Exogenous NAD+ decreases oxidative stress and protects H2O2-treated RPE cells against necrotic death through the up-regulation of autophagy

Increased oxidative stress, which can lead to the retinal pigment epithelium (RPE) cell death by inducing ATP depletion and DNA repair, is believed to be a prominent pathology in age-related macular degeneration (AMD). In the present study, we showed that and 0.1 mM nicotinamide adenine dinucleotide (NAD+) administration significantly blocked RPE cell death induced by 300 μM H2O2. Further investigation showed that H2O2 resulted in increased intracellular ROS level, activation of PARP-1 and subsequently necrotic death of RPE cells. Exogenous NAD+ administration significantly decreased intracellular and intranuclear ROS levels in H2O2-treated RPE cells. In addition, NAD+ administration to H2O2-treated RPE cells inhibited the activation of PARP-1 and protected the RPE cells against necrotic death. Moreover, exogenous NAD+ administration up-regulated autophagy in the H2O2-treated RPE cells. Inhibition of autophagy by LY294002 blocked the decrease of intracellular and intranuclear ROS level. Besides, inhibition of autophagy by LY294002 abolished the protection of exogenous NAD+ against H2O2-induced cell necrotic death. Taken together, our findings indicate that that exogenous NAD+ administration suppresses H2O2-induced oxidative stress and protects RPE cells against PARP-1 mediated necrotic death through the up-regulation of autophagy. The results suggest that exogenous NAD+ administration might be potential value for the treatment of AMD.

Western Blot. For Western blot, cells were harvested and centrifuged at 800g for 5 minutes to pellet the cell debris. Cells were lysed in RIPA buffer containing proteinase and phosphatase inhibitors. Protein concentrations of the lysates were measured with a BCA protein assay kit (Pierce, Rockford, IL, USA). The lysates were boiled in 4 × sodium dodecyl sulfate loading buffer. 30 μg protein samples were analyzed by SDS-polyacrylamide gel electrophoresis. Then, proteins were transferred to polyvinylidene fluoride membranes (Millipore, Bedford, MA, USA). The blots were blocked for 1 hour in non-fat milk in tris-buffered saline (TBS) and incubated in primary anti-body overnight at 4 °C. The blots were then rinsed thoroughly with TBS and incubated in a 1:1000 dilution of horseradish peroxidase conjugated secondary antibody in TBS for 1 hour at room temperature. HRP signals were developed by using enhanced chemiluminescence (ECL) reagent (Pierce, Rockford, IL, USA) and exposure to X-ray film. Different exposure time was used for each membrane to avoid over-exposure of the Scientific RepoRts | 6:26322 | DOI: 10.1038/srep26322 bands. The image analysis was performed using BioRad image instrument with multi-analyst software. Rabbit anti-LC3A/B (1:500, Cell Signaling Technology, Beverly, MA, USA), rabbit anti-PARP-1 (1:500, Cell Signaling Technology, Beverly, MA, USA), rabbit anti-Caspase 3 (1:500, Cell Signaling Technology, Beverly, MA, USA), mouse anti-GAPDH (1:500, Sigma, St. Louis, MO, USA) were used as primary antibodies and horseradish peroxidase (HRP)-conjugated goat anti-mouse or goat anti-rabbit IgG (1:1000, Kangcheng, Shanghai, China) as secondary antibodies.

Detection of autophagy.
For the detection of autophagy, cells were seeded in 96-well plate at a density of 5 × 10 3 cells/well with or without 300 μM H 2 O 2 or 0.1 mM NAD + treatments and incubated at 37 °C for 24 hours. Autophagy was visualized by LC3 staining using autophagy tandem sensor GFP-LC3B kit (Life Technologies, Carlsbad, CA, USA) according to the manufacturer's instructions and then images were taken using the Operetta high-content imaging system (PerkinElmer, Waltham, MA, USA) and analyzed using Harmony Software (PerkinElmer, Waltham, MA, USA). Each assay was repeated three times.

