Gomisin D (cat. no. 60546-10-3; purity, >98%), gomisin J (cat. no. 66280-25-9; purity, >98%) and gomisin O (cat. no. 72960-22-6; purity, >98%) were purchased from Wuhan ChemFaces Biochemical Co., Ltd. Chemical structure and properties are presented in Fig. 1A and Table I, respectively.

Dulbecco's modified Eagle's medium (DMEM), fetal bovine serum (FBS), phosphate-buffered saline (PBS), penicillin and streptomycin were obtained from Gibco (Thermo Fisher Scientific, Inc.). The Cell Counting Kit-8 (CCK-8) assay kit, dimethyl sulfoxide (DMSO), 3,4-dihydroxyphenylalanine (L-DOPA), α-MSH, Signal Boost Immunoreaction Enhancer kit assay, bovine serum albumin (BSA) and 4′,6-diamidino-2-phenylindole (DAPI) were obtained from Sigma-Aldrich (Merck KGaA). A cytotoxicity detection kit was purchased from Roche Diagnostics. The following were obtained from Thermo Fisher Scientific, Inc.: 5-(and-6)-chloromethyl-2′,7′-dichlorodihydrofluorescein diacetate acetyl ester (CM-H2DCFDA), APO-BrdU™ transferase dUTP nick end labeling (TUNEL) assay kit, high-capacity cDNA reverse transcription kit, SYBR Green qPCR master mix, Mammalian Protein Extraction Reagent (M-PER) and Pierce ECL western blotting substrate. The fluorescein isothiocyanate (FITC) Annexin V/Dead Cell Apoptosis Kit was purchased from Invitrogen (Thermo Fisher Scientific, Inc.). The RNeasy Mini Kit was purchased from Qiagen GmbH. The Bio-Rad protein assay kit and Mini-PROTEAN Precast Gels were obtained from Bio-Rad Laboratories, Inc. Hybond polyvinylidene difluoride membranes were obtained from Amersham (Cytiva). Antibodies against tyrosinase (cat. no. sc-15341), TRP-1 (cat. no. sc-25543), TRP-2 (cat. no. sc-25544) were obtained from Santa Cruz Biotechnology (USA). MITF (cat. no. 12590), phosphorylated (p)-PKA (cat. no. 4781), PKA (cat. no. 4782), p-CREB (cat. no. 9198), CREB (cat. no. 9197), α-tubulin (cat. no. 2144) and HRP-conjugated anti-rabbit IgG (cat. no. 7074) antibodies were obtained from Cell Signaling Technology, Inc. Alexa Fluor 488-conjugated anti-rabbit antibody (1:500; cat. no. A-21206) was obtained from Thermo Fisher Scientific, Inc.

Keratinocytes (HaCaT; cat. no. 300493; CLS Cell Lines Service GmbH) and melanocytes (B16F10; cat. no. CRL-6475 American Type Culture Collection) were cultured in DMEM supplemented with 10% (v/v) FBS and 1% (v/v) penicillin and streptomycin solution and cultured at 37°C with 5% CO₂. The experiments were performed using cells in the logarithmic growth phase. Cell growth was observed regularly and the medium was changed every 2–3 days. Cells treated with DMSO (1%) alone were used as controls in all the experiments.

For UVA and UVB exposure, keratinocytes were exposed to UVA (20 J/cm²) and UVB (50 mJ/cm²) irradiation (Bio-Link BLX-365; Vilber Lourmat Deutschland GmbH) with 5×8 W tubes that emit most of their energy with an emission peak at 365 nm (UVA) and 312 nm (UVB) (23,24).

The viability of keratinocytes irradiated with UVA (20 J/cm²) and UVB (50 mJ/cm²) after pretreatment with 10, 20, 30, 40 and 80 µM gomisin D, J and O at 37°C for 1 h was quantified. For cell viability analysis, the CCK-8 assay kit reagents were added to HaCaT keratinocytes (~2×10⁴ cells per well) and B16F10 melanoma cells (~2×10⁴ cells per well) according to the manufacturer's instructions and incubated at 37°C for 4 h. A cytotoxicity detection kit was used to determine extracellular LDH release in HaCaT keratinocyte culture medium. Absorbance was analyzed at 450 nm (CCK-8) and 490 nm (LDH) using a VICTOR Multilabel Plate Reader (PerkinElmer, Inc.).

