From the Regional Neuromuscular Diseases Unit, King's Neurosciences Centre, London, England.
This article was corrected | View correction
One of the great challenges in molecular biology is to understand the mechanisms by which a particular genetic defect gives rise to the disease or diseases associated with it. Mitochondrial myopathies are no exception. This review emphasizes some of the important mechanisms that can modify the phenotypic expression of the genetic disorders seen in mitochondrial myopathies.
Mitochondrial myopathies can manifest with a wide variety of symptoms affecting multiple organ systems, particularly the nervous system. The classic phenotypes are represented by the following: (1) Kearns-Sayre syndrome, consisting of bilateral ptosis, ophthalmoplegia, retinitis pigmentosa, ataxia, cardiac conduction defects, and an elevated level of cerebrospinal fluid protein; (2) mitochondrial encephalopathy with lactic acidosis and strokelike episodes (MELAS); and (3) myoclonic epilepsy with ragged red fibers (MERRF).
These diverse syndromes are linked by the finding of ragged red fibers (RRF) in muscle biopsy specimens, even in cases in which the muscle does not seem to be clinically involved. The RRF appearance evident on Gomori trichrome stain is due to the preferential staining of phospholipid and reflects the abnormal proliferation of mitochondria, which contain an inner membrane rich in cardiolipin. Thus, the mitochondrial myopathies were originally defined as multisystem disorders associated with RRF evident in muscle biopsy specimens.
The presence of morphological abnormalities of the mitochondria was believed to imply a functional abnormality. Mitochondria are the site of numerous metabolic pathways, but the most important pathway is the production of energy in the form of adenosine 5‘-triphosphate (ATP), by a process known as oxidative phosphorylation. It is impairment of this biochemical pathway that underlies most mitochondrial myopathies. Figure 1 shows a schematic view of the oxidative phosphorylation system. It consists of 5 multimeric protein complexes embedded to varying degrees in the mitochondrial inner membrane. Some of these complexes can accept electrons from reducing equivalents that derive from the metabolism of glucose and fatty acids. These electrons are transferred from one complex to another down an energy gradient. The energy derived from this electron transfer is used to translocate protons across the mitochondrial inner membrane, thus generating what is known as a proton motive force. This proton motive force can be discharged through complex V (ATP synthetase), enabling it to catalyze the phosphorylation of adenosine 5′-diphosphate to ATP. The multiple protein subunits that compose the oxidative phosphorylation complexes are uniquely the product of 2 distinct genomes. As shown in Table 1, most of the subunits are encoded by nuclear DNA, but 13 are encoded by mitochondrial DNA (mtDNA).
The mitochondrial oxidative phosphorylation system. NADH indicates nicotinamide adenine dinucleotide (reduced form) dehydrogenase; Q, ubiquinone; C, cytochrome c; ADP, adenosine diphosphate; and ATP, adenosine triphosphate.
Figure 2 is a schematic diagram of the mtDNA, which consists of a double circular molecule approximately 16.5 kb long. As well as containing the genes for 13 protein subunits of the oxidative phosphorylation system, it contains genes for 2 ribosomal RNAs and for 22 transfer RNAs (tRNAs).1 The compact organization of this small DNA molecule is achieved by a variety of mechanisms, some of which have consequences for the expression of mutations. The mtDNA contains few introns or noncoding regions, which increases the likelihood of deleterious mutations. Several of the genes have multiple roles, each of which may be differentially affected by mutations. For example, the tRNA genes are scattered throughout the mitochondrial genome and, in many cases, are interposed between the protein coding genes, which enables them to act as "punctuation marks," while the tRNA gene for leucine (
tRNAleu[UUR]; [UUR] refers to the anticodon sequence) contains a decatetrameric sequence that acts as a terminator of mtDNA transcription.2
Mitochondrial DNA. OH indicates origin of replication of light strand; D-loop, displacement loop; Thr, threonine; H-strand, heavy strand; Leu, leucine; Ser, serine; His, histidine; Arg, arginine; Gly, glycine; Lys, lysine; Asp, aspartic acid; OL, origin of replication of heavy strand; Tyr, tyrosine; Cys, cysteine; Asn, asparagine; Ala, alanine; Trp, tryptophan; fMET, formylmethionine; Gln, glutamine; Ile, isoleucine; 16S, ribosomal RNA 16S; Val, valine; 12S, ribosomal RNA 12S; Phe, phenylalanine; Pro, proline; Glu, glutamic acid; and L-strand, light strand. Transfer RNAs are indicated by short transverse lines and relevant amino acid abbreviation.
