Tumor samples were collected from 40 patients with colorectal carcinoma that were diagnosed by endoscopic biopsy and surgically treated in Taizhou First People’s Hospital (Taizhou, China) between December 2008 and April 2009. The diagnoses of colorectal cancer in all patients were confirmed after surgery by histopathological examination according to World Health Organization standards (34). The clinicopathological features of the patients are listed in Table I. None of the patients had received preoperative radiotherapy or chemotherapy and all patients received strict chemotherapy following surgery according to the colorectal carcinoma treatment guidelines stipulated by National Comprehensive Cancer Network (35). Tumor samples were paired with normal tissues that had been resected from a location 2 cm away from the edge of the tumor. Each sample was divided into two further samples. One was cryopreserved in liquid nitrogen within 30 min of removal for RNA extraction, and the other was formalin-fixed and embedded in paraffin for sectioning and immunohistochemical analysis. Informed written consent was obtained from each patient and the study was approved by the Human Research Ethics Committee of Taizhou First People’s Hospital.

Total RNA extraction was conducted using TRIzol reagent (Invitrogen Life Technologies, Carlsbad, CA, USA) according to the manufacturer’s instructions. In order to remove any DNA, the extracted RNA was processed using DNase I (Takara Bio, Inc., Shiga, Japan) with phenol-chloroform extraction (Kangwei Biological Company, Beijing, China). The purified RNA was dissolved in diethylpyrocarbonate water (Takara Bio, Inc.) and its concentration was measured with an ultraviolet spectrophotometer (DU640; Beckman Coulter, Miami, FL, USA), prior to storage at −80°C. Extracted RNA (1 μg) was taken from each sample for reverse transcription using Oligo dT primers (ReverTra Ace qPCR RT kit, Toyobo, Osaka, Japan). The reaction conditions were as follows: 42°C for 60 min, 75°C for 15 min, followed by −20°C at the end of the reaction. qPCR was conducted based on a 15 μl reaction system. Table II lists the primer sequences used in the analysis. The quantitative fluorescent PCR reaction system consisted of the following: 1 μl reverse transcription product, 1X SYBR-Green I Mastermix (Toyobo), and 0.5 μmol/l each of the specific forward and reverse primers. The reaction conditions were as follows: 40 cycles of 95°C for 2 min, 95°C for 15 sec and 60°C for 1 min. qPCR was conducted using the CFX96 Real-Time PCR Detection system (Bio-Rad, Hercules, CA, USA). All samples were run in triplicate to increase reliability of results. Based on the Ct values of the specimens using GAPDH as the normalizer gene, the relative quantitative method of qPCR was adopted. N=2^−ΔΔCt was considered to represent the CXCL14 relative expression level in specimens, where, ΔΔCt=(Ct_CXCL14−Ct_GAPDH) tumor − (Ct_CXCL14−Ct_GAPDH) normal.

Formalin-fixed specimens embedded in paraffin were sectioned into 4-μm samples. The slides were processed with 0.1% poly-L-lysine and tissues were dewaxed using dimethyl benzene followed by immersion in distilled water. Endogenous peroxidase activity was inhibited by incubation in a 0.3% hydrogen peroxide bath for 10 min followed by three washes with 0.01 mol/l phosphate-buffered saline (PBS; pH 7.4). Slides were then immersed in citrate antigen retrieval buffer (Zhongshan Golden Bridge Biotechnology, Beijing, China) for 1.5 min. Once the slides were cooled to room temperature, they were washed with PBS three additional times, followed by blocking in sheep serum for 2 h. The sections were incubated in primary anti-CXCL14 antibody (Abcam, Cambridge, MA, USA) diluted to 1:500 in PBS, in a humidified chamber at 4°C overnight. Sections were then incubated in horseradish peroxidase (HRP)-conjugated secondary goat anti-mouse antibodies (MaiXin Bio, Fuzhou, China) at 37°C for 30 min. The color reaction was developed using a 3,3′-diaminobenzidine kit (Zhongshan Golden Bridge Biotechnology) according to the manufacturer’s instructions. Slides were then counterstained using hematoxylin, dehydrated and sealed with neutral gum. In the negative control group, PBS was used in place of the primary antibody. As CXCL14 is located in the plasma, cells stained with brown cytoplasm were considered as positive for CXCL14. Image-Pro Plus, version 6.0 image processing software (Media Cybernetics, Rockville, MD, USA) was used to calculate the mean optical density for further statistical analyses.

