Laminaria japonica was purchased from a traditional market in Jeju, South Korea. The PSs from LJ were extracted and purified following the methods described by Zhang et al. (26) with some modifications to enhance the isolation efficiency. Briefly, the fresh LJ was washed with tap water to remove cohesive salt and chopped into 2–3 cm-long pieces and oven-dried at 65°C to a constant weight. The prepared sample was placed into a big glass beaker and mixed with double-distilled water (d-H₂O) in the ratio of 1:30 (v:v, LJ:d-H₂O), boiled for 4 h, and subsequently filtered through the two layers of nylon mesh (0.2 mm mesh). The filtrate was continuously heated and kept slightly boiling to concentrate the filtrate to one-fourth volume from the initial volume. After cooling, the remaining filtrate was mixed with trichloroacetic acid (TCA) solution (5% concentrate) at a proportion of 1:2 (v:v, filter liquid:TCA) and maintained for 2 h statically to remove protein from the liquid. After the protein was precipitated completely, the liquid fraction was centrifuged at 3,000 × g for 10 min to collect the supernatant liquid, and the sediment was discarded; then, the supernatant was transferred to another container and a 4-fold volume of absolute ethyl alcohol (v:v) was added. The mixture was kept at 4°C for 12 h and then centrifuged at 3,000 × g for 10 min to separate the precipitated crude PS. The fractionated crude PS was subsequently re-dissolved in d-H₂O and dialyzed two times using an ultrafiltration membrane (molar mass > 3,500 D Mw, DP > 20, Beijing Solarbio Science and Technology Co., Ltd., Beijing, China) against d-H₂O (10 times the sample volume) at 4°C for 48 h, changing the d-H₂O every 12 h. The dialyzed liquid was collected and lyophilized to a constant weight using a vacuum dryer (Biosafer-10A, Biosafer, Nanjing, China), which was considered the crude PSs from LJ (LJPSs). Further purification was performed using anion-exchange column chromatography. Then, 3 g of crude PSs was dissolved in 100 ml of d-H₂O and then applied to a DEAE-52 anion-exchange column (5.0 cm × 100.0 cm). A stepwise gradient NaCl aqueous solution (0, 0.2, and 0.4 mol/L) was used for elution at a flow rate of 0.5 ml/min. Three fractions were collected separately and the concentration of polysaccharide in elution was detected with the phenol–sulfuric acid method (24, 27). The absorbance peak of 0.4 mol/L was much larger than the others, therefore, the fraction eluted with 0.4 mol/L NaCl was further purified using gel-permeation chromatography with a Sephacryl S-500, yielding one homogeneous fraction LJPSs (28, 29). The column was washed with d-H₂O at a rate of 0.2 ml/min, and eluted samples were freeze-dried. The sulfate radicals were analyzed with barium chloride gelatin analysis following the method described by Li et al. (27), and the total phenol contents were determined by the liquid chromatography electrospray ionization mass spectrometric (LC-ESI-MS) method.

The monosaccharide compositions of LJPS were identified by ion chromatography (IC). The detailed operation procedure was as follows: the absorption curve was established by measuring the absorbance of monosaccharide standards. In total, 13 standard substrates were used in the present study, such as fucose (Fuc), rhamnose (Rha), arabinose (Ara), galactose (Gal), glucose (Glc), xylose (Xyl), mannose (Man), fructose (Fru), ribose (Rib), galacturonic acid (GalA), glucuronic acid (GlcA), guluronic acid (GulA), and mannuronic acid (ManA). The standard sample of each monosaccharide was accurately weighed as 10.00 mg and dissolved in 10 ml of ultrapure water in 10 ml volumetric flasks to obtain 1, 5, 8, 16, 30, 40, 50, and 60 μg/ml standard solutions. Each standard solution was analyzed by IC (Thermo Fisher Scientific, ICS5000 type, Waltham, MA, United States) and a diode-array detector (Young-Lin Co., Changsha, China). The flow phases were NaOH (200 mM) or NaAC (200 mM) and the flow rate was 1.0 ml/min. The sample injection volume was 100 μl. The standard curve for each monosaccharide was fitted according to the retention time and absorption peak. A 10 mg (± 0.005) LJPS sample was accurately weighted into a glass tube, to which 1 ml of 2.5 mol/L trifluoroacetic acid was added for 90 min acidolysis at 110°C. The acidolysis mixture was dried in a vacuum rotator, and 5 ml of sterile water was added to fully dissolve the residue. The resulting solution was centrifuged at 12,000 × g for 10 min, and the supernatant was collected to determine the components by IC following the procedure mentioned previously herein. The retention time was matched with the monosaccharide standard curve to determine the monosaccharide composition of the LJPS sample. The molar amount of each component was calculated from the peak area of each derivative.

