*Article* **Comparative Analyses of Fucoidans from South African Brown Seaweeds That Inhibit Adhesion, Migration, and Long-Term Survival of Colorectal Cancer Cells**

**Blessing Mabate 1, Chantal Désirée Daub 1, Brett Ivan Pletschke 1,\*and Adrienne Lesley Edkins 2,\***


**Abstract:** Human colorectal cancer (CRC) is a recurrent, deadly malignant tumour with a high incidence. The incidence of CRC is of increasing alarm in highly developed countries, as well as in middle to low-income countries, posing a significant global health challenge. Therefore, novel management and prevention strategies are vital in reducing the morbidity and mortality of CRC. Fucoidans from South African seaweeds were hot water extracted and structurally characterised using FTIR, NMR and TGA. The fucoidans were chemically characterised to analyse their composition. In addition, the anti-cancer properties of the fucoidans on human HCT116 colorectal cells were investigated. The effect of fucoidans on HCT116 cell viability was explored using the resazurin assay. Thereafter, the anti-colony formation potential of fucoidans was explored. The potency of fucoidans on the 2D and 3D migration of HCT116 cells was investigated by wound healing assay and spheroid migration assays, respectively. Lastly, the anti-cell adhesion potential of fucoidans on HCT116 cells was also investigated. Our study found that *Ecklonia* sp. Fucoidans had a higher carbohydrate content and lower sulphate content than *Sargassum elegans* and commercial *Fucus vesiculosus* fucoidans. The fucoidans prevented 2D and 3D migration of HCT116 colorectal cancer cells to 80% at a fucoidan concentration of 100 μg/mL. This concentration of fucoidans also significantly inhibited HCT116 cell adhesion by 40%. Moreover, some fucoidan extracts hindered long-term colony formation by HCT116 cancer cells. In summary, the characterised fucoidan extracts demonstrated promising anti-cancer activities in vitro, and this warrants their further analyses in pre-clinical and clinical studies.

**Keywords:** cancer; migration; adhesion; fucoidans; human colorectal cancer; *Ecklonia radiata*; *Ecklonia maxima*; *Sargassum elegans*

#### **1. Introduction**

Cancer is a complex, multifactorial disease characterised by the uncontrollable growth of abnormal cancerous cells [1]. Cancers may progress to invade and spread to other tissues and organs using the circulatory and lymphatic systems through metastasis [2]. Cancer has one of the highest mortality rates and significantly contributes to lower global life expectancy [3]. The cancer burden globally is expected to increase from 2020 by approximately 47%, translating to about 28.4 million new cases per year by 2040. However, the increased number of cancer cases may be affected by the social-economic status of the global populace. Additionally, the rise may be linked to increased risk factors associated with globalisation and the growing economy [4]. These risk factors may include increased processed food consumption, lack of physical activity, and increasing obesity.

Colorectal cancer (CRC) is the third most diagnosed cancer (accounting for 10% of all cases) and the second most frequent cause of cancer-related deaths (accounting for

**Citation:** Mabate, B.; Daub, C.D.; Pletschke, B.I.; Edkins, A.L. Comparative Analyses of Fucoidans from South African Brown Seaweeds That Inhibit Adhesion, Migration, and Long-Term Survival of Colorectal Cancer Cells. *Mar. Drugs* **2023**, *21*, 203. https://doi.org/ 10.3390/md21040203

Academic Editors: Yuya Kumagai, Hideki Kishimura and Benwei Zhu

Received: 4 March 2023 Revised: 20 March 2023 Accepted: 22 March 2023 Published: 24 March 2023

**Copyright:** © 2023 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

9.4% of oncological deaths). Thus, it constitutes a substantial portion of the global cancer burden [4,5]. Treatment strategies include chemotherapy, radiation therapy, surgery, or combination therapies [6]. Although surgical resection of the primary tumour in the early disease stages proves effective, patients may be diagnosed at more advanced stages [7,8]. The indiscriminate toxic effects of chemotherapeutic agents used for CRC treatment result in debilitating side effects and limit therapeutic outcomes [7,9].

With challenges of side effects, affordability, and access to current therapeutic remedies, the search for novel treatment and preventive strategies with minimal adverse effects must proceed urgently. Furthermore, marine bio-products have historically been deemed therapeutic advantages among other bio-compounds [10]. Natural compounds have gained attention over the past decades as these demonstrate targeted specific anti-cancer properties while demonstrating low toxicity [1]. The lower incidences of chronic diseases, such as heart disease, diabetes, and cancer in China and Japan have led researchers to investigate the contents of brown seaweeds, which have been used in their cuisines and medicinal applications [11]. Among the more than 3000 natural products derived from seaweeds, fucoidans have received significant attention for their most promising anti-cancer properties [1,11].

Fucoidan is a heparin-like structured, naturally derived polysaccharide compound present in the cell wall matrix of brown seaweeds [12]. This heterogeneous polysaccharide is predominantly comprised of L-fucose with smaller quantities of varying monosaccharides and sulphate, which contribute to its complex structural characteristics and have an unquestionable effect on its broad range of biological activities [1]. These biological activities include anti-oxidant [13], anti-coagulant [14], anti-thrombotic, anti-inflammatory, anti-viral, anti-lipidemic [15], anti-diabetic [16], anti-metastatic and anti-cancer activities [17].

Fucoidans have anti-cancer effects against various cancer cell lines by causing cell cycle arrest [18], inducing apoptosis [9], preventing angiogenesis [9,19], and inhibiting migration and metastasis [1]. As tumour migration is a hallmark of cancers, it is plausible to target this process to alleviate tumour progression. Moreover, fucoidans inhibit metastasis by blocking cell migration and colony formation. Fucoidan isolated from *F. vesiculosus* significantly inhibited the migration of the human colon cancer cell line HT-29 by suppressing PI3k/Akt/mTOR/p70S6K1 [19]. Whereas the treatment of colorectal carcinoma cells, DLD-1 and HCT116, with fucoidan from *Padina boryana,* proved successful in inhibiting colony formation [20].

The anti-cancer effect of fucoidan on colon cancer cell lines has been reported primarily using the commercially available *F. vesiculosus* fucoidan. However, the diversity of brown seaweeds is broad, and their bioactivities have been linked to the source of seaweed and its structural and chemical characteristics. Additionally, in addition to the limited literature on fucoidan effects on colon cancers, there is also limited literature available on the biological activities, including the anti-cancer properties of South African seaweed-derived fucoidans. However, the country harbours one of the most extensive coastlines globally, with a rich seaweed biodiversity [21]. The present study characterised fucoidan extracts from native South African brown seaweeds and linked their structural differences to their anti-cancer properties against the HCT116 cell line.

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

#### *2.1. Fucoidan Yield*

The fucoidans in this study were hot water extracted, except for the *F. vesiculosus* fucoidan, which was purchased commercially. Considerable amounts of fucoidans were successfully extracted with an average fucoidan/defatted seaweed dry weight ratio of 5.4, 5.9 and 2.2% for *E. maxima*, *E. radiata* and *S. elegans,* respectively. The resulting yields of the extracted fucoidans were within the expected range (1.1–4.8%) for water extracted fucoidans [22].

#### *2.2. Structural Analysis of Fucoidans*

#### 2.2.1. FTIR Analysis

Fucoidan extracts were structurally analysed by Fourier-transform infrared spectroscopy (FTIR) (Figure 1) and displayed similar spectra to previously characterised fucoidans [22,23]. All the profiled fucoidans displayed a spectral band between 3500 cm−<sup>1</sup> and 3200 cm−1, characteristic of polysaccharides. This peak is associated with the stretching vibrations of the O-H groups within carbohydrates. The bands in the region 2900 to 3000 cm−<sup>1</sup> observed in all the fucoidans are assigned to the C-H stretching in the pyranose ring and methyl groups associated with the fucose [24].

**Figure 1.** FTIR spectra of the fucoidans under study. The overlaid spectra were obtained from water-extracted fucoidans and the commercial *F. vesiculosus* fucoidan.

Typically, the carbonyl groups and stretching of O-acetyl groups are depicted by the peaks around 1650 cm−<sup>1</sup> [25]. Additionally, the stretching of the S=O bond linked with sulphate groups is characterised by peaks between 1210 and 1270 cm−<sup>1</sup> [22]. Stretching vibrations of the glycosidic C–O bonds within the fucoidans structures are represented by peaks close to the wavenumber 1100 cm−<sup>1</sup> [26]. Furthermore, the peaks at wavenumber 854 cm−<sup>1</sup> depict sulphate groups on fucoidans linked to carbonyl side chains [13]. This peak at around 854 cm−<sup>1</sup> was more pronounced for the *F. vesiculosus* commercial fucoidan. This suggested that *F. vesiculosus* fucoidan had a relatively higher sulphate content than the extracted fucoidans (Figure 1).

#### 2.2.2. Proton NMR Analysis of Extracted Fucoidans

The structural composition of the extracted fucoidans was also elucidated by proton NMR. The fucoidans generally exhibited chemical shifts (Figure 2) similar to several characterised fucoidans [14,27,28]. The chemical shifts in all fucoidans showed peaks at 1.28 ppm and 1.45 ppm suggesting the presence of alternating α (1–3) and α (1–4) linkages of fucose residues linked with sulphates (α-L-Fuc, α-L-Fuc (2-SO3 −) and α-L-Fuc (2,3-diSO3 −) [28]. The *F. vesiculosus* fucoidan displayed relatively more prominent peaks in the 1.28 ppm and 1.45 ppm range, suggesting a higher sulphate content than the extracted fucoidans (Figure 2). A higher sulphate content in the fucoidans may improve their biological activities.

