SLC4A11 depletion impairs NRF2 mediated antioxidant signaling and increases reactive oxygen species in human corneal endothelial cells during oxidative stress

Corneal endothelial dystrophy is a progressive disease with gradual loss of vision and characterized by degeneration and dysfunction of corneal endothelial cells. Mutations in SLC4A11, a Na+ dependent OH− transporter, cause congenital hereditary endothelial dystrophy (CHED) and Fuchs’ endothelial corneal dystrophy (FECD), the two most common forms of endothelial degeneration. Along with genetic factors, oxidative stress plays a role in pathogenesis of several corneal diseases. In this study we looked into the role of SLC4A11 in antioxidant stress response in human corneal endothelial cells (HCEnC). We found increased expression of SLC4A11 in presence of oxidative stress. Depletion of SLC4A11 using targeted siRNA, caused an increase in reactive oxygen species, cytochrome c, lowered mitochondrial membrane potential, and reduced cell viability during oxidative stress. Moreover, SLC4A11 was found to be necessary for NRF2 mediated antioxidant gene expression in HCEnC. On the other hand, over expression of SLC4A11 reduces reactive oxygen species levels and increases cell viability. Lastly, CHED tissue specimens show evidence of oxidative stress and reduced expression of NRF2. In conclusion, our data suggests a possible role of SLC4A11 in regulating oxidative stress, and might be responsible for both the etiology and treatment of corneal endothelial dystrophy.

the cells and maintains redox homeostasis 21 . Under normal conditions it is held in the cytoplasm and tightly regulated by Keap1 that causes constant degradation of NRF2 by ubiquitination 22,23 . On activation by oxidative stress and other external stimuli, it undergoes heterodimerization with small Maf proteins and translocates from the cytoplasm to the nucleus, where it binds to antioxidant responsive element 24 and mediates transcription of its target genes which include various antioxidants and detoxification enzymes 21,25,26 . Some of the cytoprotective genes regulated by NRF2 are those of NAD(P)H-quinoneoxidoreductase 1 (NQO1), heme oxygenase 1 (HO-1) and glutathione reductase (GR) 27 .
In this study, we investigated the relationship between SLC4A11 and oxidative stress in both primary and immortalized HCEnC. Using siRNA to knockdown SLC4A11 in HCEnC, we looked into the antioxidant signaling in response to oxidative stress in these cells. Our studies show that depletion of SLC4A11 in corneal endothelial cells generates increased ROS, alters mitochondrial membrane potential and results in impaired NRF2 driven antioxidant signaling. Interestingly, CHED tissue specimens obtained from patients, also exhibit signs of oxidative stress and reduced NRF2 mediated antioxidant response. This study sheds light on physiological function of SLC4A11 during oxidative stress that can lead to the development of important noninvasive therapeutic interventions to prevent corneal endothelial degeneration.

Results
Oxidative stress up-regulates SLC4A11 expression in HCEn and HEK 293 cells. Oxidative stress has been associated with pathogenesis of corneal endothelial dystrophy 28 and other corneal diseases 29 . We have earlier reported that cells expressing mutant SLC4A11 are more prone to oxidative stress 20 compared to cells expressing the wild-type protein. Thus we asked whether SLC4A11 gene itself responds to oxidative stress. We exposed HCEnC, both primary and immortalized cells, and HEK 293 cells to 500 μM of tBH as exogenous source of oxidative stress over a period of 4 h. As shown in Fig. 1, there was a significant increase in the expression of SLC4A11 and NRF2 in both primary (A) and immortalized (B) human corneal endothelial cells. Oxidative stress also induced expression of HO-1, an antioxidant gene regulated by NRF2, by nearly 4 folds in primary cells (Fig. 1a) and about 40 folds in the immortalized cells (Fig. 1b). Transcripts of SLC4A11, NRF2 and HO-1 were also significantly induced in HEK 293 cells by tBH (Fig. 1c). To test that increase of SLC4A11 expression is not tBH specific, we challenged HEK 293 cells with selenite (SN, 10 μM), as an alternative source of oxidative stress 30 . As seen in supplementary Figure S1, SN significantly increased the expression of SLC4A11, NRF2 and HO-1. Together these data indicate that SLC4A11 is an oxidative stress response gene and might have a functional role in redox control.

