Preliminary Exploration of Structure-Immunostimulatory Activity Correlation of Spherical Pectin from Chrysanthemum Tea Infusion

Abstract

The spherical pectin is an important bioactive component of chrysanthemum tea infusion, but its biological function, primary structure, and structure-activity relationship remain unclear. The present study evaluated the immunostimulatory activity of spherical pectin from Chrysanthemum morifolium Ramat. ‘Hangbaiju’ tea infusion in RAW264.7 cells and preliminarily investigated its structure-immunostimulatory activity relationship. The rhamnogalacturonan-I (RG-I) domain played a key role in the immunostimulatory activity of spherical pectin. Terminal and branched arabinose residues together accounted for 73.8% of the total arabinose residues in spherical pectin, indicating that the arabinan chains of spherical pectin were highly branched. The backbone of these arabinan chains consisted of →5)-α-Araf-(1→ repeats, and additional →5)-α-Araf-(1→ branches were linked to the backbone via α-1,3-glycosidic linkages. The spherical pectin rich in highly branched arabinan chains activated RAW264.7 cells via recognition by toll-like receptor 4 (TLR4). Molecular docking analysis revealed that →5)-α-Araf-(1→ branches in spherical pectin could bind to toll-like receptor 4/myeloid differentiation protein-2 (TLR4/MD-2) complexes and stabilize the dimer structure, which represents an important mechanism for its immunostimulatory activity. This study provides new insights into the structure-function relationship of spherical pectin.

1. Introduction

Chrysanthemum tea, widely consumed worldwide, is considered a functional beverage for the treatment of headaches, swelling, eye pain, and poisoning [1]. Several edible chrysanthemum varieties, including Chrysanthemum morifolium Ramat., Chrysanthemum indicum L., and Coreopsis tinctoria Nutt., are commonly used as raw materials for chrysanthemum tea [2]. Our previous work revealed that submicroparticles constituted 22% of the total soluble solids in chrysanthemum tea infusion prepared from Chrysanthemum morifolium Ramat. ‘Hangbaiju’ [3]. The submicroparticles isolated from C. morifolium Ramat. ‘Hangbaiju’ tea infusion had a spherical skeleton formed by esterified pectin and adsorbed 23 individual phenolic compounds [3]. The spherical pectin represented over 90% of the total content in submicroparticles isolated from C. morifolium Ramat. ‘Hangbaiju’ tea infusion [3]. Special hawthorn pectin has been reported to exhibit gut microbiota-modulating effects [4], and special lotus leaf pectin shows regulating effects on glycolipid absorption and metabolism [5]. The spherical pectin was believed to be an important contributor to the biological functions of submicroparticles isolated from C. morifolium Ramat. ‘Hangbaiju’ tea infusion. Only limited preliminary structural characteristics of this spherical pectin were reported in our previous work, such as monosaccharide composition, homogeneity, molecular weight, and conformation [3]. Further investigation is required to explore the biological functions, primary structure, and structure-activity relationship of this spherical pectin.

Polysaccharides exhibit significant immunomodulatory effects, which play an important role in defending against cancer invasion and boosting host disease resistance [6]. The mechanisms by which polysaccharides enhance immunity include promoting the growth of immune organs, activating macrophages, dendritic cells, lymphocytes, and natural killer cells, as well as increasing the release of immune-related molecules such as cytokines, antibodies, and complement molecules [7]. As previously reported, polysaccharides from C. morifolium Ramat. exhibit immune-enhancing activities by promoting the proliferation of RAW264.7 cells and upregulating the release of tumor necrosis factor α (TNF-α), interferon γ (IFN-γ), and nitric oxide (NO) [8]. It is worth mentioning that the structural characteristics of pectin determine its immunomodulatory effects on the immune system. It is generally considered that pectin comprises four distinct domains: homogalacturonan (HG), rhamnogalacturonan-I (RG-I), rhamnogalacturonan-II (RG-II), and xylogalacturonan (XGA) [9]. Our previous study reported that the spherical pectin from C. morifolium Ramat. ‘Hangbaiju’ tea infusion consisted of HG and RG-I domains at a molar ratio of 1:1.16 [3]. Previous studies [10,11] have demonstrated that pectin rich in the HG domain displays immunosuppressive activity, whereas pectin abundant in the RG-I domain exerts immunostimulatory effects. The similar proportions of HG and RG-I domains lead to unpredictable effects of the spherical pectin from C. morifolium Ramat. ‘Hangbaiju’ tea infusion on the immune system. Whether this spherical pectin possesses immunostimulatory activity remains to be verified. Our previous study inferred that the repetitive alternating arrangement of RG-I domains and polygalacturonic acid segments in the sugar chains contributed to the formation of the spherical conformation of pectin from C. morifolium Ramat. ‘Hangbaiju’ tea infusion [12]. More interestingly, in the aforementioned study [12], a soluble polygalacturonic acid obtained by acid hydrolysis of spherical pectin from C. morifolium Ramat. ‘Hangbaiju’ tea infusion was found to form a spherical conformation in aqueous solution via hydrophobic interactions, which was highly similar to that of the original spherical pectin, even though its molecular weight was considerably lower. It was hypothesized that a comparative evaluation of the immunostimulatory activities of the spherical pectin and this water-soluble polygalacturonic acid could, to some extent, reflect the contribution of the HG domains in spherical pectin to its immunostimulatory activity.

Polysaccharides are widely considered to enhance immunity mainly by binding to pattern recognition receptors (PRRs) on the surface of immune cells, thereby activating related immune signaling pathways [7]. Determining whether the spherical pectin from C. morifolium Ramat. ‘Hangbaiju’ tea infusion exerts immunostimulatory activity by recognizing PRRs, which will help to clarify the structure-immunostimulatory activity correlation. The branched neutral sugar side chains in the RG-I domain are considered vital to the immunostimulatory activity of a pectin [11]. The branching architecture confers great flexibility upon the RG-I domain, which may benefit the recognition and binding of pectin with PRRs [13]. However, the neutral sugar side chains of the RG-I domain, which comprise arabinan, galactan, and/or arabinogalactan, exhibit complex fine structures [11]. Neutral sugar side chains with distinct fine structures exhibit different immunostimulatory activities [10,14]. Therefore, a structural characterization of the RG-I domain in spherical pectin, particularly the detailed structure of its neutral sugar side chains, will help elucidate the structure-activity relationship underlying the immunostimulatory activity of spherical pectin from C. morifolium Ramat. ‘Hangbaiju’ tea infusion. Preliminary exploration of the interactions between the neutral sugar side chains in this spherical pectin and PRRs will help analyze the reasons why this spherical pectin exerts immunostimulatory activity.

