Review Mitochondrial protein import in animalsmitochondrial proteins are initially synthesized on free ribosomes as larger precursors with NH -termi-2 nal presequences which function

Ž .Biochimica et Biophysica Acta 1403 1998 12–27

Review

Mitochondrial protein import in animals

Masataka Mori ), Kazutoyo TeradaDepartment of Molecular Genetics, Kumamoto UniÕersity School of Medicine, Kuhonji 4-24-1, Kumamoto 862, Japan

Received 11 August 1997; revised 2 February 1998; accepted 19 February 1998

Keywords: Mitochondrion; Protein import; Animal

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13

2. Cytosolic factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13Ž .2.1. Heat shock cognate 70 protein hsc70 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14

Ž .2.2. dj2 HSDJrhdj-2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14Ž .2.3. Mitochondrial import stimulation factor MSF . . . . . . . . . . . . . . . . . . . . . . . . . . . 16

Ž .2.4. Targeting factor and presequence binding factor PBF . . . . . . . . . . . . . . . . . . . . . . 17

3. The Tom complex . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183.1. Tom20 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183.2. Metaxin, mammalian Tom37? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193.3. Tom34 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19

4. The Tim complex . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19

5. Mitochondrial matrix factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20Ž .5.1. Mitochondrial processing peptidase MPP and mitochondrial intermediate peptidase

Ž .MIP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20Ž .5.2. Mitochondrial hsp70 mhsp70 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21

5.3. Mitochondrial DnaJ and GrpE homologs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21Ž . Ž .5.4. cpn60 hsp60 and cpn10 hsp10 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21

6. Regulation of mitochondrial protein import . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22

Ž .7. New approaches using green fluorescent protein GFP . . . . . . . . . . . . . . . . . . . . . . . . . 22

Abbreviations: cpn, chaperonin; GFP, green fluorescent protein; hsp, heat shock protein; hsc70, heat shock cognate 70 kDa protein;MIP, mitochondrial intermediate peptidase; MPP, mitochondrial processing peptidase; MSF, mitochondrial import stimulation factor;PBF, presequence binding factor; pOTC and OTC, precursor and mature form of ornithine transcarbamylase; Tom and Tim, translocaseof the outer and inner membrane of mitochondria; TPR motif, tetratricopeptide repeat motif

) Corresponding author. Fax: q81-96-373-5145; E-mail: [email protected]

0167-4889r98r$19.00 q 1998 Elsevier Science B.V. All rights reserved.Ž .PII S0167-4889 98 00021-4

( )M. Mori, K. TeradarBiochimica et Biophysica Acta 1403 1998 12–27 13

8. Defects in mitochondrial protein import . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23

9. Concluding remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23

Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24

1. Introduction

Molecular mechanisms involved in protein importinto mitochondria is well conserved from lower eu-karyotes to higher animals, but some differencesexist. In animals as well as in lower eukaryotes, mostmitochondrial proteins are initially synthesized onfree ribosomes as larger precursors with NH -termi-2

nal presequences which function as mitochondrialtargeting and import signals, and are released intocytosolic pools. The precursors are then importedrapidly into the mitochondria and proteolytically pro-

w xcessed to the mature form 1 . Some proteins have nocleavable presequence but do have signals within themature polypeptides. The whole process of synthesisof mitochondrial proteins, their translocation, pro-cessing and folding involves factors present in thecytosol, mitochondrial membranes and matrix com-

w xpartments 2 . In the cytosol, hsc70 is involved inŽtransport and its partner molecular chaperone a DnaJ

.homolog has been identified. Presequence-specificcytosolic factors have been found in mammals. Animport-competent precursor complex was identifiedand its composition is being analyzed. These prese-quence-specific factors and cytosolic chaperone fam-ily members may participate in sorting between fold-

Ž . Žing cytosolic proteins and anti-folding mitochon-.drial proteins of the newly-synthesized proteins. In

lower eukaryotes, components of the protein importŽmachinery on the mitochondrial outer membrane the

. w x ŽTom complex 3–6 and inner membrane the Tim. w xcomplex 5–9 have been identified and character-

ized. Recently, a few mammalian homologs of theTom and Tim components and a novel Tom compo-nent have been found and their roles are being stud-ied. Imported precursors are proteolytically pro-cessed, folded and assembled to final conformationsby the action of processing peptidase and the mito-chondrial molecular chaperone system. In animals,

most studies were performed in vitro, but new ap-proaches using cultured cells and green fluorescent

Ž .protein GFP have been elaborated. Some cases ofregulated mitochondrial protein import have beenfound and defects in the import in humans result inmitochondrial diseases.

In this minireview, we summarize recent progressin studies on molecular mechanisms of protein target-ing to and importing into the mitochondria in higheranimals. Factors involved in mitochondrial protein

Table 1Factors involved in mitochondrial protein import in S. cereÕisiaeand mammals

S. cereÕisiae Mammals

Cytosol Ssa1-2p hsc70Ž .Ydj1p dj2 HSDJrhdj-2

MSFPBFTargeting factor

Outer membrane Tom20 Tom20Tom22Tom70

aTom37 MetaxinTom40Tom7Tom6Tom5

Inner membrane Tim54Tim44 Tim44Tim23 Tim23Tim22Tim17 Tim17

Matrix MPP MPPMIP

Ssc1p mhsp70Mdj1p

bMge1prmGrpE mGrpEhsp60rcpn60 hsp60rcpn60hsp10rcpn10 hsp10rcpn10

aIt is not known whether metaxin is the mammalian Tom37.bFunctional analysis not yet made.

( )M. Mori, K. TeradarBiochimica et Biophysica Acta 1403 1998 12–2714

Ž . Ž .Fig. 1. A model of protein import into the mitochondria in yeast A and mammals B . OM and IM, outer and inner membrane ofŽ .mitochondria, respectively. MPP, mitochondrial processing peptidase; MIP, mitochondrial intermediate peptidase. A Modified from Ref.

w x Ž . Ž .5 . B Metaxin Me is homologous with yeast Tom37. Involvement of mGrpE in protein import in mammals has never beendocumented.

import in mammals as well as in yeast are summa-rized in Table 1, and a model of the protein import inmammals as well as in yeast is shown in Fig. 1.

2. Cytosolic factors

Mitochondrial precursor proteins cannot be im-ported into the mitochondria in a tightly folded state,rather they must be maintained in a loosely-folded,transport-competent state in the cytosol prior to im-port. Presequences might prevent or retard folding ofthe mature portion of mitochondrial precursor pro-teins. However, some precursor proteins was isolatedin an enzymatically-active form or could be folded

w xinto an enzymatically-active form 10,11 . These re-sults indicate that the presequence portion does notalways interfere with much of the folding of themature portion.

Ž .It was thus assumed that there is a component s inthe cytosol, the function of which is to maintainnewly-synthesized precursor proteins in a conforma-tion enabling transport across mitochondrial mem-

Ž .branes. We found that a protein component s in therabbit reticulocyte lysate markedly stimulates the up-take of in vitro-synthesized pre-ornithine transcar-

w xbamylase pOTC by isolated mitochondria 12 . Re-cently, several cytosolic factors that facilitate mito-chondrial protein import have been found in mam-

Ž .mals as well as in lower eukaryotes Table 1 , includ-ing molecular chaperones and presequence-specific

w xfactors 13 .

( )2.1. Heat shock cognate 70 protein hsc70

The roles of 70-kDa cytosolic heat shock proteinŽ .hsp70 in intracellular protein traffic were first stud-ied using Saccharomyces cereÕisiae hsp70. The yeast


cytosol contains two classes of hsp70, the SSA sub-Ž . Ž .family Ssa1-4 and the SSB subfamily Ssb1-2

w x14,15 . Ssa proteins facilitate translocation of precur-sor proteins into both mitochondria and microsomes,whereas Ssb proteins seem to play a role in transla-tion.

In mammals, two hsp70 family members, hsp70and hsc70, are present and their amino acid se-

w xquences are about 87% identical. Sheffield et al. 16Žstudied effects of reticulocyte hsp70 presumably

.hsc70 on folding, aggregation and import compe-tence of a recombinant hybrid precursor protein inwhich the presequence of rat pOTC was fused tomurine cytosolic dihydrofolate reductase. Theyshowed that hsp70 prevented aggregation of the hy-brid precursor, in an ATP-dependent manner, andretarded the apparent rate and extent of its folding. Asmall amount of hsc70 was shown to be tightlyassociated with the mitochondrial outer membrane

w xfrom rat liver, but its significance is unknown 17 .Cytosolic and mitochondrial isozymes provide a

unique system to study the mechanism of proteinw xsorting. Lain et al. 18 analyzed structural features of

cytosolic and mitochondrial forms of aspartateaminotransferase. They showed that the mitochon-drial precursor form and its mature form bind hsp70Ž .probably hsc70 very early during translation,whereas the cytosolic form does not interact withhsp70. Thus, both the presequence and NH -terminal2

region of the mature portion appear to representrecognition signals for binding of hsp70.

