Mitochondrial encephalopathies
The mitochondrial genome is a circular molecule of DNA (mtDNA) comprising 16 569 base pairs. Each mitochondrion contains 2–10 copies of the genome, which encodes 22 transfer RNAs (tRNAs), two ribosomal RNAs (rRNAs), and 13 subunits of the respiratory chain (Fig. 24.1). Since spermatozoa contribute no or minimal mitochondria during fertilization, mitochondria and their genomes are essentially all maternally inherited (Figs 24.2, 24.3).
24.1 Diagram of human mtDNA.
The outer and inner circles represent the heavy and light strands respectively. The genes encoding the mitochondrial rRNAs are shown as 12 s and 16 s rRNA, those encoding the reduced form of nicotinamide adenine dinucleotide dehydrogenase (NADH) subunits as ND1–6, the cytochrome oxidase subunits as COI–III, and cytochrome b as Cyt b. Also shown are the genes encoding subunits 6 and 8 of adenosine triphosphatase (ATPase). Conventional single letter amino acid abbreviations are used to indicate the tRNA genes. The broad gray arc shows the region involved to a greater or lesser extent in most deletions of mtDNA. The blue arrows indicate the main sites of point mutations or deletions in mitochondrial disorders that affect the CNS.
A wide range of disorders affecting the central nervous system (CNS) is attributable to defective mitochondrial function. Most mtDNA defects are heteroplasmic (i.e. cells contain a mixture of mutant and wild-type mtDNA) (Fig. 24.4). As mitochondria are randomly segregated during mitosis, the proportion of mitochondria containing mutant genomes varies from cell to cell. However, the levels of mutated mtDNA tend to be highest in non-dividing cells such as neurons, and skeletal and cardiac muscle. The frequency of mutations and deletions increases with age to a much greater extent in mitochondrial than in nuclear DNA. Several factors probably contribute: much of the damage is mediated by reactive oxygen species that are generated within the mitochondrion itself; there is no ‘redundant’ non-coding mtDNA; mtDNA undergoes progressively more frequent replication (exceeding that of nuclear DNA); the responsible polymerase lacks the proof-reading and repair activities of nuclear DNA polymerases.
Tissues with a high energy demand, such as neural tissue and muscle, are particularly susceptible to impairment of mitochondrial function (Fig. 24.5). Probably because relatively few mitochondria (and therefore mitochondrial genomes) are transmitted by a mother to her progeny during embryogenesis, pronounced shifts in the degree of mitochondrial heteroplasmy can occur from generation to generation (Figs 24.6, 24.7).
24.6 Mitochondrial genetic bottleneck effect.
Rapid and marked shifts in degree of mitochondrial heteroplasmy can occur in one generation. A mitochondrial genetic bottleneck that permits only a small subset of maternal mtDNA genomes to be effectively transmitted to progeny could account for these incidences of sudden genetic drift. In this figure, the genetic bottleneck effect has resulted in a drastic shift to a state of wild-type homoplasmy for progeny 2. Given that the maternal mitochondria are approximately 50% mutant in this example, progeny homoplasmy would be unlikely if there was simply a random distribution of mtDNA genomes alone. Without the bottleneck effect, accumulated maternal mitochondrial mutations would always be transmitted in the germline. Over generations, additional accumulated mutations would result inexorably in non-functional mitochondria, non-viable oocytes or progeny, and species extinction. With random mitotic segregation, which may have a greater role in somatic cells, genetic drift is likely to be less pronounced. In contrast, the bottleneck genetic effect has the theoretical evolutionary advantage of potentially producing some progeny that have no mutant mtDNA. In the short term, deleterious mutations may accumulate in some oocytes or progeny, e.g. progeny 1, but these would be selected against.
24.7 Possible mechanisms for the mitochondrial genetic bottleneck.
Whether one or all of these mechanisms contribute to the putative genetic bottleneck remains under investigation.
In disorders with an increased level of oxidative stress, the age-related accumulation of mutations in mtDNA tends to be exacerbated. Some studies have suggested that this may impair oxidative phosphorylation and contribute to cell death in neurodegenerative diseases such as Parkinson’s disease and Alzheimer’s disease. (These neurodegenerative diseases are covered in Chapters 28 and 31). Several other diseases are either due to mutations in nuclear genes that encode mitochondrial proteins (e.g. Alpers syndrome, Menkes syndrome, and a range of defects of fatty acid, amino acid, and pyruvate metabolism), or cause secondary mitochondrial dysfunction (e.g. Batten’s disease). These too are considered separately, in Chapters 6 and 7.
