Leber Optic Atrophy

A number sign (#) is used with this entry because Leber optic atrophy, also known as Leber hereditary optic neuropathy (LHON), can be caused by mutation in multiple genes encoded by the mitochondrial genome (mtDNA).

Description

LHON presents in midlife as acute or subacute central vision loss leading to central scotoma and blindness. The disease has been associated with many missense mutations in the mtDNA that can act autonomously or in association with each other to cause the disease. The 18 allelic variants are MTND6*LDYT14459A (516006.0002); MTND4*LHON11778A (516003.0001); MTND1*LHON3460A (516000.0001); MTND6*LHON14484C (516006.0001); MTCYB*LHON15257A (516020.0001); MTCO3*LHON9438A (516050.0001); MTCO3*LHON9804A (516050.0002 ); MTND5*LHON13730A (516005.0002); MTND1*LHON4160C (516000.0002); MTND2*LHON5244A (516001.0002); MTCOI*LHON7444A (516030.0001); MTND1*LHON3394C (516000.0004); MTND5*LHON13708A (516005.0001); MTCYB*LHON15812A (516020.0002); MTND2*LHON4917G (516001.0001); MTND1*LHON4216C (516000.0003); MTND1*LHON4136G (516000.0002); MTATP6*LHON9101C (516060.0003); MTND4L*LHON10663C (516004.0002). The first 17 of these variants are summarized in Table M1, MIM12.

As pointed out by Riordan-Eva and Harding (1995), although the plethora of mtDNA mutations identified in families with LHON had resulted in confusion as to the pathogenic significance of each mutation, it had been established that the 3 primary mutations at basepairs 11778 (516003.0001), 3460 (516000.0001), and 14484 (516006.0001) are present in at least 90% of families. The correlation between the 14484 mutation and a good visual prognosis provides not only hope for affected patients, but also an approach for further research into the pathogenesis of LHON.

Yu-Wai-Man et al. (2009) provided a detailed review of LHON and autosomal dominant optic atrophy (OPA1; 165500), with emphasis on the selective vulnerability of retinal ganglion cells to mitochondrial dysfunction in both disorders.

Clinical Features

LHON patients present with midlife, acute or subacute, painless, central vision loss leading to central scotoma. Neuroophthalmologic examination commonly reveals peripapillary telangiectasia, microangiopathy, disc pseudoedema, and vascular tortuosity; these features are observed in 58% of patients with the nucleotide pair (np) 11778 mutation and occasionally in their asymptomatic maternal relatives. The mean age of onset has been variously reported from 27 to 34 years with a range of 1 to 70 years. The eyes can be affected simultaneously or sequentially, with an average interval between eyes being affected of about 2 months. The progression of each eye can range from sudden and complete vision loss to progressive decline over 2 years, with a mean progression time of about 3.7 months. The final visual acuity can range from 20/50 to no light perception, with the less severe mutations (see Table M1, MIM12) having less extreme outcomes. Thus, the most severely impaired np 11778 patients may have no light perception, the most severe np 3460 patients may retain light perception, the severe np 15257 patients will perceive hand motions, and the severe np 14484 patients will be able to count fingers. The probability of visual recovery also varies in relation to the mutation, with only 4% of np 11778 patients showing recovery an average of 36 months after onset; 22% of np 3460 patients recovering after 68 months; 28% of np 15257 patients recovering after 16 months; and 37% of np 14484 patients recovering after 16 months (Newman, 1993; Newman et al., 1991; Johns et al., 1993).

Cullom et al. (1993) found that 2 of 12 patients previously diagnosed as having tobacco-alcohol amblyopia, based on a classic clinical presentation, tested positive for known LHON mutations, one for the 11778 mutation and one for the 3460 mutation. The fact that only a few patients who abuse tobacco and alcohol develop optic neuropathy has suggested an element of individual susceptibility (Carroll, 1944). Cullom et al. (1993) proposed that susceptibility may be the result of an LHON-associated mitochondrial mutation.

Sadun et al. (2004) reported the ophthalmologic findings in 192 eyes from 96 maternally related individuals from a 7-generation Brazilian pedigree with LHON and the 11778/haplogroup J mutation. The findings demonstrated a significant influence of environmental risk factors, particularly smoking, for developing LHON and for the severity of its clinical expression. However, smoking did not correlate with the subclinical abnormalities detected in carriers. Moreover, subclinical abnormalities were equally distributed between males and females.

Mann et al. (2000) reported peripheral retinal phlebitis in a patient with LHON and the 11778 mutation. In addition to bilateral central visual loss associated with headache, the patient had vitritis, vasculitis, and optic neuritis. She was initially thought to have multiple sclerosis, but laboratory studies failed to substantiate this diagnosis or any other cause of vasculitis. The authors concluded that their report supported the theory that LHON is a neuroretinopathy with a broad spectrum of genotype-specific phenotypes.

Sadun et al. (2000) investigated the nerve fiber spectrum in optic nerves of 2 patients with LHON. Total depletion of the optic nerve fibers varied from 95 to 99%. Light and electron microscopy revealed preferential loss of the smallest axons corresponding to the P-cells, the smaller retinal ganglion cells. The authors concluded that the loss of P-cells may explain the clinical features of dyschromatopsia, central scotoma, and preservation of pupillary light response in LHON patients.

In a review of the clinical and molecular genetic aspects of LHON, Huoponen (2001) pointed out that peripapillary microangiopathy is present from the beginning and disappears as the disease progresses toward the end stages.

