Journal of Neuro-Ophthalmology, September 1990, Volume 10, Issue 3

Mitochondrial mutations in neuro-ophthalmological diseases. A review.

Mutations in the genetic material of mitochondria have been described in patients with a range of neuro-ophthalmological and neuromuscular disorders. Many cases of Leber's hereditary optic neuropathy are caused by a single point mutation, for example, and Kearns-Sayre syndrome, chronic external ophthalmoplegia, and other mitochondrial cytopathies are frequently associated with large-scale deletions of mitochondrial genes. A knowledge of the role of the mitochondrial genome and of the precise nature of these mutations is important in understanding the etiology of such diseases and is already leading to more effective therapy.

JOllnllJl of Clinical Neurcrophthalmology 10( 3): 159- 166, 1990. Mitochondrial Mutations in Neuro- ophthalmological Diseases A Review Michael A. Morris, D. phil. © 1990 Raven Press, Ltd., New York Mutations in the genetic material of mitochondria have been described in patients with a range of neuro­ophthalmological and neuromuscular disorders. Many cases of Leber's hereditary optic neuropathy are caused by a single point mutation, for example, and Kearns­Sayre syndrome, chronic external ophthalmoplegia, and other mitochondrial cytopathies are frequently associ­ated with large- scale deletions of mitochondrial genes. A knowledge of the role of the mitochondrial genome and of the precise nature of these mutations is important in understanding the etiology of such diseases and is al­ready leading to more effective therapy. Key Words: Mitochondrial DNA- Mitochondrial muta­tions- Mitochondrial cytopathy- Kearns- Sayre syn­drome- Progressive external ophthalmoplegia- Leber's hereditary optic neuropathy- Maternal inheritance. From the Institute of Medical Genetics, University Medical Centre, Geneva, Switzerland. Address correspondence and reprint r~ que~ ts to Dr. M. A. Morris, Institute of Medical Genetics, UniversIty MedIcal Cen­tre, 9, avenue de Champel, 1211 Geneva 4, Switzerland. 159 Mitochondria are rod- shaped organelles ( the name comes from the Greek for " thread- shaped granule") found in the cytoplasm of most eukary­otic cells ( the exceptions are found among proto­zoa and fungi) ( 1). They have been called the " power plants of aerobic cells" ( 2) because they are the sole site of the respiratory chain and they are responsible for providing the great majority of the energy used by the body. Uniquely among animal organelles, mitochondria contain their own ge­nome, a small, circular chromosome carrying 37 genes essential to our survival. Why do mitochon­dria have these genes, and what is their relevance to neuro- ophthalmology? THE RESPIRATORY CHAIN Living organisms need energy- a lot of it. Throughout life on Earth, from the archaebacteria to the mammals, this energy has been obtained principally by the hydrolysis of the so- called " high- energy phosphate bonds" of adenosine triphosphate ( ATP), releasing adenosine diphos­phate ( ADP), inorganic phosphate ( Pi), and free energy, which is used by the systems of the body for growth, maintenance, and useful work. Ade­nosine triphosphate is not used for energy storage, however; it must be continually produced to meet the demands of the organism. In an average mam­malian cell, an ATP molecule will last only 1- 2 min; in 1 day, a resting human produces and uses as much as 40 kg of ATP ( 3). The great majority of ATP is produced by the mitochondria, more specifically by the respiratory chain, a multimolecular system located in the inner of the two mitochondrial membranes. The chain contains five major complexes of proteins and var­ious prosthetic groups ( complexes I- V) and two 160 M. A. MORRIS THE HUMAN MITOCHONDRIAL GENOME Complex Number of Encoded by Inhibitor subunits mtDNA I NADH dehydrogenase > 25 7 rotenone II succinate dehydrogenase 4 0 - III cytochrome b g 1 antimycin IV cytochrome c oxidase 13 3 cyanide V AlP synthase 12 2 oligomycin To understand the role of mitochondrial muta­tions in disease, four major phenomena must be considered: the mode of inheritance of mitochon­drial DNA of mtDNA; the high rate of mutation of mtDNA; the phenomenon of heteroplasmy; and the distinctive pattern of expression of mitochon­drial defects. Mitochondrial inheritance is non- Mendelian. The transmission of ( mtDNA) between genera­tions can be followed by the use of restriction frag­ment- length polymorphisms ( RFLPs), distinctive