Ultrasound-assisted deep eutectic solvent extraction of ginseng folium polysaccharides: Extraction mechanism, process optimization, structural characterization, and in vitro biological activity

Graphical abstract

Abstract

This study aimed to develop a green, efficient ultrasound-assisted (UA) deep eutectic solvent (DES) extraction method for ginseng folium polysaccharides (GFPs). This approach overcomes the low yields and long extraction period of conventional techniques. Seventeen DES systems were screened, identifying choline chloride–lactic acid (ChCl-La, 1:2) as the optimal solvent. Key parameters were optimized using Plackett–Burman design followed by Box–Behnken response surface methodology, yielding optimal conditions of 30 min ultrasonic time, a liquid solid ratio of 25 mL/g, ultrasonic power of 299.16 W, and DES water content of 30.82%. The effects of three extraction methods, namely hot water extraction (HWE), UA extraction (UAE), and UA-DES, on polysaccharide characteristics were compared. Monosaccharide analysis showed that HWE-GFP was rich in glucose (Glc, 16.15%) and glucuronic acid (GlcA, 45.23%), whereas UAE and UA-DES were enriched in galactose (30.79% and 37.78%, respectively) and arabinose (21.69% and 29.80%, respectively). UA-DES-GFP had a significantly lower molecular weight (361.70 kDa) than UAE-GFP (1497.40 kDa) or HWE-GFP (3304.35 kDa). Fourier-transform infrared spectroscopy revealed no major differences in functional groups among the three GFPs. Molecular dynamics simulations of the ChCl-La system elucidated the extraction mechanism, highlighting favorable electrostatic potential energy (−20.14 eV) and hydrogen bond length (2.204 Å) interactions with GFP. In vitro bioassays showed that HWE-GFP had stronger antioxidant activity (61.97%), whereas UA-DES-GFP exhibited superior α-glucosidase inhibition (59.54%). These differences likely stem from variations in the GlcA/Glc ratio. Overall, this study provides a high-yield, green extraction strategy for GFP, along with mechanistic insights and technical guidance for utilizing ginseng leaf polysaccharides.

1. Introduction

Ginseng folium (GF) consists of the dried leaves of Panax ginseng C. A. Mey. (PG), a member of the Araliaceae family. PG, a renowned traditional Chinese medicine with a long history of use, was first documented in the Shen Nong Ben Cao Jing and is primarily distributed in northeastern China [1]. Owing to its remarkable therapeutic effects, ginseng is often called the “King of Herbs.” GF contains active ingredients similar to those in PG, including saponins, flavonoids, and polysaccharides, and its leaves contain higher levels of specific ginsenosides (such as Rb₁, Rb₂, and Rd). However, large quantities of PG stems and leaves are discarded during processing. Studies show that ginseng stems and leaves have a market value and utilization rate of only about 20% that of its roots, despite accounting for 50% of the total yield, and this proportion is increasing annually [2]. These disparities underscore major gaps in GF research and utilization. Extracting and utilizing active compounds from GF thus holds great research significance and substantial development potential.

Polysaccharides are biological macromolecules composed of one or more monosaccharide units linked by glycosidic bonds; they occur widely in plants [3]. Their high safety, non-toxicity, and water solubility have driven extensive research. In traditional practices, ginseng leaves are brewed into teas or decoctions to alleviate summer heat and irritability, underscoring the role of GF polysaccharides (GFPs) in disease treatment. Studies confirm that GFPs exhibit diverse pharmacological activities, including anti-cancer, immunomodulatory [4], antioxidant [5], and hypoglycemic [6] effects. They are also used in functional foods and health products to boost energy, regulate the spleen, and improve circulation. Efficient extraction methods are thus crucial for maximizing the biological activity of GFPs. Common extraction approaches include traditional hot water extraction (HWE), enzyme-assisted extraction (ETE) [7], and acid/alkali-assisted extraction [8]. HWE relies on heat conduction to extract active substances, but is limited due to low efficiency and high energy consumption [9]. ETE improves cell permeability via enzymatic catalysis, thereby improving extraction efficiency; however, it is costly and requires specific enzymes [10]. Furthermore, these methods require preheating the extraction vessel and depend on substrate properties. To address these limitations, many combined extraction methods have been proposed, including microwave-assisted extraction and ultrasound-assisted extraction (UAE). In recent years, the UAE has been widely used for extracting plant polysaccharides due to its high efficiency, short extraction time, and preservation of active compounds [8], [11]. It leverages cavitation to break cell walls and intercellular matrix, thereby enhancing the permeability of the extract [12]. However, despite advances and the discovery of novel extraction methods for improving the integrity and bioactivity of polysaccharides, challenges persist in solvent recoverability, reusability, and cost. Developing efficient, green, and economical methods for GFP extraction thus remains a critical bottleneck.

In recent years, deep eutectic solvents (DES), known for their green, safe properties, have demonstrated promising potential for polysaccharide extraction [13]. A DES consists of a hydrogen bond acceptor (HBA) and a hydrogen bond donor (HBD) that form a stable eutectic mixture through intermolecular hydrogen bonding [14]. This system excels in solubilizing active compounds such as polysaccharides, offering excellent stability, environmental compatibility, and cost-effectiveness. Notably, DES enables efficient polysaccharide extraction under mild conditions, preserving structural integrity and biological activity [15]. Studies show that ultrasound-assisted (UA)-DES extraction, leveraging the synergy of ultrasonication and DES, significantly enhances both the yield and bioactivity of polysaccharides [16], [17]. Ultrasonication disrupts cell walls via cavitation and improves solvent permeability, and the DES hydrogen bond network promotes target compound solubilization [18]. Reportedly, polysaccharides extracted via different methods exhibit varying bioactivities, attributable to differences in monosaccharide composition, molecular weight, functional groups, and other structural characteristics. To date, no studies have reported optimized UA-DES conditions for extracting GFPs or compared the impacts of extraction methods on GFP's chemical structure and biological activity.

