A total of 66 6-week healthy female SKH-1 hairless mice were obtained from Orient Bio, Inc. (Seongnam, Korea). The experimental groups were divided into the following 6 groups (8 mice per group), based on the body weights 7 days after accclimatization (23.32±0.93 g, range, 21.4–24.8 g/head): Intact vehicle control-unexposed (mice received vehicle-control); UVB control-UVB exposed (mice received vehicle control and UVB exposure); PCS-UVB exposed (mice received PCS 2 ml/kg and UVB exposure); PCP 100 mg/kg-UVB exposed (mice received PCP 100 mg/kg and UVB exposure); PCP 200 mg/kg-UVB exposed (mice received PCP 200 mg/kg and UVB exposure); and PCP 400 mg/kg-UVB exposed (mice received PCP 400 mg/kg and UVB exposure). During acclimatization and experimental periods, animals were kept at a temperature of 20–25°C, a humidity of 50–55% and were exposed to a 12-h light/dark cycle in polycarbonate cages (4 rats per cage). Mice had ad libitum access to standard rodent chow (Samyang Corporation, Seoul, Korea) and tap water. All procedures involving laboratory animals followed the national regulations for the usage and welfare of laboratory animals (26). The protocol of the current study was approved by the Institutional Animal Care and Use Committee of Daegu Haany University (Gyeongsan, Korea) prior to the experiment (approval no. DHU2015-066).

UVB irradiation was performed three times a week for 15 weeks at 0.18 J/cm² following the protocol of a previous study (27). Peak emission was measured at 312 nm using a UV Crosslinker system (Hoefer Scientific Instruments, San Francisco, CA, USA). Unexposed intact control mice were also exposed to the non-emitting Crosslinker system for the same duration as the UVB-exposed hairless mice.

PCS was purchased from ASYA Meyve Suyu ve Gida San. A.Ş. (Ankara, Turkey). PCS contained 2.31 mg/g ellagic acid as an active ingredient and consisted of 58.86% carbohydrate, 1.21% total protein, 0.49% fat, 27.97% water, 1.47% ash and 10% unidentified ingredients, as well as 28.03 mg/100 g sodium at proximate analysis. Health-Love Co. (Anyang, Korea) purchased raw materials of PCS and PCP from ASYA Meyve Suyu ve Gida San. A.Ş. and supplied the processed products after ingredients analysis and quality control. PCP (ASYA Meyve Suyu ve Gida San. A.Ş.) containing 1.15 mg/g ellagic acid and 0.22 mg/g punicalagin as active ingredients. All test materials were stored at 4°C in a refrigerator to protect from light and moisture. A PCP dose of 200 mg/kg was selected based on the in vitro skin whitening effect (25) and the clinical dose in humans [mouse dosage was 200 mg/kg, ~12-fold the human dosage; (1,000 mg/60 kg) ×12=200 mg/kg] and doses of 400 and 100 mg/kg were selected as the highest and lowest doses, respectively, using a ratio of 2. Similarly, a PCS dose of 2 ml/kg was selected based on a previous efficacy study investigating UVB-induced photoaging (23,24) and also based on the clinical dose in humans [mouse dosage was ~12-fold the human dosage; (10 ml/60 kg) ×12=2 ml/kg]. PCP was dissolved in distilled water to obtain concentrations of 40, 20 and 10 mg/ml, and a volume of 10 ml/kg was orally administered using a gastric gavage attached syringe, resulting in administration of 400, 200 and 100 mg/kg, respectively. PCS was diluted in distilled water in 1:4 ratios (v/v) and a volume of 10 ml/kg was orally administered, resulting in administration of 2 ml/kg of body weight, respectively. The different doses of PCP 100, 200 and 400 mg/kg, and PCS 2 ml/kg were administered to 8 mice in each PCP group once per day for 15 weeks beginning 1 h after initial UVB exposure. Equal volumes of vehicle (distilled water) were orally administered to all mice in the intact and UVB control groups.

After obtaining photos of the dorsal skin around the gluteal region, animals were sacrificed by over exdanguination through vena cava under anesthesia (3% isoflurane; Hana Pharm Co., Ltd., Hwasung, Korea) and dissected. Replicas of mouse dorsal skin were obtained using the Repliflo Cartridge kit (CuDerm Corp., Dallas, TX, USA). Impression replicas were set on a horizontal sample stand and wrinkle shadows were produced by illumination with a fixed-intensity light at a 40° angle. Black and white images were recorded using a CCD camera and analyzed using the Skin-Visiometer® VL650 (Courage and Khazaka, Cologne, Germany). To analyze skin wrinkles, the average length and average depth of wrinkles were measured using a previously described method (10).

