Crude 4-Z PCP (64.88 g) was carefully isolated and then purified from the rhizomes of P. cyrtonema Hua. by the nine-steaming and nine-basking approach. Then, 1,000 mg of crude 4-Z processed PCP was eluted through cellulose column chromatography using 0 M, 0.05 M, and 0.1 M NaCl solution to yield PCP-F0 (520 mg), PCP-F0.05 (110 mg), and PCP-F0.1 (270 mg), respectively, with a flow rate of 2.0 mL/min. The result was summarized in Supplementary Figure S1. According to DPPH radical assays, the PCP-F0.1 group at a concentration of 0.2, 0.6, and 1.0 mg/mL provided a scavenging rate of 3.7% ± 0.7, 7.1% ± 0.8, and 13.9% ± 0.9, respectively. These results were presented in Figure 3A. Additionally, according to the hydroxyl radical assays, the PCP-F0.1 group at a concentration of 0.2, 0.6, and 1.0 mg/mL provided a scavenging rate of 24.3% ± 0.2, 54.9% ± 0.8, and 80.2% ± 0.8, respectively. These results were presented in Figure 3B. Based on the results of radical assays, the PCP-F0.1 group exhibited higher radical-scavenging abilities than PCP-F0 and PCP-F0.05 groups.

The total sugar contents of PCP-F0, PCP-F0.05, PCP-F0.1, and PCP-F1 were 85.65 ± 1.30, 76.28 ± 1.77, 83.53 + 1.53, and 82.06 ± 0.35%, respectively. And the acid sugar contents of PCP-F0, PCP-F0.05, PCP-F0.1, and PCP-F1 were 4.73 ± 0.54, 5.95 ± 0.10, 6.14 ± 0.26, and 7.20 ± 0.39%, respectively (Figures 4A,B). There was no obvious absorption peak of polysaccharide PCP-F1 at 260 and 280 nm in Supplementary Figure S3, indicating that the amount of protein and nucleic acid impurity has been reduced to the lowest level, which is less than the minimum detection limit (35). The molecular weights of PCP-F0, PCP-F0.05, PCP-F0.1, and PCP-F1 were estimated as 107.51, 567.86 + 76.24, 228.53, and 37.46 kDa, respectively, based on the high-performance gel permeation chromatography (HPGPC) analysis (Figure 4C).

To understand the types of functional groups contained in molecule PCP-F1 and how they are connected, infrared spectroscopy is used. As shown in Figure 4D, the stretching vibration of O-H residues appears at 3,403, 1,626, and 1,452 cm⁻¹, which are the bending vibration peaks of-COOH (27). The absorption peaks at 1,131 cm⁻¹ represent the asymmetric stretching vibration of C-O-C (36). The absorption peak at 1,026 cm⁻¹ belongs to the C-O-H bond. In addition, the absorption peaks at 880 and 835 cm⁻¹ belong to β-configuration pyranose (30).

The monosaccharide composition of PCP-F1 was Glc, Man, Gal, Rha, and GalA with a molar ratio of 3.5: 2.5: 1.3: 1.8: 0.8 (Figure 5). By methylation and GC–MS analysis of PCP-F1, the methylation results of glycoside linkage PCP-F1 were determined, as shown in Table 1 and Supplementary Figure S4. The type of glycoside linkage PCP-F1 was determined by the MS fragment database and literature analysis (37). Methylation analysis revealed that PCP-F1 was composed of →3) Glcp, →1) Manp (3→, →3) Rhap (4→, →2) Glcp (3→, →1) Glcp (2→, →2) Glcp (6→, →2) Galp (4→, →2, 4) Manp (6→, →2) GalAp (3, 4→/→2) Galp (3, 4 → with a molar percentage of 2.4: 13.3: 13.4: 6.9: 24.2: 1.8: 18.5: 11.8: 7.7.

Due to the complexity of polysaccharide structure, the partial acid hydrolysis of PCP-F1 was performed further to determine the distribution of monosaccharides. In polysaccharide skeletons, the hydrolysis of the branch chain is a priority than that of the backbone (27). In this study, PCP-F1 was divided into four subcomponents (H-1, H-2, H-3, and H-4), and the monosaccharide components of these subcomponents are shown in Table 2. Due to the Man, Gal, and Glc being the main monosaccharide composition of H-1, they are presumed to locate in the side chain. When the concentration of TFA increases to 0.1 M, the contents of Glc and Gal in H-2 are higher than others. Meanwhile, GalA was determined in H-2. In the structurally stable H-4, the Glc, Man, and Gal were found, indicating that they should be located in the core structure of PCP-F1.