This study aimed to develop and optimize a UA-DES process for GFP extraction. We first screened numerous DES formulations to identify the optimal system. Next, the key parameters, including ultrasonication time, ultrasonication power, liquid-to-solid ratio, and water content, were optimized. The resulting GFP was characterized and compared with polysaccharides from other extraction methods to highlight structural and functional differences. Subsequently, the extraction mechanism was elucidated from the perspectives of electrostatic potential energy (EPE) and bond length using molecular dynamics simulations and density functional theory (DFT). Finally, in vitro bioactivity assays evaluated polysaccharides extracted from different methods. By integrating extraction efficiency, structural analysis, and bioactivity data, this work provides a green, high-performance method for GFP extraction and a theoretical foundation for its application in the food industry.

2. Materials and methods

2.1. Materials

Ethylene glycol (Eg), glycerol, 1,3-propanediol, 1,4-butanediol, urea, lactic acid (La), and choline chloride (ChCl) were purchased from Tianjin Chemical Reagent Co., Ltd., and monosaccharide standards were purchased from Shanghai Guoyao Chemical Reagent Co., Ltd., China. 1-Phenylen-3-methyl-5-pyrazolone (PMP) and 1,1-diphenyl-2-picrylhydrazone (DPPH) were purchased from Shanghai Aladdin Biochemical Technology Co., Ltd. and Shanghai Maclean Biochemical Co., Ltd., China, respectively. α-Glucosidase and 4-nitrophenyl-alpha-D-glucopyranoside (PNPG) were from Inokai Ltd. All other chemical reagents were analytical or chromatographic grade.

2.2. Preparation of DES system

ChCl was combined with 6 different HBDs to form DES systems with 17 different molar ratios. All systems were continuously stirred at 80°C to obtain a homogeneous and transparent solution. Considering that the viscosity of DES would complicate subsequent filtration operations, 40% water was added after the clear solution cooled to investigate the extraction efficiency. Specific molar amounts are shown in Table 1.

Table 1.

The screening and extraction results of GFP from 17 different DES systems.

HBA	HBD	Mole ratio	Yield%	Purity%
ChCl	1,4-butanediol	1:2	11.50 ± 3.15%	23.49 ± 5.45%
		1:3	11.10 ± 2.51%	20.63 ± 4.29%
ChCl	Eg	1:1	10.90 ± 2.20%	34.38 ± 7.35%
		1:2	15.30 ± 3.30%	35.33 ± 8.08%
		1:3	11.95 ± 3.50%	28.58 ± 5.91%
ChCl	glycerol	1:1	8.90 ± 2.15%	24.71 ± 5.73%
		1:2	9.80 ± 3.20%	25.08 ± 4.78%
		1:3	9.00 ± 2.50%	31.66 ± 7.51%
ChCl	1,3-glycerol	1:1	6.80 ± 2.20%	14.58 ± 3.60%
		1:2	8.20 ± 2.5%	9.80 ± 3.11%
		1:3	9.10 ± 3.00%	12.49 ± 2.59%
ChCl	La	1:1	10.61 ± 3.10%	37.59 ± 6.10%
		1:2	15.15 ± 4.10%	45.14 ± 9.52%
		1:3	11.50 ± 2.80%	32.51 ± 6.36%
ChCl	urea	1:1	7.00 ± 2.10%	23.54 ± 5.58%
		1:2	11.21 ± 3.60%	28.41 ± 6.47%
		1:3	9.61 ± 3.20%	21.14 ± 5.73%
UAE		−	10.00 ± 3.31%	30.21 ± 6.34%
HWE		−	8.10 ± 2.10%	28.51 ± 6.21%

2.3. UA-DES combination: High-efficiency extraction of ginseng leaf polysaccharides

First, weigh 1.0 g of defatted (75% alcohol) GF powder, add 20 mL of DES (40% water content), and sonicate at 40℃ for 30 min. After extraction, centrifuge the sample (to remove floating matter and other impurities), then add anhydrous ethanol to achieve an alcohol concentration of 80%, and incubate at 4°C overnight. The next day, the mixture was centrifuged at 3500 rpm for 10 min, and the precipitate was deproteinized using the Sevage method to obtain GFP. The supernatant was concentrated to obtain the recovered DES. This study uses yield and purity results to comprehensively evaluate the DES system.

Both HWE and UAE were set to a liquid–solid ratio of 1:20 and an extraction time of 120 min. The extraction temperature for HWE was controlled at 100℃, and the ultrasonic power for UAE was set to 400 W. Comparative study of HWE (A-GFP) and UAE (B-GFP) for UA-DES (C-GFP). The purity of polysaccharides was determined by the phenol–sulfuric acid method. The yield and purity of polysaccharides were calculated using formula (1). The yield and purity of the obtained GFP are shown in Table 1.

$$\begin{array}{c}\text{Yield}\left(\%\right)\mathrm{=}\text{mass of polysaccharide / weight of GF crude extract × 100\%}\hfill \\ \text{Purity}\left(\%\right)=\left[\text{C}\left(\text{mg}/\text{mL}\right)\times \text{V}\left(\text{mL}\right)\times \text{n}\right]/\text{mass of polysaccharide}\left(\text{mg}\right)\times 100\%\end{array}$$ (1)

Where C is the glucose concentration in GFP, V is the volume of GFP, and n is the dilution factor.

