Point mutations associated with Leber hereditary optic neuropathy in a Latvian population

Patients

Twelve patients, including patients in two families, F1 and F2 (Figure 1), were enrolled in this study. In the first family, the mother and son were affected, and in the second family, visual failure was observed in only one family member (LHON-4) and not in her offspring (LHON-6 and LHON-7). The current study was approved by the Ethics Committee and complied with the Declaration of Helsinki [12]; all participants provided written informed consent. Two samples (LHON-1 and LHON-2) were obtained in 2007, and other samples were obtained in 2011–2012. Information on the health state and family history of the studied participants was collected using health and hereditary questionnaires.

Ophthalmologic examination

The ophthalmologic examinations were performed in the Ophthalmology Clinic of Pauls Stradins Clinical University Hospital, Riga. The following tests were used: visual acuity and visual field examinations, fundus photography, and optical coherence tomography (OCT). Other possible causes of visual failure were excluded, such as metabolic diseases (e.g., diabetes mellitus), toxicity (alcohol), trauma, and inflammation.

Molecular analysis of the mitochondrial genome

Whole DNA was extracted from blood leukocytes using the standard phenol chloroform method [13]. A total of 72 point mutations in mtDNA that had been previously reported to be associated with LHON according to the MITOMAP database (last update in August 2012) were analyzed. The polymorphism phenotyping (PolyPhen) and Sorting Tolerant From Intolerant (SIFT) databases were used to evaluate the interspecies conservation of these mutations and the possible or probable damaging status of each mutation (PolyPhen, SIFT). This information was considered when the results were interpreted. Twenty-two mtDNA fragments were amplified using oligonucleotide primers as described previously [14-17] (Appendix 1). PCR was performed in a final volume of 12.5 µl. The master mix contained (per tube) 1.25 µl of 10x Taq Buffer containing (NH₄)₂SO₄ (Thermo Scientific, Vilnius, Lithuania), 1.25 µl of 25 mM MgCl₂ (Thermo Scientific), 0.125 µl of dNTPs (10 mM each; Thermo Scientific), 0.5 µl of 10 µM forward primer (Metabion, Martinsried, Germany), 0.5 µl of 10 µM reverse primer (Metabion), 0.25 µl of Taq DNA Polymerase (5 U/µl; Thermo Scientific), and 7.63 µl of nuclease-free water (Thermo Scientific). To each tube was added 1 µl of 20 ng/µl DNA. The conditions of the PCR were as follows: one cycle at 95 °C for 5 min; 30 cycles of 95 °C for 30 s, primer annealing for 30 s (temperatures are given in Appendix 1) and 72 °C for 30 s (for long fragments, 13 and 15, the elongation time was 2 min); and a final elongation at 72 °C for 5 min. Each fragment was evaluated using an agarose gel stained with 10 mg/ml ethidium bromide and then was purified using the ExoI and FastAP enzymes (Thermo Scientific). The fragments were sequenced in both directions using the forward and reverse primers and a 3130×l Genetic Analyzer (Applied Biosystems, Carlsbad, CA) according to the manufacturer’s protocol. For PCR fragments 13 and 15, long reads were used, and the capillary length was changed from 50 cm to 80 cm.

Analysis of mitochondrial DNA haplogroups

The mtDNA haplogroups of the participants were determined by sequencing hypervariable segment I (HVS-I) and with polymerase chain reaction-restriction fragment length polymorphism (PCR-RFLP) analysis [15].

Detection of mtDNA heteroplasmy using the SURVEYOR assay

Because sequencing is not sufficiently sensitive to detect heteroplasmy (the limit of detection is approximately 20% [18,19]), the SURVEYOR mismatch endonuclease assay was used. The limit of detection for heteroplasmy by this method is 3% [14]. The mismatch endonuclease is able to digest a DNA heteroduplex on the 3′-side of the mismatch site formed by heating and slow cooling if heteroplasmy is present in the studied mtDNA fragment. All PCR products were treated using the SURVEYOR mismatch endonuclease and analyzed with 4% polyacrylamide gel electrophoresis with staining with ethidium bromide. The SURVEYOR Mutation Detection Kit (Transgenomic, San Jose, CA) was used in this study. The reactions were performed according to the manufacturer's protocols and Bannwarth's method of detection of heteroplasmic mutations in a whole mtDNA using SURVEYOR mismatch endonuclease [14].