In this study, the molecular weight of LJPS was evaluated using the following parameters (26, 30): the number-average molecular weight (Mn), expressed as Mn = Σ(niMi)/Σni; the weight-average molecular weight (Mw), defined as Mw = Σ(niMi²)/Σ(niMi); and the Z-average molecular weight (Mz), calculated as Mz = Σ(niMi³)/(ΣniMi²) (26, 30). The polydispersity coefficient presents the range of the molecular mass distribution and can be calculated by dividing Mw by Mn or dividing M_Z by Mn. The purified LJPS samples were accurately weighed as 5.00 mg and dissolved in 1 ml of 90% dimethyl sulfoxide (DMSO) in a glass tube for an overnight incubation in a 100°C water bath, to which 3 ml of absolute ethanol (analyzed as pure, Sinopharm, Beijing, China) was added and then mixed vigorously. The mixture was centrifuged at 1,000 × g for 5 min, and the supernatant was removed. Then, the residual sediment was rinsed two times with anhydrous ethanol, and 3 ml of 0.1 M NaNO₃ (containing 0.02% NaN₃, Sigma Aldrich, Shanghai, China) was added for a 20 min incubation at 121°C. The mixture was centrifuged at 12,000 × g for 10 min, and the supernatant was collected and assessed by the gel permeation chromatography-refractive index-multiangle laser light scattering (GPC-RI-MALS; Dawn Heleos II, Wyatt Technology, Santa Barbara, CA, United States) method. The eluent was a 0.1 M NaNO₃ and 0.02% NaN₃ mixture (1:1, v/v; HPLC, Sigma-Aldrich, Shanghai, China). The flow rate was maintained at 0.4 ml/min, and the temperature of the column was maintained at 60°C. The eluent was monitored with a refractive index (RI) detector (Agilent 1260, Guangzhou, China); the analytic column included an Ohpak SB-805 HQ, Ohpak SB-804 HQ, and Ohpak SB-803 HQ (Shodex, Asahipak, Tokyo, Japan). The volume of the manually injected sample was 300 μl. Mass spectra were viewed and analyzed with ASTRA6.1 software (Wyatt Technology Corporation, Santa Barbara, CA, United States).

The methylation analysis of PS was performed according to a previous method (31, 32) with minor modifications. In brief, 3 mg LJPS powder was accurately weighed and dissolved in 1 ml d-H₂O, and added into 200 μl 2-morpholine ethane sulfonic acid (0.2 M) and 200 μl carbodiimide (500 mg/ml) solution, reacting for 2 h. The reaction mixture was subsequently added with 1 ml imidazole (4 mol/L) and divided into two equal parts, then, one of the part was mixed and reacted with 1 ml NaBH₄ (30 mg/ml) and the other was mixed and reacted with 1 ml NaBD₄ (30 mg/ml) for another 3 h until the termination of reaction by adding 200 μl acetic acid. The two reacted solutions were then dialyzed for 48 h followed by freeze-drying. The dried samples were individually dissolved in 500 μl anhydrous DMSO and reacted with 50 μl DMSO/NaOH (120 mg/ml) solution for 30 min. Then, 10 ml of CH₃I solution was added to the reaction and incubated another 10 min followed by combining with 1 ml H₂O and 500 μl of dichloromethane (DCM) solution to obtain DCM phase. The dried samples were then dissolved in 100 μl trifluoroacetic acid (2 M), reacted at 121°C for 90 min followed by evaporation to dryness at 30°C. The sample was dissolved in 50 μl of ammonia H₂O (2 M), mixed and reacted with 50 μl of 1 M NaBD₄ for 2.5 h. Further 20 μl of acetic acid was added to the mixture to end the reaction, followed with washing it two times with 250 μl of methanol, and dried under nitrogen. The obtained powder was mixed and reacted with 250 μl of acetic anhydride at 100°C for 2.5 h followed by mixed and incubated with 1 ml of H₂O for another 10 min. Finally, the sample was mixed with 500 μl of DCM, and centrifuged to discard aqueous phase, and the bottom DCM phase were tested on the machine. A 6890A-5975C gas chromatography mass spectrometry (GC-MS) system (Agilent Technologies Inc., CA, United States) equipped with a BPX70 capillary column (30 m × 0.25 mm, 0.25 μm) was used to analyze the glycosidic linkages. The gas chromatographic conditions were as follows: BPX70 chromatographic column; high purity helium gas was used as the carrier gas at a flow rate of 1.0 ml/min; the injection volume was 1 μl, and the split ratio was 10:1; the initial column temperature was 140°C, retaining this for 2 min, and procedurally increasing it to 230°C at 3°C/min, with subsequent holding for 5 min; the injection temperature was 230°C; the ion source of the mass spectrometer was set at 230°C, and the four stage bar temperature was 150°C. The resulting peaks of alditol acetates were identified based on their MS fragmentation patterns and the relative retention time in the GC spectrum. The molar ratios of individual linkage residues were estimated as the ratios of peak areas.

