Novel polysaccharide extracted from Sipunculus nudus inhibits HepG2 tumour growth in vivo by enhancing immune function and inducing tumour cell apoptosis

Abstract A novel polysaccharide was extracted from Sipunculus nudus (SNP). The molecular weight (MW) of SNP was determined to be 9223 Da by high‐performance gel permeation chromatography analyses, and the structure of the SNP repeat units was determined to be →3,4‐β‐D‐GlcpNAC (1→ and →4) ‐α‐D‐Glcp (1→ in the ratio of 15:1; →2) ‐α ‐D‐Galp ‐ (1→ as a side chain; and β‐D‐Galp‐(1→ and α‐ D‐Glcp ‐ (1→ as end groups by GC‐MS analysis and NMR assays. The effect of SNP on hepatoma HepG2‐bearing mice was analysed to verify its potential in the clinical treatment of liver cancer. A total of 90 male athymic nu/nu mice were divided into therapeutic and preventive groups and fed with different amounts of SNP. The antitumour effect of SNP on HepG2‐bearing mice and mechanism of such were studied by analysing the tumour size, spleen index, thymus index, immune factors in the blood, tumour apoptosis factors, etc. The results suggest that SNP not only increased the index of immune organs in the body, but also enhanced the secretion of immune factors, including interleukin‐2, interferon gamma and tumour necrosis factor‐alpha in the serum. SNP induced the apoptosis of tumour cells via the mitochondrial apoptosis pathway, which upregulated caspase‐3, caspase‐8, caspase‐9 and BCL2‐associated X, but downregulated B‐cell lymphoma‐2 and vascular endothelial growth factor protein expression. In conclusion, SNP inhibited tumour growth by enhancing immune function and inducing tumour cell apoptosis in HepG2‐bearing mice. Therefore, SNP may be further investigated as a promising candidate for future antitumour drugs.

traditionally used as a remedy for cancer treatment in several Asian countries.
Many polysaccharides extracted from natural resources have been proved to have effective antitumour, immunomodulation, antioxidation and anti-inflammatory activities. 3 For instance, polysaccharides from Chinese wolfberry pollen, Tetrastigma hemsleyanum, and silkworm possess excellent antioxidant and antitumour activities. [3][4][5] Sipunculus nudus is widespread in the subtidal zones and sea beds of temperate or tropical seas. Sipunculus nudus has also long been used as traditional Chinese medicine in folk remedies for the treatment of carbuncles, chronic diarrhoea, nocturia, regulating the functions of the stomach and spleen and for restoring health in debilities caused by various pathogens and ageing. 6 In vivo and in vitro studies showed that SNP administration may enhance immune function. 7 In our previous study, SNP was found to exert pro-apoptotic activity and anti-HBV activity on HepG2.2.15 cells. 8 However, the molecular weight and structure of SNP was unclear, and whether SNP has an antitumour effect on animal tumour tissue remained unknown. Therefore, the molecular weight and the structure of SNP was analysed by high-performance gel permeation chromatography (HPGPC), GC-MS and NMR assay, and the hepatoma HepG2-bearing mice model was established to investigate the various biological activities of SNP in vivo. The results demonstrated that SNP, a novel polysaccharide extracted from Sipunculus nudus, stimulates the immune function and induces cellular apoptosis to mediate its antitumour activity.

| Isolation and purification of the polysaccharides
To extract polysaccharides from Sipunculus nudus, we followed a previously described method. 8 Sipunculus nudus (diameter 6 ± 2 mm, length 10 ± 2 cm) was collected from the Xiamen market, China.
Fresh worms were washed, oven-dried and crushed into powder.
The powder was hydrolysed by trypsin at 50℃ for 5 h. The insoluble material was removed by centrifugation. The supernatant liquor was deproteinized four times using the Sevage method. Then, four volumes of cold ethanol were added to precipitate the material after standing at 4℃ overnight. Then, resulting precipitate was centrifuged (3000 × g, 10 min). After washing with ethanol three times, the precipitate was freeze-dried in vacuo and ground into a powder to produce a crude product. The crude product was subjected to a diethylaminoethyl (DEAE)-Sepharose anion exchange column

