Functional Analysis of Human Mitochondrial Receptor Tom20 for Protein Import into Mitochondria*

The mitochondrial import receptor translocase of the outer membrane of mitochondria (Tom20) consists of five segments, an N-terminal membrane-anchor segment, a linker segment rich in charged amino acids, a tetratricopeptide repeat motif, a glutamine-rich segment, and a C-terminal segment. To assess the role of each segment, four C-terminally truncated mutants of the human receptor (hTom20) were constructed, and the effect of their overexpression in COS-7 cells was analyzed. Expression of a mutant lacking the tetratricopeptide repeat motif inhibited preornithine transcarbamylase (pOTC) import to the same extent as the wild-type receptor. Thus, overexpression of the membrane-anchor and the linker segments is sufficient for the inhibition of import. Expression of either the wild-type receptor or a mutant lacking the C-terminal end of 20 amino acid residues stimulated import of pOTC-green fluorescent protein (GFP), a fusion protein in which the presequene of pOTC was fused to green fluorescent protein. On the other hand, expression of mutants lacking either the glutamine-rich segment or larger deletions inhibited pOTC-GFP import. In vitro import of pOTC was inhibited by the wild-type hTom20 and the mutant lacking the C-terminal end, but much less strongly by the mutant lacking the glutamine-rich segment. On the other hand, import of pOTC-GFP was little affected by any of the forms of hTom20. In binding assays, pOTC binding to hTom20 was only moderately decreased by the deletion of the glutamine-rich segment, whereas pOTC-GFP binding was completely lost by this deletion. Binding of pOTCN-GFP a construct that contains an additional 58 N-terminal residues of mature OTC, resembled that of pOTC. All of these results indicate that the region 106–125 containing the glutamine-rich segment of hTom20 is essential for binding and import stimulation in vivo of pOTC-GFP and for inhibition of in vitroimport of pOTC. The results also indicate that this region is important for mitochondrial aggregation. The different behaviors of pOTC and the pOTC-GFP chimera toward hTom20 mutants is explicable on the basis of the conformation of the precursor proteins.

Most mitochondrial proteins are encoded by nuclear genes, synthesized as preproteins in the cytosol, targeted to the mitochondria, and imported into the organelle. An important step in this process is the interaction of the preproteins with the outer surface of the mitochondria. Genetic and biochemical studies in yeast and Neurospora have identified a number of proteins in the mitochondrial outer membrane that are responsible for recognizing and translocating preproteins into the organelle (reviewed in Refs. [1][2][3][4]. They form a dynamic protein complex, termed the translocase of the outer membrane of mitochondria. Subunits of the complex that have been identified include the receptor components Tom20 1 (5,6), Tom22 (7,8), Tom37 (9), and Tom70 (10,11). Among these subunits, Tom20 was shown to bind to the basic amphiphilic targeting sequences of preproteins through electrostatic interactions with the acidic receptor domain (12). Together with Tom22, Tom20 of yeast mediates the import of all preproteins known to use the general import machinery of the mitochondria (13). The Tom20 and Tom70 subunits of yeast mitochondria were shown to interact via the tetratricopeptide repeat (TPR) motif in Tom20 (14).
On the other hand, little is known about the import receptors of animals. Recently, cDNA for a human homolog (hTom20) of yeast and Neurospora Tom20 was isolated (15)(16)(17). In vitro import of preproteins into isolated mitochondria was inhibited by the soluble domain of hTom20 (⌬hTom20) (17) and by anti-hTom20 (16,17). We also showed that the import of several preproteins was inhibited by ⌬hTom20 and by anti-hTom20, and found that the inhibition varied among preproteins (18). In addition to the in vitro assay of hTom20, we have assessed its role in cultured animal cells, which more closely resembles the in vivo situation. We developed an in vivo assay method in which cultured cells were cotransfected with plasmids for a preprotein and for hTom20 and showed that in pulse-chase experiments coexpression of exogenous hTom20 retarded mitochondrial import and processing of preornithine transcarbamylase (pOTC) (18). On the other hand, overexpression of hTom20 resulted in stimulated mitochondrial import of a fusion protein pOTC-GFP that consists of the presequence of human pOTC fused to green fluorescent protein (GFP) (19). Surprisingly, overexpression of hTom20 resulted in the perinuclear aggregation of mitochondria (19). In each of these assay, ⌬hTom20 had no effect.