To assess the effects of PCP and PCS on UVB-induced skin edema, changes in dorsal skin weight were examined using a previously established method (4). Dorsal skin was removed, a 6-mm diameter was delimited with the aid of a punch and this area was then weighed. Changes were measured following 15 weeks of continuous oral administration of PCP, PCS or distilled water. The results were calculated by comparing the weight of the skin between groups and expressed as g/6-mm diameter of dorsal skin.

The day after the final administration of test substances, 6 mm-diameter of dorsal back skin samples were removed and skin water contents (%) were measured using an automated moisture analyzer balance (MB23; Ohaus Corporation, Parsippany, NJ, USA), according to a previous study (28).

COL1 contents were measured following a previously established method (29). Briefly, parts of the dorsal back skin tissues were separated and tissue homogenates were prepared using a homogenization system (Model TacoTMPre; GeneResearch Biotechnology Corp., Taichung, Taiwan), an ultrasonic cell disruptor (Model KS-750; Madell Technology Corp., Ontario, CA, USA) and radioimmunoprecipitation assay buffer (Sigma-Aldrich; Merck KGaA, Darmstadt, Germany). Following the seperation of supernatants by centrifugation at 21,000 × g for 10 min at 4°C, the amount of pro-collagen type I was measured using a pro-collagen type I C peptide assay kit (Takara Bio, Inc., Otsu, Japan) according to the manufacturer's protocol and absorbance was measured at 450 nm using a microplate reader (Tecan Group Ltd., Männedorf, Switzerland).

Hyaluronan contents in the supernatant prepared from skin tissue homogenates were measured using a previously described method (30). Briefly, the fat in the sampled dorsal back skins was removed using acetone. Samples were then dried, weighed and boiled for 20 min in 50 mM Tris/HCl (pH 7.8) buffer. Subsequently, samples underwent proteolytic digestion with 1% w/v actinase E (Sigma-Aldrich; Merck KGaA) for 1 week at 40°C. Trichloroacetic acid (Sigma-Aldrich; Merck KGaA) was added to the samples at a final concentration of 10% w/v to induce deproteinization prior to centrifugation at 4,000 × g at 4°C for 20 min. The supernatants were subsequently neutralized using 10 N NaOH. Hyaluronan levels were measured using the QnE hyaluronic acid enzyme-linked immunosorbent assay (ELISA) kit (cat no. BTP-96200; Biotech Trading Partners, Encinitas, CA, USA), according to the manufacturer's protocol. Color intensity was measured at 450 nm using a spectrophotometer (OPTIZEN POP, Mecasys, Daejeon, Korea).

MPO activity was measured using the MPO kinetic-colorimetric assay, according to a previously described method (4). A total of 50 mM K₂HPO₄ buffer (pH 6.0; Sigma-Aldrich; Merck KGaA) containing 0.5% hexadecyltrimethylammonium bromide (Gibco; Thermo Fisher Scientific, Inc., Waltham, MA, USA) was used to collect and homogenize skin samples in an ice bath for 15 sec. Following centrifugation at 1,000 × g for 2 min at 4°C, the supernatant was removed. A total of 30 µl supernatant sample was mixed with 200 µl 0.05 M K₂HPO₄ buffer (pH 6.0), containing 0.167 mg/ml o-dianisidine dihydrochloride (Sigma-Aldrich; Merck KGaA) and 0.05% hydrogen peroxide. After 5 min, the absorbance of the samples was measured using a spectrophotometer at 450 nm. The MPO activity of samples was compared with a standard curve of neutrophils and protein levels in the skin homogenates was measured using the Lowry method (31). The results are presented as MPO activity (number of total neutrophils/mg protein).

The dorsal back skin tissue from the area around gluteal region was collected. Collected tissue was homogenized and processed as described by Botelho et al (32). IL-10 and IL-1β concentrations were determined using the enzyme-linked immunosorbent assay (ELISA) kit (cat no. ab108870 for IL-10; cat no. ab100705 for IL-1β; Abcam, Cambridge, UK), according to the manufacturer's instructions. Absorbance was measured at 490 nm using a microplate reader.

