Skip Navigation


Human Molecular Genetics Advance Access originally published online on December 1, 2005
Human Molecular Genetics 2006 15(1):143-154; doi:10.1093/hmg/ddi435
This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrowOA All Versions of this Article:
15/1/143    most recent
ddi435v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (8)
Google Scholar
Right arrow Articles by Koulintchenko, M.
Right arrow Articles by Lightowlers, R. N.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Koulintchenko, M.
Right arrow Articles by Lightowlers, R. N.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

© The Author 2005. Published by Oxford University Press. All rights reserved.
The online version of this article has been published under an open access model. Users are entitled to use, reproduce, disseminate, or display the open access version of this article for non-commercial purposes provided that: the original authorship is properly and fully attributed; the Journal and Oxford University Press are attributed as the original place of publication with the correct citation details given; if an article is subsequently reproduced or disseminated not in its entirety but only in part or as a derivative work this must be clearly indicated. For commercial re-use, please contact: journals.permissions@oxfordjournals.org

Natural competence of mammalian mitochondria allows the molecular investigation of mitochondrial gene expression

Milana Koulintchenko1,2,3, Richard J. Temperley1, Penelope A. Mason1,{dagger}, André Dietrich2 and Robert N. Lightowlers1,*

1Mitochondrial Research Group, Institutes of Neuroscience and Cell and Molecular Bioscience, University of Newcastle upon Tyne, Newcastle upon Tyne NE2 4HH, UK, 2Institut de Biologie Moléculaire des Plantes du CNRS, Université Louis Pasteur, 12 Rue du Général Zimmer, 67084 Strasbourg, France and 3Siberian Institute of Plant Physiology and Biochemistry of the RAS, Lermontov St 132, PO Box 1243, 664033 Irkutsk, Russia

* To whom correspondence should be addressed. Tel: +44 1912228028; Fax: +44 1912228553; Email: r.n.lightowlers{at}ncl.ac.uk

Received November 14, 2005; Accepted November 20, 2005


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 RESULTS
 DISCUSSION
 MATERIALS AND METHODS
 REFERENCES
 
Respiration, a fundamental process in mammalian cells, requires two genomes, those of the nucleus and the mitochondrion (mtDNA). Mutations of mtDNA are being increasingly recognized in disease and may play an important role in the ageing process. Accepting the vital role of mtDNA gene products, our limited knowledge concerning the details of mitochondrial gene expression is surprising. This is, in part, due to our inability to transfect mitochondria and to manipulate their genome. There have been claims of successful DNA import into isolated organelles, but most reports lacked evidence of expression and no method has furthered our understanding of gene expression. Here, we report that mammalian mitochondria possess a natural competence for DNA import. Using five functional assays, we show imported DNA can act as templates for DNA synthesis or promoter-driven transcription, with the resultant polycistronic RNA being processed (5' and 3') and excised mt-tRNA matured. Exploiting this natural competence will allow us to explore mitochondrial gene expression in organello and provides the potential for mitochondrial transfection in vivo.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
RESULTS
 DISCUSSION
 MATERIALS AND METHODS
 REFERENCES
 
The mammalian mitochondrial genome (mtDNA) encodes 13 polypeptides, all of which are believed to be essential for coupling respiration to ATP synthesis, the process of oxidative phosphorylation. In addition, the genome encodes 22 mt-tRNAs and two mt-rRNAs, which function in intramitochondrial translation. The entire 16.569 bp of human mtDNA was the first full mammalian genome to be sequenced (1Go), revealing a remarkable economy and a region of around 1.1 kb that did not contain any genes. Our knowledge of the exact molecular features of mtDNA gene expression in vivo is, however, far from complete. In vitro transcription assays identified elements in this non-coding region (NCR) that could act as promoters for the synthesis of polycistronic RNA units from both heavy (H-) and light (L-) strands (2Go,3Go), now referred to as MT-HSP1 and -LSP, respectively (4Go), and RNA mapping studies were consistent with a second heavy strand promoter (HSP2) that led to initiation around the 5'-terminusof MT–RNR1 (5Go). In organelle DNA footprinting assays of mammalian mtDNA further revealed binding sites for one of the transcription factors, Tfam, close to these promoters (6Go).

Although it is likely that mapping of the promoters in the NCR are accurate, our current inability to transfect mitochondria with functional reporter constructs has prevented final confirmation for these cis-acting elements. The lack of such a method has meant that although there are numerous elements in mtDNA that are believed to play crucial roles in regulating mtDNA expression and replication, we have been unable to challenge these hypotheses directly. Further, the sequence dependence of post-transcriptional mechanisms regulating processing, maturation and RNA stability still remains unclear (7Go). Examples are too numerous to list, but perhaps one area where this limitation is most noticeable is the current debate concerning mtDNA replication. Two hypotheses have been drawn. The first describes an asynchronous method of replication that requires a unique site for L-strand replication, proposed to lie within a group of mt-tRNA genes (8Go,9Go). The second, the strand coupled method, does not need such a unique L-strand origin (10Go). A robust mitochondrial transfection process would help resolve this kind of issue. Although of enormous benefit to increasing our basic knowledge of mtDNA maintenance and expression, the increasingly apparent role of mtDNA mutations in disease (11Go) and their association with ageing (12Go) also make mitochondrial transfection a goal of the highest priority.

RNA essential for mtDNA replication and expression is believed to be imported into mammalian mitochondria (13Go–15Go). Further, by exploiting a cryptic tRNA import mechanism, nuclear-encoded yeast tRNALys derivatives have been shown to be imported into human mitochondria and partially suppress a mitochondrial translation defect in cultured cells (16Go). The natural import of DNA into mammalian mitochondria, however, has never been shown. There have been sporadic claims of DNA uptake into isolated yeast and mammalian mitochondria using the protein import pathway or electroporation (17Go,18Go), but in only one case there has been a claim of functionality for this DNA (19Go) and there have been no further reports of the successful use of this method.

Unlike these rather inconclusive reports, plant mitochondria can be successfully transfected either by electroporation or by natural methods (20Go–22Go), resulting in template-driven RNA synthesis. Intriguingly, the import of naked DNA into plant mitochondria is reminiscent of the process of natural competence for DNA uptake that has been recognized in bacteria for almost 80 years (23Go). Although a complete molecular mechanism has not been described for any species and its significance is controversial (24Go,25Go), it is well known that a substantial number of bacteria including varieties of proteobacteria are able to naturally take up naked DNA from the environment. As mitochondria are believed to have evolved from {alpha}-proteobacteria (26Go), it is formally possible that they may have retained a crude type of this DNA import pathway. Therefore, we set out to investigate whether mammalian mitochondria were able to take up DNA by natural processes.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
DISCUSSION
 MATERIALS AND METHODS
 REFERENCES
 