Depletion of SLC4A11 causes increased ROS generation in HCEnC.
Next we attempted to determine the link between SLC4A11 expression and ROS levels in corneal endothelial cells. In order to deplete the expression of SLC4A11, primary HCEnC were transfected using siRNA against SLC4A11 and ROS was detected by flow cytometry using fluorescent ROS indicator H 2 CFDA. As shown in Fig. 2a, SLC4A11 depletion causes Up-regulation of SLC4A11 in response to oxidative stress. Both primary (a), and immortalized (b) human corneal endothelial cells and HEK 293 (c) cells were exposed to 500 μM tBH for 4 h and fold changes of the expression of SLC4A11, NRF2 and HO-1 were determined by quantitative PCR. These experiments were performed three times.
increase in ROS positive cells by one hour (91.7%), compared to cells transfected with control-siRNA (77.1%). The difference became significant by 2 h where SLC4A11-siRNA transfected cells had higher ROS positive cells (83.7%) compared to that of control-siRNA transfected cells (59.2%). We further determined if SLC4A11 depletion affects the viability of the cells in response to oxidative stress using MTT assays. Figure 2b shows that, cells depleted of SLC4A11 had decreased viability at 4 and 6 h, which become significant by 8 h (p < 0.05). Together, these results demonstrate that depletion of SLC4A11 expression induces intracellular ROS generation and decreases the viability of corneal endothelial cells in the presence of oxidative stress.

Loss of SLC4A11 alters mitochondrial membrane potential in HCEnC.
Oxidative stress is linked to mitochondrial dysfunction 31 , as mitochondria are both producers and targets of reactive oxygen species. To investigate whether depletion of SLC4A11 affects the mitochondrial membrane potential, primary corneal endothelial cells transfected with control or SLC4A11 siRNA were exposed to oxidative stress for 2 h. The cells were then incubated with the cationic dye JC-1, which localizes to normal mitochondrial membrane as red aggregates and turns into green monomer upon membrane depolarization. We found, using immunofluorescence, SLC4A11 depleted cells mostly green (depolarized) by 2 h compared to control-siRNA transfected cells in presence of oxidative stress (Fig. 3a). The ratio of red/green intensity was determined quantitatively using ImageJ software 32 and shows decreased ratio in SLC4A11 depleted cells (Fig. 3b). We further checked the mitochondrial membrane potential of these cells by flow cytometry using TMRE, another cationic dye. TMRE is a mitochondria specific cell permeable dye that readily accumulates in active mitochondria and detects the mitochondrial membrane potential in live cells. Figure 3c shows that upon exposure to oxidative stress over 2 h, SLC4A11 depleted cells have lower percentage of TMRE-positive cells (28.4%), suggesting a loss of mitochondrial membrane potential, in comparison to control cells (64.2%).
Increased levels of cytochrome c and p53 in SLC4A11 depleted HCEnC. When the trans-membrane potential of the mitochondria is affected, cytochrome c is released into the cytoplasm. The release of cytochrome c from the mitochondria initiates the apoptotic pathway in cells 33 . As there was a decrease in the mitochondrial trans-membrane potential in response to oxidative stress in SLC4A11 depleted cells, we examined the level of cytochrome c in these cells on exposure to 250 µM tBH over 2 h. There was an increase in cytochrome c production by 1 h in cells depleted of SLC4A11 (Fig. 3d v), in comparison to control cells (Fig. 3d ii), this became more significant by 2 h in SLC4A11 depleted cells (Fig. 3d vi). The fluorescence intensity was quantitated using ImageJ software 32 and the graph shows the significant reduction in the fluorescence between SLC4A11 depleted and control cells (Fig. 3e). p53 and Bax gene expression was also determined in presence of oxidative stress, and Fig. 3f shows significant increase in p53 (4 fold) and Bax (1.5 fold) expression in SLC4A11 depleted cells compared to control siRNA transfected cells.

