Advertisement
Cell, Tumor and Stem Cell Biology

Matrix Metalloproteinase-9 from Bone Marrow–Derived Cells Contributes to Survival but not Growth of Tumor Cells in the Lung Microenvironment

Heath B. Acuff, Kathy J. Carter, Barbara Fingleton, D. Lee Gorden and Lynn M. Matrisian
DOI: 10.1158/0008-5472.CAN-05-2502 Published January 2006

Abstract

The role of specific stromal-derived matrix metalloproteinases (MMPs) was analyzed in experimental metastasis assays in wild-type and either MMP-9, MMP-7, or MMP-2 null mice. MMP-9 null mice showed an 81% reduction in Lewis lung carcinoma tumor number, whereas MMP-7 null mice showed a 42% increase in tumor number, and there was no difference in tumor number in MMP-2 null mice compared with wild-type controls. Similarly, in an orthotopic model of lung cancer, 50% fewer MMP-9 null mice were able to establish tumors in the lung compared with control mice, although the size of the tumors was not different. The effect of MMP-9 on lung tumor colonization was dependent on the expression of MMP-9 from bone marrow–derived cells and is most likely contributed by neutrophils. To examine temporal effects of stromal MMP-9, bioluminescence imaging from luciferase-expressing human lung cancer–derived A549 cells revealed that there were fewer tumor cells in the lungs of MMP-9 null mice as early as 19 hours after injection compared with control mice, with no difference in subsequent growth rates. Six hours after injection of tumor cells, MMP-9 null mice showed a 4-fold increase in the percent of tumor cells undergoing apoptosis compared with control mice. We conclude that MMP-9 from the bone marrow contributes to the early survival and establishment of tumors in the lung and has no effect on subsequent growth. These results provide insights into the failure of MMP inhibitors in clinical trials in patients with late-stage lung cancer. (Cancer Res 2006; 66(1): 259-66)

  • MMP9
  • metastasis
  • lung cancer
  • apoptosis
  • bone marrow derived

Introduction

Tumors develop and progress through a multistep process involving the interaction of tumor cells and the host microenvironment. Communication between tumor cells and host components, such as immune cells, fibroblasts, and endothelial cells, contribute to the progression of cancer from its initial growth at a primary site in the body to its metastasis to distant organs ( 1). Matrix metalloproteinases (MMPs) are a family of zinc-dependent endopeptidases that have been shown to contribute to tumor progression and metastasis ( 2). Both tumor and host-derived MMPs contribute to multiple stages of tumor progression through their ability to degrade the basement membrane and components of the extracellular matrix as well as nonmatrix substrates ( 3, 4).

MMPs are expressed by tumor cells as well as by the stromal component in several different tumor types ( 5). The importance of the role of host-derived MMPs in cancer has been examined in several tumor models ( 6– 9). In a model of skin carcinogenesis, MMP-9 from bone marrow–derived cells contributed to hyperproliferation and tumor incidence ( 6). In a model of pancreatic islet tumorigenesis, MMP-2 and MMP-9 contributed to tumor growth, and stromal MMP-9 had additional effects on the angiogenic switch ( 7). In addition, stromal fibroblast–contributed MMP-11 affects the growth of transplanted breast cancer cells ( 8). Taken together, these data indicate that host-derived MMPs remain an important contributor to different stages of tumor progression.

Despite the clear evidence that MMPs play a significant role in tumor progression, the results from the clinical trials of MMP inhibitors (MMPI) indicate that MMPIs are not a useful therapeutic approach ( 10). On evaluating the clinical trial results, it is clear that our understanding of MMP biology is incomplete. To better understand the role of host MMPs in the establishment of lung tumors, we have returned to experimental metastasis assays and devised an orthotopic model of lung cancer for use in MMP null mice, specifically focusing on the role of MMP-9 in the establishment of lung tumors.