**Figure 2.** 1H NMR for the different seaweed fucoidans. Overlaid proton NMR spectra of extracted fucoidans and commercial *F. vesiculosus* fucoidan.

Vibration bands at 1.45 ppm are assigned to symmetric CH3 deformations emanating from methyl proton on C6 of fucose [14]. The peaks at 2.1 ppm are assigned to the H-6 methylated protons of *L*-fucopyranosides [27]. The peaks in the range of 3.5–4.5 ppm are characteristic of the (H2 to H5) ring protons of *L*-fucopyranosides. The exhibited peaks in the ring proton region also suggest variable fucosyl sulphates located at variable glycosidic linkages with varying monosaccharide patterns. The definitive peaks in Figure 2 at 3.3 ppm and 3.7 ppm in the fucoidans suggest the presence of hexoses, including glucose, galactose, and mannose [27]. Our results show that the spectra of *E. maxima* and *E. radiata* fucoidans

displayed more pronounced peaks at the 3.3 ppm to 3.7 ppm region than the *F. vesiculosus* and *S. elegans* fucoidans (Figure 2). This could suggest that *Ecklonia* sp. fucoidans have higher sugar content. Furthermore, the extracted fucoidans had limited to negligible uronic acid contamination as there were no chemical shifts in the region around 5.8 ppm (Figure 2) [29].

#### 2.2.3. Thermogravimetric Analysis of Fucoidans

The TGA decomposition profiles of the fucoidans validated the compounds as polysaccharides, as their decomposition started just above 200 ◦C (Figure 3), characteristic of the organic polymers [30].

**Figure 3.** Thermal gravimetric analysis (TGA) analysis of the fucoidan extracts. Superimposed thermograms for the water-extracted fucoidans from seaweeds and commercial *F. vesiculosus* fucoidan.

The TGA plots of the fucoidans showed about 20% loss in mass at a temperature of 240 ◦C, associated with the loss in moisture content through evaporation [31] and some volatile matter [32]. The most significant loss of mass (~45%) occurred between 240 ◦C and 420 ◦C, which accounted for the arbitrary depolymerisation and decomposition of organic constituents, such as carbohydrates. Notably, *F. vesiculosus* fucoidan depolymerisation and decomposition of organic matter occurred relatively more rapidly than the other fucoidans, as shown by the steeper slope in Figure 3. Its relatively low carbohydrate content may be the reason for this observation (Table 1). Above 420 ◦C, combustion of carbon black occurred. The remaining residual mass at 600 ◦C accounted for the ash content, usually containing sulphates, phosphates, and carbonates [33]. The profiles of the extracts were characteristic of previously profiled fucoidan extracts in the literature [34,35].

The structural analysis of fucoidans through FTIR, proton NMR and TGA confirmed the integrity of our extracts, as they showed comparable patterns to the commercial *F. vesiculosus* fucoidan. It was also evident that *Ecklonia* sp. fucoidans displayed similar yet unique profiles to the *S. elegans* and *F. vesiculosus* fucoidans. Our observations support the findings of Ermakova and colleagues that diverse seaweed species yield diverse fucoidan structures [36]. These differences may be caused by the survival needs of the seaweeds, influenced by their habitat. Considering the unique profiles observed within the structural analyses of the fucoidans, these were further characterised chemically to assess their composition.


**Table 1.** Composition of fucoidan structures.

Determined by <sup>a</sup> Phenol sulphuric acid method; <sup>b</sup> HPLC (RID); <sup>c</sup> Barium chloride gelatin method; <sup>d</sup> Folin-Ciocalteu method; <sup>e</sup> MegazymeTM uronic acid kit; <sup>f</sup> Bradford's assay; <sup>g</sup> TGA; <sup>h</sup> size exclusion HPLC.

#### *2.3. Composition of Fucoidans*

The fucoidans were partially characterised chemically by determining their total sugar contents, monosaccharides distributions and impurities (including protein, phenolics and uronic acids). *E. maxima* and *E. radiata* fucoidans contained high amounts of total carbohydrates, with 72.8% and 88% (*w*/*w*), respectively (Table 1). The *S. elegans* and *F. vesiculosus* fucoidans had approximately 40% (*w*/*w*) total carbohydrate content (Table 1) and were, therefore, comparatively lower than that of the *Ecklonia* sp. extracted fucoidans.

After hydrolysing the fucoidans using 2 M TFA, monosaccharides were quantified using HPLC and Megazyme kits (Table 1). The predominant monosaccharides detected in all fucoidans were fucose, glucose, galactose, and mannose. Generally, *Ecklonia* sp. fucoidans had a relatively high monosaccharide content, with glucose, fucose, galactose and mannose being the most prominent sugars (Table 1). These findings are consistent with the findings of January and colleagues, who detected considerable amounts of glucose, galactose, and mannose in their *E. maxima* fucoidan extract [37]. The commercial *F. vesiculosus* fucoidan contained relatively higher levels of fucose than that of the extracted fucoidans (Table 1). Furthermore, *S. elegans* fucoidan had notably higher mannose content than the other fucoidans (Table 1). The monosaccharide distribution of the extracts is representative of the characteristic fucoidans.

Commercial *F. vesiculosus* had the highest sulphate content (about 15%), followed by *S. elegans* fucoidan, which had 9.7% (Table 1). *S. elegans* and *F. vesiculosus* fucoidans had higher sulphate contents than the *Ecklonia* sp.-derived fucoidans (between 7–8%). The higher sulphates within *F. vesiculosus* and *S. elegans* fucoidan determined by colourimetry agreed with the structural characterisation data (Figures 1 and 2). FTIR spectra at wavenumber around 845 cm−<sup>1</sup> showed a more pronounced peak for the *F. vesiculosus* fucoidan than all extracts (Figure 1). The pronounced NMR peaks indicative of sulphates between ppm 1.2 and 1.6 were evident for the *F. vesiculosus* and *S. elegans* than for *Ecklonia* sp. fucoidans (Figure 2). Additionally, the ash content was higher within the *S. elegans* and *F. vesiculosus* fucoidans (Table 1), suggesting more sulphates among these fucoidans than the *Ecklonia* sp. fucoidans. The ash content detected from the fucoidans was between 19 and 24%, consistent with the ash contents in some characterised fucoidans [15,38]. Furthermore, the fucoidans had minimal protein and uronic acid contamination, with *S. elegans* having the highest at ~4% of each. Insignificant amounts (<2%) of phenolics were detected within the fucoidans (Table 1). The molecular weights of the fucoidans were determined by size exclusion HPLC. The molecular size of *E. maxima* fucoidan was 27.4 kDa, *E. radiata* fucoidan was 8.5 kDa, *S. elegans* fucoidan was 74.9 kDa, and *F. vesiculosus* fucoidan was 84.4 kDa. Structural and chemical characterisation data suggest that the extracted crude fucoidans were relatively pure as they showed similar profiles to the commercial *F. vesiculosus* fucoidan and other previously characterised fucoidans in literature.

#### *2.4. Fucoidans' Cytotoxicity to HCT116 Cancer Cells*

The potential cytotoxicity of all the fucoidan extracts towards the HCT116 colon cancer cell line was examined and compared to the chemotherapeutic drug 5-fluorouracil (5FU). The colon cancer cell line HCT116 was selected with the probable oral route of administration of fucoidans. For decades 5FU has played a pivotal role in the treatment of colorectal cancer [39]. Thus, it was chosen as a positive control for our experiments. The 5-FU treatments showed robust anti-cancer activity with an IC50 value of 9.9 μM. The reduction in cell viability due to treatment with fucoidan extracts was expressed as the percentage of viable cells remaining after treatment compared to the vehicle-treated control cells. Even at 2.5 mg/mL loading, none of the fucoidans displayed any significant cytotoxic effect on the HCT116 cells (Figure 4). About 4 g/day of fucoidan has been used in combination with other chemotherapeutics, including 5-FU, in colorectal cancer patients. Although patient survival was improved when fucoidan was included in the treatment, a significant observation was reduced side effects [40].

**Figure 4.** Fucoidans' cytotoxicity on HCT116 cells assessed by the resazurin assay. (**a**) Cell viability after treatment with fucoidans, (**b**) IC50 curve of 5-FU (positive control) demonstrates the compound's cytotoxic effect on HCT116 cells. The data represent values obtained from 3 biological replicates expressed as means ± SD (*n* = 3).

The lack of cytotoxicity of the fucoidans could be attributed to the large molecular sizes (Table 1), making penetration into the cells difficult. Large molecular sizes of fucoidans have been reported to limit the bio-accessibility of these compounds, posing a challenge for their applications [41]. Native *Undaria pinnatifida* fucoidan had minimal anti-tumour activity compared to its depolymerised counterpart against the human lung cancer cell line A549 [42]. This observation suggests a need for depolymerising fucoidans to increase toxicity while at the same time maintaining their bioactivities. We acknowledge that size cannot be the only determining factor, but other fucoidans' characteristic factors, including sulphation, and monosaccharide distribution, will contribute to their bioactivities.

#### 2.4.1. The Effect of Fucoidans on HCT116 Colony Formation

Having established that fucoidans did not show substantial cytotoxicity, these compounds were further tested for their ability to inhibit colony formation. This assay has been the method of choice to determine replicative cell death after ionising radiation, although it is also used to determine the effectiveness of other cytotoxic agents [43]. *S. elegans* and *F.* *vesiculosus* fucoidans were significant inhibitors of HCT116 colony formation (*p* < 0.05) (Figure 5). The positive control 4-NQO was used in this assay and showed the dose-dependent inhibition of HCT116 cell colony formation.

