Enhancement of silymarin solubility and bioactivities using betaine/ascorbyl glucoside DES

Abstract

Betaine/ascorbyl glucoside (Bet/AA-2G) deep eutectic solvent (DES) composed of betaine and ascorbyl glucoside have been successfully used to improve the solubility of silymarin in this study. Under appropriate conditions, the solubility of silymarin in a 5 wt% Bet/AA-2G DES solution increased by 17.7% and 60.9% compared to a 50 wt% 1,3-butanediol solution and a 50 wt% ethanol solution, respectively. In addition, quantum chemical simulations showed that silymarin had a strong binding capacity to Bet/AA-2G DES and can remain stable after binding. The combination of Bet/AA-2G DES with silymarin enhanced the skin permeability and antioxidant capacity of silymarin. Moreover, Silymarin extract extracted from a mixture of DES (50 wt%) (SM-DES) promoted the synthesis of type I collagen and significantly reduced the expression of MMP-1, indicating potential anti-wrinkle and firming properties. Overall, Bet/AA-2G DES showed great potential for improving the solubility and stabilizing the bioactivity of silymarin.

---

Introduction

With the increasing emphasis on a healthy environment and quality of life, medicinal plants have become the focus of increasing research in the field of cosmetics in recent years.^1,2 Silybum marianum (Milk Thistle, Asteraceae) is an important medicinal plant that has received considerable attention for its potential protective effects on the liver against damage caused by toxins, pollutants, and certain drugs.^3,4 Silymarin is a mixture of flavonolignans collectively from Silybum marianum,^5,6 which has various effects such as antioxidant, antihypertensive, anti-obesity, anti-diabetic, anti-inflammatory, and anti-carcinogenic properties.^7,8 Moreover, silymarin has great potential for application in the medical, food, and cosmetic fields. Unfortunately, the application of silymarin is limited due to its poor aqueous solubility, chemical instability, and low bioavailability.^9,10 To overcome these problems, various strategies, such as the encapsulation of silymarin in liposomes, micelles, polymeric nanoparticles, and solid lipid nanoparticles, have been used in previous studies.^9,11 For example, Lee¹² prepared silymarin liposome nanoparticles using water-soluble chitosan as a carrier and poly(r-glutamic acid) as a co-agent, and the solubility of nanocapsulated silymarin liposome particles was found to be 7.7-times higher than that of unencapsulated silymarin. Sec@PA nanoparticles, loaded into the multifunctional microneedle, were synthesized to improve the delivery efficiency of selenium and chlorin e6 on the chronic diabetic wound.¹³ OXA-DA nanoparticles were developed to improve oxaliplatin biosafety and drug efficacy in cancer therapy.¹⁴ However, these strategies are complex, costly and require further research to ensure their safety and efficacy. Therefore, it is crucial to develop a new method for improving the solubility and bioavailability of silymarin.

Deep eutectic solvents (DESs) are a class of developing solvents that have gained significant attention in recent years because of their unique properties (e.g., ecological friendliness and low toxicity) and potential applications in various fields.^15,16 In addition, DESs can improve the solubility of natural or synthetic chemicals with low water solubility.^17,18 Unlike traditional solvents, which are typically composed of a single molecular species, DESs are formed by a combination of two or more components, usually a hydrogen bond donor (HBD) and a hydrogen bond acceptor (HBA).¹⁹ Moreover, they display many advantages, including low cost, low toxicity, excellent biocompatibility, and environmental friendliness.¹⁶ In recent years, DESs have been used to extract polar and non-polar compounds such as coumarins,²⁰ triterpene saponins,²¹ anthocyanins,²² phenylethanes and phenylpropanoids,²³ polysaccharides²⁴ and phlorotannins.²⁵ Similarly, it is believed that silymarin extraction with DESs also has great application potential.

Yao found that the yield of silibinin extracted by ChCl/1,4-butanediol DES was significantly higher than that extracted with betaine/1,4-butanediol DES.²⁶ However, ChCl is limited in some applications such as food and cosmetics. Therefore, it is important to find a green and efficient DES for the extraction of silymarin. Ascorbyl glucoside (AA-2G) is a glycosylated form of L-ascorbic acid (L-AA) known for its antioxidant properties and promotion of collagen synthesis,^26,27 and can be used as an HBD. Betaine (Bet) is a naturally occurring compound with an asymmetric structure and polar functional groups that has been widely used as an HBA in the preparation of DES as a solubilizer and stabilizing agent for drugs and enzymes.^28,29 In this study, Bet/AA-2G DES was successfully synthesized for the first time and applied as the silymarin extraction solvent. The mechanism of Bet/AA-2G DES formation and interactions between the DES and silymarin were investigated by quantum chemistry calculations. In addition, the biological activities of silymarin were further studied.

Methods and materials

Materials

The silybum marianum seeds were purchased from Liaoning Fengruitiancheng Biotechnology Co., Ltd. (China), identified by Jiannan Chen (Guangzhou University of Chinese Medicine), and deposited as voucher specimens at Shenzhen Shinehigh Innovation Technology Co., Ltd R&D center.

