Passion fruit seed extract protects hydrogen peroxide-induced cell damage in human retinal pigment epithelium ARPE-19 cells

Abstract

Age-related macular degeneration (AMD) is a major cause of vision loss among adults. We investigated the protective effects of passion fruit seed extract (PFSE) and its rich polyphenol piceatannol in an AMD cell model in which human retinal pigment epithelial ARPE-19 cells were exposed to hydrogen peroxide (H₂O₂). Using a cell viability WST-8 assay, we revealed that PFSE and piceatannol increased the cellular viability of ARPE-19 cells by 130% and 133%, respectively. Moreover, PFSE and piceatannol recovered the cell viability of ARPE-19 cells, which had decreased to 60% owing to H₂O₂-induced damage, to approximately 84% and 89%, respectively. In addition, we found that the treatment of ARPE-19 cells with H₂O₂ decreased the mitochondrial and glycolytic ATP production rate to approximately 54% that of healthy control ARPE-19 cells using a Seahorse extracellular flux analyzer. Furthermore, pretreatment with PFSE and piceatannol restored the oxidative stress-induced decrease in the mitochondrial and glycolytic ATP production rate to approximately 97% and 82%, respectively. These results indicated the cytoprotective effects of PFSE and piceatannol against oxidative stress in human ARPE-19 cells by resolving the dysfunction of mitochondrial and glycolytic energy metabolism.

Introduction

Oxidative damage is associated with the pathogenesis of visual dysfunctions such as age-related macular degeneration (AMD)^1,2. AMD is a complex multifactorial disease and is one of the major causes of severe visual loss in the elderly population^1,2. Retinal pigment epithelial (RPE) cells, which are located in the outermost retinal layer, are crucial for the normal functioning and health of retinal photoreceptors and play an important role in the visual process; moreover, their oxidative damage contributes to macular degeneration². AMD is categorized into wet and dry types^1,2. Photodynamic therapy, laser photocoagulation, and antivascular endothelial growth factor (anti-VEGF) therapy are used to treat wet AMD². In particular, anti-VEGF therapy is the current standard treatment for suppressing pathological neovascularization in wet AMD². In contrast, dry AMD has no established treatment. Currently, the only Food and Drug Administration–approved treatment is pegcetacoplan, an anticomplement therapy with targeted C3 inhibition². However, it has limited therapeutic effects^2,3. Therefore, effective prevention and further treatment strategies for dry AMD need to be established.

Antioxidants have been proposed as potential preventive or curative options for dry AMD because oxidative stress is a major contributor to RPE cell degeneration². Indeed, clinical studies have revealed that antioxidants and zinc-containing supplements could substantially reduce AMD progression rate². Increasing the consumption of fruits and vegetables that contain antioxidant nutrients and phytochemicals may provide some protection against AMD². Previous study shows that AMD protection and treatment should be focused on rescuing RPE cells from oxidative damage². Some antioxidants, including vitamin C (ascorbic acid), vitamin E (α-tocopherol), curcumin, resveratrol, and catechin, have been investigated as antiaging agents^2,4. Among them, resveratrol has been extensively examined for its antioxidant and free radical scavenging properties^2,5,6.

Piceatannol (3,3′,4′,5-trans-tetrahydroxy-stilbene) is a natural stilbene derivative and a structurally similar polyphenol analog of resveratrol (3,4′,5-trans-trihydroxy-stilbene)^7,8. Several studies have revealed that piceatannol has antioxidative^8–14, anti-inflammatory^8,15–18, and neuroprotective properties^19–22. Kang et al. reported that resveratrol and piceatannol protect ARPE-19 cells from N-retinylidene-N-retinylethanolamine (A2E)- and blue light–induced photodamage²³. Additionally, Hao et al. reported the preventive function of piceatannol in hydrogen peroxide (H₂O₂)-induced ARPE-19 cell damage and apoptosis through the PI3K/Akt signaling pathway²⁴.

Passion fruit (Passiflora edulis Sims) is a tropical plant, whose seeds contain various nutrients and functional components^25,26. Passion fruit seed extract (PFSE) is rich in piceatannol, a polyphenol, and is gaining attention as a functional food^25–27. Previous studies have reported that piceatannol and/or piceatannol-rich PFSE demonstrated various biological activities such as skin protection^11,27, vasodilatation²⁸, chronic disease prevention²⁹, metabolic improvement^30–32, and anticancer effect³³. In addition, as per the results of our recent study, piceatannol-rich PFSE protects beta-amyloid-induced neuronal cell death in the differentiated human neuroblastoma SH-SY5Y cell model²².