The surface topography and microstructure of long-chain LJPS can be imaged by scanning electron microscopy (SEM). When the PS was dissolved in water or other solvents, its macromolecules with its three-dimensional structure would stretch into the unfolded chain or net status owing to the liquid tension, which can be inspected by transmission electron microscopy (TEM). The brief procedure was as follows: dried LJPS powder was scattered on a metal stub, sputtered with gold, and then examined by SEM with the model JEOL (JSM-IT100, Tokyo, Japan). The images were taken at different magnifications (× 200 and × 2,500). For TEM inspection, a well-dispersed PS solution was first prepared using sodium dodecyl sulfate (SDS, 5 μg/ml) to a final concentration of 5 μg/ml. The SDS aqueous solution was applied to dissolve LJPS as the SDS solution could reduce molecular aggregation. A droplet of LJPS solution (5 μl) was deposited on a carbon film specimen (200 mesh, Beijing Zhongjingkeyi Technology, Beijing, China) and dried at room temperature (25 ± 2°C). The specimen was examined using TEM (Tecnai G2 Spirit BIOTWIN, FEI, Hillsboro, OR, United States) at an accelerating voltage of 100 kV to observe the microstructures of LJPS.

The infrared spectrum of LJPS was measured on a Fourier transforms infrared (FT-IR) spectrometer (FT-IR 650, Tianjin Gangdong Sci. &Tech. Co., Ltd., China) in the wavenumber range of 400-4,000 cm^–1 by pressing LJPS samples (2 mg) and KBr (200 mg) into a pellet. One- and two-dimensional nuclear magnetic resonance (NMR) analyses of LJPS were carried out on an Avance Bruker III HD 600 MHz NMR spectrometer equipped with a 5 mm TCl CryoProbe at 25°C using D₂O as the solvent at a final sample concentration of 40 mg/ml. The molecular structure of LJPS was determined via the analysis of one- and two-dimensional NMR spectra, such as the ¹H-NMR, ¹³C-NMR, heteronuclear single quantum coherence (C, H-HSQC), single-bond proton–proton correlation spectroscopy (H, H-COSY), heteronuclear multiple bond coherence (H, C-HMBC), and nuclear Overhauser effect spectroscopy (H, H-NOESY).

Data were analyzed by one-way analysis of variance (ANOVA). Tukey’s family error rate was used for a one-way multiple comparison (p < 0.05) with GraphPad Software, Prism 8.0.1 (San Diego, CA, United States). Values were expressed as the mean ± SEM. The 3d structure and spatial configurations of the LJPS molecule were speculated using the online analysis GLYCAM¹ by integrating the monomer compositions, glycosidic linkages, and NMR information.

