Reinvestigation of the requirement of cytosolic ATP for mitochondrial protein import.

Protein import into mitochondria requires the energy of ATP hydrolysis inside and/or outside mitochondria. Although the role of ATP in the mitochondrial matrix in mitochondrial protein import has been extensively studied, the role of ATP outside mitochondria (external ATP) remains only poorly characterized. Here we developed a protocol for depletion of external ATP without significantly reducing the import competence of precursor proteins synthesized in vitro with reticulocyte lysate. We tested the effects of external ATP on the import of various precursor proteins into isolated yeast mitochondria. We found that external ATP is required for maintenance of the import competence of mitochondrial precursor proteins but that, once they bind to mitochondria, the subsequent translocation of presequence-containing proteins, but not the ADP/ATP carrier, proceeds independently of external ATP. Because depletion of cytosolic Hsp70 led to a decrease in the import competence of mitochondrial precursor proteins, external ATP is likely utilized by cytosolic Hsp70. In contrast, the ADP/ATP carrier requires external ATP for efficient import into mitochondria even after binding to mitochondria, a situation that is only partly attributed to cytosolic Hsp70.

Many eukaryotic proteins have to move across one or more biological membrane(s) to reach their destinations after synthesis (1). Protein translocation across the organellar membranes is mediated by translocators or translocases, which are membrane-protein complexes and provide a protein-conducting channel through which organellar proteins thread, in most cases, in unfolded conformations. Several translocation systems can actively unfold precursor proteins and achieve their vectorial movement across the membranes by using various forms of energy (2,3).
Mitochondria consist of the outer and inner membranes and two aqueous compartments, the intermembrane space and matrix. The vast majority of mitochondrial proteins are synthesized in the cytosol and imported into mitochondria. Precursors of most matrix proteins and many inner membrane proteins have a positively charged presequence that contains targeting information for mitochondria and is removed by the processing peptidase in the matrix. These proteins utilize the translocase of the outer mitochondrial membrane (TOM) 1 to cross the outer membrane and the translocase of the inner mitochondrial membrane, the TIM23 complex, for translocation across the inner membrane (4,5). The targeting information in the presequence is recognized by Tom20, a general presequence receptor subunit of the TOM complex (6). Polytopic inner membrane proteins, including members of the carrier protein family such as the ADP/ATP carrier (AAC), are not synthesized with cleavable presequences but instead contain internal mitochondrial targeting signals. Members of the carrier protein family utilize the TOM complex to cross the outer membrane and another translocase of the inner membrane, the TIM22 complex, for integration into the inner membrane (7,8). These carrier proteins are recognized by Tom70 (9 -11), another receptor subunit in the TOM complex, but some presequence-containing proteins are also recognized by Tom70 (12,13).
Mitochondrial protein import generally requires two forms of energy, ATP hydrolysis and membrane potential (⌬⌿) across the inner membrane. ⌬⌿ (negative inside) traps or even exerts a physical force to positively charged presequences or internal targeting signals, thereby pulling the translocating protein. This is essential for the initial translocation of the presequence of presequence-containing precursor proteins across the inner membrane, which is followed by engagement with mitochondrial Hsp70 (mtHsp70) (2,4,14). ⌬⌿ is also essential for the late step of insertion of polytopic inner membrane proteins into the inner membrane (7,8). ATP in the mitochondrial matrix ("internal ATP") is utilized by mtHsp70. mtHsp70 binds to unfolded segments of the incoming polypeptide at the outlet of the TIM23 import channel and facilitates the vectorial movement and unfolding of the translocating protein (15,16). Therefore, internal ATP is essential for the import of presequence-containing precursor proteins mediated by the TIM23 complex, but not for the insertion of carrier proteins mediated by the TIM22 complex.
In addition to the internal ATP, some mitochondrial proteins require ATP outside mitochondria ("external ATP") for their import (17,18). Presumably, two cytosolic chaperones, Hsp70 in the cytosol (ctHsp70) and the mitochondrial import stimulation factor (MSF), are responsible for the extramitochondrial ATP requirement (19 -22). It has been shown that ctHsp70 is important for the maintenance of import competence of some mitochondrial precursor proteins (18,21,23). However other studies reported that mitochondrial precursor proteins bound to ctHsp70 can be transferred to Tom20 of the TOM complex without hydrolysis of ATP (24,25). MSF in the rat cytosol binds to mitochondrial precursor proteins, prevents their aggregate formation, and targets them to Tom70 of the TOM complex (26). In vitro import experiments with purified MSF showed that docking of the mitochondrial protein-MSF complex to Tom70 does not need ATP hydrolysis, but subsequent release of MSF from the TOM complex/precursor protein requires it (24,25). Based on the analyses of the requirement of external ATP for the import of various mitochondrial precursor proteins (18), Lithgow et al. proposed that a group of precursor proteins that require external ATP for their import depend on MSF and enter mitochondria via Tom70 (27).