features of DNA that vary between normal indi­viduals. In this manner, it has been shown that, in humans as in many other species, an individual receives only maternal mtDNA ( 5); no paternal mi­tochondrial RFLP has ever been shown to be trans­mitted. Clinically, this has one obvious effect: he­reditary diseases caused by mitochondrial muta­tions are transmitted exclusively maternally, and THE MEDICAL GENETICS OF THE MITOCHONDRION duction of these proteins. All the remaining pro­teins of the respiratory chain and all the other pro­teins of the mitochondrion are encoded in the cell nucleus. The mitochondrial genes are packed as tightly as possible along the chromosome, with es­sentially no noncoding DNA within or between them ( in the nucleus, genes are split up by non­coding introns, which may be much longer than the expressed exons, and neighboring genes may be separated by millions of bases of apparently nonfunctional DNA). In general, the genes for the proteins and for the ribosomal RNA molecules are punctuated by the transfer RNA genes, elegantly simplifying the control of expression of mitochon­drial genes: a large part of the necessary RNA pro­cessing- the preparation of messenger RNA mol­ecules coding for single genes- is achieved merely by excising the transfer RNA portions of a single, full- length RNA transcript of the chromosome ( re­viewed in reference 4). FIG. 2. The composition of the respiratory chain com­plexes. H' H' 888e~ nl! r< sUCCinate fumarate 0, H, O The human mitochondrial chromosome is a cir­cular molecule of approximately 16,500 base pairs ( 16.5 kilobases, or kb) of DNA. For a chromosome, this is small; the sum of the human nuclear chro­mosomes is about 200,000 times this size. It has two complementary strands of DNA, differing in weight as a result of their different chemical com­position; they are termed the H ( heavy) and L ( light) strands. The mammalian mitochondrial chromosome ( Fig. 3). is a model of evolutionary economy; every nonessential feature has been eliminated. It carries just 37 genes, 28 of which are encoded in the H strand. Thirteen of the genes encode essential proteins of the respiratory chain, and the remaining 24 encode ribosomal and trans­fer RNA molecules whose sole function is the pro-smaller molecules, coenzyme Q and cytochrome c ( Fig 1). In the presence of oxygen, the first four complexes and the latter molecules use the energy derived from our food to generate a proton gradi­ent across the inner mitochondrial membrane; the fifth complex then allows these protons to flow back into the mitochondrion, while harnessing the energy of the gradient to synthesize ATP. The proton gradient, the ultimate source of so much of our metabolic energy, is generated by the transfer of charge ( as electrons) along the respira­tory chain, from complexes I and II to coenzyme Q, and then linearly through complex III, cy­tochrome c, and complexes IV and V. Inhibition or inefficiency at any stage of the chain will have an effect on ATP production. For example, rotenone prevents complex I from contributing to the proton gradient, but the chain can still function ( with re­duced efficiency) via the oxidation of succinate by complex II. However, cyanide shuts down the whole chain, by blocking it at complex IV ( Fig. 2). H' ADP + Pi ATP FIG. 1. Schematic representation of the major ele­ments of the respiratory chain of the mitochondrion. The enzyme complexes are numbered I- V. Abbrevia­tions: CoQ, coenzyme Q; Cyt c, cytochrome c. Vnl 111, No J, 1990 MITOCHONDRIAL MUTATIONS IN DISEASE 161 16.5kb FIG. 3. Genetic map of the human mi­tochondrial chromosome. The small cir­cles represent transfer RNA genes. All other genes are identified outside the NOS circle: NO, NAOH dehydrogenase sub­units ( complex I); CYB, cytochrome b subunit ( complex III); CO, cytochrome c oxidase subunits ( complex IV); and ATP, ATP synthase subunits ( complex V). Inside the circle: 0, origins of repli­cation of the heavy ( H) and light ( L) strands of the chromosome; and LHON, position 11778, the site of the mutation responsible for many cases of Leber's hereditary optic neuropathy. Mutation of the mitochondrial chromosome gives rise to heteroplasmy. Human cells contain, with a few specialized exceptions, many hundreds of mitochondria, and each mitochondrion contains several copies of its chromosome. After a single mutation event, which produces a new form of mtDNA, a cell is therefore heteroplasmic: in its cytoplasm, it