Enzymatic preparation of epigallocatechin gallate diglucoside and bioactivity assessment

Abstract

Epigallocatechin gallate (EGCG), the predominant catechin in green tea, has limited application due to its poor water solubility and instability. To address these issues, this study utilized recombinant sucrose phosphorylase to catalyze the glucosylation of EGCG, successfully synthesizing (-)-epigallocatechin gallate 4′,4″-O-α-D-diglucopyranoside (EGCG-2G). The process was optimized using response surface methodology, achieving a 97.46% conversion rate of EGCG. EGCG-2G was purified to ≥ 99% purity by semi-preparative liquid chromatography. It exhibited approximately 124-fold higher water solubility than EGCG and demonstrated significantly enhanced stability under thermal and acidic conditions (50 °C, pH = 5), with an 84.82% and 35.36% improvement over EGCG and commercial (-)-epigallocatechin-3-gallate-4′-O-α-D-glucoside (EGCG-1G), respectively. Furthermore, EGCG-2G displayed notable antioxidant, anti-inflammatory, and anti-melanogenic activities. It effectively scavenged intracellular and extracellular free radicals, reduced inflammatory cytokine levels, and inhibited melanin synthesis. Molecular docking and gene expression analyses suggested that its anti-melanogenic effect might be associated with the MC1R/cAMP/MITF signaling pathway.

Supplementary Information

The online version contains supplementary material available at 10.1186/s13568-026-02014-8.

Introduction

(-)-Epigallocatechin gallate (EGCG) is the primary catechin in green tea, constituting about 50% of its total polyphenols (Oliveira et al. 2016). It exhibits a range of notable bioactivities, including antioxidant (Shahidi et al. 2015), anti-inflammatory (Nichols et al. 2010), skin-whitening (Wang et al. 2023), and anti-cancer activities (Wang et al. 2002). However, EGCG is highly susceptible to oxidative degradation under neutral and alkaline conditions (Zeng et al. 2017), resulting in loss of activity and poor bioavailability (Dai et al. 2020). Furthermore, due to its relatively low solubility, it is not yet considered an ideal agent for pharmaceutical, commercial food, or cosmetic applications (Kim et al. 2016).

To address the aforementioned issues, enzymatic glycosylation has emerged as an effective strategy to enhance the water solubility and stability of EGCG (Andreu et al. 2023). Previous studies have demonstrated that glycoside hydrolases, such as sucrose phosphorylase (Kitao et al. 1993) and cyclodextrin glycosyltransferase (Gonzalez-Alfonso et al. 2018); and glycosyltransferases such as dextransucrase (Kim et al. 2016) can catalyze the specific binding of glucosyl groups to EGCG, leading to the formation of EGCG-glucosides (EGCG-Gs), which exhibit significantly improved water solubility and stability compared to EGCG. For instance, Noguchi et al. (2008) used α-amylase to synthesize EGCG 5′-O-α-D-glucopyranoside and EGCG 7-O-α-D-maltoside. The 5′-glucoside, in particular, demonstrated high thermal stability and solubility (Noguchi et al. 2008). Three EGCG-Gs derivatives were synthesized using dextransucrase, which demonstrated over 50-fold higher water solubility than the parent compound along with significantly enhanced UVB stability (Moon et al. 2006a, b). The glycosylation of EGCG typically yields a variety of products. The structure-activity relationships of nine EGCG-Gs were systematically investigated, confirming that the number and position of the glycosyl groups directly influence the solubility and stability of the derivatives (Kim et al. 2016). Although the stability of the (-)-epigallocatechin-3-gallate-4′-O-α-D-glucoside (EGCG-1G) is improved, practical applications require raw materials with higher resistance to heat, acid, and alkali. Kitao et al. used crude enzyme of sucrose phosphorylase from wild-type Leuconostoc mesenteroides to catalyze the synthesis of EGCG-1G and EGCG-2G, confirming that the aqueous stability of EGCG-2G is higher than that of EGCG-1G (Kitao et al. 1995), EGCG-2G demonstrates broader application potential. Although previous studies have achieved the synthesis of EGCG-2G, the proportion of EGCG-2G in the product remained relatively low (39.8%), and no efficacy investigations were conducted. Whether EGCG-2G possesses advantages in terms of biological activities such as antioxidant, anti-inflammatory, and melanin inhibition effects still lacked systematic research.

This study aims to synthesize EGCG-2G and investigate its biological activities. The catalytic conditions were systematically optimized to enhance the conversion rate of EGCG and the production proportion of EGCG-2G. EGCG-2G not only exhibits significantly improved water solubility and thermal stability compared to EGCG and its monoglycosylated derivatives, but also retains excellent antioxidant and anti-inflammatory activities. In terms of melanin inhibition, EGCG-2G effectively downregulated the expression of tyrosinase (TYR) and related proteins (TYRP1/2). Gene expression profiling indicated that this effect was accompanied by the downregulation of key components within the MC1R/cAMP/MITF signaling pathway. This research not only provides an efficient and controllable enzymatic strategy for the glycosylation modification of EGCG, but also highlights the potential of EGCG-2G as a novel skin-whitening agent, as validated through multidimensional efficacy studies.

Materials and methods

Materials

EGCG (Catalog number: PG-EG95DCL) was purchased from Chengdu Wagott Bio-tech Co., Ltd. (Chengdu, China). Cell culture reagents including high-glucose DMEM medium, RPMI-1640 medium, fetal bovine serum (FBS) and 0.25% trypsin were acquired from GibcoTM (Thermo Fisher Scientific, Waltham, MA, USA). The RNA extraction kit (Catalog number: G3013) was purchased from Wuhan Servicebio Technology (Wuhan, China). Reactive Oxygen Species (ROS, Catalog number: s0033s) Assay Kit was purchased from Beyotime Biotechnology (Shanghai, China). ELISA kits for detecting Interleukin-6 (IL-6, Catalog number: abs510003-96T), Interleukin-1 alpha (IL-1α, Catalog number: abs551001-96T), Tumor Necrosis Factor-alpha (TNF-α, Catalog number: abs510006-96T) and Prostaglandin E2 (PGE2, Catalog number: abs554161-96T) were purchased from Shanghai Universal Biotech Co., Ltd. (Shanghai, China). Unless otherwise specified, all other chemical reagents were purchased from Sinopharm Chemical Reagent (National Pharmaceutical Group Corporation, Shanghai, China). HaCaT cells were purchased from Immocell Biotechnology (Xiamen, China). B16F10 cells were obtained from BeNa Culture Collection (Beijing, China).