Tom70 is anchored to the outer mitochondrial membrane by its N-terminal transmembrane segment and exposes the Cterminal receptor domain to the cytosol. The receptor domain contains the N-terminal dicarboxylate clamp domain as well as the central core domain responsible for binding to the internal targeting signals of presequence-less precursor proteins (28,29). The clamp domain contains three tetratricopeptide repeat motifs and binds to human Hsp90 and both yeast and human ctHsp70 (29). In yeast, the docking of ctHsp70 associated with AAC to the Tom70 clamp domain is important for productive binding of AAC to the core domain of Tom70 (29). Although external ATP allows the Tom70-bound AAC to be released from Tom70 for translocation across the outer membrane, it is not clear if it is ctHsp70 that hydrolyzes ATP at this point.
In the present study, we reinvestigated the requirement of external ATP for the import of mitochondrial precursor proteins into mitochondria in vitro. By minimizing the decrease in import competency of precursor proteins during the process of ATP depletion, we could distinguish the external ATP requirement for import upon or after binding to mitochondria from the requirement before binding. The presequence-containing precursor proteins require external ATP for maintaining their import competence but do not need external ATP upon or after binding to mitochondria. This external ATP requirement for maintenance of import competence of mitochondrial precursor proteins is most likely due to ctHsp70. In contrast, AAC requires external ATP for efficient import into mitochondria even after binding to them, a situation that is only partly ascribed to ctHsp70.

EXPERIMENTAL PROCEDURES
Precursor Proteins-The precursor proteins used here were all from Saccharomyces cerevisiae except for the fusion proteins. pSu9-DHFR and pCoxIV-DHFR are fusion proteins joining mouse dihydrofolate reductase (DHFR) with the presequence of subunit 9 of Neurospora crassa F o -ATPase precursor and with the presequence of the yeast cytochrome oxidase subunit IV precursor, respectively. The genes for the Cys 7 3 Ser (⌬cys) and the Cys 7 3 Ser/Ser 42 3 Cys/Asn 49 3 Cys (SCC) mutants of pCoxIV-DHFR were constructed as described previously for the 7/188/189 and 7/42/49 mutants of pCoxIV-DHFR (30); the ⌬cys mutant lacks Gln 188 and Cys 189 of the 7/188/189 mutant, and the SCC mutant is the same as the 7/42/49 mutant. The radiolabeled precursor proteins were synthesized in a cell-free translation system with rabbit reticulocyte lysate or PURESYSTEM (Post Genome Institute Co., Ltd.) (31) in the presence of [ 35 S]methionine.
Manipulation of External ATP Levels-We improved the methods for depletion of extramitochondrial ATP (18). After synthesis of 35 S-labeled precursor proteins with reticulocyte lysate, translation was stopped by the addition of 10 g/ml RNase A and 1 mg/ml cold Met, and the translation products were depleted of ATP by treatment with apyrase or gel filtration. For apyrase treatment, the reticulocyte lysate was incubated with 25 units/ml apyrase (grade III; Sigma), which was pretreated with 1 mM phenylmethylsulfonyl fluoride, for the indicated times at 10°C. For gel filtration, the reticulocyte lysate was passed through a NAP5 column with 50 mM Hepes-KOH, pH 7.4, 50 mM KCl, and 10 mM MgCl 2 . ATP concentrations were measured as described previously (32). In parallel, mitochondria (3 mg/ml) were treated with 100 g/ml atractyloside and 5 units/ml glycerokinase for 10 min at 10°C in SHK buffer (0.6 M sorbitol, 50 mM Hepes-KOH, pH 7.2, 50 mM KCl, 10 mM MgCl 2 , 2 mM potassium Pi, 5 mM methionine, 1 mg/ml bovine serum albumin (BSA), and 10 mM glycerol). The mitochondria were washed once with SHK buffer with 100 g/ml atractyloside, suspended in import buffer (see below) containing 100 g/ml atractyloside, 2 mM NADH, and 5 mM ␣-ketoglutarate to regenerate matrix ATP, and subjected to import assays. Import Assays-In vitro protein import into isolated yeast mitochondria with ATP was performed as described previously (33). Briefly, mitochondria were isolated from the yeast strain D273-10B (34). A radiolabeled precursor protein in 10 l of the translation product was added to mitochondria (0.25 mg/ml) in 190 l of import buffer (250 mM sucrose, 80 mM KCl, 5 mM MgCl 2 , 10 mM Mops-KOH, pH 7.2, 5 mM methionine, and 1 mg/ml BSA) containing 2 mM NADH and 2 mM ATP and incubated for 8 min at 30°C. Import reactions were stopped by the addition of valinomycin to 10 g/ml, and the mitochondria were treated with 100 g/ml proteinase K for 30 min on ice. The mitochondria were re-isolated by centrifugation and washed once with 300 l of SEM buffer (250 mM sucrose, 5 mM EDTA, and 10 mM Mops-KOH, pH 7.2). The proteins were analyzed by SDS-PAGE and radioimaging with a Storm 860 image analyzer (Molecular Dynamics). The "imported proteins" represent protease-protected forms.
