MT1-MMP in breast cancer: induction of VEGF-C correlates with metastasis and poor prognosis

We used real-time RT-PCR to determine whether VEGF-C or MT1-MMP mRNA expression levels differed markedly among five breast cancer cell lines and correlated mRNA expression changes with protein levels measured by enzyme-linked immunosorbent assay (ELISA). MDA-MB-231 and MCF-7ADR cell lines had higher expression levels of VEGF-C and MT1-MMP, while neither MT1-MMP mRNA nor protein expression was detected in MCF-7 cells. VEGF-C mRNA expression correlated well with MT1-MMP mRNA expression (r = 0.914, P = 0.03; Figure 1a), and a similar correlation was found between the protein expression of VEGF-C and MT1-MMP (r = 0.945, P = 0.02; Figure 1b).

We found a statistically significant correlation between MT1-MMP mRNA expression in five breast cancer cell lines and the number of invasive tumour cells (r = 0.984. P = 0.002). VEGF-C mRNA expression was also closely associated with the number of invasive tumour cells (r = 0.979, P = 0.004).

The invasive activity of MDA-MB-231 cells into reconstituted basement membrane components was inhibited by 0, 4, 8, or 12 μg/ml anti-MT1-MMP antibody in a concentration-dependent manner. Moreover, RT-PCR showed that treatment of MDA-MB-231 with an anti-MT1-MMP antibody resulted in a concentration-dependent decrease in VEGF-C mRNA expression levels. An anti-VEGF-C antibody also inhibited the invasive activity of MDA-MB-231 cells, but did not affect MT1-MMP mRNA expression levels.

We assessed MT1-MMP and VEGF-C expression in samples from breast cancer patients. In normal mammary epithelial cells, we did not observe immunostaining for MT1-MMP and VEGF-C. However, diffuse cytoplasmic staining for VEGF-C protein was seen in tumour cells, while distinct tumour cell membrane staining of MT1-MMP was observed. We defined positive staining as immunoreactivity in >10% of the carcinoma cells [8]. Of the 106 breast cancer specimens, 55 (58.1%) had positive staining for VEGF-C and 66 (62.3%) for MT1-MMP. A significant correlation was found between the expression of VEGF-C and MT1-MMP (r = 0.458, P < 0.001).

We correlated the expression of MT1-MMP and VEGF-C with clinicopathological features of breast cancer patients (Table 1). Immunoreactivity of MT1-MMP was strongly correlated with clinical stage and axillary lymph node metastasis. VEGF-C expression was also strongly correlated with clinical stage, lymph node metastasis and tumour size. There was also a close association between VEGF-C, MT1-MMP expression and lymphatic vessel density.

Of the 106 patients examined in this study, 42 (39.6%) died due to relapse of the disease during the follow-up period. The overall survival rates for the 66 patients with positive MT1-MMP tumour staining were significantly lower than that of 40 patients with negative MT1-MMP staining (P = 0.0002), while the prognosis of 55 patients with positive VEGF-C tumour staining was poorer than with negative VEGF-C staining (P = 0.0001). Patients with tumours simultaneously positive for MT1-MMP and VEGF-C had the lowest overall survival rates. In contrast, patients with negative MT1-MMP and VEGF-C tumour expression had the highest survival rates (P = 0.0001). There was no difference in the clinical outcome between patients with positive MT1-MMP and negative VEGF-C tumour expression or patients with negative MT1-MMP and positive VEGF-C tumour expression.

Human breast adenocarcinoma cells lines (MCF-7, MDA-MB-453, MDA-MB-435, MCF-7ADR, MDA-MB-231) were stored in our laboratory. All cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with penicillin, streptomycin, 50 ng/ml ascorbic acid and 10% foetal calf serum (FCS) (Gibco BRL, Grand Island, NY, USA). Cells were maintained in a humidified incubator at 37°C with 5% CO₂. The absence of mycoplasma was confirmed using the Genprobe kit (Gen-Probe, San Diego, CA, USA).