Mammalian mitochondria possess a natural competence mechanism
To assess whether DNA can be imported, mitochondria were isolated from rat liver by differential centrifugation or further subjected to isopycnic gradient centrifugation, prior to incubation with radiolabelled double-stranded DNA substrate containing the entire major rat mtDNA NCR flanked by MT-TF and-TP (FncrP, Fig. 1A). Following an 80 min incubation, DNase-resistant full-length FncrP (~0.1–0.2 fmol) could be readily extracted from equal amounts of crude (Fig. 1B, lane 2) and Percoll-purified (Fig. 1B, lane 3) mitochondria. When mitochondria were solubilized, FncrP was degraded (Fig. 1B, lane 5). Proteinase K treatment prior to incubation prevented uptake of FncrP (Fig. 1B, lane 7), consistent with import being mediated by a protein factor(s). DNA was potentially being protected from degradation through non-specific interactions with the outer membrane or by accessing the DNase-impermeable intermembrane space. To preclude this possibility, mitochondria were subjected to increasing osmotic shock to disrupt the outer membrane before DNase treatment (Fig. 1B, lanes 9–12). Even under conditions where over 90% of the intermembrane space marker adenylate kinase was lost, protected DNA was observed. To determine whether import was common to mammalian and especially human organelles, mitochondria were isolated from human HepG2 cells and similar import experiments were performed (Fig. 1C). FncrP was again protected (Fig. 1C, lanes 2–5), unless mitochondria were solubilized with Triton X-100 (Fig. 1C, lane 6). To assess the kinetics of import, a time course was performed (Fig. 1D). FncrP was incubated with rat liver mitochondria and nucleic acids wereextracted after indicated time points. Import starts veryrapidly, but half maximal import was between 15 and 20 min.



View larger version (45K):
[in this window]
[in a new window]
  		Figure 1. DNA can be imported into human and rat mitochondria. (A) Schematic representation of the full-length insert used as template for the production of DNA import substrates used in these studies. The insert is maintained in plasmid pgfp-FncrP-luc as described in Materials and Methods. The entire rat mitochondrial NCR is indicated in bold with flanking circles representing MT-TF and MT-TP, respectively (tF and tP). The resultant 1060 bp fragment was used for all import experiments in this figure (FncrP). The two reporter genes, MT–GFP and MT–LUC, had been altered to compensate for mitochondrial codon recognition and were cloned such that MT–GFP is driven by predicted HSP1 and 2 and MT–LUC by LSP. Human MT–TR (htR) has been incorporated immediately adjacent to MT–GFP and the construct is designed such that any full-length HSP transcript would be completed with the mitochondrial transcription termination factor binding site, mtf (46Go). The htR and mtf sequences are separated by the first 14 bp of the human MT-ND4L gene (indicated by asterisk). In contrast to MT–GFP, MT–LUC terminates with a sequence-determined 30 bp poly(A) region. Constructs used as DNA import substrates throughout this study are referred to in lower case except where including the single letter code for mt-tRNA genes (FncrP, gfp, Rgfp–FncrP–luc, etc.). (B) Rat mitochondrial import assays. Radiolabelled FncrP was used as substrate for import assays with either Percoll-purified(lane 2) or crude mitochondrial preparations, untreated (lanes 4 and 8), solubilized (lane 5) or protease shaved (lane 7) as described in Materials and Methods. For mitoplast generation (lanes 9–12), intermembrane space and matrix marker enzymes were assayed after osmotic shock for the indicated times. Following DNase-treatment, protected nucleic acids from all experiments were extracted, separated through a 1% non-denaturing gel and transferred to a nylon membrane for exposure. Free probe (4000 c.p.m.) is shown in lane 1. (C) Human mitochondrial import assays. FncrP was used for import assays under import conditions identical to (B). Increasing amounts of mitochondrial protein (lanes 2–4) and detergent lysis (lanes 5 and 6) were tested. (D) Kinetics of import. Rat mitochondria were incubated with FncrP and aliquots taken after the time points indicated (lanes 1–8). Mitochondria were DNase-treated and protected DNA was analysed as in (B).

 
To determine whether there was a size dependence of the import, various constructs were used ranging from 714 (gfp) to 3644 bp (Rgfp–FncrP–luc). All species were imported over a pH range of 7.0–7.8 (Fig. 2A). Mitochondria were also capable of importing single-stranded as well as double-stranded DNA molecules of varying sequence (Fig. 2B). Mitochondrial preparations can potentially be contaminated by bacteria (27Go). To rule out the possibility that DNA was being imported by bacteria, isolation buffer (IB) was supplemented with vancomycin (5 µg/ml) and ampicillin (90 µg/ml), which were retained during import, and aliquots following incubation were analysed as mentioned earlier or spread onto LB agar. DNA import was unaffected and no colonies were visible on plates (data not shown).



View larger version (43K):
[in this window]
[in a new window]
  		Figure 2. Characterization of DNA import. (A) Affect of DNA substrate on import. Three labelled substrates (0.7–3.6 kb) were assayed for import and are schematically represented. In each case, repeat import experiments were performed in incubation buffer at either pH 7.0 or pH 7.8. Following import and DNase treatment, DNA was extracted and analysed as in Figure 1. (B) Single-stranded DNA can be imported into mitochondria. Duplex FncrP was heated and rapidly cooled to generate single-stranded template, which was used in the standard DNA import experiment. Following DNase treatment and nucleic acid extraction, protected DNA was analysed as in Figure 1. (C–E) Affect of various effectors on DNA import. Rat (C and E) or human (D) mitochondria were pre-incubated for 10 min with a variety of effectors before programming import with FncrP. Following DNase treatment, DNA was extracted and analysed as in Figure 1. (C) Import in the presence of DMSO (0.5% v/v), DIDS (0.5 mM), CAT (10 µM) and cyanide (KCN, 0.5 mM). All import reactions were performed with addition of respiratory substrates glutamate (10 mM) and malate (1 mM), although import is only mildly lower in their absence (lane 3 cf. lane 4). DMSO is a necessary control, as DIDS stock solution is prepared in this solvent. (D) Import into human mitochondria is also inhibited by DIDS. (E) Import is unaffected by the alkylating reagent NEM (2 mM) or the uncoupler FCCP (50 µM). A slight increase in mobility was occasionally seen when comparing free and protected probe. This was due to minor variability in the purity of nucleic acids following extraction.

 
Import does not require respiration or a mitochondrial membrane potential
DNA appears to be imported into intact mitochondria. To optimize the import assays and to characterize the process, potential effectors were assessed (Fig. 2C and D). Respiratory substrates were not necessary for import (Fig. 2C, lane 2 cf. lane 3), although as a minor increase was always detected, glutamate and malate were routinely added to the import medium. Import reactions did not require ATP and were unaffected by the addition of any nucleotides (data not shown). Consistent inhibition was obtained with the anion channel blocker DIDS (28Go,29Go) (Fig. 2C and D, lanes 5 and 2, respectively). VDAC is the most prominent anion channel in mitochondria, and the suggestion that VDAC might be involved in the transport process is consistent with the mechanism in plant mitochondria (21Go). However, unlike the plant system, the adenine nucleotide translocase inhibitor carboxyatractyloside (CAT) had no effect, excluding the role of the permeability transition pore in the import process. The alkylating reagent N-ethylmaleimide (NEM) also showed no effect on import (Fig. 2E), implying that the transport mechanism did not require any vicinal –SH groups. Respiratory poisons such as KCN (Fig. 2C) or oligomycin (data not shown) did not inhibit DNA import and addition of the membrane potential uncoupler FCCP had no effect (Fig. 2E).