**Figure 5.** The dose-dependent clonogenic effect of fucoidan on HCT116 cancer cells. (**a**) Visual representation of the effect of fucoidan extracts (1 mg/mL) on HCT116 colony formation; (**b**) Dosedependent effect of compounds on HCT116 cells' colony formation. The HCT116 colony cells were calculated and expressed as the means ± SD percentages (*n* = 3). The *\** shows a significant treatment difference versus the untreated control (*p* < 0.05) analysed by One-way ANOVA.

The *S. elegans* fucoidan exhibited about 40% colony formation inhibition at 0.5 mg/mL. The *F. vesiculosus* fucoidan inhibited HCT116 cell colony formation by over 50% at 0.1 mg/mL concentration (Figure 5). The inhibition by *F. vesiculosus* and *S. elegans* fucoidans may be attributed to the superior sulphate content compared to that of the *Ecklonia* sp. derived fucoidans (Table 1). A limited number of studies have reported the ability of fucoidans to decrease tumour cell survival using this assay. Nevertheless, our findings agree with those of Shin and colleagues, who reported that manganese dioxide nanoparticles coated with fucoidan decreased colony formation by a pancreatic cancer cell line [44]. Another independent study reported fucoidan inhibited colony formation of HepG2 liver cancer cells [20].

2.4.2. Fucoidans Inhibit the 2D Migration of HCT116 Cancer Cells

Fucoidans were next tested for effects on the 2-dimensional (2D) migration of human HCT116 colorectal cancer cells using the wound healing assay. The *F. vesiculosus* and *S.* *elegans* fucoidans significantly inhibited cell migration compared to the untreated control (Figure 6). *S. elegans* fucoidan showed a dose-dependent inhibition of HCT116 cell migration at all concentrations tested, with inhibition reaching up to 30% at about 0.25 mg/mL (Figure 6), while cell migration inhibition by fucoidan from *F. vesiculosus* was only significant at 0.5 mg/mL.

**Figure 6.** The effect of fucoidan extracts on 2D HCT116 cell migration. Quantified migration profiles of HCT116 cells treated with *E. maxima, E. radiata, S. elegans* and *F. vesiculosus* fucoidan extracts relative to the untreated control. The data are represented as means ± SD of biological replicates (*n* = 3). The asterisk \* represents treatment concentrations that had a statistically significant effect on the migration of the cells at *p* < 0.05 tested by One-way ANOVA.

The *Ecklonia* sp. fucoidans did not significantly inhibit HCT116 cancer cell migration, even at high concentrations (Figure 6). This observation may be linked to the high amount of sugars within their structure (Table 1). Literature has suggested that fucoidans consisting of sugars, including galactose, may provide the nutrition required for wound healing [45]. However, fucoidans with a higher sulphate concentration were associated with a better bioactivity [36,46]. Thus, we can infer that *S. elegans* fucoidan showed better inhibitory action to wound healing of the HCT116 cells due to its unique structural properties, including high sulphation (Figures 1 and 2).

#### 2.4.3. Fucoidans Inhibit HCT116 3D Spheroid Migration

Next, we tested the ability of fucoidans to prevent the migration of cells from a threedimensional sphere onto tissue culture plastics. In this assay, the fucoidans inhibited the migration of the HCT116 cells from spheres in a time-dependent manner (Figure 7a). The commercial *F. vesiculosus*, *S. elegans* and *E. radiata* fucoidans displayed comparable efficacies, showing more than 80% inhibition at 0.1 mg/mL concentration. Furthermore, inhibition of HCT116 spheroid migration by the fucoidans was dose-dependent (Figure 7b). Although *E. maxima* fucoidan showed a slightly lower inhibition potential than the other fucoidans, it still significantly (*p* < 0.05) inhibited migration from HCT116 spheres (Figure 7b).

**Figure 7.** The effect of fucoidans on the 3D HCT116 spheroid migration. (**a**) Time-dependent effect of fucoidans at a fixed concentration (0.1 mg/mL) on spheroid migration; (**b**) Quantification of the dose-dependent effect of fucoidans on 3D spheroid HCT116 migration. The data are represented as means ± SD of biological replicates of spheroids (*n* = 3). One-way ANOVA was used to compare treatments to the untreated experiments, where significance was considered at *p* < 0.05. No asterisks \* are shown since all treatments differed significantly from the untreated experiments.

Notably, the fucoidan extracts showed potency in inhibiting HCT116 cell migration during time- and dose-dependent experiments. The anionic nature, which is the common characteristic of fucoidans, could be critical in disrupting the migration of HCT116 cells. Limited reports in the literature have investigated the effects of fucoidans on spheroidbased migration. However, a study by Han and colleagues showed that tumour migration of a human colon cancer cell line (HT-29) was inhibited by fucoidan [17]. Indeed, very few investigations on the potency of chemotherapeutics on spheroid migration have been reported [47]. In addition, spheroid culture systems provide similar physicochemical environments to in vivo models, making them ideal for studying tumour migration however, their use in fucoidan studies is seldom reported. The fucoidans in the current study demonstrated their high potency in inhibiting 3D HCT116 migration from spheroids, which may be important in controlling the proliferation of colorectal cancers. Another merit of employing spheroid culture systems is that they involve cell-to-cell and cell-to-matrix interactions, which overcomes the limitations of traditional monolayer cell cultures, which are two-dimensional (2D) [47,48]. Fucoidans maybe be interfering with cell-to-matrix adhesion or even with cell-to-cell interactions.

#### 2.4.4. Fucoidans Disrupt Cancer Cell Sphere Formation

Next, the HCT116 cells were pretreated with 0.1 mg/mL and 0.5 mg/mL of fucoidans to determine whether fucoidans inhibit sphere formation. Representative morphological data of the HCT116 cell spheres pretreated with 0.5 mg/mL of *F. vesiculosus* illustrated a common observation for all fucoidans tested (Figure 8). Fucoidan treatment disrupted the formation of spheres compared to those from untreated samples (Figure 8).

**Figure 8.** A representative visual illustration of HCT116 spheroid pretreated with fucoidan. The images show spheroids before transfer to fresh medium, at t = 0, and after 24 h (t = 24). (**a**) representative sphere formed from an untreated HCT116 culture; (**b**) sphere formed from HCT116 cells pretreated with *F. vesiculosus* fucoidan at 0.5 mg/mL final concentration.

The HCT116 sphere sizes formed after pre-treatment with fucoidans were quantified (Figure 9a). All the fucoidans significantly reduced the size of spheroids formed compared to the untreated sample (Figure 9a; *p* < 0.05).

**Figure 9.** Fucoidans hinder HCT116 spheroid formation and reduce migration from spheres. (**a**) Size of HCT116 spheroids; (**b**) Distance migrated on tissue culture plastic from pretreated spheroids. The data are represented as means ± SD of biological replicates of spheroids sizes and migration (*n* = 3). The asterisk \* represents treatment concentrations that were statistically significant from the untreated cells at *p* < 0.05 tested using One-way ANOVA.

The pretreated spheroids were subsequently transferred to an untreated medium to investigate the migration of cells from the spheres back onto tissue culture plastic (Figure 9b). Interestingly, all the spheres pretreated with fucoidan showed reduced migration compared

to untreated spheroids (Figure 9b; *p* < 0.05). Therefore, pretreatment of the HCT116 cell culture indicated that fucoidans hindered spheroid formation and subsequent migration onto the tissue culture plastic matrix. In addition, the spheres which were pretreated with *F. vesiculosus* fucoidan were distorted and failed to migrate. Although investigations on spheroid formation are largely unexplored as far as the use of fucoidans is concerned, Han and colleagues reported that *F. vesiculosus* fucoidan disrupted HT-29 spheroid formation [19]. Their findings concur with our sphere formation results (Figure 8). Although this technique is a useful tool, it is limited to very few in vitro studies. However, our findings can be used as a motivation to further pursue the potential of fucoidans in in vivo and clinical settings.

#### 2.4.5. Fucoidans Inhibit HCT116 Cell Adhesion

The effect of fucoidan extracts on HCT116 cell adhesion was also investigated. The fucoidans significantly prevented the adhesion of HCT116 cells to tissue culture plastic (Figure 10).

**Figure 10.** Fucoidan inhibits the adhesion of HCT116 cancer cells. (**a**) untreated cells; (**b**) cells treated with fucoidan under light microscopy; (**c**) Quantification of HCT116 cancer cells adhesion by crystal violet. The data are represented as means ± SD of three biological replicates (*n* = 3). The asterisk \* represents treatment concentrations that were statistically significant from the untreated cells at *p* < 0.05 tested using One-way ANOVA.

The dose-dependent inhibition of HCT116 cancer cell adhesion by fucoidan was quantified by crystal violet (Figure 10c). EDTA-Na, a known chelator of metal ions required for cell adhesion, was used as a positive control. All fucoidans were efficient inhibitors of cell adhesion. Cell adhesion within cancer cells is vital for various biological processes, including cellular organisation, communication, differentiation, migration, and metastasis [49]. The cancer cell adhesion is dependent on several adhesion molecules and receptors, including integrins, selectins, glycoproteins, and proteoglycans [49]. The fucoidans may

have hindered the proper functioning of these molecules, thereby impacting the adhesion of cancerous cells. Some fucoidans prevent the adhesion of cancer cells onto the extracellular matrix (ECM). Fucoidan from *A. nodosum* inhibited the MDA-MB-231 cancerous cells adhering to fibronectin ECM [50], consistent with our findings on tissue culture plastic. Fucoidans are negatively charged polysaccharides due to their sulphated nature, which may interfere with integrins that require Mg2+ as a cofactor for adhesion [49]. Thus, it is possible to suggest our fucoidans inhibited the HCT116 cancer cells' adhesion in a similar mechanism. This observed effect of fucoidans on cell adhesion might also explain the effect of these compounds in inhibiting the formation and migration from spheres (Figures 8 and 9). However, HCT116 cells' adhesion cannot be the only process affected by fucoidans, as *E. maxima* and *S. elegans* extracts show similar anti-adhesion properties but show radically different effects in the colony formation assay (Figure 5). A complex combination of structural characteristics, including the degree of sulphation, molecular size, and carbohydrate content, should be essential to fucoidans' biological activities. The observed anticancer activities of fucoidans may be useful as a preventive/treatment strategy for CRC since they are likely to be administered orally.

