The main reason for melanin production is that L-tyrosine is hydroxylated to L-dihydroxyphenylalanine (L-DOPA) under the catalysis of TYR, and then the o-diphenol is oxidized to the corresponding L-dopaquinone (Hearing, 1993; Prota, 1992). The inhibition of TYR was assessed using the modified dopachrome technique, with L-tyrosine as the substrate (Hapsari et al., 2012; Kamkaen et al., 2007). The EBR (including WEBR and EEBR) dissolved in DMSO was prepared into an initial solution with a concentration of 2 mg/mL by PBS solution, and then the initial solution was double diluted to obtain the extract solution with a concentration of 62.5 μg/mL, 125.0 μg/mL, 250.0 μg/mL, 500.0 μg/mL, 1000.0 μg/mL and 2000.0 μg/mL, respectively, as the sample to be tested. The concentration of the positive control arbutin was 2000 μg/mL.

Four test tubes A, B, C and D were taken and 250 μL of 1 mg/mL L-tyrosine solution was added, respectively. An additional 250 μL PBS was added to tubes A and C. Add 250 μL of the sample to be tested in tubes B and D, respectively. The solution in the four test tubes was mixed and heated in a water bath at 37 °C for 10 min. After that, 250 μL of 0.07 mg/mL TYR was added to the C and D tubes, and the same volume of PBS solution was added to the A and B tubes to fill the test sample volume. The four test tubes were placed in a water bath at 37 °C for 20 min. The above test solution was prepared by combining the contents of four test tubes and then added to a quartz cuvette. An ultraviolet spectrophotometer was used to detect the absorbance at 475 nm. The TYR inhibition rate was calculated using the following formula, with arbutin serving as a positive control. TYR inhibition rate: I = [(C-A) - (D-B)]/(C-A) × 100%.

Mouse melanoma B16F10 cells were purchased from Kunming Institute Zoology, Chinese Academy of Sciences and cultured in DMEM/F12 (Gibco, Billings, MT, United States) medium supplemented with 10% fetal bovine serum (FBS) and antibiotics (100 μg/mL penicillin, 100 μg/mL streptomycin and 100 μg/mL amphotericin B), and incubated for 5% CO₂ at 37 °C. CCK-8 cytotoxicity kit (MCE, United States) was used to detect cell viability after treatment with WEBR or EEBR or positive drug arbutin. In short, 1 × 10⁶ CFU/mL B16F10 cells were inoculated into each well of a 96-well plate. After 24 h, cells were exposed to 1 μM α-MSH and EBR at varying concentrations (1.5–200 μg/mL) for an additional 48 h and the concentration of the positive control arbutin was 200 μg/mL α-MSH is a polypeptide composed of 13 amino acids. It primarily acts on melanocytes, stimulating the conversion of tyrosine in these cells into melanin, and promoting melanocyte proliferation and melanin synthesis (Goenka et al., 2019). The blank control group comprised cells without α-MSH and EBR, while the α-MSH group was composed of cells treated with α-MSH alone. After the incubation period, each well was treated with 10 μL of CCK-8 solution and incubated for 4 h. The absorbance at 450 nm was measured using a microplate reader, and the control group was 100%. The experiment was repeated six times, and the results were expressed as a percentage change relative to the control.

B16F10 cells were incubated with 1 μM α-MSH and different concentrations of EBR (1.5–200 μg/mL) for 48 h. The control group did not receive specific treatment, while the cells treated with α-MSH only constituted the α-MSH group. Arbutin was used as a positive control in the experiment and the concentration was 200 μg/mL. After washing twice with phosphate-buffered saline (PBS), the cells were lysed on ice with 1% triton X-100 solution for 10 min. Subsequently, these samples were centrifuged at 13,000 g for 10 min. The supernatant was used to measure intracellular TYR activity and the precipitate was used to quantify the melanin content of B16F10 cells (Tedasen et al., 2024). The precipitate was dissolved in 100 μL 1 mol/L NaOH at 80 °C for 30 min, and the absorbance was measured at 475 nm using a microplate reader. The absorbance of the control group was 100%. The results were reported as a percentage change relative to the control, and the experiment was set to three replicates. Subsequently, the protein concentration in the supernatant was quantified using a BCA protein quantification kit (Biosharp, Hefei, Anhui, China). In the steps, 50 μL of supernatant and 50 μL of L-DOPA (5 mM) were incubated at 37 °C for 1 h to measure the TYR inhibition activity, and then the absorbance was measured at 475 nm.

