Water-soluble COS was obtained by an enzymatic method [9] from low molecular weight chitosan (YB Bio; Gyungbuk, Republic of Korea). The lyophilized COS was dissolved in distilled water immediately before use.

MTT (#M6494) was purchased from ThermoFisher Scientific (Seoul, Republic of Korea). 5-FU (#F6627) was obtained from Sigma-Aldrich (Merck KGaA, Darmstadt, Germany). PD98059 (#9900), an inhibitor of ERK, was purchased from Cell Signaling Technology Inc. (Beverly, MA, USA).

c-Jun N-terminal kinase (JNK; #9252), phosphor-Akt (#3787), phosphor-ERK (#4370), phosphor-JNK (#9251), p38 (#9212), and phosphor-p38 (#9211) antibodies were purchased from Cell Signaling Technology Inc.; Akt (#sc-8312), ERK (#sc-93), GAPDH (#sc-47724), and phosphor-p53 (#sc-135630) antibodies were obtained from Santa Cruz Biotechnology (Santa Cruz, CA, USA); and p53 (#GTX70218) antibody was purchased from GeneTex (Irvine, CA, USA).

HCT116, HT29, and SNU-C5 cell lines (Korean Cell Line Bank; Seoul, Republic of Korea) and the SNU-C5/5-FUR cell line (Research Center for Resistant Cells; Chosun University, Gwangju, Republic of Korea) were cultured according to the supplier’s protocol at 5% CO₂, 37°C, and humidified atmosphere conditions. The mycoplasma test was done when KCLB distributed the cell lines.

Cells were seeded in triplicate wells of 96-well plates, and treated with COS (0, 1, 10, 100, and 100 μg/mL), 5-FU (0, 0.1, 1, 10, and 100 μM), and PD98059 (0, 1, 5, 10, 25, and 50 μM). Cell density were 2 × 10³ cells for HCT116, HT29, and SNU-C5 cells and 5 × 10³ cells for SNU-C5/5-FUR cells for the cell viability assay. The cells were incubated for 24 h or 3 days, and MTT solution was added to each well. After 3 h, formazan crystals were dissolved in dimethylsulfoxide (DMSO). The absorbance was read at 595 and 620 nm using a VERSAmax microplate reader (Molecular Devices Korea LLC; Seoul, Republic of Korea). The absorbance values of treated cells were compared to those of vehicle-treated cells, representing 100% cell viability.

7-week-old male athymic BALB/c mice were purchased from OrientBio (Seongnam, Republic of Korea) and housed in the Laboratory Animal Facilities, Jeju National University. The mice were housed in a specific pathogen-free environment at a standard temperature (24 ± 2°C), with controlled humidity (50 ± 5%) and 12-h light/dark cycles, and provided standard rodent food and water. All animal experiments were approved by the Jeju National University Institutional Animal Care and Use Committee (IACUC 2019-0051 and IACUC 2022-04). Tumor xenografts were generated by injecting 1 × 10⁶ SNU-C5 cells in 100 μL of Matrigel (BD Biosciences; Seoul, Republic of Korea) into the right flank of the mice as previously described [7].

At first (IACUC 2019-0051), the animals were randomly divided into 3 groups (n = 9/group) from the day of injection as follows: control mice treated with distilled water as the vehicle, and COS (100 and 500 mg/kg)-treated mice. Next (IACUC 2022-04), the tumor-bearing mice were randomly divided into 5 groups (n = 9/group) 10 days after the injection, as follows: control group treated with vehicle, COS5 (500 mg/kg), 5-FU (1 mg/kg), 5-FU + COS1 (100 mg/kg), and 5-FU + COS5. Treatment and measurements are described in Fig. 1. Resected tumor weights (g) were measured in euthanized mice.

Cells were treated vehicle, COS (100 μg/mL) and 5-FU (1 μM) for 72 h, and analyzed by the FACSCalibur^TM system (BD Biosciences) as described previously [6].

