In total, 17 pairs of BC and paracancer (PC) tissues were collected from patients (aged 55–65 years old) at Hubei Cancer Hospital (Wuhan, China) between December 2018 and December 2019. BC patients did not receive any chemotherapy or endocrine therapy prior to tumour removal. Immediately after surgery, all tissues were frozen in liquid nitrogen and maintained at −80°C until further analysis. In addition, three of the 17 pairs of BC tissues and their adjacent tissues were randomly selected for mRNA transcriptome sequencing (Illumina; PE-401-3001; Shanghai GeneChem Co., Ltd.). All clinical samples were collected with written informed consent from the patients, and this protocol was approved by the Ethics Committee of the Hubei Cancer Hospital.

Human breast epithelial cells (MCF10A) and human BC cell lines (MCF7, MDA231, MDA468 and MX1) were purchased from BeiNa Bio. MCF10A cells were maintained in RPMI-1640 medium supplemented with 10% FBS (both from Gibco; Thermo Fisher Scientific, Inc.) and 1% penicillin and streptomycin (ScienCell Research Laboratories, Inc.) in a humidified incubator with 5% CO₂ at 37°C. MCF7, MDA231, MDA468 and MX1 were maintained in Dulbecco's modified Eagle's medium (DMEM) (Gibco; Thermo Fisher Scientific, Inc.) supplemented with 10% FBS and 1% penicillin and streptomycin in a humidified incubator with 5% CO₂ at 37°C. All cell lines were confirmed to be free of mycoplasma contamination. Cycloheximide (CHX), MG132 and chloroquine, which were used for cell-based assays, were purchased from MCE. SRI-011381 (TGF-β agonist) was purchased from Selleck Chemicals.

The MMP11 overexpression (OE) plasmid and corresponding negative control (NC) plasmid were synthesized by Shanghai GeneChem Co., Ltd. Smad family member 2 small interfering (si)RNA (si-Smad2) and the corresponding NC siRNA were purchased from Guangzhou RiboBio Co., Ltd. The sequences of siRNAs are as follows: Smad2 sense, 5′-CCAGGUGGUGAAGAAUCUATT-3′, and antisense, 5′-UAGAUUCUUCACCACCUGGTT-3′; NC siRNA sense, 5′-AAAAATTCAAGACUUGGAGCU-3′ and antisense, 5′-UCUUGTTUUUUUAGCUCCAAG-3′. siRNAs at a final concentration of 50 nM and plasmids at a dose of 1.6 µg per well (12-well plates) were transfected into cells at 37°C for 48 h using Lipofectamine 3000 reagent (Invitrogen; Thermo Fisher Scientific, Inc.) according to the manufacturer's protocol. Subsequent assays were carried out after 72 h of transfection. The MMP11 lentivirus (Lv-sh) and empty control vector (Lv-NC) were constructed by Shanghai GeneChem Co., Ltd. The aforementioned recombinant lentiviruses were stably transfected into BC cell lines (MCF7 and MDA231) with an MOI of 2, and then the transfected cells were used for subsequent experiments after 3 days of puromycin selection with a final concentration of 2 µg/ml.

To assess cell viability, the cells were plated in 96-well plates at a density of 5×10³ cells/well, and then subjected to the MTT assay (Sigma-Aldrich; Merck KGaA). Then, 20 µl of MTT solution (5 mg/ml) was added to each well. After 4 h of incubation at 37°C, the culture medium in the well was removed, 150 µl dimethyl sulfoxide (DMSO) was added to each well, and the optical density was measured at 490 nm. The experiment was replicated three times independently. Each condition was plated in triplicate.

Seventy-two hours after transfection of siRNA or MMP11 lentivirus, 500 cells were inoculated into 6-well plates and grown in complete medium for two weeks. Colonies were fixed with 4% paraformaldehyde at room temperature for 15 min and stained with 0.1% crystal violet at room temperature for 30 min in 100% ethanol and then counted and photographed (D850; Nikon Corporation). All colony formation assays were conducted in triplicate.

