ORIGINAL ARTICLES
Rescue of a Mitochondrial Deficiency Causing Leber Hereditary Optic Neuropathy John Guy, MD,1,2 Xiaoping Qi, MD,1 Francesco Pallotti, MD, PhD,3 Eric A. Schon, PhD,3 Giovanni Manfredi, MD,4 Valerio Carelli, MD, PhD,5 Andrea Martinuzzi, MD, PhD,6 William W. Hauswirth, PhD,1,7 and Alfred S. Lewin, PhD7
A G to A transition at nucleotide 11778 in the ND4 subunit gene of complex I was the first point mutation in the mitochondrial genome linked to a human disease. It causes Leber Hereditary Optic Neuropathy, a disorder with oxidative phosphorylation deficiency. To overcome this defect, we made a synthetic ND4 subunit compatible with the “universal” genetic code and imported it into mitochondria by adding a mitochondrial targeting sequence. For detection we added a FLAG tag. This gene was inserted in an adeno-associated viral vector. The ND4FLAG protein was imported into the mitochondria of cybrids harboring the G11778A mutation, where it increased their survival rate threefold, under restrictive conditions that forced the cells to rely predominantly on oxidative phosphorylation to produce ATP. Since assays of complex I activity were normal in G11778A cybrids we focused on changes in ATP synthesis using complex I substrates. The G11778A cybrids showed a 60% reduction in the rate of ATP synthesis. Relative to mock-transfected G11778A cybrids, complemented G11778A cybrids showed a threefold increase in ATP synthesis, to a level indistinguishable from that in cybrids containing normal mitochondrial DNA. Restoration of respiration by allotopic expression opens the door for gene therapy of Leber Hereditary Optic Neuropathy. Ann Neurol 2002;52:534 –542
A G to A transition at nucleotide 11778 in mitochondrial DNA (mtDNA) in the gene specifying the ND4 subunit of complex I results in an arginine to histidine substitution at amino acid 340. It was the first mtDNA point mutation linked to a maternally inherited human disease, Leber hereditary optic neuropathy (LHON), a disorder blinding patients during the second and third decades of life. Since this discovery 14 years ago,1 more than 30 other pathogenic point mutations in human polypeptide-coding mtDNA genes have been described. Although mtDNA encodes 13 of the proteins needed for oxidative phosphorylation, the remainder are nuclear-encoded proteins that are synthesized on cytoplasmic ribosomes and are imported into the mitochondria, usually directed by an N-terminal mitochondrial targeting presequence.2 Thus, mutations in either mtDNA or nuclear DNA may impair mitochondrial function, resulting in human disease.3 Of all mitochondrial diseases, LHON is the most common.4 Three mtDNA mutations (G3460A, G11778A, and T14484C) account for 95% of LHON cases, with
the G11778A mutation being the most common, accounting for 50% of LHON cases.4,5 Each LHON mutation affects a different subunit of the nicotinamide adenine dinucleotide:ubiquinone oxidoreductase (complex I) in the oxidative phosphorylation pathway, where electrons first enter the electron transport chain.6 This large enzyme consists of seven subunits (ND1, 2, 3, 4, 4L, 5, and 6) encoded by mtDNA whereas the remaining 35 subunits are nuclear encoded.7 Mitochondrial oxidative phosphorylation deficiency due to mutations in complex I subunit genes is believed to play a pivotal role in development of LHON, although the precise pathophysiological events precipitating acute visual failure and cellular injury remain elusive. Each LHON mutation alters mtDNAencoded intrinsic complex I membrane proteins, but surprisingly the standard spectrophotometric assays of complex I activity in LHON cells containing the G11778A mutation in the ND4 subunit gene are reduced slightly.8 –11 Only the G3460A mutation in the ND1 subunit gene reduces complex I activity markedly.9,11,12 However, clear evidence of complex I defi-
From the Departments of 1Ophthalmology and 2Neurology, University of Florida College of Medicine, Gainesville, FL; 3Department of Neurology and Genetics and Development, Columbia College of Physicians and Surgeons; 4Departments of Neurology and Neuroscience, Cornell University, New York, NY; 5Department of Neurological Science, University of Bologna, Bologna; 6E. Medea Scientific Institute, Conegliano Research Centre, Conegliano, Italy; and 7 Department of Molecular Genetics and Microbiology, University of Florida College of Medicine, Gainesville, FL.
Received Apr 11, 2002, and in revised form Jun 20. Accepted for publication Jul 1, 2002.
534
© 2002 Wiley-Liss, Inc.