In order to facilitate the CXCL14 expression analysis, the study adopted the pLenti6.3_MCS_IRES-EGFP lentivirus expression vector (Invitrogen Life Technologies) with an internal ribosome entry site (IRES). The full-length sequence of CXCL14 (accession no. BC003513.1) was synthesized by Invitrogen and cloned into the plasmid of pLenti6.3_MCS_IRES-EGFP using Nhel and Ascl (Toyobo) double enzyme digestion. The pLenti6.3_CXCL14_IRES-EGFP recombinant plasmid was then identified by CXCL14 gene-specific PCR and sequencing analysis. The plasmid was extracted using a PureLink HiPure Plasmid Maxiprep kit (Invitrogen Life Technologies) in preparation for the lentivirus package. To obtain the lentivirus, 293T cells (ATCC, Manassas, VA, USA) at logarithmic phases were transformed with the packaged plasmids of pLP1, pLP2, pLP/VSVG and pLenti6.3_CXCL14_IRES-EGFP lentivirus expression plasmid at a ratio of 3:1 according to manufacturer’s instructions of Lipofectamine 2000 (Invitrogen Life Technologies). After 48 h, the cell culture supernatant was collected and centrifuged at 3,000 × g for 10 min to remove the residual cells and fragments. The virus supernatant was then ultracentrifugated at 50,000 × g for 2 h, and the concentration was resuspended in Opti-MEM solution (Invitrogen, Life Technologies). Following titer determination, the virus solution was stored at −80°C.

HT29 colorectal carcinoma cells (ATCC) were transfected by CXCL14 lentivirus with a multiplicity of infection (MOI) of 20, and total protein was extracted with a protein extraction kit (Shanghai Biyuntian Biological Co., Ltd, Shanghai, China) following 48 h of infection. Extracted protein (20 μg) was then separated by 10% SDS-PAGE gel electrophoresis and transferred onto polyvinylidene fluoride membranes (Millipore, Billerica, MA, USA). After blocking with 5% non-fat milk for 2 h, the membranes were then incubated with the CXCL14 monoclonal antibody (1:100,000) at 4°C overnight. Following three cycles of washing with Tris-buffered saline and Tween-20, HRP-conjugated goat anti-mouse antibodies (Beyotime Biotech, Shanghai, China) were incubated for 2 h at room temperature. A BeyoECL Plus (Beyotime Biotech) detection system was used for analysis of the final chemiluminescence reaction. The samples of HT29 colorectal carcinoma cell lines transfected with the control lentivirus were regarded as negative controls and β-actin served as an internal reference gene.

HT29 colorectal carcinoma cell lines were cultured in a 96-well plate and were transfected with CXCL14 lentivirus at an MOI of 20. Following 48-h culture, the activity of the cells was detected with a cell counting kit-8 assay kit (Beyotime Biotech). The absorbency was measured with a spectrophotometer (BioTek, Winooski, VT, USA) at a wavelength of at 450 nm according to the manufacturer’s instructions, and cell viability was calculated.

The transfected HT29 cells were digested with 0.25% pancreatic enzymes, then centrifuged at 800 × g for 10 min. Following washing, the cells were infiltrated and fixed with 70% precooled ethyl alcohol at 4°C overnight. After washing twice with PBS, the cells were digested by RNase-A (Toyobo) at 37°C for 30 min. The cells were stained with propidium iodide in an ice bath in the dark for 30 min, then filtered. Flow cytometry (BD Biosciences, San Jose, CA, USA) was utilized for analytic detection (Beckman Coulter, Brea, CA, USA).

The Wilcoxon signed-rank test was adopted for the comparison of paired samples; the Mann-Whitney U test was applied for the comparison of two groups of independent samples; the Kruskal-Wallis H test was applied for the comparison of three or more groups of independent samples; the χ² test was adopted for the comparison of inter-group rates; and the association between CXCL14 protein expression and prognosis was analyzed with a Kaplan-Meier survival curve. P<0.05 was considered to represent a statistically significant difference. All the analyses were conducted using SPSS, version 19.0 (IBM SPSS, Armonk, NY, USA).