Import of Precursor Proteins in the Absence of Hsc70 -The monoclonal antibody 1B5 was raised against Hsc70 (21). Depletion of Hsc70 (ctHsp70) from reticulocyte lysate was carried out by treatment with 1B5 antibody-conjugated Sepharose resin as described previously (21). Briefly, reticulocyte lysate without translation or a translation product was mixed with suspension of the 2-fold volume (bed volume) of 1B5 antibody-conjugated Sepharose resin in 20 mM Hepes-KOH, pH 7.4, and 80 mM potassium acetate, incubated for 4 h or 30 min, respectively, at 4°C with gentle shaking, and then centrifuged. The supernatant was found to contain Ͻ2% (reticulocyte lysate without translation) or Ͻ10% (translation product) of Hsc70 by immunoblotting. For ϩHsc70 control, reticulocyte lysate was treated with unconjugated Sepharose resin instead of 1B5 Sepharose resin. For another control, 10 mg/ml BSA in 20 mM Hepes-KOH, pH 7.4, and 80 mM potassium acetate was added to the reticulocyte lysate that had been depleted of Hsc70. 1 l of the translation product (depleted of Hsc70) was mixed with 10 l of reticulocyte lysate, reticulocyte lysate without Hsc70, or 10 mg/ml BSA in 20 mM Hepes-KOH, pH 7.4, and 80 mM potassium acetate. Then, 10 l of 5 mM ATP, 30 mM creatine phosphate, 0.4 mg/ml creatine phosphokinase, and 10 mM MgCl 2 were added, incubated for 0 -30 min at 30°C, and subjected to import assays (for 8 min at 30°C).

RESULTS
Depletion of External ATP-According to the previously adopted protocol for analyzing the role of external ATP in mitochondrial protein import (18), precursor proteins synthesized in reticulocyte lysate were first depleted of ATP with an ATPase such as apyrase, hexokinase with 2-deoxyglucose, and glycerokinase with glycerol, and then the effects of the addition of an excess amount of external ATP on protein import were analyzed. However, during the incubation of translation products with ATPase, precursor proteins may lose their import competence, probably because the functions of ATP-dependent cytosolic chaperones in reticulocyte lysate can also be inhibited by ATP depletion. Therefore, this ATP depletion protocol does not allow us to discriminate the ATP requirements before precursor binding to the mitochondrial surface from the ATP requirements upon or after binding.
We thus tried to develop an improved protocol that would allow us to deplete ATP from in vitro translation products of mitochondrial precursor proteins with minimized decrease in their import competency. To maintain import competence of mitochondrial precursor proteins, prolonged incubation of the translation products at high temperature (e.g. Ն30°C) should be avoided. We found that Ϸ1 mM ATP in the reticulocyte lysate can be decreased to 1.6 and 0.6 M ATP by incubation with apyrase for 20 min at 10°C and by gel filtration, respectively, without a significant decrease in the import competence of precursor proteins (not shown). Therefore, we adopted these procedures of ATP depletion of the translation products in the following experiments.
Incubation at a High Temperature without ATP Reduces the Import Competence of Precursor Proteins-We first asked if incubation of translation products in the absence of ATP at high temperature reduces import competence of precursor proteins. The precursor proteins we tested are precursors of alcohol dehydrogenase III (pADH), the ␤-subunit of F 1 -ATPase (pF 1 ␤), cytochrome oxidase subunit IV (pCoxIV), mitochondrial Hsp60 (pHsp60), and a fusion protein joining cytochrome oxidase subunit IV with mouse dihydrofolate reductase (pCoxIV-DHFR). Previously, pADH, pF 1 ␤, and AAC were reported to require external ATP for their import into isolated mitochondria, whereas pHsp60, pCoxIV, and pCoxIV-DHFR were imported in the absence of external ATP (18,(35)(36)(37). We synthesized the radiolabeled precursor proteins in reticulocyte lysate and divided them into two. One aliquot was first depleted of ATP by incubation with apyrase at 10°C and subsequently incubated for various times at 30°C (ϪATP). The other aliquot was first incubated for various times at 30°C and subsequently incubated with apyrase at 10°C (ϩATP). Then, both aliquots were incubated with mitochondria under the condition where ATP was present on both sides of the inner membrane to test their import abilities (Fig. 1). Prolonged incubation of the translation products at 30°C without ATP decreased the import competence of the precursor proteins, although the decrease in the import competence varied with different precursor proteins. The import abilities of pADH, pF 1 ␤, and pCoxIV-DHFR rapidly decreased to Ͻ30% after a 30-min incubation without ATP, whereas Ͼ50% of the import abilities of pCoxIV and pHsp60 still remain after a 30-min incubation without ATP.