The present study aimed to investigate the metabolic profile of piceatannol and piceatannol-rich PFSE and their protective effect on H₂O₂-induced RPE cell ARPE-19 damage.

Results

PFSE and piceatannol promote cell viability and mitochondrial energy metabolism

In this study, we examined the cytoprotective effects of piceatannol-rich PFSE and piceatannol and analyzed their effects on H₂O₂-induced RPE cell damage and energy metabolism in an AMD cell model using H₂O₂-treated ARPE-19 cells.

First, we investigated the cellular effects of PFSE and piceatannol on cell proliferation/activation and cell viability in ARPE-19 cells. Previously, several studies focused on RPE cell protective effects of the bioactivity of piceatannol in the 5–60 μM concentration range^23,24. Additionally, we have recently reported that PFSE at 0.0023 mg/mL and piceatannol at 1 μM protect beta-amyloid-induced neuronal cell death in an Alzheimer’s disease cell model²². We treated ARPE-19 cells with PFSE (0.023 mg/mL or 0.0023 mg/mL, i.e., piceatannol equivalent of 10 μM and 1 μM, respectively) and piceatannol (10 μM or 1 μM) for 25 h (Fig. 1A). PFSE and piceatannol were administered 30 min after cell seeding when the cells had adhered to the culture plate. The groups treated with 0.023-mg/mL PFSE, 0.0023-mg/mL PFSE, and 10-μM piceatannol showed an increased cellular viability compared with that of the control (solvent alone) group in ARPE-19 cells (Fig. 1B). In particular, PFSE administration at 0.0023 mg/mL in ARPE-19 cells significantly increased cellular viability to 143% compared with that of the control stage (Fig. 1B). Additionally, 0.023-mg/mL PFSE and 10-μM piceatannol increased cellular viability to 130% and 133%, respectively, compared with that of the control stage (Fig. 1B). We found that both PFSE and piceatannol enhanced cellular viability in ARPE-19 cells.

Fig. 1.

Physiological effect of PFSE and piceatannol on cell viability in human RPE cells ARPE-19. (A) Experimental timeline. Cells were treated with the indicated PFSE or Pic concentrations for 24 h. After incubation for 1 h, cell viability was assessed by WST-8 assay. (B) Cell viabilities (%) are the means of three independent experiments, with error bars showing ± standard error of the mean (SEM). Cont., control (solvent alone); PFSE10, 0.023-mg/mL PFSE (Pic equivalent of 10 μM); PFSE1, 0.0023-mg/mL PFSE (Pic equivalent of 1 μM); Pic10, 10-μM Pic; Pic1, 1-μM Pic.

Second, we examined the relationship between energy metabolism and cellular activation by using an extracellular flux analyzer XFe24 to analyze the ECAR and OCR of ARPE-19 cells with PFSE or piceatannol. The ECAR and OCR are important indicators of glycolysis and mitochondrial respiration³⁴. Figure 2 shows the timeline of the experiment. PFSE or piceatannol treatment of ARPE-19 cells resulted in higher baseline OCR and ECAR than those in the control stages (Fig. 2B, C). Cells treated with PFSE or piceatannol had an increased mitochondrial ATP production rate compared to control cells (Table 1, Fig. 2D, E). On the other hand, the glycolysis ATP production rate was comparable between control cells and cells treated with PFSE or Piceatannol (Table 1, Fig. 2D, E). In contrast, intracellular and extracellular ATP levels were similar in all experimental groups (Table 1, Fig. 2F, G. Thus, PFSE and piceatannol enhance the cellular activity of ARPE-19 cells and increase the rate of mitochondrial ATP production in the cellular energy metabolism. This finding indicates that PFSE and piceatannol may be able to prevent H₂O₂-induced RPE cell damage in the AMD mimic cell model.

Fig. 2.