To analyze the roles of hsc70 in more detail, wedeveloped a system of hsc70 depletion from the

w xreticulocyte lysate 19 . We found that hsc70 was theonly member of hsp70 family detected in the lysate.Import of pOTC into the mitochondria was muchdecreased when the precursor was synthesized in thelysate depleted for hsc70. This reduction was almostcompletely restored by re-addition of purified hsc70during pOTC synthesis. The re-addition of hsc70after pOTC synthesis and only during the importassay was not effective. These results indicate thathsc70 is required during pOTC synthesis and notduring import into mitochondria. We then studiedhsc70 requirement for mitochondrial import which

w xvaries markedly among precursor proteins 20 . pOTCsynthesized in the reticulocyte lysate forms an im-port-competent complex of about 11 S, in an hsc70-

dependent manner, and direct association of thenewly-synthesized pOTC with hsc70 became evident.Composition of this import-competent complex re-mains to be elucidated.

( )2.2. dj2 HSDJrhdj-2

hsp70 family members generally perform variouscellular functions in cooperation with DnaJ family

w xmembers as partner chaperones 21,22 . In S. cere-Ž .Õisiae, Ydj1p also termed Mas5p facilitates protein

translocation across mitochondrial and endoplasmicw xreticulum membranes 21,22 . Ydj1p is farnesylated

and is associated with the cytosolic surfaces of mem-branes. Ydj1p interacts with Ssa1p and Ssa2p andstimulates the hydrolysis of hsp70-bound ATP andthus triggers the release of hsp70-bound polypeptides.

In mammals, cDNAs for three human DnaJ ho-Ž . Ž .mologs, dj1 hsp40rHDJ-1 , dj2 HSDJrHDJ-2 and

w xbrain-specific HSJ1 have been isolated 22 . Theother two are present in various tissues and cell types.hsp40 was shown to mediate the initial folding pro-cess of the cytosolic proteins, in cooperation with

w xhsc70 and the TriC complex 23 .We studied properties of human dj2 and its role in

w xmitochondrial protein import 24,25 . Human dj2 aswell as Ydj1p has a COOH-terminal ‘CaaX box’motif common to proteins that are modified by preny-lation. Human dj2 was farnesylated, but not its mu-tant C394S in which cysteine of the ‘CaaX box’ wasmutated to serine. In the transient expression ofpOTC in COS-7 cells, pOTC was synthesized andprocessed to the mature form with an apparent half-life of 2–3 min. Coexpression of human dj2 orC394S resulted in retardation of pOTC processing.These results indicate that dj2 is involved in an early

Ž .step s of protein import into mitochondria. Our stud-ies using a system of dj2 depletion from the reticu-locyte lysate, provided unequivocal evidence that dj2,

Ž .but not dj1 hsp40 , is involved in mitochondrialw xprotein import and protein refolding 25 . The signifi-

cance of farnesylation of dj2 and its possible interac-tion with hsc70 and precursor proteins remain to bestudied. A model of possible involvement of dj2 inprotein folding and mitochondrial protein import isshown in Fig. 2. The molecular mechanisms of sort-

Ž .ing between the folding cytosolic proteins andŽ .‘anti-folding’ mitochondrial proteins of the newly


Ž . Ž .Fig. 2. Sorting between folding of cytosolic proteins A and import of mitochondrial precursor proteins B in mammals. Two DnaJŽ . Ž . Ž .homologs, dj1 hsp40rHDJ-1 and dj2 HSDJrHDJ-2 , as well as presequence-specific factors may participate in this sorting. A

w xModified from Ref. 26 . )Possible involvement of dj2 in cotranslational folding of cytosolic proteins remains to be elucidated.

synthesized proteins by the hsc70–DnaJ molecularchaperone teams require further study.

( )2.3. Mitochondrial import stimulation factor MSF

Mitochondrial precursor proteins synthesized inwheat germ lysate are imported poorly or not at all

into isolated mitochondria. A cytosolic factor thatstimulates mitochondrial import of pre-adrenodoxinwas purified from rat liver cytosol, and was named

Ž w x.MSF reviewed in Ref. 27 . MSF is a heterodimerof 32 and 34 kDa subunits, and stimulates mitochon-drial import of several precursor proteins. MSF candepolymerize and unfold aggregated pre-adrenodoxin


in an ATP-dependent manner. MSF forms a stablecomplex with urea-denatured precursors keeping them

w xin a conformation competent for import 28 . Theimport-stimulation activity of MSF is NEM-sensitive,whereas the ATP-dependent unfolding activity is

w xNEM-insensitive 29,30 .Mitochondrial precursor proteins induce ATPase

activity of MSF, irrespective of the presence or ab-w xsence of the cleavable presequence 30 . Chemically-

synthesized presequences bind to MSF and induce itsw xATPase activity 28 .

MSF shares some properties with hsp70 familyproteins, and can be regarded as a molecular chaper-one. Both proteins recognize aggregated proteins andunfold them in an ATP-dependent manner. Further-more, they form complex with unfolded proteins andstabilize the unfolded conformations. Komiya et al.w x31 showed that the precursor proteins are targeted tomitochondria via two distinct pathways: one requir-ing MSF and the other requiring hsp70. The MSF-de-pendent pathway is NEM-sensitive and requires ATPhydrolysis for the release of MSF from the mitochon-

Ž .dria presumably the import receptor , whereas thehsp70-dependent pathway is NEM-insensitive anddoes not require cytosolic ATP.

w xHachiya et al. 32 reconstituted the heterologoussystem using a purified mammalian precursor protein,MSF and either yeast intact mitochondria or intact orsolubilized outer membranes. They showed that theprecursor–MSF complex first binds to the Tom37–Tom70 subunits of the mitochondrial import receptorŽ .see below . After ATP-dependent release of MSF,the precursor is transferred from Tom37–Tom70 tothe Tom20–Tom22 subunits of the receptor, andfinally delivered to the import channel in the outer

Ž .membrane see below . Import in the absence of MSFbypasses Tom37–Tom70. A model was shown

w x w xschematically in Ref. 27 . Komiya et al. 33 ex-tended this work and reconstituted the early steps ofprecursor targeting in a system consisting of thecytosolic domains of yeast Tom20 and Tom70, bovinepre-adrenodoxin, and rat hsp70 and MSF. The Tom70domain only bound the precursor in the presence ofMSF, yielding a precursor–MSF–Tom70 complex;ATP hydrolysis by MSF released MSF and generateda precursor–Tom70 complex. In the presence of theTom20 domain, ATP caused transfer of the precursorfrom the precursor–MSF–Tom70 complex to Tom20.

The Tom20 domain alone bound the precursor in thepresence of hsp70; hsp70 itself was not incorporatedinto the resulting complex.

MSF has been identified as a member of 14-3-3family of proteins, members of which have beenassociated with many diverse intracellular functions,such as neurotransmitter biosynthesis, cell-cycle regu-

w xlations, signal transduction and exocytosis 27 . SinceS. cereÕisiae also contains 14-3-3 proteins, geneticand in vitro analyses, using a homologous system,may help to establish the role of MSF in precursortargeting to the mitochondria.

2.4. Targeting factor and presequence binding factor( )PBF

The in vitro import into the mitochondria of a34-residue synthetic peptide corresponding to the pre-sequence of ornithine aminotransferase requires a

w xcytosolic factor present in rabbit reticulocytes 34 .This factor, named targeting factor, was purified byaffinity chromatography using the synthetic prese-

w xquence peptide 35 . It has a molecular mass of 28kDa and stimulates the import of mitochondrial pre-cursor proteins and increased their binding to themitochondrial surface. However, no further character-ization has been documented.

Rat pOTC was expressed in Escherichia coli andmitochondrial protein import was reconstituted with

w xpurified recombinant pOTC 36 . A protein factor thatbinds to pOTC but not to mature OTC was namedpresequence binding factor or PBF, and was purifiedby affinity chromatography using pOTC-bound

w xSepharose 37 . PBF has a subunit molecular mass ofabout 50 kDa and markedly stimulated the import ofpurified pOTC into the mitochondria.

To further examine the roles of PBF in mitochon-drial protein import, depletion of PBF from thereticulocyte lysate was accomplished by treatmentwith pOTC-Sepharose. pOTC and two other precur-sor proteins synthesized in the lysate depleted of PBFwere poorly imported, and re-addition of the purified

w xPBF fully restored the import 38 . On the other hand,3-oxoacyl-CoA thiolase that is synthesized with no

w xcleavable presequence 39 , was imported into themitochondria in the absence of PBF. These observa-tions provide support for the existence of PBF-depen-


dent and -independent pathways of mitochondrialimport. cDNA for PBF has yet to be isolated.

