|
|
|
|
Vol. 10, Issue 7, 2461-2474, July 1999

and
*Institut für Biochemie und Molekularbiologie and
Institut für Biologie, Universität
Freiburg, D-79104 Freiburg, Germany
| |
ABSTRACT |
|---|
|
|
|---|
Two major routes of preprotein targeting into mitochondria are known. Preproteins carrying amino-terminal signals mainly use Tom20, the general import pore (GIP) complex and the Tim23-Tim17 complex. Preproteins with internal signals such as inner membrane carriers use Tom70, the GIP complex, and the special Tim pathway, involving small Tims of the intermembrane space and Tim22-Tim54 of the inner membrane. Little is known about the biogenesis and assembly of the Tim proteins of this carrier pathway. We report that import of the preprotein of Tim22 requires Tom20, although it uses the carrier Tim route. In contrast, the preprotein of Tim54 mainly uses Tom70, yet it follows the Tim23-Tim17 pathway. The positively charged amino-terminal region of Tim54 is required for membrane translocation but not for targeting to Tom70. In addition, we identify two novel homologues of the small Tim proteins and show that targeting of the small Tims follows a third new route where surface receptors are dispensable, yet Tom5 of the GIP complex is crucial. We conclude that the biogenesis of Tim proteins of the carrier pathway cannot be described by either one of the two major import routes, but involves new types of import pathways composed of various features of the hitherto known routes, including crossing over at the level of the GIP.
| |
INTRODUCTION |
|---|
|
|
|---|
Many mitochondrial precursor proteins are synthesized with
amino-terminal targeting sequences, termed presequences, that direct the proteins to the organelle and across the outer and inner membranes (Ryan and Jensen, 1995
; Schatz and Dobberstein, 1996
; Neupert, 1997
;
Pfanner et al., 1997
; Ryan and Pfanner, 1998
). In the
matrix, the presequences are typically cleaved off by the mitochondrial processing peptidase. A number of mitochondrial preproteins are not
synthesized with cleavable targeting signals. A few preproteins were
shown to contain the targeting information at the amino-terminal portion of the protein that is to carry a "noncleaved presequence" (Hurt et al., 1985
; Arakawa et al., 1990
; Rospert
et al., 1993
; Hahne et al., 1994
; Jarvis et
al., 1995
); however, most noncleavable preproteins, notably those
that are membrane proteins, seem to contain internal targeting
information distributed over various regions of the preprotein (Pfanner
et al., 1987a
; Smagula and Douglas, 1988
; Davis et
al., 1998
; Káldi et al., 1998
). Typical representatives are the members of the large family of inner membrane metabolite carriers, such as the ADP/ATP carrier (AAC).
The mitochondrial machinery for the import of presequence-containing
preproteins has been studied in detail. The presequences are typically
recognized by the surface receptor Tom20, the 20-kDa subunit of the
translocase of the outer membrane (Söllner et al.,
1989
; Ramage et al., 1993
; Brix et al., 1997
).
Subsequently, the preproteins are transferred to the general import
pore (GIP) complex where they interact with Tom22 and Tom5 and are
translocated through the import channel formed by Tom40 (Vestweber
et al., 1989
; Kiebler et al., 1993
; Dietmeier
et al., 1997
; Hill et al., 1998
). The
presequences bind to Tim23 of the translocase of the inner membrane and
in a membrane potential (
)-dependent reaction move through the
inner membrane channel, that is, the Tim core complex formed by Tim23
and Tim17 (Dekker et al., 1993
, 1997
; Emtage and Jensen,
1993
; Ryan et al., 1994
; Bauer et al., 1996
). Matrix-located heat shock protein 70 (mtHsp70) cooperates with Tim44 to
drive the completion of preprotein translocation (Kronidou et
al., 1994
; Rassow et al., 1994
; Schneider et
al., 1994
; Voos et al., 1996
; Bömer et
al., 1998
).
