Dried-cultured P. tomentosa flowers were collected from Bozhou Guoxin Pharmaceutical Co. Ltd. (Anhui, China). Live attenuated CSFV (rabbit source) was purchased from Pulike Biological Engineering INC (China). Sources of fetal bovine serum (FBS), RPMI-1640 medium, ConA, LPS, CSFV E2 protein (CSFV E2), antibodies (IgG, IgG1, IgG2a, and IgG2b), and cytokine ELISA kits (IFN-γ and IL-10) are listed in Supplementary Table S1.

All the chemicals used in this study were analytical grade. The human leukemia cell line K562, which was sensitive to NK cells, was obtained from the Shanghai Institute of Cell Biology, Chinese Academy of Sciences. The cells were cultivated in the logarithmic phase of growth in RPMI 1640 medium supplemented with penicillin (100 IU/mL), streptomycin (100 g/mL), and FBS (10%) at 37°C and 5% CO₂.

PTFP was extracted using a previously described method (21). Briefly, P. tomentosa flowers were decocted and filtered twice for 2 h and 1 h in water. Subsequently, the suspension condensed from the merged decoction was precipitated four times for 12 h in 95% ethanol. Finally, the samples were centrifuged and concentrated to a specific volume with 48% carbohydrate (D-glucose as the standard).

The morphology of PTFP was observed by SEM (Zeiss Merlin Compact, Germany). The PTFP, coated with a thin gold layer, were laid on the placode, and then the images were analyzed by Sanshu Biotech. Co., LTD (Shanghai, China) under high vacuum at a magnification of 1,000 ×, 2000 ×, 5,000 × and 10,000 × fold.

PTFP with 5 mg was weighed accurately and placed in a sealed tube. Trifluoroacetic acid solution was added to the sample for acid hydrolysis, after hydrolysis, nitrogen gas was used to dry the residual PTFP in the tube. Wash with methanol, then blow dry and repeat methanol washing 2–3 times. The residue was dissolved in deionized water and passed through 0.22 μm microporous membrane filtration. The PTFP filtrate was analyzed by high-performance anion exchange chromatography on a CarboPac PA-20 anion exchange column using a pulse current detector. Data were collected on ICS5000+ (Thermo Technology) and processed using chromeleon 7.2 CDS (Thermo Science).

PTFP was dissolved in NaNO₃ aqueous solution (0.1 mol/L), and filtered through a 0.45 μm pore size filter. Furthermore, PTFP was also dissolved in dimethyl sulfoxide (DMSO) solution containing lithium bromide at a density of 1 mg/mL, then filtered through a 0.45 μm pore size filter. The homogeneity and molecular weight of various fractions in PTFP were determined by using SEC-MALLS-RI. The weight, polydispersity index and number-average molecular weight of PTFP various fractions in NaNO₃ aqueous solution or DMSO solution were measured by DAWN HELEOS-II laser photometer. Meanwhile, three tandem columns were provided and held at 45°C by using a model column heater. A differential refractive index detector (Optilab T-rEX, Wyatt Technology Co., United States) was simultaneously connected to give the concentration of fractions and the refractive index increment (dn/dc) value. The dn/dc value of the fractions in NaNO₃ aqueous solution was measured to be 0.141 mL/g, and in DMSO solution was determined to be 0.07 mL/g. Data were acquired and processed using ASTRA6.1 (Wyatt Technology).

The ultraviolet–vis spectrum of PTFP water solution (5 mg/mL) was determined in a wavelength range of 200–1,000 nm by using a multifunctional microplate reader (Multiskan GO, Thermo Fisher Scientific, United States). Under the same conditions, pure water was used as blank control by Sanshu Biotech. Co., LTD.

Fourier Transform infrared (FT-IR) spectra of PTFP was measured by using a spectrometer (Nicolet iZ-10, Thermo Nicolet, United States). PTFP was mixed with KBr powder and then pressed into 1 mm pellets for FT-IR measurement in the range of 4,000 to 400 cm⁻¹ by Sanshu Biotech. Co., LTD.

Male and female ICR mice were obtained from Shanghai Slake Laboratory Animal Co., Ltd., China. The mice were 5 weeks old and weighing 18–22 g, then allowed to acclimatize to the environment for one week for the following experiments. The mice were maintained in groups of less than five mice per cage under stable controlled conditions (24 ± 1°C, 50 ± 10% humidity, 12/12 h light/dark cycle), and food and tap water were freely available. The study was approved by the Animal Welfare and Ethics Committee of Zhejiang University (No. 18227).