Heterologous expression and enzymatic characterization of sucrose phosphorylase

The gtfA gene (ID: QQB31661.1) encoding sucrose phosphorylase from L. mesenteroides was synthesized by Genewiz Biotech (Suzhou, China) according to codon preference of E. coli. The strain E. coli BL21(DE3) containing plasmid pET28a-gftA was utilized for gene gtfA expression. The cultivation of E. coli and the expression of recombinant sucrose phosphorylase are described in the Supporting Information. The enzyme activity was determined using the DNS method (Choi et al. 2011). The detailed experimental procedures for determining the optimal IPTG concentration, optimal temperature, optimal pH, thermal stability, pH stability, and organic solvent tolerance are provided in the Supporting Information.

Glycosylation of EGCG and optimization using response surface methodology

The glucosylation reaction was performed in 100 mL of 50 mM phosphate buffer (pH 6.5) containing 5 mM EGCG, 80 mM sucrose, and 10 U/mL sucrose phosphorylase. The reaction mixture was incubated at 25 °C with shaking at 220 rpm for 18 h, followed by enzyme inactivation through boiling for 8 min.

The transglycosylation temperature was first optimized. Considering the susceptibility of EGCG to oxidation at high temperatures (Peng et al. 2023), the reaction was conducted at 20, 25, and 30 °C with incubation at 220 rpm for 18 h. Subsequently, based on single-factor experimental results, a three-factor, three-level Box-Behnken design (BBD) was employed to optimize the reaction conditions, with the conversion rate of EGCG as the response value. The model investigated the interactive effects of EGCG concentration (X₁), sucrose concentration (X₂), and enzyme amount (X₃). The factor levels were coded in Table S1.

EGCG and its glycosylation products were determined by high-performance liquid chromatography (HPLC). The experimental data were subjected to regression and graphical analysis using Design-Expert 13 software. Analysis of variance (ANOVA) was performed to estimate statistical parameters. The significance of statistical and regression coefficients was verified through F-tests and P-values. The goodness-of-fit of the polynomial model equation was by determining both the R² coefficient and adjusted R² coefficient. Finally, evaluated the optimal parameter combinations yielding maximum response were identified to validate the model.

HPLC analysis of EGCG

The HPLC analysis was performed using a Morphling AC-C18 column (4.6 × 250 mm, 5 μm; Nanjing HeXi Biotechnology Co., Ltd.) maintained at 30 °C. The detection wavelength was set at 280 nm. The mobile phase consisted of methanol (A) and 0.15% (v/v) aqueous acetic acid solution (B) with the following gradient elution program:0–5 min: 90–85% B; 5–30 min: 85–80% B; 30–40 min: 80–60% B; 40–45 min: 60–50% B; 45–50 min: 50–10% B; 50–51 min: 10–90% B; 51–65 min: 90% B.

Isolation and purification of (-)-epigallocatechin gallate 4′,4″-O-α-D-diglucopyranoside (EGCG-2G)

The reaction products were initially separated using a Sephadex LH-20 column (25 mm × 400 mm). The column was first eluted with distilled water to remove sugars (sucrose, fructose, and glucose), followed by elution with 70% (v/v) ethanol.

The reaction products were analyzed by TLC on GF254 silica plates with ethyl acetate/acetic acid/water (3:1:1, v/v/v). Components were visualized by heating at 110 °C after spraying with anisaldehyde-sulfuric acid reagent. For thin-layer chromatography (TLC) analysis, control standards including sucrose, fructose, glucose, and unreacted EGCG were co-spotted alongside the reaction mixture on each silica plate. This allowed direct visual comparison of migration patterns and facilitated the identification of glycosylation products (EGCG-1G and EGCG-2G) based on differential Rf values relative to the substrate and sugar standards. Further purification was performed on a Morphling WD-C18 column (30 × 250 mm, 10 μm; Nanjing HeXi Biotechnology Co., Ltd.). The detection wavelength was set at 280 nm with a flow rate of 30 mL/min. The mobile phase consisted of methanol (A) and 0.25% (v/v) aqueous acetic acid solution (B), with the following gradient elution program: 0–15 min: 85% B; 15–20 min: 85–75% B; 20–60 min: 75% B.

Molecular weight determination and NMR structural characterization

To determine the molecular weight of EGCG-2G, an ultra-performance liquid chromatography coupled with quadrupole time-of-flight mass spectrometry (MALDI SYNAPT MS, Waters Corporation, USA) was employed. The mass spectrometer was operated in negative ion mode, and the mass spectra were acquired over the m/z range of 50-1500. The LC separation was as described above. The data were processed and analyzed using MassLynx software.

Nuclear magnetic resonance (NMR) spectroscopy was performed for structural characterization. Purified EGCG diglucoside (EGCG-2G, 10 mg) were dissolved in 500 µL of methanol-d4 and transferred into a 5 mm NMR tube. NMR spectra were acquired on an AVANCE III 400 system (Bruker, Germany) under the following conditions: ¹H NMR: 400 MHz frequency, 25 °C, 2 min acquisition time. ¹³C NMR: 100 MHz frequency, 25 °C, 5 h acquisition time. Heteronuclear multiple bond correlation (HMBC) experiments were conducted to establish the linkage between EGCG and glucose moieties.