Newman-Toker et al. (2003) reported 2 patients with LHON who developed subsequent worsening of visual acuity or visual field constriction several years after onset. One female patient, who presented at age 17 with 20/400 acuity in each eye, had further, painless visual field loss 8 years later; final acuities were light perception in her right eye and counting fingers at 4 feet in her left. A blood test showed that she had the 11778 mtDNA mutation. Another patient lost vision at age 35 in 1 eye. Visual acuity of 20/200 in the first eye improved over 6 months to 20/40. One year later, visual acuity in the second eye deteriorated to hand motion. One month later, the first eye deteriorated to 20/400. The patient had slow improvement in each eye over a decade to 20/30 in the right and 20/100 in the left. At age 78, acuities dropped to 20/800 in each eye. A blood test revealed the 14484 mtDNA mutation.

Although MRI of the brain and orbits are typically normal in patients with LHON, Phillips et al. (2003) described 2 patients with LHON in whom an MRI disclosed abnormal enhancement of the optic nerves or chiasm enlargement.

Barboni et al. (2005) used optical coherence tomography (OCT) to study retinal nerve fiber layer (RNFL) thickness in patients with LHON. On the basis of OCT data, the RNFL was thickened in early LHON (ELHON, when the duration of disease was shorter than 6 months) and severely thinned in atrophic LHON (ALHON, when the duration was longer than 6 months). RNFL was likely to be partially preserved in ALHON with visual recovery. The temporal fibers (papillomacular bundle) were the first and most severely affected; the nasal fibers seemed to be partially spared in the late stage the disease.

Savini et al. (2005) studied the RNFL thickness, as measured by OCT, in unaffected carriers with LHON mutations. They found a thickening of the temporal RNFL fibers in all subgroups of unaffected carriers. These differences reached statistical significance in patients carrying the 11778 mutation, whereas only a trend was detected in those with the 3460 mutation. Savini et al. (2005) concluded that their findings provided the first evidence indicating the preferential involvement of the papillomacular bundle in subclinical LHON and that males had a more diffuse involvement than females.

Reviewing the findings of Barboni et al. (2005) and Savini et al. (2005), Kerrison (2005) concluded that LHON in not necessarily a monophasic disease but may manifest a latent phase with axonal thickening and normal visual function that might precede clinically significant vision loss, an acute phase of axonal injury with clinically significant loss of visual function, and a chronic phase with spontaneous improvement of vision in some individuals and reduced likelihood or recurrence.

Ventura et al. (2007) investigated chromatic losses in asymptomatic carriers of the LHON 11778G-A mutation. Sixty-five percent of carriers had abnormal protan (303900) and/or deutan (303800) thresholds; some of those with higher thresholds also had elevated tritan (190900) thresholds (13%). Male carriers had color vision losses with the red-green pattern of dyschromatopsia typical of patients affected with LHON, which included elevation of tritan thresholds as well. This predominantly parvocellular (red-green) impairment was compatible with the histopathology of LHON, which affects mostly the papillomacular bundle. In contrast with male losses, female losses were less frequent and severe. In the most severe losses, the women had instances of diffuse defect. Ventura et al. (2007) suggested that hormonal factors may be of great importance in the pathophysiology of LHON.

Other Features

LHON patients and their maternal relatives have also been reported to manifest a variety of ancillary symptoms. Cardiac conduction defects have been noted in some families. Among the Finnish patients, preexcitation syndromes including Wolff-Parkinson-White and Lown-Ganong-Levine are common (Nikoskelainen et al., 1985). Prolongation of the corrected QT interval was also observed in an African American family with the np 11778 mutation (Ortiz et al., 1993). Various minor neurologic problems including altered reflexes, ataxia, and sensory neuropathy have been described as well as skeletal abnormalities. Interestingly, the np 11778 mutation has also been associated with multiple sclerosis in some families (Harding et al., 1992; Newman et al., 1991). In studies of 35 Japanese LHON families, Mashima et al. (1996) found that, as in Finnish families, the preexcitation syndromes, Wolff-Parkinson-White syndrome and Lown-Ganong-Levine syndrome, were found relatively commonly, being seen in 5 of 63 individuals (8%) with mtDNA mutations. Nikoskelainen et al. (1985) had found the preexcitation syndrome in 14 of 163 Finns (9%) with mitochondrial DNA mutations.

While visual loss is the primary and generally the only clinical manifestation in most LHON pedigrees with the np 11778, np 3460, np 14484, or np 15257 mutations, occasional individuals present with much more severe clinical disease with neurologic manifestations. The proband in 1 Swedish np 11778 pedigree experienced optic atrophy at age 37 and more severe neurologic disease at age 38 including bilateral lesions of the putamen on MRI, tremor, ataxia, posterior column dysfunction, dystonia, corticospinal tract dysfunction and extrapyramidal rigidity. Muscle biopsy of this individual revealed a subsarcolemmal increase in mitochondria as well as a few fibers exhibiting mitochondria with paracrystalline inclusions. These findings anticipate the wide range of clinical presentations observed in 2 LHON families with more deleterious mtDNA genotypes: an Australian pedigree harboring the MTND1*LHON4160C + MTND6*LHON14484C mtDNA haplotype and an American Hispanic family with the MTND6*LDYT14459A mutation (see Table M1, MIM12). The Australian pedigree is homoplasmic for both mutations yet includes individuals ranging from asymptomatic, through optic atrophy, to severe neurodegenerative disease. The most severe symptoms were observed in 9 of 56 maternal relatives and included headache, vomiting, focal or generalized seizures with a hemiparesis that generally resolves, and a cerebral edema (Wallace, 1970). Specific neuropathologic abnormalities were not found in 3 individuals who died, and 1 female who recovered is clinically normal as an adult but had 4 affected children (Howell et al., 1991). Other neurologic symptoms in this family included dysarthria, deafness, ataxia, tremor, posterior column dysfunction, corticospinal trait dysfunction, and skeletal deformities (Howell et al., 1991; Wallace, 1970). One branch of the pedigree harbored an additional homoplasmic mtDNA mutation, MTND1*LHON4136G, and showed milder clinical presentations. It was speculated that this second mutation may reduce the severity of the np 4160 and np 14484 mutations (Howell et al., 1991).