3. The Tom complex

Mitochondrial precursor proteins, recognized by aspecific import receptor on the mitochondrial surface,are imported through an import channel in the outermembrane.

In yeast, the receptor and channel are composed ofŽseveral non-identical subunits referred as Tom trans-

.locase of outer membrane of mitochondria subunitsw x Ž .3–6 Table 1, Fig. 1 . The receptor consists ofTom70, Tom37, Tom22 and Tom20 and can bedivided into the Tom70–Tom37 and Tom22–Tom20subcomplexes. The Tom70–Tom37 subcomplex rec-ognizes only limited numbers of precursor proteins,while the Tom20–Tom22 subcomplex directly inter-acts with majority of precursor proteins. TheTom20–Tom22 subcomplex is involved in import ofthe precursor proteins that initially interact with the

w xTom70–Tom37 subcomplex 2,4 . The receptor-bound precursors are then delivered to the proteinchannel in the outer membrane. This channel is com-posed of at least five hydrophobic proteins, Tom40,Tom7, Tom6 and Tom5. Tom40 appears to be a coresubunit and small Tom proteins seem to regulate

w xconformation of the Tom channel 40–42 .In mammals, only homologs of Tom20 and Tom37

among the nine Tom subunits in yeast have beenidentified. Surprisingly, a novel human import recep-

Ž .tor Tom34 was isolated Table 1, Fig. 1 .

3.1. Tom20

cDNA for a human homolog of yeast and Neu-w xrospora Tom20 was isolated 43–45 . Human Tom20

has an overall sequence identity of 50% with S.cereÕisiae Tom20 and 61% with N. crassa Tom20.This protein was shown to be the human counterpartof Tom20, because it can assemble with the yeast

w xreceptor complex 43 , and can complement the respi-w xratory defect of tom20-deficient yeast cells 44 . Fur-

thermore, in vitro import of precursor proteins intoisolated mitochondria was inhibited by the soluble

w xdomain of human Tom20 45 and by anti-humanw xTom20 44,45 . The inhibition by soluble Tom20

differed from 69–37% among the four precursor

w xproteins tested 46 . The inhibition by the antibodyalso differed from 80 to 40% among the precursor

w xproteins 46 . The extent to which the four precursorproteins were inhibited by soluble Tom20 coincidedwith that of inhibition by the antibody. Therefore, therequirement for Tom20 differs for different precursorproteins.

Effect of overexpression of human Tom20 onpOTC import in cultured mammalian cells was ana-

w xlyzed 46 . Co-expression of Tom20, but not its solu-ble domain, with human pOTC in COS-7 cells re-sulted in retardation of mitochondrial import andprocessing of pOTC. On the other hand, overexpres-sion of Tom20, but not its soluble domain, resulted ina stimulated mitochondrial import and processing of

Ž .a fusion precursor protein pOTC–GFP in which thepresequence of pOTC was fused to green fluorescent

Ž . w xprotein GFP 47 . This stimulation is surprisingbecause overexpression of one component of theimport receptor complex may result in import inhibi-tion. A dramatic change of mitochondria was ob-served when human Tom20 was overexpressed in

w xCOS-7 cells 47 . The mitochondria formed giganticaggregates adjacent to the nucleus. Overexpression ofthe soluble domain of Tom20 did not induce suchmitochondrial aggregation. Further studies are neededon fine structures of the mitochondrial aggregates,molecular basis of the aggregation, and functionalactivities of the aggregated mitochondria.

Mammalian Tom20 and yeast and NeurosporaTom20 consist of the NH -terminal membrane-anchor2

Ž .segment, the tetratricopeptide repeat TPR motif, acharged amino acids-rich linker segment between themembrane anchor and the TPR motif, and theCOOH-terminal acidic amino acid cluster. Functionalsignificance of these segments in Tom20 was as-sessed by expressing mutant rat Tom20 proteins intom20-deficient yeast cells and by examining theirpotential to complement the defects of respiration-

w xdriven growth and mitochondrial protein import 48 .A mutant consisting of the membrane anchor and thelinker segments, was targeted to the mitochondria andcomplemented the growth and import defects as effi-ciently as did wild-type Tom20, whereas a mutantlacking the linker segment did not do so. Thus, thelinker segment is essential for the function of ratTom20, whereas the TPR motif and the COOH-terminal acidic amino acids are not essential.


Although mammalian Tom20 can complement therespiratory defect of tom20-deficient yeast cells, there

w xappears to be differences. McBride et al. 49 foundthat fusion precursor proteins consisting of the NH -2

terminal domain of yeast Tom70 or NADH cy-tochrome b reductase fused to dihydrofolate reduc-5

tase are imported into yeast but not rat mitochondriain vitro. Substitution of yeast Tom20 with humanTom20 in yeast mitochondria prevented import of thefusion proteins. Therefore, human Tom20 may pre-vent a cryptic matrix targeting sequence from gainingaccess to the protein translocation machinery. To gainbiochemical insight into these properties of human

w xTom20, Schleiff et al. 50 expressed its cytosolicdomain as a fusion protein with glutathione S-trans-ferase and analyzed direct interactions between theimmobilized human Tom20 domain and various typesof precursor proteins. Their results show that humanTom20 interacts directly with a diverse array ofprecursor proteins, perhaps involving different do-main requirements for different classes of precursorproteins.

3.2. Metaxin, mammalian Tom37?

Metaxin, a novel gene located between the gluco-cerebrosidase and thrombospondin 3 genes in themouse, is a mitochondrial protein that extends intothe cytosol while anchored into the outer membrane

w xat its COOH-terminus 51 . In its NH -terminal re-2

gion, metaxin shows significant sequence identity toyeast Tom37. However, important structural differ-ences, apparently as the result of different mecha-nisms of targeting to membranes, also exist betweenthe two proteins. Antibodies against metaxin, whenpreincubated with mitochondria, partially inhibitedthe uptake of adrenodoxin precursor into mitochon-dria. Mice homozygous for disruption of the metaxingene have an embryonic lethal phenotype, indicatingthat this protein is necessary for the early develop-ment of the mouse embryo. It remains to be eluci-dated whether metaxin is the mammalian counterpartof Tom37.

3.3. Tom34

In S. cereÕisiae, the two receptor subcomplexesare linked by the TPR motif, a loosely defined motif

implicated in a wide range of protein–protein interac-w xtions 52 . Yeast Tom70 contains seven copies of this

motif and yeast and mammalian Tom20 contains onlyone copy. Human expressed sequence tag and cDNAdatabase were screened using a degenerate TPR se-quence present in several Tom proteins for a novel

Ž .mitochondrial translocase s . A 34 kDa proteinTom34 that functions as a subunit of the proteinimport apparatus on the outer membrane was identi-

w xfied 53 . Tom34 was located on the surface of themitochondria and was resistant to extraction underalkaline conditions. It has a large COOH-terminaldomain exposed to the cytosol. Antibodies raisedagainst this protein specifically inhibited in vitroimport of pOTC into isolated rat liver mitochondria.In addition, the recombinant soluble domain com-

Žpeted with human Tom34 for pOTC import Yano et.al., unpublished . This novel component of the pro-

tein import machinery has a 62 residue motif con-served with the Tom70 family of mitochondrial re-ceptors, but otherwise appears to have no counterpartso far characterized in the mitochondria of any otherspecies.

4. The Tim complex

Once the positively charged presequence portionshave moved across the outer membrane, they arepulled across the inner membrane by electrostaticpotential across that membrane.

In yeast, the precursor proteins are imported acrossthe inner membrane through a transmembrane chan-

Žnel formed by the Tim translocase of inner mem-.brane of mitochondria complex. The Tim complex is

Žcomposed of at least five proteins Tim54, Tim44,. w x Ž .Tim23, Tim22, and Tim17 5–9 Table 1, Fig. 1 .

