Novel Therapeutic Approaches for Leber’s Hereditary Optic Neuropathy

Abstract

Many human childhood mitochondrial disorders result from abnormal mitochondrial DNA (mtDNA) and altered bioenergetics. These abnormalities span most of the mtDNA, demonstrating that there are no “unique” positions on the mitochondrial genome that when deleted or mutated produce a disease phenotype. This diversity implies that the relationship between mitochondrial genotype and clinical phenotype is very complex. The origins of clinical phenotypes are thus unclear, fundamentally difficult-to-treat, and are usually clinically devastating. Current treatment is largely supportive and the disorders progress relentlessly causing significant morbidity and mortality. Vitamin supplements and pharmacological agents have been used in isolated cases and clinical trials, but the efficacy of these interventions is unclear. In spite of recent advances in the understanding of the pathogenesis of mitochondrial diseases, a cure remains elusive. An optimal cure would be gene therapy, which involves introducing the missing gene(s) into the mitochondria to complement the defect. Our recent research results indicate the feasibility of an innovative protein-transduction (“protofection”) technology, consisting of a recombinant mitochondrial transcription factor A (TFAM) that avidly binds mtDNA and permits efficient targeting into mitochondria in situ and in vivo. Thus, the development of gene therapy for treating mitochondrial disease offers promise, because it may circumvent the clinical abnormalities and the current inability to treat individual disorders in affected individuals. This review aims to focus on current treatment options and future therapeutics in mitochondrial disease treatment with a special emphasis on Leber’s hereditary optic neuropathy.

Introduction

Mitochondrial biology and disease

Mitochondria play a critical role in the life of the cell as they control metabolic rates, energy production, and cell death. Human mitochondrial DNA (mtDNA) is a 16.5 kilo base pair circular genome essential for maintenance of mitochondrial function and is present in many copies in most cell types. It encodes for 37 genes, 2 ribosomal RNAs, 22 tRNA’s, and 13 proteins of which 7 are components of complex I (NADH dehydrogenase), 3 of complex IV (cytochrome c oxidase), 2 of complex V (ATP synthase), and 1 of cytochrome b (a subunit of complex III) of the respiratory chain (Iyer et al., 2009a; Schon et al., 2012). Complexes I, III, and IV are the major components of the respiratory chain for generating the mitochondrial membrane potential due to their coupling of electron passage to proton pumping across the mitochondrial membrane. Energy is thus produced in the form of ~30 molecules of ATP for each molecule of glucose in a very efficient manner. Studies have shown that depletion of mtDNA to produce rho0 cells, resulting in loss of oxidative phosphorylation (OXPHOS) and subsequent loss of respiratory energy in the form of ATP (King and Attardi, 1989). Subsequent introduction of mtDNA into rho0 cells resulted in establishment of the OXPHOS activity and increase in respiratory energy (Gu et al., 1998). Thus, the number of mitochondria and the amount of mtDNA-encoded subunits influence the bio-energetic function of the mitochondrion and correlate directly with the energy requirements of any given cell type. There are ~4,900 copies of mtDNA per cell with both mutated and normal copies of mtDNA coexisting in each cell; a condition termed heteroplasmy. The presence of 100% mutated mtDNA is called homoplasmy.

Many devastating disorders arise from mutations or deletions in the mtDNA or improper energy production within the mitochondria. Classical mitochondrial diseases have mutation levels ranging from 30–100%, leading to a clear gene dosage effect (Rossignol et al., 2003). These disorders represent a large group of diseases with heterogeneous clinical and pathological symptoms characterized by improper functions of and sometimes irreversible damage to specialized cells present in tissues such as muscle, retina, and brain (Iyer et al., 2009a). The causes and mechanisms of cell death and related defects in many of these disorders, although not fully understood, likely derive from mutations in mtDNA or decline in energy levels. Clinical severity can be influenced by the percentage of normal versus abnormal mtDNA genomes present in affected cells (heteroplasmy) over time. In addition, the inability to directly manipulate the mitochondrial genome in situ has been an impediment for understanding and developing treatments for these mitochondrial disorders.