Mitochondrial DNA has other unique properties with implications for mutation expression. It is maternally inherited, ie, all mtDNA derives from the mother, and only women will pass on mtDNA to their offspring.3,4 Mitochondrial DNA exists in multiple copy numbers; typically, a mitochondrion can contain between 2 and 10 copies of mtDNA. Furthermore, mtDNA has a high rate of spontaneous mutations, resulting from its location in a hostile environment in which many free radicals are generated and from its lack of the basic repair mechanisms that are present in the nucleus.5,6
Like any other DNA, mtDNA must be replicated, and its protein coding genes must be transcribed into messenger RNA (mRNA), which then must be translated into the protein sequence. The replication and transcription of mtDNA are under the control of the nuclei genome. Transcription of mtDNA is polycistronic, ie, it generates a single length of RNA containing all the transcripts of each of the genes, and there must be a mechanism for precisely cleaving this polycistronic transcript into the individual RNAs.7 This cleavage is performed by RNA endonucleases, which are nuclear genome encoded. In contrast to replication and transcription, mtDNA is autonomous in translating its own mRNA because it has the genes for 2 ribosomal RNAs and for 22 tRNAs that constitute a complete translation system.
As shown in Table 1, only a small proportion of the subunits constituting the oxidative phosphorylation system are derived from mtDNA. The rest of the protein subunits are encoded by the nuclear genome and are therefore synthesized in the cytosol. These protein subunits must be transported into the mitochondria across the impermeable inner membrane. This involves an elaborate import machinery in which the amino terminal amino acid sequence of the proteins acts as a signal sequence that targets the proteins to receptors located at specialized regions of the mitochondrial inner membrane. These receptors specify the location to which the protein should be imported within the mitochondrion. Heat shock (also known as chaperonin) proteins facilitate the unfolding of the protein subunits, allowing them to be transported as straight lengths through the double membrane of the mitochondrion. The protein subunits are then refolded for the assembly of the oxidative phosphorylation apparatus.8 All the elements of this import process are encoded and therefore controlled by the nuclear genome.
The biosynthesis of the oxidative phosphorylation system just described allows the postulation of a number of mechanisms that could give rise to a defect of the oxidative phosphorylation system (Table 2). The literature contains examples of each of the mechanisms that causes mitochondrial disease. One of the important clinical implications of these mechanisms for the pathogenesis of mitochondrial disease is that only some of them involve a primary defect of mtDNA that might be expected to be maternally inherited. Clinicians often base diagnosis of mitochondrial disease on the fact that it is maternally inherited, but mtDNA defects secondary to a nuclear gene defect may have an autosomal pattern of inheritance.
Since the first primary mtDNA defect was discovered,9 numerous additional mtDNA mutations have been reported. This has been accompanied pari passu by a great expansion of the clinical phenotypes that are now recognized to be mitochondrial diseases (Table 3). Some of these syndromes such as Leber hereditary optic atrophy, are not associated with RRF. Some syndromes, such as maternally inherited deafness with or without diabetes, do not have central nervous system involvement, while other syndromes have selective features, such as increased sensitivity to aminoglycosides. Because of this broadening spectrum of clinical diseases, only some of which are associated with RRF, it has been suggested that these syndromes should be termed mitochondrial cytopathies, rather than myopathies, and defined as syndromes that may be associated with morphological, biochemical, or genetic abnormalities of mitochondria, singly or in combination.