Ž .Most or all except Tim44 are firmly embeddedwithin the inner membrane and appear to form pro-tein-conducting channels. One is Tim23–Tim17 sub-complex and the other is Tim54–Tim22 subcomplex.Tim44 is exposed to the matrix side and is in contactwith Tim23–Tim17 channel. Tim44 appears to func-

Ž .tion in cooperation with mitochondrial hsp70 Ssc1pand GrpE as an ATP-driven ‘import motor’ that pullsthe precursor chain across the Tim23–Tim17 channelinto the matrix space. Tim23 shows dynamic dimer–monomer formation and constitutes a part of a mem-


Ž .brane potential DC -dependent protein-conductingw xchannel 54 . Recently, the other subcomplex Tim54–

Tim22 channel was identified and appears to bespecific for the insertion of proteins into the inner

w xmembrane 8,9 .Human and Drosophila melanogaster Tim17s

w x Ž .were isolated from animals 55 Table 1, Fig. 1 . Acomparison of amino acid sequence from human, D.melanogaster and S. cereÕisiae revealed a similarityof 70 to 82%, including 46 to 62% identical aminoacid residues. Tim17 has four hydrophobic segmentsthat are predicted to function as membrane anchorsequences. The targeting and assembly of human andD. melanogaster Tim17s as well as yeast Tim17were characterized with yeast isolated mitochondria.Targeting signals in the mature proteins directed theTim17 precursors to Tom70 present on the mitochon-drial surface. The precursor was inserted into theinner membrane in a membrane potential-dependentmanner, adopted a characteristic topology and assem-bled with Tim23. The mechanisms of targeting andassembly were indistinguishable between the Tim17sfrom distinct organisms, thereby indicating a highconservation during evolution. Besides Tim17, mam-malian counterparts of Tim23 and Tim44 have beenidentified in data banks.

5. Mitochondrial matrix factors

Mitochondrial precursor proteins are transportedacross the mitochondrial membranes in an extended

Ž .state. In yeast, mitochondrial hsp70 mhsp70 func-tions as an ATP-driven motor in cooperation with

Ž .Tim44 and probably also with Mge1p mGrpE , aŽ .mitochondrial homolog of bacterial GrpE Fig. 1 .

After translocation, presequences have to be removedand the processed mature proteins have to be foldedin the matrix. Studies on protein import into thematrix of yeast mitochondria have shown that differ-ent proteins fold with the aid of different molecularchaperones. Some proteins fold with the help of

Ž .h s p 6 0 r c h a p e r o n i n 6 0 c p n 6 0 a n dŽ .hsp10rchaperonin10 cpn10 , mitochondrial ho-

mologs of the bacterial GroEL and GroES, whereasothers can fold without the aid of cpn60 and cpn10w x5 .

In higher animals, processing peptidases, mhsp70,cpn60, cpn10 and mGrpE were identified and their

Žproperties and functions have been studied Table 1,.Fig. 1 .

( )5.1. Mitochondrial processing peptidase MPP and( )mitochondrial intermediate peptidase MIP

Ž .Mitochondrial processing peptidase MPP that isresponsible for processing of mitochondrial precursorproteins was first identified in the mitochondrial ma-

w xtrix of rat liver 56 and was purified to apparentw xhomogeneity 57 . The protease contains the two

subunits of 55 and 52 kDa. The processing activity ofMPP is inhibited by metal chelators and is reactivatedby Mn2q, indicating that it is a metalloprotease.Primary structures of the two subunits of MPP wereelucidated and were found to share significant se-

w xquence homologies 58–60 .Studies using synthetic peptides corresponding to

the presequence of pre-malate dehydrogenase showedthat arginine residues present at y2 or y3 anddistant from the cleavage point are important for

w xrecognition by the enzyme 61 . Furthermore, at leastone of proline and glycine between the distal and

w xproximal arginine residues is also important 62 . Theprotease shows considerable preference for aromatic,and to some extent, hydrophobic amino acids in thePX-position. Structural requirement for recognition by1

MPP was also studied using pre-adrenodoxin, whichhas a long presequence of 58 amino acid residuesw x63 . More than 40 residues and the presence of basic

Žresidues in the distal portion 20–40 residues up-.stream of the cleavage site are required for recogni-

tion of the precursor by the peptidase. It was alsoevident that basic amino acids required for mitochon-drial targeting and those for recognition by the pepti-dase locate separately in the presequence of pre-adrenodoxin.

Some precursor proteins are processed to the ma-ture form by MPP in a single step, whereas othersincluding pOTC are processed in two successive

w xsteps via intermediates 56,64 . The first step wascatalyzed by MPP, and the second by mitochondrial

Ž . w xintermediate peptidase MIP 65 . When the octapep-tide of pOTC was deleted, or when the entire prese-quence of once-cleaved precursor was joined to the


mature portion of the twice-cleaved precursor, nocleavage was produced by either protease. Cleavageof these constructs by MPP was restored by re-insert-ing as few as two NH -terminal residues of the2

octapeptide or of the mature amino terminus of aonce-cleaved precursor.

A subset of mitochondrial proteins, including rho-danese and 3-oxoacyl-CoA thiolase, are imported intothe matrix space, yet are not processed. A syntheticrhodanese signal peptide and translated proteins wererecognized by MPP, as both inhibited the processing.The presequence of pre-aldehyde dehydrogenase con-sists of a helix-linker-helix motif but when the linker

w xportion is removed, processing no longer occurs 66 .Disruption of the helical signal sequence of rho-danese by addition of the linker did not allow forcleavage. However, with the addition of a putative

w xcleavage site, the protein was processed 67 . Thesame cleavage site was added to 3-oxoacyl-CoA thio-lase, but this protein was not processed. It seemsapparent that both the processing site and the struc-ture surrounding this site are important for MPPrecognition.

w xMIP was purified from rat liver mitochondria 68w xand the cDNA was isolated 69 . MIP, a monomer of

75 kDa, is sensitive to sulfohydryl-group reagentsand is stimulated by Mn2q, Mg2q or Ca2q ions, andreversibly inhibited by EDTA. It has similarity tomembers of a subfamily of zinc metallopeptidases.NH -terminal octapeptides function as recognition2

w xsignals for this peptidase 70 . The biological signifi-cance of the two-step processing is unknown.

( )5.2. Mitochondrial hsp70 mhsp70

w xmhsp70 was purified from mammals 71,72 aswell as yeast, and the cDNA for rat mhsp70 was

w xisolated 73 . mhsp70 is about 45% identical withcytosolic hsp70 and hsc70. Association of mhsp70with newly-imported mitochondrial protein wasshown for medium-chain acyl-CoA dehydrogenase, a

w xhomotetrameric protein 74 . Upon import into themitochondria, unfolded enzyme protein first formed atransient complex with mhsp70 and was then trans-ferred to cpn60 to complete its folding into an assem-bly-competent conformation. However, ATP-drivenprotein drawing-in machinery that is found in yeast

and is composed of Tim44 and mhsp70, has to datenot been identified in animals.

5.3. Mitochondrial DnaJ and GrpE homologs

In yeast, folding of some proteins appear to beaided by a matrix-localized complex containing

Žmhsp70rSsc1p, Mdj1p a mitochondrial DnaJ ho-. Žmolog and Mge1prmGrpErYge1p a mitochondrial

. w xGrpE homolog 5 . In mammals, cDNA encoding ratw xmGrpE was isolated 75 . The human sequence is in

the data bank. The deduced amino acid sequences ofmammalian mGrpE exhibit 26% identity with yeastmGrpE and 21% identity with E. coli GrpE. RatmGrpE mRNA is present in most, if not all organs.By contrast to other mitochondrial chaperones,mGrpE are not induced by heat shock. Roles ofmGrpE in mitochondrial protein import in animals

Ž .are unknown. Mitochondrial DnaJ homolog s fromhigher animals has not been documented.

( ) ( )5.4. cpn60 hsp60 and cpn10 hsp10

Mitochondrial cpn60 and cpn10 are structurallyand functionally related to the bacterial chaperoninsGroEL and GroES. The cDNAs for human, Chinese

w xhamster and rat cpn60 have been isolated 76,77 .Like GroEL, cpn60 from N. crassa and S. cereÕisiaeexists as a tetradecamer of 60 kDa subunits and isarranged into two attached heptameric rings with a

w xcentral cavity 13 . However, electron microscopicand chromatographic analyses indicate that mam-malian cpn60 comprises a single toroidal ring of

w xseven subunits 78 .Rat cpn10 exhibits 45% amino acid identity with

GroES, and has no cleavable presequence for mito-w xchondrial import 79 .

As expected from their close similarity to bacterialGroEL and GroES, mitochondrial chaperonins areinvolved in protein folding. Studies using yeast mito-

Ž .chondrial import function mif mutants showed thata cpn60 mutant is impaired in the assembly of pre-

w xproteins but not in translocation 13 . The require-ment for cpn60 in protein folding has also beenillustrated biochemically. The role of cpn10 in pro-tein folding was also verified in yeast. A point muta-tion in the cpn10 gene reduced the binding of cpn10


to cpn60 and resulted a temperature-sensitive pheno-type in which protein folding was impaired at non-permissive temperature.

The association of mammalian cpn60 and cpn10and their potential to refold chemically denaturedproteins in vitro indicate that the mechanisms ofchaperonin action in protein folding are conserved

w xbetween species 78,80,81 . However, because of thesingle toroidal structure of mammalian cpn60, mech-anisms may differ from those of chaperonins frombacteria and lower eukaryotes.