Nuclear Magnetic Resonance (NMR) spectroscopy is a widely used and important technique for ascertaining distinct structural attributes. It is employed for verifying monosaccharide classifications, distinguishing between α-and β-anomeric arrangements, and unraveling details about glycosidic connections (38–40). Herein, the ¹³C and ¹H NMR spectra of residues A to I were analyzed through 2D NMR. Regarding the information gleaned from the ¹³C and ¹H spectra of PCP-F1, the specific signals corresponding to different residues are outlined as follows.

Within the 1H NMR spectrum depicted in Supplementary Figure S5A, the presence of a signal at δ_H 4.62 ppm suggests the presence of a β-configured residue within the PCP-F1 structure. This signal is attributed to H-1 of residue D (38). Moreover, the regions between δ_H 4.05 and δ_H 3.43 ppm were responsible for H-2 to H-6 of residues (A, B, C, D, E, F, G, H, and I) presented in PCP-F1 (39). In ¹³C NMR spectra depicted in Supplementary Figure S5B, nine peaks at δ_C 106.47, 92.68, 104.33, 104.28, 102.29, 96.37, 106.06, 103.78, and 104.35 ppm were assigned as the anomeric signals situated at C-1 of residues A–I, respectively. In the more elevated field segment, the resonance at δ_C 166.73 ppm was attributed to the carbonyl group of the uronic acid moiety (22).

In order to establish the correlation between the signals of the anomeric carbon and their respective protons, an analysis of the HSQC spectrum depicted in Figure 6A was conducted. First, the signals at δ_H/C 4.31/104.28 and 4.51/106.06 ppm were attributable to heterologous region H-1/C-1 of the residue D and G, respectively (40). The ¹H-¹H COSY (Figure 6B) and NOESY (Figure 6C) spectrum revealed that these cross-peaks at δ_H 4.62/4.02, 4.02/3.81, 3.81/3.96, 3.96/3.79, and 3.79/3.68 ppm were consistent with the H-2 to H-6 signal of residue D at δ_H 4.02, 3.81, 3.96, 3.79, and 3.68 ppm, respectively. In addition, six cross peaks at δ_H 4.65/3.60, 3.60/3.73, 3.73/3.68, 3.68/3.55, and 3.55/3.53 ppm which were consistent with the H-2 to H-6 signal of residue G at δ_H 3.60, 3.73, 3.68, 3.55, and 3.53 ppm, respectively.

The HMBC spectrum facilitated the analysis of correlations among nine residues (A, B, C, D, E, F, G, H, and I). As depicted in Figure 6D, a reciprocal relationship was observed between the H-2 signal of residue A (δ_H 4.00 ppm) and the C-6 of residue H (δ_C 77.91 ppm), as well as between the H-6 signal of residue H (δ_H 3.43 ppm) and the C-2 of residue A (δ_C 71.76 ppm), signifying a presence of a 2,6-linkage between residue A and H. Similarly, a positive interaction was noted between the C-3 signal of residue B (δ_C 74.42 ppm) and the H-6 of residue F (δ_C 3.61 ppm), and between the C-6 signal of residue F (δ_C 71.53 ppm) and the H-3 of residue B (δ_H 3.56 ppm), indicating a 3,6-linkage between residue B and F.

Further examinations revealed the existence of 1,2-linkages between residue G and both F and H. The C-1 signal of residue G (δ_C 106.06 ppm) displayed a positive correlation with the H-2 of residue F (δ_H 3.96 ppm), and the C-2 signal of residue F (δ_C 74.92 ppm) exhibited a correlation with the H-1 of residue G (δ_H 4.65 ppm), illustrating the 1,2-linkages between residue G and F. Additionally, the C-1 signal of residue G (δ_C 106.06 ppm) showed a positive correlation with the H-2 of residue H (δ_H 3.62 ppm), and the C-2 signal of residue H (δ_C 73.09 ppm) correlated with the H-1 of residue G (δ_H 4.65 ppm), indicating 1,2-linkages between residue G and H. The linkage type between residue B and D was determined to be a 3,4-linkage based on the highly correlated H-3 signal of residue B (δ_H 3.56 ppm) and the C-4 of residue D (δ_C 73.17 ppm), as well as the correlation between the C-3 signal of residue B (δ_C 74.42 ppm) and the H-4 of residue D (δ_H 3.96 ppm). Similarly, the linkage type between residue G and D was identified as a 1,2-linkage, with the H-1 signal of residue G (δ_H 4.65 ppm) correlating with the C-2 of residue D (δ_C 77.41 ppm), and the C-1 signal of residue G (δ_C 106.06 ppm) showing correlation with the H-2 of residue D (δ_H 4.02 ppm).