Here, we report that the overexpression in COS-7 cells of the 1 The abbreviations used are: Tom20, translocase of the outer membrane of mitochondria; hTom20, human Tom20; GFP, green fluorescent protein; GST, glutathione S-transferase; pOTC and OTC, precursor and mature form of ornithine transcarbamylase; TPR, tetratricopeptide repeat; PAGE, polyacrylamide gel electrophoresis; Tricine, N-[2-hydroxy-1,1-bis(hydroxymethyl)ethyl]glycine. combined membrane-anchor and the linker segments of hTom20 is sufficient to inhibit pOTC import, whereas the region 106 -125 containing glutamine-rich segment is essential to stimulate of pOTC-GFP import and to aggregate mitochondria. Based on in vitro import and binding studies, this region is apparently involved in the binding to the presequence. A molecular model to explain difference in the interaction of substantially unfolded proteins such as pOTC and folded proteins such as pOTC-GFP to hTom20 is presented.

Antibodies-
The NcoI/blunt-ended XbaI fragment encoding Aequoria victoria GFP was excised from phGFP-S65T (CLONTECH, Palo Alto, CA), and cloned into the NcoI/blunt-ended HindIII site of the pET30a (Novagen, Madison, WI). Histidine-and S-peptide-tagged GFP was expressed in Escherichia coli from the resulting plasmid, pET30a-GFP, purified by metal chelation chromatography and used for making anti-GFP antibody. Anti-human OTC and anti-human Tom20 antibodies were prepared as described previously (19).
Cell Culture and Transfection-COS-7 cells were cultured in growth medium (Dulbecco's modified Eagle's medium plus 10% fetal calf serum) at 37°C under an atmosphere of 5% CO 2 and 95% air. The cells were transfected with 10 g of plasmids at 37°C for 4 h by the use of TransIT LT1 polyamine (Pan Vera, Madison, WI), and the cells were cultured at 37°C to allow expression (19). The transfection efficiency was about 10%.
Cell Fractionation-After 24-h culture, the cells were harvested with trypsinization, washed twice with phosphate-buffered saline, and then suspended in ice-cold hypotonic buffer (10 mM Tris-HCl (pH 7.4) containing 5 mM magnesium chloride, 1 mM dithiothreitol and 1 mM phenylmethylsulfonyl fluoride). After sonication, the suspension was centrifuged at 500 ϫ g for 10 min at 4°C, and the supernatant was used as whole cell extract. The cell extract was further centrifuged at 100,000 ϫ g for 10 min at 4°C to give the soluble fraction and the membrane fraction.
Pulse-Chase Experiments-COS-7 cells were transfected as described above. After 16-h culture, the cells were harvested with trypsinization, washed twice with methionine-free Dulbecco's modified Eagle's medium, and suspended in 1 ml of the same medium. After preincubation at 37°C for 1 h to deplete methionine, the cells were radiolabeled with 8 MBq of Pro-Mix TM containing L-[ 35 S]methionine and L-[ 35 S]cysteine (Amersham Pharmacia Biotech) for 5 min and then chased with 20 mM L-methionine in 2 ml of Dulbecco's modified Eagle's medium. At the indicated times, 0.5 ml of aliquots were withdrawn and mixed with 0.5 ml of ice-cold lysis buffer (20 mM Tris-HCl (pH 7.4) containing 4 mM EDTA, 0.2% SDS, 0.2% Triton X-100, 100 M chymostatin, 100 M pepstatin, 100 M leupeptin, and 100 M antipain). Radiolabeled proteins were immunoprecipitated with 20 l of antiserum and 200 l of a 10% suspension of protein A-Sepharose (Amersham Pharmacia Biotech), and subjected to 10% SDS-PAGE. The radioactivity in the gels was visualized and quantified using a FUJIX BAS2000 analyzer (Fuji Film Co., Tokyo, Japan).
