Mitochondrial OXPHOS participates in three major cellular functions relevant to the pathophysiology of mitochondrial disease. First, the mitochondria generate much of the energy of the cell and this process regulates cellular redox potential, mitochondrial membrane potential, ATP production, and Ca++ uptake. Second, the mitochondria generate most of the endogenous reactive oxygen species (ROS) as a toxic byproduct of OXPHOS. Third, the mitochondria integrate many of the signals for initiating apoptosis through regulating the opening of the mitochondrial permeability transition pore (mtPTP). Opening of the mtPTP results in the release of cytochrome c and apoptotic enzymes from the mitochondrial intermembrane space, precipitating programmed cell death. All three of these processes use common OXPHOS polypeptides and functions.
The OXPHOS complexes are composed of multiple polypeptides distributed between the mtDNA and nDNA. Complex I has 43 polypeptides, 7 (ND1, 2, 3, 4, 4L, 5, 6) from the mtDNA. Complex II has four nDNA subunits. Complex III has 11 subunits, 1 (CYTB) from the mtDNA. Complex IV has 13 polypeptides, 3 (COI, COII, COIII) from the mtDNA; and complex V has 16 polypeptides, 2 (ATP6, 8) from the mtDNA. In addition, the mtDNA encodes the 12S and 16S rRNAs and the 22 RNAs necessary for mitochondrial protein synthesis. The nDNA codes for all of the remaining OXPHOS complex subunits as well as the proteins necessary for their expression and assembly. The nDNA also codes for the mitochondrial inner membrane anion carriers, including the adenine nucleotide translocator (ANT), which exchanges mitochondrial ATP for cytosolic ADP; the mitochondrial Mn superoxide dismutase (MnSOD) and glutathione peroxidase (GPx1), which are involved in detoxifying mitochondrial ROS; and for proteins that constitute the mtPTP, including the ANT, the voltage-dependent anion channels (VDAC), the proapoptotic BAX family, the antiapoptotic BCL2 family, and cyclophilin D. The mechanisms for electron transport and proton pumping of each of the OXPHOS complexes are rapidly being elucidated, and this is permitting a detailed physiological understanding of some of the mitochondrial disease mutants. For example, the ATP synthases (complex V) are composed of a fixed component, the stator, which projects from the membrane into the matrix and is composed of the barrel of 3 α and 3 β subunits anchored to the inner membrane ATP6 polypeptide by an arm composed of two “b” subunits and a δ subunit. This stator interacts with a rotor composed of a wheel of about 12 ATP9 subunits attached to an axle of ϵ and γ subunits, which projects up into the 3 α and 3 β barrel. The outer rim of the ATP9 wheel has a negative amino acid Glu,58 located in the middle of the membrane. This interacts with the adjacent ATP6 subunit to form a pair of half proton channels, such that the flow of protons through the ATP6 channels causes the rotor to rotate. The spinning of the asymmetric γ subunit axle inside the 3 α and 3 β barrel causes the β subunit to move through three conformational states: binding ADP + Pi, condensing it to ATP, and releasing the ATP. Mutations distort the interaction between ATP6 and ATP9, inhibiting the rotation of the rotor and limiting ATP synthesis in the mtDNA MTATP6 gene causing retinitis pigmentosa and Leigh syndrome.
Inhibition of OXPHOS lowers mitochondrial energy output and redirects mitochondrial electrons to ROS production. Reduced energy and increased ROS, in turn, impinge on the mtPTP, stimulating it to open and initiate apoptosis. Hence, mitochondrial energy production, ROS generation, and apoptosis are all intimately interrelated in the processes of cell and tissue decline and failure.
The genetics of mitochondrial disease are particularly complex because they involve the interaction of two very different genomes, the mtDNA and nDNA. The mtDNA is maternally inherited, present in thousands of copies per cell, has a high mutation rate, and can exist within the cell as a mixture of mutant and normal mtDNAs (heteroplasmy). Heteroplasmic cells and maternal lineages can segregate the mutant and normal mtDNAs during mitosis and meiosis to give different ratios, creating variable bioenergetic defects and symptoms. The nDNA-encoded mitochondrial genes are dispersed across all of the chromosomes. Their proteins are synthesized on cytosolic ribosomes, and their polypeptides transported into the mitochondrial matrix or inner membrane by an outer (Tom) and either of two inner (Tim) membrane transport systems.
