The T414G mitochondrial DNA mutation: a biomarker of ageing in human eye

Abstract

The mitochondrial mutation T414G (mtDNA^T414G) has been shown to accumulate in aged and sun-exposed skin. The human eye is also exposed to solar harmful rays. More precisely, the anterior structures of the eye (cornea, iris) filter UV rays and the posterior portion of the eye (retina) is exposed to visible light. These rays can catalyse mutations in mitochondrial DNA such as the mtDNA^T414G, but the latter has never been investigated in the human ocular structures. In this study, we have developed a technique to precisely assess the occurrence of mtDNA^T414G. Using this technique, we have quantified mtDNA^T414G in different human ocular structures. We found an age-dependent accumulation of mtDNA^T414G in the corneal stroma, the cellular layer conferring transparency and rigidity to the human cornea, and in the iris. Since cornea and iris are two anterior ocular structures exposed to solar UV rays, this suggests that the mtDNA^T414G mutation is resulting from cumulative solar exposure and this could make the mtDNA^T414G a good marker of solar exposure. We have previously shown that the mtDNA^CD4977 and mtDNA³⁸⁹⁵ deletions accumulate over time in photo-exposed ocular structures. With the addition of mtDNA^T414G mutation, it becomes feasible to combine the levels of these different mtDNA mutations to obtain an accurate assessment of the solar exposure that an individual has accumulated during his/her lifetime.

Introduction

The human mitochondrial DNA (mtDNA), a double-stranded 16 569 bp circular DNA, can be found in up to 100 000 copies in each human cell (1–3). Compared to genomic DNA, mitochondrial DNA (mtDNA) is highly susceptible to various mutations such as deletions and point mutations (4–7). The lack of histones and the absence of nucleotide excision repair system explains, at least in part, this sensitivity to mutations (8–10). mtDNA mutations can be induced by exogenous or endogenous factors. Endogenously, ATP production by mitochondrial oxidative phosphorylation generates reactive oxygen species (ROS) (11–14). Exogenously, ROS can be induced by solar and artificial light exposure, especially by ultraviolets (UV; 100–400nm) and high-energy visible light (HEV; 400–500nm) (15–18). Oxidative stress caused by the accumulation of ROS can lead to the formation of mutagenic mtDNA damage (11,19).

The presence of multiple mtDNA copies makes mtDNA mutations heteroplasmic (mixture of wildtype and mutant mtDNA in a given cell). Accumulation of some mitochondrial mutations in aged tissues has been explained by a theoretical vicious cycle (20). Indeed, it has been proposed that endogenous and/or exogenous oxidative stress alters mtDNA integrity, which triggers respiratory function deficiency. This inefficient mitochondrial respiratory function leads to an increased production of ROS (21,22). This theory is still controversial, and the question remains as to whether mtDNA alterations accumulate with age, or accumulation of mutations causes tissue-ageing. Since mtDNA mutations are tolerated and can be catalysed by solar exposure (19), it has been proposed that they can be used as biomarkers of photo-ageing (23,24). Such biomarkers could thus be used to determine the implication of cumulative ocular solar exposition in diseases.

Skin and eye are the two main organs exposed to solar radiation. Some mtDNA large deletions have been shown to accumulate in sun-exposed skin, such as the common mtDNA deletion of 4977 bp (mtDNA^CD4977) and the 3895 bp mtDNA deletion (mtDNA³⁸⁹⁵) (25–27). We have previously measured those two mtDNA deletions in different ocular structures and we found that they both accumulate with age in the corneal stroma, the layer conferring rigidity and transparency to the cornea (28,29). This led us to hypothesise that they could be used as ocular photo-ageing biomarkers. mtDNA³⁸⁹⁵ also accumulates in the macular region of the neural retina and the retinal pigment epithelium (RPE) suggesting a role of solar exposure in the formation of this deletion (29). Another mutation, a T to G transversion (mtDNA^T414G), localised on the light strand promoter binding site of the control region, has been shown to accumulate in aged and sun-exposed skin (25).

In this study, we have developed a Q-PCR technique to precisely assess the occurrence of mtDNA^T414G. Using this technique, we have quantified mtDNA^T414G in different human ocular structures. Similar to the mtDNA³⁸⁹⁵ and mtDNA^CD4977, we found an age-dependent accumulation of mtDNA^T414G in the corneal stroma but there was no significant difference in the macula or periphery of neural retina or RPE. We found an age-dependent accumulation of mtDNA^T414G mutation in the iris. To our knowledge, the mtDNA^T414G is the first mutation to accumulate in the iris region and could potentially act as a new biomarker of photo-ageing in the iris.