Intracellular ROS levels were analyzed using CM-H2DCFDA. All procedures were performed according to the manufacturer's instructions. In brief, after the addition of gomisin D, J and O (30 µM) the cells were incubated at 37°C for 1 h. Cells were subsequently exposed to UVA (20 J/cm²) and UVB (50 mJ/cm²) at 37°C for 24 h. The HaCaT keratinocytes (~5×10⁴ cells per well) were washed with PBS and incubated with CM-H2DCFDA at 25°C for 0.5 h in the dark. Thereafter, the generation of intracellular ROS was visualized using a Carl Zeiss fluorescence microscope (Carl Zeiss AG) and measured based on a fluorescence unit using a fluorescent dye in a flow cytometer (Fit NxT Flow Cytometer; Thermo Fisher Scientific, Inc.). Data were analyzed using Attune NxT Flow Cytometer software version 3.1.2 (Invitrogen; Thermo Fisher Scientific, Inc.).

After treatment, HaCaT keratinocytes (~5×10⁴ cells per well) were trypsinized and centrifuged at 200 × g at 25°C for 3 min. Subsequently, the obtained cells were treated according to the instructions of the FITC Annexin V/Dead Cell Apoptosis Kit and TUNEL assay kit. Briefly, keratinocytes were rinsed twice with PBS, and keratinocytes in Annexin binding buffer were obtained and mixed with FITC/Annexin V (component A) and propidium iodide (PI) working solution. After incubation at room temperature for 15 min in the dark, keratinocyte apoptosis was measured using a flow cytometer (Fit NxT Flow Cytometer). The percentages of early and late apoptotic cells were calculated using Attune NxT Flow Cytometer software version 3.1.2 (Invitrogen; Thermo Fisher Scientific, Inc.). The keratinocytes in the different groups were treated with 4% paraformaldehyde solution at 25°C for 1 h and then the TUNEL assay kit was used to determine apoptosis according to the manufacturer's instructions. The nuclei were stained using DAPI mounting media (Sigma-Aldrich; Merck KGaA) for 0.5 min at room temperature. Using a Carl Zeiss fluorescence microscope (Carl Zeiss AG) three fields were observed. The percentage of TUNEL-positive cells was calculated using Attune NxT Flow Cytometer software (Invitrogen; Thermo Fisher Scientific). Apoptotic cells (TUNEL-positive cells) were determined using a Carl Zeiss fluorescence microscope (Carl Zeiss AG) and flow cytometry (Fit NxT Flow Cytometer).

Intracellular melanin content and tyrosinase activity assays were performed using a slight modification of a previously reported procedure (25). Melanocytes (~4×10⁴ cells per well) were treated with α-MSH (0.5 µM) in wells containing gomisin D, J and O (30 µM) at 37°C for 48 h. Then, the melanocytes were washed with ice-cold PBS, trypsinized using trypsin-EDTA and centrifuged at 500 × g at 25°C for 5 min. The melanocyte pellet was dissolved in a buffer containing 1N NaOH in 10% DMSO, lysed at 80°C for 1 h and then centrifuged at 20,000 × g at 25°C for 5 min. The relative melanin content was determined by measuring the absorbance at 475 nm using a VICTOR Multilabel Plate Reader. Intracellular tyrosinase activity was determined by measuring the rate of dopachrome production using L-DOPA. Melanocytes were washed with ice-cold PBS and lysed in PBS containing 1% (w/v) Triton X-100. The tyrosinase substrate L-DOPA (2 mg/ml) was prepared in phosphate lysis buffer. Each extract was placed in a well of a 96-well plate, and enzyme analysis was initiated by adding L-DOPA solution. Absorbance was measured at 475 nm using a VICTOR Multilabel Plate Reader and read every 10 min for 1 h in the dark. The end-point value of each measurement was expressed as a percentage of the control. Arbutin (A, 0.5 mg/ml) was used as a positive control.