ICR mice were raised in an environment free of specific pathogens, and separated into six groups and named: PBS, CSFV, APS, PTFP-L, PTFP-M, and PTFP-H groups, consisting of six male and six female mice each group. The experimental procedure was illustrated by schematic illustration (Figure 1). Except for the PBS group, the CSFV, APS, PTFP-L, PTFP-M, and PTFP-H groups were immunized twice by subcutaneous injection of 0.1 portion CSFV on day 6 and day 20 of the experiment. PBS and CSFV groups were gavaged with sterile PBS once a day. APS group was subjected to gavage APS (100 mg/kg) once a day. PTFP-L, PTFP-M, and PTFP-H were administered oral gavage at 30, 60, and 120 mg/kg doses of PTFP once a day, respectively. Throughout the experiment, PBS, APS, and PTFP were administered orally from day 1 to day 10 and from day 15 to day 24. On the day 34 of the experiment, mice were sacrificed by cervical dislocation, and the sera and splenocytes were reserved for further analysis.

Splenocytes were separated as previously described (22). The spleen was aseptically removed and minced using scissors. The spleen tissue was then forced to pass through a fine steel mesh, and the cell suspension was collected. Erythrocytes were removed using ammonium chloride (0.8%, w/v). The collected splenocytes (5 × 10⁵ cells/well) were spread into 96 well plates in 100 μL complete medium and treated with classical swine fever virus E2 antibody (CSFV E2), concanavalin protein (ConA), or lipopolysaccharide (LPS) at 10, 5, and 10 μg/mL, respectively, while the blank medium was used as a control. After 44 h of culture at 37°C and 5% CO₂, 50 μL MTT (2 mg/mL) was added to each well and cultured for a further 4 h. Subsequently, the cultured samples were centrifuged, and the supernatant was removed. A volume of 150 μL dimethyl sulfoxide was added to each well and incubated for 15 min. Finally, the absorbance was estimated at 570 nm using ELISA.

The stimulation index (SI) was measured according to the following equation: SI = A 570 ( CSFV E2,ConA or LPS ) / A 570 ( control ) .

The NK cell activity was evaluated by the MTT method (23). Specifically, K562 cells and splenocytes were defined as target and effector cells, respectively. K562 cells were seeded in 96-well U-bottom microtiter plates (2 × 10⁴ cells/well) in medium. Splenocytes were seeded at a density of 1 × 10⁶ cells/well, at an effector/target ratio of 50:1. After 20 h, NK cell viability was evaluated using the MTT method as follows: NK cell viability (%) = [A570_T – (A570_S–A570_E)] / A570_T × 100. A570_T, A570_S, and A570_E were the absorbance values of the target control, test sample, and effector control, respectively.

CSFV E2-specific IgG, IgG1, IgG2a, and IgG2b antibodies in sera were measured using indirect ELISA (24). Briefly, 96 well ELISA plates were coated with 100 μL of CSFV E2 antigen (2.5 μg/mL in carbonate solution) per well at 4°C overnight. ELISA plates were blocked for 2 h at 37°C. Then, 100 μL of serially diluted serum samples was placed into the plates and incubated for 2 h at 37°C. The plates were incubated with IgG, IgG1, IgG2a, or IgG2b antibodies at 1:8000, 1:6000, 1:4000, and 1:4000 dilutions, respectively. After incubation for 2 h at 37°C, 100 μL of TMB was added to the plates and 50 μL of 2 mol/L H₂SO₄ was added 15 min after incubation to stop the reaction. Between each step, the ELISA plates were washed three times. Absorbance was measured at 492 nm.

Splenocytes (5 × 10⁶ cells/well) were cultured with CSFV E2 (10 μg/mL) in 24-well plates. After incubation at 37°C in 5% CO₂ for 72 h, the plate was centrifuged (2,500 rpm, 5 min), and supernatants were reserved for measurement of IL-10 and IFN-γ using ELISA kits (25).

All data were presented as mean ± standard deviation (mean ± SD). Statistical significance of differences was evaluated by ANOVA with Tukey’s post-hoc test. The significance threshold was set at p < 0.05.

The molecular morphology of PTFP was investigated by SEM under 1,000 ×, 2000 ×, 5,000 × and 10,000 × magnifications (Figure 2). The SEM results showed that the surface of PTFP was dense and rough, with a uniform dense layered structure, small fragments, no curling, and certain parts presented stacking phenomenon, suggesting that the intermolecular force of PTFP was strong and its molecular weight was large.

The monosaccharides composition in PTFP was analyzed by HPAEC (Figures 3A,B). The specific quantification results are shown in Table 1. PTFP was primarily composed of glucose (Glc), rhamnose (Rha), galactose (Gal), arabinose (Ara), mannose (Man), and xylose (Xyl), and the molar mass ratio of their monosaccharide composition was 30.93, 29.99, 15.66, 6.95, 5.52, 4.80%. Based on the above results, it was indicated that PTFP was a heteropolysaccharide mainly composed of Glc and Rha.