has different genetic strains of mito­chondria. When such heteroplasmic cells divide, different populations emerge by variable segrega­tion of the mitochondria. In some descendent lines, mutant mitochondria become the more fre­quent form, whereas in others the normal popula­tion will predominate ( 10). In mammals, such seg­regation of mutant mitochondria between genera­tions can cause a return to homoplasmic animals after only two or three generations ( 11); this has not been demonstrated in humans, but may be the case. If so, one would predict that women with hereditary mitochondrial mutations could give rise to lines of descent differing in the severity of their disease: one line would become less severe from generation to generation, as the normal mitochon­dria became predominant, whereas the other would worsen as the mutation became prevalent. Thus, heteroplasmy provides the exception noted earlier to the rule that all offspring of a carrier fe­male will receive the mutation. The clinical expression of mitochondrial muta­tions is very variable and often unpredictable. As a result of heteroplasmy, the clinical expression of mitochondrial mutations can vary greatly. It is pre-eye LHON, Human mitochondrial DNA - I e02 eOI II ~ 03 __ 1111111: 1111111; 111. ATP8 ATP6 e03 all the offspring of a woman with such a hereditary disease will carry the mutation ( there is one excep­tion to this rule described later). The exclusively maternal transmission is per­haps explained purely physically. At fertilization, the human ovum has a high amount of cytoplasm containing tens or hundreds of thousands of mito­chondria, whereas the sperm cell possesses very little cytoplasm, with a few mitochondria gathered around the base of the tail. There is therefore a " dilution effect" of paternal mitochondria among the maternal, of one part in many thousands. It is not clear whether a fraction of a percent of our mitochondria are in fact of paternal origin or whether even these few are somehow actively ex­cluded at the moment of fertilization. Mitochondrial DNA mutates faster than nuclear DNA. Comparison of the DNA of human and other mammalian mitochondria has shown that gene sequences change at least six times more rap­idly in mitochondria than in the nucleus ( 6,7). This is almost certainly explained by a high rate of mu­tation coupled with the absence of the DNA repair systems that protect our nuclear genes. In 1987, Neckelmann and colleagues observed that this high rate of mutation would be expected to lead to a disproportionately high rate of maternally inher­ited diseases ( 8). Interestingly, Linnane and col­leagues have also suggested that mtDNA muta­tions, accumulating throughout the life of an indi­vidual, may contribute to aging and to certain degenerative diseases ( 9). I Clin Neuro- ophlhalmol, Vol. 10, No. 3, 1990 162 M. A. MORRIS dictable that mitochondrial mutations will have more effect in tissues with a high requirement for ATP; the most vulnerable tissues are, in order, the CNS ( including the retina), the type I skeletal mus­cle fibers, the heart, and the kidney ( 12). However, there is also an unpredictable element, produced by the variable proportions of normal and mutant mitochondria in different tissues: some tissues may be asymptomatic, whereas others may be sig­nificantly affected. In an excellent study of a family with a maternally inherited mitochondrial myopa­thy, Wallace and colleagues found a close correla­tion between the severity of symptoms, the range of clinical manifestations, and the degree of the biochemical defect in nine patients, all of whom presumably had the same mitochondrial mutation ( 12). Similarly, Rotig and colleagues ( 13) found that in a patient with Pearson's syndrome, which affects hematopoiesis, the exocrine pancreas, and the liver and kidneys and is caused by a mitochon­drial mutation, the more affected tissues contained a much higher proportion of mutant mtDNA than did the less affected one. Three principal forms of mitochondrial mutation have been detected in human disease. Deletions are the most common form of mito­chondrial mutation. As their name implies, dele­tions involve the loss of a section of the chromo­some; they have been found involving as little as 0.4 to as much as 8.5 kb of DNA ( 2.4- 51.5% of the chromosome). They may also involve almost any region of the chromosome, with two notable ex­ceptions: no deletions have been found that in­volve either of the origins of replication of the