| Molecular weight determination
The molecular weight and molecular weight distribution and purity of the polysaccharopeptide were analysed by HPGPC. We used 100 mg of polysaccharide, added 50 ml water for dissolution, centrifuged at 5000 rpm, placed 40 ml of the supernatant into a beaker and then added 60 ml ethanol to supernatant, shook well, placed it at 4℃ refrigerator for 1 h, 0.45µm membrane filtered, and the precipitate was dried at 80℃ oven for 4 h. An appropriate amount of the dry precipitate was used, and the solution containing about 10 mg per 1 ml was dissolved using the mobile phase. Then, 10 µL was injected into an HPLC to record the chromatogram. We used GPC software to process the data. The samples and the standard (5000, 11,600, 23,800, 48,600, 80,900 148,000, 273,000, 409,800 and 667,800, all from Sigma) were precisely weighed and prepared into a 10 mg/ml solution, centrifuged at 12,000 rpm for 10 min, the supernatant was filtered through a 0.45 µm microporous membrane, and the samples were then transferred to a 1.8-ml injection flask.
The eluent was 0.02 M NaAc, with a flow rate of 0.5 ml/min, column temperature of 50℃ and injection volume of 10 μl. The detector was a G1362A 1260 RID.

| Structural Analysis of Polysaccharide Bond
We took 10 mg of polysaccharide, added 1 ml water for dissolution, added 1 ml of 100 mg/ml carbodiimide and reacted for 2 h. Then, we added 1 ml of 2 M imidazole; divided the sample into two parts; added either 1 ml of 30 mg/ml NaBH 4 or 1 ml of 30 mg/ml NaBD 4 to the parts, separately; and reacted for 3 h. We added 100 μl acetic acid to terminate the reaction. After dialysis, the samples were freeze-dried and methylated. We added 500 μl DMSO to the lyophilized sample for dissolution, and then, 1 mg NaOH was added and incubated for 30 min. We added 50 μl iodomethane solution to react for 1 h. We added 1 ml water and 2 ml dichloromethane, then vortexed, centrifuged at 5000 rpm and discarded the aqueous phase.
We repeated the washing three times. The lower dichloromethane phase was absorbed and evaporated to dryness. We added 100 μl 2 M trifluoroacetic acid (TFA) and reacted at 121℃ for 90 min.
Solution was steam-dried at 30℃. We then added 50 μl 2 m ammonia and 50 μl 1 M NaBD4 to solution, mixed and reacted for 2.5 h at room temperature. We added 20 μl acetic acid to terminate the reaction, dried with nitrogen, washed with 250 μl methanol twice and dried with nitrogen again. We added 250 μl acetic anhydride, and the reaction was carried out at 100℃ for 2.5 h. We added 1 ml water and let solution stand for 10 min. We added 500 μl dichloromethane to solution, vortexed, centrifuged at 5000 rpm and discarded the water phase. We repeated washing three times, then took the lower dichloromethane phase and tested it on a GC-MS system (Agilent Technologies Inc.).
The chromatographic parameters were as follows: an Agilent 7890A (Agilent Technologies) was used as the chromatographic system. According to the properties of the compounds, the injection volume was 1 μl, the split ratio was 10:1, and the carrier gas was high-purity helium. The initial temperature of the column incubator was 140℃ for 2.0 min, and the temperature was raised to 230℃ at a rate of 3℃/min for 3 min.
The mass spectrum parameters were as follows: a quadrupole mass spectrometry system (Agilent 5977B; Agilent Technologies) was used in the mass spectrometry system, which was equipped with an electron bombardment ion source (EI) and Mass Hunter workstation. The analytes were detected by electron bombardment ion source (EI) in scan mode with a mass scanning range (m/z) of 30-600.

