| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
J. Biol. Chem., Vol. 280, Issue 12, 11535-11543, March 25, 2005
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
¶¶**
From the
Department of Biochemistry, La Trobe University, Melbourne 3086, Australia and the ||Department of Molecular Genetics, Kumamoto University School of Medicine, Kumamoto 860-0811, Japan
Received for publication, December 8, 2004 , and in revised form, January 7, 2005.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
400-kDa complex is well resolved via blue native polyacrylamide gel electrophoresis (BN-PAGE) and contains all subunits except Tom70 and usually Tom20 (12–14). A similar sized complex is found in mammalian mitochondria (15, 16), although genes encoding Tom5 and Tom6 counterparts have not been found in the genome.
Almost all mitochondrial proteins employ the TOM machinery for their import, since the vast majority are encoded by nuclear genes, synthesized in the cytosol, and subsequently trafficked into the organelle. Likewise, the TOM subunits must also be targeted to mitochondria and, following their integration into the membrane, are assembled into new TOM complexes. Whereas the import and assembly of individual TOM subunits has been studied in some detail, it seems that Tom40 follows a unique pathway for its biogenesis (17, 18). In vitro import studies in fungi have shown that Tom40 precursors can be arrested in stable intermediates, the first of which is a 250-kDa complex containing the Tom40 precursor associated with the recently identified sorting and assembly machinery (SAM) of the outer membrane (19–23). The SAM complex consists of a number of subunits including Sam50 (Tob55/Omp85) (19, 21, 24), Sam35 (Tob38/Tom38) (25–27), and Sam37 (Mas37/Tom37) (20). How this complex is involved in the integration of Tom40 and other 
-barrel proteins into the outer membrane is not yet known. Following its integration into the outer membrane, the Tom40 precursor binds small TOM subunits before its assembly into the TOM complex (28). This final step requires recruitment of Tom22 via an interaction with Mdm10, which associates with the SAM complex (29).
The steps before Tom40 associates with the SAM complex have not been well characterized due to the lack of detectable intermediates. However, it has been shown that Tom40 precursors require ATP for their initial targeting to mitochondria, suggesting that cytosolic chaperones may be involved (17, 30). Furthermore, it has been proposed that Tom40 precursors initially interact with TOM receptors and also the TOM complex itself before being transferred to SAM. Integration of Tom40 into the outer membrane seems to occur via the intermembrane space (IMS) side (28, 31), and this has recently been supported by studies showing that the IMS small Tim proteins facilitate this process (32, 33). These findings suggest that the Tom40 precursor must first translocate across the outer membrane, presumably through pre-existing Tom40 channels. How the Tom40 precursor is directed to the mitochondrion and engages with members of the TOM and SAM machinery is not known.
In order to address some of these questions, we investigated the biogenesis of human Tom40 (hTom40). We show that hTom40 is kept import-competent by the cytosolic chaperone Hsp90. In addition, we identify a stable association between hTom40 precursors and the TOM complex but not with the SAM complex. Deletion of an evolutionary conserved stretch of amino acids causes the precursor to bypass stable intermediate formation with the TOM complex, and instead it stalls at a complex of lower size that we show to be the human SAM complex.
|
EXPERIMENTAL PROCEDURES |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
The cDNA encoding human Sam50 (accession number AAH11681 [GenBank] was obtained from the I.M.A.G.E. Consortium. For creation of a hSam50-GFP fusion construct, the hSam50 ORF lacking its stop codon was cloned into the pE-GFP vector at SalI and NotI in frame with the GFP ORF. Similarly, for creation of a GFP-hSam50 fusion construct, the hSam50 ORF was amplified and cloned at BamHI and NotI into a modified form of the pE-GFP vector, to give in-frame C-terminal fusions to GFP. For creation of a hemagglutinin (HA)-hSam50 fusion, the hSam50 ORF was cloned at SalI and NotI of the pME-HA vector (35).