Materials and methods

All experiments performed in this study were conducted in accordance with our institution’s guidelines and the Declaration of Helsinki. The research protocols received approval by the CHU de Québec institutional committee for the protection of human subjects.

Human ocular structures isolation and DNA purification

We used a total of 64 human eyes from 38- to 94-year-old post-mortem donors unsuitable for transplantation provided by La Banque d’Yeux du Centre Universitaire d’Ophtalmologie, Québec, Canada (28–30). Transplantation exclusion criteria are based on the tissue quality criteria (e.g. scars). The eyes were enucleated no later than 1h after death and dissected upon reception, 24- to 48-h post-enucleation. The cornea, iris and neural retina were dissected from whole ocular globes and washed in 1× phosphate-buffered saline (PBS). To isolate the epithelium, stroma and endothelium, the corneas were incubated in HEPES buffer (0.01 M HEPES, pH 7.45; 0.142 M NaCl, 6.7 mM KCl and 1 mM CaCl₂) containing 2 mg/ml dispase II (Roche Applied Science) for 18h at 4°C. The epithelium and endothelium were then mechanically separated from the stroma, and the structures were washed in 1× PBS. The RPE was isolated by incubation of the posterior portion of the eye in HEPES buffer supplemented with 2 mg/ml dispase II for 45 min at 37°C. After mechanical isolation of the RPE from the choroid, RPE was washed in 1× PBS. Total DNA (genomic and mitochondrial) was then isolated using the DNeasy Blood and Tissue Kit (Qiagen) according to the manufacturer’s protocol with a RNase treatment. DNA extracted from layers containing melanin (iris, RPE and retina), were purified using MicroBiospin 6 chromatography columns (Bio-Rad Laboratories) to remove traces of melanin.

mtDNA^T414G level analysis by Q-PCR

mtDNA^T414G and total mtDNA levels were quantified using a Rotor-Gene Q real-time thermocycler (Qiagen). Primers were designed and are depicted in Table 1 and validation of the Q-PCR technique is depicted in Supplementary Figures S1 and S2. The Q-PCR was performed using Brilliant II SYBR Green (Agilent). Primers to detect total mtDNA (F5 and R4) have been used to amplify a sequence of 115 bp. A standard curve (10, 5, 1, 0.5, 0.1, 0.05, 0.01, 0.005 and 0.001 ng total DNA/reaction) was performed with total mtDNA primers for each tested sample. Using the standard curve, we derived a fitting logarithmic curve to determine the DNA amount (ng) in function of the Ct. A second set of primers (R4 and mutF3) were designed to amplify a sequence of 80 bp only when the mtDNA^T414G mutation was present (Figure 1). The amplification using those primers was done with 10–40 ng of total DNA. The amplification level of mtDNA^T414G was compared to the standard curve, and a DNA quantity corresponding to the mtDNA containing the T414G mutation was calculated. The ratio of ‘mtDNA molecules containing the T414G mutation’ to the ‘total mtDNA molecules’ was thus calculated (i.e. mtDNA^T414G/total mtDNA). Q-PCR was performed in 16 µl total volume with Brilliant II SYBR Green master mix and 0.4 µM of each primer. Q-PCR cycling conditions were as follows: 10-min hold at 95°C, followed by 40 cycles of 30 s at 95°C and 45s at 65.7°C.

Table 1.

Primers used to quantify the mtDNA^T414G/total mtDNA ratio

Primer names	Sequences	Position	Size
Forward (F5)	CCAAACCCCAAAAACAAAGAACC	348–370	23
Reverse wild type (R4)	GGAGGGGAAAATAATGTGTTAGTTGG	437–462	26
Reverse mutated (MutF3)	GATTTCAAATTTTATCTTTTGGCGTG	414–441	26

Figure 1.

Schematic representation of the Q-PCR technique used to evaluate the mtDNA^T414G/mtDNA ratio. Primer set F5/R4 is designed to amplify total mtDNA. The primer set R4/MutF3 specifically detects mtDNA molecules carrying the T414G mutation. The primer contains the T414G mutation in order to amplify only the mtDNA containing a guanine in position 414.

Statistical analysis

Statistical analysis was performed using Prism 5, GraphPad software. Since mtDNA^T414G level data did not follow a normal distribution between samples, we used two non-parametric tests to determine the significance. A Wilcoxon matched-paired rank test (two-tailed) was used to compare the mtDNA level of paired samples, such as different structures from the same subjects. When samples were not paired, a Mann–Whitney test (two-tailed) was performed to compared two conditions, e.g. to analyse the ratio mtDNA^T414G/total mtDNA of a structure in two groups of different ages.