Protein lysates of tumour tissues or cultured cells were prepared using RIPA buffer (Sigma-Aldrich; Merck KGaA) containing 1% phenylmethanesulfonyl fluoride (PMSF) and phosphatase inhibitors. Western blot analysis was performed as previously described (11). The antibodies used were as follows: MMP11 (product code ab119284; 1:1,000 dilution; Abcam), Smad2 (cat. no. 12570-1-AP; 1:1,000 dilution), Smad3 (cat. no. 25494-1-AP; 1:1,000 dilution), and β-actin (cat. no. 20536-1-AP; 1:1,000 dilution) primary antibodies and goat anti-rabbit IgG secondary antibody (cat. no. SA00001-2; 1:5,000 dilution; all from ProteinTech Group, Inc.). For immunohistochemistry (IHC), the aforementioned antibodies MMP11 and Smad2 (1:1,000 dilution) were used. The specific immunohistochemistry protocol was performed as previously described (12).

Total RNA from tissues or cultured cell lines was extracted using TRIzol^® reagent (Invitrogen; Thermo Fisher Scientific, Inc.) according to the manufacturer's instructions. RT-qPCR was carried out using the Prime-Script RT kit (TaKaRa Biotechnology Co., Ltd.). Gene expression levels were established by RT-qPCR using SYBR Premix Ex Taq (TaKaRa Biotechnology Co., Ltd.) and normalized to the endogenous level of β-actin as a control. The thermocycling conditions were as follows: Initial denaturation at 95°C for 30 sec, and then the PCR reactions: 95°C for 5 sec, 60°C for 35 sec, and 60 cycles. The primers used were as follows: MMP11 forward, 5′-CCATGTAATATCTAGATAAGGTCGGA-3′ and reverse, 5′-GAGTCAAGGTCGGGTGCGTGGGAAG-3′; β-actin forward 5′-CCAAGGCCAACCGCGAGAAGATGAC-3′ and reverse 5′-AGGGTACATGGTGGTGCCGCCAGAC-3′. All these experiments were performed in triplicate on a Bio-Rad CFX96 instrument (Bio-Rad Laboratories, Inc.), and the relative expression values were calculated using the 2^−ΔΔCq method (12).

MCF7 cells were transfected as previously indicated and then lysed in lysis buffer. The samples were centrifuged at 4000 × g at 4°C for 10 min, and the supernatants were retained. Then, the supernatants were incubated with anti-MMP11 (1 µg/ml) or IgG (1 µg/ml; product no. 6990; Cell Signaling Technology, Inc.) at 4°C for 4 h and protein A/G-Sepharose beads (10 µl) for 2 h at 4°C. The immunocomplexes were washed three times, boiled in sample buffer, and subjected to western blotting with anti-Smad2 antibody.

MCF7 cells (3×10⁶/mouse, 5/group) transfected with Lv-MMP11 or Lv-NC were injected subcutaneously into the right flank region of male BALB/c nude mice (6 weeks old; weight, ~15 g; 10 in total). All nude mice were kept in a Specific Pathogen Free (SPF) environment with controllable light (12-h light/dark cycle), temperature and humidity, with food and water available ad libitum. The mice were monitored weekly and the tumour volume was assessed; the long diameter of the tumour did not exceed 2 cm. The formula for the tumour volume was V=LxW²xπ/6 (V, volume; L, long diameter of the tumour; W, short diameter of the the tumour). Approximately 4 weeks after injection, according to institutional ethical guidelines, the mice were anesthetized with isoflurane and sacrificed by humanely cervical dislocation, and the weights of the tumours were recorded. The mouse experiments and the handling of animals were performed according to the Institutional and Animal Care and Use Committee of Hubei Cancer Hospital and the NIH Guide for the Care and Use of Laboratory Animals.

