(Received for publication, September 4, 1996, and in revised form, December 12, 1996)
From the Department of Biochemistry and Molecular
Genetics, Imperial College School of Medicine at St Mary's Hospital,
Norfolk Place, London W2 1PG, United Kingdom, and the
Max
Delbrück Center for Molecular Medicine, Robert Rössle
Strasse 10, Berlin 13122, Germany
Mitochondrial disorders are a large group of phenotypically heterogeneous diseases. An understanding of their molecular basis would benefit greatly from the ability to manipulate the mitochondrial genome and/or to introduce functional exogenous DNA into mitochondria. As a first step toward this approach, we have used electroporation to introduce a 7.2-kilobase plasmid DNA into isolated functional mitochondria. Transfer of the DNA at field strengths between 8 and 20 kV/cm was investigated by Southern blot analysis. Maximal plasmid internalization was achieved at a field strength of 14 kV/cm. The functional integrity of the mitochondria after electroporation was verified by enzymatic assays of specific mitochondrial marker enzymes and by measuring respiratory control. At field strengths above 12 kV/cm, an increasing mitochondrial destruction was observed. 12 kV/cm was found to be optimal for the most efficient plasmid internalization while still retaining the functional integrity of the mitochondria. At this field strength, about half of the internalized plasmid was found in the inner membrane or mitochondrial matrix, as determined by immunoelectron microscopy and Southern blot analysis of electroporated mitochondria treated with digitonin. We estimate that on average one plasmid molecule/mitochondrion reaches the matrix or inner membrane.
A growing number of degenerative diseases of the brain, skeletal and heart muscle, the kidney, and endocrine glands are recognized as being due to mutations in the mitochondrial genome (for reviews, see Refs. 1-3). The introduction of specific changes in the mitochondrial genome and the ability to reintroduce the modified genomes into mitochondria would facilitate a comprehensive analysis of the mechanisms of expression and replication of the mitochondrial genome, provide insight into the molecular pathology of mitochondrial diseases, and may serve as a first step toward gene therapy of mitochondrial disease (4).
Because of the high DNA transfer efficiency obtained with electroporation in bacteria and cultured cells, we decided to investigate this procedure for the introduction of DNA into mitochondria in vitro which could then be introduced into cells by microinjection (5, 6) or endocytosis (7).
In this paper we describe the development and optimization of an electroporation procedure for the introduction of DNA, up to a size of at least 7.2 kilobases, into isolated mammalian mitochondria. We demonstrate the transfer of this DNA into the inner mitochondrial membrane and/or matrix by Southern analysis after removal of the outer membrane and by electron microscopy on the intact organelles, and we have shown functionality of the manipulated organelles by preservation of their ATP-generating capacity.
All steps of mitochondrial isolation were performed at 4 °C. Mice of no particular breed or sex were starved overnight and killed by cervical dislocation. The livers were collected into homogenization buffer (70 mM sucrose, 220 mM mannitol, 1 mM EGTA, 0.5% w/v fatty acid-free bovine serum albumin, 10 mM HEPES, pH 7.4), cut into small pieces, washed three times in homogenization buffer, and homogenized by two strokes of a Potter homogenizer (1,000 rpm) at a ratio of 15 ml of buffer/3 g of tissue. The homogenate was filtered through two layers of gauze and centrifuged three times at 700 × g for 5 min to eliminate cell debris, red blood cells, and nuclei. The supernatant was centrifuged at 15,000 × g for 10 min to obtain a crude mitochondrial pellet. Mitochondria were rinsed once in one-third of the initial supernatant volume of homogenization buffer, centrifuged again at 15,000 × g for 10 min, and resuspended into respiration buffer (225 mM sucrose, 10 mM KH2PO4, 10 mM KCl, 1 mM EDTA, 10 mM Tris-HCl, pH 7.4) at a concentration of 100 mg of mitochondrial protein/ml for measurement of mitochondrial respiration or into 0.33 M sucrose for other procedures. The protein concentration was determined using the Micro BCA kit (Pierce), with bovine serum albumin as a standard.