Effect of PFSE and piceatannol on energy metabolism in human RPE cells ARPE-19. (A) Experimental timeline. Cells were seeded in an Agilent Seahorse XFe24 cell culture microplate. After incubation for 30 min, cells were treated with 0.23 mg/ml PFSE (i.e., piceatannol equivalent 10 μM) or 10 μM piceatannol for 24 h. Then, the culture medium was replaced with fresh medium. After incubation for 1 h, the culture medium was replaced with the analysis medium, and the OCR and ECAR were analyzed using an Agilent Seahorse XFe24 analyzer. (B) OCR. (C) ECAR. The OCR (pmol/min) and ECAR (mpH/min) represent the average of three independent experiments, with error bars showing ± SE. Black circle, control; black square, PFSE; black triangle, Pic. Oligomycin (final concentration: 0.5 μM); Rotenone and antimycin A (final concentration: 0.5 μM each). (D) Rate of ATP production (pmol/min) represents the average of three independent experiments, with error bars showing ± SE. Gray bar, glycoATP. White bar, mitoATP. One-way ANOVA followed by Dunnett’s multiple comparisons test. ns, not significance; ****, p < 0.0001, compared to the control group. (E) The ratio of mitoATP to glycoATP represents the average of three independent experiments. mitoATP, ATP production from mitochondria; glycoATP, ATP production from glycolysis. (F) Levels of intracellular ATP. (G) Levels of extracellular ATP. The iATP and eATP (pmol/cell) represents the average of three independent experiments, with error bars showing ± SE. One-way ANOVA followed by Dunnett’s multiple comparisons test. iATP, intracellular ATP. eATP, extracellular ATP. ns, not significance, compared to the control group.

Table 1.

Summary of PFSE and piceatannol on energy metabolism dependency, glycolysis, and mitochondrial respiration in ARPE-19 cells.

	Control	PFSE10	Pic10
Total ATP production rate (pmol/min)	1267.6	1439.8	1398.3
ATP production rate (pmol/min) mitoATP rate/glycoATP rate	941.3/326.3	1096.0/343.8	1021.3/377.0
Intracellular ATP level (pmol/cell)	16.5	16.0	15.0

PFSE and piceatannol suppress H₂O₂-induced RPE cell damage

Next, we investigated the protective effects of PFSE and piceatannol on cellular damage in an AMD cell model. The AMD cell model was prepared as shown in Fig. 3A. We treated the cells with H₂O₂ for 1 h and replaced the culture medium with fresh medium for another hour. The cell viability was reduced to 60% in this AMD cell model (Fig. 3B). Cell viability in the AMD cell model increased with PFSE and piceatannol pretreatment (Fig. 3B). Cell viability in the cell model was increased to approximately 84% and 89% after the pretreatment of PFSE at 0.023 mg/mL and 0.0023 mg/mL, respectively (Fig. 3B). Similarly, cell viability in the cell model increased to approximately 94% and 80% after the pretreatment of piceatannol at 10 μM and 1 μM, respectively (Fig. 3B). Interestingly, the AMD model cells demonstrated reduced cell damage by PFSE and piceatannol pretreatment (Fig. 3B). Furthermore, morphological observations revealed that PFSE and piceatannol treatments at 24 h do not negatively affect ARPE-19 cells (Fig. 3C). Interestingly, we also observed that PFSE and piceatannol reduces H₂O₂-induced ARPE-19 cell damage at similar levels (Fig. 3C). These results indicated that both PFSE and piceatannol can suppress H₂O₂-induced RPE cell damage in the AMD cell model.

Fig. 3.

Protective effect of PFSE or piceatannol on H₂O₂-induced human RPE cell damage. (A) Experimental timeline. Human RPE cell line ARPE-19 cells were seeded in a 96-well plate and incubated for 30 min. Cells were treated with PFSE or Pic for 24 h at indicated concentrations. After PFSE and Pic administration, the culture media was transferred to fresh media. Additionally, cells were treated with 400-μM H₂O₂ for 1 h. The culture medium was then replaced with fresh media. After incubation for 1 h, cell viability was detected by WST-8 assay. PFSE, passion fruit seed extract; Pic, piceatannol. (C) The protective effect of PFSE and Pic on H₂O₂-induced RPE cell damage. Cell viabilities (%) are averages of three independent experiments, with error bars showing ± SEM. Cont., control; HP, 400-μM H₂O₂. One-way ANOVA followed by Dunnett’s multiple comparisons test was conducted for testing significant differences. ns, not significant; *, p < 0.05, compared with the control group. (D) Morphological features were analyzed using a Leica DMi1 microscope with LAS V4.12 at 100 × magnification. The scale bar indicates 200 μm.

