Activation of Delta-Opioid Receptor Protects ARPE19 Cells against Oxygen-Glucose Deprivation/Reoxygenation-Induced Necroptosis and Apoptosis by Inhibiting the Release of TNF-α

Purpose Retinal ischemia–reperfusion injury (RIRI) is the basis of the pathology that leads to many retinal diseases and induces necroptosis and apoptosis. Tumor necrosis factor-α (TNF-α) is critically involved in necroptosis and apoptosis. Delta-opioid receptor (DOR) activation inhibits TNF-α release in our previous studies, it might prevent necroptosis and apoptosis by inhibiting the release of TNF-α. However, the role of TNF-α and DOR in necroptosis and apoptosis of retinal pigment epithelial (RPE) cells remains largely unknown. Here, we explored the mechanisms of TNF-α and DOR in necroptosis and apoptosis using an oxygen-glucose deprivation/reoxygenation (OGD/R) model of adult retinal pigment epithelial cell line-19 (ARPE19) cells. Materials and Methods ARPE19 cells were exposed to OGD/R conditions to mimic RIRI in vitro. Cell viability was quantified using the Cell Counting Kit-8 (CCK-8) assay. Morphological changes were observed by inverted microscopy. TNF-α protein levels in cell lysates were measured by enzyme-linked immunosorbent assay (ELISA). The DOR agonist TAN-67 and antagonist naltrindole (NTI) were used to pretreat cells for 1 or 2 hours before OGD24/R36 administration. Calcein acetoxymethylester/propidium iodide (Calcein-AM/PI) and Terminal deoxynucleotidyl transferase dUTP nick-end labeling (TUNEL) staining were used to detect necroptotic and apoptotic ARPE19 cells, respectively. The protein expression of DOR, p-RIP1 (RIP1), p-RIP3 (RIP3), p-MLKL (MLKL), and cleaved Caspase3 (Caspase3) was measured by western blotting. Results OGD severely damaged ARPE19 cells. Prolonged reoxygenation significantly increased TNF-α level and decreased DOR expression in ARPE19 cells. Pretreatment with the DOR agonist TAN-67 (10 µM) significantly improved ARPE19 cell viability after OGD24/R36 by reducing the number of necroptotic and apoptotic cells. Furthermore, DOR activation significantly inhibited TNF-α release and suppressed the expression of proteins related to necroptosis and apoptosis, including p-RIP1, p-RIP3, p-MLKL, and cleaved Caspase3, after OGD24/R36. This effect was reversed by the DOR antagonist NTI. Conclusion These results strongly suggest that DOR activation inhibits necroptosis and apoptosis by decreasing TNF-α release, leading to the prevention of OGD/R-induced injury in ARPE19 cells. This study provides an innovative idea for clinical treatment strategies for retinal damage and vision loss due to RIRI.