Electron Microscopy-COS-7 cells were cultured in 35-mm dishes and transfected with plasmids as described above. After culture for 24 h, the cells were fixed with 2.5% glutaraldehyde in 0.1 M cacodylate buffer (pH 7.3) for 1 h and postfixed with 1% osmium tetroxide in the same buffer for 1 h. The cells were then dehydrated and embedded in epoxy resin as described previously (22). Ultrathin sections stained with uranyl acetate and lead nitrate were inspected with a Hitachi H-7500 electron microscope (Hitachi, Tokyo, Japan).
Expression and Purification of GST-fused hTom20s-The plasmids, pGEX-2T and three types of its derivatives were transformed into TOPP 2 cells (Stratagene, La Jolla, CA). Expression of GST-fused hTom20s and absorption onto glutathione-agarose (Amersham Pharmacia Biotech) were performed as described previously (17). The absorbed proteins were eluted with 50 mM Tris-HCl (pH 8.0) containing 10 mM reduced glutathione, dialyzed in 20 mM Hepes-KOH (pH 7.4), concentrated by centricon-10 (Amicon, Beverly, MA), and kept at Ϫ80°C in the presence of 10% glycerol until use.
In Vitro Import into Isolated Mitochondria-The import mixture (total 50 l) containing rat liver mitochondria (100 g of protein) and 10 l of reticulocyte lysate containing 35 S-labeled translation products was incubated in the absence or presence of 1 nmol of GST or GST-fused hTom20s at 25°C. The reaction was stopped by diluting the import mixture into the mitochondrial isolation buffer containing 0.1 mM dinitrophenol (18). The mitochondria were reisolated by centrifugation and subjected to 10% SDS-PAGE. The radioactivity in the gels was visualized and quantified using a FUJIX BAS2000 analyzer.
Binding Assay with GST-fused hTom20s-Purified GST or GST-hTom20 delivatives (7 nmol) were absorbed onto glutathione-agarose in 1.25 ml of binding buffer (20 mM Hepes-KOH (pH 7.4), 50 mM potassium chloride, 1 mM magnesium chloride, 0.1 mg/ml bovine serum albumin) containing 62.5 l of a 50% slurry of glutathione-agarose. The agarose beads were washed three times and resuspended in 250 l of binding buffer. For assay, 20 l of GST derivative-bound agarose was diluted in 270 l of binding buffer, and then mixed with 10 l of reticulocyte lysate containing 35 S-labeled translation products for 30 min at 25°C by gentle shaking. After centrifugation, the supernatant was removed and the beads were washed once with binding buffer. Fifty l of 50 mM Tris-HCl (pH 8.0) containing 15 mM reduced glutathione were added to the wet agarose beads, the mixture was shaken gently for 1 h at 25°C, and 40 l of the supernatant was recovered after centrifugation. Twenty l of the supernatant was subjected to 10% SDS-PAGE, and the radioactivity in the gels was visualized and quantified using a FUJIX BAS2000 analyzer. Ten l were subjected to 10% SDS-PAGE, and the proteins were stained with Coomassie Brilliant Blue R-250 to check the amount of eluted proteins.
Protease Sensitivity Assay-Reticulocyte lysate (4 l) containing 35 Slabeled translation products was incubated in 16 l of 20 mM Hepes-KOH (pH 7.4) containing 50 g/ml cycloheximide and various amounts of trypsin for 10 min on ice. The reaction was stopped by adding 20 l of 20 mM Hepes-KOH (pH 7.4) containing 2 mM phenylmethylsulfonyl fluoride. Ten l of the mixture (total 40 l) were subjected to 10% SDS-PAGE, and the radioactivity in the gels was visualized.

RESULTS
Expression of C-terminally Truncated Mutants of hTom20 -cDNAs encoding C-terminally truncated mutants of hTom20 ((1-125)hTom20, (1-105)hTom20, (1-89)hTom20, and (1-73)hTom20) as well as wild-type, (1-145)hTom20 (Fig. 1A), were transfected into COS-7 cells in a potent mammalian expression vector pCAGGS. The cells were fractionated into the soluble and membrane fractions and localization of expressed proteins was examined by immunoblot analysis (Fig. 1B). All four mutants, as well as the wild-type, were recovered mostly in the membrane fraction. We previously showed that wildtype hTom20 was correctly targeted to the mitochondria in transfected COS-7 cells (19). Therefore, the four mutants also appeared to be targeted to and efficiently accumulated in the mitochondria. The results also suggest that the mitochondrial targeting signal of hTom20 is localized in the N-terminal 73 amino acid residues containing the membrane-anchor and the linker segments. The level of expression of each construct relative to endogenous hTom20 can be judged by comparison with the control (Fig. 1B).