The inner membrane carriers follow a different import route that
converges with the presequence pathway only at the level of the GIP
complex. The carrier preproteins such as AAC are preferentially recognized by Tom70 before their transfer to the GIP (Hines et al., 1990
; Söllner et al., 1990
). At the
trans side of the outer membrane, the presequence and
carrier pathways diverge (Moczko et al., 1997
; Kübrich
et al., 1998
). Recent studies led to the identification of a
number of new Tim proteins that mediate the translocation through the
intermembrane space and insertion into the inner membrane. Three
homologous small Tim proteins of the intermembrane space, Tim9, Tim10,
and Tim12, bind the carrier preproteins (Koehler et al.,
1998a
,b
; Sirrenberg et al., 1998
) and transfer them to a
translocase of the inner membrane that contains Tim22 and Tim54
(Sirrenberg et al., 1996
; Kerscher et al., 1997
).
The carrier proteins contain six membrane-spanning segments each and
are inserted into the inner membrane in a 
-dependent manner via
Tim22-Tim54. Tim23, Tim17, and Tim22 contain a homologous membrane
domain with four predicted membrane-spanning segments, and recent
evidence indicates that Tim23 and most likely Tim17 and Tim22 are also
imported via the carrier Tim pathway (Kerscher et al., 1997
;
Káldi et al., 1998
).
Little is known about the actual biogenesis of the Tim proteins that are involved in the import of carrier preproteins. For this study, we have analyzed targeting and translocation of the precursors of these Tim proteins. We report that the targeting pathways of Tim22 and Tim54 reveal a new principle of combination of different portions of the main (presequence) pathway and the special (carrier) pathway. The crossing over occurs at the level of the GIP complex. Moreover, import of the small Tims provides the first example for a preferential targeting via Tom5 and not via the trypsin-accessible domains of the larger receptors.
| |
MATERIALS AND METHODS |
|---|
|
|
|---|
Construction of Plasmids for in Vitro Transcription
The open reading frames of yeast TIM54
(Kerscher et al., 1997
), TIM22 (Sirrenberg
et al., 1996
), TIM13 (Accession No. P53299), TIM12 (Jarosch et al., 1996
), TIM10
(Jarosch et al., 1997
), TIM9 (Koehler et
al., 1998b
), and TIM8 (Accession No. Y13136) were amplified by PCR and individually cloned into pGEM-4Z (Promega, Madison, WI). Tim54
N was obtained by PCR using a downstream vector primer and a primer containing the SP6 polymerase binding site and an
18-nucleotide stretch encoding residues 39-44 of Tim54 (5'-GGA TTA GGT
GAC ACT ATA GAA ATG ATC TTT TGG TCT GTG-3').
Import of Preproteins Into Isolated Mitochondria
The yeast strains used in this study are shown in Table
1. Mitochondria were isolated from yeast
cells grown in YPG media (1% yeast extract, 2% bacto-peptone, and 3%
glycerol) according to Daum et al. (1982)
and Hartl et
al. (1987)
. Radiolabeled preproteins were obtained by in vitro
transcription and translation reactions using rabbit reticulocyte
lysate (Amersham, Arlington Heights, IL) in the presence of
[35S]methionine/cysteine (Söllner et
al., 1991
).
|
Mitochondrial in vitro import reactions were performed in BSA-containing buffer (3% [wt/vol] fatty acid-free BSA, 80 mM KCl, 5 mM MgCl2, 10 mM MOPS/KOH, pH 7.2) in the presence of 2 mM ATP and 2 mM NADH. To dissipate the membrane potential, 8 µM antimycin A, 20 µM oligomycin, and 1 µM valinomycin (Sigma, St. Louis, MO) were added to the import reaction. Reticulocyte lysate containing radiolabeled preproteins (2.5-10% [vol/vol] of import reaction) was incubated with mitochondria (25-50 µg protein) at 25°C for varying times. Valinomycin (1 µM) was added to stop import, and samples were subsequently treated with or without proteinase K (50 µg/ml) on ice for 15 min. The protease was inactivated by the addition of 1 mM PMSF, and samples were incubated for a further 10 min at 4°C.