In the present study, the immunostimulatory activities of spherical pectin from C. morifolium Ramat. ‘Hangbaiju’ tea infusion and its HG domain (a water-soluble polygalacturonic acid) were compared in RAW264.7 cells to verify the immunostimulatory function of spherical pectin and, to a certain extent, reflect the contribution of the HG domain to its immunostimulatory activity. Furthermore, the primary structure of the RG-I domain in spherical pectin was characterized by linkage and nuclear magnetic resonance (NMR) analysis. Finally, the preliminary exploration of interactions between spherical pectin and RAW264.7 cells was conducted by membrane receptor neutralization assay and molecular docking analysis. The obtained results will add to our understanding of the immunostimulatory activity of the spherical pectin and the health effects of chrysanthemum tea. In addition, the present study would facilitate the development of food applications of immune-enhancing pectin.

2. Materials and Methods

2.1. Materials and Reagents

The dried floral materials of C. morifolium Ramat. ‘Hangbaiju’ were acquired from Efuton Tea Co., Ltd. (Tongxiang, Zhejiang, China) in November 2021 and preserved at −20 °C until use. Rhamnose (Rha), galacturonic acid (GalA), galactose (Gal), arabinose (Ara), and D₂O were acquired commercially from Macklin (Guangzhou, China). Endo-polygalacturonase (EC 3.2.1.15, 5000 U) was purchased from Megazyme International Ireland Ltd. (Wicklow, Ireland). l-ethyl-3-(3-dimethylaminopropyl) carbodiimide (EDC), sodium borohydride (NaBH₄), hydrochloric acid (HCl), sodium hydroxide (NaOH), and trifluoroacetic acid (TFA), as well as all other chemicals, were of analytical grade and obtained in China. RAW264.7 cell line was obtained from the National Experimental Cell Resource Sharing Platform of China (Beijing, China). High glucose Dulbecco’s modified Eagle medium (DMEM), phosphate-buffered saline (PBS), and Penicillin-Streptomycin were purchased from Gibco Life Technologies (Grand Island, NY, USA). Fetal bovine serum (FBS) was purchased from Sigma Chemical Co. (St Louis, MO, USA). Lipopolysaccharides (LPS), Cell-Counting Kit-8 (CCK-8), and NO Assay Kit were purchased from Beyotime Institute of Biotechnology (Shanghai, China). Toll-like receptor 4 neutralizing antibody (anti-TLR4) was purchased from Thermo Fisher Scientific, Inc. (Waltham, MA, USA). A Milli-Q water purification system (Millipore, Bedford, MA, USA) was used to produce ultrapure water.

2.2. Preparation of Spherical Pectin from C. morifolium Ramat. ‘Hangbaiju’ Tea Infusion

The spherical pectin isolated from submicroparticles in C. morifolium Ramat. ‘Hangbaiju’ tea infusion was named CTI-P-PS according to our previously reported protocol [3]. C. morifolium Ramat. ‘Hangbaiju’ flower powder was immersed in demineralized water and heated in a boiling water bath for 30 min. Sequential filtration through cotton gauze and vacuum filtration afforded C. morifolium Ramat. ‘Hangbaiju’ tea infusion (CTI). Using an integrated membrane system equipped with a 10 kDa hollow fiber ultrafiltration unit, submicroparticles (CTI-P) were harvested from CTI after seven successive ultrafiltration cycles. After dephenolization with XAD-16 macroporous resin, CTI-P was precipitated by ethanol, redissolved in deionized water, and dialyzed exhaustively against deionized water using a 3.5 kDa molecular weight cutoff (MWCO) dialysis membrane to yield the purified spherical pectin (CTI-P-PS).

2.3. Separation of HG Domain from Spherical Pectin by Acid Hydrolysis

The acid hydrolysis of spherical pectin was performed according to our previously reported method [12]. CTI-P-PS was subjected to acid hydrolysis with 1 mol/L TFA at 100 °C for 3 h. After being cooled to room temperature, the resulting hydrolysate was concentrated to dryness by rotary evaporation under vacuum. The residue was reconstituted in deionized water and subjected to centrifugation at 3622× g for 20 min at room temperature. The supernatant was dialyzed against deionized water for 48 h with a dialysis bag (MWCO: 3.5 kDa) and freeze-dried. The product was named CTI-P-PS-UI.

2.4. Immunostimulatory Activity Evaluation

Immunostimulatory activity evaluation was performed according to Wang et al., with some modifications [8].

RAW264.7 cells were cultivated in complete medium consisting of high glucose DMEM, 10% FBS, and 1.1% penicillin–streptomycin, at 37 °C with 5% CO₂ in a humidified incubator.

RAW264.7 cells were seeded at 5 × 10⁴ cells/well in 100 μL culture medium into 96-well plates, then allowed to preincubate for 24 h. After replacing the supernatants with fresh complete medium, the cells were treated with 100 µL of CTI-P-PS/CTI-P-PS-UI at 1, 50, 250, or 1000 μg/mL. The cells treated with PBS and LPS (20 µg/mL) served as the blank and positive control groups, respectively. After continuous culturing for 24 h, the supernatants were removed, and the cells were washed with 200 µL of high-glucose DMEM. Complete medium (100 µL) and CCK-8 solution (10 µL) were appended, followed by another cultivation for 1 h. The absorbance was measured at 450 nm using a ReadMax 1900 ultraviolet spectrophotometer (Shanpu Biotechnology Co., Ltd., Shanghai, China). Cell viability was calculated by the following equation:

$$\mathrm{Cell} \mathrm{viability} (\%)={\displaystyle \frac{\mathrm{A}2-\mathrm{A}0}{\mathrm{A}1-\mathrm{A}0}} \times 100$$

(1)

where the A0 is the absorbance of the complete medium and CCK-8 solution. A1 is the absorbance of the blank control group. A2 is the absorbance of the positive control group or the sample treated group. Measurements were carried out on six samples.

RAW264.7 cells (100 µL, 5 × 10⁶ cells/mL) were planted onto a 96-well plate. After incubating the cells for 24 h, the supernatants were replaced with fresh complete medium and 100 µL of CTI-P-PS/CTI-P-PS-UI at 1, 50, 250, or 1000 μg/mL, followed by a further 48 h of culturing. Following the collection of the culture supernatants, NO secretion levels were quantified using the Griess method. The blank and positive controls were treated with PBS and LPS (20 µg/mL), respectively. Measurements were carried out on six samples.

