Protective effect of myricetin derivatives from Syzygium malaccense against hydrogen peroxide-induced stress in ARPE-19 cells

Materials

ARPE-19 (ATCC^® CRL-2302^TM) RPE cells (Organism: Homo sapiens; human) were purchased from American Type Culture Collection (ATCC; Manassas, VA). The species of origin (human) of the cell line (ARPE-19; passage 19; Lot number: 59,270,158) was authenticated by the supplier (ATCC) using the cytochrome C oxidase I gene (COI) assay and short tandem repeat (STR) profiling to verify the human unique DNA profile and rule out intraspecies contamination. According to ATCC, the STR analysis was performed on eight core STR loci (D5S818, D13S317, D7S820, D16S539, vWA, TH01, TPOX, CSF1PO) and gender determining locus (amelogenin). The loci were amplified using the Promega PowerPlex® 18D System (Madison, WI) and the ANSI/ATCC ASN-0002-2011 Standard for Authentication of Human Cell Lines. The amplicons were separated using capillary electrophoresis and data were analyzed using GeneMapper® software (Thermo Fisher Scientific, Waltham, MA). Myricitrin (purity >99.0%), myricetin (purity >96.0%), fetal bovine serum (FBS), o-tolidine, 2′,7′-dichlorofluorescin diacetate (DCFH-DA), and chemicals for antioxidant assays were procured from Sigma Chemical Co., Ltd. (St. Louis, MO). Glucose oxidase from Aspergillus niger was purchased from Biochemika Fluka (St. Louis, MO). Gibco® Dulbecco’s Modified Eagle Medium/nutrient mixture F12 (DMEM/F12) was purchased from Invitrogen Corporation (Carlsbad, CA). Chemicals and reagents needed for gene expression study were supplied by Qiagen (Frederick, MD). Miscellaneous reagents used were of analytical grade.

Isolation of myricetin derivatives (F2) from the S. malaccense ethanolic leaf extract

S. malaccense leaf was subjected to ethanolic extraction, and the myricetin derivative–rich fraction (F2) was isolated from the extract through a standard fractionation protocol established using high-performance liquid chromatography (HPLC) [11]. The samples were stored at − 20 °C. The samples were reconstituted with dimethyl sulfide (DMSO) at an approximate concentration and filter sterilized before use.

Determination of antioxidant properties

Antioxidant assays such as 2,2-diphenyl-1-picrylhydrazyl (DPPH), 2,2'-azino-bis(3-ethylbenzothiazoline-6-sulphonic acid (ABTS), and nitric oxide (NO) free radical scavenging assays for various samples (S. malaccense ethanolic leaf extract, myricetin derivative–rich fraction isolated from the extract (F2), and standard compounds such as myricitrin and myricetin) were performed as described in a previous report [11]. Briefly, the DPPH assay was performed by mixing and incubating various samples at different concentrations (5 µl) with 195 µl ethanolic DPPH reagent (100 mM) for 20 min and absorbance was read at 515 nm in a 96-well microtiter plate. The ABTS assay was carried out by incubating 10 µl of samples with ABTS reagent (90 µl) for 4 min. Absorbance of the mixture was measured at 734 nm in a 96-well microtiter plate. The NO assay was conducted by mixing the samples (10 µl) with 90 µl of sodium nitroprusside reagent (final concentration: 10 mM; prepared freshly in phosphate buffered solution, pH 7.3) and incubating the mixture for 2.5 h in the presence of light. This was followed by the addition of 100 µl of Griess-Ilosvay's nitrite reagent and incubation for 30 min in dark before measuring the absorbance at 540 nm using a 96-well microtiter plate. Percentage of radical scavenged (DPPH, ABTS, and NO) was calculated as stipulated below to determine the IC₅₀ value for the samples:

% of radical scavenged= Abs c o n t r o l − Abs s a m p l e Abs c o n t r o l x 100

Abs stands for absorbance. Control is the sample volume replaced with diluents.