Effect of Overexpression of Truncated hTom20s on Mitochondrial Import of pOTC-When human pOTC is transiently expressed in COS-7 cells, it is imported efficiently into the mitochondria and processed to the mature form, as revealed by cell fractionation and immunoblot analysis (18,21). The effect of overexpressing C-terminally truncated hTom20 on pOTC import and processing was analyzed by pulse-chase experiments, and the results were quantitated (Fig. 2). When the COS-7 cells expressing pOTC alone were labeled for 5 min with [ 35 S]methionine, 45% of the newly synthesized pOTC was converted to the mature form (line a). When the cells were then chased with cold methionine, the labeled pOTC was converted to the mature form. When wild-type hTom20 was coexpressed, only about 15% of newly synthesized pOTC was processed in a 5-min pulse, and pOTC was processed to the mature form much more slowly in the chase (line b). Very similar results were obtained when the four truncated mutants were coexpressed (lines c-f). These results indicate that insertion of the N-terminal half of hTom20, containing the membrane-anchor and the linker segments (region 1-73) into mitochondria, is sufficient to inhibit pOTC import and processing and therefore the domain containing the TPR motif, glutamine-rich segment, and the C-terminal segment are not essential for inhibition of import and processing in cultured cells.
Effect of Overexpression of Truncated Tom20s on Mitochondrial Import of pOTC-GFP-When the fusion protein pOTC-GFP, consisting of the presequence of human pOTC fused to GFP, is expressed in COS-7 cells, it is imported into mitochondria and processed to mature GFP (19). The effect of overexpressing truncated hTom20s on pOTC-GFP import and processing was analyzed by pulse-chase experiments (Fig. 3). When the cells expressing pOTC-GFP alone were labeled for 5 min, about 4% of newly synthesized pOTC-GFP was imported and processed (line a) and was gradually processed to mature GFP in the chase. The import and processing of pOTC-GFP appears to be rate-limiting under these conditions, probably due to the large amount expressed. In contrast to the case with pOTC, when wild-type hTom20 was coexpressed, import and processing was stimulated about 2-fold (line b). (1-125)hTom20 was also stimulatory (line c). On the other hand, coexpression of the shorter mutants resulted in slight inhibition of the import and processing (lines d-f). These results show that the region 106 -125 containing the glutamine-rich segment of hTom20 is essential for stimulation of pOTC-GFP import and processing, but the C-terminal end (region 126 -145) is not. The results also show that insertion of the N-terminal half of hTom20, containing the membrane-anchor and linker segments (region 1-73), into the mitochondrial membrane, inhibits import and processing of pOTC-GFP.
Effect of Overexpression of Truncated hTom20s on Mitochondrial Organization in the Cell-Mitochondria-targeted pOTC-GFP gives organelle-associated fluorescence and overexpression of hTom20 induces perinuclear aggregation of fluorescent mitochondria (19). The effect of C-terminal truncation of hTom20 on the alteration of the mitochondrial organization was studied (Fig. 4). When wild-type hTom20 or (1-125)hTom20 was coexpressed with pOTC-GFP in COS-7 cells, aggregation of fluorescent mitochondria was observed in almost all transfected cells (panels c-f). On the other hand, when Tom20 was truncated to (1-105)hTom20, such mitochondrial aggregation was almost completely reversed, and fluorescent mitochondria were distributed throughout the cytoplasm (panels g and h). The mitochondrial distribution with the shorter mutants (panels i-l) appeared the same as the control (panels a and b). Therefore, the region 106 -125 containing the glutamine-rich segment is essential for perinuclear mitochondrial aggregation, but the C-terminal end (region 126 -145) is not.
Effect of Overexpression of hTom20 on Mitochondrial Morphology-The effect of hTom20 overexpression on morphological alterations of mitochondria was analyzed by electron microscopy (Fig. 5). In control cells, mitochondria of various shapes were distributed throughout the cytoplasm (a and b). In the hTom20-transfected cells, on the other hand, large aggregates of mitochondria were seen adjacent to the nucleus (c and d). The mitochondrial aggregation was seen in about 10% of cells, this value being consistent with the transfection efficiency obtained in these experiments. Such mitochondrial aggregation was not seen in cells transfected with the control plasmid pCAGGS. These results are in accord with those obtained by fluorescence microscopy.