Imported DNA can associate rapidly with endogenous mtDNA
A slow migrating DNA species was visible following post-import analysis (HMW DNA, Fig. 3). This species was only formed when imported DNA contained homologous rat mtDNA sequence (see gfp alone, Fig. 2A) and did not form when mixed with isolated nucleic acids (Fig. 3A, lane 4). Inspection of ethidium bromide-stained and autoradiographed gels showed co-migration of HMW DNA with rat mtDNA isolated during mitochondrial extraction. To determine whether the imported DNA was integrating into the genome, nucleic acids were isolated following import of the radiolabelled DNA (FncrP) and DNA was cleaved by XbaI. Figure 3B shows the stained DNA following digestion and gel electrophoresis with the same gel exposed to PhosphorImage analysis (Fig. 3B, lane 2 cf. lane 4). Partially cleaved full-length mtDNA still remained, but the majority of the DNA was cut into the two expected fragments of 10.3 and 6 kb. As the entire FncrP homologous sequence is present in the 6 kb fragment, recombination at the site would have resulted in radiolabelling of this fragment. Although a strong signal can be detected, the species clearly has a slower migration than the 6 kb fragment. DNA import therefore does not lead to regular homologous recombination under our assay conditions, but the radiolabelled product possibly reflects formation of a triplex structure. The lack of recombination could suggest that conditions in the isolated mitochondria are not optimal or that recombination is a rare event. Consistent with the former possibility, we were able to increase the formation of the HMW species by import under state three respiration where ATP is being generated within the mitochondrion (data not shown).



View larger version (25K):
[in this window]
[in a new window]
  		Figure 3. Imported DNA interacts with mtDNA. Import was programmed with FncrP and following DNase treatment, protected DNA was assessed as in Figure 1. (A) A signal of high molecular weight DNA (HMWDNA) co-migrating with mtDNA is indicated. Radiolabelled DNA was also added to isolated mitochondrial nucleic acids (lane 4) or immediately prior to hot phenol extraction and precipitation (data not shown) to show that interaction is not non-specific. (B) Following extraction, mitochondrial nucleic acids were Xba1-digested. Rat mtDNA cleaves twice, generating a 10 279 bp fragment and an NCR-containing fragment of 6021 bp. Lanes corresponding to probe alone (lanes 1 and 3) or following import and Xba1 digestion (lanes 2 and 4) are shown both after ethidium bromide staining (lanes 1 and 2) and following exposure of the gel (lanes 3 and 4).

 
Imported DNA can act as a template for DNA synthesis
If this natural competence is to be exploited, it is necessary to show that imported DNA can be expressed. To address this, we first determined whether the imported DNA could perform template-driven DNA synthesis. Mitochondria were resuspended in media that have been optimized for in organellar nucleic acid synthesis (30Go). Two unlabelled DNA species were imported in separate experiments and reactions supplemented with [{alpha}-32P]dCTP. Following incubation, nucleic acids were extracted. To assess the levels of de novo synthesis, the extracted nucleic acids were hybridized to specific probes immobilized onto membranes. As shown in Figure 4, MT–RNR2 was generated, but at levels significantly below products from the NCR (presumably due to 7S DNA synthesis, lanes 1, 3 and 5). Imported DNA encompassing the reporter genes MT–LUC and MT–GFP (Rgfp–FncrP–luc) also acted as template for DNA synthesis (lane 6); however, DNA was also generated from gfp probe alone (lane 3). Thus, although imported DNA is clearly able to direct de novo DNA synthesis, we have yet to determine whether this represents DNA repair, nick translation, bona fide replication ora mixture of these processes.



View larger version (42K):
[in this window]
[in a new window]
  		Figure 4. Imported DNA substrates can direct DNA synthesis. The two unlabelled DNA substrates used in this experiment were PCR-generated from the full-length template pRgfp–FncrP–luc and are schematically represented. DNA substrates were imported independently into mitochondria and further incubated under DNA synthesis conditions as described in Materials and Methods. De novo-generated radiolabelled products were visualized by hybridization to DNA fragments corresponding to 16S rRNA (MT–RNR2), the NCR plus flanking tRNAs (FncrP) and the two reporters (MT–LUC and MT–GFP). DNA is synthesized by imported DNA substrates, but synthesis can occur independently of putative replication origins (lane 4 cf. lane 6).

 
Imported DNA requires promoter elements for transcription
To determine whether imported DNA could be transcribed, import was performed under conditions conducive to RNA synthesis in the presence of [{alpha}-32P]UTP and products hybridized as detailed earlier. Transcription required a respiratory substrate and was partially inhibited by the addition of the mtRNA polymerase inhibitor rifampicin (data not shown). Efficient mtDNA-mediated RNA synthesis was clear, with strong signals representing transcripts from rat MT–RNR1 (Fig. 5A, lanes 1, 3, 5 and 7). MT-RNR genes are believed to be transcribed from HSP1 and are synthesized at greater rates than the polycistronic unit containing the downstream mt-mRNAs (5Go). Consequently, a probe for rat MT–ND3 (an L-strand-encoded mt-mRNA) was also included on the membranes. Imported DNA substrate was transcribed, but only when the imported species contained the NCR (Fig. 5A, lanes 2 and 4 cf. lanes 6 and 8). No transcript was detected from gfp alone or when imported synchronously with the standard NCR-containing substrate, FncrP. However, when either reporter was ligated downstream from potential mtDNA promoters, transcription occurred. A signal corresponding to both MT–GFP and MT–LUC was produced when these genes flanked FncrP (Rgfp–FncrP–luc, lane 8). Transcription of MT–GFP, therefore, was being promoted from a region in FncrP, most likely from HSP(s), with MT–LUC being driven by light strand promoter (LSP).



View larger version (30K):
[in this window]
[in a new window]
  		Figure 5. Imported DNA substrates can direct RNA synthesis. Five unlabelled DNA substrates were used in these experiments, four generated from pRgfp–FncrP–luc and one from the inverted construct pluc–FncrP–gfpR. These are represented above the figure. De novo-generated transcripts were visualized by hybridization to DNA fragments corresponding to 12S rRNA (MT–RNR1), MT–ND3 and the two reporters as described for Figure 4. (A) Transcription requires the presence of FncrP. Following import of gfp alone (lane 2) or when co-imported with FncrP construct (lane 4), no signal can be detected. When the two are fused, MT–GFP is transcribed from HSP(s) (lane 6). Further, fusing MT–LUC downstream of an LSP also leads to a measurable signal (lane 8). Endogenous transcription was determined in each case by hybridization of rescued nucleic acids to components of the major polycistronic unit (MT–RNR1) and the full-length polycistronic unit (MT–ND3). (B) Comparison of promoter strength and transcript stability. In these two experiments, MT–GFP was expressed from HSP(s) and MT–LUC from LSP (lane 2) or the promoters were inverted in relation to the two reporters (lane 4).

