A total of 19 patients diagnosed with CRC between March 2015 and November 2019 at the First Affiliated Hospital of Jinan University (Guangzhou, China) were included in the present study. Patient age ranged from 49 to 86 years (mean age, 63.25 years; nine male patients and 10 female patients). Patients did not receive preoperative radiotherapy or chemotherapy and provided signed informed consent. Cancer and adjacent normal tissue samples (>10 cm away from the tumor) were stored in liquid nitrogen. This study was approved by the Ethics Committee of the First Affiliated Hospital of Jinan University (approval no. 2018038) and performed in accordance with the Declaration of Helsinki.

The immortalized cell line NCM460 that was used as the normal control and the human CRC cell lines HT29, SW620, LoVo and HCT116 were purchased from the American Type Culture Collection (STR profiling authenticated). Cells were cultured in RPMI 1640 (Beijing Solarbio Science & Technology Co., Ltd.) supplemented with 10% FBS (Gibco; Thermo Fisher Scientific, Inc.), 1% penicillin (Gibco; Thermo Fisher Scientific, Inc.), and 1% streptomycin (Gibco; Thermo Fisher Scientific, Inc.) and placed at 37˚C in a humidified incubator containing 5% CO₂. HT29 and SW620 cells were transfected with the miR-146b-5p mimic (double strand, 5'-UGAGAACUGAAUUCCAUAGGCU-3', 200 nM; Thermo Fisher Scientific, Inc.) or miR-146b-5p inhibitor (5'-GCCUAUGGAAUUCAGUUCUC-3', 200 nM; Thermo Fisher Scientific, Inc.) using Lipofectamine^™ 3000 (Invitrogen; Thermo Fisher Scientific, Inc.). For miR negative control (NC), scramble NC (5'-GUGUAACACGUCUAUACGCCCA-3'; 200 nM), mimic NC (double strand, 5'-UUCUCCGAACGUGUCACGUTT-3'; 200 nM; Thermo Fisher Scientific, Inc.) and inhibitor NC (CAGUACUUUUGUGUAGUACAA; 200 nM; Thermo Fisher Scientific, Inc.) were used. Transfection was performed at 37˚C, after which the cells were harvested after 24 h.

To generate the TRAF6 overexpression stable cell line, a 2nd-generation lentiviral system was used. Briefly, TRAF6 was amplified by PCR with the following primers: Forward, with XhoI + kozak sequence, 5'-GGACTCTCGAGGCCATGGAGTCTGCTAAACTGT-3' and reverse, with Xbal sequence- 5'CTTGTCTAGATGCAGTCGTCGAGGAATTGCTAT-3'). Subsequently, TRAF6 cDNA was inserted into a pEGFP-C1 vector (Takara Bio USA, Inc.) to generate a GFP-TRAF6 overexpression plasmid; the pEGFP-C1 vector was used as a control. The product was then inserted into the lentivirus pLVX-puro vector (Takara Bio, Inc.). This construct, together with pHelper 1.0 vector (Shanghai GeneChem Co., Ltd.) encoding the gag, pol and rev genes, and with pHelper 2.0 (Shanghai GeneChem Co., Ltd.) encoding the VSV-G gene were transfected into 293T cells (Shanghai GeneChem Co., Ltd.) for packaging for 72 h, 37˚C. Then, the lentiviruses were harvested by ultracentrifugation (Beckman SW32ti; Beckman Coulter) at 120,000 x g for 2 h at 4˚C. HT29 cells were infected with a multiplicity of infection of 10 for 48 h, then cultured in the presence of puromycin (Sangon Biotech Co., Ltd.; 2 µg/ml) for 2 weeks to screen cells stably overexpressing TRAF6. The overexpression was verified by western blotting.

Total RNA was extracted from cells or tissue samples using a TaqMan^™ MicroRNA Cells-to-CT^™ Kit (Thermo Fisher Scientific, Inc.) and was reverse-transcribed into cDNA also using this kit. The cDNA was mixed with primers and TaqMan Master Mix (2x; Thermo Fisher Scientific, Inc.) provided the kit. The sequences of the primers were as follows: TRAF6 forward, 5'-CTATTCACCAGTTAGAGGG-3', reverse, 5'-GCTCACTTACATACATACT-3' and miR-146b-5p forward, 5'-CCTGGCACTGAGAACTGAAT-3' and reverse, 5'-GCACCAGAACTGAGTCCACA-3'. U6 (forward, 5'-CGCTTCGGCAGCACATATACTA-3'; reverse, 5'-GCGAGCACAGAATTAATACGAC-3') and GAPDH (forward, 5'-GAGTCCACTGGCGTCTTC-3'; reverse, 5'-GGGGTGCTAAGCAGTTGGT-3') were used as controls. RT-qPCR reactions were performed as follows: 95˚C for 5 min, followed by 40 cycles at 95˚C for 5 sec and 60˚C for 1 min using the ABI 7500 Real-Time PCR System (Applied Biosystems). The relative expression levels were normalized to endogenous control and were expressed as 2^-ΔΔCq (24). RT-qPCR for miR-146b-5p was performed using the miRNA Detection Kit (Qiagen, Inc.). Experiments were performed in triplicate.