6. Regulation of mitochondrial protein import

When isolated or cultured animal cells are treatedwith uncouplers or rhodamine dyes that dissipate themitochondrial membrane potential, protein import isimpaired and precursor proteins accumulate in the

w xcytosol 82–85 . Under physiological conditions,however, transport of precursor proteins from thecytosol into the mitochondria is rapid and efficient in

w xintact cells 83–88 .There is evidence that mitochondrial import can be

the rate-limiting step of mitochondrial protein synthe-sis. High concentrations of hemin inhibit the importof the precursor of d-amino-levulinate synthase, thefirst enzyme of heme synthetic pathway, into mito-

w xchondria 89 . A conserved motif, termed the hemeŽ .regulatory motif HRM , was identified in prese-

quences of the erythroid d-aminolevulinate synthaseprecursors and was shown to be involved in hemininhibition of transport of these proteins into mouse

w xmitochondria in vitro 90 .w xIn a recent study, Takahashi and Hood 91 found

that different rates of mitochondrial precursor importmay underlie the two distinct mitochondrial pheno-types characterized in skeletal muscle cells, the sub-sarcolemmal and intermyofibrillar mitochondria pop-ulations. Import of precursor proteins was 3–4-foldhigher in the latter population compared with theformer populations.

Superoxide dismutase catalyzes the conversion ofsuperoxide to hydrogen peroxide and molecular oxy-gen, and is thought to function as a defense against

w xthis potentially harmful radical. Wright et al. 92found that paraquat inhibits the processing of humanmanganese-dependent superoxide dismutase precur-

sor by SF9 insect cell mitochondria. Paraquat exertsbiological effects through production of superoxideradicals and depletion of cellular reducing power. Itis interesting that mitochondrial import of a redox-re-lated enzyme is regulated by the cellular redox state.These findings also suggest that redox-modulationmay underlie different rates of precursor processingw x91 and are mentioned above.

7. New approaches using green fluorescent protein( )GFP

In animals, most studies have been performed inan in vitro system in which precursor proteins synthe-sized in reticulocyte lysates were imported into iso-lated mitochondria. Therefore, there is a need forprocedures that will enable protein import to beinvestigated in intact cells. To date, only a limitednumber of pulse-labeling and pulse-chase studies in

w xcultured cells have been reported 84,86–88 . GFPfrom the jellyfish Aequoria Õictoria yields a stronglyfluorescent signal in heterologous cell types and hasbeen used to visualize subcellular organelles andprotein translocation in living cells. Rizzuto et al.w x93,94 constructed a chimeric protein in which themitochondrial targeting presequence of cytochromeoxidase subunit 8 is fused to GFP and they showedthat this chimera is targeted to mitochondria andgives the organelle-associated fluorescence.

Ž .We constructed a chimeric protein pOTC–GFPin which the presequence of the human pOTC wasfused to GFP and we showed that it was targeted toand imported into the mitochondria with proteolyticprocessing and became strongly fluorescent in the

w xorganelle 47 . Mutant pOTC–GFP fusion proteinswith inactive mitochondrial import signals did notgive the mitochondrial fluorescence. Co-expressionof human Tom20 with pOTC–GFP resulted in perin-uclear aggregation of fluorescent mitochondria and instimulation of mitochondrial import of pOTC–GFPŽ .see above . When pOTC–GFP cDNA was microin-jected into nuclei of human fibroblasts, mitochondrialfluorescence was evident as early as 2–3 h later.Thus, GFP fusion protein can be used to visualizemitochondrial structures and to monitor mitochon-drial protein import in a single cell in real time.


8. Defects in mitochondrial protein import

Except for 13 subunits of the respiratory chainencoded by mitochondrial DNA, all other mitochon-drial proteins are encoded by nuclear genes, synthe-sized on cytosolic ribosomes and imported into themitochondria through a series of steps, as describedabove. Defects in any of the multiple steps of thiscomplex import system would be expected to cause‘mitochondrial diseases’. They may be classified intotwo types: one type includes mutations that impairmitochondrial folding and assembly of specific pro-teins. In fact, point mutations have been noted inpresequence portions of the genes encoding OTCw x w x95 , methylmalonyl-CoA mutase 96 and pyruvate

w xdehydrogenase E1a 97 . Interestingly, two missensemutations in the mature portion of ornithine amino-transferase were shown to interfere with mitochon-drial import and processing in patients with the en-

w xzyme deficiency and gyrate atrophy 98,99 . On theother hand, one type of primary hyperoxauria inpatients was shown to be due to mistargeting ofperoxisomal alanine:glyoxylate aminotransferase tomitochondria that depends on activation of a crypticmitochondrial targeting sequence by a point mutationw x100 .

Other types may include mutations that affect oneof the factors involved in one of the multiple steps ofmitochondrial protein import. In this type, mitochon-drial import of multiple mitochondrial proteins willbe impaired. Complete deficiency of these transportfactors may result in an embryonic lethal phenotype,but partial deficiency may result in mitochondrialdiseases. A fetal, systemic mitochondrial disease withdecreased mitochondrial enzyme activities, abnormalultrastructure of the mitochondria and deficiency of

w xmitochondrial cpn60 was documented 101 . The pri-mary defect remains to be elucidated.

9. Concluding remarks

Fundamental structure and function of mitochon-dria are the same among all eukaryotic cells and it isnot surprising that the system of mitochondrial pro-tein import is highly conserved from lower to highereukaryotes. In animals, however, the number and

structure of mitochondria differ among tissues andcell types, and they are responsible for various func-tions in various cell types in addition to ATP synthe-sis. Therefore, there may be unique features in mito-chondrial biogenesis in animals. In fact,presequence-specific factors were identified in thecytosol of mammalian cells and were well character-ized, but such a factor has not been found in yeast.Tom34, a novel import receptor found in mammals,appears to have no counterpart in yeast. It remains tobe elucidated whether mammalian metaxin and yeastTom37 exert the same role in precursor targeting andimport. Furthermore, evidence has been presentedthat the step of protein import is physiologicallyregulated and is rate-limiting in synthesis of somemitochondrial proteins in animals. Further studies areexpected to reveal fine differences between lower andhigher eukaryotes.

Genetic analysis is not readily applicable to animalcells and studies on import receptors and channels arebehind those in lower eukaryotes. However, with theprogress in cDNA projects, mammalian homologs ofyeast and Neurospora Tom and Tim components arebeing identified. It is now important to assess func-tions of these homologs both in vitro and in vivo. Invitro functional assays of Tom components may in-clude inhibition by antibodies, competition by solubledomains, binding of Tom components to preproteinsand cytosolic factors, and others. As an in vitro assayis not available for Tim candidates in animals, it isnecessary to develop such an import assay usingisolated mitoplasts.

In vivo assay systems are also limited in animals.Complementation of yeast cells deficient for importfactors is often useful, but this system is heterolo-gous. Attempts to assess the roles of putative importfactors by transiently co-expressing a preprotein anda factor of interest, its mutant or antisense RNA incultured cells fielded important information. Con-struction of cells depleted of these factors by stablyexpressing antisense RNAs, by using antisense oligo-nucleotides or even by gene-targeting, is awaited.Mitochondria-targeted GFP will be useful in suchefforts.

Finally, defects in mitochondrial protein importmay form a subgroup of ‘mitochondrial diseases’ or‘mitochondrial encephalomyopathies’. Studies onmolecular mechanisms of import processes will lead


to a better understanding of the molecular pathologyof untoward events.

Acknowledgements

Ž .We thank Dr. Y. Nozawa Gifu University, Japanfor recommending that we compile this minireview.M. Ohara provided helpful comments.

References

w x1 F.U. Hartl, N. Pfanner, D.W. Nicholson, W. Neupert,Mitochondrial protein import, Biochim. Biophys. Acta 988Ž .1989 1–45.

w x2 W. Neupert, Protein import into mitochondria, Annu. Rev.Ž .Biochem. 66 1997 863–917.

w x3 T. Lithgow, B.S. Glick, G. Schatz, The protein importŽ .receptor of mitochondria, Trends Biochem. Sci. 20 1995

98–101.w x4 R. Lill, W. Neupert, Mechanisms of protein import across

the mitochondrial outer membrane, Trends Cell Biol. 6Ž .1996 56–61.

w x5 G. Schatz, The protein import system of mitochondria, J.Ž .Biol. Chem. 271 1996 31763–31766.

w x6 V. Haucke, G. Schatz, Import of proteins into mitochondriaŽ .and chloroplasts, Trends Cell Biol. 7 1997 103–106.

w x7 N. Pfanner, E.A. Craig, M. Meijer, The protein importmachinery of the mitochondrial inner membrane, Trends

Ž .Biochem. Sci. 19 1994 368–372.w x8 C. Sirrenberg, M.F. Bauer, B. Guiard, W. Neupert, M.