Materials and Methods

Cell Culture

Lewis lung carcinoma cells (LLC) were obtained from Dr. John Caterina (Bethel College, McKenzie, TN) from a stock at the NIH. Luciferase expressing human A549 cells (LUC-A549) were generously provided by Dr. Jerry Shay (University of Texas Southwestern Medical Center). A549 cells were obtained from American Type Culture Collection (Manassas, VA). LLC were cultured in DMEM (Life Technologies Bethesda Research Laboratories, Long Island, NY) supplemented with 10% fetal bovine serum (FBS; Atlanta Biologicals, Norcross, GA) at 37°C, 5% CO2. LUC-A549 cells were cultured in F-12 Nutrient Mixture (HAM; Life Technologies Bethesda Research Laboratories) supplemented with 10% FBS (Atlanta Biologicals) at 37°C, 5% CO2. A549 cells were cultured in F12-K Nutrient Mixture (Kaighn's modification; Life Technologies Bethesda Research Laboratories) supplemented with 10% FBS (Atlanta Biologicals) and 1.5 g/L sodium bicarbonate at 37°C, 5% CO2.

Animal Models

Experimental metastasis assay. MMP-2 ( 11), MMP-7 ( 12), and MMP-9 ( 13) null mice were generated as described previously and backcrossed into a C57Bl/6 background to N>10. C57Bl/6 mice (Harlan, Indianapolis, IN) were used as wild-type controls. Rag2 null mice maintained on a C57Bl/6 background were purchased from Taconic (Germantown, NY) and crossed into MMP-9 null mice to obtain double Rag2/MMP-9 knockout mice. At 6 to 8 weeks of age, tumor cells were injected into the tail vein of MMP null or C57Bl/6 wild-type mice or Rag2 null and Rag2/MMP-9 null mice. After 2 weeks for the syngeneic model or 5 weeks for the immunocompromised model, the mice were sacrificed, and surface lung tumors were counted and measured.

Orthotopic model. A549 cells were injected into the lung through the trachea of 6-week-old C57Bl/6 Rag2 null or MMP-9/Rag2 null mice as a modification of the procedure described previously ( 14). The intrabronchial injection is done using a 1-in. piece of Tygon microbore tubing (0.76 mm outer diameter) attached to a 2-in., 27-gauge, blunt-end needle (Popper, New Hyde Park, NY). A small incision is made in the neck of the mouse and in the trachea. The tubing is inserted and fed down the trachea; 1 × 106 tumor cells/100 μL PBS are injected into the bronchus. The incision in the skin is closed by wound clips. A primary tumor in the right lung of the mouse is established after 5 weeks in ∼80% of the animals. Adaptation of this model to LLC cells was not successful due to the appearance of tracheal tumors that caused premature death of the mice.

Bioluminescent Imaging

Bioluminescent imaging was detected from luciferase expressing A549 cells after injection of the cells i.v. into mice. Beetle Luciferin (Promega, Madison, WI) was used as the substrate for the luciferase expressing tumor cells and injected i.v. at a concentration of 150 mg/kg in PBS, 1 to 2 minutes before imaging. Mice were anesthetized using 2% isofluorane and imaged at 2, 19, 43 hours after tumor cell injection and thereafter at weekly time points for 5 weeks using a cooled CCD camera (IVIS system, Xenogen, Alameda, CA). Exposure times ranged from 1 minute to 1 second. Images were quantified as photons/s using the Living Image software from Xenogen.

Bone Marrow Transplantation Studies

Bone marrow cells where collected from 6-week-old C57Bl/6 wild-type or MMP-9 null mice by flushing femurs and tibias with sterile PBS. Recipient mice were given 100 mg/L neomycin and 10 mg/L polymyxin B sulfate in acidified water (pH 2.7), 1 week before the transplantations and for the remainder of the study. Recipient mice, either C57Bl/6 wild type or MMP-9 null, were lethally irradiated (600 rad followed 3 hours later with 400 rad) using a cesium γ source. Four hours later, C57Bl/6 wild-type or MMP-9 null mice were injected with 1 × 106 bone marrow cells/100 μL PBS by tail vein from wild-type or MMP-9 null mice. Ten weeks after bone marrow transplantation, reconstitution of MMP-9 in the bone marrow was confirmed by gelatin zymography.