Following skin homogenization (1:3, w/w dilution) in 100 mM NaH₂PO₄ (pH 8.0; Sigma-Aldrich; Merck KGaA) containing 5 mM EDTA, 30% trichloroacetic acid (Sigma-Aldrich; Merck KGaA) was added to homogenates. Mixtures were then centrifuged twice (once at 1,940 × g for 6 min at 4°C and once at 485 × g for 10 min at 4°C) and the fluorescence of the resulting supernatant was measured using a fluorescence spectrophotometer (RF-5301PC; Shimadzu Corporation, Tokyo, Japan). Supernatant (100 µl) was mixed with 1 ml buffer 1 and 100 µl o-phthalaldehyde (1 mg/ml in methanol; Sigma-Aldrich; Merck KGaA). After 15 min, the absorbance was measured at 420 nm. A standard curve containing 0.0–75.0 µM GSH was prepared. Protein levels in the skin homogenates were measured using the Lowry method (31) and results are presented as µM GSH/mg protein. GSH levels were measured as reported previously (4).

Protein content of the homogenate (10 mg/ml in 1.15% KCl) was measured using the Lowry method (31). To determine lipid peroxidation, levels of thiobarbituric acid reactive substances were measured following a previously described method (33). Briefly, 10% trichloroacetic acid (Sigma-Aldrich; Merck KgaA) was mixed with the homogenate and this mixture was then centrifuged at 1,000 × g at 4°C for 3 min to obtain protein-free samples. Thiobarbituric acid (0.67%) was treated and the mixture was left to stand at 100°C for 15 min. Levels of malondialdehyde (MDA), an intermediate product of lipoperoxidation, were measured by determining the difference between the absorbance at 535 and 572 nm using a microplate spectrophotometer reader. The results were reported as nM/mg of protein (34).

A nitroblue tetrazolium (NBT) assay was used to determine the production of superoxide anion in dorsal back skin tissue homogenates (10 mg/ml in 1.15% KCl), following a previously described protocol (35). Briefly, NBT (1 mg/ml; Sigma-Aldrich; Merck KGaA) was added to 50 µl homogenate at 37°C for 1 h. The supernatant was then removed and reduced formazan was solubilized following addition of 120 µl 2 M KOH and 140 µl dimethyl sulfoxide. NBT reduction was measured at 600 nm using a microplate spectrophotometer reader and data were normalized to the protein content.

Total RNA in the individual dorsal back skins, sampled by skin puncher 24 h following the final administration of PCP or PCS, was extracted using TRIzol reagent (Invitrogen; Thermo Fisher Scientific, Inc.) according to a previously described method (30). RNA concentration and quality was determined using a CFX96™ Real-Time system using iTaq™ SYBR-Green (both from Bio-Rad Laboratories, Inc., Hercules, CA, USA). To remove contaminating DNA, samples were treated with recombinant DNase I (DNA-free DNA Removal kit; Ambion; Thermo Fisher Scientific, Inc.). RNA was reverse-transcribed using the High-Capacity cDNA Reverse Transcription kit (Applied Biosystems; Thermo Fisher Scientific, Inc.) following the manufacturer's instructions. The PCR cycling conditions were as follows: Initial pre-denaturation of 95°C for 1 min, denaturation for 15 sec, annealing of 55–65°C for 20 sec and extension of 72°C for 30 sec. A total of 50 cycles were performed. The β-actin mRNA level was used as a control for tissue integrity in all samples. Primer sequences for Has 1, 2 and 3, COL1A1 and 2, MMP-1, 9 and 13, Nox2, GSH reductase and β-actin are presented in Table I. For quantitative analysis, the intact control skin tissue was used as the control, and the relative expression of Has 1, 2 and 3, COL1A1 and 2, MMP-1, 9 and 13, Nox2, GSH reductase was calculated using the 2^−ΔΔCq method (36).

Samples from dorsal back skins were separated, fixed in 10% neutral buffered formalin at room temperature for 24 h, embedded in paraffin wax, cut into sections 3–4 µm thick and stained with hematoxylin and eosin (H&E) for general histopathology or Masson's trichrome (MT) to detect collagen fiber, according to previously established methods (37). The histopathological profiles of each sample were observed under a light microscope (Model 80i; Nikon Corporation, Tokyo, Japan). To quantify changes in further detail, the number of microfolds formed on the surface of epithelium (folds/mm epithelium), mean epithelial thickness (µm/epithelium) and mean numbers of inflammatory cells infiltrated in the dermis (cells/mm² of dermis) were determined in general histomorphometrical analysis using image analysis software (iSolution FL ver 9.1; IMT i-solution Inc., Vancouver, Canada) under H&E staining and in collagen fiber-occupied regions of the dermis (%/mm² of dermis) under MT staining. The histopathologist was blinded to the group distribution when conducting the analysis.

