Matrix metalloproteinases (MMP) are important regulators of angiogenesis and tumor progression by degradation of extracellular matrix. Clinical trials using MMP inhibitors have failed and recent studies suggest that MMPs may in contrast suppress tumor growth. It is not known, however, if MMPs or their inhibitors, tissue inhibitor of metalloproteinases (TIMP), can be used as therapy of established cancer. Here, adenovirus vectors carrying the human genes for MMP-9, TIMP-1, or empty controls were injected intratumorally in breast cancers established in mice supplemented with estradiol and treated with tamoxifen. Microdialysis was used to quantify MMP activity and sampling of endostatin and vascular endothelial growth factor (VEGF) in situ. We show that AdMMP-9 increased MMP activity in vivo, decreased tumor growth rate, and decreased microvessel area significantly. AdMMP-9 therapy resulted in significantly increased levels of endostatin in vivo, whereas VEGF levels were unaffected. As previously shown, tamoxifen exposure by itself increased MMP activity in all treatment groups. Moreover, the combined therapy with AdMMP-9 and tamoxifen further reduced tumor growth and increased the endostatin levels compared with either treatment alone. Gene transfer of TIMP-1 had no effects on tumor progression and counteracted the therapeutic effect of tamoxifen in our breast cancer model. This is the first report showing that overexpression of MMP-9 results in increased generation of antiangiogenic fragments, decreased angiogenesis, and therapeutic effects of established breast cancer. [Cancer Res 2008;68(9):3405–12]

Degradation and proteolysis of the extracellular matrix (ECM) are critical events in cancer cell dissemination, growth, and metastasis (1). Matrix metalloproteinases (MMP) play an important role in the remodeling and proteolysis of the ECM. The functional activity of MMPs in vivo is regulated in the extracellular microenvironment where they are activated by various proteases and inhibited by naturally occurring tissue inhibitors of metalloproteinases (TIMP; ref. 2).

The essential role of the MMP/TIMP system in the metastatic process of cancers and of tumor angiogenesis has been clearly shown. Up-regulation of MMPs and paradoxically TIMPs has been correlated with tumor aggressiveness of various cancer forms, including breast cancer (36). Depending on the tissue distribution and the expression level of individual MMPs and TIMPs, these enzymes may either promote or inhibit tumor progression. Several recent studies have shown that up-regulation of endogenous MMPs decrease angiogenesis in tumors by the production of antiangiogenic factors such as angiostatin, tumstatin, and endostatin and thereby enhance tumor regression (712). Moreover, MMP inhibition in clinical trials has failed as antitumor therapy in several cancer forms (13, 14).

The in vivo activation of MMPs and the release of proangiogenic and antiangiogenic factors occur extracellularly in the tumor microenvironment. Hence, techniques that allow for in situ investigations of these events are vital. Microdialysis, a technique that mimics a blood vessel within a tissue, allows sampling of molecules such as vascular endothelial growth factor (VEGF) and endostatin directly from interstitial space and we have previously shown that this technique may also be used to directly quantify MMP activity in tumor tissue in situ (10, 1518).

Sex steroids play an important role in breast carcinogenesis and progression and the majority of breast cancers are hormone dependent. Anti-estrogens such as the synthetic, nonsteroidal compound tamoxifen, and aromatase inhibitors are corner stones in the medical treatment of breast cancer. We recently reported that tamoxifen treatment of experimental breast cancer both in vitro and in vivo increased the expression and activity of MMP-9, whereas estradiol treatment lowered MMP-9 activity levels (10, 19). Additionally, the increased MMP-9 activity was accompanied by higher levels of interstitial endostatin and these tumors exhibited decreased angiogenesis (10, 19). This suggests that increasing the levels of MMP-9 and thereby the generation of antiangiogenic fragments are among the mechanisms that explain the antitumorigenic properties of tamoxifen.

Most studies of MMP/TIMP and tumor progression have been used in tumor cell lines overexpressing various MMPs or MMP null mice and focused on the early stages of cancer progression (9, 20). No previous studies have investigated the effects of direct treatment of established solid tumors with MMPs or TIMPs. Our previous studies demonstrating increased levels of MMP-9 by tamoxifen lead us to investigate in the present study if treatment with MMP-9 and TIMP-1 may affect tumor growth of established human breast cancer in nude mice. We show that overexpressing MMP-9 in whole tumor tissue by intratumoral injection of an adenovirus carrying the cDNA for MMP-9 induces tumor regression and the release of extracellular endostatin in vivo. The addition of tamoxifen treatment enhanced the effects of the MMP-9 gene therapy. We also show that overexpressing TIMP-1 intratumorally does not affect tumor progression but instead counteracts the beneficial effects of tamoxifen in breast cancer.

Adenoviral constructs. The human cDNA for TIMP1 was a gift from C. Richards (Department of Pathology and Molecular Medicine, Centre for Gene Therapeutics, McMaster University, Hamilton, Ontario, Canada). The human cDNA for MMP-9 (hMMP-9) was purchased from American Type Culture Collection (ATCC). Both cDNAs were cloned into the loxP-containing shuttle vector, pDC316, for viral assembly. Cotransfection of recombinant pDC316 and adenovirus genomic plasmid, pBHGloxΔE1,3Cre, subsequent purification and amplification were carried out as previously described (21).

Cell culture. The cell line 293 was used to confirm expression of hMMP9. Cells were grown on 10-mm dishes (BD Falcon) to 70% to 80% confluence in α-MEM (Sigma) supplemented with 10% fetal bovine serum (FBS; Sigma), 2 mmol/L l-glutamine (Sigma), 50 IU/mL penicillin G, 50 μg/mL streptomycin (Pen-strep, Sigma), and 5 μg/mL fungizone at 37°C in a humidified atmosphere of 5% CO2. Cells were serum starved overnight, after which cells were infected with 10 μL of adenovirus solution and allowed to incubate for a further 24 h. An empty control vector, Addl70-3, was used in parallel.

MCF-7 (HTB-22; human breast adenocarcinoma, estrogen receptor and progesterone receptor positive) cells were obtained from ATCC. Cells were cultured in DMEM without phenol red, supplemented with 2 mmol/L glutamine, 50 IU/mL penicillin G, 50 μg/mL streptomycin, and 10% FBS, at 37°C in a humidified atmosphere of 5% CO2. Cell culture medium and additives were obtained from Invitrogen unless otherwise stated.