Histologic Analysis

Formalin-fixed, paraffin-embedded tissues were dewaxed, hydrated through graded alcohols, stained with H&E, dehydrated, and coverslipped.

Immunohistochemistry

Five-micrometer formalin-fixed, paraffin-embedded sections were dewaxed, hydrated through graded ethanols, treated with 0.6% hydrogen peroxide in methanol (to destroy endogenous peroxidases), and microwaved in 10 mmol/L sodium citrate for 1 minute at high power and 9 minutes at medium power for antigen retrieval. Sections were exposed to blocking solution (10% rabbit serum/TBS) for 1 hour. Sections were incubated overnight in blocking solution with rat monoclonal anti-neutrophil (1:100 dilution; Serotec, Inc., Raleigh, NC), rabbit anti-von Willebrand factor (DAKO Corp., Carpinteria, CA) or appropriate control IgG (PharMingen, Mississauga, Ontario, Canada). Sections were washed with TBST (150 mmol/L NaCl, 10 mmol/L Tris, and 0.05% Tween 20) and incubated with biotinylated anti-rat IgG (1:1,000; Vector Laboratories, Burlingame, CA) for 1 hour. Labeled cells were visualized using an avidin-biotin peroxidase complex (Vectastain Avidin-Biotin Complex kit, Vector Laboratories) and 3,3′-diaminobenzidine tetrahydrochloride substrate (Sigma, St. Louis, MO). Sections were counterstained with hematoxylin. For neutrophil quantitation, five ×10 fields per lung were analyzed in three control and four MMP-9 null mice. Von Willebrand factor staining was assessed by counting multiple areas defined by Metamorph Imaging System (Universal Imaging Corp., Downingtown, PA) in seven mice per experimental group.

Zymography

Bone marrow cells were washed once with PBS and lysed in lysis buffer [1% Triton 100-X, 0.1% SDS, 1% deoxycholic acid, 50 mmol/L Tris (pH 7.5), 0.15 mol/L NaCl, 5 mmol/L EDTA, supplemented with 20 μg/mL leupeptin, 20 μg/mL aprotinin, 20 μg/mL phenylmethylsulfonyl fluoride] on ice for 10 minutes. Protein concentrations from bone marrow lysates were determined using the bicinchoninic acid assay (Pierce, Rockford, IL). Equal amounts of protein were loaded on 10% SDS-polyacrylamide gels containing 0.1 mg/mL gelatin and run at 100 V for 4 hours in nonreducing conditions. After electrophoresis, gels were washed with 2.5 % Triton 100-X twice for 15 minutes and incubated at 37°C in substrate buffer [50 mmol/l Tris-HCl (pH 7.5), 10 mmol/L CaCl2] overnight. Gels were stained with 0.5% Coomassie blue, 50% methanol, and 10% acetic acid and destained in 50% methanol, 10% acetic acid.

Clodronate Treatment

Liposomal clodronate was prepared as described previously ( 15). Eighty microliters of the liposomal clodronate or liposomal PBS solution was injected i.t. into C57Bl/6 wild-type mice. Twenty-four hours later, LLC cells were injected i.v. bronchoalveolar lavage fluid was collected 48 hours after liposomal clodronate administration. Mouse tracheas were cannulated with a 20-gauge, blunt-end needle attached to a 1-mL syringe, and the lungs were instilled with sterile PBS until a total lavage volume of 3 mL was collected. The bronchoalveolar lavage was centrifuged at 400 rpm for 5 minutes, and the cell pellet was resuspended in 1 mL of 3% FBS/PBS. Total cell counts were determined using a grid hemocytometer. Differential cell counts were obtained by staining cytocentrifuge slides with a modified Wright's stain (Richard Allan-Scientific, Kalamazoo, MI).

Apoptosis Analysis

Terminal deoxynucleotidyl transferase–medtaied nick-end labeling (TUNEL) was done on frozen sections using the ApoTag Fluorescein In situ Apoptosis Detection kit (Chemicon International, Temecula, CA) according to manufacturer's directions. Nuclei were counterstained with Hoechst 33258.

