**1. Introduction**

Ulcerative colitis (UC), a significant form of inflammatory bowel disease (IBD), is a chronic inflammatory disorder of the colonic mucosa. It usually starts in the rectum and spreads to the colon proximally, in a continuous manner [1]. The global prevalence of IBD is increasing every year—more than 0.3% of the population suffers from the disease [2]. The pathogenesis of IBD is unclear; however, many researches claim that it is a result of genetic defects, unhealthy lifestyles, imbalance of intestinal microbiota, or immune dysbiosis [3–5]. Typical features of UC include structural and functional impairment of the intestinal mucosa, accompanied by diarrhea, rectal bleeding, and other symptoms [6]. Currently, surgery and drugs (including corticosteroids, aminosalicylic acid, and immunosuppressive agents) are the main strategies from a clinical aspect [7]. However, in addition to the high costs and incomplete treatment issues, these drug interventions have certain side-effects

and have been seen to lose e fficacy in long-term use patients [8]. Consequently, more e fforts should be made to enrich treatment and disease prevention methods.

Numerous reports have testified to inflammatory reactions being a crucial factor causing UC [9]. The cytokines secreted by intestinal epithelial cells are important markers of the intestinal immune system and regulate their inflammatory responses [10]. Pro-inflammatory cytokines such as tumor necrosis factor (TNF)- α, interferon (IFN)-γ, and interleukin (IL)-1β can promote an inflammatory response, eventually causing colon tissue injury, while anti-inflammatory cytokines such as IL-4 and IL-10 can reduce inflammation [11]. Excessive pro-inflammatory cytokines increase the production of reactive oxygen species (ROS), which ultimately cause intestinal epithelial cell injury and aggravate pathogenesis [12]. Hence, materials with the ability to recover abnormal levels of cytokines and oxidative stress may be promising therapeutics for IBD.

Ulvan is a kind of sulphated polysaccharide located in the cell walls of *Ulva*; it represents an abundant marine resource distributed worldwide. Ulvan extracted from di fferent species of *Ulva* consists primarily of Rha (16.8–45.0%), GlcA (6.5–19.0%), Xyl (2.1–12.0%), iduronic acid (IdoA) (0.7–9.1%), Glc (0.5–6.8%), and sulphate (14.3–23.2%) and its backbone is mostly composed of α-1,4- and α-1,2,4- linked L-rhamnose, β-1,4- and terminally linked D-glucuronic acid and β-1,4-linked D-xylose [13]. Like other sulfated polysaccharides, ulvan exerts various therapeutic activities, such as antibacterial, immunostimulatory, antitumor, antioxidant, antihyperlipidemic, and anticoagulant properties [14,15]. Qi et al. (2005) reported that high sulfate content ulvan showed a stronger scavenging activity of superoxide and hydroxyl radicals, reducing power, and metal chelating ability [16]. In a research work by Li et al. (2018), the ulvan extracted from *Ulva pertusa* showed significant protection against liver damage by oxidative stress induced by a cholesterol-rich diet [17]. De Ara újo et al. (2016) found that the enzymatic digestion of ulvan extracted from *Ulva lactuca* had a vascular anti-inflammatory e ffect by decreasing TNFα and IL-1 levels [18]. Additionally, Berri et al. (2017) reported that ulvan extracted from *Ulva armoricana* was able to initiate and amplify the protective immune responses of the host, and regulate mucosal immunity against intestinal pathogens [19].

Reports on the antiinflammatory and antioxidant effects of ulvan are readily available, but few reports have been published on its protective effects in rats with DSS-induced colitis. In this study, LMW-ulvan produced by the enzymatic method was purified and characterized, and its potential to alleviate IBD symptoms in DSS-induced mice and protect the intestinal epithelial barrier were investigated.