Brunner, Import of carrier proteins into the mitochondrialŽ .inner membrane mediated by Tim22, Nature London 384

Ž .1996 582–585.w x9 O. Kerscher, J. Holder, M. Srinivasan, R.S. Leung, R.E.

Jensen, The Tim54p–Tim22p complex mediates insertionof proteins into the mitochondrial inner membrane, J. Cell

Ž .Biol. 139 1997 1663–1675.w x10 K. Murakami, F. Tokunaga, S. Iwanaga, M. Mori, Prese-

quence does not prevent folding of a purified mitochon-drial precursor protein and is essential for association with

Ž . Ž .a reticulocyte cytosolic factor s , J. Biochem. Tokyo 108Ž .1990 207–214.

w x11 F. Altieri, J.J. Mattingly, B.F. Rodriguez, J. Youssef, A.Iriarte, T.H. Wu, C.M. Martinez, Isolation and propertiesof a liver mitochondrial precursor protein to aspartateaminotransferase expressed in Escherichia coli, J. Biol.

Ž .Chem. 264 1989 4782–4786.w x12 S. Miura, M. Mori, M. Tatibana, Transport of ornithine

carbamoyltransferase precursor into mitochondria. Stimula-tion by potassium ion, magnesium ion, and a reticulocyte

Ž . Ž .cytosolic protein s , J. Biol. Chem. 258 1983 6671–6674.w x13 M.T. Ryan, D.J. Naylor, P.B. Høj, M.S. Clark, N.J.

Hoogenraad, The role of molecular chaperones in mito-

chondrial protein import and folding, Int. Rev. Cytol. 174Ž .1997 127–193.

w x14 E.A. Craig, B.D. Gambill, R.J. Nelson, Heat shock pro-teins: molecular chaperones of protein biogenesis, Micro-

Ž .biol. Rev. 57 1993 402–414.w x15 J. Rassow, O. von Ahsen, U. Bomer, N. Pfanner, Molecu-¨

lar chaperones: towards a characterization of the heat-shockŽ .protein 70 family, Trends Cell Biol. 7 1997 129–133.

w x16 W.P. Sheffield, G.C. Shore, S.K. Randall, Mitochondrialprecursor protein. Effects of 70-kilodalton heat shock pro-tein on polypeptide folding, aggregation, and import com-

Ž .petence, J. Biol. Chem. 265 1990 11069–11076.w x17 T. Lithgow, M. Ryan, R.L. Anderson, P.B. Høj, N.J.

Hoogenraad, A constitutive form of heat-shock protein 70is located in the outer membrane from rat liver, FEBS Lett.

Ž .332 1993 277–281.w x18 B. Lain, A. Iriarte, J.J. Mattingly, J.I. Moreno, C.M.

Martinez, Structural features of the precursor to mitochon-drial aspartate aminotransferase responsible for binding to

Ž .hsp70, J. Biol. Chem. 270 1995 24732–24739.w x19 K. Terada, K. Ohtsuka, N. Imamoto, Y. Yoneda, M. Mori,

Role of heat shock cognate 70 protein in import of or-nithine transcarbamylase precursor into mammalian mito-

Ž .chondria, Mol. Cell. Biol. 15 1995 3708–3713.w x20 K. Terada, I. Ueda, K. Ohtsuka, T. Oda, A. Ichiyama, M.

Mori, The requirement of heat shock cognate 70 protein formitochondrial import varies among precursor proteins and

Ž .depends on precursor length, Mol. Cell. Biol. 16 19966103–6109.

w x21 J. Rassow, W. Voos, N. Pfanner, Partner proteins deter-mine multiple functions of Hsp70, Trends Cell Biol. 5Ž .1995 207–212.

w x22 D.M. Cyr, T. Langer, M.G. Douglas, DnaJ-like proteins—molecular chaperones and specific regulators of hsp70,

Ž .Trends Biochem. Sci. 19 1994 176–181.w x23 F.U. Hartl, Molecular chaperones in cellular protein fold-

Ž . Ž .ing, Nature London 381 1996 571–580.w x24 M. Kanazawa, K. Terada, S. Kato, M. Mori, HSDJ, a

human homolog of DnaJ, is farnesylated and is involved inŽ .protein import into mitochondria, J. Biochem. Tokyo 121

Ž .1997 890–895.w x25 K. Terada, M. Kanazawa, B. Bukau, M. Mori, The human

DnaJ homologue dj2 facilitates mitochondrial protein im-Ž .port and luciferase refolding, J. Cell Biol. 139 1997

1089–1095.w x26 J. Frydman, E. Nimmesgern, K. Ohtsuka, F.U. Hartl, Fold-

ing of nascent polypeptide chains in a high molecular massŽ .assembly with molecular chaperones, Nature London 370

Ž .1994 111–117.w x27 K. Mihara, T. Omura, Cytoplasmic chaperones in precursor

targeting to mitochondria: the role of MSF and hsp70,Ž .Trends Cell Biol. 6 1996 104–108.

w x28 T. Komiya, N. Hachiya, M. Sakaguchi, T. Omura, K.Mihara, Recognition of mitochondria-targeting signals by acytosolic import stimulation factor, MSF, J. Biol. Chem.

Ž .269 1994 30893–30897.


w x29 N. Hachiya, R. Alam, Y. Sakasegawa, M. Sakaguchi, K.Mihara, T. Omura, A mitochondrial import factor purifiedfrom rat liver cytosol is an ATP-dependent conformational

Ž .modulator for precursor proteins, EMBO J. 12 19931579–1586.

w x30 N. Hachiya, T. Komiya, R. Alam, J. Iwahashi, M. Sak-aguchi, T. Omura, K. Mihara, MSF, a novel cytoplasmicchaperone which functions in precursor targeting to mito-

Ž .chondria, EMBO J. 13 1994 5146–5154.w x31 T. Komiya, M. Sakaguchi, K. Mihara, Cytoplasmic chaper-

ones determine the targeting pathway of precursor proteinsŽ .to mitochondria, EMBO J. 15 1996 399–407.

w x32 N. Hachiya, K. Mihara, K. Suda, M. Horst, G. Schatz, T.Lithgow, Reconstitution of the initial steps of mitochon-

Ž . Ž .drial protein import, Nature London 376 1995 705–709.w x33 T. Komiya, S. Rospert, G. Schatz, K. Mihara, Binding of

mitochondrial precursor proteins to the cytoplasmic do-mains of the import receptors Tom70 and Tom20 is deter-

Ž .mined by cytoplasmic chaperones, EMBO J. 16 19974267–4275.

w x34 H. Ono, S. Tuboi, The cytosolic factor required for importof precursors of mitochondrial proteins into mitochondria,

Ž .J. Biol. Chem. 263 1988 3188–3193.w x35 H. Ono, S. Tuboi, Purification and identification of a

cytosolic factor required for import of precursors of mito-chondrial proteins into mitochondria, Arch. Biochem. Bio-

Ž .phys. 280 1990 299–304.w x36 K. Murakami, Y. Amaya, M. Takiguchi, Y. Ebina, M.

Mori, Reconstitution of mitochondrial protein transportwith purified ornithine carbamoyltransferase precursor ex-

Ž .pressed in Escherichia coli, J. Biol. Chem. 263 198818437–18442.

w x37 K. Murakami, M. Mori, Purified presequence binding fac-Ž .tor PBF forms an import-competent complex with a

Ž .purified mitochondrial precursor protein, EMBO J. 9 19903201–3208.

w x38 K. Murakami, S. Tanase, Y. Morino, M. Mori, Prese-quence binding factor-dependent and -independent import

Ž .of proteins into mitochondria, J. Biol. Chem. 297 199213119–13122.

w x39 Y. Amaya, H. Arakawa, M. Takiguchi, Y. Ebina, S. Yokota,M. Mori, A noncleavable signal for mitochondrial import

Ž .of 3-oxoacyl-CoA thiolase, J. Biol. Chem. 263 198814463–14470.

w x40 K. Dietmeier, A. Honlinger, U. Bomer, P.J. Dekker, C.Eckerskorn, F. Lottspeich, M. Kubrich, N. Pfanner, Tom5functionally links mitochondrial preprotein receptors to the

Ž . Ž .general import pore, Nature London 388 1997 195–200.w x41 A. Alconada, M. Kubrich, M. Moczko, A. Honlinger, N.

Pfanner, The mitochondrial receptor complex: the smallsubunit Mom8brIsp6 supports association of receptorswith the general insertion pore and transfer of preproteins,

Ž .Mol. Cell. Biol. 15 1995 6196–6205.w x42 A. Honlinger, U. Bomer, A. Alconada, C. Eckerskorn, F.