The total phenolic content (TPC) was determined by mixing and incubating 50 µl of the samples with 50 µl of Folin-Ciocalteu phenol reagent (10% v/v) for 3 min in dark at room temperature. This was followed by the addition and incubation of the mixture with 100 µl of sodium carbonate solution (10% w/v) for 1 h. Absorbance was read at 750 nm in a 96-well microtiter plate. The average TPC value was calculated and stated as the amount of the phenolic content (mg gallic acid equivalents, GAE) per gram of sample [11]. Additionally, ferric reducing antioxidant power (FRAP) assay was performed based on the protocol described by Benzie and Strain [14] with a slight modification, using 96-well microtiter plates. Briefly, stock solutions of (a) 0.3 M sodium acetate buffer, (b) 10 mM 2,4,6-Tris(2-pyridyl)-s-triazine (TPTZ) in 40 mM hydrochloric acid and (c) 20 mM ferric chloride solution were prepared freshly prior to the assay and mixed at a ratio of 10:1:1 (a:b:c). This working solution (FRAP reagent) was protected from exposure to light. The samples (10 μl) at various concentrations were mixed with 300 μl of FRAP reagent, and the respective absorbance was measured after 4 min at 593 nm using a microplate reader. A stock solution of ferrous sulfate, Fe₂SO₄ (final concentration: 0–30 µM), was subjected to the assay to derive a standard curve to calculate FRAP values. The mean FRAP value was expressed as the Fe (II; mM) equivalent in 1 gram of sample.

Cell culture

The ARPE-19 cells were cultured according to the protocol described by the supplier (ATCC). The cells were grown in DMEM/F12 containing 10% (v/v) heat-inactivated FBS and 1% (v/v) penicillin (100 U/ml)-streptomycin (100 U/ml) solutions at 37 °C in a humidified atmosphere of 5% carbon dioxide (CO₂). Cells of passages 21–24 were used for the experiments. The cells were sub-cultured every 3–4 days when they reached 70–80% confluence.

Induction of oxidative stress using glucose oxidase treatment

ARPE-19 cells (10,000 cells/well) were allowed to adhere to the 96-well cell culture plates overnight at 37 °C in 5% CO₂ atmosphere. The cells were pretreated with various concentrations of samples (200 μg/ml or less for the extract and F2 and 40 μg/ml or less for myricitrin and myricetin) for 2 h before incubation with glucose oxidase, GO (final concentration, 12 mUnit/ml; to generate H₂O₂) for 2 h in 96-well cell culture plates. The concentrations described above were not toxic to the cells when determined using the 3,4-5-dimethylthiazol2-yl)-2,5-diphenyl tetrazolium (MTT) assay (Appendix 1). After the incubation period, the culture supernatants were removed, and the wells were rinsed with PBS (136.9 mM sodium chloride, 2.7 mM potassium chloride, 8.1 mM disodium hydrogen phosphate, 1.5 mM potassium dihydrogen phosphate, pH 7.3) and replaced with fresh media. Then, cell viability was determined using the MTT assay. MTT solution (a final concentration of 0.5 mg/ml/well) was added, and the incubation was continued for 3 h at 37 °C. DMSO (100%) solution was added to solubilize the formazon crystals. The culture plate was then placed on a plate shaker for 5 min, and the absorbance was measured at 570 nm using a microplate reader. The percentage of cell viability was determined as stipulated below:

% of cell viability= Abs treated cells − Abs blank Abs untreated cells x 100

Untreated cells = cells grown in basal media.

Blank = cells grown in sample diluent (DMSO, final concentration of 0.1%).

The protective effect of the samples was determined through three independent cell stimulation experiments (with three technical replicates in each stimulation).

Measurement of intracellular ROS

The intracellular amount of ROS was determined using the DCFH-DA fluorescent probe as described previously [15] with a slight modification. Cells were pretreated with samples (2 h) followed by GO for 2 h as described in the section above. DCFH-DA (5 μM; final concentration) was introduced 30 min before the end of the incubation period with the oxidant. The adhered cells were rinsed twice with PBS. The fluorescence intensity of the cell monolayers was measured at 495 nm (excitation wavelength) and 525 nm (emission wavelength) using a fluorescence microplate reader (Infinite® M1000 PRO series, Tecan Group Ltd., Mannedorf, Switzerland, coupled with i-control^TM data analysis software). The fold change of the ROS was determined as stipulated below:

Fold change= Fluorescence intensity of cells treated with sample(s) Flourescence intensity of untreated cells

The fold change was normalized to 10,000 cells for all GO-H₂O₂-treated cells with and without pretreatment of samples by considering the number of viable cells after the treatment based on the cell viability assay results.