Water solubility analysis and aqueous stability test of EGCG-2G

EGCG and EGCG-2G were dissolved in phosphate buffer solutions (pH 5 and 6) at a concentration of 1 mg/mL and incubated at 50 °C for 384 h (16 days). Samples were collected at regular intervals to monitor the hydrolysis process. In the stability assays, EGCG served as the primary control for EGCG-2G, and commercial EGCG-1G was included as an additional reference. All three compounds were tested under identical conditions (50 °C, pH 5.0 or 6.0), enabling direct comparison of their degradation profiles. The residual concentrations of EGCG and EGCG-2G were determined by HPLC under the aforementioned conditions. The stability data (%) were based on the relative HPLC peak area normalized to the initial sample.

An excess amount of each compound was dissolved in 100 µL of water in Eppendorf tubes and sonicated for 1 h. The solutions were then filtered through a 0.45 μm membrane, and the dissolved concentrations were determined by HPLC.

Determination of DPPH and ABTS radical scavenging activity

The 2,2-Diphenyl-1-picrylhydrazyl (DPPH) radical scavenging activity was measured according to a previously reported method with slight modifications (Villaño et al. 2007). A mixture of 100 µL of DPPH solution (50.0 µg/mL in anhydrous ethanol) and an equal volume of the sample solution was incubated at room temperature in the dark for 30 min. The absorbance was measured at 517 nm.

A₀ (absorbance of DPPH), A₁ (absorbance of samples).

The 2,2′-Azino-bis (3-ethylbenzothiazoline-6-sulfonic acid) (ABTS) radical scavenging activity was measured according to a previously reported method with slight modifications (Nenadis et al. 2004). The ABTS working solution was prepared by mixing 7 mmol/L ABTS solution with 2.45 mmol/L potassium persulfate solution at a 2:1 volume ratio, followed by 16 h reaction in the dark at room temperature. The reaction mixture containing 50 µL sample solution and 150 µL ABTS working solution was incubated in the dark at room temperature for 15 min, and the absorbance was measured at 734 nm using a microplate reader.

A₀ (absorbance of ABTS), A₁ (absorbance of samples).

The Trolox equivalent antioxidant capacity (TEAC) was determined for both assays. A standard curve was constructed using Trolox (6-hydroxy-2,5,7,8-tetramethylchromane-2-carboxylic acid) at a series of known concentrations. The TEAC value was then calculated using the following formula:

Cell culture and MTT cytotoxicity assay

The mouse melanoma B16F10 cells and human immortalized keratinocyte HaCaT cells were cultured in RPMI-1640 medium and DMEM medium, respectively, supplemented with 10% fetal bovine serum, 1 × 10⁵ U/L penicillin, and 100 mg/L streptomycin. To evaluate sample cytotoxicity (van Meerloo et al. 2011), logarithmically growing cells were seeded in 96-well plates at a density of 1 × 10⁵ cells/mL and incubated for 24 h at 37 °C in a 5% CO₂ incubator. After 24 h treatment with test samples, the supernatant was removed and replaced with 100 µL of 0.5 mg/mL MTT solution. Following 4 h incubation, the medium was aspirated and 100 µL dimethyl sulfoxide (DMSO) was added to dissolve the formazan crystals. Absorbance was measured at 490 nm.

For UVB toxicity evaluation, cells in logarithmic growth phase were seeded as described above. After 24 h incubation, cells were exposed to varying doses of UVB radiation followed by 24 h sample treatment. Cell viability was then determined using the aforementioned MTT assay protocol.

Cell viability was calculated using formula:

where A₀ and A₁ represent the average optical density (OD) values of the experimental and blank control groups at 490 nm, respectively.

Determination of ROS scavenging activity and fluorescence staining analysis

HaCaT cells were seeded in 12-well plates at a density of 2 × 10⁵ cells/mL and cultured for 24 h. After UVB irradiation, the cells were treated with samples for 24 h. Intracellular ROS levels were measured using a reactive oxygen species assay kit (Rastogi et al. 2010). Fluorescence images were acquired using an inverted fluorescence microscope (Olympus Corporation, Japan; excitation wavelength: 488 nm). The cells were then digested with trypsin, and the fluorescence intensity was measured with a microplate reader (excitation: 488 nm; emission: 525 nm). Three replicate wells were set up for each sample.

Detection of inflammatory factors

HaCaT cells were seeded in 12-well plates at a density of 2 × 10⁵ cells/mL. After 24 h of incubation, the cells were irradiated with UVB and then treated with test samples (EGCG and EGCG-2G) for 24 h. The concentrations of inflammatory cytokines (IL-6, IL-1α, TNF-α, and PGE2) in the supernatant were measured using an ELISA kit (Tan et al. 2022).

Melanin content assay

The melanin content in B16F10 cells was determined using a reported NaOH hydrolysis method (Petpiroon et al. 2022). B16F10 cells were seeded in 6-well plates at a density of 2 × 10⁵ cells/mL and cultured for 24 h. After an additional 24 h of drug treatment, the medium was discarded. The cells were lysed with 1% Triton X-100 at 4 °C for 1 h, followed by centrifugation at 12,000 rpm for 5 min to collect the pellet. After washing with PBS, 1 mL of 1 mol/L NaOH containing 10% DMSO was added. The mixture was heated in an 80 °C water bath for 1 h. Absorbance was finally measured at 405 nm. The melanin content was calculated and is presented as a percentage relative to the UVB-irradiated control group.

RT-PCR analysis of mRNA expression of related genes

Total RNA was extracted from treated cells using an RNA extraction kit. The purity and concentration of the isolated RNA samples were determined using a Nanodrop 2000 spectrophotometer (Thermo Fisher Scientific, USA). The extracted RNA was reverse-transcribed into cDNA, which was then used as a template (approximately 200 ng/µL) for amplification in a real-time PCR system (Bio-Rad, USA). The primer sequences and amplified gene information are detailed in Table S4. All experiments were performed with three technical replicates, and GAPDH was used as the reference gene for normalization.

The results were analyzed using the ΔΔCT method (Guo et al. 2023):

Molecular docking

The 2D structures of small molecule ligands were obtained from the PubChem database (http://pubchem.ncbi.nlm.nih.gov/). These 2D structures were then converted to 3D structures using ChemOffice software and saved as mol2 files. Protein targets with high-resolution crystal structures were selected from the RCSB PDB database (http://www.rcsb.org/) to serve as molecular docking receptors. The protein structures were processed using PyMOL software to remove water molecules and phosphate groups, and then saved as PDB files.