2.5. Linkage Analysis

The reduction in uronic acids in CTI-P-PS was carried out based on the method reported by Taylor and Conrad, with some modifications [15]. CTI-P-PS (40 mg) was dissolved in 20 mL of deionized water; subsequently, 300 mg of EDC was added. During reaction progress, the pH of the mixture was held at 4.75 via titration with 0.01 mol/L HCl. After two hours, ten milliliters of NaBH₄ solution (4 mol/L) was added slowly to the reaction mixture with titration by HCl (2 mol/L) for maintaining pH at 7.00. The addition of NaBH₄ should be completed in 45 min. Following agitation at ambient temperature for 2 h, the resulting mixture was dialyzed against deionized water for 48 h using a 3.5 kDa MWCO dialysis membrane and then lyophilized. The reduction was repeated until CTI-P-PS was reduced completely. Successful preparation of the reduced CTI-P-PS was verified by the absence of characteristic absorption peaks at 1740 and 1615 cm⁻¹ in attenuated total reflection (ATR)–Fourier-transform infrared (FT-IR) spectrum. FT-IR spectra were recorded on a Thermo Scientific Nicolet iS 5 spectrometer (Thermo Fisher Scientific Inc., Waltham, MA, USA) equipped with an iD5 ATR diamond accessory, over a range of 4000–400 cm⁻¹ at 4 cm⁻¹ resolution with 32 scans.

The reduced CTI-P-PS was methylated according to a previous study with some modifications [16]. An amount of 10 mg of reduced CTI-P-PS was dissolved in 2 mL of anhydrous DMSO. Following the immediate addition of 50 mg NaOH powder, 1 mL of CH3I was gradually added dropwise to the solution over 20 min. After being stirred at room temperature for 2 h, the reaction was quenched by adding 2 mL of deionized water. After evaporation under vacuum, the final mixture was dialyzed against deionized water for 48 h with a 3.5 kDa MWCO dialysis membrane and freeze-dried. The methylation process was repeated until complete methylation of reduced CTI-P-PS was achieved. Complete methylation of CTI-P-PS was confirmed by the disappearance of the OH band (3200–3700 cm⁻¹) in the ATR-FT-IR spectrum.

The methylated CTI-P-PS was converted into partially methylated alditol acetates (PMAAs) using the procedure of a previous study with some modifications [17]. Sequential acid hydrolysis of methylated CTI-P-PS was performed using 3 mL of 98% formic acid (100 °C, 6 h) and 4 mL of 4 mol/L TFA (110 °C, 6 h). After treatment with NaBH4 at room temperature for 3 h, the resulting monosaccharides were acetylated to alditol acetates by incubation with pyridine and acetic anhydride at 100 °C for 1.5 h. After adding deionized water and dichloromethane, the aqueous layer was removed, and the extraction procedure was repeated in triplicate to eliminate residual reagents. After collection, the organic phase was dried with anhydrous sodium sulfate and evaporated under vacuum. The concentrate was filtered through a 0.22 μm filter membrane for subsequent GC-MS determination. GC-MS analysis of PMAAs was performed on a TRACE DSQ II instrument (Thermo Fisher Scientific Inc.‌, Waltham, MA, USA) fitted with a TG-5SILMS capillary column (30 m × 0.25 mm, 0.25 μm). The column temperature program began at 150 °C for 2 min, followed by a ramp to 180 °C at 10 °C/min (held 2 min) and a second ramp to 260 °C at 15 °C/min (held 8 min). The ion source of the mass spectrometer was maintained at 250 °C. Peaks corresponding to PMAAs were identified via comparison of their MS fragmentation patterns against the CCRC Spectral Database for PMAA’s (https://glygen.ccrc.uga.edu/ccrc/specdb/ms/pmaa/pframe.html) (accessed on 8 July 2022) and their GC relative retention times with reported references [18,19]. The relative molar ratio was estimated by peak area corrected by molecular mass.

2.6. Separation of RG-I Domain from Spherical Pectin by Enzymatic Hydrolysis

The de-esterification of CTI-P-PS was performed according to a previous study with some modifications [20]. The pH value of the CTI-P-PS solution (10 mg/mL) was maintained at 12.00 by titration with NaOH (1 mol/L) for 2 h. The reaction mixture was neutralized to pH 7.00 using 1 mol/L HCl, followed by dialysis against deionized water for 48 h with a 3.5 kDa MWCO dialysis bag and lyophilization to afford CTI-P-PS-D (the de-esterified form of CTI-P-PS).

The enzymatic hydrolysis of CTI-P-PS-D was performed according to a previous study with minor modifications [21]. CTI-P-PS-D (200 mg) was re-dissolved in citric acid-sodium acetate buffer (0.1 mol/L) with a final concentration of 2 mg/mL and subjected to enzymatic hydrolysis by endo-polygalacturonase (1 U/mL at 40 °C for 72 h, pH 5.5) under gentle stirring. The resultant hydrolysate was heated at 90 °C for 15 min to terminate enzymatic activity. After cooling to room temperature, the hydrolysate was filtered with filter paper and dialyzed against deionized water for 72 h with a dialysis bag (MWCO: 3.5 kDa). After that, the hydrolysate was concentrated and centrifuged at 3622× g for 30 min to obtain supernatant. The supernatant was collected, lyophilized, and termed CTI-P-PS-D-E. CTI-P-PS-D-E was verified as a neutral-sugar-based glycan by analyzing monosaccharide composition according to our previous study [12].

2.7. NMR Spectroscopy Analysis

NMR spectroscopic analysis was carried out following our previously published procedure [12]. CTI-P-PS-D-E (30 mg) underwent deuterium exchange twice through lyophilization with D₂O, followed by dissolution in 500 μL D₂O. 1D NMR spectra (¹H, ¹³C, and DEPT-135) and 2D NMR spectra, including ¹H-¹H correlation spectroscopy (COSY), heteronuclear single quantum coherence (HSQC), heteronuclear multiple bond correlation (HMBC), and nuclear Overhauser effect spectroscopy (NOESY), were recorded on a 600 MHz Bruker AVANCE III HD NMR spectrometer (Bruker, Fällanden, Switzerland). MestReNova software (Mestrelab Research S.L.‌, Santiago de Compostela, Spain, v14.0) was employed to process the acquired NMR spectra.

2.8. Toll-like Receptor 4 (TLR4) Neutralization Assay

The TLR4 neutralization assay was performed according to a previous study with some modifications [22]. After being planted and preincubated for 24 h, RAW264.7 cells were treated with anti-TLR4 at 20 μg/mL for 1h. Thereafter, the harvested RAW264.7 cells were incubated with 250 μg/mL CTI-P-PS for 24 h to detect cell viability and 48 h to measure NO concentration. Cells treated with CTI-P-PS (250 μg/mL) alone served as the positive control. The negative control group was treated with PBS. The cell viability and concentration of NO were detected according to Section 2.4. Measurements were carried out on six samples.