The isolated LJPS had good aggregation and precipitation properties, and the freeze-dried powder had a uniform texture and flat surface (Figure 1A). After breaking down the outside surface, the porous or honeycombed microstructures were observed by SEM inspection (Figures 1B,C). The PS isolated from LJ consisted of 87.71% total sugar, 1.54% protein, 0.016% total phenols, and 10.73% sulfate (Table 1). The analysis of IC showed that LJPS was composed of Fuc, Rha, Ara, Gal, Glc, Xyl, Man, Fru, Rib, GalA, GluA, GlcA, and ManA, with a molar ratio of 35.71:1.48:0.28:13.16:0.55:2.97:6.92:0.58:0.41:0.14:3.16:15.84:18.79 (Figure 2 and Table 1). Thereof, the accumulated molar proportions of Fuc, Gal, Man, GlcA, and ManA reached 93.59%. In contrast, the molar percentages of Ara, Glc, Rib, and GalA (denoted with dotted lines in Figure 2) were less than 3%, and those compositions might vary considerably in different determinations.

The primary maxima of absorption bands in the IR spectra of LJPS are shown in Figure 3. FT-IR spectroscopy is an important analytical technique extensively used to study the molecular structures and conformations of macromolecules to identify the vibrations between different atoms in molecules. The transmittance spectrum with wavenumbers between 400 and 4,000 cm^–1 was considered as a reflection of the structural characteristics of the LJPS, such as the glucosidic bonds and substituted groups (34, 35). As shown in Figure 3, the absorption band at 3,600–3,200 cm^–1 was attributed to the stretching vibration of the –OH group, and the peaks in this region were indicative of the characteristic absorption for PS. In detail, the absorption peak presented at 3,432.3 cm^–1 was identified as the O–H stretching vibration (O–H group), which was the typical peak of PS. The band at 2,940 cm^–1 was C–H stretching vibration. The band at 1,646.9 cm^–1 was ascribed to the C=O stretching vibration, which are indicative of carboxyl and carbonyl groups and characteristic of uronic acid (35, 36). The strong sharp absorption band at 1,251.2 cm^–1 (S=O) stretching confirmed the existence of a significant amount of sulfate in the PSs, and the band at 1,050.0 cm^–1 was attributed to the presence of the asymmetric O=S=O stretching vibration of a sulfate group. The sharp band at 887.76 cm^–1 (C–S–O) suggested a pattern of sulfate substitution.

The molecular weights of LJPS are presented in Table 2. The Mw, Mn, and Mz of LJPS were calculated to be (1.17 ± 0.018) × 10⁴ (5.79 ± 0.006) × 10⁴, and (3.76 ± 0.012) × 10⁵ g/mol, respectively. The polydispersity coefficient of LJPS was 4.96 (Mw/Mn) or 32.11 (Mz/Mn), suggesting a relatively narrow range of molar mass distribution. In addition, the radius mean square (RMS) of LJPS was simultaneously analyzed as 26.1, 24.7, and 24.6 for the number-average radius (Rn), weight-average radius (Rw), and Z-average radius (Rz), respectively. The molecular weight characteristics of LJPS were further elucidated by the variation trend curve of molar mass (Figure 4), which displayed a rapidly decreasing laser scattering (LS) curve (red line) and gradually increasing RI line (green line), indicating that the LJPS contained relatively few proportional macromolecule PSs but had a high proportion of small molecules (the molar mass had polydispersity). The molar mass line (blue line) presented a rapid initial decrease and maintained a mild decline with an extension of the retention time, revealing a few macromolecular structures and more molecules with an intermediate molar mass in this polymer. The intersection of LS and RI curves corresponded to the value of the molar mass line of ∼10⁴, which revealed the distribution range of the molar mass of most PS molecules.

When the RMS value is more than 10 nm, taking the log(molar mass) as the horizontal coordinate and log(RMS) as the vertical coordinate to fit the conformation plot, the slope values suggested a biopolymer molecular configuration (37). As shown in Figure 4, the plot of RMS to the molar mass presented as a U-shaped structure, and the slope was 0.08 (Figure 4B), indicating that the spatial conformation of PS molecules comprised a dense crosslinking macromolecule with a highly branched structure. Correspondingly, TEM inspection verified that the unfolded LJPS molecules had a crosslinking network and/or long-chain status in SDS solution (Figure 5C). The subsequent spatial 3-D conformation speculation based on the structural parameters in this study confirmed the dense interconnected molecules with a highly branched structure in LJPS (Figure 5B).