 
The genetic environments of MT–GFP and MT–LUC differ in the large 3.6 kb template by two important factors. First, MT–GFP is transcribed from HSP(s) and secondly, the MT–GFP partial termination codon abuts onto a tRNA gene(human MT–TR) similar to most mt-mRNAs but unlike MT–LUC where the open reading frame is followed by a poly(A) sequence. To compare the relative signal intensities for each reporter when driven from either HSP(s) or LSP(s), FncrP was inverted as described in Materials and Methods. Import and transcription assays were performed with these two full-length DNA substrates and membranes were hybridized with extracted nucleic acids (Fig. 5B). An ~6.6-fold greater signal was measured for MT–GFP transcripts when compared with MT–LUC, when the former is driven by HSP(s) and the latter by the LSP (Fig. 5B, lane 2) after correcting for product size and assuming the reporter transcripts are complete. When FncrP is inverted, MT–GFP RNA signal is still 1.6-fold greater than MT–LUC. These figures are in good agreement with repeat experiments (data not shown) and confirm that the level of transcript initiated from HSP(s) is greater than that from LSP. Further, if the reporter transcripts were equally stable, the ratio of signal originating from the promoters would be inverted after switching the FncrP. If this is not the case, 6.6-fold would decrease to0.15-fold not to the observed 1.6-fold, showing that the MT–LUC transcript must be less stable than MT–GFP. Although uncertain, it is likely that the addition of MT–TR immediately downstream of MT–GFP might provide the correct environment for efficient processing and maturation, leading to increased stability.

Polycistronic RNA generated from imported DNA substrates can be processed and matured
Imported DNA substrates are transcribed, but natural polycistronic units are also processed and matured. In particular, all mt-tRNAs are matured by the addition of CCA trinucleotide at their 3'-termini (31Go,32Go). To determine whether these novel transcripts could undergo such critical modifications, we exploited human MT–TR that lies between MT–GFP and 14 bp of its natural downstream sequence (MT-ND4L) in Rgfp–FncrP–luc (Fig. 6). Human MT–TR has a strong sequence identity to the rat homologue but differs sufficiently to allow for species-specific amplification. Circularization of isolated RNA following import and nucleic acid extraction should provide a template to generate an amplicon with a region spanning the ligation break point, giving 22 bp that are not defined by primers P1 or P2. Four positions are informative regarding rat or human sequence. Therefore, following the protocol outlined in Figure 6, it would be possible to determine whether foreign mt-tRNA molecules are correctly processed (5' and 3') and whether maturation is possible.



View larger version (38K):
[in this window]
[in a new window]
  		Figure 6. Imported DNA can be transcribed and the transcript processed and matured. The complete 3.6 kb DNA template, Rgfp–FncrP–luc, was imported into mitochondria and following incubation for RNA synthesis, nucleic acids were extracted as detailed in Materials and Methods. DNA was removed and RNA circularized by T4 RNA ligase. To determine whether template-driven transcripts could be processed and matured, human MT–TR was analysed. Species-specific primer P1 was used to direct reverse transcription and used together with primer P2 to amplify across the terminal ligation site. Reverse transcription of a small circular RNA produces concatamers which when amplified will generate a ladder of products differing in size by 88–94 nt (see lane 2). The large size of the amplicon is due to additional sequences incorporated from the primers (Table 1). Primer dimers alone were generated without reverse transcription (lane1). Products were then gel-purified, cloned and sequenced. Sequences of 14 clones are given at the bottom of the figure with informative regions boxed and the 3'-termini indicating maturation.

 
Import was performed with Rgfp–FncrP–luc, RNA synthesized, rescued RNA circularized and RT–PCR amplified before cloning and sequencing of the resultant amplicon. The informative sequence of 14 clones is given in Figure 6. One clone was derived from the endogenous rat MT–TR gene. Of the 13 human sequences, eight were perfectly processed and matured, whereas the remaining five showed correct processing at the 5'-termini. Three clones were also processed accurately at the 3'-site, but CCA addition was incomplete. Two clones showed a single base inaccuracy at the 3'-site, but were also correctly matured. These data confirm that polycistronic RNA transcribed from imported DNA can be recognized by the natural mitochondrial processing and maturation machinery.

Natural competence can be used to map mtDNA promoterelements
The import of DNA substrates should enable the first direct identification of promoter elements that can drive the transcription of foreign genes in mitochondria. Previous studies have used various methods to predict two promoter elements for the rat NCR: an LSP at position 16211–16173 and an HSP at 16252–16289 (33Go). A second promoter element for the H-strand, HSP2, has been inferred, with its RNA product initiating around the start of MT–RNR1 (6Go). Assuming HSP2 is present, it is therefore likely that it lies between HSP1 and MT–RNR1. Constructs were made to localize promoters in the H-strand that could lead to transcription of MT–GFP. A control synthesis was first performed with MT–GFP transcription driven from the entire NCR (Rgfp–FncrP, Fig. 7A, lanes 1 and 2), and a signal comparable to that of the endogenous MT–ND3 was produced. When 681 bp of the entire 897 bp NCR was removed (Rgfp–Fprs), MT–GFP was still expressed well, even with slightly elevated levels (Fig. 7A, lanes 3 and 4 and unpublished results), showing HSP(s) were retained. When a further 176 bp was removed (Rgfp–Fprh2), only a very weak signal could be detected (Fig. 7A, lane 7, long exposure in lane 8), consistent with loss of the predicted HSP1. Although weak, the signal clearly represented de novo transcription, unlike from gfp alone, where no signal could be detected (Fig. 5A, lanes 2 and 4). A low efficiency promoter (HSP2) is therefore present in the 107 bp upstream of MT–GFP, which encompasses MT-TF and 40 bp of the NCR. Finally, to assess the relative efficiencies of HSPs alone as compared to LSP two constructs were made, where MT–LUC was transcribed independently (Luc–Fprs cf. FncrP–luc, Fig. 7B). Levels of import were similar (data not shown). To control for synthesis, signals were related to endogenous MT–ND3 transcription and showed a 4.6-fold greater level when driven from HSP(s).



View larger version (33K):
[in this window]
[in a new window]
  		Figure 7. Mapping of rat mtDNA promoters. DNA templates were designed to identify the promoters in the NCR. All templates used in this experiment are represented in the schematic above the images and were PCR-generated from the primer pairs in Table 1. Import and subsequent RNA synthesis were performed and analysed as detailed in Figure 5. For ease of reference, the predicted promoters retained in each construct are also listed. To confirm that import efficiency was similar for all constructs, DNA from aliquots of mitochondria post-import was analysed by Southern blot (data not shown). (A) MT–GFP is transcribed from the direction of the predicted HSPs when the NCR and flanking genes are present upstream (lane 2). Removal of MT-TP and 76% of the NCR distal to the reporter did not prevent strong transcription (lane 4). Removal of the predicted major promoter HSP1 led to diminution of the signal but not complete loss, consistent with the remaining 107 bp of upstream sequence containing a weak promoter (HSP2, lane 7). A longer exposure of both lanes 4 and 7 are shown in lanes 5 and 8, respectively. (B) HSPs alone drive greater levels of synthesis than the LSP. Synthesis of MT–LUC RNA was compared from the truncated DNA substrate Luc–Fprs containing both HSPs (lane 2) and LSP from the full-length NCR and flanking tRNAs (FncrP–Luc, lane 4). A longer exposure of both lanes 2 and 5 are shown in lanes 3 and 6, respectively.