Published online Oct 11, 2002, in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/ana.10354 Address correspondence to Dr Guy, Neuro-Ophthalmology Service, Box 100284, University of Florida, College of Medicine, Gainesville, FL 32610. E-mail:
[email protected]
ciency with all three pathogenic mutations comes from polarographic investigations, showing impairment of cellular respiration when driven by complex I–linked substrates.9 –11 How these different degrees of changes in complex I function result in the same clinical picture of almost simultaneous bilateral apoplectic visual failure during early adult life is unclear, but reductions in oxidative phosphorylation and cellular injury induced by reactive oxygen species are suspect.13,14 Unlike most other mitochondrial mutations that impair neurological and myocardial function and are often fatal, patients with LHON, though blind, have a normal life expectancy. Unfortunately, there is little propensity for spontaneous visual recovery in the G11778A LHON patients, and there is no effective therapy. One of many potential avenues for treatment is to utilize gene therapy to introduce a “normal” gene encoding the defective complex I subunit into the optic nerves of LHON patients. Although we have successfully imported exogenous genes into the nuclear genome to protect the optic nerve,15,16 these methods cannot be applied directly to similarly introduce genes into the mammalian mitochondrial genome. We therefore adapted the approach of allotopic expression,17 in which a nuclear-encoded version of a gene normally encoded by mtDNA (ND4 in this case) specifies a protein expressed in the cytoplasm that is then imported into the mitochondria.18 This approach recently has been successfully applied to a mammalian system.19 Mammalian mitochondria use a genetic code that is partially different from the universal genetic code. Simply transferring a “normal” mitochondrial ND4 gene to the nucleus would result in translation of a truncated protein, because the UGA codon that directs insertion of a tryptophan in the mitochondria is a stop codon in the nuclear genetic code. Therefore, changes in the coding sequence of a mitochondrial ND4 gene are needed to make it compatible with the “universal” nuclear code. Moreover, if mtDNA genes are to be expressed in the nucleus, promoter, enhancer, and polyadenylation signals also must be added to the complex I subunit gene, along with a mitochondrial targeting peptide to direct proper trafficking and import of the protein from the cytoplasm into the mitochondria. In this report, we describe the use of recombinant adenoassociated virus (AAV) as a vector for nuclear complementation of the G11778A mtDNA mutation to rescue the mitochondrial oxidative phosphorylation deficiency of LHON. Materials and Methods Construction of Recoded ND4F and Adenoassociated Virus Vectors To construct the fusion gene containing the mitochondrial targeting sequences (MTSs) and epitope tag, we created synthetic 80 mer oligonucleotide pairs in the nuclear genetic
code and codons prevalent in highly expressed nuclear genes to conserve amino acid sequence. The synthetic oligonucleotides were overlapped by approximately 20 complementary nucleotides serving as primers for polymerase chain reaction with the high fidelity of Pfu Turbo DNA polymerase (Stratagene, La Jolla, CA) until the entire 1,377-nucleotide nuclear-encoded ND4 gene was constructed. Using this technique, we then fused the ND4 gene in-frame to the ATP1 or aldehyde dehydrogenase (Aldh)–targeting sequences and FLAG or green fluorescent protein (GFP)18 epitope tags. Flanking XbaI (P1ND4Flag) or AflII and HindIII (AldhND4GFP) restriction sites were added for cloning into AAV vectors. Base deletions and substitutions in the reading frame were corrected using the QuickChange in vitro mutagenesis kit (Stratagene, La Jolla, CA). The entire reading frame of the P1ND4Flag fusion gene was cloned in the XbaI sites of AAV plasmid vectors pTR-UF11 (regulated by the 381bp cytomegalovirus immediate early gene enhancer 1,352bp chicken -actin promoter, exon 1 and intron 1). The AldhND4GFP was similarly constructed, but with flanking AflII and HindIII sites for cloning into pTRUF5.18 COX8GFP was constructed and inserted into pTRUF5.18 To generate mitochondrially targeted expression of P1ND4Flag and cytoplasmic-targeted expression of GFP in the same cell, we used the pTR-UF12 vector that had P1ND4Flag linked to GFP via a 637bp poliovirus internal ribosomal entry site (IRES). Both vectors have a splice donor/acceptor site from SV40 (16S/19S site) located just upstream of the coding sequence to aid in the nuclear expression of and transport of the message. Visualization of cytoplasmic GFP enabled us to identify conveniently those cells that were also expressing P1ND4Flag, which had been inserted upstream of the IRES. The plasmids were amplified and purified by cesium chloride gradient centrifugation and then packaged into recombinant AAV (rAAV) by transfection into human 293 cells using standard procedures; the rAAVs were titered by an infectious center assay.20
Cell Culture and Viral Transfection The study of the pathophysiology of mtDNA mutations has taken advantage of the use of transmitochondrial hybrid cell lines known as cybrids.21 Cybrids are created by fusion of enucleated cells from patients with mutated mtDNA, in this case the G11778A mutation, with cells that have permanently lost their mtDNA after chronic exposure to ethidium bromide. This procedure results in the production of a cell line with the mutated mtDNA of the patient and the “neutral” nuclear DNA of the host cell line. Homoplasmic osteosarcoma (143B.TK⫺)–derived cybrids containing wild-type (11778G) or mutated (11778A) mtDNA were constructed and cultured as previously reported.8 For AAV infections, cybrids at approximately 80% confluency were transfected with 1g of DNA with TransIT Transfection Reagent (Mirus, Madison, WI) or 3.0 ⫻ 107 AAV or rAAV viral particles in complete high-glucose medium. Selection in galactose was performed in 10 separate wells, with the cells treated with selective medium for 3 days. Cells were trypsinized and counted using an automated Coulter (Hialeah, FL) Z-100 particle counter.