The GC-MS analysis revealed that LJPS was predominantly comprised of 21 monomer residues linked by different glycosidic bonds, such as non-reducing terminals of t-Rhap, t-Fucp, t-Ribp, t-Xylp, t-ManpA, t-GlcpA, and t-Galp with the molar ratios of 0.47, 3.74, 0.43, 4.25, 3.44, 1.29, and 0.51%, respectively (Table 3). Other monomers compositions, such as 1,3-Fucp; 1,3-Fucp; 1,2-Xylp; 1, 3-GlcpA; 1, 2-Manp; 1, 4-ManpA; 1, 4-Galp; 1, 4-GlcpA; 1, 2, 3-ManpA; 1,3,4-GlcpA; 1, 4-GulpA; 1,2,3,4-Galp; 1,2,6-GlcpA; and 1, 2, 3, 6-ManpA with the molar ratios of 1.96, 1.86, 0.99, 4.46, 10.50, 19.07, 13.81, 18.36, 1.93, 5.89, 3.99, 0.54, 1.55, and 0.94%, respectively, were identified as the intrachain residues. The main branching points were at 1,2,3-ManpA; 1,3,4-GlcpA; 1,2,3,4-Galp; 1,2,6-GlcpA; and 1,2,3,6-ManpA comprising 1.93, 5.89, 0.54, 1.55, and 0.94%, respectively. Of those glycosidic residues, irrespective of the glycosidic bond types, seven monomer residues had relative high proportions (> 3%), namely, Fuc (7.567%), Xyl (5.25%), ManA (25.38%), GlcA (31.56%), Gal (14.85%), Man (10.51%), and GulA (3.99%). Those might be the main constituents of LJPS molecules. In addition, the degree of branching (DB) value was 24.99% for LJPS, which was the proportion of the accumulated numbers of terminal and branch residues accounting for the total amount of saccharide residues. This revealed that the LJPS was a highly branched molecule.

The linkages between glycosyl residues in LJPS were further investigated by NMR analyses, such as ¹H NMR, ¹³C NMR, and HSQC spectra (Figure 6), with the assistance COSY, HMBC, and NOESY (not shown). The ¹H NMR spectrum signals were mainly distributed in the 1.5–6.3 ppm range, the chemical shifts between δ 3.2 and 4.0 ppm were the protons of the glycosidic ring, and δ 4.09, 4.45, 4.65, 4.95, 5.42, 5.64, and 5.78 were assigned to terminal protons, whereas other signals peaks (chemical shift) were distributed in δ 2.6–3.0 and 4.3–5.5 ppm. Analysis of the ¹³C NMR spectrum revealed that the carbon signals were primarily distributed at chemical shift δ –120 ppm. The anomeric carbons were at δ 91.48, 93.49, 94.66, 97.31, 99.52, 101.1, 101.83, and 103.21 ppm, and the signals of anomeric carbons were mainly found at δ 91–105 ppm. However, the signal peaks of the chemical shift were at δ 84.61, 81.96, 81.02, 80.59, 79.55, 79.01, 78.71, 77.54, 76.70, 73.31, 71.20, 67.69, 65.80, and 62.02, which scattered in the δ 60–105 ppm range. In addition, regarding the correlations of C/H in the heteronuclear singular quantum correlation (HSQC) spectrum (Figure 6C), 18 cross signals, with the ¹H being from 4.6 to 6.5 ppm and ¹³C in the range of 90–115 ppm, further demonstrated the presence of A-R residues, which are represented in Figure 6 and summarized in Table 4. Combining the information of HMBC and NOESY analysis, the structure of this PS could be delineated as follows: the backbone of LJPS was composed of →2)-α-D-Manp-(1 → 3)-α-D-Fucp-(1 → 3)-α-D-ManpA-(1 → 2)-α-D-Fucp-(1 → 2)-α-D-Manp-(1 → 2)-α-D-Xylp-(1 → 6)-α-D-Manp-(1 → 3)-α-D-Galp-(1 → 6)-α-D-Manp-(1 → 2)-α-D-Xylp-(1 → 4)-α-D-Glcp-(1 → 6)-α-D-Glcp-(1 → 4)-α-D-Galp-(1 → 4)-α-D-Glcp-(1 → 4)-α-D-Galp-(1 →; and the branch chain included: →4)-α-D-Gulp-(1 → was linked to the backbone chain by O-2 of 2, 3)-α-D-ManpA; →3)-α-D-GlcpA-(1 → was connected with main chain by O-2 of 2, 6)-α-D-GlcpA; →4)-α-D-GlcpA-(1 → 4)-α-D-GlcpA-(1 → 4)-α-D-ManpA-(1 → 4)-α-D-ManpA-(1 → linked to the main chain by O-3 of 3,4) –α-D-GlcpA; →4)-α-D-ManpA-(1 → was linked to the backbone by O-2 of 2, 3, 6)-α-D-ManpA; →4)-α-D-GlcpA-(1 → was connected with the main chain by O-3 of 2, 3, 6)-α-D-ManpA; →2)-α-D-Manp-(1 → was linked with backbone by O-2 of 2,3,4)-α-D-Galp; and →2)-α-D-Manp-(1 → was linked to the main chain by O-4 of 2,3,4)-α-D-Galp. Therefore, based on the results of glycosidic linkage and NMR spectra analyses, the possible monomers linkage of LJPS is shown Figure 5A.