HT29 and SW620 cells were seeded on a 6-well plate at the density of 1x10⁵ cells/well. After 24 h of transfection, the cells were transferred to a 96-well plate and incubated for 24, 48, 72 and 96 h. Subsequently, CCK-8 solution (Beyotime Institute of Biotechnology; 10% v/v) was added to the cells for 4 h at 37˚C. Absorbance was measured at 450 nm using a microplate reader (Benchmark Plus; Bio-Rad Laboratories, Inc.). The cell viability was analyzed after 24, 48, 72 and 96 h.

A total of 1x10⁵ HT29 and SW620 CRC cells were digested and suspended in serum-free medium. Cells were seeded into the upper chamber of a 24-well plate, either precoated with Matrigel (for 2 h at 37˚C) or uncoated, and medium containing 10% FBS was added to the lower chamber. After 24 h at 37˚C, cells in the lower chamber were fixed with 4% paraformaldehyde for 40 min at 4˚C and stained with 0.1% crystal violet at room temperature for 30 min. Cells were observed under a light microscope and the number of cells that have invaded or migrated the lower chamber were counted.

For the colony formation assay, transfected or control HT29 and SW620 CRC cells were suspended in culture medium and were seeded on a 0.5% soft agar layer at the density of 2x10⁴ cells per well in 6-well plates. After 2-3 weeks, the supernatant was harvested and cells were stained with 4 mg/ml crystal violet (Beyotime Institute of Biotechnology) for 30 min at room temperature. The number of colonies was counted using a light microscope (magnification, x5).

CRC and adjacent non-tumor tissue samples were fixed with 4% formaldehyde (Beyotime Institute of Biotechnology) at room temperature for 12 h, embedded in paraffin and sliced into 5-7-µm sections. Sections were dewaxed with xylene and hydrated in a descending gradient of alcohol solutions from 100 to 75%. Antigen retrieval was performed by heating samples in citrate buffer (pH 6.0) for 5 min in a microwave. Sections were treated with 3% hydrogen peroxide for 10 min at room temperature to remove endogenous peroxidase and were incubated with primary antibody against TRAF6 (Abcam; cat. no. ab33915; 1:500) at 4˚C overnight and with a secondary antibody (Jackson ImmunoResearch Laboratories, Inc.; cat. no. 111-035-003; 1:1,000) at room temperature for 1 h. The images were analyzed using ImageJ version 1.8 (National Institutes of Health). For scoring TRAF6, the frequency of staining was determined using the following scale: 0, no or hardly any cells positive; 1, small fraction of cells positive; 2, approximately half of the cells positive, 3=more than half of the cells positive; 4, all or the majority of cells positive.

Proteins were extracted from CRC tissues and cells using RIPA lysis buffer (Beyotime Institute of Biotechnology) at room temperature, and the protein concentration was measured using a BCA kit (Beyotime Institute of Biotechnology). Proteins (30 µg protein/lane) were separated by SDS-PAGE on 8% gels and transferred onto a PVDF membrane as previously described (25). Membranes were blocked with 5% skimmed milk powder at room temperature for 1 h and incubated with primary antibodies against TRAF6 (monoclonal antibody; cat. no. ab33915; 1:1,000), cleaved PARP (cat. no. ab32064; 1:1,000), cleaved caspase-3 (cat. no. ab32042; 1:500), GFP (cat. no. ab290; 1:1,000) and GAPDH (cat. no. ab8245; 1:2,000) (all from Abcam), at 4˚C overnight. Membranes were then incubated with anti-mouse and anti-rabbit secondary antibodies (cat. nos. 115-035-003 and 111-035-003; 1:2,000; Jackson ImmunoResearch Laboratories, Inc.). Subsequently, the blots were visualized using an ECL kit (cat. no. P0018M; Beyotime Institute of Biotechnology) and the densitometry data were analyzed by ImageJ version 1.8.