Guy et al: Leber Hereditary Optic Neuropathy
535
Immunological Techniques For immunohistochemistry, the transfected cybrids were trypsinized and grown on glass slides. After the cells reached confluence, they were incubated for 30 minutes with 250nM of the mitochondrial-specific fluorescent dye MitoTracker Red (Molecular Probes, Eugene, OR). Immunostaining with mouse monoclonal anti–FLAG M2 antibodies (Sigma, St. Louis, MO) or anti–GFP antibodies (ClonTech, Palo Alto, CA) was performed. Secondary anti–mouse Cy5 or Cy2 and anti-rabbit Cy2 (Jackson Immunochemicals, Bar Harbor, ME) were used for immunodetection. Immunofluorescence was visualized in a Bio-Rad (Richmond, CA) confocal microscope. The collected digital images were pseudocolored red for MitoTracker, blue or green for FLAG or green for GFP then merged in red-green-blue (RGB) format for evaluation of colocalization. For Western blot analysis, sonicated proteins from total cellular lysates obtained from the transfected and restrictive media selected cells were electrophoresed through a 10% polyacrylamide gel and electrotransferred to a polyvinylidene fluoride membrane (Bio-Rad). The membrane was immunostained with mouse monoclonal anti–FLAG M2 antibodies and then with rabbit anti–mouse IgG alkaline phosphataseconjugated secondary antibodies. Immune complexes were detected by nitro-blue-tetrazolium chloride/5-bromo-4chloro-3-indolylphosphate toludine salt (NBT/BCIP).
Oxidative Phosphorylation Assays Assays of complex I (⫹III) activity were performed on P1ND4Flag and mock-transfected cybrids in whole permeabilized cells by the reduction of cytochrome c with nicotinamide adenine dinucleotide and additionally in the presence of the inhibitor rotenone.22 ATP synthesis was measured by a luciferin-luciferase assay in whole permeabilized cells using the complex I substrates malate and pyruvate or the complex II substrate succinate.23 ATP synthesis with malate and pyruvate or succinate was also measured after the addition of 10ng/ml oligomycin to test for sensitivity to low doses of a specific ATPase inhibitor.
Results Strategy for Allotopic Expression of ND4 To accomplish allotopic complementation we synthesized the full-length version of nuclear-encoded ND4 converting the “non-standard” codons read by the mitochondrial genetic system to the universal genetic code. The nucleotide sequence of the recoded ND4 was 73% homologous with the mitochondrial version of the ND4 gene, whereas the amino acid sequences encoded by both genes were identical. Therefore, our synthetic ND4 gene encodes for a “normal” ND4 protein that is identical to the ND4 protein synthesized within mitochondria. However, our recoded ND4 protein is synthesized in the cytoplasm. To direct the import of the recoded ND4 protein into the mitochondria from the cytoplasm, we added a MTS specifying the N-terminal region of either (1) the P1 isoform of subunit c of human ATP synthase (ATPc) containing
536
Annals of Neurology
Vol 52
No 5
November 2002
the entire 61–amino acid MTS plus the first five amino acids of the mature P1 polypeptide24 or (2) the Aldh containing the first 19 –amino acid MTS.25 For detection of import, we added to the C terminus of the P1ND4 gene the short FLAG epitope tag (24 nucleotides) or to the AldhND4 gene the larger GFP tag (718 nucleotides). Although we began our mitochondrial import studies first with GFP as the epitope tag, we switched to the much smaller FLAG tag. Even though GFP was successfully imported into mitochondria by a MTS fused to the N terminus thus making successful transfection easily detectable in living cell culture, when GFP was fused to the C terminus of a recoded mitochondrial gene (ATP6 or ND6) import of the fusion protein was unsuccessful.18 To achieve stable and efficient expression of the fusion gene in cells, we inserted P1ND4Flag into AAV vectors pTR-UF11 and pTR-UF12. Transgene expression in both vectors is driven by the chicken -actin promoter and cytomegalovirus enhancer. In addition, pTR-UF12 also contains an IRES linked to GFP for identification of transfected cells in living cell cultures. Thus, GFP (lacking a MTS) is expressed only in the cytoplasm, whereas the P1ND4Flag fusion protein is expressed in the mitochondria of the same cell. Unlike plasmid transfection that requires the addition of chemical reagents to facilitate DNA entry into cells and produces only transient and somewhat inefficient expression of the introduced gene, viral-mediated gene transfer permits efficient delivery of genes into cells for assays of transgene function.26 Moreover, in the case of AAV, the transferred DNA sequences may be integrated stably into the chromosomal DNA of the target cell for long-term expression of the transgene in vivo in living cells, organs, and tissues.15,16 Detection of Allotopic Expression in Cells Containing Mutated Mitochondrial DNA Homoplasmic human cybrid cells containing the mitochondria of patients harboring the G11778A