Jane K. Setlow
Brookhaven National Laboratory, Upton, NY, USA
Book Series: GENETIC ENGINEERING: PRINCIPLES AND METHODS
Genetic Engineering, published by Kluwer Academic/Plenum Publishers since 1979, has
been and continues to be one of the most widely read and authoritative review sources in
Genetic Engineering. Traditionally, each volume has contained an eclectic mix of articles on
the different aspects of genetics and genetic engineering research - all articles on topics of
This volume, as with the previous books in the series, presents state-of-the-art discussions in
genetics and genetic engineering, focusing on new genetic methodologies, including DNA
transcription, gene therapy, and plant science, among others.
Contents and Contributors
Kluwer Academic/Plenum Publishers
Hardbound, ISBN 0-306-47280-5
August 2002 , 286 pp.
Application of FLP/FRT Site-Specific DNA Recombination System in Plants; H.
Luo, A.P. Kausch. Protein Quality Control in Bacterial Cells: Integrated
Networks of Chaperones and ATP-Dependent Proteases; J.M. Flanagan, M.C.
Bewley. Regulation of the Ras-MAPK Pathway at the level of Ras and Raf; H.
Vikis, K.-L. Guan. Plant Virus Gene Vectors: Biotechnology Applications in
Agriculture and Medicine; K.-B.G. Scholthof, et al. Integrins and the
Myocardium; S.-Y. Shai, et al. Foreign DNA: Integration and Expression in
Transgenic Plants; R.M. Twyman, et al. Novel Approaches to Controlling
Transcription; T.D. Schaal, et al. The Use of DNA Polymorphisms in Genetic
Mapping; C.A. Cullis. Import of Nuclear-Encoded RNAs into Yeast and Human
Mitochondria: Experimental Approaches and Possible Biomedical Applications;
N. Entelis, et al. An Introduction to
C Metabolic Flux Analysis; W. Wiechert.
Gene Silencing - Principles and Application; C. Horser, et al. Index.
Gene Engineering: Principles & Methods
Kluwer Academic/Plenum Press (U.S.A.)
J. K. Setlow, Editor
Import of nuclear encoded RNAs into yeast and human mitochondria:
experimental approaches and possible biomedical applications.
, Kolesnikova, O.
, Kazakova, H.
, Brandina, I.
, Kamenski, P.
, Martin, R. P
& Tarassov, I.
FRE 2375 of the CNRS (MEPH), Institut de Physiologie et Chimie Biologique, 21, rue René
Descartes, 67084 Strasbourg, France.
Department of Molecular Biology, Moscow State University, Moscow 119899, Russia.
Corresponding author :FRE 2375 CNRS
"Modèles d'Etude de Pathologies Humaines",
Institut de Physiologie et Chimie Biologique,
21, rue René Descartes,
67084 Strasbourg cedex, France.
Phone: (+33) 3 90 24 14 60
FAX:(+33) 3 88 41 70 70
Mitochondria import from the cytoplasm the vast majority of proteins and some RNAs.
Although there exists extended knowledge concerning the mechanisms of protein import, the
import of RNA is poorly understood. It was almost exclusively studied on the model of tRNA
import, in several protozoans, plants and yeast. Mammalian mitochondria, which do not
import tRNAs naturally, are hypothesized to import other small RNA molecules from the
cytoplasm. We studied tRNA import in the yeast system, both in vitro and in vivo and applied
similar approaches to study 5S rRNA import into human mitochondria. Despite the obvious
divergence of RNA import systems suggested for different species, we find that in yeast and
human cells this pathway uses similar mechanisms exploiting cytosolic proteins to target the
RNA to the organelle and requiring the integrity of pre-protein import apparatus. The import
pathway might be of interest from a biomedical point of view, to target into mitochondria
RNAs that could suppress pathologic mutations in mitochondrial DNA. Yeast represents a
good model to elaborate such gene therapy approach. We describe here the various
approaches and protocols to study RNA import into mitochondria of yeast and human cells in
vitro and in vivo.
Mitochondria are DNA-containing organelles, which provide most of the cellular
energy through aerobic respiration. The activity of the mitochondrial respiratory chain relies
on the contribution of two sets of genetic information, the mitochondrial and the nuclear
genomes. The size and organization of the mitochondrial genomes vary in a remarkable
manner. On one side we find the human mitochondrial DNA (mtDNA), which is a small
circular molecule of 16569 bp coding for 2 ribosomal RNAs, 22 transfer RNAs (which, due to
the absence of initiator tRNA and to a modified genetic code, are sufficient for mitochondrial
translation), and 13 polypeptides of the respiratory chain . Yeast mtDNA occupies an
intermediate position with 85779 bp circular molecule encoding 11 polypeptides of the
respiratory chain, one ribosomal protein, 24 tRNAs (which are able, in principle, to decode all
the codons in mitochondrial DNA), 2 rRNAs, several proteins participating in RNA
processing and DNA replication (maturases, reverse-transcriptases, DNA-endonuclease) and
7 small open reading frames coding for hypothetical proteins . On the opposite extremity
we can find plant mtDNAs which can be 200-600 kb long and, coding for rRNAs, tRNAs and
respiratory complexes subunits, also contain numerous additional genes (identified or
Nuclear genes encode the vast majority of proteins that are required for mitochondrial
biogenesis and function. The mechanisms of protein translocation into mitochondria were
described in details (recently reviewed in ). Briefly, the precursors of imported proteins
have positively charged signal sequences located in amino-terminal or internal parts of the
molecule. The precursors are recognized by specific outer membrane receptors (Tom) and are
inserted into the membrane via a multi-component pore complex GIP (General Insertion
Pore). Subsequent translocation is supposed to occur at intermembrane contact sites mediated
by the inner membrane complex of Tim proteins. This process requires energy in the form of
ATP in both the cytoplasm and mitochondrial matrix as well as an electric potential (∆Ψ)
across the mitochondrial inner membrane. N-terminal signal peptides are then cleaved in the
mitochondrial matrix by specific processing peptidases. Pre-proteins can be imported into
isolated mitochondria in the presence of ATP and an ATP-regeneration system without
additional soluble protein factors.
Contrary to the case of protein import, RNA mitochondrial targeting is studied much
lesser. Targeting of nuclear encoded tRNAs into the mitochondria was found in a number of
organisms including protozoans, plants and fungi [5-8]. The complexity of the imported
tRNA pool varies from a complete set necessary for mitochondrial translation in kinetoplastid
protozoa or 19-26 tRNAs imported in some ciliate protozoa, to a single tRNA in the yeast S.
cerevisiae. The mitochondrial genomes of land plants and protists, such as Chlamydomonas,
Paramecium and Tetrahymena are deficient in a few, or, in the case of trypanosomatidae, in
all mitochondrial tRNA genes. In other cases (S. cerevisiae and M. polymorpha), the mtDNA
carries a sufficient set of tRNA genes for reading all codons (taking into account the modified
mitochondrial genetic code), still tRNAs are imported. The mechanism used by the negatively
charged molecules of nucleic to traverse the mitochondrial double membrane remains mainly
unclear and appears to differ between organisms. One system was described for
trypanosomatids, where an RNA-specific and ATP dependent bipartite receptor was found in
the mitochondrial membranes [6, 9, 10]. Another mechanism was proposed in yeast, where
the tRNA is co-targeted into mitochondria with an imported pre-protein [7, 11].
In the yeast S. cerevisiae, one of the two cytoplasmic lysine-tRNA species, tRNA
(tRK1), is partially associated with the mitochondrial matrix . The second one,
(tRK2), is localized only in the cytoplasm. The mitochondrial DNA codes for a
third lysine isoacceptor, tRNA
(tRK3), whose localization is restricted to the organelle
(Figure 1a). The import proved to be an ATP-dependent process requiring the intactness of
the mitochondrial pre-proteins import apparatus . In addition, tRK1 import in vitro
requires the presence of soluble cytosolic factors including the cytoplasmic lysyl-tRNA
synthetase (KRS) and the cytosolic precursor form of the mitochondrial lysyl-tRNA
synthetase (pre-MSK). tRK1 is imported in its aminoacylated form. On the other hand, pre-
MSK does not aminoacylate tRK1 but can form stable complexes with its aminoacylated
form and this interaction is obligatory for mitochondrial import of the tRNA [14, 15]. We
therefore hypothesized that the pre-MSK acts as a carrier of the tRNA for its mitochondrial
translocation through the protein import channel [7, 8]. However, pre-MSK alone is not
sufficient to provide tRNA import, and other factors are also needed (Figure 2).