Betaine (AR, ≥98%), ascorbyl glucoside (GR, ≥99%) and ethanol (AR, ≥99.7%) were purchased from Shanghai Adamas Reagent Co., Ltd. (China). 2,2-Diphenyl-1-picrylhydrazyl (≥97%, DPPH) was purchased from Cool Biological Engineering Co., Ltd. (China). Iron chloride hexahydrate (AR, ≥98%), potassium persulfate (AR, ≥99.5%) and L-ascorbic acid (≥99%) were purchased from Shanghai Macklin Biochemical Co., Ltd. (China). Tocopherol (AR, ≥96%) was purchased from Guangzhou Tianjun Biological Technology Co., Ltd. (China). Sodium hydroxide (AR, ≥97%), methanol (GR, ≥99.9%), potassium phosphate monobasic (AR, ≥99%), formate (AR, ≥99%), silibinin (AR, ≥98%) and isosilybin (AR, ≥98%) were purchased from Shanghai Aladdin Biochemical Technology Co., Ltd. (China). ABTS (HPLC, ≥98%) was purchased from Sigma-Aldrich Co., Ltd. (USA), and trichloroacetic acid (AR, ≥99%) was purchased from Meryer (Shanghai) Chemical Technology Co., Ltd. (China).

Preparation of DES

DES were prepared according to a previously described method,³⁰ with several modifications. Betaine and AA-2G were added to the reactor at a molar ratio of 3 : 1, an appropriate proportion of water was added, and the mixture was stirred magnetically at 80 °C for 4 h. After that, the solution was dried using a vacuum oven at 60 °C and <5 Pa for 12 h, and the residual liquid after drying was the DES. The final residual water content of the DES was 5%, which was tested by a moisture analyzer (MB27, Ohaus Instruments (shanghai) Co., Ltd, China).

Characterization of DES

The functional groups of DES were determined by Fourier transform infrared spectroscopy (FT-IR, Shimadzu, IRSpirit, Japan) at wavenumbers from 4000 to 500 cm⁻¹.³¹ The degradation temperatures of the DES were determined by differential scanning calorimetry (DSC, T.A. instruments, Q100, USA) heated at a constant heating rate of 10 °C min⁻¹ from −40 to 20 °C under a nitrogen atmosphere (flow rate of 50 mL min⁻¹).³² The structure of DES was characterized by X-ray diffraction analyzer (XRD, Ultima IV, Japan) in a 2θ range of 5°–80° with Cu Kα radiation. ¹H-nuclear magnetic resonance (¹H-NMR) spectra was preformed with a Bruker 400M, and the samples were dissolved in D₂O in advance.

Extraction of silymarin

The extraction method for silymarin was based on a previously described method,³³ and the production of lyophilized silymarin powder can be divided into six steps: pretreatment, extraction, concentration, purification, lyophilization, and recrystallization. First, silybum marianum seeds were freeze-dried to remove water, crushed using a high-speed multifunction mill, and sieved through 60 mesh sieves. Then, the oil in the silybum marianum seeds was removed by supercritical carbon dioxide (SC-CO₂, Gaoke pharmaceutical, GK-130-45-2L, China) extraction with a CO₂ flow rate of 30 L h⁻¹, a pressure of 35 MPa, a temperature of 45 °C and an extraction time of 2.5 h. Then, the treated silybum marianum seeds were extracted with an 80% (v/v) alcoholic solution with a liquid-to-solid ratio of 1 : 10 for 2 h at 60 °C. The collected silybum marianum extract was then centrifuged at 8000 rpm for 15 min at a high speed. The supernatant of the extract solution was filtered to remove impurities, in which the filtration membrane was a 0.45 μm organic phase filtration membrane. Subsequently, a concentrated silybum marianum extract was obtained by spin evaporation and purified by adsorption using an ADS-17 resin column, and the silymarin lyophilized powder was then prepared by freeze-drying this purified solution. Finally, the samples were recrystallized using a methanol solution. The sample was dissolved in a hot methanol solution and concentrated by rotary evaporation until the crystals were precipitated, and silymarin was collected by vacuum filtration. The yield of silymarin collected was about 2.1% with a purity of 80%.

Detection of silymarin content

The reference values of silymarin were measured using high-performance liquid chromatography (HPLC) according to the Pharmacopoeia of the People's Republic of China. The content of silymarin, including silybin and isosilybin, was measured by HPLC (Agilent 1200, USA) equipped with an Agilent AQ-C18 column (250 mm × 4.6 mm, 5 μm) and an ultraviolet (UV) absorption detector. Methanol-water containing 0.1% formic acid (48 : 52, v/v) was used as the mobile phase at a flow rate of 1.0 mL min⁻¹, and the column temperature was 30 °C. The injection volume was 10 μL and UV detection (287 nm) was used. The silymarin content was calculated by the external standard method and the limit of detection was 0.01 mg mL⁻¹.