Analysis of individual cases

Second family: Affected mother (LHON-4) and two unaffected daughters (LHON-6 and LHON-7)

In this family (F2), only the mother (LHON-4) exhibited clinical features of LHON, which appeared at the age of 40 years. The rare LHON-associated mutation A13637G was found in the MT-ND5 gene of this patient. Then the two daughters (LHON-6 and LHON-7) of this patient were examined ophthalmologically to assess the state of their vision. At the examination, they were 24 and 32 years old, respectively. In addition, fragment number 14 of each daughter was sequenced to determine if there was a mutation at position 13,637. The same mutation was found in both daughters even though they were phenotypically healthy at the time of the study. In the PolyPhen database, the A13637G mutation is listed as benign, and in the SIFT database, it is listed as “predict tolerated.”

Patient LHON-5

This male patient, who was 53 years old, had progressive visual impairment; over a period of 1 month, his visual acuity decreased from 0.3 to 0.05 in his right eye and from 1.0 to 0.75 in his left eye. The ophthalmological examination revealed edema, atrophy of the optic nerve, and cecocentral scotoma only in the right eye. Sequencing of the MT-ND5 gene revealed the presence of a secondary LHON-associated point mutation, A13637G, the same as that found in LHON-4.

Patients LHON-8, LHON-9, LHON-10, and LHON-11

These unrelated patients were suspected of having LHON due to significant visual impairment with an unclear etiology, but the molecular analysis of mtDNA did not confirm this diagnosis, as no LHON-associated mutations were found.

Patient LHON-12

The sex and age of onset of this study participant were characteristic of LHON. The patient was male and experienced significant visual impairment at the age of 26 years. This patient’s decrease in visual acuity was not as severe as that for the patients discussed above. Visual field impairment (cecocentral scotoma) occurred only in the right eye. Molecular analysis of the mtDNA did not reveal any LHON-associated mutations.

Analysis of the mtDNA haplogroups of the studied patients

The majority of the participants belonged to haplogroup H or U, which are abundant mtDNA genotypes in the Latvian population [20] (Table 2).

Analysis of mtDNA heteroplasmy

As described previously, all non-synonymous point mutations in the mtDNA found in the current study were present only in the homoplasmic state. The homoplasmic state of these polymorphisms was suggested by the analysis of the sequencing chromatograms and was confirmed using the SURVEYOR mismatch endonuclease assay (Figure 2).

Prevalence of Leber hereditary optic neuropathy in European populations

As described previously, there is the great probability of identifying of one of the most common with LHON-associated mutations (G11778A, G3460A, or T14484C) among individuals suspected of having LHON. In the current study, the members of one family (two affected members) with distinct clinical features of LHON harbored the G11778A mutation. None of the studied individuals possessed the G3460A or T14484C mutation. However, the low prevalence of this disease in European populations should be considered. The absence of other primary LHON-associated mutations among the studied participants may be due to the small number of patients enrolled in the current study. Furthermore, the current project started recently and represents data collected over a comparatively short period of time. These data may be insufficient for evaluating the prevalence of such a rare disease. However, the project is ongoing, and all collected data may be used to develop a LHON registry in Latvia. In Finland, which is geographically close to Latvia, 36 families with LHON were identified over a period of 34 years. The most common LHON-associated mutation was G11778A (19 families with LHON). The A3460G and C14484T mutations in the mtDNA were found less frequently (four families and one family, respectively). In the remaining 12 families with clinical manifestations of LHON, none of the three primary LHON-associated mutations were found [1]. Sixteen families with LHON were identified in an epidemiological study performed in northeast England. One of the three primary LHON-associated mutations was found in each family by sequencing of the entire mtDNA. The most common LHON-associated mutation in this study was G11778A (60%), as observed in the Finnish population. The G3460A mutation was found in 33% of families with LHON, and the T14484C mutation was found in 7% [23]. Another epidemiological study was performed in the Netherlands, and among the 63 analyzed Dutch families suspected of having LHON, 56 carried one of the primary LHON-associated mutations. In 33 cases, it was the G11778A mutation, in 11 cases, it was G3460A, and in 12 cases, it was T14484C. Seven patients did not harbor any primary LHON-associated mutations [24].