This chapter covers only those mitochondrial diseases that involve the CNS and are due to mutations of mtDNA. In most of the diseases that are included in this chapter, the mutation in mtDNA is the primary genetic defect, but the chapter also covers a few disorders in which abnormalities in mtDNA are secondary to mutations of nuclear DNA (Table 24.1). Although Leigh syndrome can be caused by point mutations or deletions of mtDNA, it more often results from mutations in nuclear DNA, and the etiology, clinical and pathologic findings are therefore described in Chapter 6. An algorithm for genetic analysis of mitochondrial disorders is shown in Figure 24.8. Mitochondrial disorders that do not involve the CNS are not considered in this book.
MITOCHONDRIAL MYOPATHY, ENCEPHALOPATHY, LACTIC ACIDOSIS, AND STROKE-LIKE EPISODES (MELAS)
ETIOLOGY OF MELAS
In over 80% of cases, MELAS is caused by an A-to-G transition at nucleotide 3243 in the mitochondrial tRNALeu(UUR) gene. This transition is within the region of termination of transcription of the mitochondrial rRNAs, but whether this is relevant to the genesis of MELAS lesions is uncertain.
• This same mutation can also cause Kearns–Sayre syndrome (KSS), at least some features of which are often present (see below).
• When only a relatively low proportion of mtDNA is mutated at nucleotide 3243, patients may develop adult-onset, non-insulin-dependent diabetes mellitus (type I diabetes) with or without deafness (DMDF).
MELAS can also be caused by a T-to-C transition at nucleotide 3271 in the tRNALeu(UUR) gene.
The lesions of MELAS probably result from defective mitochondrial metabolism, in which case strictly-speaking they are not infarcts in that the necrosis is not due to ischemia/oligemia or hypoxia. Some researchers have suggested that the stroke-like lesions are due to impaired cerebral blood flow resulting from an accumulation of mitochondria within the cerebrovascular endothelium and smooth muscle cells. However, the stroke-like lesions have been shown to be associated with hyperemia rather than reduced perfusion.
MACROSCOPIC APPEARANCES
Scattered foci of necrosis and cavitation are usually evident in the cerebral cortex (Fig. 24.9). These infarct-like lesions may be of varying ages. Although they can occur in any part of the cerebral cortex, the occipital lobe is often most severely affected. The basal ganglia, thalamus, and cerebellum are much less commonly involved.
24.9 Macroscopic appearance of brain in MELAS.
Infarct-like lesions of varying ages. (a) Recent lesions (arrows) in the frontal cortex. The distribution of necrosis in these lesions differs from that associated with hypoxia/ischemia in that the crests of the gyri are more extensively affected than the depths of sulci, and the deep laminae are largely spared. The occipital lobes (b) are often severely affected in this case by lesions of differing ages. The older lesions have produced shrinkage and brown discoloration of the cortical ribbon.
MELAS
MICROSCOPIC APPEARANCES
The necrotic foci are histologically indistinguishable from infarcts of various ages (Fig. 24.10), although their distribution does not correspond to that of any particular arterial perfusion territory (Fig. 24.10, see also Fig. 24.9) and the crests of gyri are often more extensively affected than the depths of sulci. As with infarcts, there is capillary proliferation around and within the lesions. In many cases there is also mineralization in and around the walls of blood vessels in the globus pallidus, and occasionally in the cerebral white matter, thalamus, and dentate nucleus. This may be associated with fibrous thickening of the affected blood vessels and severe narrowing of the lumen. There may be spongy vacuolation of cerebral and cerebellar white matter, but this is unusual. Histology, histochemistry, and electron microscopy of skeletal muscle reveal ‘ragged-red’ fibers containing accumulations of morphologically abnormal mitochondria, some with paracrystalline inclusions. There are also accumulations of mitochondria in the endothelium and smooth muscle of intramuscular blood vessels (Fig. 24.10).
24.10 Histology of MELAS.
(a) Infarct-like acute lesion in MELAS. This section is through the edge of the lesion. (b) The boundary between the acutely affected superficial cortex and the intact deeper cortex has a serpiginous shape. (c) Chronic lesion, which is partly cavitated. (d) ‘Punched-out’ chronic lesions within cortex that is gliotic and depleted of neurons. (e) Skeletal muscle in MELAS. Demonstration of abnormal accumulations of mitochondria in paraffin sections of muscle from a patient with MELAS by in situ