Lastly, a 2,3-linkage between residue E and I was established, as evidenced by the correlation between the C-2 signal of residue E (δ_C 77.39 ppm) and the H-3 of residue I (δ_H 3.82 ppm), along with the correlation between the H-2 signal of residue E (δ_H 3.98 ppm) and the C-3 of residue I (δ_C 88.05 ppm). A comprehensive assignment of PCP-F1 was outlined in Table 3.

The triple helix structure of PCP-F1 was characterized by Congo red experiment. Under weakly alkaline conditions, the triple helical polysaccharide can form a complex with Congo red, and compared with Congo red solution, the maximum absorption wavelength will be red-shifted (41). With the increase of concentration of alkali, the complex triple helix structure will be destroyed, leading to the decrease of the maximum absorption wavelength. As shown in Figure 7A, the maximum absorption wavelength of PCP-F1 was 496 nm. Compared with the Congo red solution, the maximum absorption wavelength of PCP-F1 polysaccharide showed a weak red-shift trend, indicating that no triple helix conformation was formed.

Monosaccharides and glycoside chains are structural components associated with polysaccharide activity. These structural characteristics include composition, conformation, molecular weight, and functional groups. Combined with these results and previous reports (38–42), the structure of PCP-F1 was tentatively determined to be a heteroglycan comprised of nine individual residues with side chains →3)-α-D-Glcp, →2)-α-D-Glcp (6→, →1)-ꞵ-D-Glcp (2→, →2)-α-D-GalAp (3,4→, →1) -ꞵ-D-Manp (3→, →2)-α-D-Glcp (3→, branched located at O-3 position of →3)-α-D-Glcp, →2)-ꞵ-D-Galp (4→, →1)-ꞵ-D-Glcp (2→, →2,4)-α-D-Manp (6→, →3)-α-L-Rhap (4→, and corresponding structural motif is depicted in Figure 7B.

Processed PCPs solution with a concentration of 1 mg/mL was prepared and tested at 25°C by Starter 3100 pH meter. Processed PCPs and PCP-F1 were, respectively, mixed with dried potassium bromide (1, 100, w/w) to ground and pressed in a vacuum at 25°C. A Nicolet 5700 IR spectrometer given corresponding spectra in the wavelength region of 4,000–400 cm⁻¹.

The uronic acid content and total sugar content of processed PCPs were measured by the anthrone-sulfuric acid and m-hydroxyphenyl methods, respectively (27, 30).

PCP-F1 solution (1 mg/mL) was prepared and scanned by a SHIMADZU spectrophotometer at a wavelength range of 200–800 nm at 25°C. Monosaccharide composition analysis was conducted employing Gas Chromatography-Single Quadrupole Mass Spectrometry (GC–MS) methodology. Briefly, PCP-F1 (20 mg) was introduced into a solution of trifluoroacetic acid (TFA) with a concentration of 2 mol/L and a volume of 2 mL. Subsequently, hydrolysis was performed at 100°C for a duration of 6 h, followed by evaporation of the solvent under reduced pressure. The acid-hydrolyzed analyte, dissolved in anhydrous DMSO (2 mL), was combined stepwise with NaOH powder (60 mg) in a flask. Following stirring at 35°C for 30 min, methylating reagent CH₃I (1 mL) was carefully added drop by drop. The reaction proceeded in darkness for 12 h, after which it was halted by introducing ultrapure water (2 mL). The resulting methylated sample was subjected to extraction using dichloromethane and subsequently subjected to analysis via GC–MS. PCP-F1 homogeneity was determined based on a previous method (22).