For trypsin treatment of accessible Tom receptor domains, mitochondrial samples in SEM buffer (250 mM sucrose, 1 mM EDTA, 10 mM MOPS/KOH, pH 7.2) were incubated with trypsin (20 µg/ml) for 20 min on ice. Trypsin was inactivated on the addition of a 30-fold excess of soybean trypsin inhibitor (Type II-S, Sigma) and samples were incubated for an additional 10 min on ice before further manipulations. For control samples, a 30-fold excess of soybean trypsin inhibitor was added to mitochondria before trypsin addition. Preproteins were imported for 15 min at 25°C before proteinase K digestion.
For import of preproteins into ssc1-3 mitochondria (Gambill
et al., 1993
), a 15 min incubation at 37°C was performed
with both wild-type and ssc1-3 mitochondria before import
studies were performed at 25°C.
Swelling of mitochondrial samples was prepared by resuspending the
mitochondrial pellets in EM buffer (1 mM EDTA, 10 mM MOPS/KOH, pH 7.2)
and incubating the samples on ice for 15 min. ATP was depleted from
mitochondrial samples and reticulocyte lysates according to Glick
(1995)
. Preproteins were imported for 15 min at 25°C before
proteinase K digestion.
After treatments, mitochondrial pellets were lysed in the appropriate detergent-containing buffer and applied to SDS or blue native polyacrylamide gels.
Accumulation of b2(167)
-dihydrofolate Reductase
Across Mitochondrial Membranes
Mitochondria (50 µg) were incubated with 1 µg purified
b2(167)
-dihydrofolate reductase (DHFR) (Dekker et
al., 1997
) for 15 min at 25°C in the presence of 2 µM
methotrexate. After accumulation, mitochondria were reisolated, washed
with SEM buffer containing 2 µM methotrexate, and finally resuspended
in BSA-containing buffer containing 2 µM methotrexate before use in
further import assays.
Blue Native Gel Electrophoresis
Blue native PAGE was performed essentially as described
previously (Schägger and von Jagow, 1991
; Schägger et
al., 1994
; Dekker et al., 1997
). Briefly, mitochondrial
pellets (25-100 µg protein) were lysed in 50 µl ice-cold digitonin
buffer (1% [wt/vol] digitonin, 20 mM Tris-HCl, pH 7.4, 0.1 mM EDTA,
50 mM NaCl, 10% [vol/vol] glycerol, 1 mM PMSF) (Blom et
al., 1995
). After a clarifying spin, 5 µl of sample buffer (5%
[wt/vol] Coomassie brilliant blue G-250, 100 mM Bis-Tris, pH 7.0, 500 mM 6-aminocaproic acid) were added, and the samples were
electrophoresed at 4°C through a 6-16% polyacrylamide gradient gel.
For immunoblotting, the native gel was soaked in blot
buffer (20 mM Tris-base, 150 mM glycine, 20% [vol/vol] methanol,
0.08% [wt/vol] SDS) before transfer onto PVDF membranes (Millipore, Bedford, MA) using the semidry blotting technique (Harlow and Lane,
1988
). Immunodecoration was performed according to standard procedures
(Harlow and Lane, 1988
), and detection was achieved using the ECL
method (Amersham). For detection of radiolabeled proteins, the dried
gel or PVDF membrane was exposed to phosphorimage storage cassettes
before phosphorimage analysis (Molecular Dynamics, Sunnyvale, CA).
Miscellaneous
Sequence alignments were generated with MegAlign (DNA Star Inc., Madison, WI) using the Clustal method and the PAM250 weight table.
