Mutation survey of the optic atrophy 1 gene in 193 Chinese families with suspected hereditary optic neuropathy

• Introduction
• Methods
• Results
• Discussion

Abstract

Purpose: Dominant optic atrophy (DOA) is the most common form of autosomal inherited optic neuropathy, mainly caused by mutations in the optic atrophy 1 (OPA1) gene. The purpose of this study was to detect OPA1 gene mutations and associated phenotypes in Chinese patients with suspected hereditary optic neuropathy.

Methods: A cohort of 193 Chinese families with suspected hereditary optic neuropathy was collected, which had been excluded from the three common primary mitochondrial DNA mutations associated with Leber hereditary optic neuropathy in our prior screening. Sanger sequencing was used to analyze variants in the coding and adjacent regions of OPA1.

Results: In this study, 11 heterozygous OPA1 mutations, among which eight were novel and three were known, were identified in 12 of the 193 families (6.2%) but in none of the 192 control individuals. These novel mutations consisted of two nonsense mutations (p.E707* and p.K797*), two missense mutations (p.T330S and p.V377I), two deletions (p.S64fs and p.L331fs), one small insertion (p.L17fs), and one splice site mutation (c.2614–2A>G). Of the 12 families, three had a family history of optic neuropathy while nine were sporadic cases. Analysis of the family members in the two sporadic cases demonstrated that one parent in each of the two families had the OPA1 mutation and mild phenotype of optic atrophy. A 4-year-old boy with severe ocular phenotype was found to be compound heterozygous for two OPA1 mutations, a p.S64fs frameshift deletion and a p.V377I missense mutation, possibly implying an additive effect.

Conclusions: This study implies that the frequency of DOA is much lower than that of Leber hereditary optic neuropathy in Chinese compared with other ethnic groups. Lack of awareness of the mild phenotype of DOA may contribute to the low frequency of OPA1-related DOA in Chinese. The phenotype associated with compound heterozygous OPA1 mutations may suggest a possible addictive effect.

Optic neuropathy, a common visual impairment, can be caused by genetic defects, toxic factors, and ocular or brain diseases [1-3]. Genetic defects may occur in nuclear genome or mitochondrial DNA (mtDNA) [4-6]. Mutations in mtDNA are known to cause Leber hereditary optic neuropathy (LHON, MIM 535000) [7]. At least eight loci (OPA1 to OPA8) have been mapped to the nuclear genome in association with optic atrophy, in which three causative genes have been identified, including the optic atrophy 1 (OPA1, OMIM 605290) gene, the optic atrophy 3 (OPA3, OMIM 606580) gene, and the transmembrane protein 126A (TMEM126A, OMIM 612988) gene [1].

Dominant optic atrophy (DOA, MIM 165500) is the most common form of hereditary optic neuropathy resulting from mutations in nuclear genome, with a prevalence of 1 in 50,000 overall [8] and as high as 1 in 35,000 in the north of England [9], similar to the frequency of LHON (1 in 25,000 to 1 in 50,000) as reported in Caucasians [1,10]. DOA typically presents as painless and slowly progressive visual loss, occurring insidiously with a mean age of onset at 6–10 years of age [1,11]. DOA exhibits marked inter- and intrafamilial variable phenotypes [9,12,13]; for instance, about 25% of patients were found to have optic atrophy with fundus examination but without a complaint of visual deterioration [11,14].

Of the three nuclear genes known to cause optic neuropathy when mutated, mutations in OPA3 and TMEM126A are extremely rare. OPA3 mutations result in optic atrophy associated with cataract [15,16], while TMEM126A mutations are responsible for an autosomal recessive form of optic atrophy [17,18]. Mutations in OPA1 account for the majority (60%–70%) of patients with DOA [19]. Thus far, at least 230 pathogenic mutations have been identified in OPA1 [20]. However, extremely variable phenotypes of DOA hindered the diagnosis and, therefore, the clinical test of OPA1 mutational screening. Until now, the genotype-phenotype correlations of OPA1 mutations with DOA were elusive [21,22]. Recently, analysis of OPA1 in some large case series of hereditary optic neuropathy in Caucasians suggested that scanning the most frequently mutated exons might be a good strategy for identifying OPA1 mutations [21,23]. Although exon 27, exon 8, and exon 15 may be mutational hot spots in Caucasians [21], the frequency of OPA1 mutation in Chinese patients has not been evaluated. Only a few Chinese families with DOA have been reported to be associated with OPA1 mutations [24-27], not to mention the mutational hot spots. In this study, Sanger sequencing was used to detect the mutation of OPA1 in 193 Chinese families with suspected hereditary optic neuropathy, in which the three common LHON-associated primary mutations of mtDNA had been excluded with prior screening.