Total RNA was isolated from each group of melanocytes using the RNeasy Mini Kit. RT was performed using the high-capacity cDNA reverse transcription kit according to the manufacturer's instructions to obtain the first strand of cDNA. The strand was then used as a template for qPCR using a Bio-Rad Chromo4™ instrument (Bio-Rad Laboratories, Inc.) and SYBR Green qPCR master mix. PCR was performed under pre-denaturation at 95°C for 5 min, denaturation at 95°C for 15 sec and annealing at 55–58°C for 30 sec. GAPDH was used as an internal reference for tyrosinase, TRP-1 and TRP-2. The relative expression levels of the target genes were calculated using the 2^−ΔΔCq method (26). Primer sequences were as follows: tyrosinase sense, 5′-ggccagctttcaggcagaggt-3′ and anti-sense, 5′-tggtgcttcatgggcaaaatc-3′; TRP-1 sense, 5′-agccccaactctgtcttttc-3′ anti-sense, 5′-ggtctccctacatttccagc-3′; TRP-2 sense, 5′-tccagaagtttgacagccc-3′ and anti-sense, 5′-ggaaggagtgagccaagttatg-3′; and GAPDH sense, 5′-aggtggtctcctctgacttc-3′ and anti-sense, 5′-taccaggaaatgagcttgac-3′.

Melanocytes (~4×10⁴ cells/well) were harvested and lysed using M-PER. All procedures were performed according to the instructions set by the manufacturer. All protein concentrations were determined using a Bio-Rad protein assay kit, according to the manufacturer's instructions. The loading buffer was then added to the protein supernatant and the solution was mixed. The mixture was boiled at 80°C for 10 min, and proteins (30 µg protein/lane) were separated via 7.5% Mini-PROTEAN Precast Gels and transferred onto a Hybond polyvinylidene difluoride membrane using a Trans-Blot SD semi-dry transfer cell. Immunodetection was performed using antibodies against tyrosinase (1:1,000), TRP-1 (1:1,000), TRP-2 (1:1,000) MITF (1:1,000), p-PKA (1:1,000), PKA (1:1,000), p-CREB (p 1:1,000), CREB (1:1,000) and α-tubulin (1:1,000) and the Signal Boost Immunoreaction Enhancer kit assay. The membrane was incubated overnight with the primary antibodies at 4°C. The secondary antibody (HRP-conjugated anti-rabbit IgG; 1:5,000) was added to the membrane and incubated at room temperature for 1 h. The protein bands were observed using an enhanced Pierce ECL western blotting substrate and semi-quantified as the ratio of the target protein band intensity to the intensity of the α-tubulin band. The band images were scanned using an ImageQuant 350 Analyzer (Amersham; Cytiva) and semi-quantified using the ImageQuant TL (version 8.1; Amersham; Cytiva).

Melanocytes (~2×10⁴ cells per well) were permeabilized using 4% paraformaldehyde solution at 25°C for 1 h and then incubated in 0.2% Triton X-100 at 25°C for 10 min. After washing with PBS, cells were blocked with 1% BSA at 25°C for 1 h and incubated with a primary antibody against p-CREB (1:200) at 4°C overnight for labeling and then incubated with a secondary antibody conjugated to Alexa Fluor 488 (1:500) for 3 h in the dark at room temperature. Subsequently, the nuclei were stained using DAPI mounting media (Sigma-Aldrich; Merck KGaA) for 0.5 min at room temperature and three images were collected using a Carl Zeiss fluorescence microscope (Carl Zeiss AG).

All assays were independently repeated at least three times. All statistical parameters are presented as the mean ± standard deviation (SD). Statistical analyses were performed using one-way analysis of variance (ANOVA) followed by Tukey's post hoc test. P<0.05 was considered to indicate a statistically significant difference.