Zymography. To quantitate transgene activity of TIMP-1, reverse zymography was used. Twenty micrograms of total protein from bronchoalveolar lavage fluid prepared from AdTIMP1-infected rats, as previously described (22), were loaded onto a 12% SDS-PAGE gel containing 0.1% gelatin (type A, 300 bloom from porcine skin, Sigma) and agarose-gelatin purified gelatinases. To obtain gelatinases, A549 cells were grown to confluence in 100-mm plates in DMEM supplemented with 2 mmol/L l-glutamine (Sigma), 50 IU/mL penicillin G, 50 μg/mL streptomycin (Pen-strep, Sigma), 5 μg/mL fungizone, and 10% fetal bovine serum (FBS). The cells were then trypsinized and resuspended in serum-free medium containing 50 ng/mL 4-aminophenylmercuric acetate (Sigma) and allowed to incubate for 48 h. The supernatants were collected, mixed with 500 μL of gelatin-agarose beads, and incubated at 4°C for 16 h. Beads were spun down and washed with buffer containing 50 mmol/L Tris (pH 7.6), 0.15 mol/L NaCl, and 0.1% Tween 20. Gelatinases were eluted from the beads with buffer containing 50 mmol/L Tris (pH 7.6), 1 mol/L NaCl, 7.5% DMSO, and 0.1% Tween 20. A total of 100 μg of gelatinase protein was copolymerized with the gelatin. Gelatin zymography for hMMP9 activity was carried out in the same manner without the addition of gelatinases to the gels. Gels were run at a constant voltage of 100 V for 1 h, removed from the apparatus, and washed in 2.5% Triton X-100 for 30 min to remove the SDS. After removal of the detergent, they were incubated in activation buffer [50 mmol/L Tris-HCl (pH 8), 10 mmol/L CaCl2, 5 μmol/L ZnSO4, 150 mmol/L NaCl] for 16 h at 37°C. Following incubation, activity was visualized by staining gels with Coomassie blue. TIMP-1 activity was visualized as a dark band against clear background and MMP9 activity was visualized as clear bands against a dark background.

Animals, oophorectomy of mice, and tumor establishment. Female athymic nude mice, BALB/c nu/nu (ages 6–8 wk), from Taconic, were housed in a pathogen-free isolation facility, with a light/dark cycle of 12/12 h and fed with rodent chow and water ad libitum. The study was approved by the Linköping University animal ethics research board. Mice were anesthetized with i.p. injections of ketamine/xylazine and oophorectomized, and 3-mm pellets containing 17β-estradiol, 0.18 mg/60 d release (Innovative Research of America), were implanted s.c. The pellets provide a continuous release of estradiol at serum concentrations of 150 to 250 pmol/L, which is in the range of physiologic levels seen in mice during the estrous cycle as previously shown (23). One week after surgery, MCF-7 cells (5 × 106 in 200 μL PBS) were injected s.c. on the flank. Tumor size was determined by measuring the length and width of the tumor every 4 days. As tumors were attached to the underlying tissue, it was not possible to measure tumor depth; therefore, tumor size was calculated during in vivo growth as tumor area. When tumors were excised, all three variables (length, width, and depth) were measured. At a tumor size of ∼40 mm2, the mice were divided into two subgroups. One of the subgroups continued with the estradiol treatment only, in the other group tamoxifen (1 mg/mouse every 2nd day s.c.) was added (day −2 before adenoviral vector administration) to the estradiol treatment. Estradiol- and estradiol+tamoxifen (E+TAM) mice were each divided into additional three subgroups, which were injected with 1 × 109 plaque-forming units (pfu) of either AdMMP-9, AdTIMP-1, or the empty control vector Addl70-3 diluted in PBS (total volume of 50 μL) intratumorally (day 0). Microdialysis and/or tumor excision were made at days 7 and at 24 after adenovirus injections.

Microdialysis for protein sampling and direct quantification of MMP-9 activity in vivo. Tumor-bearing mice were anesthetized with an i.p. injection of ketamine/xylazine and thereafter kept anesthetized by repeated s.c. injections of ketamine/xylazine. A heating lamp maintained body temperature. Microdialysis probes (CMA/20, 0.5-mm diameter, PES membrane length 4 mm, 100 kDa cutoff, CMA/Microdialysis) were inserted into tumor tissue sutured to the skin, connected to a CMA/102 microdialysis pump (CMA/Microdialysis) and perfused at 0.6 μL/min with saline154 mmol/L NaCl containing 40 mg/mL dextran (Pharmalink). The outgoing microdialysates were collected on ice and stored at −70°C for subsequent analysis. At the end of experiments, the mice were euthanized and the tumors were excised. Tumors were measured, weighed, formalin fixed, and subsequently embedded in paraffin for immunohistochemical analysis or snap frozen in liquid nitrogen for subsequent homogenization. Frozen tumor tissue was homogenized in 500 μL ice-cold at 24,000 rpm using a Polytron PT3000 (Kinematica AG); supernatants were collected and stored in −70°C until subsequent use. Total protein concentration was determined using the Bio-Rad Protein Assay with bovine serum albumin as the standard (Bio-Rad Laboratories).

To determine the intratumoral activity of MMP-9 in vivo, a quenched fluorogenic substrate specific for MMP-2 and MMP-9 (DNP-Pro-Leu-Gly-Met-Trp-Ser-Arg-OH; Calbiochem, Merck Biosciences) was used in the microdialysis infusate, as previously described in detail (10). Briefly, mice were anesthetized, and microdialysis probes (20-kDa molecular mass cutoff, 0.5-mm diameter, membrane length 4 mm, CMA/Microdialysis) were inserted intratumorally. The microdialysis probes were perfused with 50 μmol/L of the MMP substrate at 2 μL/min. Microdialysates were collected at 30-min intervals into amber tubes and analyzed using a Cary Eclipse fluorescence spectrophotometer (Varian, Inc.) with λex at 280 nm and λem 360 nm.

Quantification of MMP-9, TIMP-1, endostatin, and VEGF. Microdialysates and/or tumor homogenates were analyzed for MMP-9, TIMP-1, endostatin, and VEGF protein content using commercial quantitative immunoassay kits [MMP-9 (total) Quantikine, endostatin Quantikine, TIMP-1 Quantikine, and VEGF QuantoGlo R&D Systems]. The analyzed proteins in the tumor homogenates were normalized to the total protein content and expressed as pg/mg protein or ng/mg protein.