PFSE and piceatannol restore the reduction of ATP production rate in H₂O₂-damaged RPE cells

We analyzed the ECAR and OCR with and without PFSE or piceatannol in H₂O₂-induced AMD cell models using an extracellular flux analyzer to elucidate the association of energy metabolism and cytoprotective effects of PFSE or piceatannol on H₂O₂-induced RPE cell damage. Figure 4A shows the experimental timeline for metabolic analysis. The baseline OCR and ECAR from 0 to 40 min in the AMD cell model are significantly lower than those in the control stage, as shown in Fig. 4B, C. Conversely, PFSE or piceatannol pretreatment of ARPE-19 cells resulted in higher baseline OCR and ECAR than those in the control stages (Fig. 4B, C). The mitochondrial ATP production rate in the AMD cell model was significantly reduced to approximately 34% compared with that of the control stage (Table 2 and Fig. 4D). In contrast, the ATP production rate of glycolysis in the AMD cell model was 84% slightly lower than that in the control stage (Table 2 and Fig. 4D). Interestingly, PFSE or piceatannol pretreatment of ARPE-19 cells significantly reduced H₂O₂-induced declines in mitochondrial and glycolysis ATP production rates compared with AMD model cells (Table 2, Fig. 4D, E). Notably, pretreatment with PFSE and piceatannol restored the mitochondrial ATP production rate, which had reduced to 36% owing to oxidative stress, to approximately 71% and 63%, respectively. Moreover, pretreatment with PFSE and piceatannol restored and upregulated the glycolysis ATP production rate, which had decreased to 87% owing to oxidative stress, to approximately 145% and 117%, respectively. Furthermore, we investigated intracellular and extracellular ATP levels in the H₂O₂-induced AMD cell model with PFSE or piceatannol pretreatment. Intercellular ATP level in the AMD cell model was considerably lower than that in the control stage (Table 2 and Fig. 4F). In addition, intracellular ATP levels were lower in ARPE-19 cells pretreated with PFSE and piceatannol followed by H₂O₂ than in ARPE-19 cells treated with H₂O₂ alone (Table 2 and Fig. 4F). Therefore, PFSE or piceatannol pretreatment of ARPE-19 cells potentiated ATP level reduction due to H₂O₂ treatment. By contrast, extracellular ATP levels were similar in all experimental groups (Fig. 4G). These findings indicate that PFSE or piceatannol increases ATP consumption after H₂O₂-induced cell damage in ARPE-19 cells.

Fig. 4.

Cellular respiration property and protective effect of PFSE and piceatannol in H₂O₂-induced human RPE cell damage. (A) Experimental timeline. Cells were seeded in an Agilent Seahorse XFe24 cell culture microplate. After incubation for 30 min, cells were treated with 0.023-mg/mL PFSE (i.e., Pic equivalent of 10 μM) or 10-μM Pic for 24 h. The culture medium was then replaced with a fresh medium. Additionally, cells were treated with 400-μM H₂O₂ for 1 h. After incubation for 1 h, the culture medium was replaced with the analysis medium, and the OCR and ECAR were analyzed using an Agilent Seahorse XFe24 analyzer. (B) OCR. (C) ECAR. The OCR (pmol/min) and ECAR (mpH/min) represent the average of three independent experiments, with error bars showing ± SEM. Open circle, control; solid circle, HP alone; solid square, PFSE plus HP; solid triangle, Pic plus HP. Oligomycin (final concentration: 0.5 μM); Rotenone and antimycin A (final concentration: 0.5 μM each). OCR, oxygen consumption rate; ECAR, extracellular acidification rate. (D). The rate of ATP production (pmol/min) represents the average of three independent experiments, with error bars showing ± SEM. Gray bar, glycoATP. White bar, mitoATP. (E) The ratio of mitoATP to glycoATP represents the average of three independent experiments. mitoATP, ATP production from mitochondria; glycoATP, ATP production from glycolysis. (F) Levels of intracellular ATP. The iATP (pmol/cell) represents the average of three independent experiments, with error bars showing ± SEM. (G) Levels of extracellular ATP. The eATP (pmol/cell) represents the average of three independent experiments, with error bars showing ± SEM. One-way ANOVA followed by Dunnett’s multiple comparisons test was conducted for testing significant differences. ns, not significant; *, p < 0.05; **, p < 0.01; ****, p < 0.0001, compared with the control group. iATP, intracellular ATP; eATP, extracellular ATP.

