This study was performed in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals (22). All experimental protocols were performed in accordance with National Institutes of Health and approved by the Committee on the Ethics of Animal Experiments Defence Research (Northeast Agricultural University, Harbin, China).

A total of 20 C57BL/6 mice, 6–8 weeks old, were purchased from Jackson Laboratory (Bar Harbor, ME, USA) and housed in a temperature-controlled room (25±1°C) with artificial 12/12 h light/dark cycle. All mice could access water containing 1,2-dimethylhydrazine (3 mg/kg) to induce colon cancer. The incidence of colon tumor induced by 1,2-dimethylhydrazine was calculated by histopathology as described in immunohistochemistry assay. Experimental mice were divided into two groups (n=10/group) with free access to a regular diet (5 mg/kg) or a resistant starch diet (5 mg/kg). All mice were sacrificed for further analysis on day 120.

Ammonia in experimental mice was measured using ionization constant. pH was determined by pH meter (Mettler). Short chain fatty acids were analyzed using High Performance Liquid Chromatography (Takara, Tokyo, Japan).

Colon epithelial cells and colon tumor cells were isolated from experimental mice and cultured in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal bovine serum (FBS; Sigma-Aldrich; Merck KgaA, Darmstadt, Germany). Cells were cultured at 37°C and 5% CO₂.

RNA was reverse transcribed into cDNA at 42°C for 2 h using the High Capacity cDNA Reverse Transcription kit (Thermo Fisher Scientific, Inc., Waltham, MA, USA) according to the manufacturer's protocol. PCR amplification had preliminary denaturation at 94°C for 2 min, followed by 45 cycles of 95°C for 30 sec; the annealing temperature was reduced to 56.8°C for 30 sec and 72°C for 10 min. The reaction volume was a total of 20 μl containing 50 ng genomic cDNA, 200 μM dNTPs, 200 μM primers, and Taq DNA polymerase and SYBR-Green (both 2.5 U; Thermo Fisher Scientific, Inc.). Total RNA was extracted from colon epithelial cell and colon tumor cells by using RNAeasy mini kit (Qiagen, Inc., Valencia, CA, USA). mRNA expression levels of heat shock protein 25 (HSP25), protein kinase C-d (PKC-d), gastrointestinal glutathione peroxidase (GI-GPx), c-myc, Ras, p53, proliferating cell nuclear antigen (PCNA), claudin 1, claudin 2, mechanistic target of rapamycin kinase (mTOR), hexokinase-2 (HK-II), caspase-3, caspase-9, p53, Bcl-2 apoptosis regulator (Bcl-2), superoxide dismutase (SOD) and glutathione synthetase (GSH) in colon epithelial cell and/or colon tumor cells were measured by RT-qPCR with β-actin as an endogenous control (23) (Invitrogen; Thermo Fisher Scientific, Inc.). All the forward and reverse primers were synthesized by Invitrogen (Thermo Fisher Scientific, Inc.) (Table I). Relative mRNA expression changes were calculated by 2^−ΔΔCq (24). The results are expressed as the n-fold change compared with control.