Lottspeich, K. Dietmeier, N. Pfanner, Tom7 modulates the

dynamics of the mitochondrial outer membrane translocaseand plays a pathway related role in protein import, EMBO

Ž .J. 15 1996 2125–2137.w x43 N. Seki, M. Moczko, T. Nagase, N. Zufall, B. Ehmann, K.

Dietmeier, E. Schafer, N. Nomura, N. Pfanner, A humanhomolog of the mitochondrial protein import receptorMom19 can assemble with the yeast mitochondrial recep-

Ž .tor complex, FEBS Lett. 375 1995 307–310.w x44 I.S. Goping, D.G. Millar, G.C. Shore, Identification of the

human mitochondrial protein import receptor, huMas20p.Complementation of delta mas20 in yeast, FEBS Lett. 373Ž .1995 45–50.

w x45 B. Hanson, S. Nuttal, N.J. Hoogenraad, A receptor for theimport of proteins into human mitochondria, Eur. J.

Ž .Biochem. 235 1996 750–753.w x46 K. Terada, M. Kanazawa, M. Yano, B. Hanson, N.J.

Hoogenraad, M. Mori, Participation of the import receptorTom20 in protein import into mammalian mitochondria:analyses in vitro and in cultured cells, FEBS Lett. 403Ž .1997 309–312.

w x47 M. Yano, M. Kanazawa, K. Terada, C. Namchai, M.Yamaizumi, B. Hanson, N.J. Hoogenraad, M. Mori, Visu-alization of mitochondrial protein import in cultured mam-malian cells with green fluorescent protein and effects ofoverexpression of the human import receptor Tom20, J.

Ž .Biol. Chem. 272 1997 8459–8465.w x48 J. Iwahashi, S. Yamazaki, T. Komiya, N. Nomura, S.

Nishikawa, T. Endo, K. Mihara, Analysis of the functionaldomain of the rat liver mitochondrial import receptor

Ž .Tom20, J. Biol. Chem. 272 1997 18467–18472.w x49 H.M. McBride, I.S. Goping, G.C. Shore, The human mito-

chondrial import receptor, hTom20p, prevents a crypticmatrix targeting sequence from gaining access to the pro-

Ž .tein translocation machinery, J. Cell Biol. 134 1996307–313.

w x50 E. Schleiff, G.C. Shore, I.S. Goping, Interaction of thehuman mitochondrial protein import receptor, hTom20,with precursor proteins in vitro reveal pleiotropic specifici-ties and different receptor domain requirements, J. Biol.

Ž .Chem. 272 1997 17784–17789.w x51 L.C. Armstrong, T. Komiya, B.E. Bergman, K. Mihara, P.

Bornstein, Metaxin is a component of a preprotein importcomplex in the outer membrane of the mammalian mito-

Ž .chondrion, J. Biol. Chem. 272 1997 6510–6518.w x52 J.R. Lamb, S. Tugendreich, P. Hieter, Tetratrico peptide

repeat interactions: to TPR or not to TPR?, TrendsŽ .Biochem. Sci. 20 1995 257–259.

w x53 S.D. Nuttal, B.J. Hanson, M. Mori, N.J. Hoogenraad,hTom34: a novel translocase for the import of proteins into

Ž .human mitochondria, DNA Cell Biol. 1997 1067–1074.w x54 M.F. Bauer, C. Sirrenberg, W. Neupert, M. Brunner, Role

of Tim23 as voltage sensor and presequence receptor inŽ .protein import into mitochondria, Cell 87 1996 33–41.

w x55 U. Bomer, J. Rassow, N. Zufall, N. Pfanner, M. Meijer,¨A.C. Maarse, The preprotein translocase of the inner mito-


chondrial membrane: evolutionary conservation of target-Ž .ing and assembly of Tim17, J. Mol. Biol. 262 1996

389–395.w x56 M. Mori, S. Miura, M. Tatibana, P.P. Cohen, Characteriza-

tion of a protease apparently involved in processing ofpre-ornithine transcarbamylase of rat liver, Proc. Natl.

Ž .Acad. Sci. U.S.A. 77 1980 7044–7048.w x57 W.J. Ou, A. Ito, H. Okazaki, T. Omura, Purification and

characterization of a processing protease from rat liverŽ .mitochondria, EMBO J. 8 1989 2605–2612.

w x58 J. Kleiber, F. Kalousek, M. Swaroop, L.E. Rosenberg, Thegeneral mitochondrial matrix processing protease from ratliver: structural characterization of the catalytic subunit,

Ž .Proc. Natl. Acad. Sci. U.S.A. 87 1990 7978–7982.w x59 S. Kitada, T. Niidome, T. Nagano, T. Ogishima, A. Ito,

Ž .Molecular cloning of the smaller subunit P52 of rat livermitochondrial processing protease, Biochem. Biophys. Res.

Ž .Commun. 190 1993 289–293.w x60 V. Paces, L.E. Rosenberg, W.A. Fenton, F. Kalousek, The

beta subunit of the mitochondrial processing peptidasefrom rat liver: cloning and sequencing of a cDNA andcomparison with a proposed family of metallopeptidases,

Ž .Proc. Natl. Acad. Sci. U.S.A. 90 1993 5355–5358.w x61 T. Niidome, S. Kitada, K. Shimokata, T. Ogishima, A. Ito,

Arginine residues in the extension peptide are required forcleavage of a precursor by mitochondrial processing pepti-dase. Demonstration using synthetic peptide as a substrate,

Ž .J. Biol. Chem. 269 1994 24719–24722.w x62 T. Ogishima, T. Niidome, K. Shimokata, S. Kitada, A. Ito,

Analysis of elements in the substrate required for process-ing by mitochondrial processing peptidase, J. Biol. Chem.

Ž .270 1995 30322–30326.w x63 W.J. Ou, T. Kumamoto, K. Mihara, S. Kitada, T. Niidome,

A. Ito, T. Omura, Structural requirement for recognition ofthe precursor proteins by the mitochondrial processing

Ž .peptidase, J. Biol. Chem. 269 1994 24673–24678.w x64 F. Kalousek, J.P. Hendrick, L.E. Rosenberg, Two mito-

chondrial matrix proteases act sequentially in the process-ing of mammalian matrix enzymes, Proc. Natl. Acad. Sci.

Ž .U.S.A. 85 1988 7536–7540.w x65 G. Isaya, F. Kalousek, W.A. Fenton, L.E. Rosenberg,

Cleavage of precursors by the mitochondrial processingpeptidase requires a compatible mature protein or an inter-

Ž .mediate octapeptide, J. Cell Biol. 113 1991 65–76.w x66 K. Thornton, Y. Wang, H. Weiner, D.G. Gorenstein, Im-

port, processing, and two-dimensional NMR structure of alinker-deleted signal peptide of rat liver mitochondrial

Ž .aldehyde dehydrogenase, J. Biol. Chem. 268 199319906–19914.

w x67 M. Waltner, H. Weiner, Conversion of a nonprocessedmitochondrial precursor protein into one that is processedby the mitochondrial processing peptidase, J. Biol. Chem.

Ž .270 1995 26311–26317.w x68 F. Kalousek, G. Isaya, L.E. Rosenberg, Rat liver mitochon-

Ž .drial intermediate peptidase MIP : purification and initialŽ .characterization, EMBO J. 11 1992 2803–2809.

w x69 G. Isaya, F. Kalousek, L.E. Rosenberg, Sequence analysisof rat mitochondrial intermediate peptidase: similarity tozinc metallopeptidases and to a putative yeast homologue,

Ž .Proc. Natl. Acad. Sci. U.S.A. 89 1992 8317–8321.w x70 G. Isaya, F. Kalousek, L.E. Rosenberg, Amino-terminal

octapeptides function as recognition signals for the mito-Ž .chondrial intermediate peptidase, J. Biol. Chem. 267 1992

7904–7910.w x71 T. Leustek, B. Dalie, S.D. Amir, N. Brot, H. Weissbach, A

member of the Hsp70 family is localized in mitochondriaand resembles Escherichia coli DnaK, Proc. Natl. Acad.