Following dewaxing of the prepared skin histological paraffin wax sections, citrate buffer antigen (epitope) retrieval pretreatment was conducted (37). Briefly, a water bath was pre-heated to 95–100°C with a staining dish containing 10 mM citrate buffer (pH 6.0). Slides were immersed in the staining dish for 20 min and the lid was placed loosely. The staining dish was then kept at room temperature for 20 min to cool. Sections were immunostained using avidin-biotin complex (ABC) methods for caspase-3, cleaved poly(ADP-ribose) polymerase (PARP), nitrotyrosine (NT), 4-hydroxynonenal (4-HNE), inducible nitrogen oxide synthase (iNOS) and matrix metalloproteinase (MMP)-9 according to the results of a previous study (37). Briefly, endogenous peroxidase activity was blocked by incubation in methanol and 0.3% H₂O₂ at room temperature for 30 min and non-specific binding was blocked with 1% normal horse serum blocking solution (dilution, 1:100; Vector Laboratories, Peterborough, UK) for 1 h in a humidity chamber. Sections were incubated with primary antibodies (listed in Table II) overnight at 4°C in a humidity chamber and then incubated with biotinylated universal secondary antibody (dilution, 1:50; Vector Laboratories) using a Vectastain Elite ABC kit (dilution, 1:50; Vector Laboratories; Table II) for 1 h at room temperature in a humidity chamber acooriding to the manufacturer's instructions. Finally, sections were incubated with a peroxidase substrate kit (Vector Laboratories; Table II) for 3 min at room temperature according to the manufacturer's instructions. Between steps, all sections were rinsed in 0.01 M PBS. Cells or fibers comprising >30% immunoreactivity for each antiserum compared with intact dermal keratinocytes or dermal tissue, were regarded as positive. The mean numbers of caspase-3, PARP, NT and 4-HNE-immunoreactive epithelial cells (%, cells/100 epithelial cells) were counted using an established automated image analysis process (37). Additionally, the percentage of the dermis occupied by MMP-9 immunoreactive fibers was calculated (%/mm² dermis). The histopathologist was blinded to the group distribution when performing the analysis.

All data are presented as the mean ± standard deviation of eight hairless mice. Multiple comparison tests of the different groups were conducted. Variance homogeneity was examined using the Levene test and if no significant deviation from variance homogeneity was detected, the data were analyzed by a one-way analysis of variance followed by a Least Significant Difference multi-comparison test to determine whether differences between pairs of groups were significant. If significant deviations from variance homogeneity were detected by the Levene test, the non-parametric Kruskal-Wallis H comparison test was conducted. When a significant difference was observed in the Kruskal-Wallis H test, the Mann-Whitney U test with Bonferroni correction was used to determine whether there were significant differences between specific pairs of groups. Statistical analyses were performed using SPSS software ver. 14.0 for Windows (SPSS, Inc., Chicago, IL, USA) and P<0.05 was considered to indicate a significant difference.

To confirm the wrinkle alleviation effect of PCP, the mean length and depth of the wrinkles were measured (Fig. 1). It was determined that the mean length and depth of the wrinkles were significantly greater in the skin of the UVB-exposed control mice compared with the intact control (both P<0.01; Fig. 2). However, these increases in wrinkle formations significantly decreased following oral administration of PCP compared with the UVB control (all P<0.05). These decreases occurred in a dose-dependent manner (Fig. 2).

The effect of PCP against UVB-induced skin edema was analyzed by measuring 6-mm-diameter skin weights (edema score). The 6-mm-diameter skin weights (edema score) in mice exposed to UVB were significantly greater than those of intact controls (P<0.01). However, in mice treated with 400, 200 and 100 mg/kg PCP, mean skin weight significantly decreased (P<0.01) compared with those of the UVB-exposed mice, by −55.33, −45.68 and −37.61%, respectively (Fig. 3).

To clarify the moisturizing effects of PCP, the skin water content in 6-mm-diameter tissues was observed. Skin-water content was significantly decreased following UVB treatment (P<0.01). However, UVB-induced decreases in skin water contents were significantly reversed (P<0.01) following treatment with 100, 200 and 400 mg/kg PCP, in a dose-dependent manner. Additionally, mice administered 2 ml/kg PCS exhibited significant increases in skin water content compared with UVB control mice (P<0.01; Table III).

To confirm the moisturizing effects of PCP, changes in COL1 and hyaluronan contents were assessed. COL1 and hyaluronan contents were significantly lower in 6-mm-diameter skin tissues from UVB control mice compared with intact controls (P<0.01). However, these decreases were significantly inhibited following oral administration of all three doses of PCP (P<0.01; Table III).