Solubility test

The solubility analysis of silymarin in DES was performed according to the method described by Dong et al.³⁴ The solubility of silymarin in water, ethanol (50 wt%), 1,3-butanediol (50 wt%), DES (50 wt%), and a mixture of DES (50 wt%) and 1,3-butanediol (50 wt%) was investigated by determining the amounts of silybum marianum and isosilymarin in the supernatant.

Silymarin extract extracted from a mixture of DES (50 wt%) (SM-DES), 1,3-butanediol solution (50 wt%), and silymarin extract (SM-control) in a 1,3-butanediol solution (50 wt%) were used in subsequent experiments. The solubility of silymarin was calculated using the following formula:

where C_R, A_X, A_R, and D are the concentration of the standard (mg mL⁻¹), peak area of the test sample, peak area of the standard and dilution of the test sample, respectively.

Computational details

The Gaussian 16 program was used for all the calculations. Density functional theory (DFT) calculations were performed using the B3LYP functional and Grimme's D3BJ³⁵ version for dispersion correction. The 6-31G(d,p) basis set was used for geometry optimization and frequency calculations.³⁶ The geometries were thoroughly optimized without structural constraints. The ultimate energies of the optimized structures were estimated using the larger 6-311+G(d,p) basis set. VMD software was used to display the electrostatic potential (ESP) surfaces and interaction region indicator (IRI) statistics obtained from the Multiwfn 3.8 application. The binding energy (E_b) was calculated using the following equation:

E_b = E_Complex − (E_M1 + E_M2) (2)

where E_Complex, E_M1, and E_M2 represent the energies of the complex and interacting molecules, respectively. The combined conformational search was performed using xtb and Molclus software.

Skin penetration

The skin permeability of the samples was tested on porcine skin using the Franz diffusion cell method.³⁷ Briefly, an appropriate amount of saline was added to the receiving cell, and an appropriately sized porcine skin sample was fixed in the center of the supply and receiving cells of the Franz diffusion cell. Subsequently, 1 mL of the sample to be tested was added to the supply cell (skin surface) and the water bath was maintained at a constant temperature of 32 ± 1 °C with stirring at 350 rpm. After 24 h, the pig skin was cleaned, chopped into pieces, and put in a 5 mL centrifuge tube with 3.0 mL of ethanol for sonication for 30 min. It was then filtered through a 0.45 μm pore size membrane and HPLC was used to quantify the amount of silybin and isosilybin in the sample solution.

DPPH assay

The DPPH· scavenging capacity of the samples was determined using a method described previously³⁸ with some modifications. Briefly, SM-DES and SM-control solutions were prepared at multiple concentrations. A 0.12 mg mL⁻¹ solution was prepared by dissolving DPPH in ethanol. The DPPH ethanol solution and sample solution were then mixed in equal volumes (1 mL) in the dark. After incubation for 30 min, the absorbance of the mixture was measured at 517 nm using a UV-vis spectrophotometer (UV-1800, SHIMADZU, Japan). A trolox solution was prepared in ethanol and used as an equivalent calibration standard. The free radical scavenging activity was calculated using the following formula:where A₀, A₁, A₂, and A₃ are the absorbance values of the control, test sample, sample background, and blank background, respectively.

ABTS assay

The ABTS· scavenging experiment was based on an experiment described previously.³⁹ Briefly, ABTS was produced by mixing a 7 mM ABTS solution with 2.45 mM potassium persulfate at room temperature in the dark for 12 h before use. Subsequently the solution was diluted in ethanol to obtain an absorbance of 0.700 ± 0.020 at 734 nm. After the addition of 0.2 mL of the sample solution to 0.8 mL of a diluted ABTS solution, the absorbance at 734 nm at 30 °C was measured. The ABTS scavenging rate was calculated using the following formula:where A₀, A₁, and A₂ are the absorbance values of the control, test sample, and background, respectively.

PFRAP assay

The relative reduction capacities of the samples were evaluated according to previous studies.^40,41 The sample tube (A₁) contained the sample solution (50 μL), PBS buffer (250 μL), and potassium ferricyanide solution (250 μL). It was placed in a 50 °C water bath for 20 min, then cooled to room temperature. Subsequently, 250 μL of trichloroacetic acid solution, 1 mL of ultrapure water, and 400 μL of ferric chloride solution were added into A₁ at room temperature for 10 min, and then the absorbance value was measured at a wavelength of 700 nm. In the sample background tube (A₂), ultrapure water was used to replace potassium ferricyanide, trichloroacetic acid, and ferric chloride. A sample blank tube (A₃) was used to replace the sample with PBS buffer, and a solvent background tube (A₄) was used to replace the sample with PBS buffer based on the sample background tube (A₂). Using the ferrous ion reducing capacity of V_C (0.8 mg mL⁻¹) as a 100% reduction, the relative reduction rate was calculated as follows:

Cell viability assay

The cells used in the experiment were dermal fibroblast cells, purchased from Guangdong BioCell Biotechnology Co., Ltd. Cell viability was evaluated using the 3-(4,5-dimethylthiahiazol-z-y1)-3,5-di-phenytetrazoliumromide (MTT) assay.⁴² A 96-well cell culture plate was inoculated with cells in the logarithmic growth phase and incubated for 24 h under a humidified atmosphere of 5% CO₂ at 37 °C. Different concentrations of samples were added to each column of the 96-well plate for incubating for 24 h in a humidified atmosphere of 5% CO₂ at 37 °C. Subsequently, 10 μL of MTT solution was added to each well and incubation continued for 4 h, after which 150 μL DMSO was added to each well and the absorbance (490 nm) was determined using an enzyme marker (Tecan, Infinite E plex, Swiss) for 10 min. The cell viability was calculated using the following equation:where A_t is the absorbance of the test sample and A_N is the absorbance of the control cells.