We have found that in yeast, mitochondrial import of tRNA is a highly selective process.
Systematic analysis of mutant versions of tRK1 and tRK2 permitted us to identify "import
determinants" in the anticodon and aminoacceptor arm of tRK1 [14, 16, 17] (Figure 1a,b).
Residues C34, G1:C72, and U73, once introduced into the tRK2 backbone, were found to
determine the import of this normally non-imported tRNA, both in vitro and in vivo.
Recognition by pre-MSK and in vitro import of tRK1 (and of mutant versions) required
aminoacylation by KRS. However, some mutant versions of tRK1 and tRK2 could not be
aminoacylated by KRS and still could retained the affinity to pre-MSK and were imported.
This fact suggests that the role of aminoacylation is to induce a conformational change in the
tRNA to facilitate its binding to pre-MSK, which is essential for import [14, 16, 17]. Indeed,
some of the mutant versions were shown to be imported in a misacylated form [14, 16-18].
Thus, while aminoacylation is important for import, the aminoacylation identity of the tRNA
is less important.
The possibility to import misaminoacylated versions of tRK1 and tRK2 allowed to test
whether tRNAs of cytoplasmic origin are functional in mitochondrial protein synthesis. We
showed that a version of tRK1 with an anticodon CAU for methionine is aminoacylated by
the methionyl tRNA synthetase and is imported in its methionylated form. After in vitro
import of this tRNA charged with [
S]-methionine followed by in organello protein
synthesis, the labeled amino acid was found to be incorporated into mitochondrial
polypeptides. Furthermore, an importable version of tRK2 with an anticodon CUA
(complementary to the amber stop codon) and a strong determinant for aminoacylation with
alanine was demonstrated to be imported and to suppress an Ala-to-amber mutation in the
mitochondrial COX2 gene. Thus, imported cytoplasmic tRNAs can function in mitochondrial
protein synthesis, both in vitro and in vivo .
The flexibility of the tRNA import system in yeast makes this targeting pathway
interesting for biomedical use. Indeed, it would be attractive to complement pathological
mutations in human mitochondrial tRNAs with functional tRNAs delivered from the
cytoplasm. However, no tRNA import was detected in human cells. As a first step to develop
an artificial tRNA import system in human cells, we asked whether tRNAs can be imported
into isolated human mitochondria. We showed that mitochondria isolated from HeLa or
HepG2 cultured cells are able to internalize tRK1 in vitro and this internalization has similar
requirements as with yeast mitochondria. Surprisingly, such artificial import could be
directed by either yeast cytosolic import factors (ScIDP, for S. cerevisiae Import Directing
Proteins) or by a protein extract from human cells (HmIDP). pre-MSK was found to be the
essential factor when ScIDP were used, however, in HmIDP, it could be substituted, probably
by its human counterpart, the precursor of mitochondrial lysyl-tRNA synthetase .
Furthermore, we also demonstrated that the imported yeast tRNA was functional on the
human mitochondrial translation apparatus . These results suggest the possibility of
developing a tRNA mitochondrial import system in living human cells by expressing in the
nucleus engineered importable tRNAs. Such a system could be useful to replace
nonfunctional mutant mitochondrial tRNAs by functional tRNAs imported from the
cytoplasm and thus could provide a gene therapy approach for human mitochondrial diseases
linked to tRNA mutations.
Although human cells do not import tRNAs naturally, other small RNAs of nuclear
origin have been found to be associated with the mitochondrial compartment in vivo. In
particular, it has been reported that nuclear-coded ribosomal 5S rRNA is naturally imported
into human mitochondria in vivo [20, 21]. Recently, we have demonstrated that this import
can be reproduced in vitro with isolated human mitochondria. As for tRNA import, the import
of 5S rRNA was energy- and membrane charge-dependent. Moreover, it was also dependent
upon cytosolic import factors but distinct from those directing import of tRNA. These results
might be interpreted as mitochondrial co-targeting of 5S rRNA and a mitochondrial pre-
protein through the protein import channel .
Hereafter, we describe various methodological approaches of studying RNA import in
yeast and human mitochondria and discuss the possible applications of this targeting pathway
for biomedical purposes.
Import of tRNA into yeast mitochondria in vitro
Isolation of yeast mitochondria
We routinely use the YPH499, YPH500 or W303 strains . The cells were grown to
the density of 15-20 (O.D.600 nm) in a standard YPEG medium  containing 0.5% of
glucose. The cells were harvested and washed twice with sterile water. Two alternative
methods of cell disruption were used. One is to generate spheroplasts by zymoliase T100 
and to disrupt them in a Warring blender, the second - to broke intact cells with glass beads
by shaking. Although the first method provides better output, it is much longer and the
resulting mitochondria have, as usual, lower import capacity. We preferred disruption of the
cells by shaking in the breakage buffer (BB: mannitol, 0.6M; EDTA, 1mM; Na-PIPES,
pH6.7, 10mM; BSA, 0.3%) with sterilized glass beads of 0.5mm in a vibrogenerator V4
(Edmund Buhler) at 0°C for 5 min. All further steps are done at 0-4°C. The debris are
removed by centrifugation at 4000g for 5 min. and the crude preparation containing the
mitochondria is harvested by centrifugation at 20,000g for 30 min. Several alternative
centrifugation methods can be used thereafter: linear sucrose gradient (0.25 - 1.85M), step-
gradient of sucrose (1.85M-0.6M), self-forming percoll gradient, or 2-3 cycles of low-
(4000g, 5 min) and high- speed (15,000g, 15 min) centrifugation in BB. Yeast mitochondria
obtained by use sucrose showed lower respiration and import capacities.
The Table I shows the various approaches and markers used to characterize the purity,
integrity and functionality of mitochondria. To obtain "import-active" mitochondria we
avoided to treat them by RNases. These preparations were only verified for contamination
with ER and cytosolic contamination and for integrity of the outer membrane. Northern
analyses of mitochondrial RNA and enzymatic and Western analyses of mitochondrial protein
markers showed that the levels of contamination of mitochondria isolated by several cycles of
low- and high- speed centrifugation were similar as for the mitochondria obtained by
isopicnic centrifugation. On the other hand, the organelles obtained by several cycles of
centrifugation proved to be more active in import assays.
To decrease non-specific binding of RNA to mitochondria during tRNA import assays,
we treated the final preparation with rRNA (12S/23S rRNA of E. coli commercialized by
Boeringer, at 50 µg/ml for 10 min in ice). This step was avoided when 5S rRNA import
assays were performed, since the commercial rRNA preparation contains significant amounts
of 5S rRNA which inhibits the import. After the final round of high-speed centrifugation,
mitochondria were suspended in BB at 10 mg of mitochondrial protein per ml, aliquoted,
freezed in liquid azote and stored at -80°C. For import assays, the aliquot of 100 µl was
quickly mixed with 1 ml of BB (without EDTA, containing 1 mM MgCl
, 1 mM succinate, 1
mM α-ketoglutarate and 0.5 mM ATP) pre-heated at 37°C. The suspension was replaced in
ice and mitochondria were harvested by high-speed centrifugation. Our estimations show that
such procedure does not lead to a significant loss of integrity, membrane charge, respiration
and RNA/protein import capacities of mitochondria.
Isolation of yeast Import Directing Proteins (ScIDPs)
As indicated above, pre-MSK is one of the protein factors needed to deliver tRK1 into
mitochondrial matrix. Other factors, which are not yet identified, must also be present in the
in vitro import assay. Therefore, to isolate the mixture of proteins directing the tRNA into the
mitochondria, we used a partially purified cell extract enriched with pre-MSK. It can be
obtained in several ways. Our initial procedure [16, 25] exploited the fact that disruption of
the TOM20 gene (coding for one of the alternative pre-protein receptors, Tom20p) leads to
accumulation of the precursors of mitochondrial proteins in the cytoplasm . Thus, the
strain W303,tom20::URA3 bearing the plasmid containing the MKS1 gene was used. Another
source was to isolate protein extract from either normal YPH499/500 or W303 strains and to
supplement it with purified pre-MSKp (produced either in the Pichia pastoris expression
system from Invitrogen Inc. or in the T7 polymerase-dependent system in E. coli BL21 codon
plus (DE3)-RIL strain from Stratagen). Finally, the best results were obtained by combination
of extracts from W303,tom20::URA3 strain and addition of pure pre-MSKp.