When analyzed by glycerol density gradient centrifugation, pF 1 ␤ that was synthesized in reticulocyte lysate and incubated without ATP for 20 min at 30°C formed aggregates, whereas pF 1 ␤ that was incubated with ATP for 20 min at 30°C remained soluble as a monomer (Fig. 2). The soluble form, but not the aggregated form, of pF 1 ␤ was import-competent (Fig. 2). Therefore, ATP-dependent chaperones in reticulocyte lysate are important for preventing the aggregation of precursor proteins, thereby maintaining them in import-competent states.
Presequence-containing Precursor Proteins Can Be Taken Up by Mitochondria in the Absence of External ATP-It was previously shown that a carrier protein, AAC, requires external FIG. 1. External ATP is required for maintaining import competence of precursor proteins. pADH, pF 1 ␤, pCoxIV-DHFR, pCoxIV, and pHsp60 were synthesized in a cell-free translation system with rabbit reticulocyte lysate. The translation reaction was stopped by the addition of 10 g/ml RNase A and 1 mg/ml methionine, and the translation product was divided into two. One aliquot was incubated with 25 units/ml apyrase at 10°C for 20 min and subsequently subjected to incubation at 30°C for the indicated times (ϪATP). The other aliquot was incubated first at 30°C for indicated times and subsequently incubated with 25 units/ml apyrase at 10°C for 20 min (ϩATP). Both ϪATP and ϩATP translation products were mixed with mitochondria in import buffer and incubated at 30°C for 8 min. After proteinase K treatment, the mitochondria were re-isolated by centrifugation, and the proteins were analyzed by SDS-PAGE and radioimaging with a Storm 860 image analyzer. The input in the import assay was set to 100%. The plots represent mean values of duplicate experiments with errors of Ͻ15%. ATP for its import both before and after binding to the TOM complex, but before insertion into the inner membrane (38 -40). On the other hand, the timing of the external ATP requirement for presequence-containing precursor proteins is not clear. We thus performed a test to determine whether external ATP is essential for the import of presequence-containing precursor proteins into mitochondria before or after binding to mitochondria. pF 1 ␤ and pCoxIV, as well as AAC, were depleted of ATP by gel filtration of the translation products. Mitochondria were depleted of external ATP by preincubation with glycerokinase and atractyloside (to block transport of ADP/ATP across the inner membrane). Then, pF 1 ␤, pCoxIV, and AAC were incubated with mitochondria under the condition where ATP is present only in the matrix. Both pF 1 ␤ and pCoxIV were efficiently imported into mitochondria in the absence of external ATP, and the addition of ATP did not stimulate their import (Fig. 3, graphs for pF 1 ␤ and pCoxIV). As a control, the presence of glycerokinase, which would hydrolyze possible residual ATP, if any, outside the mitochondria, did not affect the import kinetics of pF 1 ␤ and pCoxIV either (Fig. 3, lines marked  GK). Therefore, external ATP is not required for either release from the ATP-dependent chaperones or for the subsequent process of translocation across mitochondrial membranes. This is in contrast to AAC, which required external ATP for its import into mitochondria (Fig. 3, graph for AAC).