The expression of MT1-MMP and VEGF-C transcripts was determined using real-time quantitative PCR. Briefly, total RNA was extracted using TRIzol reagent (Invitrogen, Grand Island, NY, USA) according to the manufacturer’s instructions. Total RNA (1 μg) was reverse-transcribed into single-stranded cDNA with oligo-dT₁₈ primer and SuperScript II reverse transcriptase (Invitrogen). Amplification of MT1-MMP, VEGF-C and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) as an internal control in each reaction was carried out by polymerase chain reaction (PCR) with the following primers: MT1-MMP, 5′-CCTGCATCCATCAATACTACTGC-3′ (forward) and 5′-GCGTCTGAAGAAGAAGAC AGC-3′ (reverse); VEGF-C 5′-CAGTTACGGTCTGTGTCCAGTGTAG- 3′ (forward) and 5′-GGACACACATGGAGGTTTAAAGAAG-3′ (reverse); GAPDH 5′-CCACCCATGGCAAATTCCATGGCA-3′ (forward), 5′-TCTAGACGGCAGGTCAGGTCCACC-3′ (reverse). Primers were used at a final concentration of 0.5 μM. The reaction mixture was first denatured at 95°C for 10 min followed by amplification at 95°C for 1 min, 55°C for 1 min, and 72°C for 1 min for 30 cycles, and by 72°C for 10 min. PCR products were visualised on a 2% agarose gel containing 1 μg/mL ethidium bromide. To evaluate the mRNA expression levels of MT1-MMP and VEGF-C, the ratios of MT1-MMP and VEGF-C GAPDH mRNA expression were measured by real-time quantitative PCR using a Light Cycler system and reagents (Roche Molecular Diagnostics) with the double-stranded DNA binding dye, SYBR Green 1, according to the procedure provided by the manufacturer. Standard curves were prepared for both the target gene (MT1-MMP and VEGF-C) and the internal control (GAPDH) by amplifying four logarithmic dilutions of plasmid containing the target fragment as templates. Standard dilutions were optimised to cover the relevant concentration range of target and reference RNA in the cell. The quantities of MT1-MMP and VEGF-C were determined from the standard curve and divided by the quantity of the GAPDH internal control. The quantities of MT1-MMP and VEGF-C were expressed as a fold differences relative to GAPDH from three experiments in duplicate.

Cells were plated onto 15-cm plastic dishes and grown to confluence. Plasma membrane preparations were collected with a cell scraper, and suspended in 1-ml ice-cold phosphate-buffered saline (PBS) containing protease inhibitors. The lysates were further sonicated for 5 s, and the membranes pelleted by centrifugation at 14,000 rpm for 30 min at 4°C. Pelleted membranes were washed once and resuspended in 0.5-ml complete PBS. Protein concentrations were determined with a bicinchoninic acid reagent, and samples were diluted with complete PBS to obtain starting concentrations of 1 mg/ml. The MT-MMP activity assay was performed according to the manufacturer’s protocol (Chemicon, Massachusetts, USA). Briefly, varying dilutions of the starting membrane preparations were APMA-activated and incubated for 1 h at 37°C with the MT-MMP peptide substrate (MCA-Pro-Leu-Gly-Leu-Dpa-Ala-Arg-NH2) in a 96-well plate, and read the plate on a fluorometric plate reader using a 325 nm excitation and 395 nm emission filters. Standard curves with the control peptide and MT1-MMP positive control were generated to determine the appropriate sample dilution at which the values fell in the middle of the standard curves. The experiments were completed in triplicate and repeated once.

Concentrations of VEGF-C in the medium of the cultured tumour cells were determined using a VEGF-C ELISA kit, as described in the protocol provided by the manufacturer (Bender MedSystems, San Diego, CA, USA). Briefly, cell culture medium was diluted 50-fold with sample diluent and added into wells of a microwell plate coated with anti-VEGF-C polyclonal antibody and incubated for 2 h at room temperature. Biotin-conjugated polyclonal VEGF-C antibody was added and the plate incubated for 1 h at room temperature. The plate was washed and streptavidin-HRP added. Following incubation, unbound streptavidin-HRP was removed during a wash step, and the colour reagent containing HRP added to the wells. The reaction was terminated by addition of stop solution and absorbance is measured at 450 nm in an Elx808IU Ultra Microplate Reader (BIO-TEK Instruments, Inc., Winooski, Vermont, USA). A standard curve was constructed using the VEGF-C protein standard provided in the kit. Generally, the samples were analysed using different dilutions in triplicate.

The membrane invasion culture system chamber was used to measure in vitro invasiveness of the breast cancer cells. Briefly, transwell inserts (Corning, Union City, CA, USA) with 8 μm pores were uniformly coated with 50 μg of Matrigel (Becton Dickinson, San Jose, CA, USA) and air-dried before being rehydrated. Single-tumour-cell suspensions, prepared after trypsinisation, were seeded into the upper wells at a concentration of 1 × 10⁵ per well. After a 24-h incubation in a humidified incubator at 37°C with 5% CO₂, the inserts were fixed in methanol and stained for 5 min with eosin followed by 5 min with haematoxylin. The cells on the upper surface of the inserts were wiped away with a cotton swab. The cells that had migrated through the matrix and adhered to the lower surface of the inserts were counted as nine separate fields at 40× magnification. Each cell line was tested at least twice and within a single experiment, each assay was performed in quadruplicate.

