Ultrasound-assisted extraction optimization of Fructus Tribuli polysaccharides: How stir-frying processing alters structures and enhances antihypertensive efficacy

Graphical abstract

Highlights

• •Optimized UAE of Fructus Tribuli polysaccharides (FP) via RSM achieved a high yield of 2.18%.
• •Stir-frying altered FP's structure, increasing uronic acid content and narrowing MWD.
• •SFP exhibited superior antihypertensive efficacy over FP in SHRs.
• •SFP alleviated hypertension by modulating gut microbiota, increasing SCFAs.
• •SFP may alleviate hypertension by inhibiting the vascular LPS-TLR4/MyD88 pathway.

Abstract

Fructus Tribuli (FT), the dried ripe fruit of Tribulus terrestris L., is recognized for its antihypertensive properties, which are enhanced by stir-frying without adjuvants. However, research on its polysaccharides with therapeutic potential remains limited. This study optimized the ultrasound-assisted extraction (UAE) of FT polysaccharides (FP) using response surface methodology, yielding a maximum extraction yield of 2.182 ± 0.29% under the conditions of liquid–solid ratio 20:1 mL/g, 51 min, 62 °C, and precipitation ethanol concentration of 95%. Subsequently, FP and stir-fried FT polysaccharides (SFP) were isolated, and their structural differences and antihypertensive effects were systematically compared. Structural analysis revealed notable differences between FP and SFP. Furthermore, the extraction resulted in a higher polysaccharide yield for SFP. In spontaneously hypertensive rats, both FP and SFP attenuated hypertension and vascular injury, modulated gut microbiota, increased short-chain fatty acids, and enhanced intestinal barrier function effects that were more pronounced with SFP. Mechanistically, both polysaccharides inhibited the aortic TLR4/MyD88 pathway. These results suggest that stir-frying modifies polysaccharide structure, thereby improving gut microbiome regulation, barrier protection, and vascular outcomes, highlighting the value of processing in enhancing polysaccharide efficacy. Thus, stir-frying amplifies therapeutic effects through bioactive macromolecular remodeling, advancing the understanding of Traditional Chinese medicine processing principles.

1. Introduction

Traditional Chinese medicine processing (TCMP), as a core pharmaceutical technology in the field of Chinese medicine, plays an irreplaceable role in ensuring drug quality and clinical efficacy, significantly influencing the chemical composition, therapeutic effects, and safety profile of medicinal materials [1]. With polysaccharides from natural sources gradually becoming a research hotspot and with the rapid development of the technology of glycan structure analysis, numerous studies have found that TCMP affects the polysaccharide structure, which, in turn, affects pharmacological activity [2]. Different processing approaches, including steaming, treatment with vinegar or wine, and stir frying without adjuvant materials, can markedly influence the physicochemical and structural characteristics of the extracted polysaccharides, encompassing their molecular weight, monosaccharide composition, functional groups, microstructure, and conformation [3], [4]. However, the structural analysis of polysaccharides and changes in polysaccharides after stir-frying without adjuvant materials in other TCMs have seldom been reported [5], [6], [7], [8].

Fructus Tribuli (FT), the dried ripe fruit of Tribulus terrestris L., has been frequently used as a traditional medicine for thousands of years (Fig. 1A). FT exerts antitumor, hypolipidemic, and antioxidant effects and has been reported to lower blood pressure in hypertensive rats [9]. Preliminary studies have demonstrated that FT extract can reduce systolic blood pressure (SBP) in spontaneously hypertensive rats (SHRs) by ameliorating vascular endothelial dysfunction (ED), with stir-fried Fructus Tribuli (SFT) showing higher efficacy than FT [10], [11]. Previous studies have reported the successful isolation of polysaccharides from FT [12], [13], however, their biological activities and pharmacological mechanisms remain unexplored. Our preliminary findings have demonstrated that both FT polysaccharide (FP) and SFT polysaccharide (SFP) can reduce blood pressure in SHRs, with SFP exhibiting significantly higher efficacy than FP in lowering plasma lipopolysaccharide (LPS) and interleukin (IL)-6 levels (p < 0.05).

Fig. 1.

(A) The aerial parts of Tribulus terrestris L. (B) Flowchart of FP and SFP preparation.

Essential hypertension (EH), accounting for 90–95% of hypertension cases, remains a significant global health burden. Growing evidence implicates gut microbiota dysbiosis in its pathogenesis [14], [15], [16], [17]. Studies have revealed that SHRs (a classical model of EH) exhibit reduced diversity and richness of the gut microbiota compared with normotensive Wistar-Kyoto (WKY) rats [18], [19]. Furthermore, SHRs demonstrate altered permeability of the intestinal epithelial barrier, facilitating LPS translocation into systemic circulation [20]. LPS binding to vascular endothelial Toll-like receptor 4 (TLR4) triggers MyD88/NF-κB signaling, promoting pro-inflammatory cytokine production (TNF-α, IL-6) that disrupts endothelial function and promotes vasoconstriction, thereby exacerbating hypertension [21], [22]. Conversely, gut microbiota-derived short-chain fatty acids (SCFAs) help maintain intestinal barrier integrity and exhibit anti-inflammatory effects, with studies showing an inverse correlation between SCFA levels and blood pressure[23], [24]. Concurrently, they play a crucial role in regulating blood pressure, with both clinical investigations and animal studies demonstrating an inverse correlation between circulating SCFA levels and the severity of hypertension [25].

Ultrasound-assisted extraction (UAE) has emerged as an efficient and environmentally friendly technique for the extraction of plant polysaccharides, offering notable advantages over conventional methods that often suffer from low efficiency and prolonged processing time[26]. The mechanical effects induced by ultrasonic cavitation effectively disrupt plant cell walls, facilitating solvent penetration and significantly enhancing the dissolution and extraction efficiency of polysaccharides. When performed under optimized parameters, UAE not only helps preserve the structural integrity of polysaccharides but may also enhance their bioactivity[27], [28]. To systematically optimize the extraction process, response surface methodology (RSM) serves as a reliable statistical strategy for establishing mathematical models between key process variables and extraction outcomes, enabling accurate prediction and control of polysaccharide extraction. This approach provides a solid foundation for the efficient extraction and subsequent application of polysaccharides[29].

