Cyanidin-3-Glucoside Attenuates 4-Hydroxynonenal- and Visible Light-Induced Cite This: Food Funct., 2019, 10, 2871 Retinal Damage in Vitro and in Vivo

Food & Function

View Article Online PAPER View Journal | View Issue

Cyanidin-3-glucoside attenuates 4-hydroxynonenal- and visible light-induced Cite this: Food Funct., 2019, 10, 2871 retinal damage in vitro and in vivo

Yong Wang, *†a Wentao Qi,†a Yazhen Huo,b Ge Song,a Hui Sun,a Xiaoxuan Guoc and Chengtao Wang*d

4-Hydroxynonenal (HNE) is a highly reactive end-product of lipid peroxidation reaction that leads to retinal pigment epithelial (RPE) cell damage. Cyanidin-3-glucoside (C3G), the most abundant anthocyanin in the edible parts of plants, is a nutritional supplement used for preventing retinal damage. However, the protective effect of C3G against HNE-induced RPE cell damage remains to be elucidated. The protective mechanisms of C3G on ARPE-19 cells after HNE exposure were investigated in this study. Results showed that compared with HNE-treated cells, the viability of ARPE-19 cells was significantly (P < 0.05) increased after 1 and 5 μM C3G treatment. C3G exhibited a significant (P < 0.05) inhibitory effect on the expression of senescence-associated β-galactosidase in ARPE-19 cells. VEGF levels in the C3G groups were significantly (P < 0.05) decreased relative to those of the HNE-treated group. C3G also regulated the release of Received 10th February 2019, two inflammatory mediators, namely monocyte chemoattractant protein 1 and interleukine-8, in ARPE-19 Accepted 16th April 2019 cells after HNE treatment. Furthermore, C3G attenuated retinal cell apoptosis in pigmented rabbits DOI: 10.1039/c9fo00273a induced by visible light. Therefore, our data showed that C3G has efficient protective effects on HNE- rsc.li/food-function induced apoptosis, angiogenesis, and dysregulated cytokine production in ARPE-19 cells.

Introduction light, and phagocytosis of the shed photoreceptor outer segments.4 Senescence and inflammation disturb the function of The retina is particularly vulnerable to photooxidative damage RPE, eventually leading to retinal degeneration.5 Dysfunction caused by lipid peroxidation because it has a high polyun- or progressive degeneration of RPE plays a central role in the saturated fatty acid (PUFA) content, high oxygen tension, and pathogenesis of age-related macular degeneration (AMD), high levels of light exposure.1 Lipid peroxidation products, which is one of the major causes of blindness in the elderly such as 4-hydroxynonenal (HNE) and malondialdehyde (MDA), people worldwide.6 can attack proteins, DNA, and phospholipids, resulting in the HNE is the major reactive aldehyde derived from the none- dysfunction of retinal pigment epithelium (RPE) and photo- nzymatic oxidation of n-6 PUFA, such as linoleic acid and ara- receptor cell damage.2,3 RPE is a single-cell layer in the eye chidonic acid.7 Proteomic studies showed the presence of oxi- located between the photoreceptors and vascular choroid. It dative modifications of proteins with lipid peroxidation pro- plays critical roles in the maintenance of the underlying photo- ducts namely, HNE and MDA, in RPE cells from human donor receptors by transporting nutrients from the vascular choroid, eyes.8 HNE is a highly reactive end-product of lipid peroxi- forming the blood-retinal barrier, absorbing the scattered dation reaction that can reduce lysosomal degradation and eventually induce RPE permeability and cell death.9 Anthocyanins (ACNs) are beneficial for vision because they a Academy of State Administration of Grain, Beijing 100037, stimulate regeneration of rhodopsin, inhibit lens opacities, People’s Republic of China. E-mail: [email protected] improve night vision, and inhibit photooxidative-induced bState Key Laboratory of Biomacromolecules, Institute of Biophysics, 10–12 Chinese Academy of Sciences, Beijing 100101, People’s Republic of China apoptosis in RPE cells. Dietary ACNs are widely regarded cInstitute of Quality Standard and Testing Technology for Agro-products, to possess protective eﬀects on light-induced retinal degener- Chinese Academy of Agricultural Sciences, Beijing 100081, ation. Bilberry ACNs can eﬀectively protect against blue light- ’ People s Republic of China induced N-retinylidene-N-retinylethanolamine (A2E) photo- dBeijing Engineering and Technology Research Center of Food Additives, oxidation and membrane permeabilization in RPE cells.11 The Beijing Technology & Business University (BTBU), Beijing 100048, People’s Republic of China. E-mail: [email protected] protective function of bilberry ACNs may be caused by increas- †These authors contributed equally to this work. ing the antioxidant defense mechanisms, suppressing proin-

This journal is © The Royal Society of Chemistry 2019 Food Funct.,2019,10,2871–2880 | 2871 View Article Online

Paper Food & Function flammatory cytokines and inhibiting retinal cell apoptosis and the protective eﬀect of C3G pretreatment. After the in vivo.13 Cyanidin-3-glucoside (C3G) is the most abundant designed treatments, 10 μL of CCK-8 solution was added into ACN found in blackberry and purple rice with great beneficial 96-well plates and OD 450 nm was measured by using a 2300 potential for preventing retinal degeneration.14,15 EnSpire Multimode Plate Reader (PerkinElmer, Waltham, MA,