B16F10 cells were treated with 1 μM α-MSH and different concentrations of EBR (1.5–200 μg/mL) and 200 μg/mL of arbutin for 48 h. After incubation, the cells were washed twice with PBS and then harvested. Total RNA was isolated using a Trizol reagent. The purity and concentration of the resulting RNA were evaluated using a Nanodrop spectrophotometer. A total of 1 μg of extracted total RNA was converted into cDNA using the 5X All-In-One MasterMix Reverse Transcription Kit (ABM, United States). Subsequently, the samples were amplified using the BlasTaqTM 2X qPCR MasterMix premix kit (ABM, United States). Gene expression analysis was performed using a LightCycler^® 96 SW 1.1 real-time PCR system (Roche Diagnostics, Indianapolis, IN, United States). All reactions were repeated three times. The relative fold expression levels of TYR and MITF were calculated. The threshold cycle (Cq) was determined and normalized to the average level of the housekeeping gene (GAPDH) level. The relative expression level of the gene was expressed as the calculation formula ‘2^−ΔΔCt’ and normalized by GAPDH (Livak and Schmittgen, 2001). Primers for B16F10 are shown in Supplementary Table S1. Otherwise, the dissected skin tissue was placed in the Trizol reagent to isolate RNA. The guinea pig primers involved in this experiment refer to Supplementary Table S1.

B16F10 cells were treated with 1 μM α-MSH and different concentrations of EBR (1.5–200 μg/mL) and 200 μg/mL of arbutin. After 48 h, the cells were lysed on ice for 30 min using RIPA lysis (Sigma-Aldrich, Merck, United States) buffer, and 0.1% protease inhibitor (MedChemExpress, United States) and 0.1% phosphatase inhibitor (MedChemExpress, United States) were added to it. The proteins in the supernatant were collected after centrifugation at 13,000 × g for 15 min. A BCA protein quantification kit was used for protein determination. Each quantitative protein (30 μg) was loaded into each well of the 4%–20% SDS-PAGE gel. After SDS gel electrophoresis, the protein was transferred to 0.22 μM polyvinylidene fluoride (PVDF, Millipore, Merck, Germany) membrane. Subsequently, the PVDF membrane was blocked with 5% (w/v) bovine serum albumin (BSA, BioFroxx, Germany) at room temperature for 2 h. After the blocking, the membrane was incubated with primary antibody against TYR (Bioss, Beijing, China) and MITF (Bioss, Beijing, China) (1: 1000 dilution) at 4 °C overnight. After washing with TBST buffer containing 0.1% Tween-20 (BioFoxx, Germany) for 4 times, each time for 5 min, the membrane was incubated with secondary antibody [KPL peroxidase-labeled rabbit IgG (H + L, Seracare, United States)] at room temperature for 2 h. An advanced enhanced chemiluminescence (ECL) reagent (Tanon, Shanghai, China) was used to detect the protein bands on the PVDF membrane and photographed using a chemiluminescence analysis system (Tanon5200Multi, Shanghai, China). ImageJ software was used to quantify TYR and MITF expression levels for Western blot analysis.

Three flowered guinea pigs (5 weeks of age, Hartley) were purchased from Chengdu Dossy Experimental Animals Co., ltd (Chengdu, Sichuan, China). We used a total of 9 guinea pigs and conducted three independent experiments, with each sample consisting of three guinea pigs. In the cage, they were placed under standard experimental conditions (22 °C ± 1 °C, 55% ± 5% humidity, 12 h of light and 12 h of dark cycle) and kept on a standard diet and free available water. After a week of adaptation, we used an electric haircut to cut off the back hair of female guinea pigs. Five independent square (1 cm × 1 cm) dorsal areas of each guinea pig were exposed to UVB radiation (305 nm, SANIKYO, Japan). Guinea pigs were exposed to UVB light for 3 weeks at intervals of 1 day, 1 h per day, starting from the second day after shaving (Yoshida et al., 2007). The arbutin (positive control), EBR including WEBR and EEBR were dissolved in a mixture of ethanol and water (1: 9, v/v). From the second day after the last UVB irradiation, each sample solution was applied locally with a pipette to separate the shaving area once a day for 2 weeks. After that, the guinea pigs were anesthetized and sacrificed, and the skin tissues of different samples were taken for subsequent experiments. The use of guinea pigs in research was authorized following the ethical approval application submitted by the Kunming Institute of Zoology, Chinese Academy of Sciences (Ethical registration number: IACUC-RE-2024-08–011).

