Pathogenic Mitochondrial DNA Mutations and Associated Clinical Features in Korean Patients With Leber's Hereditary Optic Neuropathy

Abstract

Purpose.: To identify the spectrum of pathogenic mitochondrial DNA (mtDNA) mutations and clinical features in Korean patients with genetically confirmed Leber's hereditary optic neuropathy (LHON).

Methods.: The medical records of 34 unrelated, genetically confirmed LHON patients were reviewed. Total genomic DNA was isolated from the peripheral blood leukocytes of the patients with suspected LHON, and primary or secondary mtDNA mutations were identified by direct sequencing. We analyzed the visual acuity (VA), color vision, RNFL thickness, and visual field (VF) at the final visit from 20 patients who were followed-up for more than 6 months after the onset of LHON.

Results.: Among 34 patients, 21 (61.8%) had the homoplasmic primary mutation, 11 (32.4%) had the homoplasmic secondary mutation, and 2 (5.9%) had the heteroplasmic primary mutation along with the homoplasmic secondary mutation. Analysis of mtDNA sequences revealed six different types of LHON-associated mutations: two primary LHON-associated primary mutations, m.11778G>A (20 patients, 58.8%) and m.14484T>C (3 patients, 8.8%), and four secondary LHON-associated mutations, which were m.3394T>C (3 patients, 8.8%), m.3497C>T (4 patients, 11.8%), m.11696G>A (4 patients, 11.8%), and m.14502T>C (2 patients, 5.9%). Secondary mutation-carrying patients demonstrated a decreased in RNFL thickness, similar to those in primary mutation–carrying LHON patients. These patients had a higher female ratio (P = 0.019), better VA (P = 0.043) and color vision (P = 0.005), as well as better VF.

Conclusions.: In addition to common primary LHON-associated mutations, our study identified secondary mtDNA mutations, which should be considered when evaluating patients with optic atrophy.

Introduction

Leber's hereditary optic neuropathy (LHON) is typically characterized by bilateral acute or subacute loss of central vision, owing to focal degeneration of the retinal ganglion cell layer and optic nerve.^1–3 It is a mitochondrial genetic disease inherited through maternal transmission that mainly affects young adult males.^1–3 To date, over 30 different types of mitochondrial DNA (mtDNA) mutations, including 14 primary mutations, have been identified in patients with LHON (in the public domain http://www.mitomap.org/MITOMAP). Based on the genetic, biochemical, and clinical features, three mtDNA point mutations, ND1 m.3460G>A, ND4 m.11778G>A, and ND6 m.14484T>C, in the genes encoding the subunits of respiratory chain complex I have been traditionally considered as high-risk LHON mutations.^4–7 These mutations represent approximately 80% to 95% of all LHON cases occurring in patients with different ethnic backgrounds.^5–9 However, no primary mutations have been identified in a small minority of diagnosed LHON cases. In one study, no primary mutations were found in 21% of LHON-affected patients from Finland.³ Recently, a growing number of mtDNA variants that are suspected to have a role in the disease expression of LHON have been reported.^10–13 The pathological significance of mtDNA mutations other than primary mutations is uncertain and considered to be intermediate or low risk. Therefore, these have been proposed as secondary mutations. Secondary mutations that are relevant for the phenotypic expression of LHON appear to have a synergistic and deleterious effect together with the primary mutations.^10–13 These secondary LHON mutations are homoplasmic, often occurring in combination with primary LHON mutations.^10–13 They are found at a lower frequency in healthy subjects compared with the LHON population; however, they may have an additional role in the pathogenesis of LHON and increase the risk of disease expression.^11,14,15 

The distribution of LHON-associated mtDNA mutations is population-specific. However, little is known about the mutation and clinical characterization for Korean patients with LHON. Only five types of mtDNA mutations have been reported in Korean patients. Restriction fragment length polymorphism-PCR (RFLP-PCR) has revealed m.11778G>A (56.1%), m.14484T>C (15.9%), and m.3460G>A (1.2%) mutations in Korean LHON patients.^16,17 In addition, secondary mtDNA mutations at nucleotide position (np) 4216 and 15257 were also reported in Korean LHON patients with or without primary mutation.^18,19 In this study, we have used DNA sequencing to analyze LHON-associated mtDNA mutations together with the clinical characterization of Korean LHON patients. 

