Download Overexpression of yeast karyopherin Pse1p/Kap121p stimulates the

Molecular Microbiology (1999) 31(5), 1499±1511
Overexpression of yeast karyopherin Pse1p/Kap121p
stimulates the mitochondrial import of hydrophobic
proteins in vivo
M. Corral-Debrinski,1 N. Belgareh,2 C. Blugeon,1 M. G.
Claros,3 V. Doye2 and C. Jacq1*
1
Ecole Normale SupeÂrieure, Laboratoire de GeÂneÂtique
MoleÂculaire URA CNRS 1302, 46 Rue d'Ulm, 75230 Paris
Cedex 05, France.
2
Institue Curie, CNRS UMR 144, 26 Rue d'Ulm, 75248
Paris Cedex 05, France.
3
Departamento de BiologõÂa Molecular y BioquõÂmica,
Facultad de Ciencias, Instituto Andaluz de BiotecnologõÂa,
Universidad de MaÂlaga, E-29071 MaÂlaga, Spain.
Summary
During evolution, cellular processes leading to the
transfer of genetic information failed to send all the
mitochondrial genes into the nuclear genome. Two
mitochondrial genes are still exclusively located in the
mitochondrial genome of all living organisms. They
code for two highly hydrophobic proteins: the apocytochrome b and the subunit I of cytochrome oxidase.
Assuming that the translocation machinery could not
ef®ciently transport long hydrophobic fragments, we
searched for multicopy suppressors of this physical
blockage. We demonstrated that overexpression of
Pse1p/Kap121p or Kap123p, which belong to the
superfamily of karyopherin b proteins, facilitates the
translocation of chimeric proteins containing several
stretches of apocytochrome b fused to a reporter mitochondrial gene. The effect of PSE1/KAP121 overexpression (in which PSE1 is protein secretion
enhancer 1) on mitochondrial import of the chimera
is correlated with an enrichment of the corresponding transcript in cytoplasmic ribosomes associated
with mitochondria. PSE1/KAP121 overexpression also
improves the import of the hydrophobic protein
Atm1p, an ABC transporter of the mitochondrial inner
membrane. These results suggest that in vivo PSE1/
KAP121 overexpression facilitates, either directly or
indirectly, the co-translational import of hydrophobic
proteins into mitochondria.
Received 9 September, 1998; revised 30 November, 1998; accepted
3 December, 1998. *For correspondence. E-mail Jacq@biologie.
ens.fr; Tel. (‡33) 144 323 546; Fax (‡33) 144 323 570.
Q 1999 Blackwell Science Ltd
Introduction
The majority of mitochondrial proteins are encoded by the
nuclear genome, translated in the cytoplasm, and translocated into the organelle. However, all the mitochondrial
genomes known so far always contain two genes, coding
for apocytochrome b and for the subunit I of cytochrome
oxidase. This speci®c feature suggests that the mitochondrial import machinery cannot cope with some properties
shared by these two proteins. Genetic engineering of the
yeast apocytochrome b coding sequence has shown that
two or more of its transmembrane segments can not be
imported easily to mitochondria. Therefore, a combination
of hydrophobicity characteristics, designated as mesohydrophobicity, distinguish proteins that cannot be mitochondrially imported (Claros et al., 1995). Usually,
proteins delivered to mitochondria have at their amino
terminal region a mitochondrial targeting sequence (mts)
characterized by positive charges and amphiphilicity,
which is related to the folding tightness of the attached
passenger protein. Hydrophobic proteins, such as the
ATPase subunit 9 of Neurospora crassa (F 0-9), have a particularly long mts spanning the two mitochondrial membranes, a feature that accelerates the unfolding of the
precursor (Matouschek et al., 1997). Early recognition of
the precursor via mts identi®cation is certainly a critical
step in the import process (Cartwright et al., 1997; George
et al., 1998). Cytosolic factors interacting directly with the
mts are involved in the delivery of the protein to the mitochondrial surface. The precursor is subsequently recognized by the translocase of the outer mitochondrial
membrane (TOM), which, together with the translocase
of the inner mitochondrial membrane (TIM), accomplishes
transportation across both membranes. These mechanisms have been the focus of intensive research, resulting
in an elaborate picture of the process as a whole (for
reviews, see Schatz, 1996; Neupert, 1997). In vitro, a
very tight coupling between precursor synthesis and
membrane transport was observed when the translation
of several precursors was performed in the presence of
mitochondria, indicating a co-translational import mechanism (Fujiki and Verner, 1991; 1993; Verner, 1993). In addition, it has been reported that ribosomes are speci®cally
associated with the mitochondrial surface (Kellems and
Butow, 1974; Kellems et al., 1974; 1975; Pon et al.,
1500 M. Corral-Debrinski et al.
1989), where they can actively translate mitochondrial
proteins and be involved in the co-translational transport
of nascent precursors (Ades and Butow, 1980; Suissa
and Schatz, 1982; Pon et al., 1989).
To test whether cellular sorting of highly hydrophobic
proteins could reveal new elements of the mitochondrial
import machinery, we initiated studies to identify proteins
that when overproduced would permit the mitochondrial import of a reporter protein fused to transmembrane
domains of the apocytochrome b. Unexpectedly, we found
that two multicopy suppressors encode Pse1p/Kap121p
and Kap123p. These highly similar proteins belong to the
karyopherin b family, which carries proteins and/or RNAs
through the nuclear pore complexes (NPC) (GoÈrlich,
1997; Rout et al., 1997; Schlenstedt et al., 1997; Seedorf
and Silver, 1997). We propose that, either directly or
indirectly, PSE1/KAP121 overexpression facilitates the
co-translational import of hydrophobic proteins.
for the S12345M and S678M constructs, the 600 N-terminal
amino-acid form of Pse1p/Kap121p (PSE1-t 600) and the
full-length protein (PSE1 ) were equally able to restore
CW02 cell respiration. To determine whether shorter versions of Pse1p/Kap121p remained active in this assay, we
constructed different truncated versions of PSE1/KAP121.
We observed that a protein composed of the 453 N-terminal
amino acids is still active (PSE1-t 453), whereas a shorter
protein with only the 265 N-terminal amino acids (PSE1-t
265) is not able to act as a multicopy suppressor
(Fig. 1A). To assess the exact role of the mitochondrial
targeting sequence (mts) in the in vivo assays, we constructed mts-less versions of some of the chimeric genes.
When the chimeric constructions S678M or S56M were
deleted for the presequence (S), no restoration of respiration for the CW02 strain was observed upon overexpression of PSE1 or PSE1-t (Fig. 1B). This result indicates
that the PSE1-driven translocation of the cytoplasmically
made chimeras requires the presence of a mts and is likely to
be dependent on the major mitochondrial import system.
