Introduction

Mitochondrial diseases represent one of the most exciting chapters of modern medicine. Knowledge of mitochondria and the relationship of mitochondrial dysfunction to human diseases has evolved during the past century, with an explosion of new information in the last decade [DiMauro and Bonilla, 2004]. During the last half of the 19th century, scientists gradually recognized the presence of subcellular organelles, and Benda coined the term mitochondrion in 1898 [Benda, 1898]. Benda was aware of threadlike granules within the cell that were barely detectable by existing methods. After this seminal observation, the metabolic role of mitochondria in cellular function was defined by a series of observations during the early part of the 20th century. The cytochrome system was described in 1925 [Keilin, 1925], and oxidation–reduction processes were described 4 years later [Warburg and Negelein, 1929]. The Krebs cycle was conceptualized in 1937, and later the phosphorylation of adenosine diphosphate (ADP) to adenosine triphosphate (ATP) was documented, together with the dependence of phosphorylation on oxygen consumption [Krebs and Kornberg, 1957]. By the middle of the 20th century, it was clear that the mitochondrion represented the intracellular domain for intermediary metabolism. In 1961, the chemiosmotic theory was proposed to explain the proton motive force that facilitates the synthesis of adenosine triphosphate [Mitchell, 1961]. In parallel with these biochemical observations, a series of studies described the ultrastructural characteristics of the mitochondrion [Palade, 1953]. The four components of the organelle were characterized, including the outer and inner membranes, the intermembranous space, and the inner matrix compartment. Subsequent studies have demonstrated that the inner mitochondrial membrane is largely impermeable to molecules of all sizes, and special adaptive mechanisms are necessary for the translocation of metabolites and proteins from the intermembranous space to the matrix. The protein importation process is energy-dependent, and requires the macromolecules to be unfolded before traversing the mitochondrial membranes and then refolded after entering the mitochondrial matrix [Bolender et al., 2008]. Surprisingly, only a handful of mitochondrial diseases have been attributed to defects of mitochondrial protein importation, and possible reasons for this are described later in this chapter.

Two important observations were made in 1963. Each observation is central to the understanding of mitochondrial diseases. The first observation recognized the presence of intramitochondrial fibers with DNA characteristics [Nass and Nass, 1963]. This observation was the first of many reports documenting DNA in mitochondria. It is realized that many human diseases are the result of mitochondrial DNA (mtDNA) mutations. The second observation was the description of ragged red fibers in biopsy specimens of skeletal muscle [Engel and Cunningham, 1963].

Because all mtDNA is derived from the ovum, mtDNA characteristics are inherited exclusively from the mother – hence the terms mitochondrial inheritance, maternal inheritance, and cytoplasmic inheritance [Fine, 1978; Giles et al., 1980]. These interchangeable terms describe a non-mendelian pattern of inheritance that characterizes human diseases resulting from mtDNA mutations. Each mitochondrion contains 2–10 copies of the mtDNA genome. Because cells have hundreds or thousands of mitochondria, more than 10,000 copies of mtDNA may exist in each cell. This genome is a small, double-stranded circular molecule containing 16,569 basepairs (bp) [Anderson et al., 1981]. The circular molecule contains a heavy and a light strand, and each strand contains its own origin of replication (Figure 37-1). Clear differences exist between the mitochondrial genome and the nuclear genome (Table 37-1). The mitochondrial genome contains no introns. The only noncoding region in mtDNA is the displacement loop (D-loop). The D-loop region contains 1000 basepairs and is the site of origin for replication of the heavy strand and the promoter regions for both light and heavy strand transcription. Because the universal genetic code does not apply to mtDNA, the mitochondrial genome requires its own transcriptional and translational factors for synthesis of mitochondrial proteins. The mitochondrial genome contains 37 genes. Thirteen genes encode structural proteins in the respiratory chain. The mitochondrial genome also contains 24 genes for protein synthesis. These genes include 2 ribosomal RNAs and 22 transfer RNAs. The mtDNA genes code for 13 messenger RNAs, and all 13 gene products are located in the respiratory chain (Table 37-2).