Labeling of Tumor Cells

CellTracker probe (CellTracker Red CMPTX; Molecular Probes, Eugene, OR; 10 μmol/L) was added to the cell culture medium at a dilution of 1:1,000. Cells were labeled for 30 minutes. Labeled cells were visualized using Axioplan 2 imaging microscope (Zeiss, Thornwood, NY) and Openlab 4.0.2 software.

Statistical Analysis

All data generated using the experimental metastasis assays were analyzed using a nonparametrical (Mann-Whitney) method. Data generated using the orthotopic model were analyzed using Fisher's exact test (Statview software, SAS Institute, Cary, NC).

Results

Host MMPs contribute to the establishment of tumors in the lung. To determine if host-derived MMPs contribute to lung tumor colonization, experimental metastasis assays were done by injecting 3 × 105 LLC cells into the tail vein of syngeneic C57Bl/6 wild-type mice or MMP-2, MMP-7, or MMP-9 null mice. Two weeks after tail vein injection, the surface lung tumors were counted and measured. In contrast to previous reports ( 16), we saw no difference in tumor number comparing control and MMP-2 null mice ( Fig. 1A ). Surprisingly, MMP-7 null mice showed a 42% increase in tumor number compared with control mice (P = 0.01), suggesting a protective effect of MMP-7 on lung colonization ( Fig. 1C). Most interestingly, MMP-9 null mice showed an 81% reduction in tumor number compared with wild-type mice (P = 0.0002; Fig. 1E). The results from the MMP-9 null mice are in agreement with those previously reported ( 17), although the effect of MMP-9 ablation on tumor number was greater in our study. There was no significant difference in tumor size in any of the MMP null mice compared with wild-type mice ( Fig. 1B, D, and F), suggesting that these MMPs are not having an effect on the growth of tumors that successfully establish in the lung. These experiments show that different host-derived MMPs can have different effects on metastasis to the lung. Because MMP-9 null mice showed the most dramatic phenotype, we chose to pursue this observation to understand the mechanism by which MMP-9 contributes to lung tumor development.

Figure 1.
Figure 1.

Host MMP contributions to lung tumor formation. A, number of surface lung tumors in control (n = 12) and MMP-2 null (n = 12) mice 2 weeks after injection of 3 × 105 LLC cells i.v. (P = not significant). B, tumor size in control and MMP-2 null mice (P = not significant). C, number of surface lung tumors in control (n = 14) and MMP-7 null (n = 14) mice 2 weeks after injection of 3 × 105 LLC cells i.v. *, P = 0.01. D, tumor size in control and MMP-7 null mice (P = not significant). E, number of surface lung tumors in control (n = 13) and MMP-9 null (n = 11) mice 2 weeks after injection of 3 × 105 LLC cells i.v. *, P = 0.0002. F, tumor size in control and MMP-9 null mice (P = not significant).

MMP-9 contributes to the establishment of a primary tumor in the lung. To test if host MMP-9 also has an effect on the establishment of primary tumors in the lung, an orthotopic model of lung cancer was established by injecting human lung carcinoma A549 cells directly into the lung via the trachea. This procedure produces distinct primary tumors in the lung in 5 weeks ( Fig. 2A ). A549 cells (1 × 106) were injected into either immunocompromised Rag2 null mice or immunocompromised Rag2/MMP-9 null mice, and tumor growth was examined after 5 weeks. Tumors develop in 82% of wild-type mice compared with 44% of MMP-9 null mice, suggesting a contribution of host MMP-9 to initial tumor take ( Fig. 2B). However, when comparing tumors that develop in either the wild-type or MMP-9 null animals, ∼50% in each case can be classified as microscopic (<0.1 mm diameter), and ∼50% are macroscopic with sizes ranging from 0.5 to 87 mm3 ( Fig. 2C). This suggests that the presence or absence of MMP-9 is not the prime determinant in tumor progression from microscopic to macroscopic tumors. These data are similar to those obtained with experimental metastasis assays, in which the initial tumor “take” or colonization is dependent on MMP-9 expression, whereas subsequent growth is not influenced by MMP-9.