Results
Respiration de®ciency of CW02 cells can be
corrected by the overexpression of PSE1/KAP121
To study the inhibitory effects of protein hydrophobicity on
the mitochondrial import process, we previously constructed
a universal code version of apocytochrome b (Claros et al.,
1995). Several chimeric proteins were designed to be cytoplasmically translated and targeted to mitochondria. They
were named according to the presence of the N. crassa
ATPase subunit 9 mts (S) and to the presence of the different apocytochrome b transmembrane helices (1±8) and
the bI4 RNA maturase (M). This last protein acts as a
reporter to complement the bI4 intron deletion and is
involved in apocytochrome b and cytochrome oxidase subunit I (COXI ) mRNA maturation (Banroques et al., 1986;
1987). We previously demonstrated that CW02 cells
expressing either S12345M or S678M chimeric proteins
are unable to grow on glycerol medium, indicating that
these proteins were not imported into the mitochondria
to ful®l the maturase function (Claros et al., 1995). To isolate multicopy suppressors of this mitochondrial import
defect, CW02 cells producing these chimeras were transformed with a genomic library constructed with a multicopy
vector (Claros et al., 1996). One of the most active suppressor genes turned out to code for a C-terminal truncated version of the 1089-amino-acid Pse1p/Kap121p
protein, a member of the karyopherin b family (Rout
et al., 1997; Seedorf and Silver, 1997). This truncated version, which contains the 600 N-terminal amino acids of
Pse1p, will be referred to hereafter as PSE1-t. The entire
PSE1 gene was obtained by PCR and tested for its ability
to restore CW02 cell respiration in the presence of S1234M,
S12345M, S56M or S678M chimeras. As shown in Fig. 1A
The C-terminal truncated form of Pse1p/Kap121p
does not complement the lethal phenotype of
the pse1 disruption and is not associated with the
NPCs
Pse1p/Kap121p is essential for yeast growth (Seedorf
and Silver, 1997). When heterozygous diploids, in which
one copy of PSE1 was replaced by TRP1, were transformed with the PSE1-t plasmid, only two viable Trp
spores were obtained per tetrad after asci dissection, indicating that the C-terminal region of Pse1p is required for its
essential function. Both full-length and truncated Pse1p
were tagged with the green ¯uorescent protein (GFP).
GFP-Pse1p rescues the lethal phenotype of the PSE1 disruption. In addition, GFP-PSE1 or GFP-PSE1-t, when
overexpressed, restore CW02 cell respiration, and thus
the fusion proteins behave as the untagged proteins
(data not shown). Western blot analysis revealed that
both fusion proteins migrated at the expected sizes and
were expressed at similar levels (Fig. 2A). Subcellular localization of GFP-Pse1p, examined by ¯uorescence microscopy, revealed a perinuclear labelling together with some
cytoplasmic and intranuclear staining. When expressed
in nup133 cells that display a constitutive NPC clustering
phenotype (Doye et al., 1994), GFP-Pse1p was localized
to one or only a few clusters around the nuclear envelope,
indicating that Pse1p/Kap121p interacts with nuclear pore
proteins. In contrast, GFP-Pse1p-t did not accumulate at
the nuclear periphery in wild-type cells and was not localized to NPC clusters in nup133 cells (Fig. 2B). Thus, the
C-terminal region of Pse1p, which is dispensable in our
mitochondrial assay, is required for both its essential
function and subcellular localization.
Q 1999 Blackwell Science Ltd, Molecular Microbiology, 31, 1499±1511
PSE1 overexpression improves mitochondrial protein delivery 1501
Fig. 1. Cytoplasmically made bI4 RNA maturase
fused to apocytochrome b fragments, with a
functional mts, complement mitochondrial
de®ciencies when PSE1/KAP121 is
overexpressed.
A. The nomenclature of the set of chimeric
proteins (upper part) is: S 1 . . . 8 M, in which S
refers to the mts of the subunit 9 of the ATPase
of N. crassa (Fo9-ts), 1 . . . . 8 to the
corresponding transmembrane helices of
apocytochrome b , and M to the bI4 RNA
maturase. CW02 cells transformed with multicopy
vectors directing the expression of the S12345M
(left) or the S678M (right) chimera do not grow
on glycerol (control). Overexpression of the entire
PSE1 gene (PSE1 ) or its truncated versions
PSE1-t 600 and PSE1-t 453, which code, respectively, for 600 and 453 amino-acid N-terminal
forms of Pse1p, restores the ability of CW02 to
grow on glycerol. In contrast, the 265 N-terminal
amino-acid form of Pse1p/Kap121p (PSE1-t 265)
is unable to rescue CW02 cell respiration.
B. PSE1-t represents the construct encoding the
600 N-terminal amino acids of Pse1p/Kap121p.
CW02 strains overexpressing PSE1 or PSE1-t
and transformed with multicopy vectors directing
the expression of the S56M or the S678M
chimera are able to grow on glycerol. In contrast,
overexpression of the 56M or 678M mts-less
sequences does not rescue CW02 respiration.
Transformant cells were grown on synthetic
complete medium containing 2% glucose until an
OD 600 of 2. Two independent clones were serially
diluted (1:5) and spotted either on synthetic
medium containing 2% glucose (glucose) or on
2% glycerol medium (glycerol). The plates were
incubated at 308C for 2 days (glucose) or 8 days
(glycerol).
Fig. 2. Expression and localization of GFP±Pse1p fusion proteins.
A. Whole cell extract from BMA41 cells expressing GFP±Pse1p and GFP±Pse1p-t proteins were analysed by Western blotting using
antibodies directed against GFP and against Nup133p. Both GFP±Pse1p and GFP±Pse1p-t are produced at similar levels and migrated at
their expected sizes.
B. Fluorescence microscopy of BMA41 and nup133 living cells expressing GFP±Pse1p and GFP±Pse1p-t proteins showed that the fulllength Pse1p fusion protein interacts with the NPC, whereas the truncated GFP±Pse1p-t is mainly cytosolic. Cells were also photographed
with Nomarski optics.
Q 1999 Blackwell Science Ltd, Molecular Microbiology, 31, 1499±1511
1502 M. Corral-Debrinski et al.
Overproduction of Kap123p but not Kap95p/Kap60p
facilitates mitochondrial import of hydrophobic
proteins
To determine whether the ability to restore CW02 cell
respiration in the presence of S12345M or S678M is
restricted to Pse1p/Kap121p, we tested two other karyopherin b family members, Kap95p and Kap123p. These
proteins have in common a role in the traf®cking of molecules across the nuclear pores and the ability to bind Ran,
an essential element of all the nucleocytoplasmic transport
processes studied so far (GoÈrlich, 1997; GoÈrlich et al.,
1997). Kap95p, known as karyopherin b1, is a soluble
transport protein that interacts with the nuclear localization
signal (NLS) receptor, Kap60p, to allow the translocation
of nuclear proteins across the NPC (for reviews, see GoÈrlich, 1997; Nigg, 1997). KAP123, the closest homologue
of PSE1/KAP121, was recently shown to be involved in
the nuclear import of ribosomal proteins, and its association to mRNA export was also suggested (Rout et al., 1997;
Schlenstedt et al., 1997; Seedorf and Silver, 1997). A
recent report showed that Pse1p/Kap121p is also involved
in the nuclear import of the transcription factor Pho4p
(Kaffman et al., 1998). Figure 3 shows the respiration ability of CW02 cells expressing the S12345M or S678M constructs, and transformed with high-copy number plasmids
carrying KAP123 or KAP95 expressed under the control
of their endogenous promoters. We also transformed CW02
cells with a high-copy number vector carrying KAP60,
alone or in combination with the KAP95 plasmid. We
observed that Kap95p overproduction did not allow CW02
cells to grow on glycerol medium, even when KAP60 was
overexpressed at the same time. Interestingly, Kap123p,
like Pse1p/Kap121p, restores CW02 respiration for all the
chimeric constructs tested, indicating that overexpression
of only a subset of karyopherin b proteins interferes with
the mitochondrial phenotype described in this study.
