This study was approved by the Ethics Committee of First Affiliated Hospital of Harbin Medical University, and written informed consent was also obtained from each patient enrolled in this research. A total of 54 paired CRC tissues and normal adjacent tissues were obtained from CRC patients who received surgical resection at First Affiliated Hospital of Harbin Medical University between July 2014 and March 2016. These 54 CRC patients have not been treated with radiotherapy or chemotherapy prior to surgical resection. All tissue samples were immediately frozen in liquid nitrogen at the time of surgery and stored at −80°C until further use.

Five human CRC cell lines (LoVo, HCT116, HT29, SW480, SW620), normal human colon epithelium cell line FHC and human embryonic kidney 293T cells (HEK293T) were acquired from American Type Culture Collection (Manassas, VA, USA). All cells were routinely grown in Dulbecco's modified Eagle's medium (DMEM; Gibco; Thermo Fisher Scientific, Inc., Waltham, MA, USA) containing 10% foetal bovine serum (FBS; Gibco; Thermo Fisher Scientific, Inc., Waltham, MA, USA), 100 U/ml penicillin (Sigma, St. Louis, MO, USA) and 100 µg/ml streptomycin (Sigma). Cell cultures were kept in a humidified environment with 5% CO₂ at 37°C.

miR-337 mimics and corresponding negative control miRNA (miR-NC) were purchased from GenePharma, Co., Ltd. (Shanghai, China). Small interfering RNA targeting Kirsten rat sarcoma viral oncogene homolog (si-KRAS) and the corresponding control siRNA (si-NC) were purchased from Ribobio (Guangzhou, China). pCMV-KRAS and blank plasmid pCMV were obtained from GeneCopoeia (Guangzhou, China). Cells were plated in 6-well clusters at a density of 60–70% confluence. Cell transfection was performed using Lipofectamine 2000 (Invitrogen; Thermo Fisher Scientific, Inc. Waltham, MA, USA) according to the manufacturer's instructions. Transfected cells were then incubated at 37°C with 5% CO₂, and the culture medium was replaced with fresh DMEM supplemented with 10% FBS at 8 h post-transfection.

Total RNA from tissue samples or cells was isolated using TRIzol reagent (Invitrogen, Carlsbad, CA, USA) according to the manufacturer's instructions. The concentration of total RNA was examined using a NanoDrop 2000 spectrophotometer (Thermo Fisher Scientific, Madrid, Spain). To detect miR-337 expression, complementary DNA (cDNA) was prepared using TaqMan MicroRNA Reverse Transcription kit (Applied Biosystems, Carlsbad, CA, USA). Real-time PCR was then conducted using a TaqMan MicroRNA PCR kit (Applied Biosystems) on an ABI PRISM 7900 Sequence Detection system (Applied Biosystems). To quantify KRAS mRNA expression, total RNA was converted into cDNA using PrimeScript RT Reagent kit (Takara Biotechnology Co., Ltd., Dalian, China), followed by quantitative PCR with SYBR Premix Ex Taq™ (Takara Biotechnology Co., Ltd.). U6 and GAPDH served as internal control for miR-337 and KRAS, respectively. Primers used in this study were as follows: miR-337 forward, 5′-ACACTCCAGCTGGGCTCCTATATGATGC-3′ and reverse 5′-ACTCCACGACACCAGTTGAG-3′; U6 forward, 5′-CTCGCTTCGGCAGCACA-3′ and reverse 5′-AACGCTTCACGAATTTGCGT-3′; KRAS forward, 5′-ATTCCTTTTATTGAAACATCAGCA-3′ and reverse 5′-TCGGATCTCCCTCACCAAT-3′; GAPDH forward, 5′-GACTCATGACCACAGTCCATGC-3′ and reverse 5′-AGAGGCAGGGATGATGTTCTG-3′. Relative gene expression was calculated using comparative 2^−∆∆Ct method (22).