chromosome, the sites at which the copying of the mtDNA is initiated. Presumably, any such muta­tion would be selected against and lost almost im­mediately. Interestingly, about one third of dele­tions detected involve exactly the same 5- kb re­gion, stretching from the ATPase8 gene to ND5, so eliminating some genes of complexes I, IV, and V, plus some transfer RNAs ( Fig. 3). The DNA flank­ing this region has a distinctive repetitive sequence that seems to be responsible for its disproportion­ate vulnerability ( 14- 17). Duplications of segments of the mitochondrial genome have also been found: these produce chro­mosomes that are larger than normal, as a result of a second copy of a region having been inserted. Two cases with duplications of up to 7 kb have been reported ( 18) and in some malignancies, a large number of dimeric or larger mitochondrial chromosomes have been observed ( 19). The dele­terious effects are probably due not to the presence of two copies of some genes, but to disruption of "',,/ 111, No, J, 1990 the gene into which the duplication has been in­serted. Deletions and duplications have so far been found to be invariably heteroplasmic. Although the relative proportions of normal and abnormal mitochondria vary from tissue to tissue, some of the former are always present; evidently the cell requires some normal mitochondria to produce sufficient ATP to survive. Point mutations often have much less severe consequences than deletions. They involve the substitution of a single base pair of the chromo­some, generally leading to a structural change and so a functional change in a single protein. To date, Leber's hereditary optic neuropathy ( LHON) is the only disease for which a causative mitochondrial point mutation has been identified. Mitochondrial mutations have been found in a range of human diseases, two classes of which are of particular interest to the neuro- ophthalmologist ( Fig. 4). MITOCHONDRIAL MUTATIONS IN MITOCHONDRIAL CYTOPATHIES The term " mitochondrial cytopathy" covers a range of disorders, sometimes seen as clinically distinct. The diseases included are not clearly de­fined, with very variable manifestations and con­siderable overlap of features. A detailed clinical discussion is beyond the scope of this review, but the subject has been comprehensively covered elsewhere ( 20- 24). Essentially, the hypothesis is that conditions such as mitochondrial myopathy, chronic progressive external ophthalmoplegia ( CPEO), mitochondrial encephalomyopathy, and syndromes such as that of Keams- Sayre, MELAS ( mitochondrial myopathy, encephalopathy, lactic acidosis, and stroke- like episodes), and MERRF ( mitochondrial epilepsy with ragged red fibers) are different expressions of one etiological entity. In all these disorders there are structurally and biochem- Disease Type of mutation Comments Leber's hereditary Point mutation Homoplasmic optic neuropathy (+ heteroplasmic?) Mitochondrial Deletions + Heteroplasmic cytopathies duplications Pearson's syndrome Deletions Reference 35 Malignancy Dimerization Reference 19 FIG. 4. Diseases associated with characterized muta­tions of mitochondrial DNA. MITOCHONDRIAL MUTATIONS IN DISEASE 163 ically defective mitochondria and some forms of neuromuscular impairment. Beyond these fea­tures, as Petty et al. observed ( 22), a variety of defects have been identified, but there is no con­sistent association with syndromes. Mitochondrial cytopathies are caused by severe mitochondrial mutations. The concept of a com­mon etiology has been strongly supported by mtDNA analysis; starting in February 1988, with the article of Holt and colleagues ( 25), many groups have reported deletions or duplications of the mitochondrial chromosome in mitochondrial cytopathies. Approximately 90 cases of mitochon­drial cytopathy with mitochondrial deletions have been reported, plus two with duplications ( 14­16,24,26- 34). Remarkably, there is little correlation between the size of a deletion or the genes affected by it and the clinical symptoms. For example, the smallest deletion yet described was found in a patient di­agnosed as having encephalomyopathy and MERRF, with myoclonic epilepsy, progressive ataxia, deafness, and stroke- like episodes; DNA analysis showed that 400 base pairs ( 0.4 kb) of a single gene ( NOS) were deleted ( 30). Yet many pa­tients with CPEO have a deletion more than 10 times as large, the common 5- kb deletion ( 15), and a further patient with the very same 5- kb deletion