The high mtDNA mutation rate has created a large number of neutral or near-neutral population-specific polymorphisms, which have accumulated as sequential mutations on radiating maternal lineages as women migrated out of Africa and occupied Europe, Asia, and the Americas. It is essential to have a detailed understanding of natural mtDNA variation in order to discriminate pathogenic mutations from nonpathogenic polymorphisms. To this end, the analysis of mtDNA polymorphisms has allowed the reconstruction of the origin and radiation of Homo sapiens. According to the mtDNA, humans originated in Africa about 150,000 years before present (YBP). Currently about two-thirds of all African mtDNAs belong to an ancient African-specific group of mtDNA haplotypes, designated macrohaplogroups L1 + L2. About one-third of African mtDNAs are distinctive and designated L3. A subset of the people with the L3 mtDNAs left Africa and migrated to Asia and later to Europe. European populations developed nine distinctive mtDNA lineages distributed between two macrohaplogroups (+/−, −/−) defined by the presence or absence of a DdeI restriction site at np 10394 and an AluI site at np 10397. The +/− haplogroups are I, J, and K, while the −/− haplogroups are H, T, U, V, W, and X. The age of European mtDNAs is about 40,000 to 50,000 YBP. Asian mtDNA haplogroups also fall into two similar macrohaplogroups: −/− and +/+. The −/− macrohaplogroup encompasses Asian haplogroups like A, B, and F, while the +/+ macrohaplogroup encompasses other haplogroups such as C, D, E, and G. Asia was occupied about 60,000 to 70,000 YBP. From Asia, haplogroups A, B, C, and D migrated either through Siberia or along the Siberian coast to populate the Americas in two to four migrations starting 20,000 to 30,000 YBP. It also appears that a small group of individuals came to America from Europe, carrying haplogroup X and arriving in central North America about 15,000 to 30,000 YBP.
The high mtDNA mutation rate also generates many nonneutral mutations that disrupt essential mitochondrial gene functions, resulting in mitochondrial disease. Mitochondrial genetic disease can also result from nDNA mutations, because both genomes code for genes essential for the assembly and maintenance of the mitochondria as well as OXPHOS.
mtDNA mutations can either be base substitution mutations or rearrangement mutations. Base substitution mutations can, in turn, either be missense mutations affecting polypeptides or protein synthesis mutations affecting the tRNA or rRNA genes. mtDNA missense mutations have been associated with a number of clinical phenotypes, the most common of which are LHON and NARP (neurogenic muscle weakness, ataxia, and retinitis pigmentosa), which can also present as Leigh syndrome.
LHON is a maternally inherited, late-onset, acute, optic atrophy. While most LHON families manifest only optic atrophy, some patients present with additional or alternative clinical symptoms including pediatric-onset dystonia associated with bilateral striatal necrosis. Some LHON family members can also manifest a multiple sclerosis-like optic neuritis. More than 90 percent of European and Asian LHON cases result from three mtDNA missense mutations. These include a G to A mutation in the MTND4 gene at nucleotide 11778 (MTND4*LHON11778A), which converts the highly conserved arginine at codon 340 of the ND4 gene to a histidine. This mutation accounts for about 50 percent of European cases and about 95 percent of Asian LHON patients. In addition, the MTND1*LHON3460A (ND1 Ala52Thr) and MTND6*LHON14484C (ND6 Met64Val) mutations each account for about 15 percent of European cases. A number of rare mutations also appear to cause LHON. These include MTND6*LHON14482G, which changes the same codon as MTND6*LHON14484C; MTND6*14498T; MTND2*LHON5244A; MTND5*LHON13730A; and MTATP6*LHON9101C.
In addition to these LHON mutations, additional mutations can cause both LHON and dystonia (LDYT). The best characterized of these is the MTND6*LDYT14459A mutation, which changes a highly conserved alanine at codon 72 of ND6 to a valine. Rarer mutations associated with LDYT include MTND6*LDYT14596A and MTND4*LDYT11696G.
