Mitochondrial Encephalomyopathies

Historical Background

The history of mitochondria can be traced back at least to 1850 when Kolliker described a regular array of subcellular particles in muscle sarcomeres. Fifty-two years were to pass before Luft and colleagues implicated mitochondria as the cause of a human disease. A 39-year-old woman had florid symptoms of hypermetabolism thought initially to be due to thyroid disease. However, antithyroid medication and partial thyroidectomy were ineffective. Studies of biopsied muscle would prove to be more informative. Isolated mitochondria were loosely coupled in vitro by polarographic measurements, and the number of organelles was vastly increased by morphological criteria. Many mitochondria were abnormally large with tightly packed cristae ( Figure 41.1 ). Later, Shy and colleagues would use the terms pleoconial and megaconial when describing similar mitochondrial abnormalities in the biopsied skeletal muscle specimens from some children with limb weakness.

The term mitochondrion was coined by Benda in 1898. The specialized function and the unique structure of mitochondria emerged in the 20th century. Keilin described the cytochrome system and Warburg found evidence of a terminal oxidase system. He also observed that nicotine adenine dinucleotide behaved as a hydrogen acceptor in oxidation-reduction reactions. Flavoproteins were found to act as intermediary carriers of electrons between the reduced form of nicotine adenine dinucleotide and cytochrome systems. Krebs and Kornberg discovered that phosphorylation in aerobic systems depended on the consumption of oxygen, hence the concept of “oxidative phosphorylation.”

Krebs and Kornberg continued their seminal observations in the 1930s, showing that citrate was formed from pyruvate and oxaloacetate. These observations led to the discovery of the tricarboxylic acid cycle (the TCA cycle, now better known as Krebs cycle), which converted pyruvate to carbon dioxide and water. These investigators and others showed that oxidation of intermediates of the tricarboxylic acid cycle was coupled to the generation of adenosine triphosphate.

The importance of the mitochondria as the source of cellular energy was being established by these seminal observations. Additional experiments explained the relationships between fuel combustion to carbon dioxide and water, the generation of reducing equivalents in the Krebs cycle, and the re-oxidation of these equivalents, NADH and FADH ₂ , in the respiratory chain coupled with the synthesis of ATP. The concept of oxidation-phosphorylation was established and the chemiosmotic hypothesis of Mitchell in 1961 proposed the proton-motive force as the primary source of energy for ATP synthesis in complex V. The rudimentary structure of the respiratory chain now was complete with the five enzyme complexes. More critical observations would follow over the next 50 years, characterizing in greater detail the subunit structure of the respiratory chain complexes, the critical shuttling roles of coenzyme Q10 and cytochrome c, and the dual roles of nuclear DNA and mitochondrial DNA in the synthesis of the many subunits that composed the five enzyme complexes.

In parallel with these brilliant biochemical observations, a series of studies was being published describing the ultrastructural characteristics of the mitochondrion. Four components of the organelle were characterized: the outer and inner membranes, the intermembranous space, and the inner mitochondrial compartment matrix. A series of studies subsequently demonstrated that the inner mitochondrial membrane was largely impermeable to molecules of all sizes, and special adaptations were necessary for the translocation of metabolites and proteins from the intermembranous space to the mitochondrial matrix. These observations set the stage for an understanding of the complex importation processes that are necessary both to translocate solutes across the inner mitochondrial membrane and to move proteins from the cytoplasmic environment to the intramitochondrial environment. Protein translocation is elaborate and energy-dependent, requiring the proteins to be unfolded before traversing the mitochondrial outer and inner membranes and then refolded after entering the mitochondrial matrix. The leader sequence that guides the protein through the mitochondrial membrane to the matrix is cleaved by a mitochondrial-specific protease. Only 13 proteins are synthesized within the mitochondrial matrix by protein-coding genes located in the mitochondrial genome.

In 1963, two important observations were made simultaneously and independently. The first observation described the presence of intramitochondrial fibers with DNA characteristics. Twenty-five years would elapse before this observation would take on clinical significance. The second observation was the description of abnormal mitochondria under the light microscope: Engel and Cunningham exploited a modification of the Gomori trichrome stain and found conglomerations of abnormal organelles in the cross section of muscle fibers that they called “ragged-red fibers” (RRF) ( Figure 41.1 ). The RRF almost invariably contained structurally abnormal mitochondria as seen under the electron microscope. These morphological criteria later were displaced by biochemical and genetic criteria for the classification of mitochondrial diseases.