To clarify the anti-inflammatory effects of PCP, MPO activity, as well as IL-1β and IL-10 levels, were assessed. Skin MPO activity and IL-1β levels were significantly higher in UVB-exposed control mice than intact control mice (P<0.01). MPO activity and IL-1β levels were significantly inhibited following oral administration of PCS (P<0.01) and all doses of PCP (P<0.01) in a dose-dependent manner, compared with the UVB control mice (Table IV). IL-10 levels in UVB-exposed control mice were significantly decreased compared with those in unexposed intact control mice (P<0.01). However, significant (P<0.05) and dose-dependent increases in skin IL-10 levels were observed following administration of all three doses of PCP compared with untreated UVB-exposed mice (Table IV).

To determine the anti-oxidative effects of PCP, GSH contents and MDA levels were analyzed. GSH contents were significantly lower in the UVB-exposed mice compared with the intact control (P<0.01; Table V). Furthermore, significant increases in skin MDA levels, indicating increased lipid peroxidation and superoxide anion production, were detected in UVB-exposed mice compared with unexposed control mice (P<0.01; Table V).

Skin GSH contents were significantly increased in all PCP-administered hairless mice compared with UVB control mice (P<0.01; Table V). This increase occurred in a dose-dependent manner. Furthermore, skin MDA levels and superoxide anion production were significantly decreased in PCP-treated mice compared with UVB control mice (P<0.01; Table V).

Although no significant changes in skin Has 1, 2, and 3 or in COL1A1 and 2 mRNA levels (relative to the control) were identified in UVB-exposed control mice compared with unexposed intact control mice, significant increases (P<0.01) in skin Has 1, 2, and 3 and COL1A1 and 2 mRNA levels were detected in mice receiving the three doses of PCP, compared with UVB control mice, suggesting that PCP increases hyaluronan and COL1 synthesis (Table VI).

Significant increases in skin MMP-1, −9 and −13 and Nox2 mRNA levels were detected in UVB-exposed control mice compared with unexposed intact mice (P<0.01), indicating increased MMP activity and ROS-dependent inflammation in UVB-exposed control mice. However, levels of skin MMP and Nox2 mRNA were significantly inhibited following administration of all three doses of PCP compared with UVB control mice (P<0.01; Table VI).

Levels of skin GSH reductase mRNA were significantly decreased in UVB-exposed mice compared with intact control mice (P<0.01). However, significant dose-dependent increases in skin GSH reductase mRNA expression were detected in mice receiving 100, 200 and 400 mg/kg PCP compared with UVB control mice (P<0.01; Table VI).

When the mean epithelial thicknesses of UVB-exposed mice and intact control mice were compared, it was determined that UVB-exposed control mice exhibited higher mean epithelial thicknesses of dorsal back skin tissues compared with intact control mice. In addition, there was a cellular infiltration of inflammatory cells into the epidermis, an abnormal accumulation of collagen in the dermis and the formation of microfolds on the surface of the epithelial lining was observed (Fig. 4).

These findings were confirmed by histomorphometric analysis; significant increases in the number of epithelial surface microfolds, mean epithelial thickness, numbers of dermal infiltrating inflammatory cells and the percentages of collagen fiber-occupied dermal regions were detected in UVB-exposed mice compared with unexposed intact vehicle controls (P<0.01). However, significant dose-dependent decreases in UVB-induced histopathological dermis sclerosis and inflammatory signs were observed in mice orally administered 100, 200 and 400 mg/kg PCP compared with UVB control mice (P<0.01). At an oral dose of 2 ml/kg, PCS was also associated with significant decreases in epithelial surface microfolds, mean epithelial thickness, numbers of dermal infiltrating inflammatory cells and percentages of collagen fiber-occupied dermal regions compared with the UVB control (P<0.05; Table VII).

Notable increases in the number of cells immunolabeled for oxidative stress (NT and 4-HNE) and apoptosis (caspase-3 and PARP) markers were detected among epidermal keratinocytes on the dorsal back skin tissues in UVB-exposed control mice, with marked increases in dermal MMP-9 immunoreactivity (Fig. 5). These changes were confirmed by histomorphometric analyses; significant increases in epithelial NT-, 4-HNE-, caspase-3- and PARP-immunoreactive cells as well as dermal MMP-9 immunoreactivity, were observed in UVB-exposed control mice compared with unexposed control mice (P<0.01; Table VIII). However, significant decreases in oxidative stress and apoptosis markers and dermal MMP-9 immunoreactivity were observed in mice treated with 2 ml/kg PCS and 100, 200 and 400 mg/kg PCP compared with UVB-exposed mice (P<0.01; Table VII). In particular, PCP induced clear dose-dependent decreases in epithelial NT-, 4-HNE-, caspase-3- and PARP-positive cells, as well as dermal MMP-9-immunoreactive cells (Table VIII and Fig. 5).