| NMR Analysis
The method of NMR analysis in Shi et al. was followed. 9

| Cytotoxicity evaluation
Cell counting kit (CCK)-8 was used for cytotoxicity evaluation of SNP. Human hepatocyte cells (LO 2 ) were first counted, and approximately 10 4 cells per well were seeded in a 96-well cell culture plate (Corning Inc.). Then, after incubation at 37℃ in a humidified atmosphere with 5% CO 2 for 12 h, the culture medium was

| Experimental design
A total of 90 athymic nu/nu mice were used in this study, which were divided into three groups: treatment, prevention and positive groups. For the treatment experiment (marked as a), 40 mice were inoculated with tumour cells. When the tumours reached millet size, mice were screened as treatment animal models (marked as a).
These animals were randomly divided into four groups (n = 6 mice/ group): vehicle control group (saline, 0.2 ml/mouse, referred to as a Ctrl) and three treatment groups at various dosages (SNP at 50, 100 and 200 mg/kg, referred to as a SNP-50, a SNP-100 and a SNP-200, respectively). For the preventive experiment (marked as b), another 40 mice were randomly divided into four groups: a vehicle control group (saline, 0.2 ml/mouse, referred to as b Ctrl) and three precondition groups at various dosages (SNP at 50, 100 and 200 mg/kg, referred to as b SNP-50, b SNP-100 and b SNP-200, respectively). After preconditioning for 1 month, these animals also received tumour xenotransplantation. When the tumours reached millet size, mice were screened as preventive animal models (n = 6/group). After the two animal models were established, the drugs were administered intragastrically once daily for 16 consecutive days and samples were collected. Ten mice inoculated with tumour cells were fed with APS after tumour reach millet size as positive control group (Astragalus polysaccharides (APS), 100 mg/kg, referred to as APS). The experiment was carried out in batches.

| Sample collection
At the end of the experimentation, the mice were euthanized by the inhalation of CO 2 at a rate of 10-30% of the volume of the euthanasia chamber per min. The blood samples were collected from orbit for the determination of cytokines. The tumours, thymuses and spleens of all animals were removed and weighed. The tumour volumes were measured and calculated using the following equation: volume = w × ow 2 / 2, where w is the maximal width (mm) and ow is the maximal orthogonal width (mm). The tumour inhibition rate and organ index were calculated using the following equations: tumour inhibition rate (by weight) = (mean tumour weight of control group -mean tumour weight of treatment group)/mean tumour weight of control group × 100%; tumour inhibition rate (by volume) = (mean tumour volume of control group -mean tumour volume of treatment group) / mean volume of control group × 100%; organ index = organ weight (mg) / body weight (g) × 100%.

| Histopathological analysis
Tumour tissues of all mice were fixed in 4% paraformaldehyde solution for 24 h and then embedded into paraffin blocks. Contiguous 100% ethanol (twice for 2 min) for dehydration and then immersed in xylene twice for 2 min to make the section transparent. Then, the slides were mounted on cover glass using cedarwood oil and the samples were imaged under a light microscope for morphological observation.

| Immunofluorescence staining
The tissue sections were hydrated, heated in a microwave for 10-

| ELISA
The blood samples of mice in each group were centrifuged at 1000 ×g for 10 min, and the serum was collected. The cytokines, including IL-2, IFNγ and TNFα, were detected using the relevant commer-

| Reverse transcription-quantitative PCR (RT-qPCR)
Total RNA from tumour samples was extracted using TRIzol ® reagent (Invitrogen; Thermo Fisher Scientific, Inc.), and gene expression was analysed via RT-PCR. The primers were synthesized by Shanghai Shenggong Bioengineering Co., Ltd. as shown in Table S1.
cDNA was synthesized using a ReverTra Ace ® qPCR RT kit (Toyobo Life Science), followed by qPCR using SYBR Green Real-Time PCR Master Mix-Plus (Toyobo Life Science). The expression of β-actin was used as the internal control.

| Western blotting
Tumour tissue (≥50 mg) in each group was harvested, minced and Co., Ltd.). The bands were semiquantified using Image-Pro Plus 6.0 software Media Cybernetics, Inc.), which was used to analyse the protein levels.

| Statistical analysis
Data from ≥3 independent experiments are presented as the mean ± standard deviation (SD). Statistical significance of difference was performed with Student's t test and two-factor without replication (analysis of variance: ANOVA). The data were considered to be significant at p < 0.05.