The American Hispanic family reported by Novotny et al. (1986) harbored a Native American mtDNA and was heteroplasmic for the MTND6*LDYT14459A mutation (Jun et al., 1994). Maternal relatives in the pedigree ranged from normal, through adult-onset optic atrophy, to pediatric dystonia associated with bilateral striatal necrosis. One interesting feature of this pedigree is that LHON predominated in the earlier generations while dystonia predominated in the more recent generations. The phenotype associated with dystonia and striatal lucencies may be considered part of a spectrum of LHON (see 500001).

Funalot et al. (2002) reported 3 unrelated patients with LHON harboring mtDNA mutations at position 3460 of the MTND1 gene and positions 14459 and 14484 of the MTND6 gene. In addition to visual loss, each patient developed a complicated neurologic syndrome resembling Leigh syndrome (256000). Features included gaze palsy, hearing loss, spastic ataxia, cerebellar ataxia, rigidity, hyperreflexia, and multiple hyperintensities in the brainstem.

Gropman et al. (2004) reported a family with a homoplasmic 14459G-A mtDNA mutation of the ND6 gene (516006.0002) and a broad spectrum of clinical manifestations. The proband was a 3-year-old girl with anarthria, dystonia, spasticity, and mild encephalopathy, whose MRI revealed bilateral, symmetric basal ganglia lucencies associated with cerebral and systemic lactic acidosis. Her maternal first cousin presented with a limp and mild hemiparesis along with similar MRI findings with a much milder phenotype. Gropman et al. (2004) investigated additional family members with the mutation and found both asymptomatic and symptomatic individuals with variable clinical and laboratory features, confirming the heterogeneous phenotype of homoplasmic 14459G-A mtDNA mutations, even within the same family.

Jaros et al. (2007) reported a 39-year-old woman with severe complicated LHON who developed progressive gait and sensory disturbances 5 years after onset of subacute bilateral visual failure. Visual symptoms included loss of acuity, central scotomata, optic atrophy, and nystagmus. She also had symmetric pyramidal-pattern lower limb weakness, hyperreflexia, and distal loss of vibratory sensation. Brain MRI showed symmetric high T2 signals in the substantia nigra, pons, and dorsal columns of the spinal cord. After an unexpected death, postmortem examination showed myelin loss and macrophage activation in the posterior columns of the upper spinal cord and neurodegeneration at multiple levels. Molecular analysis detected a homoplasmic 3460G-A mutation in blood and spinal cord. Her mtDNA haplotype H and HLA-DR8 status did not explain the severe phenotype.

La Morgia et al. (2008) reported 6 individuals from 2 unrelated Italian families with LHON confirmed by the findings of mutations in the ND1 and ND4 genes, respectively. All 6 individuals had recurrence of myoclonus as an extraocular feature, and EEG/EMG studies showed myoclonus to be of cortical origin. One family had cosegregation of LHON with psychiatric problems. Four of 6 patients had evidence of increased CSF lactate, which was not found in 6 mutation-matched LHON patients without myoclonus or in controls, suggesting greater bioenergetic impairment in those with myoclonus. Mitochondrial sequence analysis identified other nonsynonymous variants, which La Morgia et al. (2008) postulated may play a role in the LHON-plus phenotype.

Biochemical Features

Larsson et al. (1991) observed that all mutations associated with LHON have been missense mutations in Complex I, III, and IV polypeptides, suggesting that the disease results from a defect in the respiratory chain. Analysis of respiratory Complex I from patients harboring the MTND4*LHON11778A mutation suggests that this defect may occur during the interaction of the NADH generating enzymes with Complex I. Polarographic studies using NADH-linked substrates revealed a 55% reduction in the respiration rate of muscle mitochondria (Larsson et al., 1991) and a 77% reduction in lymphoblast mitochondrial respiration (Majander et al., 1991). Direct assays of NADH:ubiquinone oxidoreductase in skeletal muscle (Larsson et al., 1991) and lymphoblast mitochondria failed to detect a deficiency.

Studies on Complex I in MTND1*LHON3460A patients has revealed a marked deficiency in Complex I activity. Rotenone sensitive NADH:ubiquinone oxidoreductase activity was reduced about 80% in both lymphoblast mitochondria (Majander et al., 1991) and platelet mitochondria (Howell et al., 1991).

A Complex I deficiency was also observed in the Australian pedigree harboring the MTND1*LHON4160C + MTND6*LHON14484C mtDNA haplotype. Analyzing platelet mitochondrial, the NADH: Coenzyme Q oxidoreductase in 4 family members was reduced an average of 79%, while Complexes III and IV showed no significant reduction (Parker et al., 1989).

Danielson et al. (2002) investigated the possibility that the LHON mutation confers a pro-apoptotic stimulus and tested the sensitivity of osteosarcoma-derived cybrid cells carrying the most common and severe mutations (11778 and 3460) to cell death induced by FAS (134637). They observed that LHON cybrids were sensitized to FAS-dependent death. Control cells that carried the same mitochondrial genetic background (the J haplogroup) without the pathogenic 11778 mutation were no more sensitive than other controls, which indicated that increased FAS-dependent death in LHON cybrids was induced by the LHON pathogenic mutations. The type of death was apoptotic by several criteria.