2.4. Optimization of extraction conditions for the UA-DES system

2.4.1. Single factor experiment

After screening the optimal UA-DES system, single-factor analysis was conducted on the optimal system to obtain polysaccharides with higher purity. This study investigated five relevant factors affecting the extraction process in response to polysaccharide yield: water content (0, 10, 20, 30, 40, 50%), liquid solid ratio (5, 10, 15, 20, 25, 30 mL/g), ultrasonic time (10, 20, 30, 40, 50 min), and ultrasonic power (100, 200, 300, 400, 500 W).

2.4.2. PBD experiment

The optimization conditions for PBD were selected based on single factor experiments. A factorial design was used to conduct 12 experiments at two levels with four variables (water content X1, liquid solid ratio X2, ultrasonic time X3, and ultrasonic power X4). The yield (T) of GFP was used as the response value. Design-Expert 13.0 was used to further identify the key factors affecting GFP extraction.

2.4.3. BBD-RSM experiment optimization

Based on single factor and PBD results screening, the main factors affecting the ginseng leaf extraction rate were X2 (liquid solid ratio), X3 (ultrasonic power), and X1 (water content). A three-factor, three-level Box-Behnken response surface design was performed using Design Expert 13.0. As shown in Table S2, each independent variable was assigned a code of −1 (low), 0 (medium), or + 1 (high).

2.5. FT-IR structural analysis of the optimal UA-DES system

The prepared DES system and each ligand were dispersed in dry potassium bromide. Fourier transform infrared (FT-IR) spectra of each sample were obtained in the range of 4000 ∼ 400 cm⁻¹ and the changes in internal structure were analyzed.

2.6. Structural differences between different extraction methods

2.6.1. Differences in monosaccharide composition

The monosaccharide composition was determined using the complete acid hydrolysis method studied previously [19]. In simple terms, the polysaccharide is first completely hydrolyzed with trifluoroacetic acid, then derivatized with PMP under alkaline conditions, followed by neutralization with hydrochloric acid. Finally, excess PMP is extracted with chloroform to obtain the final product to be tested. Detection conditions: Instrument: Agilent 1260 system; Detection wavelength: 245 nm; Mobile phase: 18% acetonitrile phosphate buffer (pH = 6.85); flow rate: 0.9 mL/min.

2.6.2. Differences in FT-IR spectra

Polysaccharides obtained by three different extraction methods were mixed with a certain amount of dried KBr (1:20) in a mortar and ground, then compressed into tablets. Finally, FT-IR spectroscopy was used to obtain the sample spectra. The range is 4000 ∼ 400 cm⁻¹, the resolution is 4 cm⁻¹, and the scan is performed 32 times.

2.6.3. Differences in molecular weight

The molecular weight distribution of polysaccharides was determined using high-performance gel permeation chromatography (HPGPC). The method was based on our previous method for detection [20]. The general conditions were as follows: the chromatographic column was TSK-Gel G-5000PW_xl, the mobile phase was Watson's water, the flow rate was 1 mL min⁻¹, the standard was dextran (T2000 ∼ T50), and the standard curve was: Y = −0.4407X + 5.7314, R² = 0.9909.

2.7. Molecular dynamics simulation of the DES system

This study uses quantum mechanical methods to obtain the molecular model of the optimal DES system. The selected molecular model was generated by Material Studio (MS) simulation. Gal was used as the target compound for simulation. Then, Dmol3 in the density function theory (DFT) of MS was used for calculation. Finally, the structure was optimized and the binding energy and EPE were calculated. The binding energy is calculated according to formula (2).

$$\text{W}={\text{E}}_{\text{N}}-{\text{E}}_0$$ (2)

In the formula, E_N is the total energy of the free state, and E₀ is the crystal energy.

2.8. Evaluation of in vitro bioactivity

2.8.1. Evaluation of antioxidant activity experiments

The antioxidant capacity of A-GFP, B-GFP, and C-GFP was determined by the methods in the literature [21]. In simple terms: different polysaccharides are prepared into solutions of corresponding concentrations (5, 4, 3, 2, 1 mg mL⁻¹). Different concentrations of polysaccharide solutions (20 μL) were mixed with 200 μL of 0.1 mmol L⁻¹ DPPH and reacted in the dark for 30 min. The absorbance was measured at 517 nm. Each sample was measured in triplicate. Vitamin C was used as the positive control.

2.8.2. In vitro hypoglycemic activity study

The in vitro hypoglycemic activity of A-GFP, B-GFP and C-GFP was investigated using an α-glucosidase inhibition assay. Minor modifications were made based on the literature [22]. Polysaccharides were prepared into solutions of different concentrations (10, 5, 4, 3, 2, 1 mg mL⁻¹). 100 μL of each solution was reacted with 50 μL of α-glucosidase (2 U/mL) at 37°C for 5 min. Then add 50 μL of PNPG (5 mmol/L), continue the reaction for 10 min, and finally add 50 μL of sodium carbonate solution (1 mol/L) to terminate the reaction. Acarbose was used as a positive control, and the absorbance was recorded at 405 nm. Each sample was measured in triplicate. The inhibition rate was calculated according to formula (3).

$$\text{Inhibition rate}\left(\%\right)=\left[1-\left(\text{A}1-\text{A}2\right)/\text{A}0\right]\times 100\%$$ (3)

Where A1 is the absorbance of GFP, A2 is the background absorbance, and A0 represents the absorbance of the blank group.