SDS-PAGE of larger proteins (e.g., Tim54 and Tim22) was performed
according to Laemmli (1970)
, and urea SDS-PAGE (Ito et al., 1980
) was used for the analysis of the small Tim proteins.
| |
RESULTS |
|---|
|
|
|---|
The Preproteins of Tim22 and Tim54 Require Different Surface Receptors for Import
The preproteins of Tim22 and Tim54 were synthesized in vitro
in rabbit reticulocyte lysate in the presence of
[35S]methionine/cysteine and incubated with isolated
yeast wild-type mitochondria. In the presence of a membrane potential
across the inner membrane, the preproteins were transported to a
protease-protected location (Figure 1A,
lanes 1-3 and 5-7). On dissipation of the 
, import was blocked
(Figure 1A, lanes 4 and 8).
|
To determine whether the proteins were correctly imported and
assembled, we used blue native electrophoresis of digitonin-lysed mitochondria, which allows a separation of the mitochondrial
translocase complexes (Dekker et al., 1996
-1998
): the large
Tom complex, termed the GIP complex (Figure 1B, lane 1); the Tim core
complex containing Tim23 and Tim17 (Figure 1B, lane 2); and a
~300-kDa complex containing Tim22, termed the carrier translocase
(Figure 1B, lane 3). We found that both imported Tim22 and Tim54 were
efficiently assembled into the 300-kDa translocase complex in a

-dependent manner (Figure 1B, lanes 4-6 and 8-10).
By a pretreatment of the mitochondria with trypsin, the cytosolic
domains of the import receptors Tom20, Tom22, and Tom70 are removed
(Alconada et al., 1995
; Dietmeier et al., 1997
).
The import of both Tim22 and Tim54 into trypsin-treated mitochondria was inhibited, indicating a dependence on one or more of these surface
receptors (Figure 1C, lane 2). Such pretreatment also inhibited the
import of the outer membrane protein porin and a matrix-targeted fusion
protein between the presequence of Fo-ATPase subunit 9 and
the entire dihydrofolate reductase (Su9-DHFR), which mainly uses
Tom20 (Hines et al., 1990
; Moczko et al., 1994
;
Pfanner et al., 1997
). The import of AAC, which mainly uses
Tom70 as its receptor (Hines et al., 1990
; Söllner
et al., 1990
), was also inhibited (Figure 1C). It has been
shown previously that there is a differential dependence on cytosolic
ATP and cofactors for the targeting of preproteins to either Tom20 or
Tom70 (Hachiya et al., 1995
; Komiya et al., 1996
,
1997
). We asked how the depletion of cytosolic ATP affected the import
of Tim22 and Tim54 and found a strong difference between both
preproteins. Although the import of Tim54 was inhibited by the ATP
depletion, the import of Tim22 was unchanged (Figure 1D, lane 2). As a
control, we show that depletion of cytosolic ATP also inhibited the
import of the preprotein AAC but not Su9-DHFR (Figure 1D) (Wachter
et al., 1994
).
We thus examined the possibility that Tim22 and Tim54 interacted with
different receptors by using mitochondria isolated from yeast strains
lacking TOM20 or TOM70 genes, respectively.
Because a lack of Tom20 causes a reduction in the mitochondrial levels of Tom22 and thus indirectly a reduction of the mitochondrial membrane
potential (Lithgow et al., 1994
; Gärtner et
al., 1995b
; Hönlinger et al., 1995b
), we used a
tom20
strain where TOM22 was put on a
high-copy number plasmid to restore the mitochondrial levels of Tom22
and the membrane potential (Hönlinger et al., 1995b
).