Based on this evidence, we hypothesized that SFP may enhance antihypertensive efficacy by modulating the gut microbiota and ameliorating LPS/TLR4-mediated intestinal pathology. To validate this hypothesis, FP and SFP were isolated from raw and stir-fried FT, respectively, based on the optimized UAE conditions established by RSM. The obtained polysaccharides were then characterized in terms of content, molecular weight distribution, and monosaccharide composition. Stir frying–induced structural changes were systematically compared. The mechanisms underlying the efficacy of FP/SFP in alleviating EH were further investigated based on our findings of the blood pressure–lowering effects and microbiota-modulating activities of FP and SFP in SHRs.

2. Materials and methods

2.1. Materials and reagents

FT (2209300132) was purchased from Bozhou Huqiao Pharmaceutical Co., Ltd. (Anhui, China). A nitric oxide (NO) kit (20230527) and total protein quantitative assay kit (A045-2–2) were purchased from Nanjing Jiancheng Bioengineering Institute (Nanjing, China), and a D-lactic acid kit (D-LA, BC5355) was bought from Beijing Solarbio Science & Technology Co., Ltd. (Beijing, China). ET-1 (L220607703), Ang II (L230220891), endothelial nitric oxide synthase (eNOS, L230529479), and LPS (L230618650) kits were procured from Cloud-Clone Corp. Wuhan (Wuhan, China). Rat TNF-α (GER0004) and IL-6 (GER0001) kits, and Occludin (GB111401), Claudin-1 (GB11032), and β-actin (AC220227007) antibodies were purchased from Wuhan Servicebio Technology Co., Ltd. (Wuhan, China).

TLR4 polyclonal antibody (19811–1-AP), GAPDH (10494–1-AP) antibody, goat anti-mouse immunoglobulin (Ig)G (H + L) (SA00001-1), and goat anti-rabbit IgG (H + L) (SA00001-2) were purchased from Proteintech Group (Wuhan, China). TLR4 antibody (sc-293072) was purchased from Santa Cruz Biotechnology (Santa Cruz, CA, USA). MyD88 (AB219413), IκBα (AB32518), and NF–κB p65 (AB32536) antibodies were obtained from Abcam (Cambridge, UK). IKKβ antibody (#2678) was obtained from Cell Signaling Technology (Danvers, MA, USA). Standard acetic acid (D2313946), propionic acid (D2308396), butyric acid (C2328179), isobutyric acid (F2229483), pentanoic acid (E2318806), and isovaleric acid (K2215881) were purchased from Shanghai Aladdin Biochemical Technology Co., Ltd. (Shanghai, China).

2.2. UAE optimization

2.2.1. Single-factor experiments

The effects of liquid–solid ratio, ultrasonic time, ultrasonic temperature, and precipitation ethanol concentration on the yield of FP were determined in a one-way test by weighing 10 g of FT powder. The one-factor test conditions were as follows: liquid-to-solid ratio of 5:1, 10:1, 15:1, 20:1, 25:1; ultrasonic time of 20 min, 30 min, 40 min, 50 min, 60 min; ultrasonic temperature of 40 ℃, 50 ℃, 60 ℃, 70 ℃, 80 ℃; precipitation ethanol concentration of 75%, 80%, 85%, 90%, 95%.

2.2.2. RSM experiments

According to the Box-Behnken principle, four factors, namely, the desired liquid–solid ratio, ultrasonic time, ultrasonic temperature, and precipitation ethanol concentration have been organized as varying conditions that carried out the experimental design shown in Table S1. The combined impact of these three factors on yield was mainly examined[30].

2.3. SFT and FP preparation

Briefly, cleaned FT was placed in a preheated stir-frying machine (MK-30, Jiangyin Zhutang Mingke Machinery Factory, Jiangyin, China) and stir-fried at 200 ± 5℃ for 11 min at 60 r·min⁻¹.

After the FT and SFT powders were extracted using ultrasonication for 50 min with 20 volumes of pure water at 60 ℃, the products were filtered, and the filtrates were collected. The extraction process was repeated; the filtrates were combined and concentrated under reduced pressure. Ethanol (95%) was added to the concentrated extract to 85%, stirred, and left undisturbed at 4 °C for 12 h. After filtering, the precipitate was dried until it was odorless. The process of precipitation was repeated, and the obtained precipitate was dried. These sequential extractions with ethanol were conducted to eliminate pigments and small, fat-soluble molecules. The precipitates that were obtained were resuspended in 800 mL of pure water and mixed with Sevage reagent in a 4:1 ratio (solution: Sevage reagent). The mixture was shaken vigorously for 30 min and allowed to separate. The upper layer of the solution was retained. This process was performed 8 times. The final solution was dialyzed using a 3.0 kDa dialysis bag for 48 h. The sugar solution in the bag was collected and freeze-dried to obtain crude FP and SFP.

2.4. Analysis of the total polysaccharides in FP and SFP

2.4.1. Determination of polysaccharide, uronic acid and protein content

The phenol–sulfuric acid method was used to quantify the total polysaccharides in FP and SFP. The highest absorbance was observed at 490 nm, and D(+)-Glucose was used as an anhydrous reference material. The test solution or reference solution (1 mL) was added to a 10-mL test tube and mixed with 6% phenol and concentrated sulfuric acid. The reaction was terminated by boiling for 30 min and cooling on ice to room temperature.

Uronic acid content was determined using the m-hydroxydiphenyl method. The absorbance at 525 nm was measured using D-galacturonic acid as a reference standard. Protein content was determined at 595 nm using the Coomassie Brilliant Blue method and a total protein quantitative assay kit, following the specific procedures outlined in Table S2.

2.4.2. Ultraviolet (UV) and Fourier-transform infrared (FT-IR) spectroscopy

The UV spectra of FP and SFP were acquired using a UV–visible spectrophotometer (Thermo Co. Ltd., MA, USA). The UV–visible absorption spectra were recorded in a wavelength range of 200–400 nm.

FT-IR spectroscopy was performed using a Frontier FT-IR spectrometer (PerkinElmer, Inc., MA, USA). Dried FP and SFP were mixed with KBr powder for FT-IR spectroscopy in the wavelength range of 4000–400 cm⁻¹.

2.4.3. Molecular weight analysis using size-exclusion chromatography with multiangle light scattering and refractive index (SEC-MALS-RI)

Samples were dissolved in 0.1 M aqueous NaNO₃ containing 0.02% NaN₃ at a concentration of 1 mg/mL and filtered through a 0.45-μm filter. Molecular weight and homogeneity were determined using SEC-MALS-RI using a DAWN HELEOS-II laser photometer (Wyatt Technology Co., CA, USA) equipped with two tandem columns (300 × 8 mm, Shodex OH-Pak SB-805 and 803; Showa Denko K.K., Tokyo, Japan) and simultaneously connected to a differential refractive index detector (Optilab T–rEX, Wyatt Technology Co., CA, USA). The mobile phase was 0.02% NaN₃ in 0.1 M NaNO₃ (A), and isocratic elution was used at 45℃. The flow rate was 0.6 mL/min.