did not have CPEO but Pearson's syndrome ( Mc­Kusick catalogue 26056), which is not an ophthal­mological disease ( 35). The citing of heteroplasmy to explain these extraordinary symptomatic varia­tions seems valid. In this last patient, for example, 80- 90% of mitochondria in the affected tissues ( bone marrow, leukocytes, gut) carried the dele­tion, compared with only 50% in the healthy mus­cle ( 13). Furthermore, it is generally found that the DNA of leukocyte mitochondria from patients with mitochondrial cytopathies is predominantly normal. The mutual association of mitochondrial cyto­pathies and mitochondrial DNA deletions is not absolute. Two cases of DNA duplication have al­ready been mentioned ( the diagnoses for the two patients were Keams- Sayre syndrome and mito­chondrial myopathy), and there are several further reports indicating the involvement of other mech­anisms. Ozawa and colleagues ( 31) analyzed the mtDNA of a mother and daughter, both with CPEO. Both patients had deletions of mtDNA ( in skeletal mus­cle) but, very surprisingly, the mother had a 2.5- kb deletion and the daughter a 5- kb deletion, with 1.2 kb of deletion in common. The nature of the basic defect here is unknown but could perhaps involve a nuclear gene with some role in controlling mtDNA replication. The MERRF family described by Wallace and colleagues ( 12) has no detectable loss of mtDNA in muscle or in lymphoblasts but presents all the characteristics expected of a severe mitochondrial mutation, including heteroplasmy and maternal transmission: Wallace et al. suggest that either a very small deletion or a point mutation must be present. The fact that both complexes I and IV are defective implies that the mutation is either at a point where complex I and IV genes are in prox­imity ( for example, between C03 and ND3, Fig. 3) or that a ribosomal or transfer RNA, involved in the synthesis of several mitochondrial proteins, is affected. In a detailed genetic study, Harding and col­leagues ( 36) found that 18% of 71 patients had rel­atives " definitely affected with a similar disorder." Including analysis of other published reports, they observed that about 90% of familial mitochondrial cytopathies were maternally transmitted and that only about half the offspring of affected females were themselves affected. Familial mitochondrial cytopathies were therefore significantly less fre­quent than isolated cases but were almost all com­patible with mitochondrial transmission by a het­eroplasmic mother. However, 10% of familial cases were paternally transmitted, implying mutations of nuclear chromosomes. As so many mitochon­drial proteins are encoded in the nucleus, includ­ing most components of the respiratory chain, it would not be surprising if some mitochondrial cy­topathies were caused by nuclear mutations and transmitted in Mendelian fashion; the biochemical defects could be essentially identical to those caused by mitochondrial mutations, although one might expect less symptomatic variation ( between tissues and between individuals), for variable het­eroplasmy would not occur. MITOCHONDRIAL MUTATIONS IN LHON Leber's hereditary optic neuropathy has long been considered the " classic" maternally transmit­ted disease. However, its etiology has also been the cause for much debate. It was thought that not all offspring of carrier females inherited the disease ( 37), and indeed this has not been formally dis­proved, but the long- term follow- up observations by Seedorff of families first studied by Lundsgard showed that, at least in some cases, the transmis­sion rate could reach 100% ( 38). Transplacental in­fection was also considered as a possible cause ( 39). Furthermore, there has been a long- standing J Clin Neuro- ophthalmol. Vol. 10. No. 3. 1990 164 M. A. MORRIS debate about the role of defects of cyanide metab­olism ( notably of the enzyme rhodanese) in LHON ( 40- 45). The discovery by Wallace and colleagues in 1988 of a point mutation of mtDNA causing LHON in nine of 11 independent pedigrees and its subse­quent confirmation seemed to herald the end of the confusion ( 46- 48). The mutation of a single base, at position 11778 of the mitochondrial chro­mosome, in the gene ND4, leads to the alteration of one amino acid of a subunit of respiratory chain complex I, substituting a histidine for an arginine. This apparently minor change produces a detect­able deficiency of complex I function ( 49) and, ul­timately, LHON. However, much remains