Detection of ROS.
For the detection of intracellular ROS, cells were seeded in 96-well plate at a density of 5 × 10 3 cells/well with or without 300 μM H 2 O 2 or 0.1 mM NAD + and incubated at 37 °C for 24 hours. Intracellular ROS was detected with CellROX ® Reagent (Life Technologies, Carlsbad, CA, USA) according to the manufacturer's instructions and then images were taken using the Operetta high-content imaging system (PerkinElmer, Waltham, MA, USA) and analyzed using Harmony Software (PerkinElmer, Waltham, MA, USA) to determine the ROS levels. Each assay was repeated three times.
Detection of apoptosis and necrotic death. For the detection of apoptosis and necrotic death, cells were seeded in 96-well plate at a density of 5 × 10 3 cells/well with or without 300 μM H 2 O 2 or 0.1 mM NAD + and incubated at 37 °C for 24 hours. Apoptosis and necrotic death was detected with annexin V fluorescence apoptosis detection kit (BD bioscience, Bedford, MA, USA) and then images were taken using the Operetta high-content imaging system (PerkinElmer, Waltham, MA, USA) and analyzed using Harmony Software (PerkinElmer, Waltham, MA, USA) to determine the proportion of apoptotic cells or necrotic cells. Each assay was repeated three times.
Statistical Analysis. All experiments were repeated at least three times independently. All data were expressed as means ± SEM and statistical analysis was performed using GraphPad Prism5. The data were analyzed by one-way ANOVA with post hoc test or unpaired t test, as appropriate (*p < 0.05, **p < 0.01, and ***p < 0.001).  (Fig. 1A,B). Then, we also investigated the effect of different concentration of NAD + on the viability of RPE cells by CCK8 assay to determine the concentration of NAD + we treated on the RPE cells. The results showed that no significant difference was observed in the RPE cells treated without NAD + and in the RPE cells treated with 0.01 mM NAD + , and in the RPE cells treated with 0.1 mM NAD + . However, the cell viability was significantly decreased in the cultured RPE cells treated with 1 mM or 20 mM NAD + compared that in the RPE cells treated without or with 0.01 mM, or 0.1 mM NAD + (Supplementary Figure S1A, H 2 O 2 treatment did not affect NAD + biosynthesis. NAD + depletion reportedly causes cell death under cytotoxic stress 19 . Given the protection of exogenous NAD + against the death of RPE cell induced by H 2 O 2 , we supposed that H 2 O 2 might induce RPE cell death by affecting NAD + production. To test this hypothesis, we first examined the gene expression of seven enzymes involved in NAD + biosynthesis in the RPE cells treated with or without 300 μM H 2 O 2 using quantitative RT-PCR. The PCR results showed that only the expression of NMNAT1 was significantly decreased in the RPE cells treated with 300 μM H 2 O 2 compared with that in the RPE cells treated without H 2 O 2 . No significant differences were observed in the expressions of other six enzymes in the RPE cells between H 2 O 2 -treated group and control group ( Fig. 2A). Furthermore, we measured NAD + levels and NAD + /NADH ratios in the RPE cells treated with or without 300 μM H 2 O 2 . The result showed that no significant differences were observed in the NAD + levels, NADH levels or NAD + /NADH ratio in the RPE cells between H 2 O 2 -treated group and control group (Fig. 2B-D). The above results suggest that H 2 O 2 -induced RPE cell death has not been through the inhibition on NAD + biosynthesis or NAD + depletion.

NAD + reduced intracellular and intranuclear ROS levels in the H 2 O 2 treated-RPE cells.