2.4.4. Monosaccharide composition analysis using high-performance anion-exchange chromatography (HPAEC-PAD)

FP and SFP were completely hydrolyzed, and their monosaccharide composition was analyzed using a previously published method [31]. Briefly, FP and SFP were hydrolyzed at 121℃ for 2 h after dissolving with 1 mL of 2 M trifluoroacetic acid. After hydrolysis, the sample was dried under a stream of nitrogen, washed 3 times with methanol, and blow dried. The residue was redissolved in deionized water and filtered through 0.22-μm microporous filtering film for analysis. Polysaccharide content was determined using HPAEC linked to a pulsed amperometric detector (PAD; Dionex ICS 5000 + system). H₂O (A)/0.1 M NaOH (B)/0.1 M NaOH, 0.2 M NaAc (C) was used as the mobile phase at a flow rate of 0.5 mL/min and using gradient elution to obtain the monosaccharides. The elution gradients are shown in Table S3. A Dionex™ CarboPac™ PA20 (150 × 3.0 mm, 10 μm) was used at 40℃. Details for the 13 monosaccharide standards are presented in Table S4 and S5.

2.4.5. Scanning electron microscopy (SEM)

SEM (QuantaTM 450PEG, FEI Technology Co., USA) was used to determine the surface morphological features of FP and SFP. FP and SFP powders were sputter-coated with gold under reduced pressure and observed at magnifications of 300×, 600×, and 1000 × .

2.5. In vivo antihypertensive activity

2.5.1. Animals

Eleven-week-old healthy male SHRs (n = 32) and male WKY rats (8) weighing 250 ± 10 g were provided by Beijing Vital River Laboratory Animal Technology Co., Ltd. (license number: SCXK2021-0006, Beijing). All rats were housed in our laboratory at Shandong University of Traditional Chinese Medicine (SDUTCM) at constant humidity and temperature in a specific–pathogen-free grade rearing room. All animals were exposed to a 12-h/12-h daily light/dark cycle and provided access to food and water ad libitum (ethical approval number SDUTCM202101005). All animal experiments were performed following the guidance of the Experimental Animal Ethics Committee of SDUTCM. All animal experiments also complied with the National Research Council’s Guide for the Care and Use of Laboratory Animals.

2.5.2. Experimental design

After 1 week of adaptation, 32 SHRs were randomly assigned to the following four groups: 1) model group (SHR, n = 8), 2) valsartan (VAL) group (14.4 mg/kg, n = 8), 3) FP group (400 mg/kg, n = 8), and 4) SFP group (400 mg/kg, n = 8) [32]. Next, 8 WKY rats were assigned to the control (WKY) group. Rats in the model and control groups were administered an equal volume of distilled water by gavage at a fixed time between 8:00 AM and 10:00 AM daily for 5 weeks. Body weights were measured every 7 days to adjust the dosing amount accordingly.

2.5.3. Blood pressure measurements

After drug administration, SBP was measured at a consistent time point 2 h later every Sunday using a fully automated noninvasive blood pressure monitor (BP-2000 AUL, Visitech Systems, Inc., NC, USA).

2.5.4. Enzyme-linked immunosorbent assay (ELISA)

Blood samples were collected from the abdominal aorta of rats after 5 weeks of drug administration. The blood was centrifuged at 3000 rpm for 10 min to separate the serum. Serum IL-6, TNF-α, NO, ET-1, eNOS, D-LA, Ang Ⅱ, and LPS levels were determined using the corresponding ELISA kits.

2.5.5. Histological evaluation using hematoxylin and eosin (H&E) and immunohistochemical (IHC) staining

Blood vessels and colon tissues were fixed in 10% paraformaldehyde, dehydrated in an automatic dehydrator, and embedded in paraffin. Wax blocks were sectioned using a slicer. H&E staining was used to analyze blood vessels and colon lesions. Masson staining was used to determine vascular collagen deposition. Occludin and Claudin-1 expression in the intestines and TLR4 levels in the blood vessels were determined using a previously described method [10].

2.5.6. Western blotting

Total proteins from blood vessels were extracted and separated using sodium dodecyl sulfate–polyacrylamide gel electrophoresis for western blotting. The protein bands were incubated overnight with the primary antibody at 4℃. Immunoblotting was performed with species-matched secondary antibodies. The blots were treated with enhanced chemiluminescence reagent and visualized using a KQ imager. All images were analyzed using ImageJ software.

2.5.7. Quantitative real-time polymerase chain reaction (qRT-PCR)

Following anesthesia and euthanasia, the rat thoracic cavity was opened to expose the thoracic aorta. A segment from the aortic arch to the abdominal aortic bifurcation was excised. After removing perivascular adipose and connective tissues and rinsing the lumen with ice-cold saline, approximately one‑third of the vessel was immersed in RNA stabilization solution for 24  h and stored at –80 ℃.

Total RNA from aortic tissues was extracted using SparkZol Reagent (Sparkjade®, Jinan, China). About 50  mg of aortic tissue was minced, placed in a pre‑chilled nuclease‑free tube with 0.5  mL TRIzol and grinding beads, and homogenized on ice using a high‑throughput tissue homogenizer (CK-5F-3D, Wuhan Servicebio Technology Co., Ltd., Wuhan, China) until the lysate became transparent. After 10  min at room temperature, 0.2  mL chloroform was added, followed by vortexing for 40  s and incubation for 3  min. The sample was centrifuged at 4 ℃ and 12,000  rpm for 10  min, and the upper aqueous phase was collected as total RNA. The whole RNA was reverse-transcribed into cDNA using a SPAPKscript Ⅱ RT Plus kit (with gDNA Eraser). Real-time PCR was performed using 2 × SYBR Green qPCR Mix (with ROX) and cDNA templates from the LightCycler® 480 Real-time PCR system. A list of the forward and reverse primers that were used is listed in Table S6 (designed and synthesized by Wuhan Servicebio Technology Co., Ltd.). RNA enrichments were calculated using the 2^−ΔΔct method. Data were analyzed as fold induction compared with the control.