Determination of H₂O₂ levels

The level of H₂O₂ in the culture supernatants of the cells was determined using a protocol described previously [16] with a slight modification. The culture supernatants of incubations with oxidants (with and without pretreatment of samples) were collected. Tolidine (1% w/v) stock solution was prepared by dissolving 10 mg tolidine in 1 ml of absolute ethanol. The working solution comprised 2 units/ml peroxidase and 0.15% of tolidine (from the tolidine stock) in a total volume of 10 ml, 0.2 M sodium acetate buffer, pH 5.0. Fifty microliters of the sample (culture supernatants collected after the incubation) were added to 50 μl of the working solution in a 96-well microtiter plate. The mixture was incubated for 1 h at room temperature, and the absorbance was measured at 625 nm using a microplate reader. A standard curve derived using the H₂O₂ standard (0–2,500 μM) was used to determine the amount of H₂O₂ present in the culture supernatants of various treated samples.

Gene expression assessment

Samples that showed statistically significant protective effects were subjected to gene expression study. ARPE-19 cells (7.3 × 10⁵ cells/T25 flask) were seeded and allowed to adhere overnight before proceeding with the selected treatments and gene expression analysis. Cells grown in basal media (untreated) served as control. The RNA extraction, determination of the purity and integrity of the RNA samples, synthesis of first-strand cDNA, and quantitative real-time reverse transcriptase-PCR (RT–PCR) were performed according to the method described previously [12]. Briefly, RNA was extracted using Qiazol reagent and miRNeasy Mini Kit. The purity of the RNA samples was determined on Nano-Drop® ND-1000 UV-Vis Spectrophotometer (NanoDrop Technologies, Inc., Wilmington, DE) based on optimum absorbance ratio of 230/260 nm and 260/280 nm. Concentration and integrity of the RNA samples were analyzed on Agilent 2000 Bioanalyzer (Agilent Technologies Inc., Santa Clara, CA) using Agilent RNA 6000 kit. First strand cDNA was synthesized using RNA samples with high quality of RNA integrity number (RIN) ranging from 9-10. The reverse transcription procedure was performed using RT2 First strand kit (Qiagen) with optimized thermal cycling condition on BioRad C1000™ (Bio-Rad Laboratories Inc., Hercules, CA). Quantitative RT-PCR was conducted using custom RT2 profiler PCR array (CAPH12983) supplied by Qiagen which works based on SYBR Green detection chemistry on Applied Biosystem 7500-Fast Real-Time PCR (Thermo Fisher Scientific Inc., Waltham, MA) system. The program run was performed as stipulated below:

Step1 (1 cycle): 95 ˚C for 10 min.

Step 2 (40 cycles): 95 ˚C for 15 sec followed by 60 ˚C for 1 min. The genes investigated in this study are listed in Table 1. The gene expression data were analyzed using the comparative C_T (2^-∆∆Ct) scheme [17]. Final data that were normalized against human 18SrRNA (endogenous control) are expressed as the n-fold of regulation.

Statistical analysis

Results are expressed as the mean ± standard deviation (SD) unless otherwise stated. The values were computed from triplicate measurements using Microsoft Excel 2007 (Redmond, WA). All data were analyzed using the one-way ANOVA Tukey’s post hoc test using SPSS software version 17 (Chicago, IL). In statistical analysis, * denotes p<0.05, ** denotes p<0.01, and *** denotes p<0.001. In the paired t test, # denotes p<0.05, ## denotes p<0.01, and ### denotes p<0.001.

Antioxidant property and TPC of S. malaccense leaf extract and its active fraction (F2)

Table 2 depicts the TPC, the concentration of the sample needed to scavenge 50% of the radicals (IC₅₀), i.e., the DPPH, ABTS, and NO and FRAP levels of myricetin, myricitrin, leaf extract of S. malaccense, and its active fraction, F2. Standard myricitrin and myricetin were also included in this study as a comparison to the S. malaccense leaf extract and F2. Myricetin and myricitrin are classified as flavonols, a subgroup of flavonoids, and are known for their good radical scavenging property [18,19]. In addition, the aglycone myricetin is a major metabolic product of the flavonoid glycoside, myricitrin [20,21].