Cells were treated with vehicle, COS (100 μg/mL), 5-FU (1 μM), or PD98059 (10 μM), followed by Western blotting as described previously [6,7].

Cell lysates were subjected to gel electrophoresis with a polyvinylidene difluoride membrane. The membranes were incubated with primary antibodies and the appropriate secondary antibody (#PI-2000 and #PI-1000; Vector Laboratories Inc., Burlingame, CA, USA). Protein bands were detected using the Azure^TM c300 and quantified using the AzureSpot analysis software (version 14.2; Azure Biosystemc Inc., Dublin, CA, USA).

All data were compiled from a minimum of 3 replicate experiments and are expressed as the mean ± SEM. A p-value of < 0.05, as determined using the Student’s t-test or one-way analysis of variance followed by a post-hoc test (MS Excel, 2016), indicates a statistically significant difference.

The IC₅₀ values of SNU-C5 cells were significantly lower than those of other CRC cells (Fig. 2A). Cell viability of SNU-C5 cells significantly reduced at 10 (85.8% ± 4.3%, p = 0.020) and 100 μg/mL (69.1% ± 4.0%, p = 0.014) COS, respectively. SNU-C5 cells proliferated up to 1.62 ± 0.01 and 2.62 ± 0.04-fold at 48 and 72 h as compared with the vehicle-treated condition after 24 h of incubation (Fig. 2B). Exposure to 10 μg/mL COS, the minimum effective dose, inhibited cell proliferation by 0.72 ± 0.01, 1.11 ± 0.01, and 2.03 ± 0.04-fold at 24, 48, and 72 h compared to vehicle-treated condition (p < 0.001).

A COS dose of 100 and 500 mg/kg was adopted for the in vivo experiment (Fig. 3). Body weights of the mice were similar among the groups throughout the experiment. Each group showed similar tumor volumes at the beginning, which were not different among the groups until day 30. On day 40, the tumor volume in the vehicle, COS 100 mg/kg, and COS 500 mg/kg group were 1156.4 ± 141.9, 1161.0 ± 251.2, and 686.0 ± 116.3 mm³ (p = 0.144). However, COS treatment at 500 mg/kg was considerably reduced in tumor volume compared with vehicle-treated mice (p = 0.012). Similar results were also observed in tumor weights (0.70 ± 0.08, 0.67 ± 0.15, and 0.46 ± 0.08 g), in which COS at 500 mg/kg significantly reduced tumor weights compared with the vehicle-treated group (p = 0.010). Taken together, the anti-proliferative effects of COS were observed at higher doses.

The IC₅₀ value of SNU-C5/5-FUR cells treated with 5-FU was significantly higher than that of other 5-FU-treated CRC cells (Fig. 4A). SNU-C5 cells treated with 1 μM (94.9% ± 1.0%, p = 0.001) 5-FU showed significantly lower cell viability whereas other CRC cells did not show considerable changes at this concentration. Co-treatment with 5-FU and COS did not induce any further effects in CRC cells, but showed significant additive effects in SNU-C5 cells (p < 0.001) (Fig. 4B). The combined effects of 5-FU and COS were evaluated in SNU-C5 cells using 1 μM 5-FU as the minimally effective dose. SNU-C5 cells proliferated up to 1.49 ± 0.19 and 2.23 ± 0.13-fold at 48 and 72 h compared with vehicle-treated cells following a 24 h incubation. 5-FU treatment inhibited proliferation by 1.45 ± 0.08 and 2.03 ± 0.11-fold at 48 and 72 h compared with vehicle-treated cells. Co-treatment with 5-FU and 10 or 100 μg/mL COS inhibited proliferation by 1.08 ± 0.07 (p = 0.001) and 0.95 ± 0.05-fold (p < 0.001) at 48 h, and 1.55 ± 0.06 and 1.36 ± 0.03-fold (p < 0.001/each) at 72 h, respectively.