3. Results and discussion

3.1. Yield and purity of different DES systems

The composition of DES can be determined by assessing the composition of HBA and HBD [23]. During the synthesis of DES, different types of HBD and HBA can form solvents with different properties (such as polarity and color). Ultimately, we identified 17 state-stable DES for synthesis (Table 1). As most pure DES systems exhibit slow reaction efficiency at room temperature, which results in incomplete dissolution, increasing the temperature and water content during the dissolution process can accelerate the reaction. Most experiments have demonstrated that adding water to the DES system can increase the diffusion coefficient of the medicinal material and reduce the viscosity of the system. Some studies have also indicated that > 50% water can disrupt the hydrogen bonds between the structures [17]. Therefore, when examining the yield and purity of DES in different systems, 40% water was added to the screen for the optimal system. The results revealed that the polysaccharide extraction rate was the highest (15.30%) when HBD was Eg and the molar ratio was 1:2; however, the polysaccharide purity was the highest (45.14%) when HBD was La and the molar ratio was 1:2. This difference may be attributed to the fact that acidic DES can disrupt the molecular bonds in polysaccharides, which releases more polysaccharides. Furthermore, the yield and purity of polysaccharides obtained from higher proportions of HBA were found to be lower than those from the lower proportions of HBA. This may be attributed to the higher proportion of HBA, leading to a higher viscosity of the DES system, which, in turn, reduces the fluidity of the extraction system and further affects the diffusion effect of polysaccharides. Therefore, based on the comparison presented in Table 1, this study adopted ChCl-La (1:2) as the research system for subsequent extraction processes.

3.2. Effects of various factors on polysaccharide extraction yield

3.2.1. Effect of water content

Numerous studies have demonstrated that, in DES systems, water content can not only alter the system’s viscosity but also affect the extraction and mass transfer efficiency of substances [24]. The higher the water content, the lower the viscosity of the DES system; however, this does not mean that the water content can be increased indefinitely. Although lower viscosity DES has strong permeation and mass transfer capabilities, excessively high-water content can disrupt the hydrogen bonds in DES, thereby disrupting its intrinsic hydrogen-bond network. The high extraction yield observed in the UA-DES system may be attributed to the presence of a large number of hydrogen bond networks and the interaction of van der Waals forces among different components, which is more conducive to the extraction and separation of active ingredients from plants. Therefore, in this study, we investigated the effect of water content on the optimal UA-DES system for GFP extraction. The results are shown in Fig. 1A. The yield varied with the water content. For instance, within the 0–30% range, the yield of polysaccharides gradually increased with increasing water content and decreasing system viscosity. However, when the water content reached 40–50%, the extraction rate of polysaccharides decreased significantly. This fluctuation may be attributed to the addition of a large amount of water, which disrupted the hydrogen bond structure of the system, ultimately weakening the hydrogen bond interaction within the system and thus negatively impacting the extraction process. Based on this result, the optimal water content of the UA-DES extraction system was initially determined to be 30%.

Fig. 1.

The results of four single-factor experiments that affect the extraction of GFP by UA-DES (A: Water content; B: Liquid solid ratio; C: Ultrasonic time; D: Ultrasonic power).

3.2.2. Effect of liquid–solid ratio

An appropriate liquid–solid ratio can not only maximize the extraction of active ingredients from plants but also avoid the waste of raw materials and reagents. The effect of the liquid–solid ratio on the GFP extraction rate is shown in Fig. 1B. At a liquid–solid ratio of 5–20 mL/g, the extraction rate increased gradually with a slight effect. This effect may be attributed to the low liquid–solid ratio that led to incomplete diffusion of the active components from the herbs, resulting in reduced solubility of GFP in the system. Only a small amount of the polysaccharide dissolved in the DES system, with most remaining in the cells. With increasing liquid–solid ratio, most of the GFP was completely dissolved, and the extraction rate reached its peak at 25 mL/g. Subsequently, a further increase in the liquid–solid ratio can lead to the continuous dilution of the DES system and a reduction in the extraction rate. Finally, considering the amount of the solvent and the extraction effect, the optimum liquid–solid ratio was finally confirmed to be 25 mL/g.

3.2.3. Effect of ultrasonic time

Fig. 1C demonstrates the effect of ultrasound time on the extraction of GFP from the DES system. This result indicates that the extraction rate increases within 10–30 min, peaking at 30 min. When the sonication time is extended to 50 min, the polysaccharide yield begins to decline gradually. This effect may be attributed to the dual effects of ultrasonic cavitation on mass transfer and polysaccharide stability. The active ingredients in plant species require sufficient time to dissolve and diffuse; therefore, extraction time usually has a promoting effect on the extracted active ingredients [25]. When the ultrasonic treatment time is inadequate, the solvent and medicinal materials may not come into complete contact and may not dissolve completely in the solution, resulting in insufficient extraction of the active ingredients. Prolonged ultrasonic waves, while accelerating the dissolution and diffusion of substances, also generate reactive components such as free radicals and oxides. These resultant components can disrupt the chemical bonds among the polysaccharides, causing polysaccharide decomposition and degradation, which reduces the yield of polysaccharides [26]. Ultimately, based on the experimental outcomes, the optimal sonication time was determined to be 30 min, at which point, the polysaccharide yield was highest (15.97%).