2.9. Molecular Docking

Molecular docking analysis was conducted according to an earlier report with minor modifications [23]. The target proteins (2Z64 and 3VQ2) employed in molecular docking analysis were retrieved from the RCSB Protein Data Bank (RCSB PDB, https://www.rcsb.org/). As putative ligands, galactan and arabinan segments were constructed in ChemDraw 20 (PerkinElmer, Waltham, MA, USA) and optimized by energy minimization before being used for molecular docking analysis. AutoDock Vina on AMDOCK (Alpha Technology (Tianjin) Co., Ltd.‌, Tianjing, China, v1.5.2) was employed to conduct molecular docking analysis. PyMOL (Schrödinger, New York, NY, USA‌, v2.1.0) was utilized for visualizing proteins and related molecular structures.

2.10. Statistical Analysis

Duncan’s multiple-range test and Student’s t-test (p < 0.05) were used to analyze significant differences among the parameter means, and all statistical analyses were conducted with SPSS 19.0 (SPSS Inc., Chicago, IL, USA).

3. Results and Discussion

3.1. Comparative Analysis on Immunostimulatory Activities of Spherical Pectin and Polygalacturonic Acid

Macrophages are important to the innate immune response and mediate host defense [8]. The experiments performed with RAW 264.7 cells (a macrophage-like cell line) are good indicators of the immunobiological activities of polysaccharides [24]. Numerous studies have shown that plant polysaccharides can regulate the immune system and do not exhibit obvious cytotoxicity, and have almost no side effects [25]. The immunostimulatory activities of spherical pectin from C. morifolium Ramat. ‘Hangbaiju’ tea infusion and its HG domain (obtained by acid hydrolysis) were compared in RAW 264.7 cells.

As shown in Figure 1A, CTI-P-PS at 1–1000 µg/mL increased the cell viability of RAW 264.7 cells significantly, which demonstrated that CTI-P-PS effectively promoted the proliferation of RAW 264.7 cells without toxic effects. NO, a signaling molecule explosively released from stimulated and activated macrophages, can kill microorganisms, parasites, and tumor cells to protect the body from external adverse [8,25]. As shown in Figure 1B, compared with the blank group, the RAW 264.7 cells treated with CTI-P-PS at 50, 250, and 1000 μg/mL secreted more NO. It was demonstrated that CTI-P-PS could activate the RAW 264.7 cells. The RAW 264.7 cells activated by CTI-P-PS displayed irregular-shaped structures with visible long pseudopodia (Figure S1). These results showed that CTI-P-PS possessed immunostimulatory activity. The polygalacturonic acid (CTI-P-PS-UI), derived from spherical pectin and devoid of RG-I domains, exhibited morphological characteristics in aqueous solution highly similar to those of the original spherical pectin, including analogous particle size distribution, average hydrodynamic diameter, and spherical morphology [12]. As shown in Figure 1A,B, CTI-P-PS-UI possessed extremely weak proliferation-promoting activity and could not stimulate RAW 264.7 cells to secrete NO. It was demonstrated that the polygalacturonic acid hydrolyzed from spherical pectin (MW = 1.9 kDa) did not possess immunostimulatory activity. A previous study reported that commercially available polygalacturonic acid (MW = 93.9 kDa) exhibited no immunostimulatory activity, whereas its enzymatic hydrolysate (MW = 2.6 kDa) showed distinct immunostimulatory activity, indicating that higher-molecular-weight polygalacturonic acid has a lower probability of exerting immunostimulatory activity [14]. The polygalacturonic acid region (HG domain) derived from spherical pectin exhibited no immunostimulatory activity in its free state (i.e., in CTI-P-PS-UI), indicating that it is highly unlikely to exert immunostimulatory activity when embedded within the spherical pectin. More importantly, the immunostimulatory activity of hydrolyzed polygalacturonic acid products may vary depending on the source of the original pectin. For the spherical pectin derived from C. morifolium Ramat. ‘Hangbaiju’ tea infusion, its immunostimulatory activity is more likely attributable to the RG-I domains with neutral sugar side chains. The subsequent research focused on the primary structure and structure-immunostimulatory activity correlation of the RG-I domain.

3.2. Chemical Structure of RG-I Domain in Spherical Pectin

3.2.1. Linkage Analysis of Spherical Pectin

The spherical pectin (CTI-P-PS) was directly subjected to linkage analysis for preliminary revelation of the linkage sequence among its monosaccharide residues, especially neutral sugar residues in the RG-I domain. First, the GalA residues of CTI-P-PS were fully reduced to Gal. After being fully methylated, CTI-P-PS was hydrolyzed, reduced, and acetylated to obtain PMAAs, followed by GC-MS measurement. The mass spectrograms of PMAAs are shown in Figure S2. The sugar residues and the relative molar ratios of PMAAs were summarized in

Table 1. GC-MS results of PMAAs for CTI-P-PS.

Retention Time (min)	PMAA	Residue		Major Mass Fragment (m/z)	Relative Molar Ratio (%)
5.47	2,3,5-Me3-Ara	T-Araf		43, 71, 87, 101, 117, 129,145,161	9.1
6.79	2,5-Me2-Ara	1,3-Araf		43, 71, 87, 99, 117, 129, 233	0.2
7.33	2,3-Me2-Ara	1,5-Araf		43, 71, 87, 99, 101, 117, 129, 189	5.2
8.42	2-Me-Ara	1,3,5-Araf		43, 85, 99, 117, 127, 141, 159, 172, 201, 217, 261	6.1
8.60	3-Me-Rha	1,2,4-Rhap		43, 87, 101, 129, 143, 159, 172, 189, 203	6.9
8.23	2,3,4,6-Me4-Gal	T-Galp		43, 60, 71, 87, 101, 117, 129, 145, 161, 205	0.4
9.09	2,3,6-Me3-Gal	1,4-Galp		43, 59, 71, 87, 99, 101, 113, 117, 129, 131, 143, 161, 173, 233	65.4
9.29	2,4,6-Me3-Gal	1,3-Galp		43, 87, 101, 117, 129, 161, 173, 203, 217, 233, 277	1.2
9.70	2,3,4-Me3-Gal	1,6-Galp		43, 60, 71, 87, 99,101, 117, 129, 161, 189, 233	0.4
9.79	2,6-Me2-Gal	1,3,4-Galp		43, 87, 117, 129, 143, 159, 185, 203, 217, 231, 245,261,305	1.7
9.98	3,6-Me2-Gal	1,2,4-Galp		43, 60, 87, 99, 113, 129, 173, 189, 233	1.0
10.3	2,3-Me2-Gal	1,4,6-Galp		43, 85, 99, 101, 117, 127, 142, 159, 187, 201, 261, 305	1.3
10.55	2,4-Me2-Gal	1,3,6-Galp		43, 87, 99, 117, 129, 189, 233, 305	0.9
10.83	2-Me-Gal	1,3,4,6-Galp		43, 60, 97, 117, 139, 333	0.1

Green pentagram: arabinose (Araf); green triangle: rhamnose (Rhap); yellow circle: galactose (Galp).