The skin tissue was incubated in 2 mol/L NaBr at 37 °C for 5 h to separate the epidermis from the dermis. To measure the melanin content, the epidermis was dissolved in 250 μL of 1 mol/L NaOH containing 10% DMSO and treated at 80 °C for 1 h and centrifuged 13, 000 × g for 30 min. The absorbance of the supernatant was measured at 475 nm. For the determination of TYR activity, the isolated epidermis was placed in a 1% Triton X-100 (Beyotime, Shanghai, China) containing 0.1 mM PMSF and ground in a grinding machine. The supernatant was collected by centrifugation at 13,000 g for 30 min at 4 °C. Subsequently, the protein concentration in the supernatant was quantified using a BCA protein quantification kit (Biosharp, Hefei, Anhui, China). In the steps, 50 μL of supernatant and 50 μL of L-DOPA (5 mM) were incubated at 37 °C for 1 h to measure the TYR inhibition activity, and then the absorbance was measured at 475 nm.

The dissected skin tissue was placed in the RIPA lysis (Sigma-Aldrich, Merck, United States) solution of 0.1% protease inhibitor and 0.1% phosphatase inhibitor, ground for 20 min in a grinder, and centrifuged at 13,000 g at 4 °C for 30 min. To maintain the protein concentration, the protein concentration in the supernatant was quantified using a BCA protein quantification kit (Biosharp, Hefei, Anhui, China). The following Western blot steps are shown in Materials and Methods 2.7. ImageJ software was used to quantify TYR and MITF expression levels for Western blot analysis.

Tissue sections were fixed with 4% paraformaldehyde, dewaxed, and dehydrated, and then washed with phosphate-buffered saline (PBS, pH = 7.2) to remove excess reagents. The sections were then incubated for 15 min in boiling citrate buffer (10 mM, pH 6.0) to restore the antigenic sites. Next, the slices were treated with 3% H₂O₂ for 25 min and washed three times in PBS (pH 7.4) on a horizontal shaker (Qilinbeier, Jiangsu, China) to inhibit endogenous peroxidase activity. This was followed by a 15-min blocking step with 2% BSA at 37 °C. Subsequently, the sections were incubated overnight at 4 °C with purified rabbit polyclonal antibodies diluted 1:1,000, followed by washing with PBS (pH 7.4). The sections were then exposed to a secondary antibody (KPL peroxidase-labeled rabbit IgG [H + L, Seracare, United States]) for 50 min at room temperature. The immunoperoxidase reaction was carried out using 3,3-diaminobenzidine (Servicebio, Wuhan, China), and the sections were counterstained with hematoxylin. The primary antibody was substituted in the negative control with an antigen-free antibody. Images were acquired using a slide scanning system (Teksqray, Shenzhen, China) and ImageJ software was used to quantify TYR and MITF expression levels in guinea pig skin tissues.

The 0.03 g insect extraction of water and ethanol was dissolved with 0.6 mL 60% ethanol and then ultrasound for 5 min and centrifuged at 12,000 g for 10 min at 25 °C. The supernatant was diluted to 0.6 mL with 60% ethanol. The 2 mg/mL tannic acid standard solution was diluted to 0.025, 0.05, 0.1, 0.2, 0.3, 0.4 mg/mL. The total phenol content of the EBR (The concentration was 0.39–25.00 μg/mL) was determined by the total phenol detection kit, according to the manufacturer’s instructions (Norminkoda, Wuhan, China) and mixed well at room temperature for 30 min, and the absorbance was read at 750 nm by ultraviolet spectrophotometer, and the formula: △A = A_{determination} - A_control, △A’ = A_standard - A_blank. The standard curve ‘y = kx + b’ was drawn with the concentration of tannic acid as abscissa and △A′ as ordinate, and △A was brought into the formula to obtain x. Total phenol content (mg/g mass) = x × V1/(W × V1/V) × D (V: Add extract volume, 0.6 mL; V1: sample volume in the reaction: 0.05 mL; W: sample quality: g; D: dilution multiple, undiluted is 1).

