In total, 107 patients, who were diagnosed with CRC at The Second Xiangya Hospital of Central South University, were involved in the present study. The characteristics of these patients are shown in Table I. Normal control tissues were collected from 115 patients without CRC. The tissues were embedded in formalin-fixed paraffin. The follow-up study was performed for 60 months. All of the experiments were approved by the Ethics Committee of the Second Xiangya Hospital of Central South University and informed consent was provided by the patients themselves or their relatives.

Protein expression was assayed by immunohistochemistry using the paraffin-embedded sections from the patients. The sections were first deparaffinized by xylene (Shanghai Sangon, Shanghai, China). Then, the sections were incubated with rabbit polyclonal anti-fibronectin antibody (1:500, ab2413; Abcam, Cambridge, MA, USA) at 4°C for 24 h. After washing with TBST buffer (Shanghai Sangon) for three times, the sections were incubated with goat anti-rabbit IgG H&L (HRP, 1:500, ab6721; Abcam) at room temperature for 1 h. Then, protein immunostaining was performed using DAB Plus substrate (Thermo Fisher Scientific, Waltham, MA, USA) following the manufacturer's instructions. The scoring of the stained tumor cells were divided into five grades: 0 (positive cells <5%), 1 (positive cells between 5 and 25%), 2 (positive cells between 26 and 50%), 3 (positive cells between 51 and 75%), and 4 (positive cells between 75 and 100%). The cells were observed using a light microscope (Axioskop; Zeiss, Germany). We designated grade 1 and 2 as low expression of fibronectin and grade 3 and 4 as high expression of fibronectin for the prognosis analysis.

CCD-18Co (normal colon cells), HT-29 (CRC cells), CaCo2 (CRC cells) and SW480 (CRC cells) cell lines were obtained from the American Type Culture Collection (ATCC; Rockville, MD, USA). The cells were cultured in RPMI-1640 medium (Gibco-BRL, Paisley, UK) supplemented with 10% fetal bovine serum (Hyclone, Logan, UT, USA) at 37°C in 5% CO₂. The fibronectin siRNA (sc-29315) and control siRNA (sc-37007) were purchased from Santa Cruz Biotechnology (Santa Cruz, CA, USA). Transfection was performed using Lipofectamine^® 2000 (Invitrogen Life Technologies, Shanghai, China) following the manufacturer's instructions. Negative small interfering RNAs (siRNAs) and scrambled siRNA were synthesized by Shanghai Sangon.

RNA isolation and reverse transcription were performed using RNAiso Plus kit and PrimeScript™ II First Strand cDNA Synthesis kit (both from Takara, Dalian, China) according to the manufacturer's instructions, respectively. The qPCR primers were: fibronectin: forward, 5′-TTATG ACGAC GGGAA GACCT-3′ and reverse, 5′-GCTGG ATGGA AAGAT TACTC-3′; GADPH forward, 5′-AATCC CATCA CCATC TTCCA-3′ and reverse, 5′-TGGAC TCCAC GACGT ACTCA-3′; caspase-3 forward, 5′-TTAAT AAAGG TATCC ATGGA GAACA CT-3′ and reverse, 5′-TAGAG TTCTT TTGTG AG-3′; p53 forward, 5′-AACGG TACTC CGCCA CC-3′ and reverse, 5′-CGTGT CACCG TCGTG GA-3′; PARP forward, 5′-AGGCT GCTTT GTCAA GAA-3′ and reverse, 5′-CTTGC TGCTT GTTGA AGAT-3′; Bax forward, 5′-ACCAA GCTGA GCGAG TGTC-3′ and reverse, 5′-ACAAA GATGG TCACG GTCTG CC-3′; cytochrome c forward, 5′-CGTCG CATTC CAGAT TATCC A-3′ and reverse, 5′-CAACT ACGGA TATAT AAGAG CCAAA ACTG-3′; and NF-κB forward, 5′-ACCTG AGTCT TCTGG ACCGC TG-3′ and reverse 5′-CCAGC CTTCT CCCAA GAGTC GT-3′. The reaction was performed on Applied Biosystems^® ABI 7500 system (Thermo Fisher Scientific). The reaction system SYBR^® Premix Ex Taq™ II (Takara) was used according to the manufacturers instructions. The reaction condition was: 95°C for 15 min and 40 cycles of 95°C for 10 sec and 60°C for 10 sec. Melting curve analysis was used to determine the specific amplification.