Import of the preprotein of Tim22 was strongly inhibited in
tom20
mitochondria but practically unchanged in
tom70
mitochondria in comparison to wild-type
mitochondria (Figure 2, left panel). In
contrast, the import of Tim54 was strongly inhibited in
tom70
mitochondria, but only mildly affected in
tom20
mitochondria (Figure 2, middle panel). Furthermore,
the presequence-containing preprotein Su9-DHFR displayed import
characteristics similar to those of Tim22 in its receptor requirements
(Figure 2, right panel). In the absence of a membrane potential across
the inner membrane, import of the preproteins was inhibited in all
cases (Figure 2, lanes 5, 10, and 15), confirming the specificity of
the import processes into the mutant mitochondria.
|
In addition, we used mitochondria from a yeast strain lacking the small
subunit Tom5 of the GIP complex. Tom5 is resistant to a treatment with
trypsin and functions after the surface receptors at the entry site of
the import pore where the presequence and carrier routes converge
(Dietmeier et al., 1997
). The import of both Tim22 and Tim54
along with Su9-DHFR were inhibited in tom5
mitochondria
(Figure 2, bottom panels). We conclude that the preprotein of Tim54 is
directed into mitochondria preferentially by Tom70 in an ATP-dependent
reaction, whereas the preprotein of Tim22, like the
presequence-containing preprotein Su9-DHFR, uses Tom20 as receptor in a
manner independent of cytosolic ATP. Both import pathways join at the
entry site of the GIP that includes Tom5.
Different Tim Pathways for the Preproteins of Tim22 and Tim54
Koehler et al. (1998a)
and Sirrenberg et al.
(1998)
showed that mutations in Tim components of the carrier pathway,
including Tim10, caused a strong reduction in the mitochondrial level
of Tim22, suggesting that the preprotein of Tim22 itself was imported via the carrier Tim pathway, as shown previously for the homologous proteins Tim23 and Tim17 (Dekker et al., 1997
; Kerscher
et al., 1997
; Káldi et al., 1998
). We used
two assays to determine which Tim pathway was used by the preprotein of
Tim54 in comparison with the import of Tim22.
Accumulation of Chemical Amounts of a Presequence-Containing
Preprotein in Translocation Contact Sites
A matrix-targeted preprotein, consisting of a portion of
precytochrome b2 and the entire dihydrofolate
reductase [b2(167)
-DHFR], can be prepared as a soluble
species in large amounts. After stabilization of the DHFR moiety with
methotrexate, the preprotein can be accumulated in the mitochondrial
import sites, spanning both the Tom machinery and the Tim23-Tim17
machinery (Dekker et al., 1997
). Thereby the subsequent
import of preproteins using the Tim23-Tim17 pathway is impaired in
mitochondria with accumulated b2(167)
-DHFR. The import
of carrier proteins is only slightly affected because the Tom complexes
are approximately four times more abundant than the Tim23-Tim17
complexes, and thus most Tom complexes are not occupied by the
b2(167)
-DHFR preprotein (Dekker et al.,
1997
). We accumulated methotrexate-bound b2(167)
-DHFR in
mitochondria and tested the import kinetics of different preproteins.
Although the import of Tim22 was only slightly inhibited (Figure
3A), a significant reduction in import
was observed for the preprotein of Tim54 (Figure 3B). For comparison,
the import of AAC was only slightly affected (Figure 3C), but the
import of the Rieske Fe/S protein, a typical presequence-containing
preprotein, was inhibited (Figure 3D). Furthermore, the analysis of
inner-membrane insertion and assembly of the imported preproteins by
blue native electrophoresis revealed that Tim54 assembly was reduced in
mitochondria containing accumulated preprotein (Figure 3E).
|
A Point Mutation in TIM23
Mitochondria from the yeast mutant tim23-2 carry
an amino acid substitution in Tim23 that causes a labilization of the
Tim23-Tim17 complex and thus a reduction in preproteins imported via
Tim23-Tim17 (Bömer et al., 1997a
; Dekker et
al., 1997
). The import of Tim22 into tim23-2
mitochondria occurred with wild-type rates, like the import of the
ADP/ATP carrier (Figure 4A, first and
third panels). The import of Tim54 into tim23-2
mitochondria was partly reduced, comparable to the import of the Fe/S
protein (Figure 4A, compare second and fourth panels, lanes 5-7 with
lanes 1-3; Figure 4B, compare columns 2 and 4 [showing the
degree of inhibition]).