Defects of mtDNA secondary to a nuclear mutation result in abnormalities of mtDNA replication and lead to multiple deletions of mtDNA10- 14 or a depletion in the total mtDNA content.15 Multiple deletions of mtDNA mean that a given individual has several populations of mtDNA containing deletions of variable length or in different positions in the mtDNA. The primary mtDNA defects as listed in Table 2 are the focus of the remainder of this review, and particular reference will be made to the single deletions as a model for describing the various properties of primary mtDNA defects and how these modify the expression of the mutation.
Figure 3 shows the result of Southern blot analysis of mtDNA from a healthy person and from 3 patients with mitochondrial disease. Whole cellular DNA, ie, nuclear and mitochondrial DNA, has been extracted from muscle homogenate, and this has been digested with a restriction endonuclease that cuts the circular molecule of mtDNA at just 1 site, thereby generating a single straight length of mtDNA. The digest has been run on a gel and mtDNA alone is visualized by using radioactive probes for mtDNA-specific sequences. The healthy person has a single band that corresponds to the 16.5-kb size of whole mtDNA. The abnormal pattern seen in the 3 patients with mitochondrial cytopathy illustrates several important properties of mtDNA single deletions. Each of the affected patients has an extra band of mtDNA visible in addition to the normal band. This second band is a smaller molecular mass because some of the mtDNA has been deleted. The different persons have different-sized deletions, but each person has only 1 size of deletion. Indeed, the size of the deletion is identical in any tissue of a given person in whom it is found, hence the term single deletion. The result from the Southern blot analysis also shows that the patients have normal mtDNA and mtDNA with a deletion. This is because only some of the multiple copies of mtDNA within the mitochondria are affected by a deletion. Thus, mtDNAs are described as being heteroplastic for the deletion, and the proportion of normal to abnormal mtDNA, ie, the degree of heteroplasmy, can vary from patient to patient.
Southern blot analysis of mitochondrial DNA from a healthy person and from 3 patients with mitochondrial cytopathy.
Figure 4 shows the site and extent of the single deletions found in a series of 33 patients with mitochondrial cytopathy. The deletions are long, in some cases as long as 7 kb, which represents almost half the 16.5-kb size of the normal genome. Almost half of the patients have an identical deletion, and this has been termed the common deletion. Many of the remaining patients have a deletion that clusters around the common deletion. This deletion "hot spot" is flanked by a 13–base pair (bp) repeat sequence, and it is hypothesized that when the double-stranded mtDNA separates during replication, the upstream base pair repeat subsequently misaligns with the downstream 13-bp repeat, leading to the generation of a redundant fragment of mtDNA that is degraded.16- 18 This results in an mtDNA genome smaller than the normal genome from which it derived.
Map of the deletions found in 33 patients with mitochondrial myopathy. 12S indicates ribosomal RNA 12S; 16S, ribosomal RNA 16S; and bars, tRNA genes.
Several clinical and biochemical observations have been made in association with the single deletions of mtDNA. The single mtDNA deletions seem to be associated with 2 main clinical phenotypes. The first phenotype is ophthalmoplegia with or without additional features that may constitute elements of a partial or full Kearns-Sayre syndrome phenotype.19,20 The second phenotype is that of Pearson syndrome in which, at birth or shortly afterward, catastrophic multiorgan failure with metabolic acidosis develops that affects the bone marrow, gastrointestinal system, kidneys, and liver. This syndrome is usually fatal, but infants have survived with resolution of the renal, hepatic, and gastrointestinal failure, but with residual abnormalities identical to Kearns-Sayre syndrome.21 This is a dramatic example of how the phenotype in mitochondrial disease can change, but lesser degrees of change can be seen with aging. Mitochondrial disease can appear de novo later in life, or additional features of the disease can emerge.22 An additional clinical observation that has been evident in patients with mitochondrial disease due to a single mtDNA deletion is that they do not have a family history of mitochondrial disease. A primary mtDNA defect, such as the single deletions, would normally be expected to be maternally inherited.