Electroporation5 µg of pCMV plasmid (Clontech) was
added to 100 µl of the mitochondrial suspension of which 20 µl was
transferred into a cold electroporation cuvette (0.1-cm gap cuvette;
Bio-Rad) containing 30 µl of 0.33 M sucrose.
Electroporation was carried out using a Bio-Rad Gene
PulserTM at a capacitance of 25 microfarads, a resistance
of 400
, and a range of field strengths varying between 8 and 20 kV/cm. Electroporated mitochondria were removed quickly from the
cuvette. The cuvette was washed once with 50 µl of 0.33 M
sucrose, which was added to the electroporated mitochondria. Freshly
electroporated mitochondria were centrifuged at 15,000 × g for 4 min at 4 °C, and the pellet was used for all
subsequent experiments, with the exception of the enzymatic assays for
which it was frozen in liquid nitrogen. Nonelectroporated mitochondria
controls were treated identically.
The mitochondrial pellet was
resuspended in 500 µl of 0.33 M sucrose and centrifuged
at 15,000 × g for 6 min. It was resuspended in 100 µl of DNase buffer (0.33 M sucrose, 2 mM
magnesium acetate, 10 mM Tris-HCl, pH 7.5) and incubated
for 1 h at room temperature with DNase I (Life Technologies, Inc.)
at a final concentration of 20 µg/ml to digest any noninternalized
DNA. DNase was subsequently inactivated by the addition of 4 µl of
0.5 M EDTA, 10-min incubation at 65 °C, and treatment
with 1 mg/ml proteinase K (Merck) for 4 h at 37 °C. 2 µl of
10% SDS was added to achieve mitochondrial lysis. DNA was extracted
with phenol/chloroform (8) and precipitated with 0.10 volume of 3 M sodium acetate, 3 volumes of 100% ethanol, and 20 ng of
carrier yeast tRNA at 20 °C overnight. The DNA was resuspended in
TE buffer (10 mM Tris-HCl, pH 8.0, 1 mM EDTA)
and electrophoresed in a 0.8% agarose gel. The separated DNA was
transferred overnight to a Hybond N membrane (Amersham Corp.). Blots
were baked for 2 h at 80 °C before a 3-h prehybridization in
CHURCH buffer (1.68% w/v NaH2PO4, 5.41% w/v
Na2HPO4, 0.7% w/v SDS, and 1% v/v denatured
salmon sperm DNA) at 65 °C. Blots were hybridized overnight at
65 °C with pCMV
or pRSmt22 (containing the whole mouse
mitochondrial DNA) plasmids labeled with 32P using a
Megaprime DNA labeling kit (Amersham). Blots were washed at 65 °C
for 15 min consecutively in 2 × SSC, 1 × SSC, and 0.2 × SSC containing 0.1% v/v SDS before autoradiography. For
quantification, DNA band intensities were measured on a PhosphorImager
(Molecular Dynamics) using a known amount of pCMV
plasmid and
endogenous mitochondrial DNA (mtDNA) as internal control.
The mitochondrial pellet was resuspended in 10 µl of respiration buffer. Mitochondria were added to an oxygen electrode chamber maintained at 30 °C, containing 2.4 ml of respiration buffer. The final concentrations of added substrates were: 2.5 mM malate+glutamate, 5 mM succinate, or 0.6 mM NADH. Oxidation was measured for 3 min before the addition of 180 µmol of ADP. The respiration rate was measured until the added ADP was exhausted. Mitochondrial coupling was determined by measuring the respiratory control ratio (RCR)1 and produced ATP/consumed oxygen (P/O) ratio. The RCR is the ratio of the oxygen consumption rate in the presence of ADP (state III) to the oxygen consumption rate when ADP is exhausted (state IV). Coupled liver mitochondria have an RCR of approximately 2-10. The P/O ratio is an estimation of the number of coupling sites along the respiratory chain, measured by the number of molecules of ADP used (i.e. ATP synthesized) per oxygen atom consumed. For a NADH-linked substrate (malate+glutamate), the theoretical P/O ratio is 3 and for FADH2-linked substrate it is 2. In practice, P/O ratios are between 2.5 and 3.0 for NADH-linked substrates and between 1.5 and 2.0 for FADH2-linked substrates.