Introduction
Retinal ischemia-reperfusion injury (RIRI) is the basis of the pathology that leads to progressive visual loss and blindness in many retinal diseases, such as glaucoma [1], diabetic retinopathy [2], and retinal vascular occlusion disease [3]. Te pathogenesis of RIRI is complex, and there are currently four main types of injury mechanisms: infammatory reactions, cellular death, excitatory amino acids, and reactive oxygen species [4,5]. Among them, cellular death is the key to impairments in vision and retinal function [6,7]. Retinal pigment epithelial (RPE) cells are located between Bruch's membrane of the choroid and the retinal neurepithelium and perform a variety of functions, such as transporting nutrients from the choroid to photoreceptors, expelling metabolic waste, engulfng detached photoreceptor outer segments, and maintaining the stability of retinal appendages by removing fuid from the retinal sites of the inner photoreceptor layer [8][9][10]. RPE cell death occurs in RIRI, ultimately causing progressive and irreversible degeneration of retinal ganglion cells and vision loss [11]. Tus, protecting RPE cells can be considered an efective strategy to prevent the devastating damage of RIRI events.
RIRI alters neuronal tissue function by inducing neuronal cell death [12]. In RIRI, RPE cells mainly undergo two types of programmed cell death, necroptosis, and apoptosis, and these two types of death do not exist in isolation. Tey often occur together or one after the other [7]. Tumor necrosis factor-α (TNF-α) is well known to be the major trigger of necroptosis and apoptosis [13]. When TNF-α binds to the TNF receptor on the cell membrane, it further activates the downstream necroptosis and apoptosis pathways [14]. When receptor-interacting protein kinase 1 (RIP1) is activated, it can recruit receptor-interacting protein kinase 3 (RIP3), causing RIP3 phosphorylation. Once RIP3 is activated, mixed lineage kinase domain-like protein (MLKL) is assembled and phosphorylated. Eventually, necrosomes are formed, which triggers necroptosis. Apoptosis mainly occurs through the activation of Caspase 8 via the apoptotic pathway, which further activates Caspase 3 (cleaved Caspase 3) to cause apoptosis [15,16]. Terefore, inhibiting TNF-α is particularly important for reducing necroptosis and cell apoptosis.
Te delta-opioid receptor (DOR) is an oxygen-sensitive protein sensitive to ischemia and hypoxia [17]. Several experiments in vivo and in vitro support the essential role of DOR in the fght against ischemic and hypoxic injury [18,19]. Furthermore, it has been reported that DOR is widely distributed in the retina and plays an important role in protecting the retina [20]. Peng et al. demonstrated that in RIRI, the upregulation of DOR expression corrected the redox imbalance in the impaired retina and had a protective efect on the retina [21]. In addition, previous studies in our laboratory have confrmed that DOR activation can inhibit the increase in TNF-α caused by cerebral ischemia-reperfusion and protect neurons from death in vivo and in vitro [22]. As the retina is an extension of the central nervous system, whether DOR activation in the retina could also play a similar protective role in inhibiting necroptosis and apoptosis in RPE cells induced by RIRI has not been examined.
Te oxygen-glucose deprivation/reoxygenation (OGD/ R) cell model can successfully mimic the acute restriction of metabolites and oxygen supply during ischemia-reperfusion injury [23]. Tus, (1) OGD/R model was used to simulate RIRI in adult retinal pigment epithelial cell line-19 (ARPE19) cells; (2) agonists or antagonists of DOR were administered to examine whether DOR activation could inhibit OGD/R-induced necroptosis and apoptosis by inhibiting the release of TNF-α in ARPE19 cells and exploring the mechanism.

Materials and Methods
2.1. Cell Culture. ARPE19 cells were purchased from the National Collection of Authenticated Cell Cultures (Shanghai, China) and maintained in DMEM/F12 (Gibco, USA) containing 10% fetal bovine serum (FBS) (Gibco, USA) and 1% penicillin/streptomycin (Beyotime, China) at 37°C with 5% CO 2 . Te medium was changed every other day and regularly tested for mycoplasma contamination. Cells in the logarithmic growth phase were used for subsequent experiments.

Cell
Culture Treatment and OGD/R. Cells in the logarithmic growth stage were randomly assigned to fve groups and subjected to diferent treatments as follows: (1) control group; (2) OGD/R group; (3) agonist group (Tan-67 + OGD/ R); (4) antagonist group (naltrindole (NTI) + OGD/R); and (5) coadministration group (NTI + Tan-67 + OGD/R). TAN-67 and NTI were dissolved in phosphate-bufered saline(PBS) (pH 7.2) as stock solutions (10 mM and 100 mM for each) and stored at −20°C. Immediately before use, the stock was quickly thawed and diluted to the appropriate concentration with a normal culture medium. In the agonist and antagonist groups, ARPE19 cells were pretreated with TAN-67 or NTI alone for one hour. In comparison, cells in the coadministration group were incubated with NTI for one hour before TAN-67 was added for one hour followed by OGD/R. OGD/R was used as an in vitro model of ischemia-reperfusion injury. First, ARPE19 cells were washed twice with PBS (HyClone, USA) and cultured in DMEM without glucose and FBS in a hypoxia chamber that was fushed with a 95% N 2 /5% CO 2 gas mixture at 3 L/min for 15 min at room temperature. Ten, the chamber was placed in a thermostatic incubator (Termo Fisher Scientifc, USA) for oxygen and glucose deprivation. After OGD, the cells were returned to normal DMEM/F12 under normoxic conditions and underwent reoxygenation. Cells in the control group were cultured under normal conditions for the corresponding times.