Cells were seeded at a density of 3×10³ cells/well in 96-well plates. After transfection, cells were grown at 37°C with 5% CO₂ for 0, 24, 48 and 72 h. At indicated time-point, CCK8 assay was performed to determine cell proliferation. In brief, 10 µl CCK-8 reagent (Dojindo Laboratories, Kumamoto, Japan) was added into each well and incubated for another 2 h at 37°C. The optical density (OD) value was detected using a plate reader (ELx808 Bio-Tek Instruments, USA) at a wavelength of 450 nm. Three independent experiments were performed in triplicate.

Matrigel (BD Biosciences, San Jose, CA, USA)-coated Transwell inserts containing a membrane with 8-µm pores (Costar; Corning Inc., Corning, NY, USA) were used to perform cell invasion assays. In brief, 1×10⁵ cells suspended in 200 µl FBS-free DMEM medium were plated on top of Transwell insert, whereas DMEM medium containing 20% FBS was placed in the lower chamber as a chemoattract. After 20-h incubation, cells remaining on the upper membrane of the filter were removed by wiping with a cotton swab. The invasive cells on the lower membrane surface were fixed in 20% methanol and stained with 0.1% crystal violet. Photographs were taken, and the number of invasive cells was counted in five randomly selected fields under an inverted microscope (×200; Olympus, Tokyo, Japan). The assay was independently repeated thrice.

After 48 h of transfection, transfected cells were harvested, washed with phosphate-buffered saline (PBS) and fixed with 80% ice-cold ethanol in PBS. Cell apoptosis was detected using Annexin V-FITC apoptosis detection kit (Invitrogen Corp., CA, USA). In brief, cells were re-suspended in 300 µl 1X binding buffer, followed by staining with 5 µl FITC-Annexin V and 5 µl propidium iodide for 20 min in the dark at room temperature. Cell apoptosis was recognised by a FACScan flow cytometer (Becton-Dickinson, San Jose, CA, USA), and data were analysed using CellQuest^® software (Becton-Dickinson, Heidelberg, Germany).

To determine the potential targets of miR-337, bioinformatics analysis was performed with the publicly available algorithms: TargetScan (http//www.targetscan.org), miRanda (http://www.microrna.org/microrna/getExprForm.do) and PICTA (http://pictar.mdc-berlin.de/).

Luciferase reporter vector containing the wild-type (pGL3-KRAS-3′-UTR-Wt) or mutant (pGL3-KRAS-3′-UTR-Mut) 3′-UTR of KRAS were synthesised and confirmed by GenePharma Co., Ltd. HEK293T cells were seeded into 24-well plates at a density of 60–70% confluence. After incubation overnight, cells were transfected with miR-337 mimics or miR-NC, together with pGL3-KRAS-3′-UTR-Wt or pGL3-KRAS-3′-UTR-Mut using Lipofectamine 2000. Cells were collected 48 h after transfection, and luciferase activities were examined using a dual-luciferase reporter system (Promega, Madison, WI, USA) according to the manufacturer's protocols. Firefly luciferase activities were normalised to Renilla luciferase activities. Each assay was performed in triplicate.

Total protein was extracted from tissue samples or cells using radioimmunoprecipitation assay cell lysis buffer (Beyotime, Shanghai, China). Protein concentrations were quantified using BCA assay kit (Pierce, Rockford, IL, USA). Equal amounts of protein were separated by 10% SDS PAGE, transferred onto PVDF membrane (Millipore, Billerica, MA, USA) and then blocked in Tris-buffered saline with Tween (TBST) buffer with 5% fat-free milk for 2 h at 37°C. Subsequently, the membranes were incubated with primary antibodies: mouse anti-human monoclonal KRAS (sc-30; 1:1,000 dilution; Santa Cruz Biotechnology, CA, USA), mouse anti-human monoclonal p-AKT (sc-271966; 1:1,000 dilution; Santa Cruz Biotechnology), mouse anti-human monoclonal AKT (sc-81434; 1:1,000 dilution; Santa Cruz Biotechnology), mouse anti-human monoclonal p-ERK (sc-81492; 1:1,000 dilution; Santa Cruz Biotechnology), mouse anti-human monoclonal ERK (sc-514302; 1:1,000 dilution; Santa Cruz Biotechnology), and mouse anti-human monoclonal GAPDH antibody (sc-47724; 1:1,000 dilution; Santa Cruz Biotechnology). Following washing in TBST, the membranes were further probed with goat anti-mouse horseradish peroxidase-conjugated secondary antibody (sc-2005; 1:5,000 dilution; Santa Cruz Biotechnology) at room temperature for 1 h. Protein bands were visualised by applying ECL Protein Detection kit (Pierce Biotechnology, Inc., Rockford, IL, USA) and analysed using Quantity One software (Bio-Rad Laboratories, Inc., Hercules, CA, USA). GAPDH was used as the loading control.