Overproduction of Pse1p/Kap121p can usher an excess
of the ABC transporter Atm1p into mitochondria
Restoration of CW02 cell respiration by overexpression of
PSE1 and apocytochrome b domains fused to the maturase
implies the complete translocation of the chimera into the
mitochondrial matrix. It was therefore important to determine
the impact of PSE1 overexpression on the transport of
hydrophobic proteins which are naturally imported into mitochondria. For that purpose, we chose Atm1p, a recently
described ABC transporter of the mitochondrial inner membrane (Leighton and Schatz, 1997). Atm1p contains six
putative transmembrane fragments, and its mesohydrophobicity level is very high although still compatible with
an in vivo mitochondrial import (Fig. 4A). To con®rm the
mitochondrial localization of Atm1p, we puri®ed mitochondria from cells expressing a C-terminal myc-tagged
version of Atm1p produced from a multicopy vector (Fig.
4B). Puri®ed mitochondria and mitoplasts were enriched
for Atm1p as they were for the mitochondrial Abf2 protein
(Caron et al., 1979; Dif¯ey and Stillman, 1991), in contrast
they lacked the nucleoporin Nup133p (Doye et al., 1994).
Fig. 3. PSE1/KAP121, KAP123, KAP95, KAP60
and the mitochondrial import of hydrophobic
chimeras. Experiments similar to those described
in Fig. 1 were carried out in CW02 cells
transformed with S12345M (left) or S678M (right)
plasmids and overproducing Kap123p, Kap95p,
Kap60p, or the combination of Kap95p and
Kap60p. Two independent clones for each
transformation were serially diluted and grown on
glucose (for 2 days) or glycerol (for 8 days).
Q 1999 Blackwell Science Ltd, Molecular Microbiology, 31, 1499±1511
PSE1 overexpression improves mitochondrial protein delivery 1503
Fig. 4. PSE1 or PSE1-t overexpression improves the mitochondrial delivery of the ABC transporter Atm1p.
A. Mitochondrial proteins used in this study are distributed according to their mesohydrophobicity values (which is an average regional
hydrophobicity value determined over a 60±80 residue window) and to the hydrophobicity of the most hydrophobic 17 residue segment
(< H17 >) using the MITOPROT II software (Claros and Vincens, 1996). The hatched area represents the border between proteins which can
(lower part, left) or cannot be mitochondrially imported. Atm1p is highly hydrophobic, whereas Abf2p is very hydrophilic. Both Atm1p and
Abf2p are encoded by the nuclear DNA whereas Cox1p and Cytbp are encoded by the mitochondrial genome.
B. Mitochondria were prepared from CW04 cells overexpressing both PSE1 and Atm1p-myc (Leighton and Schatz, 1997). SDS±PAGE was
performed on 30 mg of proteins from ®ve steps in the mitochondrial preparation: whole yeast, low-speed supernatant, crude mitochondria
(Yaffe, 1991), sucrose gradient-puri®ed mitochondria (Sargueil et al., 1990) and mitoplasts obtained by osmotic shock (Yaffe, 1991). As
control, crude mitochondria from wild-type cells was used (CW04 ). Immunoblotting was performed with the following antibodies: monoclonal
antibody Myc-9E10, polyclonal antibody against Abf2p, and polyclonal antibodies against the nuclear envelope protein Nup133p. No signal
was detected in crude mitochondria prepared from wild-type cells (CW04 ), con®rming the identity of the revealed 76 kDa polypeptide with the
protein expressed from the ATM1-myc plasmid.
C. Mitochondria were puri®ed from wild-type cells (CW04 ), or from cells overproducing the Atm1p-c-myc tagged protein and transformed with
the pFL44L multicopy vector (control) or with multicopy vectors carrying PSE1 or PSE1-t. Proteins (30 or 10 mg) from untreated mitochondria
or 30 mg of proteins from proteinase K-treated mitochondria (PK; 0.4 or 0.2 mg ml 1 ) were submitted to Western blots analyses using the
monoclonal antibody 9E10 against the myc epitope, the monoclonal 4B12-A5 antibody against Cox2p and polyclonal antibody against Abf2p.
D. Densitometric analysis of eight experiments performed with three independent mitochondria puri®cations allowed us, after normalization of
the Atm1p-myc signal with Cox2p signal, to determine the fraction of myc-tagged Atm1p which was successfully translocated into the
mitochondria. We clearly observed that Atm1p is more sensitive to proteolysis in mitochondria puri®ed from control cells (control) than from
cells overproducing Pse1p (PSE1 ) or its truncated form (PSE1-t ). This difference was signi®cant according to the paired Student's t-test
(P < 0.0004).
To assess PSE1 overexpression effect on Atm1p mitochondrial translocation, puri®ed mitochondria from cells
producing an excess of Atm1p alone (control) or in combination with Pse1p (PSE1 ) or its truncated form (PSE1-t )
were treated with proteinase K (PK) and submitted to
Western blot analyses (Fig. 4C). The amounts of Cox2p
Q 1999 Blackwell Science Ltd, Molecular Microbiology, 31, 1499±1511
and Abf2p were successively determined and used to
normalize the Atm1-myc signal for each mitochondrial preparation. Cox2p, encoded by the mitochondrial genome, is
located in the inner mitochondrial membrane and Abf2p,
encoded by the nuclear genome, is a highly hydrophilic protein which binds mitochondrial DNA in the mitochondrial
1504 M. Corral-Debrinski et al.
matrix (Caron et al., 1979; Dif¯ey and Stillman, 1991).
Results from eight independent experiments unambiguously
showed that overproduction of Pse1p or Pse1p-t stimulates
by four- to sixfold the mitochondrial import of Atm1p, as
measured by the accessibility of the tagged protein to proteolysis (Fig. 4, C and D). Noteworthy, we observed that
wild-type amounts of Pse1p led to the mitochondrial delivery of only 6% of the plasmid-borne Atm1p. In contrast,
<30% were successfully imported into the organelle
when cells overproduced Pse1p/Kap121p or its truncated
form (Fig. 4, C and D). Therefore, the mitochondrial import
machinery is not able to translocate an excess of the hydrophobic Atm1p protein in wild-type conditions, but when
cells overproduced Pse1p/Kap121p or its truncated form
the ef®ciency of this translocation is signi®cantly improved.
PSE1/KAP121 overexpression also facilitates the
mitochondrial translocation of apocytochrome b
transmembrane domains
For a better de®nition of the effect of PSE1/KAP121 overexpression on the mitochondrial import of highly hydrophobic proteins, we puri®ed mitochondria from cells expressing
the chimeric protein S56HA1. The precursor (pS56HA1)
and the mature form of the chimera (m56HA1) were equally
enriched in crude mitochondria and in mitochondria puri®ed on sucrose density gradients (Fig. 5A). The presence
of m56HA1 polypeptide, obtained after the cleavage of the
presequence by the matrix processing peptidase (MPP),
indicates that the precursor recognizes the TOM complex
and engages in the translocation process. Both pS56HA1
and m56HA1 polypeptides were equally abundant in mitochondria from wild-type cells (control) and from cells overproducing either Pse1p (PSE1 ) or its truncated form
(PSE1-t ). This is true for chimeras expressed from a multicopy vector (Fig. 5B) or from a low-copy number vector
(Fig. 5C). Thus, the steady-state levels of the precursor
and its ability to recognize the TOM complex do not depend
on PSE1 overexpression. When mitochondria were treated
with proteinase K (PK) (Fig. 5B and C), we clearly observed
that m56HA1 was much less accessible to proteolysis in
mitochondria puri®ed from cells overproducing Pse1p
(PSE1 ) or its truncated version (PSE1-t ) compared with
control cells, as observed for the Atm1p-myc protein.