 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
 DISCUSSION
MATERIALS AND METHODS
 REFERENCES
 
In this article, we report that isolated mammalian mitochondria can import DNA by a natural competence mechanism and that this DNA can act as a template for both DNA and RNA syntheses. Until now, a detailed knowledge of the mechanisms that underlie mammalian mtDNA expression has remained elusive due in major part to the absence of any transfection protocol. One of the main obstacles to this process has been to deliver a large polyanion such as a nucleic acid across two membranes and against a membrane potential. Attempts have been made to overcome this by utilizing either the protein import machinery or by electroporation (17Go,18Go). In the former, a pre-protein was conjugated to a short 24mer oligonucleotide (single or double-stranded) via a 3'-terminal cysteine. Access to a mitochondrial environment insensitive to phosphodiesterase required a membrane potential and could be prevented by pre-shaving the mitochondria with proteases. No function was demonstrable, but follow-up by Seibel et al. (34Go) showed that this technique could render a DNA fragment of up to 322 bp insensitive to added DNase. Collombet et al. optimized a procedure for electroporation of mitochondria, with plasmid DNA becoming resistant to added DNase. As this construct did not include any element that would be likely to drive DNA or RNA synthesis, the authors did not measure function of the potentially imported DNA, although electron microscopy and Southern blot analysis were consistent with DNA uptake. A similar method was used to electroporate RNA into rat liver mitochondria (35Go). Although RNA became resistant to RNase, no functionality could be demonstrated. This is clearly not the case with tRNA, as yeast tRNALys derivatives can functionally complement human MT-TK mutations when expressed in human cells (16Go). To date, we are aware of only one report of electroporated DNA being functional in isolated mammalian mitochondria (19Go). Transcription of foreign DNA was also shown when mitochondria isolated from cells devoid of mtDNA (rho0 cells) were electroporated. This is surprising,as the essential transcription factor Tfam is known to be unstable in the absence of mtDNA and is likely to be present in vanishingly low quantities (36Go). We are unaware of further reports for this method. Unlike mammalian mitochondria, there is strong supportive data for the successful transfection of isolated plant mitochondria (20Go). In addition, asmentioned earlier, isolated plant mitochondria have been shown to possess natural competence (21Go). Intriguingly, this mechanism shares some characteristics with the mammalian DNA import process but also differs in important areas. For example, the plant system is inhibited by CAT and is unable to import single-stranded DNA. These differences are currently being explored.

We believe our data clearly demonstrates the natural competence of mammalian mitochondria, but we have only begun to dissect the import process. The inhibition by DIDS suggests that VDAC might be involved in the transport of DNA across the outer membrane. This is consistent with the diameter of B-form DNA (~2 nm) and the size of the aqueous pore in VDAC (2.5–3 nm). Further, DNA can be transferred through reconstituted mammalian VDAC (37Go). It is more difficult to explain how such a large polyanion could be transported through the highly impermeable inner membrane. We were unable to show competition with any form of nucleotide or tRNA, or any inhibition with CAT, precluding the facile role of nucleotide transporters. The lack of inhibition by alkylating reagents also concurs with import through a pore rather than a transporter. More than 40 species of bacteria are known to be naturally competent, with many examples of proteobacteria taking up DNA from their environment (25Go,38Go). Studies have focused on the complex processes that underlie natural DNA import. The COM gene family has been implicated in Bacilli, with COMEC orthologues having been identified in numerous species as part of the DNA translocase (39Go). This protein acts as the aqueous channel that facilitates import, but no obvious orthologue is present in higher eukaryotes.

The data reported in our article show that isolated mammalian mitochondria can take up naked DNA and that the imported DNA is functional. Our inability to show that import is driven by a membrane potential or ATP is surprising and might suggest that the import process is somehow due to an artefact of mitochondrial isolation, but it is difficult to ascertain what artefact could result in such import. One intriguing implication of natural competence, therefore, is that mitochondria in vivo might be able to import DNA delivered to the organelle. Any delivery mechanism must be able to cross the plasma membrane and then target mitochondria where the cargo needs to be released. There is an ideal candidate that satisfies most of these criteria. DQAsomes are lipophilic cations that can be condensed with DNA, access the cytosol and localize to mitochondria as a consequence of their charge (40Go,41Go).

Why have mammalian mitochondria retained such a mechanism? Clearly, first there has been an evolutionary migration of genes from the evolving endosymbiont to the nucleus and wehave not determined whether the transport we see is unidirectional. A second prospect is that the import mechanism has become tailored to the transport of other similar molecules but still retains the capability of DNA import. Third, DNA import might have been retained as a monitor for cytosolic DNA invasion. This is a particularly intriguing possibility, as mitochondria are known to play an important role in apoptosis. Finally, although surprising, it is interesting to note that an equally unexpected retention of a tRNA import mechanism has also been reported in human mitochondria (16Go).

The aim of our study was to investigate the natural competence of mammalian mitochondria, to show unequivocally that imported DNA was functional and to take advantage of this process for defining the roles of specific mtDNA sequences in organello. Our report is intended to highlight the potential of this approach for investigating fundamental questions of mitochondrial gene expression. Our constructs contain two modified reporter genes whose transcripts can also be used to assess in organellar translation. For these first studies, we have focused on demonstrating that naked DNA can be imported into isolated mammalian mitochondria and that the imported DNA is functional. We have defined functionality of imported DNA as being able to act as template for nucleic acid synthesis and have limited our post-transcriptional studies to demonstrating RNA processing and maturation. For clarity of this first report, we have specifically excluded any reference to translation of these transcripts. Indeed, we have yet to determine whether the resultant transcripts can be translated in organello, although we intend to explore this possibility in due course.

The work to date has led to several important observations. First, we have demonstrated RNA synthesis from a foreign DNA substrate in isolated organelles. Human MT–TR is faithfully 5'-processed from larger polycistronic transcripts and 3'-processing is accurate within two residues. All processedspecies examined showed evidence of maturation. The 5'-terminus of MT–TR abutted the foreign MT–GFP, suggesting that accurate processing is consistent with the original speculation that folding of the tRNAs acts as punctuation points inpolycistronic RNA units (31Go). Second, imported DNA templates appear to transcribe reporter genes at greater levels from HSPs than from LSP (4.6-fold, Fig. 7B). This approximates to the 2–3-fold weaker induction reported from human LSP (42Go) and reinforces the suggestion that reconstituted and in vitro transcription systems lack an important factor(s), as this differential is lost in these systems (43Go). Thirdly, a low level HSP, HSP2, has been localized to within 107 bp upstream from the start of rat MT–RNR1. S1 mapping and kinetic studies have previously identified a species initiating from around the start of MT–RNR1 in various species, and methylation studies have indicated that a promoter element may lie within MT-TF (6Go). Finally, it is known that mt-RNAs can be very rapidly degraded and post-transcriptional processes can play critical roles in mitochondrial gene expression. We have shown that MT–LUC RNA is more unstable than MT–GFP (Fig. 7B). In summary, natural competence of mammalian mitochondria is a striking observation. Although we are currently unable to describe the molecular mechanisms underlying import, the potential for investigating mitochondrial gene expression processes is clear.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
 DISCUSSION
 MATERIALS AND METHODS
REFERENCES
 