The IC₅₀ values (DPPH, ABTS, and NO) of the S. malaccense extract were adapted from our previous study [11]. F2 showed statistically significant improvement in its antioxidant property (based on its reduced IC₅₀ and increased FRAP values) when compared to the ethanolic leaf extract (Table 2). F2 was able to scavenge DPPH and NO radicals efficiently and showed good ferric reducing antioxidant power that was comparable to the effect exerted by myricetin. Myricitrin and myricetin showed better DPPH and ABTS scavenging activities than the leaf extract of S. malaccense and F2. Nevertheless, F2 was a better nitric oxide scavenger compared to myricitrin and myricetin.

GO-H₂O₂-induced stress model

The ARPE-19 cells were preincubated for 2 h with the samples before GO treatment to allow H₂O₂ production. The protective effect of the test samples was monitored using the cell viability (MTT) assay as well as changes in intracellular ROS and H₂O₂ levels in culture supernatants. Figure 1 shows the cytoprotective effects of the test samples to improve and restore the viability of the ARPE-19 cells that were exposed to GO-H₂O₂-induced oxidative stress. Two hours of 12 mUnit of GO treatment resulted in 45% reduction in cell viability (Figure 1). Interestingly, myricetin and F2 that were preincubated with the cells before the GO treatment showed statistically significant protection in a dose-dependent manner against GO-H₂O₂-induced cell death (Figure 1). Myricetin (40 μg/ml) showed protection by restoring the viability of the cells by 88% (p<0.001) followed by F2 (200 μg/ml) by 91% (p<0.001). Myricitrin and S. malaccense leaf extract exhibited statistically significant protection (p<0.001) of the retinal cells against GO-H₂O₂-induced cell death. However, the effect was not dose-dependent. F2 at a concentration of 50 μg/ml contained 39 µg of myricitrin (calculated based on liquid chromatography mass spectrometry, LCMS data of 77% volume [11]). F2 improved the viability of the retinal cells but not as effectively as myricitrin at 40 μg/ml. However, after the concentration of F2 was increased to 100 µg/ml (myricitrin content, 77 µg), the fraction was able to improve the cell viability up to 75%, which was similar to the effect of myricitrin (40 μg/ml). At 200 µg/ml, F2 (myricitrin content, 154 µg) enhanced the cell viability up to 91%, which was higher than the effect exerted by the S. malaccense leaf extract (Figure 1).

The GO-H₂O₂-treated cells showed an approximate 12-fold increase in the intracellular ROS level (Figure 2) with the presence of approximately 1,950 μM H₂O₂ in the culture supernatant (Figure 3). These values were statistically significantly reduced when the cells were preincubated with the samples for 2 h before the introduction of GO-H₂O₂-induced stress (Figure 2 and Figure 3). Myricetin, F2, and myricitrin attenuated intracellular ROS (Figure 2) and H₂O₂ generation (Figure 3) statistically significantly (p<0.001) in a dose-dependent manner. The S. malaccense leaf extract scavenged ROS at concentrations ≤25 µg/ml (p<0.001) and reduced the H₂O₂ levels statistically significantly at all concentrations tested.

Gene expression study

The preincubation of myricetin (40 μg/ml) and F2 (200 μg/ml) for 2 h showed statistically significant protection of the ARPE-19 cells against GO-H₂O₂-induced stress (Figure 1). Thus, the molecular mechanism of these compounds was assessed using the quantitative real-time RT–PCR technique. Table 1 depicts the list of genes investigated in this study. Figure 4 shows the gene expression results for the GO-H₂O₂-treated cells in the presence or absence of preincubation with F2 and myricetin.

GO-H₂O₂-induced stress caused statistically significant upregulation of the MAPK1, NFE2L2, SOD2, GPx1, CASP3, Bax, HSPB1, IFN-γ, NOS2, and LOX genes (Gene ID, OMIM, RefSeq Accession and UniGene numbers are listed in Table 1) and downregulation of the MAPK14 gene when compared to the control (untreated retinal cells). The introduction of oxidative stress did not result in changes in the expression of the SOD1, CTSD, CRP, and TLR2 (Gene ID, OMIM, RefSeq Accession and UniGene numbers are listed in Table 1) genes. Treatment of F2 along with GO caused enhanced upregulation of the NFE2L2 (p<0.001) and SOD2 (p<0.001) genes. F2 reversed the expression of the NOS2 (downregulation; p<0.01) gene compared to the GO-H₂O₂-treated cells. F2 treatment did not affect the pattern of gene expression influenced by GO-H₂O₂-induced stress for the MAPK1, MAPK14, GPx1, CASP3, Bax, HSPB1, IFN-γ, and LOX genes.