unexplained about the etiology of LHON. Only a proportion of LHON families have this mutation; combining published cases and those of seven families analyzed at this Institute indicates that only 29 of 47 families have the mutation at position 11778 ( 62%) ( 46- 48,50- 52), implying that there is at least one other mtDNA mutation. Furthermore, there is convincing evi­dence in one family of LHON caused by a nuclear gene ( the disease was transmitted by a male), with the primary defect probably being of the enzyme rhodanese ( 42). It has been shown in the author's laboratory that the affected members of this family do not have the 11778 mutation. The considerable problem of the relative fre­quencies of affected males and females is still out­standing: approximately six of seven sufferers are male ( 37,53,54). One explanation for this is the sec­ondary involvement of a gene on the X chromo­some, modifying the effect of " the Leber's gene." However, there is no direct evidence for such a gene, and a preliminary genetic linkage analysis of the Xchromosome in three LHON families in Aus­tralia has made its existence unlikely ( 55). No other mechanism for this difference is apparent; it is con­ceivable that there is simply a different tolerance to the complex I deficiency in the male and female retina. Most groups have found the LHON mutation to be homoplasmic, but Holt et al. ( 52) have recently reported that almost all of their patients with the 11778 mutation are heteroplasmic, with a propor­tion of normal mtDNA, and that the relative pro­portions of mutant and normal DNA correlate with the risk of clinical LHON development. If this is confirmed in other families it may simplify expla­nations of the variability of expression of LHON. Wallace et al. have also suggested that such a het­eroplasmic point mutation may be involved, in , \',,/ 10, Nul, 1490 families with LHON plus infantile bilateral striatal necrosis ( 46). THE FUTURE In conclusion, it is evident that mtDNA muta­tions are responsible for a number of neuro­ophthalmological disorders, but what practical help can the geneticist offer the clinician now, and what does the future hold? One option available now is diagnosis. In the author's laboratory, detection of the LHON muta­tion 11778 has been offered as a diagnostic aid for approximately 1 year; while the test is not infalli­ble, for only about 60% of familial cases involve this mutation, a positive result has an obvious value. The test requests only a small blood sample. For the mitochondrial cytopathies, DNA analysis of muscular biopsy samples may become a routine diagnostic technique, and when more cases have been described, it may be that a knowledge of the degree of heteroplasmy will be a useful prognostic guide. It should be mentioned that an accurate family history can be a valuable diagnostic aid. Some groups are using a more detailed knowl­edge of the defect to improve the specificity of therapy. Wallace relates three such examples: the use of coenzyme Q in Keams- Sayre syndrome, ascorbate and menadione in a complex III defi­ciency, and riboflavin in a complex I deficiency ( 56) and Shoffner et al. ( 16) noted a marked improve­ment in a patient with a complex I defect treated with coenzyme QI0 and succinate ( which enters the chain at complex II). Cagianut and colleagues ( 42) emphasized the importance of diet in families with defects of the respiratory chain and included a useful list of con­traindicated foodstuffs that contain cyanide or in­hibit its metabolism. Although the article refers specifically to defects of cyanide metabolism, the avoidance of certain foods may have practical value in other cases. The LHON patients at the author's Institute are so advised, and they are also strongly advised against smoking, which has often been cited as being involved in the onset and se­verity of LHON. The future? Some of the outstanding puzzles have been discussed earlier. For example, it is clearly necessary to determine the remaining mu­tations in familial LHON. It is to be hopes that more frequent analysis of mtDNA in mitochon­drial cytopathies will reveal associations between the nature of the mutation and the degree of het­eroplasmy and the prognosis for the patient. Fi- MITOCHONDRIAL MUTATIONS IN DISEASE 165 nally, the precise identification of defects will be sure to lead to more effective and more specific metabolic therapies. In the very long term, an un­derstanding of the processes ( which may be ran­dom, selective, or both) leading to variable hetero­plasmy may permit their manipulation and so pro­vide the hope of a genuine cure. Acknowledgment: I am indebted to Professor E. Engel and Dr. C. D. Delozier- Blanchet for their invaluable en­couragement and numerous discussions, and to Profes­sor A. Safran, Dr. T. Stautz, Dr. B. Cagianut, and N. Malik for many interesting discussions and for access to DNA samples of patients. REFERENCES 1. Cavalier- Smith T. Eukaryotes with no mitochondria [ News and Views). Nature 1987; 326: 332- 3. 