For RNA interference studies, two complementary oligonucleotides (sense, 5'-GATCCGGTGATGACGCACTTCCAATTCAAGAGATTGGAAGTGCGTCATCACCTTTTTTGGAAA-3'; antisense, 5'-TTTCCAAAAAAGGTGATGACGCACTTCCAATCTCTTGAATTGGAAGTGCGTCATCACCGGATC-3') were designed specifically for hSam50 depletion using pSilencer vectors according to the manufacturer's instructions (Ambion Inc.). The oligonucleotides were annealed and cloned into pSilencer 3.0-H1 at BamHI and HindIII sites. A control plasmid containing a scrambled sequence (36) was also employed. All clones were verified by DNA sequencing. Plasmids expressing human Tom20 (37), Tom22 (38), and p-OTC (39) were published previously.
Cell Culture and Transfection—Cells were cultured in Dulbecco's modified Eagle's medium (Invitrogen) containing 10% (v/v) fetal calf serum at 37 °C under an atmosphere of 5% CO2 and 95% air. Cells growing on coverslips in 2-cm wells or in 9-cm plates were transfected with2or10 µg of plasmid DNA respectively, using Lipofectamine 2000, according to manufacturer's instructions (Invitrogen). For microscopic analysis, cells were incubated with 80 nM MitoTracker Red CMXRos (Molecular Probes, Inc., Eugene, OR) for 30 min before mounting coverslips onto glass slides. For immunofluorescence, cells were fixed in paraformaldehyde, washed in phosphate-buffered saline containing 0.1% (v/v) Triton X-100, and incubated with appropriate antibodies (36). Cells were visualized using an Olympus BX-50 fluorescence microscope with a x100 oil immersion objective. Images were captured with a SPOT RT 3CCD camera (Diagnostic Instruments) and processed using SPOT Advanced software (Version 3.4).
In Vitro Transcription and Translation—Plasmid DNA or PCR products (40) were used for in vitro transcription using SP6 RNA polymerase (Promega). In vitro translation of the RNA transcripts using rabbit reticulocyte lysates (Promega) in the presence of [35S]methionine/cysteine (TRANS-Label; ICN Biomedicals) was performed as previously described (40). In some cases, hTom40 was in vitro transcribed/translated using a coupled system (Promega).
In Vitro Import and Assembly Assays—Mitochondria were isolated from tissue culture cells (HEK 293T, HT1080, or HeLa) grown as monolayers according to Ref. 16. For in vitro protein import studies, 35S-labeled proteins were incubated with isolated mitochondria at 20 or 37 °C for various times as indicated in the figures. Samples were subjected to various treatments as previously described (40). BN-PAGE was performed according to Simpson et al. (41). All mitochondrial pellets were resuspended in digitonin-containing buffer (1% (w/v) digitonin, 50 mM NaCl, 10% (v/v) glycerol, 20 mM bis-Tris, pH 7.0). BN-PAGE antibody shift assays were undertaken as previously described (16).
Inhibition Assays with C-90 and C-Bag Proteins—Plasmids coding for the hexahistidine-tagged C-terminal domain of Hsp90 (C-90) and the C-terminal domain of Bag-1 (C-Bag) were obtained from Jason Young (Munich) (42). Purification of hexahistidine-tagged recombinant proteins was carried out using nickel-nitrilotriacetic acid-agarose under native conditions according to the manufacturer's instructions (Qiagen). Purified proteins were subsequently dialyzed in import buffer minus sucrose (5 mM MgOAc, 80 mM KOAc, 1 mM dithiothreitol, 0.1 mM ADP, 10 mM sodium succinate, 20 mM Hepes-KOH, pH 7.4). Inhibition assays were performed according to Young et al. (42). Inhibitory recombinant protein (C-90 or C-Bag) or control bovine serum albumin (20 µM final concentration) was added to [35S]hTom40 in 150 µl of import buffer in the presence of 2.5 mM ATP. After preincubation at 37 °C for 5 min, mitochondria were added and incubated for various times. After import, mitochondria were reisolated, washed, and solubilized in 1% (w/v) digitonin buffer and subjected to BN-PAGE (40).