Table 2.

Summary of PFSE and piceatannol on energy metabolism dependency, glycolysis, and mitochondrial respiration in H₂O₂-damaged ARPE-19 cells.

	Control	HP	PFSE10 + HP	Pic10 + HP
Total ATP production rate (pmol/min)	941.2	504.0	908.8	770.8
ATP production rate (pmol/min) mitoATP rate/glycoATP rate	614.9/326.3	218.9/285.1	434.1/474.7	388.4/382.4
Intracellular ATP level (pmol/cell)	15.5	9.3	6.3	4.0

Note. HP, Hydrogen peroxide (H₂O₂).

Discussion

AMD is associated with several environmental and genetic risk factors that are connected to increased oxidative stress³⁵. Among them, age is the major risk factor³⁵. Oxidized nucleic acid, protein, and lipid accumulation in aged cells and tissues are the evidence for an overall increase in oxidative stress with aging³⁵. Damage to the RPE due to oxidative stress is considered to be an important factor in AMD pathology³⁵. RPE cells are constantly exposed to oxidative stress, which may result in the accumulation of damaged nucleic acids, proteins, lipids, and organelles, including mitochondria³⁵. Oxidative stress ultimately involves excessive reactive oxygen species (ROS) production mainly in mitochondria³⁵. ROS then cause mitochondrial damage and cell death. Mitochondrial dysfunction generally occurs in patients with AMD³⁵. Previously, two different research groups have revealed reduced mitochondrial function and ATP production in RPE cells from patients with AMD compared with those from healthy controls^35–37. These results indicate the involvement of mitochondrial damage and dysfunction in AMD pathology.

This study provides evidence that dependent changes in energy metabolism, i.e., enhanced glycolysis, mitochondrial citric acid cycle, and oxidative phosphorylation, may explain the cytoprotective effect of PFSE and piceatannol against H₂O₂-induced cell damage in ARPE-19 cells. Numerous studies have revealed piceatannol as a potent antioxidant⁸. Additionally, several previous studies have reported the cytoprotective effect of piceatannol against H₂O₂-, A2E-, and blue light–induced cell damage in ARPE-19 cells^23,24. Our findings indicate that PFSE and piceatannol not only act as antioxidants but also increase the metabolic dependence of mitochondrial oxidative phosphorylation in cellular energy metabolism, i.e., enhance mitochondrial function. Notably, PFSE and piceatannol did not recover H₂O₂-induced ATP reduction. This phenomenon raised the possibility that intercellular ATP is consumed to repair H₂O₂-induced cell damage. Future studies need to investigate whether PFSE and piceatannol directly enhance glycolysis, mitochondrial biogenesis, mitochondrial activity, or functional regulation by removing mitochondrial oxidative stresses.

Overall, our findings indicate that PFSE and piceatannol exhibit cytoprotective effects by maintaining glycolysis, mitochondrial function and suppressing H₂O₂-induced mitochondrial dysfunction in ARPE-19 cells. This effect may be exerted by PFSE and piceatannol by modulating cellular dependence on glycolysis and mitochondrial respiration, i.e., tricarboxylic acid cycle and oxidative phosphorylation, in energy metabolism (Fig. 5).

Fig. 5.

Overview of the cytoprotective effects of piceatannol-rich PFSE in H₂O₂-induced AMD model ARPE-19 cells.

In conclusion, piceatannol-rich PFSE can be a promising functional food in maintaining glycolysis and mitochondrial function in RPE cells as well as in preventing and treating AMD pathology.

Methods

Reagents

We prepared piceatannol-rich PFSE as previously described³³. Piceatannol was obtained from Tokyo Chemical Industry (Tokyo, Japan). We stored PFSE as 100-mg/mL stocks in dimethyl sulfoxide (DMSO, Sigma-Aldrich; Merck KGaA, Darmstadt, Germany) at − 20 °C. Piceatannol was stored as 100-mM stocks in DMSO at − 20 °C. H₂O₂ was purchased from FUJIFILM Wako Pure Chemical (Osaka, Japan).