Results

We developed a Q-PCR-based technique to precisely evaluate the mtDNA^T414G/total mtDNA ratio in the different structures of human eye. To specifically assess this ratio, we used three primers forming two sets. Two primers are placed on each side of the mutation (R4 in position 348–370 and F5 in position 437–462) in order to amplify the fragment proportionally to the total mtDNA levels in the sample. In combination with R4 primer, a third primer recognises the sequence near the mutation (mutF3 in position 414–441). This primer contains two mismatched nucleotides including the G at position 414 (Table 1). Using precise temperature that allows annealing of the primer mutF3 only when the T414G mutation is present, we could selectively amplify mutated mtDNA. Using the Q-PCR data, we derived the ‘mtDNA molecules containing T414G mutation/total mtDNA molecule’ (mtDNA^T414G/mtDNA) ratio. This strategy corrects for the variation of mtDNA copy in each cell.

mtDNA^T414G occurrence in the human eye

We quantified the mtDNA^T414G/total mtDNA ratio in the different ocular structures, i.e. cornea, iris and neural retina, from 7 different subjects. mtDNA^T414G level was significantly higher in the iris when compared to cornea and neural retina (P < 0.02). Indeed, the median mtDNA^T414G levels for the iris, cornea and neural retina were 0.051%, 0.021% and 0.012%, respectively (Figure 2). No statistical difference was found between the cornea and the neural retina.

Figure 2.

mtDNA^T414G/total mtDNA ratio analysis in different ocular structures. Total DNA was harvested from the cornea, iris and neural retina isolated from 7 human subjects. The mtDNA^T414G/total mtDNA ratio was derived from each ocular structure by Q-PCR. A higher mtDNA^T414G level has been found in the iris (0.051%), when compared to the cornea (0.021%) and the neural retina (0.012%). According to Wilcoxon matched-paired rank test, the mtDNA^T414G/total mtDNA ratio was significantly different between the iris and the cornea and between the iris and the neural retina (*P value < 0.02) but not between the cornea and the neural retina.

mtDNA^T414G accumulates with age in the human iris

The higher level of mtDNA^T414G in the iris led us to investigate the potential accumulation of this mutation with age in this ocular structure. We used 26 different individuals that were arbitrary divided into two groups ranging from 38 to 69 years old for the younger group, and from 70 to 94 years old for the second group. mtDNA^T414G/total mtDNA ratio was significantly higher in the human iris of the older group (P < 0.02). Median mtDNA^T414G values were 0.052% and 0.091% for the younger and the older groups, respectively (Figure 3). This result indicates an age-related accumulation of mtDNA^T414G in human iris.

Figure 3.

Age-related accumulation of mtDNA^T414G mutation in the iris. Total DNA was harvested from the iris of 13 subjects ranging from 38 to 70 years old and 13 subjects ranging from 70 to 94 years old. mtDNA^T414G/total mtDNA ratio was determined using Q-PCR. We found a significant higher mtDNA^T414G/total mtDNA ratio in the older group (70–94 years old), when compared to the younger 38- to 70-year-old group, with median values of 0.091% and 0.052%, respectively. A Mann–Whitney test was used (*P value < 0.02).

mtDNA^T414G analysis in corneal layers

Human cornea is composed of three cellular layers, i.e. epithelium, stroma and endothelium. We assessed the mutation levels in those three layers from 9 subjects (Figure 4). We found significantly more mtDNA^T414G in the corneal stroma in comparison to the corneal epithelium and endothelium (P < 0.04), with median values of 0.043%, 0.003% and 0.013%, respectively.

Figure 4.

mtDNA^T414G/total mtDNA ratio analysis in the different corneal cellular layers. The mtDNA^T414G/total mtDNA ratio was analysed by Q-PCR in the corneal epithelium, stroma and endothelium of 9 human subjects with a median age of 78 years old. We found that the mtDNA^T414G mutation is concentrated in the stroma (0.043%), when compared to the epithelium (0.003%) and the endothelium (0.013%). According to Wilcoxon test, the mtDNA^T414G/total mtDNA ratio is significantly different between the corneal stroma and the epithelium and between the stroma and the endothelium (*P value < 0.04).

mtDNA^T414G accumulates with age in the corneal stroma

Similar to the analysis in the iris, we evaluated mtDNA^T414G occurrence in corneal stroma according to age (Figure 5). The younger group was chosen between 38 and 69 years old with 14 subjects, whereas the older group contains 23 subjects from 70 to 94 years old. There was more variation in mtDNA^T414G/total mtDNA ratio in the older group. Nonetheless, there were significantly (P < 0.005) more mtDNA^T414G mutations in the older group (median 0.017%) when compared with the younger group (median 0.007%). As observed in the iris, mtDNA^T414G mutation accumulates in corneal stroma with age.