Determination of MMP-1 inhibition and type I procollagen

Cells were first exposed to UVB irradiation and then seeded in 96-well plates. After 24 h of treatment, cell culture supernatants were collected to measure type I procollagen and MMP-1 levels. The levels of MMP-1 (Neobioscience Technology Co., Ltd.) and type I procollagen (BOSTER Biological Technology Co., Ltd.) were evaluated according to the enzyme-linked immunosorbent assay (ELISA) kit protocol.

Statistical analysis

Each experiment was repeated three times, and the results were presented as the mean standard deviation. SPSS 22.0 was used for statistical analysis. P < 0.05 indicated significant differences.

Results and discussion

Characterization of Bet/AA-2G

The formation of DES and interactions between the two molecules were observed using FT-IR, DSC, XRD and ¹H-NMR. As shown in Fig. 1A, the peaks appearing from 3380 to 3340 cm⁻¹ corresponded to hydroxyl (–OH) stretching, and those from 1710 to 1630 cm⁻¹ corresponded to the stretching of the carbonyl group (C=O), which were consistent with the characteristic peaks of betaine and AA-2G reported in many previous studies.^28,29,43 Compared with the individual components, these bands shifted to lower wavenumbers, indicating proton migration between betaine and AA-2G. Moreover, a shift of the peaks in the broad region between 2500 and 3400 cm⁻¹ of Bet/AA-2G DES was observed, confirming that Bet/AA-2G DES was successfully prepared through hydrogen bond interactions between betaine and AA-2G.

Fig. 1 FT-IR spectra (A), DSC (B), XRD (C), and ¹H-NMR (D) of Bet/AA-2G and components.

The glass transition temperatures of the Bet/AA-2G DES, individual betaine, and AA-2G were determined by DSC. As shown in Fig. 1B, the melting of DES occurred between −40 and −20 °C with a decreasing and then slightly increasing heat flow value, and the melting point of Bet/AA-2G DES was approximately −20 °C. Compared to betaine and AA-2G, the melting point of Bet/AA-2G DES was significantly lower. Meanwhile, it can be observed that betaine and AA-2G had a typical crystal structure (Fig. 1C). However, the XRD curve showed an amorphous phase for the Bet/AA-2G DES, which indicated that betaine and AA-2G did not exist as individual components, and the Bet/AA-2G DES was synthesized by some interactions. On the other hand, in the ¹H-NMR spectra (Fig. 1D), the corresponding peaks to betaine and AA-2G were also shown in the Bet/AA-2G curve, yet there was a slight peak shift. This meant that there were no chemical interactions between betaine and AA-2G, but hydrogen bonding interactions.⁴⁴ Above all, the Bet/AA-2G DES was synthesized successfully, attributed to the hydrogen bonding interactions between betaine and AA-2G, which was consistent with the results of previous studies.^30,45

Optimal structure of Bet/AA-2G

The optimized structure of Bet/AA-2G was shown in Fig. 2A. The OH group of AA-2G was hydrogen-bonded to the COOH group of betaine (1.71 and 1.69 Å, respectively), and the binding energy (E_b) was approximately −12.37 kcal mol⁻¹, indicating that the Bet/AA-2G DES was successfully synthesized and the system could stably exist. The optimized structure of the DES verified the interaction between the betaine and AA-2G groups, which was consistent with the infrared analysis. The electrostatic attraction between betaine and AA-2G was further analyzed by measuring the electrostatic potential.

Fig. 2 Optimized structure (A), electrostatic potential (B), and interaction region indicator (C) of Bet/AA-2G.

The electrostatic potentials of the molecular van der Waals surfaces were calculated, the electron density isosurface coloring was “blue–white–red”, and these colors are used to indicate the distribution of the electrostatic potential on the van der Waals surface. As shown in Fig. 2B, the carboxyl group of betaine was electronegative, whereas the region around the nitrogen atom was electropositive. The hydroxyl group of AA-2G was electropositive, whereas the carboxyl group was electronegative. The electropositive region of betaine and the electronegative region of AA-2G attracted each other to form a stable DES structure after the synthesis of Bet/AA-2G.

The IRI of the Bet/AA-2G DES was analyzed using Multiwfn. The IRI reflects the intermolecular interaction energy, and different colors on the isosurface indicate different types and intensities of interactions: green indicated the van der Waals action region, red indicated some steric hindrance, and blue indicated a significant attractive effect.^46,47 In Fig. 2C, the green flake areas between betaine and AA-2G indicate van der Waals forces and the blue areas indicated hydrogen bonding, which exhibited a strong attractive force. Overall, the Bet/AA-2G DES could exist stably due to these interactions.