| Molecular weight determination
The molecular weight of SNP was determined by HPGPC. The logarithm of polysaccharide standards and retention time (RT) are plotted in Figure 1A-C. The molecular weight (MW) of SNP is 922 3D (Table 1) according to the RT of SNP ( Figure 1D) and standard curve calculating.

| Structural analysis of polysaccharide bond
The results of GC-MS analysis after SNP methylation are shown in Figure S1, and the statistics of quantitative analysis are shown in Table S2. The results showed that there were 16 different molecular bonds, but the main structure was composed of galactose and glucose.

| NMR analysis
To obtain the structural information of SNP, one-dimensional NMR 13 C-NMR and distortionless enhancement by polarization transfer (DEPT)-135)) and two-dimensional NMR (correlation spectroscopy (COSY), heteronuclear single-quantum coherence (HSQC), heteronuclear multiple-bond correlation (HMBC) and total correlation spectroscopy (TOCSY)) were used to determine the chemical shifts of the H and C of the main sugar residues and to infer the connection sequence of each sugar residue.
Most of the hydrogen spectra signals of SNP were in the range of   (Table S3). Combined with the methylation results and literature reports, 9,11,12 we inferred that sugar residue A was β-D-Galp -(1→).  (Table S3). The chemical shifts of C-1 and C-2 shifted to the low field, indicating that the residues were substituted at C-1 and C-2 of the sugar ring. Combined with the methylation results and reports, 13 we inferred that residue B was →2)α -D-Galp -(1→).    (Table S3). The chemical shifts of C-4 and C-3 shifted to the low field, indicating that the residues were substituted at the C-4 and C-3 positions of the sugar ring. Combined with findings of previous reports, 16,17 we inferred that the residue E was →3,4)β -D-

Glcpnac (1→).
Based on the coupling signals of heterocephalic hydrogen with carbon or heterocephalic carbon with hydrogen in the HMBC remote correlation spectrum, we further inferred the interconnection sequence of each sugar residue. The HMBC correlation spectrum of SNP is shown in Figure S6, from which the following can be found: Based on the analysis of the methylation results (Table S3)

| SNP suppresses tumour growth in HepG2 tumour-bearing mice
The tumour weight and volume were measured, and the corresponding inhibition rates were calculated simultaneously. In the therapeutic experiment, the tumour weights of the a SNP-100 and a SNP-200 groups were significantly lower compared with those of the a Ctrl group (both p < 0.01; Figure 3A, left). The tumour volumes of all SNP-treated groups were significantly smaller compared with those of the a Ctrl group (all p < 0.01; Figure 3B, left). These results suggested that SNP has a certain therapeutic effect on the tumours of HepG2 tumourbearing mice. In the preventive experiment, the tumour weights of all SNP-related groups were significantly lower compared with the b Ctrl group (all p < 0.01; Figure 3A

| SNP enhances the secretion of cytokines
The antitumour function of SNP may be associated with the im- In the therapeutic experiment, the IL-2 and IFNγ levels in the a SNP-100 and a SNP-200 groups were significantly upregulated compared with those of the a Ctrl group (both p < 0.01), but this was not found in the a SNP-50 group ( Figure 6A,B, left). However, the production of TNFα in all SNP-related groups was upregulated compared with the a Ctrl group (all p < 0.01; Figure 4C, left).
In the preventive experiment, the secretions of IL-2, IFNγ and TNFα in all SNP-related groups were higher compared with those of the b Ctrl group (p < 0.01 and p < 0.05 for IFNγ and TNFα levels of b SNP-50 group, respectively; Figure 6A  Using a high-throughput sequencing assay, we identified that the gene expression levels of Cyr61, Hsp90 and vascular endothelial growth factor (VEGF) were decreased in all SNP-related groups (data not shown). Th mRNA and protein expression levels of cysteine-rich angiogenic inducer 61 (Cyr61), heat-shock protein 90 (Hsp90) and VEGF were detected ( Figure 7D-F) and found to be downregulated in all SNP-related groups in a dose-dependent manner. The results suggested that the a SNP-50 and b SNP-100 groups were most similar to the APS-positive group. According to these results, we concluded that the mechanism of SNP antitumour function may be exerted, at least partially, by blocking the actions of CYR61, HSP90 and VEGF.