2.5.8. Gut microbiota analysis

The contents of the cecum of rats were collected and stored at –80℃. Total DNA was obtained after total genomic DNA extraction from the microbial community according to the instructions in the Magnetic Soil and Stool DNA kit (Tiangen, Beijing, China). The V3–V4 variable region of the bacterial 16S rRNA gene was selected as the PCR amplification target, and the amplification sequence was CCTAYGGGRBGCASCAG, GGACTACNNGGGTATCTAAT. Libraries were prepared using an NEB Next® Ultra™ II FS DNA PCR-free Library Prep kit (New England Biolabs, MA, USA), which were then quantified using Qubit and Q-PCR. DNA sequencing was performed using a NovaSeq 6000 platform (Novogene, Beijing, China). Final data analyses were conducted using the Novogene Co., Ltd. platform (https://cn.novogene.com), and the Wilcoxon rank-sum test was used to statistically analyze bacterial microflora–related data.

α-diversity indices (Observed OTUs, Shannon, Simpson, Chao1, Goods_coverage, Dominance, Pielou_e) were calculated in QIIME 2 after rarefying OTU tables to 10,000 sequences per sample. β-diversity analysis was conducted using both weighted and unweighted UniFrac distances in QIIME 2. Principal component analysis (PCA) was performed on centered log-ratio (CLR) transformed abundance data using the ade4 package in R (version 4.0.3). Principal coordinate analysis (PCoA) was based on Bray-Curtis distance matrices with statistical testing via PERMANOVA (Adonis, 999 permutations), visualized using ggplot2.

2.5.9. Extraction and analysis of SCFAs

Cecal contents (approximately 0.2 g) were mixed with 3 mL of 0.005 M aqueous NaOH solution, vortexed for 30 s, sonicated for 3 min, and centrifuged at 14,000 rpm for 10 min at 4 °C. Next, 2 mL of the supernatant was pipetted and added to 0.2 mL of 50% sulfuric acid and 2 mL of diethyl ether. The mixture was shaken vigorously for 5 min and centrifuged for 10 min at 12,000 rpm (4 °C). The supernatants were filtered through a 0.22-µm filter.

Samples were analyzed using gas chromatography linked to a flame ionization detector (FID) equipped with an Inter Cap Wax column (30 m × 0.32 mm × 0.25 μm). For sample heating, the temperature was initially set to 135℃ for 1 min and then increased to 140℃ at a rate of 20℃/min. The temperature was then increased to 170℃ at a rate of 3℃/min and maintained for 2 min. FID detector settings: temperature = 300℃; split ratio = 5:1; final injection volume = 2 μL.

2.6. Statistical analysis

Data are expressed as mean ± standard deviation, and all experiments were performed in triplicate. GraphPad Prism 9.5.0 was used for all statistical analyses. Statistical significance was determined using one-way analysis of variance. Results were considered significant at p < 0.05.

3. Results and discussions

3.1. UAE optimization of FP

3.1.1. Single-factor optimization

Fig. 2 illustrates the influence of four critical parameters: liquid–solid ratio, ultrasonic time, ultrasonic temperature, and precipitation ethanol concentration on FP yield.

Fig. 2.

Effect of different extraction parameters on the yield of FP. (A) liquid–solid ratio; (B) ultrasonic time; (C) ultrasonic temperature; (D) precipitation ethanol concentration.

As illustrated in Fig. 2A, the polysaccharide yield initially increased and then decreased as the liquid–solid ratio varied from 5:1 to 25:1 (mL/g). A moderate increase in solvent volume enhances the concentration gradient between solid and liquid phases, facilitating polysaccharide dissolution; however, excessive solvent may reduce mass transfer efficiency [33]. The maximum yield of 2.08 ± 0.03% was achieved at a liquid–solid ratio of 20:1 (mL/g).

As shown in Fig. 2B, the yield reached its peak at an extraction time of 50 min. Prolonging the extraction beyond this point resulted in a decline in yield, likely due to polysaccharide chain scission induced by prolonged ultrasonication or thermal degradation caused by accumulated mechanical energy[34].

Fig. 2C clearly demonstrates the significant influence of extraction temperature on the FP yield. Increasing the temperature from 30 °C to 60 °C led to a rise in yield from 1.67 ± 0.01% to 1.99 ± 0.02%, indicating that elevated temperatures promote molecular motion and polysaccharide dissolution. However, to avoid structural degradation of polysaccharides at higher temperatures, 60 °C was selected as the optimal extraction temperature[34].

From Fig. 2D, it is observed that the polysaccharide yield gradually increased with higher ethanol concentration used for precipitation. The most notable increase occurred between 80% and 85%, beyond which the yield stabilized at a high level from 85% to 95%. Therefore, an ethanol concentration of 90% was chosen for subsequent experiments to balance recovery efficiency and cost-effectiveness.

3.1.2. RSM optimization

Based on the results of single-factor experiments, response surface methodology (RSM) was employed using a Box-Behnken design (BBD) with four factors and three levels, generated via Design-Expert software (Version 13.0). A total of twenty-nine experimental runs were conducted, and the corresponding yield of FP was determined experimentally, as shown in Table S7. The obtained data were fitted to a quadratic polynomial model, as expressed as:

Y = 2.14 + 0.04*A + 0.06*B + 0.03*C + 0.08*D-0.01*A*B + 0.02*A*C-0.02*A*D + 0.02*B*C-0.04*B*D + 0.02*C*D-0.14*A²-0.11*B²-0.12*C2.

The ANOVA results (Table S8) indicated that the regression model was highly significant (p < 0.01), with an F-value of 40.74. The R² (0.9760) and R²_Adj (0.9742) values demonstrated excellent model fit. The lack of fit was not significant (p > 0.05), confirming the model's reliability. The significance order of the four variables on the response was D > B > A > C, with the BD interaction exhibiting a highly significant effect.

The distinct curvature of the surfaces indicates considerable and sensitive influences of these variables on the extraction efficiency (Fig. 3). Based on the model generated by Design-Expert software, the predicted optimal extraction conditions were as follows: liquid–solid ratio of 20.46:1 mL/g, ultrasonic time of 51.01 min, temperature of 62.32 °C, and ethanol concentration of 95%. Under these conditions, the predicted FP yield was 2.227%. Considering practical operational precision, the parameters were adjusted to a liquid–solid ratio of 20:1 mL/g, ultrasonic time of 51 min, temperature of 62 °C, and ethanol concentration of 95% for experimental validation. The actual average yield obtained was 2.182 ± 0.29% (n = 3), which closely agrees with the predicted value, demonstrating that the model accurately reflects the extraction behavior of FP and shows potential for industrial application.

Fig. 3.

3D response surface plots and contoured maps of the interaction of different variables on the extraction yield of FP.