Analysis of Cell Viability.
Cell viability was monitored with a Cell Counting Kit-8 (CCK-8) assay (Dojindo, Japan) according to the manufacturer's suggestions. Briefy, 100 μl of media containing 5 × 10 3 ARPE19 cells were seeded in 96-well plates. Following modeling, 10 μl of CCK-8 solution was added to each well, and the plates were placed in a thermostatic incubator for half an hour and then assayed using a microplate reader (Termo Fisher Scientifc, USA) at a wavelength of 450 nm. Cell viability is expressed as the percentage of cell viability compared to the control group.

Measurement of TNF-α Protein
Levels. TNF-α levels in cell lysates were measured by anti-human TNF-αenzymelinked immunosorbent assay (ELISA) (R&D Systems, USA) according to the manufacturer's instructions. In brief, protein samples were extracted from cells by sonication. Te samples (100 μL/well) were added to 96-well plates and incubated at 37°C for 90 minutes. After the samples were washed with washing bufer three times, HRP (100 μL/well) was added to the wells and incubated at 37°C for 30 min. Ten, the wells were washed three times with washing bufer and developed with TMB (100 μL/well) for 15 min. After the termination solution was added, the absorbance was measured at 450 nm with a microplate reader. Te concentration of TNF-α was calculated by a TNF-α standard curve.

Western Blot Analysis.
Western blotting was performed according to a standard method, as described previously [22]. Te following primary antibodies were used in this study: RIP1  and goat anti-mouse IgG (1 : 4000) were purchased from Sigma. Te protein bands were quantifed by densitometry using ImageJ software, and protein expression was normalized to the expression of the internal control.

Calcein-AM and PI Staining.
Viable or necroptotic ARPE19 cells were stained with calcein acetoxymethylester (Calcein-AM) or propidium iodide (PI) according to the kit instructions (Beyotime Biotechnology, China). Te primary steps are briefy described. Te culture medium was discarded, and the wells were washed twice with PBS. Te dye solution was composed of Calcein-AM, PI, and detection bufer at a volume ratio of 1 : 1:1000. It was added to six-well plates and incubated at 37°C for 30 min in the dark. At the end of the incubation, the wells were washed twice with PBS, and Hoechst 33258 was added and incubated at 37°C for 10 min in the dark. Finally, images were taken by fuorescence microscopy (Nikon, Japan). Te number of PI-positive cells was analyzed using ImageJ software.

TUNEL
Staining. ARPE19 cell apoptosis was determined using an in situ terminal deoxynucleotidyl transferase dUTP nick-end labeling nick-end labeling (TUNEL) assay (Beyotime Biotechnology, China). Te specifc processes were performed according to the instructions. Te cells were fxed in 4% paraformaldehyde for 30 min at room temperature. After being washed three times with PBS, the cells were permeabilized with 0.3% Triton X-100 in PBS for 5 min at room temperature. Te enzyme solution (TdT) and labeling solution (dUTP) were mixed with the TUNEL detection solution at a dilution of 1 : 9. TUNEL detection solution was added to each well and incubated at 37°C in the dark for 1 h. Before detection, the wells were washed twice with PBS, and Hoechst 33258 was added and incubated at 37°C for 10 min in the dark. At the end of the incubation, images were obtained by fuorescence microscopy, and the analysis was performed using ImageJ software.