Increased oxidative stress is thought as the major cause for RPE cell death. Then, we speculated that the protection of NAD + against H 2 O 2 -induced RPE cell death may associated with the alteration of intracellular oxidative stress. To test this hypothesis, we investigated the effect of NAD + on the alteration of intracellular reactive oxygen  (Fig. 3A,B). Furthermore, we analyzed the intranuclear ROS level in the RPE cells. The result showed that treatment of 300 μM H 2 O 2 significantly increased intranuclear ROS level in the RPE cells. While, intranuclear ROS level significantly decreased in the RPE cells treated with 0.1 mM NAD + and 300 μM H 2 O 2 compared with that in RPE cells treated with 300 μM H 2 O 2 alone (Fig. 3A,C).

NAD + protected RPE cells from H 2 O 2 -induced necrotic death by inhibiting the activation of PARP-1.
Hydrogen peroxide has been shown to induce apoptosis in RPE cells 12,20,21 . However, recent study reports that necrotic cell death is a major mechanism for RPE cell death in response to oxidative stress 13 . To determine the effect of NAD + on the two major types of cell death induced by H 2 O 2 , we first performed annexin V/propidium iodide (PI) staining assay detected by high-content screening system (HCS). Cells were judged to be viable if double negative, as early apoptotic if positive for annexin V alone, as necrotic if positive for PI and as necrotic or late apoptotic if double positive. The result showed that no significant difference in the percentage of cells positive for Annexin V was observed in the RPE cells treated with 0.1 mM NAD + and 300 μM H 2 O 2 compared with that in the RPE cells treated with 300 μM H 2 O 2 alone, or compared with that in the control RPE cells for 24 hours (Fig. 4A,B). However, the percentage of cells positive for PI showed a significant increase in the RPE cells treated with 300 μM H 2 O 2 for 24 hours. While, the percentage of cells positive for PI was significantly decreased in the RPE cells treated with 0.1 mM NAD + and 300 μM H 2 O 2 for 24 hours compared with that in the RPE cells treated with 300 μM H 2 O 2 alone. However, no significant difference in the percentage of cells positive for PI was observed in the RPE cells treated with 0.1 mM NAD + and 300 μM H 2 O 2 and in the control RPE cells for 24 hours (Fig. 4A,C). Similarly, the percentage of cells positive for PI and Annexin V was also significantly increased in the RPE cells treated with 300 μM H 2 O 2 alone compared with that in the RPE cells treated with 0.1 mM NAD + and 300 μM H 2 O 2 or in the control RPE cells for 24 hours. No significant difference in the percentage of cells positive for PI and Annexin V was observed in the RPE cells treated with 0.1 mM NAD + and 300 μM H 2 O 2 and in the control RPE cells for 24 hours (Fig. 4A,D).
Poly(ADP-ribose) polymerase-1 (PARP-1) activation is thought as a hallmark of oxidative stress-induced necrotic cell death 22 . To determine the effect of NAD + on the PARP, we investigated the expression of PARP-1 and its active form by western blot. The result showed that the expression of cleaved PARP-1 (active form of PARP-1) significantly increased in RPE cells treated with 300 μM H 2 O 2 compared with that in the control RPE cells. However, the expression of cleaved PARP-1 significantly decreased in the RPE cells treated with 0.1 mM NAD + and 300 μM H 2 O 2 compared with that in the RPE cells treated with 300 μM H 2 O 2 alone. No significant difference was observed in the expression of cleaved PARP-1 between the RPE cells treated with 0.1 mM NAD + and 300 μM H 2 O 2 and control RPE cells. Besides, no significant difference was observed in the expression of PARP-1 among the three groups ( Fig. 4E-G). Furthermore, we investigated the expression of caspase-3, apoptosis-related marker,    by the western blot assay. The result showed that no obvious alterations in the expressions of caspase-3 and cleaved caspases-3 were observed among the three groups (Fig. 4E,H,I).