The direct oxidant H2O2 had been used in most previous USA). The LDH activity was determined by using a commercial studies that induced RPE cell oxidative damage models.16,17 kit according to the manufacturer’s instructions. We preferred HNE because RPE cells are susceptible to the type of photooxidative damage induced by lipid peroxidation.18 Cell apoptosis measurement In our previous studies, the HNE levels were significantly ARPE-19 cell apoptosis was detected by using the Annexin- increased in bisretinoid-burdened rod outer segments after V-PI/FITC apoptosis detection kit according to the manufac- blue light irradiation in vitro and in the retinal cells after turer’s instructions. In brief, after the indicated treatments, visible light exposure in vivo.14,19 Therefore, the present work the cells were washed, harvested, and resuspended in 100 μL investigated the protective eﬀects of C3G against HNE-induced Annexin binding buﬀer. The cells were stained with 5 μL damage in ARPE-19 cells and visible light-induced retinal Annexin V-FITC and 5 μL propidium iodide solution, and then damage in vivo. the reaction system was incubated at room temperature in the dark. Flow cytometry analysis was performed by using a FACScan flow cytometer (Becton Dickinson, Franklin Lakes, Materials and methods NJ, USA). Chemicals and reagents Senescence-associated β-galactosidase activity C3G was obtained from Nanjing Jingzhu Bio-technology, Ltd Senescence was assayed by using the senescence-associated (Nanjing, Jiangsu, China). HNE was obtained from BioVision β-galactosidase staining kit according to our previous study.15 ’ Research Products (Mountain View, CA, USA). Dulbecco s The treated ARPE-19 cells were washed twice with PBS and ’ ’ modified Eagle s/Ham s F12 medium (DMEM/F12) was then fixed with 2% formaldehyde and 0.2% glutaraldehyde in obtained from Invitrogen (Carlsbad, CA, USA). Fetal bovine PBS at room temperature for 4 min. Afterward, the cells were serum (FBS) was obtained from Biological Industries Co. washed with PBS and incubated in the dark for 8 h at 37 °C (BioInd, Israel). Penicillin and streptomycin were obtained with the fresh β-galactosidase staining solution and photo- from Gibco Life Technologies (Grand Island, NY, USA). Cell graphed under a light microscope. The senescent cell percen- Counting Kit-8 (CCK-8) was obtained from Dojindo Molecular tage was calculated as the number of blue-stained cells divided Technologies, Inc. (Kumamoto, Japan). Lactate dehydrogenase by the number of total cells. (LDH) kit and senescence-associated β-galactosidase staining kit were obtained from Beyotime (Shanghai, China). Annexin Enzyme-linked immunosorbent assay (ELISA) V-FITC apoptosis detection kit was obtained from BD After C3G and HNE treatments, the cell culture supernatants Biosciences Pharmingen (San Diego, CA, USA). Deionized were collected and subsequently centrifuged at 1000g for water was produced by using a Milli-Q water-purification 5 min. The expression levels of monocyte chemoattractant system (Millipore, Billerica, MA, USA). All other chemicals and protein 1 (MCP-1), interleukine-8 (IL-8), and vascular endo- reagents were obtained from Sigma-Aldrich (St Louis, MO, thelial growth factor (VEGF) in the supernatants were USA). measured via the corresponding commercial ELISA kits (CUSABIO, Wuhan, Hubei, China), following the manufac- Cell culture and treatment turer’s instructions. The absorbance was measured at 450 nm The ARPE-19 cells were obtained from the American Type by using a 2300 EnSpire Multimode Plate Reader. Culture Collection (Manassas, VA, USA) and cultured according to our previous studies.14,15 They were cultured in DMEM/F12 Real time quantitative polymerase chain reaction (RT-qPCR) with 10% FBS in a humidified incubator at 37 °C in 5% CO2, Total RNA was extracted with Trizol (Tiangen Biotech, Beijing, −1 −1 supplemented with 100 U mL penicillin and 100 μgmL China). Reverse transcription was carried out by using the streptomycin. The cells were seeded in 6-well or 96-well plates High Capacity cDNA Reverse Transcription Kit (Applied and cultivated with serum-containing DMEM/F12 until conflu- Biosystems, Foster City, CA, USA). Real-time PCR was per- ence. Then, they were washed once with the serum-free formed with the QuantStudio 3D Digital PCR (Thermo Fisher medium before C3G and HNE were added in the fresh serum- Scientific, Waltham, MA, USA) by using SYBR Green master free medium as indicated. After pretreatment of ARPE-19 cells mix (Tiangen Biotech) following the manufacturer’s instruc- with C3G for 4 h, HNE dissolved in ethanol was incubated tions. The gene expression in each sample was normalized −ΔΔ with the cells for another 24 h. (2 Ct)toβ-actin expression and expressed as fold change against the control group. The primer sequences used were Cell viability assays as follows: for MCP-1,5′-CTCATAGCAGCCACCTTCATTC-3′ Cell viability was measured on the basis of the CCK-8 and LDH (forward) and 5′-TCACAGCTTCTTTGGGACACTT-3′ (reverse); for enzyme activity to determine the cytotoxicity of HNE treatment IL-8,5′-GACATACTCCAAACCTTTCCACCC-3′ (forward) and