mutation in mtDNA transfected with rAAV containing the P1ND4Flag fusion gene expressed the fusion polypeptide (Fig 1). The ATPc MTS directed the allotopically expressed ND4F polypeptide into mitochondria. Immunocytochemistry to detect the FLAG epitope inserted at the C terminus of the imported ND4 showed a typical punctate mitochondrial pattern that colocalized with the mitochondrion-specific dye MitoTracker Red, thus implying the recoded ND4Flag was imported into mitochondria (Fig 2). Cells transfected with P1ND4Flag in AAV vector UF-11 showed mitochondrially targeted FLAG (see Fig 2D) colocalized with MitoTracker Red (see Fig 2A) in the merged panel (see Fig 2J). Cells transfected with P1ND4Flag in AAV vector UF-12 that contained the IRES linked to GFP showed mitochondrially targeted FLAG and
Fig 1. Illustration and immunoblotting of the P1ND4FLAG construct in UF-11. Diagram showing the nuclear-encoded ND4 in AAV vector UF-11 (top). Cellular infection with this construct should result in the synthesis of a 52kD polypeptide, the molecular weight of the ND4Flag. (bottom) Western blot of ND4Flag-transfected G11778A cybrids (lanes 1– 4) shows a 52kD band consistent with expression of the ND4Flag fusion polypeptide (lanes 2, 3), whereas the control (untransfected cells; lanes 5– 8) shows no staining with the anti–FLAG antibody. Stained gel shows corresponding protein loading with successive 1 log unit dilutions (bottom half). Overloading of lane 1 by cellular protein is readily apparent by the absence of any discrete pattern of protein bands in the stained gel. This is in contrast with lane 2 in which discrete bands are best seen and the intensity of anti-Flag immunostaining was optimized. CMV ⫽ cytomegalovirus; TR ⫽ terminal repeat; CBA ⫽ chicken B actin.
cytoplasmic GFP in the same cell. Cells mock- transfected with AAV vector UF-11 driving GFP expression in the place of the P1ND4Flag gene exhibited diffuse cytoplasmic staining of GFP only (see Fig 2H). Last, when ND4 with the Aldh MTS was linked to GFP, rather than to FLAG, the ND4GFP fusion did have a punctate staining pattern mimicking import into mitochondria (see Fig 2I), but relatively poor colocalization of GFP with MitoTracker Red (see Fig 2I) suggested this fusion protein was not imported. Allotopic ND4 Improves Cybrid Cell Survival Although P1ND4Flag was expressed and imported into mitochondria, would allotopic complementation with this protein improve the defective oxidative phosphorylation of LHON? To answer this question, homoplasmic cybrid cells harboring mutant mtDNA (ie, 100% G11778A derived from a patient with LHON inserted into a neutral nuclear background) were transfected with rAAV containing the P1ND4Flag or mocktransfected with the same AAV plasmid lacking the allotopic insert and expressing GFP (UF-11). Immediately after the transfection, cells were grown in glucoserich media for 3 days and then placed in glucose-free
media containing galactose as the main carbon source for glycolysis. This media forces the cells to rely predominantly on oxidative phosphorylation to produce ATP.27 Cells harboring complex I mutations have a severe growth defect compared with wild-type cells in such medium.26 We found that cybrid cell survival after 3 days in the glucose-deficient galactose media was threefold greater for the allotopically transfected P1ND4Flag cybrids than were the cybrids transfected with the mock AAV ( p ⬍ 0.001; Fig 3A). Apparently, in the mutated cybrids this selection enriched for cells that expressed higher levels of P1ND4Flag, suggesting these cells likely had improved oxidative phosphorylation. Oxidative Phosphorylation Deficiency Rescued by Allotopic ND4 Consistent with the finding that spectrophotometric assays of complex I activity do not discriminate between wild-type cells and G11778A mutant cybrids,9 –11,28,29 transfection with P1ND4Flag did not increase complex I activity (see Fig 3B). These results are in accord with published observations that the impact of the G11778A LHON mutation on complex I–specific activity in cell
Guy et al: Leber Hereditary Optic Neuropathy
537
Figure 2
538
Annals of Neurology
Vol 52
No 5
November 2002
lines appears to be mild.11,29 Therefore, we focused on changes in ATP synthesis using malate and pyruvate as complex I substrates for oxidative phosphorylation (see Fig 3C).10 It has been shown that respiration of G11778A cell lines is reduced with complex I substrates but may be increased with complex II substrates due perhaps to compensatory regulation of the nuclearencoded complex II.9,10,30 Consistent with these observations, we found that relative to the wild-type cell line with normal mtDNA, cybrid cells containing the LHON G11778A mutation in mtDNA showed a 60% reduction in the rate of complex I–dependent ATP synthesis ( p ⬍ 0.005).31–33 Moreover, using the complex II substrate succinate that bypasses the mutated complex I, we found that ATP synthesis in G11778A cybrids increased fivefold (82 nm ATP/min/106 cells with succinate vs 15 nm ATP/min/106 cells with malate and pyruvate; p ⬍ 0.02). However, in the wild-type cell line containing normal mtDNA, the rates of ATP synthesis obtained with either complex I or complex II substrates were virtually identical (30.8 nm ATP/min/106 cells with succinate vs 31.4 nmATP/min/106 cells with malate and pyruvate). Although complex II–dependent ATP synthesis was actually increased more than twofold ( p ⬍ 0.05) in our LHON cybrids relative to the wildtype cell line, this finding was likely compensatory as previously demonstrated.9,10,30 We therefore focused our attention on the main problem, the deficiency in complex I–dependent ATP synthesis induced by the G11778A mutation in the mitochondrial gene for complex I. Such substantial reductions in ATP synthesis likely contribute to the development of optic neuropathy in LHON patients with the G11778A mutation, but would allotopic expression of a normal ND4 gene rescue the substantial deficiency in complex I–dependent ATP synthesis of LHON cybrids? Indeed, relative to G11778A cybrids transfected with a mock AAV vector lacking the P1ND4Flag gene, P1ND4Flag complemented G11778A cybrids showed a threefold increase in the rate of complex I–dependent ATP synthesis. This degree of recovery led to levels of ATP synthesis that were virtually indistinguishable from the corresponding wild-type cell line containing normal mtDNA. Although the level of transfection by AAV containing P1ND4Flag
Š
is somewhat variable, as shown by higher standard deviations obtained with this construct, the differences between P1ND4Flag and mock-transfected cybrids were statistically significant ( p ⬍ 0.02); thus, P1ND4Flag has a major impact on ATP synthesis. In contrast, when the AldhND4GFP construct was tested, cytoplasmic expression of ND4 had no impact on ATP levels, as predicted by the lack of mitochondrial import (see Fig 2I). Discussion We provide evidence here that ATP synthesis dependent on complex I substrates is substantially reduced in transmitochondrial cybrids containing homoplasmic levels of G11778A-mutated mtDNA and more importantly that complementation with a normal ND4 gene can rescue this severe deficiency in oxidative phosphorylation even in the presence of the mutated mtDNA. This successful restoration of complex I–dependent respiration to a cell line still harboring the LHON G11778A mutation by allotopic expression of a normal ND4 subunit gene may be a promising step in a genebased treatment for this blinding disorder. However, before this form of therapy can be applied to patients, several important issues need to be addressed, not the least of which is the precise pathophysiology causing LHON. Recent studies showing neuronal cell lines are more severely affected by the G11778A mtDNA mutation imply a relative vulnerability of neurons to this mutation,34 but why then is only the optic nerve destroyed in patients with the LHON mutation? Clearly, optic nerve degeneration is a feature of several different human diseases that include Leigh’s syndrome, infantile bilateral striatal necrosis, Friedreich’s ataxia, and dominant optic atrophy (OPA1), as well as LHON. Although each of these diseases affecting the optic nerve has in common mutated mitochondrial proteins, the selective vulnerability of the optic nerve in LHON remains somewhat a mystery.35 This key issue remains elusive in large part because of the absence of an LHON animal model to investigate the pathogenic effects of the G11778A mtDNA mutation in vivo in tissues rather than isolated in vitro in cultured cells. The lack of an LHON animal model is truly unfortunate; without one we are unable to test whether allotopic
Fig 2. Immunocytochemistry of G11778A Leber hereditary optic neuropathy cybrids. The cellular localization of mitochondria visualized by MitoTracker Red (A–C) and FLAG visualized by indirect immunofluorescence using antibodies to FLAG (D–F), green fluorescent protein (GFP; G–I), and the merged images (J–L). Cells were transfected with P1ND4Flag inserted into the UF-11 adenoassociated virus (AAV) vector (column 1), the parent UF-11 vector (with no mitochondrial targeting sequence [MTS]; column 2), and AldhND4GFP inserted into UF-5 (column 3). Indicative of mitochondrial import, cells transfected with P1ND4Flag show mitochondrially targeted FLAG colocalizes with MitoTracker Red (J). In contrast, cells mock-transfected with the same AAV vector driving GFP expression in the place of the P1ND4Flag gene and lacking a mitochondrial targeting sequence exhibit diffuse cytoplasmic staining of GFP only (H). This was not imported into mitochondria (K). Another construct, ND4 linked to GFP with the aldehyde dehydrogenase (Aldh) MTS, exhibited a punctate staining pattern (I), the relatively poor colocalization of GFP with MitoTracker Red (L) suggested this ND4GFP fusion protein was not imported. Maps of the constructs used are shown below the micrographs.