Fig. 37-1 Mitochondrial DNA contains 16,569 basepairs and 37 genes.

These genes include 2 ribosomal RNA, 22 transfer RNA, and 13 structural reading frames. Digestion of human mtDNA with a single endonuclease (PvuII) cleaves the circular molecule in one place and produces a linearized version. CO, cytochrome-c oxidase; Cyt, cytochrome; ND, NADH-CoQ reductase; O_L and O_H, origin of replication for L and H strands; P_H1, P_H2, primary transcripts of the H strand starting at H1, H2; P_L, light (I-strand promoter).

(From DeVivo DC: The expanding clinical spectrum of mitochondrial diseases. Brain and Development. 15;1–22, 1993. Elsevier Science Publishers, 1993.)

Table 37-1 Comparison of the Human Nuclear and Mitochondrial Genomes

Table 37-2 Structural (Mit) Gene Products Encoded by Mitochondrial DNA

The mitochondrial genotype is homoplasmic if all mtDNA genomes are identical. Conversely, the genotype is heteroplasmic if the genomes represent a mixture of wild-type mtDNAs and mutated mtDNAs. The phenotype is determined by the proportion of mutated genomes; when this proportion exceeds a threshold, the biologic behavior of the cell, the tissue, and indeed the individual changes, reflecting the impaired energy state. The threshold effect is a relative concept influenced by several factors, such as the age of the patient and the energy demands of any specific tissue or organ. For example, brain and muscle cells have high energy demands, as do the tissues of the developing child. In these situations, the threshold for phenotypic expression of a pathogenic mutation is lower. The threshold may vary from 60 to 70 percent in chronic progressive external ophthalmoplegia due to single large-scale mtDNA deletions, to 90–95 percent in the syndromes of mitochondrial encephalomyopathy with lactic acidosis and strokelike episodes (MELAS) and myoclonus epilepsy and ragged red fibers (MERRF).

Replicative segregation is a biologic concept that refers to the stochastic redistribution of the mtDNA genomes during mitochondrial and cell divisions. The random segregation of the mitochondrial genomes during replication influences the oxidative capability of the cellular progeny. The concepts of threshold effect and replicative segregation have provided some explanations for the variable phenotypic expression of maternally transmitted human diseases.

Inheritance Patterns

One of the many exciting features of mitochondrial diseases is the pattern of inheritance of these diverse conditions. The unique dual influence of the nuclear and mitochondrial genomes on the respiratory chain has captivated students of mitochondrial diseases. A biochemical defect involving the respiratory chain may be transmitted either by mendelian or by nonmendelian patterns. The strictly maternal inheritance of mitochondria determines the pattern of vertical transmission of mtDNA mutations. Clinical conditions related to mtDNA point mutations are transmitted from the mother to all her male and female progeny, but only daughters pass the condition to succeeding generations (Figure 37-2). This genetic profile is reminiscent of mendelian inheritance, including autosomal-dominant and X-linked patterns, but both genders are equally affected and there is no father-to-son transmission. Expression of the maternally inherited genetic defect is determined by replicative segregation and by the threshold effect. These biologic principles are demonstrated in several neurologic diseases associated with mtDNA mutations, including Leber’s hereditary optic neuropathy (LHON), MERRF, MELAS, and the syndrome of neuropathy, ataxia, retinitis pigmentosa/maternally inherited Leigh’s syndrome (NARP/MILS).

Fig. 37-2 Inheritance of a mitochondrial DNA HaeII polymorphism in a 33-member three-generation family.

The typical mitochondrial DNA cleavage pattern is depicted in the open symbols, and the atypical cleavage pattern is depicted in the black symbols. NT, not tested.

(From Giles RE, Blanc H, Cann HM, et al. Proc Natl Acad Sci USA 1980;77:6715.)

Single deletions of mtDNA generally occur sporadically, as is the case with Kearns–Sayre syndrome, progressive external ophthalmoplegia, and Pearson’s syndrome.