Data and Statistical
Analysis. Te data are presented as the mean ± SEM and analyzed by Student's t-tests or oneway ANOVA followed by post hoc comparisons with Dunnett's post hoc test and Fisher's LSD test. Statistical analysis was conducted using SPSS software (version 23; IBM Corp., Armonk, NY, USA). A P value <0.05 was considered statistically signifcant.

Diferent Durations of Oxygen and Glucose Deprivation Have Diferent Efects on ARPE19 Cell Viability and
Morphology. To construct a stable OGD/R model in ARPE19 cells, the diferent duration and extents of OGD/R (12 h, 18 h, and 24 h) were examined. We fxed the reoxygenation time at 24 h. Ten, CCK-8 and inverted microscopy were used to examine cellular viability and morphological changes, respectively. As shown in Figure 1(a), cell viability gradually decreased with prolonged oxygen and glucose deprivation time. Compared with that in the control group, cell viability decreased to approximately 85% in the OGD12/R24 group ( # P < 0.05), 78% in the OGD18/R24 group ( ### P < 0.001), and 60% in the OGD24/ R24 group ( ### P < 0.001), respectively. As shown in Figure 1(b), the morphology of ARPE19 cells in the control group was fusiform, and the cells aggregated in a colony form. With increasing oxygen and glucose deprivation durations, foating cells with spherical or round shapes (white arrows) and cellular debris, probably from dead cells, increased. Few cells remained attached to the cell culture plate, while most cells were rounded and free-foating in the OGD24/R24 group.
Tese results demonstrate that various durations of oxygen-glucose deprivation caused diferent degrees of injury in ARPE19 cells. Moreover, OGD24/R24 markedly decreased cell viability to approximately 60% and resulted in considerable alterations in cell morphology, consistent with the injury conditions of OGD/R modeling.

Diferent Durations of Reoxygenation Have Diferent Efects on DOR and TNF-α Protein Levels in ARPE19 Cells.
It has been reported that diferent reoxygenation time points also had diverse efects on cells in an OGD/R model [24,25].
Tus, we examined the TNF-α and DOR protein levels in ARPE19 cells to determine whether various durations of reoxygenation have diferent efects on ARPE19 cells. TNF-α protein levels in cell lysates were measured by ELISA (Figure 2(a)), and the levels in the control group were extremely low. After OGD/R, the levels of TNF-α increased signifcantly ( ### P < 0.001); moreover, the TNF-α levels in the OGD24/R36 group were signifcantly higher than those in the OGD24/R24 group ( * P < 0.05). Western blotting showed abundant DOR expression in the control group; however, DOR expression showed a relative decrease after OGD/R (Figures 2(b), 2(d)). Te protein expression of DOR in the OGD24/R24 group and OGD24/R36 group decreased by approximately 30% and 40% compared with that in the control group (Figures 2(c), 2(e)).
We can conclude that under fxed oxygen and glucose deprivation times, diferent reoxygenation times have distinct efects on ARPE19 cells. In the OGD24/R36 group, TNF-α protein levels increased the most, and DOR protein expression decreased the most. Here the OGD24/R36 group was used as the modeling condition for subsequent experiments.

Te Efects of the DOR Agonist TAN-67 and the Antagonist NTI on the Viability of ARPE19 Cells Induced by OGD24/R36.
To examine the impacts of DOR protein activation and antagonism on the cell viability after OGD24/R36, ARPE19 cells were pretreated with diferent concentrations of the DOR agonist TAN-67 (50, 30, 10 µM) or the antagonist NTI (50, 30, 10 µM) for one hour. CCK-8 was used to examine cellular viability after OGD24/R36 (Figure 3). Te results showed that OGD24/R36 resulted in a signifcant decline in cell viability ( ### P < 0.001). Pretreatment with the DOR agonist TAN-67 increased cell viability compared with that in the OGD24/R36 group ( * P < 0.05, * * P < 0.01, * * * P < 0.001). Among them, 10 µM TAN-67 had the best efect on increasing the viability of ARPE19 cells. Treatment with 10 μM NTI resulted in a substantial decrease in cell viability ( * P < 0.05), while no obvious changes were observed when the cells were treated with 50 μM and 30 μM NTI compared with that in the OGD24/R36 group. Tese results suggested that DOR activation could improve the survival of ARPE19 cells, and the opposite efects were observed in response to DOR inhibition.