Analysis of the background haplotypes of patients with LHON and LDYT mutations has shown that the common mutations have arisen repeatedly in the population, consistent with the high rate of mtDNA mutation. In three independent MTND6*LDYT14459A families, the mutation was associated with Native American haplogroup D, African haplogroup L, and European haplogroup I mtDNAs. Similarly, in a study of European LHON patients, with the MTND1*LHON3460A and MTND4*LHON11778A mutations, both mutations were found to be associated with a variety of different European haplogroups, although 37 percent of the MTND4*LHON11778A cases were found on a haplogroup J background. By contrast, when the MTND6*LHON14484C cases were examined, 80 percent were found to be associated with European haplogroup J. This latter association was not due to common descent because haplogroup J LHON patients and controls are randomly interspersed in the haplogroup J lineage. Hence, it appears that the European haplogroup J mtDNA increases the probability that the MTND6*LHON14484C, and to a lesser extent the MTND4*LHON11778A, mutations will cause clinical blindness. This might be due to one or more of the three amino acid substitutions associated with haplogroup J: MTND1*LHON4216G, MTND5*LHON13708A, and MTCYB*LHON15257A.
Biochemical analysis has revealed that the MTND6*LDYT14459A mutation is associated with a 55 percent reduction in complex I-specific activity, but only a minor reduction in the respiration rate on complex I substrates. This biochemical defect has been correlated with an altered binding of the complex I electron acceptor, coenzyme Q10 (CoQ10). A similar biochemical defect is found in the MTND1*LHON3460A mutation, and ND1 has also been proposed to be involved in CoQ10 binding. The MTND4*LHON11778A mutation has no significant reduction in complex I enzyme activity, although it is associated with a 30 percent to 50 percent reduction in complex I respiration rates, as well as altered rotenone binding, which implicates CoQ10 interactions. Finally, the MTND6*LDYT14459A mutation exhibits only a mild inhibition of respiration, but it is located in the same gene as the MTND6*LHON14484C mutation, in a region between codons 26 and 72, encompassing six LHON and LDYT mutations. Hence, it is likely that all of these mutations affect electron transfer from complex I to CoQ10. The mechanism by which inhibition of CoQ10 reduction by complex I could cause optic atrophy is still unclear. The pathophysiology could be the simple result of energy deficiency. Alternatively, the inhibition of the electron transport chain could direct more electrons into ROS production, increasing the tendency for retinal ganglion cell apoptosis or, alternatively, promoting the destruction of nitric oxide (NO) from retinal vessels causing spasm and ischemia.
Retinitis pigmentosa and Leigh syndrome are most frequently associated with three missense mutations: MTND6*NARP8993G, MTND6*NARP8993C, and MTATP6*FBSN9176C. However, these phenotypes can also result from the more severe mtDNA protein synthesis mutations such as MTTK*MERRF8344G and MTTL1*MELAS3243G.
The MTATP6*NARP8993G/C mutations are the best characterized. These mutations are invariably heteroplasmic, with the mutant and normal mtDNAs segregating along the maternal lineage. This random genotypic variation results in striking clinical variation among maternal relatives, with symptoms ranging from unaffected, to mild salt-and-pepper retinitis pigmentosa, to macular degeneration and olivopontocerebellar atrophy, to lethal Leigh syndrome.
The biochemical defect of the MTATP6*NARP8993G mutation has been extensively analyzed and found to be associated with a 24 percent to 53 percent reduction in ADP-stimulated respiration and a 30 percent reduction in ADP/O ratios. Correlation of these data with current information on the mechanism of the ATP synthase (complex V) indicates that the MTATP6*NARP8993G mutation adds an additional positive charge to the ATP6 stator subunit, increasing its attraction to the Glu58 in the ATP9 rotor subunit, inhibiting the rotation of the rotor. Other potentially pathogenic ATP6 mutations may impinge on these same interactions. The inhibition of the ATP synthase would limit ATP production, but also inhibit the electron transport chain. This could stimulate ROS production and increase damage due to oxidative stress.
Leigh syndrome, frequently with retinitis pigmentosa and ophthalmoplegia, has also been shown to result from nDNA mutations. This syndrome, as well as other more severe mitochondrial diseases, is associated with the degeneration of the basal ganglia of the brain. Nuclear gene variants identified in Leigh syndrome patients include mutations in the structural genes of the pyruvate dehydrogenase (E1α gene), various complex I subunit genes, the complex II flavoprotein gene, and the putative complex IV assembly gene, SURF-1. Nuclear DNA mutations affecting mitochondrial functions have also been associated with pediatric encephalomyopathy including basal ganglia degeneration, dystonia and deafness, hereditary spastic paraplegia, and Friedreich ataxia.
