Six-week-old male C57BL/6J mice (weight is ~20 g; Kyudo Co., Ltd., Kumamoto, Japan) were randomly allocated into four experimental groups: Group C (n=7), the control group, was fed standard chow (SC) and normal water; Group D (n=10) was fed SC and water containing 1% dextran sulfate sodium (DSS; MW 5,000; Wako Pure Chemical Industries, Ltd., Osaka, Japan); Group F (n=7) was fed a high-fructose diet (HFD; Oriental Yeast Co., Ltd., Tokyo, Japan) and normal water; and Group FD (n=10) was fed an HFD and water containing 1% DSS. The mixture containing 1% DSS and HFD was supplied for 2 weeks, without normal drinking water. The composition of HFD was 60% fructose, 20.7% casein, 0.3% methionine, 5.0% lard, 9.249% cellulose, 4.5% vitamins and minerals, 0.25% choline bitartrate, and 0.001% tertiary butylhydroquinone. Each day, the mice were visually assessed to investigate the presence of loose stools, blood in the feces or decreased mobility, the latter considered the disease activity index (DAI) score, to evaluate the clinical activity. The DAI scores were determined based on the following criteria: Changes in body weight, stool consistency and fecal occult blood or visible bleeding. These parameters were assessed on a scale of 0-4 (20-22). The colon tissues were fixed in 10% formalin, embedded in paraffin, and stained with hematoxylin and eosin to evaluate the pathological activity. The pathological score was used to evaluate the histological activity (23). In addition, a section of the colon tissue was stored at -80°C in order to analyze the gene expression. All animals were housed under standard conditions (in a 24°C-controlled environment with a 12-h light/dark cycle) and received humane care in compliance with the institutional guidelines. The present study was approved by the institutional Animal Care and Use Committees of Kagoshima University (Kagoshima, Japan) and was performed in accordance with the Committees' guidelines for animal experiments.

The feces of the mice that were fed SC (Group C) and those fed with the HFD (Group F) were collected; 10 µg of feces from each group was diluted in 400 µl of distilled water and the level of fructose in the feces was measured using an EnzyChrom Fructose Assay kit (cat. no. EFRU-100; BioAssay Systems LLC, Hayward, CA, USA), a Bio-Rad iMark Microplate Reader and the Microplate Manager software program v6 (Bio-Rad Laboratories, Inc., Hercules, CA, USA).

Total RNA (0.5 µg) was reverse transcribed using a PrimeScript RT Reagent kit (Takara Bio Inc., Otsu, Japan) at 37°C for 15 min and 85°C for 5 sec. Each reaction mixture for RT (total volume 20 µl) consisted of total RNA (adjusted to 500 ng, 2.5 µl), oligo dT PCR primer (50 µM, 1 µl), 5X PrimeScript buffer (4 µl), PrimeScript RT enzyme mix (1 µl), random 6 mers (100 µM, 1 µl) and RNase free distilled H₂O (10.5 µl). The synthesized cDNA was amplified by RT-qPCR using SYBR^® Premix Ex Taq™ II (Takara Bio Inc.). Each reaction mixture for PCR (total volume 20 µl) consisted of cDNA (adjusted to 500 ng, 2 µl), forward and reverse PCR primer (10 µM, 0.8 µl each), SYBR Premix Ex Taq II (10X, 10 µl), ROX reference dye (50X, 0.4 µl) and distilled H₂O (6 µl). The cycling conditions were as follows: One cycle at 95°C for 30 sec followed by 35 cycles each at 95°C for 5 sec and 60°C for 34 sec. Data were collected and analyzed using the ABI Prism software program v2.3 (Applied Biosystems; Thermo Fisher Scientific, Inc., Waltham, MA, USA). The following primers were used: Mouse interleukin (IL)-1β, IL-10, IL-6, transforming growth factor-β (TGF-β), tumor necrosis factor-α (TNF-α), GAPDH, occludin, claudin-1 and zonula occludens-1 (ZO-1) from Takara Bio Inc., and Sigma-Aldrich; Merck KGaA (Darmstadt, Germany), listed in Table I, and human occludin, claudin-1 and ZO-1 (Table II).