SW480 cells were collected at the treatment time-points with reagents as suggested in each experiment. Cells were lysed with RIPA buffer (Thermo Fisher Scientific) and 15% electrophoresis (Thermo Fisher Scientific) was conducted under 90 V for 2 h. After undergoing electrophoresis, the proteins were transferred to polyvinylidene difluoride membranes. Primary antibodies, rabbit polyclonal anti-fibronectin antibody (1:500), rabbit polyclonal caspase-3 antibody (1:500, ab2302), rabbit polyclonal NF-κB antibody (1:500, ab7971), rabbit polyclonal p53 antibody (1:500, ab1431), rabbit polyclonal PARP (1:500, ab6079), rabbit polyclonal Bax antibody (1:500, ab53154), rabbit polyclonal cytochrome c antibody (1:500, ab90529) and rabbit polyclonal GAPDH antibody (1:500, ab9485) (all from Abcam) were incubated with the membranes for 24 h at 4°C. After washing with TBST buffer three times, the secondary antibody goat anti-rabbit IgG H&L (HRP) was incubated at room temperature for 1 h. Immunostaining was carried out using DAB Plus substrate and chemiluminescence system (Amersham Biosciences, Freiburg, Germany). The results were analyzed using chemiluminescence Molecular Imager^® ChemiDoc™ XRS+ system (Bio-Rad Laboratories, Inc., Hercules, CA, USA).

In order to analyze the viability of the cells, the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay was performed. After undergoing transfection, the cells were seeded in 96-well plates at a density of 1×10⁴ cells/well for 24, 48 and 72 h. Consequently, 20 µl MTT (5 mg/ml; Sigma-Aldrich, St. Louis, MO, USA) was added to the wells and incubated at 37°C for 4 h. Then, the supernatant was removed, and the cells were dissolved in 200 µl of dimethyl sulfoxide (Sigma-Aldrich). The optical density was observed at 490 nm via a spectrophotometer (SpectraMax Plus 384; Molecular Devices, Sunnyvale, CA, USA). The experiments were performed in triplicate.

Transwell assays were performed using a 24-well insert (Corning, Inc., Corning, NY, USA) to analyze the effect of fibronectin on CRC cells. After transfection, the cells (1×10⁴ cells/well) were seeded in the top of the chambers in triplicate for 48 h. The lower chambers with 10% fetal bovine serum were co-cultured for another 72 h. For the invasion assay, extracellular matrix gel (BD Biosciences, Bedford, MA, USA) was used. The cells located on the top surface of the membrane were discarded and the cells on the bottom surface were stained with 0.1% crystal violet (Shanghai Sangon). The number of cells on each insert were calculated in five visual fields randomly and calculated using a light microscope (Axioskop; Zeiss).

The cell cycle distribution was assessed by flow cytometry. The control and transfected cells (48 h after transfection) were collected and washed with PBS buffer (Shanghai Sangon). After fixation in 70% ethanol, the cells were stained using 20 µg/ml propidium iodide (PI) containing 20 µg/ml RNase (DNase-free) (both from Shanghai Sangon) for 2 h. Then the cells were assessed using flow cytometer Partec-PAS (Partec GmbH, Muenster, Germany). The cells in the G₀/G₁, S, G₂/M, and sub-G₁ phases were defined via Multicycle Cell Cycle software (Phoenix Flow System, San Diego, CA, USA).

Female nude mice (6 to 8-weeks old) were provided by The Second Xiangya Hospital of Central South University. To obtain the tumor in vivo model, equal numbers of SW480 cells (1×10⁶) or SW480 cells transfected with control siRNAs and fibronectin-siRNA were injected into the nude mice (four mice per group). After injection, the mice were maintained under a 12-h light/12-h dark cycle and fed with standard mouse diet for four weeks. The tumor volume was measured and calculated as (width² × length)/2. The mice were then sacrificed using sodium pentobarbital (Sigma-Aldrich) and the tumor tissues were collected for further analysis. The study was performed following the recommendations of the Guide for the Care and Use of Laboratory Animals of The Second Xiangya Hospital of Central South University.

All the results are presented as mean ± standard deviation. Data analyses were processed via the two-tailed Student's t-test and the log-rank test. P<0.05 was considered as significant. SPSS v17.0 software was used to perform these analyses (SPSS Inc., Chicago, IL, USA).