Enzymatic AssaysThe frozen mitochondrial pellet was diluted to a concentration of 1.72 mg of protein/ml in 0.33 M sucrose.
Citrate synthase activity was measured as described previously by Srere (9) in the presence of 0.1% Triton X-100.
Cytochrome c oxidase activity was measured in a 10 mM potassium phosphate buffer, pH 7.0, containing 0.08 mM cytochrome c reduced immediately before measurements with few grains of sodium hydrosulfite. The activity was determined as a decrease in absorbance at 550 nm after the addition of about 10 µg of protein of the mitochondrial suspension.
Monoamine oxidase activity was measured at 250 nm in a 50 mM potassium phosphate buffer, pH 7.5, containing 2.5 mM benzylamine. The reaction was started by the addition of about 150 µg of mitochondrial protein.
Analysis of Electroporation-induced Mitochondrial DamageThe effect of electroporation on the integrity of
mitochondria was investigated by estimation of the relative changes (in percent) of enzyme activity and mtDNA amount associated with the mitochondrial pellet after exposure to increasing field strengths compared with the values of mitochondria before electroporation (100%). Each graph (see Figs. 2, 3, 4 and 5, a and
b) is constructed using the mean of two or three estimations
at each indicated field strength derived from three independent
electroporation experiments using the indicated range of increasing
field strengths. The overall standard deviation from the mean of all
single estimations at each field strength was calculated from all
values of each parameter. It was ±4.8% for cytochrome c
oxidase, ±2.8% for citrate synthase, ±4.6% for monoamine oxidase,
and ±18.5% for mtDNA. For RCRs, standard deviation values were
±6.4% and ±3.1% and for P/O ratios ±3.9% and ±5.0% with
malate+glutamate or succinate as substrate, respectively. For the
O2 consumption of mitochondria using malate+glutamate, succinate, or NADH as substrate the overall standard deviation was
±2.0, ±2.3, and ± 0.7 µmol of O2 consumed/min/mg
of protein, respectively.
Digitonin Solubilization of the Outer Membrane
Several samples of mitochondria were collected after electroporation by centrifugation for 10 min at 10,000 × g. They were resuspended in 1 ml of 0.8 M sucrose, 10 mM Tris-HCl, pH 7.5, 1 mM EDTA, 0.1% w/v fatty acid-free bovine serum albumin and purified further on a continuous 1.0-2.0 M sucrose gradient by centrifugation at 80,000 × g for 30 min at 4 °C. The band of pure mitochondria was removed, diluted with 2 volumes of TE, pH 8.0, and centrifuged for 10 min at 10,000 × g at 4 °C. Pure mitochondria were resuspended in 0.33 M sucrose at a concentration of 100 mg of protein/ml. A 2.4% (w/v) digitonin solution in 0.33 M sucrose was added to obtain a concentration of 0.30 mg of digitonin/mg of mitochondrial proteins. Digitonin treatment was carried out for 15 min at 4 °C without shaking and terminated by the addition of 500 µl of 0.33 M sucrose. The homogenate was centrifuged for 15 min at 15,000 × g, and the mitoplast pellet was washed once with 500 µl of 0.33 M sucrose. After a further 15-min centrifugation at 15,000 × g, the mitoplast pellet was used for DNA extraction after DNase treatment or for enzymatic assays, as described previously.
Immunoelectron MicroscopypCMV was labeled for electron
microscopy with digoxigenin-11-dUTP (Boehringer Mannheim) by nick
translation, using conditions that prevent degradation of the DNA
molecule (10). The labeled DNA was purified on a Sephadex G-50 column.