SLC4A11 is required for NRF2 mediated antioxidant response in HCEnC.
Oxidative stress is regulated by antioxidant signaling, and mediated by various genes whose expression depends on the transcription factor NRF2. Since expression of both SLC4A11 and NRF2 were up-regulated by oxidative stress (Fig. 1), we evaluated the NRF2 expression in SLC4A11-siRNA transfected cells exposed to oxidative stress. Immortalized human corneal endothelial cells were transfected with SLC4A11-siRNA and the level of SLC4A11 was checked. A significant decrease in SLC4A11 expression was found compared to cells transfected with control-siRNA (Fig. 4a). These cells were then exposed to 500 μM tBH for 4 h, and the gene expressions monitored. NRF2 expression was significantly lower in SLC4A11 depleted cells, compared to control cells under oxidative stress conditions (Fig. 4b). Consistent with low NRF2 expression, SLC4A11 depleted cells showed significantly reduced expression of NRF2 transcriptional target genes HO-1, and NQO1, compared with control cells under oxidative stress (Fig. 4c). The depletion of SLC4A11, however, did not affect the phosphorylation of the IkB and MAPK pathway proteins like ERK, JNK and p38 (Fig. 4d) in immortalized human corneal endothelial cells. Determination of ROS and cell viability in SLC4A11 depleted human corneal endothelial cells. ROS induced by oxidative stress was analyzed by flow cytometry using H 2 CFDA fluorescent probe. HCEnC were transfected with SLC4A11 targeted siRNA or control siRNA and exposed to 250 μM tBH, washed with 1X PBS, and ROS was determined (a). Cell viability was determined using MTT assay (b). Untreated cells were used as controls in each case. These experiments were performed twice.
In order to ensure if primary corneal endothelial cells behave in a similar way, SLC4A11 was depleted from primary cells using siRNA. The depletion of SLC4A11 is confirmed as shown in Fig. 5a. Similar to immortalized cells, we found a significant decrease in expression of NRF2 in these SLC4A11 depleted cells (Fig. 5b) The expressions of antioxidant genes were also checked and as shown in Fig. 5c, there were a significant decrease in expression of HO-1, and NQO1, in SLC4A11-siRNA transfected cells compared to control-siRNA transfected cells. Mitochondrial activity in SLC4A11 depleted human corneal endothelial cells exposed to oxidative stress. HCEnC were transfected with SLC4A11 targeted siRNA or control siRNA and exposed to 250 μM tBH over 2 h, washed with 1X PBS, incubated with JC-1 and mitochondrial membrane potential was determined by microscopy (a), and ratio of red/green fluorescence intensities were quantitated (b). The mitochondrial membrane potential was also determined by flow cytometry after cells were incubated with TMRE following SLC4A11 depletion and exposure to oxidative stress (c). The expressions of cytochrome c were imaged using fluorescence microscopy (d) and mean fluorescence intensities were quantitated (e). The expression of p53 and Bax genes was determined by quantitative PCR (f). Data are the representative of two different experiments, each performed with duplicates.

SLC4A11 overexpression increases cell viability, reduces ROS generation and protein nitration
in cells in response to oxidative stress. As depletion of SLC4A11 causes a reduced expression of antioxidant genes, we wanted to determine the effects of SLC4A11 on the sensitivity of cells to oxidizing agents. In order to do so, HEK 293 cells were transfected with wt-SLC4A11 plasmids (wt-SLC) or empty vectors. For the overexpression assays, HEK 293 cells were used to minimize the effect of endogenous protein expression. The cells were exposed to 250 μM tBH or 10 μM SN for defined time periods, after which, the cell viability was determined using MTT assay. Cells transfected with wt-SLC were significantly more viable and resistant to tBH (Fig. 6a) or SN (Supplementary Fig. S2) mediated cell death while cells expressing empty vectors were not so.  Next, we determined the effect of SLC4A11 overexpression on the ROS level by transfecting cells with wt-SLC or empty vectors and exposing them to 250 μM tBH for 2 h. The ROS level was detected by a quantitative assay of carboxy-DCF fluorescence. Non-transfected cells, treated in similar way, were used as the negative control. As shown in Fig. 6b, cells transfected with wt-SLC caused reduced levels of ROS, although not significant, compared to cells transfected with empty vectors in presence of oxidative stress. We then wanted to check if SLC4A11 overexpression could inhibit protein nitration in these cells. Cells transfected with wt-SLC or empty vectors were exposed to tBH for 2 h, washed and stained with anti-3-nitrotyrosine (3-NT) antibody and imaged under a fluorescent microscope. Cells transfected with the empty vector has increased expression of 3-NT as evident from Fig. 6c. Taken together, these data imply that overexpression of SLC4A11 protects cells against oxidative stress.