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

*Fucus vesiculosus* fucoidan (Cat. No. F5631) was purchased from Sigma-Aldrich (St. Louis, MO, USA). The analytical kits used in this study were purchased from MegazymeTM (Bray, WC, Ireland). The other reagents were purchased from Sigma-Aldrich, MERCK, Flucka Saarchem (Darmstadt, HE, Germany), and Celtic Diagnostic and Life Technologies (Cape Town, South Africa).

#### *3.1. Sampling and Seaweed Processing*

The brown seaweeds, *Ecklonia radiata* and *Sargassum elegans,* were harvested between February and March 2019 from Kelly's beach in Port Alfred (coordinates 33◦36 36.8424" S; 26◦53 23.4996" E) in the Eastern Cape province, South Africa. *Ecklonia maxima* seaweed was kindly donated by the HIK-Abalone farm located in Hermanus, Western Cape province, South Africa. Most of the *E. radiata* seaweed was collected as beach cast. However, some were harvested together with the *S. elegans* from rock pools. The beach cast and rockpool collected *E. radiata* were mixed and processed as a single batch. The harvested seaweeds were stored on ice during transportation to the laboratory. Upon arrival at the laboratory, the seaweed was washed 3× with distilled water, cut into smaller pieces and oven-dried at 40 ◦C for 72 h. The dried seaweed was pulverised using a coffee grinder, and the resulting powder was stored at room temperature until use.

#### *3.2. Hot Water Extraction*

The seaweeds were defatted, and pigments were extracted using a high methanol percentage mixture, with a solvent ratio of 4:2:1 for MeOH: CHCl3: H2O [51,52]. The fucoidans were hot water extracted as described by Lee and co-workers with minor modifications [53]. A mass of 15 g dry defatted seaweed powder was suspended in 450 mL of distilled water in a ratio of 1:30 (*w*/*v*). The mixture was heated to 70 ◦C with agitation overnight. The extracted fucoidan yield was expressed as a percentage of the dry defatted seaweed weight (% dry wt).

#### *3.3. Structural Validation of Extracted Fucoidans*

#### 3.3.1. Fourier Transform Infrared Spectrometry (FTIR) Analysis

A hundred milligrams of ground fucoidan was scanned using Fourier-transform infrared spectroscopy (FTIR) on a 100 FT-IR spectrometer system (Perkin Elmer, Wellesley, MA, USA). The signals were automatically recorded by averaging four scans over 4000–650 cm<sup>−</sup>1. The baseline and ATR corrections for penetration depth and frequency variations were performed using Spectrum One software (version 1.2.1) (Perkin Elmer, Wellesley, MA, USA).

#### 3.3.2. NMR Spectroscopy Analysis

Fucoidan samples (10 mg) were dissolved in 1 mL D2O. After centrifugation at 13,000× *g* for 2 min, any insoluble matter was removed by filtering the supernatant through a 0.45-μm filter. The deuterium-exchanged samples were subjected to 1H-NMR analysis, and spectra were recorded at 23 ◦C using a 400 MHz spectrometer (Bruker, Fällanden, Switzerland) with Topspin 3.5 software (Bruker, Billerica, MA, USA).

#### 3.3.3. Thermogravimetric Analysis

Fucoidans were subjected to thermogravimetric analysis using a Pyris Diamond model thermogravimetric analyser (PerkinElmer®, Shelton, CT, USA). Samples of 4 mg fucoidan were analysed in an aluminium crucible. Pure nitrogen (purity of 99.99%) was used as the carrier gas during all the experiments to reduce the mass transfer effect. The gas flow rate was at 20 mL/min. The fucoidans were heated from 30 ◦C to 900 ◦C at a heating rate of 30 ◦C/min. A separate blank using an empty tray was run for baseline correction. Lastly, the mass loss relative to the temperature increment was automatically recorded, and the derivative thermogram (DTG) was plotted using GraphPad Prism version 6.

#### *3.4. Chemical Characterisation of Fucoidans*

Using *L*-fucose as a standard, the phenol-sulphuric acid method estimated the total sugar content within the fucoidans [54]. The total reducing sugar content in 2 M TFA partially hydrolysed fucoidans was quantified using the dinitrosalicylic acid (DNS) assay [55]. Furthermore, the protein contamination was measured using Bradford's method, utilising bovine serum albumin (BSA) as a standard [56]. The sulphate content in formic acid (60% *v*/*v*) desulphurised fucoidan was measured using a barium chloride–gelatin method as described previously [57], which was scaled down to microtitre volumes.

Polyphenols within the fucoidans were quantified using a modified Folin–Ciocalteu method with gallic acid as a standard [58]. Moreover, quantitative analyses of *L*-fucose, D-fructose, D-galactose, D-xylose, L-arabinose, and D-mannose in the fucoidans were performed using high-performance liquid chromatography (HPLC) method [16]. A Shimadzu HPLC (RID) instrument (Kyoto, Japan) and a Fortis Amino column (Fortis Technologies Ltd., Cheshire, UK) was utilised in the HPLC method. The ash contents in fucoidans were derived from derivative thermogravimetry (DTG) data.

#### *3.5. Determination of Fucoidans Molecular Weights by HPLC*

The molecular weights of fucoidans were determined using size exclusion highperformance liquid chromatography with a refractive index detector (HPLC-RID). The fucoidan extracts were separated using a Shodex OHpak SB-806M HQ (8.0 mm I.D. × 300 mm) column (Showa Denko, Tokyo, Japan) according to the manufacturer's recommendations. The mobile phase (0.1 M NaNO3 aq) used was filtered through 0.22 μm nylon membranes (Membrane solutions, Auburn, USA). The flow rate was adjusted to 0.6 mL/min, the column temperature was at 30 ◦C, and the sample injection volume was 20 μL. Pullulan standards (Shodex, Tokyo, Japan) were used to construct the standard curve for interpolating fucoidan molecular weights.

#### *3.6. Cell Culture*

The HCT116 human colon cancer cell line was from the American Type Culture Collection (ATCC CCL-247). The cell line was cultured in Dulbecco's Modified Eagle's Medium (DMEM) with GlutaMAX™-I, supplemented with 10% (*v*/*v*) fetal bovine serum (FBS) and 1% (*v*/*v*) sodium pyruvate. The cell culture was maintained at 9% CO2 in a humidified incubator at 37 ◦C.

#### *3.7. Cytotoxicity Screening*

The susceptibility of the HCT116 cell line to the fucoidan extracts was determined using an optimised resazurin assay [16]. Briefly, cells were seeded at a density of

<sup>1</sup> × <sup>10</sup><sup>5</sup> cells/well in DMEM growth medium in a 96-well plate. After the cells adhered to the plate matrix overnight, they were treated with varying doses of fucoidan (0.1 mg/mL to 2.5 mg/mL). The anti-cancer agent 5-fluorouracil (5-FU) in a concentration range of 0.0064 μM to 2500 μM was included as a positive control for cytotoxicity. The experiments were incubated for 72 h before treatment with resazurin. Cell viability was measured by fluorescence (excitation = 560 nm and emission = 590 nm).

#### *3.8. Clonogenic Assay*

HCT116 cells were seeded at a density of 1.5 × 103 cells/mL in a six-well plate and allowed to adhere overnight. The cells were treated with fucoidan extracts or 4 nitroquinoline 1-oxide (4NQO), which was used as a positive control. The cultures were incubated at 37 ◦C for 48 h, upon which half the volume of spent medium was removed and replaced with fresh medium lacking treatment. The cultures were incubated, and the medium changed every two days until individual colonies of at least 50 cells/colony were visible. The medium was removed, and the cells were washed once with PBS. The cells were fixed for 10 min by a 3:1 methanol to the acetic acid mixture. The fixative was removed, and the plate was allowed to air dry for 2 min. The HCT116 cell colonies were stained with 5% (*w*/*v*) crystal violet in methanol for 4–6 h, washed three times in PBS and rinsed in water. The plates were air-dried, and the images were captured using a ChemiDoc™-XRS (BioRad, Hercules, CA, USA). The cells were solubilised completely using 1 M acetic acid, and the absorbance was read at a wavelength of 590 nm. The % colony formation was expressed as percentiles relative to the untreated experiments.