Methods

This study was performed according to the principles of the Declaration of Helsinki, after approval by the institutional review board of Seoul St. Mary's Hospital, College of Medicine, The Catholic University of Korea, Seoul, Korea (KC14RISE0241). 

Subjects

The medical records of 34 unrelated genetically confirmed LHON patients at Seoul St. Mary's Hospital from January 2009 to May 2014 were reviewed. The age, sex, family history, age at LHON onset, and initial/final visual acuity (VA) were evaluated for each patient. Acute or subacute visual loss with typical peripapillary microangiopathy were classified as type I clinical features. Bilateral optic atrophy with temporal retinal nerve fiber layer (RNFL) thinning and central scotoma were classified as type II clinical features. After thorough ophthalmologic evaluations, subjects were excluded on the basis of any of the following criteria: a history of retinal disease, including diabetic or hypertensive retinopathy, a history of eye trauma or surgery, a history of systemic or neurologic disease that may affect the optical coherence tomography (OCT) and visual field (VF) results, and any relevant drug history. 

Mutational Analysis

After obtaining informed consent from each patient, blood samples were taken for mtDNA extraction and analysis. Total genomic DNA was isolated from peripheral blood leukocytes using the QIAmp DNA Mini Kit (Qiagen, Hamburg, Germany). The presence of the m.3460G>A, m.11778G>A, and m.14484T>C mutations and secondary mutations in genes mitochondrially encoded NADH dehydrogenase 1 (MT-ND1), mitochondrially encoded NADH dehydrogenase 4 (MT-ND4), mitochondrially encoded NADH dehydrogenase 6 (MT-ND6), and mitochondrially encoded cytochrome B (MT-CYB) were screened by direct sequencing using an ABI prism Big Dye Terminator cycle sequencing Ready Reaction kit V 3.1 (Applied Biosystems, Foster City, CA, USA) and analyzed on an ABI Prism 3100 Genetic Analyzer (Applied Biosystems). All sequences were analyzed against mitochondrial reference sequence NC_012920. We performed in silico analysis using the software Polyphen-2 (in the public domain http://genetics.bwh.harvard.edu/pph2/) and Sorting Intolerant from Tolerant (SIFT; in the public domain http://sift.jcvi.org/) to assess if the substitutions were predicted as potentially pathogenic.^20,21 

Ophthalmologic Examination

Each participant underwent a comprehensive ophthalmologic assessment, which included the measurement of best-corrected visual acuity (BCVA), a color vision test, slit-lamp biomicroscopy, a dilated stereoscopic examination of the optic nerve head and fundus, a VF test, and optical coherence tomography (OCT). Clinical examinations were performed by an experienced neuro-ophthalmologist. 

We analyzed the ophthalmologic results at the final visit from 20 patients who were followed-up for more than 6 months after the onset of LHON. When both eyes were eligible for the study, we analyzed a worse-seeing eye in each patient. We used the 24-plate edition from the Ishihara test book published in 1997 by Kanehara & Co., Ltd. (Tokyo, Japan), and the score was set as the number of plates identified out of the first 12 screening plates and five diagnostic plates.²² 

The peripapillary RNFL thickness was measured using a spectral-domain Cirrus HD-OCT (software version 6.0, Carl Zeiss Meditec, Dublin, CA). Spectral-domain OCT imaging was obtained from a 3-dimensional data set from an optic cube scan, which was composed of 200 A-scans from each of 200 B-scans that covered a 6-mm² area centered on the optic disc. Following the creation of the RNFL thickness map from the cube data set, the software then automatically determines the center of the disc and extracts a circumpapillary circle (1.73-mm radius) from the cube data set, which it then uses to determine RNFL thickness. To avoid confusion, clock-hour sectors were named according to their location as either superior (S), inferior (I), temporal (T), or nasal (N). The following OCT parameters were used in the analysis: S, I, T, N, and overall RNFL thickness. Image quality was assessed by an experienced examiner who was unaware of the patient's identity and condition. 

We performed at least two standard automated perimetry tests (Humphrey Visual Field Analyzer; Zeiss/Humphrey Systems, Dublin, CA, USA) 24–2, using the Swedish interactive threshold algorithm (SITA) strategy. The VF defect had to be present at the same location in at least two consecutive, reliable VFs for the result to be included in the study. The results from the VF tests were considered reliable, if the fixation losses were less than 20% and false positive and false negative rates were less than 15%. 