Figure 2.
Figure 2.

Host MMP-9 contributes to establishment of a primary tumor in the lung. A, histology of lung 5 weeks after orthotopic injection of 1 × 106 A549 cells (H&E). T, tumor. B, % control (n = 17) and MMP-9 null mice (n = 18) that establish a primary tumor in the lung after orthotopic injection of A549 cells. *, P = 0.02, two-tailed Fisher's exact test. C, % tumors in control and MMP-9−/− mice that were classified as microscopic (micro; <0.1 mm3) or macroscopic (macro; 0.5-87.5 mm3).

MMP-9 from the bone marrow contributes to lung tumor formation. MMP-9 is expressed in the stromal component in human lung tumors, including localization in inflammatory cells ( 18, 19), and inflammatory cells have been shown to be a source of MMP-9 in several cancer models ( 6, 7, 20, 21). To determine if MMP-9 from bone marrow–derived cells plays a role in lung metastasis, bone marrow transplantation assays were done. Lethally irradiated mice received either 1 × 106 wild-type or MMP-9 null bone marrow cells i.v. into the tail vein. Ten weeks after transplantation, 1 × 105 LLC cells were injected i.v. into wild-type mice receiving either wild-type or MMP-9 null bone marrow cells or MMP-9 null mice receiving either wild-type or MMP-9 null bone marrow cells. Two weeks later, lungs were harvested, and surface lung tumors were counted. Bone marrow reconstitution was determined by gelatin zymography to detect MMP-9 from bone marrow cell lysates in transplanted mice ( Fig. 3A ). MMP-9 was present in wild-type mice that received wild-type bone marrow and absent in MMP-9 null mice receiving MMP-9 null bone marrow. Two of the five wild-type mice receiving MMP-9 null bone marrow showed complete reconstitution, and three showed partial reconstitution. MMP-9 was detected in the bone marrow of all MMP-9 null mice that received wild-type bone marrow but at different levels. The number of lung nodules that developed in MMP-9 null mice reconstituted with MMP-9 null bone marrow was 14% of those that developed in wild-type mice reconstituted with wild-type bone marrow ( Fig. 3B), similar to the results obtained with intact wild-type or MMP-9 null mice ( Fig. 1E). In mice reconstituted with bone marrow of a different genotype, the number of lung nodules was directly proportional to the level of MMP-9 in the bone marrow ( Fig. 3B). Thus, the effect of MMP-9 in the experimental metastasis assay was the result of MMP-9 contributed by bone marrow–derived cells.

Figure 3.
Figure 3.

MMP-9 levels from bone marrow–derived cells correlate with lung tumor formation. A, gelatin zymography of bone marrow cell lysates showing MMP levels in representative samples of wild-type (WT) and MMP-9 null (MMP−/−) mice after reconstitution with either MMP-9-positive or MMP-9 null bone marrow cells. ND, not detected; M, medium; H, high levels of bone marrow MMP-9. Duplicate zymograms in the presence of EDTA confirm metal-dependent enzymatic activity. B, quantitation of surface lung tumors in control or MMP-9 null mice reconstituted with MMP-9-positive or MMP-9 null bone marrow cells.

MMP-9 contributes to the early establishment of tumors in the lung and has no effect on tumor growth. To further examine the effect of host MMP-9 on tumor growth and establishment in the lung, the experimental metastasis assay was done using luciferase expressing human A549 cells (LUC-A549) in immunocompromised Rag2 null and Rag2/MMP-9 null mice. Because the tumor cells express luciferase, tumor growth rates in the lung could be measured over time by measuring luciferase expression by bioluminescence. LUC-A549 cells (2 × 106) were injected i.v. into the tail vein of Rag2 null and Rag2/MMP-9 null mice. Bioluminescent images were obtained 2, 19, and 43 hours and weekly up to 5 weeks after injection of tumor cells ( Fig. 4A ). Lung-localized luciferase activity was identical in wild-type and MMP-9-null mice at the 2-hour time point, but as early as 19 hours and continuing throughout the experiment, there was a statistically significant decrease in the bioluminescent signal in MMP-9 null compared with wild-type mice ( Fig. 4B). After 5 weeks, mice were sacrificed, and surface lung tumors were counted and measured. MMP-9 expression was confirmed by immunohistochemistry in the A549 cells and in stromal cells in the lungs of control but not MMP-9 null mice (data not shown). MMP-9 null mice showed a 97% reduction in tumor number compared with wild-type mice ( Fig. 4C). Of the tumors that developed, there was no significant difference in tumor size between control and MMP-9 null mice ( Fig. 4D).

