A first dataset with access number NM_002421 was obtained from the UCSC database (https://genome.ucsc.edu/). The dataset contained 24 samples of normal colorectal tissues and 81 samples of colorectal cancer tissues. The next dataset with the access number X05231 contained 18 samples of colorectal cancer and 18 samples of matched adjacent non-tumor tissues. Then, the last dataset, which included 22 normal colorectal tissues and 215 colorectal sample tissues, was obtained using the Cancer Genome Atlas (TCGA) database (http://tcga-data.nci.nih.gov/tcga/). The expression levels of MMP-1 in these datasets were statistically analyzed.

The following human colorectal cancer cell lines were obtained from Sangon Biotech Co., Ltd.: Lovo, HT-29, SW-480, HCT-116, Caco-2 and SW-620. All of these cells were cultured at 37°C and 5% CO₂ in Dulbecco's modified Eagle's medium (DMEM; Gibco; Thermo Fisher Scientific, Inc.) containing 100 units/ml penicillin, 100 units/ml streptomycin and 10% foetal bovine serum (FBS; HyClone; GE Healthcare Life Sciences); the FBS was inactivated by incubating at a constant 56°C for 30 min.

We collected 49 paraffin-embedded colorectal cancer tissues and matched adjacent noncancerous tissues from patients with colorectal cancer from the First Affiliated Hospital of Xi'an Jiaotong University from January 2014 to December 2015. All patients were diagnosed with primary colorectal cancer without distant metastases. All patients underwent surgery for the first time for the treatment of colorectal cancer. No patients received radiotherapy or chemotherapy before the operation. All patients provided consent for the use of their samples, and this use was approved by the Institute Research Medical Ethics Committee of the First Affiliated Hospital of Xi'an Jiaotong University.

An RNA interference lentivirus containing a puromycin marker from GeneChem, Inc. was obtained; the vector can steadily downregulate the expression of MMP-1. The targeted sequence of the shRNA was UGAACAUCACCACUGAAGGUGUAGC (25), and its efficiency was demonstrated by western blotting. A control lentivirus that did not target anything, but carried a puromycin marker as well, was also obtained. In the present study, HT-29 and SW-480 cells were infected with shRNA lentiviruses targeting MMP-1. Successfully transfected cells were selected in the aforementioned DMEM containing 1 µg/ml puromycin and were then expanded for the subsequent experiments.

Cell proliferation was assessed via a Cell Counting Kit-8 (CCK-8; Beyotime Institute of Biotechnology). A total of 2,000 cells in 200 µl of medium were placed in each well of 96-well plates and were cultured in the abovementioned complete medium. At the instructed time-point, 20 µl of CCK-8 was added to each well, which was followed by 2 h of incubation. Then, the OD 450 absorbance was detected by a microplate reader (BioTek Instruments, Inc.).

Cell colony formation was examined by a colony formation assay. A total of 2,000 cells in complete medium were inoculated in 60-mm plates and were then incubated at 37°C in a 5% CO₂ environment for two weeks. Cells were dyed with 0.2% dissolved crystal violet after fixation with methanol for half an hour. The images were obtained using a chemiluminescence imager (Bio-Rad Raboratories, Inc.). Then, the number of colonies that contained >50 cells were counted. Each assay was repeated three times.

TRIzol reagent (Sigma-Aldrich; Merck KGaA) was used to purify the total RNA from cells according to the manufacturer's instructions. Then, the total RNA was reverse-transcribed using SYBR Green (Takara Biotechnology Co., Ltd.) and an ABI 7500 instrument (Thermo Fisher Scientific, Inc.). Primers were designed according to the reported sequences (26): MMP-1 forward, 5′-AAATGCAGGAATTCTTTGGG-3′ and reverse, 5′-ATGGTCCACATCTGCTCTTG-3′; β-actin forward, 5′-TGGCACCCAGCACAATGAA-3′ and reverse, 5′-CTAAGTCATAGTCCGCCTAGAAGCA-3′. The real-time PCR data were normalized to endogenous levels and were carried out in triplicate.

Tissue samples were formalin-fixed and paraffin-embedded, sliced into 4-µm sections and mounted on slides. Graded ethanol was used to rehydrate the sections after deparaffinization in xylene at room temperature. After washing with phosphate-buffered saline (PBS), the sections were placed in 3% hydrogen peroxide for 20 min to inhibit endogenous peroxidase. Antigen retrieval was carried out by soaking the slides in 0.01 M citrate buffer (pH 6.0) and heating them for 30 min in a microwave. After incubation with a primary MMP-1 antibody (dilution 1:500; ab137332; Abcam) at 4°C overnight, the slides were washed with PBS and incubated with biotinylated secondary antibody for half an hour at 37°C. 3,′3-Diaminobenzidine tetrahydrochloride (DAB) was applied as a chromogen, and the sections were counterstained with hematoxylin. PBS was substituted for the primary antibody and used as a negative control.