The above results demonstrated that inhibition of NAD + on the oxidative stress-induced activation of PARP-1 protected RPE cells from necrotic death. NAD + up-regulated autophagy in the RPE cells treated with H 2 O 2 . Autophagy reportedly suppresses necrotic cell death 22 . To identify whether autophagy is involved in the protection of NAD + against H 2 O 2 -induced RPE cells necrotic death, we first transfected RPE cells with autophagy sensor LC3B-GFP to examine the formation of autophagosome by HSC. The result showed that the RPE cells treated with 0.1 mM NAD + and 300 μM H 2 O 2 showed significant increases in GFP-LC3B autophagic puncta compared with that in RPE cells treated with 300 μM H 2 O 2 alone and in the control RPE cells (Fig. 5A,B). To confirm the positive effect of autophagy, we had treated RPE cell with rapamycin, a well-known autophagy inducer, which was also used as a positive control for autophagy. The result showed that GFP-LC3B autophagic puncta significantly increased in RPE cells treated with rapamycin (Supplementary Figure S2). Furthermore, we investigated the activation of widely used autophagosome marker LC3B by western blot. The result showed that in H 2 O 2 -and NAD + -treated RPE cells, the ratio of LC3B-II/LC3B-I significantly increased compared with that in the RPE cells treated with 300 μM H 2 O 2 alone and in the control RPE cells (Fig. 5C,D). In addition, we characterized the autophagic response of RPE cells using transmission electron microscopy (TEM). In accordance with the result of HSC, the TEM result showed that number of autophagosome significantly increased in the RPE cells treated with 0.1 mM NAD + and 300 μM H 2 O 2 compared with that in the RPE cells treated with 300 μM H 2 O 2 alone and in the control RPE cells (Fig. 5E). These results demonstrated that NAD + treatment enhanced autophagy in the RPE cells exposed to H 2 O 2 .

Blocking autophagy inhibited the decrease of intracellular and intranuclear ROS level mediated by NAD + in the H 2 O 2 treated-RPE cells.
Based on the results described above, we speculated that the influence of NAD + on the decrease of ROS might be mediated by its up-regulation on autophagy. To test this hypothesis, we first examined the influence of LY294002, an inhibitor of autophagy, on the viability of REP cells treated with 0.1 mM NAD + and 300 μM H 2 O 2 by CCK-8 assay. The result showed that the viability of REP cells treated with 300 μM H 2 O 2 , 0.1 mM NAD + and 15 μM LY294002 significantly decreased compared with the viability of REP cells treated with 0.1 mM NAD + and 300 μM H 2 O 2 and the viability of REP cells treated with 15 μM LY294002 alone (Fig. 6A,C). Furthermore, we investigated the influence of LY294002 on the autophagy with autophagy sensor LC3B-GFP. The result showed that the addition of 15 μM LY294002 into the RPE cells treated with 0.1 mM NAD + and 300 μM H 2 O 2 significantly decrease GFP-LC3B autophagic puncta compared with that in the RPE cells treated with 300 μM H 2 O 2 and 0.1 mM NAD + and that in the RPE cells treated with 15 μM LY294002 alone (Fig. 6B,D).
Then, we investigated the influence of LY294002 on the intracellular and intranuclear ROS level in the RPE cells treated with 300 μM H 2 O 2 , 0.1 mM NAD + . The result showed that the addition of 15 μM LY294002 into the RPE cells treated with 0.1 mM NAD + and 300 μM H 2 O 2 significantly increased intracellular and intranuclear ROS level compared with that in the RPE cells treated with 300 μM H 2 O 2 and 0.1 mM NAD + or that in the REP cells treated with 15 μM LY294002 alone (Fig. 6E-G). The results indicated that NAD + decreased intracellular and intranuclear ROS level in H 2 O 2 -induced RPE might mediate through the up-regulation of autophagy. Furthermore, to determine the effect of inhibition of autophagy on the activation of PARP-1, we investigated the expression of PARP-1 by western blot. The result showed that addition of 15 μM LY294002 into the RPE cells treated with 0.1 mM NAD + and 300 μM H 2 O 2 significantly increased the expression of cleaved PARP-1 (active form of PARP-1) compared with that in the RPE cells treated with 300 μM H 2 O 2 and 0.1 mM NAD + or that in the REP cells treated with 15 μM LY294002 alone. No significant difference was observed in the expression of cleaved PARP-1 between the RPE cells treated with 0.1 mM NAD + and 300 μM H 2 O 2 and the REP cells treated with 15 μM LY294002 alone. Similarly, no significant difference was observed in the expression of PARP-1 among the three groups ( Fig. 6H-J).