2872 | Food Funct.,2019,10,2871–2880 This journal is © The Royal Society of Chemistry 2019 View Article Online

Food & Function Paper

5′-CCAGACAGAGCTCTCTTCCATCAG-3′ (reverse); for β-actin,5′- Results GGATGCAGAAGGAGATCACTG-3′ (forward) and 5′-CGATCCAC- ACGGAGTACTTG-3′ (reverse). HNE-induced ARPE-19 cytotoxicity As shown in Fig. 1A and B, various concentrations of HNE (10 Pigmented rabbit treatment with C3G and exposure to visible to 100 μM) were applied to the ARPE-19 cells, and 30 μM HNE light induced a significant reduction in cell viability when com- A total of 24 pigmented rabbits weighing 2.5–3.0 kg were pared to the untreated control cells via CCK-8 and LDH obtained from the Animal Center of Beijing Kaiyuan Co. enzyme activity assays (P < 0.05). Cell apoptosis was also sig- μ (Beijing, China) and maintained for 1 week before the experi- nificantly increased by 17.4% ± 2.9% after 30 M HNE treat- μ ments. The experiments were performed according to the ment (P < 0.05; Fig. 1C). Thus, 30 M HNE was used to induce Animal Management Rules of the Ministry of Health of the ARPE-19 cell oxidative damage in this study to establish the People’s Republic of China (documentation no. 55 (2001), HNE-induced ARPE-19 cell damaged model. Ministry of Health of P.R. China). The research protocol was Inhibitory effect on HNE-induced cell death approved by the Ethical Committee for Animal Experimentation of the First Hospital Affiliated to the General We pretreated C3G to the ARPE-19 cells 4 h before HNE ff Hospital of the Chinese People’s Liberation Army. All rabbits exposure to explore whether C3G has protective e ects against μ were individually housed in standard cages and placed in a HNE-induced ARPE-19 cell damage. After 1 and 5 M C3G room with a 12 h/12 h light/dark cycle at 22 °C–25 °C and treatment, the viability of ARPE-19 cells was significantly 55%–60% humidity. All rabbits were provided with a standard increased when compared with only HNE-treated cells (P < maintenance diet (Beijing KEAO XIELI FEED Co., Ltd, China) 0.05; Fig. 2A). In addition, HNE-induced LDH releases were μ and water ad libitum. significantly diminished by 1 and 5 M C3G treatment (P < μ The rabbits were randomly divided into three groups as 0.05; Fig. 2B). After 1 and 5 M C3G treatment, the cell apopto- follows: the control group (no light exposure and vehicle sis ratios induced by HNE were also significantly decreased by administration, n = 8), the retinal damage model group (light 26.3% ± 2.1% and 51.4% ± 3.6%, respectively, measured by exposure and vehicle administration, n = 8), and the C3G using flow cytometry (P < 0.05; Fig. 2C and D). −1 group (light exposure and administration of C3G, 50 mg kg Inhibitory effect on senescence-associated β-galactosidase −1 day , n = 8). C3G was dissolved in PBS and administered activity orally to the rabbits. The rabbits were treated with C3G for 2 β weeks of pre-illumination and 1 week of post-illumination The positive staining of -galactosidase was significantly μ until sacrifice. greater after 30 M HNE treatment (P < 0.05; Fig. 3), and the The light exposure method was adopted from our previous proportion of the positively stained cells reached 52.0% ± μ studies.13 In brief, after acclimatization to the dark (60–100 lx) 4.8%. At concentrations of 1 and 5 M, C3G exhibited a signifi- ff for 24 h, the nonanesthetized rabbits were exposed to four cant (P < 0.05) inhibitory e ect on the expression of senes- β diffused cool-white fluorescent lights at 18 000 ± 1000 lx for cence-associated -galactosidase. The proportion of the posi- μ 2 h in cages with reflective interiors. After light exposure, the tively stained cells decreased to 33.2% ± 1.5% in the 5 M C3G rabbits were placed in the dark for 24 h. group. Inhibitory effect on VEGF overexpression Histologic analysis In Fig. 4, a significant (P < 0.05) aggravation of VEGF release The rabbits were sacrificed, and the eyeballs were quickly enu- after HNE treatment indicated that HNE induced the over- cleated and fixed in a fixative solution containing 2.5% glutar- expression of VEGF in ARPE-19 cells. In the present study, the ffi aldehyde and 2% paraformaldehyde. Twelve para n- inhibition of VEGF overexpression by C3G in ARPE-19 cells was μ embedded 4 m thick sections were cut from the optic disc of observed. At 1 and 5 μM, the VEGF levels in the C3G groups each eye, stained with hematoxylin and eosin (HE) and viewed were significantly (P < 0.05) decreased by 60.1% ± 2.4% and with a light microscope (Leica, Heidelberg, Germany). The 70.7% ± 2.3%, respectively, relative to those of the HNE-treated neurosensory retina thickness was measured within group. 250–2750 μm (counted at 500 μm intervals) of the superior and inferior edges to the optic nerve head. The mean neuro- Modulatory effect on the protein and mRNA expression of sensory retina thickness was calculated from 12 sections for MCP-1 and IL-8 each eye. Our results show that the exposure of ARPE-19 cells to 30 μM HNE significantly diminished the basal production of MCP-1 Statistical analysis and IL-8 (P < 0.05; Fig. 5A and B), suggesting that HNE led to The differences between the groups were analyzed via one-way ARPE-19 cell dysfunction. However, 5 μM C3G pretreatment to ANOVA with Origin version 8.5, followed by Tukey’s tests. ARPE-19 cells for 4 h led to a significant increase of MCP-1 P < 0.05 was considered statistically significant. Data were pre- and IL-8 release compared with that of the HNE-treated group sented as the mean ± standard deviation. (P < 0.05). Consistent with their protein expression, the MCP-1