#### **2. Results and Discussion**

#### *2.1. Characterization of LMW-Ulvan*

#### 2.1.1. Determination of Molecular Weight of LMW-Ulvan

The average molecular weight of ulvan was 1068.2 kDa according to a previous research by Chi et al. (2020) [20]. Because of its high molecular weight, ulvan usually shows poor biological absorption compared to oligosaccharide [13]. Consequently, the ulvan was degraded by ulvan lyase and the purified LMW-ulvan was prepared using ion exchange chromatography (Figure 1A). The molecular weight distribution of LMW-ulvan was then measured (Figure 1B). According to the standard curve obtained from dextran standards with di fferent molecular weights (logMw = −0.3643x + 10.049, R<sup>2</sup> = 0.991) and retention time of the polysaccharide peak, the weight-average molecular weight of LMW-ulvan was calculated to be 2.56 kDa.

#### 2.1.2. Monosaccharide Composition Analysis of LMW-Ulvan

The LMW-ulvan generated in this study was mainly composed of 57.23% Rha, 28.76% Xyl, 7.42% GlcA, and 1.77% Glc (Figure 2). The monosaccharide composition of LMW-ulvan was slightly di fferent from that of ulvan, whose molar percentage of Rha, Xyl, GlcA, IdoA, and Glc was 56.50%, 20.34%, 16.95%, 2.82%, and 3.39%, respectively [20]. The lyase used in this study belonged to the

PL25 family, which cleaves the (1→4) glycosidic bond between 3-*O*-sulfated Rha and GlcA [21]. After enzymatic hydrolysis, unsaturated uronic acid was formed in a β-elimination reaction [22]. Consequently, the characterization of ulvan lyase might be the reason for a significant decrease in percentage of GlcA.

**Figure 1.** Purification of LMW-ulvan using a DEAE column (**A**) and its gel permeation chromatogram (**B**).

**Figure 2.** High performance liquid chromatography (HPLC) analyses of monosaccharide standards (**A**) and LMW-ulvan (**B**). Lactose is the internal standard.

2.1.3. Methylation and Gas Chromatography-Mass Spectrometer (GC-MS) Analysis of LMW-Ulvan

The integrated structure of LMW-ulvan, including linkages and sulfate substitution position, were further determined by methylation analysis and GC-MS. The completeness of methylation of LMW-ulvan and desulfated LMW-ulvan (dsLMW-ulvan) was confirmed by FT-IR spectroscopy with the disappearance of OH bands (data not shown). As shown in Table 1, LMW-ulvan mainly consisted of (1→3,4)-linked Rha, and (1→4)-linked Xyl with small amounts of (1→4)-linked Rha residues. Sulfate substitution was at C-3 of (1→4)-linked Rha because of the distinctive increase of amounts of (1→4)-linked Rha and significantly decrease of the amounts of (1→3,4)-linked Rha in dsLMW-ulvan. Besides, according to the characterization of ulvan lyase belonged to the PL25 family, which cleaves the (1→4) glycosidic bond between 3-*O*-sulfated Rha and GlcA [21]. Thus we deduced that there is 4)-GlcAp-(1 in the LMW-ulvan. The linkages and sulfate substitution position of LMW-ulvan are similar to polysaccharide from *Ulva* species [23].


**Table 1.** Profile of partially *O*-methylated alditol acetates obtained by methylation analysis of LMW-ulvan.

> Desulfation of the LMW-ulvan: dsLMW-ulva, rhamnose: Rha, xylose: Xyl.