3.2.4. Effect of ultrasonic power

Under optimal conditions for other factors, the effect of ultrasonic power was further investigated, and the results are presented in Fig. 1D. Within a certain range, the extraction rate of GFP increased with increasing ultrasonic power, reaching its highest level at 300 W. Beyond 300 W, the extraction rate reduced gradually. The increased extraction rate in the former case can be attributed to the synergistic effect of ultrasonic cavitation and mechanical–thermal interaction, which promotes the pulverization of medicinal herb particles, increases the surface area of the extract, and facilitates solvent penetration into the medicinal cell interior, thereby accelerating the polysaccharide-extraction process [27]. The possible reason for the decrease in the latter extraction rate could be that the continuous cavitation effect produces excessive bubbles, which, in turn, hinder the extraction and penetration of the solution as well as the transmission of ultrasonic energy [28]. Therefore, in this study, we selected 300 W of ultrasonic power as the basis for further optimization of GFP.

3.3. Key factors influencing Plackett–Burman design (PBD) screening and GFP extraction

PBD is a statistical method for screening key factors. It identifies significant variables at low cost and with minimal experiments, making it widely used for optimizing polysaccharide extraction processes [29]. Building on single-factor experiments, this study applied PBD to screen key factors affecting GFP extraction. The data in Table S1 were subjected to PBD regression analysis, and the following optimal extraction equation for GFP was obtained:

$$\text{T}=+12.31+0.7250\text{X}1+0.9667\text{X}2+0.8867\text{X}3-0.2050\text{X}4$$

The variance results are shown in Table 2. The model predicted a P-value of < 0.01, indicating the significance of the regression model and that it can be used to identify key factors.

Table 2.

UA-DES extracted the results of the differential analysis for the PBD test in GFP.

Source	Sum of Squares	df	Mean Square	F-value	p-value
Model	27.46	4	6.86	8.20	0.0089	P < 0.01
A-%	6.31	1	6.31	7.53	0.0287	P < 0.05
B-g/mL	11.21	1	11.21	13.39	0.0081	P < 0.01
C-W	9.43	1	9.43	11.26	0.0121	P < 0.05
D-min	0.5043	1	0.5043	0.6021	0.4632
Residual	5.86	7	0.8375
Lack of Fit	5.71	6	0.9519	6.29	0.2960
Pure Error	0.1513	1	0.1513
Cor Total	33.32	11

In a Pareto plot, factors exceeding the t-value reference line significantly affect the response. In a semi-normal plot, greater deviation from the straight line indicates higher importance [30]. The factors influencing GFP yield (Fig. 2A and B) were as follows: X2 (liquid-to-solid ratio), X3 (ultrasonic power), X1 (water content), and X4 (ultrasonic time). Notably, X2, X3, and X1 positively correlated with the response value, exhibiting a significant impact. Thus, PBD identified X2, X3, and X1 as key factors for subsequent optimization of GFP.

Fig. 2.

Pareto chart and Half-Normal plot showing the effects of four variables on GFP extraction rate: X1 (water content), X2 (liquid solid ratio), X3 (ultrasonic power) and X4 (ultrasonic time).

3.4. Box–Behnken design (BBD)-response surface methodology (RSM) optimization

Based on single-factor and PBD experiments, three significant factors were selected for optimization using RSM: liquid-to-solid ratio (20, 25, 30 mL/g), ultrasonic power (200, 300, 400 W), and water content (20, 30, 40%). Polysaccharide yield served as the response variable. A BBD-RSM was implemented using 17 experiments based on a three-factor, three-level principle. Data from Table S2 were analyzed in Design Expert 13.0, yielding the following model equation for polysaccharide yield:

$$\begin{array}{c}T=16.33+0.0413X1-0.0225X2+0.3463X3+0.2825X1X2\\ +0.2700X1X3-0.5625X2X3-2.53X{1}^2-2.24X{2}^2-2.31X{3}^2\end{array}$$

where T represents the polysaccharide yield, and X1, X2, and X3 represent liquid-to-solid ratio, ultrasonic power, and water content, respectively. Analysis of variance (ANOVA) results for the regression model are presented in Table 3. The F-value of the model was 81.63 (P < 0.001), indicating high statistical significance. The lack-of-fit F-value was 1.70 (P = 0.3030; >0.05), confirming the model's accuracy and rationality. Among linear terms, only water content (P < 0.05) significantly affected yield in the DES system, whereas ultrasonic power and liquid-to-solid ratio did not. RSM ranked these factors from largest to smallest, as follows: water content > liquid-to-solid ratio > ultrasonic power. Fig. 3 illustrates factor importance and interactions of different factors on the polysaccharide extraction rate, with X2–X3 interactions significantly affecting the extraction rate. This likely stems from water content’s direct influence on the hydrogen bond network's integrity and overall extraction efficiency. Response surface and contour plots, combined with ANOVA, predicted optimal conditions: liquid-to-solid ratio of 25.57 mL/g, ultrasonic power of 299.16 W, and water content of 30.82%. For practical verification of the applicability and operability, conditions were adjusted to 25.50 mL/g (liquid–solid ratio), 300 W (ultrasonic power), and 31.00% (water content). Experimental yield under these conditions was validated, and the GFP extraction was found to be 16.55%, closely matching the predicted value of 16.31%. These results validate the reliability of GFP extraction using DES optimized by BBD-RSM.