Oxidative stress in CHED patients. Although presence of oxidative stress has been manifested in several corneal disorders including FECD and keratoconus 28,34,35 , there are no reports of the same for CHED. Our next experiments were on tissue specimens that were obtained from patients with CHED (n = 5), who underwent penetrating keratoplasty (PK) at our Institute. As shown in Fig. 7a, specific staining for 3-NT was observed in the corneal tissue specimens of the CHED patients (i, ii), while there was minimal or no staining for 3-NT in control cadaveric corneas (iii, iv). We next measured expression of oxidative stress markers in HCEnC-DM complex isolated from 5 CHED patients who underwent Descemet's stripping endothelial keratoplasty (DSEK). While the expression of SLC4A11 (mutated in these tissues) did not show a major change, there was significant reduction in expression of NRF2 (Fig. 7b) accompanied by reduced expression of HO-1, ferritin (FRT), GR, and to some extent of NQO1 in CHED patients (Fig. 7c). Together, these data provide strong evidence for oxidative stress and reduced antioxidant genes level in CHED patients.

Discussion
Oxidative stress has been increasingly reported to be involved in the pathogenesis for several ocular diseases 36 including corneal endothelial dystrophies 13,28 . There are earlier reports regarding oxidative stress induced degeneration of cells in corneal endothelial dystrophies. Although the incidence of CHED is low in western countries, in India it is accounted for 21 percent for all pediatric keratoplasties 37 . On the other hand, FECD is more common in North America and account for 10% of all corneal transplants 38 . The mutations in SLC4A11 gene have been found associated for both FECD and CHED [2][3][4] . SLC4A11 mutations cause dysfunction of endothelial cells leading to thickening of the DM and corneal edema. Recently, we found that cells expressing mutant SLC4A11 are more Figure 6. Overexpression of SLC4A11 in HEK 293 cells. Cells were transfected with wild-type. SLC4A11 plasmid or empty vectors and exposed to 250 µM of tBH. Cell viability (a), ROS generation (b) and Nitrotyrosine staining (c) were determined. For nitrotyrosine staining, cells were stained using antinitrotyrosine antibody, followed by Alexafluor 488 secondary antibody. Cells were imaged both under bright field and fluorescence and images were captured by Olympus IX73 using 20X objective. The experiments a, and b, were done three times in duplicates and experiment c was performed two times.
prone to oxidative stress 20 . SLC4A11 null mice also exhibit signs of CHED such as corneal edema, swelling of endothelial cells and thickening of DM 39 .