The 0.03 g insect extract was dissolved in 0.6 mL of 60% ethanol and then sonicated for 5 min, followed by centrifugation at 12,000 g for 10 min at 25 °C. The supernatant was diluted to 0.6 mL with 60% ethanol. The 10 mg/mL rutin standard solution was diluted to 0.005, 0.01, 0.02, 0.039, 0.078, 0.156, 0.3125, 0.625, and 1.25 mg/mL. The total flavonoid content of the EBR (The concentration was 0.39–25.00 μg/mL) was determined by the total flavonoid content detection kit, according to the manufacturer’s instructions (ZCi Bio, Shanghai, China) and mixed well at 37 °C water bath for 45 min, centrifuged at 10,000 g for 10 min to take the supernatant, and the absorbance was read at 510 nm by a microplate reader (Epoch Etock; Biotek, United States), and calculated using the formula: △A = A_{determinationm} - A_control, △A’ = A_standard - A_blank. The standard curve ‘y = kx + b’ was drawn with the concentration of rutin as abscissa and △A′ as ordinate, and △A was brought into the formula to obtain x. Total flavonoids content (mg/g fresh weight) = x × V/W (V: Add extract volume, 0.6 mL; W: sample quality: g).

The appropriate amounts of samples including water extracts and ethanol extracts (0.5 mL) were added to precooled methanol/acetonitrile/water solution (2:2:1, v/v), vortex mixing, low temperature ultrasonic 30 min, −20 °C standing 10 min, 14,000 × g centrifugation for 20 min at 4 °C, the supernatant was vacuum dried, 100 μL solution (acetonitrile: water = 1:1, v/v) was added to the mass spectrometry analysis, vortex, 14,000 × g centrifugate to 15 min at 4 °C, the supernatant was injected for analysis.

The phenols and flavonoids in B. rhynchopetera Fairmaire were imported into the PubChem database (https://pubchem.ncbi.nlm.nih.gov/) to obtain their SMILES structures. They were used to predict targets on the SwissTargetPrediction platform (http://www.swisstargetprediction.ch/), and targets with scores greater than 0.2 were selected. Potential targets related to melanin deposition were retrieved from the GeneCards database (https://www.genecards.org) (correlation score >3), OMMI database (https://www.omim.org), and DisGeNET database (https://www.disgenet.org) with ‘Melanin deposition’, ‘TYR inhibition’ as the keywords (Xu J. et al., 2024). The intersection of disease targets and component targets was taken as the potential effect of EBR on improving melanin deposition, and the Venn diagram was drawn by an online mapping website (https://www.bioinformatics.com.cn) (Tang et al., 2023).

Through the above process, we identified the targets of active ingredients related to melanin deposition and TYR inhibition from B. rhynchopetera Fairmaire. To illustrate the relationship between the primary protein targets of the drug, a protein-protein interaction (PPI) network was generated using the STRING website (https://cn.string-db.org) and visualized using Cytoscape 3.10.1 software. In addition, a complex target-pathway-network was created using Cytoscape 3.10.1 software. Compounds, proteins, and signal pathways are represented as nodes in the graph network, while the interactions between compounds, proteins, and signal pathways are shown as edges. The nodes in the graphical network represent many substances, targets, and related diseases. The edges between them show the connection between the composite target and the target disease. After analyzing the values of ‘degree’, ‘closeness’, and ‘betweenness’, the target with a comprehensive score greater than the average of 0.900 and a degree greater than the average is selected as the key target.

DAVID database (https://david.ncifcrf.gov) was used to perform Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis on the potential targets of the EBR to improve melanin deposition and inhibit TYR activity. Enrichment analysis was performed to predict the main targets and possible biological activity mechanisms of the active components of the EBR to improve melanin deposition and play a whitening role. GO enrichment includes molecular function (MF), biological process (BP) and cellular component (CC) and these analyses are carried out on the Wei Sheng Xin platform (https://www.bioinformatics.com.cn).