Myricetin downregulated the expression of the MAPK1 (p<0.001), GPx1 (p<0.001), CASP3 (p<0.01), Bax (p<0.01), HSPB1 (p<0.05), IFN-γ (p<0.01), and LOX (p<0.001) genes when compared to the GO-H₂O₂-treated cells. Treatment of myricetin did not show statistically significant changes in the expression of the MAPK14 and NOS2 genes that were affected by GO-H₂O₂-induced stress. It did not upregulate the NFE2L2 and SOD2 genes further as observed with the F2 treatment.

Antioxidant capacity of myricetin, myricitrin, leaf extract of S. malaccense, and its active fraction

The pure standard compounds (myricitrin and myricetin) showed higher TPC compared to F2 and the crude extract as the values were determined per gram of the sample and reflected the purity of the respective samples. Free radical scavenging activity is mainly attributed to the presence of phenolic compounds, such as phenolic acids, phenolic diterpenes, and flavonoids, in plants [22]. In the present study, myricetin exhibited a better radical (DPPH and ABTS) scavenging property than myricitrin (Table 2). This result concurs with another report that showed the aglycone myricetin was more active than its glycosidic form (myricitrin or myricetin-3-rhamnoside) when assessed for antioxidant property in methyl linoleate, a non-polar lipid model widely used in antioxidant testing [23]. Myricetin is generally more potent than its glycosides due to the presence of more hydroxyl groups in the benzene ring structures, namely, the presence of a 3-hydroxyl group (C-ring) and a 5-hydroxyl group (A-ring), a 3′,4'-dihydroxy structure (B-ring), and a 2,3-double bond in combination with a 4-oxo group (C-ring) [24,25]. The weaker antioxidant property of myricitrin has been attributed to steric hindrance contributed by the rhamnoside attached to the flavonol at the C3 position [23]. Nevertheless, myricitrin showed higher ferric reducing power compared to myricetin.

The higher TPC in F2 compared to the extract is attributed to the predominant presence of myricitrin (77% based on LCMS analysis [11]). Myricetin derivatives isolated from the ethanolic leaf extract of S. malaccense (F2) clearly showed an improved and considerable antioxidant property when compared to the extract. F2 reflected lower TPC, DPPH, and ABTS scavenging activities in contrast to the standard compounds (myricetin and myricitrin). The presence of other myricetin derivatives that contain arabinofuranoside and glucoside in addition to rhamnoside (myricitrin) could possibly caused a concerted effect in attenuating the scavenging property. The FRAP potential of F2 was higher than that of myricetin but lower than that of myricitrin. However, the NO scavenging property of F2 was stronger than myricetin and myricitrin (Table 2). This clearly reflected the presence of the synergic antioxidant potential of the isolated compounds. The varied antioxidant property of F2 indicates that antioxidant activity is not necessarily always correlated to the phenolic content of the compounds [26,27]. The antioxidant property of phenolic compounds is dependent on their modes of action (ability to scavenge, inhibit, or chelate), as well as their stereochemical structures [28].

Protective effect of S. malaccense leaf extract and myricetin derivative–rich fraction against oxidative stress in ARPE-19 cells

Retinal cells are often subject to high risk of oxidative stress due to ROS, particularly H₂O₂ as they are exposed to oxygen tensed-environment [29,30]. H₂O₂ is constantly generated through oxygen metabolism in mitochondria, light irradiation of melanin, and outer segment phagocytosis of shed photoreceptors [31-33]. H₂O₂ easily crosses cell membranes, confines into various parts of cells, and breaks down to hydroxyl radicals, which contribute to the significant cell cytotoxic condition [31,34,35]. The oxidant can be added directly in pulse or through enzymatic (GO) production in experimental models. However, the latter has been proposed to be useful as it generates a sustained level of H₂O₂ using glucose in the medium as substrate. This would be favorable for establishing a steady chronic or mild stress condition which can be followed by downstream measurement of the injury in cells in addition to evaluating the protective effects of antioxidants [36]. ARPE-19 is a common retina model used to assess oxidative stress particularly H₂O₂-induced injury to cells [37,38].