Gel chromatography was used to analyze the molecular weight of PTFP (Figures 3C,D). As shown in the Figure 3C, the molar masses ranged from 1.0 × 10⁵ to 1.0 × 10⁷ (g/mol) with symmetric signal peaks, suggesting that PTFP possessed a homogeneous relative molecular weight distribution. The specific molecular weights of PTFP are displayed in Table 2. Weight-average molecular weight (Mw) is the main indicator reflecting the mass of macromolecular substances. The Mw of PTFP was 667.02 kDa, indicating that PTFP possessed a large molecular weight. Polydispersity is the ratio of Mw to number-average molecular weight (Mn), represents the distribution range of molecular weight, and polydispersity of PTFP was 10.23, demonstrating the dispersion of molecular weight distribution of components in PTFP. Figure 3D displayed the molecular configuration of PTFP, and conformation plot slope was 0.07 ± 0.00, implying that the molecular configuration of PTFP may be a spherical structure. Moreover, the chain conformation of PTFP did not alter obviously with changes in molecular weight, which may indicate that PTFP had stable molecular conformation.

The component and chemical bonds of PTFP were analyzed by ultraviolet and infrared spectrum (Figure 4). Compared with the blank control, the PTFP sample had absorption in the wavelength range of 200–400 nm, refreshing that the PTFP sample may contain small amounts of pigments, proteins, and nucleic acids (Figures 4A–C). As shown in the Figure 4D, the absorption band was the absorption peak of -OH stretching vibration at 3600–3200 cm⁻¹, which was the characteristic absorption peak of sugars. The absorption peak at 3413.01 cm⁻¹ was the stretching vibration O-H, representing the characteristic peak of sugars. The small absorption peak at 2927.61 cm⁻¹ was attributed to C-H shearing vibration and bending vibrations of free sugars. The C=O stretching vibration at 1630.01 cm⁻¹ suggested the existence of acetylation and uronic acid in PTFP. Furthermore, there was the absorption peak at 1038.99 cm⁻¹, attributed to the stretching vibration of C-O.

The SI changes of splenocyte induced by ConA, LPS, and CSFV E2 are shown in Figure 5. SI is the main index reflecting the proliferation of splenocyte. Under ConA induction, compared with CSFV group, SI of splenocyte in PBS, APS and PTFP groups was significantly increased (Figure 5A) (p < 0.05 or p < 0.01), as was the case under LPS induction (Figure 5B) (p < 0.05, p < 0.01, or p < 0.001), suggesting that CSFV vaccine-immunized represented inhibitory effect on the proliferation of splenocyte after ConA or LPS stimulation, while APS and PTFP could promote splenocyte proliferation. Under CSFV stimulation, SI of splenocyte in APS and PTFP groups was significantly elevated compared to that in the CSFV group (Figure 5C) (p < 0.05, p < 0.01, or p < 0.001), whereas SI of splenocyte in PBS group showed a downward trend (p < 0.05), indicating that after stimulation with CSFV E2, proliferation of splenocytes from CSFV immunized mice could be promoted, with APS and PTFP showing remarkable effects. These data demonstrated that each dose of PTFP could significantly promote the proliferation of spleen cells induced by ConA, LPS, and CSFV E2 in mice immunized with CSFV, of which 60 mg/kg had the best effect that was equal to APS.

Figure 6A presents changes in NK cell viability, APS and PTFP at the three doses significantly promoted the activity of NK cells in CSFV-immunized mice (p < 0.05, p < 0.01, or p < 0.001). The findings revealed that all doses of PTFP could significantly activate NK cells in mice immunized with CSFV, and this effect was equivalent to that of APS (p > 0.05).

Considering that cytokines play a pivotal role in the immune response and to explore the interaction between the Th1 cytokine IFN-γ and Th2 cytokine IL-10, we evaluated the effects of PTFP on the secretion of cytokines from CSFV E2-stimulated splenocytes. Compared with CSFV group, IFN-γ and IL-10 levels in PBS group were significantly elevated (p < 0.05, or p < 0.001). Moreover, APS and PTFP at doses of 60 and 120 mg/kg significantly increased IFN-γ levels (Figure 6B) (p < 0.001). APS and PFTP at three doses also significantly improved the ability of CSFV E2-stimulated splenocytes in CSFV-immunized mice to secrete IL-10 (Figure 6C) (p < 0.001).

As the nature of the antigen may play a pivotal role in the mediation of immune responses induced by an adjuvant, we explored the adjuvant effects of PTFP on antigen-specific humoral immune responses to CSFV E2. The CSFV E2-specific IgG, IgG1 antibody titers in CSFV-immunized mice were distinctly increased by APS and PTFP at each dose (Figures 7A,B) (p < 0.05, p < 0.01, or p < 0.001). Furthermore, APS and PTFP at 60 mg/kg significantly enhanced the IgG2a and IgG2b titers (Figures 7C,D) (p < 0.05). There was no obvious difference in the titers of CSFV E2-specific antibodies in sera between the three doses of PTFP, whereas the middle-dose group had the best effect equivalent to APS (p > 0.05).