Te DOR Agonist TAN-67 Reduced the Number of Necroptotic and Apoptotic ARPE19 Cells after OGD24/R36.
We further examined whether the DOR agonist TAN-67 could inhibit necroptosis and apoptosis in ARPE19 cells after OGD24/R36. Calcein-AM/PI and TUNEL staining were used to examine necroptosis and apoptosis in ARPE19 cells, respectively. As shown in Figure 4(a), Calcein-AM staining (green) represents live cells, while PI staining (red) represents necroptotic cells. Based on the results shown in Figure 4 Taken together, these results suggest that the DOR agonist TAN-67 reduces the number of necroptotic and apoptotic ARPE19 cells after OGD24/R36.

Te DOR Agonist TAN-67 Suppressed TNF-α Levels and the Expression of Necroptotic and Apoptotic Proteins in ARPE19 Cells Induced by OGD24/R36.
To test whether the key cytokine TNF-α is involved in the mechanism triggering necroptosis and apoptosis, ELISA was performed to examine the levels of TNF-α in ARPE19 cells. As shown in Figure 5(a) the levels of TNF-α in the OGD24/R36 group were increased compared with those in the control group ( ### P < 0.001). Pretreatment with TAN-67 inhibited the OGD24/R36-induced increase in TNF-α ( * P < 0.05), but this efect was eliminated by NTI pretreatment. NTI pretreatment alone also showed no signifcant efects compared with the OGD24/R36 group.