Data are expressed as the mean ± standard deviation. All statistical analyses were performed using Student's t-test or one way ANOVA test with SPSS software (version 18.0; SPSS, Inc., Chicago, IL, USA). Student-Newman-Keuls was used to compare between two groups in multiple groups. P<0.05 was considered to indicate a statistically significant difference.

To investigate the expression pattern of miR-337 in CRC, we measured its expression in 54 paired CRC tissues and normal adjacent tissues using RT-qPCR. As shown in Fig. 1A, miR-337 was significantly downregulated in CRC tissues compared with that in normal adjacent tissues (P<0.05). Subsequently, we explored the association between miR-337 expression and clinicopathological characteristics of CRC patients. The colorectal cancer tissues were divided into either the miR-337 low-expression group (n=28) or the miR-337 high-expression group (n=26) using the median expression level of miR-337 as a cut-off. As shown in Table I, low expression levels of miR-337 were significantly correlated with lymph node metastasis (P=0.002), distant metastasis (P=0.014) and TNM stage (P<0.0001, significant difference), whereas no correlation was found with patient gender (P=0.276), age at diagnosis (P=0.177), location (P=0.394) or tumour size (P=0.279).

Moreover, miR-337 expression level was also detected in CRC cell lines (LoVo, HCT116, HT29, SW480, SW620) and normal human colon epithelium cell line FHC. The expression level of miR-337 was reduced in CRC cell lines compared with that in FHC (Fig. 1B, P<0.05). These results suggested that miR-337 may play important roles in CRC formation and progression.

To observe the effects of miR-337 on CRC cells, miR-337 mimics or miR-NC were transfected into HCT116 and SW480 cells. After transfection, RT-qPCR was performed to evaluate the transfection efficiency and found that miR-337 was markedly upregulated in HCT116 and SW480 cells after transfection with miR-337 mimics (Fig. 2A, P<0.05). CCK-8 assay was utilised to examine the effect of miR-337 overexpression on CRC cell proliferation and revealed that upregulation of miR-337 obviously suppressed the HCT116 and SW480 cell proliferation (Fig. 2B, P<0.05). To evaluate the effect of miR-337 on the invasion capacity of CRC cells, cell invasion assay was applied in HCT116 and SW480 cells which were transfected with miR-337 mimic or miR-NC. As shown in Fig. 2C, miR-337 overexpression decreased the invasive abilities of HCT116 and SW480 cells (P<0.05). Moreover, we used flow cytometry analysis to explore the effect of miR-337 on CRC cell apoptosis. Results showed that the restoration expression of miR-337 increased cell apoptosis in HCT116 and SW480 cells (Fig. 2D, P<0.05). These data suggested that miR-337 may serve as a tumour suppressor in CRC.