In this study we report SLC4A11 as an oxidative stress response gene required for proper NRF2 activation under oxidative stress conditions in human corneal endothelial cells. We found increased ROS level and decreased cell viability in SLC4A11 deficient cells in presence of oxidative stress. We have earlier reported increased accumulation of ROS in cells expressing mutant SLC4A11 20 . Higher levels of H 2 O 2 were earlier detected in FECD corneal endothelial cells 40 . Liu et al. have previously reported decrease in cell viability in SLC4A11-shRNA transfected human corneal endothelial cells 7 to 10 days post transfection compared to control cells 16 . Although ROS is generated in response to various stimuli, increased levels of ROS can cause damage to DNA, lipids or proteins and is also an important contributor for mitochondrial dysfunction. In accordance to our earlier report of lowered mitochondrial membrane potential in cells expressing mutant SLC4A11, we found altered mitochondrial membrane potential in SLC4A11 depleted cells as evident from green monomers formed from incubation with JC-1 dye. Altered mitochondrial structure and decreased activity were previously reported in corneal fibroblasts obtained from patients with type 2 granular corneal dystrophy 41 . High occurrence of mitochondrial DNA lesion was also found in the DM of FECD patients compared to the controls that indicate incompetent mechanism of removal of damaged mitochondrial DNA in FECD 18 . Increased cytochrome c, caspase activation and apoptotic cell death, are associated with loss of mitochondrial membrane potential. In our study, we detected increased cytochrome c expression along with increase in the expression of p53 and Bax, a pro apoptotic gene, in cells depleted of SLC4A11. This is in agreement with previous studies that showed increased level of p53 40 and Bax 16 in FECD and SLC4A11 depleted corneal endothelial cells respectively. Increased caspase-3 activation in cells expressing mutant SLC4A11 has been reported earlier 20 along with caspase 3, 7, and 9 in SLC4A11 depleted human corneal endothelial cells 16 . p53 has been reported to play a role in antioxidant signaling by suppression of NRF2 activation and antioxidant gene expression 42 . Since NRF2 is the master regulator for antioxidant signaling 43 , we checked the expression of NRF2 and antioxidant genes in SLC4A11 depleted cells. We saw significant reduction in the expression of NRF2 and NRF2 mediated antioxidant genes in cells depleted of SLC4A11. Bitar et al. have reported reduced NRF2 translocation and degradation of DJ-1, a NRF2 stabilizer, in FECD corneal endothelial cells 19 . We also observed decreased NRF2 and antioxidant gene expression in CHED tissue specimens which is similar to what has been observed for FECD 28 . Taken together, our results indicate a generalized down-regulation of the antioxidant genes in cells with depleted SLC4A11 expression. This is important and gives an indication that inducing the expression of antioxidant genes might help in overcoming the oxidative stress in corneal endothelial cells. Very recently, there is a promising report where glafenine has been able to correct the trafficking defect of SLC411 mutants and increased its functionality 44 . SLC4A11 was initially reported as a borate transporter and as borate has been shown to have mitogenic effect mediated by MAPK 5 , we checked if SLC4A11 depletion has any effect on MAPK and NfkB pathways. Our results show that there was no difference in the activation of NF-κB and MAPK proteins in cells depleted of SLC4A11 compared to control cells. There are earlier reports stating that over expression of SLC4A11, then known as NaBC1, in HeLa cells helps in cell proliferation in presence of borate 5 . We also over expressed SLC4A11 in HEK 293 cells and looked into the response of these cells towards oxidative stress and found that there was significant increase in cell viability and reduced ROS in these cells compared to control cells. SLC4A11 overexpression also reduced the levels of protein nitration in HEK 293 cells. The mechanism by which SLC4A11 regulates antioxidant signaling is in scope for future study and remains to be determined. We hypothesize that a positive regulatory feed-back loop might be present between SLC4A11 and NRF2 that helps to mobilize cell's antioxidant machinery. SLC4A11 might regulate optimum activation of NRF2 in corneal endothelial cells by direct interaction or by promoting protein stabilization.