According to the results of KEGG pathway enrichment, the pathways most closely related to melanin deposition and the TYR metabolism mechanism were identified, and the corresponding active components were molecularly docked. The top four core compounds were selected as therapeutic small molecules. The SDF format of the needed compounds was retrieved from the PubChem database and converted to the MOL format using PyMOL software. To determine the 3D structure of the protein, the SWISS-MODEL database (https://swissmodel.expasy.org) was used to retrieve the PDB format export, and then PyMOL software was employed to remove excess chains, including ligands, dehydrated and hydrogenated proteins, and drug molecules, and to set the docking box. Semi-flexible docking was performed on the identified molecules using Mgltools software (version 1.5.6) and AutoDock Vina software (Eberhardt et al., 2021; Trott and Olson, 2010) to find the binding energy and position of each component of B. rhynchopetera Fairmaire. The docking state with a binding energy of <−7.0 kcal/mol was considered to be good (Zhu et al., 2024), and the molecular docking results were visualized using PyMOL software.

By comparing the binding energies of quercetin, luteolin, myricetin, apigenin, and TYR, as well as AKT1 and EGFR using molecular docking, we selected TYR and the aforementioned four compounds for MD simulation. To better understand the complex interaction processes between the TYR protein and the four candidate small molecules, this study employs GROMACS software (version 2024.2) to predict the dynamic changes between the biomacromolecule and the ligands (Lemkul, 2018). To ensure the MD simulation runs correctly, the protein and ligands must be converted into GROMACS-compatible force field files for the simulation. The OPLS-AA/L all-atom force field was selected, and the TIP3P model was used as the solvent. The boundaries of the simulation box were set to 1 nm × 1 nm × 1 nm. In the simulation box, one molecule of TYR protein and one molecule of the ligand were included, while the remaining space was filled with water molecules (spc216). Subsequently, sodium and chloride ions were added to neutralize the system’s charge, and energy minimization was performed to optimize the system components.

In the initial state of the protein-ligand complex, there might be excessively short distances between atoms, conflicts, or unnatural interactions. If the dynamics simulation is started directly, these unreasonable conditions may destabilize the system or even cause the simulation to fail. Through energy minimization and equilibration steps, these issues can be mitigated, allowing the system to relax under mild conditions and avoid drastic structural changes gradually. Thus, hydrogen atoms in the ligand must be constrained before proceeding with the NVT (constant Number of particles, Volume, and Temperature) and NPT (constant Number of particles, Pressure, and Temperature) calculations. The MD simulation was executed under the NPT ensemble for a duration of 50 nanoseconds at 300 K (using C-rescale for temperature control) and 1.0 bar (Childers and Daggett, 2018). The system maintained constant volume and used the Smooth Particle Mesh Ewald (PME) method for efficient long-range electrostatic interactions.

Finally, MD simulations were performed to obtain curves representing protein-ligand interactions, such as the RMSD and RMSF of the protein and ligand. For analyzing the dynamics of the protein-ligand interactions, interaction energy calculations should not be considered part of a regular simulation (Tedasen et al., 2024). Therefore, the average short-range Coulombic interaction energy and the short-range Lennard-Jones energy were used. The total interaction energy, however, is useful in this case.

All performed experiments were undertaken in triplicate (n = 3) and the results were conducted utilizing GraphPad Prism 6.01 (GraphPad Software, San Diego, CA, United States). Statistical differences were assessed using a one-way ANOVA. The data are presented as mean ± standard deviation (mean ± SD), with statistical significance defined as *p < 0.05, **p < 0.01.

The potential of EBR as a whitening agent to inhibit melanin formation was evaluated by measuring TYR inhibitory activity. As shown in Figure 1, compared with the blank control, there were significant differences in different concentrations of WEBR and EEBR (p < 0.01). With the increase in different concentrations, the inhibition rate of WEBR and EEBR on TYR activity increased gradually. When the concentration was 1000.0 μg/mL, the inhibition rate of WEBR and EEBR on TYR activity could reach more than 50%.

As shown in Figure 2A, compared to the blank control (PBS), the cell viability of WEBR and EEBR was above 90% in the range of 1.5–200 μg/mL. When the concentration reached 100 μg/mL, the cell viability reached 100%, indicating that the EBR should be safe and reliable in the field of cosmetics.