Combining all structural parameters involving the monosaccharide composition, glycosidic type, sequences, and linking position of branched chains to fit the potential molecular structure and morphology, the structural formula indicated that the LJPS molecule was a long-chain PS with 176 monomers and highly branched structural characteristics (Figure 5A and Supplementary Figure 1). The spatial conformation results showed that there were 32 potential spatial conformations with different folding angles and torsions (Supplementary Figure 2). Additionally, the 3-D spatial conformation video showed that the long-chain LJPS isolated in this study was a denser interconnected macromolecule with highly branched and helix domains (Figure 5B and Supplementary Video 1). The profile from the side (Figure 5B1) and front (Figure 5B2) were well matched with the scanning images of TEM (Figure 5C), showing the overall shape and spatial conformation of the long-chain LJPS. The structural characteristics and chain spatial conformation confer the bioactivities of PSs (38). However, owing to the complex compositions, highly branched structure, large molecular weight, viscosity, and aggregating properties in solution, it was difficult to characterize the spatial conformation of PS molecules. In this study, we preliminarily explored the spatial structure of long-chain LJPS; however, the relationship between this unique molecular conformation and its biological activity needs to be further studied.

Various concentrations of LJPS were administered to the mouse macrophage RAW 264.7 cells for 24 h to determine the potential cytotoxicity of LJPS. As shown in Figure 7A, no significant cytotoxicity was observed after treatment for 24 h, although the highest concentration of LJPS, 50 μg/ml, seemed to result in a slight reduction in cell viability (p = 0.069). Therefore, further experiments with RAW 264.7 cells were performed with < 10 μg/ml LJPS. Next, we investigated how LJPS regulates the activation of macrophages with and without LPS endotoxin stimulation by measuring NO production, as well as inflammatory cytokines, such as IL-6 and TNF-α. As shown in Figure 7B, LJPS at both 10 and 25 μg/ml significantly induced NO production without the LPS stimulus, whereas LPS-induced NO production was significantly decreased by LJPS treatment at both 1 and 10 μg/ml concentrations in mouse macrophage cells (Figure 7B). However, 25 μg/ml LJPS resulted in a significant increase in NO production regardless of LPS challenge. This indicated the biphasic regulatory property of LJPS in anti-inflammation and immunomodulation only within a certain range of concentrations (≤ 10 μg/ml). LJPS treatment did not significantly change the mRNA expression of either Il-6 or Tnf-α regardless of LPS stimulation in RAW 264.7 cells (Figure 7C). Unlike the biphasic regulatory effect of LJPS on NO production, LJPS itself did not induce the production of the proinflammatory cytokines TNF-α and IL-6 in RAW 264.7 cells. However, LJPS significantly attenuated LPS-induced TNF-α and IL-6 production, compared with that in the LPS only treated group (Figure 7D), suggesting an anti-inflammatory capacity and its translational regulation.