#### *2.2. LMW-Ulvan Relieves the Manifestations of IBD Induced by DSS*

To evaluate the anti-colitis effects of LMW-ulvan, the DSS was used to establish a model to mimic human UC and determine the therapeutic effect of LMW-ulvan [24]. Preliminary tests revealed that 30 mg/kg of LMW-ulvan showed no protective effects in IBD rats, while 50 mg/kg of LMW-ulvan began to show protective effects in IBD. Besides, 100 and 120 mg/kg of LMW-ulvan were more effective. Thus, 50 and 100 mg/kg were chosen for subsequent studies. The results of the acute toxicity experiment showed that an LMW-ulvan dose less than 1200 mg/kg showed no significant damage to ICR mice. Consequently, both doses used in this study were safe. The clinical symptoms of IBD in mice were comprehensively determined by changes in body weight (Figure 3A), colon length (Figure 3B), and DAI (disease activity index) score (Figure 3C). Body weight of the mice in the N group showed a gradual upward trend during the experiment period. In contrast, the body weight of the mice in the M group revealed a downward trend, especially after the fifth day. The loss in weight of the mice was significantly improved on treatment with LMW-ulvan of both dosages. There was no significant difference between the LP group and the PC group. Another indicator reflecting the severity of intestinal inflammation is colon length, which is restored with improvement in inflammation [25,26]. The colon length of mice in the M group was significantly shorter than that of the N group (Figure 3B). However, after treatment with LMW-ulvan, the colon was distinctly longer than that of the M group and similar to that of the PC group. In addition, the DAI score were also significantly reduced in the LMW-ulvan treatment group compared with that of the M group on the sixth day (Figure 3C). Spleen and thymus index was measured as an indirect marker of inflammation in all the experiment groups. As shown in Figure 3D,E, DSS decreased the thymus index and increased the spleen index significantly, reflecting on worse immune activity. LMW-ulvan was observed to improve the immune system because the spleen and thymus index of the HP group recovered to normal levels and showed no significant difference as compared to that of the N group. Besides, Figure 3E also showed that the LMW-ulvan of both dosages could significantly relieve splenomegaly induced by DSS.

#### *2.3. LMW-Ulvan Reverses the Histological Injury of Colonic Epithelium Caused by DSS*

Due to severe colitis, mice in the M group exhibited mucosal degeneration and necrosis, distortion of crypts, loss of goblet cells, and mucosal and submucosal infiltration of mononuclear inflammatory cells, as compared to that of the N group (Figure 4A,B). However, in all mice in the LP group, infiltration and aggregation of mononuclear inflammatory cells were observed in the submucosa of the colon, with improvement observed in epithelial degeneration and necrosis. In mice from the HP group, a nearly intact structure with few inflammatory cells as that of the N group was exhibited. This result indicates that LMW-ulvan could reduce the inflammatory infiltration and damage caused by DSS.

**Figure 3.** Assess the severity of IBD (Inflammatory bowel disease) in each group: (**A**) Change in body weight, (**B**) colon length of mice in each group, (**C**) DAI (disease activity index) score, (**D**) thymus index, and (**E**) spleen index. N: normal group, M: model, PC: positive control, LP: low dose LMW-ulvan (50 mg/kg), and HP: high dose LMW-ulvan (100 mg/kg).

**Figure 4.** H&E staining of representative colons from different treatment groups (200×): (**A**) N group, (**B**) M group, (**C**) PC group, (**D**) LP group, and (**E**) HP group. N: normal group, M: model, PC: positive control, LP: low dose LMW-ulvan (50 mg/kg), and HP: high dose LMW-ulvan (100 mg/kg).

#### *2.4. E*ff*ects of LMW-Ulvan on Cytokines in Colon Tissue and Serum*

It is well known that an abnormal inflammatory response is one of the most striking features of IBD [27]. The result showed that DSS increased the levels of IL-1β and IFN-γ significantly and decreased the level of IL-4, when compared to the N group (Figure 5). After administration of LMW-ulvan, the levels of IL-1β, IFN-γ, and IL-4 in colon tissue and serum all recovered to normal levels. IL-1β is a kind of pro-inflammatory cytokine and its active form is regulated by caspase 1 secreted from the NLRP3 inflammasome [28]. In this study, LMW-ulvan reduced the level of IL-1β in the colon and serum, which indicated its possibility to suppress NLRP3 inflammasome activation. Activated Th cells are classified as type 1 of T helper cells (Th1) and type 2 of T helper cells (Th2) according to different sorts and bioactivities of cytokines [29]. Usually, Th1 mediates a cell-mediated immune response, while Th2 mediates a humoral immune response [30]. IFN-γ and IL-4 are representative cytokines to coordinate immunity secreted by Th1 and Th2, respectively [31]. The IFN-γ/IL-4 ratio can also reflect the balance of Th1/Th2. The results showed that IFN-γ/IL-4 ratios in serum of the N, M, PC, LP, and HP groups were similar to that in the colon and were about 1.0, 2.5, 1.1, 1.2, and 1.1, respectively. Consequently, DSS disturbs the balance of Th1/Th2 toward to Th1 by increasing the value of IFN-γ and reducing the secretion of IL-4. The results indicated that LMW-ulvan could inhibit the Th1 cell response and improve the Th2 cell response, which is beneficial in alleviating the inflammatory damage caused by DSS.