As shown in Figure 2B, as expected, α-MSH treatment significantly increased the melanin content in B16F10 cells. Compared with the α-MSH group, the concentration of EBR at 100 and 200 μg/mL exhibited a significant inhibitory effect on the melanin synthesis of B16F10 cells. The content of melanin eliminated by 200 μg/mL WEBR was 35.040% ± 1.683%, and the content of melanin eliminated by 200 μg/mL EEBR was 25.325% ± 4.094%, respectively, showing a certain concentration dependence. Furthermore, different concentrations of EBR treatment had a significant inhibitory effect on TYR activity in B16F10 cells (Figure 2C). Under a concentration range of 1.5–200.0 μg/mL, the inhibitory effect gradually increased with increasing concentration, exhibiting a concentration-dependent effect.

Melanin production is regulated by enzyme cascades including TYR. The expression of TYR is regulated by the transcription factor MITF (Liu and Fisher, 2010). Western blot analysis showed that the protein expression levels of TYR and MITF were significantly downregulated after α-MSH treatment (Figure 2D; Supplementary Figure S1). It indicated that the EBR significantly inhibited the level of TYR and MITF protein.

qPCR was used to detect the expression of TYR and MITF mRNA in B16F10 cells induced by the EBR. The expression of β-actin mRNA remained stable under varying concentrations of EBR, confirming that no other factors interfered with the expression of TYR and MITF mRNA during the experiment (Figures 2E,F). Compared with α-MSH, the expression of TYR and MITF mRNA was significantly inhibited by different concentrations of EBR for 48 h, especially with EEBR. The expression of TYR and MITF mRNA decreased with increasing EBR concentration. It is speculated that the EBR inhibits the production of melanin by inhibiting the activity of TYR and MITF gene.

The effect of EBR on melanin deposition in guinea pig skin was detected by measuring melanin content and TYR activity. The melanin content of the control treated with UVB alone was 100%. The results showed that the deposition of melanin decreased with an increase in the concentration of the EBR, indicating a concentration-dependent effect (Figure 3A). Treated with arbutin and the EBR, the melanin content of arbutin was 53.120% ± 6.237%, the content of WEBR was 57.070% ± 1.047%, and the content of EEBR was 49.895% ± 2.539%. Compared with the control group, the effect of EEBR on eliminating melanin deposition in guinea pig skin was slightly stronger than that of arbutin, indicating that the EBR eliminated melanin deposition.

The TYR activity treated with only UVB was assigned to 100%. The results showed that the inhibitory activity of TYR increased with the increase in the concentration of EBR, exhibiting a concentration-dependent effect (Figure 3B). When the concentration of the EBR was 1000 μg/mL, the inhibition rate of TYR by the WEBR was 55.893% ± 2.672%, and the inhibition rate of TYR by the EEBR was 63.780% ± 5.792%. At the same time, the inhibition rate of TYR by arbutin at 1000 μg/mL was 64.483% ± 3.485%. The TYR inhibition rate of EEBR was equivalent to that of arbutin.

To elucidate the inhibitory effect of EBR on melanin synthesis, we determined the expression of TYR and MITF proteins by Western blot analysis (Figure 3C) and mRNA levels by qPCR (Figures 3F,G). Compared with UVB treatment, arbutin and EBR inhibited the expression of TYR and MITF in guinea pig epidermal cells. The expression of TYR and MITF mRNA decreased with increasing EBR concentration. The inhibition of protein expression is consistent with the inhibition of gene expression (Figure 3C; Supplementary Figure S1). It is speculated that the EBR inhibits melanin production by suppressing the expression of TYR and MITF genes, thereby reducing the protein levels of TYR and MITF in the epidermis of guinea pigs, which in turn inhibits cell melanin production.

To evaluate the content of melanin in the epidermis of guinea pigs treated with EBR, the results of immunohistochemical antibody identification showed that compared with the samples treated with UVB irradiation alone, the positive expression of epidermis treated with arbutin and EBR including WEBR and EEBR decreased (the positive expression of DAB was brownish yellow) (Figures 3D,E). The results showed that the activity of TYR and MITF protein was decreased, which could further indicate that the EBR can eliminate the increase of melanin content stimulated by UVB.

