**4. Discussion**

In this study, antioxidative properties of ascorbic acid and astaxanthin were evaluated based on H2O2-induced and UVB-induced oxidative stress models within ARPE-19 cells. Studies have found that H2O2 and UVB have di fferent e ffects on cells regarding oxidative stress. First, even directly adding H2O2 in the cell culture medium results in a short-term exposure because its concentration decreases rapidly in the presence of the cells. H2O2 can penetrate the cell easily, but it is also reduced rapidly by the antioxidative mechanism [29]. On the other hand, UVB has a lingering e ffect on the cells by directly damaging DNA, causing gene mutation, and modifying gene expression, and enzyme activity along with increasing ROS level [36]. UVB-induced damage is mediated by two di fferent pathways. sne is by ROS generated immediately after the irradiation and the other is by reactive nitrogen species in the later time point [37]. As a result, even with a single and momentary exposure to UVB, the viability of the exposed cells decreases in the course of time [36].

Based on the precedent research, viability change of ARPE-19 cells was evaluated after 24 h for a H2O2 model and a UVB model [38]. H2O2 (0–0.8 mM) was exposed to ARPE-19 cells and their viability was dose-dependently reduced and intracellular ROS level was increased. UVB also reduced the cell viability and increased the intracellular ROS level but the H2O2 seemed to decrease the viability exponentially. One explanation can be that because H2O2 not only produces ROS, but it also a ffects junctional integrity of the RPE cell [39], weakening the cell adhesion to the bottom of the well—the cell viability assay result may have been a ffected. This can also explain the lower cell viability at 0.4 mM of H2O2 than 40 mJ/cm<sup>2</sup> UVB even though cells with H2O2-induced oxidative stress have a lower ROS level.

Within the condition of sublethal and lethal doses of H2O2 and UVB, antioxidative potencies of ascorbic acid and astaxanthin were evaluated. Although their antioxidative properties had been studied and have moved on to clinical level with patients with retinal diseases (Age-Related Eye Disease Study [AREDS] and Carotenoids in Age-Related Maculopathy Italian Study [CARMIS]) [40,41], there are controversies about whether they have a protective e ffect on cellular oxidative stress model. In one study, ascorbic acid did not have a protective effect on Fenton-reaction-mediated oxidative stress model of ARPE-19 but it rather decreased the cell survival ratio at a low concentration (0.1–1 mM) compared to the group without ascorbic acid [42]. This was also the case for a *tert*-butyl hydroperoxide (t-BOOH)-induced oxidative stress model. In a study by Kagan et al., ascorbic acid (0.02–0.2 mM) also decreased the cell viability of ARPE-19 with oxidative stress induced by t-BOOH [43]. The effect of t-BOOH in porcine RPE also could not be diminished by ascorbic acid [44]. In our study, however, ascorbic acid increased the viability of the cells even at a low concentration where studies mentioned above sugges<sup>t</sup> it decreases the viability and this was confirmed within two different oxidative stress models mediated by H2O2 and UVB. Although the central mechanism of t-BOOH to induce oxidative stress is by generating alkyl radicals [45], H2O2 is the central redox signaling molecule in general [30] forming hydroxy radicals [46], which can react intracellularly to generate various radicals including alkyl radical [47]. Considering H2O2 model reproduces more general situation of oxidative stress, and UVB model mediates H2O2 as the central signaling molecule [48], our result based on both models is more convincing.