|
and Fe/S protein was
severely inhibited (Figure 4C, bottom panels). The import of Tim22,
which is imported via the carrier Tim route, was not affected by the
ssc1-3 mutation (Figure 4C, top left panel), as observed
with the precursors of Tim17 and Tim23 (Bömer et al.,
1997b
N, which is synthesized from
the methionine at residue 39 (Figure 5A). The import of Tim54
N into
a protease-protected location of mitochondria was completely blocked
(Figure 5B, bottom panel, lanes 5-7). The binding of the mutant
preprotein to mitochondria, however, was still possible (Figure 5B, top
panel, lanes 5-8). The binding of preproteins to isolated mitochondria
typically includes two fractions: a specific binding to receptors and a nonspecific binding to the organelle surface (Pfanner et
al., 1987b
N bound to
mitochondria included specific binding to receptors, we compared its
association with both wild-type mitochondria and mitochondria lacking
Tom20 or Tom70. The binding of Tim54
N to mitochondria was reduced by
~60% tom70
mitochondria (Figure 5C, column 6),
comparable to the binding of full-length Tim54 (Figure 5C, column 3),
whereas for tom20
mitochondria, an increase in the
binding of Tim54
N was observed (Figure 5C, column 2 vs. 5). We
conclude that a significant fraction of Tim54
N bound to mitochondria
specifically interacts via Tom70, indicating that the positively
charged amino-terminal segment is dispensable for this receptor
interaction. The amino-terminal segment, however, is required for
translocation of Tim54 into mitochondria, suggesting that both the
amino-terminal portion of Tim54 and a more C-terminal region are
involved in the import of this preprotein.
|
A Role for Tom5 in Targeting of the Small Tims to Mitochondria
Tim9, Tim10, and Tim12 are the three currently known small Tim
proteins that are homologous to each other. A search in the yeast
genome revealed the presence of two additional small Tim proteins,
termed Tim8 and Tim13, with significant homology to the other three
Tims, including a complete conservation of four cysteine residues
(Figure 6A). The preproteins of the small
Tims were synthesized in reticulocyte lysates and labeled with
[35S]methionine/cysteine. Like the known small Tims, Tim8
and Tim13 were transported to a protease-protected location in
mitochondria (Figure 6B, lane 1). After swelling of the mitochondria
that led to an opening of the intermembrane space with an efficiency of ~80-90% (Bömer et al., 1997a
,b
), each small Tim
became accessible to protease (Figure 6B, lane 2), demonstrating that
Tim8 and Tim13 also were located in the intermembrane space. As
control, immunodecoration showed that endogenous Tim10 along with the
intermembrane space protein cytochrome b2 were
accessible to protease after swelling but not the matrix-located
protein Mge1 (Figure 6B, Immunodecoration).
|
We then investigated the targeting principles of the small Tims. No
obvious mitochondrial targeting signals are observed in any of the
small Tim sequences (Figure 6A), and our unpublished results along with
those of Koehler et al. (1998b)
showed that the small Tim
proteins do not require a membrane potential for their insertion into
the intermembrane space. Surprisingly, a pretreatment of the
mitochondria with trypsin did not inhibit the import of the small Tims,
as shown here with the preproteins of Tim9, Tim10, and Tim13 (Figure
6C, columns 2, 4, and 6), indicating that they did not strictly require
mitochondrial surface receptors. Indeed, import of the small Tims was
not affected by a deletion of TOM70 (Figure 6D,
tom70
) and was only slightly impaired, in the case of
Tim9, by a deletion of TOM20 (Figure 6D,
tom20
). We wondered whether the small Tim proteins showed
a requirement for Tom5 as receptor because Tom5 is resistant to trypsin
treatment and can suffice as a receptor under bypass import conditions
(Dietmeier et al., 1997
). Indeed, the import of the small
Tims was strongly inhibited in tom5
mitochondria (Figure
6D, tom5
). Previous work showed that the
tom5
mitochondria are selectively deficient in Tom5,
whereas the other Tom proteins are present in normal amounts and the
GIP complex is not altered (Dietmeier et al., 1997
; Dekker et al., 1998
). Although the import of preproteins in
tom5
mitochondria is generally reduced compared with
wild-type mitochondria (Dietmeier et al., 1997
), the small
Tim proteins showed an even stronger reliance on Tom5 for their import
compared with Tim22 and Tim54 (Figure 2). We conclude that Tom5, but
not the trypsin-accessible surface receptors, plays an important role
in the targeting of the small Tims into mitochondria.