The cells were grown overnight to 10-15 (O.D. 600 nm) in 1-2 liters of a standard
YPD medium , harvested by centrifugation, washed several times with sterile water and
suspended in 2-5 ml of HKED buffer (10 mM HEPES (pH6.7), 50mM KCl, 1 mM EDTA, 5
mM DTT, protease inhibitor cocktail (Boheringer Mannheim), 0.1 mM DIFP, 0.1 mM PMSF,
10% glycerol). Cells were broken by ultrasounds (4 times 2 min at maximal efficiency) and
the debris were removed by centrifugation (15,000 rpm, 10 min). For elimination of
endogenic nucleic acids, polyethylenimin (polymin P, polyethylenediamine ) was
stepwise added. To completely remove nucleic acids, one needs to add 80 µl of 50% w/v
aqueous solution of polyethylenimin (Sigma-Aldrich) per 10g of cells. The precipitate was
removed by centrifugation (20,000 g, 10 min). The crude protein preparation was separated
by differential ammonium precipitation with steps of 30-40-50-60-70% of saturation and the
precipitates were dissolved in HKMD buffer, differing from HKED by the absence of EDTA
and preasence of 5 mM MgCl
. and dialized against HKMD containing 50% of glycerol.
Fraction 30-40% (of saturation) contained pre-MSK, while fraction 50-60% - all other import
factor(s). Combination of fractions 40% and 60%, or fraction 60% and pure pre-MSK
provided the best tRK1 import efficiency. ScIDPs were stored at -20°C for moths without loss
Preparation of labeled RNAs
In our in vitro import assays we used either natural tRNAs or T7 transcripts.
Individual S. cerevisiae tRNAs
isoacceptors were partially purified by column
chromatography (DEAE cellulose, BD-cellulose, RPC5). Final purification was achieved by
preparative polyacrylamide gel electrophoresis in a denaturing 12-13% 40cm-long gel.
Individual bands were identified by either partial sequencing or dot-hybridization with
To obtain an in vitro tRK1 transcript, as well as all other tRNA- in vitro constructs, the
tRNA genes were PCR-cloned under control of a T7 promoter and a BstNI site was
introduces at the 3'-terminus of the coding region, which gives rise to the CCA-OH3'
sequence. The resulting plasmids were BstNI-digested and transcribed in vitro by T7 RNA
polymerase. Versions containing U in 5'-terminal positions, that is unfavorable for
transcription by T7 RNA polymerase, were cloned so that the tRNA sequence was preceded
by a hammerhead ribozyme sequence. Transcription of these constructs gives rise to a
"tranzyme" molecules  with a autocatalytic activity that liberates the corresponding
tRNAs bearing a correct terminal nucleotide.
tRNAs and T7-trascripts were gel-purified, dephosphorylated by Calf Intestine
Phospatase (CIP), polynucleotide kinase-labelled with γ-[
P]-ATP, gel-purified for the nd
time and, after denaturation at 94°C for 2 min, renatured at 30°C in the presence of 0.5 mM
. As stressed above, tRK1 import is aminoacylation-dependent [15, 16]. This stands
true for the majority of importable in vitro synthesized versions of tRK1, tRK2 and tRK3 [14,
17, 19], though with some exceptions. We therefore aminoacylated tRK1 or the T7-transcripts
prior the import assays. Aminoacylation was performed by either purified cytoplasmic lysyl-
tRNA synthetase (KRS), or by the recombinant KRS produced in the T7 polymerase-
dependent system in E. coli BL21 codon plus (DE3)-RIL strain. Aminoacylation conditions
used were as described elsewhere . After aminoacylation, KRS was removed by treatment
with phenol (equilibrated with 0.1M Na-acetate, pH4.5) and tRNAs were precipitated with
In vitro import assay and its quantification
The import assay (100 µl) contained isolated mitochondria (50 µg of mitochondrial
protein), 3 pmoles of 5'-end [
P]-labeled and pre-aminoacylated tRNA, 10 µg of ScIDPs
(fraction 60%), and 10 ng of recombinant pre-MSK in the "import buffer", IB:
0.44 M mannitol
20 mM HEPES-KOH (pH 6.8)
20 mM KCl
2.5 mM MgCl
1 mM ATP
5 mM DTT
0.5 mM PMSF
0.1 mM DIFP
0.1 mM L-lysine (is dispensable for versions that do not require aminoacylation)
0.5 mM phosphoenol pyruvate (alternatively, 0.5 mM creatine phosphate)
4 units of pyruvate kinase (alternatively, 5 units creatine kinase)
The import assay was carried out at 30°C for 15-20 min, which corresponds to the
logarithmic phase of RNA uptake . Thereafter, the external RNA was removed by
addition of a mixture of nucleases (20 u/ml of micrococcal nuclease, 50 µg/ml of RNase A
and 25 u/ml of phosphodiesterase) followed by an incubation at 20°C for 5 min in the
presence of 1 mM CaCl
. The mixture was then diluted 5 times with the "RNase-stop" buffer
(BB with 2 mM of EDTA and 4 mM of EGTA) and mitochondria were harvested by
centifugation. To remove all traces of externally associated RNAs, mitoplasts were generated
by hypotonic shock: the mitochondrial pellet was suspended in 100 µl of BB and diluted 10
times in HEPES, 10 mM, pH6.8. After incubation for 10 min in ice, sucrose was added to
0.25M, mitoplasts were harvested by centrifugation and washed twice with BB. Mitoplasts
were then lysed in 1% SDS, 0.1 M sodium acetate (pH 4.8) and 0.05% diethyl pyrocarbonate
at 100°C for 1 min and mtRNA was phenol-extracted at 60°C for 5 min, placed in ice for 5
min and ethanol precipitated. RNAs were analyzed by denaturing gel-electrophoresis and
imported RNA was detected by scanning in a Phosphor-Imager (Fuji, Bas -2000) (Fig. 1b).
The amount of imported RNA was determined by comparison of the band density of
the protected RNA isolated from the mitoplasts after the import assay and an aliquot of the
input labeled RNA. To quantify import we always scanned the band corresponding to the full-
length RNA. Various controls (listed in the Table II) demonstrate that there was no
contamination with nuclear and cytosolic RNAs after RNase treatment and mitoplast
isolation. For fully aminoacylated tRK1, 2% of the added RNA was internalized. This amount
is in agreement with the import efficiency found in vivo, where 2-3% of the total cellular
tRK1 is associated with mitochondrial matrix . From one tRNA version to another the
efficiency vary among 0.1 and 5% [14, 17]. When the efficiency was less, we considered that
the RNA was not imported.
Among natural tRNA tested (we have used a dozen of individual yeast cytosolic
tRNAs, human tRNA mixture and individual bacterial tRNAs), only tRK1 was imported,
therefore this experimental system can model events occuring in vivo. On the other hand,
analysis of in vitro synthesized mutant versions of tRK1 and tRK2 [14, 16-18] demonstrated
1. selectivity of tRNA import can be thoroughly modified and mutant versions of non-
imported tRNAs can be imported if they contain "import determinants", localised in the
anticodon loop (C34) and aminoacceptor stem (G1:C72 pair and the U73 base) (see also
2. aminoacylation identity of the tRNA can be changed by introducing mutations in
sequence without affecting its import efficiency (tRNAs aminoacylable with Lys-,
Met- and Ala-tRNA synthetases were found to be imported);
3. imported cytoplasmic-type tRNAs are functional in mitochondrial translation.
Import of tRNA into yeast mitochondria in vivo
In vivo, tRK1 represents a major cytoplasmic tRNA
, essential for cytosolic
translation [12, 30, 31]. 2% of the total tRK1 is associated with mitochondrial matrix, which
is comparable to the normal content of other mitochondrial tRNAs.
The tRNA genes in eukaryotes contain an internal promoter recognized by the RNA
polymerase III, while flanking regions are less important, but can influence expression
efficiency. To express mutant versions of tRK1 and tRK2, we used the flanks of one of the
tRK1 gene copies, cloned in centromeric vectors of the pRS serie , which provided the
level of expression comparable to the normal expression of tRK1, though varying from one
version to another. The flank sequences used correspond to the gene encoded in the
chromosome VII and originally PCR-cloned from the Y109 plasmid :
5'-flank: GAGAGGTCAGATTTCCAATAACAGAATA 1st tRNA base....
3'-flank: ... 73rd tRNA base TTCTTTTTTTTTTTAAAACACGATG
To quantify the import efficiency, we used Northern analysis. In vivo import efficiency
was calculated as the ratio between the signal obtained with the total cellular RNA and
mitochondrially extracted one. To normalize the amount of mitochondrial RNA, hybridization
with host mitochondrial tRNA (tRK3, tRNA
) was used as control. The
oligonucleotide probes used to identify various RNAs are indicated in Table II. To search for
mutant tRNA versions, we used shorter oligonucleotides, optimized case-by-case, which do
not recognize the wild-type versions.