Antibodies—Antibodies specific for Tom20 and Tom22 were raised in rabbits (38). Antibodies to hTom40 and hSam50 were raised against hexahistidine-tagged recombinant protein, expressed in E. coli and purified using nickel-nitrilotriacetic acid-agarose (Qiagen) chromatography under denaturing conditions. Antigens were precipitated by dialysis against phosphate-buffered saline and injected into rabbits using Freund's complete adjuvant and (for subsequent boosts) Freund's incomplete adjuvant. Antibodies against cytochrome c were purchased from BD Biosciences.
Miscellaneous—Tris-Tricine SDS-PAGE was performed as previously described (43). Radiolabeled preproteins were detected using PhosphorImage storage technology and quantitated using ImageQuant software (Amersham Biosciences). Western blotting was performed using a semidry transfer method (44). Immunoreactive proteins from blots were detected using horseradish peroxidase-coupled secondary antibodies and SuperSignal West Pico chemiluminescent substrate (Pierce). Images were obtained using a ChemiGenius chemiluminescence detection system (SynGene). The hTom40 topology prediction was performed based on a hidden Markov method using the program PRED-TMBB (available on the World Wide Web at bioinformatics2.biol.uoa.gr/PRED-TMBB/input.jsp) (45).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
500-kDa complex (termed "I"; lanes 1 and 2), whereas after longer incubations, they assemble in a smaller 100-kDa complex (II) before finally accumulating in a 380-kDa complex (lane 4). The latter complex is identical in size to the mature TOM complex as verified by immunodecoration with anti-Tom40 and anti-Tom22 antibodies (lanes 5 and 6). In some cases, hTom40 precursors were also found in a poorly resolved, high molecular weight species. Since Western blot analysis shows that pre-existing Tom40 molecules are found in a single resolvable TOM complex at 380 kDa (lane 5), the other complexes observed during the import of Tom40 precursors most likely represent kinetically trapped assembly intermediates (28). Indeed, we found that [35S]hTom40 could be arrested specifically at the 500-kDa intermediate I complex after incubating with mitochondria for 4 min at 30 °C (Fig. 1B, lanes 1 and 7). After performing a chase at 20 °C, the radiolabel in the 500-kDa complex was reduced in intensity, accompanied by some increase in the intensity of complex II. In contrast, performing the chase at 37 °C resulted in a loss in intensity of the 500-kDa species with the corresponding assembly of Tom40 precursors into complex II and, more strongly, into the mature TOM complex (Fig. 1B, lanes 7–12). The profile of hTom40 assembly was verified in mitochondria isolated from other tissue culture cell lines (data not shown). Thus, we conclude that in human cells, Tom40 precursors first assemble into a 500-kDa intermediate before their maturation into the TOM complex via a 100-kDa intermediate. In contrast, in yeast, Tom40 precursors first stably associate with a protease-resistant SAM complex of 250 kDa, followed by its membrane insertion into a 100-kDa complex before its final assembly into the TOM complex (17, 28). The assembly pattern found for hTom40 resembles that of its yeast counterpart except that the size of the initial assembly intermediate I is significantly larger. Whereas this size difference is puzzling, it has been suggested that the SAM complex in yeast observed by BN-PAGE is the core complex, since additional SAM subunits may dissociate from the complex following membrane solubilization (29). Thus, the first assembly intermediate seen for hTom40 precursors might be an association with a more stable SAM complex present in human mitochondria. However, unlike the SAM-arrested Tom40 seen in yeast, the hTom40 precursor arrested at assembly intermediate I was accessible to externally added protease (Fig. 1C, compare lanes 1 and 2). In contrast, when hTom40 was imported for longer times and assembled into the TOM complex, it was largely resistant to the addition of external protease, with the reduction in size of the complex corresponding to the shaving of the cytosolic domain of Tom22 (16, 38). The intermediate complex I might therefore represent the hTom40 precursor associated with the TOM complex.