Increased oxidative stress is a prominent feature in corneal endothelial dystrophies, particularly for FECD. However, there is no direct evidence for the presence of oxidative stress in CHED which results from mutations of SLC4A11. We show here that increased oxidative stress is present in the corneal sections of CHED patients. Nitrotyrosine staining has been observed earlier in FECD and keratoconus corneas, but was absent in bullous keratopathy, although it has clinically similar signs as FECD 29 . To our knowledge, this makes our study to be the first to report the presence of oxidative stress in CHED. We also observed decreased expression of NRF2 and NRF2 target genes in CHED patients. Since SLC4A11 is induced by oxidative stress, reduced NRF2 expression may explain presence of oxidative stress in CHED patients due to diminished or total loss of SLC4A11 functions owing to mutations of SLC4A11. Although there is significant reduction of a few antioxidant genes in CHED patients, we found that is not the case for NQO1 which suggest that there might be some redundant mechanism of NQO1 regulation involving other transcription factors 45 . Increased oxidative DNA damage, decreased expression of NRF2 and peroxiredoxins has also been reported in FECD tissues 46 . Although mutations in SLC4A11 are associated with both CHED and FECD, one is autosomal recessive 2 while the other is autosomal dominant disease 3 . Moreover, apart from SLC4A11, several other genes like ZEB1 47 , COL8A2 48 , TCF4 49 , LOXHD1 50 have also found to be associated with FECD. Thus, defect in antioxidant signaling due to SLC4A11 mutation might affect in a diverse pattern, which can also be the reason for delayed manifestation of the disease in FECD. It will be highly interesting to find out the extent of oxidative stress and difference in the expression of antioxidant gene between these two forms of corneal endothelial dystrophy.
Taken together, our study suggests a novel role of SLC4A11 in oxidative stress response in HCEnC. These findings can be made therapeutically beneficial to prevent or at least delay degeneration of corneal endothelium by enhancing antioxidant defenses and protecting the cells from reactive oxygen species. Currently, surgical interventions in form of DSEK or PK are the only modes of treatment for corneal endothelial dystrophy. In this context, our findings that deregulated NRF2 dependent antioxidant signaling due to SLC4A11 depletion may provide new possibilities for noninvasive therapeutic interventions.

Materials and Methods
Tissue procurement. Tissue procurement for research purposes was approved by the Institutional Review Board of L.V. Prasad Eye Institute, Hyderabad, India and the research followed the tenets of the Declaration of Helsinki. The tissue specimens were collected after obtaining informed consent from the patients.

Culture of primary human corneal endothelial cells. Primary corneal endothelial cells were cultured
from human donor corneas sourced from the Ramayamma International Eye Bank (LV Prasad Eye Institute, Hyderabad, TS). Briefly, the DM was carefully peeled, chopped into smaller pieces and subjected to collagenase (1 mg/ml, Sigma, St. Louis, MO) digestion for 2 h at 37 °C. The cells were collected by centrifugation following trypsin-EDTA (0.25%) treatment for 5 min. The cells were cultured on FNC (Athena, Baltimore, MD) pre-coated plastic dishes in medium containing Opti-MEM, 8% FBS, 0.08% chondroitin sulphate, 200 mg/L calcium chloride, 5 ng/ml hEGF and 1X antibiotic/antimycotic. Cells were cultured until they reached confluence and were split using trypsin-EDTA 51  RNA Isolation and Quantitative PCR. Quantitative real time PCR (QPCR) was done to evaluate mRNA expression of SLC4A11, NRF2, and several other genes in human corneal endothelial cells. Total RNA was isolated using TRIzol (Invitrogen, Carlsbad, CA) and was reverse transcribed using cDNA synthesis kit according to the manufacturer's details (Eurogentec, Belgium). QPCR was performed on ABI 7900HT (Applied Biosystems, Foster City, CA) using the Sybr green PCR master mix (Thermo Fisher, Waltham, MA). Relative quantities of mRNA expression of different genes were normalized using the 2 −ΔΔCt method using 18 s rRNA as the housekeeping gene. The sequences of the primers used are in Table 1. supplemented with 5% heat-inactivated fetal bovine serum and 0.5% penicillin streptomycin at 37 °C. Cells were transiently transfected with SLC4A11 plasmid or empty vectors using lipofectamine 3000 as descried earlier 20 . The wild-type SLC4A11 plasmid was a kind gift from Dr. J.A. Bonanno, Indiana University, USA 6 .