Table 3.

UA-DES extracts the results of the difference analysis for the BBD test in GFP.

Source	Sum of Squares	df	Mean Square	F-value	p-value
Model	81.51	9	9.06	81.63	< 0.0001	significant
A-g/mL	0.0136	1	0.0136	0.1227	0.7364
B-W	0.0040	1	0.0040	0.0365	0.8539
C-%	0.9591	1	0.9591	8.64	0.0217
AB	0.3192	1	0.3192	2.88	0.1337
AC	0.2916	1	0.2916	2.63	0.1490
BC	1.27	1	1.27	11.41	0.0118
A2	26.92	1	26.92	242.61	< 0.0001
B2	21.05	1	21.05	189.72	< 0.0001
C2	22.44	1	22.44	202.23	< 0.0001
Residual	0.7767	7	0.1110
Lack of Fit	0.4358	3	0.1453	1.70	0.3030	not significant
Pure Error	0.3409	4	0.0852
Cor Total	82.29	16
R2			0.9906
R2adj			0.9784
R2pre			0.9088
C.V. %			2.56
AP			21.2572

Fig. 3.

Contour lines and response surfaces under the influence of various intersecting factors (A: X1 liquid solid ratio, B: X2 ultrasonic power, C: X3 water content).

3.5. Fourier-transform infrared (FT-IR) analysis of the optimal system

To elucidate interactions in the ChCl-La system, FT-IR spectroscopy was used to characterize the spectral features of individual components and their mixture (Fig. 4A). The ChCl spectrum showed sharp characteristic peaks at 3256.35 cm⁻¹ and 3013.85 cm⁻¹, assigned to O-H and C-H stretching vibrations, respectively. The peak at 1482.10 cm⁻¹ corresponded to the C-N single bond. The multiple absorption peaks in the 1300–1000 cm⁻¹ range were attributed to peaks such as C-O stretching, C-C, and C-N bond vibrations in quaternary ammonium salts of choline molecules [31]. The La spectrum exhibited a sharp carbonyl (C=O) stretching peak at 1748.84 cm⁻¹. In the ChCl-La system, the O-H stretching band shifted to 3282.41 cm⁻¹ and broadened (blue shift), indicating that the DES system altered intermolecular hydrogen bonding between LA and ChCl [32]. These spectral changes confirm a restructured molecular environment in ChCl-La, strengthening the hydrogen bond network, enhancing polysaccharide solubility, and enabling efficient GFP extraction.

Fig. 4.

A: FT-IR spectra of various DES systems(Chcl; La; Chcl + La); B: Monosaccharide composition results of three GFPs; C: FT-IR spectra of three GFPs; D: Molecular weight distribution of the three polysaccharides HPGPC (A-GFP: HEW; B-GFP: UAE; C-GFP: UA-DES).

3.6. Differences in polysaccharide structure from different extraction methods

3.6.1. Differences in yield and purity

The extraction yields for A-GFP and B-GFP were low (8.10% and 10.00%, respectively, whereas C-GFP reached 15.30%, which is 1.53–1.88 times higher (Table 1). Polysaccharide purity was highest with UA-DES extraction (45.14%) and lowest with HWE (28.51%). This superior performance of the DES system likely stems from its disruption of intermolecular hydrogen bonds, van der Waals forces, and EPE. These mechanisms not only disrupt the hydrogen bond network linking polysaccharides to cell walls but also enhance polysaccharide solubility and extraction efficiency.

3.6.2. Differences in monosaccharide composition

Monosaccharides are the basic units of polysaccharides, and their types and proportions directly influence the physicochemical properties, spatial configuration, and biological activities (including antioxidant and immunomodulatory effects) of polysaccharides. Accurately determining monosaccharide composition is essential for elucidating polysaccharide properties and structure, as well as understanding structure–activity relationships [33]. The monosaccharide compositions and proportions of the three polysaccharides are presented in Fig. 4B and Table 4. Notably, the three extraction methods altered monosaccharide proportions but not their types. This confirms that GFP comprises Man, Rha, GlcA, Glc, Gal, and Ara. Notably, the monosaccharide composition ratio of C-GFP differed significantly from those of A-GFP and B-GFP. In terms of proportions, A-GFP exhibited the highest GlcA content (45.23%), possibly due to high-temperature conditions disrupting the original polysaccharide structure. The Gal and Ara percentages in C-GFP were markedly higher than in A-GFP, likely attributable to strong hydrogen bonding in the DES system. In DES, acidic components (such as malic acid and La) and ionic components (such as ChCl) significantly modify the polysaccharide structure, altering molecular arrangements and monosaccharide proportions [15]. Furthermore, researchers have reported that intermolecular interactions in DES disrupt glycosidic bonds, converting six-membered ring monosaccharides (Gal) into five-membered ring forms (Ara) [34]. This explains why A-GFP shows the highest Glc and GlcA proportions, with the lowest levels of the other four monosaccharides. These results suggest that tailoring DES composition could enable targeted modifications of polysaccharide monosaccharide profiles for future applications.

Table 4.

The proportions of monosaccharides of GFP obtained by three different extraction methods (A-GFP: HEW; B-GFP: UAE; C-GFP: UA-DES).