3.2. Characterization of FP and SFP

3.2.1. Total polysaccharide content and molecular weight of FP and SFP

The flow chart for the preparation of FP and SFP is shown in Fig. 1B. Crude polysaccharides were extracted using ultrasonication, as this process can avoid the degradation of heat-sensitive polysaccharides and afford higher yields compared with traditional hot–water extraction [35], [36]. The total polysaccharide contents of FP and SFP were determined using the phenol–sulfuric acid method. The standard regression equation for glucose was y = 16.293x + 0.0753, R² = 0.9991 (where y is the absorbance and x is the glucose concentration). The total polysaccharide content of FP and SFP was 52.41% ± 2.13% and 63.55% ± 1.91%, respectively.

To enable a direct comparison of the polysaccharide fractions obtained from FT before and after processing, this study applied the optimal UAE parameters determined for FP to the extraction of SFP. Although the processing technique might influence the optimal extraction conditions by altering cell wall structure, the state of constituent existence, or matrix density[37], the primary aim of this research was to compare the differences in yield, chemical composition, and bioactivity between FP and SFP obtained under a standardized extraction procedure. This strategy of controlling for a single variable is often employed in studies comparing chemical composition or activity before and after processing[38], [39].

The total uronic acid and protein contents of FP and SFP are shown in Table S9. The standard regression equation for D–galacturonic acid was y = 8.7782x + 0.008, R² = 0.9959 (where y is the absorbance and x is the glucose concentration).

3.2.2. Analysis of FP and SFP using UV and FT-IR spectroscopy

The UV spectra of FP and SFP (Fig. 4A) were recorded over a wavelength range of 200–400 nm for the determination of proteins and nucleotides. However, no significant absorption peaks were observed at 260 and 280 nm, indicating the absence of proteins or nucleotides in FP and SFP.

Fig. 4.

Characteristics of FP and SFP. (A) UV spectra of FP and SFP. (B) FT-IR spectra of FP and SFP. (C) SEC-MALS-RI diagrams of FP and SFP. (D) HPLC chromatograms of the monosaccharides in FP and SFP and the standards.

Fig. 4B shows the infrared spectra of FP and SFP. The absorption bands at 3300 and 2900 cm⁻¹ are typical for polysaccharides and are associated with O–H and C–H stretching vibrations in polysaccharides, respectively [5]. The absorption band at 1738 cm⁻¹ can be attributed to the C=O stretching vibration in the uronic acid of SFP, which is not present in FP. This feature, potentially introduced or exposed during the stir-frying process, may contribute to enhanced polysaccharide solubility and altered bioactivity[40]. The extensive absorption peak at approximately 1661 cm⁻¹ corresponds to the bound water, whereas the peak at 1412 cm⁻¹ corresponds to C–H deformation vibration. The absorption peaks at 1248 cm⁻¹ and 1072 cm⁻¹ are associated with the C–O–C and C–O–H groups. A prominent signal is noted at around 1027 cm⁻¹, which corresponds to the stretching vibrations of C–O from the glycosidic bonds or the skeletal modes of the pyranoid rings, indicating the characteristics of polysaccharides. The weak absorption peak at 758 cm⁻¹ corresponds to the skeletal modes of the pyranoid [41], [42].

3.2.3. Molecular weight determination of FP and SFP

Molecular weight analysis of the polysaccharides revealed two peaks for FP, with absolute molecular weights of 151 kDa for peak 1 and 39.86 kDa for peak 2, whereas SFP exhibited a single peak, indicating an absolute molecular weight of 89.41 kDa (Fig. 4C and Table 1). Additionally, FP and SFP had different polydispersity indices (PDIs), a parameter reflecting the breadth of polymer molecular weight distribution. Higher PDIs indicate broader distribution. The PDI of FP was 2.982 ± 0.10, which decreased by 20.57% after stir-frying, demonstrating that SFP had a more uniform molecular-weight distribution after processing. These findings align with the results reported by Yao et al [7].

Table 1.

Contents and molecular parameters of the specific polysaccharides in FP and SFP.

Sample	Content /%	peak	Mw/kD	Mn/kD	Mw/Mn
FP	52.41 ± 2.13	peak 1	151	103	1.46
FP		peak 2	39.86	36.37	1.10
SFP	63.55 ± 1.91	peak 1	89.41	56.32	1.59

3.2.4. Monosaccharide composition of FP and SFP

As shown in Fig. 4D and Table S10, FP was mainly composed of the 6 monosaccharides Rha, Ara, Gal, Glc, Xyl, and Man in a 0.76: 3.23: 3.74: 1.00: 0.47: 0.37 ratio. SFP was mainly composed of the 7 monosaccharides Rha, Ara, Gal, Glc, Xyl, Man, and GalA in a 0.72: 2.67: 2.62: 1.00: 0.90: 0.37: 0.93 ratio. The contents of Gal and Ara decreased after stir-frying, whereas those of Xyl and GalA increased. Specifically, Gal content decreased by 34.39% and GalA content increased to 9.91%, consistent with the results from FT-IR analysis. Based on these findings, we hypothesized that stir-frying may have oxidized the C-6 hydroxyl group of Gal to a carboxyl group, forming GalA [2], [43]. These findings provide a scientific basis for further investigations into the unique pharmacological effects of SFP.

3.2.5. SEM analysis

FP showed a lamellar structure with a smooth surface and a complete, uniform structure, whereas SFP showed an irregular fragmented structure with more rough and fine spherical particles attached to the surface (Fig. 5). These findings indicate that the structure and morphology of FP changed significantly after stir-frying, the lamellar structure of FP was destroyed after heating and stir-frying, and the lamellar structure of SFP exhibited a fragmented state.

Fig. 5.

Scanning electron microscopy of the polysaccharides under different magnifications.

3.3. Administration of FP and SFP inhibits TLR4/MyD88 signaling and reduces endothelial vessel damage in SHRs

3.3.1. Administration of FP and SFP lowers the blood pressure and improves vascular endothelial function in SHRs

To investigate the antihypertensive effects of FP and SFP, blood pressure was monitored in all experimental groups. A significant reduction in SBP was observed in rats during the 1–5-week course of FP and SFP administration compared with SHRs (p < 0.01, 0.05) (Fig. 6), with this effect remaining stable. Meanwhile, after 4–5 weeks of administration of the test compounds, a significant difference in antihypertensive effect was noted between rats in the FP and SFP groups (p < 0.05). SFP showed superior efficacy to FP, consistent with previous findings that validate the processing-enhanced effect of stir-frying[10], [11].

Fig. 6.