Haplogroup association with LHON

In some studies of European populations, an association between LHON and haplogroup J has been found [1,15]. Moreover, in some cases, the clinical manifestations and penetrance of LHON have been found to be milder in families belonging to haplogroup H than among families belonging to haplogroup J [1,25]. Additionally, one clinical study showed that rare LHON-associated point mutations may be deleterious or, conversely, beneficial in the context of different mitochondrial haplogroup backgrounds [26]. It is important to consider the distribution of the mitochondrial haplogroups in the region under study to evaluate the mitochondrial genetic background and its possible influence on the penetrance of LHON. Haplogroup J is more common in the Italian population than in the Latvian population (14.3% versus 6.4%, respectively) [20,27]. Haplogroup J is found in the Finnish population at the same proportion as in the Latvian population (6.3% and 6.4%, respectively) [20,28]. Although haplogroup J is not common in these populations, it could be overrepresented among individuals with LHON due to the possible influence on the clinical manifestation of this disease. Based on phylogenetic analysis, in regions where haplogroup J is not common, there is an association between LHON and other definitive geographic region-specific haplogroups, e.g., M7b for the Chinese population [29]. This result may indicate that various combinations of selectively neutral polymorphisms in mtDNA may be associated with the penetrance of this disease. However, the more abundant haplogroups in the region under study should also be considered because of the possibility of masking the association of LHON with comparatively rare haplogroups, especially if the sample size is not large enough. For example, haplogroups H and U are more frequent in the Latvian population, and therefore, there is a high probability that individuals with LHON in this region may belong to these haplogroups instead of haplogroup J, which is rare [20]. Published data on the association of LHON with specific haplogroups are more abundant for the three common LHON-associated mutations, especially for G11778A [15]. Currently, there are insufficient data on the effect of the mitochondrial haplogroup background on the severity and age of onset of LHON in patients harboring the only rare LHON-associated mutation in European populations. In the current study, the majority of affected individuals belonged to haplogroups H and U, and none of subjects belonged to haplogroup J; therefore, it is not possible to compare the penetrance and severity of LHON between the two mtDNA phylogenetic branches: haplogroup H and haplogroup J.

In epidemiology studies of LHON in European populations, there were many sporadic cases of this disease and families without primary LHON-associated mutations; these patients harbored other non-synonymous polymorphisms in the mitochondrial genome [1,24]. Many of these newly identified polymorphisms are included in the MITOMAP database as predictably associated with the development of LHON. The majority of these polymorphisms are localized in MT-ND genes (49 (71%) out of 69 mutations), which encode the subunits of NADH dehydrogenase, which is part of OXPHOS (oxidative phosphorylation) respiratory chain complex I.

The previously reported rare LHON-associated mutation in the MT-ND5 gene A13637G was found in two unrelated individuals with LHON in the current study. These individuals belonged to the mitochondrial haplogroup U5b, which is listed as associated with this polymorphism in the PhyloTree database. According to the PhyloTree database, the A13637G polymorphism is also associated with two Asian haplogroups: M1a3 and N1c. This association may indicate that this polymorphism arose in different populations due to independent mutational events [30]. In addition, the A13637G polymorphism was found in one Chinese family that harbored the primary LHON-associated mutation G3460A. This family belonged to the haplogroup M10a1, which has not been shown to be associated with the A13637G polymorphism according to the PhyloTree database [31].

In the study of LHON in a Chinese population, the non-synonymous polymorphism T6253C was also observed [29]. This same polymorphism was found in patient LHON-3 in the current study. According to the MITOMAP database, this point mutation was previously reported to be associated only with prostate cancer [32,33]. According to the PhyloTree database, the T6253C polymorphism is also associated with various mitochondrial haplogroups: L1c2, M13’46’61, D5b1, and H15.

The PolyPhen and SIFT databases, which are widely used to predict probable deleterious effects of non-synonymous polymorphisms, did not confirm the pathogenicity of the A13637G and T6253C mutations. In light of all this information, these non-synonymous polymorphisms in the mtDNA may not be sufficient for the development of LHON. However, we could not exclude the possibility that these polymorphisms may be possible risk factors for LHON that could contribute to the clinical manifestation of this disease under certain conditions (e.g., stress, trauma) or when accompanied by other risk factors (e.g., a definitive nuclear background, excessive alcohol consumption). Further investigations and additional reports on the association between the molecular and clinical characteristics are necessary to evaluate the significance of these mutations in the development of this complex mitochondrial disease.

Conclusions

In a Latvian population, one family affected with LHON with a primary LHON-associated mutation, G11778A, was identified. Additionally, the A13637G mutation was identified in the mtDNA of two affected unrelated participants in this study, and the T6253C mutation was found in one patient. The analysis of the A13637G and T6253C mutation using the SIFT, PolyPhen data analysis software, and PhyloTree databases revealed that these polymorphisms are not conserved among species and are widely distributed among different mitochondrial haplogroups. Further investigations and additional reports are necessary to confirm the roles of these polymorphisms in the manifestation of LHON.