Ophthalmoplegia and ptosis, with mitochondrial myopathy, have been associated with mtDNA rearrangements in 83 percent of KSS and 47 percent of CPEO cases, and with autosomal dominant PEO in about 6 percent of cases; mtDNA tRNA mutations occur in at least 4 percent of KSS and 14 percent of CPEO. Mitochondrial myopathy is a distinctive muscle pathology involving ragged-red muscle fibers (RRFs) resulting from the degeneration of muscle fibers accompanied by the proliferation of abnormal muscle mitochondria.
The KSS and CPEO cases resulting from rearrangements are generally spontaneous, isolated cases. The severity of the disease, and hence the age of onset and number of organs involved, appears to depend on the number and distribution of rearranged mtDNAs at conception, and the relentless, preferential propagation of the rearranged mtDNAs. Once the mutant mtDNAs exceed expression thresholds, symptoms appear and progress. High proportions of deleted mtDNAs inhibit cell replication. Consequently, most bone marrow stem cells with high proportions of mutant mtDNAs stop replicating, and only normal bone marrow cells grow. As a result, most CPEO and KSS patients lack rearranged mtDNAs in their blood cells, requiring that definitive diagnostics involving mtDNA rearrangements be done on a postmitotic tissue such as muscle.
If all bone marrow cells contain rearranged mtDNAs, then all the stem cells stop growing, resulting in children with pancytopenia, a condition known as Pearson marrow syndrome. These children have rearranged mtDNAs in their circulating blood cells, and often die unless they have repeated transfusions. In rare cases, the pancytopenia spontaneously reverts, but the children subsequently develop KSS.
A milder mtDNA rearrangement syndrome manifests only as maternally inherited diabetes mellitus and deafness, with occasional individuals experiencing stroke-like episodes. The reason for the mildness of these symptoms is less clear.
mtDNA rearrangements can include deletions, duplications, or both. The deletions most commonly occur in the two-thirds of the mtDNA between the origins of replication, OH and OL. Duplications generally occur in association with deletions and are reciprocal to the deletion, sharing a common breakpoint. This suggests that the duplications might give rise to deletions. The duplicated molecules appear to have a replicative advantage in cultured cells and hence may have a higher predilection for maternal transmission. Indeed, the maternally inherited diabetes and deafness syndrome family harbored a reciprocal 10.4-kb deletion and 6.1-kb duplication.
Mitochondrial myopathy, frequently with ophthalmoplegia and ptosis, can also result from nDNA mutations that destabilize the mtDNA. Mutations in the thymidine phosphorylase gene result in mitochondrial neurogastrointestinal encephalomyopathy (MNGIE). Defects in at least two nDNA loci result in the multiple deletion syndrome, AD-PEO, and nDNA defects have been implicated in the lethal mtDNA depletion syndrome.
Finally, ophthalmoplegia, ptosis, and mitochondrial myopathy can also be the presenting clinical symptoms in patients harboring mtDNA tRNA mutations. The most common of these is the MELAS mutation, MTTL1*MELAS3243C. However, multiple additional tRNA mutations have also been associated with CPEO, including MTTN* CPEO5692G, MTTN*CPEO5703G, MTTI*CPEO4298A, MTTL2*CPEO12308G, MTTL2* CPEO12311G, and so on.
The association between mtDNA tRNA mutations and ophthalmoplegia, ptosis, and mitochondrial myopathy may provide some insights into the pathophysiology of these symptoms. Because essentially all mtDNA rearrangements remove at least one tRNA or rRNA gene, then it seems that ophthalmoplegia, ptosis, and mitochondrial myopathy are most commonly caused by defects in mitochondrial protein synthesis. This has been supported by biochemical analysis of cybrids containing mtDNA deletions and the MTTL*MELAS3243G and MTTK*MERRF8344G tRNA mutations, which reveal a severe defect in mitochondrial protein synthesis.
Mitochondrial diseases generally have a delayed onset and progress with age. This has been correlated with the age-related decline in mitochondrial function in association with the accumulation of somatic mtDNA mutations. It has been hypothesized that the accumulation of these somatic mtDNA mutations is the result of ROS damage, and that the damaged mtDNAs are selectively amplified. Experimental data continue to support this hypothesis, both as it relates to mitochondrial disease and aging.