The protein and small molecule structures were prepared using AutoDock tools. Hydrogen atoms were added to the protein while water molecules were removed. For the small molecule ligands, hydrogen atoms were added and rotatable bonds were assigned. The docking grid box coordinates were then determined. Molecular docking was performed using AutoDock Vina 1.1.2 software to explore protein-ligand interactions. The optimal docking conformation was identified by comparing the docking scores. Finally, PyMOL and Discovery Studio 2019 software were used for visualization and analysis of the interaction patterns between the compounds and key amino acid residues.

In the molecular docking studies, the parent compound EGCG served as the reference ligand for comparison. Both EGCG and EGCG-2G were prepared and docked into the MC1R binding site using the identical computational protocol (including protein preparation, grid definition, and docking parameters in AutoDock Vina). The binding energy, predicted binding pose, and interaction patterns of EGCG-2G were systematically compared with those of EGCG to directly evaluate the effect of glucosylation on the ligand-receptor interaction.

Results

Expression and enzymatic properties of recombinant sucrose phosphorylase

The recombinant strain E. coli BL21(DE3) harboring the plasmid pET-28a-gftA was cultured to prepare crude enzyme extracts. As shown in Fig. S1, a distinct protein band at 56 kDa was observed in the cell lysate supernatant, confirming the successful expression of the recombinant sucrose phosphorylase. The enzyme was induced with varying concentrations of IPTG, and the highest enzymatic activity was achieved at 0.6 mM IPTG (Fig. 1a). The optimal reaction temperature and pH for the recombinant SPase were determined to be 40 °C and 6.5, respectively (Fig. 1b, c). The enzyme activity remained stable (maintaining > 95% activity after 6 h) at temperatures ≤ 35 °C. However, activity progressively decreased at higher temperatures, with a dramatic decline observed at 45 °C, where nearly all activity was lost within 3 h (Fig. 1d). The pH stability tests revealed that recombinant SPase maintained > 75% relative residual activity within pH 6.0–8.0, but activity decreased rapidly below pH 6.0 (Fig. 1e). Organic solvent stability assays (Fig. 1f) demonstrated that whole-cell suspensions in 10% methanol, ethanol, and DMSO solutions showed 57.19%, 58.85%, and 60.58% higher enzyme activity, respectively, compared to buffer controls after 18 h. Crude enzyme preparations exhibited more modest activity increases of 6.37%, 3.49%, and 10.68%.

Fig. 1.

Enzymatic properties of recombinant sucrose phosphorylase expressed in E. coli BL21(DE3). a Effect of IPTG concentration (0–1.6 mM) on enzyme activity. b Optimal reaction temperature (25–65 °C) and c optimal pH (5.0–8.0). d Thermal stability at 35, 40, and 45 °C and e pH stability (pH 5.0–8.0). f Residual activity of crude enzyme and whole-cell preparations after 18 h incubation in methanol (0–40%), ethanol (0–40%), and DMSO (0–50%)

Optimization of the transglycosylation reaction

As shown in Fig. S2, the proportion of EGCG-2G increased with temperature and reached 47.0% at 30 °C. Therefore, 30 °C was selected as the catalytic temperature to enhance the yield of EGCG-2G. A three-factor, three-level Box-Behnken design (BBD) was employed to systematically investigate the effects of EGCG concentration (X₁), sucrose concentration (X₂), and enzyme concentration (X₃) on the transglycosylation efficiency, with the conversion rate of EGCG as the response value. The experimental results for both crude enzyme and whole-cell catalytic systems were presented in Table 1, while the corresponding three-dimensional response surface plots were illustrated in Fig. 2a (crude enzyme system) and Fig. 2b (whole-cell system).

Table 1.

Experimental design and results of box-behnken design (BBD) for EGCG conversion to EGCG-Gs

	Coded level			Conversion (%)
run	X1	X2	X3	Crude enzyme	Whole-cell
1	1	600	30	96.325	96.100
2	5.5	450	30	93.522	92.264
3	5.5	600	20	92.158	93.068
4	5.5	600	40	89.935	90.941
5	1	450	20	95.195	95.260
6	5.5	450	30	93.383	92.141
7	1	300	30	88.836	90.557
8	10	300	30	80.831	86.889
9	10	450	20	86.641	89.300
10	5.5	300	40	87.594	89.957
11	1	450	40	94.175	94.646
12	5.5	300	20	78.218	84.350
13	5.5	450	30	94.551	92.084
14	5.5	450	30	92.623	92.269
15	10	450	40	90.598	92.900
16	10	600	30	91.837	91.868
17	5.5	450	30	93.315	92.548

Fig. 2.

Three-dimensional response surface plots depicting the interactive effects of EGCG concentration (X₁), sucrose concentration (X₂), and enzyme concentration (X₃) on the conversion rate of EGCG to glycosylated products (EGCG-Gs). EGCG concentration (1–10 g/L), sucrose concentration (300–600 g/L), enzyme concentration (20–40 U/mL). Reaction conditions: 30 °C, pH 6.5, 18 h. a Crude enzyme-catalyzed system; b whole-cell-catalyzed system. The models were generated using a Box-Behnken design (BBD). Experimental conditions are detailed in Table 1 and Table S1

For the crude enzyme system, the relationship between the variables and EGCG conversion to EGCG-Gs was described by the regression equation:

where Y represented the percentage conversion of EGCG to EGCG-Gs (%), X₁ denoted the EGCG concentration (g/L), X₂ indicated the sucrose concentration (g/L), and X₃ stood for the sucrose phosphorylase concentration (U/mL). Analysis of variance (ANOVA) demonstrated excellent model fit for the crude enzyme system, with a determination coefficient (R²) of 0.9878, adjusted R² of 0.9721, and a highly significant model F-value of 62.93 (p < 0.0001), as detailed in Table S2. The model predicted a maximum conversion efficiency of 96.44% under optimal conditions of 2.5 g/L EGCG, 500 g/L sucrose, and 30 U/mL enzyme, which was experimentally validated to be 97.46 ± 0.45%, showing excellent agreement with the predicted value. Under the optimized reaction conditions, EGCG-2G constituted 81.68% of the total glycosylated products, while the monoglucoside derivative EGCG-1G accounted for the remaining 18.32%. Similarly, for the whole-cell catalytic system, the regression equation was established as:

where Y, X₁, X₂, and X₃ represented the same variables as previously defined.