Colon tumor cells were homogenized in a radioimmunoprecipitation assay buffer (Sigma-Aldrich; Merck KgaA) and centrifuged at 6,000 × g at 4°C for 10 min. Protein concentration was measured with a bicinchoninic acid protein assay kit (Thermo Fisher Scientific, Inc.). A total of 10 μg/lane protein was were separated in a 12% SDS assay and then transferred onto polyvinylidene fluoride membranes (EMD Millipore, Billierica, MA, USA). Membranes were blocked in 5% BSA (Sigma-Aldrich; Merck KgaA) for 1 h at 37°C and subsequently incubated with the following primary antibodies: HSP25 (ab202846), PKC-d (ab182126), GI-GPx (ab137431), c-myc (ab32071), Ras (ab52939), P53 (ab1431), PCNA (ab18197), claudin1 (ab15098), claudin2 (ab53032), mTOR (ab2732) and HK-II (ab24937), caspase-3 (ab13847), caspase-9 (ab202068), Bcl-2 (ab59348), SOD (ab13533), GSH (ab26255), DDIT3 (ab179823), Beclin1 (ab62557), CHOP (ab10444), BIP (ab108615), caspase-12 (ab62484), ATF-4 (ab23760), BACE1 (ab2077), eIF2α (ab5369) and β-actin (ab8227) for 12 h at 4°C. All primary antibodies were used at a dilution of 1:1,000 and purchased from Abcam (Cambridge, UK). The membranes were then incubated with horseradish peroxidase (HRP)-conjugated goat anti-rabbit immunoglobulin G (IgG) monoclonal secondary antibodies (1:2,000; cat. no. PV-6001; OriGene Technologies, Inc., Beijing, China) for 24 h at 4°C. An enhanced chemiluminescence substrate ECL Select™ (Roche Diagnostics, Basel, Switzerland) was used to analyze the protein expression (Olympus BX51; Olympus Corporation, Tokyo, Japan).

Colon tumor cells (1×10⁶) were transfected with 100 pmol of siRNA targeting eukaryotic translation initiation factor 2α (eIF2α) with siRNA-vector as control (both from Applied Biosystems; Thermo Fisher Scientific, Inc.) using a Cell Line Nucleofector kit L. (Lonza Group, Ltd., Basel, Switzerland). Cells were cultured in 2.5 ml DMEM containing 10% FBS 6-well plates for 24 h. All siRNAs were synthesized by Invitrogen (Thermo Fisher Scientific, Inc.) including siRNA-eIF2α (eIF2α, L-015389) or siRNA-vector (scramble, D-001810). Cells were used for the subsequent assays after 48-h transfection.

Colon tumor cells isolated from experimental mice and cultured in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal bovine serum (FBS; Sigma-Aldrich; Merck KgaA) at a 37°C humidified atmosphere of 5% CO₂. Cell colonies growing on Matrigel^® were loosely detached by dispase treatment for 5 min, washed 3 times with PBS. Cells were resuspended in DMEM medium containing 20% FBS. Cells were maintained on 1% agar-coated and allowed to differentiate for another 18 days. Cells were then fixed with 10% formalin for 1 h at 37°C. Next following stained with the 60% Oil Red O in isopropanol as working solution for 10 min. The proportion of Oil Red O-positive cells was determined by counting stained cells under a light microscope.

Eukaryotic translation initiation factor 2-α kinase 3 (PERK) activity in colon tumor cells was analyzed by recombinant glutathione S-transferase-PERK (536-1,116 amino acids) with 6-His-full-length human eIF2α as a substrate (25). eIF2α activity was analyzed by stimulation of eIF-2-α kinase GCN2 in colon tumor cells (26). AMPK activity was determined by transient transfection assays as described previously (27).

Immunohistochemistry and immunofluorescent staining were performed according to the standard procedures (28). Paraffin-embedded colon tumor tissues sections were prepared and epitope retrieval was performed for further analysis. The paraffin sections were treated with hydrogen peroxide (3%) for 10–15 min, which subsequently blocked by a regular blocking solution for 10–15 min 37°C for immunohistochemistry. Colon tumor cells were cultured and stained with microtubule associated protein 1 light chain 3 α (MAP1LC3A), translocase of outer mitochondrial membrane 20 (TOMM20) for observation of microtubules and/or microvessels, Neo (Nase), and calreticulin (Invitrogen; Thermo Fisher Scientific, Inc.), NRP-2 (ab129050; Abcam), Apaf-1 (ab2001; Abcam), Bad (ab32445; Abcam) for immunofluorescent staining. All antibodies were used at a dilution of 1:1,000. Cells were then incubated with goat anti-rabbit IgG H&L (HRP) (1:2,000; ab205718; Abcam) for 1 h at 37°C. All fluorescent samples were visualized with a confocal fluorescence microscope (Leica TCS SP8; Leica Corporation, Wetzlar, Germany).