The intensity degrees of positive signals were determined as 0, none; 1, weak; 2, moderate; 3, intense; and 4, strongly intense; and the percentage degrees of the number of positive cells were recorded as 0, 0%; 1, 1–25%; 2, 26–50%; 3, 51–75%; and 4, 76–100%. The present study defined the outcome by multiplying the aforementioned two scores to achieve the final score of IHC staining. A score of <3 was categorized as ‘negative’, and a score of ≥3 was categorized as ‘positive’ in the following statistical analysis. Two pathologists evaluated the scores separately to exclude subjectivity.

Cells were harvested, washed with PBS and lysed with radioimmunoprecipitation assay buffer (RIPA; Beyotime Institute of Biotechnology) to extract the total proteins. After determining the protein concentrations using a BCA assay kit (Thermo Fisher Scientific, Inc.), an equal amount of 20 µg protein from each sample was resolved via 10% SDS-PAGE and then the proteins were transferred to polyvinylidene difluoride membranes (EMD Millipore). After submerging the membranes in 5% non-fat dry milk for 2 h to block non-specific binding, they were incubated with a primary antibody in TBST at 4°C overnight. The primary antibodies used were as follows: MMP-1 (dilution 1:500; ab6721), p-Akt (dilution 1:1,000; ab64148), c-myc (dilution 1:1,000; ab32072), E-cadherin (dilution 1:1,000; ab40772), N-cadherin (dilution 1:1,000; ab76057), vimentin (dilution 1:1,000; ab92547), Twist1 (dilution 1:1,000; ab50581), and GAPDH (dilution 1:1,000; ab9485; all from Abcam). The PVDF membranes were subsequently washed with PBS and were then incubated with a Goat Anti-Rabbit horseradish peroxidase (HRP)-conjugated secondary antibody (dilution 1:2,000; ab6721; Abcam). The proteins were developed with a chemiluminescent substrate (Thermo Fisher Scientific, Inc.), and the expression was normalized to GAPDH.

Approximately 5×10⁵ cells were seeded into each well of 6-well plates containing complete medium and the cells were cultured overnight in an incubator set at 37°C and 5% CO₂. When the cell monolayers reached 100% confluence, a yellow pipette tip was used to scrape them. After washing with 1X PBS, the cells were placed back in the incubator. The breadth of the scratched areas was measured at 0, 24, 48 and 72 h under a fluorescence microscope (CKX41; Olympus Corp.).

To estimate the migration ability of the cells, individual groups of cells (approximately 1×10⁵ cells/well) suspended in 200 µl of serum-free medium were placed into the upper chambers of an 8-µm-pore Transwell (Corning Inc.) coated with Matrigel matrix. All the upper chambers were placed in 24-well plates after the lower chambers were filled with 500 µl of medium containing 20% FBS. Next, the chambers were incubated at 37°C in a 5% CO₂ atmosphere for 24 h, swabbing the upper chambers with cotton swabs in succession. The cells located on the underside of the upper chambers were fixed using 4% paraformaldehyde and stained using 0.1% crystal violet at 25°C for 15 min. The migrated cells were counted in five randomly selected visual fields at an ×200 magnification. The same procedures were applied in the migration assay, however, the upper chambers were replaced with uncoated chambers. All assays were carried out in triplicate.

A total of six female BALB/c nude mice (age 6 weeks, weight 18–22 g) were purchased from the SLAC Laboratory Animal Center and divided into two groups with three mice in each group. The use of animals was approved by the Institute Research Medical Ethics Committee of The First Affiliated Hospital of Xi'an Jiaotong University.

The SW-480 cells treated with a negative control or with an MMP-1 knockdown were harvested and inoculated into the right forelimb of nude mice (1×10⁶ cells/mouse). Every week, the tumor was measured using Vernier calipers and the volume was computed by the following formula: Tumor volume (mm³)=length (mm) × width²(mm²)/2. All assays were carried out in triplicate. All animals were euthanized by cervical dislocation at the 3rd week. Then, the tumors were resected, weighed and fixed with paraffin for subsequent experiments.

All data were statistically analyzed using GraphPad Prism 6.0 (GraphPad Software) and SPSS 18.0 software. All in vitro assays were repeated three times, and the data are presented as the mean ± SD. Differences between experimental groups were evaluated with Student's t-tests or one-way analysis of variance (ANOVA) followed by Fishers' least significant difference test (LSD). Statistical significance was defined as P<0.05.

