Mitochondria were free-living independent organisms with their own unique DNA and membranes about 2 billion years ago, when they developed a symbiotic relationship with the evolving proeukaryocytes, and the two have lived more or less harmoniously as eukaryotic cells ever since. The cell depends on its mitochondria for the generation of oxidative enzymes of aerobic energy production, and the mitochondria depend upon the cell’s nuclear DNA to provide the majority of the subunits for its respiratory chain complexes.

The first probable human visualization of mitochondria was in 1841, when intracellular granules were described using the crude optics of early microscopes. In 1890 Altmann (2) recognized the ubiquitous nature of “bioblasts” as elementary organs living in cells, and in 1898 Benda named the organelle the “mitochondrion” (3). The ultrastructure of mitochondria was described in 1952 by Palade (4), and mitochondrial DNA (mtDNA) was defined by Nass and Nass in 1963 (5). Luft was the first to describe a mitochondrial disease in 1962, deciphering a deficit in oxidative phosphorylation in a patient with a now rare disease (6). Ragged-red fibers were described in the muscle biopsy and recognized as associated with mitochondrial disorders in 1963 by Engel and Cunningham (7). In the early 1970s, several disorders of brain and muscle involving deficiencies in respiratory chain enzymes were described, and in 1973, DiMauro and DiMauro provided the first examples of myopathies due to isolated deficiencies of muscle carnitine and carnitine palmitoyltransferase (8). By 1981, the complete genomic sequence of human mtDNA was documented (9). In 1984 phenotype–genotype correlations were being made by Pavlakis et al. (10), and the concept of mitochondrial syndromes was founded. In 1988, major advances occurred with the description by Holt et al. (11) of human encephalomyopathies associated with large deletions of mtDNA and the description by Wallace et al. (12) of Leber hereditary optic neuritis as due to a point mutation at nucleoprotein (np)11778 in mtDNA, a clinical disorder subsequently shown to be a syndrome associated with several additional mtDNA point mutations in other patients, as is the case with most mitochondrial disorders. Hundreds of point mutations or deletions of mtDNA have been described since 1990, and new genetic varieties are continuously being discovered. Mitochondrial diseases also can result from mutations in the nDNA that encodes respiratory chain subunits or parts of mitochondrial membranes,
even with a normal mtDNA genome. In 2003, Servidei showed that large-scale mtDNA rearrangements, rather than mtDNA deletions or point mutations, are the basis of some mitochondrial diseases (13). Nearly all of the distinctive clinical and pathologic patterns of mitochondrial diseases were initially thought to be singular diseases but now are recognized as syndromes in which the same phenotype can be produced by multiple mitochondrial genotypes.

The biochemical aspects of mitochondrial structure and function are reviewed more extensively by Lodish et al. (14) and Schapira (15). The primary function of mitochondria is to generate the high-energy phosphate bond in ATP by phosphorylation of adenosine diphosphate (ADP). The energy required for this reaction is derived from oxidation of the metabolic products of carbohydrates, fatty acids, and proteins. These processes of energy generation are collectively referred to as oxidative phosphorylation. Energy production is the only function of mitochondrial DNA; unlike the nuclear DNA of the cell, mtDNA does not determine any features of anatomic structure of the body, growth, or tissue differentiation.

Mitochondrial oxidative phosphorylation involves electron transport among five respiratory chains of enzymes, termed respiratory complexes. Complex I is nicotinamide adenine dinucleotide diaphorase; complex II is succinate dehydrogenase; complex III is ubiquinone or cytochrome b oxidase; complex IV is cytochrome c oxidase; complex V is ATP synthase. Coenzyme Q10 helps to mediate transport from complex I to complexes III and IV.