#### *3.9. Wound Healing Assay*

A volume of 500 <sup>μ</sup>L/well of HCT116 cells were seeded at 7 × 105 cells/mL into 24 well plates. The cells were allowed to adhere and grow to 100% confluence overnight at 37 ◦C. A wound was made down the centre of the well with a pipette tip. After wounding, the floating cells were removed, and fresh medium without or with varying doses of fucoidan (0.1–0.5 mg/mL) treatments was added. Pre-migration images of the wounds were taken at 4× magnification. The plates were further incubated at 37 ◦C for 12 h, whereafter, images were taken at the same position as the premigration images. The images were analysed on ImageJ using a wound healing plugin [59], and wound closure was calculated using the formula below where the percentage wound area was calculated relative to the wound size at t = 0:

Wound closure = %wound area(t = 0) − %wound area (t = 12)

#### *3.10. Sphere-Based Tumour Migration Assays*

HCT116 cells were resuspended in an appropriate volume of Dulbecco's Modified Eagle's Medium (DMEM) (final concentration of 1 × 104 cells/10 <sup>μ</sup>L) for the formation of spheres in an optimised hanging drop method [60]. About 5 mL of sterile PBS was pipetted into the bottom portion of the tissue culture plate (100 mm diameter) to create a humidified environment., Multiple 10 μL culture drops were deposited inside the lid of the culture dish. The lid with the hanging drops was placed back on top of the PBS-containing dish, taking care to avoid disturbing the droplets. The plate was incubated for 48 h at 37 ◦C to allow the spheres to grow. The spheres were then transferred to a 24-well plate prefilled with 300 μL medium and respective treatments with compounds ranging from 0.1 mg/mL to 0.5 mg/mL. In some experiments, the HCT116 culture (at a density of 10,000 cells/10 μL) was pretreated with 0.5 mg/mL and 0.1 mg/mL of the compounds during sphere formation. Untreated or pretreated spheres transferred to the adherent plate were allowed to adhere by incubation at 37 ◦C for 4 h. Images were taken at 4× magnification and this time was taken as t = 0 (4 h post-seeding of spheres into adherent plates). The spheroids were monitored over time for t = 24 h. The areas of migration were quantified using Fiji/ImageJ (Version 1.53f51). The results presented were represented by three experimental biological replicates. The data were normalised to the initial size of each spheroid at time 0 to determine cell

migration from the spheroid. The migration of cells from spheres was calculated as follows:

Distance migrated = area measured at t = 24 hrs − area at t = 0 h

#### *3.11. Cell Adhesion Assays*

HCT116 cells were seeded at a density of 6 × <sup>10</sup><sup>4</sup> cells/well in a 96-well plate and treated with varying concentrations of fucoidan (0.1–0.4 mg/mL) or left untreated as the control. The culture was incubated at 37 ◦C with 9% CO2 for 8 h. The spent medium was decanted from the plate, and the adhered cells were washed thrice with 1 × sterile PBS (137 mM NaCl, 2.7 mM KCl, 10 mM Na2HPO4 and 2 mM KH2PO4, pH = 7). The cells were fixed by adding 100 μL/well of a 3:1 methanol: acetic acid solution and incubated at room temperature for 15 min. The methanol-acetic acid mixture was washed off using sterile distilled H2O and blotted dry on a paper towel. The cells were stained with 0.1% (*w*/*v*) crystal violet dye (40 μL/well) and incubated at room temperature for 20 min with gentle agitation at 30 rpm. The crystal violet dye was discarded, and the plate was washed four times with distilled water. A volume of 100 μL/well of 1% (*w*/*v*) SDS was added, the plate incubated overnight, and the optical density was then measured at a wavelength of 590 nm.

#### **4. Conclusions**

The extracted compounds were unique in composition, with *Ecklonia* sp. fucoidans having a relatively high carbohydrate content compared to *S. elegans* and commercially purchased *F. vesiculosus*. Furthermore, the *Ecklonia* sp. fucoidans had a comparatively low degree of sulphation compared to the other fucoidans, despite having comparatively lower molecular weights. Their low molecular weight could have had an impact on the anti-adhesion and anti-spheroid migration of the HCT116 cancer cells. Although slightly larger, the *S. elegans* and *F. vesiculosus* fucoidans had a relatively higher sulphate content than the *Ecklonia* sp. fucoidans, which may have enhanced the anti-cancer activities of these fucoidans observed by the anti-colony formation, anti-adhesion and anti-spheroid migration of the HCT116 cells. Our study reaffirms that molecular weight and sulphation of fucoidan, along with other properties, may be important for biological activity. This study also showed that fucoidans differ in structure and activity depending on the type (genus and species) of seaweed. Although there are structural differences between the fucoidans studied, their anti-cancer effects suggest some potential health benefits of seaweed fucoidans that warrant further analysis.

**Author Contributions:** Conceptualisation, B.M., C.D.D., A.L.E. and B.I.P.; Investigation, B.M. and C.D.D.; writing—original draft preparation, B.M. and C.D.D.; writing—review and editing, B.M., C.D.D., A.L.E. and B.I.P.; supervision, A.L.E. and B.I.P.; project administration, A.L.E. and B.I.P.; funding acquisition, B.I.P. and A.L.E. All authors have read and agreed to the published version of the manuscript.

**Funding:** B.M. was funded by the German Academic Exchange Service (DAAD) In-Region Scholarship (grant no. 57408782). C.D.D. received financial support for this study from the Pearson Young Memorial scholarship. A.L.E is supported by the National Research Foundation of South Africa (Grant Numbers 98566 and 105829), and both A.L.E. and B.I.P. are supported by Rhodes University (RRG). This research was supported in part by KelpX.

**Institutional Review Board Statement:** Not applicable.

**Data Availability Statement:** Data are available upon request.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


**Disclaimer/Publisher's Note:** The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

### *Article* **Chitooligosaccharides Derivatives Protect ARPE-19 Cells against Acrolein-Induced Oxidative Injury**

**Cheng Yang 1, Rongrong Yang 1, Ming Gu 1, Jiejie Hao 1,2, Shixin Wang 1,3,\* and Chunxia Li 1,2,3,\***


**Abstract:** Age-related macular degeneration (AMD) is the leading cause of vision loss among the elderly. The progression of AMD is closely related to oxidative stress in the retinal pigment epithelium (RPE). Here, a series of chitosan oligosaccharides (COSs) and *N*-acetylated derivatives (NACOSs) were prepared, and their protective effects on an acrolein-induced oxidative stress model of ARPE-19 were explored using the MTT assay. The results showed that COSs and NACOs alleviated APRE-19 cell damage induced by acrolein in a concentration-dependent manner. Among these, chitopentaose (COS–5) and its *N*-acetylated derivative (N–5) showed the best protective activity. Pretreatment with COS–5 or N–5 could reduce intracellular and mitochondrial reactive oxygen species (ROS) production induced by acrolein, increase mitochondrial membrane potential, GSH level, and the enzymatic activity of SOD and GSH-Px. Further study indicated that N–5 increased the level of nuclear Nrf2 and the expression of downstream antioxidant enzymes. This study revealed that COSs and NACOSs reduced the degeneration and apoptosis of retinal pigment epithelial cells by enhancing antioxidant capacity, suggesting that they have the potential to be developed into novel protective agents for AMD treatment and prevention.

**Citation:** Yang, C.; Yang, R.; Gu, M.; Hao, J.; Wang, S.; Li, C. Chitooligosaccharides Derivatives Protect ARPE-19 Cells against Acrolein-Induced Oxidative Injury. *Mar. Drugs* **2023**, *21*, 137. https:// doi.org/10.3390/md21030137

Academic Editors: Yuya Kumagai, Hideki Kishimura and Benwei Zhu

Received: 30 January 2023 Revised: 13 February 2023 Accepted: 20 February 2023 Published: 22 February 2023

**Copyright:** © 2023 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

**Keywords:** chitosan oligosaccharide; *N*-acetylated chitosan oligosaccharide; ARPE-19; oxidative stress; Nrf2

#### **1. Introduction**

Age-related macular degeneration (AMD) is one of the main causes of irreversible central visual loss in the elderly worldwide [1,2]. In 2020, the number of AMD patients worldwide was 19.6 million, and this number was predicted to be 288 million in 2040 [1,3,4]. According to the symptoms, AMD can be divided into dry and wet forms with 80% and 20% prevalence, respectively [5]. Current treatments for wet AMD include laser therapy and VEGF antibody injection (such as Eylea) [6], while there is no preventive therapy for dry AMD [7]. Therefore, it is necessary to develop effective agents to prevent or cure dry AMD.

Numerous studies have indicated that AMD pathogenesis was related with chronic oxidative stress and the inflammation of retinal pigment epithelial (RPE) cells, which could lead to the eventual degeneration of the RPE [8–11]. Jin et al. [12] suggested that retinal pigment epithelium cell apoptosis was induced by ultraviolet and hydrogen peroxide via AMPK signaling. *Melissa officinalis* L. extracts and resveratrol were reported to improve cell viability and decrease reactive oxygen species (ROS) generation in RPE cells to prevent AMD [10,13]. These results demonstrated that the inhibition of RPE cell damage induced by ROS could prevent the process of AMD [14]. Therefore, antioxidation could be an effective strategy to protect RPE cells for the amelioration of early AMD.

Chitin is extracted mainly from the shells of crabs, shrimps and insects and is one of the most abundant natural biopolymers [15–17]. Chitosan oligosaccharides (COSs) are the degraded product of chitin or chitosan and consist of glucosamine linked by β-1,4-glycosidic bonds, possess various biological effects, including anti-inflammatory, antimicrobial, immunomodulatory, antioxidant, and anticancer activities [18–21]. Fang demonstrated that COS attenuated oxidative-stress related retinal degeneration in a dose-dependent manner in a rat model [22]. Xu found that COSs protected against Cu(II)-induced neurotoxicity in primary cortical neurons by interfering with an increase in intracellular reactive oxygen species (ROS) [23]. Our previous study indicated that peracetylated chitosan oligosaccharide (PACOs) pretreatment significantly reduced lactate dehydrogenase release and reactive oxygen species production in PC12 cells [24]. In addition, Guo's group indicated that the antioxidant properties of chitosan were inversely related to its molecular weight (Mw) [25]. We performed a preliminary screening of structurally related compounds, and COSs and NACOs showed excellent antioxidant activity with the potential to prevent AMD. In the present study, we investigated the effect of a series of chitosan oligosaccharides and their *N*-acetylated derivatives on RPE cell damage and explored the possible mechanisms of action. The results showed that chitosan oligosaccharides had an excellent capacity for protecting RPE cells from acrolein-induced oxidative stress.