Electroporation was performed with digoxigenin-labeled pCMV
plasmid
(or unlabeled control plasmid) as described previously. Mitochondria
were then washed once in 0.33 M sucrose, pelleted at
15,000 × g in a microcentrifuge for 1 min, and
resuspended immediately in 500 µl of fixation buffer (0.1 M sodium phosphate, pH 7.4, 0.18 M sucrose, 1%
w/v glutaraldehyde) after careful removal of all of the supernatant.
After fixing for 1 h at room temperature, mitochondria were
pelleted as above. All traces of supernatant were removed, and the
mitochondria were washed twice for 1 h at room temperature in 500 µl of 0.1 M sodium phosphate buffer, pH 7.4. The
mitochondrial pellet was immersed for 3 h in a mixture of 1.8 M sucrose and 20% polyvinylpyrrolidone according to
Tokuyasu (11) and cryosectioned. Immunolabeling and detection with
anti-digoxin antibody (Sigma) and anti-mouse gold (Dianova) were
performed according to Griffiths (12).
Based on the similar size of Escherichia coli and mitochondria and the high efficiency of bacterial transformation by electroporation, we chose to investigate this procedure for the transformation of mitochondria.
Southern blot analysis was used to determine the effect of increasing
field strengths on the internalization of the pCMV plasmid into
mitochondria by electroporation (Fig. 1a).
Following electroporation at field strengths between 0 and 20 kV/cm,
mitochondria were treated with DNase I to remove any noninternalized
plasmid and were pelleted before DNA extraction and electrophoresis.
The Southern blot was subsequently hybridized with labeled probes for
both mtDNA (pRSmt22) and pCMV
(Fig. 1a), and the
intensities of the bands were quantified with a PhosphorImager (Figs.
1b and 2).
As shown in Fig. 1, increasing amounts of plasmid are rendered DNase-resistant by increasing field strengths of electroporation. No DNase protection is achieved without electroporation. The amount of internalized plasmid increases with field strengths up to 14 kV/cm (approximately of 4.4 ng of internalized plasmid at 14 kV/cm) and decreases at higher field strengths (Fig. 1, a and b). The majority of internalized plasmid, up to 14 kV/cm, is in the open circular conformation. At higher field strengths, most of it is linear (Fig. 1, a and b). No supercoiled plasmid was detected at any of the applied field strengths. Beginning at 12 kV/cm and increasing with higher field strengths, electroporation-induced damage of the plasmid is also visible as a smear (Fig. 1a). Plasmid DNA may be more susceptible to such damage than the endogenous mtDNA as it is not protected by the mitochondrial membranes.
The amount of mtDNA remains practically constant up to a field strength
of about 12 kV/cm (Fig. 2). At 12 kV/cm, intact mtDNA is reduced to
83% of the amount presents in nonelectroporated mitochondria. At
higher field strengths, a dramatic decrease of the mtDNA was observed.
At 20 kV/cm it is reduced to 20% compared with nonelectroporated
mitochondria (Fig. 2). This decrease is most likely due to an
electroporation-induced mitochondrial destruction, with a release of
the mtDNA into the electroporation medium and its subsequent DNase I
digestion. Electroporation-induced damage to the mtDNA does not appear
likely as judged from the absence of observable degradation on the
Southern blot, after mtDNA probe hybridization. To quantify the amount
of internalized pCMV plasmid in relation to intact mitochondria the
amount of endogenous mtDNA was used for reference (Fig.
1b).
These results demonstrate that it is possible to transfer DNA of plasmid size into mitochondria by electroporation. However, the extent of damage to the mitochondria inflicted by this procedure, as suggested by the decrease of endogenous mtDNA, was a major concern. We therefore investigated the influence of electroporation on the structural and functional integrity of mitochondria by studying its effect on the mitochondrial association of marker enzymes and on the respiratory capacity of the organelles.
The following marker enzymes, specific for particular mitochondrial compartments, were used: monoamine oxidase and cytochrome c oxidase for the outer and the inner mitochondrial membranes, respectively, and citrate synthase for the matrix. For this analysis, the mitochondria were pelleted, and the enzyme activity and DNA content associated with the mitochondrial pellet were determined at the indicated field strengths.