Effect of FP and SFP on the blood pressure of SHRs. Changes in SBP. Compared with the WKY group ^#p < 0.05, ^##p < 0.01; compared with the SHR group *p < 0.05, ^**p < 0.01; compared with the FP group ^▲p < 0.05, ^▲▲p < 0.01, n = 8.

Next, we performed histological analysis of aortic tissues to determine whether the blood pressure reduction was associated with improved vascular morphology. H&E staining (Fig. 7A) revealed a significant thickening of blood vessel walls, shedding of endothelial cells, and an increase in the number of smooth muscle layers, along with disordered cell arrangement in SHRs compared with normal rats. The vascular morphology and structural damage were alleviated after FP administration. In the SFP group, only a slight thickening of the vessel walls was observed, accompanied by occasionally incomplete intimal pathology. As shown in Fig. 7B, collagen deposition in the SHR group was significantly enhanced compared with that in the WKY group (p < 0.01). The FP and SFP groups exhibited reduced collagen deposition compared with that in the SHR group (p < 0.05 or p < 0.01). Additionally, the decrease in collagen deposition in the SFP group was significantly better than that in the FP group (p < 0.05).

Fig. 7.

Effect of FP and SFP on vascular endothelial cells. (A) Representative graphs (×200, ×400) of H&E staining of the blood vessels: shedding of endothelial cells (red arrow), damage and incompleteness of the vascular endothelium (black arrow), and smooth muscle cell hypertrophy (yellow arrow), n = 3. (B) Representative photographs (×200) of Masson’s staining of the thoracic aortas. Average optical density of the blue areas from Masson’s staining, n = 3. (C) Changes in serum Ang Ⅱ, eNOS, NO, and ET-1 levels, n = 8. Compared with the WKY group, ^#p < 0.05, ^##p < 0.01; compared with the SHR group *p < 0.05, ^**p < 0.01; compared with the FP group ^▲p < 0.05, ^▲▲p < 0.01.

ET-1 secreted by endothelial cells is a potent vasoconstrictor. Prolonged exposure to high concentrations of ET-1 can lead to vascular wall thickening and tissue fibrosis. In contrast, NO is a vasodilator that inhibits platelet aggregation. These two substances are key antagonistic factors in regulating vascular tone and smooth muscle cell proliferation, wherein their dynamic balance plays a key role in maintaining vascular homeostasis [44]. We further measured the levels of key regulators of vascular modulation. Significantly lower eNOS and NO (p < 0.01) levels and higher Ang II and ET-1 (p < 0.01) levels were determined in the blood vessels in the SHR group compared with those in the WKY group (Fig. 7C). Both FP and SFP groups exhibited increased serum eNOS and NO (p < 0.01) levels and decreased Ang II and ET-1 (p < 0.01) levels compared with those in the SHR group. Furthermore, compared with those in the FP group, the SFP group demonstrated elevated serum eNOS and NO levels and decreased Ang II and ET-1 levels (p < 0.01, 0.05, respectively).

3.3.2. Administration of FP and SFP inhibits TLR4/MyD88 signaling and reduces endothelial vessel damage in SHRs

TLR4, highly expressed in vascular endothelium, serves as the primary receptor for LPS. Its activation by LPS triggers the MyD88 signaling pathway, leading to phosphorylation and nuclear translocation of NF-κB p65. This cascade upregulates pro-inflammatory cytokines such as TNF-α and IL-6, which impair endothelial function and promote ET-1 secretion. Concurrently, vascular smooth muscle cells are stimulated to produce Ang II via angiotensin-converting enzyme-dependent pathways. The resulting NO/ET-1–Ang II imbalance sustains increased vascular resistance and remodeling, thereby exacerbating hypertension [21], [22]. Findings from IHC staining indicated a significant increase in TLR4 expression in the blood vessels in the model group compared with normal blood vessels (p < 0.01). Both FP and SFP groups exhibited lower TLR4 expression compared with that in the SHR group (p < 0.01). Moreover, the overall effective rate in the SFP group was significantly higher than that in the FP group (p < 0.05) (Fig. 8A).

Fig. 8.

Effect of FP and SFP on the TLR4/MyD88 signaling in blood vessels. (A) Representative images from immunohistochemical staining of TLR4 in blood vessels (×200). (B, C, E) Western blotting to determine the levels of TLR4, MyD88, NFκB p65 and its phosphorylation, IκBα and its phosphorylation, and IKKβ and its phosphorylation in blood vessels. (D) Changes in serum TNF-α and IL-6 levels. (F) Relative expression of TLR4, MyD88, NFκB p65, IκBα, and IKKβ mRNA. Compared with the WKY group, ^#p < 0.05, ^##p < 0.01; compared with the SHR group, *p < 0.05, ^**p < 0.01; compared with the FP group, ^▲p < 0.05, ^▲▲p < 0.01, n = 3.

Western blotting indicated a significant increase in TLR4 and MyD88 expression in the blood vessels in the model group compared with normal blood vessels (Fig. 8B), with the administration of FP or SFP downregulating this expression. The blood vessels of model rats exhibited increased phosphorylation of IKKβ, IκBα, and NF-κB p65 (Fig. 8C). The administration of FP and SFP inhibited the phosphorylation of these proteins. Furthermore, SFP could significantly downregulate TLR4 expression and inhibit IKKβ phosphorylation (p < 0.05) compared with that by FP. Additionally, both FP and SFP treatments reduced serum TNF-α and IL-6 levels in SHRs (p < 0.01, 0.05, respectively) (Fig. 8D). At the mRNA level, the administration of FP and SFP decreased the expression of the aforementioned five proteins in the blood vessels (p < 0.01), with SFP exhibiting a significantly pronounced inhibitory effect than FP (Fig. 8F).

Therefore, FP and SFP administration inhibited the activation of the TLR4/MyD88 pathway in blood vessels, alleviated vascular endothelial injury, and reduced the release of inflammatory cytokines.

3.4. FP and SFP administration improve intestinal barrier function and reduce LPS entry into the systemic circulation

3.4.1. FP and SFP decrease fecal LPS levels

Studies have revealed that SHRs demonstrate altered permeability of the intestinal epithelial barrier, facilitating LPS translocation into systemic circulation [20]. Histological analysis via H&E staining assessed the impact of FP and SFP on colonic morphology in SHRs. Fig. 9A revealed incomplete colonic tissue structure, severely damaged mucosal epithelium, extensive cell shedding, decreased number and disordered arrangement of goblet cells, and marked infiltration of inflammatory cells in the SHR group compared with those in normal animals. However, administration of FP and SFP significantly improved colonic injury in SHRs. Colonic tissues from the SFP group exhibited clear stratification and smooth and intact mucosal epithelium; however, occasional cell shedding was observed. In addition to the observation of tissue structure, the expression of tight junction proteins (TJPs) in the intestines was determined using various methods. Both IHC (Fig. 9B-D) and qRT-PCR (Fig. 9F) revealed that FP and SFP administration upregulated the protein and mRNA expression of Occludin and Claudin-1 in the intestine.