Leber’s hereditary optic neuropathy (LHON)

Being one of the most extensively studied mitochondrial disorders, Leber’s hereditary optic neuropathy (LHON) was first described by Theodre Leber in 1871 as blindness due to death of the optic nerve cells. Although many disorders are attributed to defects in the mtDNA, this review is focused on understanding the pathology and current treatment strategies for LHON disease with an emphasis towards novel therapeutic approaches for reversing visual loss in optic nerve cells. LHON usually begins in early childhood with rare cases appearing later in adulthood. In more than half of the LHON patients, blurring and clouding of vision in one or both the eyes is the first symptom of eye disorder, with intervals between affected eyes ranging from days to months. Over time, vision in both eyes worsens with a severe loss of visual acuity and color vision. This loss of vision is caused due to the death of cells in the optic nerve within the retinal ganglion cell layer while the surrounding epithelial cells and photoreceptors are spared. This disorder mainly affects central vision needed for reading, driving, and facial recognition. In some patients, central vision spontaneously improves in a small percentage of cases over a period of six-twelve months after onset of visual loss (summarized in Yu-Wai-Man et al., 2011).

The most important factor for visual recovery in patients with LHON is influenced by reduction in the heteroplasmic burden of abnormal mtDNA in affected cells over time. The abnormal mtDNA are primarily caused by mutations in the mitochondrial genome affecting the respiratory chain complexes. The first mtDNA point mutation identified to correlate with LHON disease is a G11778A nucleotide change within the MT-ND4 gene (Wallace et al., 1988). Other point mutations that have since been identified in different regions around the mitochondrial genome (G3460A in MT-ND1 gene and 14484T>C in MT-ND6 gene) causing LHON disease in the patient or their family members (Chinnery et al., 2001; Howell et al., 1991). Over the years, other LHON related mutations have been discovered (Achilli et al., 2012; Brown et al., 1995; Fauser et al., 2002; Gropman et al., 2004; Howell et al., 1998; Johns and Neufeld, 1993; Kim et al., 2002; Lamminen et al., 1995) (summarized in Table 1), and extensively documented in the human mitochondrial genome database: MITOMAP (http://www.mitomap.org). Results indicate that LHON disease manifests itself when the abnormal mtDNA is homoplasmic with >95% in most cases; although a minority of LHON patients are carriers with <14% of heteroplasmic mutant mtDNA (Man et al., 2002). The three mutations highlighted above and considered as the primary causes of LHON vary significantly in clinical severity due to mtDNA heteroplasmy in the mitochondria of retinal ganglion cells. Two of the LHON mutations, G3460A and G11778A, lead to reduction in enzyme activity of NADH while the T14484C mutation exhibits near normal activity of complex I (Brown et al., 2000; Man et al., 2002). In addition to abnormalities within the mitochondrial genome, there has been evidence of increased oxidative stress and apoptosis within the optic system which could potentially spark LHON disease (Koilkonda and Guy, 2011).

Table 1.

Summary of Mitochondrial DNA Mutation Associated with LHON.

Electron Transport	Genes	LHON MUTATION	References
Chain Complex
NADH:quinone	ND1	MTND1*LHON3460A	Howell N et al., 1991
oxidoreductase	ND2	MTND1*LHON3733A	Achilli A et al., 2012
(CI)	ND3	MTND1*LHON4171A	Kim JY et al., 2002
	ND4	MTND4*LHON11778A	Wallace DC et al., 1998
	ND4L	MTND4L*LHON10663C	Brown MD et al., 1995
	ND5	MTND6*LHON14484C	Chinnery PF et al., 2001
	ND6	MTND6*LHON14495G	Chinnery PF et al., 2001
		MTND6*LHON14568T	Fauser S et al., 2002
		MTND6*LHON14502C	Achilli A et al., 2012
		MTND6*LDYT14459A	Gropman A et al., 2004
		MTND6*LHON14482A	Achilli A et al., 2012
		MTND6*LHON14482G	Howell N et al., 1998
quinone:cytochrome c	CytB	MTCYB*LHON14831A	Fauser S et al., 2002
oxidoreductase (CIII)
cytochrome c oxidase	COX1	MTCO3*LHON9804A	Johns DR et al., 1993
(CIV)	COX2
	COX3
F0F1 complex	Atp6	MTATP6*LHON9101C	Lamminen T et al., 1995
(ATP synthase, CV)	Atp8