Up to a field strength of 12 kV/cm, the activity of both monoamine oxidase and citrate synthase decreases slightly to about 83% relative to nonelectroporated mitochondria (Fig. 3). This indicates a slight destructive effect up to this field strength. At higher field strengths, the citrate synthase activity falls to 60% at 18 kV/cm and then to a residual 18% at 20 kV/cm. In contrast, monoamine oxidase activity decreases to only 79% at 18 kV/cm and then falls to 60% of nonelectroporated mitochondria at 20 kV/cm (Fig. 3). The difference between the two enzymes suggests a leakage of the matrix enzyme through the destabilized mitochondrial membranes rather than a complete destruction of the mitochondrial structures up to about 18 kV/cm. The increase of cytochrome c oxidase activity between a field strength of 10 and 14 kV/cm to a plateau value of about 120% between 14 and 18 kV/cm (Fig. 3) is probably also due to this electroporation-induced membrane destabilization allowing increased access of the substrate (reduced cytochrome c) to the enzyme. At 20 kV/cm, a steep fall in cytochrome c oxidase activity to 50% is observed, reaching practically the level of monoamine oxidase activity (Fig. 3) and suggesting a massive destruction of the mitochondrial structures.
The pattern of release of mtDNA from the mitochondria with increasing field strength correlates well with the decrease of citrate synthase activity (Fig. 2). This supports the hypothesis of an initial leakage of the mitochondrial matrix during electroporation since both citrate synthase and mtDNA are located in the matrix compartment.
To assess the functional integrity of mitochondria during electroporation, we measured mitochondrial respiration using two different substrates, malate+glutamate and succinate, in the presence of ADP. Succinate is a direct substrate for complex II, whereas malate+glutamate produce NADH in the mitochondrial matrix, which is a direct substrate for complex I. NADH cannot be used directly as a substrate on intact and coupled mitochondria as it is unable to cross the inner membrane. Respiration with NADH on intact mitochondria is very low and can therefore be used as an indicator of mitochondrial integrity.
Fig. 4 shows a linear decrease in the rate of oxygen consumption up to a field strength of 14 kV/cm using both malate+glutamate and succinate as substrate. At 14 kV/cm, the oxygen consumption rate is reduced to 71 and 53% of that of control mitochondria using malate+glutamate and succinate as substrate, respectively. At higher field strengths, the rate of oxygen consumption drops dramatically with both substrates indicating total loss of function above 14 kV/cm.
Using NADH as substrate, mitochondrial respiration is very low up to 10 kV/cm (between 1.7 and 2.4 µmol of O2 consumed/min/mg of protein), indicating complete membrane integrity. Oxygen consumption increases from 12 kV/cm onward, reaching a peak at 14 kV/cm (6.2-fold increase compared with control mitochondria) indicating membrane destabilization. It drops at higher field strengths in accordance with malate+glutamate- or succinate-fueled respiration (Fig. 4), again indicating mitochondrial destruction. This is in agreement with the severe structural membrane damage above field strengths of 14 kV/cm determined by the enzymatic assays.
Mitochondrial coupling was used as the most sensitive functional assay, by measuring the changes of the P/O ratio and RCR under the influence of electroporation (Fig. 5, a and b). With increasing field strengths, a similar decrease of the RCR is observed with either malate+glutamate or succinate as substrate. The RCRs indicate good coupling (RCR = 3.3 and 2.1 with malate+glutamate or succinate as substrate, respectively) up to 12 kV/cm. The P/O ratio is also conserved up to this field strength, but it drops dramatically above this value.
In conclusion, the maximum field strength up to which isolated mitochondria can be electroporated while still maintaining their structural and functional integrity is 12 kV/cm.