Biochemical studies have attempted to correlate the site(s) of the respiratory chain defect with the site and extent of the deletion of the mtDNA. Deletions that only erase genes for complex I subunits may have a defect of the respiratory chain that is confined to complex I, but this is not invariably so. Conversely, extensive deletions that involve subunits of several respiratory chain complexes may have apparently normal respiratory chain function. Some of these conundrums may merely reflect the limitations of biochemical analysis and access to suitable tissue, but nevertheless, there is often disparity between the site and size of the mtDNA deletion and the biochemical defect that arises.23
These genotype-phenotype observations have generated a number of questions about why the single deletions are sporadic and why there is variability in the clinical and biochemical phenotype that can change with time.
The single deletions are heteroplasmic, and the percentage of mtDNA harboring a single deletion can vary from person to person. Postmortem studies have shown that this percentage of mutant mtDNA can vary in different tissues in a given person.20,24,25 In 1 such case (Table 4), the percentage of mutant mtDNA varied from 4% for smooth muscle to 50% for skeletal muscle. The percentage of mutant mtDNA, or the percentage of heteroplasmy, was constant in multiple samples from most tissues; 1 notable exception was the brain. This difference was believed to be related to the different phylogenetic origins of the cerebrum and cerebellum. Variable heteroplasmy may be one way to explain why different tissues are affected in different people. The percentage of heteroplasmy refers to the relative proportion of normal vs mutant mtDNA. In the expression of these mutations, a more important factor is likely be the absolute number of each species of mtDNA. A given tissue may have 50% mutant mtDNA, but if there has been an increase in the copy number of mtDNA, the absolute amount of normal mtDNA may be normal. Little is known about the absolute numbers of mutant and normal mtDNA. Also of relevance to this scenario is the question of whether mutant mtDNA is functionally dominant over wild-type mtDNA, so that even if there were a normal, or near normal, absolute amount of wild-type mtDNA, the presence of mutant mtDNA might still result in its abnormal expression. In situ hybridization experiments attempting to address the issue of which mtDNA species is functionally dominant have given conflicting answers.26,27
Another important phenomenon that affects the expression of mtDNA mutations is the phenomenon of complementation. Because mtDNA contains a minimum set of 22 tRNAs, allowing it to be autonomous in translation of the genome, and because these tRNAs are scattered throughout the genome, it follows that a deletion of any appreciable size will result in loss of tRNAs and, consequently, render the genome unable to translate genes that remain intact. Such a genome might, however, achieve successful translation if it could obtain the missing tRNAs from adjacent normal mtDNA. This is the process of complementation. A number of factors could, in theory, influence the ability of abnormal mtDNA to complement for the missing tRNAs. A longer deletion will result in more missing tRNAs and thus require more complementation for normal translation to occur. The position of the deletion might determine how many tRNA genes are lost. The percentage of heteroplasmy might be important. There could be a threshold above which there is insufficient normal genome to allow complementation for missing tRNAs. The distribution of mutant mtDNA within individual mitochondria might influence this threshold value, as illustrated in Figure 5, which shows heteroplasmy for a deletion at 50%. Figure 5, top, shows segregation of mutant and normal mtDNA in separate mitochondria; in this situation complementation is impossible. On the other hand, in Figure 5, bottom, the 2 genomic populations are mixed inside individual mitochondria, and complementation might therefore be possible. Little is known about the distribution of heteroplasmic populations of mtDNA within mitochondria. A percentage figure of the amount of mutant mtDNA in a sample from a given person actually represents a composite of the percentages present in individual mitochondria in that sample, and such a figure does not consider how the mutant mtDNA genomes are distributed among the mitochondria. Thus, while threshold percentage mutant mtDNA values for complementation are quoted in the literature,28 this conceals the fact that persons with identical percentages of a given mtDNA mutation may have a completely different distribution of the mutation within mitochondria.