Early morphological observations gradually gave rise to another fundamental biological principle related to mitochondria. Mitochondrial ribosomes resemble bacterial ribosomes under the electron microscope. This observation led Ephrussi and colleagues in 1949 to suggest that mitochondrial morphogenesis was controlled by extrachromosomal or cytoplasmic genetic factors. The observation of a yeast mutation causing a deficiency of respiratory function and inherited in a non-Mendelian fashion was central to the development of the unique theory put forward by Ephrussi and colleagues. Schatz et al. in 1964 found DNA in yeast mitochondria and this observation led him to propose that the yeast organelle might be self-replicating and relatively independent of the nuclear genome. A series of observations followed and culminated in the report by Fine in 1978 that all mitochondrial DNA was derived from the unfertilized ovum. This hypothesis was validated two years later by Giles and colleagues when they were able to trace the inheritance of a mitochondrial DNA polymorphism in a 33-member three-generation family. The typical mitochondrial DNA cleavage pattern and the atypical mitochondrial DNA cleavage pattern could be traced from one generation to the next, clearly documenting the non-Mendelian pattern of maternal inheritance. Later, it would be shown that clinical conditions related to mitochondrial DNA (mtDNA) mutations were transmitted from the mother to all of her male and female progeny, but only the daughters would be able to pass the condition to succeeding generations. This genetic profile was similar to Mendelian inheritance including autosomal dominant and X-linked patterns, but differed from Mendelian inheritance because both genders were equally affected and there was no evidence of father-to-child transmission. Wallace and colleagues later would add two other important concepts to explain the variable expression of maternally inherited genetic defects: namely, replicative segregation and threshold effect.

The historical stage now was set for the two dramatic publications that appeared in 1988 and that introduced the medical community to the presence of mitochondrial DNA mutations as a new cause of human diseases. Holt et al. reported large-scale deletions of mitochondrial DNA in patients with a mitochondrial myopathy and Wallace et al. described a point mutation in the mitochondrial protein-coding gene for subunit 4 of complex I ( ND4 ) in a family with Leber’s hereditary optic neuropathy (LHON). More than 200 additional mitochondrial DNA point mutations would be described over the next 25 years in addition to numerous mitochondrial DNA large-scale rearrangements. The number of mitochondrial DNA mutations continues to increase, and now exceeds 250 at the time of this chapter revision.

These mutations affect all aspects of the mitochondrial genome, which had been fully characterized and sequenced in 1981 by Anderson and colleagues. The double-stranded molecule contains 16,569 base pairs. Thirty-seven genes are represented including two ribosomal RNAs, 22 transfer RNAs, and 13 protein-coding genes. This molecular machinery allows for the intramitochondrial synthesis of 13 subunits of the respiratory chain. Seven of these subunits are located in complex I, one in complex III, three in complex IV, and two in complex V ( Figure 41.2 ). The simplicity of the mitochondrial genome is shown in Figure 41.2 and is characterized by the circularity, the absence of introns, and the presence of one asymmetric replication origin.

Epidemiology

Epidemiological studies have confirmed the assumption of clinicians that this group of diseases is underestimated. In fact, epidemiological studies limited to mtDNA-related disorders in adults have shown a minimum prevalence of approximately 1 in 5000, making them among the most common genetic disorders and a major burden for society. More interesting to pediatricians, screening of umbilical cord blood for the 10 most common pathogenic mtDNA point mutations revealed that around 1 in 200 infants harbors a mutation. As the prevalence of mtDNA-related disorders is far less than 1 in 200, it must be concluded that the mutation load in most carriers remains below the pathogenic threshold (see below). This concept is also in keeping with the results of a supersensitive next-generation sequencing study showing that low-level changes (0.2–2.0% heteroplasmy [the coexistence of mutated and wild-type mtDNAs within a cell—see the following]) are present in blood and skeletal muscle of clinically unaffected individuals, a phenomenon dubbed “universal heteroplasmy.” The possibility exists, therefore, that clinically silent mtDNA mutations could expand over time and manifest as de novo mitochondrial diseases.

Beginning in 1995, mutations in nuclear genes affecting the respiratory chain were identified at an escalating pace, largely thanks to linkage analysis, homozygosity mapping, monochromosomal transfer techniques, gene complementation studies, and, more recently, whole exome or mito-exome sequencing. As Mendelian mitochondrial disorders are more common in infants and children than in adults, as exemplified by Leigh syndrome (LS), next-generation sequencing is particularly important for the molecular diagnosis of pediatric disorders.

Classification

Our earlier classification of mitochondrial diseases was based on the biochemical site of dysfunction, for example substrate transport, substrate utilization, Krebs cycle, respiratory chain, and protein importation. However, this scheme proved inadequate and clumsy when assessing patients with various clinical manifestations. The discovery in 1988 of disease-causing mtDNA mutations and, starting in 1995, the description of myriad diverse Mendelian mitochondrial disorders, forced a revision of the classification, which is today rationally and comprehensively based on molecular criteria.