Cells (1×10⁷) were collected from the experimental mice, and fixed with 70% ethanol for 2 h at 30°C. Fixed cells were rehydrated in PBS for 5 min and incubated in RNase A (1 mg/ml) for 30 min at 37°C. The cells were then subjected to PI/RNase staining followed by flow cytometric analysis using a FACScan instrument (Becton Dickinson, Mountain View, CA, USA) and Cell Quest software (Becton Dickinson).

Colon tumor cells were isolated from experimental mice and trypsinized and collected for apoptosis analysis. The cells were adjusted to 5×10⁶ cells/ml with phosphate-buffered saline, labeled with Annexin V-fluorescein isothiocyanate (FITC) and propidium iodide (PI) using an Annexin V-FITC kit, and analyzed with a FACScan flow cytometer (both from BD Biosciences, Franklin Lakes, NJ, USA). The treatments were performed in triplicate, and the percentage of labeled cells undergoing apoptosis in each group was determined and calculated using FCS Express™ 4 IVD software 1.0 (De Novo Software, Glendale, CA, USA).

All data were presented as mean + standard error with three independent experiments. Statistical significance was analyzed using two tailed Student's t-test between groups. Unpaired data was analyzed by variance. P<0.05 was considered to indicate a statistically significant difference.

In order to investigate the benefits of resistant starch diets for metabolism of experimental mice, body weight and metabolic characteristic of colon tissues were analyzed. A resistant starch diet decreased body weight compared with a regular diet (Fig. 1A). Weight of the proximal and distal colon was also increased by resistant starch diets in experimental mice (Fig. 1B and C). The results demonstrated that digesta weight in the caecum, proximal colon and distal colon was significantly increased by the resistant starch diet compared with a regular diet (Fig. 1D). The digesta pH in the caecum, proximal colon and distal colon was downregulated in mice fed with resistant starch compared with a normal diet (Fig. 1E). Free ammonia, pH and short chain fatty acids (SCFA) in feces were decreased in mice fed with resistant starch (Fig. 1F–H). These results suggest that a resistant starch diet improves body weight and metabolic characteristics of colon tissues.

Anti-tumorigenesis efficacy of resistant starch was investigated in colon tissues induced by 1,2-dimethylhydrazine. Results showed that resistant starch suppressed tumor formation in experimental mice treated by 1,2-dimethylhydrazine (Fig. 2A). Histopathology demonstrated that resistant starch significantly inhibited incidence of colon tumor in mice model induced by 1,2-dimethylhydrazine (Fig. 2B). The gene and protein expression levels of carcinogenesis-associated genes, HSP25, PKC-d and GI-GPx, were downregulated in colon epithelial cells from mice receiving resistant starch compared with the standard starch group (Fig. 2C and D). However, the gene and protein expression levels of oncogenes c-myc, Ras and pro-apoptosis gene p53 were upregulated by resistant starch in colon epithelial cells (Fig. 2E and F). Immunofluorescence demonstrated that microtubule and tumor vessels were inhibited in colon tissues in mice in the resistant starch group compared with the control group (Fig. 2G). Immunohistochemistry demonstrated that the expression levels of MAT-1 and NRP-2 were downregulated by resistant starch in colon tumor tissues (Fig. 2H). These results indicated that a resistant starch diet inhibits tumorigenesis in colon tissues induced by 1,2-dimethylhydrazine.

Tumor cell proliferation and differentiation has a vital role in tumorigenesis. Thus, the effect of resistant starch on colon tumor cell proliferation and differentiation was determined. Colon tumor cell proliferation was suppressed by resistant starch compared with the control group in 1,2-dimethylhydrazine-induced mice (Fig. 3A). RT-qPCR and western blot analysis demonstrated that the expression levels of PCNA, claudin 1 and claudin 2 were decreased in colon tumor cells in resistant starch-fed mice compared with the control group (Fig. 3B and C). The results also demonstrated that resistant starch inhibited colon tumor cell differentiation in colon tissues (Fig. 3D). RT-qPCR and western blot analysis demonstrated that expression levels of mTOR and HK-II were reduced by resistant starch compared with the control diet in colon tumor cells (Fig. 3E and F). The resistant starch diet increased S phase arrest of colon tumor cells, and downregulation of mTOR and HK-II expression levels compared with the normal diet group (Fig. 3G). Long-term survival of experimental mice was prolonged by the inclusion of resistant starch in the diet compared with the control diet group (Fig. 3H). These results suggest that resistant starch diets can inhibit the proliferation and differentiation of colon tumor cells by arresting the cell cycle.