Ascorbic acid neutralized the effect of the oxidative stress inducer in both H2O2 and UVB model but astaxanthin only did so in UVB-induced stress model. Li et al. [38] investigated the effect of astaxanthin on ARPE-19 cells against oxidative stress with H2O2. They incubated ARPE-19 cells with 20 μM astaxanthin for different lengths of time (6, 12, and 24 h) and then exposed to 200 μM H2O2 for 24 h. The cell viability increase was time-dependent, and they suggested that 24 h was the optimal time for astaxanthin treatment. In the current study, we incubated ARPE-19 cells with ascorbic acid or astaxanthin for 6 h and then exposed to H2O2 with antioxidants for 24 h. Our findings that astaxanthin did not show significant effect on the cell viability with H2O2 exposure while it did show increased viability with UVB irradiation may be due to different lengths of time from those of Li et al. [38] for astaxanthin treatment. There is also another possibility that even if astaxanthin, an extremely lipophilic compound, was dissolved in DMSO, it may be possible that the efficiency was lower than the actual concentration in an aqueous environment such as cell culture medium.

The synergistic effect of ascorbic acid and astaxanthin was also evaluated in this study. The combination of ascorbic acid and astaxanthin showed better antioxidative effect compared to each drug alone. There are few reports that investigate the effective antioxidant action of ascorbic acid and astaxanthin in combination. In a study by Guerra et al. [49], the association of astaxanthin with ascorbic acid greatly improved neutrophil phagocytic capacity and decreased ROS with pro-inflammatory cytokines. They suggested that the astaxanthin/ascorbic acid system mimics the recycling system of vitamin E/vitamin C. Astaxanthin provides cell membranes with potent protection against free radicals or other oxidative attack. Moreover, previous studies confirm that astaxanthin has a large capacity to neutralize free radicals or other oxidant activity in the nonpolar zones of phospholipid aggregates, as well as along their polar boundary zones [50]. Ascorbic acid, in turn, promotes antioxidant effects mainly in a water-phase microenvironment. The exact mechanism other than reducing the intracellular ROS production is unknown through this study, but it is assumed that the two antioxidants exhibited synergistic effects through the mechanism identified in previous experimental studies.

An elevated level of ROS is always observed in the diabetic retina. Given that the administration of antioxidants in animal models preserves retinal capillaries from hyperglycemia-induced degeneration, ROS are considered a major causative factor involved in DR development [51,52]. Meanwhile, diabetes induces mitochondrial damage in the retina and its capillary cells and mitochondrial dysfunction is also considered to play a significant role in the development of DR [53,54]. The results of this study showed that antioxidants treatment resulted in significantly improved cell viability which is perhaps due to the improved mitochondrial function, improved cellular attachment performance, and increased growth rate of the cells. However, further investigation will be required to determine more precise mechanisms and effects of antioxidants on ARPE-19 cells.

Although the role of retinal endothelial cells comprising inner blood-retinal barriers (BRB) is important in DR development, the role of RPE cells comprising outer BRB is also crucial. The flow of nutrients materials, metabolites, ions, proteins, and water flux to and from the retina is regulated by these two BRBs and disruption of inner and outer BRB causes retinal hyperpermeability and development of diabetic macular edema, which is a leading cause of vision loss in DR [55]. Therefore, current study using ARPE-19 cells is thought be closely related to BRB breakdown and the pathogenesis of DR. Oxidative damage to cells is commonly modeled using treatment with H2O2 for a long time [56,57], however, very little is known about the role of UV irradiation on ARPE-19 cells and especially on the development of DR. A recent population-based study investigated the association between daily sunlight exposure duration and DR [58]. The authors suggested that the risk of DR was 2.66 times higher in the group with ≥ 5 h of daily sunlight exposure than in the group with less exposure after adjusting for risk factors such as duration of diabetes, serum hemoglobin A1c level, hypertension, and dyslipidemia. Although a lot of evidence is still lacking, the results of the current study can be used as evidence that the e ffects of oxidative stress induced by H2O2 or UVB irradiation on the development of DR as well as the antioxidants can reduce disruption of outer BRB, which is represented by the viability of ARPE-19 cells.

This study successfully established an oxidative stress model of RPE which can be used to test potential antioxidative compounds. Since it has been known that oxidative stress is closely linked to diabetic retinopathy, identifying antioxidants like ascorbic acid and astaxanthin used in this study can also be beneficial to patients with diabetic retinopathy.