Bioinformatics analyses revealed that the expression of MMP-1 was significantly increased in colorectal carcinoma samples (Fig. 1A-C). The results of immunohistochemistry revealed that the expression levels of MMP-1 protein were significantly increased in 28/49 colorectal cancer tissues compared with 11/49 adjacent non-tumor tissues (Fig. 1D and E). To further investigate the association between the expression level of MMP-1 protein and clinicopathological features, a chi-square test and a two independent samples t-test were performed to assess the relationship between MMP-1 and the clinical characteristics of colorectal cancer patients. The P-values revealed that high expression of MMP-1 was associated with the TNM stage (P<0.01) as well as with lymphatic metastasis (P<0.01; Table I). These results demonstrated that increased MMP-1 expression was related to poor diagnosis in colorectal carcinoma.

In the next experiment, it was explored whether the high expression of MMP-1 influenced the prognosis of colorectal cancer patients. The outcome of Kaplan-Meier analyses revealed that a high expression level of MMP-1 was related to poor prognosis in both overall survival (P<0.01) and recurrence-free survival (P<0.01; Fig. 1F and G).

This study examined the expression levels of MMP-1 protein in various colorectal cancer cell lines via western blotting. Although the expression levels were variable in these cell lines, MMP-1 was expressed in most of them (Fig. 2A and B). A similar expression pattern was also revealed when measuring the mRNA of MMP-1 by means of real-time PCR. According to the aforementioned results, the HT-29 and SW-480 cell lines were selected for the next experiments.

The HT-29 and SW-480 cell lines were stably transfected with an shMMP-1 lentivirus and an empty vector as a control. To verify the efficiency of infection, real-time PCR was performed after transfection (Fig. 2C and D). To further determine the effect of transfection, the expression levels were assessed by western blotting (Fig. 2E). All of the aforementioned results revealed that the expression levels of MMP-1 protein decreased after the cells were transfected with lentivirus.

Having knocked down the expression of MMP-1 in the HT-29 and SW480 cell lines, the role of MMP-1 in the progression of colorectal carcinoma was investigated. CCK-8 assays revealed that downregulation of MMP-1 expression attenuated the proliferative capability of colorectal cell lines (Fig. 2F and G). In addition, the number of HT-29 and SW-480 colonies was significantly decreased after the expression of MMP-1 was knocked down, indicating that MMP-1 enhances the colony formation capability of these cell lines (Fig. 2H and I).

Subsequently, Transwell assays were performed to evaluate the influence of MMP-1 on the invasive ability of colorectal cells. The migration and invasion experiments demonstrated that downregulated expression of MMP-1 attenuated the migratory and invasive capabilities of colorectal cell lines (Fig. 3A-C). Then, wound healing assays were carried out to measure the influence of MMP-1 on the migration capability of the cells (Fig. 3D-F). The outcomes demonstrated that silencing of MMP-1 attenuated the migration potential of colorectal cells.

To further verify the tumorigenic ability of MMP-1 in vivo, mice were subcutaneously inoculated with shMMP-1-treated HT-29 cells and control HT-29 cells. The subcutaneous tumor nodules resected from the mice with shMMP-1-treated HT-29 cells grew slower than those resected from the empty vector group (Fig. 4A-E). In addition to tumor volume and weight, the common proliferation marker, Ki-67, was assessed via western blotting. The results demonstrated the suppressed expression of Ki-67 in tumor nodules derived from shMMP-1-treated cells compared to tumor nodules derived from control cells (Fig. 4F and G). All of the evidence revealed in Fig. 4 demonstrated that interfering with the expression of MMP-1 in colorectal cancer can inhibit the growth of tumors in vivo.

To reveal the latent mechanism through which MMP-1 enhances the progression of colorectal carcinoma, western blotting was used to identify the potential signaling pathway associated with the downregulation of MMP-1. This study revealed that the Akt signaling pathway was involved in both the HT-29 and SW-480 cell lines transfected with shMMP-1. The western blot results revealed that the expression of phosphorylated Akt as well as c-Myc was suppressed after the downregulation of MMP-1 (Fig. 5), while the expression of total Akt was not significantly different (data not shown). Furthermore, the knockdown of MMP-1 suppressed EMT in colorectal carcinoma.

In HT-29 and SW-480 cells, the downregulation of MMP-1 induced a significant decrease in N-cadherin, vimentin and Twist1 expression and a simultaneous increase in E-cadherin expression compared to cells transfected with the empty vector (Fig. 5).

Collectively, MMP-1 promoted activation of the Akt signaling pathway and EMT in colorectal carcinoma.