The R ‘limma’ Bioconductor package was used to screen the differentially expressed genes (DEGs) between KIRC and adjacent tissues based on the following criteria: Fold change (FC), |log2(FC)|>1; and false discovery rate (FDR) <0.05. Adjusted P<0.05 was used to define a gene as a DEG. An online database (https://www.UniProt.org) was used to predict MMP11-related molecules. The Database for Annotation, Visualization and Integrated Discovery (DAVID) v6.8 (david-d. ncifcrf.gov/) was used to analyze functional enrichment among DEGs. In addition, only those Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways with P≤0.05 and ≥10 enriched genes were considered significant. Online database analysis (http://kmplot.com) was used to further investigate the relationship between MMP11 and the prognosis of BC patients.

All statistical analyses were performed using SPSS 22.0 (IBM Corp.), and figures were produced using GraphPad Prism 6.0 (GraphPad Software, Inc). For parametric variables, paired and unpaired Student's t-test and one-factor ANOVA followed by Tukey's post hoc test were utilized to analyse the results of the animal experiments. Survival was assessed using the Kaplan-Meier method and analysed using the log-rank test. Univariate analyses were carried out using the Cox proportional hazards model. All experimental data are expressed as the mean ± SD of at least five independent experiments. In addition, the BC data set of The Cancer Genome Atlas (TCGA) (http://cancergenome.nih.gov/) of the University of North Carolina TCGA Genome Characterization Center was used for clinical correlation analysis (13). The differences were considered statistically significant when P<0.05.

To explore gene expression patterns, transcriptome sequencing of clinical BC samples and their adjacent tissues was performed. The sequencing results revealed that MMP11 was the most upregulated gene (Fig. 1A). To confirm the gene overexpression of MMP11, 34 paired of BC and adjacent PC tissues were used. Based on the RT-qPCR analysis, MMP11 mRNA was overexpressed in tumour tissues (Fig. 1B). Western blotting was also performed in paired BC and PC tissues, and comparison of the expression levels revealed that MMP11 was highly expressed in tumour tissues (Fig. 1C). In addition, the expression level of MMP11 was significantly higher in the four BC (MCF7, MDA231, MDA468, MX1) cell lines than in the normal human breast epithelial cell line (MDA10A) (Fig. 1D and E).

First, 34 BC specimens were assessed by RT-qPCR and IHC, and it was revealed that the mRNA and protein expression levels of MMP11 were basically the same (Fig. 2A). Then, patient clinicopathological characteristics were analysed to determine the association between MMP11 expression and clinical factors. The results revealed that only tumour stage was significantly associated with MMP11 expression (P=0.037) (Table I). Additionally, Kaplan-Meier survival analysis confirmed that high MMP11 expression was associated with poor OS in BC patients (Fig. 2B). To further expand the clinical samples and analyse the clinical significance of MMP11, the BC dataset from TCGA was downloaded and analysed. As revealed in Table II, a systematic analysis revealed that high MMP11 expression was significantly associated with progesterone receptor status (P=0.045), HER2 receptor status (P=0.027), tumour stage (P=0.003), and pathologic stage (P=0.004) in BC. Moreover, Kaplan-Meier survival analysis confirmed that the OS rate of patients with low MMP11 expression was significantly higher than that of patients with high MMP11 expression (P=0.008) (Fig. 2C). Online database analysis (http://kmplot.com) also confirmed that high expression of MMP11 was closely related to poor OS in BC (P<0.001) (Fig. 2D) (14).

To explore whether MMP11 affects the biological behaviour of BC, cell proliferation and colony formation assays were applied. First, it was determined that OE plasmid transfection significantly increased the mRNA and protein expression of MMP11 in MCF7 and MDA231 cells compared with that in the control group (Fig. 3A). Further results revealed that MMP11 overexpression significantly increased the proliferation and colony formation of MCF7 and MDA231 cells (Fig. 3B and C). In addition, to further demonstrate the role of MMP11 in BC, MMP11 Lv-shRNA was used to determine whether MMP11 knockdown could inhibit BC cell development. Lv-shRNA significantly reduced the expression of MMP11 mRNA and protein (Fig. 3D). Moreover, knockdown of MMP11 markedly inhibited the proliferation and colony formation of MCF7 and MDA231 cells (Fig. 3E and F). Therefore, these results indicated that overexpression of MMP11 promoted the proliferation of BC cells.