To predict the potential binding interaction between miR-146b-5p and TRAF6, TargetScan (26) and starBase (27) tools were applied. In starBase (http://starbase.sysu.edu.cn/index.php), we used ‘miRNA-mRNA’ analysis with the input of miRNA (has-miR-146b-5p) and the target of TRAF6. The results showed significant interactions between miR-146b-5p and TRAF6 in multiple databases, such as RNA22(28), miRmap (29), PicTar (https://pictar.mdc-berlin.de/) and TargetScan. Subsequently, the results were verified in Targetscan (http://www.targetscan.org/vert_72/) with the input of human gene symbol ‘TRAF6’ and the matching sequence was shown. For the mRNA expression of TRAF6 in different types of cancer, the GEPIA tool (30) was used, with the search term of ‘TRAF6’ for expression profile, the parameters selected were: ‘ANOVA analysis of log-scale’ and ‘match TCGA normal and GTEx data’.

A potential binding site between miR-146b-5p and TRAF6 was predicted by starBase v2.0(27). The 3' UTR sequence of TRAF6 mRNA containing the wild-type or mutant miR-146b-5p-binding site (TRAF6 3'-UTR-wt and TRAF6 3'-UTR-mut, respectively; ~300 nucleotide long both) was synthesized by Shanghai GenePharma Co., Ltd. and was cloned into the pmirGLO vector (Promega Corporation). Cells were co-transfected with the negative control and miR-146b-5p mimic with Lipofectamine 3000. The dual luciferase reporter gene assay was conducted to detect luciferase activity (firefly and Renilla luciferase activities) after 48 h in a BioTek Synergy 2 luminometer (BioTek Instruments Inc.) according to the manufacturer's instructions. Firefly luciferase activities were normalized against Renilla luciferase activities to determine relative fluorescence intensity.

HT29 cells stably overexpressing TRAF6 or transfected with an empty vector were subcutaneously injected into 15 immunodeficient BALB/c nude mice (Laboratory Animal Center of Jinan University; male; age, 4-6 weeks; weighing, 18-22 g); five mice per group; 2 mice were found dead in the miR-146b-5p group because of tumor burden; thus, 3 mice are shown for each group. The mice were housed in specific pathogen free facilities under a 12 h light/dark cycle, and temperature-(25˚C) and humidity-controlled (50-60%) conditions. The miR-146b-5p mimic or negative control was injected into the tumor four times with 1-week interval. When tumor volume reached 50 mm³, the volume was measured every 5 days. After ~25 days, when tumors reached 15x15 mm in size in the control group, the mice were euthanized by intraperitoneal injection of 4% pentobarbital (180 mg/kg) and subcutaneous tumors were collected and weighed. Death was confirmed by cervical dislocation. The related experiment was performed at the Laboratory Animal Center. The animal experiments were approved by the Laboratory Animal Ethics Committee of Jinan University (approval no. 2019231). All animal procedures followed the guidelines issued by the China Animal Protection Association.

The data are presented as the mean ± standard deviation and analyzed using SPSS 21.0 statistical software (IBM Corp.). Cancer and adjacent normal tissue samples were compared using paired t-tests. Differences between two groups were compared using unpaired t-tests, and differences between multiple groups were determined using one-way ANOVA followed by Tukey's post hoc test. To assess correlation, the Pearson correlation analysis was performed. P<0.05 was considered to indicate a statistically significant difference.

We evaluated the miR-146b-5p expression level in tissue sample from patients with CRC by RT-qPCR analysis (Fig. 1A). Expression of miR-146b-5p was significantly upregulated in CRC compared with normal tissue. Furthermore, RT-qPCR analysis revealed that miR-146b-5p was overexpressed in CRC cells (Fig. 1B) compared with the normal cell line. miR-146b-5p may therefore be involved in CRC progression.

