WO1997041150A2

WO1997041150A2 - Gene therapy for mitochondrial dna deffects using peptide nucleic acids

Info

Publication number: WO1997041150A2
Application number: PCT/GB1997/001102
Authority: WO
Inventors: Douglas Matthew TURNBULL; Robert Neil LIGHTOWLERS; Robert William TAYLOR
Original assignee: The University Of Newcastle Upon Tyne
Priority date: 1996-04-27
Filing date: 1997-04-22
Publication date: 1997-11-06
Also published as: GB9608803D0; WO1997041150A3

Abstract

The invention concerns a method for selectively preventing the replication of mitochondrial DNA using complementary peptide nucleic acids; and the peptide nucleic acids for use in the method. Additionally novel targeting means are described for ensuring the targeting of the peptide nucleic acids into mitochondria.

Description

GENE THERAPY FOR MITOCHONDRIAL DNA DEFECTS USING PEPTIDE NUCLEIC ACIDS

The invention relates to a method for selectively preventing replication and/or expression of selected mitochondrial DNA; and peptide nucleic acids adapted to bind to selected parts of the mitochondrial genome.

Mitochondrial DNA (mtDNA) is the only extrachromosomal DNA in humans. It is a small (16.5kb) circular genome which encodes 13 polypeptides, 2 rRNAs and 22 tRNAs. The peptides encoded are all essential members ofthe mitochondrial respiratory chain and are synthesised within the organelle.

The human mitochondrial genome has evolved to show remarkable economy of organisation, containing only a short section, the D-loop, which does not contain any coding information. This region does, however, contain sequences important for the initiation and regulation of both transcription and replication.¹ Other unique features of mtDNA are that it is almost completely inherited from the mother,² and that there is more than one copy of the mitochondrial genome per mitochondrion.

The major function of mitochondria is to generate ATP from the oxidation of reduced cofactors. Essential for this process is the mitochondrial respiratory chain, a system which consists of five multisubunit complexes. The first four of these (complexes I-IV) are responsible for the electron transfer and proton pumping functions, while the fifth (complex V) is the ATP synthetase. These respiratory chain complexes are composed of between four and greater than thirty polypeptides, of which only a small proportion (seven for complex I, one for complex III, three for complex IV and two for complex V) are encoded by the mitochondrial genome. Thus, while mtDNA is crucial for maintaining a fully functional mitochondrion, it is the nuclear DNA that is responsible for encoding the majority of intramitochondrial proteins.

Defects of this genome are now recognised as important causes of disease and may take the form of point mutations or rearrangements. Defects of mtDNA are the primary genetic lesion in most patients with mitochondrial cytopathies.^3"5 Patients with these defects may present at any age with symptoms that vary from fatal lactic acidosis in infancy to a dementing illness in adulthood.⁶

The first described abnormality of the human mitochondrial genome was a mtDNA deletion,⁷ subsequently many other mutations of this genome have been reported.^8"11 Rearrangements of mtDNA are found in patients suffering from Kearns Sayre syndrome or with chronic progressive external ophthalmoplegia. In many of these patients there is a mtDNA deletion of the same size, in the same region of the genome, the so called "common" deletion.¹⁷ Point mutations of mtDNA have also been identified involving either a protein coding gene as described in Leber's hereditary optic neuropathy (LHON)^8,9 or more commonly affecting one of the tRNA genes, as for example in patients with the MERRF and MELAS syndromes, ^{lo π} or in patients with pure myopathy.¹³ The pathogenic nature of these mtDNA mutations has been confirmed by studies in which the mutated mtDNA has been transferred into cells lacking mtDNA (rho°).¹⁴ Under such circumstances the biochemical abnormality is transferred to this host cell.^{15 16} Interestingly, both deletions and point mutations of mtDNA have been shown to occur in normal individuals, although at a much lower level (<0.1% of total mtDNA), with the proportion increasing with age.¹⁷ This has raised speculation that defects of mtDNA are implicated in the aetiology of ageing and a variety of neurodegenerative diseases.^{17 18} Despite major advances in our ability to establish a diagnosis in patients with mtDNA defects, treatment for the majority of patients is supportive. Many patients progressively deteriorate resulting in severe disability and death.