|
Human Tom40 Precursors Require Hsp90 for Import—Yeast Tom40 translated from reticulocyte lysate has been shown to require ATP for its import (17, 47), thus suggesting that, at least in this in vitro system, mammalian chaperones are required. We sought to address the role of this ATP requirement using entirely mammalian components in an in vitro import. Following translation of hTom40, ATP was depleted from the lysate using the ATP-diphosphatase, apyrase, and added to mitochondria also depleted of ATP. A time course of import and assembly (Fig. 2A) showed that in the absence of ATP, hTom40 precursors accumulated at the high molecular weight position and did not form a stable 500-kDa intermediate. After 30 min of incubation, however, the hTom40 precursor assembled into the TOM complex but to a lesser extent than in the presence of ATP. The high molecular weight complex (labeled with an asterisk) resembles that of the ADP/ATP carrier (AAC) precursor, which, in the absence of ATP, forms a high molecular weight complex with Tom70 oligomers as well as other components of the TOM machinery (48, 49) and also Hsp90 (42). We therefore wondered whether the hTom40 precursor requires Hsp70 or Hsp90 to assist with its targeting and assembly into the mitochondrial membrane. To test this, we performed competition analyses using either the purified domain of C-Bag, which competitively binds to Hsc70, or the C-terminal domain of Hsp90, which serves as a dominant negative mutant by binding to Tom70 and preventing productive interaction with native Hsp90 (42). The addition of these purified recombinant proteins (Fig. 2B) to lysates containing hTom40 prior to the addition of mitochondria led to a decrease in the formation of the high molecular weight complex in comparison with the bovine serum albumin control (Fig. 2C). In addition, the amount of hTom40 at intermediate I was lower in the presence of C-Bag, as was the yield of fully assembled hTom40 (lanes 5–8), indicating the involvement of Hsp70 in Tom40 targeting. Impairment of Hsp90 function, however, led to stalling of hTom40 precursors at the TOM machinery (Intermediate I; Fig. 2C, lanes 9–12). This indicates that Hsp90 is also required for driving the assembly of hTom40, perhaps by maintaining the precursor in an appropriate conformation. Loss of Hsp90 function may not prevent targeting of hTom40 to mitochondria but may lead to improper folding of the precursor, thereby blocking its import and assembly pathway. We therefore conclude that Hsp90 and, to a lesser extent, Hsp70, play an important role in the import of hTom40 into mitochondria.
|
|
|
-barrel (Fig. 5A), although the C-terminal end may also form an additional transmembrane strand. The N-terminal end of rat Tom40 was found experimentally to face the cytosol, and the C-terminal end was found to face the intermembrane space (52). The N-terminal end of hTom40 is proline-rich and lacks sequence similarity between fungal Tom40 proteins. We made a hTom40 construct lacking amino acids 2–44 and fused it to GFP. Fluorescence microscopy of COS-7 cells expressing hTom40
2–44-GFP revealed that this construct was still able to target to mitochondria, displaying fluorescence distribution identical to that of the full-length hTom40-GFP fusion (Fig. 5B). Expression of the truncated fusion protein was confirmed via Western blot analysis (data not shown). To determine if this N-terminal region is important for assembly of hTom40 into the TOM complex, mitochondrial in vitro import and assembly assays were performed. As can be seen (Fig. 5C), hTom402–44 was still able to assemble into the TOM complex like hTom40, with the same assembly intermediates being observed. Thus, this N-terminal region in hTom40 is dispensable for its mitochondrial targeting and subsequent assembly with pre-existing wild type Tom40 and other TOM subunits within the complex. This finding is somewhat surprising, given that fusion of GFP to the N terminus of hTom40 blocks mitochondrial targeting. Furthermore, residues 40–44 are predicted to form part of the first transmembrane -strand (Fig. 5A). These results suggest that this is probably not the case. It has, however, been shown that deletion of an evolutionary conserved region within the N-terminal portion of Neurospora Tom40 (residues 51–60) prevents its efficient assembly into the TOM complex (51). These residues are predicted to form a loop between the first two transmembrane -strands (Fig. 5A). We therefore made a similar construct where we deleted residues 72–87 in hTom40. These residues are highly conserved across all species (51). Human Tom4072–87 fused to GFP was still capable of mitochondrial targeting when expressed in cells (Fig. 5B); however, when imported into mitochondria in vitro, the protein (lacking GFP) was unable to assemble into the TOM complex (Fig. 5C, lanes 9–12). A complex, however, was resolved at 200 kDa (marked 
), comparable in size with yeast Tom40 precursors arrested at the SAM intermediate (28). Furthermore, this complex containing hTom4072–87 was not accessible to externally added proteinase K (lane 12), indicating that it might be in the IMS. A similar stalled complex was observed in the assembly of the Neurospora Tom40 mutant precursor (51) and is consistent with its being in association with the SAM complex. Indeed, this complex was accessible to protease in mitochondria containing an opened intermembrane space as a result of hypotonic swelling (Fig. 5D, lane 4). The precursor of ornithine transcarbamylase that assembles into a trimer of 120 kDa following its import into the matrix was not protease-accessible in these mitochondria, thus demonstrating that swelling led specifically to rupture of the outer membrane (Fig. 5D, lanes 5 and 6).