	Man	Rha	Glca	Glc	Gal	Ara
A-GFP	2.24%	3.61%	45.23%	16.15%	19.65%	13.11%
B-GFP	4.32%	5.50%	25.96%	11.74%	30.79%	21.69%
C-GFP	5.62%	6.95%	8.71%	11.13%	37.78%	29.80%

3.6.3. Differences in FT-IR spectra

FT-IR spectroscopy is a key method for analyzing polysaccharide structures, as it simultaneously records responses across multiple wavelengths [35]. The FT-IR spectra of A-GFP, B-GFP, and C-GFP are shown in Fig. 4C. All three polysaccharides exhibited typical absorption peaks for polysaccharides with substantial peak overlap, indicating strong structural similarities. Seven characteristic peaks were evident: the O-H stretching vibrations at 3395.31, 3381.75, and 3380.20 cm⁻¹; C-H stretching vibrations at 2954.18, 2935.79, and 2928.45 cm⁻¹; C=O stretching vibrations around 1750 cm⁻¹; –OH and –CH₂ stretching vibrations around 1620 and 1420 cm⁻¹, respectively; C-O-C and pyran ring peaks at 1100–1000 cm⁻¹; α-glycosidic bond peak around 769 cm⁻¹; and a weak β-glycosidic bond peak around 840 cm⁻¹ [19]. The most notable difference was the carboxyl absorption band near 920 cm⁻¹, which appeared prominently in A-GFP and B-GFP but was absent in C-GFP. This aligns with variations in monosaccharide composition, confirming that different extraction methods yield polysaccharides with distinct functional groups.

3.6.4. Differences in molecular weight

Molecular weight is a defining structural feature of polysaccharides, closely linked to their solubility, branching, and biological activity [36]. Polysaccharides are broadly classified by molecular weight into high- and low-molecular-weight polysaccharides. Typically, high-molecular-weight polysaccharides have a relative molecular weight of > 100 kDa, exhibit good gelation and water retention properties, but exhibit poor solubility. In contrast, low-molecular-weight polysaccharides have good solubility and biological activity, attributable to their ability to easily cross the cell walls to exert effects [37]. The molecular weight distributions of the three extracted polysaccharides are presented in Fig. 4D and Table 5. Notably, A-GFP displayed four peaks, with Peak 1 at 3304.35 kDa and Peak 2 at 1138.54 kDa. Its maximum molecular weight substantially exceeded those of B-GFP (1497.41 kDa) and C-GFP (361.71 kDa), by factors of 2.2 to 9.1. These results suggest that HWE promotes polysaccharide polymerization, whereas ultrasonic and DES extractions induce cavitation and chemical bond disruption, breaking glycosidic linkages and causing polysaccharide degradation [38]. Overall, the tested extraction methods significantly alter polysaccharide molecular weight, affecting their distribution.

Table 5.

The molecular weights (HPGPC) of GFP obtained by three different extraction methods(A-GFP: HEW; B-GFP: UAE; C-GFP: UA-DES).

	Peak1(KDa)	Peak2(KDa)	Peak3(KDa)	Peak4(KDa)
A-GFP	3304.35	1138.54	28.61	11.25
B-GFP	1497.40	12.57
C-GFP	361.70	91.92	32.32	10.91

3.7. Simulation study of the extraction method

Monosaccharide composition analysis revealed C-GFP as a heteropolysaccharide dominated by Gal. Hence, Gal was used as a reference standard in this experiment. DFT calculations compared binding energies and bond lengths between the DES system with Gal and H2O with Gal, elucidating molecular-level differences between UA-DES and HWE extractions.

Interactions within the systems are shown in Fig. 5. The HBD was the –COOH group in La, and the HBA was the –OH group in ChCl. Simulations revealed a strong bond between –OH and –COOH (length: 1.762 Å) in the ChCl-La system. Upon Gal binding in UA-DES, this bond length increased to 2.204 Å, compared to 1.883 Å in HWE. Binding energies (Table 6) of the ChCl-La system and HWE were − 20.14 and − 14.97 eV, respectively, indicating that ChCl-La forms stronger hydrogen bonds with GFP and exhibits lower binding energy. These findings validate the optimized UA-DES system for GFP extraction.

Fig. 5.

The results of molecular simulation and bond lengths for the Chcl-La and HWE systems.

Table 6.

The binding energies between DES and HWE and the characteristic monosaccharide (Gal).

E0 (HBA)	E0 (HBD)	EN(HBA + HBD)	W (Ha)	W (Ev)
−686.51 (Gal)	1093.44(ChCl + La)	−1780.69	−0.74	−20.14
−686.51 (Gal)	−75.96 (H2O)	−763.02	−0.55	−14.97

EPE analysis further characterized non-covalent interactions in multiple systems, including the ChCl-La system, enabling direct visualization of charge distributions (Fig. 6) [39], [40]. EPE maps depicted the ChCl-La system and Gal conformations (Fig. 6A–C). Chlorine and oxygen atoms showed electronegative regions, while nitrogen and hydrogen atoms displayed positively charged regions. Polysaccharide hydrogen atoms expanded the positively charged region around chloride ions, with mutual attraction of negative regions stabilizing the ChCl-La system. In contrast, HWE (Fig. 6D) represented greater fluctuations in positive and negative charge regions than those of the UA-DES system, suggesting instability of extracted polysaccharides. Overall, this study demonstrates a green, efficient GFP extraction method aligned with green chemistry principles.

Fig. 6.

EPE of Chcl-La and HWE systems (A:ChCl-La; B: ChCl-La + Gal; C:Gal; D:H₂O + Gal).