Fig. 9.

Effect of FP and SFP on intestinal barrier. (A) Representative graph (×200) of H&E staining of the colon, n = 3. (B) Representative IHC images for Occludin in the colon and ileum (×200), n = 3. (C) Representative IHC images for Claudin-1 in the colon and ileum (×200), n = 3. (D) Occludin and Claudin-1 expression in the brownish-yellow areas of IHC staining. (E) Changes in serum LPS and D-LA levels, n = 8. (F) Relative expression of Occludin and Claudin-1 mRNA, n = 8. (G) Changes in SCFAs in the feces, n = 3. Compared with the WKY group, ^#p < 0.05, ^##p < 0.01; compared with the SHR group, *p < 0.05, ^**p < 0.01; compared with the FP group, ^▲p < 0.05, ^▲▲p < 0.01.

Mechanistically, LPS, a complex lipid molecule situated in the cell wall of Gram-negative bacteria, primarily triggers downstream inflammatory signaling pathways by binding to TLR4[45]. Additionally, SHRs showed activation of the TLR4/MyD88 pathway, which may have been activated via excessive LPS production in the intestines, thereby impairing intestinal barrier function and leading to intestinal inflammation. This finding is consistent with the results from the determination of serum indicators, where serum LPS and D-LA levels in SHRs were significantly increased (p < 0.01) (Fig. 9E). The significant elevation in serum LPS provides a direct mechanistic link between the observed intestinal barrier damage and the downstream inflammatory signaling that exacerbates hypertension, underscoring the central role of gut-derived LPS in this pathological cascade.

3.4.2. FP and SFP increase fecal SCFAs levels

Carbohydrates are crucial metabolic substrates for SCFA production by bacteria. In contrast to the effects of LPS, SCFAs primarily contribute to maintaining the integrity of the intestinal mucosal barrier, improving the equilibrium of the gut microbiota, and suppressing inflammatory responses in the intestinal tract [46], [47]. SCFA levels (acetic acid, propionic acid, butyric acid, isobutyric acid, valeric acid, and isovaleric acid) in the feces of SHRs were significantly reduced compared with those in WKY rats (p < 0.01, 0.05). FP administration increased butyric acid, isovaleric acid, and valeric acid levels (p < 0.01, 0.05), whereas, compared with the FP group, SFP administration significantly increased propionic acid and butyric acid levels (p < 0.05) (Fig. 9G). Of particular note, butyric acid serves as the preferred energy source for colonic epithelial cells, and its elevated level holds direct implications for repairing the observed colonic mucosal damage.

3.4.3. FP and SFP administration regulate the intestinal microbiota in SHRs

The effects of FP and SFP on the gut microbiota of SHRs were evaluated using 16S rRNA sequencing. The Venn diagram indicates a decrease in the intestinal microbial diversity in SHRs, whereas an increase is noted in the FP and SFP groups (Fig. 10A). The α-diversity of sample colonies was analyzed to calculate the Dominance and Observed feature indices of the gut microbiota (Fig. 10B). The Dominance value for the SHR group increased significantly, whereas the Observed feature value decreased significantly compared with those for the WKY group (p < 0.01). After FP treatment, both indicators were restored to normal levels. Surprisingly, SFP treatment resulted in indicator levels better than those in the WKY group, suggesting the significant protective effect of SFP on the diversity of intestinal flora in SHRs.

Fig. 10.

16S rRNA sequencing analysis of gut microbiota after FP and SFP intervention. (A) Venn diagram. (B) Dominance and Observed feature indices. (C) PCA and PCoA at the genus level in the experimental groups. (D) Analysis of the differences in the abundances of the top 20 intestinal flora between the WKY, SHR, and SFP groups at the genus level. Compared with the WKY group, ^#p < 0.05, ^##p < 0.01; compared with the SHR group, *p < 0.05, ^**p < 0.01; compared with the FP group, ^▲p < 0.05, ^▲▲p < 0.01, n = 5.

In PCA and PCoA, the similarity in species composition is directly correlated with the distance between sample points: the closer the points, the more similar the community structures, whereas greater distances indicate larger differences in community composition[48]. PCA and PCoA of β-diversity revealed significant structural differences in microbial communities among the experimental groups. The SHR group exhibited clear segregation from the WKY group in multivariate space, indicating characteristic changes in the composition of the gut microbiota in hypertensive models (Fig. 10C). Notably, both FP and SFP groups were clustered separately from the SHR group while remaining distinct from the WKY control, suggesting that intervention-induced microbiota remodeling rather than complete restoration to a normotensive state. This finding highlights the protective effects of FP and SFP against hypertension-induced dysbiosis [49].

From a genus perspective (Fig. 10D), a significant gut microbiota dysbiosis was observed in the SHR group, the hallmark of which was an increase in pathogenic bacteria coupled with a reduction in beneficial species. Specifically, compared with the normal group, the model group exhibited a significant increase in the abundance of conditionally pathogenic bacteria such as Bacteroides and Desulfovibrio, while the abundance of beneficial bacteria, including Akkermansia and Butyricimonas was markedly decreased. SFP intervention significantly ameliorated this dysbiotic profile. It selectively enhanced the abundance of several beneficial bacteria, such as Blautia and Bifidobacterium, while simultaneously suppressing the growth of potentially harmful bacteria, including Desulfovibrio and Lachnospiraceae_NK4A136_group. This finding is consistent with the results from the serum index assay.

SFP administration significantly increased the abundance of the Akkermansia genus in the gut of SHRs. This genus belongs to the Verrucomicrobiota phylum and plays a role in maintaining intestinal barrier integrity, regulating metabolic homeostasis, and modulating immune balance [50]. Previous studies have demonstrated that the abundance of A. muciniphila is markedly reduced in hypertensive animal models and patients and is accompanied by the downregulated expression of intestinal TJPs and elevated levels of circulating pro-inflammatory cytokines[51], [52], [53]. Furthermore, enrichment of A. muciniphila promotes the production of SCFAs, including propionic and butyric acids, which activate the GPR41 receptor to inhibit the renin–angiotensin system, thereby exerting antihypertensive effects[54].