|
72–87 might represent the first evidence of such a complex having conserved function. Human mitochondria contain the proteins metaxin 1 and 2 that share limited homology to Sam37 and -35, respectively (25, 53, 54), although metaxin 1 does not form a discrete complex on BN-PAGE (55). A human Sam50 homologue has been identified at the level of protein sequence (21), but not characterized at the functional level. In order to determine whether hSam50 is located in mitochondria, we fused it to GFP and expressed it in COS-7 cells. Both N- and C-terminal GFP-tagged hSam50 fusions were indeed targeted to mitochondria (Fig. 6A). Although clearly mitochondrial, some cells expressing the GFP fusions also displayed a background cytosolic localization. To eliminate GFP as a potential hindrance for proper insertion into the outer membrane, we employed a construct expressing hSam50 with an N-terminal HA tag. The HA-hSam50 fusion was also targeted to mitochondria, although some cytosolic fluorescence was still evident (Fig. 6A). Continued hSam50 expression led to impaired mitochondrial morphology, and transfected cells had a short lifetime, indicating that the fusion protein was toxic (data not shown). We generated antibodies to recombinant hSam50 purified from bacterial inclusion bodies. Western blot analysis of mitochondrial extracts detected a specific band at 52 kDa, corresponding to the size of its in vitro translation product (data not shown). Treatment of intact mitochondria with proteinase K did not significantly reduce the signal of hSam50 or that of the control matrix Hsp70 but did lead to clipping of a cytosolically exposed domain of Tom22 (Fig. 6B). Proteinase K treatment of mitochondria following rupture of the outer membrane by osmotic shock did, however, lead to the loss of Sam50 signal as well as the controls Tom40, Tom22 (using antibodies against the intermembrane space domain), and cytochrome c while leaving matrix Hsp70 protected as expected (Fig. 6C). Human Sam50 nevertheless displayed characteristics of it being membrane-integrated, since it was not extracted into the supernatant fraction following alkali washing (Fig. 6C, compare lanes 4 and 5). The control integral membrane proteins Tom40 and Tom22 were also not extracted, but the peripheral membrane protein cytochrome c and matrix-soluble mtHsp70 were found in the supernatant fraction. Finally, we showed through two-dimensional BN-PAGE and Western blotting of digitonin-lysed mitochondria, that hSam50 is found in a protein complex of 200 kDa, distinct from the 380-kDa TOM complex (Fig. 6D). We conclude that hSam50 is a mitochondrial outer membrane protein exposed to the intermembrane space and potentially in complex with itself or additional subunits.