Keeping in mind that a generally accepted definition of mitochondrial diseases restricts these diseases to defects of the respiratory chain and oxidative phosphorylation (OXPHOS), they fall broadly into two categories: those due to mutations in mtDNA, which cause maternally inherited or sporadic disorders; and those due to mutations in nDNA, which show Mendelian inheritance.

A more practical way of classifying these disorders is to divide them into three groups: mtDNA defects; defects of nDNA that affect the respiratory chain directly or indirectly; and defects of mtDNA maintenance (also known as defects of intergenomic communication) ( Table 41.1 ). We separate out the defects of mtDNA maintenance for two reasons: (i) they involve the mitochondrial genome directly, causing multiple mtDNA deletions or mtDNA depletion; and (ii) although unequivocally Mendelian with regard to inheritance, these disorders share much of the clinical heterogeneity of primary mtDNA-related diseases because the polyploid mtDNA is involved in both conditions.

Pathogenesis and Genetics

There are two main disturbances of cell function associated with mitochondrial diseases: failure to generate sufficient ATP to carry out cellular work; and failure to neutralize reactive oxygen species (ROS) with resulting damage to structural elements of the cell. All categories of mitochondrial diseases, as shown in Table 41.1 , impair ATP production. Clinically, these patients exhibit classical symptoms of energy failure such as weakness, hypotonia, reduced physical stamina, muscle atrophy, heart failure, and limitation of eye movements with ptosis. Developmental processes also may suffer as shown by failure to thrive with short stature, reduced muscle bulk, and deceleration of head growth.

Accumulation of ROS damages tissue membranes, promotes genetic mutations, and contributes to cell death. Symptoms associated with this insidious tissue damage are more difficult to appreciate in a correlative sense but reflect the overall degenerative process. The failing heart, muscular atrophy, and basal ganglia damage likely reflect this microchemical attack on tissue structures, suggesting that the pathogenesis of human mitochondrial diseases directly affecting oxidation-phosphorylation mechanisms is intensified by the combination of energy failure and the accumulation of ROS.

Before proceeding with the description of mitochondrial diseases, it may be useful to remind the readers of the bona fide mitochondrial diseases that do not belong to this category because they are not due to defects of the respiratory chain and OXPHOS. They belong in three major categories: defects of fatty acid oxidation (FAO); defects of pyruvate metabolism; and defects of the Krebs cycle (except for succinate dehydrogenase [SDH, complex II], which belongs both to the Krebs cycle and to the respiratory chain) ( Figure 41.3 ). FAO defects are discussed in Chapter 40 .

The molecular genetics of mitochondrial diseases is complex because it is determined by two genomes, nDNA and mtDNA. The respiratory chain is composed of five complexes containing about 85 subunits, all of which are encoded by nDNA except 13 that are encoded by mtDNA ( Figure 41.2 ). As a result, a defect of the respiratory chain may be genetically determined as an autosomal trait or as a maternal trait. The autosomal conditions may be dominant, recessive, or X-linked traits. Maternally transmitted traits are non-Mendelian and obey the rules of mitochondrial genetics. There are three biological rules that explain mitochondrial genetics. The first is maternal inheritance . Sperm mtDNA is actively degraded after fertilization; as a result, all mtDNA is maternally inherited. Mitochondrial DNA mutations, therefore, are transmitted from mother to her male and female progeny, but only her daughters will transmit the mutation to their children.

The second rule is heteroplasmy , with the related concept of threshold effect . Each mitochondrion contains two to ten mitochondrial DNA copies, and each cell contains hundreds or thousands of mitochondria. Therefore, any given cell may contain as many as several thousand mitochondrial DNA copies. If an mtDNA mutation exists, some of the molecules will harbor the mutation whereas other molecules will be normal. This mixture of mutated and wild-type mtDNAs is referred to as heteroplasmy. Despite recent evidence of “universal heteroplasmy” (see above), for all practical purposes all mtDNAs in normal cells are wild-type, a condition called homoplasmy. An increasing percentage of mutated mtDNAs threatens the biological integrity of the cell until the mutation load reaches a theoretical threshold and the cellular phenotype is transformed from normal to diseased. Pathogenic mtDNA mutations usually are heteroplasmic whereas neutral polymorphisms (nondeleterious mtDNA mutations) are homoplasmic, but there are many exceptions to this rule, notably the three mutations commonly associated with LHON. As the threshold varies inversely with the oxidative demands of the tissue, oxidative-rich tissues are more vulnerable to the presence of pathogenic mtDNA mutations.