. The main observed derivatives were derived from Rha (6.9%), Gal (72.4%), and Ara (20.6%). The Rha residues in CTI-P-PS only comprised 1,2,4-Rhap (6.9%). The Gal residues from Gal and GalA residues in CTI-P-PS mainly comprised T-Galp (0.4%), 1,4-Galp (65.4%), 1,3-Galp (1.2%), 1,6-Galp (0.4%), 1,3,4-Galp (1.7%), 1,2,4-Galp (1.0%), 1,4,6-Galp (1.3%), 1,3,6-Galp (0.9%), and 1,3,4,6-Galp (0.1%). According to our previous study [3], the monosaccharide composition analysis of CTI-P-PS revealed that Gal and GalA accounted to 11.6% and 52.5% of the monosaccharide residues of CTI-P-PS. It is easy to infer that the GalpA residues in CTI-P-PS are 1,4-GalpA. The Ara residues in CTI-P-PS mainly comprised T-Araf (9.1%), 1,3-Araf (0.2%), 1,5-Araf (5.2%), and 1,3,5-Araf (6.1%).

It is believed that the most important advantage of glycans to activate immune cells is their promotion of the multivalent immune recognition between glycans and PRRs [23]. The rich branches of glycans were the important structural bases of the multivalent immune recognition between glycans and PRRs. The number of terminal and branched neutral sugar residues accounted to 27.5% of monosaccharide residues of CTI-P-PS, suggesting that the RG-I domain of CTI-P-PS possessed rich neutral-sugar-based branches. In particular, the number of terminal and branched Ara residues accounted to 73.8% of the Ara residues of CTI-P-PS, suggesting that arabinan chains of CTI-P-PS were multibranched. Therefore, the rich neutral-sugar-based branches, especially the multibranched arabinan chains of CTI-P-PS, might be the important structural features for its immunostimulatory activity. However, the primary structures of branch chains, including galactan, arabinan, and/or arabinogalactan, are unclear at present. More data were needed to elucidate the fine structures of the RG-I domain in CTI-P-PS.

3.2.2. NMR Analysis of Spherical Pectin

NMR spectroscopy enables the characterization of key structural features of polysaccharides, such as monosaccharide composition, anomeric configuration (α/β), glycosidic linkage patterns, and residue sequence [26]. The spherical pectin (CTI-P-PS) was directly subjected to NMR analysis using cryogenic probes to try to obtain effective NMR signals. The ¹H-NMR, ¹³C-NMR, HSQC, COSY, HMBC, and NOESY spectra of CTI-P-PS are shown in Figure S3.

As seen in Figure S3A, the signals at 52.7/3.73 ppm and 20.0/2.11 ppm were attributed to methyl and acetyl groups attached to GalpA residues. The signal at 170.5 ppm in the ¹³C-NMR spectrum was assigned to C6 of GalpA residues. The chemical shift in C6 in the GalpA residue was less than 175 ppm, which demonstrated that the carboxyl group in the GalpA residue was esterified. The anomeric carbon of the ¹³C NMR spectrum for CTI-P-PS showed the signal at 100.3 ppm corresponding to C1 of the GalpA residue. As illustrated in the HSQC spectrum (Figure S3B), the anomeric signal (C1/H1) at 100.3/4.88 was assigned to the GalpA unit (residue a). In the COSY spectrum (Figure S3C), the off-diagonal cross-peak a1-a2 at 4.88/3.63 defined the position of H2 of residue a. Using the same approach, the chemical shifts in H3, H4, H5, C2, C3, C4 and C5 of residue a were assigned as 3.92, 4.38, 5.03, 67.8, 68.1, 78.9 and 70.1 ppm, respectively. In general, pyranose sugars exhibiting anomeric proton signals above 4.70 ppm typically present an α-anomeric configuration, whereas those below this chemical shift correspond to the β-configuration [27]. The residue was identified as α-configuration. Literature data showed that the C4 signal of unsubstituted GalpA residues usually appeared at approximately 71 ppm, while glycosylation at this carbon would result in a downfield shift of 6–10 ppm [27]. The C4 hydroxyl group of residue a was substituted by a sugar moiety, as supported by its C4 chemical shift of 78.9 ppm. Therefore, the residue was 1,4-α-GalpA(OMe). Generally, the HMBC spectrum was utilized to identify the glycosidic linkage sequence via anomeric cross-peaks, with the NOESY spectrum providing further validation of the structural assignments [26]. Owing to the weak signal intensity in the HMBC spectrum (Figure S3D), no long-range correlations were observed between H1 and C4, or between C1 and H4. However, a distinct NOE correlation between H1 and H4 was detected at 4.88/4.38 in the NOESY spectrum (Figure S3E). It was tentatively confirmed that esterified GalpA residues in the HG domain were repeatedly connected via α-1,4-glycosidic linkages, which was consistent with the linkage analysis result of CTI-P-PS.

The anomeric carbon of the ¹³C NMR spectrum (Figure S3A) for CTI-P-PS showed the signals at 102.5 and 104.3 ppm corresponding to C1 of Galp residues, and the signals at 107.0, 107.4, and 109.1 ppm corresponding to C1 of Araf residues. According to the HSQC spectrum (Figure S3B) and the COSY spectrum (Figure S3C), the positions of C1, C2, H1, and H2 of 3 Galp residues (b, c, and d) and 3 Araf residues (e, f, and g) were defined. However, the positions of C3, C4, C5, C6, H3, H4, H5, and H6 of them could not be defined due to low abundance and overlapping of signals. The signals of Rha residues were too weak to be detected. The C/H chemical shifts in neutral monosaccharide residues in CTI-P-PS were difficult to assign based on the present NMR spectra. It is difficult to directly elucidate the linkage patterns, sequence, and precise structure of the RG-I domain using the spherical pectin sample, due to its high viscosity resulting from a high GalpA content.

3.2.3. NMR Analysis of Neutral-Sugar Based Glycan

To obtain the fraction containing a small amount of GalpA residues and abundant neutral monosaccharide residues for characterizing the linkage patterns and sequences of neutral monosaccharide residues by NMR spectroscopy, CTI-P-PS was subjected to enzymatic hydrolysis with endo-polygalacturonase. Compared with CTI-P-PS, the content of GalpA residues of CTI-P-PS-D-E decreased a lot (Figure S4), suggesting that the neutral-sugar-based glycan was successfully obtained. CTI-P-PS-D-E solution at 100 mg/mL exhibited good fluidity and low turbidity, indicating that this fraction was suitable for elucidating the linkage patterns of neutral-sugar residues in CTI-P-PS by NMR spectroscopy.

One-dimensional spectra (¹H-NMR, ¹³C-NMR, and DEPT-135) shown in Figure 2A and two-dimensional spectra (HSQC and COSY) shown in Figure 2B,C were used for the assignment of as many hydrogen and carbon signals as possible.