Isolation of mitochondria
Mitochondria were prepared from the livers of Wistar rats, 200–250 g, dispatched by cervical dislocation and used at most 4 h postmortem. The liver was transferred directly into 25 ml of ice-cold IB [220 mM mannitol, 70 mM sucrose, 10 mM HEPES–KOH pH 7.4, 1 mM MgCI2, 1 mM EGTA, 1% (w/v) BSA]. All of the following steps were implemented on ice in a 4°C room. The liver was minced and suspended in 30 ml of IB, PMSF (0.33 mM) and homogenized by six strokes of a Potter–Elvehjem, glass–teflon homogenizer. Following centrifugation (400g, 4 min), the pellet was discarded and the supernatant centrifuged at 8600g for 10 min. The mitochondrial pellet was carefully resuspended and layered onto a Percoll gradient for further purification or resuspended in 4.5 ml IB and the differential centrifugation repeated. Mitochondria were finally washed and resuspended in IB minus BSA prior to estimating concentration by Bradford assay. For Percoll purification, mitochondria were layered onto 3.9 ml 50% (v/v) Percoll/gradient buffer [500 mM sucrose, 20 mM Mops–KOH pH 7.2, 10 mM EGTA, 2% (v/v) ethanol] with a cushion of 60% (w/v) sucrose and subjected to 100 000g (rmax) for 30 min. Mitochondrial fraction was removed and 4 vol of 0.5x gradient buffer added to allow harvesting and final suspension. For the isolation of human mitochondria, human HepG2 cells (1–2.5x108) were grown to ~80% confluency, stripped from flasks, washed in PBS and resuspended in 0.6 M mannitol, 10 mM Tris–HCl, 1 mM EGTA, 0.1% (w/v) BSA and 1 mM PMSF before homogenization by 15 strokes of a hand-held glass–teflon homogenizer. Mitochondria were prepared by repeat differential centrifugation as detailed earlier, except the high-speed pass was at 15 000g.

Generation of vectors and templates for mitochondrial DNA import
A fragment incorporating basepairs 15 330–67 of the rat mitochondrial genome (44Go) was PCR-amplified with primers SpeI NCR-F and NcoI NCR-R (Table 1) using standard techniques. This PCR product, containing the entire NCR and the natural flanking rat MT-TF and -TP, was then cleaved with SpeI and NcoI and cloned into the pBSSK+ based pcoxIII-Luc plasmid (45Go), yielding an intermediate plasmid pBS-Luc. Separately, a 920 bp fragment containing genes encoding mitochondrial GFP (MT–GFP), human MT-TR/partial ND4L [human mtDNA basepairs 10 405–10 482 (1Go)] and a final 36 bp containing the human mTERF binding site (human mtDNA basepairs 3227–3262) was amplified from pMag-1 (kindly provided by P. Seibel, University of Leipzig, Germany) with primers GFP-NCRL-F and GFP-R (Table 1) and cloned into the Spe1 site of pBS-Luc. The correct orientation of the cloned insert was confirmed, with the final plasmid pgfp-NCR-luc constructed to transcribe MT–GFP from HSP(s) and MT–LUC from LSP. To generate a second plasmid where the orientation of NCR and flanking mt-tRNAs were switched with respect to the two reporter genes, primers NcoI NCR-F and SpeI NCR-R were used (Table 1). Following PCR, the amplicon was cut with SpeI/NcoI and ligated into NcoI/partial SpeI digested pgfp-NCR-Luc to produce vector pluc-NCR-gfp.


View this table:
[in this window]
[in a new window]
  		Table 1. Oligodeoxynucleotides used for generating constructs
 
Production of DNA substrates for mitochondrial import
All DNA substrates for import, either radiolabelled or unlabelled, were PCR generated from the permutation of primers listed in Table 1, using one of the constructs described earlier as template. The main probe used for import assays was the 1060 bp DNA fragment encompassing the rat mtDNA NCR and flanking genes (FncrP). Amplification was from 10 ng of pgfp-NCR-luc, 0.5 µM primers Spe1 NCR-F/Nco1 NCR-R in a standard 30-cycle PCR reaction. To obtain a radioactive probe, 50–100 ng of this PCR product was transferred to a new reaction mixture with the same primer pair and cold dCTP replaced by 50 µCi of [{alpha}-32P]dCTP (3000 Ci/mmol). Following a single cycle with a 10 min elongation, 0.2 mM cold dCTP was added prior to a final 5 min elongation. A similar protocol was used to generate all radiolabelled probes, with the primers substituted as required. All unlabelled probes were generated by a standard 30-cycle PCR amplification using the relevant primer pairs as detailed (Table 2). For single-stranded template production, duplex template was heated at 95°C for 5 min and immediately chilled on ice.


View this table:
[in this window]
[in a new window]
  		Table 2. DNA constructs used in import experiments
 
Mitochondrial import assays and manipulations
A standard reaction contained 400 µg (2 mg/ml) mitochondrial protein suspended in import buffer (220 mM mannitol, 70 mM sucrose, 20 mM Tris–HCl pH 7.4, 1 mM EDTA, 1 mM EGTA, 10 mM glutamate, 1 mM malate). Import was initiated with 5–10 ng of 32P-labelled DNA (100–200 000 c.p.m.) and the reaction maintained at 30°C for 80 min under mild shaking. Following addition of 150 µg DNaseI; 10 mM MgCl2, the incubation was continued for 20 min. Mitochondria were subsequently washed thrice in l ml IB; 10 mM EDTA/EGTA. Nucleic acids were extracted from mitochondrial pellets after resuspension in 1 vol of 10 mM Tris–HCl pH 7.5, 0.15 M NaCl, 1 mM EDTA and 1% (w/v) SDS with an equal volume of phenol:chloroform at 55°C for 5 min. Following centrifugation at 20 000g for 10 min, nucleic acids were recovered from the aqueous phase by ethanol precipitation in 0.3 M NaOAc pH 5.2, fractionated by electrophoresis on a 1% (w/v) agarose gel and transferred onto a nylon membrane (Hybond N, GE Healthcare) for phosphorimage analysis followed by autoradiography. In all cases, the efficiency of total mitochondrial nucleic acids extraction from each aliquot following import was confirmed by comparing ethidium bromide staining in each lane.

To produce mitoplasts by osmotic shock after DNA import, mitochondria were harvested (5 min at 10 000g), resuspended in 1 ml of 5 mM Tris–HCl pH 7.5 and incubated on ice for the indicated time periods. Following addition of an equal volume of 2x IB without BSA, mitoplasts were harvested before resuspension in import buffer prior to DNase treatment and nucleic acid extraction as mentioned earlier. For mock treatment, IB lacking BSA was used instead of 5 mM Tris–HCl. Equal volumes of supernatant and resuspended organelles from each time point were assayed for intermembrane space (adenylate kinase) and matrix (citrate synthase) enzymatic markers as described (35Go). For mitochondrial solubilization, 1% (v/v) Triton X-100 was added after import and before DNase treatment. To shave mitochondrial outer membrane proteins prior to import, 20 µg of proteinase K was added to 400 µg mitochondrial protein for 30 min at 4°C in import buffer and the reaction terminated by the addition of 1 mM PMSF.