The third rule is mitotic segregation , which stipulates that the proportion of mutant mtDNAs can shift in daughter cells at division. This stochastic shift in mutated mtDNAs during mitosis can result in segregation of the mutation and is invoked as an explanation for the phenotypic heterogeneity among patients with the same mtDNA point mutation, including siblings. For example, the m.3243A>G mutation, commonly referred to as the MELAS mutation, may produce several different phenotypes, one reflecting the classical MELAS presentation (mitochondrial encephalomyopathy, lactic acidosis, and stroke-like episodes), others reflecting different clinical pictures such as progressive external ophthalmoplegia (PEO), myopathy, cardiomyopathy, and diabetes mellitus and deafness (see Case Example 41.1 ). Presumably, these phenotypic differences are determined by the segregation of the A3243G point mutation during mitosis, although the influence of modifying mtDNA or nDNA genes cannot be excluded. Other examples of phenotypic heterogeneity and genotypic homogeneity are shown in the morbidity map of the human mitochondrial genome ( Figure 41.4 ). This figure also demonstrates examples of phenotypic homogeneity and genotypic heterogeneity, including the diverse point mutations resulting in Leigh syndrome (LS) or in MELAS.

Clinical Features

There are over 250 pathogenic mtDNA point mutations, most of which affect mitochondrial protein synthesis while a smaller number affect protein-coding genes. Pure myopathies and neuropathies are unusual, but neuromuscular symptoms are common as part of multisystemic phenotypes. Multisystemic disorders are illustrated by Kearns-Sayre syndrome (KSS), Pearson syndrome, MELAS, myoclonus epilepsy and RRF (MERRF), and neuropathy, ataxia, and retinitis pigmentosa/maternally inherited LS (NARP/MILS) ( Table 41.2 ). Two of these diseases (KSS and Pearson syndrome) are due to large-scale deletions; two (MELAS and MERRF) are caused by tRNA mutations that impair mitochondrial protein synthesis globally; and the remaining two diseases (NARP and MILS) are due to mutations in a single protein-coding gene.

Mitochondrial genetics provides some explanation for the clinical heterogeneity of these diseases. For example, the degree of heteroplasmy of the same mutation in the gene encoding ATPase 6 (m.8993T>G) is clearly related to the severity and age at onset of NARP, a disorder of young adults with ~70% mutation load, and MILS, a severe neurodegenerative disease of infancy or young children with mutation loads >90%.

A conundrum that lies at the heart of mitochondrial genetics regards the pathogenesis of typical heteroplasmic diseases. Simply put, why is KSS different from MELAS and MELAS different from MERRF, when all three conditions result from impaired mitochondrial protein synthesis and ATP production? It would be logical to expect phenotypic homogeneity rather than syndromic individuality and yet—despite occasional overlapping presentations—this is not the case. There is immunohistochemical evidence that the different mutations that characterize the three syndromes have distinct spatial distribution in the brain: the single deletions of KSS are more abundant in the choroid plexus; the most common m.8344A>G MERRF mutation predominates in dentate and olivary nuclei; and the typical m.3443A>G mutation abounds in subpial arterioles. Although these preferential brain localizations do correlate with typical symptomatology (e.g. MELAS is ultimately a mitochondrial angiopathy), the question then becomes what directs the different mutant mtDNAs to selective regions of the brain.

Large-Scale Rearrangements of mtDNA

These genetic abnormalities include single large-scale deletions and duplications of mtDNA, i.e. each patient harbors a single species of partially deleted mtDNA. The rearrangements generally are several thousand base pairs in length ( Figure 41.5 ).

Kearns-Sayre Syndrome

KSS is defined by onset before age 20 years, progressive ophthalmoplegia, and pigmentary retinopathy. In addition to this triad of symptoms, at least one additional symptom must be present: cardiac conduction block, cerebellar syndrome, or cerebrospinal fluid (CSF) protein concentration >100 mg/dL. Patients with KSS exhibit many of the classical signs and symptoms of mitochondrial DNA-related disorders such as short stature, dementia, sensorineural hearing loss, isolated endocrinopathies including diabetes mellitus, hypothyroidism, hypoparathyroidism, and growth hormone deficiency. Surprisingly, seizures are very infrequent and commonly are associated with hypoparathyroidism. Brain MRI shows diffuse white matter abnormalities ( Figure 41.6D , KSS). Muscle histology is abnormal, with ragged-red fibers and a mosaic pattern of cytochrome c oxidase (COX)-negative fibers. Mitochondrial DNA deletions are easily demonstrated in muscle, and are less likely to be found in blood DNA extracts.

KSS almost always is sporadic and a multicenter retrospective study of 226 families has shown that the risk of a carrier woman to have affected children is very small but finite (1 in 24 births) while disproving the idea that the risk increases with maternal age. There are rare examples of women with the symptoms of KSS who have had children with Pearson’s marrow/pancreas syndrome and the same mitochondrial DNA deletion as the mother.

Point Mutations in mtDNA

Over 250 pathogenic point mutations pepper the small mtDNA molecule ( Figure 41.4 ). We will consider separately those that impair mitochondrial protein synthesis globally (mutations in rRNA and in tRNA genes) and those that affect protein-coding genes.

Point Mutations in Protein-Coding Genes