2. Lehninger AL. The mitochondrion. New York: WA Benjamin, 1964: 7. 3. Stryer L. Biochemistry. 2nd ed. San Francisco: WH Freeman and Co, 1981: 241. 4. Tzagoloff A, Myers AM. Genetics of mitochondrial biogen­esis. Ann Rev Biochem 1986; 55: 249- 85. 5. Giles RE, Blanc H, Cann HM, Wallace DC. Maternal inher­itance of human mitochondrial DNA. Proc Natl Acad Sci USA 1980; 77: 671~ 9. 6. Brown WM, George M, Wilson AC. Rapid evolution of an­imal mitochondrial DNA. Proc Natl Acad Sci USA 1979; 76: 1967- 71. 7. Miyata T, Hayashida H, Kikuno R, et aJ. Molecular clock of silent substitution: at least six- fold preponderance of silent changes in mitochondrial genes over those in nuclear genes. ! Mol EvoI1982; 19: 28- 35. 8. Neckelmann N, Li K, Wade RP, Shuster R, Wallace DC. eDNA sequence of a human skeletal muscle ADP/ ATP translocator: lack of a leader peptide, divergence from a fibroblast translocator eDNA, and coevolution with mito­chondrial DNA genes. Proc Nat! Acad Sci USA 1987; 84: 7580- 4. 9. Linnane AW, Marzuki S, Ozawa T, Tanaka M. Mitochon­drial DNA mutations as an important contributor to aging and degenerative diseases. Lancet 1989; 1: 642- 5. 10. Linnane AW, Haslam JM, Lukins HB, Nagley P. The bio­genesis of mitochondria in microorganisms. Ann Rev Micro­bioi 1972; 26: 163- 98. 11. Ashley MY, Laipis PJ, Hauswirth WW. Rapid segregation of heteroplasmic bovine mitochondria. Nucleic Acids Res 1989; 17: 73~ 31. 12. Wallace DC, Zheng X, Lott Mr, et al. Familial Mitochon­drial Encephalomyopathy ( MERRF): genetic, pathophysio­logical, and biochemical characterization of a mitochondrial DNA disease. Cell 1988; 55: 601- 10. 13. R6tig A, Colonna M, Cormier V, et aJ. Intramolecular re­combinations of the human mitochondrial genome in Pear­son's syndrome. Cytogenet Cell Genet 1989; 51: 1069- 70. . 14. Moraes CT, DiMauro S, Zeviani M, et aJ. Mitochondnal DNA deletions in progressive external ophthalmoplegia and Keams- Sayre syndrome. N Engl! Med 1989; 320: 1293- 9. 15. Schon EA, Rizzuto R, Moraes CT, et aJ. A direct repeat is a hotspot for large- scale deletion of human mitochondrial DNA. Science 1989; 244: 346- 9. 16. Shoffner JM, Lott Mr, Voljavec AS, et aJ. Spontaneous Keams- Sayre/ chronic external ophthalmoplegia plu~ syn­drome associated with a mitochondrial DNA deletion: a slip- replication model and metabolic therapy. Proc Natl Acad Sci USA 1989; 86: 7952-- 6. 17. Wallace DC, Lott Mr, Voljavec AS, Shoffner JM. Mitochon­drial DNA deletions in Keams- Sayre syndrome can be as­sociated with direct repeats. Cytogenet Cell Genet 1989; 51: 1101- 2. 18. Poulton J, Deadman ME, Gardiner RM. Tandem direct du­plications of mitochondrial DNA in mitochondrial myopa­thy: analysis of nucleotide sequence and tissue distribu­tion. Nucleic Acids Res 1989; 17: 10223- 9. 19. Robberson DL, Gay ML, Wilkins CEo Genetically altered human mitochondrial DNA and a cytoplasmic view of ma­lignant transformation. In: Arrighi FE, Rao PN, Stubble­field E, eds. Genes, chromosomes, and neoplasia. New York: Raven Press, 1981: 12~ 56. 20. Egger I, Lake BD, Wilson J. Mitochondrial cytopathy. A multisystem disorder with ragged red fibres on muscle bi­opsy. Arch Dis Child 1981; 56: 741- 52. 21. Egger I, Wilson J. Mitochondrial inheritance in a mitochon­drially mediated disease. N Engl! Med 1983; 309: 142-- 6. 22. Petty RKH, Harding AE, Morgan- Hughes JA. The clinical features of mitochondrial myopathy. Brain 1986; 109: 91~ 8. 23. R6tig A, Bonnefont JP, Colonna M, et al. Les remaniements du genome mitochondrial dans les deficits energetiques de l'enfant: de nouvelles maladies de systeme? Med Sci 1989; 5: 459- 71. 24. Rutledge SL, Johns DR, Hurko O. Mitochondrial DNA analysis in a mitochondrial cytopathy overlap syndrome. Am! Hum Genet 1989; 45: A216. 25. Holt II, Harding AE, Morgan- Hughes JA. Deletions of mus­cle mitochondrial DNA in patients with mitochondrial my­opathies. Nature 1988; 331: 717- 9. 