However, the pattern of inheritance for most mitochondrial diseases is not maternal. Rather, classic mendelian inheritance patterns apply to all disorders affecting biochemical pathways other than the respiratory chain, including defects of fatty acid oxidation, pyruvate utilization, and the Krebs cycle. Also, not surprisingly, several defects of the respiratory chain are inherited by mendelian inheritance because most subunits of respiratory chain complexes are encoded by nuclear DNA (nDNA) genes. Interestingly, many abnormalities of mtDNA are also due to mutations in nDNA genes; these mendelian diseases associated with mtDNA defects are due to faulty communications between the two genomes and are usually called defects of intergenomic signaling. For example, syndromes of progressive external ophthalmoplegia with multiple deletions of mtDNA are inherited as autosomal-dominant or, more rarely, autosomal-recessive traits. Another defect of intergenomic signaling is due to a quantitative defect of mtDNA (mtDNA depletion). The nuclear gene defects in these cases alter the biologic integrity or the replication of the mitochondrial genome and predispose the patient to multiple mtDNA deletions or mtDNA depletion that characterize these clinical syndromes [Spinazzola and Zeviani, 2009].

Metabolic Disturbances

The principal function of mitochondria is the oxidation of substrates and the synthesis of ATP. The primary oxidizable substrates include pyruvate, fatty acids, ketone bodies, and amino acids. Mitochondria also play a role in the intracellular sequestration of calcium and in the detoxification of ammonia in the urea cycle. Biochemical defects involving these pathways are associated with distinctive metabolic disturbances. A biochemical defect altering pyruvate metabolism directly or indirectly causes an elevation of pyruvic acid. Pyruvate is in equilibrium with lactate and alanine. As a result, there is lactic acidosis that is proportionate or disproportionate to the elevation of pyruvate, depending on the associated effect of the biochemical defect on the oxidation-reduction potential. If the oxidation–reduction potential is unaffected by the biochemical defect, the lactate and pyruvate elevations are proportional and the lactate/pyruvate ratio is normal (<20). In contrast, if the oxidation–reduction potential is disturbed by a primary defect involving the respiratory chain, the lactate values are disproportionately elevated and the lactate/pyruvate ratio is increased (>20). Patients with pyruvate dehydrogenase (PDH) deficiency typically have elevated lactate, pyruvate, and alanine concentrations and a normal lactate/pyruvate ratio. In contrast, patients with cytochrome-c oxidase (COX) deficiency have elevated lactate, pyruvate, and alanine concentrations, but the lactate/pyruvate ratio is increased. These metabolic observations serve as clues to the underlying biochemical defect and may be helpful when all other clinical and metabolic factors are considered simultaneously.

Defects of fatty acid metabolism may be associated with elevated free fatty acids, hypoketonemia, hypocarnitinemia, and dicarboxylic aciduria. Patients with biochemical defects involving beta oxidation typically have hypoketotic hypoglycemia, dicarboxylic aciduria, and secondary carnitine deficiency. Medium-chain acyl coenzyme A dehydrogenase (MCAD) deficiency is the classic example. The hypocarnitinemia associated with defects of beta oxidation typically is accompanied by a disturbance of the ratio of free carnitine to esterified carnitine. The disproportionately high esterified carnitine fraction is characteristic of the secondary carnitine deficiency states [Tein, 1995, 2003]. Defects of the carnitine cycle produce a different metabolic profile. As with defects of beta oxidation, affected patients may have hypoketotic hypoglycemia during fasting. However, carnitine cycle defects are not associated with dicarboxylic aciduria as a rule. The disturbances of carnitine also are more variable. Defects of the membrane carnitine transporter system and the carnitine-acylcarnitine translocase system are typically associated with low serum and tissue carnitine concentrations, whereas defects of carnitine palmitoyltransferase II (adult form) and carnitine palmitoyltransferase I are associated with normal or high serum and tissue carnitine concentrations. In contrast, the infantile multiorgan form of carnitine palmitoyltransferase II deficiency is associated with low serum and tissue carnitine concentrations [Tein, 2003; Ohkuma et al., 2009].