The analysis of gut microbiota revealed a state of dysbiosis in the SHR group, characterized by a significant increase in the Dominance index and a decrease in Observed features. Biologically, this pattern signifies a less stable microbial community that is more vulnerable to perturbation. The high dominance typically reflects an ecosystem driven by strong selective pressure, such as disease, which allows opportunistic taxa to thrive. Concurrently, the low richness indicates a loss of diversity and functional redundancy. The SFP intervention effectively reversed this trend by promoting a shift towards a richer and more balanced community structure, which is associated with greater ecosystem stability and functional resilience [55], [56], [57].

Spearman’s correlation analyses (Fig. 11) revealed that LPS levels predominantly demonstrated positive correlations with Desulfovibrio and Bacteroides and negative correlations with Bifidobacterium and Enterorhabdus. Notably, Desulfovibrio is a Gram-negative bacterium; its elevated intestinal abundance directly stimulates excessive LPS production. As a sulfate-reducing bacterium, it metabolizes dietary sulfur compounds to generate H₂S, which exhibits significant toxicity to the intestinal mucosal barrier[58]. Furthermore, SHRs exposed to SFP demonstrated a marked reduction in plasma LPS concentrations. SCFA levels were primarily positively correlated with Akkermansia, Blautia, Parabacteroides, and Butyricimonas. Blautia can synthesize acetate from hydrogen and carbon dioxide, a product that is utilized as a secondary energy substrate by intestinal epithelial cells while also providing energy to muscle and brain tissues. It also exhibits pathogen-inhibiting and anti-inflammatory activities[59]. Thus, a possible role of SFP in regulating the intestinal flora may be by inhibiting LPS production by decreasing the abundance of Desulfovibrio and other Gram-negative bacteria, while increasing SCFA concentration by increasing the abundance of Blautia and other SCFA-producing bacteria.

Fig. 11.

Spearman’s correlation heatmap between the intestinal flora and acetic acid, propionic acid, butyric acid, isobutyric acid, valeric acid, isovaleric acid, LPS, TNF-α, and IL-6 levels. *p < 0.05, ^**p < 0.01, n = 5.

In conclusion, SFP was found to modulate the gut microbiota to suppress excessive LPS production, consequently reducing the binding of LPS to vascular TLR4 and interfering with the activation of TLR4 signaling. This cascade downregulates the expression of downstream proteins and corresponding mRNAs, including MyD88, IKKβ, IκBα, and NF-κB p65, ultimately decreasing the levels of inflammatory mediators and attenuating vascular endothelial injury. Concurrently, SFP increases the levels of intestinal SCFAs by restructuring the microbial communities, which enhance the expression of intestinal Occludin and Claudin-1 (protein and mRNA), thereby strengthening the integrity of the gut barrier. This mechanism further restricts LPS and D-LA translocation into the systemic circulation, contributing to a reduction in blood pressure. Collectively, SFP could modulate two regulatory axes, the gut microbiota–LPS–vascular inflammation axis and the SCFA–gut barrier axis, highlighting its role as a multidimensional intervention to alleviate hypertensive vascular damage and pathological gut hyperpermeability and opening up a pathway for the development of novel perspectives for the management of EH.

4. Conclusion

In this study, FP was efficiently extracted via the ultrasonic method optimized through RSM. Although RSM indicated that 95% ethanol yielded the maximum polysaccharide extraction rate, a concentration of 85% ethanol was employed for the final preparation. This decision was based on the rationale of selective precipitation and the documented association between medium molecular weight and enhanced bioactivity, with the aim of obtaining higher-purity polysaccharides and enriching these potentially more active fractions for pharmacological evaluation. It took into comprehensive consideration the extraction yield, component purity, and the influence of molecular weight distribution on bioactivity, seeking an optimal balance between mass and quality.[60], [61], [62].

Our findings confirm that the frying step during the extraction process significantly affects the content, molecular weight, monosaccharide composition (uronic acid content and Gal/Ara ratio), and apparent structure of FT polysaccharides. The antihypertensive effect of SFP was significantly higher than that of FP, and this potentiating effect may be closely related to enrichment of uronic acid, optimization of the monosaccharide composition, or narrowing of the molecular weight distribution triggered by stir-frying. These findings suggest that structural modification is the core material basis of shelling that enhances the potentiating effect. However, the in-depth characterization of the structural changes in polysaccharides before and after processing is still lacking. Accordingly, structural changes in polysaccharides following processing will be further investigated using methylation analysis, ¹H NMR, ¹³C NMR, HSQC, and HMBC. Furthermore, studying the conformational relationships of polysaccharides after processing is an important research direction that we will pursue to provide a structural biology basis for the precise regulation of processing.

SFP modulated the intestinal flora in SHRs, inhibited LPS, increased SCFA production, and improved intestinal barrier function. Thus, the entry of LPS into the systemic circulation was reduced, the activation of LPS/TLR4 of intestinal origin was inhibited, vascular endothelial injury was decreased, and the blood pressure was lowered in SHRs (Fig. 12). Our findings also indicate that frying without excipients can alter the structure of the active polysaccharides of TCM, explaining the principle of concoction to enhance the efficacy of traditional Chinese herbal medicine from the perspective of polysaccharides, and providing further theoretical support for the development of TCMP.

Fig. 12.

Mechanism diagram of raw and stir-fried Fructus Tribuli polysaccharides.

Ethics statement

This study was approved by the Experimental Animal Welfare Ethics Committee of Shandong University of Traditional Chinese Medicine (NO. SDUTCM202101005), and conducted in accordance with the Helsinki Declaration. Animal experiments were conducted in accordance with the National Institutes of Health Animal Use Guidelines and experimental protocols were approved by the Experimental Animal Welfare Ethics Committee of Shandong University of Traditional Chinese Medicine.

CRediT authorship contribution statement

Shuai Wang: Writing – original draft, Visualization, Methodology, Formal analysis, Data curation, Conceptualization. Haiping Sun: Writing – review & editing, Formal analysis, Data curation. Xiyu Lv: Writing – review & editing, Formal analysis, Data curation. Jingna Li: Investigation, Conceptualization. Defeng Du: Investigation, Conceptualization. Fei Guo: Methodology, Investigation. Chao Zhang: Writing – review & editing, Supervision, Project administration, Funding acquisition, Conceptualization.

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.

No conflict of interest exits in the submission of this manuscript, and manuscript is approved by all authors for publication. Fig. 12 was created by Figdraw (https://www.figdraw.com).

Appendix A. Supplementary data

The following are the Supplementary data to this article: