Albizia julibrissin Durazz. leaf and pod were procured at the Bozhou Chinese herb market in Anhui, China. Shanghai Aladdin Bio-Chem Technology Co., LTD. supplied the trifluoroacetic acid (TFA). The Guangzhou Hongquan Biological Technology Company, situated in Guangzhou, China, produced the human serum. Erythrocyte extraction was performed utilizing Alsever’s solution, originally sourced from male sheep and rabbits. The essential reagents, all of HPLC quality, were obtained from commercial vendors. The entire experimental process adhered strictly to the guidelines set by the National Institutes of Health for the Care and Use of Laboratory Animals. Ethical approval was also obtained from Yanbian University’s Ethics Committee, with approval number YBU-2018-090401.

Polysaccharides from A. julibrissin leaf and pod were extracted utilizing an ultrasonic extractor (JM-15D-28, Skymen, China). The extraction process involved varying four parameters: extraction time (20 min to 60 min), temperature (40°C to 80°C), liquid solid ratio (10 mL/g to 50 mL/g), and ultrasonic power (100 W to 300 W). Based on the results of the single-factor test, the preliminary ranges for extraction temperature (X₁, °C), liquid-to-solid ratio (X₂, mL/g), ultrasonic time (X₃, min), and ultrasonic power (X₄, W) were determined. The extraction process conditions were optimized using RSM (Hao et al., 2023). Additionally, the yield of polysaccharide served as the response value, with each experiment replicated three times. Subsequently, the experimental data obtained from the Box–Behnken Design (BBD) were analyzed using Design Expert 13.0 software (Li et al., 2018).

The crude polysaccharides underwent precipitation using 80% ethanol, followed by deproteinization via the Sevag method (Yuan et al., 2024). Further purification involved loading the polysaccharides onto a DEAE-52 cellulose column (2.6 cm × 60 cm) and eluted using 0.5 mol/L NaCl solution at a flow rate of 1 mL/min, subsequently dialyzing through a membrane with a 8,000 Da cutoff.

Chemical composition and purity analysis were conducted on the polysaccharide constituents of leaf and pod before (crude-AJLP and crude-AJPP) and after (AJLP and AJPP) purification. The measurement of the total polysaccharide content was performed using the phenol-sulfuric acid approach, with absorbance precisely quantified at 490 nm. For quantifying polyphenol and flavonoids content, the Folin–Ciocalteu method and aluminum trichloride assay were employed, where the absorbance was precisely measured at 530 nm. Furthermore, the Bradford method employing Coomassie blue G250 dye was utilized to evaluate the protein content, with absorbance meticulously measured at a wavelength of 595 nm. Meanwhile, the content of uronic acid was determined with precision using the meta-hydroxydiphenyl method, with absorbance accurately quantified at 520 nm (Niu et al., 2017; Duan et al., 2024; Kanlayavattanakul et al., 2019).

This test involved mixing each polysaccharide fraction with an appropriate amount of KBr, then compressing the mixture into tablets, and using fourier transform infrared spectroscopy (FTIR, Thermo Science Nicolet iS20, Thermo Fisher, United States) to scan the wavenumber spectrum from 4,000 to 400 cm⁻¹.

The molecular weight (Mw) of the polysaccharides was assessed using high-performance gel permeation chromatography (GPC) equipped with a RID-20A detector (Shimadzu, Japan). The samples were diluted to a concentration of 2 mg/mL for analysis. Chromatographic Conditions: A Shodex sugar KS-804 column (7 μm, 8.0 mm × 300 mm) was utilized, maintained at a temperature of 40°C. The mobile phase consisted of ultrapure water, pumped through the system at a flow rate of 1.0 mL/min, and the sample was injected with a volume of 20 μL. A calibration curve, established using dextran standards of known molecular weights, was employed to determine the Mw of the polysaccharides.

The monosaccharide composition was assessed and modified employing a method previously documented in (Liu et al., 2019). A 10 mg polysaccharide sample was hydrolyzed using 2 mol/L TFA, subsequently derivatized with a solution of 0.3 mol/L NaOH, PMP-Methanol, and 0.3 mol/L HCl. To purify the derivative, the mixture was washed three times with chloroform to eliminate impurities from the aqueous layer. Subsequently, the sample was analyzed using high-performance liquid chromatography (HPLC) equipped with a 1,410 UV detector (Hitachi Primaide, Japan). The HPLC analysis was conducted on a Supersil ODS2 column (2.5 μm, 4.6 mm × 250 mm) maintained at 35°C. The elution process occurred at a flow rate of 0.8 mL/min, and the derivative was detected at a wavelength of 245 nm. The mobile phase was composed of a mixture of phosphate buffered saline (PBS) at pH 6.8 and acetonitrile in a ratio of 83:17 (v/v).

The microscopic morphological properties of the polysaccharides was investigated through the use of a Gemini SEM 360 (ZEISS, Oberkochen, Germany). The sample powder was placed on the specimen stage and then coated with a thin layer of gold using an ion sputtering device. This allowed for the examination of the samples at various magnifications, including 200×, 500×, 1,000 × and 2000×.

The assessment of the thermal stability of the polysaccharide sample was conducted using a Thermal gravimetric analyzer TG 209 F1 Libra (NETZSCH-Gerätebau GmbH, Selb, Germany). In detail, a quantity of 5 mg of the sample was placed within a platinum crucible, followed by heating at a rate of 20°C/min, ranging from 30 to 800°C.

The polysaccharide samples’ anticomplementary activity was assessed through a hemolysis test, following a previously reported method (Wang et al., 2022). For this test, a dilution of normal human serum pool (NHSP) was prepared in GVB-Ca²⁺/Mg²⁺ buffer, using a specified ratio. The polysaccharide samples underwent subsequent dilution in the identical buffer and were preincubated with NHSP. Subsequently, sensitized sheep erythrocytes (2.0 × 10⁹ cells/mL) or rabbit erythrocytes (5.0 × 10⁸ cells/mL) were added. Upon completion of the reaction, the mixture underwent centrifugation, followed by precise measurement of the supernatant’s optical density at a wavelength of 540 nm using a microplate reader (Infinite 200 PRO, Tecan, Switzerland). Ultimately, the CH₅₀ and AP₅₀ was employed to evaluate the anticomplementary activity of the tested fraction.

The outcomes of three parallel experiments (n = 3) were presented as mean values accompanied by their standard deviations (± SD). Significant statistical differences (p < 0.05) were determined through ANOVA analysis utilizing GraphPad Prism 6.0 software.

Utilizing the conventional single factor variation approach, during each individual factor experiment, all parameters remain constant except for the one being specifically investigated. Figure 1 demonstrated the influence of four extraction parameters: extraction temperature (X₁, °C), liquid to solid ratio (X₂, mL/g), ultrasonic time (X₃, min), and ultrasonic power (X₄, W), on the yield of A. julibrissin polysaccharide. The findings revealed that the most favorable individual condition for maximizing leaf polysaccharide yield was ultrasonic time of 50 min, extraction temperature of 70°C, liquid to solid ratio of 40 mL/g, and ultrasonic power of 250 W. Analogously, for pod polysaccharides, the optimal single parameter was 40 min, 70°C, 30 mL/g, 200 W.

The results for each of the two parts demonstrated a trend of initial increase followed by a decrease. In particular, as the temperature continued to increase, there was a noticeable decline in the extraction yield. This could potentially be attributed to excessively high temperatures disrupting the structural stability of polysaccharides, thereby promoting their oxidation and thermal degradation (Zhang et al., 2024). At higher liquid to solid ratio, the extraction yield decreased, this could be due to an increased solvent volume leading to a greater internal diffusion distance, subsequently reducing hydration and impacting mass transfer (Zheng et al., 2016). Eventually, as the ultrasound time and power increased, a declining trend was observed in the extraction yield. When prolonged ultrasound time coupled with excessive power was applied, it intensified the thermal effect and mechanical action, leading to polysaccharide molecule degradation and consequently lowering the yield (Zhang W. et al., 2023). In addition, an interesting phenomenon observed was that leaf parameters were higher in terms of ultrasonic time, liquid to solid ratio, and ultrasound power. This may be due to the fact that the cell wall structure of leaf and the binding form of polysaccharides are more complex compared to pod, requiring higher extraction time, liquid to material ratio, and ultrasound power to overcome these interactions.