DOA and LHON are the most common hereditary optic neuropathies in the general population [4,33,34], with similar prevalence in Caucasians [1,9]. Ferre et al. reported identification of genetic defects in 440 of 980 cases (45%) with suspected hereditary optic neuropathy, including OPA1 mutations in 295 (30%) patients, mtDNA mutations in 131 (13%) patients, and OPA3 mutations in 14 (1.4%) patients [21]. These results suggest that OPA1 mutations are the most common cause for patients with suspected hereditary optic atrophy. However, there is no molecular epidemiological study for DOA in Chinese populations even though there are large case series of such studies for Chinese patients with LHON [28]. It is unclear whether DOA is rare in Chinese or may not be easily recognized in clinics compared to LHON. Clinically, it may be difficult to distinguish DOA from LHON in many cases, especially in the atrophic phase [35,36]. In our previous report of a molecular epidemiological analysis of 903 families with optic neuropathy, the three primary mutations (G11778A, T14484C, and G3460A) in mtDNA were identified in 346 families (38.3%) [28]. In this study, mutations in OPA1 were detected in only 12 of the 193 families (6.2%). Since patients with one of the three primary mtDNA mutations (G11778A, T14484C, and G3460A) were not included in the cohort, the actual detection rate of OPA1 mutations in Chinese patients with hereditary optic neuropathy should be even lower, indicating a significantly lower frequency of OPA1 mutations (<6.2%) and a comparatively higher frequency of mtDNA mutations (38%) in Chinese patients with suspected hereditary optic neuropathy compared with French Caucasians (30% for OPA1 and 13% for mtDNA) [21]. Lack of awareness of the mild phenotype of DOA may contribute to the low frequency of OPA1-related DOA in Chinese and relatively high frequency of OPA1 mutations in sporadic cases, which is implied by the segregation analysis of mutations in the family of le1432 and le2062. A family history showing autosomal dominant inheritance is a key indicator leading to clinical diagnosis of DOA and genetic analysis of OPA1. For the probands from the 12 families with OPA1 mutations identified in this study, however, most (9/12, 75%) were sporadic cases (nine out of 155 sporadic cases). Similar frequency of OPA1 mutations in singleton cases was also mentioned in other studies [16,21,24,37]. Since individuals who harbor OPA1 mutations may have mild phenotypes that the individuals are unaware of, it is important to keep OPA1 in mind for patients with suspected hereditary optic neuropathy without a family history and evaluate the family members of these singleton cases carefully. Moreover, such evaluation and genetic consultation to family members may be crucial since more severe phenotypes might be induced by toxic materials, such as alcohol, smoking, and some drugs, as seen in two patients who had taken medicine for tuberculosis in this study.

Although OPA1 mutations contribute to the most (60%–70%) cases with DOA [19], the genetic diagnosis of DOA is still a challenge. OPA1 mutations are spread over 30 exons [21]. In addition, about a quarter of the mutations are located in the intronic regions adjacent to exon-intron boundaries, which usually affect splicing [20]. Under this circumstance, a cost-effective method for identifying mutations responsible for DOA is of great value. Screening the most frequently mutated exons of OPA1 initially would be a feasible strategy. Although exons 27, 8, and 15 were suggested to be regions of mutation hot spots in a previous study by Ferre et al. [21], we are unable to suggest ethnic-specific mutation hot spots based on limited number of mutations identified in our cohort of Chinese patients. The c.2708_2711del mutation in exon 27 was present in approximately 27% of families in previous studies. The same mutation was found in three probands in our study, which accounts for 25% among the OPA1-positive cases, further confirming that it is a mutational hot spot [21,24,37]. Interestingly, three novel mutations, discovered in three unrelated patients, arose from a region less than 10 bp around the 5′ end of exon 10, which may imply another mutational hot spot. Thus far, this is the largest series of patients screened for OPA1 in an Asian population. As our results provide evidence of ethnic variations in the mutation frequency of OPA1, analyzing causative genes and their exons in a systematic and population-specific fashion is necessary, especially in the context of genetic counseling for different ethnic groups.

To our knowledge, the patient with confirmed compound heterozygosity of OPA1 mutations in our study is the fourth case reported so far [38,39], which represents a rare event. All cases with compound heterozygous OPA1 mutations reported thus far had more severe phenotypes than their parents with a single heterozygous OPA1 mutation, which may suggest an additive effect [39]. Although the mother is a possible case of non-penetrance in optic atrophy, the c.1129G>A (p.V377I) mutation is still likely to be pathogenic. First, the mutation is rare as it is not found in 384 control chromosomes or the Single Nucleotide Polymorphism database. Second, the mutation is located in an evolutionarily conserved region and is predicted to be probably damaging by PolyPhen-2. Third, the severity of the phenotype of the son could, at least partly, reflect the addictive effect of the mutation from his mother. However, it is not the only interpretation of this compound heterozygous case. A c.190_194del (p.S64fs) mutation was identified in the father, presumably causing premature termination. However, the phenotype of the father was much milder than the other probands with frameshift deletions or nonsense mutations in our study. This may suggest the existence of modifier alleles at other loci. If the son did not inherit the beneficial modifier alleles from his father, this could also result in the severity of the son’s phenotype. The effect of modifier alleles may also explain the marked intrafamilial phenotypical heterozygosity between the probands and other mutation carriers, as shown in le1432 and le2062.

In summary, this study implies that the frequency of DOA is much lower than that of LHON in Chinese compared with other ethnic groups. Lack of awareness of the mild phenotype of DOA may contribute to the low frequency of OPA1-related DOA in Chinese. Further analysis of OPA1 in individuals with mild visual impairment and temporal disc pallor may be helpful in disclosing the real frequency of DOA in Chinese. Routine clinical test of OPA1 variations in such cases may also enhance the care of eyes through consultation and pretreatment, especially for individuals who are unaware of visual problems but harbor OPA1 mutations.