Equally important observations can be made by measuring circulating ketone body concentrations. Ketone bodies normally are negligible in the fed state and elevated in the fasting state. Ketone bodies are formed primarily in the liver from the metabolism of fatty acids and are exported to extrahepatic tissues that are capable of metabolizing these substrates. Classically, hypoketotic hypoglycemia develops in patients with defects of fatty acid metabolism during fasting. Serum ketone body values also are helpful in distinguishing the biochemical defects associated with congenital lactic acidosis. The serum ketone body concentrations in these various defects reflect the intracellular concentration of acetyl coenzyme A. Acetyl coenzyme A is the pivotal metabolite in mitochondrial metabolism. Acetyl coenzyme A may be formed from the decarboxylation of pyruvate, fatty acids, or ketone bodies. A fraction of the acetyl coenzyme A pool may be shunted into the β-hydroxy-β-methylglutaryl coenzyme A cycle in the liver to form ketone bodies. Defects involving the pyruvate dehydrogenase complex prevent the conversion of pyruvate to acetyl coenzyme A. Therefore, ketone bodies cannot be formed, and the increased pyruvate concentrations are shunted disproportionately into the gluconeogenic pathway. As a result, patients with pyruvate dehydrogenase deficiency have elevated pyruvate, lactate, and alanine concentrations, relative resistance to hypoglycemia during fasting, and a disproportionately low serum ketone body concentration, particularly during fasting. In contrast, patients with pyruvate carboxylase deficiency fail to convert pyruvate to oxaloacetate. As a result, acetyl coenzyme A concentrations increase, and there is a paradoxical formation of ketone bodies in the setting of the lactic acidosis. The deficiency of oxaloacetate and the secondary decrease in tissue aspartate concentrations also cause a paradoxical increase in the cytoplasmic oxidation–reduction potential and a decrease in the intramitochondrial oxidation–reduction potential. Therefore, patients with pyruvate carboxylase deficiency have an increased lactate/pyruvate ratio and a decreased β-hydroxybutyrate/acetoacetate ratio [De Vivo et al., 1977; Wang and De Vivo, 2009]. This metabolic profile is distinctive for pyruvate carboxylase deficiency. Similar, although less striking, observations may be seen with defects of the Krebs cycle or the respiratory chain as it relates to the paradoxical presence of ketosis in the setting of lactic acidosis.

Biochemical defects involving the urea cycle, specifically ornithine transcarbamoylase deficiency and carbamoyl phosphate synthetase deficiency, are associated with hyperammonemia. The urine orotic acid concentrations are increased in ornithine transcarbamoylase deficiency and decreased in carbamoyl phosphate synthetase deficiency.

Combinations of metabolic abnormalities also may exist. For example, defects involving the metabolism of long-chain fatty acids may be associated with the distinctive patterns of free fatty acidemia, dicarboxylicaciduria, hypocarnitinemia, hyperammonemia, and lactic acidosis. This profile is particularly evident in defects of the trifunctional enzyme protein and very-long-chain acyl coenzyme A dehydrogenase that are located in the inner membrane of the mitochondrion [Di Donato and Taroni, 2003]. These patients also have increased tissue concentrations of long-chain acylcarnitine, which may explain the organ toxicity associated with the defects of long-chain fatty acids.

Histopathologic Disturbances

The tissue reactions in mitochondrial diseases may be informative and direct the clinician’s attention to the primary metabolic defect, or they may be modest or absent. A normal muscle biopsy result does not rule out a mitochondrial disease. The skeletal muscle biopsy specimen often looks normal in defects of pyruvate carboxylase, pyruvate dehydrogenase, the Krebs cycle, and the urea cycle. The biopsy tissue may appear normal, with some defects involving fatty acid metabolism, including the adult form of carnitine palmitoyltransferase II, and several defects involving the respiratory chain. A lipid storage myopathy signifies a defect in fatty acid metabolism. Lipid accumulation is common in defects of beta oxidation, including glutaric aciduria type II and the infantile multiorgan form of carnitine palmitoyltransferase II deficiency. This tissue abnormality may be particularly striking in patients with the tissue-specific form of short-chain acyl coenzyme A dehydrogenase deficiency. In fact, several patients previously described with the myopathic carnitine deficiency syndrome [Engel and Angelini, 1973] later proved to have muscle-specific short-chain acyl coenzyme A dehydrogenase deficiency. Lipid storage in muscle is massive in two conditions due to defects of neutral triglyceride lipases. The first, also known as neutral lipid storage disease with ichthyosis (NLSDI), or Chanarin–Dorfman syndrome, characterized by weakness, hepatopathy, steatorrhea, and ichthyosis, is due to mutations in the GCI-58 gene. The second, known as neutral lipid storage disease with myopathy (NLSDM), is characterized by juvenile- or adult-onset myopathy, sometimes associated with cardiopathy, and is due to mutations in the PNPLA2 gene [Ohkuma et al., 2009].