Guy et al: Leber Hereditary Optic Neuropathy
539
expression with a normal ND4 can rescue the oxidative phosphorylation deficiency of LHON and thus reverse or prevent blindness in vivo. This paucity of animal models is not restricted to LHON but is common to most diseases associated with mutations in mtDNA.6 In fact, the only animal model with mutated mtDNA is a transgenic mouse created with a deletion of almost half of the entire mouse mitochondrial genome.36 Even if it were available to us, this animal is not well suited for our work on ND4 and LHON, as not only is the ND4 subunit gene deleted but three other subunits of complex I (ND4L, ND5, and ND6), one subunit of complex IV (COX 3), and two subunits of complex V (ATP 6 and ATP 8) are deleted as well.37 Thus, potential rescue of this animal would involve allotopically expressing all the deleted genes. Though technically possible, the deletion of six additional genes encoding transfer RNA genes involved in mitochondrial protein synthesis almost assures that even such an arduous task would not correct the multiple oxidative phosphorylation deficiencies of this “mito-mouse.” Therefore, currently we cannot test our strategy for allotopic rescue in animals. Who then, if anyone, should get AAV-mediated gene therapy with a normal ND4? Only 50% of male and 10% of female patients harboring the G11778A LHON mutation actually develop loss of vision.1,38,39 Clearly then, not every person with the G11778A mutation is a candidate for allotopic AAV-ND4 gene therapy. Injection of this gene into the visual system involves a very small risk for visual loss from the procedure itself. This potential risk may not be acceptable to patients with the G11778A mutation who do not have visual loss and may never actually develop LHON. Conversely, allotopically expressed ND4 gene therapy may be warranted for patients debilitated by the visual loss of LHON, but would an injection of the
Fig 3. Bar graphs of Leber hereditary optic neuropathy cybrid cell growth in selective media, complex I and complex V assays. (A) Cell survival, after 3 days of media selection, of G11778A cybrids and wild-type cells transfected with P1ND4Flag compared with the mock-transfected cells (mean ⫾ standard deviation [SD]; n ⫽ 10). (B) Bar graph showing complex I (⫹III) activity in whole lysed cells. Results are expressed as the total cellular complex I activity subtracted by the value obtained after the addition of the complex I inhibitor rotenone, giving the mitochondrial component of complex I activity (mean ⫾ SD; n ⫽ 3). (C) Bar graph showing the rate of ATP synthesis in permeabilized cells with pyruvate and malate serving as electron donors. Results are total ATP levels detected in a luciferin-luciferase assay and in the presence of oligomycin, an inhibitor of the mitochondrial ATP synthase (mean ⫾ SD). Wt ⫽ wild type; LHON ⫽ Leber hereditary optic neuropathy.
540
Annals of Neurology
Vol 52
No 5
‹
November 2002
normal ND4 gene restore visual function after it has deteriorated? The timing of AAV-mediated allotopic ND4 gene therapy after visual loss may be the key to its success, because once the optic nerve has degenerated too few cells may remain to be rescued.35 Transgene expression mediated by the AAV vector typically occurs quickly within days in vitro, but it takes 2 to 4 weeks in vivo and perhaps even longer for mitochondrial complementation of ND4 to deliver a clinically significant effect on cellular respiration. However, this time frame appears to be well within the limits of delivery to LHON patients who have lost vision in one eye but still retain good vision in the fellow eye. This companion eye usually becomes affected sequentially in 75% of patients, with a delay of approximately 2 months.39 The window for intervention is much narrower for the rest, who lose vision simultaneously in both eyes. Even for them, allotopic gene therapy may be beneficial if delivered quickly after visual loss and before atrophy of the optic nerve sets in, typically 6 to 8 weeks after loss of vision. AAV has been used successfully as a vector to deliver normal genes to the visual system,15 even forestalling or reversing blindness in animals,16,40 thus suggesting that rescue of LHON by allotopic ND4-mediated gene therapy is also feasible. Because AAV also has been used to deliver genes to other body organs and tissues, therapy with another allotopically expressed gene (ATP6) may be more apropos for other less common but lethal mitochondrial disorders such as maternally inherited Leigh’s syndrome and NARP that are currently untreatable.19 Although no other therapeutic option has shown such promise,41 can we move directly from rescue of a mitochondrial defect in cultured human cells to the treatment of mitochondrial disease in humans? Can afflicted patients afford to wait while we contemplate?