Current Therapeutic Approaches in Mitochondrial Disease Treatment

Pharmacological therapy

Treatments for mitochondrial disorders are very limited and mainly supportive while the disorders progress over the years. Most therapies are limited to vitamins and cofactors like CoenzymeQ₁₀ (CoQ₁₀) folic acid, vitamin B12, riboflavin, L-carnitine, and creatine; electron acceptors like vitamin C; free radical scavengers like idebenone, EPI-743, vitamin E, alpha lipoic acid, and curcumin; inhibitors of toxic metabolites (DCA or dicholoroacetate); and customized “mitococktail” tailored for individual patients (DiMauro and Mancuso, 2007; Parikh et al., 2009). Most of the evidence supports the use of a combination of mitococktails with antioxidant supplements to increase mitochondrial respiration and simultaneously scavenge free radicals to reduce reactive oxygen species and toxic acyl coenzyme A (acyl CoA) molecules produced in mitochondria diseases (Parikh et al., 2009). Some supplements also act as alternate energy fuels and possibly bypass the block within the respiratory chain complexes. The evidence for these and other treatments is still developing as no long-term benefits of any antioxidant therapy has been proven curative, while the benefits of treatment in individual patients remain limited and variable.

Idebenone treatment for LHON

As is widely known, oxidative stress increases free radical production and can lead to apoptosis. Therefore, free radical scavengers like CoQ₁₀, idebenone, and vitamin E have been prescribed by clinicians as part of treatment regimens for many different mitochondrial disorders. CoQ₁₀ is a longer-chain quinone analogue and is often used to treat patients with mitochondrial disorders. There is growing evidence in the literature supporting the beneficial effects of idebenone treatment in LHON (Klopstock et al., 2013). Idebenone, a short chain benzequinone, is structurally similar to CoQ₁₀ and protects the mitochondria from lipid peroxidation (Mashima et al., 1992). Unlike CoQ₁₀, idebenone can bypass mitochondrial complex I inhibition by transferring electrons from the cytosol to complex III, to stimulate energy production and decrease lactate production (Haefeli et al., 2011). Initial reports in 1992 were limited to isolated case studies where researchers reported improvement in both eyes for a 10 year old child with G11778A mutation, after treatment with idebenone for a year (Mashima et al., 1992). Yet the study was inconclusive given the early onset of the disease thus raising the possibility of spontaneous recovery (Cortelli et al., 1997). Mashima et al. (2000) reported a larger anecdotal study of LHON combined with vitamin B2, vitamin C, and idebenone therapy. Their findings indicated a consistent limited improvement in recovery of symptoms in some cases, whilst accelerating the rate of spontaneous recovery. It was conceivable that inadequate amount of substances reached the mitochondria to cause for these experimental failures. Most recently in 2011, the result of a large scale randomized placebo-controlled trial (RHODOS, Rescue of Hereditary Optic Disease Outpatient Study; clinical trial number: NCT01421381) for idebenone treatment in LHON was reported (Klopstock et al., 2011). Eighty five patients carrying mostly mtDNA mutations (G3460A, G11778A, and T14484C) were treated with 900 mg doses of idebenone per day for 24 weeks with no adverse side effects. The primary endpoint was the visual recovery of blindness. The results indicated that patients at early stage of the disease with different visual acuity and at highest risk of vision loss in the least affected eye were most likely to benefit from the idebenone treatment. The treatment was least effective in patients with T14484C mtDNA mutations due to high spontaneous vision recovery in these patients. On the other hand, the treatment with idebenone versus placebo showed 20% improvement in patients who were unable to read any letter on the chart at the beginning of the study. Based on this study, it is difficult to arrive at conclusions on whether the treatment will benefit the entire LHON population. However, the findings from the RHODOS trial and a subsequent similar non-randomized trial for LHON (Carelli et al., 2011) are promising and offer hope for at least some of the LHON patients at early stages of the disease.

