The study protocol was approved by the Medical Ethics Committee of the Third Hospital of Hebei Medical University (Shijiazhuang, China). All methods used in the present study were performed according to the International Ethical Guidelines for Biomedical Research Involving Human Subjects, and informed consent was obtained from all participants prior to the study.

A total of 105 CRC patients whose cancer was removed at The Third Hospital of Hebei Medical University were recruited between January and December 2010. The mean age was 56.36±9.19 year (range, 38–78 years) with 30 females and 75 males. All participants were first diagnosed with CRC without any other malignancies and the diagnosis was conformed by pathological examination. No participants had received radio- or chemotherapy or targeted therapy prior to surgery. Complete clinicopathological data and follow-ups were recorded. TNM Classification of Malignant Tumors was performed according to Union for International Cancer Control/American Joint Committee on Cancer gastric cancer staging system (13). The follow-ups ended in December 2015. Paraffin-embedded samples of tumor tissues and adjacent mucosa (≥2 cm from the edge of the caner with no cancer cells verified) were obtained for the detection of ANXA9 protein expression. Additional fresh samples of tumor tissues and adjacent mucosa were collected from 20 participants who had undergone surgery at The Third Hospital of Hebei Medical University (Shijiazhuang, China) between March and October 2016. These samples were maintained at −80°C and the expression level of ANXA9 proteins were detected by western blotting.

Human CRC cell lines, Caco-2, HCT116, SW620, SW480 and L_OV_O were purchased from the Cell Resource Center of Life Sciences (Shanghai, China). Cells from the HT-29 cell line were preserved and incubated at the Scientific Research Center of The Third Hospital of Hebei Medical University (Shijiazhuang, China). Lipofectamine™ 2000 Reagent (Invitrogen; Thermo Fisher Scientific, Inc., USA), Gibco Dulbecco's modified Eagle's medium (DMEM) and Gibco fetal bovine serum (FBS; Thermo Fisher Scientific, Inc.) were applied. In addition, Thiazolyl Blue Tetrazolium Bromide (MTT), TRIzol (Invitrogen; Thermo Fisher Scientific, Inc.), and the quantitative polymerase chain reaction (qPCR) and protein extraction kits (cat. no. PROTTOT) were obtained from Sigma-Aldrich (Merck KGaA, Darmstadt, Germany). The PCR primers and ANXA9-siRNA were designed and synthesized by Sangon Biotech, Co., Ltd. (Shanghai, China). ANXA9 (cat. no. sc-373934), ADAM17 (cat. no. sc-390859), MMP-9 (cat. no. sc-12759), TIMP-1 (cat. no. sc-365905), E-cadherin (cat. no. sc-71008), N-cadherin (cat. no. sc-59987) and β-actin (cat. no. sc-8432) antibodies were obtained from Santa Cruz Biotechnology, Inc. (Dallas, TX, USA).

Paraffin specimens were deparaffinized and rehydrated, and immunohistochemical staining of surfactant proteins (S–P) (cat. no. sc-80621) was performed according to the manufacturer's instruction (Santa Cruz Biotechnology, Inc.) Five random visual fields (magnification, ×400; 100 cells per field) for each section were evaluated by pathologists. Expression levels of the ANXA9 protein were determined as positive if yellow or brown plasmids were observed in the cytoplasm or on the cell membrane. Expression positivity was scored as follows: i) Darkness of staining (transparent, 0; light yellow, 1; brownish-yellow, 2; and brown, 3; ii) ratio of positive to negative cells (positivity of 0%, 0; ≤10%, 1; 11–50%, 2; 51–75%, 3; and >75%). The two scores were added and ≤2 was considered to be negative (−) and >2 was considered to be positive (+).