In the present study, oxidative stress was induced in ARPE-19 cells using glucose oxidase (to generate H₂O₂) to mimic the pathogenesis of DR. A pretreatment period of the samples for 2 h before the introduction of GO-H₂O₂-induced stress was used in view of the relatively short half-life of myricetin at a higher pH range. Myricetin, an aglycone, is a major metabolic product of the flavonoid glycoside, myricitrin [20,21] and therefore, was used as a standard for comparison in the present study. The stability of bioflavanoids has been shown to be predominantly affected by the pH of the environment [39]. Myricetin and myricitrin have been shown to be unstable at neutral and alkaline pH compared to acidic pH when incubated for 3 h at 37 °C. Myricetin has a shorter half-life (0.51 h) than myricitrin (13.08 h) at pH 6.8 [39] and is more rapidly degraded. A separate study reported a half-life and a degradation rate of 2.11 ± 0.34 h and 0.3343 ± 0.0547 h⁻¹, respectively, for myricetin in phosphate buffer at pH 6.8 [40]. An incubation period of 2 or 4 h was used for myricetin in a previous study to assess the independent potential of the compound to induce transcription factor Nrf2 (phase 2 detoxifying enzyme) in ARPE-19 cells [41]. Thus, in the present study the preincubation period of myricetin, myricitrin, F2 containing myricitrin was optimized to 2 h, followed by an additional 2 h incubation with GO-H₂O₂ to induce oxidative stress. Although the flavonoid compounds used in this study have shown antioxidant properties, one should not rule out the possibility that after a longer period of incubation, they may behave as prooxidants. This can be attributed to the stress level contributed by the compound’s structural characteristics and the concentration used [42,43]. The focus of the present study was to assess the direct protective effects of the bioflavonoids against oxidative stress in retinal cells. However, future investigations should include a longer incubation period to confirm the possible prooxidant effect.

The viability of retinal cells was affected statistically significantly by GO-H₂O₂ induced stress (Figure 1). This was attributed to the cytotoxic effect of H₂O₂ (oxidant) released through the GO enzymatic reaction [36]. The protection against stress-induced cytotoxicity conferred by the test samples (concentration range of 40 or 50 μg/ml) in descending order is as follows: myricetin (40 μg/ml) > myricitrin (40 μg/ml) > S. malaccense leaf extract (50 μg/ml) > F2 (50 μg/ml; Figure 1). F2 exerted better protection of retinal cell viability than the extract at higher concentrations due to the presence of pure myricetin derivatives. The pattern of protection corresponded with the antioxidant properties of the samples (Table 2) as discussed previously in which myricetin was shown to be the most potent antioxidant followed by F2, myricitrin, and the leaf extract. This concurred with the ability of the samples to attenuate intracellular ROS in ARPE-19 cells (Figure 2) and H₂O₂ level in the culture supernatants (Figure 3), effectively. The consistent lower protective effect exerted by myricitrin when compared to myricetin could be attributed to the presence of an additional bulky glycoside group on the compound that could hinder its reaction with ROS.

The findings of the in vitro investigation may not be literally extrapolated to the in vivo condition. Although F2 showed effective protection at a higher concentration, 200 µg/ml (myricitrin content, 154 µg [11]), upon metabolic digestion, it will be degraded to myricetin and other minor components [20,21]. Therefore, the effective dose of the metabolites is likely to be considerably lower. Administration of myricitrin to rats (100 mg/rat; n=5) has been shown to result in urinary excretion of myricetin (1.46 mg) and 3,5-dihydroxy-phenylacetic acid (0.48 mg) as the metabolites in 24 h [21]. It is feasible to speculate that myricetin will be bioavailable at a far lower dose approximately 1.5% (3 µg/ml) from the consumption of F2 at a concentration of 200 µg/ml containing 154 µg of myricitrin. Myricetin was shown to be statistically significantly effective in protecting retinal cells against oxidative stress even at a very low concentration (2.5 µg/ml) as shown in the present study. Thus, even after metabolic digestion to yield myricetin, the compound could greatly attenuate stress-induced complications through direct inhibition or activation of the endogenous protective mechanism. Myricetin is also well-known for its numerous health benefits which include antioxidant, antihyperglycemic, anti-inflammatory, antihypertensive, and anticancer properties [12,44,45]. Therefore, F2 has potential as an alternative therapeutic agent to manage DR in addition to offering other medicinal values.