When considering diseases due to mtDNA defects it is essential to consider the phenomenon known as heteroplasmy. This describes the presence of both mutant and wild-type mtDNA in the same tissue, cell and possibly within the individual mitochondrion.^19,20 Furthermore, the proportion of mutant to wild- type genome is seen to vary between tissues and also throughout life.²¹ It also seems that it is only once a given threshold level of mutant mitochondrial genomes has been reached that the abnormality becomes biochemically apparent,²⁰ the threshold level being dependent upon the mutation and the tissue involved.^3,10'²²

Regardless of major advances in our understanding of mtDNA defects at the genetic and biochemical level, there is no satisfactory treatment for the vast majority of patients. This is, perhaps, not surprising since the defect involves the final common pathway of oxidative metabolism and it is, therefore, impossible to bypass the defect by giving alternative metabolic fuels. While objective improvements on clinical examination and in biochemical parameters have been reported with treatments such as ubiquinone, L- carnitine, ascorbic acid, menadione and dichloroacetate,^23*24 in our experience, patients' symptoms continue to progress, often leading to severe disability and death. In the absence of any specific biochemical treatment, it is important to consider other possibilities such as gene therapy for patients with mtDNA disorders (25). Gene therapy typically involves placing a correct copy of a defective gene into a cell. This is achieved using a transfection vector. However, in the instance of mtDNA defects the same rational is not so easily copied because mitochondrial transfection vectors are not yet available. Thus those skilled in the art are currently focusing their attention on novel mitochondrial transfection vectors and/or ways to deliver mtDNA into mitochondria.

For example, one group of workers (26) have coupled double stranded mtDNA covalently to a short mitochondrial leader peptide, so generating a chimera that can enter mitochondria via a protein import pathway. This technique was successful and it was notable that translocation of the chimera into the mitochondria occurred with high efficiency and it was also independent of the size of passenger DNA. Notably, no experiments were undertaken to transport single stranded DNA into mitochondria, nor were any experiments undertaken to show binding of any transport materials to the mitochondrial DNA, indeed the use of double stranded DNA would tend to preclude this sort of investigation.

However, the development of gene therapy must take into account several key factors in the relationship between mtDNA mutations and disease. Firstly, there are multiple copies of the mtDNA in individual mitochondria and thus up to several thousand within an individual cell. Secondly, in the majority of patients with mtDNA defects both mutant and wild-type genomes are present in the same cell - the phenomenon known as intracellular heteroplasmy. Thirdly, in the presence of heteroplasmy there seems to be a threshold effect and only when a certain level of mutant mitochondrial genomes is reached does the disease become biochemically and clinically apparent. The mutant molecule appears to be extremely recessive since levels of >85% of mutant are usually required before any biochemical dysfunction is found (25).

Diseases characterised by mtDNA defects are therefore characterised by a mutant threshold level. Given this information it has been suggested that an alternative approach to gene therapy might involve the delivery of antisense material (26) to mitochondria with a view to preventing expression of defective mitochondrial genes and thus selectively favouring the expression of wild-type mtDNA. Whilst this approach seems straight forward and likely to bring success we have repeatedly failed to selectively inhibit defective mtDNA expression using antisense material (26) in the form of standard synthetic oligodeoxynucleotides i.e. a strand of nucleic acid linked by phosphate bonding.

We are unsure as to why our experiments involving antisense oligodeoxynucleotide material have been unsuccessful but somewhat surprisingly we have found that the same experiments can be conducted with success if we choose to use peptide nucleic acids. We were surprised to achieve success using peptide nucleic acids especially as we used the same sequence of nucleic acids which simply had a different backbone but we have found that our success is repeatabie and reliable as will be presented hereinafter.

Peptide nucleic acids comprise naturally occurring nucleobases or other nucleobase-binding moieties which are coherently bounded to a polyamide back bone. Peptide nucleic acids are known to bind to complementary DNA and RNA strands. Peptide nucleic acids, and their production, is described in US 5,539,082.

However, it is notable in the literature (36) that peptide nucleic acids do not passively defuse into biological cells via the lipid membrane and therefore our experiments are also involved in a way of successfully transporting peptide nucleic acids into biological cells.

Based on these key facts, we believe it will be possible to treat patients with heteroplasmic mtDNA defects by selectively inhibiting the replication of the mutant mtDNA by sequence complementary peptide nucleic acids. If this inhibition was maintained for a sufficient period of time then levels of wild- type mtDNA would increase relative to those of the mutant mtDNA. Assuming irreversible tissue damage has not been done, it will be possible to couple reversal of the genetic and clinical defect. The replication of human mtDNA may give us an unique opportunity for such a strategy since it is initiated at two different origins of replication and this results in the formation of single-stranded mtDNA during much of the replication process (27). Thus, during the single stranded phase of mtDNA replication there is the opportunity for binding of sequence specific peptide nucleic acids which inhibit replication.