Wong et al. (2002) created cybrids using a neuronal precursor cell line, NT2, containing mitochondria from patient lymphoblasts bearing the LHON mutations 11778 and 3460. The undifferentiated LHON-NT2 mutant cells were not significantly different from the parental cell control in terms of mtDNA/nDNA ratio, mitochondrial membrane potential, reactive oxygen species (ROS) production, or the ability to reduce the reagent Alamar blue. Differentiation of NT2s resulted in a neuronal morphology, a neuron-specific pattern of gene expression, and a 3-fold reduction in mtDNA/nDNA ratio in both mutant and control cells; however, the differentiation protocol yielded 30% less LHON cells than controls, indicating either a decreased proliferative potential or increased cell death of the LHON-NT2 cells. Differentiation of the cells to the neuronal form also resulted in significant increases in ROS production in the LHON-NT2 neurons versus controls, which was abolished by rotenone (a specific inhibitor of complex I). Wong et al. (2002) inferred that the LHON genotype may require a differentiated neuronal environment in order to induce increased mitochondrial ROS, which may be the cause of the reduced NT2 yield. They hypothesized that the LHON degenerative phenotype may be the result of an increase in mitochondrial superoxide which is caused by the LHON mutations, possibly mediated through neuron-specific alterations in complex I structure.

In fibroblasts derived from 16 patients with hereditary optic neuropathy, including either LHON, OPA1 (165500), or OPA3 (165300), Chevrollier et al. (2008) found a common coupling defect of oxidative phosphorylation resulting in reduced efficiency of ATP synthesis. LHON fibroblasts showed a mean decrease of 39% in complex I activity compared to controls. OPA1 and OPA3 fibroblasts showed normal complex I activities, but a mean decrease of 25% in complex IV activity and a mean 60% increase in complex V activity. Resting respiration was about twice as high in all LHON, OPA1, and OPA3 fibroblasts compared to controls, reflecting a proton leak or electron slip. However, all mutant cell lines used a greater proportion of routine respiratory capacity compared to controls, suggesting a compensatory mechanism. The energy defect was most pronounced in fibroblasts from patients with additional neurologic symptoms.

Molecular Genetics

While LHON is traditionally considered to be familial, many individuals represent isolated cases. The proportion of cases with family histories have been reported to be 43% for np 11778, to be 78% for np 3460, to be 57% for np 15257, and to be 65% for np 14484 (Newman et al., 1991; Johns et al., 1993). Families homoplasmic for these common mutations generally exhibit reduced penetrance, with the percentage of affected relatives in np 11778 families ranging from 33 to 60%, for np 3460 from 14 to 40%, for np 14484 from 27 to 80%, and for np 15257 from 27 to 80%. The common mutations also show a strong male bias in Europeans, ranging from 80% for np 11778 to 33-67% for np 3460, 68% for np 14484, and 75-100% for np 15257 (see Table M1, MIM12) (Newman et al., 1991; Johns et al., 1993). Interestingly, in Asia, greater than 90% of LHON patients harbor the np 11778 mutation, yet only 58% of the patients are males (Mashima et al., 1993).

Estimates of recurrence risks differ between sexes and vary among published reports. Studies based on multiple families have suggested recurrence risks for males to be between 50 and 60%, with one study that followed males to age 50 suggesting a risk of 83%. The comparable risk for women ranges from 8 to 32%. However, the prevalence of singleton families confirmed by molecular testing indicates that these values are over-estimated. Using genetic analysis as the starting point, one Australian study proposed that the risk of visual loss for males with the np 11778 mutation is 20% and for females is 4% (Mackey and Buttery, 1992; Newman, 1993).

In familial cases of LHON, all affected individuals are related through the maternal lineage, consistent with the inheritance of human mtDNA (Giles et al., 1980; Case and Wallace, 1981). However, the incomplete penetrance of the clinical phenotype obscures the strict maternal transmission of the mtDNA, and the strong male basis of expression in Europeans (Newman et al., 1991) has frequently led to the erroneous conclusion that the disease results from an X-linked recessive mutation. In fact, most if not all LHON cases are associated with specific mtDNA mutations that occur in isolation or together.

Seventeen mtDNA missense mutations have been proposed to contribute to LHON (see Table M1, MIM12), though to varying degrees. Five of these are generally felt to be 'primary' mutations, the presence of which greatly increases the probability of blindness. Each disease mutation is designated by the gene followed by an asterisk (*), a phenotypic descriptor (LDYT means LHON plus dystonia), the nucleotide number, and the disease-associated base. Listed in order from highest to lowest disease-causing potential, these are MTND5*LDYT14459A (Jun et al., 1994), MTND4*LHON11778A (Wallace et al., 1988), MTND1*LHON3460A (Huoponen et al., 1991; Howell et al., 1991), MTND6*LHON14484C (Johns et al. (1992, 1993); Mackey and Howell, 1992; Howell et al., 1991), and MTCYB*LHON15257A (Brown et al., 1991; Johns and Neufeld, 1991). Three additional mutations may also be primary, but require confirmation; these are MTND5*LHON13730A (Howell et al., 1993); MTCO3*LHON9438A and MTCO3*LHON9804A (Johns and Neufeld, 1993). Nine other mutations have been found at increased frequencies in LHON patients, but generally in conjunction with one of these primary mutations. Hence, these are felt to be 'secondary' mutations which may interact with the primary mutation to increase the probability of clinical expression. Among the more important of these mutations are MTND5*LHON13708A (Brown et al., 1992; Johns and Berman, 1991); MTND1*LHON3394C (Brown et al., 1992); MTCO1*LHON7444A (Brown et al., 1992); MTND1*LHON4160C (Howell et al., 1991); and MTND2*LHON5244A (Brown et al., 1992).

The criteria for ranking the severity of the primary mutations include (a) range of clinical manifestations with mild being LHON alone and the more severe involving LHON plus other neurologic disease; (b) specificity for the disease meaning the proportion of the 'normal' population that harbors the mutation; (c) association with specific mtDNA lineages with the more severe mutations being rapidly eliminated by selection and hence appearing on multiple different haplotypes; (d) co-occurrence with secondary LHON mutations with the more severe mutations able to cause LHON alone while the milder mutations require interaction with secondary mutations to cause disease; (e) heteroplasmy, with the severe mutations appearing repeatedly and hence more likely to be recent and heteroplasmic; (f) amino acid conservation with the more severe variants altering more conserved amino acids; (g) penetrance with the more severe mutations affecting a greater proportion of the maternal relatives; and (f) spontaneous recovery with the milder mutations being more prone to visual recovery (Wallace and Lott, 1993; Newman et al., 1991; Brown et al., 1992; Huoponen et al., 1993; Johns et al., 1993).

