Blue

  •  
  • DTT

    dithiothreitol

  •  
  • ER

    endoplasmic reticulum

  •  
  • GST

    glutathione transferase

  •  
  • IMS

    intermembrane space

  •  
  • Mia40

    an essential protein of the mitochondrial IMS (synonyms YKL195W, FMP15 and Tim40)

  •  
  • OxDTT

    oxidized form of DTT

  •  
  • PDI

    protein disulfide-isomerase

  •  
  • RP-HPLC

    reverse-phase HPLC

  •  
  • TCEP

    tris-(2-carboxyethyl)phosphine

  •  
  • TFA

    trifluoroacetic acid

  •  
  • TOM

    translocase of the outer membrane

References

References
1
Koehler
 
C. M.
 
New developments in mitochondrial assembly
Annu. Rev. Cell Dev. Biol.
2004
, vol. 
20
 (pg. 
309
-
335
)
2
Neupert
 
W.
Herrmann
 
J. M.
 
Translocation of proteins into mitochondria
Annu. Rev. Biochem.
2007
, vol. 
66
 (pg. 
863
-
917
)
3
Pfanner
 
N.
Wiedemann
 
N.
 
Mitochondrial protein import: two membranes, three translocases
Curr. Opin. Cell Biol.
2002
, vol. 
14
 (pg. 
400
-
411
)
4
Koehler
 
C. M.
 
The small Tim proteins and the twin Cx3C motif
Trends Biochem. Sci.
2004
, vol. 
29
 (pg. 
1
-
4
)
5
Pfanner
 
N.
Wiedemann
 
N.
Meisinger
 
C.
Lithgow
 
T.
 
Assembling the mitochondrial outer membrane
Nat. Struct. Mol. Biol.
2004
, vol. 
11
 (pg. 
1044
-
1048
)
6
Bauer
 
M. F.
Rothbauer
 
U.
Muhlenbein
 
N.
Smith
 
R. J.
Gerbitz
 
K.
Neupert
 
W.
Brunner
 
M.
Hofmann
 
S.
 
The mitochondrial TIM22 preprotein translocase is highly conserved throughout the eukaryotic kingdom
FEBS Lett.
1999
, vol. 
464
 (pg. 
41
-
47
)
7
Lu
 
H.
Allen
 
S.
Wardleworth
 
L.
Savory
 
P.
Tokatlidis
 
K.
 
Functional TIM10 chaperone assembly is redox-regulated in vivo
J. Biol. Chem.
2004
, vol. 
279
 (pg. 
18952
-
18958
)
8
Allen
 
S.
Lu
 
H.
Thornton
 
D.
Tokatlidis
 
K.
 
Juxtaposition of the two distal CX3C motifs via intrachain disulfide bonding is essential for the folding of Tim10
J. Biol. Chem.
2003
, vol. 
278
 (pg. 
38505
-
38513
)
9
Webb
 
C. T.
Gorman
 
M. A.
Lazarou
 
M.
Ryan
 
M. T.
Gulbis
 
J. M.
 
Crystal structure of the mitochondrial chaperone TIM9.10 reveals a six-bladed alpha-propeller
Mol. Cell
2006
, vol. 
21
 (pg. 
123
-
133
)
10
Ivanova
 
E.
Ball
 
M.
Lu
 
H.
 
Zinc binding of Tim10: evidence for existence of an unstructured binding intermediate for a zinc finger protein
Proteins: Struct. Funct. Bioinf.
2007
, vol. 
69
  
11
Lu
 
H.
Golovanov
 
A. P.
Alcock
 
F.
Grossmann
 
J. G.
Allen
 
S.
Lian
 
L. Y.
Tokatlidis
 
K.
 
The structural basis of the TIM10 chaperone assembly
J. Biol. Chem.
2004
, vol. 
279
 (pg. 
18959
-
18966
)
12
Lutz
 
T.
Neupert
 
W.
Herrmann
 
J. M.
 
Import of small Tim proteins into the mitochondrial intermembrane space
EMBO J.
2003
, vol. 
22
 (pg. 
4400
-
4408
)
13
Herrmann
 
J. M.
Hell
 
K.
 