To further investigate the biological importance of MMP11 in BC, we used an online database (https://www.UniProt.org) to predict MMP11-related molecules and performed Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) analyses (https://david-d.ncifcrf.gov) (15,16). The biological processes of MMP11-related molecules were mainly enriched in ‘cell-cell signalling’, ‘regulation of cell proliferation’, and ‘cell-cell adhesion’ (Fig. 4A). In addition, certain signalling pathways associated with breast carcinogenesis were identified to be related to MMP11 in the KEGG pathway analysis; these pathways included the ‘TGF-β signalling pathway’, ‘PPAR signalling pathway’, ‘apoptosis’ pathway and ‘pathways in cancer’ (Fig. 4B). Then, the TGF-β signalling pathway, which was the most closely related to cell growth and proliferation in our analysis, was investigated. Smad2 and Smad3 are two key molecules in the TGF-β signalling pathway (17). The present results revealed that MMP11 overexpression increased the expression of Smad2 and Smad3 (Fig. 4C), while knockdown of MMP11 inhibited Smad2 and Smad3 expression (Fig. 4D), and this inhibition could not be increased by TGF-β inducers (Fig. 4E and F), which revealed that MMP11 promoted the expression of Smad2 and Smad3 independent on TGF-β.

To verify the interaction of Smad2 and Smad3 with MMP11, first, changes in the mRNA expression of Smad2 and Smad3 were detected after MMP11 overexpression by RT-qPCR and it was revealed that there was no significant difference in the mRNA expression of Smad2 and Smad3 between the MMP11 OE group and the NC group (data not shown). It was further attempted to use anti-MMP11 antibodies to immunoprecipitate Smad2 or Smad3 from MCF7 cells, which exhibit endogenous expression of MMP11, Smad2 and Smad3. Notably, it was revealed that Smad2 was detected in the MMP11 immune complex by western blot analysis (Fig. 5A). Therefore, it was speculated that MMP11 could affect the protein degradation of Smad2. To further verify our conjecture, a CHX chase experiment was performed. The results revealed that transfection of MCF7 cells with MMP11 Lv-sh led to a marked reduction in the half-life of Smad2 compared with that of Lv-NC transfection (Fig. 5B), indicating that the decrease in MMP11 expression promoted the degradation of Smad2. Furthermore, MCF7 cells were transfected with MMP11 Lv-sh and then treated with the proteasome inhibitor MG132 or autophagy inhibitor chloroquine. The present results revealed that MG132 effectively abrogated the effect of MMP11 Lv-sh on Smad2 protein degradation, whereas chloroquine had no such effect (Fig. 5C and D), which confirmed that MMP11 inhibited Smad2 degradation through the ubiquitin proteasome pathway. Finally, the MMP11 OE plasmid and Smad2-siRNA were co-transfected into MCF7 cells, and the MMP11 overexpression-mediated induction of cell proliferation was significantly inhibited by si-Smad2 (Fig. 5E), which verified that Smad2 was a critical downstream factor that mediated the ability of MMP11 to promote BC cell proliferation.

To further clarify the biological role of MMP11 in vivo, MCF7 cells stably transfected with Lv-NC or Lv-sh of MMP11 were implanted as xenografts in BALB/c nude mice (Fig. 6A). Compared with those in the Lv-NC group, the tumour weight and growth of the Lv-sh group were significantly inhibited (Fig. 6B and C). According to IHC, RT-qPCR and western blot assay results, the expression of MMP11 and Smad2 in implanted tumours of the Lv-sh group was significantly lower than that in the Lv-NC group (Fig. 6D-F). These results further indicated that MMP11 played an essential role in promoting BC through Smad2.