The MTND6*LDYT14459A mutation (516006.0002) results in the most severe phenotype (see Table M1, MIM12). It was identified by Jun et al. (1994) in the large Hispanic family reported by Novotny et al. (1986) that showed variable clinical manifestations ranging from normal, through late-onset optic atrophy, to early-onset dystonia accompanied by bilateral basal ganglial degeneration (500001). This G-to-A transition is a new mutation that arose on the Native American haplogroup D mtDNA background. It was not found on any of 38 related mtDNA haplotypes nor in 310 control mtDNAs representing the major ethnic groups. The mutation is heteroplasmic in some maternally related family members and converts the moderately conserved alanine at position 72 in MTND6 to a valine. An alanine is found in this position in all mammals, Xenopus, and sea urchin, whereas a serine is present in all other species that have been examined. When the mutation approaches homoplasmy, the penetrance is high, with 48% of maternal relatives manifesting pediatric dystonia, 10% LHON, and 3% LHON plus dystonia (Novotny et al., 1986; Wallace et al., 1985) (see Table M1, MIM12).

The next most severe mutation and the most common cause of LHON is MTND4*LHON11778A (516003.0001). It accounts for more than 50% of European cases and 95% of Asian cases, but has not been found in controls (Wallace et al., 1988; Newman et al., 1991). Whereas most individuals with this mutation present with LHON (Newman et al., 1991), 1 patient experienced central vision loss at 37 years of age and cerebellar-extrapyramidal tremor and left-side rigidity associated with bilateral basal ganglial lesions at 38 years of age(Larsson et al., 1991). The mutation has arisen repeatedly on different mtDNA lineages (Singh et al., 1989), and is occasionally found with other LHON mutations (Huoponen et al., 1993). It is frequently heteroplasmic (Lott et al., 1990), converts a highly conserved arginine to a histidine, is about 82% penetrant in males and shows only a 4% spontaneous recovery rate (see Table M1, MIM11) (Newman et al., 1991; Wallace and Lott, 1993; Johns et al., 1993).

The MTND1*LHON3460A (516000.0001) mutations account for about 35% of European LHON, and has not been identified in controls (Huoponen et al., 1991; Howell et al., 1991). It has been observed on several mtDNA lineages, occasionally co-occurs with other LHON mutations, is generally homoplasmic, changes a moderately conserved alanine to a threonine, is expressed in 69% of males, and exhibits a 22% spontaneous recovery rate (see Table M1, MIM12) (Howell et al., 1991; Howell et al., 1992; Huoponen et al., 1991; Huoponen et al., 1993; Johns et al., 1993).

The fourth primary mutation is MTND6*LHON14484C (516006.0001). This mutation accounts for about 20% of European LHON patients, has not been observed in 250 controls (Johns et al., 1992), and is commonly associated with specific mtDNA lineages, often in association with MTND5*LHON13708A, MTCYB*LHON15257A, or MTND1*LHON3394C. It has been homoplasmic in every case but one (Mackey and Howell, 1992), changes a weakly conserved methionine to a valine, has a penetrance in males of 82%, and a visual recovery rate of 37% (see Table M1, MIM12) (Johns et al., 1993).

The mildest primary mutation is MTCYB*LHON15257A (516020.0001). This is found in about 15% of LHON patients, and in 0.3% of the general population (Brown et al., 1992). The mutation has been observed on the same mtDNA lineage, usually together with the MTND5*LHON13708A and MTND6*LHON14484C mutations in all but 1 case (Howell et al., 1993). This mutation is consistently homoplasmic, changes a highly conserved aspartate to an asparagine, has a penetrance in males of 72%, and a probability of visual recovery of 28% (see Table M1, MIM12) (Johns et al., 1993).

Five secondary mutations of particular note are MTND5*LHON13708A, MTND1*LHON3394C, MTCO1*LHON7444A, MTND2*LHON5244A and MTND1*LHON4160C. The first 3 mutations are consistently homoplasmic and occur on specific mtDNA lineages prone to LHON. The MTND5*LHON13708A mutation changes a moderately conserved alanine to a threonine, is frequently associated with MTND6*LHON14484C, MTCYB*LHON15257A, and MTND1*LHON3394C mutations, and is found in about 15% of patients and in 4% of controls (Brown et al., 1992; Johns and Berman, 1991). The MTND1*LHON3394C mutation changes a highly conserved tyrosine to a histidine, is commonly associated with MTND6*LHON14484C in French Canadians, and has also been found in 1% of the general population (Brown et al., 1992; Johns et al., 1992). The MTCOI*LHON7444A mutation converts the termination codon of MTCOI to lysine. This extends the polypeptide by 3 charged amino acids, changes the protein electrophoretic mobility, and diminishes the cytochrome c oxidase specific activity 35%. The mutation is found in about 9% of patients, and also in 1% of the general population (see Table M1, MIM12)(Brown et al., 1992).

The MTND1*LHON4160C and MTND2*LHON5244A mutations have been observed in individual families and appear to be relatively recent mutations. MTND1*LHON4160C converts a highly conserved leucine to a proline and is associated with the MTND6*LHON14484C mutation (Howell et al., 1991). This combination is associated with amblyopia in more than 80% of family members and with neurodegenerative disease in many individuals. One branch of the family also harbors the MTND1*LHON4136G mutation, which as has been proposed to ameliorate some of the symptoms (Wallace, 1970; Howell et al., 1991). The MTND2*LHON5244A mutation occurred on a MTND6*LHON14484C + MTCYB*LHON15257A haplotype. It changed a highly conserved glycine to a serine (Brown et al., 1992) and probably was an important contributor to the disease in this patient.