Pyruvate dehydrogenase is one of the most complex enzymes known, with a molecular weight of greater than 7 million. It contains multiple subunits of three catalytic enzymes (E₁, E₂, and E₃), two regulatory polypeptides, and five different coenzymes. The pyruvate dehydrogenase complex mediates the oxidation of pyruvate to acetyl coenzyme A to carbon dioxide, to generate reduced electron carriers through a series of nine reactions, carried out within the mitochondrial matrix by the citric acid cycle of Krebs. These reactions are carried out on the inner mitochondrial membrane by the electron transport chain, a set of electron carriers grouped into five multienzyme complexes. The free energy released during oxidation is stored in an electrochemical proton gradient across the inner mitochondrial membrane. The movement of protons back across the mitochondrial membrane is then coupled with the synthesis of ATP from ADP and phosphate. Coenzyme Q10 subserves the transport of electrons from complex I to complexes III and IV.

Transport of free fatty acids into mitochondria as fuel is mediated by carnitine. Within the mitochondria, the fatty acids are uncoupled from carnitine, a carrier molecule, and react with coenzyme A (CoA) to form an acyl-CoA. By a series of four sequential reactions (beta-oxidation), each molecule of acyl-CoA is oxidized to form one molecule of acetyl-CoA and an acyl-CoA shortened by two carbon atoms. Because most defects of beta-oxidation present with organicaciduria, these conditions are considered in Chapter 1 in the section dealing with organic acidurias. Amino acids are oxidized within the mitochondria; after transamination to their respective ketoacids, the amino acids alanine, aspartic acid, and glutamic acid enter the citric acid cycle.

In addition to its principal function in energy metabolism, mtDNA has additional functions during development. It is important for neuroblast polarity by modulating calcium homeostasis in microtubules and also regulates the bcl2 gene for apoptosis. Damage to mtDNA and loss of the mitochondrial membrane potential are demonstrated in apoptotic cell death (16).

The genetics of mitochondrial cytopathies is complex. The mitochondrion has its own DNA (mtDNA) on a single, circular chromosome of about 16.5 kilobases (kb). The mtDNA has an intimate relation with the cell’s nuclear DNA (nDNA), and for each of the five respiratory complexes, the majority of the subunits are encoded by nDNA, not mtDNA. Complex I consists of 41 subunits, of which only 7 are encoded in the mtDNA; the other 34 are encoded in the nDNA. Complex II has only 4 subunits, all of which are encoded in nDNA. Complex III has 10 subunits, 1 encoded by mtDNA and 9 by nDNA. Complex IV has 13 subunits, 3 encoded by mtDNA and 10 by nDNA. Complex V has 12 subunits, 2 encoded by mtDNA and 10 by nDNA. The nDNA encodes the vast majority of the mitochondrial proteins, including all proteins present in the outer membrane and the matrix. The respiratory complexes are located on the inner membrane.

Because in the formation of the zygote, almost all mitochondria are contributed by the ovum, mtDNA is maternally inherited (17). Maternal inheritance appears to be operative in a small but significant proportion of inherited mitochondrial diseases; in the remainder transmission is as an autosomal recessive trait or is sporadic or unpredictable (18). Because the nuclear genome contributes subunits to the respiratory complexes, some mitochondrial diseases may follow a Mendelian pattern of inheritance, in which the metabolic defect is due to defective subunits encoded by mutation in nDNA, with preservation of normal mtDNA. Inheritance in these mutations is nearly always autosomal recessive. Some cases of Leigh encephalopathy provide a good example of this phenomenon. Nine
different genetic defects have been documented in this syndrome, five of which involve mtDNA and four of which involve nDNA. If a point mutation in mtDNA is involved, the obligatory transmission is maternal, though not involving the X chromosome of nDNA.

Homoplasmy refers to the situation in which all mtDNA molecules within the cell are identical, the normal condition. If mutations in mtDNA occur, two populations of mtDNA then coexist, the heteroplasmic condition. Most pathogenic mtDNA mutations are heteroplasmic. In this situation the ratio of normal to abnormal mtDNA determines the clinical expression as well as the severity of the mitochondrial disease. A 100% abnormal mtDNA is incompatible with life. Differences in this ratio of heteroplasmy among cells in different organs determine the clinical expression in each organ system. Heteroplasmy may occur not only at the level of the cell, but also at the level of the individual mitochondrion, intramitochondrial heteroplasmy.