Statistical Analysis

Statistical analysis was performed using SPSS software (version 19.0; SPSS, Inc., Chicago, IL, USA). The χ² test and Mann-Whitney U test were used to compare data from the primary and secondary mutation groups. A P value of less than 0.05 was regarded as statistically significant. 

Results

Genetic Analysis

Among 34 patients, 21 (61.8%) had the homoplasmic primary mutation, 11 (32.4%) had the homoplasmic secondary mutation, and 2 (5.9%) had the heteroplasmic primary mutation along with the homoplasmic secondary mutation. 

The mtDNA mutations identified in each patient are summarized in Table 1. Six different kinds of mtDNA mutations were detected in 34 unrelated patients with bilateral optic atrophy. In addition to the two common LHON-associated primary mutations, m.11778G>A (20 patients, 58.8%) and m.14484T>C (3 patients, 8.8%), four secondary mutations were also identified, m.3394T>C (3 patients, 8.8%), m.3497C>T (4 patients, 11.8%), m.11696G>A (4 patients, 11.8%), and m.14502T>C (2 patients, 5.9%). In silico analysis of the secondary mutations generated differing results. Polyphen-2 predicted that all four substitutions would be benign and SIFT predicted that two of the secondary mutations (m.3394T>C, m.3497C>T) would not be tolerated (Table 1). We found no patients carrying the primary mutation m.3460G>A. 

Table 1

 

Sequence Analysis of Mutations in Korean Patients With LHON

	Numbers	Gene Locus	Mutation	Codon	Mutation Type	MITOMAP Status	SIFT	Polyphen-2	References
Primary	19	MT-ND4	m.11778G>A	p.Arg340His	Homoplasmic	Confirmed	Deleterious	Probably Damaging
	2	MT-ND6	m.14484T>C	p.Met64Val	Homoplasmic	Confirmed	Deleterious	Benign
Secondary	3	MT-ND1	m.3394T>C	p.Tyr30His	Homoplasmic	Reported/unclear	Not tolerated	Benign	7 (American)
									11 (Japanese)
									12, 15 (Chinese)
									37 (Finnish)
	3	MT-ND1	m.3497C>T	p.Ala64Val	Homoplasmic	Reported/secondary	Not tolerated	Benign	11 (Japanese)
									38 (Thai)
	4	MT-ND4	m.11696G>A	p.Val313Ile	Homoplasmic	Reported/possibly synergistic	Tolerated	Benign	10, 32, 33 (Chinese)
									31 (Dutch)
	1	MT-ND6	m.14502T>C	p.Ile58Val	Homoplasmic	Reported/possibly synergistic	Tolerated	Benign	39 (Chinese)
Combined	1	MT-ND4	m.11778G>A	p.Arg340His	Heteroplasmic +	Confirmed + reported/secondary			38 (Thai)
		MT-ND1	m.3497C>T	p.Ala64Val	homoplasmic
	1	MT-ND6	m.14484T>C + m.14502T>C	p.Met64Val p.Ile58Val	Heteroplasmic + homoplasmic	Confirmed + reported/possibly synergistic			13, 40 (Chinese)
Total	34

Patients with homoplasmic secondary mutations had common ocular features, with a pale optic disc and decreased peripapillary RNFL thickness in both eyes (Table 2; Fig. A–D). Two patients were shown to have homoplasmic secondary mutations and heteroplasmic primary mutations. The coexistence of mutations MT-ND1 m.3497C>T/MT-ND4 m.11778G>A and MT-ND6 m.14502T>C/MT-ND6 m.14484T>C was detected in each patient (Table 3, cases 12 and 13). 

Table 2

 

Clinical Findings of Korean Patients Carrying Homoplasmic Secondary LHON Mutations

Case No.	Age	Sex	Mutation	Mutation Type	Age of Onset, y	Type I or II	Family History	Final BCVA		Spontaneous Recovery
								Right	Left
1	59	F	m.3394T>C	Homoplasmic	58	II	−	20/32	20/25	+
2	57	F	m.3497C>T	Homoplasmic	19	II	+	20/200	20/1000	−
3	26	M	m.3394T>C	Homoplasmic	26	I	−	20/80	20/160	−
4	42	F	m.11696G>A	Homoplasmic	42	I	−	20/200	20/125	−
5	59	F	m.11696G>A	Homoplasmic	Unclear	II	−	20/32	20/25	+
6	7	F	m.11696G>A	Homoplasmic	6	II	−	20/160	20/80	−
7	66	F	m.3394T>C	Homoplasmic	Unclear	II	−	20/40	20/32	+
8	64	F	m.14502T>C	Homoplasmic	4	II	+	20/32	20/32	+
9	30	M	m.3497C>T	Homoplasmic	Unclear	II	−	20/40	20/32	+
10	14	F	m.3497C>T	Homoplasmic	14	I	−	20/25	20/32	+
11	51	M	m.11696G>A	Homoplasmic	Unclear	II	+	20/100	20/200	−

 M, male; F, female.