It is therefore an object of the invention to provide material that selectively binds to mtDNA and more specifically that it is engineered to selectively bind to defective mtDNA with a view to preventing replication and/or expression of said defective DNA. It is yet a further object of the invention to provide a method for the prevention of defective mtDNA replication and/or expression which method essentially involves blocking the replication and/or expression of said defective DNA by the binding thereto of the material of the invention.

In its broadest aspect the invention concerns the use of peptide nucleic acids to prevent replication and/or expression defective mtDNA.

According to a first aspect of the invention there is therefore provided a peptide nucleic acid strand comprising a plurality of preselected nucleic acids having at least one peptide bond in said strand which are adapted to bind to at least a part of at least one mitochondrial gene.

Ideally most, if not all, neighbouring nucleic acids are linked by peptide bonds.

In a preferred embodiment of the invention said peptide nucleic acids are selected so as to bind to a defective mitochondrial gene and more particularly a part of a defective gene which includes a mutation or polymorphism, which mutation or polymoφhism ideally is thought to have biochemical consequences.

In yet a further preferred embodiment of the invention said peptide nucleic acid comprises between 5 and 20 nucleic acids and more preferably between 10 and 15 nucleic acids.

In a preferred embodiment of the invention said peptide nucleic acid is attached to or linked to a mitochondria targeting peptide so as to provide a PNA-peptide construct. Ideally, said PNA and said mitochondrial targeting peptide are linked theretogether using a linker.

It will be apparent to those skilled in the art that the presequence of any nuclear encoded mitochondrial gene will be sufficient to target PNA sequences into mitochondria.

In a yet further preferred embodiment of the invention said targeting peptide comprises an N-terminal region of human cytochrome c oxidase subunit VIII (a nuclear-encoded inner mitochondrial membrane protein), and most preferably the 25 N-terminal amino acids thereof. More preferably still said targeting peptide comprises the aforementioned N-terminal amino acid region joined to a further selected number of amino acids from the N-terminus of the mature protein and ideally a 4 further amino acids.

More preferably still said transport peptide comprises the sequence peptide shown in Figure 5.

According to a second aspect of the invention there is provided a method for selectively preventing replication and/or expression of selected mtDNA which method comprises the binding to said DNA of a complementary strand of peptide nucleic acid.

In a preferred method of the invention said peptide nucleic acid comprises a strand of selected nucleic acids which nucleic acids are selected so as to be complementary to a pre-determined part of at least one mitochondrial gene.

More specifically, said nucleic acids are selected so as to be complementary to a part of said gene which includes a mutation or polymoφhism which has deleterious biochemical consequences.

In a preferred embodiment of the invention said peptide nucleic acid is attached to or linked to a mitochondrial targeting peptide so as to provide a PNA-peptide construct. Ideally, said PNA and said mitochondrial targeting peptide are linked theretogether using a linker.

It will be apparent to those skilled in the art that the presequence of any nuclear encoded mitochondrial gene will be sufficient to transfer PNA sequences into mitochondria.

In a yet further preferred embodiment of the invention said mitochondria targeting peptide comprises an N-terminal region of human cytochrome c oxidase subunit VIII (a nuclear-encoded inner mitochondrial membrane protein), and most preferably the 25 N-terminal amino acids thereof. More preferably still said targeting peptide comprises the aforementioned N- terminal amino acid region joined to a further selected number of amino acids from the N-terminus of the mature protein and ideally a 4 further arnino acids.

An embodiment of the invention will now be described by way of example only with reference to the following figures wherein:-

Fig. 1. Replication run-off from human single-stranded mtDNA template by mitochondrial DNA polymerase. Lanes 1 and 4 show control reactions. Lanes 2 and 3 show replication products generated in the presence of lOμg.ml¹ aphidicolin and 5 μM ddTTP respectively. Products were sized by comparison to a standard DNA sequencing ladder.

Fig. 2. Specific inhibition of mutant "delete" template replication by a sequence-specific PNA. (A) Schematic representation of template production (PCR primers indicated by small arrows) and expected size of replication products in the presence and absence of the 14mer PNA, PNA-DELETE. (B) Phosphorimage of the replication products generated in the presence of increasing concentrations (0-0.2 μM) of PNA-DELETE. Lanes 1 to 6, wild type template; lanes 7 to 12, delete template. A truncated product generated by inhibition of replication due to PNA-DELETE is only visible in reaction lanes containing the mutant template (lanes 8 to 12).