Ragged red fibers are present in many respiratory chain diseases, including many nDNA defects. Ragged red fibers represent the morphologic counterpart of large-scale rearrangements of mtDNA or point mutations affecting transfer RNA (tRNA) genes and some, but not all, protein-coding genes.

The pathologic process in the brain is distinctive in many mitochondrial diseases. Three patterns are common. The first is a widespread insult to the brain tissue, with resulting microcephaly and ventricular dilatation. This pattern is seen in defects of the urea cycle, several defects associated with congenital lactic acidosis, and some of the defects associated with fatty acid metabolism. Malformations also may be seen, including agenesis of the corpus callosum, ectopic displacement of the olivary nuclei, and cystic destruction of the basal ganglia. This pattern is particularly striking in infants with pyruvate dehydrogenase E1 alpha-subunit deficiency. The second pattern is best typified by the neuropathologic changes associated with Leigh’s syndrome [Leigh, 1951]. These patients have a symmetric subcortical distribution of tissue injury, with a particular predilection for the basal ganglia, thalamus, brainstem, and cerebellar roof nuclei. Microscopic features include loss of brain cells, proportionate loss of myelin, reactive astrocytosis, and proliferation of the cerebral microvessels. In some patients, particularly those with MELAS, multifocal encephalomalacia develops. This pathologic condition typically is located in the posterior aspect of the cerebral hemisphere. The third pattern is a spongy encephalopathy with loosening and rarefaction of the neuropil. The spongy encephalopathy is the histopathologic counterpart of a defect in cerebral energy metabolism and is commonly seen in patients with Kearns–Sayre syndrome. Magnetic resonance imaging (MRI) may reflect these patterns of brain-tissue injury with a hyperintense signal of the central white matter on the T2-weighted image (Figure 37-3). Strokelike lesions in the posterior cerebral hemisphere are typical of MELAS. Signal hyperintensities involving the putamen, globus pallidus, and caudate nuclei are characteristic of Leigh’s syndrome. A diffuse signal abnormality involving the central white matter is typical of Kearns–Sayre syndrome. Intracranial calcifications also are seen in these conditions. Basal ganglia calcifications are most commonly associated with Kearns–Sayre syndrome and MELAS.

Fig. 37-3 Characteristic cranial abnormalities in mitochondrial encephalomyopathies on magnetic resonance imaging.

The images represent T2-weighted images from four different patients. A, Bilateral putaminal hyperintensities in a child with cytochrome-c oxidase-associated Leigh’s syndrome. B, Bilateral globus pallidus hyperintensities in a child with a familial Leigh’s syndrome phenotype. C, A left posterior cerebral hyperintensity in a child with MELAS. D, Bilateral white-matter hyperintensities in a child with Kearns–Sayre syndrome.

Classification of Mitochondrial Diseases

Several classification schemes of mitochondrial diseases are based on different criteria [DiMauro et al., 1985; Moraes et al., 1991; Shoffner et al., 1990]. The original attempts to classify mitochondrial diseases by clinical or morphologic criteria were unsatisfactory. The phenotypic heterogeneity of mitochondrial diseases complicated the clinical classification efforts. Similarly, the lack of histopathologic uniformity undermined the morphologic classification efforts. A biochemical classification of mitochondrial diseases was introduced in 1985 [DiMauro et al., 1985]. The clinical conditions were subclassified according to the primary site of the biochemical defect. In 1988, primary mutations of the mitochondrial genome were first recognized [Holt et al., 1988; Wallace et al., 1988]. These findings were particularly relevant to respiratory chain defects because this metabolic pathway is under the dual genetic influence of the mitochondrial and nuclear genomes. In recent times, the term mitochondrial diseases has been increasingly restricted to defects of one biochemical pathway, the respiratory chain [DiMauro and Schon, 2003], and further subgrouped as nuclear DNA defects and mtDNA defects (Box 37-1). However, in this review, the more general, biochemical classification for defects resulting from mutations in nuclear DNA is used, whereas for defects of the respiratory chain, a genetic classification is used.