This journal is © The Royal Society of Chemistry 2019 Food Funct.,2019,10,2871–2880 | 2873 View Article Online

Paper Food & Function

Fig. 1 Cytotoxicity of HNE on ARPE-19 cells. (A) and (B) Cell viabilities were measured by using CCK-8 and LDH enzyme activities, respectively. (C) Cell apoptosis was determined by Annexin V/PI staining. Data were expressed as the mean ± standard deviation (n = 3). *P < 0.05 indicates significantly different from the normal cells. Bars with different letters are significantly different from each other (P < 0 05).

and IL-8 mRNA levels were also significantly decreased after HNE din-3-glucoside), C3G has been the most effective ACN in treatment (P < 0.05; Fig. 5C and D). After 5 μM C3G treatment, terms of antioxidant, anti-angiogenic, and anti-ageing pro- the cells expressed significantly higher amounts of MCP-1 and perties in the visible light-induced RPE damage model IL-8 mRNA than those only treated with HNE (P <0.05). in vitro.15 Our previous study confirmed that C3G protected RPE cells against light damage by inhibiting the photooxida- Protective effect on neurosensory retina thickness tion and photodegradation of bisretinoids.14 In addition, our We determined whether the retinal structure was protected by previous study showed that C3G attenuated retinal damage via C3G supplementation in vivo. At 7 days after light exposure, a the activation of the Nrf2/HO-1 pathway and NF-κB suppres- 23 significant decrease in the neurosensory retina thickness was sion in vivo. Previous reports have shown that ACNs such as observed in the light-exposed rabbits (P < 0.05; Fig. 6). After C3G were identified in the ocular tissues and plasma by oral 24,25 C3G administration, the thinning of the neurosensory retina administration and ACN species were distributed in the 26 was significantly increased by 16.23% ± 5.6% (P < 0.05), imply- muscle tissues of the outer layer of the retina, suggesting the ing that C3G treatment protected the retinal cells from light- possible protective effect of oral ACNs for retinal diseases. In induced damage. addition, high concentrations of protocatechuic and ferulic acids, the major human metabolites of C3G, may explain the short-term increase in plasma and retina antioxidant activity Discussion observed after C3G consumption.27 Previous studies indicated that light irradiation triggered In recent years, dietary ACNs have received increasing atten- the dysfunction of RPE cells, thereby leading to retinal photo- tion as a potential nutritional supplement for preventing receptor cell apoptosis, eventually decreasing the retina – retinal degeneration.20 22 Among the four ACNs (i.e., C3G, thickness.28,29 In the present study, C3G treatment significantly pelargonidin-3-glucoside, delphinidin-3-glucoside, and malvi- increased the neurosensory retina and outer nuclear layer thick-

2874 | Food Funct.,2019,10,2871–2880 This journal is © The Royal Society of Chemistry 2019 View Article Online

Food & Function Paper

Fig. 2 Inhibitory effect of C3G on HNE-induced ARPE-19 cell death. (A) and (B) Cell viabilities were measured via CCK-8 and LDH enzyme activities, respectively. (C) and (D) Apoptosis was measured via flow cytometry using Annexin V/PI staining. ARPE-19 cells were pretreated with C3G (1 and 5 μM) for 4 h, and then 30 μM HNE was challenged for 24 h in the serum-free medium. Data were expressed as the mean ± standard deviation (n = 3). #P < 0.05 indicates significantly different from the normal cells; *P < 0.05 indicates significantly different from the model cells. Bars with different letters were significantly different from each other (P < 0 05).

ness, suggesting that C3G can protect retinal cells from 0.05). However, the level of HNE in the rabbits treated with − − apoptosis. 50 mg kg 1 day 1 C3G was significantly decreased than that in − Damage to retinal proteins due to modifications induced by the light-exposed rabbits (1.23 ± 0.15 ng mg 1 vs. 2.14 ± 0.17 − lipid peroxidation products, such as HNE, has been detected ng mg 1, respectively; P < 0.05). Phagocytosis of the rod outer in vivo as a result of light irradiation.30 In this study, HNE segment constituents damaged by the reactive lipid aldehyde levels in retinas were significantly increased in the light- HNE is likely to be a source of material resistant to lysosomal exposed rabbits compared with those in the control group degradation, finally resulting in abnormally increased lipofus- − − (2.14 ± 0.17 ng mg 1 vs. 1.08 ± 0.08 ng mg 1, respectively; P < cin.31 Lipofuscin and its dominant fluorophore A2E in drusen