The expression levels of CXCL14 mRNA in colorectal carcinoma were analyzed using RT-qPCR, with GAPDH as an internal reference. The melting curves of CXCL14 and GAPDH were sharply defined curves with a narrow peak, indicating that the established PCR method effectively amplified the target genes (Fig. 1A). As seen in Fig. 1B, the relative level of CXCL14 expression in tumor specimens was 0.502, which was significantly lower than the level in paired normal tissues (P<0.01). The CXCL14 protein expression level was further analyzed by immunohistochemical analysis. The expression of CXCL14 protein was observed in the cytoplasm (Fig. 1C). In normal tissues, there was strong positive staining, while the tumor tissues expressed CXCL14 at a low level, if at all. Semi-quantitative analyses of immunohistochemical results indicated that the expression level of CXCL14 in colorectal carcinoma tissue was lower than that in normal tissues (Fig. 1D, P<0.01). The mean optical densities of CXCL14 in tumor specimens and normal cells were 0.5411 and 0.1769, respectively.

Further analysis of the correlation between CXCL14 expression and the clinicopathological features of colorectal carcinoma indicated that the expression of CXCL14 protein is significantly correlated with tumor location, lymphatic metastasis and clinicopathological stages (P<0.05; Table II). The relative expression level of CXCL14 in colorectal carcinoma tumor tissue with lymphatic metastasis was 0.146 (0.006~0.061), which is significantly lower than the level in tumor tissue without lymphatic metastasis [0.215 (0.004~0.120)] (P<0.05). The relative expression levels in colorectal carcinoma tissues at pathological stages III and IV were significantly lower than those of specimens at pathological stages I and II (P<0.05), whose relative expression levels were 0.146 (0.005~0.061) and 0.215(0.004~1.200), respectively. No significant correlation was identified between the expression levels of CXCL14 and age, gender, infiltration degree or tumor marker (P>0.05). According to the relative values of the mean optical density of CXCL14 proteins in paired tumor and normal tissues, specimens with a relative value ≥1/3 were set as the high expression group and specimens with a relative value <1/3 were denoted as the low expression group. Kaplan-Meier survival analysis demonstrated that the five-year survival rate of the high expression group was 72.7%, while in the low expression group it was 11.1% (Fig. 2). Therefore, CXCL14 may be an independent factor that positively affects prognosis.

To analyze the effect of CXCL14 on colorectal carcinoma cells, a lentivirus that overexpressed CXCL14 was constructed. The PCR product (~336 bp) was obtained from the constructed plasmid of recombinant pLenti6.3_CXCL14_IRES-EGFP. Nhel and Ascl double enzyme digestions also revealed that the CXCL14 gene had been successfully cloned into the vectors (Fig. 3A). DNA sequencing further verified the successful construction of a recombinant plasmid. The CXCL14-overexpressing lentivirus was further packaged by infecting 293T cells for 48 h. The virus titer was detected through the enhanced green fluorescent protein (EGFP) carried by the virus. As demonstrated in Fig. 3B, the cells expressing EGFP were detected in cells infected with 2×10⁻⁸ ml virus and the titer of the virus was calculated as 1.2×10⁹ TU/ml. To confirm the expression of CXCL14, HT29 colorectal carcinoma cells were transfected with the CXCL14 lentivirus for 48 h followed by CXCL14 protein detection with western blotting. Compared with the lentivirus control, CXCL14 expression increased markedly in the cells transfected with the CXCL14 lentivirus (Fig. 3C).

Ectopic expression of CXCL14 led to a significant reduction in the viability of the HT29 cells, compared with the group transfected with the control lentivirus (Fig. 4A; P<0.05). The cell viabilities in the CXCL14-overexpression and control groups were 96.47±2.53% and 36.56±5.47%, respectively. Further FCM analysis indicated that the CXCL14-overexpression group presented a lower percentage of cells in the G₁ stage (67.46±0.92%) compared with the control group (82.34±0.75%) (P=0.02); and a significantly higher percentage of cells in the S stage (36.47±0.59) compared with the control group (21.97±0.64%) (P=0.01) (Fig. 4B). No significant difference between the two groups was observed in the percentages of cells in the G₂/M stage. These results suggest that the inhibition of HT29 cell proliferation by CXCL14 overexpression was mainly due to cell-cycle arrest at the G₀/G₁ stage (Fig. 4B and C).