The anti-apoptosis effects of resistant starch were also analyzed in experimental mice with 1,2-dimethylhydrazine-induced colon tumors. The resistant starch diet promoted apoptosis of colon tumor cells (Fig. 4A). Cellular structure demonstrated that mitochondria exhibited different degrees of damage in colon tumor cells (Fig. 4B). Pro-apoptosis gene and protein expression levels of caspase-3 and caspase-9 were upregulated in colon tumor cells (Fig. 4C and D). However, anti-apoptosis gene and protein expression levels of p53 and Bcl-2 were downregulated in colon tumor cells in resistant starch-fed experimental mice compared with the control treatment (Fig. 4E and F). Immunohistochemistry and immunofluorescence demonstrated that resistant starch increased Apaf-1 and Bad expression levels in colon tumor tissues in experimental mice compared with control mice (Fig. 4G and H). These results suggest that resistant starch in the diet can enhance apoptosis of colon tumors through mitochondrial apoptotic pathway in mice with colon tumors induced by 1,2-dimethylhydrazine.

Changes of oxidative stress in colon tumors and colon epithelial cells was investigated. Resistant starch diets reduced the mRNA and protein levels of SOD and GSH in colon tumor cells compared with mice that received the control diet (Fig. 5A and B). However, mRNA and protein levels of SOD and GSH were increased by resistant starch in colon epithelial cells compared to regular diet (Fig. 5C and D). AMP-activated protein kinase (AMPK) activity was increased in colon tumor cells from mice fed with resistant starch compared with the regular diet (Fig. 5E). Potential mechanisms were demonstrated as resistant starch increased the expression levels of DNA damage-inducible transcript 3 protein and Beclin 1 in colon tumor cells (Fig. 5F). The results also demonstrated that resistant starch promoted mitophagy and reticulophagy of colon tumor cells (Fig. 5G and H). Collectively, the results indicated that the resistant starch diet enhanced-activated autophagy in colon tumors through upregulation of autophagy genes.

The effect if resistant starch on ER stress and its potential mechanism in inhibition of tumor cells growth was investigated. The resistant starch diet increased the expression levels of eIF2α, ATF-4 and secretase-β (BACE1) in colon tumor cells compared with cells from control mice (Fig. 6A). Total protein expression levels of eIF2α and PERK were increased by resistant starch in colon tumor cells (Fig. 6B). eIF2α and PERK activity was increased by the resistant starch diet compared with the regular diet in mice induced by 1,2-dimethylhydrazine (Fig. 6C). The resistant starch diet increased the expression levels of DNA damage-inducible transcript 3 protein (CHOP), binding immunoglobulin protein (BIP) and caspase-12 in colon tumor cells (Fig. 6D). In vitro assays demonstrated that downregulation of expression levels of eIF2α using siRNA suppressed PERK activity in colon tumor cells (Fig. 6E). Expression levels of ATF-4 and BACE1 were reduced by knockdown of eIF2α compared with control siRNA in colon tumor cells (Fig. 6F). Knockdown of eIF2α exhibited inhibitory effects on CHOP, BIP and caspase-12 expression in colon tumor cells isolated from mice fed by resistant starch (Fig. 6G). Result also demonstrated that knockdown of eIF2α inhibited resistant starch-induced apoptosis of colon tumor cells (Fig. 6H). Collectively, these results suggest that the resistant starch diet regulated ER stress-dependent PERK activity through upregulation of eIF2α activity in colon tumor cells in mice induced by 1,2-dimethylhydrazine.