The amounts of Cox2p and Abf2p were determined and
used to normalize the m56HA1 signal for each mitochondrial preparation. In contrast to m56HA1, we did not
observe any difference in the proportion of Abf2p in PKtreated mitochondria when Pse1p/Kap121p or its truncated
form were overproduced (Fig. 5B and C). Densitometric
analyses clearly showed that PK-treated mitochondria
from cells overexpressing PSE1 or its truncated form accumulate ®ve- to sevenfold more of the m56HA1 protein than
PK-treated mitochondria from control cells, independently
of the vector used to direct S56HA1 synthesis (Fig. 5B and
C). These results indicate that in wild-type cells most of the
chimera undergoes the initial steps of mitochondrial import,
but <90% gets stuck across the membranes. In this
blocked situation, the most hydrophobic part of the protein
remains in the cytoplasm, whereas the mts is cleaved by
the MPP in the mitochondrial matrix (Fig. 5D, I). When
Pse1p or its truncated form is overexpressed, the import
Fig. 5. Overexpression of PSE or PSE1-t improves the mitochondrial import of the S56HA1 chimera.
A. Mitochondria were prepared from CW04 cells overexpressing both S56HA1 and PSE1. SDS±PAGE was performed on 30 mg of proteins
from ®ve steps in the mitochondrial preparation: whole yeast, low-speed supernatant, crude mitochondria (Yaffe, 1991), sucrose gradientpuri®ed mitochondria (Sargueil et al., 1990) and mitoplasts obtained by osmotic shock (Yaffe, 1991). As control, crude mitochondria from
wild-type cells was used (CW04 ) Immunoblotting was performed with the following antibodies: monoclonal 12CA5 antibody against the HA1
epitope, polyclonal antibody against Abf2p, and polyclonal antibodies against the nuclear envelope protein Nup133p. No signal was detected
in mitochondria puri®ed from the wild-type strain (CW04 ), con®rming that the revealed polypeptides of 22 kDa and 14 kDa, respectively, were
synthesized from the S56HA1 plasmid. Both precursor and mature forms of the chimera were equally abundant in crude mitochondria and in
mitochondria puri®ed on a sucrose density gradient. As expected, in mitoplasts from crude mitochondria (CM) and puri®ed mitochondria (PM),
we only detected the mature form.
B. Mitochondria from CW04 cells producing the epitope-tagged S56HA1 protein encoded by a multicopy vector in combination with either the
pFL44L vector (control) or PSE1 gene (PSE1 ) or its truncated version (PSE1-t ) were submitted to Western blot analyses. Left: 32 or 16 mg
from untreated mitochondria or 32 mg of mitochondria treated with proteinase K (PK; 0.4 or 0.2 mg ml 1 ). Control of PK digestion was carried
out in the presence of 1% Triton X-100 (Triton). The following antibodies were used: monoclonal antibody 12CA5 against the HA1 epitope
(upper panel), monoclonal antibody against Cox2p (middle panel) and polyclonal antibody against Abf2p (lower panel). Right: densitometric
analyses using the TINA software were performed to calculate the ratio between m56HA1 and Cox2p signals in intact mitochondria (±PK) or in
PK-treated mitochondria (‡PK). As a control, ratios were also calculated between the Abf2p and Cox2p signals for both intact and PK-treated
mitochondria. Six independent mitochondria puri®cations, with at least three blot analyses per preparation, were used for quanti®cations.
C. Experiments similar to those presented in B were conducted using CW04 cells expressing the hydrophobic construct S56HA1 from a lowcopy vector (three HA1 epitopes were fused to the apocytochrome b domains in this vector). The graphic on the right presents densitometric
analyses performed with three independent mitochondria puri®cations and at least three blot analyses per preparation. The difference in
m56HA1 amounts, fully translocated into the mitochondrial matrix, in normal cells and in cells overexpressing PSE1 or PSE1-t is signi®cant
according to the paired Student's t-test, P < 0.0004 with the multicopy vector and P < 0.0007 with the low-copy number vector.
D. Schematic representation of mitochondrial import intermediates. (I) the hydrophobic passenger protein can be trapped en route to the
matrix. This does not prevent the cleavage of the mitochondrial targeting sequence (S) by the mitochondrial protease (MPP), but leaves the
rest of the protein accessible to the action of the proteinase K (PK). (II) A fraction of the hydrophobic protein is completely translocated to the
matrix, thus preventing its PK-induced proteolysis.
Q 1999 Blackwell Science Ltd, Molecular Microbiology, 31, 1499±1511
PSE1 overexpression improves mitochondrial protein delivery 1505
Q 1999 Blackwell Science Ltd, Molecular Microbiology, 31, 1499±1511
1506 M. Corral-Debrinski et al.
Fig. 6. S56M mRNA level is increased in mitochondria-bound ribosomes from cells overexpressing PSE1 or PSE1-t.
A. Northern blots were performed with 15 mg of total RNAs prepared from CW04 cells co-transformed with the S56M low-copy number vector
and pFL44L (control) or PSE1 (PSE1) or PSE1-t (PSE1-t) plasmids. Filters were stained with a methylene blue solution (top) and successively
probed with the maturase oligonucleotide and the ACT1 coding sequence.
B. RNAs were puri®ed from mitochondria-bound ribosomes (M) and from free cytoplasmic ribosomes (R) fractions obtained from wild-type
cells (CW04), CW04 cells co-transformed with the S56M low-copy number vector and pFL44L (control) or with either PSE1 (PSE1 ) or PSE1-t
(PSE1-t ) plasmids. Before hybridization, membranes were stained with a methylene blue solution (top). Northern blot analyses were
performed using successively ACT1, COXIII and maturase probes. Maturase probe revealed a 1.8 kb species corresponding to the S56M
mRNA, longer exposure times revealed the signal for the apocytochrome b 3.6 kb RNA precursor, which contains the bI4 intron (not shown).
C. Densitometric analyses of three independent experiments were performed using phosphorimager and TINA software systems. Ratios
between the S56M and COXIII signals are represented.
process is successful for more than 50% of the precursor,
as measured by the amount of the mature form resistant to
proteolysis (Fig. 5D, II).
PSE1/KAP121 overexpression leads to an enrichment
of the S56M transcript in the ribosome population
associated with mitochondria
It was previously shown that mitochondria-associated ribosomes can be recovered by isolating mitochondria in the
presence of Mg2‡ and cycloheximide (Kellems et al., 1975;
Ades and Butow, 1980; Suissa and Schatz, 1982).
These ribosomes were found to be enriched in mRNA species coding for imported mitochondrial proteins, and it has
been suggested that they were probably bound to mitochondria by the nascent chains of these proteins (Ades
and Butow, 1980; Suissa and Schatz, 1982; Pon et al.,
1989; Cartwright et al., 1997; George et al., 1998). The
process controlling this asymmetric distribution is, for the
time being, largely speculative. In connection with these
observations, we studied the in¯uence of PSE1/KAP121
overexpression on the speci®c localization of the S56M
mRNA coding for the hydrophobic chimeric protein. We
®rst observed that the steady-state level of the S56M
mRNA was not modi®ed upon PSE1/KAP121 overexpression (Fig. 6A). We then puri®ed free cytoplasmic ribosomes (Fig. 6B, R) and mitochondria-bound ribosomes
(Fig. 6B, M) to analyse their respective mRNA contents.