The above results, which show that pF 1 ␤ and pCoxIV can be imported into mitochondria in the absence of external ATP, may be due perhaps to the slow leakage of ATP from the matrix by an uncharacterized mechanism. Indeed, slow and inefficient import of AAC into mitochondria even after glycerokinase treatment of the mitochondria (Fig. 3, lines marked GK) could reflect such slow leakage of ATP from the matrix. To test this possibility, we performed two-step import experiments for presequence-containing precursor proteins as well as for AAC (Fig.  4). Mitochondria were first incubated with oligomycin (to inhibit F 1 F o -ATPase), atractyloside (to inhibit ADP/ATP transport across the inner membrane), and NADH (to maintain ⌬⌿), and pSu9-DHFR, pF 1 ␤, pADH, pHsp60, and AAC were depleted of ATP by gel-filtration of the translation products. Then the precursor proteins were incubated with the mitochondria with (Fig. 4, bars 1-4) or without (bars 5-8) external ATP (first incubation) for 5 min at 25°C. The mitochondria were isolated by centrifugation and subjected to incubation with (Fig. 4, bars 2-4 and 6 -8) or without ATP (bars 1 and 5) in the matrix, but they were incubated in the absence of external ATP (bars 1-3 and 5-7), for 10 min at 25°C (second incubation). Because there is no ATP on both sides of the mitochondria during the first incubation (Fig. 4, bars 5-8), the precursor proteins recovered with mitochondria should not require external ATP for their binding to mitochondria. Therefore, the subsequent chase of the bound precursor proteins into the matrix in the presence of ATP only in the matrix will allow us to estimate the amount of productive binding during the first incubation in FIG. 3. Import of pF 1 ␤ and pCoxIV, but not of AAC, does not depend on external ATP. The translation products of pF 1 ␤, pCoxIV, and AAC were depleted of ATP by passing through a NAP5 gel-filtration column that was equilibrated with 50 mM Hepes-KOH, pH 7.4, 50 mM KCl, and 10 mM MgCl 2 . The residual ATP in the translation products was reduced from 1 mM to 0.6 M. Mitochondria (3 mg/ml) were incubated with 100 g/ml atractyloside and 5 units/ml glycerokinase in an import buffer containing 10 mM glycerol at 10°C for 10 min. The mitochondria were re-isolated by centrifugation, washed with import buffer with glycerol, and suspended in import buffer with 100 g/ml atractyloside, 2 mM NADH, 10 mM glycerol, and 5 mM ␣-ketoglutarate. The ATP-depleted mitochondrial suspension (200 l) was pre-incubated in the absence of ATP (ϪATP) or presence of 1 mM ATP (ϩATP) or 5 units/ml glycerokinase (GK) at 20°C for 2 min. Then, the mitochondrial suspension received the ATP-depleted translation product (10 l) and was incubated at 20°C for indicated times. After proteinase K treatment, the mitochondria were recovered by centrifugation, and the proteins were analyzed by SDS-PAGE and radioimaging. The input in the import assay was set to 100%.
FIG. 4. Two-step import assay reveals that pSu9-DHFR, pHsp60, and pADH, but not AAC, bound to mitochondria can be chased into mitochondria independently of external ATP. Prior to the first incubation, mitochondria (200 g/ml) were incubated with 10 g/ml oligomycin, 100 g/ml atractyloside, and 2 mM NADH in an import buffer containing 10 instead of 5 mM MgCl 2 and 2 mM potassium Pi for 10 min at 25°C and subsequently received 2 mM ATP after incubation for 5 min at 25°C (ATP out ϩ/ATP in Ϫ/⌬⌿ ϩ) or incubated with 5 units/ml apyrase for 5 min at 25°C (ATP out Ϫ/ATP in Ϫ/⌬⌿ ϩ). For the first incubation in two-step import, the translation product depleted of ATP by gel filtration (5% v/v) was incubated with mitochondria in the presence of 2 mM ATP (ATP out ϩ/ATP in Ϫ/⌬⌿ ϩ) or with 25 units/ml apyrase (ATP out Ϫ/ATP in Ϫ/⌬⌿ ϩ) for 10 min at 25°C. Then, the mitochondria were re-isolated by centrifugation, washed with import buffer containing 10 g/ml oligomycin and 100 g/ml atractyloside, and re-suspended in the same buffer. The mitochondria were subjected to the second incubation with 10 g/ml valinomycin (ATP out Ϫ/ATP in Ϫ/⌬⌿ Ϫ), 10 g/ml valinomycin, 5 mM ␣-ketoglutarate, 2 mM NADH (ATP out Ϫ/ATP in ϩ/⌬⌿ Ϫ), 5 mM ␣-ketoglutarate, 2 mM NADH (ATP out Ϫ/ATP in ϩ/⌬⌿ ϩ), or 2 mM ATP, 5 mM ␣-ketoglutarate, 2 mM NADH ATP out ϩ/ATP in ϩ/⌬⌿ ϩ) for 10 min at 25°C. The imported proteins were analyzed by SDS-PAGE and radioimaging after treatment of the mitochondria with 100 g/ml proteinase K on ice for 20 min. the absence of external ATP. Fig. 4 shows that pSu9-DHFR, pADH, and pHsp60 bound to the mitochondria in the absence of ATP on both sides of the mitochondria, and the bound forms were chased into the matrix by the replenishment of internal ATP. The presence of ⌬⌿ during the second incubation did not affect the chase of pSu9-DHFR, or it affected only moderately the chase of pHsp60 and pADH (Fig. 4), because ⌬⌿ is mainly required for the initial binding to mitochondria. The different effects of ⌬⌿ during the second incubation on different precursor proteins may reflect the difference in stabilities of the productive translocation intermediates generated in the absence of matrix ATP and ⌬⌿ during the first incubation. On the other hand, AAC that bound the mitochondria in the absence of external ATP during the first incubation was not efficiently chased into mitochondria by the second incubation, irrespective of the presence of ⌬⌿ or ATP in the matrix, unless external ATP was supplied (Fig. 4).