To investigate the function of miR-146b-5p in CRC cells, two CRC cell lines (HT29 and SW620) were transfected with an miR-146b-5p mimics and inhibitor. All the controls (scramble NC, inhibitor NC and mimic NC) showed no effect on the expression of miR-146b-5p in both cell lines (Fig. S1), thus scramble NC was used for the following experiments. The results from RT-qPCR confirmed the efficiency of transfection with miR-146b-5p mimic and inhibitor (Fig. 1C and D). The inhibitory efficiency was 60% in the inhibitor-transfected group compared with the control group (Fig. 1D). Furthermore, the results from colony formation assay demonstrated that miR-146b-5p mimics significantly promoted whereas miR-146b-5p inhibitor significantly inhibited the colony growth of HT29 (Fig. 1E) and SW620 cells (Fig. 1F). To confirm the impact of miR-146b-5p on the expression of protein related to the cell apoptotic pathway, cell lysates were subjected to western blotting. As presented in Fig. 2A and B, the miR-146b-5p inhibitor markedly increased the expression of cleaved PARP and cleaved caspase-3 proteins, which are markers of apoptosis. Furthermore, the results from CCK-8 assays demonstrated that proliferation of HT29 and SW620 cells was significantly increased by the miR-146b-5p mimics, but significantly decreased by the miR-146b-5p inhibitor (Fig. 2C and D). The Transwell assays revealed that miR-146b-5p mimics significantly increased cell migration and invasion, whereas miR-146b-5p inhibitor decreased them (Fig. 2E-H). Taken together, these results indicated that miR-146b-5p may promote the development of CRC.

According to TCGA and GTEx databases, TRAF6 was downregulated in CRC tissue samples; compared with in the normal tissue group, the TPM value of TRAF6 was significantly decreased in the colon adenocarcinoma tumor group (Fig. 3A). Immunohistochemistry was then performed to evaluate the expression of TRAF6 in CRC and normal tissue samples (Fig. 3B) and western blotting was used to determine TRAF6 expression in CRC cell lines (Fig. 3C). The results demonstrated that TRAF6 expression was significantly decreased in CRC tissue and CRC cell lines compared with normal tissue and cell lines, respectively (Fig. 3B and C).

Correlation analysis revealed a negative correlation between TRAF6 expression and miR-146b-5p expression in the collected CRC tissues (Fig. 4A). Furthermore, bioinformatics analysis was performed using Targetscan (26) and starBase (27) and found that a sequence of 3' UTR of TRAF6 matched miR-146b-5p (Fig. 4B), indicating that TRAF6 may be a target of miR-146b-5p. To investigate whether miR-146b-5p would bind to TRAF6 in CRC cells, the sequence of TRAF6 containing the potential binding site was cloned into the pmirGLO vector. The results demonstrated that luciferase activity was decreased upon transfection of TRAF6 3'-UTR-wt and miR-146b-5p but was unaffected upon transfection of TRAF6 3'-UTR-mut and miR-146b-5p (Fig. 4B-D) in both HT19 and SW620 cells. Furthermore, TRAF6 expression was significantly decreased following transfection with miR-146b-5p mimics and significantly increased after transfection with miR-146b-5p inhibitor (Fig. 4E and F). These findings demonstrated that TRAF6 may be a target of miR-146b-5p.

To further elucidate the role of the miR-146b-5p-TRAF6 axis in CRC, CRC cells were transfected with a TRAF6-expressing vector and with miR-146b-5p. The expression of GFP-TRAF6 in both cell lines was determined by western blotting with GFP antibody and the overexpression was confirmed (Fig. S2). As presented in Fig. 5A and B, TRAF6 overexpression significantly inhibited the effect of miR-146b-5p on colony formation. Western blotting demonstrated that TRAF6 overexpression induced the cleavage of PARP and caspase-3 (Fig. 5C and D). In addition, CCK-8 (Fig. 5E and F) and Transwell (Fig. 5G and H) assays demonstrated that overexpression of miR-146b-5p potentiated the malignant characteristics of CRC cells, and that these effects were abrogated following TRAF6 overexpression.

To determine the interaction between miR-146b-5p and TRAF6 in vivo, HT-29 cells stably overexpressing TRAF6 were generated via lentivirus infection. The western blotting results demonstrated that the transfection was efficient (Fig. 6A). A murine xenograft tumor model was then established to evaluate the role of the miR-146b-5p-TRAF6 axis in tumor growth. Tumor samples were harvested 25 days after tumor cell injection. As presented in Fig. 6B, miR-146b-5p overexpression significantly promoted tumor growth. Furthermore, miR-146b-5p expression in each group was determined by RT-qPCR (Fig. 6C) and tumor weight (Fig. 6D) and volume (Fig. 6E) were evaluated. The data showed that miR-146b-5p significantly promoted tumor growth in vivo, whereas overexpression of TRAF6 significantly abolished the tumor growth induced by miR-146b-5p (Fig. 6C-E). The results demonstrated that the effects of miR-146b-5p were reversed in cells stably overexpressing TRAF6 (Fig. 6). Taken together, these findings demonstrated that miR-146b-5p may target the tumor suppressor TRAF6 to promote CRC.