Fig. 3. Specific inhibition of MERRF template replication by a sequence-specific PNA. (A) Schematic representation of template production and sizes of expected replication products in the presence and absence of the 1 lmer PNA, PNA-MERRF. Note that as both replication assays are initiated from identical primers, the truncated product generated in experiments using either mutant or wild type templates will be identical in size (approx. 215bp). (B) Phosphorimage of the replication products generated in the presence of increasing concentrations (0-9.2μM) of PNA-MERRF. Lanes 1 to 9, wild type template; lanes 10 to 18, MERRF template. A truncated replication product is only apparent in the reactions containing the mutant template (lanes 11 to 18). (C) Incoφoration of radiolabel into a 215bp truncated replication product for both the wild type (•) and mutant (■) templates (n~5, mean + SD). This confirms the sequence-specific inhibition of only the mutant MERRF template. (D) Selective inhibition of MERRF mtDNA replication in the presence of the wild type template. Templates (total amount lOng) were mixed to mimic levels of mtDNA heteroplasmy ranging from 0-100% mutant mtDNA. In the presence of 4.6μM PNA-MERRF, an increasing amount of truncated product is generated as the amount of mutant template in the reaction mix is increased. The histogram highlights the concomitant increase of incorporation of [ α -³²P] dCTP into a 215bp truncated replication product, with the increasing levels of MERRF template (Δ) in the reaction mix. The % wild type template in each replication reaction is shown

(^■)•

Fig. 4. Effect of E.coli SSB on inhibition of MERRF template replication by PNA-MERRF. (A) Phosphorimage of inhibition of MERRF template replication by 4.6μM PNA-MERRF with increasing amounts of SSB (0-1000ng). PNA-MERRF and SSB were added simultaneously to the replication run-off assays. At the ratio of template: SSB that gives maximal stimulation of polymerase γ activity (50ng SSB, lane 3), there is still substantial (75-80%) inhibition of MERRF template replication by PNA- MERRF (n=3, mean + SD).

Fig. 5. Shows a PNA-MERRF-peptide construct.

Fig. 6. Shows uptake of PNA and a PNA-peptide construct into biological cells i.e. myotubes, and more specifically into myotubes and into mitochondria. Materials and Methods

In the following section reference is made to inhibition of DNA replication, however it is of note that a reduction in the amount of DNA, by virtue of a reduction in DNA product will have an inhibitory effect on the corresponding level of expression and so an indirect effect on expression and also a direct effect when the material of the invention binds to DNA during protein production.

A region of mtDNA from a patient encompassing the deletion breakpoint was amplified by PCR using two pairs of oligonucleotide primers to generate wild type and mutant templates. Oligonucleotides L8283 (nucleotides 8283-8301) and H8582 (nucleotides 8565-8582) were used to amplify a 300 base pair region of wild type mtDNA, whilst oligonucleotides L8233 (nucleotides 8233-8253) and H 13559 (nucleotides 13541-13559) were used to amplify a 350 base pair region of mutant mtDNA; both light (L) strand primers were 5'-biotinylated. Thirty cycles of amplification were performed with lOOng of

DNA from patient, 60 pmol of each appropriate primer, 20 nmol of each 2'- deoxynucleoside-5'-triphosphate (dNTP) and 2.5 units of thermostable DNA polymerase in a buffer containing 75 mM Tris-HCl, pH 9.0, 20mM (NH₄)₂S0₄, 1.5mM MgCl₂ and 0.01% Tween-20 in a total volume of lOOμl. Samples were subjected to the following PCR conditions: denaturation at 94° C for 1 min, annealing at 55° C for 1 min, and extension at 72° C for 1.5 min; the final extension proceeded for 8 min. Amplification mixtures were diluted with lml of water, and concentrated to a volume of 40 μl using a Centricon-100 microconcentrator (Amicon) to remove unincoφorated primers. Biotinylated PCR products were bound to streptavidin-coated magnetic beads (Dynal A.S.) according to the manufacturers instructions and the non-biotinylated single strand recovered after melting the DNA duplex with 0.1M NaOH. Templates were precipitated, washed with 70% ethanol and resuspended in water. The size of the single strand DNA templates was confirmed by trace labelling with [α-^3SS] dATP, and running an aliquot on a 6% denaturing poly aery iamide gel against a control sequence ladder.