| SNP regulates the expression levels of apoptosis-related proteins (caspase-3, caspase-8, caspase-9, Bax, Bcl-2 and VEGF)
It was reported that low VEGF levels downregulated the antiapoptotic factor B-cell lymphoma-2 (Bcl-2), 21  pro-apoptotic factor BCL2-associated X (Bax). 22 The expression levels of apoptosis-related proteins (caspase-3, caspase-8, caspase-9, Bax, Bcl-2 and VEGF) were detected via immunofluorescence analysis to investigate the antitumour mechanism of SNP. The relative expression levels of these proteins were analysed by comparing the fluorescence intensities in the tumour tissues of different groups In our previous study, the flow cytometry assay was used to validate the pro-apoptosis activity of SNP on HepG2.2.15 cells. 8 Another study also reported the different potential immunologic effects of SNP. 6 We identified several possible mechanisms to explain the antitumour activity of SNP on hepatoma HepG2-bearing mice, which involves immunologic function. shows that polysaccharide could be absorbed by membrane transport. 23 Polysaccharide extracted from Astragalus membranaceus (APS) was found to exert a marked inhibitory effect on human solid tumours by potentiating the immune system, inducing apoptosis and inhibiting cancer cell growth. 24 APS exerted a therapeutic effect on hepatocellular carcinoma H22-bearing mice. 25 APS could upregulate the level of nitric oxide (NO) and tumour necrosis factorα (TNFα), which acted as inducers of tumour cell apoptosis. 26 So, APS was selected as the positive control in the present study. The tumour inhibition rates of SNP on hepatoma HepG2-bearing mice were similar to those of the APS group. The thymus and spleen indices are considered to reflect the immunological function of the organism. 27 In the present study, the thymus and spleen indices were significantly increased, and the morphological changes were evident in the tumour tissue after treatment with SNP, thus suggesting that SNP may inhibit HepG2 tumours via an enhanced immunological effect.
Cytokines and molecular pathways play pivotal roles in the immunomodulatory activity of organisms and in tumour development, and cytokines are the essential immunological response components that implement the antitumour function. 28,29 IL-2 is an essential immunomodulatory cytokine that reflects cellular immunity; thus, an increase in IL-2 level suggests the enhancement of cellular immunity in patients with cancer. 27,29 In the present study, IL-2 levels were significantly increased in all SNP-related groups, which indicated that SNP enhanced the immune function. IFNγ is a critical cytokine having both negative and positive immunomodulatory effects under different circumstances. 30 A role of IFNγ for tumour rejection has been demonstrated by its cytotoxic activity on some tumour cells, upregulating MHC expression and thereby increasing tumour cell recognition and elimination, inducing expression of angiogenesis inhibitors, like IP-10, by tumour cells. 31 TNFα is another pleiotropic cytokine that was initially described as antitumorigenic. 32 TNFα could naturally be bound to TNF receptor on cell surface, thereafter activate TRADD (the dead region protein), phosphorylate to JNK protein, increase the expression of p53 transcription factor and thus promote the transcription of BAX gene. 33,34 Ganoderma lucidum polysaccharides (GL-PS) increased the concentration of serum IL-2, TNFα and IFNγ and enhanced the cytotoxic activity of natural killer cells and T cells, inhibited glioma growth and prolonged the survival of rats. 35 In the present study, the IFNγ and TNFα levels were significantly upregulated after SNP treatment in mice (especially at the concentrations of 100 and 200 mg/kg), indicating that SNP could strengthen the immune response by promoting the expression levels of IFNγ and TNFα.