For analyses of tRNA import in vivo, the RNA was isolated with the Trizol reagent
(BRL) from the whole cells or from nuclease-treated mitochondria purified on sucrose-
gradient and after generation of mitoplasts. Treatment with nucleases and mitoplasting were
done as after the in vitro import assay. To check aminoacylation in vivo, we used acidic
conditions of RNA isolation. This protocol  reveled suitable to detect the levels of
aminoacylation by the lysine, the methionine and the alanine . Figure 3 shows an example
of such experiment. A version of tRK1 with the anticodon CAU, as mentioned above, can be
aminoacylated by the methionyl-tRNA synthetase in vitro (Fig. 3a). We expressed the same
version in vivo and found that the efficiency of its import was similar to that of tRK1 (Fig.
3b). Analyzing mitochondrial RNA by Northern hybridization after the electrophoresis in
acidic conditions clearly show a difference of migration of aminoacylated form for
lysynylated tRNAs and for the mutant one (Fig. 3c). Furthermore, the distance between the
bands corresponding to deacylated and aminoacylated versions are similar to that for tRNA
One can suggest therefore that the mutant tRNA is expressed and imported in vivo in its
methionylated form. The same approach was successfully applied to mutant versions of tRK1
and tRK2 that can be misacylated with alanine .
Import of RNAs into human mitochondria in vitro
As mentioned above, human mitochondria import 5S rRNA [20, 21], MRP RNA and
RNase P RNA , but do not import tRNAs. We have shown that isolated human
mitochondria can import in an artificial way yeast tRK1 and several of its mutant versions
. The features of this internalization are similar to those found previously for yeast
organelles. We next demonstrated that likely yeast they can specifically import human 5S
rRNA, which can model the in vivo import pathway . These results signify that the import
pathway is very flexible and can be re-introduced into one biological system from another.
Isolation of mitochondria
Mitochondria for import assays were isolated from either HeLa or HepG2 cells,
cultured in DMEM medium with non-essential amino acids, Earl's salts and L-Glutamate
(Sigma), buffered with sodium bicarbonate and supplemented with 10% of Foetal Calf Serum
at 37°C and 5% CO
. The cells were harvested by PBS-EDTA, washed twice with PBS and
disrupted either in a Dounce homogenizer (pestle A), or in a Warring blender at medium
speed 2 x 10 sec in BB (see above), the debris were removed by centrifugation at 2000g for 5
min at 4°C and mitochondria were harvested by centrifugation at 15000g for 20 min. To
purify mitochondria were either subjected to percoll self-forming gradient or to several cycles
of low- (2000g for 5 min) and high speed (11000g for 10 min) centrifugation in BB. These
approaches give similar extents of purity, as judged by criteria listed in Table I. Disruption of
the cells with the Warring blender lead to a higher output (100% of disrupted cells) and
permits to obtain mitochondria intact to >90% (as judged by the citrate synthase assay).
Mitochondria did not contain significant contamination by cytosolic membranes. Treatment
of mitochondria by nucleases was done only for isolation of mitochondrial RNA, but not to
prepare the organelles for import assays.
Isolation of human Import Directing Proteins (HmIDPs)
To isolate HmIDPs, HepG2 cells were harvested in PBS containing 1 mM EDTA,
washed with PBS, suspended in NPMD buffer (20 mM Na-Phosphate buffer, pH6.5, 150 mM
NaCl, 1 mM MgCl
, 5 mM DTT) containing the cocktail of protease inhibitors, and disrupted
by ultra-sounds (4 times 2.5 min at maximal frequency). Cellular debris were removed by
centrifugation (4000g, 10 min), nucleic acids were removed by polyethylenimine treatment
(see above), proteins were precipitated by ammonium sulfate (70% of saturation) and dialysed
against the NPMD buffer. HmIDPs were fractionated by differential ammonium sulfate
precipitation and fractions precipitating at 30, 40, 50, 60 and 70% of saturation were dialyzed
against NPMD or HKM buffer containing 50% of glycerol. The fraction 30-40% was used for
5S rRNA import assays and combination of fractions 30-40% and 50-60% - for tRNA import
Isolation of RNAs
Import substrates were isolated mostly in a similar way as for import into yeast
mitochondria. Synthetic RNAs were excised from a 40cm-long denaturing gel permitting
single base resolution, and were refolded by one cycle of heat denaturing - refolding in the
presence of 0.5 mM Mg
. We identified 5' terminal nucleotides of the tranzymes by P1
nuclease hydrolysis  with subsequent TLC . To verify 3'-termini of synthetic tRNAs,
we used aminoacylation assays. These controls permit to affirm that tRNA transcripts used in
import assays had predicted termini and were correctly folded.
To purify human and yeast 5S rRNA, human 5.8S rRNA and tRNAs, total RNA of
HepG2 cultured cells was prepared with Trizol-reagent (BRL) and separated on 10%
denaturing polyacrylamide gels. Bands corresponding to the needed RNA were excised and
extracted . RNAs were checked for purity by electrophoresis and Northern analysis.
Before import assays, all RNAs were fully 5'-end
P-labeled by T4 polynucleotide kinase,
gel-purified and refolded.
Up to the treatment with nucleases, the procedure was the same as for yeast
mitochondria. The artificial import of tRNAs could be directed by either ScIDPs or HmIDPs
(fractions described above). 5S rRNA import was better directed by HmIDPs, fraction 40%
. After treatment with a mixture of nucleases, mitoplasts were generated by treatment
with digitonin, at 100 µg per 1 mg of mitoprotein. After incubation for 15 min in ice,
mitoplasts were harvested by centrifugation and washed twice as described . Mitoplast
quality was controlled by Western analysis with antibodies against an outer membrane
protein, porin (Calbiochem) (Table I). In average, we observed a loss of 50-60 % of porin
with respect to intact mitochondria. Mitoplasts were then lysed in 1% SDS, 0.1 M sodium
acetate (pH 4.8) and 0.05% diethyl pyrocarbonate at 100°C, mtRNA was phenol-extracted at
60°C, separated by denaturing gel-electrophoresis and import was quantified by scanning in a
Phosphor-Imager (Fuji, Bas -2000). Figure 1b shows an example of import assays for a panel
of yeast tRNA
versions performed with either yeast (YPH499) of human (HepG2) isolated
mitochondria. When the assays were done in the presence of ScIDPs, similar selectivities of
import were observed for human and yeast organelles. On the other hand, when the assays
were done in the presence of HmIDPs, selectivity of import were slightly different , that
may be explained by different affinity of pre-MSK and its human counterpart to mutant tRNA
Quantification of RNA import in vitro and in vivo
As deduced by the densitometry of the Northern-hybridization signals obtained for an
internal RNA, it appears that generation of mitoplasts and nuclease treatment results in its
partial degradation (in our tests, the signal obtained for mitoplasts represented 60% of the
signal obtained for mitochondria). This effect is probably due to a partial disruption of
mitochondria during generation of mitoplasts. We therefore used the corresponding
coefficient (k=1.67) to estimate real import efficiencies .
To quantify 5S rRNA import in vivo, one can use densitometry of Ethidium Bromide
stained gels. The exact amount of the 5S rRNA in mitoplast preparation was determined in
comparison with an aliquot of pure RNA (we used either 5S or 5.8S rRNA). It appears that
0.9% of the total cellular 5S rRNA is associated with mitochondrial matrix .
The number of mitochondria per cell is variable depending on the cell cycle stage and
is estimated, in average, as 400 [39, 40]. Taking the total number of 5S rRNA molecules in
the cell as 3.6 x 10
, one can deduce that approximately 80 molecules are associated with
For in vitro import, by comparing by Phosphor-imaging signals corresponding to the
imported RNA and to an aliquot of labeled RNA that served as an import substrate, we could
calculate the import efficiency in vitro was 1±0.1% for 5S rRNA and 2±0.1% for tRK1 .
Such values are similar to these found in vivo for tRK1 in yeast  and for 5S rRNA in
human cells .
Translational activity of imported tRNAs
Two approaches, the first - in vitro and the second - in vivo, were used to determine if
the imported tRNAs were active in mitochondrial translation. The in vitro approach was
applied to both yeast and human mitochondria. We exploited the fact that one version of
tRK1, containing a mutation in the anticodon (U35A) could be misacylated by the methionyl-
tRNA synthetase [18, 42]. The principle of the approach was to import into isolated
mitochondria tRK1(CAU) aminoacylated with
S-methionine and to determine if the labeled
amino acid was incorporated into mitochondrial translation products.