Taken together, the present results indicate that external ATP is required for presequence-containing proteins only prior to the binding to mitochondria, whereas external ATP is essential for AAC after binding to the mitochondria. Therefore, the previously observed requirement of external ATP for the import of presequence-containing mitochondrial precursor proteins (18) reflects solely the decrease in import competency of precursor proteins before engaging the TOM complex.
Hsc70 in the Cytosol Is Responsible for the Maintenance of Import Competency-Which cytosolic chaperones are responsible for the maintenance of import competency of mitochondrial precursor proteins? Because ctHsp70 has been suggested to play such a role in mitochondrial protein import, we tested the effects of depletion of ctHsp70 (Hsc70) from the translation products on import of pCoxIV-DHFR into mitochondria. After synthesis of pCoxIV-DHFR in reticulocyte lysate, Ͼ90% of Hsc70 was immuno-depleted from the translation products by the immobilized anti-Hsc70 monoclonal antibody (1B5). Most of pCoxIV-DHFR was not co-immunoprecipitated with Hsc70, suggesting that the binding of pCoxIV-DHFR to Hsc70, if it occurs at al, is dynamic. Then, the translation products of pCoxIV-DHFR depleted of Hsc70 was diluted 10-fold with untreated reticulocyte lysate, Hsc70-depleted lysate, or buffer containing BSA and incubated for various times at 30°C. The remaining import competence of pCoxIV-DHFR was assessed by subsequent incubation with mitochondria in the presence of ATP on both sides of the mitochondria. Fig. 5A (graph labeled WT) shows that incubation of pCoxIV-DHFR with reticulocyte lysate depleted of Hsc70 resulted in a moderate decrease in its FIG. 6. The effects of Hsc70 depletion and external ATP on the import of AAC and pCoxIV-DHFR are well correlated with each other. pCoxIV-DHFR and AAC were translated in reticulocyte lysate, treated with 1B5-Sepharose beads (ϪHsc70) or Sepharose beads (ϩHsc70) and subjected to gel filtration to deplete ATP. The translation products were then subjected to import assays in the presence of 2 mM ATP (ϩATP) or with 25 units/ml apyrase (ϪATP) at 25°C. The imported proteins were analyzed by SDS-PAGE and radioimaging after the treatment of the mitochondria with 100 g/ml proteinase K on ice for 20 min. The plots represent mean values of duplicate experiments with errors of Ͻ5%.

FIG. 5. Destabilized mutants of pCoxIV-DHFR depend more on Hsc70
in their import into mitochondria. A, wild-type (WT), ⌬cys mutant, and SCC mutant DHFR were translated in reticulocyte lysate and depleted of Hsc70. 1 l of the translation product was incubated with 10 l of reticulocyte lysate with Hsc70 (ϩHsc70; treated with control Sepharose beads) and 10 l of reticulocyte lysate without Hsc70 (ϪHsc70; treated with 1B5-Sepharose beads) or 10 mg/ml BSA and 10 l of 5 mM ATP, 30 mM creatine phosphate, 0.4 mg/ml creatine phosphokinase, and 10 mM MgCl 2 . The reaction mixtures were pre-incubated for the indicated times at 30°C, and subjected to import assays for 8 min at 30°C. B, the translation products were diluted 10-fold with 20 mM Hepes-KOH, pH 7.4, and 50 mM KCl and digested with the indicated concentrations of trypsin for 10 min on ice. The reaction was stopped by the addition of 4 mg/ml trypsin inhibitor and analyzed by SDS-PAGE followed by quantification by radioimaging. Gray bars, full-length forms (p); black bars, proteaseresistant fragments likely corresponding to the DHFR domain (f). import ability, which is comparable to the incubation with buffer containing BSA.