The HT-29 cells were exposed to fructose (0, 0.01, 0.1, 1 or 10 mM), and total protein was extracted from the cells using RIPA buffer (Sigma-Aldrich; Merck KGaA). The cell lysates were analyzed for protein content using a Lowry Assay (Bio-Rad Laboratories, Inc.). Cell lysates containing equal quantities of protein (10 µg) from each group were separated on 10% SDS polyacrylamide gels (Wako Pure Chemical Industries, Ltd.) and electroblotted onto polyvinylidene fluoride membranes. (Bio-Rad Laboratories, Inc.). Following blocking overnight at 4°C with 5% non-fat milk, the blots were probed with primary antibodies for 1 h at room temperature. The membranes were then incubated with the following primary antibodies: Anti-occludin antibody (1:1,000; cat. no. 71-1500; Invitrogen; Thermo Fisher Scientific, Inc.), anti-claudin 1 antibody (1:1,000; cat. no. ab15098; Abcam, Cambridge, MA, USA), anti-ZO-1 antibody (1:1,000; cat. no. 40-2200; Thermo Fisher Scientific, Inc.), and β-actin antibody (1:1,000; cat. no. 4970; Cell signaling Technology, Inc., Danvers, MA, USA). The membranes were washed with Tris buffered saline (Bio-Rad Laboratories, Inc.) with 0.1% Tween-20 (Wako Pure Chemical Industries, Ltd.) three times. Following incubation with horseradish peroxidase-conjugated secondary antibodies (1:1,000; cat. no. 7074; Cell Signaling Technology, Inc.) for 1 h at room temperature, the membrane was visualized for the reaction using electrogenerated chemiluminescence western blotting detection reagents (GE Healthcare Life Sciences, Shanghai, China). Expression level quantification was performed using Image J v1.48 (National Institutes of Health, Bethesda, MD, USA), and β-actin was used as an internal control.

DNA was extracted from fecal samples and standard strains using ISOFECAL for Beads Beating (Nippon Gene Co., Ltd., Tokyo, Japan) for quantification by RT-qPCR analysis. First, 20 µg of feces was diluted in 200 µl of distilled water. The standard strains were used for each group: Bifidobacterium longum JCM 1217 for the Bifidobacterium group, 06TCa19 for the Lactobacillus group, Clostridium coccoides JCM 1395 for the C. coccoides group, and Bacteroides fragilis JCM 11019 for the Bacteroides-Prevotella group. The Bifidobacterium subspecies and the C. coccoides, Bacteroides-Prevotella and Lactobacillus groups were detected using SYBR-Green RT-qPCR analysis. The following primers were used (Sigma-Aldrich; Merck KGaA): Bifidobacterium longum, Lactobacillus, C. coccoides, Bacteroides fragilis (Table III). Gene quantification was performed using an ABI Prism 7700 sequence detection system (Applied Biosystems; Thermo Fisher Scientific, Inc.). To determine the total bacteria count, fecal samples were stained with 4′,6-diamidino-2-phenylindole solution (Dojindo Molecular Technologies, Inc., Kumamoto, Japan), and the bacteria were counted using a hemocytometer for bacteria (AS ONE Corporation, Osaka, Japan) under a fluorescence microscope (Carl Zeiss AG, Oberkochen, Germany).

The human HT-29 colon cancer cell line (European Collection of Cell Cultures, no. EC910722201) was obtained from Dainippon Sumitomo Pharma Biomedical (Osaka, Japan) and maintained at 37°C in 5% CO₂ and McCoy's 5A medium supplemented with 10% heat-inactivated fetal bovine serum (Gibco; Thermo Fisher Scientific, Inc.), streptomycin, and penicillin.

The epithelial permeability to macromolecules was determined using fluorescein isothiocyanate (FITC)-dextran. The 1.5×10⁵ cells/well of HT-29 cells were seeded on the upper chamber with 8-µm pores and cultured until they reached full confluence in monolayers. The cells in the upper chamber were exposed to fructose (0.001, 0.01, 0.1, 1 or 10 mM) and then incubated with FITC-dextran (5 mg/ml) for 24 h at 37°C in a CO₂ incubator. The fluorescence of FITC-dextran in the lower chambers was sequentially measured following exposure to fructose (0, 0.5 and 1 h) 24 h at 37°C in a CO₂ incubator. The data were calculated as follows: Apparent permeability (cm/s) = (rate of change in cumulative mass of FITC transferred to lower chamber)/[(surface area of cell culture insert) × (initial concentrations of radio-labeled substances)] (24). This assay was conducted five times.

An LDH Cytotoxicity Detection kit was used (Takara Bio, Inc.) to examine the cytotoxicity of fructose towards HT-29 cells that were exposed to various concentrations of fructose (0, 0.001, 0.01, 0.1, 1 or 10 mM) for 0.5, 1 or 2 h. An ELISA was performed using a Bio-Rad iMark Microplate Reader and the Microplate Manager software program (Bio-Rad Laboratories, Inc.). This LDH assay was conducted five times.

All data are expressed as the mean ± standard error of the mean. Differences between two groups were analyzed using the Mann-Whitney U test. Differences among multiple groups were compared using the Kruskal-Wallis test and one-way analysis of variance (IBM SPSS v23.0; IBM Corp., Armonk, NY, USA). P<0.05 was considered to indicate a statistically significant difference.

First, the fecal concentrations of fructose were measured and the results of Groups C and F were compared using an ELISA following 14 days of feeding with normal chow and an HFD, respectively. The median concentrations of fructose in Groups C and F were 864 and 1,679 µM, respectively. The level of fructose in the stools of Group F was significantly higher than that in Group C (Fig. 1).