ACT1 mRNA and cytochrome oxidase subunit III (COXIII )
mRNA were considered as internal controls for free cytoplasmic ribosomes and mitochondria-bound ribosomes
fractions respectively. The 1.8 kb S56M mRNA was mainly
found in mitochondria-bound ribosomes, as expected for a
mitochondrially targeted protein (Ades and Butow, 1980;
Suissa and Schatz, 1982). This transcript is poorly represented in the free cytoplasmic ribosome fraction and only
detectable by RT-PCR (data not shown). When PSE1 or
its truncated form was overexpressed, we consistently
observed a three- to fourfold increase of the S56M mRNA
level in mitochondria-bound ribosomes (Fig. 6, B and C).
These data suggest that PSE1/KAP121 overexpression
might, somehow, enhance the asymmetric subcellular
localization of a mRNA coding for a highly hydrophobic
protein fused to a mitochondrial targeting sequence.
Further studies are required to assess the generality of
this phenomenon.
Discussion
Stuck hydrophobic precursors can be driven into
mitochondria when PSE1/KAP121 is overexpressed
We have previously studied the inhibitory effects on mitochondrial import of diverse combinations of apocytochrome b transmembrane helices (1±8). When these
hydrophobic domains were fused to the mts of F 0 subunit
9 of N. crassa (S) and maturase (M) to create chimeric
proteins such as S12345M, S678M or S56M, the ability
to complement a mitochondrial maturase de®ciency was
totally abolished (Claros et al., 1995). Here, we demonstrated that very little of the S56M hybrid is actually delivered to the mitochondrial matrix because the precursor
Q 1999 Blackwell Science Ltd, Molecular Microbiology, 31, 1499±1511
PSE1 overexpression improves mitochondrial protein delivery 1507
gets stuck between both membranes. This transport block
does not preclude the N-terminal cleavage of the precursor
by the matrix processing protease MPP. Such intermediates, spanning both mitochondrial membranes, have
been widely used in vitro to study the mitochondrial import
machinery and they have also been detected in vivo (Pon
et al., 1989; Schulke et al., 1997). Our genetic screen
revealed that overproduction of either Pse1p/Kap121p or
its C-terminal truncated versions improves the ability of
several combinations of apocytochrome b fragments fused
to the maturase to complement mitochondrial maturase
de®ciency. Biochemical analyses demonstrated that this
effect is unambiguously due to a more ef®cient delivery
of the chimeric protein into the mitochondrial matrix. The
fact that the Pse1p/Kap121p's effect entirely relies on the
presence of the mts indicates that its overproduction
acts in collaboration with the classical TOM/TIM mitochondrial import machinery (for reviews, see Neupert, 1997;
Schatz, 1996). Moreover, PSE1/KAP121 overexpression
can stimulate mitochondrial translocation of Atm1p, a
recently characterized hydrophobic protein (Leighton and
Schatz, 1997). These data strongly suggest that PSE1/
KAP121 overexpression actually improves, directly or
indirectly, a limiting step in the mitochondrial delivery of
hydrophobic proteins.
Connections between karyopherins b and the
mitochondrial import process
Structurally, Pse1p/Kap121p belongs to the karyopherin b
superfamily of proteins which bind Ran (GoÈrlich et al.,
1997; Rout et al., 1997; Schlenstedt et al., 1997; Seedorf
and Silver, 1997). All of these proteins share the same
set of motifs in their amino-terminal regions: the CRIME
motifs, the acidic domain and the ARM1±ARM2 domains
(Fornerod et al., 1997; Schlenstedt et al., 1997). Among
the karyopherins, Kap123p shares the closest homology
with Pse1p/Kap121p and showed similar effects on the
mitochondrial import of hydrophobic proteins. In contrast,
overproduction of Kap95p alone or in combination with
Kap60p did not have any effect on mitochondrial import,
con®rming that this phenotype is restricted to a subset of
karyopherins b. In the movement of macromolecules in
and out of the nucleus, two different functions have been
proposed for Pse1p/Kap121p and Kap123p. These karyopherins could either control the nuclear import of ribosomal proteins (GoÈrlich et al., 1997; Rout et al., 1997;
Schlenstedt et al., 1997) or be involved in the nuclear
export of mRNAs (Seedorf and Silver, 1997). A recent
report showed that Pse1p/Kap121p is also involved in the
nuclear import of the transcription factor Pho4p (Kaffman
et al., 1998). How could PSE1/KAP121 overexpression
modify the nucleocytoplasmic traf®c? It seems that karyopherins might act in dynamic competition to transport
Q 1999 Blackwell Science Ltd, Molecular Microbiology, 31, 1499±1511
more or less distinct classes of substrates (Rout et al.,
1997). In that respect, amino-terminal deletions of karyopherin b1(Kutay et al., 1997) and Kap123p (Schlenstedt et
al., 1997) were shown to block bidirectional traf®c across
the NPC. In our system, a dynamic competition might be
in¯uenced by Pse1p/Kap121p overproduction and speci®c pathways of the nucleocytoplasmic traf®c may be modi®ed, particularly when the truncated form, poorly localized
to the nuclear envelope, is overexpressed.
PSE1 overexpression might improve the co-translational
import of hydrophobic proteins
Because we observed that PSE1/KAP121 overexpression does not quantitatively modify the steady-state levels
of either the hybrid mRNA or the corresponding protein,
we can discard the possibility that Pse1p/Kap121p could
control the transcription or translation of the chimera.
Instead, high levels of Pse1p/Kap121p lead to a successful translocation of the hydrophobic protein into the mitochondria. It is then possible that translation might be
modi®ed in a way that prevents aggregation of the apocytochrome b helices, bene®cially maintaining unfolded
structures required for the import. This is reminiscent of
a long-standing hypothesis in which mitochondrial import
could be a co-translational process (Ades and Butow,
1980; Suissa and Schatz, 1982; Verner, 1993). Although
controversial, this process is generally envisaged either
for subclasses of mitochondrially imported proteins (Neupert, 1997) or during special growth conditions (Reid and
Schatz, 1982; Schatz and Butow, 1983). Recent reports
strengthen this hypothesis because both the nascent polypeptide-associated complex (NAC) and Mft52p, a protein
that binds both the presequence and Tom20p, co-operate
in initiating mitochondrial protein targeting in vivo (Cartwright et al., 1997; George et al., 1998). Therefore, hydrophobic proteins delivery to the mitochondria could actually
rely on a tight coupling between both translation and
import processes. It has been shown that mRNAs for different imported mitochondrial polypeptides are enriched
in mitochondria-bound polysomes fractions (Ades and
Butow, 1980; Suissa and Schatz, 1982). In our hands, mitochondria-bound ribosomes puri®ed from cells overproducing Pse1p/Kap121p are actually enriched in the mRNA
coding for the hydrophobic protein S56M. Further studies
are required to generalize this phenomenon to additional
hydrophobic protein messengers. Nevertheless, we postulate that the enrichment of mRNA species coding for hydrophobic proteins in mitochondria-bound ribosomes could
favour the co-translational import of this kind of protein
into the organelle.