Discussion
Clinically, treatment of AMD remains insufficient. NAD + has been proposed to be a novel and inexpensive cytoprotective agent for its protection against cell death induced by oxidative stress. However, the effect of NAD + on the oxidative stressed-RPE cell death is unknown. The questions are addressed in the present study. NAD + improved the viability of RPE cell treat with H 2 O 2 . NAD + , a hydride-accepting and hydride-donating coenzyme, has central roles in cellular metabolism and energy production. Exogenous NAD + supplement has been reported to protect cell from death induced by oxidative stress 17,23 . Besides, study on the light-induced retinal damage shows that NAD + treatment attenuates photoreceptor degeneration and RPE cell death triggered by zinc toxicity, a result from pathologic light exposure 24,25 . All these studies suggest that NAD + might be a novel cytoprotective agent against oxidative stress or zinc toxicity. In line with these studies, our results showed that 0.1 mM NAD + administration to 300 μM H 2 O 2 -treated RPE cells significantly improved the cell viability (Fig. 1).
A number of studies have shown that oxidative stress or zinc toxicity can induce DNA damage and cell death by causing intracellular NAD + depletion or ATP depletion [24][25][26][27] . In the present study, neither NAD + depletion by measurement of intracellular NAD + levels and NAD + /NADH ratios nor inhibition of NAD + synthesis by measurement of gene expression of seven enzymes involved in NAD + biosynthesis was detected in the RPE cells treated with 300 μM H 2 O 2 (Fig. 2). In vitro study has shown that NAD + treatment can prevents the decline of ATP levels 28  NAD + protected RPE cell from necrotic death induced by H 2 O 2 . Oxidative stress is considered to play significant role in the pathogenesis of AMD. It triggers RPE cells progressive damage by accumulation of ROS which contributes to protein mis-folding and evoking functional abnormalities during RPE cellular senescence 29 . In the present study, we showed that NAD + treatment significantly decreased not only intracellular ROS   (Fig. 3). Therefore, it is possible that the protection of NAD + on the RPE cell death may associate with its decline in intracellular ROS. Previous studies implicate that apoptosis is a main mechanism for oxidative stress-induced RPE cell death 30,31 . However, recent study of the systematic analysis of RPE cell death induced by oxidative stress indicates that necrotic death is a major type of cell death in RPE cells in response to oxidative stress 13 . In the present study, annexin V/PI staining assay detected by HCS showed that the percentage of cells positive for PI staining significantly increased in the RPE cells treated with 300 μM H 2 O 2 for 24 hours. Also, we investigated the expression of caspase-3, apoptosis-related marker, by the western blot assay. The results showed that no obvious alterations in the expressions of caspase-3 and cleaved caspases-3 were observed. Our result further supports the idea that necrotic cell death is a major death form for RPE in response to oxidative stress (Fig. 4).
Poly (ADP-ribose) polymerase-1 (PARP-1), chromatin-binding and transcription-related protein, reportedly mediates oxidative stress-induced cellular injury and necrotic cell death 32 . In the present study, we showed that the expression of cleaved PARP-1 significantly increased in RPE cells treated with 300 μM H 2 O 2 compared with that in control RPE cells. No significant difference in the expression of PARP-1was observed between RPE cells without or with H 2 O 2 treatment. The results suggest that oxidative stress induced the necrotic death of RPE cells mediated by activation of PARP-1.