The remaining LHON mutations are of ambiguous significance. MTCYB*LHON15812A mutation converts a moderately conserved valine to a methionine and is consistently found with MTND6*LHON14484C and MTCYB*LHON15257A mutations on a specific mtDNA lineage (Brown et al., 1992). The MTND1*LHON4216C and MTND2*LHON4917G mutations alter poorly and highly conserved amino acids, respectively, and are in somewhat higher frequencies in LHON patients (Johns and Berman, 1991).

Chinnery et al. (2001) described 2 LHON pedigrees that harbored the same novel point mutation within the MTND6 gene at position 14495. The mutation was heteroplasmic in both families, and sequencing of the mitochondrial genome confirmed that the mutation arose on 2 independent occasions. Protein modeling studies indicated that the 7 known mutations in the MTND6 gene that cause LHON lie in close proximity in a hydrophobic cleft or pocket. The authors concluded that this was the first evidence for a relationship between a specific structural domain within a mitochondrial respiratory chain subunit and a specific disease phenotype. They suggested that the MTND6 gene be sequenced in all patients with clinical LHON who do not harbor one of the 3 primary LHON mutations at basepair 11778 (MTND4), 3460 (MTND1), or 14484 (MTND6).

Fauser et al. (2002) sequenced the complete mitochondrial DNA in 14 LHON patients with typical clinical features but without a primary mtDNA mutation. The results suggested that the mutation at np 15257 should be included in a routine screening, as well as the ND6 gene (516006), a hotspot for LHON mutations. Fauser et al. (2002) suggested that screening for the secondary LHON mutations at np 4216 and np 13708 might also help in making the diagnosis of LHON, as these changes seem to modify the expression of LHON mutations.

The male bias and incomplete penetrance of LHON in Europeans has led to the hypothesis that blindness results when two factors coincide, a maternally inherited mtDNA mutation and an X-linked recessive mutation (308905). In a model based on this hypothesis, the penetrance for females was estimated at 0.11 +/- 0.02, and the frequency of the X-linked gene was estimated at 0.08 (Bu and Rotter, 1991). Support for this model was obtained from X-chromosome linkage studies which revealed a linkage between LHON susceptibility and the DXS7 chromosomal marker, with a LOD score of 2.32 (Vilkki et al., 1991). However, this linkage has not been confirmed by other groups (Chen et al., 1989; Chen and Denton, 1991; Carvalho et al., 1992; Sweeney et al., 1992).

Oostra et al. (1994) described the distribution of 7 different mtDNA mutations and the associated clinical findings in 334 LHON patients belonging to 29 families. Mutations described only in LHON at nucleotide positions 11778, 3460, and 14484 were found in 15, 2, and 9 families, respectively. In 3 families, none of these mutations was found. Mutations described in LHON but also in controls at nucleotide positions 15257, 13708, 4917, and 4216 were found in 1, 10, 3, and 12 families, respectively. Combinations of mtDNA mutations were found in most families. In 11 families in which only the 11778 mutation was found, affected males had a mean age of onset of 29.2 years and a mean visual outcome of 0.113. Observations in groups of patients with other mutations indicated that the clinical severity is dependent on the mitochondrial genotype.

Mackey et al. (1996) screened 159 LHON families of northern European origin living in Australia, New Zealand, the United Kingdom (including Ireland), the Netherlands, Denmark, and Finland. These pedigrees comprised more than 12,000 maternally related individuals and more than 1,500 affected individuals. In the 159 families, 153 (97%) carried 1 of the 3 previously identified primary LHON mutations at nucleotides 3460 (13% of the 159 LHON families), 11778 (69%), or 14484 (14%). The primary mutation was not identified in the other families. The 15257 mutation occurred in 6 of the 159 LHON families. However, in every one of these instances, it was associated with 1 of the 3 established LHON mutations: 11778 (4 of 78 families), 3460 (1 of 14 families), and 14484 (1 of 23 families). Because it did not occur in isolation of an established primary LHON mutation, the results did not support a primary pathogenic role for the 15257 mutation.

Liu et al. (2011) investigated the molecular pathogenesis of LHON in 6 Han Chinese families in which 9 (6 males/3 females) of 86 matrilineal relatives exhibited variable severity and age of onset of optic neuropathy. The average age of onset was 20 years. Molecular analysis of mtDNA in these families identified the homoplasmic ND5 12338T-C mutation (516005.0011) and a distinct set of variants belonging to the Asian haplogroup F2. The 12338T-C mutation was present in the maternal lineage of the 6 pedigrees and not in 178 Chinese controls.

Population Genetics

Carelli et al. (2006) evaluated the mtDNA of 87 index cases with LHON sequentially diagnosed in Italy, including an exceedingly large Brazilian family of Italian maternal ancestry. The results revealed that the large majority of the LHON mutations in affected Italian families are due to independent mutational events; only 7 pairs of families and 3 triplets of families showed identical haplotypes. Thus, the study confirmed that the preferential association of the LHON mutations 11778/ND4 (516003.0001) with haplogroups J1 and J2 and 14484/ND6 (516006.0001) with haplotype J1 is attributable not to founder events but to a true mtDNA background effect. In the case of the 11778/ND4 mutation, such a role of the mtDNA background was narrowed to the subclades J1c and J2b, which both, intriguingly, harbor unique combinations of amino acid changes in cytochrome b (516020). Carelli et al. (2006) reinvestigated the genealogies of the families with identical haplotypes and were able to reconnect 3 pairs of families, including the Brazilian family and its Italian counterpart, into extended pedigrees. The survey of the 2 control region sites that were heteroplasmic in the Brazilian family showed triplasmy in most cases, but there was no evidence of the tetraplasmy that would be expected in the case of mtDNA recombination.