CCK-8 and LDH assays were performed to investigate the cytoprotective effects of gomisin D, J and O on keratinocytes. HaCaT keratinocytes were treated with gomisin D, J and O under UVA, UVB or non-irradiation conditions. The keratinocytes were exposed to 10, 20, 40, 80 µM for 1 h, respectively, aim at determining the UVA or UVB irradiation according to the result of cell viability using CCK-8 assay. Compared with the control group, UVA or UVB irradiation reduced the viability of the keratinocytes. By contrast, gomisin D and J (at concentrations of 10, 20, 40 and 80 µM) dose-dependently enhanced the viability of keratinocytes. Gomisin O (at concentrations of 10, 20, 40 and 80 µM) did not change the viability of keratinocytes. The CCK-8 analysis showed that gomisin D, J and O did not significantly change keratinocyte viability at a concentration of 30 µM (Fig. 1B and C). Therefore, 30 µM gomisin D, J and O were used in follow-up experiments (Fig. S1A-C). Keratinocyte viability decreased when keratinocytes were cultured under UVA or UVB irradiation compared with the control group. However, under UVA or UVB irradiation, it was found that the viability of keratinocytes significantly increased after gomisin D and J treatment compared with the UVA or UBV-irradiated groups. Gomisin O did not alter keratinocyte viability after UVA or UVB irradiation (Fig. 1B). Keratinocyte damage was further monitored by measuring extracellular LDH release. LDH is released from the cytoplasm of keratinocytes into the culture medium as a result of keratinocyte membrane damage and lysis. An increase in LDH activity in keratinocyte culture supernatants is correlated with the rate of cytotoxicity (27). The results demonstrated that UVA and UVB irradiation markedly promoted LDH release compared with the control group. However, gomisin D significantly reduced LDH release and gomisin J markedly reduced LDH release under UVA or UVB irradiation, compared with the UVA and UVB only groups, respectively (Fig. 1C). The aforementioned experimental results showed that the antiproliferative and cytotoxic effects of UVA or UVB irradiation on keratinocytes were alleviated by gomisin D, J and O, in that order, thereby mitigating the damage caused by UVA or UVB irradiation on keratinocytes.

Oxidative stress is associated with the development and progression of UVA- or UVB-irradiated keratinocyte damage. High levels of ROS are formed in keratinocytes due to redox state imbalance (14,15). To explore the effect of gomisin D, J and O on endogenous ROS levels in UVA- or UVB-irradiated keratinocytes, fluorescence microscopy and flow cytometry were used to analyze the intracellular fluorescence intensity of the probe CM-H2DCFDA. As a result of fluorescence microscopy, images showed slight staining in the control and gomisin D-, J- and O-treated keratinocytes and marked staining in UVA- and UVB-irradiated keratinocytes. Meanwhile, keratinocytes subjected to UVA or UVB irradiation showed reduced CM-H2DCFDA staining after gomisin D and J treatment (Fig. 2A). According to the quantified results of flow cytometry, compared with the control, the intracellular ROS level of keratinocytes was markedly elevated after irradiation with UVA or UVB only compared with the control group. In gomisin D and J treatment groups, the intracellular ROS level significantly declined under UVA or UVB irradiation compared with the UVA- or UVB-irradiated groups. Intracellular ROS levels in UVA- or UVB-irradiated cells (after gomisin treatment) were inhibited at higher levels by gomisin D than gomisin J at the same concentration (Fig. 2B). These results indicated that gomisin D and J significantly inhibited UVA- or UVB-irradiated keratinocyte damage by inhibiting endogenous ROS production.

UVA and UVB irradiation of keratinocytes has been reported to lead to the inhibition of proliferation, ROS generation and apoptosis (28). To detect apoptosis, which is a reliable indicator of keratinocyte damage, keratinocytes were stained with Annexin V and PI (Annexin V/PI) and subjected to a TUNEL assay and analyzed by fluorescence microscopy and flow cytometry. Annexin V/PI flow cytometry was used to quantify apoptotic keratinocytes. The proportion of viable keratinocytes in gomisin D and J pretreatment was clearly higher than that in UVA- or UVB-irradiated keratinocytes, indicating that gomisin D and J inhibited UVA- and UVB-induced apoptosis (Fig. 3A). Quantified data from flow cytometry analysis showed that early and late apoptotic keratinocyte rates were markedly upregulated in UVA- or UVB-irradiated keratinocytes compared with the control group, whereas gomisin D and J effectively reversed these changes (Fig. 3B). UVA and UVB irradiation increased TUNEL nuclear staining activity, whereas gomisin D and J decreased the TUNEL nuclear staining activity rate under UVA and UVB irradiation. Gomisin D, J and O (30 µM) did not increase the TUNEL nuclear staining intensity compared with the control keratinocytes. As shown in Fig. 3C and D, it was confirmed that the level of apoptosis was suppressed the most by treatment with gomisin D, followed by gomisin J and then gomisin O under UVA or UVB irradiation. Overall, these results confirmed that the apoptosis of keratinocytes irradiated with UVA and UVB was inhibited by treatment with gomisin D, J and O in that order.