Top, Mitochondria showing segregation of wild-type and mutant mitochondrial DNA (mtDNA). Bottom, Mitochondria showing nonsegregation of wild-type and mutant mtDNA. In both cases, the overall proportion of mutant to wild-type mtDNA is 50%.
Indirect evidence that complementation can occur comes from biochemical, in situ hybridization, and in vitro mitochondrial translation experiments and seems to suggest that it depends on some of the aforementioned factors.23,29,30 Direct evidence for complementation would come from the isolation of "fusion proteins." In the case of the "common" deletion, the break points occurred in the ATPase 6 and the ND5 genes, and this generated a new sequence of mtDNA with a premature stop codon. This novel sequence would only be translated if complementation occurred, and this would generate a fusion protein of just over 1-kd molecular weight. A fusion transcript has been isolated, but no fusion protein has yet been found.26,28 This may reflect difficulty in isolating what may be a transient unstable protein. If a fusion protein were produced in vivo, this might interfere with the assembly of the normal respiratory chain complexes. This is one way in which a deletion might result in a variety of biochemical defects that are not necessarily directly related to the genes that have been lost.
Factors have been discussed that might affect the assembly and, therefore, the function of the respiratory chain. In addition to defects in the energy supply, however, energy requirements must be considered. Different tissues have different energy requirements, so for a disease to be expressed in a particular tissue, the balance between the energy demands and energy supply of that tissue is important. The central nervous system has a constant high energy demand and therefore only a small decrease in the energy supply could result in tissue abnormality. Skeletal muscle has a low basal demand for energy, but because this demand is greatly increased during exercise, then symptoms may occur. The balance between energy demand and supply does not remain constant throughout life because the efficiency of the respiratory chain, even in normal mitochondria, declines with age because of the accumulation of spontaneous mtDNA deletions with time.31 Thus, a given tissue initially may have a sufficient energy supply, but as the person ages and the energy supply declines, symptoms attributable to the declining function of that tissue may occur.
Another factor governing the susceptibility of a particular tissue is its mitotic activity. Figure 6 shows the result of a Southern blot analysis of 3 mtDNA samples from the same patient. In the first lane, there is 50% heteroplasmy for a single deletion in mtDNA extracted from skeletal muscle. The second and third lanes show mtDNA extracted from fibroblast and myoblast cultures derived from the same patient. The same deletion is present in the fibroblast cultures, but at a much lower percentage, while no abnormal mtDNA is present in the myoblast cultures. This illustrates that while the proportion of mtDNA containing a deletion may be fixed in a postmitotic tissue such as skeletal muscle, in a dividing cell line, or in cell culture, there is the opportunity for the mtDNA mutation to be selected out. Thus, postmitotic tissue, such as in the central nervous system and skeletal muscle, may be more vulnerable to the accumulation of mutant mtDNA. This phenomenon may also account for the changing phenotype with time. The most dramatic example of this, namely neonates with Pearson syndrome maturing to a Kearns-Sayre syndrome phenotype, is explicable on the basis that the cells of the tissues that are initially severely affected, ie, the hemopoietic tissues and the liver, are continuously dividing and thus have the opportunity to select out mtDNA containing a deletion. For the same reason, the mtDNA single deletions may not be found in leukocytes because this mutation has been selected out of the hemopoietic cell line.9 The rapidly dividing germ cell line also has an opportunity to select out this mutation, and this may account for the sporadic rather than inherited occurrence of this mutation. How does this selection process occur? The mtDNA containing a deletion may actually have a replicative advantage over wild-type mtDNA because of its smaller size, but set against that is the biochemical disadvantage that the mutation confers on a particular cell. This explanation does not seem to apply to all types of mtDNA mutation; tRNA and protein gene point mutations are not selected out in the hemopoietic cell line and are maternally inherited, but they can result in just as much biochemical deficit as the single deletions.