In the anomeric region of the ¹³C-NMR spectrum (Figure 2A), the signals at 92.2, 96.1, 99.1, and 99.0 ppm were assigned to anomeric carbons of GalpA residues. According to the HSQC spectrum (Figure 2B), 4 possible GalpA residues in CTI-P-PS-D-E were revealed. As seen, anomeric signals (C1/H1) at 92.2/5.24, 96.1/4.53, 99.1/4.99, and 99.0/5.03 were respectively attributed to residues A, B, C, and D. In the ¹³C-NMR spectrum (Figure 2A), the specific signals at ~175 ppm were unambiguously assigned to the carboxylic carbon (C6) of residues A-D. In the COSY spectrum (Figure 2C), the off-diagonal cross-peaks including A1-A2, B1-B2, C1-C2, and D1-D2 defined the positions of H2 of residues A, B, C, and D. More off-diagonal cross-peaks were searched for defining the positions of H3, H4, H5, C2, C3, C4, and C5 of residues A, B, C and D.

The signal at 98.5 ppm in the ¹³C-NMR spectrum (Figure 2A) was assigned to the anomeric carbons of Rhap residues. Based on the HSQC spectrum (Figure 2B), the anomeric signal (C1/H1) at 98.5/5.17 was assigned to residue E. Residue E was in the α-configuration, as the chemical shift in its anomeric proton was greater than 4.70 ppm. The specific signal at 1.16 ppm in the ¹H-NMR spectrum (Figure 2A) and the specific signal at 16.5 ppm in the ¹³C-NMR spectrum (Figure 2A) were unambiguously assigned to the H-methyl (H6) and C-methyl (C6) of residue E, respectively. The off-diagonal cross-peak of E1-E2 in Figure 2C defined the position of H2 of residue E as 4.23 ppm. According to the HSQC spectrum, C2/H2 at 81.6/4.23 defined the position of C2 as 81.6 ppm. The off-diagonal cross-peaks at 4.23/3.87, 3.87/3.66 in the COSY spectrum defined the positions of H3 and H4 of residue E as 3.87 and 3.66 ppm, respectively. According to the HSQC spectrum, the chemical shift in C4 of residue E was identified as 80.1 ppm, while the position of C3 could not be defined since the signal of C3/H3 was of low intensity. The off-diagonal cross-peak of 3.94/1.16 in Figure 2C defined the position of H5 of residue E as 3.94 ppm. The signal in the HSQC spectrum at 71.8/3.94 defined the position of C5 as 71.8 ppm.

The signals at 102.9 and 104.4 ppm in the ¹³C-NMR spectrum (Figure 2A) were assigned to anomeric carbons of Galp residues. According to the HSQC spectrum (Figure 2B), anomeric signals (C1/H1) at 102.9/4.40, 102.9/4.46, and 104.4/4.56 were respectively attributed to residues F, G, and H. The residues F, G, and H were identified as β-configurations for their anomeric protons below 4.70 ppm. In the COSY spectrum (Figure 2C), the off-diagonal cross-peaks, including F1-F2, G1-G2, and H1-H2, defined the positions of H2 of residues F, G, and H. More off-diagonal cross-peaks were used to determine the carbon and proton positions of residues F, G, and H. The chemical shifts in C3, H4, and C4 for residues G and H were not assigned, as the corresponding signals were weak. The chemical shifts in H5, H6a, H6b, C5, and C6 in residues F, G, and H could not be assigned due to weak corresponding signals.

The signals at 107.1, 107.5 and 109.2 ppm in the ¹³C-NMR spectrum (Figure 2A) were assigned to anomeric carbons of Araf residues. According to HSQC spectrum (Figure 2B), anomeric signals (C1/H1) at 107.1/5.07, 107.5/5.01 and 109.2/5.16 were respectively attributed to residues I, J and K. For furanose monosaccharides like Araf, the H1 chemical shift is less diagnostic for anomeric configuration, as both α- and β-anomers usually exhibit signals above 5 ppm [27]. Furanose-type sugars were generally identified by comparing experimental spectroscopic data with reported literature values [27]. The literature revealed that chemical shift in C1 of α-Araf was 109.2 ppm whereas that of C1 of β-Araf was 103.1 ppm [27]. The residues I, J and K were identified as α-configurations since the chemical shifts in C1 of Araf residues in CTI-P-PS-D-E were closer to 109.2 ppm. In the COSY spectrum (Figure 2C), the off-diagonal cross-peaks including I1-I2, J1-J2 and K1-K2 defined the positions of H2 of residues I, J and K. Additional off-diagonal cross-peaks were analyzed to determine other carbon and proton positions of residues I, J, and K.

There were 4 GalpA residues (residues A, B, C and D), 1 Rhap residue (Residue E), 3 Galp residues (residues F, G and H) and 3 Araf residues (residues I, J and K) detected by NMR analysis. All defined chemical shifts in residues A-K were summarized in

Table 2. ¹H and ¹³C NMR assignments for CTI-P-PS-D-E.

Residue	Chemical Shift (ppm)
	C1/H1	C2/H2	C3/H3	C4/H4	C5/H5a, H5b	C6/H6a, H6b
(A) 4-α-GalpA	92.2/5.24	68.2/3.76	68.8/3.93	77.9/4.35	71.3/4.71	175.1
(B) 4-β-GalpA	96.1/4.53	71.3/3.42	72.2/3.68	77.1/4.30	74.2/3.99	174.4
(C) 1-α-GalpA	99.1/4.99	68.2/3.65	69.4/3.84	70.7/4.19	72.2/4.72	175.1
(D) 1,4-α-GalpA	99.0/5.03	68.2/3.70	68.8/3.93	78.8/4.52	71.3/4.76	175.1
(E) 1,2,4-α-Rhap	98.5/5.17	81.6/4.23	n.d./3.87	80.1/3.66	71.8/3.94	16.5/1.16
(F) 1,?-β-Galp (1) *	102.9/4.40	70.7/3.46	72.2/3.60	n.d./n.d.	n.d./n.d.	n.d./n.d., n.d.
(G) 1,?-β-Galp (2) *	102.9/4.46	69.9/3.58	n.d./3.69	n.d./n.d.	n.d./n.d.	n.d./n.d., n.d.
(H) 1,4,?-β-Galp *	104.4/4.56	72.2/3.60	n.d./3.69	77.7/4.09	n.d./n.d.	n.d./n.d., n.d.
(I) 1-α-Araf	107.1/5.07	80.8/4.05	76.5/3.87	83.9/3.95	61.1/3.74, 3.64
(J) 1,3,5-α-Araf	107.5/5.01	80.8/4.05	83.9/3.88	82.2/4.13	66.5/3.80, 3.72
(K) 1,5-α-Araf	109.2/5.16	82.2/4.13	76.7/3.93	83.9/4.05	66.5/3.86, 3.80

n.d.: Not determined, due to low abundance. *: Residues that cannot be fully identified.