Immunohistochemistry and scoring of tumor vessels. Formalin-fixed, paraffin-embedded tumors were cut in 3-μm sections, deparaffinized, and subjected to von Willebrand immunohistochemistry (rabbit antihuman von Willebrand; dilution 1:500; DakoCytomation) and counterstained with Mayer's hematoxylin. Negative controls did not show staining. In a blinded manner, 10 high-power fields (×200) were examined per section of three different tumors in each group. Vessel quantification of tumor sections was conducted as described previously using a Nikon microscope equipped with a digital camera (24). Percentage of area stained positively for von Willebrand factor was assessed using Easy Image Measurement software (Bergstrom Instruments).

Statistical analysis. Data are expressed as means ± SE. ANOVA and two-tailed Student's t test was used for simple comparison of values where appropriate. All statistical tests were two-sided. P < 0.05 was considered statistically significant.

Vector function and gene transduction in tumor tissue. Bronchoalveolar lavage fluids prepared from virus-injected (AdMMP-9 and AdTIMP-1) rats were loaded on zymogram gels. Figure 1A shows MMP-9 activity on a zymogram and Fig. 1C depicts a reverse zymogram of TIMP-1 function.

Figure 1.

Efficient gene transfer of AdMMP-9 and AdTIMP-1. The transgene activity of AdMMP-9 and AdTIMP1 was quantitated by zymography and, in the case of TIMP-1, reverse zymography. Bronchoalveolar lavage fluids prepared from intratracheally infected rats were loaded on zymogram gels. To verify that gene transfer was also efficient in tumor tissue, fresh frozen tumors were homogenized and analyzed for protein content 1 wk after virus injections. Mice were oophorectomized and supplemented with a physiologic level of estradiol. MCF-7 cells were injected s.c., and tumors were formed on the right hind flank. One group of mice continued with estradiol only, and in the other group tamoxifen treatment was added to the estradiol treatment (E+TAM). AdMMP-9, AdTIMP-1, or Addl70-3 (empty control vector), 1 × 109 pfu, diluted in PBS was injected into tumors of similar sizes (∼40 mm2). A, MMP-9 activity on zymograms. B, there was a significant increase of MMP-9 protein content after intratumoral injections of AdMMP-9. Tamoxifen treatment by itself significantly increased the MMP-9 levels in all adenovirus treatment groups. Columns, mean (n = 4–5 in each group); bars, SE. C, TIMP-1 activity on reverse zymograms. D, AdTIMP-1 injections into tumors resulted in a significant increase of TIMP-1 protein in tumor tissue in both estradiol- and E+TAM–treated animals. Columns, mean (n = 4–5 in each group); bars, SE.

Figure 1.

Efficient gene transfer of AdMMP-9 and AdTIMP-1. The transgene activity of AdMMP-9 and AdTIMP1 was quantitated by zymography and, in the case of TIMP-1, reverse zymography. Bronchoalveolar lavage fluids prepared from intratracheally infected rats were loaded on zymogram gels. To verify that gene transfer was also efficient in tumor tissue, fresh frozen tumors were homogenized and analyzed for protein content 1 wk after virus injections. Mice were oophorectomized and supplemented with a physiologic level of estradiol. MCF-7 cells were injected s.c., and tumors were formed on the right hind flank. One group of mice continued with estradiol only, and in the other group tamoxifen treatment was added to the estradiol treatment (E+TAM). AdMMP-9, AdTIMP-1, or Addl70-3 (empty control vector), 1 × 109 pfu, diluted in PBS was injected into tumors of similar sizes (∼40 mm2). A, MMP-9 activity on zymograms. B, there was a significant increase of MMP-9 protein content after intratumoral injections of AdMMP-9. Tamoxifen treatment by itself significantly increased the MMP-9 levels in all adenovirus treatment groups. Columns, mean (n = 4–5 in each group); bars, SE. C, TIMP-1 activity on reverse zymograms. D, AdTIMP-1 injections into tumors resulted in a significant increase of TIMP-1 protein in tumor tissue in both estradiol- and E+TAM–treated animals. Columns, mean (n = 4–5 in each group); bars, SE.

Close modal

One week after intratumoral injections of the vectors, tumors were excised, snap frozen in liquid nitrogen, and homogenized. Protein levels of MMP-9 and TIMP-1 were analyzed in all treatment groups. There was a significant increase of MMP-9 protein content after intratumoral injections of AdMMP-9 compared with control, E+Addl70-3 versus E+AdMMP-9 (4.67 ± 0.56 versus 350 ± 51.5 ng/mg protein; P = 0.0005; n = 4–5 in each group; Fig. 1B). In the tamoxifen-treated tumors, the increase after AdMMP-9 was more pronounced, E+TAM+Addl70-3 versus E+TAM+AdMMP-9 (18.5 ± 4.4 versus 858 ± 125 ng/mg protein; n = 4–5 in each group, P = 0.0006; Fig. 1B). In line with our previous data (10, 19), tamoxifen treatment by itself significantly increased the MMP-9 levels in all adenovirus treatment groups, E+Addl70-3 versus E+TAM+Addl70-3 (4.67 ± 0.56 versus 18.5 ± 4.4 ng/mg protein; n = 4–5 in each group, P = 0.02; Fig. 1B), E+AdMMP-9 versus E+TAM+AdMMP-9 (350 ± 51.5 versus 858 ± 125 ng/mg protein; n = 4–5 in each group, P = 0.01; Fig. 1B), and E+TIMP-1 versus E+TAM+TIMP1 (2.45 ± 0.45 versus 8.4 ± 1 ng/mg protein; n = 4–5 in each group P = 0.002; Fig. 1B). The injection of AdTIMP-1 significantly decreased the levels of MMP-9 compared with control tumors, E+Addl70-3 versus E+AdTIMP-1 (4.67 ± 0.56 versus 2.4 ± 0.55 ng/mg protein; n = 4–5 in each group, P = 0.02; Fig. 1B).