While identification of the mtDNA mutations associated with neuro-ophthalmologic disease has defined this field, efforts to understand the pathophysiology of these diseases and to develop effective metabolic therapies have been hampered by an inadequate understanding of mitochondrial physiology, the variability in clinical presentations, difficulty in studying the affected tissues, and concerns about experimental drug toxicity to patients. These limitations are being overcome by the application of transgenic mouse technology for the development of mouse models of mitochondrial disease. Three targeted nuclear gene mutations have created animals that have been informative. The first of these was the inactivation of the heart/muscle isoform of the ANT (ANT1). These mice develop classical hypertrophic cardiomyopathy and a mitochondrial myopathy with RRFs and the massive proliferation of giant mitochondria. Because elimination of the ANT would limit ADP/ATP exchanged across the mitochondrial inner membrane, this would limit the availability of ADP to the ATP synthase. Inhibition of the ATP synthase would result in hyperpolarization of the electrochemical gradient (ΔΨ), which would inhibit the electron transfer chain, redirecting a greater percentage of mitochondrial electrons into generating ROS. The increased ROS would damage mitochondrial membranes, proteins, and mtDNA. These predictions have been confirmed by demonstrating that the hydrogen peroxide production rates of mitochondria from ANT1-deficient mice were maximal in heart and muscle. Moreover, the antioxidant enzyme GPx1 was induced threefold in the mitochondria of both tissues and MnSOD was induced fifteenfold in muscle, but not in heart. The high ROS production in heart coupled with the lack of compensating MnSOD induction would be expected to create imbalance between pro-oxidants and antioxidants, resulting in oxidative stress. Consistent with this expectation, the level of heart mtDNA rearrangement mutations was greatly elevated. Hence, inherited OXPHOS defects can also increase oxidative stress and stimulate the accumulation of somatic mtDNA mutants.
The importance of mitochondrial ROS in the pathophysiology of mitochondrial disease was further confirmed in mice lacking the mitochondrial MnSOD. These animals die at about 8 days of age of dilated cardiomyopathy and the massive accumulation of fat in the liver. Biochemical analysis has shown that this results from the inactivation of the iron-sulfur centers in respiratory complexes I, II, and III, as well as in the tricarboxylic acid (TCA) cycle enzyme aconitase. Hence, acute ROS toxicity shuts down the electron transport chain and the TCA cycle, starving the cell of energy and resulting in cardiac failure. This same model has been evoked to explain the pathophysiology of Friedreich ataxia.
This interpretation has been confirmed by treating the MnSOD-deficient mice with the catalytic antioxidant drug, MnTBAP. This drug completely rescued the cardiomyopathy and fatty liver of the mutant mice. However, MnTBAP does not cross the blood-brain barrier, and the animals developed a movement disorder reminiscent of dystonia and Parkinson disease in association with a spongiform encephalopathy. Clearly, the pathophysiology of mitochondrial encephalomyopathy must involve both inhibited energy generation and increased ROS production.
A third knockout mouse involves the inactivation of the gene for the mitochondrial transcription factor Tfam. Animals with a systemic defect in Tfam died in utero of severe mtDNA depletion. In mice where the mutation is confined to heart, the animals die at about 20 days of age of a cardiomyopathy. Hence, defects in mitochondrial biogenesis also can cause mitochondrial myopathy.
While these nuclear mutations are proving useful in understanding the pathophysiology of mitochondrial disease, they lack the stochastic element of pathogenic mtDNA mutations. To create models of mtDNA disease, efforts are being made to introduce mtDNA mutations isolated in cultured cells into mouse female embryonic stem cells by mitochondrial cybrid transfer. These mutant female stem cells are then being used to create mouse maternal lineages harboring the mutant mtDNAs. Using the mtDNA mutation imparting resistance to the mitochondrial ribosome inhibitor chloramphenicol (CAP), chimeric mice have been established in which up to 50 percent of the kidney mtDNAs harbor the mutant mtDNAs. Efforts are now underway to achieve female germ line transmission of the mutant mtDNAs. Because the CAPR mutation causes a partial reduction in complexes I and IV, a heteroplasmic CAPR mouse may be a good model for human heteroplasmic mtDNA protein synthesis mutation diseases.