Figure 4.
Figure 4.

MMP-9 contributes to the early establishment of tumors in the lung. A, in vivo bioluminescent images of representative control and MMP-9 null mice injected with 2 × 106 luciferase expressing A549 cells (LUC-A549) i.v. into the tail vein. B, points, mean tumor growth rates in control (n = 10) and MMP-9 null (n = 10) mice at the time points indicated in (A) determined by bioluminescence (photons/s) from the LUC-A549 cells; bars, SD. *, P < 0.05. C, number of surface lung tumors in control (n = 6) and MMP-9 null (n = 10) mice 5 weeks after injection of 2 × 106 LUC-A549 cells i.v. *, P = 0.0007. D, surface lung tumor size (mm) in control and MMP-9 null mice (P = not significant).

To further examine the early stages of the experimental metastasis assay, Rag2 null and Rag2/MMP-9 null mice were sacrificed 24 hours after injection of A549 cells i.v. into the tail vein, and the lungs were examined at the histologic level. Clusters of tumor cells could be seen in the MMP-9 null mice and the wild-type mice ( Fig. 5A and B ). Quantitation of the tumor cells in the Rag2 null and Rag2/MMP-9 null lungs shows a 72% reduction in the number of tumor cells in the Rag2/MMP-9 null mice at the 24-hour time point compared with the control mice (P = 0.02; Fig. 5C). This confirms the bioluminescence data, where decreased luciferase activity was seen in the lungs of MMP-9 null mice 19 hours after injection of tumor cells compared with control mice. Thus, the data obtained by bioluminescence measurements and gross and histologic examination support the conclusion that the lack of MMP-9 affects the initial steps of tumor formation in the lung and not subsequent tumor growth.

Figure 5.
Figure 5.

MMP-9 null mice show fewer tumor cells in the lung 24 hours after tumor cell inoculation compared with control mice and both groups show neutrophil infiltration. Histologic (H&E) analysis of lungs in control (A) and MMP-9 null (B) mice 24 hours after i.v. injection of A549 cells into the tail vein. Clusters of tumor cells (arrowheads). C, quantitation of tumor cell clusters in the lungs of control (n = 7) and MMP-9 null mice (n = 6) 24 hours after tumor cell inoculation. *, P = 0.018. Immunohistochemical analysis of neutrophils 24 hours after tumor cell inoculation in control (D) and MMP-9 null (E) mice. Arrows, neutrophils in the lungs. F, neutrophils in the lungs of noninjected Rag2 null mice.

Cellular source of MMP-9. MMP-9 in the lung tumor microenvironment is contributed by bone marrow–derived cells ( Fig. 3). The absence of lymphocytes in the Rag2 null mice, which are deficient in mature B and T cells, did not seem to influence the effect of MMP-9 ablation on lung tumor formation because we observed similar results in the experimental metastasis assay done with immunocompetent and immunodeficient mice ( Fig. 1E and Fig. 4C). To examine other bone marrow–derived cells that may be contributing MMP-9, we treated mice with liposomal clodronate to eliminate alveolar macrophages from the lung. Liposomal clodronate reduced alveolar macrophages by an average of 32%, but we saw no reduction in tumor number in clodronate treated mice compared with control mice treated with liposomal PBS (data not shown). However, we did observe an abundant influx of neutrophils in tumor-bearing lungs, which have been shown to be a predominant source of MMP-9 in inflamed lung tissue ( 22). Immunostaining for a neutrophil marker revealed a marked elevation in neutrophils in tumor-bearing wild-type and MMP-9 null mice ( Fig. 5D and E) compared with noninjected Rag2 null mice, where few neutrophils were observed in the lung ( Fig. 5F). The neutrophils were clustered adjacent to tumor cells. There was no difference in the number of neutrophils in wild-type versus MMP-9 null mice 6 hours after tumor inoculation (28.2 ± 7.5 versus 28.5 ± 3.8 neutrophils/×10 field, respectively), showing that the lack of MMP-9 did not alter neutrophil influx at this early time point. These data suggest that rapid influx of bone marrow–derived neutrophils after tumor cell inoculation contributes the majority of host-derived MMP-9 in our model.