Table 3

 

Clinical Findings of Korean Patients Carrying Combined Primary and Secondary LHON Mutations

Case No.	Age	Sex	Mutation	Mutation Type	Age of Onset, y	Type I or II	Family History	Final BCVA		Spontaneous Recovery
								Right	Left
12	43	M	m.11778G>A + m.3497C>T	Heteroplasmic + homoplasmic	23	II	−	20/1000	20/20000	−
13	21	M	m.14484T>C + m.14502T>C	Heteroplasmic + homoplasmic	5	II	−	20/50	20/50	−

Figure.

 

(A) Patient number 1 was a 59-year-old female with a homoplasmic m.3394T>C mutation. Her final visual acuity was 20/32 in the right eye and 20/25 in the left eye. (B) Patient number 2 was a 57-year-old female with a homoplasmic m.3497C>T mutation. Her final visual acuity was 20/200 in the right eye and 20/1000 in the left eye. (C) Patient number 4 was a 42-year-old female with a homoplasmic m.11696G>A mutation. Her final visual acuity was 20/200 in the right eye and 20/125 in the left eye. (D) Patient number 8 was a 64-year-old female with a homoplasmic m.14502T>C mutation. Her final visual acuity was 20/32 in both eyes.

Clinical Features

The male:female ratio in this study was 21:13. The overall proportion of LHON patients with a family history of the disease was 38.2%. In 10 patients who had type I clinical features, with acute or subacute visual loss and peripapillary telangiectatic microangiopathy, the mean age of LHON onset was 25.2 years. Both Tables 2 and 3 summarize the clinical findings from patients carrying homoplasmic secondary LHON mutations and combined primary and secondary LHON mutations. 

We analyzed the ophthalmologic results at the final visit from 20 patients who were followed-up for more than 6 months after the onset of LHON. We then compared the clinical data from patients with primary and secondary LHON mutations (Table 4). In the secondary mutation group, there was a higher proportion of type II clinical features (P = 0.020) and lower male to female ratio (P = 0.019). No significant difference was found in the OCT parameters, including those in the temporal area, between the two groups. However, the BCVA was significantly better in patients with the secondary mutation than in those with the primary mutation (P = 0.043). The color vision deficiency was also less prominent in patients with the secondary mtDNA mutation compared with those with the primary mtDNA mutation (P = 0.005). Both the VF index and MD of the VF test were significantly different between the two groups (P = 0.013 and P = 0.024). 

Table 4

 

Comparisons of Clinical Data of Patients With Primary and Secondary Mutation of LHON

	Primary Mutation, N = 12	Secondary Mutation, N = 8	P Value
Type I : Type II	3:1	1:7	0.020*
Sex ratio (M:F)	10:2	2:6	0.019*
BCVA, logMAR	0.72 ± 0.54	0.41 ± 0.48	0.043†
Color vision (mean Ishihara score)	3.00 ± 4.20	12.11 ± 6.66	0.005†
RNFL, μm
Superior RNFL	112.17 ± 39.65	98.38 ± 32.41	0.487†
Nasal RNFL	64.58 ± 11.65	62.38 ± 14.20	0.231†
Inferior RNFL	109.67 ± 44.79	93.13 ± 45.49	0.375†
Temporal RNFL	49.75 ± 16.06	40.38 ± 8.88	0.097†
Overall RNFL	84.00 ± 25.94	73.25 ± 19.84	0.512†
VF
MD, dbs	−18.11 ± 11.06	−7.57 ± 10.96	0.024†
VFI	45.44 ± 33.45	79.33 ± 32.43	0.013†

 VFI, visual field index.

*  χ² test (Fisher exact test in the case of presence of cell with frequency of lower than 5).

†  Mann-Whitney U test. Data are expressed as the mean ± SD.