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

#### *2.1. Characterization of Chitooligosaccharides and N-Acetylated Chitooligosaccharides*

According to the previous method, chitooligosaccharides (COSs) and *N*-acetylated chitooligosaccharides (NACOs) were prepared via enzymatic hydrolysis [24] and acetylated modification [26].

The crude products were isolated and purified to provide monomers with different degree of polymerization (Figure 1A). NACOs were purified by column chromatography using graphitized carbon black as the stationary phase and ethanol–water as the mobile phases. This purification method was simple and efficient with the elimination of the tedious operation process of desalting, compared with gel exclusion and ion exchange purification methods [27]. The purity of these compounds was analyzed via HPLC (LC-10AD, Shimadzu, Kyoto, Japan) [28,29]. As shown in Figure 1B, the purity of COS (COS–2~6) and NACO (N–2~6) monomers was above 95%.

The structures of COSs and NACOSs were characterized using a quadrupole time of flight (Q-TOF) mass spectrometer, and nuclear magnetic resonance (NMR) and Fouriertransform infrared spectroscopy (FT-IR) analysis (Figure 1C–E). The Q-TOF MS analysis (positive ion mode) of COSs and NACOs samples are shown in Figure 1C and Table 1.

For the FT-IR spectra (Figure 1D), the bands at 3370 cm<sup>−</sup>1, 2876 cm−1, and 1073 cm−<sup>1</sup> were corresponded to stretching vibrations of O-H, C-H and C-O, respectively. The spectra of *N*-acetyl chitosan oligosaccharides showed the characteristic absorptions of 1649 cm<sup>−</sup>1, 1549 cm−<sup>1</sup> and 1314 cm−1, which were attributed to amide I, II and III bands of amide, respectively [30]. Moreover, there was no 1735 cm−<sup>1</sup> band (-C(=O)O-) in the *N*-acetylated chitosan oligosaccharide, indicating no acetylation on the OH groups of COSs.

The structures of COSs and NACOSs were also characterized via NMR. Taking COS–3 and N–3 as examples, the 13C NMR signals in spectra (Figure 1E) were assigned in Table 2. Compared to COS–3, acetyl signal peaks appeared in the 13C NMR spectrum of N–3, with 174.8 ppm attributed to C=O, and peaks at 22.6~22.3 ppm attributed to CH3 of acetyls.

**Figure 1.** Characterization of COSs and NACOs. (**A**) Schematic structures of COSs (**a**) and NACOs (**b**). (**B**) HPLC chromatograms of COSs (**a**) and NACOs (**b**). (**C**) MS spectra of COSs (**a**) and NACOs (**b**). (**D**) IR spectra of COSs and NACOs. (**E**) 13C NMR (125 MHz, D2O) of COS–3 and N–3.


**Table 1.** The MS data of COS and NACO samples.

**Table 2.** The 13C NMR date of COS–3 and N–3.


#### *2.2. Protective Effect of COSs and NACOs against Acrolein-Induced Cell Death*

The cytotoxicity of COS and NACO monomers (DP 2, 3, 4, 5, 6) was tested via MTT assay in ARPE-19 cells. After 24 h of incubation with 1 mM COSs or NACOs, the MTT test showed that both COSs and NACOs exhibited no significant cytotoxicity (Figure S1A). In addition, the effect of COSs and NACOs on the viability of ARPE-19 cells was tested with different concentrations (200, 400, 800 μM). It showed that COSs and NACOs did not affect cell proliferation. (Figure S2).

Acrolein, a major component of the gas phase of cigarette smoke and also a product of lipid peroxidation in vivo, has been shown to be a mitochondrial toxicant related to mitochondrial dysfunction [31]. Therefore, acrolein-induced cellular oxidative mitochondrial dysfunction in retinal pigment epithelial (RPE) cells had been used as a cellular model to evaluate antioxidants and mitochondrial protecting agents [32–34]. Here, AREP-19 cells were pretreated with different concentrations of COSs or NACOs (200, 400 and 800 μM) for 48 h, and then treated with 75 μM acrolein for 24 h, and cell viability was measured using the MTT test.

As shown in Figure 2, the cells exposed to 75 μM acrolein showed a significant decrease in cell viability (about 50%) compared to the untreated control group. However, after pretreatment with COSs or NACOs at 200, 400, and 800 μM for 48 h before acrolein exposure, the cell viability increased significantly. Furthermore, COSs and NACOs exhibited similar protective activity which was dose-dependent. In addition, we also prepared peracetylated chitosan oligosaccharides (PACOs) [24], but PACOs had a certain cytotoxicity to ARPE-19 (Figure S1B). Glucosamine pentamer (COS–5) and *N*-acetylated chitopentaose (N–5) showed the highest protective activity, which indicated that the pentaose skeleton may be the suitable structure for binding to the receptors or targets, such as heparin core pentasaccharide, for anticoagulant activity [35].

**Figure 2.** Protective effect of COSs and NACOs against acrolein-induced ARPE-19 cell death. The cells were pretreated with 200, 400, 800 μM COSs (**A**) or NACOs (**B**) for 48 h and then treated with 75 μM acrolein for an additional 24 h. Cell viability was analyzed using the MTT method. Values are mean <sup>±</sup> SD of five separate experiments. ## *<sup>p</sup>* < 0.01 vs. control (no acrolein, no COSs or NACOs); \* *p* < 0.05, \*\* *p* < 0.01 vs. acrolein.

#### *2.3. Protective Effect of COS–5 and N–5 against Acrolein-Induced Oxidative Stress*

The involvement of oxidative-stress-triggered apoptosis in retinal endothelial cells was considered as the leading cause of AMD [2,36,37]. In this study, RPE cells were stimulated by acrolein to induce oxidative stress. It was evaluated for the capacity of COS–5 and N–5 to prevent oxidative-stress-induced cell death and the imbalance of the antioxidant system. Initially, the effects of COS–5 and N–5 on acrolein-induced ROS generation (Figure S3) and MMP decline (Figure S4) in ARPE-19 cells were evaluated at different concentrations (200, 400, 800 μM). The results showed that there were no significant differences between 400 and 800 μM. Thus, all subsequent experiments were performed with the 400 μM dose. Then, intracellular and mitochondria ROS accumulation, GSH level, and GPx and SOD activities were measured (Figure 3).

ROS are natural by-products of aerobic respiration. ROS can be controlled by various cellular antioxidant compounds and enzymes, and their overproduction would lead to cell death [38]. Compared with the control, intracellular and mitochondria ROS levels were significantly increased to about 270% and 276% after acrolein exposure, respectively (Figure 3A,B). However, pretreatment with COS–5 or N–5 at the concentration of 400 μM reduced acrolein-induced ROS production significantly. GSH is one of the most important endogenous small molecule antioxidants. As shown in Figure 3C, the intracellular GSH level was decreased significantly after acrolein exposure (about 49%). Pretreatment with COS–5 or N–5 could successfully inhibit the decrease in GSH content induced by acrolein, which increased by 39% and 41% (*p* < 0.01), respectively. GSH peroxidase (GPx) and SOD activity was decreased to 30% and 55% after acrolein treatment (Figure 3D,E), respectively. The activities of antioxidant enzymes (GPx and SOD) significantly enhanced after COS–5 or N–5 treatments. These results indicated that the excellent antioxidant activity of COSs and NACOs played a critical role in protecting cells against acrolein-induced oxidative damage.

**Figure 3.** COS–5 and N–5 against acrolein-induced oxidative stress. ARPE-19 cells were treated with 400 μM COS–5 or N–5 for 48 h and then treated with acrolein for an additional 24 h. Cellular ROS generation (**A**), ROS level in mitochondria (**B**), GSH level (**C**), GPx (**D**) and SOD activities (**E**). The data expressed as ratio relative to controls. Values are mean ± SD of five separate experiments. ## *p* < 0.01 vs. control (no acrolein, no COS–5 and N–5); \*\* *p* < 0.01 vs. acrolein.

In this study, we found that chitooligosaccharides and their derivatives could protect APRE-19 cells from acrolein oxidative damage by improving their antioxidant capacities. However, without acrolein exposure, COS–5 or N–5 pretreatment did not affect these antioxidant biomarkers when compared to control cells (Figure 3). This is an interesting phenomenon. ROS at a low level play important roles as signaling molecules in normal physiology. Navdeep et al. [39] found that the mitochondrial complex III ROS was essential for T cell activation both in vitro and in vivo. It is a huge advantage that the antioxidant activities of N–5 and COS–5 were selective, and they did not affect ROS balance and ROSmediated signaling pathways in normal cells. The data above showed that COS–5 or N–5

has the potential to be studied further and developed into a novel therapeutic agent for the treatment of AMD.

#### *2.4. COS–5 and N–5 Improved Mitochondrial Function in Acrolein-Treated ARPE-19 Cells*

Mitochondria are the main sites of oxidant generation, and are easily affected by oxidants, resulting in mitochondrial dysfunction and apoptosis. We examined mitochondrial function by assaying cellular and mitochondrial ROS production, and mitochondrial membrane potential MMP. The results of cellular (Figure 3A) and mitochondrial ROS production is shown in Figure 3B. MMP is an important index of mitochondrial function, which could be evaluated using a JC-1 fluorescent probe. As shown in Figure 4, MMP was decreased to about 45% by acrolein (75 μM, 24 h), which was consistent with previous reported results [32]. MMP was significantly increased after pretreatment with COS–5 or N–5. Similarly, COS–5 or N–5 did not affect mitochondrial function of normal ARPE-19 cells.