Tumor cell survival in the lung 6 hours after inoculation is dependent on MMP-9. To determine if the absence of MMP-9 leads to reduced cell survival in the lung at the early stages of tumor cell colonization in the experimental metastasis assay, apoptosis was measured by TUNEL, 6 and 20 hours after i.v. injection of CellTracker-labeled A549 cells. Using fluorescence, the percentage of tumor cells that were undergoing apoptosis could be measured in sections from control and MMP-9 null mice ( Fig. 6A-F ). Five random fields from each mouse lung were examined to determine the percentage of tumor cells undergoing apoptosis. When apoptosis was examined at 6 hours, 8% of tumor cells in MMP-9 null mice were undergoing apoptosis compared with only 2% of tumor cells in control mice (P = 0.01; Fig. 6G). However, at 20 hours, more tumor cells in the control mice (8%) were undergoing apoptosis compared with MMP-9 null mice (1%; P = 0.04; Fig. 6H). These data confirm that very early after injection of tumor cells (6 hours), the cells in the MMP-9 null mice are undergoing more apoptosis, resulting in fewer tumors in the MMP-9 null mice at later time points compared with control mice.

Figure 6.
Figure 6.

MMP-9 null mice show significantly more tumor cells undergoing apoptosis 6 hours after injection of tumor cells. A-G, TUNEL analysis at 6 hours after injection of CellTracker-labeled A549 cells in control (A-C) and MMP-9 null (D-F) mice. Apoptotic cells (green, A and D). CellTracker-labeled tumor cells (red, B and E). C and F, localization of apoptotic tumor cells (yellow) by a merge of TUNEL staining (green) and CellTracker staining (red). Nuclei (blue, Hoechst). G, % tumor cells in control (n = 5) and MMP-9 null (n = 6) mice undergoing apoptosis 6 hours after injection of A549 cells i.v. *, P = 0.01. H, percentage of tumor cells in control (n = 5) and MMP-9 null (n = 4) mice undergoing apoptosis 20 hours after injection of A549 cells i.v. *, P = 0.04.

Discussion

The failure of MMPI clinical trials prompted us to analyze the discrepancy between the clinical results and the numerous preclinical reports showing an advantage of MMP inhibition. We reproduced the results of Itoh et al. ( 17), showing that the lack of host MMP-9 reduced LLC lung nodules, and showed that this result was consistent across systems with an 81% inhibition with LLC cells in syngeneic mice and a 97% reduction in A549 cells in immunocompromised mice. However, of the three MMPs examined, MMP-9 was the only one to show this effect, and MMP-7 showed an opposing protective role in the LCC experimental metastasis assay. Recently, protective roles for MMP-3 and MMP-8 in carcinogenesis models have been shown ( 23, 24), and MMPIs have been shown to have negative effects in some specific animal models ( 25, 26) and clinical trials ( 27). These results underscore the complexity of MMP biology and the importance of understanding the specificity of a targeted cancer therapy.