**Figure 5.** Levels of inflammatory cytokines in serum and colon tissues. (**A**) interleukin (IL)-1β level in serum. (**B**) IL-4 level in serum. (**C**) interferon (IFN)-γ level in serum. (**D**) IL-1β level in colon. (**E**) IL-4 level in colon. (**F**) IFN-γ level in colon. N: normal group, M: model, PC: positive control, LP: low dose LMW-ulvan (50 mg/kg), and HP: high dose LMW-ulvan (100 mg/kg).

#### *2.5. E*ff*ects of LMW-Ulvan on Malondialdehyde (MDA), GPx, and Catalase (CAT) in Colonic Tissue*

Excessive inflammation augments oxidative stress, which has been associated with extensive infiltration of immune cells, and activation of NF-κB signaling and ER stress, thereby causing tissue damage and exacerbation of a patient's condition [12,32,33]. As shown in Figure 6, DSS significantly enhanced the MDA level, but downregulated the enzyme activity of GPx and CAT. LMW-ulvan at doses of 50 and 100 mg/kg significantly reduced (*p* < 0.05) the MDA level and increased the enzyme activity of GPx and CAT. Specifically, the activity of GPx and CAT of the HP group was restored to that of the normal group. MDA, an indicator reflecting the accumulation of ROS in response to oxidative damage, is one of the most important products of lipid peroxidation and could damage the mucosa [34], while GPx is an important component of the glutathione antioxidant system, which forms an antioxidant barrier in the gu<sup>t</sup> mucosa [35]. Based on the above analysis, the protective effect of LMW-ulvan on IBD damage might be attributed to an enhanced antioxidant defense system.

**Figure 6.** The level of MDA (**A**), activity of GPx (**B**), and activity of CAT(**C**). N: normal group, M: model, PC: positive control, LP: low dose LMW-ulvan (50 mg/kg), and HP: high dose LMW-ulvan (100 mg/kg).

#### *2.6. E*ff*ect of LMW-Ulvan on mRNA Expression of Colonic Epithelial Tight Junction Protein*

To ascertain whether LMW-ulvan treatment can protect the integrity of the intestinal barrier, the expression of claudin, occluding, and tight junction protein 1 (ZO-1) was detected at both protein and mRNA level (Figure 7). The expression of three tight junction components in the M group was reduced significantly (*p* < 0.05), while the expression in the LMW-ulvan treatment groups was increased to di fferent degrees compared with that of the N group. For ZO-1 and claudin, the expression in both HP and LP groups was significantly increased and the low dose LMW-ulvan showed a similar e ffect as that of 5-aminosalicylic acid (5-ASA) (Figure 7C,D). For occludin, the expression level in both HP and LP groups showed no significant di fference with that of the N group (Figure 7B). Claudin, occludin, and ZO-1 are the main components of tight junction, which is the most significant structure in the intestinal mucosal mechanical barrier [10]. The tight junction can e ffectively prevent enteric pathogens, antigens, and other substances from entering the intestinal mucosa, preventing the activation of immune cells and abnormal immune responses and maintaining stability of mucosal barrier function and intestinal permeability [36]. The improvement of LMW-ulvan on DSS-induced colitis might be closely related to its protective e ffect on the intestinal mucosal barrier by increasing the expression of tight junction proteins. Thus, LMW-ulvan could further inhibit the abnormal immune response in the intestinal mucosa, thus ameliorating mucosal barrier function and intestinal mucosal permeability.

**Figure 7.** The expression of colonic epithelial tight junction protein. ( **A**) Representative western blotting images for colonic epithelial tight junction protein. (**B**) The relative mRNA content of claudin. ( **C**) The relative mRNA content of occludin. ( **D**) The relative mRNA content of ZO-1. N: normal group, M: model, PC: positive control, LP: low dose LMW-ulvan (50 mg/kg), and HP: high dose LMW-ulvan (100 mg/kg).