The whole-cell system exhibited even better model performance, with an R² of 0.9981, adjusted R² of 0.9958, and a remarkably significant F-value of 418.33 (p < 0.0001), as shown in Table S3. The model predicted maximum conversion of 96.15% at optimal conditions of 2.5 g/L EGCG, 520 g/L sucrose, and 27 U/mL enzyme, which was experimentally confirmed to be 94.14 ± 0.38%, demonstrating good correlation with model predictions.

Purification and characterization of EGCG-2G

The transglucosylation of EGCG catalyzed by sucrose phosphorylase was confirmed through TLC (using anisaldehyde-sulfuric acid staining, Fig. 3a), HPLC analysis (Fig. 3b) and mass spectrometry. The major product exhibited a molecular ion peak at m/z 782.1971 [M-H]⁻ (Fig. 3c), corresponding to EGCG with two glucosyl moieties (EGCG-2G). Following purification by HPLC, the major EGCG-Gs fraction was obtained as EGCG-2G with a purity of ≥ 99%, the structure was unequivocally determined by comprehensive NMR analysis (¹H, ¹³C, and HMBC). NMR Characterization: ¹H NMR (400 MHz, MeOD) δ 6.92 (s, 2 H), 6.56 (s, 2 H), 6.01–5.93 (m, 2 H), 5.59–5.51 (m, 1H), 5.13 (d, J = 3.7 Hz, 1H), 5.01 (s, 1H), 4.99 (d, J = 3.8 Hz, 1H), 4.17 (dddd, J = 9.8, 7.3, 5.0, 2.4 Hz, 2 H), 3.87–3.78 (m, 4 H), 3.72 (ddd, J = 11.8, 9.1, 5.1 Hz, 2 H), 3.55 (ddd, J = 9.5, 3.9, 2.7 Hz, 2 H), 3.46–3.36 (m, 2 H), 3.00 (dd, J = 17.4, 4.6 Hz, 1H), 2.86 (dd, J = 17.5, 2.5 Hz, 1H) (Fig. S3). ¹³C NMR (100 MHz, MeOD) δ 166.9, 158.0, 157.9, 157.0, 151.9, 151.5, 139.4, 137.1, 135.1, 127.8, 110.3, 107.0, 105.6, 104.9, 99.3, 96.7, 95.9, 78.3, 75.2, 75.1, 75.0, 75.0, 73.5, 73.3, 71.1, 70.9, 70.4, 62.3, 62.2, 26.8 (Fig. S4).

Fig. 3.

Analysis of EGCG glycosylation products catalyzed by recombinant sucrose phosphorylase. a TLC plate stained with anisaldehyde-sulfuric acid. Lane assignments: 1, sucrose; 2, fructose; 3, glucose; 4, EGCG; 5, crude reaction mixture; 6, EGCG-Gs after Sephadex LH-20 purification; 7, purified EGCG-2G. b HPLC chromatograms of EGCG, the crude reaction mixture, and purified EGCG-2G (detection at 280 nm). c Negative-ion ESI mass spectrum of EGCG-2G

Comparative analysis of the NMR spectra between the starting material EGCG and the product revealed characteristic signals corresponding to two additional glucosyl moieties in the product: δH 5.13 (d, J = 3.7 Hz, 1H), 4.99 (d, J = 3.8 Hz, 1H), 4.17 (m, 2 H), 3.87–3.78 (m, 4 H), 3.72 (m, 2 H), 3.55 (m, 2 H), 3.46–3.36 (m, 2 H) in the ¹H NMR spectrum, and δC 105.6, 104.9, 75.2, 75.1, 75.0, 75.0, 73.5, 73.3, 71.1, 70.9, 62.3, 62.2 in the ¹³C NMR spectrum. The signals at δH 5.13 (d, J = 3.7 Hz, 1H), 4.99 (d, J = 3.8 Hz, 1H) and δC 105.6, 104.9 were identified as the anomeric protons and carbons of the two glucosyl units, respectively.

In the 2D NMR spectrum, signals at δH 5.54 (m, 1H), 5.01 (s, 1H), and 3.00 (dd, J = 17.4, 4.6 Hz, 1H) / 2.86 (dd, J = 17.5, 2.5 Hz, 1H) were identified as belonging to a single spin system based on the ¹H-¹H COSY spectrum (Fig. S5). According to their chemical shifts, these protons were assigned as H-3, H-2, and H-4, respectively. Additionally, signals at δH 5.13 (d, J = 3.7 Hz, 1H) and 4.99 (d, J = 3.8 Hz, 1H) showed correlations with H-2′′′ and H-2′′′′ in the COSY spectrum, leading to their assignment as H-1′′′ and H-1′′′′, respectively. The coupling constants indicated that both glucose units adopt the α-configuration.