Southern blot analysis of mitochondrial DNA from skeletal muscle and from fibroblast and myoblast cultures from the same patient.
Many of the factors that influence the expression of mtDNA single deletions, such as heteroplasmy, complementation, and tissue susceptibility in terms of energy demand and mitotic activity, also apply to other types of mtDNA mutation. There are, however, additional mechanisms that may apply to tRNA mutations resulting from their alternative functions. Thus, as well as causing loss of tRNA function, mutations of some tRNAs may result in the loss of punctuation between the protein genes or, in the case of the
loss of the termination site for mtDNA transcription. However, this still leaves several questions unanswered. Single base pair substitutions in a variety of positions in the same
tRNAleu(UUR) gene can result in quite a variation in the phenotype. Classically,
tRNAleu(UUR) mutations are said to result in the MELAS phenotype, but the correlation between the two is not very strong as there are a wide variety of other phenotypes. Even within this same gene, differing positions of the point mutation can result in different phenotypes, ranging from deafness, with or without diabetes, to cardiomyopathy, MELAS, and MERRF (Table 5). Even the nucleotide 3243 A to G base pair substitution can give rise to distinctly different phenotypes. Extrapolating from the single deletion experience might lead to the supposition that differing tissue involvement for the
tRNAleu(UUR) (3243) mutation could be explained on the basis of variable heteroplasmy for the mutation in the different tissues. However, postmortem studies have shown that unlike the single deletions, the proportion of this mutation is constant in all tissue sampled.32 In addition to lacking an explanation for how diverse phenotypes can emerge from mutations within the same gene, the converse situation is also difficult to explain. The mutations that have been associated with Leber hereditary optic atrophy illustrate how different point mutations affecting different genes can cause a similar phenotype (Table 6).
Accepted for publication June 19, 1997.
I thank John B. Clark, PhD, for helpful comments during the preparation of the manuscript.
The data used in preparing Tables 5 and 6 were derived from the MITOMAP; Mitochondrial Human Genome Database, Emory University, Atlanta, Ga (http://www.gen.emory.edu/mitomap.html), in December 1995.
Figures 2, 4, and 5 are used with the permission of John A. Morgan-Hughes, MD.
Reprints: Michael R. Rose, MD, Regional Neuromuscular Diseases Unit, King's Neurosciences Centre, Denmark Hill, London SE5 9RS, England.
Thank you for submitting a comment on this article. It will be reviewed by JAMA Neurology editors. You will be notified when your comment has been published. Comments should not exceed 500 words of text and 10 references.
Do not submit personal medical questions or information that could identify a specific patient, questions about a particular case, or general inquiries to an author. Only content that has not been published, posted, or submitted elsewhere should be submitted. By submitting this Comment, you and any coauthors transfer copyright to the journal if your Comment is posted.
* = Required Field
Disclosure of Any Conflicts of Interest*
Indicate all relevant conflicts of interest of each author below, including all relevant financial interests, activities, and relationships within the past 3 years including, but not limited to, employment, affiliation, grants or funding, consultancies, honoraria or payment, speakers’ bureaus, stock ownership or options, expert testimony, royalties, donation of medical equipment, or patents planned, pending, or issued. If all authors have none, check "No potential conflicts or relevant financial interests" in the box below. Please also indicate any funding received in support of this work. The information will be posted with your response.
Some tools below are only available to our subscribers or users with an online account.
Download citation file:
Web of Science® Times Cited: 13
Customize your page view by dragging & repositioning the boxes below.
Enter your username and email address. We'll send you a link to reset your password.
Enter your username and email address. We'll send instructions on how to reset your password to the email address we have on record.
Athens and Shibboleth are access management services that provide single sign-on to protected resources. They replace the multiple user names and passwords necessary to access subscription-based content with a single user name and password that can be entered once per session. It operates independently of a user's location or IP address. If your institution uses Athens or Shibboleth authentication, please contact your site administrator to receive your user name and password.