| |
DISCUSSION |
|---|
|
|
|---|
The biogenesis of the Tim proteins of the carrier import route
neither follows one of the known major import pathways for mitochondrial preproteins (Figure 7A,
routes I and II) nor fits into a common new mechanism. Three different
targeting principles seem to be necessary to import the Tim components
of the carrier route (Figure 7B).
|
The preprotein of Tim22 is the first preprotein found that preferentially uses the receptor Tom20 for targeting to the outer membrane but follows the Tim route for carrier proteins (Figure 7B, route III). Tim22 is a quite hydrophobic protein and therefore would have been a typical candidate for binding to Tom70 like the carrier preproteins, but import signals in Tim22 that are not yet defined direct the preprotein to Tom20, which usually functions as a receptor for presequence-containing preproteins.
In contrast, Tim54 carries an amino-terminal noncleaved translocation
sequence that is positively charged like mitochondrial presequences,
yet Tim54 preferentially depends on Tom70 for its targeting to the
outer membrane (Figure 7B, route IV). After translocation through the
GIP, Tim54 then follows the typical route for presequence-containing preproteins until it reaches the Tim23-Tim17 core complex. The amino-terminal positively charged sequence of Tim54 is required for
translocation of the preprotein into the mitochondria, whereas interaction with the outer membrane can occur without this sequence. This indicates that the remainder of Tim54, which includes two predicted hydrophobic segments (Kerscher et al., 1997
),
contains a signal for its interaction with Tom70. Although carrying an abundance of positive residues, the amino-terminal sequence of Tim54 is
not predicted to form an amphipathic
-helix because of the presence
of several helix-breaking prolyl residues, distinguishing this sequence
from typical mitochondrial presequences (Roise et al., 1986
;
Von Heijne, 1986
) and providing an explanation for the only weak
dependence on Tom20. Tim54 branches from the main import route at the
level of Tim23 before a strict requirement for matrix Hsp70 becomes
crucial. It has been observed that a hydrophobic signal anchor
following the positively charged amino-terminal region of a preprotein
minimizes its requirement for mtHsp70. This is apparently due to the
sorting/membrane insertion activity of the hydrophobic segment at an
early stage of translocation of an unfolded preprotein (Glick et
al., 1993
; Voos et al., 1993
; Stuart et al.,
1994
; Gärtner et al., 1995a
, 1995b
). In agreement with
this model, the first predicted hydrophobic segment of Tim54 is located
immediately after the positively charged sequence (Kerscher et
al., 1997
).