In affected members of a 3-generation Chinese family exhibiting high penetrance and expressivity of visual impairment due to LHON, Qu et al. (2006) identified the homoplasmic 11778G-A mutation as well as a novel secondary homoplasmic mutation, 4435A-G, belonging to the Asian haplogroup D5.

In a European multicenter study of 3,613 individuals with LHON from 159 different families, Hudson et al. (2007) found evidence that clinical penetrance of the 3 most common mtDNA mutations is influenced by mtDNA haplotype group. The risk of visual failure was greater when the 11778G-A or 14484T-C mutations were present in haplotype subgroups J2 and J1, respectively, and when the 3460G-A mutation occurred in haplotype K. In contrast, the risk of visual failure was slightly decreased (OR = 0.79) when 11778G-A was present on haplotype H.

By examining data from a population-based study (1970-2004), Puomila et al. (2007) estimated that the prevalence of LHON in Finland is 1 in 50,000, and that 1 in 9,000 Finns is a carrier of 1 of the 3 LHON primary mutations (MTND4, 11778G-A; MTND1, 3460G-A; and MTND6, 14484T-C).

Spruijt et al. (2006) investigated the genotype/phenotype correlation of the 3 major LHON mutations in the Dutch population. They found that the specific mtDNA mutation did not influence disease penetrance (50% in male subjects; 10-20% in female subjects). Regardless of the acuteness of disease onset, more than 50% of patients with the MTND6 14484C mutation exhibited partial recovery of vision, whereas only 22% of the MTND4 11778A carriers and 15.4% of the MTND1 3460A carriers recovered. The recovery did not take place within the first year after onset and was uncommon after 4 years. In general, onset of LHON is very acute but might be more gradual in 11778A carriers and in children. Spruijt et al. (2006) concluded that the LHON genotype influences the recovery of vision and disease onset but is unrelated to age, acuteness of onset, or gender.

By studying the penetrance of LHON in 1,859 individuals from 182 Chinese families (including 1 from Cambodia) with the MTND4 11778G-A mutation (516003.0001), Ji et al. (2008) found that mitochondrial haplogroup M7b1-prime-2 was associated with increased risk of visual loss, whereas the M8a haplogroup was associated with decreased risk of visual loss. Further sequence analysis suggested that the M7b1-prime-2 effect was due to variation in the MTND5 gene, and that the M8a effect was due to variation in the MTATP6 gene.

Inheritance

Mitochondrial DNA point mutations are exclusively maternally inherited (nonmendelian).

Imai and Moriwaki (1936) suggested that LHON might be cytoplasmically transmitted, a hypothesis again enunciated by Ronne (1944, 1945). Wallace (1970) described a large Australian pedigree in which amaurosis and neurologic disease were maternally transmitted, leading him to suggest that the disease was transmitted by a cytoplasmic slow virus (Wallace, 1970). These and related historical studies were summarized by Erickson (1972), who concluded that LHON was maternally transmitted.

Wallace and co-workers demonstrated that human mtDNA was maternally inherited and suggested that maternally transmitted diseases might be due to mtDNA mutations (Giles et al., 1980; Case and Wallace, 1981). This hypothesis was elaborated by Egger and Wilson (1983). The maternal transmission of LHON was further documented by Nikoskelainen et al. (1984, 1987), who emphasized a possible mtDNA etiology. This hypothesis was confirmed by Wallace et al. (1988) who demonstrated that the majority of LHON families harbored the same mtDNA mutation at MTND4*LHON11778A, regardless of mtDNA sequence background (Singh et al., 1989).

Oostra et al. (1997) described 2 LHON pedigrees atypical with respect to sex, age of onset, interval between the eyes becoming affected, course of the disease, concomitant disorders, additional test results, final visual acuity, and/or results of mtDNA analysis. Furthermore, the pedigrees did not suggest maternal inheritance. One pedigree had an affected grandmother and granddaughter through an unaffected carrier; the females in all 3 generations had the 11778 mutation. In the second pedigree, a grandfather and grandson through 1 of his daughters were affected and carried the 11778 mutation; the affected grandmother and the unaffected daughter likewise carried the 11778 mutation, suggesting that the grandson inherited the mutation from his mother.

Incomplete Penetrance

While mtDNA mutations now explain the maternal transmission of LHON, the incomplete penetrance and male bias in Caucasian expression remain enigmatic. One hypothesis is that the expression of LHON is the product of both an mtDNA mutation plus an X-linked recessive allele (Bu and Rotter, 1991). This hypothesis was supported by the report of the cosegregation of the DXS7 X-chromosome marker with LHON (Vilkki et al., 1991, see 308905), though this has not been corroborated by other studies (Chen et al., 1989; Chen and Denton, 1991; Carvalho et al., 1992; Sweeney et al., 1992).

Environmental effects have also been hypothesized to play a role in LHON expression. Heavy tobacco smoking was proposed to be a possible factor by Wilson (1963, 1965), and this idea generalized to cyanide intoxication which for genetic reasons was not adequately detoxified to thiocyanate (Adams et al., 1966; Wilson, 1965). This hypothesis was extended to propose that rhodanase (180370) deficiency might contribute to LHON (Cagianut et al., 1981), but rhodanase deficiency has not been consistently documented in LHON patients (Whitehouse et al., 1989). Recently, heavy tobacco smoking has again been noted in some LHON patients (Cullom et al., 1993).

In a large, multicenter epidemiologic study of 196 affected and 206 unaffected mutation carriers from 125 LHON pedigrees confirmed by genetic analysis, Kirkman et al. (2009) found a strong and consistent association between visual loss and smoking, independent of gender and alcohol intake, leading to a clinical penetrance of 93% in men who smoked. There was a trend towards increased visual failure with alcohol, but only with heavy intake. Based on these findings, Kirkman et al. (2009) concluded that asymptomatic carriers of a LHON mtDNA mutation should be strongly advised not to smoke and to moderate alcohol intake.