Single-strand templates were incubated in a buffer containing 20mM Tris- HCl, pH 7.5, lOmM MgCl₂ 14mM 2-mercaptoethanol, 150mM KCl, ImM ATP, 100 μM each of dATP, dGTP and dTTP, in the presence of lOOnM specific priming oligonucleotide. Replication of the wild type template was primed by the oligonucleotide L8333 (nucleotide positions 8333-8354); replication of the delete template was primed by the oligonucleotide L8283. Replication run-off using these oligonucleotides generates full length products of 250 base pairs for the wild type, and 300 base pairs for the delete template respectively. Samples were incubated at 70° C for 3 min and cooled to 37° C (to allow hybridization of the oligonucleotide). Following the addition of PNA, lOμM [α-³²P] dCTP (10μCi;3000Ci/mmH) and 1.5μg enzyme fraction, reaction mixtures were incubated at 37° C for 60 min. Incubation was terminated by the addition of aqueous phenol. Replication products were precipitated with 0.1 volume 3M Sodium Acetate, pH 5.2 and 2 volumes of ethanol in the presence of lOμg E.coli carrier tRNA, and resuspended in 4 μl sample buffer (95% formamide, 20mM EDTA, 0.05% bromophenol blue, 0.05% xylene cyanole). Samples were heated at 90° C for 3 min and separated on 6% denaturing polyacrylamide gels. Dried gels were exposed to a Phosphorimager cassette (Molecular Dynamics) and labelled products quantified using ImageQuant software.

An 875 base pair region of mtDNA encompassing the tRNA^Lys gene was amplified as described above using muscle DNA from a patient with the A8344G MERRF mutation and oligonucleotide primers L8154 (nucleotide positions 8154-8171) and H9028 (nucleotide positions 9028-9008); annealing was performed at 40° C. PCR products were subcloned into a pCRll plasmid vector (Invitrogen). Individual bacterial clones were isolated, and plasmids sequenced (15) to test for the presence of the A8344G mutation. A MERRF template was PCR-amplified using 200ng plasmid DNA and primers H8593 (nucleotide positions 8575-8593) and L8244 (nucleotide positions 8244-8264). Wild type template encompassing the tRNA^Lys gene was amplified from muscle DNA (previously sequenced to verify the absence of the A8344G mutation), using primers H8563 (nucleotide positions 8563-8545) and L8294 (nucleotide positions 8294-8314); both H strand primers were 5'-biotinylated. Single strand templates for the replication run-off assay were generated as previously described.

Replication run-off from both the MERRF and its corresponding wild type template used the same oligonucleotide primer, H8563. This generates a full length MERRF replication product of 31 1 base pairs, and a full length wild type replication product of 270 base pairs.

Uptake of PNA and a PNA-peptide construct into cells was investigated using the following techniques. The aforementioned 11-mer PNA-MERRF (with a 5' biotin group) was linked to a peptide component comprising 29 amino acids, corresponding to the 25 N-terminal amino acids of human cytochrome c oxidase submit VIII (a nuclear-encoded inner mitochondrial membrane protein) presequence and 4 amino acids of the amino terminus of the mature protein, as shown in Figure 5. Cultured human myotubes were then incubated with either PNA or the aforementioned PNA-peptide construct and uptake into the cells and mitochondria was monitored by fluorescence microscopy. Streptavidin fluorescence was used to monitor uptake into cells and Mito Tracker (TM), Molecular Probes was used to monitor uptake into mitochondria. Myotubes were incubated in serum free medium for 12 hours with either PNA (20μM) or construct (lOμM), fixed in 2.5% paraformaldehyde, permeabilised with 0.5% Triton in fixative and labelled. The images shown in Figure 6 were then obtained.

Results

An in vitro mtDNA replication run-off assay was used. All assays were initiated by the addition of a mitochondrial fraction containing DNA polymerase activity (28). The templates were single stranded DNA replication human mtDNA and replication was primed by relevant oligonucleotides. Aphidicolin, at a concentration that causes substantial inhibition of all nuclear DNA polymerases (lOμg.ml ¹), had no effect on the polymerase activity. This confirmed that this fraction contained only the mitochondrial polymerase (Fig. 1).