In an elegant series of experiments, Mihara and colleagues (Hachiya
et al., 1995
; Komiya et al., 1996
, 1997
)
demonstrated that cytosolic cofactors from rabbit reticulocyte lysate
are important determinants for the selection of import receptors by
preproteins. By using purified preproteins and cytosolic chaperones,
they showed that preproteins bound to the mitochondrial import
stimulation factor are imported via Tom70 in an ATP-dependent reaction,
whereas preproteins interacting with cytosolic Hsp70 preferentially use Tom20 in an ATP-independent manner. In agreement with these
observations we found that the import of Tim22 via Tom20 did not
require cytosolic ATP, whereas the import of Tim54 via Tom70 was
ATP-dependent. Because we used complete rabbit reticulocyte lysate, the
full set of cytosolic chaperones was present during the import
reaction, suggesting that the preproteins of Tim22 and Tim54 were bound to different chaperones before their delivery to the mitochondria. Some
preproteins bound to Tom70 seem to be transferred to Tom20 before their
insertion into the GIP, i.e., they require both Tom70 and Tom20 for
import (Keil and Pfanner, 1993
; Keil et al., 1993
; Hachiya
et al., 1995
; Hönlinger et al., 1995a
;
Komiya et al., 1997
), whereas other preproteins can be
directly transferred from Tom70 to the GIP complex (Söllner
et al., 1990
; Steger et al., 1990
). The
preprotein of Tim54 mainly behaves like the latter preproteins. Tim54
bound to Tom70 does not strictly need to be transferred to Tom20 but
can be directly transferred to the GIP complex in the absence of Tom20.
A third and quite short import pathway is followed by the small Tim
proteins (Figure 7B, route V). The small Tims are directly translocated
into the intermembrane space without an insertion into the inner
membrane because their import does not require an inner membrane
potential. Other intermembrane space proteins that only use the Tom
machinery are the cytochrome c heme lyase and the cytochrome
c1 heme lyase. These preproteins strongly require Tom20 for
their transfer to the GIP complex (Lill et al., 1992
; Steiner et al., 1995
). For the import of the small Tims,
however, the trypsin-accessible surface domains of import receptors
were dispensable, including Tom20, Tom22, and Tom70. Here, the
trypsin-resistant Tom5 that typically functions as the link between the
receptor and the general import pore (Dietmeier et al.,
1997
) plays the crucial role. Tom5 seems to represent the first Tom
protein that interacts with the majority of preproteins of the small
Tims. Besides the import of apocytochrome c (Stuart et
al., 1990a
,b
), the import of the small Tims is therefore one of
the simplest mitochondrial membrane translocation mechanisms known to date.
In addition to the various targeting pathways, this study led to two
additional pieces of information. 1) Two new homologues of the small
Tims, termed Tim13 and Tim8, were identified and found to be located in
the intermembrane space. The four cysteines that were suggested to be
of functional importance for the small Tims by formation of
Zn-finger-like motifs are fully conserved in both Tim8 and Tim13.
These small Tims were also identified independently as mitochondrial
intermembrane space proteins interacting with Tim9 and Tim10 (Koehler
et al., 1999
). 2) Blue native electrophoresis provides a
simple and efficient method to assess the correct assembly of in vitro
imported Tim22 and Tim54 into a ~300-kDa complex. This complex also
contains peripherally associated Tim12 (Koehler et al.,
1998a
,b
; Sirrenberg et al., 1998
; our unpublished results) and is termed the carrier translocase.
In summary, we conclude that multiple mechanisms exist for targeting and membrane translocation of mitochondrial preproteins. Depending on the preprotein, distinct pieces of the known major import routes are combined to yield novel pathways. This includes crossing over of pathways at the level of the GIP complex of the outer membrane. For special preproteins like the small Tims, Tom5, which typically functions at the second or third stage of import, can become the first level import component and may thus have receptor-like functions.
| |
ACKNOWLEDGMENTS |
|---|
We thank Drs. Elizabeth Craig, Bernard Guiard, Michiel Meijer, and Falk Martin for yeast strains and preproteins, and Dr. Wolfgang Voos for helpful discussion. We are grateful to Hanne Müller for expert technical assistance. This work was supported by the Deutsche Forschungsgemeinschaft, Sonderforschungsbereich 388 Freiburg, the Fonds der Chemischen Industrie, and a long-term fellowship from the Alexander-von-Humboldt Stiftung to M.T.R.
| |
FOOTNOTES |
|---|
Corresponding author. E-mail
address: pfanner{at}uni-freiburg.de.
| |
REFERENCES |
|---|
|
|
|---|