Localization of Internalized Plasmid DNAFurther questions to be addressed concerned the localization of the introduced DNA, in particular whether or not the DNA reached the mitochondrial matrix, and the extent to which contaminating nonmitochondrial structures contributed to the DNase protection effect. Electron microscopy and selective solubilization of the outer mitochondrial membrane in conjunction with sucrose gradient purification of the electroporated mitochondria were applied to answer these questions.
The electron micrographs in Fig. 6 are examples of
ultrathin sections from mitochondria electroporated at 12 kV/cm with
digoxigenin-labeled pCMV plasmid and stained with monoclonal
anti-digoxin and a gold-labeled secondary antibody. Internalized
plasmid molecules can be seen at the mitochondrial periphery
(panels a-c) and in the inner mitochondrial membrane or
matrix (panels e-g). The resolution of immunoelectron microscopy does not allow a more exact distinction between these two
compartments. Plasmid DNA signals could be detected in 63 out of a
total 303 mitochondria (Table I). This is statistically significant (p < 0.001) compared with a no-pulse
control in which 5/303 mitochondrial sections contained a signal. Of
these 63 sections, signals localized in the inner membrane/matrix were
detected in 37 sections (58%) and in 3 sections of the no-pulse
control (Table I).
|
To exclude the possibility that free digoxigenin-11-dUTP would account for the observed signal a control electroporation was performed using plasmid DNA which was nick translated in the absence of digoxigenin-11-dUTP and supplemented with this nucleotide after nick translation. This DNA did not result in any significant signal, showing that only labeled DNA plasmid molecules (as opposed to free digoxigenin-11-dUTP) were responsible for the immunoreactive signal.
DNA plasmid signals were also found associated with nonmitochondrial structures (endoplasmic reticulum, peroxisomes, Golgi-like membranes, but no nuclei) present in the mitochondrial preparation (panel d). 33 signals in nonmitochondrial structures were counted in the viewing fields containing the 303 analyzed mitochondria (Table I). Since this plasmid is also not digested by DNase I, it probably contributes to the protected plasmid DNA detected by Southern blot analysis (Fig. 1) to about 30%.
To establish further the predominantly mitochondrial internalization and in particular inner membrane/matrix localization of the transferred plasmid DNA, we purified the electroporated mitochondria (12 kV/cm) on a sucrose gradient and subsequently treated them with digitonin and DNase I before extracting DNA for Southern blot analysis. The gradient purification resulted in a mitochondrial preparation virtually free of contaminating membrane structures (less than 5%). Digitonin is a specific detergent that selectively solubilizes the outer mitochondrial membrane. Treatment with digitonin was therefore used to determine the amount of plasmid internalized into the mitochondrial matrix. Complete destruction of the outer membrane after digitonin treatment was verified by the absence of monoamine oxidase activity in digitonin-treated mitochondria (results not shown).
Fig. 7 demonstrates the internalization of plasmid
DNA into the highly purified nontreated and digitonin-treated
mitochondria. The internalized plasmid was quantified by radioimaging,
and the amount of internalized plasmid in treated and nontreated
samples was compared after normalization with the respective mtDNA
amount. 43% of the total amount of internalized plasmid was found to
be localized to the matrix with similar amounts of both linear and open
circular plasmid forms.
We have shown that exogenous plasmid DNA with a size of 7.2 kilobases can be introduced into isolated mitochondria by electroporation at an optimal electroporation field strength of about 12 kV/cm. At this field strength, a significant internalization of plasmid into mitochondria is observed by Southern blot analysis. This is not the field strength of maximum internalization (maximum at 14 kV/cm) but represents the maximum applicable field strength at which minimal mitochondrial damage is observed.
Electroporation can induce damage of DNA at high field strengths (13). At 12 kV/cm we observed conversion of the supercoiled DNA to open circular and linear forms. As mitochondria possess specific ligases and topoisomerases (14), the reconstitution of the supercoiled form of the internalized plasmid can be expected.