Glycosidic linkage analysis of PCP-F1 was conducted with a slight modification to a previously documented method (35, 37, 52). The initial step involved reducing the uronic acid to a neutral sugar prior to methylation analysis. To achieve this, 10 mg of dried PCP-F1 was fully dissolved in anhydrous DMSO (2 mL). Subsequently, NaOH (60 mg) was added, and the reaction mixture was stirred at 35°C for a span of 2 h. Following this, a slow addition of 1 mL of CH₃I took place. The reaction proceeded under dark conditions for 12 h. Termination of the reaction occurred by adding ultrapure water (2 mL). The resultant methylated sample underwent extraction using dichloromethane and was washed thrice with ultrapure water. Complete methylation of PCP-F1 was confirmed by the disappearance of peaks in the infrared spectrum within the range of 3,200–3,700 cm⁻¹. The ensuing procedure involved hydrolyzing the mixture using 2 mL of 2 M TFA at 100°C for a period of 6 h. Post removal of excess TFA, NaBH₄ (30 mg) and a solution of 0.05 M NaOH (1 mL) were sequentially introduced. After a reaction time of 12 h, 100 μL of acetic acid was added, followed by solvent removal under vacuum. Subsequently, a mixture of pyridine and acetic anhydride (1 mL each) was added, sealed, and stirred at 90°C for an additional 2 h. The reacted solution was subjected to extraction using dichloromethane and subsequently analyzed using the GC–MS technique.

According to a previous method (27), partial acid hydrolysis of PCP-F1 was performed. 20 mg PCP-F1 was hydrolyzed with 2 mL TFA for 1 h (0.05 M, 100°C). TFA was removed, and four volumes of ethanol (95%) were added to the hydrolysate. The sample was stored at 4°C overnight for centrifugation (5,000 rpm, 5 min), the supernatant was collected and named H-1, and the precipitation was hydrolyzed again with 2 mL TFA (0.1 M, 100°C) for 1 h. Similarly, after removing the TFA with methanol, the supernatant was obtained and named H-2. The precipitation was separated and hydrolyzed with 2 mL TFA with a concentration of 0.5 mol/L at 100°C for 1 h to get H-3 and the final precipitated H-4. The H-1, H-2, H-3, and H-4 samples were completely hydrolyzed using 2 mL TFA (2 M, 100°C) for 6 h to further analysis.

Nuclear magnetic resonance (NMR) measurements were performed using 0.6 mL of D₂O dissolved with 20 mg of dry polysaccharide PCP-F1 at 25°C. A 600 Hz NMR spectrometer was utilized to record the 1D ¹H NMR and ¹³C NMR, as well as the 2D ¹H-¹H COSY, HSQC, HMBC, and NOESY spectra of the PCP-F1.

According to a previous method (41), the Congo red experiment of PCP-F1 was performed. The polysaccharide solution (1 mL, 1 mg/mL) and Congo Red solution (100 μM) at the ratio of 1: 1 (v/v) were mixed with different concentrations of NaOH (1 mL, 0, 0.05, 0.10, 0.15, 0.20, 0.25, 0.30, 0.40, and 0.50 M) at 25°C for 30 min, and the final solution was determined at a wavelength of 400–600 nm.

The capacity of scavenging DPPH radicals of processed PCPs and purified 4-Z PCPs, including PCP-F0, PCP-F0.05, PCP-F0.1, and PCP-F1, were detected according to a previous report with slight modifications (36). Briefly, first, fresh DPPH (0.2 mM in methanol, 2.0 mL) and PCPs solutions (2.0 mL) at various concentrations (0, 0.2, 0.4, 0.6, 0.8, and 1.0 mg/mL) were prepared. Second, the reaction mixture of prepared DPPH and PCPs solutions at various concentrations was thoroughly stirred and hatched at 25°C for 30 min under dark conditions. The UV-2550 UV spectrophotometer recorded the absorbance of the mixture at a wavelength of 517 nm. Finally, the activity of DPPH radical scavenging was calculated.

The capacity of scavenging hydroxyl radicals of processed PCPs and purified 4-Z PCPs, including PCP-F0, PCP-F0.05, PCP-F0.1, and PCP-F1, were measured refer to a previous report (53). First, PCPs solutions (1.0 mL) at various concentrations (0.2, 0.4, 0.6, 0.8, 1.0, 2.0, 4.0, 6.0, 8.0, and 10.0 mg/mL) were prepared. Second, the reaction mixture of prepared sample solutions (1 mL), FeSO₄ solution (1 mL, 6 mM), H₂O₂ solution (1 mL, 6 mM), and salicylic acid (1 mL, 2 mM in ethanol) was mixed thoroughly and incubated at 37°C for 30 min. The UV-2550 ultraviolet spectrophotometer determined the absorbance of the mixture to a wavelength of 510 nm. Finally, the activity of hydroxyl radical scavenging was measured.