. As documented in previous reports [18,27], the C1 chemical shifts in free reducing α-GalpA and β-GalpA were 93.1 and 96.9 ppm, respectively, whereas the C1 signals of non-reducing GalpA residues in pectin were typically observed at 99–102 ppm. The residue A was identified as α-GalpA in the free reducing form and the residue B was identified as β-GalpA in the free reducing form. The C4 chemical shifts in residues A and B showed a downfield shift of approximately 6 ppm relative to 71 ppm, indicating that the C4 hydroxyl groups of residues A and B were substituted by sugar residues. The residues A and B were 4-α-GalpA and 4-β-GalpA, respectively. The residues C and D were identified as α-GalpA with the non-free reducing form. The C4 hydroxyl group of residue D was substituted by a sugar residue. The residues C and D were 1-α-GalpA and 1,4-α-GalpA, respectively. Through comparing chemical shifts with reference [27], the residues E was identified as 1,2,4-α-Rhap. Based on the comparison of chemical shifts with those reported in reference [27], the residues F, G and H were identified as non-reducing β-Galp. The substitution sites of residues F and G could not be identified because the chemical shifts in H4, H5, H6a, H6b, C4, C5 and C6 of residues F and G were unable to identified. The residues F and G were tentatively recorded as 1,?-β-Galp (1) and 1,?-β-Galp (2), respectively. The C4 hydroxyl group of residue H was substituted by a sugar residue. The chemical shifts in H5, H6a, H6b, C5 and C6 of residue H were also unable to identified. The residue H were tentatively recorded as 1,4,?-β-Galp. According to chemical shift comparisons with reference [27], the residues I, J and K were tentatively identified as 1-α-Araf, 1,3,5-α-Araf and 1,5-α-Araf, respectively.

The inter-residue sequences of CTI-P-PS-D-E were determined by analyzing the anomeric proton and carbon cross-peaks in the HMBC spectrum (Figure 3A), and the NOESY spectrum (Figure 3B) was used to verify and supplement the structural assignments. The structural sequences 1–6 of CTI-P-PS-D-E were tentatively deduced and schematically depicted in Figure 3C. In the HMBC spectrum, key long-range correlations were observed at 4.99/77.9 (C H1–A C4) and 99.1/4.35 (C C1–A H4), demonstrating that GalpA residues C and A are connected by an α-1,4-glycosidic linkage (sequence 1). The distinct correlations at 5.03/77.1 (D H1–B C4) and 99.0/4.30 (D C1–B H4) revealed that GalpA residues D and B are connected by an α-1,4-glycosidic linkage (sequence 2). The long-range correlations at 5.03/78.8 (D H1–D C4) and 99.0/4.52 (D C1–D H4) in the HMBC spectrum indicated that GalpA residue D was repeatedly linked via an α-1,4-glycosidic linkage (sequence 3). The long-range correlations at 5.03/78.8 (D H1–E C2) and 99.0/4.52 (D C1–E H2) indicated that GalpA residue D was linked to Rhap via an α-1,2-glycosidic linkage (sequence 4). In the HMBC spectrum, key long-range correlations were observed at 4.56/77.7 (H H1–H C4) and 104.4/4.09 (H C1–H H4). Meanwhile, the NOESY spectrum showed an NOE correlation at 4.56/4.09 (H H1–H H4). These results demonstrated that Galp residue H was repeatedly connected via a β-1,4-glycosidic linkage (sequence 5). However, the complete linkage pattern of galactan in spherical pectin was unable to elucidate the failed identification of Galp residues (F, G, and H). In the HMBC spectrum, the key long-range correlations at 5.01/66.5 (J H1–J C5) and 107.5/3.80 (J C1–J H5a) demonstrated that Araf residue J was repeatedly linked through an α-1,5-glycosidic linkage. In the HMBC spectrum, the characteristic long-range correlations at 5.16/66.5 (K H1–K C5) and 109.2/3.86 (K C1–K H5a) revealed that Araf residue K was repeatedly linked via an α-1,5-glycosidic linkage. In the HMBC spectrum, the long-range correlations at 5.07/83.9 (I H1–J C3) and 107.1/3.88 (I C1–J H3) indicated that Araf residue I was linked to Araf residue J via an α-1,3-glycosidic linkage. In the HMBC spectrum, the long-range correlation at 5.16/83.9 (K H1–J C3) indicated that Araf residue K was linked to Araf residue J via an α-1,3-glycosidic linkage. The Araf residues I, J, and K are linked as sequence 6, shown in Figure 3C.

The NOE correlation at 4.56/3.66 (H H1–E H4) in the NOESY spectrum suggested that Galp residue H was linked to Rhap residue E via a β-1,4-glycosidic linkage. In the HMBC spectrum, the long-range correlation at 107.5/3.66 (J C1–E H4) suggested that Araf residue J was linked to Rhap residue E via an α-1,4-glycosidic linkage. Based on the above HMBC and NOE correlations, the structural information of sequences 1–6, and the primary structure of pectin reported in reference [9], the putative structure of the RG-I domain of CTI-P-PS is schematically illustrated in Figure 3C. The sequence of →4)-α-GalpA-(1→2)-α-Rhap-(1→4)-α-GalpA-(1→2)-α-Rhap-(1→4)-α-GalpA-(1→ was regarded as the backbone of the RG-I domain of CTI-P-PS, and the branches of incomplete galactan (sequence 5) and arabinan (sequence 6) were located at C4 of 1,2,4-α-Rhap (residue E).

A previous study reported that an acidic polysaccharide with an abundant arabinan moiety consisting of 1-Araf, 1,5-Araf, and 1,3,5-Araf exhibited macrophage activating activity [28]. The arabinan branches in the RG-I domain of spherical pectin from C. morifolium Ramat. ‘Hangbaiju’ tea infusion exhibited a relatively high content, well-defined structure, and immune activation potential. The abundant multibranched arabinan chains, which have a backbone composed of →5)-α-Araf-(1→ repeats linked by α-1,5-glycosidic linkages and additional →5)-α-Araf-(1→ branches attached via α-1,3-glycosidic linkages, were regarded as the key structures enabling spherical pectin from chrysanthemum tea infusion to recognize PRRs and activate RAW264.7 cells.