To elucidate the underlying mechanisms by which miR-337 affects the biological functions of CRC cells, we investigated the potential targets of miR-337 using bioinformatics analysis. Among these candidate genes, KRAS was selected for further validation because it was aberrantly observed in CRC and plays key roles in CRC occurrence and development (23). As shown in Fig. 3A, two miR-337 putative binding sites were identified covering the nucleotides 3532-3528 (site 1) and 4448-4454 (site 2) of KRAS 3′-UTR. To assess whether miR-337 could directly target the 3′-UTR of KRAS, luciferase reporter assay was conducted in HEK293T cells cotransfected with luciferase reporter vector and miR-337 mimics or miR-NC. Results revealed that upregulation of miR-337 led to a significant decrease in pGL3-KRAS-3′-UTR site 1 Wt and pGL3-KRAS-3′-UTR site 2 Wt luciferase activity in HEK293T cells compared with miR-NC cells (Fig. 3B, P<0.05). Mutation of the two miR-337 binding sites (site 1 Mut or site 2 Mut) restored the normal luciferase activity of KRAS-3′-UTR in HEK293T cells.

To further confirm whether miR-337 can affect the endogenous level of KRAS, miR-337 mimics or miR-NC were transfected into HCT116 and SW480 cells to investigate the mRNA and protein levels of KRAS using RT-qPCR and western blot analysis. Results showed that enforced expression of miR-337 significantly reduced the mRNA (Fig. 3C, P<0.05) and protein (Fig. 3D, P<0.05) levels of KRAS in HCT116 and SW480 cells. Overall, KRAS gene is a direct target of miR-337 in CRC.

To further examine the association between miR-337 and KRAS, we measured KRAS mRNA and protein expression in CRC tissues and normal adjacent tissues using RT-qPCR and western blot analysis. KRAS was dramatically upregulated at both mRNA (Fig. 4A, P<0.05) and protein (Fig. 4B, P<0.05) levels in CRC tissue compared with normal adjacent tissues. Spearman's correlation analysis further demonstrated that miR-337 expression level was negatively correlated with the expression level of KRAS mRNA in CRC tissues (Fig. 4C; r=-0.6648, P<0.0001).

KRAS was identified as a direct target of miR-337. Therefore, we hypothesised that miR-337 affecting the biological functions of CRC cells might be achieved by KRAS knockdown. To confirm this hypothesis, si-KRAS was used to decrease KRAS expression level in HCT116 and SW480 cells. After transfection, western blot analysis showed that KRAS protein expression was obviously downregulated in HCT116 and SW480 cells transfected with si-KRAS (Fig. 5A, P<0.05). Functional assays demonstrated that KRAS knockdown could repress proliferation (Fig. 5B, P<0.05) and invasion (Fig. 5C, P<0.05) and promoted apoptosis (Fig. 5D, P<0.05) of HCT116 and SW480 cells, which was similar to the effects induced by miR-337 overexpression. The tumour-suppressing effects of miR-337 on CRC cells depends, at least in part, on its direct target KRAS.

To further evaluate whether KRAS is responsible for the functional roles of miR-337 in CRC, rescue experiments were performed in HCT116 and SW480 cells transfected with miR-337 mimics with or without pCMV-KRAS. Western blot analysis showed that KRAS protein was obviously inhibited by miR-337 overexpression in HCT116 and SW480 cells, whereas upregulation of KRAS by pCMV-KRAS abolished the suppression of KRAS caused by miR-337 mimics (Fig. 6A, P<0.05). Function investigation revealed that upregulation of KRAS reversed the tumour-suppressing effects of miR-337 on CRC cell proliferation (Fig. 6B, P<0.05), invasion (Fig. 6C, P<0.05) and apoptosis (Fig. 6D, P<0.05). miR-337 served as a tumour suppressor in CRC, at least in part, by directly targeting KRAS.

Previous studies reported that activation of KRAS can trigger several significant pathways, including the AKT and ERK pathways (24,25). Therefore, we measured the protein expression level of p-AKT, AKT, p-ERK and ERK in HCT116 and SW480 cells after transfection with miR-337 mimics or miR-NC. The data of western blot analysis indicated that ectopic expression of miR-337 reduced the p-AKT and p-ERK expression levels and negatively regulated AKT and ERK pathways (Fig. 7). These results suggested that miR-337 exerts tumour-suppressing roles in CRC cells by directly targeting KRAS and indirectly regulating AKT and ERK pathways.