COS at 500 mg/kg was adopted for the in vivo experiment (Fig. 5). Body weights of the mice in 5-FU-injected groups were decreased, and thus, the experiments were terminated 3 weeks after treatment.

Each group showed similar tumor volumes at the beginning. Single and co-treatment with 5-FU and COS significantly reduced tumor volumes on day 21 (p = 0.002). COS5 (359.3 ± 41.6 mm³, p = 0.019) and 5-FU (381.8 ± 32.3 mm³, p = 0.026) treatment reduced tumor volumes compared with vehicle-treated mice (479.5 ± 34.3 mm³). Co-treatment of COS1 (271.6 ± 35.1 mm³, p < 0.001) or COS5 (305.2 ± 33.4 mm³, p < 0.001) with 5-FU showed further significant reductions in tumor volumes. Co-treatment with 5-FU and COS1 showed considerable changes (p = 0.017) compared to the 5-FU group, but the changes were not significantly different from the COS5 group (p = 0.062). Co-treatment with 5-FU and COS5 did not show significant changes compared to the COS5 or 5-FU groups. Similar results were also observed for tumor weights (0.39 ± 0.04, 0.29 ± 0.04, 0.26 ± 0.02, 0.20 ± 0.03, and 0.24 ± 0.03 g, respectively, for vehicle, COS5, 5-FU, 5-FU + COS1, and 5-FU + COS5 group). All treated groups showed considerable reductions in tumor weight (p = 0.039, p = 0.004, p = 0.001, p = 0.001 as previously described in each group vs. vehicle-treated group).

As COS enhanced the anti-cancer effects of 5-FU in SNU-C5 cells, the cells were incubated with vehicle, COS (100 μg/mL), 5-FU (1 μM), and COS plus 5-FU, then analyzed by flow cytometry (Fig. 6). COS treatment resulted in a slight change in cell death compared with vehicle- or 5-FU-treated cells. 5-FU treatment markedly increased the ratio of apoptosis and necrosis in SNU-C5 cells with or without COS. Increased cell fraction was observed in total apoptosis (4.14 ± 0.04, 5.60 ± 0.08, 14.46 ± 0.35, and 13.26% ± 0.47% for each treatment described above; p < 0.001), and necrosis (1.16 ± 0.06, 1.04 ± 0.06, 2.15 ± 0.12, and 1.65% ± 0.09% for each treatment described above; p < 0.001) after 5-FU treatment.

Cell cycle analysis revealed a significant change in the G0/G1 and G2/M phases in COS-treated SNU-C5 cells. 5-FU treatment considerably increased the fraction of cells in the S phase, and the fraction was decreased with COS treatment regardless of 5-FU treatment (38.76 ± 1.49, 29.76 ± 0.31, 53.58 ± 1.19, and 49.82% ± 0.19% for each treatment described above; p < 0.001). COS treatment without 5-FU significantly increased the fraction of cells in the G0/G1 phase (p = 0.010), but this was slightly decreased with 5-FU (p = 0.184) (43.53 ± 1.96, 51.08 ± 0.52, 39.08 ± 3.01, and 35.87% ± 0.95% for each treatment described above; p = 0.002). COS treatment without 5-FU slightly increased the fraction of cells in the G2/M phase (17.71 ± 0.54 vs. 19.16% ± 0.36%; p = 0.044), but 5-FU co-treatment showed a marked increase (7.34 ± 1.83 vs. 14.53% ± 0.99%; p = 0.013).