The limiting point of these experiments is the amount and concentration of the labeled
amino acid charging the imported tRNA. A standard import assay was done with 3 pmoles of
labeled tRNA and the efficiency of import was 1-5%, depending on the RNA version.
tRK1(CAU) import efficiency was 2±0.1%. tRK1(CAU) can be aminoacylated to 50% by the
yeast methionyl-tRNA synthetase, both in vitro and in vivo . 10 fmoles of
delivered by the imported tRNA were not sufficient to detect mitochondrially synthesized
polypeptides. We therefore used higher amounts of mitochondria and of the charged tRNA.
The assay of import was performed in the volume of 1 ml of IB (see above) and contained 0.5
mg of mitochondrial protein and 200 pmoles of methionylated (to 50%) tRK1(CAU). Such
ratio does not increase significantly the import efficiency, but permits to deliver 10 pmoles of
S-methionine attached to the imported tRNA into the mitochondria. The import assay was
carried as described above and mitochondria were pelleted by a brief centrifugation. They
were next placed in conditions permitting in organello translation by suspension in the final
volume of 1 ml in:
Mannitol 600 mM
KCl 150 mM
ATP 4 mM
GTP 0.5 mM
Phosphoenol pyruvate 5 mM
Pyruvate kinase 10 units
α-Ketoglutarate 5 mM
Tris-HCl (pH7.2) 20 mM
Amino acid mixture (-Met) 0.1 mM each
Bovine serum albumine 3 mg
Cycloheximide 100-500 µg
Protein synthesis was carried out at 30°C for yeast and 37°C for human mitochondria.
As negative control, we used the yeast cytoplasmic tRNA
fully methionylated by
methionine; as a positive one, we used 10-50 pmoles of
S-methionine. After incubation non-
labeled methionine was added to 0.1 mM and the mixture was incubated for additional 5 min
to achieve translation of the mitochondrial polypeptides. In parallel, the same reactions were
performed in the presence of Chloramphenicol, at 0.5 mg/ml, which is sufficient to
completely abolish mitochondrial translation. After achivement of the chase reaction,
mitochondria were pelleted, suspended in the Laemmli sample buffer and the proteins were
analyzed by a standard 12% or gradient (10-20%) SDS gel-electrophoresis and
autordiography. Alternatively, one can check for incorporation of the labeled amino acid into
mitochondrial translation products by measuring total radioactivity in the TCA-precipitated
material. To this end, the aliquots of the "translation" suspension were individualized each 5
min, chase reactions were done separately for each aliquot, mitochondria were pelleted,
dissolved in 0.05% Triton-X100 and proteins were precipitated by addition of cold TCA to a
final concentration of 5%. After a 15 min heating at 70°C, the precipitates were collected by
suction on Millipore filters (0.45µm), which were then washed several times with cold 5%
TCA, once with ethanol, dried and the radioactivity was determined in the scintillation
counter. Finally, mitochondrial proteins can be solubilized in 0.1% Triton X100 after import
and in vitro translation and treated with antibodies against mitochondrially synthesized
polypeptides (we used antibodies against I-III subunits of cytochrome c oxidase from
Molecular Probes). Reaction with antibodies was performed for 1-2h at 8°C under shaking,
the immunocomplexes were precipitated by Protein-A Sepharose beads (30 min at 4°C) and
the precipitated material counted by Cerenkov. All these approaches applied to yeast and
human mitochondria, clearly demonstrated that tRK1(CAU) participated in the organellar
To test involvement of imported tRNAs in mitochondrial translation in vivo, we used
advantages of yeast genetics. For this purpose, a strain was designed, which contained a non-
sense mutation Ala
->stop (UAG) in the COXII gene (coded by the mitochondrial DNA).
This strain (HM4) was constructed by a biolistic transformation technique and was
characterized by a respiratory deficiency (inability to grow on media containing the glycerol
as the carbon source) [18, 43]. For suppression analysis, we constructed a version of tRK2
that contained two mutations in the anticodon (U34C and U36A), which results in recognition
by the mutant version the UAG stop codon and provides its import capacity (C34 being an
import determinant ). Additionally, two mutations were introduced in the aminoacceptor
stem (C3G and C70U, resulting in a G3:U70 base pair), which provide the version with an
alanine aminoacylation identity . After transformation, the yeast cells were screened for
the rescue of respiration (on glycerol medium YPEG) and for the presence of plasmid
carrying the mutant tRK2 gene (SC-Uracyl). It is noteworthy that only a minor part of the
selected cells were proven to manifest "true suppression", all the other being results of various
reversion events. In our screening with the tRK2(C34, A36, G3:U70) only 5% of the clones
growing onto the double-selective medium corresponded to suppression events. Introduction
of two additional mutations in the same tRNA (U1G and A72C) resulted in a higher output
(20%), probably due to a better import efficiency. To select 'true" suppressing clones, all the
growing cells were tested by plasmid loss technique, based on counterselection of the URA3
marker on media containing 5-Fluoro-orotate . The suppressing clones were characterized
by a loss of the glycerol growth after the plasmid loss. An additional control was to check for
the presence of the correct full-length Cox2p in the rescued clones by Western analysis. These
experiments permitted to demonstrate in vivo that imported cytosolic-type tRNAs are able to
participate in mitochondrial translation and to cure respiratory deficiency due to mutations in
mitochondrial DNA . On the other hand, this system of suppression provides a tool to
further study the mechanisms of import, for example to identify other import factors in yeast
by direct genetic screens.
Studying the involvement of pre-protein import in RNA import
Formation of low-ATP import intermediates
tRNA import into yeast mitochondria and 5S rRNA import into human mitochondria
revealed to depend on pre-protein translocation machinery. The working hypothesis is that
imported pre-proteins (pre-MSK for tRK1 and another non-identified pre-protein for 5S
rRNA) serve as carriers to target the RNAs towards the mitochondrial membranes and,
possibly, to translocate them across the membranes. If the mechanism of this cotranslocation
is still completely non-understood, the mechanisms of pre-protein translocation are studied in
much more details. One can therefore exploit this knowledge for studying involvement of pre-
protein import channel in RNA internalization by mitochondria.
One convenient system to block translocation intermediates on the mitochondrial
surface is to dissipate the electrochemical potential across the inner membrane in a reversible
way. This will result in a correct interaction between the pre-protein and the outer membrane
but will not lead to its translocation into the matrix. The pre-protein will then be blocked at
the outer membrane in a relatively stable manner. Such a reversible dissipation of the
membrane charge can be caused in "low-ATP conditions" . To this end, mitochondria
were pre-treated with apyrase (1 unit per ml, 10 min at 20°C), pelleted and the RNA import
assay was performed as described above, but the IB lacked ATP and import was performed in
the presence of 6 mM ADP and 20 µM oligomycin. We have done this assay with labeled
tRK1 and both yeast and human mitochondria and, in both cases, the import of the RNA was
arrested. On the other hand, the RNA remains associated with the mitochondria, probably via
the pre-protein (pre-MSK), which anchors in the outer membrane receptor and GIP (General
Insertion Pore) . This association may be checked by pelleting mitochondria, washing
with BB and achievement of import by suspension of mitochondria in IB with 1 mM ATP and
without oligomycin. Mild treatment with trypsin or proteinase K (20 µg/mg mitoprotein;
20°C, 10 min) are sufficient to completely remove the labeled RNA from the mitochondrial
surface. One can suggest to use a similar approach to characterize protein-protein and RNA-
protein contacts in GIP leading to internalization of the RNA.
Blocking the GIP by an non-imported pre-protein analogue
It is difficult to assume how the RNA/pre-protein complex penetrates into the
mitochondria via the same channel as pre-proteins. It may occur that the interaction is strong
but very local and is not disrupted during the translocation. This hypothesis still needs to be
verified. To be internalized by the mitochondria, the most of pre-proteins are at least partially
unfolded during the import, especially by the translocation apparatus localized in the inner
membrane and are re-folded in the matrix by interaction with a mitochondrial chaperone,
HSp60/Cpn10/mtHsp70 complexes . One can exploit this well-established fact to directly
prove and study involvement of the GIP and inner membrane pore in RNA translocation. We
constructed a synthetic pre-protein analogue that cannot be (even partially) unfolded (Fig. 4).