AdTIMP-1 injections into tumors resulted in a significant increase of TIMP-1 in tumor tissue, E+Addl70-3 versus E+AdTIMP-1 (2.3 ± 0.48 versus 68.4 ± 10.3 ng/mg protein; n = 4–5 in each group, P = 0.0008; Fig. 1D). A similar significant increase was found in the tamoxifen-treated animals, E+TAM+Addl70-3 versus E+TAM+AdTIMP-1 (3 ± 0.47 versus 95 ± 9.6 ng/mg protein; n = 4–5 in each group, P < 0.0001; Fig. 1D).

AdMMP-9 injection resulted in increased MMP-9 activity in tumors in vivo. Although gene transduction successfully induced elevated levels of total protein, we investigated if the high protein levels of MMP-9 also resulted in increased MMP-9 activity in tumor tissue in vivo. One week after adenovirus injections, microdialysis catheters were inserted into tumor tissue and perfused with a quenched fluorescent MMP-9 substrate. During perfusion, the substrate will passively diffuse from the inner side of the microdialysis membrane into the tissue. When active MMP-9 is present in the tissue, the substrate will be enzymatically cleaved and fluorescence will develop. The cleaved substrate diffuses from the tissue across the microdialysis membrane into the outgoing perfusion fluid, which can be directly analyzed in a fluorescence spectrophotometer. The developed fluorescence will thereby directly correlate with the MMP-9 activity in vivo. During the experiment, a steady state was reached after 30 minutes in all groups except the control E+Addl70-3. As previously shown, tamoxifen by itself induced an increase in the MMP activity, as measured by relative fluorescence units (RFU) at 120 minutes, E+Addl70-3 versus E+TAM+Addl70-3 (23760 ± 2769 versus 32060 ± 1044 RFU; n = 3–4 in each group, P = 0.048, Fig. 2D). AdMMP-9 injection resulted in a significant increase in MMP activity both in estradiol and E+TAM–treated tumors, E+Addl70-3 versus E+AdMMP-9 (23,760 ± 2,769 versus 39,003 ± 1,955 RFU; n = 3–4 in each group, P = 0.01, Fig. 2D) and E+TAM+Addl70-3 versus E+TAM+AdMMP-9 (32,060 ± 1,044 versus 45,957 ± 1,490; n = 3–4 in each group, P = 0.001, Fig. 2D). AdTIMP-1 resulted in a nonsignificant increase compared with control, E+Addl70-3 versus E+AdTIMP-1 (2,3760 ± 2,769 versus 30,849 ± 1,704 RFU; n = 3–4 in each group, P = 0.07; Fig. 2).

Figure 2.

MMP activity 1 wk after adenovirus injections into tumor tissue. Mice were treated with hormones and virus vectors were injected intratumorally as described in Fig. 1. One week after virus injections, mice were anesthetized and subjected to intratumoral microdialysis. A fluorescent substrate for MMP-9 was perfused in a microdialysis catheter inserted in tumor tissue. The outgoing microdialysates were collected every 30 min and the fluorescence was immediately analyzed in a fluorescence spectrophotometer. Developed fluorescence thereby reflects the MMP activity directly in the tissue in vivo. A, MMP-9 activity in estradiol-treated animals over time. B, MMP-9 activity in E+TAM–treated animals over time. C, MMP-9 activity in both estradiol- and E+TAM–treated animals over time. D, MMP-9 activity at 120 min in all treatment groups. Tamoxifen induced an increase in the MMP-9 activity by itself. Treatment of AdMMP-9 caused a significant increase in the activity of MMP-9 in both in estradiol- and E+TAM–treated tumors. Columns, mean (n = 3–4 in each group); bars, SE.

Figure 2.

MMP activity 1 wk after adenovirus injections into tumor tissue. Mice were treated with hormones and virus vectors were injected intratumorally as described in Fig. 1. One week after virus injections, mice were anesthetized and subjected to intratumoral microdialysis. A fluorescent substrate for MMP-9 was perfused in a microdialysis catheter inserted in tumor tissue. The outgoing microdialysates were collected every 30 min and the fluorescence was immediately analyzed in a fluorescence spectrophotometer. Developed fluorescence thereby reflects the MMP activity directly in the tissue in vivo. A, MMP-9 activity in estradiol-treated animals over time. B, MMP-9 activity in E+TAM–treated animals over time. C, MMP-9 activity in both estradiol- and E+TAM–treated animals over time. D, MMP-9 activity at 120 min in all treatment groups. Tamoxifen induced an increase in the MMP-9 activity by itself. Treatment of AdMMP-9 caused a significant increase in the activity of MMP-9 in both in estradiol- and E+TAM–treated tumors. Columns, mean (n = 3–4 in each group); bars, SE.

Close modal

Intratumoral gene transfer of AdMMP-9 induced tumor regression whereas AdTIMP-1 induced tumor progression in MCF-7 tumors in nude mice. Tumor growth was followed during 24 days after virus injections. MCF-7 tumors require estrogen for growth in nude mice and therefore the estrogen treatment was not discontinued during the experiment. As expected and as previously shown, tamoxifen induced reduced tumor volume in the control groups at the end of treatment, E+Addl70-3 versus E+TAM+Addl70-3 (624 ± 84 versus 89 ± 5.2 mm3; n = 4–8 in each group, P = 0.0031; Fig. 3D). In the estrogen-exposed group, AdMMP-9 injections significantly reduced tumor growth over time and at sacrifice the tumor volume was significantly lower in AdMMP-9–treated tumors compared with control tumors, E+Addl70-3 versus E+AdMMP-9 (624 ± 84 versus 128 ± 27 mm3; n = 4–8 in each group, P = 0.0002; Fig. 3D). Tumors injected with AdTIMP-1 exhibited a similar growth curve compared with Addl70-3–treated tumors and there was no significant difference in tumor volume at sacrifice (Fig. 3D).

Figure 3.