26. Harding AE, Holt 11, Morgan- Hughes JA. Deletions of mus­cle mitochondrial DNA in patients with mitochondrial my­opathies. Am! Hum Genet 1988; 43: A86. 27. Johns DR, Drachman DB, Hurko O. Identical mitochondrial DNA deletion in blood and muscle. Lancet 1989; 1: 393- 4. 28. Johns DR, Hurko O. Preferential amplification and molec­ular characterization of junction sequences of a pathoge­netic deletion in human mitochondrial DNA. Genomics 1989; 5: 623- 8. 29. Lestienne P, Ponsot G. Keams- Sayre syndrome with mus­cle mitochondrial DNA deletion. Lancet 1988; 1: 885. 30. Saifuddin- Noer A, Marzuki S, Trounce I, Byrne E. Mito­chondrial DNA deletion in mitochondrial encephalomyop­athy. Lancet 1988; 2: 1253- 4. 31. Ozawa T, Yoneda M, Tanaka M, et aJ. Maternal inheritance of deleted mitochondrial DNA in a family with mitochon­drial myopathy. Biochem Biophys Res Commun 1988; 154: 1240- 7. 32. Poulton J, Deadman ME, Gardiner RM. Duplications of mi­tochondrial DNA in mitochondrial myopathy. Lancet 1989; 1: 236- 40. 33. Zeviani M, Moraes CT, DiMauro S, et al. Deletions of mi­tochondrial DNA in Kearns- Sayre syndrome. Neurology 1988; 38: 1339- 46. 34. Zeviani M, Moraes CT, Shanske S, et al. Deletion of mito­chondrial DNA in Keams- Sayre syndrome. Am! Hum Genet 1988; 43: A100. 35. R6tig A, Colonna M, Bonnefont JP, et al. Mitochondrial DNA deletion in Pearson's marrow/ pancreas syndrome. Lancet 1989; 1: 902- 3. 36. Harding AE, Petty RKH, Morgan- Hughes JA. Mitochon­drial myopathy: a genetic study of 71 cases. ! Med Genet 1988; 25: 528- 35. 37. Van Senus AHC. Leber's disease in the Netherlands. Doc OphthalmoI1963; 17: 1- 162. 38. Seedorff T. The inheritance of Leber's disease: a genealog­ical follow- up study. Acta Ophthalmol ( Copenh) 1985; 63: 1~. 39. Erickson RP. Leber's optic atrophy, a possible example of maternal inheritance. Am! Hum Genet 1972; 24: 348- 9. 40. Wilson J. Leber's hereditary optic atrophy: a possible defect of cyanide metabolism. Clin Sci 1965; 29: 50~ 15. 41. Cagianut B, Rhyner K, Furrer W, Schnebli HP. Thiosul- JClin Neuro- ophthalmol, Vol. 10, No. 3. 1990 166 M. A. MORRIS phate- sulphur transferase ( rhodanese) deficiency in Leber's hereditary optic atrophy. Lancet 1981; 2: 981- 2. 42. Cagianut B, Schnebli HP, Rhyner K, Furrer J. Decreased thiosulfate sulfur transferase ( rhodanese) in Leber's hered­itary optic atrophy. Klin Wochenschr 1984; 62: 85(}.. 4. 43. Poole CjM, Kind PRN. Deficiency of thiosulphate sul­phurtransferase ( rhodanese) in Leber's hereditary optic neuropathy. Br Med J1986; 292: 1229- 30. 44. Berninger TA, Meyer LV, Siess E, et al. Leber's hereditary optic atrophy: further evidence for a defect of cyanide me­tabolism? Br JOphthalmol 1989; 73: 314- 6. 45. Whitehouse DB, Poole CjM, Kind PRN, Hopkinson DA. Rhodanese isozymes in three subjects with Leber's optic neuropathy. JMed Genet 1989; 26: 113-- 5. 46. Wallace DC, Singh G, Lott MI, et al. Mitochondrial DNA mutation associated with Leber's hereditary optic neurop­athy. Science 1988; 242: 1427- 30. 47. Yoneda M, Tsuji S, Yamauchi T, et al. Mitochondrial DNA mutation in family with Leber's hereditary optic neuropa­thy. Lancet 1989; 1: 107~ 7. 48. Singh G, Lott MI, Wallace DC. A mitochondrial DNA mu­tation as a cause of Leber's hereditary optic neuropathy. N Engl JMed 1989; 320: 1300- 5. 49. Parker WD, Oley CA, Parks JK. A defect in mitochondrial " · . f 10, No. 3, 19911 electron- transport activity ( NADH- coenzyme Q oxi­doreductase) in Leber's hereditary optic neuropathy. N Engl JMed 1989; 320: 1331- 3. 50. Vilkki J, Savontaus M- L, Nikoskelainen EK. Genetic heter­ogeneity in Leber hereditary optic neuroretinopathy re­vealed by mitochondrial DNA polymorphism. Am J Hum Genet 1989; 45: 20~ 11. 51. Hotta Y, Hayakawa M, Saito K, et al. Diagnosis of Leber's optic neuropathy by means of polymerase chain reaction amplification. Am JOphthalmoI1989; 108: 601- 2. 52. Holt IJ, Miller DH, Harding AE. Genetic heterogeneity and mitochondrial DNA heteroplasmy in Leber's hereditary op­tic neuropathy. JMed Genet 1989; 26: 739- 43. 53. Waardenburg PJ. Some remarks on the clinical and genetic puzzle of Leber's optic neuritis. JGenet Hum 1969; 17: 479­96. 54. Nikoskelainen E. New aspects of the genetic, etiologic, and clinical puzzle of Leber's disease. Neurology 1984; 34: 1482- 4. 55. Chen JD, Cox I, Denton MJ. Preliminary exclusion of an X- linked gene in Leber optic atrophy by linkage analysis. Hum Genet 1989; 82: 203- 7. 56. Wallace DC. Mitochondrial DNA mutations and neuromus­cular disease. Trends Genet 1989; 5: 9- 13.