This work was supported by grants from the National Institutes of Health (EY12335, J.G.; EY11123, NS36302, WWH; and NS28828, E.A.S.), Muscular Dystrophy Association (E.A.S. and F.P.), Foundation Fighting Blindness (W.W.H.) and Research to Prevent Blindness (J.G.). We thank M Wilson for editorial work in preparation of the manuscript.
References 1. Wallace DC, Singh G, Lott MT, et al. Mitochondrial DNA mutation associated with Leber’s hereditary optic neuropathy. Science 1988;242:1427–1430. 2. Hartl FU, Neupert W. Protein sorting to mitochondria: evolutionary conservations of folding and assembly. Science 1990; 247:930 –938.
3. Schon EA. Mitochondrial genetics and disease. Trends Biochem Sci 2000;25:555–560. 4. Chinnery PF, Johnson MA, Wardell TM, et al. The epidemiology of pathogenic mitochondrial DNA mutations. Ann Neurol 2000;48:188 –193. 5. Carelli V, Ghelli A, Bucchi L, et al. Biochemical features of mtDNA 14484 (ND6/M64V) point mutation associated with Leber’s hereditary optic neuropathy. Ann Neurol 1999;45: 320 –328. 6. Wallace DC. Mitochondrial diseases in man and mouse. Science 1999;283:1482–1488. 7. Sazanov LA, Peak-Chew SY, Fearnley IM, et al. Resolution of the membrane domain of bovine complex I into subcomplexes: implications for the structural organization of the enzyme. Biochemistry 2000;39:7229 –7235. 8. Vergani L, Martinuzzi A, Carelli V, et al. MtDNA mutations associated with Leber’s hereditary optic neuropathy: studies on cytoplasmic hybrid (cybrid) cells. Biochem Biophys Res Commun 1995;210:880 – 888. 9. Majander A, Huoponen K, Savontaus ML, et al. Electron transfer properties of NADH:ubiquinone reductase in the ND1/ 3460 and the ND4/11778 mutations of the Leber hereditary optic neuroretinopathy (LHON). FEBS Lett 1991;292: 289 –292. 10. Larsson NG, Andersen O, Holme E, et al. Leber’s hereditary optic neuropathy and complex I deficiency in muscle. Ann Neurol 1991;30:701–708. 11. Brown MD, Trounce IA, Jun AS, et al. Functional analysis of lymphoblast and cybrid mitochondria containing the 3460, 11778, or 14484 Leber’s hereditary optic neuropathy mitochondrial DNA mutation. J Biol Chem 2000;275:39831– 39836. 12. Cock HR, Cooper JM, Schapira AH. Functional consequences of the 3460-bp mitochondrial DNA mutation associated with Leber’s hereditary optic neuropathy. J Neurol Sci 1999;165: 10 –17. 13. Esposito LA, Melov S, Panov A, et al. Mitochondrial disease in mouse results in increased oxidative stress. Proc Natl Acad Sci USA 1999;96:4820 – 4825. 14. Brown MD. The enigmatic relationship between mitochondrial dysfunction and Leber’s hereditary optic neuropathy. J Neurol Sci 1999;165:1–5. 15. Guy J, Qi X, Muzyczka N, et al. Reporter expression persists 1 year after adeno-associated virus-mediated gene transfer to the optic nerve. Arch Ophthalmol 1999;117:929 –937. 16. Guy J, Qi X, Hauswirth WW. Adeno-associated viral-mediated catalase expression suppresses optic neuritis in experimental allergic encephalomyelitis. Proc Natl Acad Sci USA 1998;95: 13847–13852. 17. Gray RE, Law RH, Devenish RJ, et al. Allotopic expression of mitochondrial ATP synthase genes in nucleus of Saccharomyces cerevisiae. Methods Enzymol 1996;264:369 –389. 18. Owen R IV, Lewin AP, Peel A, et al. Recombinant adenoassociated virus vector-based gene transfer for defects in oxidative metabolism. Hum Gene Ther 2000;11:2067–2078. 19. Manfredi G, Fu J, Ojaimi J, et al. Rescue of a deficiency in ATP synthesis by transfer of MTATP6, a mitochondrial DNAencoded gene, to the nucleus. Nat Genet 2002;30:394 –399. 20. Hauswirth WW, Lewin AS, Zolotukhin S, et al. Production and purification of recombinant adeno-associated virus. Methods Enzymol 2000;316:743–761. 21. King MP, Attardi G. Mitochondria-mediated transformation of human rho0 cells. In: Attardi GM, Chomyn A, eds. Mitochondrial biogenesis and genetics. Vol 264. San Diego: Academic Press, 1996:313–334.
Guy et al: Leber Hereditary Optic Neuropathy
541
22. Trounce IA, Kim YL, Jun AS, et al. Assessment of mitochondrial oxidative phosphorylation in patient muscle biopsies, lymphoblasts, and transmitochondrial cell lines. Methods Enzymol 1996;264:484 –509. 23. Manfredi G, Spinazzola A, Checcarelli N, et al. Assay of mitochondrial ATP synthesis in animal cells. Methods Cell Biol 2001;65:133–145. 24. Higuti T, Kawamura Y, Kuroiwa K, et al. Molecular cloning and sequence of two cDNAs for human subunit c of H(⫹)ATP synthase in mitochondria. Biochim Biophys Acta 1993; 1173:87–90. 25. Ni L, Heard TS, Weiner H. In vivo mitochondrial import. A comparison of leader sequence charge and structural relationships with the in vitro model resulting in evidence for cotranslational import. J Biol Chem 1999;274:12685–12691. 26. Bai Y, Hajek P, Chomyn A, et al. Lack of complex I activity in human cells carrying a mutation in MtDNA-encoded ND4 subunit is corrected by the Saccharomyces cerevisiae NADHquinone oxidoreductase (NDI1) gene. J Biol Chem 2001;276: 38808 –38813. 27. Reitzer LJ, Wice BM, Kennell D. Evidence that glutamine, not sugar, is the major energy source for cultured HeLa cells. J Biol Chem 1979;254:2669 –2676. 28. Andreu AL, Tanji K, Bruno C, et al. Exercise intolerance due to a nonsense mutation in the mtDNA ND4 gene. Ann Neurol 1999;45:820 – 823. 29. Hofhaus G, Johns DR, Hurko O, et al. Respiration and growth defects in transmitochondrial cell lines carrying the 11778 mutation associated with Leber’s hereditary optic neuropathy. J Biol Chem 1996;271:13155–13161. 30. Yen MY, Lee HC, Liu JH, et al. Compensatory elevation of complex II activity in Leber’s hereditary optic neuropathy. Br J Ophthalmol 1996;80:78 – 81.
542
Annals of Neurology
Vol 52
No 5
November 2002
31. Yen MY, Lee JF, Liu JH, et al. Energy charge is not decreased in lymphocytes of patients with Leber’s hereditary optic neuropathy with the 11,778 mutation. J Neuroophthalmol 1998; 18:84 – 85. 32. Majander A, Lamminen T, Juvonen V, et al. Mutations in subunit 6 of the F1F0-ATP synthase cause two entirely different diseases. FEBS Lett 1997;412:351–354. 33. Lodi R, Taylor DJ, Tabrizi SJ, et al. In vivo skeletal muscle mitochondrial function in Leber’s hereditary optic neuropathy assessed by 31P magnetic resonance spectroscopy. Ann Neurol 1997;42:573–579. 34. Wong A, Cavelier L, Collins-Schramm HE, et al. Differentiation-specific effects of LHON mutations introduced into neuronal NT2 cells. Hum Mol Genet 2002;11:431– 438. 35. Carelli V, Ross-Cisneros FN, Sadun AA. Optic nerve degeneration and mitochondrial dysfunction: genetic and acquired optic neuropathies. Neurochem Int 2002;40:573–584. 36. Inoue K, Nakada K, Ogura A, et al. Generation of mice with mitochondrial dysfunction by introducing mouse mtDNA carrying a deletion into zygotes. Nat Genet 2000;26:176 –181. 37. Nakada K, Inoue K, Hayashi J. Mito-mice: animal models for mitochondrial DNA-based diseases. Semin Cell Dev Biol 2001; 12:459 – 465. 38. Newman NJ. Leber’s hereditary optic neuropathy. New genetic considerations. Arch Neurol 1993;50:540 –548. 39. Harding AE, Sweeney MG, Govan GG, et al. Pedigree analysis in Leber hereditary optic neuropathy families with a pathogenic mtDNA mutation. Am J Hum Genet 1995;57:77– 86. 40. Acland GM, Aguirre GD, Ray J, et al. Gene therapy restores vision in a canine model of childhood blindness. Nat Genet 2001;28:92–95. 41. Turnbull DM, Lightowlers RN. A roundabout route to gene therapy. Nat Genet 2002;30:345–346.