The aggregates are composed of many round mitochondria of various sizes, and practically no mitochondria were seen outside the aggregates. Although all the mitochondria were confined to the aggregates, not all appeared to be in physical contact. No evidence was seen for fusion of mitochondria. The mitochondrial aggregates were close to nucleus but was not in direct contact. No prominent structures other than mitochondria were found in the aggregates.
Effect of GST-hTom20s on in Vitro Import into Isolated Mitochondria-To analyze the roles of hTom20 in mitochondrial protein import in vitro, we expressed a GST fusion protein containing the entire cytosolic domains of hTom20 (GST-(25-145)hTom20) and two containing deletions (GST-(25-105)hTom20 and GST-(25-125)hTom20). In all fusions, the predicted transmembrane region (region 1-24) was omitted. When in vitro synthesized pOTC was incubated with isolated rat liver mitochondria, it was efficiently imported into the mitochondria and processed to the mature form with time ( Fig.   FIG. 4. Effect of overexpression of C-terminally truncated hTom20s on aggregation of mitochondria. COS-7 cells were grown on coverslips in 35-mm culture dishes. One g of pCAGGS-pOTC-GFP was co-transfected with 1 g of pCAGGS (a and b), pCAGGS-(1-145)hTom20 (c and d), pCAGGS-(1-125)hTom20 (e and f), pCAGGS-(1-105)hTom20 (g and h), pCAGGS-(1-89)hTom20 (i and j) or pCAGGS-(1-73)hTom20 (k and l). Fluorescence images of living cells cultured for 24 h were directly photographed as described previously (19). Bars: a, c,  e, g, i, and k, 10 m; b, d, f, h, j, and l, 100 m. 6A). pOTC import was not inhibited by GST, but was markedly inhibited by 20 M GST-(25-145)hTom20. GST-(25-125)hTom20 was as effective as GST-(25-145)hTom20, whereas GST-(25-105)hTom20 was much less inhibitory (Fig.  6, B and C). Therefore, the region 106 -125 which includes the major part of the glutamine-rich segment of hTom20 is critical for pOTC import. On the other hand, there was no measurable effect of GST-(25-145)hTom20 or the other deleted proteins on pOTC-GFP import (Fig. 6, B and C). These results suggest that not only the presequence portion but also the mature portion of the precursor protein is involved in the interaction with hTom20.
Binding of Precursor Proteins to GST-hTom20s-Recently, Schleiff et al. (23) measured the direct interaction between hTom20 and precursor proteins by using the cytosolic portion of hTom20 fused to GST. We employed their system to analyze the binding of pOTC and pOTC-GFP to the C-terminally truncated hTom20 s. 35 S-Labeled pOTC or pOTC-GFP synthesized in vitro was incubated with glutathione-agarose beads prebound with GST or GST-hTom20 proteins, and the precursor proteins, and GST fusions were then eluted with reduced glutathione. About 80% of GST and GST-hTom20 fusions that were applied to the binding assay, were recovered in the eluate (data not shown). Eight percent of applied pOTC was bound to GST-(25-145)hTom20 (Fig. 7). pOTC binding was decreased as hTom20 was C-terminally truncated from 145 to 125. It was further decreased as hTom20 was further truncated from 125 to 105, but a moderate amount of binding still remained (about one-fourth of that to (25-145)hTom20). Slightly less (about 6%) of pOTC-GFP bound to GST-(25-145)hTom20. pOTC-GFP binding decreased as hTom20 was truncated. However, in sharp contrast with pOTC, the binding of pOTC-GFP was almost completely lost when hTom20 was deleted to (25-105)hTom20. When pOTCN-GFP in which the presequence plus 58 residues of mature OTC was fused to GFP, was used, it behaved very similarly to pOTC. We have previously shown that this construct is imported more efficiently than pOTC-GFP, but upon processing it does not become fluorescent, suggesting that, unlike pOTC-GFP, this construct remains unfolded after removal of the presequence (19). These results indicate that both the C-terminal segment and the glutaminerich region are involved in precursor binding, but the region 106 -125 containing the glutamine-rich segment is critical for the binding of pOTC-GFP, but was less critical for binding of pOTC and pOTCN-GFP.