Molecular mechanism of F2 and myricetin in protecting ARPE-19 cells against oxidative stress

The mixture of myricetin derivatives present in F2 was found to exert a more prominent and consistent protective effect compared to myricitrin only (Figure 1, Figure 2, Figure 3). Myricitrin was expected to reflect similar effects in the gene expression study. However, myricetin, a major metabolic product of myricitrin, showed considerably better protection against oxidative stress in the retinal cells compared to all the tested compounds and extract in the present study. Therefore, the molecular mechanism of F2 and myricetin (instead of myricitrin) in preventing GO-H₂O₂-induced stress in retinal cells was assessed. An F2 concentration of 200 µg/ml (myricitrin content, 154 µg) was selected for the purpose of determining the molecular mechanism as this concentration showed higher restoration of cell viability (91%) which was comparable to the protective effect (88%) exerted by myricetin (40 µg/ml).

The antioxidant adaptive mechanism in retinal cells is crucial and thus, serves as a good cellular model to be explored [46]. SOD is one of the major antioxidant enzymes responsible for protecting RPE cells [47]. The cytosolic (CuZnSOD; SOD1) and mitochondrial (MnSOD; SOD2) enzymes dismutate superoxides into oxygen and H₂O₂ [48,49]. H₂O₂ is degraded to water (H₂O) by catalase (CAT) or glutathione peroxidase (GPx) [48]. GO-H₂O₂-induced oxidative stress caused upregulation of SOD2 and GPx1 when compared to the untreated ARPE-19 cells (Figure 4). Presumably, the cells activated the internal antioxidant system immediately to combat stress factors. The introduction of stress along with F2 or myricetin pretreatments showed different response patterns in the expression of antioxidant enzyme genes. F2 upregulated SOD2 (p<0.001) to exert its protective effect, but myricetin did not affect the expression of SOD2 induced by GO-H₂O₂-induced stress. However, myricetin downregulated GPx1 gene expression. The compound possibly scavenged or inactivated the radicals directly, thus attenuating the activation of the internal antioxidant defense mechanism. Expression of the SOD1 gene was not statistically significantly affected in the presence of stress with or without the samples (myricetin or F2). However, F2 activated SOD2 to manage the oxidative stress in ARPE-19 cells.

The nuclear factor erythroid 2-related factor 2-antioxidant response element (Nrf2/ARE) pathway plays an important role in protecting RPE cells through activation of the endogenous protection mechanism against oxidative stress [15,50]. Nrf2 also targets and interacts with apoptosis-related pathways [50,51]. F2 showed statistically significant upregulation of NFE2L2 indicating activation of the Nrf2 pathway to exert its protective effect (Figure 4). This finding corresponds to previous reports that showed the protective effects of bioactive compounds, such as simvastatin, 6-shogaol, and astaxanthin, against H₂O₂-induced stress through the activation of the Nrf2 pathway [15,52,53]. Upregulation of NFE2L2 which indicates activation or expression of Nrf2 was not observed with myricetin treatment against GO-H₂O₂-induced stress (Figure 4) in the present study. This appears to contradict a previous report that showed myricetin induces the expression of Nrf2 independently when incubated for 2 or 4 h in ARPE-19 cells [41]. In the present study, the effect of myricetin alone was not assessed; instead, the effects of GO-H₂O₂ in the presence of myricetin were assessed. Pretreatment of myricetin (40 µg/ml) showed a negligible presence of H₂O₂ in culture supernatant (Figure 3). It is possible that myricetin exerted a direct inhibitory effect on GO-H₂O₂ and thus, conferred exogenous protection before the compound could activate the endogenous phase 2 enzyme protective mechanism (Nrf expression) in the cells during the incubation period.

The cytotoxic effect of steady H₂O₂ triggers apoptosis in cells due to DNA damage [54,55]. Generally, the MAPK1 (ERK) or MAPK14 (p38MAPK) signal transduction pathway is activated instantly in the presence of oxidative stress. However, activation of the pathways reported previously showed contradictory results of either inhibition or further enhancement of cell death in the retinal cell model [56-58]. In the present study, GO-H₂O₂-induced ARPE-19 cell death was accompanied by the upregulation of MAPK1 and the downregulation of MAPK14 (Figure 4). Myricetin showed downregulation of the MAPK1 gene, and this inverse expression of the gene suggested a counter-protective mechanism of myricetin in the stress model. However, F2 and myricetin did not exhibit statistically significant changes in the expression of the MAPK14 gene. Caspase 3 (CASP3) was activated twofold with the GO-H₂O₂ cytotoxicity treatment (Figure 4), and this concurred with a previous report that confirmed CASP3 activation and its association with apoptosis in ARPE-19 cells caused by H₂O₂ [59]. Myricetin normalized the upregulation of CASP3 caused by the stress. HSPB1 was activated upon GO-H₂O₂ treatment in this study, which concurred with a previous report [58], but it was downregulated by myricetin. Myricetin also showed an expression pattern of the Bax gene (a proapoptotic protein) that implied elimination of the basal apoptosis of the cells during the preincubation for 2 h. CTSD is a lysosomal enzyme which is involved in the phagocytosis of photoreceptor outer segment in RPE. The protein has been recognized as an important marker in the pathogenesis of several vitreoretinal diseases which includes DR [60]. However, GO-H₂O₂ treatment with or without the presence of the samples (F2 or myricetin) did not statistically significantly affect the expression of the CTSD gene when compared to the untreated control.