Conservation of mitochondrial integrity is also of great importance since injection of only a small number of partially uncoupled mitochondria into human fibroblasts was found to be detrimental to cell viability (15). At a field strength of 12 kV/cm the mitochondria appear morphologically intact by immunoelectron microscopy, and the values of the coupling indicators are still in the normal range (RCR and P/O values of > 3 and 2 with malate+glutamate and succinate, respectively), despite some membrane destabilization.
The introduced foreign DNA has to reach the mitochondrial matrix to allow its replication and gene expression. The combined evidence from electron microscopy and digitonin treatment demonstrates the presence of about half of the internalized plasmid in the inner membrane/matrix compartment.
303 sections of the mitochondria electroporated with a 12-kV/cm pulse were analyzed, of which 37 (12%) contained an inner membrane/matrix-localized plasmid DNA signal. Since each detected signal represents antibody bound to a single plasmid molecule, 12% of mitochondrial sections contain internalized plasmid DNA that may be localized to the matrix. It should be noted that the immunoelectron microscopy will only detect plasmid molecules that are on the surface of a section. Consequently, the total number of molecules in a section is likely to be underestimated. Given that each section only represents approximately one-tenth of the total mitochondrial volume, these data can only provide a rough estimate of the number of plasmid molecules present in a whole mitochondrion. Assuming a uniform distribution of plasmid molecules among the mitochondria and that each section represents a single mitochondrion, the observed frequency of 12% supports the presence of one plasmid molecule/mitochondrion (p < 0.05). Given that there is an average of five copies of endogenous mtDNA/mitochondrion, it can be postulated that plasmid DNA is present at approximately 20% of the level of mtDNA.
Several techniques have been tried for the introduction of foreign DNA into mitochondria (16-19). Although successful in plant cells and yeast, biolistic delivery has so far not been applicable to mammalian mitochondria. Early attempts to introduce DNA into "mega"-mitochondria of liver cells by systemic delivery in encapsulating liposomes (20) have not been continued, probably because of the low efficiency of this procedure. More recently Seibel et al. (21) have exploited the protein import machinery of mitochondria to translocate a DNA/targeting-peptide chimera into isolated mitochondria following earlier experiments by Vestweber and Schatz (22). They delivered a 322-base pair DNA fragment covalently conjugated to the 32-amino acid presequence of rat ornithine transcarbamylase into isolated rat liver mitochondria. 37% of the conjugated fraction associated with mitoplasts after digitonin treatment was completely translocated into the matrix.
The electroporation technique described here is so far the most effective means for introduction of exogenous DNA of a size of at least 7.2 kilobases into the matrix of mammalian mitochondria. This method should allow the effective introduction of functional nucleic acid sequences into viable mitochondria in vitro and thereby be of great value for the elucidation of mitochondrial gene expression. We also believe that this technique may constitute a first step toward a novel approach of gene therapy for mitochondrial disorders. We are, however, aware of the multitude of problems to be solved before this is likely to become feasible. Among them are the need to adapt this method for the size of mitochondrial DNA and/or to construct a "minimal" functional mitochondrial genome incorporating a corrective sequence, the demonstration of stable replication and expression of such construct in mitochondria, and finally the ability to introduce such manipulated mitochondria into cells in vitro and in vivo.
We have chosen the X-linked ornithine transcarbamylase deficiency as a model for such an approach and have synthesized a gene encoding human ornithine transcarbamylase according to mitochondrial codon usage (4). We have recently inserted this gene sequence into the cloned mouse mitochondrial genome between two contiguous tRNAs, giving it the potential to be processed from the primary mitochondrial transcript.2 We plan to introduce this construct by the described electroporation method into mitochondria and to test for expression using an in organello expression system (23). If successful, we will attempt correction of this defect by introduction of mitochondria containing this modified genome into ornithine transcarbamylase-deficient hepatocytes or oocytes from spf or spf-ash mice (24, 25) by microinjection (5) or by endocytosis (7).
We thank Drs. Heiko Müller, Bettina Erdmann, and Catherine Godinot for initial help and advice on this work. A patent for this mitochondrial gene therapy approach is pending.