The expression levels of fibronectin were shown to be upregulated in the CRC tissues in comparison with that in the normal colon tissues using qPCR (P<0.05, Fig. 1A). Higher expression levels of fibronectin were also noted in the CRC cell lines (P<0.05, Fig. 1B). Western blot analysis indicated that the fibronectin blot was remarkably denser in the CRC samples (Fig. 1C). In the CRC cell lines, the protein levels of fibronectin showed a significant increase (Fig. 1D). In addition, as shown in the immunohistochemistry results, the expression level of fibronectin was obviously higher in the CRC tissues than that in the non-cancerous, normal colorectal tissues (Fig. 1E). The fibronectin protein expression was assessed in 107 CRC tissue samples by immunohistochemistry. In the CRC tissues, 54.21% (58/107) of the cases had high fibronectin expression (grade >2). Additionally, we observed a marked difference in fibronectin expression in tumors with different TNM stage (P=0.0025) and distant metastasis (P=0.0013) (Table I).

In short, the survival analysis showed that the survival time of patients whose fibronectin expression was low (grade 1–2) was significantly longer than that of those whose fibronectin expression was high (grade 3–4) (log-rank test, P<0.05, Fig. 2A). Patients with metastatic lymph nodes whose fibronectin expression was low showed a longer overall survival time than that in the patients whose fibronectin expression was high (Fig. 2B, log-rank test, P<0.05). These outcomes indicate that fibronectin expression is significantly related to the survival rate of CRC patients.

Based on our preliminary data of the high expression level of fibronectin in SW480 cell lines, we used RNAi technique to silence this gene in the SW480 cells. Fibronectin-siRNA (si-fibronectin) effectively suppressed fibronectin expression in the SW480 cells (Fig. 3). First, si-fibronectin was successfully transfected into the SW480 cells as indicated by fluorescent labels (Fig. 3A). Meanwhile the efficiency of the transfection had a time-dependent effect (Fig. 3B). Although no obvious decrease in expression of fibronectin was found after si-fibronectin transfection of SW480 cells for 12 h, fibronectin mRNA and protein levels were significantly suppressed at 16, 24 and 48 h (Fig. 3B and C).

The effects of fibronectin on CRC SW480 cells were assessed using siRNA approach. After transfection with si-fibronectin, cell proliferation was significantly decreased compared to that observed in the control and si-control groups (Fig. 4A). In addition, fibronectin knockdown also suppressed cell migration (Fig. 4B) and cell invasion (Fig. 4C) of the SW480 cells. These results indicate the suppression of cell activities by fibronectin knockdown.

Effects of the silencing of fibronectin on the suppression of cell progression have been proven. To demonstrate the effect of fibronectin knockdown on the cell cycle, flow cytometry was performed. The results for the SW480 cells are shown in Fig. 5A after transfection with si-fibronectin, si-control and in the untreated control. After transfection with si-fibronectin, the percentage of cells in the S phase increased to 40.54% which was significantly higher compared to the control and si-control groups (Fig. 5B). The percentage of cells in the G₀/G₁ phases after si-fibronectin transfection was decreased to 42.62% (Fig. 5B). Hence, these results showed that the si-fibronectin transfection suppressed CRC cell viability. The cells could not proliferate normally by arrest in the S phase of the cell cycle and S-phase arrest may induce cell death and morphological changes during progression of CRC.

After understanding the effects of si-fibronectin transfection on the cell cycle, we explored whether the NF-κB/p53-apoptosis signaling pathway is regulated by fibronectin. The qPCR results showed that caspase-3, p53, PARP, Bax and cytochrome c were significantly increased after si-fibronectin transfection while only NF-κB was significantly depressed by si-fibronectin transfection (Fig. 6A). Similar results were observed for the protein expression levels. Caspase-3, p53, PARP, Bax and cytochrome had higher expression after si-fibronectin transfection while NF-κB had lower expression following si-fibronectin transfection (Fig. 6B).

To investigate the effects of fibronectin on tumor growth in vivo, the nude mice were injected using equal numbers of SW480 cells (1×10⁶) or SW480 cells transfected with si-control or si-fibronectin. After 4 weeks, tumors significantly appeared in the mice. The results showed that si-fibronectin inhibited tumor growth in vivo (Fig. 7A and B). The expression of fibronectin was measured using qPCR and western blotting. Both at the RNA (Fig. 7C) and protein levels (Fig. 7D), the fibronectin expression was significantly decreased in the tumor tissue after transfection with si-fibronectin.

The effects of si-fibronectin on expression of NF-κB/p53-apoptosis signaling pathway in nude mice were assayed by qPCR and western blot analysis. The result was very similar to the results in vitro. qPCR indicated that caspase-3, p53, PARP, Bax and cytochrome c were significantly increased by si-fibronectin transfection in the tumor tissues from the nude mice while NF-κB was decreased after transfection (Fig. 8A). The result of western blotting also showed that caspase-3, p53, PARP, Bax and cytochrome c had higher expression after si-fibronectin transfection while NF-κB had lower expression after si-fibronectin transfection (Fig. 8B).