Considering the possible role of mRNA binding proteins
in the transport of mRNAs through the cytosol, some relationships between mRNA export and mitochondrial protein
1508 M. Corral-Debrinski et al.
transport were proposed (Lithgow et al., 1997). In this
respect, it is worth remembering that a mutation in Npl3p,
an essential protein involved in mRNA traf®cking between
the nucleus and the cytoplasm (Bossie et al., 1992; Flach
et al., 1994; Singleton et al., 1995), restores the mitochondrial delivery of the ATPase subunit b, in which the targeting signal was corrupted (Ellis and Reid, 1993).
Interestingly, the mammalian ATPase b mRNA was located
in the vicinity of mitochondria. A role for this localization in
fast tracking the sorting of the precursor protein to its mitochondrial destination has been suggested (Egea et al.,
1997; Izquierdo and Cuezva, 1997; Lithgow et al., 1997).
We therefore favour the hypothesis that PSE1/KAP121
overexpression can, either directly or indirectly, improve
the coupling between mRNA-speci®c localization/translation and mitochondrial import processes. Studies aimed
at generalizing mRNA-speci®c delivery and translation in
the vicinity of mitochondria for other hydrophobic proteins
and at characterizing sequences responsible for the asymmetric distribution of mitochondrial transcripts are currently
in progress.
Experimental procedures
Yeast strains and growth conditions
All of the S. cerevisiae strains used in this study are isonuclear
with the strain W303-1B (MATa, leu2-3; ura 3-1; trp1-1, ade1-2,
his3-11, can1-100 ). CW02 and CW04 strains were obtained
as described previously (Conde and Fink, 1976; Dujardin et
al., 1980). CW02 cells do not possess the bI4 maturase activity and are therefore unable to respire because the splicing of
the aI4 intron in the precursor of COXI messenger does not
occur. CW04 cells used for mitochondria puri®cation possess
a functional bI4 RNA maturase.
For PSE1 disruption, the BMA41 diploid strain was used
(MATa/MATa, ura3-1/ura3-1, ade2-1/ade2-1, leu2-3,112/
leu2-3,112, his3-11,15/his3-11,15, Dtrp1/ Dtrp1 ). The diploid
strain disrupted for one pse1 locus was called MCD01
(MATa/MATa , ura3-1/ura3-1, ade2-1/ade2-1, leu2-3,112,
leu2-3,112, his3-11,15/his3-11,15, Dtrp1/ Dtrp1, PSE1/
pse1::TRP1 ). For GFP-Pse1p localization, the nup133
strain was used (Doye et al., 1994). Strains were grown at
308C in rich medium (1% yeast extract, 2% peptone, 2% glucose and 30 mg ml 1 adenine) in glycerol medium (1% yeast
extract, 2% peptone and 2% glycerol) or synthetic medium
containing 2% glucose supplemented with the appropriate
nutritional requirements. For solid medium, 2% agar was
added. Yeast cells were transformed using a simpli®ed lithium
method (Gietz et al., 1992).
Plasmids and DNA manipulations
All bacteriological analyses and DNA manipulations were performed as previously described (Sambrook et al., 1989). Oligonucleotides used in this study are shown in Table 1.
Vectors used to express the S 1. . .8 M chimera [in which S
refers to the mts of the subunit 9 of the ATPase of N. crassa
(Fo9-mts); 1±8, to the corresponding transmembrane helices
of apocytochrome b; and M to the bI4 RNA maturase] derive
from the YEpJB1-21-2 plasmid (Banroques et al., 1986; 1987).
Constructions devoid of the Fo9-mts were obtained by cloning
sequences coding for apocytochrome b helices in the YEpBJ120-2 plasmid (Claros et al., 1995; 1996). To construct the
multicopy vector expressing S56HA1, DNA encoding the nineamino-acid epitope of in¯uenza virus haemagglutinin HA1
was inserted after helices 5 and 6 of the apocytochrome
b sequence. For the low-copy number vector expressing
S56HA1, three HA1 epitopes were added after helices 5
and 6 of the apocytochrome b sequence. The screen for multicopy suppressors improving the delivery of highly hydrophobic
proteins into the mitochondria was performed as previously
Table 1. Oligonucleotides used in this study.
58 Primer
58±38
38 Primer
58±38
PCR product
length (bp)
AUG position (bp)
STOP position (bp)
PSE1/KAP121
gataaaggatccaatagc
ctatcatggactatctg
gtctttggatccatcatttc
cttatccattgccgc
3970
‡ 365
‡ 3550
KAP123/YRB4
cgcatagtcgacaattca
caatgtaccaataattttta
ctat
tcgtcagtcgacttacattt
gtacaagaccaatatccg
gggc
4370
‡ 405
‡ 3744
KAP95
gcccatggatcctgttcc
gggggtggttgccaagt
ttcacta
cattatggatccgggtac
ctgctcattcgcattaccat
acat
3960
‡ 680
‡ 3260
KAP60
ccactaacgataagacat
ctgctattagcc
cgcctgttagatctttcag
ctgtggatatt
2547
‡ 588
‡ 2210
GFP
tccatccacgtgatggtg
agcaagggcgaggagc
tgttca
gccgctcacgtggtaca
gctcgtccatgccgaga
gtgatc
714
COXIII
tattacctgcgattaaggc
atgatgactat
Maturase
atctaataggcttaacacc
tgatc
‡ 13
None, in frame with PSE1
Q 1999 Blackwell Science Ltd, Molecular Microbiology, 31, 1499±1511
PSE1 overexpression improves mitochondrial protein delivery 1509
described (Claros et al., 1996). The sequence coding for the
453 N-terminal amino acids of Pse1p/Kap121p was cloned
using the unique NaeI site present in the original pFL44LPSE1-t vector. Gene sequences coding for full-length Pse1p,
its 265 N-terminal amino-acid form, Kap123p, Kap60p and
Kap95p were obtained by a 16 cycle PCR ampli®cation,
using as template yeast genomic DNA and the rTth DNA polymerase XL (Perkin-Elmer Cetus) (Barnes, 1994). The DNA
from PCR products was cloned in both high-copy number
(pFL44L) and low-copy number (pFL38) vectors (Bonneaud
et al., 1991).
PSE1 gene disruption and subcellular localization of
the corresponding protein
Two kilobases of the PSE1 coding sequence, including the
AUG codon, was removed and replaced by the TRP1 gene.
This sequence was used to disrupt the PSE1 chromosomal
locus of the BMA41 strain. Tryptophan prototrophs obtained
were checked by Southern blot and PCR analyses to con®rm
that PSE1 was correctly disrupted. The MCD01 strain obtained
was sporulated and dissected. Each tetrad yielded only two
viable spores, which were trp1 . To obtain the GFP±PSE1
and GFP±PSE1-t fusion genes, the GFP-coding sequence
of pEGFP-N3 (Clonetech Laboratories) was ampli®ed by
PCR using the Pfu protein and cloned in frame at the unique
Pml I site of the PSE1 and PSE1-t vectors, located 27 nucleotides upstream of the PSE1 AUG codon. Therefore, the initiation codon for these fusion proteins is the GFP one, PSE1
ORF being in frame with the last amino acid of the GFP and
nine additional amino acids before the authentic PSE1 AUG
(AIARTVTTSI).
Living cells expressing the GFP-tagged proteins were
observed by ¯uorescence with a chilled CCD camera (Hamamatsu Photonics) as previously described (Doye et al., 1994).