Chopped, trapped or tacked – protein translocation into the IMS of mitochondria
Trends Biochem. Sci.
2005
, vol. 
30
 (pg. 
205
-
211
)
14
Terziyska
 
N.
Lutz
 
T.
Kozany
 
C.
Mokranjac
 
D.
Mesecke
 
N.
Neupert
 
W.
Herrmann
 
J. M.
Hell
 
K.
 
Mia40, a novel factor for protein import into the intermembrane space of mitochondria is able to bind metal ions
FEBS Lett.
2005
, vol. 
579
 (pg. 
179
-
184
)
15
Chacinska
 
A.
Pfannschmidt
 
S.
Wiedemann
 
N.
Kozjak
 
V.
Sanjuán Szklarz
 
L. K.
Schulze-Specking
 
A.
Truscott
 
K. N.
Guiard
 
B.
Meisinger
 
C.
Pfanner
 
N.
 
Essential role of Mia40 in import and assembly of mitochondrial intermembrane space proteins
EMBO J.
2004
, vol. 
23
 (pg. 
3735
-
3746
)
16
Mesecke
 
N.
Terziyska
 
N.
Kozany
 
C.
Baumann
 
F.
Neupert
 
W.
Hell
 
K.
Herrmann
 
J. M.
 
A disulfide relay system in the intermembrane space of mitochondria that mediates protein import
Cell
2005
, vol. 
121
 (pg. 
1059
-
1069
)
17
Naoe
 
M.
Ohwa
 
Y.
Ishikawa
 
D.
Ohshima
 
C.
Nishikawa
 
S.
Yamamoto
 
H.
Endo
 
T.
 
Identification of Tim40 that mediates protein sorting to the mitochondrial intermembrane space
J. Biol. Chem.
2004
, vol. 
279
 (pg. 
47815
-
47821
)
18
Milenkovic
 
D.
Gabriel
 
K.
Guiard
 
B.
Schulze-Specking
 
A.
Pfanner
 
N.
Chacinska
 
A.
 
Biogenesis of the essential Tim9–Tim10 chaperone complex of mitochondria: site-specific recognition of cysteine residues by the intermembrane space receptor Mia40
J. Biol. Chem.
2007
, vol. 
282
 (pg. 
22472
-
22480
)
19
Sideris
 
D. P.
Tokatlidis
 
K.
 
Oxidative folding of small Tims is mediated by site-specific docking onto Mia40 in the mitochondrial intermembrane space
Mol. Microbiol.
2007
, vol. 
65
 (pg. 
1360
-
1373
)
20
Schafer
 
F. Q.
Buettner
 
G. R.
 
Redox environment of the cell as viewed through the redox state of the glutathione disulfide/glutathione couple
Free Radical Biol. Med.
2001
, vol. 
30
 (pg. 
1191
-
1212
)
21
Hwang
 
C.
Lodish
 
H. F.
Sinskey
 
A. J.
 
Measurement of glutathione redox state in cytosol and secretory pathway of cultured cells
Methods Enzymol.
1995
, vol. 
251
 (pg. 
212
-
221
)
22
Ostergaard
 
H.
Tachibana
 
C.
Winther
 
J. R.
 
Monitoring disulfide bond formation in the eukaryotic cytosol
J. Cell Biol.
2004
, vol. 
166
 (pg. 
337
-
345
)
23
Lu
 
H.
Woodburn
 
J.
 
Zinc binding stabilizes mitochondrial Tim10 in a reduced and import-competent state kinetically
J. Mol. Biol.
2005
, vol. 
353
 (pg. 
897
-
910
)
24
Truscott
 
K. N.
Wiedemann
 
N.
Rehling
 
P.
Muller
 
H.
Meisinger
 
C.
Pfanner
 
N.
Guiard
 
B.
 