We generated two templates, one wild type and one encompassing the 4977 base pair "common" deletion (30). Light strand primers were biotinylated, facilitating isolation of heavy strand template for the replication assay (Fig. 2A). We synthesised a 14-mer peptide nucleic acid (PNA-DELETE) which was complementary to the deletion breakpoint sequence (5' CTGCCAATGGTGAG3') in the mutant mtDNA template. Only the first 7 bases of this PNA were complementary to the wild type template. Inhibition of replication following addition of PNA-DELETE would result in the formation of a truncated replication product (Fig. 2A). Using lOng of single- stranded human templates, we found that in a range of only 0.01-0.2μM (1 : 1- 20:1 molar excess) of the 14-mer PNA-DELETE, replication of the mutant template was inhibited specifically (>80% inhibition) and an appropriately sized truncated product generated (Fig. 2B). There was no inhibition of replication of the wild-type template at the same concentrations of 14-mer PNA-DELETE (Fig. 2B).

We extended these investigations to determine if we could inhibit replication in a sequence-specific manner when the mutant mtDNA only contained a point mutation. It has been shown that a single base mismatch in a PNA/DNA duplex can alter the thermostability of the complex by between 8-20° C (31). In view of the clinical importance of the heteroplasmic A8344G tRNA^Lys (MERRF) mutation (32), we wished to determine if a PNA could selectively inhibit replication of this mutation. In order to generate homoplasmic mutant template for our in vitro studies, an 875 base pair fragment of mtDNA amplified from muscle DNA of a patient with the A8344G tRN A^Lys mutation was cloned and the homoplasmic nature confirmed by sequencing (33). PCR primers around this region were designed to amplify templates (Fig. 3A); heavy strand primers were biotinylated to generate light strand templates. We synthesised an 11-mer peptide nucleic acid (PNA-MERRF) which was complementary to the light strand MERRF sequence (5' AGAGAGCCAAC 3') with the mutant base at nucleotide position, 8344 underlined. Replication run-off from MERRF mtDNA templates was inhibited by up to 75% in the presence of PNA-MERRF (0.092-9.2μM) (Figs. 3B and 3C) with the formation of a shortened replication product. Identical concentrations of PNA-MERRF (up to 1000- fold molar excess) did not inhibit replication of the wild type template (Figs 3B and 3C). To confirm this sequence-specific inhibition and to mimic the situation in vivo, defined ratios of wild type and mutant templates were mixed and the replication run-off assay performed in the presence of 4.6 μM PNA-MERRF, a concentration at which its near maximal inhibitory effect occurs. Whilst any truncated replication product generated by the wild type template would be the same size as that generated by the MERRF template (Fig. 3A), no product of the appropriate size is visible with 100% wild type template (Fig. 3D lane 1). Furthermore, as the proportion of MERRF template in the mix is gradually increased, there is a concomitant increase in the formation of a truncated replication product (Fig. 3D). mtDNA replication involves extensive formation of single-stranded mtDNA. Whilst we believe this will give us a unique opportunity to inhibit mtDNA replication, the single- stranded DNA will be associated in vivo with several proteins predominantly with the mitochondrial single-stranded binding protein (SSB), which may influence the binding of an anti-genomic PNA. Human mitochondrial SSB has been purified to homogeneity from mitochondria and has been shown to have many analogous properties to the E. coli protein (34). SSB stimulates DNA replication in vitro by mitochondrial DNA polymerase γ prepared from a variety of species (35). To determine the effect of SSB on human mtDNA synthesis in vitro increasing amounts of E.coli SSB was added to the replication run-off assay. E.coli SSB stimulated the replication of both the delete and MERRF templates with maximal stimulation occurring at a molar ratio of 1:75 (mtDNA template; SSB tetramer), consistent with the SSB tetramer exhibiting an average binding size of 46 nucleotides. Using concentration of E.coli SSB where template was saturated inhibition of mtDNA replication by PNA-MERRF was still measured (Fig. 4, lane 3). Even in the presence of a large molar excess of E.coli SSB, greater than 50% inhibition of replication was apparent in the presence of PNA. MERRF (Fig. 4, lane 7). Thus even in the presence of high concentration of DNA binding proteins such as SSB, the antigenomic PNA still selectively inhibits replication of mutant mtDNA.

In order to ensure that PNA adapted to bind to mitochondrial DNA can be taken into cells, and more specifically into mitochondria, we produced a PNA-peptide construct and then monitored the uptake of this construct into myotubes. The results are shown in Figure 6. Panels A, C and E represent the control experiments and show the presence of mitochondria in the myotubes. Similarly, Panel B also represents a control in that it shows the background fluorescence. Panel D shows uptake of PNA and Panel F shows uptake of PNA-peptide construct. These results show that whilst there is, suφrisingly, some non-specific uptake of PNA into cells, there is a large amount of staining in and around the nuclear region. However, as can be clearly seen in Panel F the construct co-localisers with mitochondria (compare with Panels F and E) whilst no nuclear staining is apparent.