**Figure 4.** Protective effect of COS–5 and N–5 against acrolein-induced ARPE-19 mitochondrial dysfunction. The cells were pretreated with 400 μM COS–5 or N–5 for 48 h and then treated with 75 μM acrolein for an additional 24 h. The effects of COS–5 or N–5 on mitochondrial membrane potential were tested using the JC-1 method. Data are red/green (590/530 nm) fluorescence ratios. The data are expressed as ratio relative to controls. Values are mean ± SD of five separate experiments. ## *p* < 0.01 vs. control (no acrolein, no COS–5 and N–5); \*\* *p* < 0.01 vs. acrolein.

Zhou [40] found that COS could entered into cells in a dose-dependent and timedependent manner, and COS was localized preferentially in the mitochondria. However, it was not reported whether NACOSs could enter into cells. Here, the localization of N–5 in ARPE-19 cells was detected by confocal microscopy using the FITC-labeled N–5 (N5- FITC). After treatment with N5-FITC (100 μM) for 3 h, a green fluorescence was observed around the mitochondria, while nearly no fluorescence was found in control cells (Figure 5), suggesting that N–5 could enter into ARPE-19 cells and localize in the mitochondria. These data indicated that the intracellular localization of chitooligosaccharides was not affected by the introduction of acetyl group into amino. Taken together, the results demonstrated that N–5 could localize in the mitochondria and protect ARPE-19 cells against mitochondrial dysfunction and apoptosis induced by oxidative stress.

#### *2.5. N–5 Promoted Nrf2 Nuclear Translocation and Increased Antioxidant Enzyme Expression*

Nuclear transcription factor Nrf2 plays a key role in regulating the expression of phase II detoxification enzymes and antioxidant enzymes. Under normal physiological conditions, Nrf2 was present in the cytoplasm coupled with the negative regulatory protein Kelch Ech-associated protein 1 (Keap1), which interacted with Nrf2 and acted as an adaptor protein, maintaining Nrf2 at a low level and allowing it to be continuously degraded by the proteasome in a ubiquitin-mediated process [41]. When cells were exposed to oxidative stress, Nrf2 in the cytoplasm was released from the negative regulatory protein Keap-1 and translocated to the nucleus, then bonded to an antioxidant response element (ARE). Then, a variety of genes, including glutathione reductase (GR), heme oxygenase-1 (HO-1), catalase (CAT), NAD(P)H Quinone oxidoreductase-1 (NQO-1), and γ-glutamyl cysteine ligase (GCL) were regulated to resist the cell damage caused by oxidative stress [42].

**Figure 5.** The intracellular localization of N–5 in ARPE-19 cells. The cells are stained with FITC and MitoTracker Red CMXRos. Red: MitoTracker Red CMXRos (**a**,**d**), Green: FITC (**b**,**e**), Merge images (**c**,**f**). Images were captured with confocal microscope. Scale bar: 8 μm.

We determined the effect of oxidative stress induced by acrolein on Nrf2 nuclear translocation in the ARPE-19 cell. Due to the similar activity of N–5 and COS–5, as well as the easy preparation of N–5, we focused on N–5 in subsequent experiments. As Figure 6A,B shows, the level of Nrf2 protein in the nucleus significantly decreased after acrolein damage, similar to a published report [43], while pretreatment with N–5 significantly increased the level of nuclear Nrf2, indicating that N–5 could promote Nrf2 nuclear translocation.

Meanwhile, we detected the effect of N–5 on the transcription of genes downstream of Nrf2. The mRNA expression of HO-1 and NQO-1 were performed via qRT-PCR. As shown in Figure 6C,D, the mRNA levels of HO-1 and NQO-1 were significantly reduced in ARPE-19 cells treated with acrolein, and upregulated significantly when pretreated with N–5. These results suggested that N–5 could activate the Nrf2-ARE pathway in ARPE-19 cells, enhance Nrf2 protein nuclear translocation and upregulate the expression of phase II metabolizing enzymes (such as HO-1 and NQO1) to alleviate acrolein-induced oxidative injury.

Oxidative damage of RPE cells was a major factor in the pathogenesis of AMD, and protecting RPE from oxidative damage and death has become a trend in the treatment and prevention of AMD disease. COS and their derivatives were well-known for their free radical scavenging potential by interrupting radical chain reactions to inhibit oxidative damage [44]. The antioxidant activity of chitosan increased with decreasing Mw [45]. Li et al. [46] reported that COS had strong antioxidant activities such as hydroxyl and superoxide radical scavenging activity and reducing power. Qu [47] found that chitooligosaccharides had a certain radical scavenging activity in vitro, and they protected mice from oxidative stress, increased the activity of SOD, catalase, and GPx significantly in mice on a high-fat diet. However, there are fewer reports on *N*-acetylated oligochitosan

with the same repeated unit as chitin. Several high-purity chitosan oligosaccharides and their *N*-acetylated derivatives were prepared in this study, and their protective effect on retinal pigment epithelial cells was studied. Similar to other antioxidants such as curcumin analogs [32], luteolin [48], naringenin [49], or tocopherol [31], chitooligosaccharide monomers also had good protective activity, and COS–5 and N–5 showed the best activities.

**Figure 6.** N–5 promoted Nrf2 nuclear translocation and upregulated antioxidant enzyme expressions in the ARPE-19 cell model of acrolein damage. The cells were pretreated with 400 μM N–5 for 48 h and then treated with 75 μM acrolein for an additional 24 h, and mRNA and protein levels were analyzed. Western blot image of nuclear Nrf2 (**A**) and quantification of Western blots (**B**), mRNA expression of HO-1 (**C**) and NQO1 (**D**). The data expressed as ratio relative to controls. Values are mean <sup>±</sup> SD of three separate experiments. ## *<sup>p</sup>* < 0.01 vs. control (no acrolein, no N–5); \*\* *<sup>p</sup>* < 0.01 vs. acrolein.

Further studies found that acetyl group introduction did not affect the protective effect of chitooligosaccharides. Subsequent study found that N–5 could enhance the antioxidant capacity of ARPE-19 cells, via reducing ROS production, increasing the GSH level, and enhancing SOD/GPx enzyme activity. In addition, N–5 could localize in mitochondria, increase MMP, reduce mitochondrial dysfunction and cellular damage, and enhance Nrf2 nuclear translocation and the transcription of downstream antioxidant enzyme (HO-1 and NQO1). Interestingly, N–5-mediated antioxidant properties were selective and associated with the oxidative stress state. N–5 does not inhibit ROS production and ROS-mediated signaling pathways in the normal cells. The above results indicated that *N*-acetylated chitooligosaccharides may have a potential application in anti-AMD degenerative diseases.

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

#### *3.1. Materials*

Chitosan (deacetylation > 95%) was purchased from Jinhu Crust Product Corp (zi bo, Shandong, China). Chitosanase fermented by Renibacter ium sp.QD1 was obtained from the Ocean University of China. Acrolein was purchased from Xiya Reagent (Chengdu, China). MitoTracker Red CM-H2Xros and Trizol Reagent were purchased from Invitrogen (Foster City, CA, USA). PrimeScript RT-PCR Kit was purchased from TaKaRa (Dalian, China). The reduced glutathione (GSH) assay kit was purchased from the Nanjing Jiancheng Bioengineering Institute (Nanjing, China). The MTT cell proliferation and cytotoxicity detection kits, phenyl methane sulfonyl fluoride (PMSF), reactive oxygen species (ROS) detection kit, mitochondrial membrane potential (MMP) detection kit, BCA protein assay kit, CuZn/Mn-SOD assay kit (WST-8), cellular glutathione peroxidase (GPx) assay kit, Nuclear and Cytoplasmic Protein Extraction kit, PVDF membranes, and BCIP/NBT Alkaline Phosphatase Color Development kit were purchased from the Beyotime Institute of Biotechnology (Shanghai, China). Nrf2 XP Rabbit mAb and Histone H3 XP Rabbit mAb were purchased from Cell signaling technology (Boston, MA, USA). All other reagents were obtained from Sigma-Aldrich (Saint Louis, MO, USA), unless otherwise stated.

#### *3.2. Chitosan Oligosaccharide (COSs) Preparation and Purification*

The COSs were prepared via the enzymatic hydrolysis of chitosan and purified with gel filtration chromatography according to a previously reported method [24]. In brief, chitosan (10 g) was added to 80 mL of distilled water, then 1.5 mL of chitosanase solution (10 U/mL) was added. The mixture was stirred at 50 ◦C for 24 h, and the pH of the reaction mixture was adjusted to 5~6 with HCl solution (4 mol/L) during the hydrolysis process. The hydrolysate was adjusted to pH 8~9 with NaOH solution (1 mol/L) and filtered to remove insoluble parts. The filtrate was concentrated and precipitated by adding a fourfold volume of ethanol at 4 ◦C overnight. The precipitate was collected via centrifugation for 15 min at 8000 rpm, and then lyophilized to yield powdered products, and identified as a COS mixture.

The COS mixture (200 mg) was dissolved in 2 mL of 0.1 M NH4HCO3, and then filtered with a microporous membrane (0.22 μm) to obtain a clear solution. The filtrate was loaded on a Bio Gel P6 column (2.6 × 110 cm) that was connected to an AKTA UPC100 purification system (GE Healthcare, Fairfield, CT, USA) equipped with an online refractive index detector. The column was eluted with 0.1 M NH4HCO3 solution at a flow rate 0.5 mL/min. Eluents (8 mL/tube) were collected using a fraction collector to afford the pure dimers, trimers, tetramers, pentamers, and hexamers of the COSs. The COSs were analyzed using the high-performance liquid chromatography (HPLC), mass spectra, nuclear magnetic resonance (NMR) and Fourier-transform infrared spectroscopy (FT-IR) methods [24].