Figure 5.

Age-related accumulation of mtDNA^T414G in the corneal stroma. The mtDNA^T414G/total mtDNA was analysed in total DNA harvested from the stroma of 14 subjects ranging from 38 to 70 years old and 23 subjects ranging from 70 to 94 years old. mtDNA^T414G/total mtDNA ratio was determined by Q-PCR. We found significantly more mtDNA^T414G mutation in the older group (0.017%) than in the younger one (0.007%). A Mann–Whitney test has been used (*P value < 0.01).

mtDNA^T414G in human neural retina and RPE

We investigated mtDNA^T414G/total mtDNA ratio more precisely in neural retina and RPE by discriminating macular to non-macular regions (Figure 6). The mutation was quantified in peripheral or macular regions of 7 subjects (i.e. retina macula, retina periphery, RPE macula and RPE periphery). We found no significant difference between the macular or peripheral region of the neural retina (median 0.027% and 0.025%, respectively) or the RPE (0.011% and 0.014%, respectively).

Figure 6.

mtDNAT^414G/total mtDNA ratio analysis in the macular and peripheral regions of the neural retina and the RPE. After dissection of the neural retina (RET) and the RPE of 7 patients with a median age of 60, the macular (mac) and peripheral (periph) regions were separated. Total DNA was extracted and mtDNA^T414G/total mtDNA ratio was determined using Q-PCR. Analysis of the mtDNAT^414G/total mtDNA ratio shows no significant difference between the macular and peripheral regions of both neural retina (0.027% and 0.025%, respectively) and RPE (0.011% and 0.014%, respectively).

Discussion

Development of a new way to precisely quantify mtDNA^T414G level

Previous studies have used different techniques to measure the presence of mtDNA^T414G in cells such as denaturant gradient gel electrophoresis (DGGE) (31) or denaturing high performance liquid chromatography WAVE analysis (25,26) and sequencing-based technics (32–34). In this study, we developed a fast and accurate Q-PCR-based technique to precisely quantify the mtDNA^T414G/total mtDNA ratio. This method allows specific evaluation of mtDNA copies that contains mtDNA^T414G. Analysis has been performed on fresh post-mortem ocular tissue, which prevented any positive or negative selection of this mutation that could have happened with cellular culture.

The T to G transversion at position 414 of the mtDNA has been previously reported in skin where it was shown to be related to age and cumulated sun exposure (25,26). Those studies concluded that mtDNA^T414G could be used as a photo-ageing biomarker. However, there is still no mechanistic explanation for the induction of this transversion mutation.

mtDNA^T414G in cornea, iris and retina

We found that mtDNA^T414G accumulates in the iris and the cornea, which are the most anterior structures of the human eye and thus the most sun-exposed structures. This suggests a potential involvement of sunlight in the formation of this mutation. The fact that the retina is protected from harmful UV rays (35,36) might explain the lower occurrence of mtDNA^T414G in this ocular structure. We have previously studied large mtDNA deletions known to be induced by sunlight in human ocular structures (28,29). In line with our data showing a protection of the retina against mtDNA^T414G, we found lower levels of mtDNA deletions in the retina.

mtDNA^T414G accumulation in corneal stroma

As previously described for the large mtDNA deletions of 3895 bp and 4977 bp (28,29), we found an accumulation of mtDNA^T414G in the corneal stroma (Figure 4). Indeed, the mtDNA^T414G levels were three times higher in the corneal stroma than in the endothelium and 14 times higher than in the epithelium. In the skin, a correlation between mtDNA^T414G mutation and mtDNA³⁸⁹⁵ deletion has been described (25). It is possible that the mtDNA^T414G is also correlated to the mtDNA³⁸⁹⁵ deletion in corneal stroma. In fact, the mtDNA³⁸⁹⁵/total mtDNA ratio we previously shown in corneal stroma (0.062%) (29) is in the same order of magnitude than the mtDNA^T414G/total mtDNA ratio we found in this study (0.043%). Further analysis would be necessary to validate the interconnection between the mtDNA³⁸⁹⁵ and the mtDNA^T414G mutations in human corneal stroma.