This protein represented a conjugate between the N-terminal 32 amino acid residues of a
mitochondrially targeted pre-protein, ornitine transcarbamoilase of rat (OTC) and a molecule
of Bovine Pancreas Trypsin Inhibitor (BPTI). The synthetic oligopeptide corresponding to the
mitochondrial targeting signal of OTC (with the following sequence:
MLSNLRILLNKALRKAHTSMVRNFRYGKPVQC) was coupled via the thiol group of its
single Cysteine residue to free NH
-groups of BPTI using the heterobifunctional reagent MBS
(m-Maleimidobenzoyl succinimide ester) and the conjugate was purified by gel-filtration on
Superdex 200. The resulting conjugate OTC-BPTI, due to the signal N-terminal peptide of
OTC, is directed towards the mitochondrial insertion pore GIP, but remains blocked within
the pore, since the BPTI molecule cannot be unfolded due to three disulfide bridges
stabilizing its 3D-structure [46, 47].
To block the GIP, isolated mitochondria were suspended in IB without IDPs and
labeled RNA, the conjugate was added and the mixture was incubated at 30°C (yeast
mitochondria) or 37°C (human mitochondria) for 10 minutes. In this case IB lacked DTT, in
order to avoid reducing of disulfide bridges of BPTI. Mitochondria were then harvested by
centrifugation, suspended in IB (also without DTT) and the import assay was performed as
described above. For a pre-protein import assay, we used
S-labeled pre-MSK synthesized in
a coupled transcription-translation system commercialized by Promega. We found that both
yeast and human mitochondria were able to pre-MSK in the same conditions as used for RNA
OTC-BPTI had a strong inhibitory effect on pre-MSK import in vitro, which became
detectable already at 10 nmoles of the conjugate per mg of mitoprotein and was complete at
100 nmoles/mg. The conjugate was also found to inhibit import of tRK1 into yeast
mitochondria and tRK1 and 5S rRNA into human mitochondria, independently of the origin
of IDPs used (ScIDPs or HmIDPs). One can note that BPTI by itself, used as a negative
control, had a weak inhibitory effect onto import of both pre-proteins and RNA (10-15% at
200 nmoles per mg of mitoprotein). It may be due to some contamination present in
commercial BPTI (we used the Sigma-Aldrich preparation), which interfered with import,
since Bovine Serum Albumin, used at the same concentration, did not cause any inhibition.
However, at 200 nmoles of OTC-BPTI, inhibition of pre-protein and RNA import was
complete, therefore, one can conclude that the inhibitory effect was specific to translocation
machinery. These experiments permitted to demonstrate that the pre-protein import channel is
involved in RNA delivery into mitochondria, both in the case of tRK1 import into yeast
organelles and in the case of 5S rRNA internalization by human mitochondria .
Prospects and potential biomedical application of RNA import
Most of currently available information on the mechanisms of RNA import into
mitochondria comes from experiments with yeast, human and trypanosomatid models.
Comparison of various requirements of RNA import in vitro [6, 19] exhibits both similar
features (dependence on ATP, the need of outer-membrane receptors, the membrane charge)
and differences (need of soluble factors and pre-protein import apparatus in yeast and human
cells and pre-protein independent import that requires an RNA-specific receptor in
trypanosomatids). In comparison with the pre-protein import mechanisms the available data
are still very limited. Several important questions are to be resolved for each particular
system. In yeast, by exploiting the various approaches described above, it would be necessary
to identify other (than pre-MSK) essential factors of tRK1 import and to determine the
mechanistic steps of the translocation process. In human cells similar in vitro approaches can
be used to identify factor(s) of 5S rRNA mitochondrial targeting.
Studies of RNA mitochondrial import in vivo, in yeast, as well as in other organisms
(various trypanosmatids, Tetrahymena) demonstrate that the specificity of this pathway can be
thoroughly modified and RNA species normally non-imported may be directed into the
organelle. Such flexibility enables us to propose the use of RNA import pathway to
complement mutations in human mitochondrial DNA, which are known as an important
source of neuromuscular diseases . This idea is strongly supported by the findings that
one can establish, in an artificial manner, tRNA import into isolated human mitochondria
(which normally do not import tRNAs) and to model in vitro 5S rRNA import [18, 19].
Furthermore, the possibility to cure mutations in mitochondrial DNA by suppressor tRNAs
addressed form the cytoplasm was demonstrated in yeast in vivo . Finally, it was
demonstrated that human cells can import a mutant version of 5S rRNA . An important
step towards the setting up models of gene therapy based on RNA import would be to
demonstrate that tRNAs with various aminoacylation identities and mutant forms of 5S rRNA
containing additional sequences can be imported in transgenic human cells. If this reveals to
be possible, one can imagine various manners to use RNA import for gene therapy purposes.
The most evident way would be to direct into mitochondria tRNAs replacing the
mutant tRNA species. For this moment, nearly 80 various pathogenic point mutations were
found in 17 out of 21 human mitochondrial tRNAs . Among the cases studied, the most of
such mutations result in poor aminoacylation, hypomodification or low stability, which results
in inhibition of mitochondrial protein synthesis [20, 50-55]. Our results of in organello
translation (see above) confirm that tRNAs of "cytoplasmic type" can participate in
translation in human mitochondria . Therefore, mutations in human mitochondrial tRNAs
could be suppressed by functional tRNAs imported from the cytoplasm.
Nearly 50 point mutations associated with human pathologies were detected in protein
coding genes localized in mitochondrial DNA . These mutations were localized in
NADH-dehydrogenase subunits 1,2,5-6, Cytochrome c oxidase subunits 1-3, ATPase 6 and
Cytochrome b genes. In the majority of cases, they represent missense mutations, only three
cases are nonsense mutations in COX1, COX3 and ND4 genes, one case described creates a
frameshift in CYTB gene. One can propose to cure these mutations by importing suppressor
tRNAs from the cytoplasm. Our study on yeast  clearly demonstrate that such approach is
realistic. The main problem to resolve would be to construct importable versions of tRNAs
with the needed aminoacylation identity.
Another way to exploit mitochondrial targeting of RNA would be to use imported
molecules as vehicles to deliver into the mitochondria additional oligoribonucleotides with
therapeutic activity. Taking into account stringent structural requirements of tRNA import in
yeast [14, 17], it would be difficult to use this way. On the other hand, it may occur that 5S
rRNA, normally imported into human mitochondria, may represent a better candidate to the
role of vector. Taking into account the availability of an in vitro import assay , it would
be important now to understand if the highly structured 5S rRNA molecule may be extended
without loosing its import capacity. If it would become possible to import this molecule
containing additional oligonucleotides as extensions or insertions, one can suggest using them
as agents interfering with mitochondrial gene expression.
It was recently hypothesised that certain pathological tRNA mutations have an
inhibitory effect onto mitochondrial translation by formation of non-productive complexes
with their cognate aminoacyl-tRNA synthetases . This dominant effect cannot be
bypassed in the potentially heteroplasmic environment of mitochondria. One can propose
therefore to verify this hypothesis by suppressing such mutations with short anti-sense
oligoribonucleotides complementary to the mutant tRNAs and delivered via RNA import
The second possibility would be to inhibit the replication of mutant mitochondrial
genomes without affecting the wild-type ones, thus "switching" the heteroplasmic equilibrium
from a pathological to a healthy one. Inhibition of replication of mutant mitochondrial DNA
was described in a cell-free systems, by using Peptide Nucleic Acids (PNA) oligonucleotides
complementary to mutant mitochondrial DNA molecules [57, 58]. However, treating living
cells with PNA did not provided inhibition . This effect may be explained by a low
accessibility of mitochondrial DNA during replication in vivo . To verify this hypothesis,
one can suggest to exploit RNA import pathway to deliver into mitochondria
oligoribonucleotides complementary to the mutant mitochondrial DNAs.
This work was supported by the CNRS, Université Louis Pasteur, Moscow State
University, AFM (Association Française contre les Myopathies), INTAS (International
Association for promotion of cooperation with scientists from the New Independent States of
the former Soviet Union), HFSP (Human Frontier Science Program) and RFBR (Russian
Foundation for Basic Research), N.E. was supported by the CNRS, the Université Louis
Pasteur and HFSP, H.K. was supported by UNESCO-MCBN and FEBS (Federation of
European Biochemical Societies), O.K. was supported by INTAS, FEBS and AFM, I.B. and
P.K. were supported by INTAS.