Whole cell extracts from BMA41 strains expressing the GFP±
Pse1p or GFP±Pse1p-t proteins were prepared using glass
beads. Twenty micrograms of protein was applied on 8%
SDS-polyacrylamide gels. Immune sera were used at dilutions
1:250 for af®nity-puri®ed anti-Nup133p (Doye et al., 1994)
and 1:2000 for polyclonal anti-GFP (Clontech).
Mitochondria puri®cation
Mitochondria were puri®ed either by successive centrifugations or sucrose density gradients as described previously
(Sargueil et al., 1990; Yaffe, 1991). To determine the fraction
of the C-terminally tagged proteins which have been translocated into the organelle, mitochondria were treated with
200 or 400 mg ml 1 of proteinase K (PK) at 08C for 15 min. As
control for protease accessibility, PK digestion was performed
in the presence of 1% Triton X-100, a detergent that disrupts
both mitochondrial membranes. The proteolysis of the tagged
protein con®rms its location, somewhere in the organelle, in a
protease-sensitive form.
Western blot analyses
Proteins were resolved in SDS±PAGE and transferred to nitrocellulose (Sambrook et al., 1989). Filters were probed with the
Q 1999 Blackwell Science Ltd, Molecular Microbiology, 31, 1499±1511
following antibodies: mouse monoclonal antibody 12CA5
against the HA1 epitope (1:5000); mouse monoclonal antibody 9E10 against the myc epitope (1:5000); rabbit polyclonal
antibody against Abf2p (1:10 000); mouse monoclonal antibody 4B12-A5 against Cox2p (1:10 000). Immunoreactive
bands were visualized with either anti-mouse or anti-rabbit
IgGs coupled to horseradish peroxidase followed by ECL
detection (Amersham International) according to the manufacturer's instructions. Densitometric analyses were performed
on independent mitochondrial preparations. Speci®c signals
for m56HA1 and Atm1p-myc were normalized for each sample,
using the Cox2p or the Abf2p signals ( TINA software). In all
the experiments performed, both normalized results were
very similar, suggesting that Abf2p and Cox2p quantities
were equivalent in all the mitochondria preparations tested.
The amounts of m56HA1 and Atm1p-myc-tagged protein
resistant to proteolysis in mitochondria from wild-type cells
or from cells overproducing Pse1p were compared with a
paired Student's t-test using a con®dence level of 95%.
Puri®cation of free and mitochondria-bound
ribosomes
Puri®cation of mitochondria-bound ribosomes and free cytoplasmic ribosomes was performed as described previously
(Ades and Butow, 1980; Suissa and Schatz, 1982). Total
RNA from both fractions and from whole cells were obtained
by a hot-phenol method (Schmiit et al., 1990). Northern blots
were performed using 15 mg of RNA electrophoresed through
denaturing formaldehyde agarose gels, after transfer nylon
membranes were stained with methylene blue solution (Sambrook et al., 1989). Probes used were the 1 kb EcoRI±Hin dIII
fragment of the ACT1 ORF, COXIII and Maturase oligonucleotides shown in Table 1, which revealed, respectively, ACT1
mRNA, COXIII mRNA and the S56M hybrid transcript. Phosphorimager system and TINA software were used to compare
the relative abundance of the S56M hybrid mRNA.
Acknowledgements
We would like to thank A. Delahodde, Suzanne Joyce and E.
Fabre for friendly discussions and critical reading of the
manuscript. We are grateful to J. P. diRago for the gift of the
BMA41 strain and to Dr G. Schatz for the gift of the ATM1-cmyc vector. This work was supported by CNRS (URA1302)
by the EÂcole Normale SupeÂrieure and by a grant from the
AFM (French Association for Myopathies, MNV 96).
References
Ades, I.Z., and Butow, R.A. (1980) The products of mitochondrial-bound cytoplasmic polysomes in yeast. J Biol Chem
255: 9918±9924.
Banroques, J., Delahodde, A., and Jacq, C. (1986) A mitochondrial RNA maturase gene transferred to the yeast
nucleus can control mitochondrial mRNA splicing. Cell
46: 837±844.
Banroques, J., Perea, J., and Jacq, C. (1987) Ef®cient splicing of two yeast mitochondrial introns controlled by a
nuclear-encoded maturase. EMBO J 6: 1085±1091.
1510 M. Corral-Debrinski et al.
Barnes, W.M. (1994) PCR ampli®cation of up to 35 kb DNA
with high ®delity and high yield from l templates. Proc
Natl Acad Sci USA 91: 2216±2220.
Bonneaud, N., Ozier-Kalogeropoulos, O., Li, G., Labouesse,
M., Minvielle-Sebatia, L., and Lacoutre, F. (1991) A family
of low and high copy replicative, integrative and singlestranded S. cerivisiae/E. coli shuttle vectors. Yeast 7:
609±615.
Bossie, M.A., DeHoratius, C., Barcelo, G., and Silver, P.A.
(1992) A mutant nuclear protein with similarity to RNA
binding proteins interferes with nuclear import in yeast.
Mol Biol Cell 3: 875±893.
Caron, F., Jacq, C., and RouvieÁre-Yaniv, J. (1979) Characterization of a histone-like protein extracted from yeast mitochondria. Proc Natl Acad Sci USA 76: 4265±4269.
Cartwright, P., Beilharz, T., Hansen, P., Garrett, J., and Lithgow, T. (1997) Mft52, an acid-bristle protein in the cytosol
that delivers precursor proteins to yeast mitochondria. J
Biol Chem 272: 5320±5325.
Claros, M.G., and Vincens, P. (1996) Computational methods
to predict mitochondrially imported proteins and their transit peptides. Eur J Biochem 241: 779±786.
Claros, M.G., Perea, J., Shu, Y., Samatey, F.A., Popot, J.-L.,
and Jacq, C. (1995) Limitations to in vivo import of hydrophobic proteins into yeast mitochondria: the case of a cytoplasmic synthesized apocytochrome b. Eur J Biochem
228: 762±771.
Claros, M.G., Perea, J., and Jacq, C. (1996) Allotopic expression of yeast mitochondrial maturase: To study mitochondrial import of hydrophobic proteins. Methods Enzymol
264: 389±403.
Conde, J., and Fink, G. (1976) A mutant of S. cerevisiae
defective for nuclear fusion. Proc Natl Acad Sci USA 73:
3651±3655.
Dif¯ey, J.F.K., and Stillman, B. (1991) A close relative of the
nuclear, chromosomal high-mobility group protein HMG1 in
the yeast mitochondria. Proc Natl Acad Sci USA 88: 7864±
7868.
Doye, V., Wepf, R., and Hurt, E.C. (1994) A novel nuclear
pore protein Nup133p with distinct roles in poly (A‡) RNA
transport and nuclear pore distribution. EMBO J 13:
6062±6075.
Dujardin, G., Pajot, P., Groudinsky, O., and Slonimski, P.
(1980) Long range control circuits within mitochondria
and between nucleus and mitochondria. Mol Gen Genet
179: 469±482.
Egea, G., Izquierdo, J.M., Martin, R.C.S., and Cuezva, J.M.
(1997) Mrna encoding the b-subunit of the mitochondrial
F1-ATPase is a localized mRNA in rat hepatocytes. Biochem J 322: 557±565.
Ellis, E.M., and Reid, G.A. (1993) The Saccharomyces cerivisiae MTS1 gene encodes a putative RNA-binding protein
involved in mitochondrial protein targeting. Gene 132:
175±183.