When pCoxIV-DHFR was digested with trypsin, the presequence was cleaved off easily whereas the DHFR domain resisted degradation, suggesting that the DHFR domain is tightly folded (Fig. 5B, graph labeled WT). We now asked if mutations that destabilize the structure of DHFR affect the requirement of Hsc70. The mutations in the DHFR domain, namely Cys 7 3 Ser, which we have designated ⌬cys, and Cys 7 3 Ser/Ser 42 3 Cys/Asn 49 3 Cys, which we call SCC, made the DHFR domain more sensitive to trypsin digestion, suggesting that these mutations indeed destabilize the structures of the DHFR domain ( Fig. 5B and Ref. 28). These destabilized mutants of pCoxIV-DHFR showed more significant decrease in import competence after incubation with reticulocyte lysate depleted of Hsc70 or buffer containing BSA (Fig. 5A). These results suggest that Hsc70 is the major chaperone responsible for the maintenance of import competence of pCoxIV-DHFR and that a less stable precursor protein depends more on Hsc70 to remain import competent.
Hsc70 Is Not Essential for the Requirement of External ATP for the Import of AAC-Is it only Hsc70 that is responsible for the external ATP requirement of the mitochondrial protein import? We thus compared the external ATP dependence of mitochondrial protein import between conditions with and without Hsc70 in the absence of preincubation at 30°C (Fig. 6). When we compare the import with and without Hsc70, pCoxIV-DHFR was imported into mitochondria independently of the presence of Hsc70. When we compare the import with and without ATP, import of pCoxIV-DHFR into mitochondria was slightly suppressed in the absence of external ATP, irrespective of the presence of Hsc70. However, this small effect of ATP depletion could be due merely to the slight loss of the import competence of pCoxIV-DHFR during the process of ATP depletion from the translation product.
Depletion of Hsc70 moderately affected the import rate of AAC into mitochondria (Fig. 6), which is consistent with the suggested important role of Hsc70 in the import of AAC (29). However, even after the depletion of Hsc70, AAC was still imported into mitochondria to some extent, and this residual import was suppressed by the depletion of external ATP. Of course, we cannot eliminate the possibility that this inefficient import reflects the role of a residual amount of Hsc70 that escaped the Hsc70 depletion treatment with anti-Hsc70 antibody. However, an alternative explanation could be that not only Hsc70 but also other ATP-dependent factor(s) may be involved in the import of AAC into mitochondria.
To further examine the role of ctHsp70 in the AAC import, we performed two-step import of AAC into mitochondria in the presence or absence of Hsc70 (Fig. 7). For depletion of Hsc70, AAC synthesized in reticulocyte lysate was treated with the immobilized anti-Hsc70 monoclonal antibody. Then AAC with and without Hsc70 was incubated with isolated mitochondria, which had been treated with oligomycin (to inhibit F 1 F o -ATPase) and NADH (to maintain ⌬⌿), in the presence or absence of external ATP at 25°C for 10 min (first incubation). The mitochondria were isolated by centrifugation and subjected to incubation with or without external ATP but in the absence of ATP in the matrix for 10 min at 25°C (second incubation). When external ATP was present during the first incubation, AAC was imported into mitochondria without accumulation of the surface-bound intermediate so that the imported AAC did not increase (or chase was not observed) during the second incubation. When external ATP was absent during the first incubation, imported AAC increased (or bound AAC was chased FIG. 7. Two-step import of AAC suggests that Hsc70 is not the only factor responsible for the external ATP requirement. AAC was translated in reticulocyte lysate, treated with 1B5-Sepharose beads (ϪHsc70) or Sepharose beads (ϩHsc70), and subjected to gel filtration to deplete ATP. The translation products were incubated in import buffer containing 10 instead of 5 mM MgCl 2 , 2 mM potassium Pi, 2 mM NADH, and 10 g/ml oligomycin with mitochondria in the presence of 2 mM ATP (ϩATP) or with 5 units/ml apyrase (ϪATP) for 10 min at 25°C (1st incubation). The mitochondria were recovered by centrifugation, washed once with import buffer, and subjected to a second incubation (2nd incubation) with 2 mM ATP (ϩATP) or 5 units/ml apyrase (ϪATP) for 3-10 min at 25°C. The import buffer for the wash and second incubation contained 10 g/ml oligomycin. The imported proteins were analyzed by SDS-PAGE and radioimaging after treatment of the mitochondria with 100 g/ml proteinase K on ice for 20 min.