The effects of an anti-MT1-MMP antibody (Chemicon, Massachusetts, USA) and anti-VEGF-C antibody (Santa Cruz Biotechnology, Dallas, Texas, USA) on the invasive and proteolytic activity of tumour cells were examined using the membrane invasion assay. Briefly, log-phase cell cultures of MDA-MB-231 were harvested. A 200-μL cell suspension was added to transwell inserts and incubated with anti-MT1-MMP antibody at concentrations of 0, 4, 8, or 12 μg/ml for 24 h. Invasive cells were counted as described above. VEGF-C mRNA levels in cells incubated with various amounts of anti-MT1-MMP antibody were analysed by real-time RT-PCR as described above. Similarly, the effects of anti-VEGF-C antibody on invasive activity and MT1-MMP mRNA expression were also examined.

Formalin-fixed, paraffin-embedded samples obtained from 106 invasive ductal carcinomas and normal tissues from benign breast disease procedures free from pathological changes (n = 43) were retrieved from the pathology files of Nanfang Hospital, Southern Medical University (Guangzhou, China). All breast cancer patients underwent mastectomy with an axillary dissection during 2004 and none received pre-operative radiation or chemotherapy. The patients with histologically positive axillary lymph nodes or a primary tumour > 1 cm in diameter received post-operative chemotherapy. All patients were treated with adjuvant anti-oestrogen therapy if oestrogen receptor or progesterone receptor. All women were followed-up after surgical treatment at 6-month intervals for a mean period of 71.41 months. All patients provided informed consent according to a protocol approved by the ethics committee of the institute. Details of breast cancer patients are provided in Table 2.

To identify the expression of MT1-MMP and VEGF-C in the breast cancer tissue specimens, immunohistochemical analysis was performed. Briefly, paraffin-embedded breast cancer samples were cut into 4-μm sections. Slides were incubated for 12 h at 37°C, de-waxed in a histological clearing agent and hydrated. Endogenous peroxidase activity was blocked by incubation in 3% hydrogen peroxide for 15 min. Non-specific binding was prevented by incubation in 1.5% normal rabbit serum (VEGF-C) or 3% normal mouse serum (MT1-MMP) for 15 min in a humidified chamber in PBS. Slides were incubated with anti-VEGF-C (1.14 μg/ml; Dallas, Texas, USA) or MT1-MMP (5 μg/ml; Oncogene, Cambridge, MA, USA) for 60 min at room temperature. The primary antibody was detected using a biotinylated horse anti-rabbit (VEGF-C) or horse anti-mouse (MT1-MMP) secondary antibody for 30 min at room temperature and the peroxidase was introduced using a streptavidin conjugate. The slides were washed thoroughly with PBS between each stage in the procedure. The antibody reaction was visualised using a fresh substrate solution containing an aminoethyl carbazole substrate kit (Sigma, St. Louis, MO, USA). The sections were counterstained with haematoxylin, dehydrated and mounted in glycerol–vinyl–alcohol. For the negative controls, the primary antibody was replaced with mouse IgG.

A monoclonal mouse anti-human D2-40 antibody (Zymed, Grand Island, NY, USA) was used for the staining of lymphatic vessels. The D2-40 antibody detects a fixation-resistant epitope on a 40 kDa O-linked sialoglycoprotein expressed in lymphatic endothelium but not blood vessels, and can be used to assess lymphangiogenesis specifically in conventionally processed formalin-fixed and paraffin-embedded tissue specimens [60, 61]. The procedure for immunohistochemical staining of D2-40 is similar to that for MT1-MMP and VEGF-C. Sections from a previously studied case of tonsilla known to express D2-40 were used as positive controls.

Two pathologists who were unaware of the clinical data evaluated the immunohistochemical staining. To evaluate MT1-MMP and VEGF-C protein expression, the results were graded as follows; (+), >10% of the neoplastic cells were stained; (±), <10% of the neoplastic cells were stained; (−), neoplastic cells were not completely stained. In this study, (−) and (±) were classified as negative.

Lymphatic vessel density (LVD) was determined as suggested by Weidner et al.[62]. The immunostained sections were scanned by light-microscopy at low magnification (40×) and the areas of tissue with the greatest number of distinctly highlighted microvessels (‘hot spots’) were selected. LVD was then determined by counting all immunostained vessels at a total magnification of (200×) from five areas for each case.

All statistical calculations were carried out using the SPSS statistical software. The relation between two different values was evaluated using the Pearson correlation coefficient. Correlations between VEGF-C, MT1-MMP expression and clinicopathological features were assessed by the Chi-square test. Survival curves were generated by the Kaplan-Meier method and the difference between the curves was assessed using the log-rank test. P < 0.05 was accepted as indicative of statistical significance.

Department of Scientific Research approved the study and the publication of our paper.