At the time of cell division, mitochondria and mtDNA are randomly partitioned into the two daughter cells, mitotic segregation. The proportion of normal and abnormal mtDNA inherited by each daughter cell is not necessarily equal, and a threshold effect of mtDNA mutation may influence clinical expression including age of onset of symptoms (19).

As an example of heteroplasmy, in the nucleotide (nt)-8993 T→C mutation that substitutes leucine for arginine, if less than 70% of the mtDNA shows this point mutation, the patient has the NARP (nystagmus, ataxia, and retinitis pigmentosa) syndrome; if the mutation involves 90% or more, the patient presents with Leigh encephalopathy. Between 70% and 90% mutated mtDNA produces mixed or variable clinical expression.

Another class of mtDNA abnormalities is large-scale mtDNA rearrangements, such as kilobase deletions of the mitochondrial chromosome. Patients may harbor a 5- or 7.5-kb mtDNA deletion (the most common varieties), with striking differences in the proportion of deleted mtDNA molecules within the total mtDNA population. Sometimes as little as 2% mitochondria with significant deletions may be enough to render the individual symptomatic. High levels of these large-scale mtDNA deletions frequently occur in patients with Kearns-Sayre and PEO (progressive external ophthalmoplegia) syndromes (20). Approximately one-third of Kearns-Sayre patients harbor the same kind of mtDNA deletion, often designated the common 5-kb deletion, but occasional patients with Kearns-Sayre syndrome have larger, 11-kb deletions, with correspondingly more severe neurologic, muscular, and cardiac manifestations (21,22).

Many of the common point mutations are now known, particularly in mitochondrial myopathies with ragged-red fibers, in mitochondrial cardiomyopathies, and in some mitochondrial encephalopathies, especially Leigh encephalopathy, but new mutations are published weekly in this rapidly expanding field of genetics. A table summarizing the known mtDNA point mutations and deletions in mitochondrial encephalomyopathies is updated quarterly in the journal Neuromuscular Disorders (23). Many mitochondrial point mutations are expressed in both striated muscle and central nervous system (CNS). For example, Kearns-Sayre syndrome is clinically mainly a myopathy but also involves the visual system of the brain, demonstrated during life by abnormally slow visual-evoked potentials.

Most genetic laboratories that study mtDNA employ batteries of several point mutations that they screen for the common, well-documented mutations, but cannot test the entire mitochondrial genome. At times, point mutations are found in nucleotide sequences that are not evolutionarily conserved, do not specify highly conserved amino acid residues, and/or are not associated with an amino acid substitution. The interpretation of such defects or polymorphic variants, in the context of clinical and pathologic presentation, often is problematic and uncertain.

Because mitochondria are present in almost every cell, the wide variety of clinical presentations is not surprising. Mitochondrial disorders can be initially classified by clinical criteria in accordance with the organ systems most affected. One principle of mitochondrial diseases in general is that rarely is only one organ affected, and the clinical involvement of multiple organ systems is an important feature that allows the clinician to suspect a mitochondrial disease. The combination of symptoms and signs referable to the central and/or peripheral nervous system and striated muscle is the most frequent. Clinical manifestations are extremely variable, even with a known pathogenic point mutation and even among affected members of the same family. Phenotype–genotype correlations often are poor for the clinical identification of specific mtDNA point mutations or at times for specific mitochondrial syndromes (24,25,26). Multiple genotypes of both mtDNA and nDNA defects may produce similar clinical and pathologic presentations, hence the common patterns of mitochondrial disorders and syndromes.

Each of the mitochondrial syndromes has some constant clinical features that raise suspicion of the diagnosis and justify investigations to confirm or refute this provisional diagnosis. For instance, some patients fit a classical clinical and pathologic phenotype for MELAS (mitochondrial encephalopathy with lactic acidosis and strokelike episodes) but have an mtDNA point mutation that is more
typical for MERRF (myoclonic epilepsy with ragged-red fibers), or they show a mixed MELAS/MERRF phenotype (27,28) or MERRF/Kearns-Sayre phenotype (29). Many mitochondrial cytopathies cannot be classified under a single clinical eponym.