Tumor growth after intratumoral delivery of AdMMP-9 or AdTIMP-1. Mice were treated with hormones and virus vector injections as described in Fig. 1. Tumor growth was followed by measuring the length and width of the tumors every 4 d. As tumors were to some degree attached to the underlying tissue, tumor area was followed during growth. At the end of the experiment, tumors were excised and the length, width, and depth were measured. A, tumor growth in estradiol-treated animals over a period of 24 d. B, tumor growth in E+TAM–treated animals over a period of 24 d. C, tumor growth in all treatment groups over a period of 24 d. D, tumor volume at sacrifice, day 24. Tamoxifen reduced tumor volume in the control group. The treatment with AdMMP-9 induced significant tumor regression compared with controls and was almost as effective as tamoxifen treatment. The combined treatment with tamoxifen and AdMMP-9 significantly reduced the final tumor volume compared with control tamoxifen alone. Tumors injected with AdTIMP-1 exhibited a similar growth curve compared with Addl70-3–treated tumors and there was no significant difference in tumor volume at sacrifice. AdTIMP-1 counteracted to some extent the therapeutic effects of TAM. Points and columns, mean (n = 4–8 in each group); bars, SE.

Figure 3.

Tumor growth after intratumoral delivery of AdMMP-9 or AdTIMP-1. Mice were treated with hormones and virus vector injections as described in Fig. 1. Tumor growth was followed by measuring the length and width of the tumors every 4 d. As tumors were to some degree attached to the underlying tissue, tumor area was followed during growth. At the end of the experiment, tumors were excised and the length, width, and depth were measured. A, tumor growth in estradiol-treated animals over a period of 24 d. B, tumor growth in E+TAM–treated animals over a period of 24 d. C, tumor growth in all treatment groups over a period of 24 d. D, tumor volume at sacrifice, day 24. Tamoxifen reduced tumor volume in the control group. The treatment with AdMMP-9 induced significant tumor regression compared with controls and was almost as effective as tamoxifen treatment. The combined treatment with tamoxifen and AdMMP-9 significantly reduced the final tumor volume compared with control tamoxifen alone. Tumors injected with AdTIMP-1 exhibited a similar growth curve compared with Addl70-3–treated tumors and there was no significant difference in tumor volume at sacrifice. AdTIMP-1 counteracted to some extent the therapeutic effects of TAM. Points and columns, mean (n = 4–8 in each group); bars, SE.

Close modal

In the group of mice treated with a combination of E+TAM, the addition of AdMMP-9 significantly reduced the final tumor volume compared with control vector, E+TAM+Addl70-3 versus E+TAM+AdMMP-9 (89 ± 5.2 versus 41 ± 6 mm3; n = 4–8 in each group, P = 0.001; Fig. 3D). Treatment with E+TAM+AdTIMP-1 increased tumor volume significantly compared with the control E+TAM+Addl70-3 (160 ± 10 versus 89 ± 5.2 mm3; n = 4–8 in each group, P = 0.0033; Fig. 3D).

Release of soluble extracellular VEGF in vivo after AdMMP-9 and AdTIMP-1 therapy. Previous studies have implicated that increased MMP-9 levels lead to increased release of soluble bioactive VEGF from the ECM (25). However, no previous study has sampled VEGF directly from its bioactive site, the extracellular space in vivo after MMP treatment. We have previously used microdialysis as a technique for sampling molecules, including VEGF from tissues in situ (1518, 2628). In this study, we performed microdialysis 1 week after viral infection and measured the extracellular VEGF. In line with our earlier data, the addition of tamoxifen decreased soluble VEGF in all treatment groups, E+Addl-70-3 versus E+TAM+Addl-70-3 (17.8 ± 1.6 versus 11.1 ± 0.95 pg/mL; P = 0.0072; n = 5 in each group; Fig. 4) and E+AdMMP-9 versus E+TAM+AdMMP-9 (24.5 ± 4.9 versus 11.3 ± 1.45 pg/mL; n = 5 in each group, P = 0.02; Fig. 4). In the AdTIMP-1 group, there was a slight but nonsignificant decrease after tamoxifen addition.

Figure 4.

Levels of extracellular VEGF in vivo in tumors after adenovirus injections. Mice were treated with hormones and virus vector injections as described in Fig. 1. One week after virus injections, mice were anesthetized and subjected to intratumoral microdialysis for sampling of VEGF directly in vivo in tumor tissue. In the estradiol-treated animals, the injection of AdMMP-9 caused a slight but nonsignificant increase of VEGF. The addition of tamoxifen decreased extracellular VEGF in all treatment groups. In the AdTIMP-1 group, there was a slight but nonsignificant decrease of extracellular VEGF after tamoxifen addition. Columns, mean (n = 5 in each group); bars, SE.

Figure 4.

Levels of extracellular VEGF in vivo in tumors after adenovirus injections. Mice were treated with hormones and virus vector injections as described in Fig. 1. One week after virus injections, mice were anesthetized and subjected to intratumoral microdialysis for sampling of VEGF directly in vivo in tumor tissue. In the estradiol-treated animals, the injection of AdMMP-9 caused a slight but nonsignificant increase of VEGF. The addition of tamoxifen decreased extracellular VEGF in all treatment groups. In the AdTIMP-1 group, there was a slight but nonsignificant decrease of extracellular VEGF after tamoxifen addition. Columns, mean (n = 5 in each group); bars, SE.

Close modal

In the estradiol-treated animals, the injection of AdMMP-9 caused a slight but nonsignificant increase of VEGF, E+Addl70-3 versus E+AdMMP-9 (17.8 ± 1.6 versus 24.5 ± 4.9 pg/mL; n = 5 in each group, P = 0.19; Fig. 4).

AdMMP-9 treatment increased endostatin levels in vivo. Endostatin is a 20-kDa COOH-terminal cleaved fragment of collagen XVIII, which inhibits angiogenesis by decreasing endothelial cell proliferation and migration and by increasing endothelial cell apoptosis (2931). As endostatin is generated in the extracellular space exclusively, it is imperative to quantify this protein in this compartment. Microdialysis allows for such measurements and this technique was used for sampling of endostatin 1 week after adenovirus tumor injections.

The addition of tamoxifen resulted in a significant increase of endostatin in all treatment groups, E+Addl70-3 versus E+TAM+Addl70-3 (288 ± 15 versus 415 ± 22 pg/mL; n = 5–7 in each group, P = 0.0029; Fig. 5), E+AdMMP9 versus E+TAM+AdMMP9 (489 ± 64 versus 721 ± 53 pg/mL; n = 5–7 in each group, P = 0.019; Fig. 5), and E+AdTIMP-1 versus E+TAM+AdTIMP-1 (263 ± 32 versus 433 ± 48 pg/mL; n = 5–7 in each group, P = 0.019; Fig. 5).

Figure 5.

Levels of endostatin in tumors in vivo after adenovirus injections. Mice were treated with hormones and virus vector injections as described in Fig. 1. One week after virus injections, mice were anesthetized and subjected to intratumoral microdialysis for sampling of soluble endostatin directly in tumor tissue in vivo. The addition of tamoxifen resulted in a significant increase of endostatin in all treatment groups. AdMMP-9 increased the levels of endostatin in a similar fashion as tamoxifen. The increase of endostatin was more pronounced in the TAM+AdMMP-9–treated animals. Columns, mean (n = 5–7 in each group); bars, SE.

Figure 5.

Levels of endostatin in tumors in vivo after adenovirus injections. Mice were treated with hormones and virus vector injections as described in Fig. 1. One week after virus injections, mice were anesthetized and subjected to intratumoral microdialysis for sampling of soluble endostatin directly in tumor tissue in vivo. The addition of tamoxifen resulted in a significant increase of endostatin in all treatment groups. AdMMP-9 increased the levels of endostatin in a similar fashion as tamoxifen. The increase of endostatin was more pronounced in the TAM+AdMMP-9–treated animals. Columns, mean (n = 5–7 in each group); bars, SE.

Close modal

AdMMP-9 increased the levels of endostatin in a similar fashion as tamoxifen treatment, E+Addl70-3 versus E+AdMMP-9 (288 ± 15 versus 489 ± 64 pg/mL; n = 5–7 in each group, P = 0.04; Fig. 5). The increase of endostatin was more pronounced in the tamoxifen-treated animals, E+TAM+Addl70-3 versus E+TAM+AdMMP-9 (415 ± 22 versus 721 ± 53 pg/mL; n = 5–7 in each group, P = 0.0008; Fig. 5).

AdMMP-9–treated tumors exhibited decreased microvessel area. To investigate the functional effects of the MMP-9–induced increase of endostatin levels, microvessel area was scored in tumor sections from all treatment groups. Tumors were collected at the end of treatment at day 24. Tumors, which exhibited high levels of endostatin 1 week after adenovirus injections, had a decreased microvessel area at the end of therapy. Treatments with tamoxifen or AdMMP-9 caused a similar decrease in microvessel area compared with control tumors, E+Addl70-3 versus E+TAM+Addl70-3 (1.72 ± 0.2 versus 0.3 ± 0.033%; n = 9 in each group, P < 0.0001; Fig. 6) and E+Addl70 versus E+AdMMP-9 (1.72 ± 0.2 versus 0.2 ± 0.03%; n = 9 in each group, P < 0.0001; Fig. 6).

Figure 6.

Microvessel density after adenovirus injections. Mice were treated with hormones and virus vector injections as described in Fig. 1. At day 24, the animals were sacrificed and tumors were excised and fixed in formalin. Tumor sections were stained with the anti–von Willebrand factor, and the vessel area was measured on tumor sections. Tumor vessel density quantification was conducted in a blinded manner. Ten randomly selected areas of three different tumors in each group were counted. Representative sections are shown. Tamoxifen and AdMMP-9 treatment both induced a similar decrease in microvessel area compared with control tumors. Columns, mean; bars, SE.

Figure 6.

Microvessel density after adenovirus injections. Mice were treated with hormones and virus vector injections as described in Fig. 1. At day 24, the animals were sacrificed and tumors were excised and fixed in formalin. Tumor sections were stained with the anti–von Willebrand factor, and the vessel area was measured on tumor sections. Tumor vessel density quantification was conducted in a blinded manner. Ten randomly selected areas of three different tumors in each group were counted. Representative sections are shown. Tamoxifen and AdMMP-9 treatment both induced a similar decrease in microvessel area compared with control tumors. Columns, mean; bars, SE.

Close modal

E+AdTIMP-1–treated tumors showed a nonsignificant decrease in microvessel area compared with E+Addl70-3 (0.34 ± 0.06 versus 1.72 ± 0.2%; n = 9 in each group, P = 0.09; Fig. 6).

MMP activity is generally discussed in terms of the ability of MMPs to induce tissue destruction that promotes tumor invasion and angiogenesis. However, here we show that MMP-9 can induce quite the opposite effect. In our model of established breast cancer in nude mice, overexpression of MMP-9 by adenoviral gene transfer caused tumor regression and decreased angiogenesis. A possible mechanism behind this finding may be increased release of antiangiogenic fragments such as endostatin, which was found in the present study. Contrary to AdMMP-9 gene transfer, we show that injection of AdTIMP-1 into the breast cancer tumors had no effect on tumor growth compared with estradiol-treated animals injected with the control vector, Addl70-3. Cotreatment with AdMMP-9 and tamoxifen had a synergistic effect on tumor regression compared with either treatment alone, whereas AdTIMP-1 counteracted the beneficial effects of tamoxifen on tumor regression and caused tumor progression compared with the control animals treatment with Addl70-3+TAM.

When using viral gene transfer, one obstacle may be the difficulty in measuring the efficacy of the infection by detection of bioactive protein expression within the tissue. In the present study, we do not only detect increased levels of the proteins after gene delivery but we can also in a unique manner directly quantify the activity of the adenovirally delivered MMP-9. By perfusing microdialysis catheters with the fluorescent substrate for MMP-9, we show that the injection AdMMP-9 directly into the tumor induced increased MMP activity. The in vivo MMP activity after AdMMP-9 gene delivery was enhanced by tamoxifen, and animals treated with the combination of AdMMP-9 and tamoxifen exhibited a synergistic effect on tumor regression. This may be explained by the ability of tamoxifen to induce cell cycle arrest and thus slow cell proliferation in breast cancer cells. Genes delivered via adenoviruses remain extrachromosomal and a cell cycle arrest will thereby prolong gene expression. A prolonged gene expression by tamoxifen may also explain that tumors treated with AdTIMP-1 counteracted the therapeutic tamoxifen effects on breast cancer growth and instead caused tumor progression despite adequate antiestrogen doses of tamoxifen.

MMP activity has generally been positively associated with increased invasiveness of cancers and the angiogenic switch by releasing bioactive molecules such as VEGF from the ECM and thereby enhancing tumor angiogenesis and progression (4, 5, 25, 32). However, no previous study has directly measured VEGF released in tumor tissue in vivo after MMP-9 treatment. We have previously shown that microdialysis is a valid technique for sampling of extracellular soluble VEGF in vivo in both normal and cancerous breast tissue in animals as well as in humans (1518, 26, 28). In the present study, we performed microdialysis and sampled VEGF 1 week after AdMMP-9 and AdTIMP-1 injections. As previously shown, the addition of tamoxifen in all treatment groups, Addl70-3, AdMMP-9, and AdTIMP-1, caused a decrease of extracellular soluble VEGF in situ. Interestingly, we found, as previously suggested (25), a slight but nonsignificant increase of extracellular VEGF after AdMMP-9 intratumoral injections in animals without tamoxifen treatment. This increase, however, was not of enough biological importance to enhance angiogenesis in these tumors because all of the AdMMP-9–injected tumors exhibited tumor regression and decreased angiogenesis.

An increasing number of recent studies have suggested that MMPs may negatively regulate tumor progression and angiogenesis through the release of angiogenic inhibitors. MMP-9–deficient mice lacking the antiangiogenic fragment tumstatin exhibit accelerated tumor growth and increased angiogenesis when inoculated with Lewis lung carcinomas (9). Moreover, decreased tumor growth and angiogenesis have been shown in MMP-9–overexpressing mice, effects that were attributed to increased production of angiostatin (33, 34). We have previously found that tamoxifen induces increased MMP-9 activity, which resulted in increased endostatin generation in breast cancer in vivo and thereby decreased angiogenesis and tumor regression (10, 19). Previous studies have investigated knockout MMP-9 mice or MMP-9–overexpressing mice and cell lines and modulations of endogenous MMP-9. To our knowledge, no previous study has investigated direct treatment of established solid tumors with MMP-9. Our results in the present study confirm previous reports of a beneficial effect of endogenous MMP-9; therefore, it is plausible that this effect may also be found in established tumors treated with MMP-9. Our results suggest that one of the mechanisms involved in tumor regression may be the release of endostatin after MMP-9 treatment. In our model, intratumoral AdMMP-9 injection resulted in a significant increase of released interstitial endostatin. The levels of endostatin were almost doubled after AdMMP-9 injections and although some VEGF was simultaneously released, the biological consequence in the AdMMP-9–injected tumors was decreased angiogenesis and tumor regression. Moreover, our study shows that elevating MMP-9 levels in established late-stage tumors may be an effective tumor treatment and not only beneficial in very early stage cancer disease as previously suggested (35).

Tumor progression in animals treated with AdTIMP-1 strengthens the notion of a beneficial therapeutic effect of AdMMP-9 in our breast cancer model. TIMP-1 is an endogenous natural inhibitor of MMP-9, and in tumors injected with AdTIMP-1, the levels of MMP-9 protein analyzed by ELISA were decreased compared with Addl70-3–treated controls. The role of TIMP-1 in cancer biology is complex and contradictory. Similar to MMPs, both tumor-protective actions and tumor-enhancing action have been reported (20, 3638). Our results in the present study are in line with recent studies emphasizing TIMP-1 as a tumor-enhancing factor in cancer progression (20, 36, 37).

Taken together, our data illustrate the complex role of the MMP/TIMP system in tumor biology. MMPs may facilitate ECM degradation, thereby promoting tumor invasion and angiogenesis; however, they may also block angiogenesis by producing antiangiogenic fragments such as endostatin. These latter effects may explain the negative results in clinical trials using MMP inhibitors (35, 39). For successful targeting of the MMP/TIMP system in future cancer therapy, a better understanding of the function of specific MMPs and TIMPs in specific tumors is necessary. Exploring MMP/TIMP effects at different tumor stages as well as a possible dose-response relationship of tumor promotion/inhibition is also needed. Hitherto, the main focus on MMP function has been on their modulation of tumor invasiveness and angiogenesis and to a much lesser extent on their protective roles in cancer progression. It is clear that these effects are potent and need much more attention in future research. Elevating levels of MMP-9 and thereby the generation of angiogenesis inhibitors either directly by gene therapy, as shown in this article, or protein delivery or indirectly via small molecular inducers of MMP-9, may be one promising future therapeutic approach against breast cancer that should be explored.