The optimum ultrasound-assisted extraction parameters for polysaccharides from A. julibrissin leaf and pod were optimized using response surface methodology. The quadratic polynomial function representing the leaf (Equation 1) and pod (Equation 2) polysaccharides was expressed in order as follows:

The analysis of variance (ANOVA) for the quadratic polynomial model extracted from leaf and pod polysaccharides (Tables 1, 2). As displayed, the p-values of the model were all less than 0.0001, which meant that the model had significance. Meanwhile, the values of the regression coefficient (R²) of leaf and pod were 0.9901 and 0.9849, respectively. The experimental data exhibited a high degree of precision and reliability, with R²Adj above 0.90 and C.V. below 10%, indicating a strong correlation between the observed and predicted values. And, the model equation demonstrated adequate predictive capability for the extraction yield of polysaccharides, as evidenced by the negligible lack of fit (Ji et al., 2021). Consequently, these findings imply that the model was applicable for exploring the design space. Moreover, four interaction term coefficients (X₁X₂, X₁X₄, X₂X₃, and X₂X₄) significantly affected leaf polysaccharide yield, while three (X₁X₂, X₂X₃, and X₃X₄) influenced pod polysaccharide yield (p < 0.05). All quadratic term coefficients (X₁², X₂², X₃², and X₄²) significantly impacted both yields (p < 0.001). The utilization of two-dimensional contour (2D) plots and three-dimensional response surface (3D) charts, depicted in Figures 2, 3, aided in elucidating the relationship between the yield of leaf and pod polysaccharides, which was influenced by four factors, and its independent and dependent variables.

The ultrasound-assisted extraction conditions for polysaccharides differed according to the different parts being extracted. Specifically, the optimal settings for extracting leaf polysaccharides comprised an extraction temperature at 70°C, a liquid-to-solid ratio of 40 mL/g, an ultrasonic time of 50 min, and an ultrasonic power setting of 249 W. Similarly, for pod polysaccharides, the optimal conditions were 70°C, 28 mL/g, 40 min and 201 W, respectively. Under the circumstances, the polysaccharide yield from leaf, and pod were 1.07 ± 0.20%, and 8.32 ± 0.27%, respectively, showing increments of up to 0.41 and 2.33% when compared to the traditional water extraction and alcohol precipitation method. Additionally, ultrasonic-assisted extraction technique decreased both the extraction duration and processing temperature, suggesting its efficacy in obtaining leaf and pod polysaccharides. The significantly lower yield of leaf polysaccharides compared to pod polysaccharides (p < 0.05) may result from advanced leaf lignification, which converts soluble polysaccharides into insoluble lignocellulose complexes, thereby reducing extractability. These findings align with report of Bai et al. (2022) on different parts of ginseng polysaccharide.

Table 3 presented the total sugar, protein, uronic acid, polyphenol and flavonoid contents of crude polysaccharides extracted from leaf and pod of A. julibrissin before and after purification. As anticipated, significant differences in chemical composition were observed among the polysaccharides from various parts. The total sugar content of pod crude-AJPP was greater than that of crude-AJLP, indicating that the pod of A. julibisin had a richer polysaccharide content than the leaf. Similar results have also been reported by Shi et al. (2024), which may be due to the higher content of water-soluble impurities and pigments in the leaf of A. julibrissin. Besides, the protein content of crude-AJLP was higher than that of crude-AJPP, and it contained small amounts of polyphenol and flavonoid. These common impurities can affect the identification of polysaccharide structures and the study of their biological activities, so the Sevag method, DEAE-52 cellulose and dialysis were used for purification and separation. After purification, the presence of these impurities was not detected and the uronic acid content was increased. And they contained over 90% total sugar content and were suitable for subsequent experimental determinations.