It can therefore be seen that it is possible to arrange for the uptake of PNA to biological cells and moreover to monitor the co-localisation of this uptake with mitochondria.

The gene therapy of mtDNA disorders presents many different problems to that of nuclear genetic defects, the need to develop such treatment is of major importance. In patients with mtDNA defects the genetic diagnosis can be established in most patients but there is no effective treatment. We have developed a novel strategy based upon the particular properties of mtDNA replication and have been able to show that we can selectively impair replication in a sequence specific manner and moreover we have been able to show that PNA can be introduced in cells, and more specifically into mitochondria. Thus, this work represents the first crucial step in establishing a potential gene therapy for patients with mtDNA defects.

REFERENCES

1. Anderson, S. et al. Sequence and organization of the human mitochondrial genome. Nature 1981 ; 290: 457-465.

2. Giles, R. E., Blanc, H., Cann, H. M., Wallace, D. C, Maternal inheritance of human mitochondrial DNA. Proc Natl Acad Sci 1980; 77:

6715-6719.

3. Morgan-Hughes, J. A., et al. The molecular pathology of respiratory- chain dysfunction in human mitochondrial myopathies. Biochim Biophys Acta 1990; 1018: 217-222.

4. Wallace, D. C, Lott, M. T., Shoffner, J. M., Brown, M. D., Diseases resulting from mitochondrial DNA point mutations. J Inher Metab Dis 1992, 15: 472-479.

5. Lightowlers, R. N. Neurologic complications of inherited mitochondrial abnormality and neurologic consequences of inborn errors of metabolism. Curr Opin Neurol Neurosuerg 1993; 6: 429-436.

6. Munnich, A. et al. Clinical aspects of mitochondrial disorders. J Inherit Metab Dis 1992; 15: 448-455.

7. Holt, I. J., Harding, A. E., Morgan-Hughes, J. A. Deletions of muscle mitochondrial DNA in patients with mitochondrial myopathies. Nature 1988; 331 : 717-719. 8. Wallace, D. C, et al. Mitochondrial DNA mutation associated with Leber's hereditary optic neuropathy. Science 1988; 242: 1427-1430.

9. Howell, N. et al. Leber Hereditary optic neuropathy: identification of the same mitochondrial NDl mutation in six pedigrees. Am J Human Genet 1991; 49: 939-950.

10. Shoffner, J. M., et al. Myoclonic epilepsy and ragged red fiber disease (MERRF) is associated with a mitochondrial DNA, tRNA^Lys mutation. Cell 1990; 61: 931-937.

11. Goto, Y., Noneka, I., Horai, S. A mutation in tRNA^leu gene associated with the MELAS subgroup of mitochondrial encephalomyopathies. Nature

1990; 348: 651-654.

12. Schon, E. A., et al. A direct repeat is a hotspot for large-scale deletion of human mitochondrial DNA. Science 1989; 244: 346-349.

13. Bindoff, L. A., et al. Abnormal processing associated with a novel tRNA mutation in mitochondrial DNA. J Biol Chem 1993; 268: 19559-

19564.

14. King, M. P., Attardi, G. Human cells lacking mtDNA: repopulation with exogenous mitochondria by complementation. Science 1989; 246: 500- 503.

15. Chomyn, A. et al. In vitro genetic transfer of protein synthesis and respiration defects to mitochondrial DNA-less cells with myopathy-patient mitochondria. Mol Cell Biol 1991; 11: 2236-2244.

16. King, M. P., Koga, Y., Davidson, M., Schon, E. A. Defects in mitochondrial protein synthesis and respiratory chain activity segregate with the tRNA^Leu(UUR) mutation associated with mitochondrial myopathy, encephalopathy, lactic acidosis and stroke-like episodes. Mol Cell Biol 1992; 12: 480-490.

17. Wallace, D. C, Mitochondrial genetics: a paradigm for aging and degenerative diseases? Science 1992; 256: 628-632.

18. Cortopassi, G. A., Shibata, D., Soong, N-W., Arnheim, N. Proc Natl Acad Sci 1992; 89: 7370-7374.

19. Holt, I. J., Harding, A. E., Petty, R. K. H., Morgan-Hughes, J. A. A new mitochondrial DNA heteroplasmy. Am J Hum Genet 1990; 46: 428-433.

20. Boulet, L., Kaφati, G., Shoubridge, E. A. Distribution and threshold expression of the tRNA^lys mutation in skeletal muscle of patients with myoclonic epilepsy and ragged red fibres (MERRF). Am J Hum Genet 1992; 51: 1187-1200.

21. Larsson N-G et al. Progressive increase of the mutated mitochondrial DNA fraction in Kearns-Sayre Syndrome. Pediatr. Res 11990; 28: 131-136.

22. Rotig, A., Cormier, V., Blanche, S. et al. Pearson's marrow-pancreas syndrome, a multisystem mitochondrial disorder in infancy. J Clin Invest

1990; 86: 1601-1608. 23. Przyrembel, H. Therapy of mitochondrial disorders. J Inher Metab Dis 1987; 10: 129-146.

24. Eleff, S. et al. 3 IP NMR study of improvement in oxidative phosphorylation by vitamins K3 and C in a patient with a defect in electron transport at complex III in skeletal muscle. Proc Natl Acad Sci USA 1984; 81: 3529-3533.

25. Boulet, L., Kaφati, G., Shoubridge, E. A. Am J Hum Genet 1992; 51 : 1187; Chomyn, A., et al. Proc Natl Acad Sci USA 1992; 89: 4221.

26. Seibel, P., Trappe, J., Villani, G., Klopstock, T., Papa, S., Reichmann, H. Transfection of mitochondria: strategy towards a gene therapy of mitochondrial DNA diseases. Nucleic Acids Research 1995; 23: 10-17.

27. Clayton, D. A. Int Rev Cytol 1992; 141: 217.

28. Wong, T. W., Clayton, D. A. J Biol Chem 1985; 260: 11530; Wong, T. W., Clayton, D. A. Cell 1985; 42: 951.

29. Bovine liver mitochondria were isolated as previously described [REF] , and purified on sucrose gradients. 2g of bovine liver mitochondrial protein was subjected to chromatographic fractionation as described (7).

30. Schon, E. A., et al. Science 1989; 244: 346.

31. Orum, H. et al. Nucleic Acids Res 1993; 21 : 5332. 32. Enriquez, J. A., Chomyn, A., Attardi, G. Nature Genet 1995; 10: 47.

33. Sanger, F., Nicklen, S., Coulson, A. R. Proc Natl Acad Sci USA 1977; 74: 5463.

34. Curth, U., et al. Eur J Biochem 1994; 221: 435.

35. REFs: stimulation of DNA polymerase γ by SSB.

36. Wittung, P., et al. SEPS 1995; 375: 27-29.

Claims

1. A peptide nucleic acid strand comprising a plurality of preselected nucleic acids, having at least one peptide bond in said strand, which are adapted to bind to at least a part of at least one mitochondrial gene.

2. A peptide nucleic acid according to Claim 1 wherein a plurality of neighbouring nucleic acids are linked by peptide bonds.

3. A peptide nucleic acid according to Claims 1 or 2 wherein said nucleic acids are selected so as to be complementary to a part of a mitochondrial gene including a mutation or polymoφhism of clinical consequence.

4. A peptide nucleic acid according to any preceding claim comprising between 5 and 20 nucleic acids.

5. A peptide nucleic acid according to any preceding claim comprising between 10 and 15 nucleic acids.

6. A peptide nucleic acid according to any preceding claim wherein said peptide nucleic acid is attached to a mitochondria targeting peptide so as to provide a PNA-peptide construct.

7. A peptide nucleic acid according to Claim 6 wherein said PNA and said targeting peptide are linked theretogether using a linker.

8. A peptide nucleic acid according to Claim 6 or 7 wherein said targeting peptide comprises a presequence of a nuclear encoded mitochondrial gene.

9. A peptide nucleic acid according to Claims 6 to 8 wherein said targeting peptide comprises an N-terminal region of human cytochrome c oxidase subunit VIII.

10. A peptide nucleic acid according to Claim 9 wherein said targeting peptide comprises the 25 N-terminal amino acids of said oxidase.

11. A peptide nucleic acid according to Claims 9 or 10 wherein said targeting peptide comprises the said N-terminal joined to a further selected number of amino acids from the N-terminus of the mature protein.

12. A peptide nucleic acid according to Claim 11 wherein said preselected number comprises 4 amino acids.

13. A peptide nucleic acid according to any preceding claim comprising the sequence peptide shown in Figure 5.

14. A method for selectively preventing replication and/or expression of selected mitochondrial DNA which comprises the binding to said DNA of a complementary strand of peptide nucleic acid.

15. A method according to Claim 14 wherein said strand of peptide nucleic acid comprises a peptide nucleic acid according to Claims 1 to 13.