Protease Sensitivity of Precursor Proteins-Although pOTC and pOTC-GFP have the same presequence, they differ in mitochondrial import both in vivo and in vitro, and in binding to the hTom20 mutants. In order to determine whether this difference is due to a different tendency of the precursor proteins to fold, we compared protease sensitivity of these two proteins (Fig. 8). When pOTC (about 40 kDa) was treated with increasing concentrations of trypsin, pOTC was digested without formation of a trypsin-resistant fragment. On the other hand, when pOTC-GFP (about 31.5 kDa) was digested under the same conditions, a fragment of 28 kDa, that apparently corresponded to mature GFP, was formed. These results suggest that the mature portion of pOTC remained unfolded, whereas that of pOTC-GFP (namely GFP) was folded into a trypsin-resistant folded conformation. DISCUSSION The results obtained from the present experiments with cultured cells show that amino acids 106 -125 of hTom20, which contain the glutamine-rich segment, is essential for stimulation of pOTC-GFP import, but the C-terminal end (region 126 -145) seems not to be. The essential region has no "consensus" sequence, but is moderately homologous with the corresponding regions of yeast and Neurospora Tom20 (15)(16)(17). Overexpression of the N-terminal half of hTom20 (region 1-73), containing the membrane-anchor and charged linker segments, and its insertion into the mitochondrial outer membrane, inhibits import and processing of both pOTC and pOTC-GFP, suggesting that this portion of hTom20 can disrupt the import machinery. The linker segment of 45 amino acid residues contains 17 positively charged residues (Arg plus Lys) and 6 negatively charged residues (Glu plus Asp). This linker segment may be important for the observed dominant-negative effect in transfected cells. However, this effect may require that the linker segment be attached to the membrane-anchor segment After 24-h culture, the cells were fixed with 2.5% glutaraldehyde, postfixed with 1% osmium tetroxide, dehydrated, and embedded in epoxy resin. Ultrathin sections stained with uranyl acetate and lead nitrate were observed electron microscopically. Bars, 1 m. so that it can associate with other components of the translocase and disrupt its function, since overexpression of the soluble domain of hTom20 (domain 30 -145) containing the linker segment does not inhibit pOTC import in cultured cells (18). Although the TPR motif of Tom20 of yeast was reported to increase interaction of Tom20 with the Tom70-Tom37 complex, its mutation did not inactivate the receptor function of Tom20 (14). Likewise, in this study, we found that the TPR motif of hTom20 was not essential for the dominant negative effect.
We have previously shown that overexpression of hTom20 results in perinuclear aggregation of mitochondria in addition to affecting mitochondrial import and processing (19). Whether the effect on import and perinuclear mitochondrial aggregation are related is not known. However, Tom20 depletion as well as its overexpression is known to induce altered morphology of mitochondria. Tom20-deficient N. crassa cells contain mitochondria that are highly deficient in cristae (24), suggesting that the appropriate number of Tom20 molecules is important in maintaining the normal morphology of mitochondria.
Recently, Iwahasi et al. (25) investigated the functional domain of rat Tom20. They expressed mutant rat Tom20 proteins in ⌬tom20 yeast cells and examined their ability to complement the defects of respiration-driven growth and mitochondrial protein import. Based on their results, they concluded that the N-terminal region containing the membrane-anchor and the linker segments is essential for the function of rat Tom20, whereas the TPR motif, glutamine-rich segment and the Cterminal segment are not. The region essential for the Tom20 function in the complementation study is nearly identical to that required for the dominant-negative effect in this study. In the current study, a distinct region (106 -125) is required for the stimulation of pOTC-GFP import. In a heterologous system, some subtle interaction between mammalian Tom20 and the yeast translocase complex may be overlooked, as suggested by the fact that when yeast Tom20 was substituted with hTom20 in yeast mitochondria, hTom20 worked somewhat differently from the endogenous yeast Tom20 (26).