Cell culture

The human retinal pigment epithelium cell line ARPE-19 was obtained from the American Type Culture Collection. We cultured ARPE-19 cells in DMEM/Ham’s F-12 medium (FUJIFILM Wako Pure Chemical) containing 10% heat-inactivated fetal bovine serum, 100 units/mL of penicillin, and 100 μg/mL of streptomycin in an incubator at 37 °C in an atmosphere with 5% CO₂ and 100% relative humidity.

Cell viability by WST-8 assay

Cell viability assays were performed as previously described³⁸. We determined cell viability using the WST-8 (Cell Counting Kit-8) cell proliferation assay (Dojindo, Tokyo, Japan). Briefly, we seeded the cells into 96-well plates (4,000 cells per well) in triplicate and incubated for 30 min; further, we treated them with PFSE or piceatannol and solvent (DMSO, as a negative control; final concentration of 0.2%) for 24 h. We administered PFSE at 0.023 mg/mL (piceatannol equivalent of 10 μM) or 0.0023 mg/mL (piceatannol equivalent of 1 μM) and piceatannol at 10 μM or 1 μM concentrations. Subsequently, the culture medium was replaced with a fresh medium. Additionally, we treated the cells with 400 μM of H₂O₂ and water (negative control). We transferred the culture media to fresh media 1 h after the H₂O₂ treatment. After an incubation for 1 h, we added WST-8 reagent to each well and placed the plate in a 5% CO₂ incubator at 37 °C for an additional 1 h. We measured the absorbance at 450 nm on a Tecan microplate reader (Mannedorf, Switzerland).

Cellular respiration analysis

We performed cellular respiration analysis as previously described³⁸. The cells were dissociated using Accutase® and suspended in DMEM/Ham’s F-12 medium. The cells were then seeded in an Agilent Seahorse XF24 cell culture microplate (Agilent Technologies, Santa Clara, CA) (4 × 10⁴ cells per well). After incubation for 30 min, cells were treated with 0.023-mg/mL PFSE (i.e., piceatannol equivalent 10 μM) or 10-μM piceatannol for 24 h. Further, the culture medium was replaced with fresh DMEM/Ham’s F-12 medium. After incubation for 1 h, the culture medium was replaced with XF DMEM medium (Cat#103,575–100, Agilent Technologies), and the oxygen consumption rate (OCR) and extracellular acidification rate (ECAR) were analyzed using an Agilent Seahorse XFe24 analyzer (Agilent Technologies). The analysis medium contained 2-mM Seahorse XF glutamine solution (Agilent Technologies), 1-mM Seahorse XF pyruvate solution (Agilent Technologies), and 25-mM glucose. Additionally, we used a Seahorse XF Real-Time ATP Rate Assay Kit (Cat# 103,592–100, Agilent Technologies) to analyze the ATP production rate, following the manufacturer’s protocol.

Measurement of intracellular/extracellular ATP levels

We measured the intracellular and extracellular ATP levels using the CellTiter-Glo® 2.0 Cell Viability Assay (Promega, WI). Briefly, the cells were seeded into 96-well plates (2,500 cells per well) in triplicate and incubated for 30 min. They were treated with 0.023-mg/mL PFSE (piceatannol equivalent of 10 μM) or 10-μM piceatannol and solvent (DMSO, as a negative control; final concentration of 0.2%) for 23 h. After treatment with supplements, we replaced the culture medium with a fresh medium. In addition, the cells were treated with 400-μM H₂O₂ and water (negative control) for incubation at 1 h. Luminescence was measured using a Tecan microplate luminometer (Mannedorf, Switzerland).

Statistical analysis

We used GraphPad Prism 9 software for statistical analyses. We determined the significance of differences among groups using Student’s t-test and one-way analysis of variance (ANOVA) followed by Dunnett’s multiple comparison test. P-values < 0.05 were considered significant.

Author contributions

A.S. conceived and designed the research. H.U. and A.S. performed the research and acquired the data. H.U., S.K., Y.M., S.M., and A.S. analyzed and interpreted the data. A.S. wrote the main manuscript text. H.U. and A.S. prepared Figs. 1–5. All authors reviewed the manuscript.

Data availability

All data generated or analysed during this study are included in this published article.

Declarations

Competing interests

Akira Sato received a research grant from Morinaga ＆ Co., Ltd. All other authors declare that they do not have any competing interest.