FIG. 8. Cytosolic factors for AAC import was required upon translation of AAC with eukaryotic translation systems but not with prokaryotic translation systems. AAC and pSu9-DHFR were synthesized with PURESYSTEM or reticulocyte lysate in the presence of [ 35 S]methione at 30°C. Experiments for bars 5-8 contained 250 g/ml cycloheximide and 10% (v/v) reticulocyte lysate. The translation products were incubated with mitochondria at 25°C for 3 or 10 min in the import buffer containing 10 instead of 5 mM MgCl 2 , 2 mM potassium Pi, 2 mM NADH, 1% instead of 1 mg/ml BSA, and 2 mM ATP. Experiments for bars 3, 4, 7, and 8 contain 2% (v/v) of reticulocyte lysate, and those for bars 11 and 12 contains 2% (v/v) of the solution for PURESYSTEM. The imported proteins were analyzed as in Fig. 6. P, PURESYSTEM; R, reticulocyte lysate. into the mitochondria) during the second incubation only when external ATP was present. This ATP-dependent chase of the AAC intermediate during the second incubation was observed even when Hsc70 was depleted. This is consistent with our interpretation that an ATP-dependent factor(s) other than Hsc70 is involved in the import of AAC into mitochondria.
To characterize the ATP-dependent factor required for the AAC import, we took advantage of the use of PURESYSTEM, a cell-free translation system reconstituted from purified components including ribosomes, translation factors, aminoacyl-tRNA synthetases, T7 RNA polymerase, and tRNAs, but free of any chaperones (31). pSu9-DHFR and AAC were synthesized with PURESYSTEM (Fig. 8, bars 1-4) or reticulocyte lysate (bars 9 -12) in the presence of [ 35 S]methionine and subjected to import into isolated yeast mitochondria. Although pSu9-DHFR synthesized with PURESYSTEM and reticulocyte lysate was imported into mitochondria at similar efficiencies, the import of AAC synthesized with PURESYSTEM into mitochondria was much less efficient than that of AAC synthesized with reticulocyte lysate. Interestingly, import of PURESYSTEM-made AAC was not stimulated by the presence of reticulocyte lysate (Fig. 8, bars 3, 4), or, conversely, that of reticulocyte lysatemade AAC was not inhibited by the presence of PURESYSTEM during the import (Fig. 8, bars 11 and 12). The import efficiency of AAC was not improved even if it was synthesized with PURESYSTEM in the presence of reticulocyte lysate from the beginning (Fig. 8, bars 5-8). These results suggest that the ATP-dependent factors, including Hsc70 in the reticulocyte lysate, are required during the synthesis of AAC as well as during the import and are incompatible with the prokaryotic protein translation system. Hsc70 and other cytosolic factor(s) in reticulocyte lysate may interact with eukaryotic ribosomes, but not with prokaryotic ribosomes, to maintain the import competence of newly synthesized AAC. DISCUSSION In the present study, we re-examined the requirement of external ATP for the import of various mitochondrial precursor proteins into isolated yeast mitochondria. By minimizing the decrease in import competency of precursor proteins during the process of ATP depletion, we could assess the external ATP requirement of precursor proteins for their import into mitochondria upon or after, but not prior to, binding to the mitochondria. Although pADH and F 1 ␤ were reported to require external ATP (18), we found here that they do not require external ATP for their import into mitochondria and that the previously observed requirement of external ATP arose from the decrease in their import competency before binding to mitochondria. This means that, because MSF requires ATP to release the substrate precursor protein at the docking site of the TOM complex, Tom70 (24,25), the presequence-containing precursor proteins studied here do not require MSF in the reticulocyte lysate for their import into mitochondria either. Previous in vitro studies showed that MSF was required for mitochondrial protein import if artificially denatured precursor proteins were pre-loaded onto MSF before incubation with isolated mitochondria (24,25). However, the role of MSF in vivo still remains unclear because the yeast MSFs, Bmh1p and Bmh2p, are not essential for yeast cell growth (6). Because depletion of Hsc70 in the translation products led to a decrease in the import competence of pCoxIV-DHFR fusion proteins (Fig. 5), the external ATP is most likely utilized by Hsc70, but not by MSF, to maintain precursor proteins in import-competent states.
In contrast to the presequence-containing precursor proteins, AAC requires external ATP for translocation across the outer membrane after binding to mitochondria. This requirement of external ATP may well be connected to the requirement of Hsc70 for productive binding of AAC to Tom70 (29). However, even after depletion of Ͼ90% of Hsc70 from the AAC translation products, AAC could be still imported into mitochondria to some extent. On the other hand, Young et al. showed that a mutation in the clamp domain of Tom70 that abrogates binding to Hsc70 inhibited the productive binding of AAC to Tom70 in its import into mitochondria completely (29). Although we cannot rule out the possibility that a small residual amount of Hsc70 in reticulocyte lysate could still manage to mediate the AAC import, there may be a factor other than Hsc70 in reticulocyte lysate that interacts with AAC upon its synthesis on the ribosome and targets it to the clamp domain of Tom70, thereby substituting for Hsc70 when Hsc70 is depleted. Clearly, further analysis is necessary to address this question, and such a study is underway in our laboratory.