This journal is © The Royal Society of Chemistry 2019 Food Funct.,2019,10,2871–2880 | 2875 View Article Online

Paper Food & Function

Fig. 3 Inhibitory effect of C3G on HNE-induced ARPE-19 cell senescence. Senescence was assayed by using the senescence-associated β-galactosidase staining kit. ARPE-19 cells were pretreated with C3G (1 and 5 μM) for 4 h, and then 30 μM HNE was challenged for 24 h in the serum-free medium. Data were expressed as the mean ± standard deviation (n = 3). #P < 0.05 indicates significantly different from the normal cells; *P < 0.05 indicates significantly different from the model cells.

exert phototoxic effects in RPE cells (aging and dysfunction), apoptosis of ARPE-19 cells treated with different concen- which may result in the pathogenesis of AMD.32,33 Thus, high trations of HNE, and then 30 μM HNE was chosen. HNE levels lead to the apoptotic degeneration of RPE cells, Animal studies demonstrated increased protein modifi- which is one of the molecular bases for AMD.34 cation via reactive aldehyde HNE in the RPE layer upon light HNE promotes cell proliferation at low concentrations, but irradiation, which is an early event that precedes photo- it depletes intracellular antioxidants at high concentration, receptor cell apoptosis.30 Thus, the formation of protein modi- leading to necrosis or apoptosis.35 It has been reported that in fication via HNE is an indicator for assessing the effect of anti- the plasma and erythrocytes of humans, the normal physio- oxidants against light-induced retinal damage.19,23 In our pre- logical concentration of HNE is 0.1–10 μM.36 But in biological vious study, treatment with C3G (200 and 500 μM) before membranes exposed to oxidative stress, the concentration of irradiation with 430 nm light markedly decreased HNE-BSA HNE will be between 10 μM and 1.0 mM.37 Based on two pre- levels in the all-trans-retinal-laden rod outer segments irra- vious studies,38,39 we further determined the cell viability and diated at 430 nm.14 In addition, the level of HNE was upregu-

2876 | Food Funct.,2019,10,2871–2880 This journal is © The Royal Society of Chemistry 2019 View Article Online

Food & Function Paper

lated after light exposure but attenuated by C3G in vivo.23 Thus, a possible mechanism for the protective eﬀects of the C3G in this work could be that C3G inhibited HNE formation in ARPE-19 cells. Hytti et al. also revealed that dietary polyphenols such as quercetin, fisetin and luteolin protected ARPE-19 cells from HNE-induced oxidative stress and inflammation,38,39 and these compounds reduced the activation of MAPKs and CREB in stress conditions. Thus, there may be several mechanisms of the protective action of C3G against HNE-induced cytotoxicity, and its antioxidant action and regulation of inflammation and redox signaling may be of special significance. Nevertheless, this assumption needs to be confirmed in future works.

Exposure to oxidants, such as H2O2 and tert-butylhydroper- oxide, can stimulate senescence-associated β-galactosidase 40 Fig. 4 Inhibitory effect of C3G on HNE-induced VEGF overexpression activity in RPE cells. Here, our result indicates that 30 μM in ARPE-19 cells. ARPE-19 cells were pretreated with C3G (1 and 5 μM) HNE induced ARPE-19 cell senescence, a phenomenon that for 4 h, and then 30 μM HNE was challenged for 24 h in the serum-free has been implicated in the early pathogenesis of AMD. The n medium. Data were expressed as the mean ± standard deviation ( =3). possible mechanisms behind C3G alleviation of RPE cellular #P < 0.05 indicates significantly different from the normal cells; *P < 0.05 indicates significantly different from the model cells. senescence include telomere shortening inhibition and anti-oxidative stress activity.41 The RPE layer is the source of VEGF that is implicated in angiogenesis and is crucial to retinal microenvi- ronment.42 Excessive light exposure induces VEGF overexpression

Fig. 5 Effects of C3G on (A, C) MCP-1 and (B, D) IL-8 (A, B) proteins, and (C, D) mRNA levels in ARPE-19 cells after HNE treatment. ARPE-19 cells were pretreated with C3G (1 and 5 μM) for 4 h, and then 30 μM HNE was challenged for 24 h in the serum-free medium. Data were expressed as the mean ± standard deviation (n = 3). #P < 0.05 indicates significantly different from the normal cells; *P < 0.05 indicates significantly different from the model cells.

Paper Food & Function

Fig. 6 Effect of C3G on the neurosensory retina thickness in the pigmented rabbits. (A) Representative micrographs of HE-stained sections of rabbit retinas. Scale bar: 50 μm. (B) Neurosensory retina thickness was measured within 250–2750 μm (counted at 500 μm intervals) from the optic nerve to the superior and inferior sides of the retina. Data were expressed as the mean ± standard deviation (n = 8). #P < 0.05 indicates significantly different from the control group; *P < 0.05 indicates significantly different from the retinal damage model group.