mtDNA^T414G accumulates with age in corneal stroma and iris

We found an age-related accumulation of the mtDNA^T414G in the corneal stroma and in the iris. The highest levels of mtDNA^T414G found in our study were in the iris of older group (up to 0.17%). Since it has been shown that the mtDNA^T414G mutation accumulates with age and sun exposure (25) and that we found it concentrated in solar UV-exposed ocular structure (corneal stroma and iris), it can be postulated that we could use this mutation as a biomarker of ocular photo-ageing. We have previously shown that the mtDNA³⁸⁹⁵ is virtually absent from the iris (29), which is in contradiction with the results in this study showing an accumulation of mtDNA^T414G in this ocular structure. There must be a mechanism unrelated to the mtDNA³⁸⁹⁵ deletion involved in the generation of mtDNA^T414G mutation. Since we did not find any difference in mtDNA^T414G levels between the most light-exposed region of the retina, i.e. the macula and the non-macula (Figure 6), it was irrelevant to investigate the accumulation of this mutation in the retina as a function of age in an attempt to determine a photo-ageing marker in the eye.

Involvement of ROS in mtDNA^T414G formation

We found similarities (e.g. corneal stroma) and differences (e.g. iris, retina) in mutation induction when comparing previously analysed deletion (i.e. mtDNA^CD4977 and mtDNA³⁸⁹⁵) with the mtDNA^T414G (28,29). The induction mechanism involved in these mutations is still under investigation. We know that the mtDNA^CD4977 and the mtDNA³⁸⁹⁵ deletions are both flanked with short, repeated GC-rich sequences and it has been hypothesised that oxidative DNA damage at those sites catalyse the occurrence of the deletions.

We found an accumulation of mtDNA^T414G in the corneal stroma and the iris, two anterior ocular structures. Those structures are directly exposed to solar irradiation. One major role of the cornea is to filter a large portion of the UVB wavelengths (280–315 nm) and some of the UVA wavelengths (315–400 nm) (37,38). The iris is exposed to UV wavelengths that are not filtered by the cornea, mainly UVA. Exposure to UVA leads to the formation of exogenous ROS, which can damage mtDNA (15). However, for the mtDNA^T414G mutation, there is no site that can be favourable for the formation of oxidative damage and oxidative stress does not seem to have any implication in the formation of this mutation (26,39). Nonetheless, previous publications linked the accumulation of mtDNA^T414G mutation to sun exposure in skin and we show that this mutation is concentrated in the most sun-exposed structures of human eyes. This suggests that there is a relation between sun exposure and mtDNA^T414G accumulation in ocular tissue.

mtDNA^T414G as a new photo-ageing biomarker in corneal stroma and iris

Ocular cumulative solar exposure is difficult to assess. The measurement of ocular solar exposure is challenging especially because multiple components contribute to an individual’s exposure at any given time, including wearing prescription glasses, sunglasses or contact lenses; extent of shade coverage; length of forelock hairs; protrusion of the brow; eyelid anatomy; posture; activity, leisure and occupation; day of the year, latitude, elevation; environmental condition (air quality, cloud cover). Cumulative solar exposure is suspected or demonstrated in many ocular pathologies, including uveal melanoma, age-related macular degeneration (AMD) or Fuch’s endothelial corneal dystrophy, but evidence is mainly epidemiological derived.

Due to inter-individual variability, it can be difficult to use a single mitochondrial mutation as a reliable marker of cumulative solar exposure. We have previously shown that certain mtDNA deletions (mtDNA^CD4977 and mtDNA³⁸⁹⁵) accumulate over time in photo-exposed ocular structures (28,29). With the addition of mtDNA^T414G mutation, it becomes feasible to combine the levels of these different mtDNA mutations to obtain an accurate assessment of the solar exposure that an individual has accumulated during his lifetime. This assessment could help determine the implication of cumulative sun exposure in ocular pathologies for which sun exposure is suspected, such as AMD, Fuchs’ endothelial corneal dystrophy, uveal melanoma or glaucoma.

Funding

This work was supported by a Grant from the Canadian Institutes of Health Research (CIHR, MOP-133719) to P.J.R. P.J.R. is a research scholar from the Fonds de Recherche du Québec – Santé (FRQ-S).

Conflict of interest statement. The authors declare that they have no conflict of interest.

Authors’ contributions. A.-S.G.: experiment conception and design, data collection, analysis and interpretation of data, manuscript writing, and critical revision; M.M.D.: experiment conception and design, data collection, analysis and interpretation of data and critical revision; P.J.R.: experiment conception and design, funding acquisition, critical revision.