1 Anderson, S., Bankier, A.T., Barrell, B.G., de Bruijn, M.H., Coulson, A.R., Drouin, J.,
Eperon, I.C., Nierlich, D.P., Roe, B.A., Sanger, F., Schreier, P.H., Smith, A.J., Staden,
R. and Young, I.G. (1981) Nature 290, 457-65.
2 Foury, F., Roganti, T., Lecrenier, N. and Purnelle, B. (1998) FEBS Lett 440, 325-31.
3 Attimonelli, M., Altamura, N., Benne, R., Brennicke, A., Cooper, J.M., D'Elia, D.,
Montalvo, A., Pinto, B., De Robertis, M., Golik, P., Knoop, V., Lanave, C., Lazowska,
J., Licciulli, F., Malladi, B.S., Memeo, F., Monnerot, M., Pasimeni, R., Pilbout, S.,
Schapira, A.H., Sloof, P. and Saccone, C. (2000) Nucleic Acids Res 28, 148-52.
4 Pfanner, N. and Geissler, A. (2001) Nat Rev Mol Cell Biol 2, 339-49.
5 Schneider, A. (1994) Trends Cell Biol. 4, 282-286.
6 Schneider, A. and Marechal-Drouard, L. (2000) Trends Cell Biol 10, 509-13.
7 Tarassov, I.A. and Martin, R.P. (1996) Biochimie 78, 502-10.
8 Tarassov, I.A., Entelis, N. S., Martin, R. P. (1999) in Mitochondrial diseases, methods
and models (Lestienne, P., ed.), pp. 303-316, Springer, Heidelberg.
9 Adhya, S., Ghosh, T., Das, A., Bera, S.K. and Mahapatra, S. (1997) J Biol Chem 272,
10 Rubio, M.A., Liu, X., Yuzawa, H., Alfonzo, J.D. and Simpson, L. (2000) Rna 6, 988-
11 Entelis, N.S., Kolesnikova, O.A., Martin, R.P. and Tarassov, I.A. (2001) Adv Drug
Deliv Rev 49, 199-215.
12 Martin, R.P., Schneller, J.M., Stahl, A.J. and Dirheimer, G. (1979) Biochemistry 18,
13 Tarassov, I., Entelis, N. and Martin, R.P. (1995) J Mol Biol 245, 315-23.
14 Entelis, N.S., Kieffer, S., Kolesnikova, O.A., Martin, R.P. and Tarassov, I.A. (1998)
Proc Natl Acad Sci U S A 95, 2838-43.
15 Tarassov, I., Entelis, N. and Martin, R.P. (1995) Embo J 14, 3461-71.
16 Entelis, N.S., Krasheninnikov, I.A., Martin, R.P. and Tarassov, I.A. (1996) FEBS Lett
17 Kazakova, H.A., Entelis, N.S., Martin, R.P. and Tarassov, I.A. (1999) FEBS Lett 442,
18 Kolesnikova, O.A., Entelis, N.S., Mireau, H., Fox, T.D., Martin, R.P. and Tarassov,
I.A. (2000) Science 289, 1931-3.
19 Entelis, N.S., Kolesnikova, O.A., Dogan, S., Martin, R.P. and Tarassov, I.A. (2001) J
Biol Chem, in press.
20 Magalhaes, P.J., Andreu, A.L. and Schon, E.A. (1998) Mol Biol Cell 9, 2375-82.
21 Yoshionari, S., Koike, T., Yokogawa, T., Nishikawa, K., Ueda, T., Miura, K. and
Watanabe, K. (1994) FEBS Lett 338, 137-42.
22 Sikorski, R.S. and Hieter, P. (1989) Genetics 122, 19-27.
23 Rose, M.D., Winston, F., Hieter, P. (1990) Methods in yeast genetics. A laboratory
course manual., Cold Spring Harbor Laboratory, Cold Spring Harbor, NY.
24 Gasser, S.M., Daum, G. and Schatz, G. (1982) J Biol Chem 257, 13034-41.
25 Tarassov, I.A. and Entelis, N.S. (1992) Nucleic Acids Res 20, 1277-81.
26 Moczko, M., Ehmann, B., Gartner, F., Honlinger, A., Schafer, E. and Pfanner, N.
(1994) J Biol Chem 269, 9045-51.
27 Von der Haar, F. (1979) Methods in Enzymology LIX, 257267.
28 Fechter, P., Rudinger, J., Giege, R. and Theobald-Dietrich, A. (1998) FEBS Lett 436,
29 Kern, D., Dietrich, A., Fasiolo, F., Renaud, M., Giege, R. and Ebel, J.P. (1977)
Biochimie 59, 453-62.
30 Martin, R.P., Schneller, J.M., Stahl, A.J. and Dirheimer, G. (1977) Nucleic Acids Res
31 Martin, R.P., Sibler, A.P., Gehrke, C.W., Kuo, K., Edmonds, C.G., McCloskey, J.A.
and Dirheimer, G. (1990) Biochemistry 29, 956-9.
32 Nelbock, P., Stucka, R. and Feldmann, H. (1985) Biol Chem Hoppe Seyler 366, 1041-
33 Varshney, U., Lee, C.P. and RajBhandary, U.L. (1991) J Biol Chem 266, 24712-8.
34 Puranam, R.S. and Attardi, G. (2001) Mol Cell Biol 21, 548-61.
35 Silberklang, M., Gillum, A.M. and RajBhandary, U.L. (1977) Nucleic Acids Res 4,
36 Edqvist, J., Grosjean, H. and Straby, K.B. (1992) Nucleic Acids Res 20, 6575-81.
37 Maniatis, T., Fritsch, E. F., Sambrook, J. (1982) Molecular cloning: a laboratory
manual., Cold Spring Harbor Laboratory, Cold Spring Harbor, NY.
38 Kun, E., Kirsten, E. and Piper, W.N. (1979) Methods Enzymol 55, 115-8.
39 Attardi, G. and Ojala, D. (1971) Nat New Biol 229, 133-6.
40 Attardi, G., Costantino, P., England, J., Lederman, M., Ojala, D. and Storrie, B. (1973)
Acta Endocrinol Suppl 180, 263-93.
41 Lewin, B. (1983) Genes, John Wiley & Sons, New York, Chichester, Brisbane,
42 Aphasizhev, R., Senger, B. and Fasiolo, F. (1997) Rna 3, 489-97.
43 Mireau, H., Cosset, A., Marechal-Drouard, L., Fox, T.D., Small, I.D. and Dietrich, A.
(2000) J Biol Chem 275, 13291-6.
44 Hou, Y.M. and Schimmel, P. (1988) Nature 333, 140-5.
45 Sollner, T., Rassow, J. and Pfanner, N. (1991) Methods Cell Biol 34, 345-58.
46 Seibel, P., Trappe, J., Villani, G., Klopstock, T., Papa, S. and Reichmann, H. (1995)
Nucleic Acids Res 23, 10-7.
47 Vestweber, D., Brunner, J. and Schatz, G. (1989) Biochem Soc Trans 17, 827-8.
48 Wallace, D.C. (1999) Science 283, 1482-8.
49 Kogelnik, A.M., Lott, M.T., Brown, M.D., Navathe, S.B. and Wallace, D.C. (1998)
Nucleic Acids Res 26, 112-5.
50 Enriquez, J.A., Chomyn, A. and Attardi, G. (1995) Nat Genet 10, 47-55.
51 Flierl, A., Reichmann, H. and Seibel, P. (1997) J Biol Chem 272, 27189-96.
52 Hao, H. and Moraes, C.T. (1997) Mol Cell Biol 17, 6831-7.
53 Helm, M., Florentz, C., Chomyn, A. and Attardi, G. (1999) Nucleic Acids Res 27,
54 Jacobs, H.T. and Holt, I.J. (2000) Hum Mol Genet 9, 463-5.
55 Yasukawa, T., Suzuki, T., Ishii, N., Ohta, S. and Watanabe, K. (2001) Embo J 20,
56 Kelley, S.O., Steinberg, S.V. and Schimmel, P. (2000) Nat Struct Biol 7, 862-5.
57 Chinnery, P.F., Taylor, R.W., Diekert, K., Lill, R., Turnbull, D.M. and Lightowlers,
R.N. (1999) Gene Ther 6, 1919-28.
58 Chinnery, P.F., Taylor, R.W., Diekert, K., Lill, R., Turnbull, D.M. and Lightowlers,
R.N. (2000) Gene Ther 7, 813.
59 Muratovska, A., Lightowlers, R.N., Taylor, R.W., Turnbull, D.M., Smith, R.A., Wilce,
J.A., Martin, S.W. and Murphy, M.P. (2001) Nucleic Acids Res 29, 1852-1863.