In addition to their photoprotective properties, the effects of gomisin D, J and O on melanin synthesis in melanocytes were investigated. The viability of melanocytes was assessed via a CCK-8 assay following treatment with gomisin D, J and O (at concentration of 10, 20, 40, 80 µM). Gomisin D, J and O treatment did not alter the viability of melanocytes at 10, 20 and 40 µM for 48 h (Fig. S2A-C). On measuring the intracellular tyrosinase activity, it was found that α-MSH treatment elevated intracellular tyrosinase activity compared with the control group, whereas gomisin D suppressed intracellular tyrosinase activity in a dose-dependent manner compared with the α-MSH group. However, gomisin J and O treatment did not result in a significant difference in intracellular tyrosinase activity (Fig. S3A-C). Prior to investigating the biological potential of gomisin D, J and O in α-MSH-induced melanogenesis, the viability of B16F10 melanocytes after treatment with gomisin D, J and O (30 µM) with or without α-MSH treatment was assessed using a CCK-8 assay. Gomisin D, J and O (30 µM) did not alter cell viability compared with the controls, with or without α-MSH (Fig. 4A). Therefore, 30 µM gomisin D, J and O were used in follow-up experiments. α-MSH is a commonly known melanogenic agent that increases intracellular melanin content by activating adenylate cyclase by binding to the melanocortin 1 receptor (28). To investigate the effect of gomisin D, J and O on melanogenesis in melanocytes, the intracellular melanin content was determined by visual observation and biochemical measurements. As shown in Fig. 4B, the intracellular melanin content was significantly increased by α-MSH treatment compared with the control group, whereas co-treatment with gomisin D showed a significant reduction of the intracellular melanin content compared with that of the α-MSH treatment group (Fig. 4B). Next, the effect of gomisin D, J and O on the intracellular tyrosinase content in MSH-stimulated melanocytes was determined. As expected, the intracellular tyrosinase activity of α-MSH-stimulated melanocytes increased compared with the control group, whereas that of α-MSH-stimulated melanocytes treated with gomisin D significantly decreased compared with the α-MSH group (Fig. 4C). These results indicated that gomisin D significantly inhibited the intracellular melanin content and tyrosinase activity without altering cell viability.

Melanogenesis markers (tyrosinase, TRP-1 and TRP-2) are activated by MITF (23,24). Therefore, the effects of gomisin D, J and O on the protein and mRNA expression levels of MITF, tyrosinase, TRP-1 and TRP-2 were examined; the melanocytes were treated with gomisin D, J and O in the presence or absence of α-MSH. As shown in Fig. 5A-C, the mRNA levels of tyrosinase, TRP-1 and TRP-2 were significantly increased by α-MSH treatment compared with the control group. By contrast, compared with the α-MSH treatment, gomisin D significantly downregulated the mRNA levels of tyrosinase, TRP-1 and TRP-2. In addition, gomisin D, J and O alone had no significant effects on tyrosinase, TRP-1 and TRP-2 mRNA levels. Western blotting showed that α-MSH treatment increased the protein expression levels of MITF, tyrosinase, TRP-1 and TRP-2 compared with the control group, and the levels of these proteins were downregulated by co-treatment with gomisin D compared with the α-MSH group (Fig. 5D and E). These results suggested that gomisin D had the potential to promote anti-melanogenesis effects, which may play a role in the downregulation of melanogenesis markers.

Phosphorylation of CREB and PKA upregulates MITF expression, which can enhance melanogenesis in melanocytes (23). PKA and CREB phosphorylation was detected by western blotting to further explore the potential regulatory mechanisms underlying the effects of gomisin D, J and O on α-MSH-induced melanogenesis. Western blot analysis indicated that gomisin D effectively inhibited the increase in the phosphorylation levels of PKA and CREB caused by α-MSH treatment. In addition, gomisin D, J and O alone had negligible effects on the phosphorylation levels of PKA and CREB (Fig. 6A and B). Immunofluorescence analysis was performed to observe the phosphorylation of CREB in α-MSH-stimulated melanocytes. Immunofluorescence analysis also revealed a clear decrease in the phosphorylation levels of CREB in gomisin D compared with that in the α-MSH treatment group (Fig. 6C). These results indicated that gomisin D suppressed melanogenesis in melanocytes, at least in part, through PKA and CREB phosphorylation.