Box 37-1 Biochemical Genetics Classification of Mitochondrial Diseases

Inherited Conditions

Nuclear DNA Defects

Defects of substrate transport

Defects of substrate utilization

Defects of the Krebs cycle

Defects of oxidation–phosphorylation coupling

Defects of the respiratory chain

Defects of protein importation

Defects of intergenomic signaling

Mitochondrial DNA Defects

Sporadic large-scale rearrangements

Transmitted large-scale rearrangements

Point mutations affecting structural genes

Point mutations affecting synthetic genes

Acquired Conditions

Infectious (e.g., Reye’s syndrome)

Toxic (e.g., MPTP (mitochondrial permeability transition pore))

Drugs (e.g., zidovudine)

Aging

General Clinical Features

From the clinical perspective, mitochondrial diseases can be divided into the following three major categories: defects of fatty acid oxidation, defects of pyruvate metabolism, and defects of the respiratory chain. Most mitochondrial diseases can be subsumed under these major categories.

Defects of Fatty Acid Oxidation

These conditions have specific clinical and metabolic signatures [Tein, 2003; DiDonato and Taroni, 2004]. Patients with defects of fatty acid oxidation experience metabolic decompensation during fasting. Organs that depend on fatty acid oxidation are primarily affected, including skeletal and cardiac musculature. Fatty acids accumulate in the liver in most of these defects and are shunted into auxiliary pathways, including omega oxidation. Omega oxidation occurs in the cytoplasm and accounts for the formation of dicarboxylic acids. Dicarboxylicaciduria is particularly evident in defects of beta oxidation (Figure 37-4). Acylglycine and acylcarnitine esters also accumulate in patients with defects of beta oxidation, offsetting the sequestration of coenzyme A that occurs with accumulating acyl coenzyme A thioesters. Depletion of serum and tissue carnitine stores may result. Increased bound carnitine fractions are characteristic of mitochondrial beta-oxidation defects. Absolute decreases of free and total carnitine fractions are particularly distinctive in carnitine cycle defects (see Figure 37-4). Defects of fatty acid oxidation result in the underproduction of acetyl coenzyme A and the resulting impairment of Krebs cycle activity and hepatic ketogenesis. As a result, many patients with defects of fatty acid oxidation have hypoketotic hypoglycemia. This metabolic state is associated with neurologic symptoms of altered consciousness. The precise mechanism of the cerebral dysfunction that is associated with defects of fatty acid oxidation is unknown. Decreased availability of ketone bodies and glucose deprives the brain of its two principal metabolic fuels. Fatty acids may enter the brain, but they are poorly metabolized. The accumulation of fatty acid intermediates in brain tissue also may have deleterious effects on cellular function.

Fig. 37-4 Schematic representation of fatty acid oxidation.

This metabolic pathway is divided into the carnitine cycle (A), the inner mitochondrial membrane system (B), and the mitochondrial matrix system (C). The carnitine cycle includes the plasma membrane transporter, carnitine palmitoyltransferase I, carnitine-acylcarnitine translocase system, and carnitine palmitoyltransferase II. The inner mitochondrial membrane system includes the very-long-chain acyl-CoA dehydrogenase and the trifunctional protein with three catalytically active sites. Long-chain acylcarnitines enter the mitochondrial matrix by the action of the carnitine palmitoyltransferase II to yield long-chain acyl-CoAs. These thioesters undergo one or more cycles of chain shortening, catalyzed by the membrane-bound system. Chain-shortened acyl-CoAs are degraded further by the matrix beta oxidation system. Medium-chain fatty acids enter the mitochondrial matrix directly and are activated to the medium-chain acyl-CoAs before degradation by the matrix beta oxidation system. AD, acyl-CoA dehydrogenase; CoA, coenzyme A; CPT, carnitine palmitoyltransferase; EH, 2-enoyl-CoA hydratase; HD, 3-hydroxyacyl-CoA dehydrogenase; KT, 3-ketoacyl-CoA thiolase; LC, long chain; MC, medium chain; SC, short chain; T_L, carnitine-acylcarnitine translocase; T_P, carnitine transporter; VLC, very long chain.