The expression of proinflammatory cytokine IFN-γ was upregulated in ARPE-19 cells approximately by 11-fold upon exposure to GO-H₂O₂-induced oxidative stress (Figure 4), and this expression was reduced statistically significantly in the myricetin-treated cells. Introduction of IFN-γ is known to be able to stimulate nitric oxide synthase and thus, NO production in RPE cells [61]. Inducible nitric oxide synthase (iNOS) is believed to be involved in the breakdown of the blood–retinal barrier in DR [62]. In addition, in vivo iNOS expression has been shown to induce apoptosis in the retina and thus, retinal degeneration in ischemic proliferative retinopathy [63]. In this study, upregulation of NOS2 led to activation of IFN-γ in GO-H₂O₂-induced oxidative stress (Figure 4). Interestingly, F2 downregulated NOS2 which supported its NO scavenging property as described previously. The findings indicate that F2 attenuated the cell death caused by GO-H₂O₂-induced oxidative stress partly through the downregulation of the NOS2 gene though it did not affect the expression pattern of the other apoptosis-related factors (the MAPK1, CASP3, HSPB1, and Bax genes; Figure 4). The focus of the gene expression analysis in the present study was mainly on investigating the effect of H₂O₂-induced cytotoxicity followed by apoptosis in DR. However, the present data are not sufficient to clearly rule out the possible occurrence of necrosis in the cells due to the injury induced by GO-H₂O₂ and thus, warrants further investigations.

Lysyl oxidase (LOX) is an enzyme responsible for maintaining the stability and integrity of the extracellular matrix (ECM) of retinal cells. The activity and expression of LOX were shown to be increased upon exposure to vascular endothelial growth factor (VEGF), a proangiogenic agent closely related to retinopathy [64]. In this study, GO-H₂O₂-induced oxidative stress also upregulated LOX indicating disorganization of the ECM (Figure 4). Myricetin downregulated the expression of the LOX gene, which is indicative of myricetin’s protective activity against GO-H₂O₂-induced oxidative stress compared to F2 treatment.

CRP is another inflammatory marker believed to be associated with the pathogenesis of DR [65]. Whereas TLR2 has been recently recognized as a DR marker as its mRNA and protein expression was significantly elevated in the hyperglycemic condition. TLR2 plays a vital role in inflammation and innate immune responses. The rise in the expression of TLR2 was linked to inflammation and ROS [66]. However, the expression of the CRP and TLR2 genes for the samples did not show any statistically significant differences compared to the untreated or GO-H₂O₂ treated controls.

In conclusion, myricetin derivatives isolated from leaf extract of S. malaccense (F2) showed considerable potential as potent antioxidants and exerted good protective effect against GO-H₂O₂-induced oxidative stress in ARPE-19 cells. The derivatives deactivated the nitric oxide producer (NOS2) and activated the expression of Nrf2, as well as the antioxidant enzyme (SOD2) to manage the oxidative stress and exert a protective effect at the transcriptional level. In addition, myricetin, the potential metabolic product of F2, prevented apoptosis of the retinal cells by controlling various proapoptotic factors and inflammatory markers at the molecular level. However, the possible molecular mechanisms of F2 and myricetin against GO-H₂O₂-induced oxidative stress using an in vitro model in the present study must be further validated through protein expression, as well as in vivo studies. Nevertheless, the present study has clearly reflected that myricetin derivatives of S. malaccense leaf extract have the ability to protect retinal cells against oxidative stress and therefore, may be used as a potential agent for managing DR.