As shown in Fig. 4a, the 2D NMR (HMBC) spectrum demonstrated key correlations: the anomeric proton at δH 5.13 (d, J = 3.7 Hz, 1H) showed a cross-peak with the oxygen-bearing quaternary carbon (δC 139.4) of the benzoyl group, confirming the attachment of one glucosyl unit at the 4″-position. The second anomeric proton at δH 4.99 (d, J = 3.8 Hz, 1H) correlated with the oxygen-bearing quaternary carbon (δC 135.1) of the flavonoid B-ring, establishing the glycosylation at the 4′-position. These results conclusively determined the structure of the compound (Fig. 4c) and its key HMBC and ¹H-¹H COSY correlations (Fig. 4b), providing definitive evidence that the sucrose phosphorylase-catalyzed transglycosylation successfully introduced two glucosyl moieties at the 4′- and 4″-positions, yielding the EGCG-2G product. The structure of the purified major product was unequivocally identified as (-)-epigallocatechin gallate 4′,4″-O-α-D-diglucopyranoside (EGCG-2G) [systematic IUPAC name: (2R,3R)-2-(3,4,5-trihydroxy-4′-O-α-D-glucopyranosylphenyl)-3,4-dihydroxy-2 H-1-benzopyran-3,5,7-triol-3-yl 3,4,5-trihydroxy-4″-O-α-D-glucopyranosylbenzoate] by NMR and mass spectrometry. The full HMBC spectrum is provided in Fig. S6.

Fig. 4.

Structural characterization of purified EGCG-2G. a Key region of the HMBC spectrum highlighting correlations between anomeric protons (δH 5.13 and 4.99) and the quaternary carbons of the galloyl (δC 139.4) and B-ring (δC 135.1). b Key HMBC and ¹H-¹H COSY correlations. c Chemical structure of EGCG-2G

Analysis of water solubility and aqueous solution stability of EGCG-2G

The water solubility of three EGCG glycosylation products was 49, 55, and 114 times higher than that of EGCG after glucosylation (Moon et al. 2006). In this study, the solubility of EGCG in aqueous solution was 5.1 mM (Kim et al. 2016), while that of EGCG-2G reached 633.7 mM, approximately 124.25 times higher than EGCG, demonstrating a significant improvement in solubility. The stability of EGCG and its glycosylated derivative EGCG-2G was investigated under elevated temperature (50 °C) and different pH conditions. The results demonstrated that EGCG-2G exhibited significantly superior stability in aqueous solution compared to EGCG. In phosphate buffer solution at pH 5, EGCG-2G maintained 91.87% of its original content after 384 h (16 days) at 50 °C, whereas only 7.05% of EGCG remained, representing an 84.82% improvement in stability (Fig. 5a). Similarly, in buffer solution at pH 6, EGCG-2G showed remarkable stability with 75.01% retention, in contrast to merely 0.23% for EGCG, corresponding to a 74.78% enhancement in stability (Fig. 5b). Furthermore, when commercially available EGCG-1G and EGCG-2G were subjected to 50 °C for 340 h, EGCG-2G demonstrated 35.36% greater stability than EGCG-1G (Fig. 5c).

Fig. 5.

Aqueous stability of EGCG and EGCG-2G under thermal and acidic conditions. Relative residual concentrations (%) of a EGCG and EGCG-2G in phosphate buffer (pH 5.0) at 50 °C, b EGCG and EGCG-2G in phosphate buffer (pH 6.0) at 50 °C, and c commercial EGCG-1G and synthesized EGCG-2G in phosphate buffer (pH 5.0) at 50 °C over 16 days (384 h)

Antioxidant activity analysis of EGCG-2G

The antioxidant activities of EGCG and its glycosylated derivative EGCG-2G were systematically evaluated through DPPH and ABTS radical scavenging assays, along with intracellular ROS inhibition experiments. As shown in Fig. 6a, both compounds demonstrated concentration-dependent DPPH radical scavenging capacity, with EGCG exhibiting stronger activity (IC50 = 20.90 ± 6.73 µM) compared to EGCG-2G (IC50 = 37.39 ± 8.53 µM). In the DPPH assay, with a Trolox standard curve provided in Figure S7, EGCG and EGCG-2G showed TEAC values of 2.04 and 1.14, respectively. In the ABTS assay (Fig. 6b), EGCG-2G demonstrated slightly superior activity at lower concentrations, as reflected by its IC50 of 33.51 ± 8.14 µM compared to 43.93 ± 7.74 µM for EGCG. While EGCG displayed enhanced scavenging at elevated concentrations, the overall difference remained small, narrowing the gap in their efficacy. In the ABTS assay, with a Trolox standard curve provided in Fig. S8, EGCG-2G and EGCG showed TEAC values of 8.16 and 6.22, respectively.

Fig. 6.

Comparative evaluation of antioxidant efficacy between EGCG and EGCG-2G. a DPPH radical scavenging activity. b ABTS radical scavenging activity. c Inhibition of intracellular ROS generation in UVB-irradiated (30 mJ/cm²) HaCaT keratinocytes (*p < 0.05, **p < 0.01, ***p < 0.001)

In the UVB-induced oxidative stress model using HaCaT cells (30 mJ/cm²), EGCG demonstrated more potent ROS inhibition (47.88 ± 4.13% ROS level) than EGCG-2G (59.87 ± 2.14%). However, EGCG-2G still maintained significant antioxidant activity, markedly reducing ROS levels compared to the UVB-only control group (p < 0.001, Fig. 6c).

Effects of EGCG-2G on UVB-induced inflammatory factors inHaCaTCells

To assess the biosafety of EGCG-2G, its cytotoxic effects on HaCaT cells were evaluated using the MTT assay (Fig. 7a). Experimental data revealed no significant cytotoxicity within the concentration range of 20–80 µM, and this safe dosage range was therefore selected for subsequent inflammatory factor analysis. To establish a UVB-induced HaCaT keratinocyte injury model (Fig. 7b), the MTT assay results demonstrated that a UVB irradiation dose of 30 mJ/cm² maintained cell viability at 83.65 ± 1.8% (p < 0.001 compared to the control group). This dose was determined as the optimal irradiation condition for subsequent experiments, as it effectively induced oxidative damage while ensuring cell survival.

Fig. 7.