In this study, the contents of total phenols and total flavonoids were characterized. Tannic acid was used as a standard to determine the total phenol content. The linear equation of standard tannic acid was y = 2.5204x + 0.0253; R ² = 0.9976 (Supplementary Figure S3A). Using the standard tannic acid calibration curve equation, the total phenol content obtained from the WEBR and EEBR is shown in Figures 4A,B. The total phenol content of EEBR was higher than that of WEBR, and with the increase of concentration, the total phenol content of EEBR and WEBR increased gradually, indicating that both EEBR and WEBR contained phenolic compounds. Rutin was used as a standard to determine the content of total flavonoids. The standard linear equation was y = x + 1E-16, R ² = 1 (Supplementary Figure S3B). As shown in Figure 4C, with the increase of EEBR concentration, the flavonoid content gradually increased, while the WEBR did not detect any total flavonoid content, which may be attributed to the insolubility of flavonoids in water, resulting in the inability to detect the presence of flavonoids in the WEBR.

To determine the total phenols and total flavonoids in the EBR (WEBR and EEBR), we carried out UHPLC-Q-TOF MS analysis. The results showed that the response intensity and retention time of each chromatographic peak overlapped, indicating that the variation caused by instrument error was minimal throughout the experiment, which ensured the repeatability and reliability of the data. The main peaks observed in the chromatogram were attributed to positive and negative modes using ESI-MS (Supplementary Figure S4). The OPLS-DA analysis was performed on the peaks extracted from the WEBR and the EEBR to examine the concentration and distribution of metabolites in the groups and the differences between them. The results showed that samples from different groups (WEBR and EEBR) could be distinguished, while samples from the same group had a more concentrated metabolite distribution, indicating that the collected data could be used to identify different compounds between groups (Supplementary Figure S5A, 5B. After the combination of positive and negative ions, a total of 1535 compounds were identified, including 1003 positive ion compounds and 535 negative ion compounds.

Phenols and flavonoids are recognized as natural secondary metabolites that have been demonstrated to possess potential benefits, including antioxidant, antibacterial, and anti-inflammatory effects. Many phenols and flavonoids have been reported to inhibit TYR activity, which in turn inhibits melanin production (Gillbro and Olsson, 2011; Zhu and Gao, 2008), and have anti-aging properties due to their antioxidant and anti-inflammatory properties (Ferreira et al., 2021; Silva et al., 2017). The biological activity of the extract is closely related to the content of total flavonoids and phenols. Therefore, in this study, UHPLC-Q-TOF MS was used to characterize the possible compounds in the EBR. The results showed that there were 1538 compounds in the EBR, including phenols, flavonoids, carboxylic acids and derivatives, benzene and substituted derivatives, organooxygen compounds, organooxygen compounds, steroids, and steroid derivatives, etc (Figure 4D). Total phenols accounted for 2.48%, and flavonoids accounted for 7.43%. Among them, 54 kinds of flavonoids and 28 kinds of phenolic compounds were identified (Supplementary Table S2).

The Venn diagram of 161 compound targets and 954 disease targets was drawn. The intersection part was a total of 41 targets of phenols and flavonoids of EBR acting on melanin and TYR (Figure 5A). These intersection targets were imported into the string database and screened the value of ‘degree’, ‘closeness’, and ‘betweenness’ to obtain the interaction diagram between the target and the target, as shown in Figure 5B and Supplementary Figure S6. Based on Cytoscape 3.10.0 software, a compound-target-disease network was constructed to predict the relationships between the active components, targets, and diseases of Blaps rhynchopetera Fairmaire, and the key compounds inhibiting melanin deposition and TYR inhibition were identified (Figure 5C). The network consists of 225 nodes and 933 edges, with an average degree of 8.29. Each edge represents the interaction between chemical components and targets, and the higher the degree, the more critical its role in the interaction network. According to the compound-target-disease network, quercetin had the highest number of docking targets (72), followed by myricetin (56), luteolin (34), apigenin (32), etc. Among the 161 targets, EGFR had the most docking compounds (21), followed by AKT1 (19) and TYR (14), indicating that these components and targets may play a crucial role in the process by which B. rhynchopetera Fairmaire exerts its skin whitening effect.