RNA Interference
Determination of reactive oxygen species by flow cytometry. For intracellular ROS determination in HCEnC, transfected cells exposed to tBH or media for defined time periods were washed in 1X PBS and incubated with 10 μM dicholorodihydrofluoresceindiacetate (H 2 CFDA; Invitrogen) for 20 min in dark at 37 °C. Cells were further washed in 1X PBS, and immediately analyzed by flow cytometry using BD Facsaria II (BD Biosciences, San Jose, CA). Immunocytochemistry analysis. For immunocytochemistry analysis, 1 × 10 4 human corneal endothelial cells were seeded on coverslips and transfected with control or SLC4A11 siRNA as described above. The cells were then exposed to oxidative stress, washed in PBS and fixed in 4% paraformaldehyde for 15 min. For immunofluorescence analysis of pIkB, MAPK, Nitrotyrosine, or cytochrome c, staining was done as described 53 . Cells were stained with rabbit anti-pERK, anti-pJNK, anti-pp38 or anti-pIkB antibodies (1:100, Cell Signaling Technology, Beverly, MA), mouse anti-nitrotyrosine (1:100, Novus Biologicals, Littleton, CO), or mouse anti-cytochrome c antibody (1:100, BioLegend, SanDiego, CA) for 45 min followed by Alexafluor 488 secondary antibody (1:500, Molecular probes, Eugene, OR) for 30 min. Images were captured on a fluorescent microscope (Olympus IX73) using a 20X objective. The fluorescence intensities were measured by ImageJ software.
Determination of mitochondrial membrane potential. Primary human corneal endothelial cells were cultured overnight on coverslips, transfected with control or SLC4A11 siRNA and exposed to 250 μM of tBH for defined time periods. The cells were then washed with 1X PBS and incubated with cationic dye JC-1 (final concentration 2 μM; Santa Cruz, Dallas, TX) for 20 min at 37 °C. The cells were washed with 1X PBS and coverslips were inverted onto slides and live cells were imaged to detect JC-1 staining under fluorescent microscope (Olympus IX73) using 20X objective. All images were acquired at equal exposure time to compare the fluorescence. The red and green fluorescence intensities of the images were measured by ImageJ using ROI Manager tool. To analyze the mitochondrial membrane potential by flow cytometry, cells after being exposed to oxidative stress were incubated with tetramethylrhodamine ethyl ester (TMRE; final concentration 200 nM; Santa Cruz, Dallas, TX) for 20 min at 37 °C and fluorescence was determined by flow cytometry using BD Facsaria II (BD Biosciences, San Jose, CA).
Cell viability Assay. Cell viability was determined quantitatively using a 3-[4,5-Dimethylthiazole-2yl]-2,5-diphenyltetrazolium bromide (MTT) (Invitrogen) cleavage assay. In brief, 1 × 10 4 cells were plated in 96 well plates and transfected with SLC4A11-siRNA, control siRNA or SLC4A11 plasmid, empty vectors. Cells were exposed to tBH or SN as mentioned above. After washing with 1x PBS, cells were incubated with 5 mg MTT per ml of culture media for 2 h at 37 °C. The supernatant was discarded and stop solution (DMSO) was added to each well to dissolve the formazone crystals. Absorbance was measured at 595 nm in SpectraMax M3 Reader (Molecular Device).
Nitrotyrosine Staining of Tissues. Archived formalin fixed, paraffin embedded corneas with confirmed clinical diagnosis of CHED (n = 5) were obtained along with three normal cadaveric corneas unsuitable for transplantation. 5 μm sections were mounted on slides and were deparafinized, hydrated and permeabilized, then rinsed with 1X PBS. The sections were incubated with mouse anti-3 nitrotyrosine antibody (1:50, Novus Biologicals, Littleton, CO), for 1 h, washed and reincubated with Alexafluor 488 secondary antibody (1:250, Molecular probes, Eugene, OR) for 1 h and further washed. The sections were counterstained with DAPI (Abcam, Cambridge) and observed under fluorescent microscope (Olympus IX73) using 20X objective and imaged using Olympus DP71 camera.
Statistics. Statistical analysis was performed using either an unpaired t test or one way Anova (Prism, GraphPad Software). p values less than 0.05 was considered significant.  Table 1. List of primers.