Solubility of silymarin

DES have the advantages of a wide range of polarities, a stable structure, and good solubility in many water-insoluble or insoluble substances.⁴⁸ Ethanol is a traditional solvent used for the effective extraction of active ingredients and 1,3-butanediol is a water-soluble solvent commonly used in the cosmetics industry.⁴⁹ Based on this, the solubility of silymarin in water, ethanol (50 wt%), 1,3-butanediol (50 wt%), DES (50 wt%), and a mixture of DES (50 wt%) and 1,3-butanediol (50 wt%) were compared in this section. The solubility of silymarin in water was tested, and the results showed that silymarin was insoluble (not detected) in water. As shown in Fig. 3, the maximum solubility of silymarin in 1,3-butanediol solution (50 wt%) was achieved when the material–liquid ratio was 1 : 10, whereas silymarin was also able to dissolve in the DES (50 wt%) (2.74 mg mL⁻¹), suggesting that silymarin was insoluble in water but soluble in DES. Interestingly, the maximum solubility of silymarin was observed in a mixture of 5 wt% DES (50 wt%) and 95 wt% 1,3-butanediol solution (50 wt%) in the four solvents when the material–liquid ratios were 1 : 5, 3 : 10, and 1 : 2, suggesting that DES played a role in promoting the solubility of silymarin in the 1,3-butanediol solution. The solubility of silymarin in the mixture of DES and the 1,3-butanediol solution was 17.7% higher than that in the 1,3-butanediol solution without DES (50 wt%) and 60.9% higher than that in the ethanol solution (50 wt%) at a material–liquid ratio of 1 : 2, which may be related to the structure of the DES. In contrast to a previous study,²⁶ Bet/AA-2G DES produced a larger yield of silymarin than ChCl/1,4-butanediol and betaine/1,4-butanediol, implying that it was more suited for silymarin solubilization.

Fig. 3 Solubility of silymarin in different solvents. (A) Ethanol solution (50 wt%); (B) 1,3-butanediol solution (50 wt%); (C) DES (50 wt%); (D) mixture of 5 wt% DES (50 wt%) and 95 wt% 1,3-butanediol solution (50 wt%).

Molecular interactions between silymarin and DES

As shown in Fig. 4A, the carbonyl and hydroxyl groups in the DES bound to the hydroxyl group in silybin (1.97 and 2.06 Å), and similarly, the DES bound to isosilybin by hydrogen bonding and van der Waals forces (1.79 and 1.78 Å). The optimal structures of silybin-DES and isosilybin-DES had binding energies of −11.93 and −11.21 kcal mol⁻¹, respectively, indicating that silymarin had a strong binding capacity with DES and can remain stable after binding. The electrostatic potential of SM-DES was calculated to analyze the electrostatic attraction between silymarin and DES.

Fig. 4 Optimized structure (a) and electrostatic potential (b) of SM-DES.

The electrostatic potentials of the SM-DES systems (including silybin DES and isosilybin DES) were calculated. As shown in Fig. 4B, the electropositive region of DES and the electronegative region of silymarin were attracted to each other, while the electronegative region of DES and the electronegative region of silymarin were also attracted to each other, forming a stabilized SM-DES structure.

The IRI of SM-DES was analyzed using Multiwfn. As shown in Fig. 5, silymarin and DES were mainly bound via van der Waals forces and hydrogen bonds. Interestingly, there were differences in the intermolecular forces between silybin DES and isosilybin SM. Compared with silymarin-DES, isosilybin-SM has more van der Waals forces as their intermolecular interactions. Overall, the IRI results suggested that silymarin and DES were stably bound by hydrogen bonding and van der Waals forces, thus increasing the solubility of silymarin.

Fig. 5 Interaction region indicator figures of SM-DES.

Skin penetration

It is essential for bioactive ingredients in cosmetics to exert their effects by maintaining an excellent capacity for transdermal absorption. The permeation of SM-DES through porcine skin was evaluated using a vertical Franz diffusion cell. As shown in Fig. 6, the peak times for the silybin and isosilybin standards were approximately 18.52 and 26.49 min, respectively, which were similar to those reported in previous studies.⁵⁰ Compared to the silymarin in SM-control after skin penetration (<0.01 mg mL⁻¹), silymarin was detected in SM-DES after skin penetration and the transdermal flux of silymarin was 0.17 mg cm⁻², which suggested that DES significantly increased the permeability of silymarin through porcine skin. This may be because DES can dissolve phospholipids on cell membranes and disrupt the structure of cell membranes, thereby increasing the permeability of cell membranes.⁵¹ This meant that SM-DES can exert its biological activity in deeper layers of the skin.

Fig. 6 HPLC chromatogram of silymarin. (A) Silybin standard; (B) isosilybin standard; (C) SM-DES; (D) SM-control.

