Cytogenetic and Genome Research · 1999 · 12 citations
Leber hereditary optic neuropathy (LHON): a mitochondrial disease with unresolved complexities
AbstractLeber hereditary optic neuropathy, which seems to be the most common inherited eye disease, has, until recently, baffled investigators. Notwithstanding the finding of a mitochondrial DNA (mtDNA) mutation at nucleotide position (np) 11,778 by Wallace et al. (1988), there are still many questions to be answered. These include the relationship with diminished penetrance, the uneven sex distribution, the influence and nature of environmental factors, the possibility of additional (X-linked?) genes, the importance of heteroplasmy, and the significance of the mitochondrial haplotype background.The disease was named after Theodor Leber (Professor of Ophthalmology in Tübingen), who was the first to describe it fully and correctly as an entity in 1871. It was originally published under the name of Leber(s) disease, or Leber’s optic atrophy (LOA), but has since been renamed Leber hereditary optic neuropathy (LHON), which is now universally accepted. Exhaustive clinical and epidemiological descriptions of LHON can be found in three monographs: (1) Ruth Lundsgaard (1944) described all patients with LHON that she could find in Denmark (101, belonging to 20 families). (2) Asseman (1958) reviewed in detail the findings on 86 patients from 17 pedigrees from the region around Lille in northern France. (3) Van Senus (1963) provided the most comprehensive published analysis, involving 352 patients (202 of whom he saw himself) from 27 pedigrees in the Netherlands.Very careful studies have since been undertaken in Denmark (Seedorff, 1985), Finland (Nikoskelainen et al., 1987, 1995; Lamminen et al., 1997), England (Harding et al., 1995; Riordan-Eva and Harding, 1995; Chalmers et al., 1996), the USA (Novotny et al., 1986), Australia (Wallace, 1970), the Netherlands (Stehouwer and Went, 1982; Bruyn et al., 1991, 1992; De Vries et al., 1996; Oostra, 1996; Oostra et al., 1996), and Japan (Imachi, 1969; Nakamura et al., 1992).Howell (1997) does not believe that extra-ophthalmological abnormalities form an essential part of LHON. Mackey et al. (1996), in their multicenter analysis of mitochondrial DNA (mtDNA) in LHON pedigrees also exclude the “LHON-plus” pedigrees. On the other hand, De Weerdt and Went (1971), Novotny et al. (1986), Palan et al. (1989), and, most notably, the Finnish group (see Nikoskelainen et al., 1995), have presented evidence that extra-ophthalmological, particularly neurological and ECG abnormalities, belong to the spectrum of manifestations of LHON, even in the absence of optic atrophy. De Weerdt and Went (1971) presented nine patients (six from Van Senus’ 1963 pedigrees and three from other publications) with a multiple sclerosis (MS)-like picture from families with classical LHON, eight of whom were female. Riordan-Eva and Harding (1995) found that among 11 women with LHON carrying the 11,778 mtDNA mutation, 5 had an MS-like illness. Nikoskelainen et al. (1995) also believed that an MS-like picture could be the result of an mtDNA LHON mutation, but not that it is specifically restricted to females.It is known that over 90% of LHON patients carry one of the three primary mtDNA mutations (Riordan-Eva and Harding, 1995; Mackey et al., 1996). We still need to explain why only 50% of males and 5–10% of females in the matrilineal relation are affected by optic atrophy. This approximately 6:1 male-to-female ratio seems to be different only in Japan, where Imachi (1969) found that 61% of 911 LHON patients in 188 pedigrees were male. Nakamura et al. (1992) disagreed with Imachi’s findings, but studied only a small number of patients and did not give arguments for their disagreement. It is interesting that Black et al. (1995), in analyzing pooled data from six countries, observed that the male excess in LHON pedigrees with the ND1/3460 mutation is significantly lower than that usually quoted for LHON.A further complicating factor is that the overall penetrance in Australian LHON families has decreased in recent generations to 20% in males and 4% in females (Howell, 1997); this trend also appears, although it is not specifically mentioned, in the large Dutch pedigree reproduced by Bruyn et al. (1992).Various authors, such as Van Senus (1963), have searched, without much success, for environmental factors determining the acute onset of the optic atrophy. Palan et al. (1989) analyzed Van Senus’s data on 293 patients for whom the year of onset was known. Pooling the data into five-year periods between 1915 and 1975, Palan et al. observed a very high peak between 1940 and 1944, followed by a dip in 1945–1949, suggesting some possible influence of malnutrition during the German occupation of the Netherlands in World War II.Riordan-Eva and Harding (1995) discussed the different reports involving the possible effect of tobacco or alcohol abuse (and secondary vitamin B deficiencies) on optic nerve disease, suggesting that the background might be an abnormality of cyanide metabolism caused by a deficiency of the enzyme rhodanese (thiosulfate-sulfur transferase). Conflicting as the studies were in this respect, Riordan-Eva and Harding found it prudent to advise subjects at risk of developing LHON to refrain from tobacco smoking and heavy alcohol consumption.Van Senus (1963) and Howell (1997) believe that head trauma or physical exertion may occasionally precipitate the onset of optic atrophy.There is conflicting evidence on whether an additional X-linked locus is involved in LHON (Harding et al., 1995; Howell, 1997). Harding et al. (1995) discussed unsuccessful attempts to confirm linkage to DXS7, but presented evidence that suggested it would be worthwhile exploring the X-linked hypothesis further. In a more recent article, the same UK group present evidence against X-linkage. It should be mentioned that in a large Dutch family with an X-linked optic atrophy, close linkage was found (a lod score without