in RPE cells, which increases the risk of AMD.43 In the present C3G pretreatment. The data demonstrated that C3G attenuated study, HNE can induce VEGF overexpression in RPE cells, HNE-induced apoptosis and angiogenesis in ARPE-19 cells. whereas its production was reduced in the presence of C3G. Our study also provided evidence that C3G regulated the MCP-1 and IL-8 are the major cytokines synthesized by RPE release of cytokines MCP-1 and IL-8. Furthermore, C3G attenu- cells, and dysregulation of the chemokines MCP-1 and IL-8 ated visible light-induced retinal degeneration in vivo. The role was implicated in the pathogenesis of neovascular AMD.44 of C3G in modulating angiogenesis and inflammation-related Challenge with 30 μM HNE significantly inhibited the pro- factors may be the mechanisms through which dietary C3G duction of both MCP-1 and IL-8, indicating that the lipid peroxi- can prevent ocular damage. This finding suggests that C3G is dation product HNE results in RPE cell dysfunction.45 We effective in preventing HNE-induced damage in RPE cells and demonstrated here that C3G could protect ARPE-19 cells from may be suitable as a dietary supplement for reducing the risk HNE-induced cytotoxicity and adjust the release of two inflamma- of AMD. tory mediators, namely MCP-1 and IL-8. However, the related mechanisms are still unknown. A recent study showed that sequestosome 1/p62 as an autophagy substrate was a significant Conflicts of interest factor in inflammatory responses and upregulated apoptosis in RPE cells.46 Previous research indicated that C3G regulated the There are no conflicts of interest to declare. autophagy signaling pathway in vitro and in vivo.47,48 Thus, the protective effects of C3G on apoptosis and release of inflammatory mediators in RPE cells may be related to regulating the auto- Acknowledgements phagy signaling pathway in this work. However, this hypothesis remains to be confirmed in future studies. This study was financially supported by the fund of the Beijing In summary, our study clearly showed that the cytotoxic Engineering and Technology Research Center of Food effects of HNE in ARPE-19 cells were significantly rescued by Additives, Beijing Technology & Business University (BTBU),