Tissue and cell lysates were prepared with lysis buffer [1% Triton X-100, 150 mM NaCl, 10 mM Tris-HCl (pH 7.4), 1 mM ethylene diamine tetraacetic acid (EDTA), 1 mM ethylene glycol-bis(β-aminoethylether)tetraacetic acid (pH 8.0), 0.2 mM Na₃VO₄, 0.2 mM phenylmethylsulfonyl fluoride, and 0.5% NP-40]. The samples were rinsed in ice-cold lysis buffer for 20 min followed by centrifugation for 10 min at 7,104 × g at 4°C. The supernatant was collected and the bicinchoninic acid assay was performed for quantitation of protein. Equal quantities of protein (60 µg) from each sample were separated in 10% dodecyl sulfate, sodium salt-polyacrylamide gel electrophoresis gels and electrotransferred to polyvinylidene difluoride membranes (100 V, 2 h). The membranes were blocked in 5% non-fat milk for 2 h at room temperature, followed by incubation in diluted antibodies at 4°C overnight. The following antibodies were used: Mouse anti-ANXA9 (1:200, cat. no. sc-373934), mouse anti-ADAM17 (1:200, cat. no. sc-390859), mouse anti-MMP-9 (1:400, cat. no. sc-12759), mouse anti-TIMP-1 (1:200, cat. no. sc-365905), mouse anti-E-cadherin (1:200, cat. no. sc-71008), mouse anti-N-cadherin (1:800, cat. no. sc-59987), mouse anti-β-actin (1:200, cat. no. sc-8432) all Santa Cruz Biotechnology, Inc. Following three rinses with Tris-Hcl, NaCl and Tween-20 (TBST), blots were incubated with peroxidase-conjugated donkey anti-mouse antibody (1:2,000; cat. no. AB10085; Jackson ImmunoResearch Laboratories, Inc., West Grove, PA, USA) for 2 h at room temperature. After three rinses with TBST and one with TBS, the optical density (OD) of the bands was detected using an enhanced chemiluminescence detection system. The concentration of proteins in the samples was then determined by comparing the OD of the samples to the standard curve.

All cell lines were cultured in DMEM supplemented with 10% FBS, 100 U/ml penicillin and 0.1 mg/ml streptomycin. Cells were maintained at 37°C in an incubator saturated with 5% CO₂. Cells were dissociated with 0.25% trypsin containing 0.02% EDTA and were passaged. Cells in the exponential growth phase were used for the experiments.

Design and synthesis of the sequence of siRNA targeting ANXA9 was performed as below: siRNA, 5′-GCAGUCUACAAACACAAUUtt-3′ and non-specific control siRNA (NS-siRNA), 5′-UUCUCCGAACGUGUCACGUtt-3′. HCT116 cells were transplanted into 6-well plates 24 h prior to transfection (density of 1×10⁶ cells/ml). Plasmid transfection was performed using Lipofectamine™ 2000 according to the manufacturer's instructions after samples were washed with serum- and antibody-free DMEM. Efficiency of transfection and ANXA9 suppression was detected 24 h after transfection. Experimental samples were divided into three groups according to the transfection status, which were the Lipofectamine™ 2000 transfected group (blank group), NS-siRNA transfected group (NS-siRNA group) and the ANXA9-siRNA transfected group (ANXA9-siRNA group).

Total cellular RNA in the tissue specimens and cells of different groups was extracted using TRIzol (Invitrogen; Thermo Fisher Scientific, Inc.) according to the manufacturer's protocol. qPCR was performed in a total volume of 20 µl containing 1 µl reverse transcription product as a template for PCR, 2X UltraSYBR mixture (10 µl; Applied Biosystems; Thermo Fisher Scientific, Inc.), 10 µmol/l per 1 µl primer, 8 µl DNase/RNase-Free water. The primer sequences used in PCR are presented in Table I. PCR was performed over 35 cycles as follows: Initial denaturation at 95°C for 5 min, denaturation at 95°C for 30 sec, annealing at 60°C for 30 sec and elongation at 72°C for 30 sec. Fluorescence was detected at the end of each cycle. The specificity of the products was confirmed by melting curve analysis. GAPDH served as an endogenous reference to standardize relative expression levels for data analysis to calculate the expression levels.

Cells were incubated in 96-well plates at a density of 1×10⁵ cells/ml. When the cell density reached 70–80% confluence, ANXA9-siRNA or NS-siRNA was transfected. The cells were plated in 6 replicate wells per cell density. Following incubation at 37°C for 20 h, 20 µl (5 mg/ml) MTT was added for another 4-h incubation and discarded, followed by the addition of 150 µl DMSO in each well and gentle shaking at room temperature for 15 min. The OD value was measured at a wavelength of 490 nm using a microplate spectrophotometer. Each treatment was performed in triplicate.