A final hypothesis to explain the male bias could be that LHON is hormonally influenced by androgens. In an anecdotal case, a heteroplasmic carrier of the MTND4*LHON11778A mutations experienced precipitous vision loss after androgen therapy (Wallace and Lott, unpublished data; Newman et al., 1991).

Mapping

Allelic variants map to specific nucleotides in the mitochondrial DNA.

Diagnosis

Optic Atrophy; See description of phenotype (above).

Clinical Management

See description of phenotype (above). No proven therapeutic agents have been found.

Gene Therapy

Guy et al. (2002) found that cybrid cells containing the G11778A mutation (516003.0001), found in 50% of LHON cases, showed a 60% reduction in the rate of complex I-dependent ATP synthesis compared to wildtype cells. Using 'allotopic expression,' a technique in which a mitochondrial gene is expressed in the nucleus and the protein product is then imported back to the mitochondria, Guy et al. (2002) transfected a fusion ND4 subunit gene into cybrids containing the G11778A mutation. Cybrid cell survival after 3 days was 3-fold greater for the allotopically transfected cells, and these cells showed a 3-fold increase in the rate of complex I-dependent ATP synthesis, to a level indistinguishable from that in normal cybrids. Guy et al. (2002) suggested that this rescue of a severe oxidative phosphorylation deficiency held promise for development of gene therapy for mitochondrial disorders.

Qi et al. (2007) explored a treatment paradigm for LHON. They augmented mitochondrial antioxidant defenses to rescue cells with the G11778A mutation in mtDNA. An adeno-associated viral (AAV) vector containing the human mitochondrial superoxide dismutase (SOD2; 147460) gene increased LHON cell survival relative to controls. Qi et al. (2007) concluded that protection against mitochondrial oxidative stress might be useful for the treatment of LHON, and that gene therapy with antioxidant genes might protect patients with LHON against visual loss.

History

LHON was named for German ophthalmologist Theodor Leber (1840-1917) (pronounced LAY-ber) and was recognized as a familial neuro-ophthalmologic disease in the late 1800's (Leber, 1871; von Graefe, 1858). Subsequently, many affected pedigrees were reported from Europe, North America, Asia, and Australia (Carroll and Mastaglia, 1979; de Weerdt and Went, 1971; Hiida et al., 1991; Livingstone et al., 1980; Lundsgaard, 1944; Mackey and Buttery, 1992; Morlet, 1921; Muller-Jensen et al., 1978; Nakamura et al., 1992; Newman et al., 1991; Newman et al., 1991; Nikoskelainen et al., 1987; Plauchu et al., 1976; Seedorff, 1968; Seedorff, 1985; van Senus, 1963; Waardenburg, 1924; Wallace, 1970; Went, 1972). A major focus of the earlier studies was the elucidation of the mode of inheritance of LHON. Perplexing features of the disease transmission included the exclusive matrilineal transmission and predilection for males to be affected. Most forms of genetic and epigenetic transmission together with environmental factors have been considered (Adams et al., 1966; Cagianut et al., 1981; Erickson, 1972; Nikoskelainen, 1984; van Senus, 1963; Wallace, 1970; Whitehouse et al., 1989; Wilson (1963, 1965)).

Animal Model

To develop an animal model system for study of oxidative injury to the optic nerve, Qi et al. (2003) designed hammerhead ribozymes to degrade superoxide dismutase-2 (SOD2; 147460) mRNA, thereby decreasing mitochondrial defenses against reactive oxygen species (ROS). Several potential ribozymes were analyzed in vitro. The one with the best kinetic characteristics was cloned into a recombinant adeno-associated virus (rAAV) vector for delivery and testing in cells and animals. The rAAV ribozyme was then injected into the eyes of DBA/1J mice, and the effect on the optic nerve was evaluated by ocular histopathologic examination. The AAV-expressing ribozyme decreased SOD2 mRNA and protein levels by as much as 85%, increased cellular superoxide, reduced mitochondrial membrane potential, and culminated in the death of infected cell lines by apoptosis without significantly altering complex I and III activity, somewhat spared in the most common LHON mutation (G11778A; 516003.0001) although ATP synthesis is markedly reduced. When inoculated into the eyes of mice, the AAV-expressing ribozyme led to loss of axons and myelin in the optic nerve and ganglion cells in the retina, the hallmarks of optic nerves examined at autopsy of patients with LHON. The striking similarity of the mouse model's optic neuropathy to the histopathology of LHON patients is evidence supporting ROS as a key factor in the pathogenesis of LHON.

Qi et al. (2007) described expression of the wildtype ND4 gene in the mouse visual system and produced a mouse model of Leber hereditary optic neuropathy by delivery of a nuclear-encoded version (R340H) of the mutant human ND4 gene. The mutant ND4 disrupted mitochondrial cytoarchitecture, elevated ROS, induced swelling of the optic nerve head, and induced apoptosis, with a progressive demise of ganglion cells in the retina and their axons comprising the optic nerve. In contrast, ocular expression of the wildtype ND4 subunit in mice appeared safe, suggesting that it might be useful for treatment of patients with LHON.

To create an animal model of LHON, Ellouze et al. (2008) introduced the human ND4 gene harboring the 11778G-A mutation (516003.0001), responsible for 60% of LHON cases, into rat eyes by in vivo electroporation. The treatment induced the degeneration of retinal ganglion cells, which were 40% less abundant in treated eyes than in control eyes. This deleterious effect was also confirmed in primary cell culture, in which both RGC survival and neurite outgrowth were compromised. Importantly, RGC loss was clearly associated with a decline in visual performance. A subsequent electroporation with wildtype ND4 prevented both RGC loss and the impairment of visual function. Ellouze et al. (2008) concluded that their data provided the proof of principle that optimized allotopic expression can be an effective treatment for LHON, and that they opened the way to clinical studies of other devastating mitochondrial disorders.