Anti-inflammatory effects of EGCG-2G in UVB-irradiated HaCaT keratinocytes. a Cytotoxicity of EGCG-2G determined by MTT assay. b Determination of optimal UVB irradiation dose (30 mJ/cm²). Secretion levels of inflammatory cytokines in the supernatant after UVB irradiation and EGCG-2G treatment: c IL-6, d IL-1α, e TNF-α, and f PGE2 (*p < 0.05, **p < 0.01, ***p < 0.001)

The anti-inflammatory efficacy of EGCG-2G in UVB-irradiated HaCaT cells was evaluated by measuring the secretion levels of inflammatory factors IL-6, IL-1α, TNF-α, and PGE2. These inflammatory cytokines may bind to their respective receptors, thereby activating or inhibiting signaling pathways associated with melanogenesis. This regulation could either promote or suppress the expression of genes involved in melanin formation, ultimately modulating the process of skin pigmentation (Fu et al. 2020). After establishing the UVB-induced (30 mJ/cm²) HaCaT keratinocyte injury model, increasing concentrations of EGCG-2G (20–80 µM) exhibited a dose-dependent enhancement in the suppression of IL-6, IL-1α, TNF-α, and PGE2 expression. At a concentration of 80 µM, EGCG-2G significantly inhibited UVB-induced inflammatory responses, reducing the secretion levels of IL-6, IL-1α, TNF-α, and PGE2 by 32.5% (p < 0.001), 39.2% (p < 0.001), 34.0% (p < 0.001), and 31.9% (p < 0.001), respectively (Fig. 7c–f). These results confirmed that EGCG-2G, generated through glycosylation modification of EGCG, retained significant anti-inflammatory activity, validating its potential as a novel anti-inflammatory agent for nutraceutical and cosmetic applications.

Effects of EGCG-2G on melanogenesis in B16F10Cells

To establish a UVB-induced B16F10 cell injury model (Fig. S9), MTT assay results demonstrated that a UVB irradiation dose of 30 mJ/cm² maintained cell viability at 86.01 ± 1.36% (p < 0.001 compared to the control group). This dose was determined as the optimal irradiation condition for subsequent experiments, as it effectively induced oxidative damage while ensuring cell survival. To assess the biosafety of EGCG-2G, its cytotoxic effects on B16F10 cells were evaluated using the MTT assay (Fig. S10). Experimental data revealed no significant cytotoxicity within the concentration range of 20–80 µM, and this safe dosage range was therefore selected for subsequent melanin content analysis.

After establishing the UVB-induced (30 mJ/cm²) melanogenesis model in B16F10 cells, the inhibitory effect of EGCG-2G on UVB-stimulated melanin production was evaluated. Within the concentration range of 20–80 µM, EGCG-2G exhibited a dose-dependent inhibition of melanin synthesis. At the highest concentration tested (80 µM), EGCG-2G significantly reduced melanin content by 43.79% (p < 0.001) (Fig. 8a), confirming its potent anti-melanogenic activity.

Fig. 8.

Inhibitory effects of EGCG-2G on melanogenesis in UVB-stimulated B16F10 melanoma cells. a Relative melanin content after UVB (30 mJ/cm²) and 24 h sample treatment (20–80 µM). b mRNA expression of melanogenesis-related genes after same treatment. c Molecular docking of EGCG-2G with MC1R

Upon ultraviolet (UV) irradiation, keratinocytes synthesize and secrete α-melanocyte-stimulating hormone (α-MSH), which binds to the melanocortin 1 receptor (MC1R) on melanocytes. This ligand-receptor interaction activates MC1R, triggering an increase in intracellular cyclic AMP (cAMP) concentration. The elevated cAMP level enhances the expression of microphthalmia-associated transcription factor (MITF). MITF in turn transcriptionally upregulates the expression of key melanogenic enzymes such as tyrosinase (TYR), tyrosinase-related protein 1 (TYRP1), and tyrosinase-related protein 2 (TYRP2), thereby promoting melanin biosynthesis (Herraiz et al. 2021; Mahrous et al. 2025). To explore the potential molecular mechanism underlying the inhibitory effects of EGCG-2G on melanogenesis, transcriptional changes in key regulators (MC1R, MITF, TYR, TYRP1/2) of the MC1R/cAMP/MITF signaling pathway were systematically evaluated. RT-qPCR analysis (Fig. 8b) demonstrated that, compared to the blank control group, EGCG-2G treatment significantly downregulated the expression levels of melanogenesis-related marker genes (p < 0.001), with MC1R, MITF, TYR, TYRP1, and TYRP2 mRNA expression showing dose-dependent suppression. Notably, under 80 µM EGCG-2G treatment, the inhibition rates of these genes reached 77.07%, 57.46%, 69.48%, 68.01%, and 72.75%, respectively. These findings provide important molecular-level evidence supporting the potential application of glycosylated EGCG derivatives in skin-whitening formulations.

Molecular docking analysis of EGCG-2G with MC1R

MC1R is a seven-transmembrane G protein-coupled receptor on melanocytes, activated by α-MSH (Yardman-Frank et al. 2021). Due to its higher expression in malignant melanoma compared to normal organs and tissues, it is widely used as a biomarker for melanoma and is regarded as a potential target for selectively delivering therapeutic agents against melanoma (Li et al. 2019). The ubiquitination and degradation of MC1R mediated by the membrane-anchored E3 ligase MGRN1, via the adaptor protein Attractin, reduces receptor surface levels and attenuates downstream eumelanin synthesis (Parashara et al. 2025). Molecular docking was performed to predict the binding interactions of both EGCG and EGCG-2G with the MC1R protein. As shown in Fig. 8c, EGCG-2G formed hydrogen bonds with MC1R residues TYR37, LYS57, ARG42, LEU55, MET386, and ASP240, while establishing van der Waals interactions with residues ASN313, GLY53, GLN31, ARG52, PHE335, and ARG67. In comparison, EGCG exhibited hydrogen bonding with residues GLU268 and ARG265, along with van der Waals forces involving residues LYS53, GLY47, ALA48, ASP272, SER112, THR111, and LEU45. The binding energy of EGCG-2G to MC1R was − 9.9 kcal/mol, and that of EGCG was − 8.9 kcal/mol. These in silico results predicted that EGCG-2G could bind to the MC1R protein with high affinity, providing a supportive molecular hypothesis that the observed anti-melanogenic effect of EGCG-2G might involve interaction with MC1R.