Because each of the mitochondrial syndromes may be a similar clinical expression of multiple genetic defects, and the ratio of normal to abnormal mitochondria is variable between different patients even in the same family and between different organs, the phenotype–genotype correlations often are poor. Myopathology–genotype correlations are somewhat more reliable, but even the muscle biopsy does not always predict the precise genetic mutation.

The neurologic manifestations of mitochondrial diseases are many, but the most common in infancy are seizures, including infantile spasms and myoclonic epilepsy, visual impairment, deafness, central dysphagia, apnea and respiratory failure, diffuse muscular hypotonia, and global developmental delay. In older children and young adults, neurologic deficits include dystonia, cerebellar deficits, spastic diplegia, memory loss, cognitive deficits, progressive hearing loss, dementia, and autism. Regression and loss of skills may be seen as with any neurodegenerative disease. Peripheral neuropathy also occurs in mitochondrial disease and may contribute to the hypotonia and weakness. Because extraocular muscles have more mitochondria per unit volume of tissue than do most other muscles, they often are involved and produce progressive external ophthalmoplegia of all six muscles. Respiratory insufficiency may result from myopathic involvement or from central respiratory failure or a combination.

Lactic acidosis is a prominent feature of many mitochondrial diseases but not all. It is universally present in MELAS syndrome and helps to define this entity but is not present in MERRF and Kearns-Sayre syndromes. It is present in most cases of Leigh encephalopathy but not all; lactic acidosis is more likely in those cases with mtDNA mutations than in cases with nDNA defects. In some patients, the serum lactate may be normal but cerebrospinal fluid (CSF) lactate is elevated. When drawing serum lactate levels, it is important to also measure pyruvate. The most common cause of high serum lactate is that the tourniquet has been left on the arm too long while searching for a good vein; the resulting acute anaerobic state causes pyruvate to become converted to lactate; hence the lactate is high but the pyruvate is very low. In mitochondrial cytopathies, by contrast, if the lactate is elevated, the pyruvate is normal or also high (30). Simultaneous measurement of serum pyruvate will thus remove all doubt about whether an elevated lactate is merely a biological artifact. Other differential diagnoses of high serum lactate are shown in Table 2.1.

Imaging of the brain is an important diagnostic study in patients with mitochondrial cytopathy involving encephalopathy. Characteristic lesions may be seen with both computed tomography (CT) and magnetic resonance imaging (MRI) in the regions of the basal ganglia in particular (Fig. 2.1).

Mitochondrial syndromes can be broadly divided into two large categories: those in which the muscle biopsy exhibits ragged-red fibers and those that show many other features of mitochondrial myopathy but not ragged-red fibers. Ragged-red fibers are muscle fibers seen in frozen section and stained with modified Gomori trichrome stain that show irregularly shaped, bright red deposits in the subsarcolemmal and intermyofibrillar sarcoplasm, contrasting with the green myofibrils. The red color with this histochemical stain is derived from the strongly lipophilic chromotrope-2R in the stain, which has a particularly strong affinity for phospholipids; mitochondrial membranes contain a large amount of the phospholipid sphingomyelin. Electron microscopic examination of the ragged-red zone reveals large numbers of closely packed normal and abnormal mitochondria.

A majority of adults presenting clinically with mitochondrial cytopathy have a ragged-red fiber syndrome with combined complex I and IV deficiencies, whereas most infants and children symptomatic with mitochondrial
disease do not have this combination, hence the designation non–ragged-red fiber syndromes. A common combination defect in early childhood is one of complexes III and V, which is unusual in adults (31). Infants with complex I and IV defects may not show ragged-red fibers histologically by trichrome stain until 3 years of age or older; hence this marker is not entirely reliable in infant muscle biopsies.