Expression assays for imported DNA templates
To analyse the expression of the imported DNA, 500 ng of PCR-generated unlabelled DNA template was used in each import reaction with either 2 mg (RNA synthesis) or 1 mg (DNA synthesis) mitochondrial protein. To minimize manipulations, all mitochondrial import and expression assays (500 µl final volume) were performed in base transcription medium (75 mM mannitol, 25 mM sucrose, 100 mM KCl, 10 mM Tris–HCl, 10 mM KH2PO4, 0.05 mM EGTA, 10 mM glutamate, 1 mM malate pH 7.4) described in (30Go). After import, the medium was supplemented with 10 mM ADP, 10 mM glutamate, 2.5 mM malate, 1 mg/ml BSA, 5 mM MgCl2 and 40 µCi [{alpha}-32P]UTP (GE Healthcare 800 Ci/mmol) and continued for another 3 h at 37°C under rotation. DNase (100 µg) was added for the final 30 min of incubation. Mitochondria were then harvested and washed twice in 1 ml of 10% (v/v) glycerol, 10 mM Tris–HCl, pH 6.8 and 0.15 mM MgCl2 (30Go) to remove unincorporated nucleotides. Nucleic acids were extracted as described earlier and ethanol precipitated.

For analysis of expressed products, DNA fragments (1 µg) PCR-amplified from a variety of genes using the primers detailed in Table 1 were separated by electrophoresis through a 1% (w/v) agarose gel and membranes prepared by Southern transfer. Hybridizations were performed at 42°C in minimal amounts of binding buffer [6x SSPE, 50% (v/v) deionized formamide, 5x Denhardt solution, 0.5% (w/v) SDS, 100 µg/ml salmon sperm DNA] with the entire nucleic acid extract for 72 h, followed by two sequential 10 min washing with 2x, 1x and 0.2x SSC; 0.5% (w/v) SDS at room temperature (2x) and 42°C (1x and 0.2x). Membranes were then exposed to PhosphorImage screen, followed by autoradiography.

For DNA synthesis, the same procedures were used after mitochondrial uptake, but the medium was further supplemented with dATP, dGTP and TTP (500 µM each) and radiolabelled UTP was substituted with 40 µCi [{alpha}-32P]dCTP (3000 Ci/mmol). Import was less efficient in the base transcription medium, but import efficiencies for all experiments were monitored by extracting nucleic acids from 50 µl aliquots prior to expression and subjecting the unlabelled DNA to Southern transfer and hybridization with the relevant radiolabelled probe.

RNA-circularization assays
Nucleic acids were extracted from mitochondria after using the Rgfp–FncrP–luc construct in the import-transcription procedure. DNA was removed by two successive RNase-free DNase treatments and remaining RNA was ligated with 20 U of T4 RNA ligase (Promega) for 3 h at 37°C in the recommended buffer. Following phenol extraction and precipitation, circularized RNA was used as a template for reverse transcription with P1 primer (2 pmol) according to manufacturer's instructions.

Two microlitres were then used in a standard 30-cycle PCR reaction with primers P1 and P2 (Table 1). Following amplification, the products were analysed on 4% (w/v) agarose gel (NuSieve GTG, Cambrex), DNA fragments corresponding to single and double copies of amplified cDNA were eluted by melting at 70°C and following phenol/chloroform extraction, cloned directly into pGEM-T Easy (Promega). Resultant plasmids were sequenced using M13-For universal primer.


   ACKNOWLEDGEMENTS
 
Professor Peter Seibel (University of Leipzig, Germany) is very gratefully thanked for his kind gift of pMAG-1. Paul Smith is thanked for help with the original experiments and Zofia Chrzanowska-Lightowlers for constructive criticism of the article. This work was supported by Wellcome Trust Grant 075536 (R.N.L.), the European Commission (QLG1-Ct-2001-00966, A.D. and R.N.L.) and the Royal Society (A.D. and R.N.L.). Funding to pay the Open Access publication charges for this article was provided by The Wellcome Trust.

Conflict of Interest statement. The authors have no conflict of interest to declare.


   FOOTNOTES
 
{dagger} Present address: Laboratory of Molecular Gerontology, National Institute on Aging, National Institutes of Health, Baltimore, MD 21224, USA. Back


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
 DISCUSSION
 MATERIALS AND METHODS
 REFERENCES
 

  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. et al. (1981) Sequence and organization of the human mitochondrial genome. Nature, 290, 457–465.[CrossRef][Medline]

  2. Chang, D.D. and Clayton, D.A. (1984) Precise identification of individual promoters for transcription of each strand of human mitochondrial DNA. Cell, 36, 635–643.[CrossRef][ISI][Medline]

  3. Bogenhagen, D.F., Applegate, E.F. and Yoza, B.K. (1984) Identification of a promoter for transcription of the heavy strand of human mtDNA: in vitro transcription and deletion mutagenesis. Cell, 36, 1105–1113.[CrossRef][ISI][Medline]

  4. Brandon, M.C., Lot, M.T., Nguyen, K.C., Spolim, S., Navathe, S.B., Baldi, P. and Wallace, D.C. (2005) Mitomap—a human mitochondrial genome database. http://www.mitomap.org.

  5. Montoya, J., Gaines, G.L. and Attardi, G. (1983) The pattern of transcription of the human mitochondrial rRNA genes reveals two overlapping transcription units. Cell, 34, 151–159.[CrossRef][ISI][Medline]

  6. Cantatore, P., Daddabbo, L., Fracasso, F. and Gadaleta, M.N. (1995) Identification by in Organello footprinting of protein contact sites and of single-stranded DNA sequences in the regulatory region of rat mitochondrial DNA. Protein binding sites and single-stranded DNA regions in isolated rat liver mitochondria. J. Biol. Chem., 270, 25020–25027.[Abstract/Free Full Text]

  7. Gagliardi, D., Stepien, P.P., Temperley, R.J., Lightowlers, R.N. and Chrzanowska-Lightowlers, Z.M. (2004) Messenger RNA stability in mitochondria: different means to an end. Trends Genet., 20, 260–267.[CrossRef][ISI][Medline]

  8. Wong, T.W. and Clayton, D.A. (1985) In vitro replication of human mitochondrial DNA: accurate initiation at the origin of light-strand synthesis. Cell, 42, 951–958.[CrossRef][ISI][Medline]

  9. Clayton, D.A. (1992) Transcription and replication of animal mitochondrial DNAs. Int. Rev. Cytol., 141, 217–232.[ISI][Medline]

  10. Yasukawa, T., Yang, M.Y., Jacobs, H.T. and Holt, I.J. (2005) A bidirectional origin of replication maps to the major noncoding region of human mitochondrial DNA. Mol. Cell, 18, 651–662.[CrossRef][ISI][Medline]

  11. McFarland, R., Taylor, R.W. and Turnbull, D.M. (2002) The neurology of mitochondrial DNA disease. Lancet Neurol., 1, 343–351.[CrossRef][ISI][Medline]