HCT116 cells were formed into mono-layer suspension (density, 1×10⁶ cell/ml) and seeded in each well of 6-well culture plates. The cells were transfected with ANXA9-siRNA or NS-siRNA at 60–70% confluence and cultured until 100%. The culture medium was removed and the cells were rinsed with phosphate-buffered saline (PBS). This monolayer was then scored using a sterile pipette tip to form scratches and was rinsed with PBS again to remove any dislodged cells. Wound closure was visualized using a microscope. The procedures were performed in triplicate.

Each well in the upper chamber was coated with 100 µl Matrigel (BD Biosciences, Franklin Lakes, NJ, USA) under ultraviolet light. HCT116 cells were suspended and plated at 1×10⁶ cells/ml in 6-well plates. These cells were cultured until 60–70% confluence for transfection. After a 24-h incubation at 37°C, 200 µl cells were extracted from each group and plated in the upper chamber of a Transwell, while DMEM was added to the lower chamber. After removing any excess Matrigel and non-invading cells from the upper chamber, the Transwell membranes were fixed in methanol for 10 min and stained with crystal violet for 30 min at 37°C. Cells on the underside of the membranes that had invaded the Matrigel were counted under an inverted microscope (Carl Zeiss AG, Oberkochen, Germany). The treatment was repeated three times.

All of the data was analyzed by SPSS 26.0 statistical software (IBM Corp., Armonk, NY, USA). Quantitative data was represented as the mean ± standard deviation, and the deviation between groups was analyzed using one way analysis and Dunnett t-test and variation analysis. Categorical data were expressed as percentages and analyzed using the χ² test. Kaplan-Meier analysis and COX's proportional hazard regression model were utilized to investigate the prognostic factors of the ANXA9 protein. P<0.05 was considered to indicate a statistically significant difference.

The IHC result demonstrated that the positive rate of ANXA9 protein expression in CRC tissue samples was higher than that in the adjacent mucosa with 76.19% (80/105), and 16.19% (17/105), respectively (χ²=76.041; P<0.001), as illustrated in Fig. 1A. Similarly, the result of western blot analysis demonstrated that the ANXA9 expression level was higher in CRC tissues compared with the adjacent mucosa (P<0.001) (Fig. 1B).

The result demonstrated that a higher ANXA9 positive rate presented deeper-infiltration and lymphatic metastasis in the CRC tissue samples (P<0.05), and no significant correlation was identified between ANXA9 and other clinicopathological parameters (P>0.05). The results were illustrated in Table II.

The association between ANXA9 expression levels and prognosis was analyzed and presented using a Kaplan-Meier survival curve (Fig. 2). The data demonstrates that the overall survival rate was lower in the patients with positive ANXA9 expression compared with those with negative ANXA9 expression (P=0.005). According to Cox's proportional hazards regression model presented in Table III, the present study illustrated that ANXA9 expression level was an independent risk factor in CRC prognosis (P=0.022), and other independent risk factors, including lymphatic metastasis, differentiation and distant metastasis (P=0.017, 0.021 and 0.026, respectively).

As a result of western blotting, different levels of ANXA9 protein were detected in six CRC cell lines, among which the highest expression level of ANXA9 protein was demonstrated in HCT116 cells, and thus was selected for subsequent experiments (Fig. 3).

The result of western blotting demonstrated that after a 48-h transfection with 20 µmol/l ANXA9-siRNA, the ANXA9 expression level was downregulated more significantly when compared with the NS-siRNA and blank groups (Fig. 4).

The cell activity of the ANXA9-siRNA group varied with different concentrations and durations (Fig. 5A). Following transfection with ANXA9-siRNA (20 µmol/l) for 48 h, the cell activity of the ANXA9-siRNA group (49.64±5.82%) was significant lower than that in the NS-siRNA group (98.62±9.69%) and the blank group (100±11.24%; P<0.05), as demonstrated in Fig. 5B.

Results of the wound healing assay (Fig. 6A) and Transwell assay (Fig. 6B) demonstrate that, following ANXA9-siRNA transfection, the migration and invasion of HCT116 cells treated with ANXA9-siRNA were significantly decreased when compared with the NS-siRNA group and the blank group (P<0.05), as shown in Fig. 6.

Expression levels of ADAM17 and MMP-9 mRNA and proteins were significantly downregulated, while TIMP-1 and E-cadherin mRNA and protein expression levels were significantly upregulated in HCT116 cells following ANXA9-siRNA transfection (P<0.05), and no obvious variation was observed in N-cadherin following ANXA9-siRNA transfection (P>0.05; Fig. 7).