From the results of in vitro import and binding assays, both the glutamine-rich segment-containing region (region 106 -125) and the C-terminal end of hTom20 (region 126 -145), especially the former region were shown to be involved in the binding to precursors. These results are in close agreement FIG. 6. Effect of GST-fused hTom20s on mitochondrial import of pOTC and pOTC-GFP. A 35 S-labeled reticulocyte lysate translation product was incubated in the absence (control) or presence of 20 M of GST or GST-fused hTom20s at 25°C. A, at the indicated times, the import reaction was stopped and subjected to 10% SDS-PAGE. The amount of pOTC present in 20% of import mixture, representing the input radioactivity for each import assay, is shown as 20%. The radioactive polypeptides were visualized by image plate analysis. B, the radioactive mature OTC and mature OTC-GFP on the SDS-polyacrylamide gels, shown in panel A, were quantitated by image plate analysis. The percent import represents the amount of mature protein compared with the input precursor. C, the radioactive amount of mature OTC (15-min incubation) and mature OTC-GFP (40-min incubation) were quantitated by image plate analysis, and the percent import is shown. Values are represented by means Ϯ S.E. of three independent experiments.
with those obtained on cultured cells. Schleiff et al. (23) showed recently that a predicted "glutamine face" (region 104 -114) and an extreme C-terminal cluster of acidic residues (region 141-145) have effects on binding of some precursor proteins. Stimulation of pOTC-GFP import seems to be attributable to the predicted glutamine face, most of which is included in the region 106 -125. The critical necessity of this region for binding to the presequence of pOTC-GFP was shown through the binding assay. In addition, the binding studies suggest that there may be additional interactions between hTom20 and precursor proteins through mature protein domains adjacent to the cleavable presequence. Thus, we have previously shown that the import velocity of pOTCN-GFP was almost the same as that of pOTC but was about 2.5-fold higher than that of pOTC-GFP (19). This is consistent with the ability of both pOTC and pOTCN-GFP to bind to a similar degree to region 25-105 of hTom20, whereas pOTC-GFP does not bind to this region. The interaction between the region 25-105 of hTom20 and the N-terminal portion of mature OTC seems to facilitate the import of pOTC. Previous study on an artificial precursor fused with dihydrofolate reductase has shown that destabilization of the precursor facilitates its import (27). The low import velocity of pOTC-GFP may be attributable to the stability of the GFP domain, whereas the high import velocity of pOTC and pOTCN-GFP may be due to their instability, in which the N-terminal mature portion of OTC may work as an intramolecular chaperone as shown for the F 1 -ATPase ␤-subunit (28).
Based on the in vitro assays and the experiments in cultured cells, we propose the following hypothesis (Fig. 9). Among the preproteins, pOTC (and pOTCN-GFP) is unfolded and when bound to hTom20, it is rapidly imported into the mitochondria. However, pOTC unbound to hTom20 or bound to nonfunctional receptor forms an import-incompetent aggregate. On the other hand, the GFP portion of pOTC-GFP is rapidly folded, and the preprotein remains import-competent prior to binding to the functional receptor and is imported slowly because the folded GFP must be unfolded during translocation. The difference in tightness of binding of pOTC (and pOTCN-GFP) appears to be due to additional interaction with the N-terminal residues present in mature OTC. FIG. 9. Model for the interaction between precursors and hTom20. This model presents the different domains of hTom20 as in Fig. 1 and shows the likely sites of interaction of pOTC/pOTCN-GFP and pOTC-GFP with hTom20. The observed, more rapid import of pOTC and pOTCN-GFP is explained by there precursors being largely unfolded. This unfolded, import-competent conformation, is prone to aggregation giving an import-incompetent state. The import-competent form interact with hTom20 via the presequence and an adjacent segment of the mature protein. Proteins which are in a more tightly folded conformation such as pOTC-GFP are imported more slowly. The pOTC-GFP interacts with the glutamine-rich segment and C-terminal segment of hTom20 only via its presequence. S-labeled reticulocyte lysate translation product was incubated with 40 g/ml cycloheximide and 0, 5, or 50 g/ml trypsin for 10 min on ice. Proteolysis was stopped by adding phenylmethylsulfonyl fluoride to the final concentration of 1 mM. Samples were subjected to 10% SDS-PAGE, and the radioactivity in the gels was visualized by fluorography.