Discussion

This study successfully synthesized diglucosylated EGCG (EGCG-2G) through dual glycosylation modification catalyzed by sucrose phosphorylase, achieving conversion rates of 97.46 ± 0.45% (using crude enzyme extract) and 94.14 ± 0.38% (using whole cells). Structural characterization confirmed a purity exceeding 99%. The water solubility of EGCG-2G was approximately 124 times higher than that of EGCG. EGCG-2G demonstrated significantly superior stability compared to both EGCG and commercial EGCG-1G, retaining 91.87% of its activity after 16 days of treatment at 50 °C and pH 5, thereby providing a stable active ingredient for cosmetic applications.

Efficacy studies revealed that EGCG-2G not only retained the antioxidant (scavenging DPPH/ABTS/ROS free radicals) and anti-inflammatory activities (inhibiting inflammatory factors such as IL-6 and TNF-α) of the parent compound but also exhibited anti-melanogenic effects. These effects were accompanied by the downregulation of key genes within the MC1R/cAMP/MITF signaling pathway, including MC1R, MITF, TYR, TYRP1, and TYRP2, thereby suggesting a correlation with the inhibition of melanin production.

Firstly, we investigated the organic solvent stability of whole cells and crude enzyme solutions. It was found that when suspended in 10% organic solvent, the enzymatic activity of both the crude enzyme solution and whole cell solution increased. The enhanced activity in whole-cell suspensions might be attributed to improved membrane permeability and reduced peptidoglycan cross-linking by low-concentration organic solvents, facilitating substrate access (Siziya et al. 2021). At higher concentrations, the toxicity of organic solvents reduced cell viability, which led to a decline in enzyme activity with increasing solvent concentration. However, the cell membrane in the whole-cell system provided a protective effect. Consequently, higher residual enzyme activity was maintained in the whole-cell system compared to the crude enzyme extract (León et al. 1998).

The stability of EGCG-2G was found to be significantly superior to that of both EGCG and EGCG-1G. EGCG-2G retained a high retention rate of ≥ 75% after 16 days of treatment at elevated temperature, demonstrating that glycosylation significantly improved the stability of EGCG against degradation. The improved stability of EGCG-2G could be attributed to the protective effect of the additional glucosyl moiety, which shielded the reactive phenolic hydroxyl groups from degradation while maintained the compound’s bioactive properties. The significantly enhanced aqueous stability of EGCG-2G effectively reduced degradation during product processing and storage, improved bioavailability, and extended its durability in food and cosmetic formulations.

The observed difference in antioxidant capacity between EGCG and EGCG-2G is likely due to the steric hindrance from the glucosyl group, which may partially hinder the radical-scavenging ability of the phenolic hydroxyl groups (Xiao et al. 2021). Importantly, despite this moderate reduction in activity, EGCG-2G retained considerable free radical scavenging capacity, confirming that the glycosylation modification did not fundamentally disrupt its antioxidant structure. These findings indicated that while glycosylation affected the antioxidant properties to some degree, EGCG-2G maintained sufficient biological activity.

EGCG-2G effectively suppressed the secretion of key pro-inflammatory mediators (IL-6, IL-1α, TNF-α, and PGE2) in a dose-dependent manner, with the highest inhibition observed at 80 µM. These results indicate that the glycosylation modification of EGCG does not compromise its inherent anti-inflammatory properties. Instead, EGCG-2G retains the ability to mitigate UVB-induced inflammatory responses, likely through modulating cytokine release linked to melanogenesis-related signaling pathways. Based on the observed broad-spectrum inhibition of UVB-induced inflammatory mediators (IL-6, IL-1α, TNF-α, and PGE2) by EGCG-2G in this study, we propose that its anti-inflammatory mechanism may be similar to that of its parent compound EGCG. EGCG has been demonstrated to suppress NF-κB p65 activation and nuclear translocation, thereby downregulating iNOS mRNA expression and subsequently reducing NO production (Song et al. 2006). This mechanism provides significant protection against UVB-induced damage. Given that EGCG-2G exhibited comparable anti-inflammatory effects to EGCG in our experimental system, we hypothesize that the glycosylated derivative may function through a similar pathway, potentially modulating the NF-κB signaling cascade to alleviate UVB-induced inflammatory responses. The data support the conclusion that EGCG-2G is a promising candidate for further development as an anti-inflammatory agent in cosmetic and nutraceutical products aimed at alleviating UV-induced skin inflammation.

According to molecular docking energy evaluation criteria, binding energies below − 5.0 kcal/mol indicated significant ligand-receptor binding activity, while values below − 7.0 kcal/mol suggested strong binding capability, with lower energies corresponding to higher affinity and more stable complex conformations. Our study revealed that EGCG-2G demonstrated a binding energy of -9.9 kcal/mol with MC1R, superior to EGCG’s -8.9 kcal/mol, indicating that glycosylation modification enhanced the interaction strength between EGCG and MC1R. These results demonstrated that EGCG-2G could stably bind to the MC1R protein with high affinity, providing molecular-level evidence that EGCG-2G might inhibit melanin synthesis through modulation of MC1R protein activity.

In summary, EGCG-2G achieved synergistic optimization of water solubility and stability through glycosylation modification, and its biological activities were thoroughly investigated. The anti-melanogenic activity was associated with downregulation of the MC1R/cAMP/MITF pathway genes, and molecular docking provided a plausible model for its potential interaction with MC1R. This study provides a scientific basis for developing stable and highly functional bioactive ingredients with significant potential for applications in both the food and cosmetic industries.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

BY, NZ and JW conceived and designed the experiments. BY, YH, JD and MY carried out the experiments. XY prepared figures. BY, YH, JD, MY and JW analyzed the data. BY and NZ wrote the paper. All authors approved the final manuscript.

Funding

This work was funded by the National Natural Science Foundation of China (52472098), and the Jiangsu Province Industry-University-Research Collaboration Project (BY20240724).

Data availability

No datasets were generated or analysed during the current study.

Declarations

Competing interests

The authors declare no competing interests.