Antioxidant activity

Excessive free radical production severely damages the skin microenvironment.⁵² Antioxidants are the primary means of protecting the skin from destruction by free radicals. To evaluate the antioxidant capacity of SM-DES, the DPPH and ABTS radical scavenging rates and PFRAP were tested separately. The variations in the antioxidant capacities of SM-DES and SM-control with concentration were shown in Fig. 7. The results showed that there was a significant difference in the antioxidant capacities of SM-DES and SM-control. Specifically, the IC₅₀ values of SM-DES and SM-control in the DPPH radical scavenging experiment (Table 1) were 0.19% (95%CI: 0.15–0.24%) and 1.82% (95%CI: 1.50–2.20%), respectively. Moreover, the IC₅₀ values of SM-DES and SM-control in the ABTS radical scavenging assay were 0.025% (95%CI: 0.016–0.036%) and 0.091% (95%CI: 0.067–0.12%), respectively, while the IC₅₀ values of SM-DES and SM-control were 0.88% (95%CI: 0.062–1.94%) and 5% (95%CI: 2.43–7.42%), respectively, in the PFRAP assay. These results suggested that the antioxidant capacity of DES was much higher than that of SM-control, which could be explained by two factors. On the one hand, the addition of DES increased the solubility of silymarin in the system. On the other hand, the combination of DES and silymarin may exert a synergistic effect, with the combination of silymarin and DES being able to produce a more stable structure and better antioxidant activity.

Fig. 7 Relationship between the sample concentration and antioxidant capacity: DPPH radical scavenging of (A) SM-control and (B) SM-DES; ABTS radical scavenging of (C) SM-control and (D) SM-DES; relative reduction rates of (E) SM-control and (F) SM-DES.

Table 1 IC50 values of DPPH, ABTS, and PFRAP

IC50/%	SM-control	95%CI	SM-DES	95%CI
DPPH	1.82	1.50–2.20	0.19	0.15–0.24
ABTS	0.091	0.067–0.12	0.025	0.016–0.036
PFRAP	5.00	2.43–7.42	0.88	0.62–1.94

---

Cytotoxicity

Excellent biosafety is a prerequisite for cosmetic ingredients. Therefore, the cell viability of SM-DES was determined. As shown in Fig. 8, SM-DES did not harmfully affect the dermal fibroblasts at concentrations below 0.0875%. In particular, the cell viability was higher than 85% at a concentration of 0.0875%. Therefore, the following experiments were tested by SM-DES at a concentration below 0.0875%.

Fig. 8 Cytotoxicity of SM-DES.

MMP-1 inhibition

The overexpression of matrix metalloproteinases (MMPs) leads to the significant degradation of collagen and elastin and accelerates skin aging, with matrix metalloproteinase 1 (MMP-1) being the most predominant enzyme for the degradation of type I.^53,54 The anti-wrinkle ability of SM-DES was evaluated by measuring MMP-1 levels. As shown in Fig. 9, MMP-1 expression in the model was 100%, indicating the successful modelling of the experiment. Compared to the model, MMP-1 expression in the positive control (PC) was significantly decreased (P < 0.05), indicating that the experiment was effective. SM-DES at concentrations of 0.075%, 0.050%, and 0.020% showed a significant decrease in MMP-1 expression compared to the model group, with inhibition rates of 92.92% (P < 0.05), 81.94% (P < 0.05), and 66.80% (P < 0.05), respectively, indicating that SM-DES could effectively inhibit the expression of MMP-1 with an anti-wrinkle ability.

Fig. 9 Cytotoxicity and MMP-1 inhibition of SM-DES.

Collagen type I content

Type I collagen is one of the main components of the extracellular matrix of the dermis, and the firming and anti-wrinkle efficacy of the samples to be tested can be assessed by measuring the amount of type I collagen secreted by fibroblasts. The effect of different concentrations of SM-DES on type I collagen content was determined using ELISA.⁵⁵ As shown in Fig. 10, the type I collagen content of SM-DES at concentrations of 0.1, 0.4, and 0.8 mg mL⁻¹ increased significantly by 53% (P < 0.05), 102% (P < 0.05), and 153% (P < 0.05), respectively, compared with that of SM-control, which meant that SM-DES can promote type I collagen synthesis, thereby exerting anti-wrinkle and tightening effects.

Fig. 10 Type I collagen synthesis capacity of SM-DES. *P < 0.05, **P < 0.01.

Conclusions

In this study, solubilization, molecular interactions, and biological activity of silymarin in DES were investigated. The Bet/AA-2G DES exhibited superior solubilization properties compared to conventional solvents. Molecular modeling studies revealed strong binding interactions between silymarin and the DES, confirming the enhanced solubility observed experimentally. Additionally, DES significantly improved the skin permeability of silymarin, indicating its potential for topical applications. Moreover, silymarin dissolved in DES exhibited enhanced antioxidant activity compared to SM-control. In addition, SM-DES significantly reduced the expression of MMP-1 and promoted the synthesis of type I collagen, suggesting anti-wrinkle and firming properties. Overall, DES is a promising solvent for enhancing the solubility and bioactivity of hydrophobic compounds, and silymarin dissolved in DES has a wide potential range of applications in food, cosmetics, and biomedicine.