|
250-kDa complex observed for hTom4072–87 represents an association with the SAM complex. After import of radiolabeled hTom4072–87 or hTom40 (as control), mitochondria were isolated and subjected to antibody shift analyses using BN-PAGE. As can be seen (Fig. 7C), antibodies to hTom40 shifted all complexes containing hTom4072–87, including the stalled complex (labeled ), whereas antibodies against Tom22 and metaxin 1 did not, similar to that of preimmune serum control (lanes 6, 8, and 10). Antibodies to Sam50 specifically shifted the stalled complex, indicating that Sam50 is also present in this complex. These antibodies did not shift the 100-kDa Tom40 precursor intermediate or the fully assembled TOM complex (lane 4). These results show that Sam50 is indeed involved in the assembly of Tom40 in humans, thus demonstrating the conserved function of this protein in the biogenesis of Tom40.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
-barrel structure, and like other members of this family, it contains cryptic targeting and assembly information. Not only must human Tom40 act as a protein translocase channel, but like its fungal counterpart, it also must form a stable association with other TOM subunits (at least Tom22 and Tom7 in humans) (15, 16). The process by which Tom40 associates with single spanning, presumably 
-helical, membrane proteins is not known. Furthermore, the way in which the Tom40 precursor is targeted to the mitochondrial surface is also not understood; nor is the process of SAM-mediated protein folding and membrane insertion. In order to understand some of these processes in more detail, we undertook a study into the biogenesis of human Tom40. Such a study enables us to analyze the degree of conservation between fungal and mammalian cells, which differ considerably in their biology, including a number of processes that occur directly at the cytosolic-mitochondrial outer membrane interface (42, 56–58). In this report, we identify an intermediate complex containing hTom40 precursors stably associated with the TOM machinery. This intermediate is observed before hTom40 precursor integration and assembly into the outer membrane. Whereas the intermediate is stable to digitonin solubilization and native electrophoresis, the precursor is accessible to external protease treatment, indicating that it is at least partially exposed to the cytosol. In yeast, it has been shown that the Tom40 precursor initially interacts with the TOM complex, since blocking the TOM channel with a matrix-targeted precursor prevents its association with the SAM complex (19). The finding that a precursor can stably interact with the TOM machinery and be resolved by BN-PAGE has seldom been reported. In most cases, this has been shown by blocking precursor translocation via attachment to a folded protein domain (48). Truscott et al. (46) did, however, show by BN-PAGE that the precursor of the AAC can stably interact with the TOM complex in mitochondria from yeast cells containing mutant small Tim proteins of the IMS. These proteins are crucial to enable efficient translocation of the AAC precursor into the IMS (59). However, a very small amount of TOM-arrested AAC precursor was also observed in mitochondria isolated from wild type cells (46). The stronger association observed for AAC and hTom40 precursors might reflect the fact that both are predicted to interact with the TOM machinery in a nonlinear form (17, 49).
Depletion of ATP from the in vitro import system caused the hTom40 precursor to be arrested in a new large molecular weight complex on the outer mitochondrial face. We believe that this represents a macromolecular complex consisting of the hTom40 precursor with members of the TOM machinery as well as cytosolic chaperones. This is consistent with an earlier report suggesting that ATP is required for release of Neurospora Tom40 precursors from cytosolic chaperones (17). Furthermore, similar large molecular weight complexes consisting of Tom70 oligomers and other components have been previously observed by BN-PAGE with the AAC precursor in the absence of ATP (48, 49). Up to now, the cytosolic factors involved in Tom40 precursor import have not been identified. Our experimental data support a role of Hsp90, and possibly Hsp70, in delivery of hTom40 to the mitochondrial surface. Hsp70 inactivation caused a decrease in the import efficiency of hTom40. In contrast, inactivation of Hsp90 function still led to binding of the hTom40 precursor to the TOM complex; however, its further progression along the assembly pathway was blocked. This suggests that the precursor is not in a proper folding state for translocation. This would be consistent with the finding that denatured proteins bind and block the TOM channel, whereas cytosolic chaperones can counteract this effect (60). Given these findings, Hsp90 and Hsp70 are probably involved in binding the hTom40 precursor in the cytosol, thereby maintaining it in an import competent state. Such a process is also consistent with the function of Hsp70/Hsp90 in the import of other precursor molecules into mitochondria (42, 61).