In the endoplasmic reticulum (ER) stress response, TNFα exerts its role via the PERK-elF2α-ATF4-CHOP axis. 36 ATF4 can activate DDIT3 (CHOP) activity and inhibit Bcl-2/Bcl-xl activity, ultimately promoting cell apoptosis. 37 In inflammation and cancer, TNFα inhibits the activity of PPARγ via IκBα-mediated signalling. 38 IκBα usually exists with nuclear factor kappa-B (NF-κB) as heterodimers in the cytoplasm. When IκBα is separated from the heterodimer, NF-κB is activated and then enters the nucleus to bind to DNA, which promotes the transcription of cytokines. 39 In the present study, both the mRNA and protein expression levels of ATF4, DDIT3 and IκBα were markedly increased after SNP treatment in mice, suggesting that the SNP may be involved in two immunomodulatory signalling pathways to inhibit HepG2 tumour progression: the TNFα/ATF4/ DDIT3 and TNFα/IκBα/PPARγ signalling pathways.
The present findings demonstrated that the gene expression levels of Cyr61, Hsp90 and VEGF decreased after SNP treatment, as detected using high-throughput sequencing assay (data not shown). Cyr61 is associated with cell survival, proliferation and differentiation and is usually upregulated in human tumours. 40 The overexpression of Cyr61 can activate mitogen-activated protein kinase (MAPK) signalling and inhibit p53 activity, 41 as well as promote vascular proliferation by upregulating VEGF. 42 Usually, growth factor receptors, such as EGFR, or signal transduction proteins, including PI3K and AKT, are upregulated in tumour cells, 43,44 which are stabilized by HSP90. 45 HSP90 can induce the expression of VEGF. 46 It was reported that Hsp90 47 and VEGF 48 are important targets for anticancer therapeutics administered over a long time period. In the present study, the protein expression levels of CYR61, HSP90 and VEGF decreased after SNP treatment in mice (except at an SNP concentration of 50 mg/kg). Therefore, we concluded that SNP suppressed tumour growth in HepG2 tumourbearing mice by inhibiting the expression levels of CYR61, HSP90 and VEGF.
Along with the aforementioned signal molecules, ATF4 inhibits Bcl-2/Bcl-xl activity, thus promoting cell apoptosis. 37 HSP90 can also modulate tumour cell apoptosis via AKT, TNFR and NF-κB functions. 49,50 VEGF at a low level can reduce Bcl-2, increase the Bax/Bcl-2 ratio and enhance the expression of the Bax oligomer. 51 Once the Bax oligomer binds to the mitochondrial membrane, the permeability of the mitochondrial membrane can be enhanced, which causes the translocation of Cyt-C from the mitochondria to the cytoplasm. 52 In the cytoplasm, the apoptotic body, formed from Apaf-1, Cyt-C and pro-caspase-9, induces the hydrolytic activation of caspase-9, which can cause a cascade reaction and cleavage of cell DNA, ultimately leading to cell apoptosis. 53 Pollen polysaccharides from Chinese wolfberry (WPPs) decreased the levels of AKT, p-AKT and Bcl-2 proteins and increased expression of Bax, caspase-3 and caspase-9 in DU145 cells, which eventually promoted apoptosis. 4 In the current study, SNP significantly promoted the protein expression levels caspase-3, caspase-8, caspase-9 and Bax in a dose-dependent manner in tumour tissues of mice. Contrary to the pro-apoptotic factors, Bcl-2 and VEGF protein expression levels were inhibited. These results demonstrated that SNP induced apoptosis via the mitochondrial apoptosis pathway in tumour cells.

| CON CLUS IONS
In conclusion, a novel polysaccharide was extracted from Sipunculus nudus (SNP). The molecular weight (MW) of SNP was 9223 Da according to HPGPC analysis, and the structure of SNP repeat units was determined by GC-MS and NMR assays. The findings suggested that SNP exerts its antitumour activity by influencing immunoregulation and inducing the mitochondrial apoptosis of tumour cells.
Thus, SNP may be investigated in the future as a promising candidate as an antitumour drug.

| INS TITUTIONAL RE VIE W BOARD S TATE M E NT
The experiments were performed in accordance with the guide-

CO N FLI C T O F I NTE R E S T
The authors have declared that they have no competing interests.

DATA AVA I L A B I L I T Y S TAT E M E N T
Data are available in a publicly accessible repository.