3.8. In vitro activity analysis

3.8.1. Evaluation of in vitro antioxidant activity

Prolonged oxidative stress damages cells and organs, contributing to various diseases and apoptosis. 2,2-Diphenyl-1-picrylhydrazyl (DPPH) radical scavenging rates for A-GFP, B-GFP, and C-GFP are shown in Fig. 7A. All three polysaccharides exhibited strong scavenging activity (54.12–61.97%) in a concentration-dependent manner. From 1 to 3 mg, the scavenging ability rose steeply with increasing concentration, and plateaued at 5–10 mg, indicating the concentration dependence of these polysaccharides against DPPH scavenging within the test range. Scavenging rates followed the order A-GFP > C-GFP > B-GFP, with A-GFP peaking at 61.97%. Polysaccharides exert antioxidant activity by scavenging free radicals and ions by direct binding. Generally, this ability correlates positively with uronic acid content. Polysaccharides with higher uronic acid content have stronger scavenging abilities, but this is not the only influencing factor. Studies have associated Ara, Man, and Glc with enhanced antioxidant capacity [41]. A-GFP’s superior performance likely stems from its highest proportion of uronic acid and glucose, coupled with a narrow molecular weight distribution. Between B-GFP and C-GFP, C-GFP showed stronger activity due to elevated Gal and Ara levels, along with its much lower molecular weight range than that of A-GFP and B-GFP. Lower molecular weight increases exposed surface area, facilitating greater free radical binding and its scavenging. These findings show that polysaccharide antioxidant capacity often depends on its structural characteristics, such as monosaccharide composition and ratios, molecular weight, and glycosidic bond type [42]. Hence, different extraction methods can tailor polysaccharides for specific applications.

Fig. 7.

The DPPH scavenging rate and α-glucosidase inhibitory rate of GFP obtained through three different extraction methods (A-GFP: HEW; B-GFP: UAE; C-GFP: UA-DES; A: DPPH scavenging rate; B: α-glucosidase inhibitory rate).

3.8.2. Evaluation of in vitro hypoglycemic activity

α-Glucosidase plays a key role in intestinal starch and glycogen digestion/absorption, regulating carbohydrate metabolism. Inhibiting its activity delays glucose metabolism and absorption, exacerbating diseases such as glucose metabolism disorders [43]. This study assessed the hypoglycemic effects of A-GFP, B-GFP, and C-GFP via α-glucosidase inhibition assays (Fig. 7B). The higher the inhibition rate, the better the hypoglycemic effect. Inhibition rates of A-GFP, B-GFP, and C-GFP were 50.16%, 55.87%, and 59.54%, respectively. Studies report positive correlations of Man, Rib, and Gal with hypoglycemic activity, but negative correlations for Ara and GlcA. In C-GFP, the content of GlcA was significantly lower than that in the other polysaccharides, while that of Gal was significantly higher. The differences in hypoglycemic activity may be attributed to these variations in the proportion of monosaccharides. Overall, GFPs show strong potential as hypoglycemic adjuvants with promising development prospects.

4. Conclusion

HWE of polysaccharides often presents limitations of low yields, long extraction times, and incomplete structural preservation of polysaccharides, necessitating sustainable and more effective alternatives. This study, therefore, established a UA-DES extraction approach for GFP extraction, leveraging non-covalent interactions and cavitation effects. Optimization identified ChCl-La as the optimal solvent for extraction at the following parameters: liquid-to-solid ratio, 25.57 mL/g; ultrasonic power, 299.16 W; water content, 30.82%; and extraction time, 30 min. FT-IR confirmed DES system formation, with UA-DES yield being 1.53–1.88 times higher than HWE and UAE. Building upon this optimization, structural analyses revealed method-specific differences in the chemical structure of polysaccharides, reflecting subtle changes in the polysaccharides and synergistic effects of combining ultrasound with the DES system. These effects enhance cell wall penetration and hydrogen bond solvation. DFT calculations of optimized EPE and bond length revealed that the established UA-DES system is an efficient and environmentally optimized extraction method. Finally, in vitro assays demonstrated strong antioxidant and hypoglycemic activities of all three polysaccharides.

Although the UA-DES method exhibited extremely strong polysaccharide extraction capabilities and achieved the highest α-glycosidase inhibition reflecting hypoglycemic activity, the antioxidant activity was slightly lower than that of HWE, likely due to Glc/GlcA ratios. This study did not explore and evaluate solvent recovery and other compounds; therefore, future work should assess DES recyclability and its environmental impact. Additionally, the versatility and practicality of UA-DES extraction should be explored for other plant polysaccharides, along with its economic sustainability in industrial and food applications.

CRediT authorship contribution statement

Weijie Yang: Writing – original draft, Methodology, Data curation, Conceptualization. Shuang Jiang: Writing – original draft, Visualization, Validation, Formal analysis, Data curation. Xiaotian Wu: Writing – review & editing, Supervision, Funding acquisition. Yuxin He: Investigation, Data curation. Changcheng Peng: Methodology. Chenglu Zhu: Validation, Data curation. Ruitong Du: Methodology. Huiyue Yuan: Methodology, Formal analysis. Shunfeng Lan: Methodology. Zejun Liu: Methodology. Yimai Liu: Validation. Haixue Kuang: Supervision, Methodology. Zhibin Wang: Writing – review & editing, Supervision, Methodology, Funding acquisition, Conceptualization.

Funding

All work was supported by Heilongjiang Provincial Key Research Plan (Grant №. GA22B012).

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

The following are the Supplementary data to this article:

Data availability

Data will be made available on request.