Discussion

In this study, we report genetic and clinical characterization of LHON in 34 unrelated Korean patients. Sequence analysis of the mitochondrial genome revealed six different kinds of LHON-associated mtDNA mutations in patients with bilateral visual impairment due to optic atrophy as a sole clinical manifestation. The two common primary LHON-associated mutations, m.11778G>A and m.14484T>C were identified in 58.8% (20 patients) and 8.8% (3 patients) of the patients, respectively. The four secondary mtDNA mutations, m.3394T>C, m.3497C>T, m.11696G>A, and m.14502T>C, which may be associated with LHON, were found in 8.8% (3 patients), 11.8% (4 patients), 11.8% (4 patients), and 5.9% (2 patients) of patients, respectively. These secondary mtDNA mutations have been previously reported in other populations, but reported here for the first time in Korean LHON patients.^10–13 

The seven complex I genes form a large proportion of the mitochondrial genome, comprising 38% of the total mtDNA. Complex I defects are increasingly being recognized as important causes of respiratory chain disease.^23,24 The most common biochemical abnormalities seen in patients with LHON are complex I defects. Mutations in mtDNA may cause alterations of the electron transport components, which can compromise the normal electron flow. 

This study found that the m.11778G>A mutation in the MT-ND4 gene was the most prevalent LHON-associated mutation responsible for 58.8% of Korean LHON patients, which was consistent with studies from other countries.^5,6,14 The m.14484T>C mutation in the MT-ND6 gene, an another common primary LHON-associated mutation, had been reported to be the second most prevalent mtDNA mutation type, with an incidence of 15.9% (13/82) in a previous Korean LHON study.¹⁶ In this study, only two patients carried the homoplasmic m.14484T>C mutations. We found no patients carrying the primary mutation m.3460G>A. The frequencies of three primary mtDNA mutations for LHON in the Asian populations were demonstrated in Table 5.^8,16,25–30 The prevalences have been reported to be 77% to 95% for the m.11778G>A mutation, 5% to 22% for the m.14484T>C mutation, and 0% to 4% for the m.3460G>A mutation. The relatively lower contribution of the m.11778G>A mutation to LHON was an interesting finding in our study. 

Table 5

 

Frequencies of the Three Primary mtDNA Mutations for LHON in Asia

References	Ethnic Population	No. of LHON Proband	Mutation			Other Reported Mutations
			m.11778G>A Number (%)	m.14484T>C Number (%)	m.3460G>A Number (%)
Mashima et al.8	Japanese	68	59 (86.8)	6 (8.8)	3 (4.4)	m.3394T>C, m.4216T>C, m.7444G>A, m.9438G>A, m.13708G>A
Yamada et al.25	Japanese	72	63 (88)	6 (8)	3 (4)
Jia et al.26	Chinese	346	312 (90.2)	30 (8.7)	4 (1.1)
Yen et al.27	Chinese	25	23 (92)	2 (8)	0 (0)	m.3394T>C, m.3497C>T, m.4216T>C
Ji et al. 28	Chinese	1164	-	-	16 (1.4)
Liang et al. 29	Chinese	1218	-	59 (4.8)	-	m.14502T>C, m.14459G>A
Sudoyo et al.30	Southeast Asian	19	18 (94.7)	1 (5.3)	0 (0)	m.3316G>A
Kim et al.16	Korean	60	46 (76.7)	13 (21.7)	1 (1.6)
Present study	Korean	34	20 (58.8)	3 (8.8)	0 (0)	m.3394T>C, m.3497C>T, m.11696G>A, m.14502T>C