Food & Function Paper the National Natural Science Foundation of China (No. model in pigmented rabbits, Molecules, 2015, 20, 22395– 31471591), and the Special Funds of Basic Research of Central 22410. Public Welfare Institute (No. ZX1908). 14 Y. Wang, H. J. Kim and J. R. Sparrow, Quercetin and cyanidin-3-glucoside protect against photooxidation and photodegradation of A2E in retinal pigment epithelial cells, Exp. References Eye Res., 2017, 160,45–55. 15 Y. Wang, D. Zhang, Y. Liu, D. Wang, J. Liu and B. Ji, The 1 S. G. Jarrett and M. E. Boulton, Consequences of oxidative protective effects of berry-derived anthocyanins against stress in age-related macular degeneration, Mol. Aspects visible light-induced damage in human retinal pigment Med., 2012, 33, 399–417. epithelial cells, J. Sci. Food Agric., 2015, 95, 936–944. 2 D. T. Organisciak and D. K. Vaughan, Retinal light damage: 16 N. M. M. Saviranta, L. Veeroos, L. J. Granlund, Mechanisms and protection, Prog. Retinal Eye Res., 2010, V. H. Hassinen, K. Kaarniranta and R. O. Karjalainen, Plant 29, 113–134. flavonol quercetin and isoflavone biochanin A differentially 3 M. Enoki, S. Shinto, Y. Matsuoka, A. Otsuka, S. Kaidzu, induce protection against oxidative stress and inflam- M. Tanito, T. Shibata, K. Uchida, A. Ohira, M. Yamatoa and mation in ARPE-19 cells, Food Res. Int., 2011, 44, 109–113. K.-I. Yamada, Lipid radicals cause light-induced retinal 17 X. Cao, M. Liu, J. Tuo, D. Shen and C.-C. Chan, The effects degeneration, Chem. Commun., 2017, 53, 10922–10925. of quercetin in cultured human RPE cells under oxidative 4 O. Strauss, The retinal pigment epithelium in visual func- stress and in Ccl2/Cx3cr1 double deficient mice, Exp. Eye tion, Physiol. Rev., 2005, 85, 845–881. Res., 2010, 91,15–25. 5 M. van Lookeren Campagne, J. LeCouter, B. L. Yaspan and 18 J. J. Hunter, J. I. W. Morgan, W. H. Merigan, D. H. Sliney, W. Ye, Mechanisms of age-related macular degeneration J. R. Sparrow and D. R. Williams, The susceptibility of the and therapeutic opportunities, J. Pathol., 2014, 232, 151– retina to photochemical damage from visible light, Prog. 164. Retinal Eye Res., 2012, 31,28–42. 6 I. Bhutto and G. Lutty, Understanding age-related macular 19 Y. Wang, L. Zhao, C. Wang, J. Hu, X. Guo, D. Zhang, degeneration (AMD): Relationships between the photo- W. Wu, F. Zhou and B. Ji, Protective effect of quercetin and receptor/retinal pigment epithelium/Bruch’s membrane/ chlorogenic acid, two polyphenols widely present in edible choriocapillaris complex, Mol. Aspects Med., 2012, 33, 295– plant varieties, on visible light-induced retinal degener- 317. ation in vivo, J. Funct. Foods, 2017, 33, 103–111. 7 J. Z. Nowak, Oxidative stress, polyunsaturated fatty acids- 20 W. Huang, Z. Yan, D. Li, Y. Ma, J. Zhou and Z. Sui, derived oxidation products and bisretinoids as potential Antioxidant and anti-inflammatory effects of blueberry inducers of CNS diseases: focus on age-related macular anthocyanins on high glucose-induced human retinal degeneration, Pharmacol. Rep., 2013, 65, 288–304. capillary endothelial cells, Oxid. Med. Cell. Longevity, 2018, 8 F. Schutt, M. Bergmann, F. G. Holz and J. Kopitz, Proteins 2018, 1862462. modified by malondialdehyde, 4-hydroxynonenal, or 21 J. M. Silvan, M. Reguero and S. de Pascual-Teresa, A protec- advanced glycation end products in lipofuscin of human tive effect of anthocyanins and xanthophylls on UVB- retinal pigment epithelium, Invest. Ophthalmol. Visual Sci., induced damage in retinal pigment epithelial cells, Food 2003, 44, 3663–3668. Funct., 2016, 7, 1067–1076. 9 T. U. Krohne, N. K. Stratmann, J. Kopitz and F. G. Holz, 22 E. Pawlowska, J. Szczepanska, A. Koskela, K. Kaarniranta Effects of lipid peroxidation products on lipofuscinogen- and J. Blasiak, Dietary polyphenols in age-related macular esis and autophagy in human retinal pigment epithelial degeneration: Protection against oxidative stress and cells, Exp. Eye Res., 2010, 90, 465–471. beyond, Oxid. Med. Cell. Longevity, 2019, 2019, 9682318. 10 H. Matsumoto, Y. Nakamura, S. Tachibanaki, S. Kawamura 23 Y. Wang, Y. Z. Huo, L. Zhao, F. Lu, O. Wang, X. Yang, and M. Hirayama, Stimulatory effect of cyanidin 3-glyco- B. P. Ji and F. Zhou, Cyanidin-3-glucoside and its phenolic sides on the regeneration of rhodopsin, J. Agric. Food acid metabolites attenuate visible light-induced retinal Chem., 2003, 51, 3560–3563. degeneration in vivo via activation of Nrf2/HO-1 pathway 11 Y. P. Jang, J. L. Zhou, K. Nakanishi and J. R. Sparrow, and NF-B suppression, Mol. Nutr. Food Res., 2016, 60, Anthocyanins protect against A2E photooxidation and 1564–1577. membrane permeabilization in retinal pigment epithelial 24 W. Kalt, J. B. Blumberg, J. E. McDonald, M. R. Vinqvist- cells, Photochem. Photobiol., 2005, 81, 529–536. Tymchuk, S. A. E. Fillmore, B. A. Graf, J. M. O’Leary and 12 S. M. Yacout and E. R. Gaillard, The anthocyanins, oenin P. E. Milbury, Identification of anthocyanins in the liver, and callistephin, protect RPE cells against oxidative stress, eye, and brain of blueberry-fed pigs, J. Agric. Food Chem., Photochem. Photobiol., 2017, 93, 590–599. 2008, 56, 705–712. 13 Y. Wang, L. Zhao, F. Lu, X. Yang, Q. Deng, B. Ji and 25 D. Xiao, A. Sandhu, Y. Huang, E. Park, I. Edirisinghe and F. Huang, Retinoprotective effects of bilberry anthocyanins B. M. Burton-Freeman, The effect of dietary factors on via antioxidant, anti-Inflammatory, and anti-apoptotic strawberry anthocyanins oral bioavailability, Food Funct., mechanisms in a visible light-induced retinal degeneration 2017, 8, 3970–3979.