Data availability

Data available upon request from the authors.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This study was supported by the National Natural Science Foundation of China (21905069, U21A20307), the Shenzhen Science and Technology Innovation Committee (Grant No. ZDSYS2019 0902093220279, KQTD20170809110344233, GXWD20201230155427003-20200821181245001, GXWD20201230155427003-20200821181809001, ZX20200151), and the Department of Science and Technology of Guangdong Province (Grant No. 2020A1515110879).

References

1. I. Maleš, S. Pedisić, Z. Zorić, I. Elez-Garofulić, M. Repajić, L. You, S. Vladimir-Knežević, D. Butorac and V. Dragović-Uzelac, J. Funct. Foods, 2022, 96, 105210.
2. D. c Hao and P. g Xiao, Chin. Herb. Med., 2020, 12(2), 104–117.
3. M. Macit, G. Duman, A. Cumbul, E. Sumer and C. Macit, J. Drug Delivery Sci. Technol., 2023, 83, 104378.
4. M. Koushki, R. F. Yekta and N. Amiri-Dashatan, J. Funct. Foods, 2023, 104, 105502.
5. M. Fallah, A. Davoodvandi, S. Nikmanzar, S. Aghili, S. M. A. Mirazimi, M. Aschner, A. Rashidian, M. R. Hamblin, M. Chamanara and N. Naghsh, Biomed. Pharmacother., 2021, 142, 112024.
6. A. A. Elateeq, Y. Sun, W. Nxumalo and A. M. M. Gabr, Biocatal. Agric. Biotechnol., 2020, 29, 101775.
7. V. K. Shankar, A. Police, S. Ajjarapu and S. N. Murthy, Int. J. Pharm., 2023, 630, 122431.
8. M. Bijak, Molecules, 2017, 22, 1942.
9. Z. Zhang, X. Li, S. Sang, D. J. McClements, L. Chen, J. Long, A. Jiao, J. Wang, Z. Jin and C. Qiu, Food Res. Int., 2022, 156, 111314.
10. S. S. Kesharwani, V. Jain, S. Dey, S. Sharma, P. Mallya and V. A. Kumar, J. Drug Delivery Sci. Technol., 2020, 60, 102021.
11. Y. Wang, A.-J. Yuan, Y.-J. Wu, L.-M. Wu and L. Zhang, J. Funct. Foods, 2023, 100, 105384.
12. J.-S. Lee, E. S. Kim and H. G. Lee, Colloids Surf., B, 2017, 154, 171–177.
13. L. Yang, D. Zhang, W. Li, H. Lin, C. Ding, Q. Liu, L. Wang, Z. Li, L. Mei, H. Chen, Y. Zhao and X. Zeng, Nat. Commun., 2023, 14.
14. P. Sun, Z. Li, D. Zhang, W. Zeng, Y. Zheng, L. Mei, H. Chen, N. Gao and X. Zeng, Chin. Chem. Lett., 2024, 35, 108346.
15. K. A. Omar and R. Sadeghi, J. Mol. Liq., 2023, 384, 121899.
16. Z. Usmani, M. Sharma, M. Tripathi, T. Lukk, Y. Karpichev, N. Gathergood, B. N. Singh, V. K. Thakur, M. Tabatabaei and V. K. Gupta, Sci. Total Environ, 2023, 881, 163002.
17. F. Nourizadeh, M. Mokhtarpour, Z. V. Ziaee, M. Khorsandi, A. Sadrmousavi and H. Shekaari, J. Mol. Liq., 2022, 367, 120433.
18. O. Alioui, W. Sobhi, M. Tiecco, I. M. Alnashef, A. Attoui, A. Boudechicha, K. K. Yadav, A. M. Fallatah, N. Elboughdiri and B.-H. Jeon, J. Mol. Liq., 2022, 359, 119149.
19. A. Prabhune and R. Dey, J. Mol. Liq., 2023, 379, 121676.
20. W. D. Aryati, A. Nadhira, D. Febianli, F. Fransisca and A. Mun’im, J. Res. Pharm., 2020, 24, 389–398.
21. A. A. Petrochenko, A. Orlova, N. Frolova, E. B. Serebryakov, A. Soboleva, E. V. Flisyuk, A. Frolov and A. N. Shikov, Molecules, 2023, 28, 3614.
22. X. Fu, D. Wang, T. Belwal, J. Xie, Y. Xu, L. Li, L. Zou, L. Zhang and Z. Luo, LWT - Food Science and Technology, 2021, 144, 111220.
23. A. N. Shikov, V. M. Kosman, E. V. Flissyuk, I. E. Smekhova, A. Elameen and O. N. Pozharitskaya, Molecules, 2020, 25, 1826.
24. C.-V. Nicolae, B. Trică, S.-M. Doncea, I. C. Marinaş, F. Oancea and D. Constantinescu-Aruxandei.