#### **3. Materials and Methods**

#### *3.1. Materials and Reagents*

*Ulva pertusa* was collected from the coast near Weihai, China, in 2019. Ulvan lyase was provided by the Applied Microbiology Laboratory (Ocean University of China). DSS (36–50 kDa) and 5-aminosalicylic acid (5-ASA) were purchased from MP Biomedicals (Solon, CA, USA) and Sigma-Aldrich (St. Louis, MO, USA), respectively. ELISA detection kits for IL-1β, IL-4, IFN-γ, malonic dialdehyde (MDA), catalase (CAT), and glutathione peroxidase (GPx) were obtained from Dakewe Biotech Corporation (Beijing, China). Primary antibodies specific for β-actin, ZO-1, occluding, and claudin-1 were purchased from Protein Tech Group (Wuhan, China). The BCA Protein Assay Kit was purchased from Beijing Solarbio Science & Technology Co., Ltd. (Beijing, China). All other reagents were of analytical grade.

#### *3.2. Enzymatic Preparation and Purification of LMW-Ulvan*

The ulvan was extracted according to a method described by Qiao et al. (2020) [37]. It was then degraded though an enzymatic method described by Chi et al. (2020) [20] and crude enzymatic fractions with average molecular weight of 1–5 kDa were obtained by nanofiltration membrane system. The obtained fraction was purified using ÄKTAprime plus (GE Healthcare, Woburn, MA, USA) equipped with DEAE-Sepharose CL-6B column (1.6 × 10 cm). The fraction was pre-equilibrated with ultrapure water and then eluted with a linear gradient of 0–1 M NaCl at 1.0 mL/min. The purified fraction was then dialyzed, lyophilized, and nominated as LMW-ulvan.

#### *3.3. Characterization of LMW-Ulvan*

The total sugar content of LMW-ulvan was determined by phenol-sulfuric acid method [38]. Sulfate content was determined after hydrolysis of trifluoroacetic acid [39]. The molecular weights of LMW-ulvan were measured according to a method described by Ye et al. (2019) [40]. The Fourier transform infrared (FI-IR) spectra of LMW-ulvan were recorded using the Magna-IR560 spectrometer (Nicolet Instrument Corp., Madison, WI, USA) according to research by Cui et al. (2019) [41]. Its monosaccharide composition was measured by reversed-phase HPLC (Agilent Technologies, Santa Clara, CA, USA) after pre-column derivatization, according to a method described by Yu et al. (2017) [22].

#### *3.4. Methylation Analysis of LMW-Ulvan*

Desulfation of the LMW-ulvan was achieved according to Falshaw and Furneaux (1998) [42]. Briefly, the LMW-ulvan was converted to the pyridinium salt form by dialysis against pyridinium hydrochloride (0.1 M, adjusted to pH 6.8), then against distilled water, and finally lyophilised. The resulting materials was dissolved in 89:10:1 *v*/*v* Me2SO-MeOH-pyridine, and heated for 4 h at 100 ◦C. After cooling, the desulfated product was recovered by dialysis, freeze-dried, and designated as dsLMW-ulvan.

Methylation analysis was performed as the reported method with some modification (Li et al., 2019) [23]. Briefly, 1 mg sample (LMW-ulvan or dsLMW-ulvan) dried by P2O5 was dissolved in 2.0 mL DMSO, and then 100 mg anhydrous NaH was added under nitrogen atmosphere. The mixture was stirred at 25 ◦C for 2 h under nitrogen atmosphere, and then 1 mL CH3I was added dropwise in an ice-cold water bath. The mixture was incubated in the dark at 25 ◦C, stirring for 3 h. Finally, 1 mL distilled water was added to terminate the reaction, and then extracted with CHCl3. The extract was washed with distilled water and evaporated to dryness. The methylated polysaccharide was converted into partially methylated alditol acetates, which were analysed by GC-MS (Thermo Fisher Scientific, MA, USA). The GC-MS analysis designed for methylation analysis was performed according to a report by Lin et al. (2016) [43].