Grant support: Swedish Research Council grant 60294601 (C. Dabrosin); and Swedish Cancer Society grants 060036, 070012, and 070049 (C. Dabrosin).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1
Woodhouse EC, Chuaqui RF, Liotta LA. General mechanisms of metastasis.
Cancer
1997
;
80
:
1529
–37.
2
Visse R, Nagase H. Matrix metalloproteinases and tissue inhibitors of metalloproteinases: structure, function, and biochemistry.
Circ Res
2003
;
92
:
827
–39.
3
Chang C, Werb Z. The many faces of metalloproteases: cell growth, invasion, angiogenesis and metastasis.
Trends Cell Biol
2001
;
11
:
S37
–43.
4
Duffy MJ, Maguire TM, Hill A, McDermott E, O'Higgins N. Metalloproteinases: role in breast carcinogenesis, invasion and metastasis.
Breast Cancer Res
2000
;
2
:
252
–7.
5
Liotta LA, Stetler-Stevenson WG. Metalloproteinases and cancer invasion.
Semin Cancer Biol
1990
;
1
:
99
–106.
6
Stetler-Stevenson WG, Yu AE. Proteases in invasion: matrix metalloproteinases.
Semin Cancer Biol
2001
;
11
:
143
–52.
7
Cornelius LA, Nehring LC, Harding E, et al. Matrix metalloproteinases generate angiostatin: effects on neovascularization.
J Immunol
1998
;
161
:
6845
–52.
8
Ferreras M, Felbor U, Lenhard T, Olsen BR, Delaisse J. Generation and degradation of human endostatin proteins by various proteinases.
FEBS Lett
2000
;
486
:
247
–51.
9
Hamano Y, Zeisberg M, Sugimoto H, et al. Physiological levels of tumstatin, a fragment of collagen IV α3 chain, are generated by MMP-9 proteolysis and suppress angiogenesis via αV β3 integrin.
Cancer Cell
2003
;
3
:
589
–601.
10
Nilsson UW, Dabrosin C. Estradiol and tamoxifen regulate endostatin generation via matrix metalloproteinase activity in breast cancer in vivo.
Cancer Res
2006
;
66
:
4789
–94.
11
Lopez-Otin C, Matrisian LM. Emerging roles of proteases in tumour suppression.
Nat Rev Cancer
2007
;
7
:
800
–8.
12
Martin MD, Matrisian LM. The other side of MMPs: protective roles in tumor progression.
Cancer Metastasis Rev
2007
;
26
:
717
–24.
13
Coussens LM, Fingleton B, Matrisian LM. Matrix metalloproteinase inhibitors and cancer: trials and tribulations.
Science
2002
;
295
:
2387
–92.
14
Overall CM, Kleifeld O. Tumour microenvironment—opinion: validating matrix metalloproteinases as drug targets and anti-targets for cancer therapy.
Nat Rev Cancer
2006
;
6
:
227
–39.
15
Dabrosin C. Variability of vascular endothelial growth factor in normal human breast tissue in vivo during the menstrual cycle.
J Clin Endocrinol Metab
2003
;
88
:
2695
–8.
16
Dabrosin C. Positive correlation between estradiol and vascular endothelial growth factor but not fibroblast growth factor-2 in normal human breast tissue in vivo.
Clin Cancer Res
2005
;
11
:
8036
–41.
17
Dabrosin C, Margetts PJ, Gauldie J. Estradiol increases extracellular levels of vascular endothelial growth factor in vivo in murine mammary cancer.
Int J Cancer
2003
;
107
:
535
–40.
18
Garvin S, Dabrosin C. Tamoxifen inhibits secretion of vascular endothelial growth factor in breast cancer in vivo.
Cancer Res
2003
;
63
:
8742
–8.
19
Nilsson UW, Garvin S, Dabrosin C. MMP-2 and MMP-9 activity is regulated by estradiol and tamoxifen in cultured human breast cancer cells.
Breast Cancer Res Treat
2007
;
102
:
253
–61.
20
Partridge JJ, Madsen MA, Ardi VC, et al. Functional analysis of MMPs and TIMPs differentially expressed by variants of human HT-1080 fibrosarcoma exhibiting high and low levels of intravasation and metastasis.
J Biol Chem
2007
;
282
:
35964
–77.
21
Ng P, Parks RJ, Cummings DT, et al. A high-efficiency Cre/loxP-based system for construction of adenoviral vectors.
Hum Gene Ther
1999
;
10
:
2667
–72.
22
Bonniaud P, Margetts PJ, Kolb M, et al. Adenoviral gene transfer of connective tissue growth factor in the lung induces transient fibrosis.
Am J Respir Crit Care Med
2003
;
168
:
770
–8.
23
Dabrosin C, Palmer K, Muller WJ, Gauldie J. Estradiol promotes growth and angiogenesis in polyoma middle T transgenic mouse mammary tumor explants.
Breast Cancer Res Treat
2003
;
78
:
1
–6.
24
Schor AM, Pendleton N, Pazouki S, et al. Assessment of vascularity in histological sections: effects of methodology and value as an index of angiogenesis in breast tumours.
Histochem J
1998
;
30
:
849
–56.
25
Bergers G, Brekken R, McMahon G, et al. Matrix metalloproteinase-9 triggers the angiogenic switch during carcinogenesis.
Nat Cell Biol
2000
;
2
:
737
–44.
26
Bergman Jungestrom M, Thompson LU, Dabrosin C. Flaxseed and its lignans inhibit estradiol-induced growth, angiogenesis, and secretion of vascular endothelial growth factor in human breast cancer xenografts in vivo.
Clin Cancer Res
2007
;
13
:
1061
–7.
27
Dabrosin C. Microdialysis—an in vivo technique for studies of growth factors in breast cancer.
Front Biosci
2005
;
10
:
1329
–35.
28
Dabrosin C, Chen J, Wang L, Thompson LU. Flaxseed inhibits metastasis and decreases extracellular vascular endothelial growth factor in human breast cancer xenografts.
Cancer Lett
2002
;
185
:
31
–7.
29
Abdollahi A, Hahnfeldt P, Maercker C, et al. Endostatin's antiangiogenic signaling network.
Mol Cell
2004
;
13
:
649
–63.
30
O'Reilly MS, Boehm T, Shing Y, et al. Endostatin: an endogenous inhibitor of angiogenesis and tumor growth.
Cell
1997
;
88
:
277
–85.
31
Abdollahi A, Hlatky L, Huber PE. Endostatin: the logic of antiangiogenic therapy.
Drug Resist Updat
2005
;
8
:
59
–74.
32
Stetler-Stevenson WG, Hewitt R, Corcoran M. Matrix metalloproteinases and tumor invasion: from correlation and causality to the clinic.
Semin Cancer Biol
1996
;
7
:
147
–54.
33
Pozzi A, LeVine WF, Gardner HA. Low plasma levels of matrix metalloproteinase 9 permit increased tumor angiogenesis.
Oncogene
2002
;
21
:
272
–81.
34
Pozzi A, Moberg PE, Miles LA, et al. Elevated matrix metalloprotease and angiostatin levels in integrin α1 knockout mice cause reduced tumor vascularization.
Proc Natl Acad Sci U S A
2000
;
97
:
2202
–7.
35
Zucker S, Cao J, Chen WT. Critical appraisal of the use of matrix metalloproteinase inhibitors in cancer treatment.
Oncogene
2000
;
19
:
6642
–50.
36
Kopitz C, Gerg M, Bandapalli OR, et al. Tissue inhibitor of metalloproteinases-1 promotes liver metastasis by induction of hepatocyte growth factor signaling.
Cancer Res
2007
;
67
:
8615
–23.
37
Lipton A, Ali SM, Leitzel K, et al. Elevated plasma tissue inhibitor of metalloproteinase-1 level predicts decreased response and survival in metastatic breast cancer.
Cancer
2007
;
109
:
1933
–9.
38
Miyagi M, Aoyagi K, Kato S, Shirouzu K. The TIMP-1 gene transferred through adenovirus mediation shows a suppressive effect on peritoneal metastases from gastric cancer.
Int J Clin Oncol
2007
;
12
:
17
–24.
39
Moore MJ, Hamm J, Dancey J, et al. Comparison of gemcitabine versus the matrix metalloproteinase inhibitor BAY 12-9566 in patients with advanced or metastatic adenocarcinoma of the pancreas: a phase III trial of the National Cancer Institute of Canada Clinical Trials Group.
J Clin Oncol
2003
;
21
:
3296
–302.