To determine the protection of NAD + against necrotic death of RPE cells, we investigated the effect of NAD + on the activation of PARP-1 and necrotic death of RPE cells induced by oxidative stress. The results showed that 0.1 mM NAD + significantly decreased the necrotic death of RPE cell induced by 300 μM H 2 O 2. In addition, western bolt result showed that NAD + significantly decreased the expression of cleaved PARP-1. The results suggested that exogenous NAD + might block the oxidative stress-induced activation of PARP-1 and subsequently necrosis of RPE cells.

Up-regulated autophagy by exogenous NAD + protected RPEC cells from death induced by
Autophagy, a highly conserved cellular degradation pathway for the clearance of damaged or superfluous proteins and organelles, can be induced by various cellular factors and multiple cellular stresses 33 . Although accumulation of cellular oxidative stress and increased generation of ROS have been proposed to induce autophagy [34][35][36] , recent study shows that the increased oxidative stress/ROS does not result in stimulation of autophagy or mitophagy in cardiomyocytes 37 . In the present study, we also did not detect any induction of autophagy in the RPE cells treated with H 2 O 2 through the assays of LC3B immunoblotting, LC3B-GFP puncta formation and autophagosome formation by TEM (Fig. 5). The results suggested that the increased oxidative stress/ROS might not be essential for the induction of autophagy in cultured RPE cells treated with H 2 O 2 . However, the increased ratio of LC3B-II/LC3B-I, increased GFP-LC3B positive cells and autophagosome structure were observed in cultured RPE cells treated with 0.1 mM NAD + and 300 μM H 2 O 2 . The results suggested that exogenous NAD + induced autophagy in the H 2 O 2 treated-RPE cells. Mammalian target of rapamycin (mTOR) pathway has been known as a key regulator of autophagy 38,39 . Studies have been shown that SIRT1 positively regulates autophagy by inhibition of mTOR pathway 40,41 . However, in our study, no significant difference in the SIRT1 expression was observed in the RPE cells treated with 0.1 mM NAD + and 300 μM H 2 O 2 compared with that in the RPE cells treated with 300 μM H 2 O 2 alone, or in the control RPE cells for 24 hours (data not shown). Study on cardiac hypertrophy shows that exogenous NAD + elevates cellular NAD + levels and activates SIRT3. SIRT3 activation stabilizes the activity of the LKB1-AMPK signaling pathway and blocks activity of mTOR 42 . Whether the induction of autophagy by exogenous NAD + in the H 2 O 2 -treated RPE cells is through activation of SIRT3 and subsequently the inhibitory effect on mTOR pathway needs further investigated.
Autophagy is implicated to play pro-survival function in necrotic cell death 22 . A recent report shows that up-regulated autophagy protects cardiomyocytes from oxidative stress-induced toxicity 37 . Besides, autophagy has been proposed to serve to reduce ROS/oxidative stress level by removal of damaged mitochondria or oxidized proteins 43,44 . It is postulated that a breakdown in the recycling capacity of autophagy may be associated with accumulation of proteins and damaged organelles which are a general observation in the aging RPE as well as in AMD 45 . In line with those reports, the present study showed that treatment with LY294002, inhibitors of autophagy, resulted in significant decrease in cell viability in the RPE cell treated with 0.1 mM NAD + and 300 μM  (Fig. 6). Our results suggested that up-regulation of autophagy might be the major mechanism underlying the protection of NAD + on the RPE cell against necrotic death induced by H 2 O 2 . However, whether the reduced ROS/oxidative stress level by up-regulated autophagy is through the removal of damaged mitochondria or oxidized proteins needs further investigated.

Conclusion
In conclusion, this study has provided the first evidence that up-regulated autophagy by NAD + reduced the oxidative stress in RPE cells. In turn, it protects RPE cells from PARP-1 mediated-necrotic cell death induced by oxidative stress (Fig. 7). The results suggest that exogenous NAD + administration may be a novel, inexpensive, and effective treatment for preventing RPE cell death in AMD pathology. Further animal studies are required to explore this treatment value.