Flach, J., Bossie, M., Vogel, J., Corbett, A., Jinks, T., Willins,
D.A., and Silver, P.A. (1994) A yeast RNA-binding protein
shuttles between the nucleus and the cytoplasm. Mol Cell
Biol 14: 8399±8407.
Fornerod, M., Deursen, J., Baal, S., Reynolds, A., Davis,
D., Murti, K.G., Fransen, J., and Grosveld, G. (1997)
The human homologue of yeast CRM1 is in a dynamic
subcomplex with CAN/Nup214 and a novel nuclear pore
component Nup88. EMBO J 16: 807±816.
Fujiki, M., and Verner, K. (1991) Coupling of protein synthesis and mitochondrial import in a homologous yeast in
vitro system. J Biol Chem 266: 6841±6847.
Fujiki, M., and Verner, K. (1993) Coupling of cytosolic protein
synthesis and mitochondrial protein import in yeast. J Biol
Chem 268: 1914±1920.
George, R., Beddoe, T., Landl, K., and Lithgow, T. (1998)
The yeast nascent polypeptide-associated complex initiates
protein targeting to mitochondria in vivo. Proc Natl Acad
Sci USA 95: 2296±2301.
Gietz, D., St. Jean, A., Woods, R.A., and Schiestl, R.H.
(1992) Improved method for high ef®ciency transformation
of intact yeast cells. Nucleic Acids Res 20: 1425.
GoÈrlich, D. (1997) Nuclear protein import. Curr Opin Cell Biol
9: 412±419.
GoÈrlich, D., Dabrowski, M., Bischoff, F.R., Kutay, U., Bork,
P., Hartmann, E., Prehn, S., and Izaurralde, E. (1997) A
novel class of Ran GTP binding proteins. J Cell Biol 138:
65±80.
Izquierdo, J.M., and Cuezva, J.M. (1997) Control of the translational ef®ciency of b-F1-ATPase mRNA depends on the
regulation of a protein that binds the 38 untranslated region
of the mRNA. Mol Cell Biol 17: 5255±5268.
Kaffman, A., Rank, N.M., and O'Shea, E.K. (1998) Phosphorylation regulates association of the transcription factor
Pho4 with its import receptor Pse1/Kap121. Genes Dev
12: 2673±2683.
Kellems, R.E., and Butow, R.A. (1974) Cytoplasmic type 80S
ribosomes associated with yeast mitochondria. III.
Changes in the amount of bound ribosomes in response
to changes in metabolic state. J Biol Chem 249: 3304±
3310.
Kellems, R.E., Allison, V.F., and Butow, R.A. (1974) Cytoplasmic type 80S ribosomes associated with yeast mitochondria. II. Evidence for the association of cytoplasmic
ribosomes with the outer mitochondrial membrane in situ.
J Biol Chem 249: 3297±3303.
Kellems, R.E., Allison, V.F., and Butow, R.A. (1975) Cytoplasmic type 80S ribosomes associated with yeast
mitochondria. IV. Attachment of ribosomes to the outer
membrane of isolated mitochondria. J Cell Biol 65: 1±14.
Kutay, U., Izaurralde, E., Bischoff, F.R., Mattaj, I.W., and
GoÈrlich, D. (1997) Dominant-negative mutants of importinb block multiple pathways of import and export through
the nuclear pore complex. EMBO J 16: 1153±1163.
Leighton, J., and Schatz, G. (1997) An ABC transporter in
the mitochondrial inner membrane is required for normal
growth of yeast. EMBO J 14: 188±195.
Lithgow, T., Cuezva, J.M., and Silver, P.A. (1997) Highways
for protein delivery to the mitochondria. Trends Biochem
Sci 22: 110±113.
Matouschek, A., Azem, A., Ratliff, K., Glick, B.S., Schimd, K.,
and Schatz, G. (1997) Active unfolding of precursor proteins during mitochondrial protein import. EMBO J 16:
6727±6736.
Neupert, W. (1997) Protein import into mitochondria. Annu
Rev Biochem 66: 683±717.
Nigg, E.A. (1997) Nucleocytoplasmic transport: signals,
mechanisms and regulation. Nature 386: 779±787.
Q 1999 Blackwell Science Ltd, Molecular Microbiology, 31, 1499±1511
PSE1 overexpression improves mitochondrial protein delivery 1511
Pon, L., Moll, T., Vestweber, D., Marshallsay, B., and Schatz,
G. (1989) Protein import into mitochondria: ATP-dependent translocation activity in submitochondrial fraction
enriched membrane contact sites and speci®c proteins.
J Cell Biol 109: 2603±2616.
Reid, G.A., and Schatz, G. (1982) Import of proteins into
mitochondria. Yeast cells grown in the presence of carbonyl
cyanide m-chlorophenylhydrazone accumulate massive
amounts of some mitochondrial precursor polypeptides.
J Biol Chem 257: 13056±13060.
Rout, M.P., Blobel, G., and Aitchison, J.D. (1997) A distinct
nuclear import pathway used by ribosomal proteins. Cell
89: 715±725.
Sambrook, K.J., Fritsch, E.F., and Maniatis, T. (1989) Molecular Cloning. Cold Spring Harbor, NY: Cold Spring Harbor
Laboratory Press.
Sargueil, B., Hatat, D., Delahodde, A., and Jacq, C. (1990) In
vitro and in vivo analyses of an intron-encoded DNA endonuclease from yeast mitochondria. Nucleic Acids Res 18:
5659±5665.
Schatz, G. (1996) The protein import system of mitochondria.
J Biol Chem 271: 31763±31766.
Schatz, G., and Butow, R. (1983) How are proteins imported
into mitochondria? Cell 32: 316±318.
Schlenstedt, G., Smirnova, E., Deane, R., Solsbacher, J.,
Q 1999 Blackwell Science Ltd, Molecular Microbiology, 31, 1499±1511
Kutay, U., GoÈrlich, D., Ponstingl, H., and Bischoff, F.R.
(1997) Yrb4p, a yeast Ran-GTP-binding protein involved
in import of ribosomal protein L25 into the nucleus.
EMBO J 16: 6237±6249.
Schmiit, M.E., Brown, T.A., and Trumpower, B.L. (1990)
A rapid and simple method for preparation of RNA from
Saccharomyces cerevisiae. Nucleic Acids Res 18: 3091.
Schulke, N., Sepuri, N.B., and Pain, D. (1997) In vivo zippering of inner and outer mitochondrial membranes by a
stable intermediate. Proc Natl Acad Sci USA 94: 7314±
7319.
Seedorf, M., and Silver, P.A. (1997) Importin/karyopherin
protein family members required for mRNA export from
the nucleus. Proc Natl Acad Sci USA 94: 8590±8595.
Singleton, D.R., Chen, S., Hitomi, M., Kumagai, C., and Tartakoff, A.M. (1995) A yeast protein that bidirectionally
affects nucleocytoplasmic transport. J Cell Sci 108: 265±
272.
Suissa, M., and Schatz, G. (1982) Import of proteins into
mitochondria. J Biol Chem 257: 13048±13055.
Verner, K. (1993) Co-translational protein import into mitochondria: an alternative view. Trends Biochem Sci 18:
364±371.
Yaffe, M.P. (1991) Analysis of Mitochondrial function and
assembly. Pasadena, CA: Academy Press.