Discussion
Based on this research, the following conclusions can be drawn: (1) the viability of ARPE19 cells decreased signifcantly with prolonged oxygen-glucose deprivation. (2) Te more extended the reperfusion time after oxygen-glucose deprivation, the more TNF-α and the less DOR protein expression in the cells. (3) DOR activation signifcantly reduces the release of TNF-α in ARPE19 cells after OGD24/ R36, which can decrease the number of necroptotic and apoptotic cells and inhibit related protein expression from exerting a protective efect on ARPE19 cells.
OGD/R is a classic in vitro model for simulating ischemia-reperfusion injury and has been used in ischemia-reperfusion injury experiments to study the brain, heart, and kidney in vitro [26][27][28]. Previous studies have shown that diferent oxygen and glucose deprivation durations have robust impacts on cell survival [29,30]. Terefore, we examined the duration of oxygen and glucose deprivation and the survival rates of ARPE19 cells at different time points by CCK-8 assays. Te cell survival rate in response to OGD24/R24 was 60%, and most cells showed obvious damage, which was consistent with the conditions of model damage. It has also been reported that cells release more infammatory factors with the prolongation of reperfusion. [31,32]. We then determined the time course of changes in TNF-α after reoxygenation, and our results verifed this point of view. Te amount of TNF-α released at      Data are presented as means ± SEM, and group diferences were analyzed by one-way ANOVA with Dunnett's post hoc test. ## P < 0.01, ### P < 0.001 compared with the control group; * P < 0.05, * * P < 0.01, * * * P < 0.01, compared with the OGD24/R36 group. 36 hours of reperfusion was higher than that at 24 hours, and reperfusion for 36 hours induced more damage to the expression of DOR protein in ARPE19 cells. Based on the relationship between TNF-α and both necroptosis and apoptosis and the efect of activated DOR in ischemia-reperfusion, we chose OGD24/R36 as the optimum time for this study.
Necroptosis and apoptosis have distinctly diferent characteristics in terms of pathological features [33]. First, the integrity of the necroptotic cell membrane is severely damaged, the organelles are swollen and disintegrated, and chromatin in the nucleus lacks obvious morphological changes. Apoptosis is characterized by an intact cell membrane, intracellular chromosome condensation, nuclear fragmentation, and the formation of DNA bands [34]. Tese distinct diferences can also be clearly distinguished by light microscopy, electron microscopic observation, and specifc staining methods [35,36]. In this study, calcein-AM/ PI staining and TUNEL staining were used to examine necroptosis and apoptosis in ARPE19 cells, respectively. PI only penetrates damaged cell membranes and releases red fuorescence when embedded in double-stranded DNA, so it has been used to identify necroptotic cells in recent studies [37,38]. TUNEL is a commonly used method to label apoptotic cells by specifcally detecting DNA breaks that occur during apoptosis. Consistently, our results showed that the number of necroptotic and apoptotic ARPE19 cells induced by OGD24/R36 was also markedly diminished after pretreatment with the DOR agonist TAN-67, whereas the DOR antagonist NTI blocked this protective efect.
Current research suggests that TNF-α is a common promoter of both necroptosis and apoptosis [14]. Many previous studies, including ours, have confrmed that DOR activation can inhibit the release of TNF-α in cerebral ischemia-reperfusion injury and markedly attenuate the infammatory response [22,39]. Similarly, the results of this study demonstrated that DOR activation inhibited the ODG24/R36-induced increase in TNF-α levels in ARPE19 cells. Te reduction in TNF-α may downregulate the activation of the necroptotic marker proteins p-RIP1, p-RIP3, and p-MLKL or the apoptotic marker protein cleaved Caspase3 downstream of the signaling pathway. Terefore, the DOR agonist TAN-67 reverses the OGD24/R36-induced increases in necroptosis-and apoptosis-associated proteins. Our previous studies have shown that DOR activation suppresses infammation and is closely related to the MAPK/ p38 pathway [40]. It has also been reported that TNF-α release is associated with the activation of the nuclear factor kappa-B (NF-κB) pathway. And DOR activation can reduce TNF-α release by inhibiting NF-κB pathway activation, thereby mitigating infammation [41]. In the future, we will also validate the mechanisms associated with the inhibition of the infammatory response in ARPE19 cells by agonists of DOR in OGD/R injury.
Previous studies have shown that the DOR protein is abundantly expressed in the human retinal pigment epithelium, and the administration of DOR agonists can signifcantly alleviate RPE cell damage and maintain the function and integrity of the retinal pigment epithelium in diabetic retinopathy [22]. As an essential component of the retinal epithelium, RPE cells play a key role in maintaining visual circulation, barrier formation, and material transport in the retina [42]. When pigment epithelial cells are injured, the retina sufers a disruption in water-electrolyte transport and a considerable accumulation of oxygen-free radicals and infammatory factors that eventually damage the retinal ganglion cells resulting in vision loss or blindness [43]. Terefore, inhibiting RPE cell death is critical to the therapeutic and prognostic efects of retinal diseases. In the present study, DOR activation by TAN-67 dramatically afected the inhibition of necroptosis and apoptosis in ARPE19 cells after OGD/R and could decrease ARPE19 cell injury. Consistent with our recent study (data not shown), DOR activation may play a critical role in ameliorating retinal ganglion cell injury by rescuing OGD/R-induced human RPE cell damage. DOR could be a new target for the treatment of visual loss or blindness after RIRI.

Conclusion
In summary, the present study demonstrates that DOR activation can reverse the increase in TNF-α induced by OGD/R while alleviating TNF-α-induced necroptosis and apoptosis in ARPE19 cells. Tis fnding indicates that DORmediated retinal protection is closely correlated with the downregulation of the TNF-α pathway. Tis study provides new insights into the clinical treatment of retinal damage and vision loss caused by RIRI.

Data Availability
Te data used to support the fndings of this study are available from the corresponding author upon request.