Mitochondrial import of the ADP/ATP carrier: the essential TIM complex of the intermembrane space is required for precursor release from the TOM complex
Mol. Cell. Biol.
2002
, vol. 
22
 (pg. 
7780
-
7789.
)
25
Dyall
 
S. D.
Agius
 
S. C.
de Marcos-Lousa
 
C.
Trezeguet
 
V.
Tokatlidis
 
K.
 
The dynamic dimerization of the yeast ADP/ATP carrier in the inner mitochondrial membrane is affected by conserved cysteine residues
J. Biol. Chem.
2003
, vol. 
278
 (pg. 
26757
-
26764
)
26
Wilcox
 
A. J.
Choy
 
J.
Bustamante
 
C.
Matouschek
 
A.
 
Effect of protein structure on mitochondrial import
Proc. Natl. Acad. Sci. U.S.A.
2005
, vol. 
102
 (pg. 
15435
-
15440
)
27
Ivanova
 
E.
Jowitt
 
T.
Lu
 
H.
 
Assembly of the mitochondrial Tim9–Tim10 complex: a multi-step reaction with novel intermediates
J. Mol. Biol.
2008
, vol. 
375
 (pg. 
229
-
239
)
28
Chau
 
M. H.
Nelson
 
J. W.
 
Direct measurement of the equilibrium between glutathione and dithiothreitol by high performance liquid chromatography
FEBS Lett.
1991
, vol. 
291
 (pg. 
296
-
298
)
29
Rothwarf
 
D. M.
Scheraga
 
H. A.
 
Equilibrium and kinetic constants for the thiol-disulfide interchange reaction between glutathione and dithiothreitol
Proc. Natl. Acad. Sci. U.S.A.
1992
, vol. 
89
 (pg. 
7944
-
7948
)
30
Meisinger
 
C.
Pfanner
 
N.
Truscott
 
K. N.
 
Isolation of yeast mitochondria
Methods Mol. Biol.
2006
, vol. 
313
 (pg. 
33
-
39
)
31
Lundstrom
 
J.
Holmgren
 
A.
 
Determination of the reduction–oxidation potential of the thioredoxin-like domains of protein disulfide-isomerase from the equilibrium with glutathione and thioredoxin
Biochemistry
1993
, vol. 
32
 (pg. 
6649
-
6655
)
32
Krause
 
G.
Lundstrom
 
J.
Barea
 
J. L.
Pueyo de la Cuesta
 
C.
Holmgren
 
A.
 
Mimicking the active site of protein disulfide-isomerase by substitution of proline 34 in Escherichia coli thioredoxin
J. Biol. Chem.
1991
, vol. 
266
 (pg. 
9494
-
9500
)
33
Aslund
 
F.
Berndt
 
K. D.
Holmgren
 
A.
 
Redox potentials of glutaredoxins and other thiol-disulfide oxidoreductases of the thioredoxin superfamily determined by direct protein–protein redox equilibria
J. Biol. Chem.
1997
, vol. 
272
 (pg. 
30780
-
30786
)
34
Leuenberger
 
D.
Curran
 
S. P.
Wong
 
D.
Koehler
 
C. M.
 
The role of Tim9p in the assembly of the TIM22 import complexes
Traffic
2003
, vol. 
4
 (pg. 
144
-
152
)
35
Freedman
 
R. B.
 
The formation of protein disulphide bonds
Curr. Opin. Struct. Biol.
1995
, vol. 
5
 (pg. 
85
-
91
)
36
Sevier
 
C. S.
Kaiser
 
C. A.
 
Conservation and diversity of the cellular disulfide bond formation pathways
Antioxid. Redox Signal.
2006
, vol. 
8
 (pg. 
797
-
811
)
37
Jessop
 
C. E.
Chakravarthi
 
S.
Watkins
 
R. H.
Bulleid
 
N. J.
 
Oxidative protein folding in the mammalian endoplasmic reticulum
Biochem. Soc. Trans.
2004
, vol. 
32
 (pg. 
655
-
658
)
38
Hatahet
 
F.
Ruddock
 
L. W.
 
Substrate recognition by the protein disulfide isomerases
FEBS J.
2007
, vol. 
274
 (pg. 
5223
-
5234
)