SNU-C5 cells were treated with vehicle, COS (100 μg/mL), 5-FU (1 μM), and COS plus 5-FU for 24 h, then subjected to Western blotting to identify changes in relevant proteins (Fig. 7). Treatment with COS resulted in significant changes in MAPK, p53, and Akt activation (phosphor/total). Among the MAPK signaling pathways, ERK activation was significantly increased by COS treatment with (p = 0.042) or without (p < 0.001) 5-FU treatment (62.04 ± 10.06 in COS, 31.64 ± 7.90 in 5-FU, and 53.24 ± 8.03-fold in 5-FU + COS group compared to vehicle; p < 0.001). Activation of p38 was significantly increased (4.21 ± 0.35, 3.96 ± 0.51, and 4.18 ± 0.28-fold in each group described above; p < 0.001), but the activation of JNK was slightly decreased in all groups. p53 activation was significantly increased by 5-FU treatment, and co-treatment with COS decreased p53 activation (0.94 ± 0.04, 1.43 ± 0.08, and 1.14 ± 0.04-fold; p < 0.001).

Because the anti-cancer effects of COS were ERK-dependent, we checked ERK activation in other CRC cells (Fig. 8). HCT116 and HT29 cells showed faster proliferation than SNU-C5 and SNU-C5/5-FUR cells. The expressions of ERK and pERK were higher in SNU-C5 and SNU-C5/5-FUR cells than in HCT116 and HT29 cells. COS (100 μg/mL) treatment did not affect the proliferation of HT29 and SNU-C5/5-FUR cells. The relative proliferation of vehicle- and COS (100 μg/mL)-treated HCT116 cells was calculated as 4.27 ± 0.23-fold and 2.90 ± 0.20-fold, respectively, on day 3 (p < 0.001). ERK activation was not considerably changed in HCT116 (1.22 ± 0.06-fold; p = 0.082) and HT29 (0.61 ± 0.03-fold; p = 0.054) cells but significantly decreased in SNU-C5/5-FUR (0.78 ± 0.01-fold; p = 0.027) cells.

An ERK inhibitor (PD98059) treatment decreased the viability of HCT116 and HT29 cells in a dose-dependent manner. SNU-C5 and SNU-C/5-FUR cells showed slightly increased viability at a low dose and decreased viability at a high dose of PD98059 (Fig. 9A).

SNU-C5 cells proliferated up to 1.82 ± 0.09 and 2.52 ± 0.10-fold at 48 and 72 h as compared with the vehicle-treated condition after 24 h of incubation. Exposure to 5 μM PD98059 increased the relative cell proliferation by 1.08 ± 0.04, 2.04 ± 0.10, and 2.83 ± 0.16-fold at 24 (p = 0.043), 48 (p = 0.011), and 72 h (p = 0.036) compared to vehicle-treated condition. Exposure to 10 μM PD98059 slightly increased the proliferation without significance. The relative proliferation of vehicle-treated SNU-C5 cells was calculated as 4.14 ± 0.19-fold on day 3, which was significantly decreased to 2.34 ± 0.07-fold by COS (100 μg/mL) treatment (p < 0.001). Treatment with 5 (3.09 ± 0.17-fold; p = 0.007 vs. vehicle) and 10 μM (2.82 ± 0.19-fold; p = 0.040 vs. vehicle) PD98059 and COS enhanced the proliferation. However, cell proliferation was inhibited by 1.46 ± 0.10-fold (p = 0.001) at 50 μM (Fig. 9B).

After treating SNU-C5 cells with the vehicle, COS (100 μg/mL), and COS plus PD98059 (10 μM) for 3 days, (Fig. 9C), ERK activation (p < 0.001) was significantly increased in cells treated with COS compared to vehicle treatment (1.60 ± 0.06-fold; p = 0.001) but decreased with PD98059 co-treatment (0.44 ± 0.04-fold; p < 0.001 vs. vehicle and vs. COS). p38 activation (p = 0.003) was decreased by PD98059 treatment (0.48 ± 0.10-fold; p = 0.007 vs. vehicle, p = 0.002 vs. COS) but similar in cells treated with COS (1.19 ± 0.07-fold; p = 0.077). The activation of JNK and p53 did not change considerably in cells treated with COS and PD98059.