#### *3.3. N-Acetylated Chitooligosaccharide (NACOs) Preparation and Purification*

The NACOs were prepared via the acetylation of COSs according to a previously reported method [26]. Briefly, the dried COS mixture (1 g) and NaHCO3 (756 mg) were added to methanol–water solution (8:1; *v*/*v*, 35 mL) with stirring, and 5 mL of acetic anhydride was added dropwise at 0 ◦C with stirring. After stirring for 4 h at room temperature, the NACO mixture solution was filtered to remove insoluble parts and the reaction completion was monitored using TLC (*n*-propanol:water, 2:1, *v*/*v*). The filtrate was concentrated and lyophilized to obtain the NACO mixture powder.

Then, the NACO mixture (500 mg) was dissolved in 2 mL of water, and filtered with a microporous membrane (0.22 μm). The filtrate was loaded on a graphitization of carbon black column (2.6 × 20 cm) that was connected to an AKTA UPC100 purification system (GE Healthcare, Fairfield, CT, USA). After loading the sample, the column was eluted with the following gradient of water and ethanol with a gradient of solvent B (ethanol) as follows: 0% B for 3 CV (column volume), then up to 60% B over 5 CV. Eluents (10 mL/tube) were collected using a fraction collector and monitored using TLC (*n*-propanol:water, 2:1, *v*/*v*). Pure dimers, trimers, tetramers, pentamers, and hexamers of the NACOs were pooled and lyophilized. The NACO samples were identified via HPLC chromatogram, mass spectra, nuclear magnetic resonance (NMR) and Fourier-transform infrared spectroscopy (FT-IR) analysis.

#### *3.4. MTT Assay for Cell Viability*

The ARPE-19 (human retinal pigment epithelial) cell line was purchased from ATCC (CRL2302) and cultured in a DMEM-F12 medium supplemented with 10% fetal bovine serum, 0.348% sodium bicarbonate, 2 mM L-glutamine, 100 μg/mL of streptomycin, and 100 U/mL of penicillin. The cell culture was maintained at 37 ◦C in a humidified atmosphere of 95% air and 5% CO2 [50]. ARPE-19 cells were used within 10 generations, and the medium was changed every two days. COSs and NCOSs were dissolved with PBS buffer, filtered through a sterile 0.22 μm filter, and diluted with complete culture medium to different concentrations for the cell experiments.

The ARPE-19 cells were seeded in 96-well plates at 5 × 104 cells per well and incubated overnight. After incubation with different concentrations of COSs or NCOSs for 48 h, the cells were treated with 75 μM acrolein for 24 h. Cell viability was measured via MTT cell proliferation and a cytotoxicity detection kit (Beyotime). After 4 h of incubation with MTT, the solubilization buffer was added to each well and incubated at 37 ◦C overnight. The optical densities were read at 555 nm using a SpectraMax M5 plate reader (Molecular Devices, Sunnyvale, CA, USA).

#### *3.5. Antioxidant Enzyme Activities, ROS Generation, and Intracellular GSH Levels Assay*

The GSH level, superoxide dismutase (SOD) activity, and GPx activity were determined using commercial assay kits [51]. Briefly, cells were placed in 6-well plates at a density of 5 × <sup>10</sup><sup>5</sup> cells per well. After 12 h, the cells were treated for 48 h with 400 <sup>μ</sup>M of COS–5 or N–5 and then for 24 h with or without 75 μM acrolein. After treatment, the cells were washed twice with PBS, and then the antioxidant enzyme activities and GSH level in the cells were detected.

Moreover, the ROS levels in PRE cells and mitochondria exposed to acrolein were determined using fluorescent probe. In brief, cells were plated in 96-well plates at a density of 2.5 × <sup>10</sup><sup>4</sup> cells per well for 12 h. ARPE-19 cells were treated with 400 <sup>μ</sup>M of COS–5 or N–5 for 48 h, and then incubated with or without 75 μM acrolein for another 24 h. The ROS level in PRE cells was determined by the 2 , 7 -dichlorofluorescein diacetate (DCFH-DA) method using a SpectraMax M5 plate reader (Molecular Devices, San Jose, CA, USA) at a 488 nm excitation wavelength and a 525 nm emission wavelength [52]. The ROS generation in mitochondria was detected using MitoTracker Red CM-H2Xros at a 579 nm excitation wavelength and a 599 nm emission wavelength.

#### *3.6. Confocal Imaging*

ARPE-19 cells were cultured on glass-bottom cell culture dishes at a density of <sup>2</sup> × 104 cells per well for 12 h. The cells were incubated with 25 nM MitoTracker Red CMXROS at 37 ◦C for 30 min. Thereafter, the cells washed three times with PBS to remove unbound probes. Then, the cells were incubated with FITC (100 μM) or FITC-labeled N–5 (100 μM) for 3 h at 37 ◦C. Cellular uptake was terminated by washing the cells three times with PBS. Finally, the cells were observed under a Nikon A1 confocal microscope (Nikon Corporation, Tokyo, Japan). The green fluorescence of FITC was measured at Ex495/Em525, and the red fluorescence of MitoTracker Red CMXRos was measured at Ex578/Em599 [53].

#### *3.7. Mitochondrial Dysfunction Evaluation*

Mitochondrial membrane potential (MMP) was detected in live ARPE-19 cells using a cationic fluorescent indicator JC-1, according to the manufacturer's instructions. Briefly, APRE-19 cells were seeded at a density of 2.5 × 104 cells per well in a 96-well plate. After 12 h, the cells were exposed to 400 μmol/mL of N–5 for 48 h. After treatment with 75 μmol/mL of acrolein for 24 h, the cells were treated with JC-1 for 30 min at 37 ◦C, washed with PBS, and observed under the fluorescence microscope. The Δψm of ARPE-19 cells in each treatment group was calculated as the fluorescence ratio (590 to 530 nm) [54].

#### *3.8. Western Blot*

Western blot was performed as in previously described methods [55] and each Western blot was repeated at least three times. Nuclear proteins were prepared using a Nuclear and Cytoplasmic Protein Extraction Kit, and nuclear Nrf2 was analyzed using Western blot. Briefly, the lysates were homogenized and centrifuged at 13,000 ×g for 15 min at 4 ◦C. The supernatants were collected, and the protein concentrations were determined using the BCA Protein Assay kit. Equal amounts (20 μg) of each protein sample were loaded on 10% SDS-PAGE gels, electrophoresed, transferred to PVDF membranes, and blocked with 5% non-fat milk. The membranes were incubated with anti-Nrf2 (1:1000) and anti-histone H3 (1:1000) at 4 ◦C overnight, and then incubated with anti-mouse secondary antibodies at room temperature for 1 h. Protein bands were visualized using a BCIP/NBT Alkaline Phosphatase Color Development Kit. Signals were quantified using ImageJ software (Version 1.52b, NIH, Baltimore, MD, USA), and defined as the ratio of target protein to histone H3.

#### *3.9. Real-Time PCR*

Real-time PCR was performed using a previously described method [56]. Total RNA was extracted from the cells using Trizol reagent according to the manufacturer's protocol. Reverse transcription was performed using the PrimeScript RT-PCR Kit followed by semiquantitative real-time PCR using specific primers. The primer sequences are listed in Table 3.

**Table 3.** Primer sequences.


#### *3.10. Statistical Analysis*

All quantitative experiments were repeated at least 3 times independently. Data are presented as mean ± SD. Data were analyzed by one-way analysis of variance (ANOVA) with Tukey's multiple comparison post hoc test using GraphPad Prism 8.0 Statistics Software (Graphpad Software, Inc., La Jolla, CA, USA). A *p* value of < 0.05 was considered statistically significant.

#### **4. Conclusions**

In conclusion, our study demonstrated that chitosan oligosaccharides (COSs) and their *N*-acetylated chitooligosaccharides (NACOs) exhibited excellent protection effects on acrolein-induced ARPE-19 cell damage. Among the monomers, COS–5 or N–5 pretreatment significantly reduced reactive oxygen species production, raised the intracellular level of GSH and the activity of SOD and GSH-Px, and attenuated the loss of mitochondrial membrane potential. Further study indicated that the N–5 could localize in the mitochondria and promote Nrf2 nuclear transfer and the expression of downstream phase II detoxification enzymes. These results suggest that COSs and NACOs might be promising antagonists against acrolein-induced APRE-19 cell death.

**Supplementary Materials:** The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/md21030137/s1, Figure S1: The cytotoxicity of COSs, NACOs and PACOs in ARPE-19 cell; Figure S2: Effects of COSs and NACOs on the proliferation of ARPE-19 cells; Figure S3: COS–5 and N–5 against acrolein-induced oxidative stress; Figure S4: Protective effect of COS–5 and N–5 against acrolein-induced ARPE-19 mitochondrial dysfunction.

**Author Contributions:** Conceptualization: C.Y. and C.L.; investigation: C.Y., R.Y. and M.G.; project administration: C.L.; formal analysis: C.Y., R.Y., J.H. and C.L.; data curation: C.Y.; resources: C.Y., R.Y. and M.G.; writing—original draft: C.Y.; writing—review and editing: S.W. and C.L.; supervision: S.W. and C.L.; funding acquisition: S.W. and C.L. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was supported in part by programs of the Shandong Major Science and Technology Project (2021ZDSYS22), National Natural Science Foundation of China (U21A20297), Shandong Provincial Natural Science Foundation (ZR2021QH144), National Science and Technology Major Project for Significant New Drugs Development (2018ZX09735004).

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** Not applicable.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


**Disclaimer/Publisher's Note:** The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.