Based on the DAVID database, GO and KEGG enrichment analysis was performed on 41 potential melanin deposition and TYR inhibition targets of Blaps rhynchopetera Fairmaire, and 260 items of biological process (BP) (p < 0.05), 44 items (p < 0.05) cellular component (CC), and 72 items (p < 0.05) of molecular function (MF) were obtained. According to the number of genes, the top 10 were selected for GO function analysis and the top 20 were selected for KEGG enrichment analysis (Figures 6A,B). GO function analysis indicated that the biological processes mainly involve phosphorylation, negative regulation of the apoptotic process, response to xenobiotic stimulus, and negative regulation of gene expression, among others, which occur primarily in the plasma membrane, membrane, cytoplasm, nucleus, and other cellular compartments (Figure 6A). KEGG enrichment showed that pathways in cancer, PI3K-Akt signaling pathway, MAPK signaling pathway, etc., Were ranked at the top, suggesting that the B. rhynchopetera Fairmaire may exert antioxidant, anti-melanin, and anti-TYR through these pathways (Figure 6B). Based on KEGG enrichment analysis and the ‘Component-Target-Disease’ network of B. rhynchopetera Fairmaire in Figure 5C, we further screened the top four key components (degree value >55) and the top 11 key targets (degree value >20) according to the degree value, and constructed the ‘Component-Target-Disease’ interaction network of B. rhynchopetera Fairmaire (Figure 6C).

To verify the reliability of the screened compound-target interaction, molecular docking analysis selected the top three proteins. According to the results of the ‘Component-Target-Disease’ interaction network analysis from Figure 5C, the targets TYR, Akt1 and EGFR with higher degree values were selected for molecular docking simulation with the main active components quercetin, myricetin, luteolin and apigenin of Blaps rhynchopetera Fairmaire. The results of molecular docking show that the binding activity between the receptor and the ligand is expressed by the binding energy, and the binding energy <0 kJ/mol can be used to determine the spontaneous binding ability of the ligand and the receptor. The smaller the binding energy, the higher the affinity of the receptor and ligand, the greater the possibility of action. The binding energy <−5.0 kJ/mol indicates that the binding activity is high, and the binding energy <−7.0 kJ/mol indicates that the binding activity is strong. The results showed that the binding energies of active components including quercetin, myricetin, luteolin and apigenin in B. rhynchopetera Fairmaire to target protein TYR, AKT1, and EGFR were all <−7.0 kJ/mol, indicating that the binding ability was very high, as shown in Figure 7A and visualized by PyMOL software which shown in Figure 7B.

Based on the results of molecular docking, we selected small molecules quercetin, luteolin, myricetin, and apigenin with strong binding energy and receptor TYR for MD simulation to further verify the dynamic characteristics of their interactions, including RMSD, RMSF, Cour-SR and LJ-SR interaction energy to describe the stability of ligand-receptor in 0–50 ns, as shown in Figure 8, Supplementary Figure S8 and Table 1. The RMSD of quercetin, luteolin, myricetin and apigenin were binding to TYR remained stable below 5 nm, indicating that they were in a consistent state during the simulation. In the presence of TYR protein molecules, it can be seen from the RMSD peak diagram that quercetin, luteolin, myricetin and apigenin fluctuate smoothly, indicating that they can be combined with TYR stably, and the overall carbon skeleton fluctuation state of TYR is also stable (Figures 8B,D,F,H). RMSF represents the degree of freedom of each atom in the molecule. RMSF analysis of the four sets of trajectories shows that there is no significant fluctuation in the key residues of TYR. The RMSF values of the key residues in the active pocket were analyzed. The RMSF values of the key residues of quercetin, luteolin, myricetin, and apigenin interacting with TYR were at a relatively low level, with an average value of about 0.5 nm, indicating that the key residues lost flexibility due to the binding of TYR to small molecule ligands, making it stable (Figures 8B,D,F,H). In addition, to quantify the strength of the interaction between quercetin, luteolin, myricetin, and apigenin with TYR, Cour-SR, and LJ-SR were used to calculate the nonbonded interaction energy. As shown in Table 1 and Supplementary Figure S8, the total average interaction energy (kJ/mol) (Average Cour-SR interaction energy (kJ/mol) + Average LJ-SR interaction energy (kJ/mol) from high to low are myricetin-TYR, apigenin-TYR, luteolin-TYR and quercetin-TYR, respectively. Therefore, myricetin is more likely to bind to TYR receptors and bind more stably.