EPI-743 treatment for LHON

Since there was no effective drug for severe blindness in LHON, the FDA-approved drug EPI-743 was evaluated in 5 patients with LHON who were actively losing vision on an emergency trial basis. EPI-743 drug was primarily used to evaluate the safety and efficacy in these patients (Sadun et al., 2012). EPI-743 is structurally similar to CoQ₁₀ and idebenone with a modified benzene ring to improve efficiency by 1,000 to 10,000 fold than either drug in accepting electrons to reduce oxidative stress while improving mitochondrial function (Esposti et al., 1996). EPI-743 works by interacting with the enzyme NADPH quinone reductase (NQO1) to form stable hydroquinones with excellent antioxidant properties (Siegel et al., 1997). The initial results from EPI-743 appear promising where 4 out of 5 LHON patients with different mtDNA mutations showed an improvement in visual recovery based on various tests, like visual acuity, field, color vision, and other metrics (Sadun et al., 2012). These findings constitute the first proof of concept studies that demonstrated a reversal of loss of vision despite death of optic nerve cells. Although the mechanism of action is currently under investigation, these results provide hope for conducting a large scale controlled multicenter study in the LHON population.

Future Therapeutic Approaches in Mitochondrial Disease Treatment

Given the causative nature of multiple mtDNA mutations in LHON, a definitive treatment may in theory be achieved by replacement of mutated mitochondrial genomes. Existing methods for manipulation of mitochondrial genomes are very limited, and can be classified in two groups: direct and indirect.

Adenoviral mediated mitochondrial gene therapy for LHON

In the indirect approach, also termed allotopic expression, the gene of interest or therapeutic DNA sequence is transfected into the nucleus by successfully modifying a virus (adenovirus-associated virus or AAV) which directs the production of proteins normally expressed in the mitochondria. These proteins are then imported to the mitochondria by specific mitochondrial targeting sequence that ensures its uptake into the mitochondria thus complementing the genetic defect or mutant mtDNA sequence present inside the mitochondria. Possibly due to their high hydrophobicity, to date only four mitochondrial genes have been corrected in this manner, in plants (Pineau et al., 2005), yeast (Roucou et al., 1999), and in mammalian cell culture (Guy et al., 2002; Oca-Cossio et al., 2003; Zullo et al., 2005). Allotopic expression technique has been used to complement the mutated ND4 gene in a cybrid cell line containing the homoplasmic LHON G11778A mutation resulting in normalization of ATP production and respiration, to levels identical to healthy cybrid cells (Guy et al., 2002). Using the same allotopic approach a second therapeutic gene superoxide dismutase (SOD2) was used to dramatically reduce oxidative stress in LHON cybrid cultures (Qi et al., 2007b).

Two independent groups demonstrated that allotopic expression of the mutant subunit of human ND4 gene (G11778A) caused mitochondrial dysfunction, increased oxidative stress, and induced cell death of retinal ganglion cells in both rat and mouse models of LHON. At the same time introduction of the normal wild-type ND4 gene into mitochondria did not cause additional toxicity and appeared to be safe in both rodent models of LHON. These results, although limited to just one mtDNA mutation, could still be useful for treating LHON patients carrying the G11778A mutant mtDNA in one or both eyes in the near future (Qi et al., 2007a; Guy et al., 2009; Ellouze et al., 2008). For these treatments to be translated from bench to the patients, safety and efficacy trials need to be conducted. Furthermore, additional research needs to be conducted to demonstrate that allotopically expressed proteins are fully incorporated into the electron chain subunits present in the inner membrane of the mitochondria or it could result in increased free radical production. Furthermore, the complex issue of controlling expression of allotopic mitochondrial genes would have to be solved before these techniques would be feasible in humans. At the time of this review, clinical trials (e.g., NCT01267422) are currently underway to assess the safety and efficacy of rAAV2-ND4 treatment of LHON with G11778A mutation in humans.