This FTIR spectra absorption regions can facilitate the identification of polysaccharide functional groups (Zheng et al., 2024; Tang et al., 2024). The FT-IR spectra of AJLP and AJPP was presented in Figure 4. The approximate peaks at 3400 cm⁻¹ and 2,900 cm⁻¹ indicated O-H and C-H stretching vibrations, respectively. The spectral range from 1750 cm⁻¹ to 1,600 cm⁻¹ showed peaks indicative of C=O vibrations, hinting at the inclusion of uronic acid in the polysaccharide structure (He et al., 2016). Peaks observed between 1,500 and 1,200 cm⁻¹ suggested C-H bond deformation vibrations. Additionally, multiple peaks in the 1,200 to 1,000 cm⁻¹ range for AJLP and AJPP implied the presence of glycosidic linkages and pyran rings within the polysaccharide structure. Furthermore, the peaks observed within the 900 to 700 cm⁻¹ range could potentially be associated with β-glycosidic bonds and α-glycosides (Zhang et al., 2018; Zhang et al., 2020). There was no significant difference in the infrared spectra among the AJLP and AJPP, which also was consistent with the previous report related to the Angelica sinensis (Oliv.) Diels polysaccharides (Zou et al., 2022), Hibiscus manihot L. polysaccharides (Shi et al., 2024) and different parts of ginseng polysaccharides (Bai et al., 2022).

The Mw can reflect the molecular chain information, exhibiting a profound correlation with the biological functionality of polysaccharides (Liang et al., 2023; Chen et al., 2024). In Figure 5 and Table 4, all of AJLP and AJPP showed two peaks, suggesting that they were composed of two polymers with distinct Mw. The broad peaks observed for AJLP and AJPP were due to the overlap of different polymer types. The Mw of the first fragment (peak 1) of AJLP (2.323 × 10⁵ Da) is greater than that of the corresponding fragment of AJPP (8.051 × 10⁴ Da). Similarly, the Mw of the second fragment (peak 2) also followed the same descending order. The findings indicated notable variations in the molecular weights of AJLP and AJPP, likely due to their origin from distinct parts and functions. High molecular weight polysaccharides may better facilitate the formation of membrane structures, participating in leaf support and water retention. Low molecular weight polysaccharides have higher water solubility, facilitating rapid metabolism and energy release during pod germination. Additionally, the polydispersity index, denoted as Mw/Mn, enables assessment of the uniformity in molecular weight distribution through the polydispersity coefficient. Table 4 revealed that the polysaccharide samples exhibited a polydispersity index between 1.0 and 1.1, suggesting a high level of homogeneity among the polysaccharides.

Variations in monosaccharide compositions and their proportions significantly impact the structure and properties of polysaccharides (Zhang S. et al., 2023; Kong et al., 2023). The results of the experiment were depicted in Figure 6 and summarized in Table 5. It was noteworthy that the monosaccharide species composing AJLP and AJPP were the same, consisting of mannose, rhamnose, glucuronic acid, galacturonic acid, glucose, galactose, xylose and arabinose, but the proportions were obviously different. And the content of glucose, rhamnose, and galactose was relatively high, but the contents of glucuronic acid and xylose were the lowest. These differences in polysaccharide constituents could stem from distinct parts. Although the types of monosaccharide components remain unchanged, shifts in their content may affect the physicochemical characteristics and structure of these polysaccharides, potentially influencing their biological activities thereafter.

Figure 7 illustrated the surface morphology of AJLP and AJPP captured by SEM at various magnifications: 200×, 500×, 1,000 × and 2000×. AJLP and AJPP displayed a disordered accumulation of flaky structures and rough surfaces, featuring prominent folds along the edges. In contrast, the microstructure of AJLP and AJPP was different, and the size of their flaky structure increasing sequentially from AJLP and AJPP. The differences in surface morphology might ascribe to the different degrees of branching in the two polysaccharides fractions (Zhu et al., 2021). A flatter surface could be observed under high magnification, the strong intermolecular interaction might be beneficial to exerting its bioactivity (Qu et al., 2019). Reports indicated that the dense structure is beneficial for polysaccharides to resist oral and gastrointestinal digestion, enabling them to reach the large intestine where they can be utilized by gut microbiota (Wang et al., 2025). These studies have found that polysaccharides extracted from the leaf and pod of A. julibrissin agro-byproducts have the potential to be developed into functional foods, health products, and medicines.