(Modified from Pons R, De Vivo DC: Primary and secondary carnitine deficiency syndromes. J Child Neurol 1995;10:1. With the kind assistance of Dr. Horst Schulz.)

Many patients with specific defects of fatty acid oxidation have been described since 1973, when the adult form of carnitine palmitoyltransferase type II and the myopathic form of carnitine deficiency were described [DiMauro and DiMauro, 1973; Engel and Angelini, 1973]. About half of all cases involve medium-chain acyl coenzyme A dehydrogenase deficiency. In 1975, the first patient with systemic carnitine deficiency was described [Karpati et al., 1975]. Many patients with systemic carnitine deficiency have been restudied and found to have medium-chain acyl coenzyme A dehydrogenase deficiency [DiDonato and Taroni, 2004]. The fatty acid oxidation defects can be subdivided into defects of the carnitine cycle, defects of the inner mitochondrial membrane, and defects of beta oxidation (see Figure 37-4). The specific clinical conditions are discussed later in this chapter.

Defects of Pyruvate Metabolism

Pyruvate is the end product of glycolysis. This metabolite can be reduced to lactate or transaminated to alanine in the cytoplasm. Otherwise, pyruvate is translocated across the mitochondrial membrane (Figure 37-5). In the mitochondrial matrix, pyruvate is carboxylated to oxaloacetate, or decarboxylated and activated to acetyl coenzyme A. The first reaction is catalyzed by pyruvate carboxylase, and the second is catalyzed by the pyruvate dehydrogenase complex. The two reaction products – oxaloacetate and acetyl coenzyme A – condense to form citrate as the primary reactant in the Krebs cycle. Defects of pyruvate metabolism involve the pyruvate dehydrogenase complex, pyruvate carboxylase, or several enzymes in the Krebs cycle [De Meirleir, 2002; De Vivo et al., 2002]. These defects are associated with elevated serum and tissue concentrations of pyruvate, lactate, and alanine. The lactate/pyruvate ratio is relatively preserved because the oxidation–reduction potential is maintained. This generalization, although not absolute, is helpful in attempts to distinguish defects of pyruvate metabolism from defects of the respiratory chain (discussed later). The Krebs cycle is central to intermediary metabolism. Complete biochemical defects of this pathway are probably incompatible with life. Four partial defects of the Krebs cycle have been described and are discussed later in this chapter.

Fig. 37-5 A schematic overview of intermediary metabolism.

Fatty acid oxidation is illustrated in detail in Figure 37-4. Pyruvate metabolism involves two reactions: pyruvate carboxylase and pyruvate dehydrogenase. Condensation of these two reaction products produces citrate at entry point of the Krebs cycle. Reducing equivalents are generated in the Krebs cycle and reoxidized by the respiratory chain. The five respiratory chain complexes contain approximately 80 polypeptides. Thirteen polypeptides are encoded by mitochondrial DNA, as illustrated in the upper right. The 13 mitochondrial DNA gene products are noted in parentheses above complexes I, III, IV, and V. ADP, adenosine diphosphate; ATP, adenosine triphospate; CO, cytochrome-c oxidase; CoA, coenzyme A; Cyt, cytochrome; ND, NADH-CoQ reductase; O_L and O_H, origin of replication for L and H strands; P_H1, P_H2, primary transcripts of the H strand starting at H1, H2.

(From De Vivo DC. The expanding clinical spectrum of mitochondrial diseases. Brain and Development. 15;1–22, 1993. Elsevier Science Publishers, 1993.)

Clinical Features of Mitochondrial Diseases