Four secondary mtDNA mutations, which were all listed on MITOMAP, were identified in this study. Interestingly, the homoplasmic m.11696G>A mutation was the second most prevalent LHON-associated mtDNA mutation identified in our study (4 unrelated patients, 11.8%). The homoplasmic m.11696G>A mutation in the MT-ND4 gene associated with LHON has been identified in a large Dutch family and in some Chinese pedigrees.^10,31–33 This mutation, which is located within a predicted transmembrane region, results in the replacement of an arginine with a histidine and is currently classified as provisional according to the MITOMAP website.³⁴ The two types of secondary mtDNA mutations in the MT-ND1 gene were detected in seven patients. The homoplasmic m.3394T>C and m.3497C>T mutation were found in three unrelated patients, respectively. The ND1 protein is involved in the first step of the electron transport chain of oxidative phosphorylation. The m.3394T>C and m.3497C>T mutations resulted in the substitution of a histidine for a tyrosine at position 30 and a valine for an alanine at position 64 of the ND1 protein. These two mutations occur within a highly conserved region in the human ND1 protein.^34–36 Mutations in this highly conserved region are assumed to alter the structure and function of the ND1 protein. The m.3394T > C mutation has been reported to be associated with LHON in American, Japanese, Finnish, and Chinese populations.^7,11,12,37 This mutation is currently considered to have an unclear status on the MITOMAP resource. The m.3497C>T mutation was first reported to be associated with Japanese LHON patients in combination with the m.11778G>A mutation.¹¹ The homoplasmic m.3497C>T LHON-associated mutation is categorized as provisional on the MITOMAP resource.^11,38 The m.14502T>C mutation in the MT-ND6 gene was also identified in our study. The m.14502T>C mutation causes substitution of a highly conserved isoleucine to valine, at position 58 in the ND6 protein.³⁴ The homoplasmic m.14502T>C mutation, which was first detected in three unrelated Chinese families, has been suggested as a possible pathogenic target.³⁹ The coexistence of a homoplasmic m.14502T>C mutation with a heteroplasmic m.14484T>C mutation was present in one patient (Table 3). The m.14502T>C mutation may exert a synergistic effect with the nearby primary m.14484T>C mutation.^13,40,41 

Interestingly, a considerable number of Korean LHON patients carried secondary LHON-associated mtDNA mutations. In silico analysis using SIFT, predicted two of the secondary mutations (m.3394T>C, m.3497C>T) detected in our study as not tolerated. Whether these mutations are pathogenic LHON mutations is unclear. However, the occurrence of homoplasmic m.3394T>C and m.3497C>T in the MT-ND1 gene and the MT-ND4 m.11696G>A mutations in several genetically unrelated patients with bilateral optic atrophy strongly suggested their involvement in the pathogenesis of visual impairment (Table 2). Our findings support the notion that the homoplasmic m.11696G>A, m.3394T>C, and m.14502T>C mutations could be the causative mutations of LHON.^10,12,39 

We did not observe any significant differences in the RNFL thickness between patients with primary and secondary mutations (Table 4). Patients carrying a secondary mutation demonstrated a decreased RNFL thickness, compatible with optic atrophy, similar to those in primary mutation–carrying LHON patients (Fig., Table 4). These patients had a higher female ratio, better VA and color vision as well as better VF (Table 4). 

The pathophysiology of LHON is complex, and incomplete penetrance and variable expressivity, in terms of severity and age at onset of visual impairment, are typical features of LHON, found even in matrilineal relatives carrying an identical LHON mtDNA mutation. They implicate that secondary modulating factors such as nuclear modifier genes, environmental factors and mitochondrial variants/haplotypes are necessary in the phenotypic manifestation of the optic neuropathy.^{15,28,42–46} In our study, 13 (38.2%) patients had a family history of visual loss related to optic neuropathy, the remaining 21 (61.8%) patients were sporadic. However, extended pedigree analyses for maternally-transmitted LHON patients have not been performed enough to demonstrate the phenotypic manifestations, such as the penetrance, age at onset, and disease severity. In addition, sequence analysis of the entire mitochondrial genomes is necessary to delineate the phenotypic manifestation of mitochondrial variants in its full aspect. Further research of pedigree analyses of the penetrance and expressivity in the context of mitochondrial variants/haplotypes will conclusively demonstrate the influence of the mtDNA mutations in the phenotypic manifestation of LHON. 

Based on our results, Korean patients with LHON exhibited a wide spectrum of mtDNA mutations responsible for visual impairment. This research suggests that the evaluation of patients with optic atrophy should include both the primary and secondary LHON-associated mtDNA mutations. Bidirectional DNA sequencing is recommended for the genetic confirmation of suspected LHON patients. 

Acknowledgments

Supported by grants from Basic Science Research Program through the National Research Foundation of Korea (NRF; Daejeon, Korea) funded by the Ministry of Education (MOE; Sejong, Korea) (NRF-2013R1A1A2006801). 

Disclosure: H.R. Yum, None; H. Chae, None; S.Y. Shin, None; Y. Kim, None; M. Kim, None; S.H. Park, None 

Footnotes

 HRY and HC contributed equally to the work presented here and should therefore be regarded as equivalent authors.