Paper Food & Function

26 S. Yamada, N. Zaima, Y. Yoshimura, S. Inaba, T. Fujimori, 37 H. Esterbauer and P. Ramos, Chemistry and pathophysiol- T. Sogon and T. Moriyama, Visualization of the distribution ogy of oxidation of LDL, Rev. Physiol., Biochem. Pharmacol., of anthocyanin species in mice eyeball by matrix-assisted 1996, 127,31–64. laser desorption/ionization mass spectrometry imaging, 38 M. Hytti, N. Piippo, A. Salminen, P. Honkakoski, Rapid Commun. Mass Spectrom., 2018, 32, 380–384. K. Kaarniranta and A. Kauppinen, Quercetin alleviates 27 C. Czank, A. Cassidy, Q. Zhang, D. J. Morrison, T. Preston, 4-hydroxynonenal-induced cytotoxicity and inflammation P. A. Kroon, N. P. Botting and C. D. Kay, Human metab- in ARPE-19 cells, Exp. Eye Res., 2015, 132, 208–215. olism and elimination of the anthocyanin, cyanidin-3-glu- 39 M. Hytti, N. Piippo, E. Korhonen, P. Honkakoski, coside: a C-13-tracer study, Am. J. Clin. Nutr., 2013, 97, 995– K. Kaarniranta and A. Kauppinen, Fisetin and luteolin 1003. protect human retinal pigment epithelial cells from oxi- 28 M. Sasaki, K. Yuki, T. Kurihara, S. Miyake, K. Noda, dative stress-induced cell death and regulate inflammation, S. Kobayashi, S. Ishida, K. Tsubota and Y. Ozawa, Sci. Rep., 2015, 5, 17645. Biological role of lutein in the light-induced retinal 40 A. L. Yu, R. Fuchshofer, D. Kook, A. Kampik, degeneration, J. Nutr. Biochem., 2012, 23, 423–429. H. Bloemendal and U. Welge-Luessen, Subtoxic oxidative 29 Q. Deng, Y. Wang, C. Wang, B. Ji, R. Cong, L. Zhao, stress induces senescence in retinal pigment epithelial P. Chen, X. Zang, F. Lu, F. Han and F. Huang, Dietary sup- cells via TGF-beta release, Invest. Ophthalmol. Visual Sci., plementation with omega-3 polyunsaturated fatty acid-rich 2009, 50, 926–935. oils protects against visible-light-induced retinal damage 41 Y. Liu, M. Liu, Q. Chen, G.-M. Liu, M.-J. Cao, L. Sun, Z. Lu in vivo, Food Funct., 2018, 9, 2469–2479. and C. Guo, Blueberry polyphenols ameliorate visible light 30 M. Tanito, M. H. Elliott, Y. Kotake and R. E. Anderson, and lipid-induced injury of retinal pigment epithelial cells, Protein modifications by 4-hydroxynonenal and 4-hydroxy- J. Agric. Food Chem., 2018, 66, 12730–12740. hexenal in light-exposed rat retina, Invest. Ophthalmol. 42 J. R. Sparrrow, D. Hicks and C. P. Hamel, The retinal Visual Sci., 2005, 46, 3859–3868. pigment epithelium in health and disease, Curr. Mol. Med., 31 C. Brandstetter, F. G. Holz and T. U. Krohne, Complement 2010, 10, 802–823. component C5a primes retinal pigment epithelial cells for 43 J. S. Penn, A. Madan, R. B. Caldwell, M. Bartoli, inflamrnasome activation by lipofuscin-mediated photooxi- R. W. Caldwell and M. E. Hartnett, Vascular endothelial dative damage, J. Biol. Chem., 2015, 290, 31189–31198. growth factor in eye disease, Prog. Retinal Eye Res., 2008, 27, 32 J. R. Sparrow and M. Boulton, RPE lipofuscin and its role 331–371. in retinal-pathobiology, Exp. Eye Res., 2005, 80, 595–606. 44J.Lechner,M.Chen,R.E.Hogg,L.Toth,G.Silvestri, 33 S. Montserrat-de la Paz, M. Carmen Naranjo, B. Bermudez, U. Chakravarthy and H. Xu, Peripheral blood mononuclear S. Lopez, W. Moreda, R. Abia and F. J. G. Muriana, cells from neovascular age-related macular degeneration Postprandial dietary fatty acids exert divergent inflamma- patients produce higher levels of chemokines CCL2 (MCP-1) tory responses in retinal-pigmented epithelium cells, Food and CXCL8 (IL-8), J. Neuroinflammation, 2017, 14,42. Funct., 2016, 7, 1345–1353. 45 S. Qin and G. A. Rodrigues, Differential roles of AMPK 34 J. Gautam, S. Banskota, S. Shah, J.-G. Jee, E. Kwon, alpha 1 and AMPK alpha 2 in regulating 4-HNE-induced Y. Wang, D. Y. Kim, H. W. Chang and J.-A. Kim, 4- RPE cell death and permeability, Exp. Eye Res., 2010, 91, Hydroxynonenal-induced GPR109A (HCA(2) receptor) acti- 818–824. vation elicits bipolar responses, G(alpha i)-mediated anti- 46E.Korhonen,N.Piippo,M.Hytti,J.M.T.Hyttinen, inflammatory effects and G(beta gamma)-mediated cell K. Kaarniranta and A. Kauppinen, SQSTM1/p62 regulates the death, Br. J. Pharmacol., 2018, 175, 2581–2598. production of IL-8 and MCP-1 in IL-1beta-stimulated human 35 R. Vatsyayan, P. Chaudhary, A. Sharma, R. Sharma, retinal pigment epithelial cells, Cytokine, 2019, 116,70–77. P. C. R. Lelsani, S. Awasthi and Y. C. Awasthi, Role of 47 T. Wan, S. Wang, M. Ye, W. Ling and L. Yang, Cyanidin-3- 4-hydroxynonenal in epidermal growth factor receptor- O-beta-glucoside protects against liver fibrosis induced by mediated signaling in retinal pigment epithelial cells, Exp. alcohol via regulating energy homeostasis and AMPK/auto- Eye Res., 2011, 92, 147–154. phagy signaling pathway, J. Funct. Foods, 2017, 37,16–24. 36 S. Srivastava, B. L. Dixit, J. Cai, S. Sharma, H. E. Hurst, 48 K. Ogawa, Y. Kuse, K. Tsuruma, S. Kobayashi, A. Bhatnagar and S. K. Srivastava, Metabolism of lipid M. Shimazawa and H. Hara, Protective effects of bilberry peroxidation product, 4-hydroxynonenal (HNE) in rat and lingonberry extracts against blue light-emitting diode erythrocytes: Role of aldose reductase, Free Radicals Biol. light-induced retinal photoreceptor cell damage in vitro, Med., 2000, 29, 642–651. BMC Complementary Altern. Med., 2014, 14, 120.