3.3. Preliminary Exploration of Interactions Between Spherical Pectin and Macrophages

3.3.1. The Pathway for Spherical Pectin Activating Macrophages

TLR4, in the form of the TLR4/myeloid differential protein-2 (TLR4/MD-2) complex, is an important kind of PRR located in macrophages [29]. Plant polysaccharides, including taro polysaccharide, Curcuma xanthorrhiza polysaccharide, Tinospora cordifolia polysaccharide, apple polysaccharide, safflower polysaccharide, Angelica sinensis polysaccharide, astragalus polysaccharide, Lycium barbarum polysaccharide, Dioscorea batatas polysaccharide, ginseng polysaccharide, and Platycodon grandiflorum polysaccharide, can activate immune cells by directly interacting with TLR4 [22,25,29]. Gal, glucose (Glc), and Ara are the most prevalent monosaccharide residues in TLR4-related immunostimulatory polysaccharides [29]. CTI-P-PS possesses neutral sugar side chains consisting of Ara and Gal residues, indicating its potential to activate RAW264.7 cells via interaction with TLR4. Anti-TLR4 was used to block the TLR4 of RAW264.7 cells in the present study. As seen in Figure 4A,B, although blocking TLR4 could not affect the proliferation-promoting activity of CTI-P-PS, it reduced the NO production promoted by CTI-P-PS. The results confirmed that CTI-P-PS activated RAW264.7 cells through recognition with TLR4. It was found that β-1,4-glycosidic linkages between Gal residues were associated with TLR4 signaling [29]. A previous study demonstrated that α-1,5-linked arabinose oligosaccharides activated macrophages via the TLR4 signaling pathway [30]. In the spherical pectin from C. morifolium Ramat. ‘Hangbaiju’ tea infusion, the relative abundance of Ara residues was significantly higher than that of Gal residues. Moreover, the abundant α-1,5-linked arabinose branches present in the spherical pectin from C. morifolium Ramat. ‘Hangbaiju’ tea infusion has been clearly characterized and structurally elucidated. Therefore, α-1,5-linked arabinose branches are considered to play a crucial role in the recognition process between spherical pectin and TLR4.

3.3.2. Molecular Docking Analysis of Arabinan Segments with TLR4/MD-2 and the TLR4/MD-2 Dimer

Polysaccharides abundant in galactan branches linked by β-1,4-glycosidic linkages have been reported to be multivalently recognized by macrophage surface receptors, which induces receptor accumulation and spatial clustering, triggers TLR4 dimerization in the plasma membrane, and subsequently promotes a cascade of downstream signaling reactions [23]. It was speculated that the spherical pectin from C. morifolium Ramat. ‘Hangbaiju’ tea infusion could activate RAW264.7 cells through the recognition of TLR4 by →5)-α-Araf-(1→ chains. The repeating unit →5)-α-Araf-(1→ was used as the representative active chain to investigate the interaction between TLR4 and spherical pectin. The →4)-β-Galp-(1→ segment was set as a positive control in the molecular docking analysis. The three-dimensional molecular structures of the repeating units →4)-β-Galp-(1→ and →5)-α-Araf-(1→ were constructed as linear homopolysaccharide segments with degrees of polymerization (DP) ranging from 4 to 10. Upon binding with agonistic ligands, the extracellular domains of receptors undergo dimerization, which further recruits specific adaptor proteins to the intracellular domains and initiates a downstream signaling cascade [31]. Therefore, receptor aggregation and dimerization are regarded as essential events in the process of immune activation. In the present study, the extracellular crystal structures of TLR4/MD-2 and the TLR4/MD-2 dimer (PDB codes: 2Z64 and 3VQ2, respectively) were both employed as receptors for molecular docking analysis. The docking affinities of galactan segments (Gal₄ to Gal₁₀) and arabinan segments (Ara₄ to Ara₁₀) with the extracellular crystal structures of TLR4/MD-2 and TLR4/MD-2 dimer were visualized using a heatmap (Figure 4C). As seen, the binding energies of both galactan and arabinan segments with the TLR4/MD-2 dimer were lower than those with TLR4/MD-2. This observation supports that →5)-α-Araf-(1→ forms a more stable docking complex with the TLR4/MD-2 dimer than with TLR4/MD-2. It is suggested that →5)-α-Araf-(1→ may play an important role in promoting the dimerization of TLR4/MD-2 and stabilizing the structure of the TLR4/MD-2 dimer. When the DP was greater than 6, the binding energies of the optimal docking poses between arabinan segments and the TLR4/MD-2 dimer were all below −9.0 kcal/mol (Figure 4C). A previous study demonstrated that α-1,5-linked arabinose oligosaccharides with a DP of 5 or 6 exert immunostimulatory effects via the TLR4 signaling pathway [30]. It is suggested that α-1,5-linked arabinose side chains with a DP higher than 6 possess the ability to interact with TLR4 and further activate macrophages. Taking the structure of Ara₉ as an example, the optimal docking pose between Ara₉ and the TLR4/MD-2 dimer at the lowest binding energy was further analyzed. As shown in Figure 4D, in the optimal binding pose, the arabinan segment (Ara₉) was located between two horseshoe-like leucine-rich repeats (LRRs) of TLR4/MD-2 dimer. It is speculated that the arabinan branches of spherical pectin can bind to two TLR4/MD-2 complexes via LRRs on the cell membrane, stabilize the TLR4/MD-2 dimer structure, and subsequently trigger the intracellular downstream signaling cascade, ultimately leading to the activation of RAW264.7 cells.

4. Conclusions

The present study preliminarily investigated the immunostimulatory activity and structure-activity relationship of the spherical pectin that forms the submicroparticle skeleton in C. morifolium Ramat. ‘Hangbaiju’ tea infusion. The immunostimulatory activity of spherical pectin from C. morifolium Ramat. ‘Hangbaiju’ tea infusion is more likely attributable to its RG-I domains with neutral sugar side chains. The arabinan side chains of the spherical pectin were highly branched. The backbone of these arabinan chains consisted of →5)-α-Araf-(1→ repeats, and additional →5)-α-Araf-(1→ branches were linked to the backbone via α-1,3-glycosidic linkages. The spherical pectin activated RAW264.7 cells through recognizing with TLR4. Molecular docking analysis demonstrated that →5)-α-Araf-(1→ branches with a DP higher than 6 in the spherical pectin were capable of binding to two TLR4/MD-2 complexes via LRRs on the cell membrane and stabilizing the TLR4/MD-2 dimer structure, which may represent an important mechanism underlying the immunostimulatory activity of spherical pectin. It was firmly believed that these results possessed significance and importance for other researchers to understand the primary structure of spherical pectin from chrysanthemum tea infusion and to explore the structure-immunostimulatory activity correlation of pectin.

One limitation of this study is that high-purity galactan was not successfully obtained from the spherical pectin, so its structure could not be elucidated. Furthermore, although the polygalacturonic acid hydrolyzed from spherical pectin was used to preliminarily exclude the contribution of the HG domain to immunostimulatory activity, changes in the degree of polymerization and degree of esterification may also account for the absence of activity. Therefore, it cannot be fully confirmed that the HG domain is unrelated to the immunostimulatory activity of the original spherical pectin. More experimental verification and deeper research are in progress by our research group.