No significant difference in the rates of body weight change were found among Groups C, D and F; however, the rate of body weight change in Group F tended to be lower compared with those in Groups C and D. The rate of body weight change in Group FD was significantly decreased compared with that in the other groups on day 14 (Fig. 2A). The colon length in Group FD was significantly shorter than in the other groups (Fig. 2B). In addition, the pathological and DAI scores of Group FD were significantly increased compared with those of Groups C, D and F; furthermore, the scores of Group F were significantly increased compared with those of Group C (Fig. 2C and D). On pathological examination of the colon, mild inflammatory cell infiltration was observed in the mucosal layer, whereas the epithelial cells were normal in Group F. In Group FD, the mucosa was broken by erosion or ulcers, and infiltration of inflammatory cells into the submucosal layer was observed (Fig. 2E).

The gene expression levels of TNF-α, IL-6 and IL-1β in Groups D and FD were significantly increased compared with those in Group C. The gene expression levels of IL-6 and IL-1β in Group FD were significantly increased compared with those in Group D; however, the expression of TNF-α in Groups D and FD did not differ to a statistically significant extent. The gene expression levels of IL-10 and TGF-β in Group FD were significantly increased compared with those in Group D (Fig. 3).

The expression of occludin in Group F was decreased compared with that in Group C. The expression of occludin in Group FD was significantly decreased compared with that in Group C. Otherwise, the expression levels of ZO-1 and claudin-1 in the large intestine of these groups did not differ to a statistically significant extent (Fig. 4).

The numbers of intestinal bacteria, including the Lactobacillus group, Bifidobacterium group, C. coccoides group and Bacteroides-Prevotella in each of the groups were measured. The total numbers of bacteria in the colon of Group F and FD were increased compared with those of Groups C and D. The total counts in each group were as follows: Group C, 4.3×10¹⁰ CFU/ml; Group D, 4.4×10¹⁰ CFU/ml; Group F, 5.8×10¹⁰ CFU/ml; and Group FD, 6.2×10¹⁰ CFU/ml (Fig. 5A). The counts (and percentages) of Bacteroides-Prevotella in the four groups were as follows: Group C, 2.3×10⁸ CFU/ml (1.02%); Group D, 2.2×10⁸ CFU/ml (0.93%); Group F, 6.1×10⁸ CFU/ml (1.59%); and Group FD, 6.2×10⁸ CFU/ml (1.90%). The counts of C. coccoides were as follows: Group C, 3.0×10⁹ CFU/ml (13.0%); Group D, 5.2×10⁹ CFU/ml (21.48%); Group F, 1.1×10¹⁰ CFU/ml (28.9%); and Group FD, 7.7×10⁹ CFU/ml (23.72%). The counts of the Lactobacillus group were as follows: Group C, 5.7×10⁸ CFU/ml (2.5%); Group D, 9.6×10⁸ CFU/ml (3.9%); Group F, 3.0×10⁹ CFU/ml (7.9%); and Group FD, 4.1×10⁹ CFU/ml (12.7%). Finally, the counts of the Bifidobacterium group were as follows: Group C, 2.0×10⁷ (0.08%); Group D, 1.0×10⁷ (0.04%); Group F, 4.1×10⁷ (0.1%); and Group FD, 4.9×10⁷ (0.15%). Excessive fructose intake affected the numbers and percentages of the Bacteroides-Prevotella group, C. coccoides, the Lactobacillus group and the Bifidobacterium group (Fig. 5B).

The gene expression levels of occludin, ZO-1 and claudin-1 of HT-29 cells following exposure to several concentrations of fructose were measured using RT-qPCR analysis, and the gene expression of occludin was significantly decreased by fructose exposure in a dose-dependent manner. The gene expression of ZO-1 and claudin-1 did not differ in vitro (Fig. 6A). The expression of tight junction proteins was also examined by western blotting. The expression of occludin in HT-29 cells was significantly decreased by fructose exposure in a dose-dependent manner, whereas no differences in the expression ZO-1 or claudin-1 were observed (Fig. 6B and C).

The effect of fructose on paracellular permeability to macromolecules was examined. The HT-29 cells were exposed to various concentrations of fructose (0, 0.001, 0.01, 0.1, 1 or 10 mM) for several sequential periods (0, 0.5 or 1 h). Fructose enhanced the paracellular permeability of the cells in a concentration- and time-dependent manner (Fig. 7A). Furthermore, the cytotoxicity of fructose to cells was examined. The HT-29 cells were exposed to various concentrations of fructose (0, 0.001, 0.01, 0.1, 1 or 10 mM) for several periods (0.5, 1 or 2 h). Fructose injured epithelial cells in a concentration- and time-dependent manner (Fig. 7B).