TGA serves as a thermal analysis method, enabling the evaluation of crucial properties of polysaccharides, including their thermal stability and decomposition temperature (Dou et al., 2023; Lin et al., 2024). According to Figure 8, the thermal degradation of AJPP proceeded in two stages, whereas AJLP underwent degradation in four stages. During the initial stage, a minor weight reduction was observed at approximately 90°C, which may be due to the evaporation of water, both free and bound (Li et al., 2019). Variations were observed in the quantity of bound water, ranging between 5.10 and 9.06%. Notably, AJLP exhibited the highest weight loss (9.06%) at this stage, suggesting superior water retention capabilities. The second stage witnessed the greatest weight loss, primarily due to the depolymerization of polysaccharide chains and the decarboxylation of acid groups (Abubakar et al., 2024). During the third and fourth stages, the weight change of AJLP tended to stabilize, mainly because of the cleavage of sugar rings and the gradual development of a carbonized structure in the polysaccharide sample (Li et al., 2024; Xiao et al., 2019). The order of final weights is AJLP (39.96%) followed by AJPP (31.53%), indicating a variation in thermal stability, with AJLP exhibiting the highest and AJPP the lowest. Collectively, all polysaccharide fractions demonstrated commendable thermal stability and are applicable to high-temperature food processing. This disparity could potentially stem from variations in Mw, advanced structures and different parts. This finding has the potential to broaden the utilization of agro-byproducts (AJLP and AJPP) within the food industry.

The hemolytic assay reveals the drug’s ability to inhibit complement activation, evidenced by its suppressive effect on red blood cell lysis induced by complement system activation. An investigation was conducted to evaluate the impact of AJLP and AJPP on the activation of the complement system via both the classical (CP) and alternative (AP) pathways, with heparin (CH₅₀: 0.143 ± 0.028 mg/mL; AP₅₀: 0.148 ± 0.037 mg/mL) serving as a positive control comparator. A pool of normal human serum served as the complement source for both the CP and AP. All results were presented in Table 6.

In an assay measuring anticomplementary activity through the CP, a 1:80 dilution of NHSP induced a 97.83 ± 0.76% hemolysis in antibody-sensitized sheep erythrocytes (EAs). The concentrations necessary to achieve 50% inhibition of hemolysis in EAs (CH₅₀) were found to be 0.122 ± 0.004 mg/mL for AJPP. In an assay measuring anticomplementary activity via the AP, a 1:8 dilution of NHSP induced a lysis of 96.90 ± 0.83% in rabbit erythrocytes (ERs). The AP₅₀ values, which represent the concentrations causing 50% inhibition of hemolysis in ERs, were determined to be 0.070 ± 0.003 mg/mL for AJLP. AJLP demonstrated no inhibitory activity against complement activation via the CP, while AJPP showed no such activity against activation through the AP. In CP, the polysaccharides obtained from the leaf were comparable to the positive drug heparin in terms of complement inhibitory capacity. Notably, the polysaccharides derived from the pod exhibited significantly higher anticomplementary activity than heparin in AP. The results of this experiment were encouraging, the often-neglected agro-byproducts (leaf and pod) possessed inhibitory abilities comparable to or even superior to those of heparin as a positive control drug. It was interesting to note that the complement inhibition by the leaf and pod was via a single pathway, which will enable them to have a more specific therapeutic effect on related diseases caused by a certain pathway. Even in the field of functional foods and health products, AJLP and AJPP have great potential for development. This offers a fresh perspective on the holistic development and utilization of Traditional Chinese Medicine, ensuring efficient use of active ingredients and generating substantial economic benefits.