Recombinant protein mediated mitochondrial gene therapy for LHON

In the direct approach, DNA is delivered to the mitochondrial matrix and is used to express proteins or RNAs. Delivery of mtDNA in yeast was achieved using a biolistic (biological ballistic) gun (Anziano and Butow, 1991) but it has thus far failed in mammalian cells, and would be of little use in LHON. Although DNA conjugated to dequalinium was shown to concentrate in the vicinity of mitochondria, there was no proof of mitochondrial delivery (D’Souza et al., 2003). Subsequent studies focused on peptide motifs called MLS (Mitochondrial Leader Sequence) that are involved in mitochondrial import of many cytoplasmic proteins (Wiedemann et al., 2004). Results indicated their capability in directing other cargoes to mitochondria, including PNAs (Peptide Nucleic Acids) (Chinnery et al., 1999) and oligonucleotides (Flierl et al., 2003). The work of Gaizo and colleagues further supported the use of PTDs (Protein Transduction Domains) for targeted protein delivery to mitochondria (Del Gaizo et al., 2003; Del Gaizo and Payne, 2003). PTDs are short amphipathic helices consisting of highly basic amino acids, such as lysine and arginine, which when tagged to proteins allowed them to enter into cells of choice. Studies later demonstrated that a fluorescently tagged protein containing a PTD and an MLS sequence could enter the mitochondria of cells and tissues of choice (Del Gaizo et al., 2003; Del Gaizo and Payne, 2003). Studies have also shown that protein transduction domain sequences could be used to directly transfer DNA into cells and animals. Initial studies showed that engineered PTDs with cationic liposomes improved gene transfer (Hyndman et al., 2004), while others constructed branched peptides containing PTDs capable of binding DNA and delivering to cells (Hashida et al., 2004). The PTD linked proteins administered orally, or by injections into the portal vein in mouse, have ensured their capability of entering or delivering DNA into animals (Cai et al., 2006). These findings suggest that the PTD approach may engender a new era of gene therapy modalities with a high probability of human therapeutic use. Although these indirect and direct methods provided initial glimpses of therapeutic possibilities, there have been no reports of any viable full length mitochondrial DNA delivery technology for development of gene therapy for LHON or other mitochondrial disorders.

We have recently developed a novel mitochondrial transduction technology “protofection” that requires recombinant engineered protein vehicle for introduction of full length human mtDNA into mitochondria (Iyer et al., 2009b; Khan and Bennett, 2004). TFAM (transcription factor A, mitochondrial) is an abundantly expressed protein present in mitochondria that protects the mtDNA from oxidative stress by binding to it in a non-sequence specific manner (Kanki et al., 2004). The “protofection” protein TFAM has been engineered to contain an N-terminal protein transduction domain (PTD) -- stretch of 11 arginines -- a strong mitochondrial localization sequence (MLS) derived from mitochondrial super oxide dismutase 2 (SOD2) gene sequence, yielding the final recombinant protein rhTFAM (Iyer et al., 2009b).

We successfully demonstrated introduction of labeled rhTFAM and healthy mtDNA complexed with rhTFAM into homoplasmic LHON cybrid cells containing the G11778A mutation within 39 minutes and ~110 minutes (Iyer et al., 2012a). Further results in LHON cybrid cells, demonstrated an increase in mitochondrial genome replication, transcription, translation, and respiration initiated within a week when the complex was introduced into the mitochondria. We also observed the activation of the mitochondrial biogenesis (creation of new mitochondria) program in these human LHON cybrid cells. As further proof of application of this approach towards introduction of mtDNA, we also conducted studies where pathogenic LHON mtDNA carrying the G11778A mutation was introduced and expressed into human neural progenitor stem cells devoid of their own mtDNA, without affecting their self renewal properties or neuronal differentiation potential. These seminal studies demonstrated the potential for delivering and transcribing the introduced mtDNA into mitochondria in the short term in human neural stem cell models (Iyer et al., 2009a; 2012b). It is expected that this mitochondrial genome manipulation approach based on introduction of exogenous normal or pathogenic mtDNA provides hope for LHON patients afflicted with other mutations in the mitochondrial genome (Figure 1).

Figure 1.

Mitochondrial genome manipulations that result in (A) cell based models for understanding LHON pathophysiology (Iyer et al., 2009a; 2012b) and (B) therapeutic approaches for LHON treatment (Iyer et al., 2012a; 2009b). MTD, mitochondrial transduction domain; TFAM, transcription factor A, mitochondrial; ddC, dideoxycytidine; hNP, human neural progenitor.

In an in vivo study, we also injected rhTFAM into the tail vein of normal adult mice and assayed for motor endurance and increase in mitochondrial respiration. Significant increase in mitochondrial complex I respiration in brain and muscle mitochondria isolated from rhTFAM treated mice (Iyer et al., 2009b) was observed. While our preliminary findings are encouraging for the therapeutic potential of systemically delivered rhTFAM to increase mitochondrial function in vivo, much work remains to be done in translating the results into long term studies using mtDNA complexed with rhTFAM in LHON disease mouse models.

Recent studies have shown another form of gene therapy, where the mitochondrial genome is replaced in monkey oocytes by transferring the nucleus from the mother’s egg to an enucleated, mitochondrial-replete egg of the donor using a specialized spindle replacement technique (Tachibana et al., 2013). The study showed that reconstructed oocytes with replaced mitochondria were capable of supporting normal fertilization, produced embryonic stem cell lines, and also produced healthy offspring. Although this approach proposed mitochondrial genome replacement, the end goal is an option to prevent mtDNA disease transmission in affected families and not as a curative strategy for people already suffering from LHON diseases.

In summary, although gene therapy holds great promise for treatment for LHON patients with mtDNA mutations spanning the entire mitochondrial genome, a lot of work needs to be done before ongoing research can be translated into clinical trials and eventual treatments. The delivery of vectors to affected tissues and change in levels of mtDNA heteroplasmy must be regulated. Also, the duration and dosage of gene therapy must be regulated because current novel approaches have yet to optimize the long term effects and maintenance of the genome of interest in cells or animal models. It is also important to conduct safety and efficacy trials in pre-clinical animal models to test for immunological and other side effects before initiation of testing in human patients. The multiple gene therapy strategies we have highlighted are indeed exciting developments that with further refinement and optimization have potential for development of LHON treatment strategies.

Conclusion

The current studies indicate that the mitochondrial genome can be manipulated and lead to improvement in mitochondrial function in in vitro and in vivo models. Future coordinated efforts between scientists and clinicians are necessary to translate these findings towards development of therapies for LHON patients. Equally important is involvement of patients and caregivers in this process. As researchers, it is important that we update patients and caregivers on the basic principles of mitochondria genetics along with providing current information on various treatment options available. For example, it would be useful to understand that if the mother is homoplasmic for LHON mtDNA mutation, transfer of the genetic mutation to her children is highly possible; but the degree of clinical severity would depend on the percentage of heteroplasmic mutant mtDNA molecules present in her affected child and might not necessarily reflect the symptoms present in the mother. Thus, an understanding of the mechanism of mitochondrial DNA transmission in humans is important in the context of mitochondrial genome manipulation at different stages in early development as it undergoes a “bottleneck” effect where the transmission of mutation may be magnified or diminished when mtDNA transits from the fertilized oocyte into the preimplantation embryo, the fetus, and offspring (St John et al., 2010).

With the advent of increased genetic testing and next generation sequencing approaches, rare mtDNA mutations affecting LHON are being identified (Sundaresan et al., 2010). However, many aspects of the complex etiology of LHON remain poorly defined at present. The incomplete penetrance of LHON disease is that ~50% of males and ~10% of females who harbor one of the three primary mutations actually develop the optic neuropathy. This result clearly indicates mtDNA mutation is insufficient on its own for disease manifestation (Carelli et al., 2013). The identification of the secondary factors modulating the phenotypic expression of LHON is currently an area of intense research. A better understanding of the relationship between mitochondrial genetics, biogenesis, respiration, and optic nerve cell death is also needed to clarify the still unclear pathophysiology of LHON for the development of effective therapies in the future.