recombination of 4.19) at the MAOB locus Xp11.4, where DXS7 is also situated (Assink et al., 1997).Since Leber’s first publication, there has been widespread agreement that the disease is never transmitted through affected males. Notwithstanding the observation that usually no more than 50% of sons of females in affected pedigrees are affected themselves (Van Senus, 1963, and others), various authors have suggested cytoplasmic inheritance as a causative factor in the disease (Kitashima, 1930; Imai and Moriwaki, 1936; Rønne, 1944). This was supported by studies from Finland of mitochondrial area measurements (Nikoskelainen et al. 1984). Nikoskelainen et al. (1987) suggested that all sons and daughters of female carriers inherit the trait. And, more recently, Harding et al. (1995) showed that 140 unaffected matrilineal relatives from 37 LHON families all had the same mtDNA mutation as the index case.Wallace et al. (1988) were the first to describe an actual mutation (G→A) in the mtDNA at np 11,778 of the ND4 subunit of NADH-ubiquinone oxidoreductase (Complex I; see Fig. 1). Subsequently, LHON families were found with mutations at np 3,460 (subunit ND1) and np 14,484 (subunit ND6). In a multicenter publication from Australia, the UK, the Netherlands, Denmark, and Finland (Mackey et al. 1996), the results of a search for these three so-called primary mutations have been pooled. In a total of 159 multigeneration pedigrees, 110 were found to harbor the 11,778 mutation, 21 the 3,460, and 23 the 14,484. There was no evident difference in the distribution of these three mutations between the countries. In five families, none of the three mutations could be found.The authors discussed the possible role of “secondary” mutations, which are also found in control populations (albeit at a much lower frequency), in contradistinction to the three primary mutations, which are not found in controls. Mackey et al. (1996) specifically selected a “secondary” LHON mutation at np 15,257, which is situated in the cytochrome b gene. They concluded that the increased frequency of this mutation in LHON pedigrees can be explained solely in terms of population history and genetics. This conclusion is contradicted by Hofmann et al. (1997a), based on a study in Germany of a number of 55 LHON index cases. In this group there were 24 carriers of the 11,778 mutation (three times in association with 15,257), 5 carriers of 3,460, and 21 carriers of 14,484 (six times in association with 15,257), whereas in five instances the 15,257 mutation was present without any of the established primary mutations. Hofmann et al. (1997a) also discussed the influence of 24 other mutational polymorphic mtDNA sites—combined in haplogroups—on the chance of a “primary” LHON mutation occurring. This work is based on their experience reported in a more extensive publication (Hofmann et al. 1997b) and also supported by haplotype studies of mtDNA in LHON from Italy (Torroni et al. 1997). To show the complexity of the problem, I quote the last paragraph of Hofmann et al. (1997b): “We propose that, in contrast to the primary mutations, which are obviously LHON specific, the so-called secondary LHON mutations are not pathognomonic for LHON but seem to create a susceptibility basis for certain neurodegenerative disorders in a way that is, so far, unknown. The term secondary LHON mutations, therefore, should be dropped, at least with respect to 4216, 4917, and 13708.”A further complicating factor with the mtDNA studies is heteroplasmy, the presence in about 15% of patients of a mixture of normal and mutant mtDNA in the leukocytes of both the patients and their maternal relatives (discussed by Riordan-Eva and Harding, 1995, and Harding et al., 1995). Lamminen et al. (1997) believed that heteroplasmy might indicate a recent occurrence of the mtDNA mutation, arguing further that the 11,778 mutation seems to have arisen independently several times in the Finnish population. However, to quote Howell (1997), “it can be safely asserted that heteroplasmy of the primary mutation is not a major factor because there are large, well characterized LHON families in which this mutation is homoplasmic and penetrance is incomplete.” However, since only leukocytes are usually studied, there may be differences in the degree of heteroplasmy in the target tissue (e.g., optic nerve). One indication that such a difference exists comes from the findings of De Vries et al. (1996), who observed 12% wild-type mtDNA in the optic nerve of a deceased patient, whereas no wild-type mtDNA was present in his fibroblasts (see also page 83 of Harding et al., 1995).In the family studied by De Vries et al. (1996), the three primary LHON mutations were absent, but all persons studied carried two new mutations: a T→A transition of ND6/14596 and an A→G transition of ND4/11696; this last mutation was heteroplasmic in 8 of the 13 persons studied (between 10% and 23% wild-type mtDNA). Neither of these mutations was found in 100 control persons, and there was no clear-cut relationship between the presence of heteroplasmy and the optic atrophy (three of seven persons with optic atrophy were heteroplasmic, versus five of six without optic atrophy).The only therapeutic approach for LHON that has been tried is neurosurgical exploration of the chiasmal area, with loosening of the arachnoidal adhesions. In Japan, Imachi et al. (1964, 1969) have reported favorable results of craniotomy followed by weekly intrathecal injections of vitamins B1, B2, and B12. However, these results seem to be invalidated by the studies of Asseman (1958) in France and Van Senus (1963) in the Netherlands.The problems outlined above are intended to make clear that, as Harding et al. (1995) stated at the end of their article, “LHON, in many ways the prototype of mitochondrial inheritance, may well turn out to be more etiologically complex than other human diseases associated with defects of mtDNA.”This article is dedicated to the memory of Prof. Meera Khan.
https://doi.org/10.1159/000015370