60 Holt, I.J., Lorimer, H.E. and Jacobs, H.T. (2000) Cell 100, 515-24.
61 Heitzler, J., Marechal-Drouard, L., Dirheimer, G. and Keith, G. (1992) Biochim
Biophys Acta 1129, 273-7.
62 Wlodawer, A., Deisenhofer, J. and Huber, R. (1987) J Mol Biol 193, 145-56.
63 Gould, G.W., East, J.M., Froud, R.J., McWhirter, J.M., Stefanova, H.I. and Lee, A.G.
(1986) Biochem J 237, 217-27.
64 Ragan, C.I., Wilson, M. T., Darley-Usmar, V. M., Lowe, P. N. (1987) in
Mitochondria, a practical approach (Darley-Usmar, V.M., Rickwood, D., Wilson, M.
T., ed.), pp. 79-112, IRL Press, Oxford, Washington DC.
65 Robinson, J.B., Brent, L.G., Sumegi, B. and Spere, P.A. (1987) in Mitochondria: a
practical approach. (Darley-Usmar, V.M., Rickwood, D., Wilson, M. T., ed.), pp. 153-
170, IRL Press, Oxford, Washington DC.
66 Kunkele, K.P., Juin, P., Pompa, C., Nargang, F.E., Henry, J.P., Neupert, W., Lill, R.
and Thieffry, M. (1998) J Biol Chem 273, 31032-9.
67 Rickwood, D., Wilson, M. T., Darley-Usmar, V. M. (1987) in Mitochondria, a
practical approach (Darley-Usmar, V.M., Rickwood, D., Wilson, M. T., ed.), pp. 1-16,
IRL Press, Oxford, Washington DC.
Legends to figures
Figure 1. (a) Cloverleaf structures of the three yeast tRNAs
. The arrows indicate
various mutations introduced into tRK1 and tRK2 to study import determinants [14, 16, 17].
The corresponding numbers of the mutant versions are indicated in parentheses (trN, for T7
transcripts, rN, for transzyme constructions). Introduction of C34 or G1:C72 confer to tRK2,
normally non-imported, an import capacity (versions tr2 and tr93). Replacement of the
anticodon arm of tRK1 by that of tRK2 gives rise to the version tr7 which cannot be imported.
Simultaneous introduction of U34 and U1:A72 mutations in tRK1 (version r8) also abolish
import . Other mutations in D-arm, aminoacceptor stem and the anticodon loop result in
variation of import efficiency (versions tr25, tr40, tr8, tr9, tr30). tRK3, coded for by
mitochondrial DNA, is normally restricted to the mitochondrial compartment. However we
found that the corresponding T7 transcript (tr3) can be internalized by mitochondria in vitro
. (b) Examples of in vitro import assays. Autoradiographs of 5'-end labeled RNAs
protected from nucleases after incubation with isolated human (HepG2 cell line) or yeast
(YPH499 strain) mitochondria (M) in the presence of ScIDPs are presented. "Input", the
aliquot (1%) of labeled tRK1 used in the assay. tRNAs (tRK1, tRK2, yeast cyto-tRNA(Met),
tRNA(Phe) de E. coli), T7-transcripts (trN) or tranzymes (rN) used as substrates are indicated
above the autoradiographs. The tr13 was either lysynylated (K) or methionylated (M) before
the import. aa, aminoacylated with KRS, da, without pre-aminoacylation.
Figure 2. Hypothetical mechanism of tRK1 targeting into yeast mitochondria [7, 8,
11, 13, 15]. tRK1 is aminoacylated by the cytoplasmic lysyl-tRNA synthetase (KRS) and
becomes a target of competition between cytoplasmic translation factors (EF1 is indicated as
a possible candidate, although there is no direct experimental data proving its implication) and
the import factor, which is the cytoplasmic precursor of mitochondrial lysyl-RNA synthetase
(pre-MSK). This interaction, helped by other protein factors that are not yet identified, directs
the lysinylated form of tRK1 to the GIP (general insertion pore) via interaction with the outer
membrane receptor Tom20p. The tRNA and the pre-protein are next translocated into the
matrix in a ∆Ψ- and ATP-dependent manner. The first 32 amino acids of pre-MSK are
removed upon this translocation, while tRK1 remains aminoacylated and is able to participate
in mitochondrial translation .
Figure 3. Analysis of tRK1(CAU) import and aminoacylation in vivo. (a) Cloverleaf
structures of yeast tRK1 and tRNA
. Anticodon loops are outlined, the U35A mutation in
tRK1 is indicated by an arrow. One can see that anticodon loop of tRK1(CAU) is identical to
that of tRNA
. (b) In vivo detection of tRK1(CAU). Results of dot hybridization  with
oligonucleotide probes (bellow the autoradiographs) are presented. Above autoradiographs
the source of tRNA is indicated: C, cytoplasmic; M, mitochondrial. The first three blots
correspond to RNA prepared from the YPH499 yeast cells transformed with the tRK1(CAU)-
containing plasmid, the forth one - to non-transformed YPH499 cells, as indicated
("transformed" and "wild type"). (c) Analysis of the aminoacylation level of tRK1(CAU) in
vivo . Results of Northern-hybridization after electrophoresis in acidic conditions  are
presented. Hybridization probes are indicated bellow the autoradiographs. To see the
difference between aminoacylated and deacylated forms, the native tRNA preparation was
treated with 0.2 M Tris(HCl), pH 9.0 for 1 to 10 min, as indicated. "aa" and "da", positions of
the aminoacylated and deacylated forms, respectively.
Figure 4. Structure of OTC-BPTI conjugate used to block the GIP channel. N-
terminal sequence of OTC sequence is schematically shown as zigzag line. "+" indicate
positive residues of the signal peptide. BPTI was drawn from the known crystal structure 
by MOLSCRIPT program. Black circles indicate three disulfide bridges providing non-
unfoldable properties to the conjugate.
Table I. Characterization of isolated yeast mitochondria.
Northern hybridizationcytoplasmic tRNAs
U2 and U3 snRNAs
[Heitzler, 1992 #575]
Contamination with nuclear
and cytoplasmic fractions
2+-dependent ATPase<0.5% of total
[Gould, 1986 #992]
Enzymatic assaysmonoamine oxidase (OM)
citrate synthase (matrix)
>90% of activity
<5% in isotonic
[Ragan, 1987 #539]
[Robinson, 1987 #1121]
Northern hybidizationmitochondrial tRNAs>95% of total[Heitzler, 1992 #575]
Integrity of mitochondria
Western analysisATP-ADP carrier (OM, IS)
bands of predicted
[Sollner, 1991 #545]
by oxygen electrode
substrates, succinate, ADPRCR 3-5[Rickwood, 1987 #530]
Functionality of mitochondria
protein importpre-MSKp (TnT translation)protection of >30%
of added labeled
[Sollner, 1991 #545]
Table II. Oligonulcleotide probes used to detect RNA species by Northern hybridization.
tRNA oligonucleotide probeT°C of hybridization
yeast mt tRNAGly
yeast mt tRNAIle
yeast mt tRNAMet
yeast 5S rRNACATTATTTATTTATGAGA40
yeast U4 snRNACACAATCTCCGGACGAATCCTC30
yeast U6 snRNATCATCCTTATGCAGGG30
yeast U5 snRNAAAGTTCCAAAAAATATGGCAAGC30
human cyto tRNA
human mt tRNA
human mt tRNA
human mt tRNA
human 5S rRNACCCGACGTTGCTTAACTTC40
human U2 snRNAGAGTGGACGGAGCAAGCTCCTATTCCATCTCC50
human U3 snRNACGCTACCTCTCTTCCTCGTGGTTTTCGGTGCTCTACA55
U - A
U - A
G - C
U U U G
A A A C
G - C
A - U
G - C
A - U
A - U
A - U
G - C
G - C
G - C
C - G
C - G
U - G
U - A
U - A
C C C C
G G G G
C G C G
G C G C
G - C
C - G
C - G
U - A
G - C
C C C C
G G G G
C U C G
G A G C
U - A
U - G
C - G
A - U
U - A
G - C
G - C
G - C
U - A
C - G
tRK1, 1% of input
G - U
G - C
G - C
G - C
C - G
C - G
C - G
G A G C C
C U C G G
C G C G
G C G C
C - G
A - U
G - C
G - C
G - C
G - C
G - C
C - G
C - G
U - G
U - A
U - A
U - A
A - U
U - A
G - C
A - U
C G C G
G C G C
C C C C
G G G G