Whereas a strong association was observed between Tom40 precursors and the TOM complex in human mitochondria, we were unable to detect a resolvable complex containing the hTom40 precursor bound to the SAM machinery. These observations led us to characterize the mammalian SAM complex. We found that the human Sam50 family member is indeed mitochondrial and, like its yeast counterpart, resolves as a complex of 200 kDa. Mitochondria depleted of hSam50 caused a decreased efficiency in assembly of hTom40 precursors, supporting the function of the SAM complex in -barrel membrane protein assembly seen in yeast (19–21, 23, 48). Whereas we were unable to detect a hTom40-SAM intermediate by BN-PAGE, we did observe a 100-kDa complex similar to yeast. Assembly of hTom40 from the 100-kDa complex to the TOM complex is dependent on the levels of free hTom22 as demonstrated from our overexpression data. This is consistent with the function of Tom22 in assembly of the TOM complex (8, 16).
When we looked at targeting and assembly, it was surprising to find that fusion of GFP at the N terminus of hTom40 abolished mitochondrial targeting, yet the first 44 residues of hTom40 were dispensable for its biogenesis. It is possible that attachment of the GFP -barrel domain to the N terminus altered the overall structure of hTom40, thereby preventing mitochondrial recognition. Whereas hTom40 precursors lacking residues 72–87 were still targeted to mitochondria, they did not form any stable assembly intermediates and instead accumulated at the SAM complex. Residues 72–87 therefore contribute to membrane integration, folding, and/or assembly of hTom40. Since this region is predicted to form a loop structure (Fig. 5A), shortening it might disrupt packing of transmembrane -strands, leading to precursor stalling at the SAM complex. In addition, deletion of these residues seemed to abolish stable intermediate formation with the TOM complex, indicating that this predicted loop might be involved in binding to the TOM machinery. Alternatively, its deletion might alter the precursor structure such that it is more compact so that it translocates more efficiently across the Tom40 channel. Whereas in vitro reconstitution experiments found that the first 165 residues of rat Tom40 are not required for channel formation (52), our studies nevertheless show that at least some of these residues are essential for Tom40 biogenesis.
In conclusion, it seems that for fungi, a stable intermediate is seen with the SAM machinery, whereas in mammals it is seen earlier, at the TOM machinery. This may indicate that upon translocation across the TOM machinery, human Tom40 is in a conformation more amenable for its insertion into the outer membrane, and hence a stable SAM intermediate is not observed. Yeast Tom40 precursors, however, might not present such a barrier for their translocation across the outer membrane but may require additional conformational changes at the SAM complex. Indeed, yeast Tom40 precursors might require further interaction with the SAM machinery due to the existence of additional TOM subunits (Tom5 and Tom6) seemingly lacking in mammalian mitochondria. For yeast, assembly of the TOM complex has been shown to also require the involvement of Mdm10 (29), a protein that seems to be present only in fungi, possibly due to its additional role in attachment of mitochondria to the actin cytoskeleton in these cells (62).
|
FOOTNOTES |
|---|
These authors contributed equally to this work.
¶ Recipient of Australian postgraduate research awards.
** To whom correspondence may be addressed. Tel.: 61-3-9479-2196; Fax: 61-3-9479-2467; E-mail: N.Hoogenraad{at}latrobe.edu.au. To whom correspondence may be addressed. Tel.: 61-3-9479-2156; Fax: 61-3-9479-2467; E-mail: M.Ryan{at}latrobe.edu.au.
1 The abbreviations used are: TOM, translocase of the mitochondrial outer membrane; AAC, ADP/ATP carrier; BN-PAGE, blue native polyacrylamide gel electrophoresis; GFP, green fluorescent protein; IMS, intermembrane space; ORF, open reading frame; SAM, sorting and assembly machinery; bis-Tris, 2-[bis(2-hydroxyethyl)amino]-2-(hydroxymethyl)propane-1,3-diol; Tricine, N-[2-hydroxy-1,1-bis(hydroxymethyl)ethyl]glycine; hTom40, human Tom40.
|
ACKNOWLEDGMENTS |
|---|
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