Ž .Sci. U.S.A. 86 1989 7805–7808.w x72 L.A. Mizzen, C. Chang, J.I. Garrels, W.J. Welch, Identifi-

cation, characterization, and purification of two mam-malian stress proteins present in mitochondria, grp 75, amember of the hsp 70 family and hsp 58, a homolog of the

Ž .bacterial groEL protein, J. Biol. Chem. 264 1989 20664–20675.

w x73 T.J. Webster, D.J. Naylor, D.J. Hartman, P.B. Høj, N.J.Hoogenraad, cDNA cloning and efficient mitochondrialimport of pre-mtHSP70 from rat liver, DNA Cell Biol. 13Ž .1994 1213–1220.

w x74 T. Saijo, W.J. Welch, K. Tanaka, Intramitochondrial fold-ing and assembly of medium-chain acyl-CoA dehydroge-

Ž .nase MCAD . Demonstration of impaired transfer ofK304E-variant MCAD from its complex with hsp60 to the

Ž .native tetramer, J. Biol. Chem. 269 1994 4401–4408.w x75 D.J. Naylor, N.J. Hoogenraad, P.B. Høj, Isolation and

characterisation of a cDNA encoding rat mitochondrialGrpE, a stress-inducible nucleotide-exchange factor ofubiquitous appearance in mammalian organs, FEBS Lett.

Ž .396 1996 181–188.w x76 S. Jindal, A.K. Dudani, B. Singh, C.B. Harley, R.S. Gupta,

Primary structure of a human mitochondrial protein homol-ogous to the bacterial and plant chaperonins and to the65-kilodalton mycobacterial antigen, Mol. Cell. Biol. 9Ž .1989 2279–2283.

w x77 D.J. Picketts, C.S. Mayanil, R.S. Gupta, Molecular cloningof a Chinese hamster mitochondrial protein related to the‘chaperonin’ family of bacterial and plant proteins, J. Biol.

Ž .Chem. 264 1989 12001–12008.w x78 P.V. Viitanen, G.H. Lorimer, R. Seetharam, R.S. Gupta, J.

Oppenheim, J.O. Thomas, N.J. Cowan, Mammalian mito-chondrial chaperonin 60 functions as a single toroidal ring,

Ž .J. Biol. Chem. 267 1992 695–698.w x79 M.T. Ryan, N.J. Hoogenraad, P.B. Høj, Isolation of a

cDNA clone specifying rat chaperonin 10, a stress-induci-ble mitochondrial matrix protein synthesised without a

Ž .cleavable presequence, FEBS Lett. 337 1994 152–156.w x80 D.J. Hartman, N.J. Hoogenraad, R. Condron, P.B. Høj,

Identification of a mammalian 10-kDa heat shock protein,a mitochondrial chaperonin 10 homologue essential forassisted folding of trimeric ornithine transcarbamoylase in

Ž .vitro, Proc. Natl. Acad. Sci. U.S.A. 89 1992 3394–3398.w x81 T.H. Lubben, A.A. Gatenby, G.K. Donaldson, G.H.

Lorimer, P.V. Viitanen, Identification of a groES-like


chaperonin in mitochondria that facilitates protein folding,Ž .Proc. Natl. Acad. Sci. U.S.A. 87 1990 7683–7687.

w x82 L. Anderson, Identification of mitochondrial proteins andsome of their precursors in two-dimensional elec-trophoretic maps of human cells, Proc. Natl. Acad. Sci.

Ž .U.S.A. 78 1981 2407–2411.w x83 T. Morita, M. Mori, F. Ikeda, M. Tatibana, Transport of

carbamyl phosphate synthetase I and ornithine transcar-bamylase into mitochondria. Inhibition by rhodamine 123and accumulation of enzyme precursors in isolated hepato-

Ž .cytes, J. Biol. Chem. 257 1982 10547–10550.w x84 R. Jaussi, P. Sonderegger, J. Fluckiger, P. Christen,

Biosynthesis and topogenesis of aspartate aminotransferaseisoenzymes in chicken embryo fibroblasts. The precursorof the mitochondrial isoenzyme is either imported intomitochondria or degraded in the cytosol, J. Biol. Chem.

Ž .257 1982 13334–13340.w x85 J. Kolarov, B.D. Nelson, Import and processing of cy-

tochrome b – c1 complex subunits in isolated hepatomaascites cells. Inhibition by Rhodamine 6G, Eur. J. Biochem.

Ž .144 1984 387–392.w x86 M. Mori, T. Morita, F. Ikeda, Y. Amaya, M. Tatibana, P.P.

Cohen, Synthesis, intracellular transport, and processing ofthe precursors for mitochondrial ornithine transcarbamy-lase and carbamoyl-phosphate synthetase I in isolated hep-

Ž .atocytes, Proc. Natl. Acad. Sci. U.S.A. 78 1981 6056–6060.

w x87 A.L. Horwich, W.A. Fenton, F.A. Firgaira, J.E. Fox, D.Kolansky, I.S. Mellman, L.E. Rosenberg, Expression ofamplified DNA sequences for ornithine transcarbamylasein HeLa cells: arginine residues may be required for mito-chondrial import of enzyme precursor, J. Cell Biol. 100Ž .1985 1515–1521.

w x88 G. Isaya, W.A. Fenton, J.P. Hendrick, K. Furtak, F.Kalousek, L.E. Rosenberg, Mitochondrial import and pro-cessing of mutant human ornithine transcarbamylase pre-

Ž .cursors in cultured cells, Mol. Cell. Biol. 8 1988 5150–5158.

w x89 K. Yamauchi, N. Hayashi, G. Kikuchi, Translocation ofdelta-aminolevulinate synthase from the cytosol to themitochondria and its regulation by hemin in the rat liver, J.

Ž .Biol. Chem. 255 1980 1746–1751.w x90 J.T. Lathrop, M.P. Timko, Regulation by heme of mito-

chondrial protein transport through a conserved amino acidŽ .motif, Science 259 1993 522–525.

w x91 M. Takahashi, D.A. Hood, Protein import into subsar-colemmal and intermyofibrillar skeletal muscle mitochon-

dria. Differential import regulation in distinct subcellularŽ .regions, J. Biol. Chem. 271 1996 27285–27291.

w x92 G. Wright, V. Reichenbecker, T. Green, G.L. Wright, S.Wang, Paraquat inhibits the processing of human man-ganese-dependent superoxide dismutase by SF-9 insect cell

Ž .mitochondria, Exp. Cell Res. 234 1997 78–84.w x93 R. Rizzuto, M. Brini, P. Pizzo, M. Murgia, T. Pozzan,

Chimeric green fluorescent protein as a tool for visualizingŽ .subcellular organelles in living cells, Curr. Biol. 5 1995

635–642.w x94 R. Rizzuto, M. Brini, G.F. De, R. Rossi, R. Heim, R.Y.

Tsien, T. Pozzan, Double labelling of subcellular structureswith organelle-targeted GFP mutants in vivo, Curr. Biol. 6Ž .1996 183–188.

w x95 M. Grompe, D.M. Muzny, C.T. Caskey, Scanning detec-tion of mutations in human ornithine transcarbamoylase bychemical mismatch cleavage, Proc. Natl. Acad. Sci. U.S.A.

Ž .86 1989 5888–5892.w x96 F.D. Ledley, R. Jansen, S.U. Nham, W.A. Fenton, L.E.

Rosenberg, Mutation eliminating mitochondrial leader se-quence of methylmalonyl-CoA mutase causes muto meth-ylmalonic acidemia, Proc. Natl. Acad. Sci. U.S.A. 87Ž .1990 3147–3150.

w x97 F. Takakubo, P. Cartwright, N.J. Hoogenraad, D.R. Thor-burn, F. Collins, T. Lithgow, H.H. Dahl, An amino acidsubstitution in the pyruvate dehydrogenase E1 alpha gene,affecting mitochondrial import of the precursor protein,

Ž .Am. J. Hum. Genet. 57 1995 772–780.w x98 G. Inana, C. Chambers, Y. Hotta, L. Inouye, D. Filpula, S.

Pulford, T. Shiono, Point mutation affecting processing ofthe ornithine aminotransferase precursor protein in gyrate

Ž .atrophy, J. Biol. Chem. 264 1989 17432–17436.w x99 T. Kobayashi, H. Ogawa, M. Kasahara, Z. Shiozawa, T.

Matsuzawa, A single amino acid substitution within themature sequence of ornithine aminotransferase obstructsmitochondrial entry of the precursor, Am. J. Hum. Genet.

Ž .57 1995 284–291.w x100 P.E. Purdue, J. Allsop, G. Isaya, L.E. Rosenberg, C.J.

Danpure, Mistargeting of peroxisomal L-alanine:glyoxylateaminotransferase to mitochondria in primary hyperoxaluriapatients depends upon activation of a cryptic mitochondrialtargeting sequence by a point mutation, Proc. Natl. Acad.

Ž .Sci. U.S.A. 88 1991 10900–10904.w x101 A. Huckriede, E. Agsteribbe, Decreased synthesis and

inefficient mitochondrial import of hsp60 in a patient witha mitochondrial encephalomyopathy, Biochim. Biophys.

Ž .Acta 1227 1994 200–206.

Documents

Review Mitochondrial protein import in animalsmitochondrial proteins are initially synthesized on free ribosomes as larger precursors with NH -termi-2 nal presequences which function