  12. Trifunovic, A., Wredenberg, A., Falkenberg, M., Spelbrink, J.N., Rovio, A.T., Bruder, C.E., Bohlooly, Y.M., Gidlof, S., Oldfors, A., Wibom, R. et al. (2004) Premature ageing in mice expressing defective mitochondrial DNA polymerase. Nature, 429, 417–423.[CrossRef][Medline]

  13. Li, K., Smagula, C.S., Parsons, W.J., Richardson, J.A., Gonzalez, M., Hagler, H.K. and Williams, R.S. (1994) Subcellular partitioning of MRP RNA assessed by ultrastructural and biochemical analysis. J. Cell Biol., 124, 871–882.[Abstract/Free Full Text]

  14. Puranam, R.S. and Attardi, G. (2001) The RNase P associated with HeLa cell mitochondria contains an essential RNA component identical in sequence to that of the nuclear RNase P. Mol. Cell. Biol., 21, 548–561.[Abstract/Free Full Text]

  15. Magalhaes, P.J., Andreu, A.L. and Schon, E.A. (1998) Evidence for the presence of 5S rRNA in mammalian mitochondria. Mol. Biol. Cell, 9, 2375–2382.[Abstract/Free Full Text]

  16. Kolesnikova, O.A., Entelis, N.S., Jacquin-Becker, C., Goltzene, F., Chrzanowska-Lightowlers, Z.M., Lightowlers, R.N., Martin, R.P. and Tarassov, I. (2004) Nuclear DNA-encoded tRNAs targeted into mitochondria can rescue a mitochondrial DNA mutation associated with the MERRF syndrome in cultured human cells. Hum. Mol. Genet., 13, 2519–2534.[Abstract/Free Full Text]

  17. Vestweber, D. and Schatz, G. (1989) DNA-protein conjugates can enter mitochondria via the protein import pathway. Nature, 338, 170–172.[CrossRef][Medline]

  18. Collombet, J.M., Wheeler, V.C., Vogel, F. and Coutelle, C. (1997) Introduction of plasmid DNA into isolated mitochondria by electroporation. A novel approach toward gene correction for mitochondrial disorders. J. Biol. Chem., 272, 5342–5347.[Abstract/Free Full Text]

  19. Yoon, Y.G. and Koob, M.D. (2003) Efficient cloning and engineering of entire mitochondrial genomes in Escherichia coli and transfer into transcriptionally active mitochondria. Nucleic Acids Res., 31, 1407–1415.[Abstract/Free Full Text]

  20. Farre, J.C. and Araya, A. (2001) Gene expression in isolated plant mitochondria: high fidelity of transcription, splicing and editing of a transgene product in electroporated organelles. Nucleic Acids Res., 29, 2484–2491.[Abstract/Free Full Text]

  21. Koulintchenko, M., Konstantinov, Y. and Dietrich, A. (2003) Plant mitochondria actively import DNA via the permeability transition pore complex. EMBO J., 22, 1245–1254.[CrossRef][ISI][Medline]

  22. Placido, A., Gagliardi, D., Gallerani, R., Grienenberger, J.M. and Marechal-Drouard, L. (2005) Fate of a larch unedited tRNA precursor expressed in potato mitochondria. J. Biol. Chem., 280, 33573–33579.[Abstract/Free Full Text]

  23. Griffith, F. (1928) Significance of Pneumococcal types. J. Hyg., 27, 113.

  24. Lorenz, M.G. and Wackernagel, W. (1994) Bacterial gene transfer by natural genetic transformation in the environment. Microbiol. Rev., 58, 563–602.[Abstract/Free Full Text]

  25. Chen, I. and Dubnau, D. (2004) DNA uptake during bacterial transformation. Nat. Rev. Microbiol., 2, 241–249.[CrossRef][ISI][Medline]

  26. Andersson, S.G., Zomorodipour, A., Andersson, J.O., Sicheritz-Ponten, T., Alsmark, U.C., Podowski, R.M., Naslund, A.K., Eriksson, A.S., Winkler, H.H. and Kurland, C.G. (1998) The genome sequence of Rickettsia prowazekii and the origin of mitochondria. Nature, 396, 133–140.[CrossRef][Medline]

  27. Beattie, D.S., Basford, R.E. and Koritz, S.B. (1967) Bacterial contamination and amino acid incorporation by isolated mitochondria. J. Biol. Chem., 242, 3366–3368.[Abstract/Free Full Text]

  28. Thinnes, F.P., Florke, H., Winkelbach, H., Stadtmuller, U., Heiden, M., Karabinos, A., Hesse, D., Kratzin, H.D., Fleer, E. and Hilschmann, N. (1994) Channel active mammalian porin, purified from crude membrane fractions of human B lymphocytes or bovine skeletal muscle, reversibly binds the stilbene-disulfonate group of the chloride channel blocker DIDS. Biol. Chem. Hoppe Seyler, 375, 315–322.[ISI][Medline]

  29. Han, D., Antunes, F., Canali, R., Rettori, D. and Cadenas, E. (2003) Voltage-dependent anion channels control the release of the superoxide anion from mitochondria to cytosol. J. Biol. Chem., 278, 5557–5563.[Abstract/Free Full Text]

  30. Enriquez, J.A., Perez-Martos, A., Lopez-Perez, M.J. and Montoya, J. (1996) In organello RNA synthesis system from mammalian liver and brain. Methods Enzymol., 264, 50–57.[Medline]

  31. Ojala, D., Montoya, J. and Attardi, G. (1981) tRNA punctuation model of RNA processing in human mitochondria. Nature, 290, 470–474.[CrossRef][Medline]

  32. Nagaike, T., Suzuki, T., Tomari, Y., Takemoto-Hori, C., Negayama, F., Watanabe, K. and Ueda, T. (2001) Identification and characterization of mammalian mitochondrial tRNA nucleotidyltransferases. J. Biol. Chem., 276, 40041–40049.[Abstract/Free Full Text]

  33. Gudikote, J.P. and Van Tuyle, G.C. (1996) Rearrangements in the shorter arc of rat mitochondrial DNA involving the region of the heavy and light strand promoters. Mutat. Res., 356, 275–286.[CrossRef][ISI][Medline]

  34. Seibel, P., Trappe, J., Villani, G., Klopstock, T., Papa, S. and Reichmann, H. (1995) Transfection of mitochondria: strategy towards a gene therapy of mitochondrial DNA diseases. Nucleic Acids Res., 23, 10–17.[Abstract/Free Full Text]

  35. McGregor, A., Temperley, R., Chrzanowska-Lightowlers, Z.M. and Lightowlers, R.N. (2001) Absence of expression from RNA internalised into electroporated mammalian mitochondria. Mol. Genet. Genomics, 265, 721–729.[CrossRef][ISI][Medline]

  36. Larsson, N.G., Oldfors, A., Holme, E. and Clayton, D.A. (1994) Low levels of mitochondrial transcription factor A in mitochondrial DNA depletion. Biochem. Biophys. Res. Commun., 200, 1374–1381.[CrossRef][ISI][Medline]

  37. Szabo, I., Bathori, G., Tombola, F., Coppola, A., Schmehl, I., Brini, M., Ghazi, A., De Pinto, V. and Zoratti, M. (1998) Double-stranded DNA can be translocated across a planar membrane containing purified mitochondrial porin. FASEB J., 12, 495–502.