25. E. D. Obluchinskaya, O. N. Pozharitskaya, V. A. Shevyrin, E. G. Kovaleva, E. V. Flisyuk and A. N. Shikov, Mar. Drugs, 2023, 21, 263.
26. J. Yao, L. Xiao, C. Li, B. Wang, Y. Chen, X. Yan and Z. Cui, J. Mol. Liq., 2022, 363, 119768.
27. W. Zhang, Q. Huang, R. Yang, W. Zhao and X. Hua, J. Funct. Foods, 2021, 82, 104481.
28. F. M. Fuad and M. M. Nadzir, Ind. Crops Prod., 2023, 192, 116069.
29. F. M. Fuad and M. M. Nadzir, J. Mol. Liq., 2022, 360, 119392.
30. M. Hu, G. Feng, L. Xie, X. Shi, B. Lu, Y. Li, S. Shi and J. Zhang, J. Mol. Liq., 2023, 382, 121977.
31. P. V. Dangre, A. D. Tattu, S. P. Borikar, S. J. Surana and S. S. Chalikwar, Int. J. Biol. Macromol., 2021, 171, 514–526.
32. I. M. Aroso, J. C. Silva, F. Mano, A. S. D. Ferreira, M. Dionísio, I. Sá-Nogueira, S. Barreiros, R. L. Reis, A. Paiva and A. R. C. Duarte, Eur. J. Pharm. Biopharm., 2016, 98, 57–66.
33. H. T. Çelik and M. Gürü, J. Supercrit. Fluids, 2015, 100, 105–109.
34. Q. Dong, J. Cao, L. Chen, J. Cao, H. Wang, F. Cao and E. Su, Ind. Crops Prod., 2022, 177, 114455.
35. S. Grimme, S. Ehrlich and L. Goerigk, J. Comput. Chem., 2011, 32, 1456–1465.
36. O. A. Vydrov and G. E. Scuseria, J. Chem. Phys., 2006, 125.
37. A. L. S. Oliveira, D. Valente, H. R. Moreira, M. Pintado and P. Costa, Colloids Surf., B, 2022, 218, 112779.
38. X. Sun, C. Liu, A. M. Omer, L.-Y. Yang and X.-k Ouyang, Int. J. Biol. Macromol., 2019, 132, 487–494.
39. X. Huang, X. Huang, Y. Gong, H. Xiao, D. J. McClements and K. Hu, Food Res. Int., 2016, 87, 1–9.
40. A. Mokrani, S. Cluzet, K. Madani, E. Pakina, A. Gadzhikurbanov, M. Mesnil, A. Monvoisin and T. Richard, Int. J. Mass Spectrom., 2019, 445, 116192.
41. A. Mokrani, S. Krisa, S. Cluzet, G. Da Costa, H. Temsamani, E. Renouf, J.-M. Mérillon, K. Madani, M. Mesnil and A. Monvoisin, Food Chem., 2016, 202, 212–220.
42. J. S. Bang and S.-Y. Choung, J. Photochem. Photobiol., B, 2020, 209, 111946.
43. D. Kwaśniewska and J. Kiewlicz, J. Mol. Liq., 2022, 354, 118917.
44. Z. Pan, Y. Bo, Y. Liang, B. Lu, J. Zhan, J. Zhang and J. Zhang, J. Mol. Liq., 2021, 326, 115283.
45. B. Li, T. Xiao, S. Guo, Y. Wu, R. Lai, Z. Liu, W. Luo and Y. Xu, Int. J. Pharm., 2023, 637, 122880.
46. C. Sun, Y. Zhang, Y. Dai, J. Xing, Y. Qu, Y. Wang and J. Gao, J. Mol. Liq., 2023, 371, 121108.
47. C. Zhang, Y. Liu, F. Song and J. Wang, J. Mol. Liq., 2022, 361, 119665.
48. I. Bashir, A. H. Dar, K. K. Dash, V. K. Pandey, U. Fayaz, R. Shams, S. Srivastava and R. Singh, Sustainable Chem. Pharm., 2023, 33, 101102.
49. R. A. Abdel-Rahem, J. Surfactants Deterg., 2014, 17, 353–362.
50. S. Gilabadi, H. Stanyon, D. DeCeita, B. A. Pendry and E. Galante, J. Herb. Med., 2023, 37, 100619.
51. J. Cao, R. Wu, Q. Dong, L. Zhao, F. Cao and E. Su, ChemBioChem, 2020, 21, 672–680.
52. K. P. Anthony and M. A. Saleh, Antioxidants, 2013, 2, 398–407.
53. S. Han, T. G. Lim, J. E. Kim, H. Yang, D. K. Oh, J. H. Yoon Park, H. J. Kim, Y. K. Rhee and K. W. Lee, J. Cell. Biochem., 2017, 118, 3756–3764.
54. S. Jang, J. Chun, E. M. Shin, H. Kim and Y. S. Kim, J. Pharm. Invest., 2012, 42, 33–39.
55. A. C. Weihermann, M. Lorencini, C. A. Brohem and C. M. De Carvalho, Int. J. Cosmet. Sci., 2017, 39, 241–247.

---

Footnote

† Both authors contributed equally to this work.

---