In addition to showing selectivity in the effect of MMP-9 on lung colonization, we showed that MMP-9 contributes to lung metastasis specifically at the earliest stages of colonization in the lung and has no effect on subsequent tumor growth. Bioluminescence measurements indicated that the LUC-A549 cells reached the lungs of wild-type and MMP-9 null mice in equal proportions, but within 19 hours, there was an 88% reduction in photons/s in the MMP-9 null mice compared with controls. Histologic analysis confirmed the reduction in the number of A549 tumor cells in the lungs of MMP-9 null mice 24 hours after injection compared with wild-type mice (72%). Despite the difference in tumor number, there was no difference in tumor size between control and MMP-9 null mice in either of the experimental metastasis models. The LUC-A549 cells clearly showed that the increase in bioluminescence with time, an indication of the rate of tumor growth, was virtually identical between 48 hours and 5 weeks in wild-type and MMP-9-null mice. These results are consistent with results reported from an experimental metastasis model of i.v. injected T-cell lymphoma cells and liver metastasis, where administration of an inhibitor specific to the gelatinases reduced liver metastasis only when the gelatinase inhibitor was given 1 hour before tumor cell inoculation with no effect when it was given 1 day after injection of tumor cells ( 28). We conclude that host MMP-9 contributes to the earliest stages of establishment of tumors in the lung and has no effect on subsequent tumor growth.

It is intriguing that host MMP-9 influences tumor establishment in tumor cells that express MMP-9 (A549 cells) as well as those that do not (LLC cells). Explanation for this include the possibility that A549 cells express MMP-9 that is not activated, or that they do not express CD44, a known receptor for MMP-9 ( 29). However, A549 cells have been reported to express CD44 ( 30), and we have shown in preliminary studies that A549 cells produce active MMP-9 as measured using a proteolytic beacon ( 31) designed to be selective for MMP-9 (data not shown). We conclude that tumor-produced MMP-9 does not compensate for the lack of host-derived MMP-9 in this model system.

The lack of an effect of MMP-9 on tumor growth differs with previous reports of both positive and negative effects of MMP-9 on tumor angiogenesis. MMP-9 can promote tumor angiogenesis and subsequent tumor growth in several systems ( 7, 20), most likely through a mechanism of releasing vascular endothelial growth factor from the matrix allowing increased accessibility to its receptor ( 7). However, MMP-9 can also inhibit angiogenesis by generating endogenous angiogenesis inhibitors (angiostatin and tumstatin), and consequently, MMP-9 has been shown to be associated with reduced tumor growth and angiogenesis ( 32, 33). In our studies, the tumors formed with either the LLC or A549 cells were <2 mm in diameter, a size at which angiogenesis is not required ( 34). We found no difference in angiogenesis as measured by von Willebrand factor staining in the tumors of control and MMP-9 null mice ( Fig. 1E; data not shown). Therefore, the absence of a difference in tumor size between wild-type and MMP-9 null mice in our model may be due to the tumors not growing large enough to undergo an angiogenic switch.

Bone marrow transplantation studies revealed that MMP-9 from bone marrow–derived cells contributes to the difference in tumor number that was observed in the experimental metastasis assay. Inflammatory cells derived from the bone marrow are a source of MMP-9 in human tumors, including lung tumors ( 18, 19, 35, 36), and we observed MMP-9 immunoreactivity in stromal cells in our model. We speculated that T cells may be required for the MMP-9 effect. T lymphocytes release MMP-9 ( 37, 38), and T-lymphocyte membranes in contact with interstitial macrophages, but not alveolar macrophages, stimulated the release of MMP-9 from the macrophages ( 39). However, we saw that the absence of MMP-9 has a similar effect on tumor number in the lung in wild-type and Rag2 null mice, which are deficient in mature B and T lymphocytes. Ablating alveolar macrophages by administration of liposomal chlodronate also did not affect tumor number in the lung, suggesting that macrophages may not be the predominant source of MMP-9 in the experimental metastasis assay. We note that the administration of liposomal chlodronate reduced the number of alveolar macrophages by 32% rather than causing complete ablation, but the zymography data of MMP-9 expression in the bone marrow cell lysates indicate that a small reduction in MMP-9 from the bone marrow is sufficient to cause a concomitant reduction in tumor number. It is likely that MMP-9 from neutrophils is contributing to the phenotype in the experimental metastasis assay. We observed an influx of neutrophils in the lung within 6 hours and at 24 hours after injection of tumor cells, and neutrophils are a predominant source of MMP-9 ( 22). In addition, there is evidence that neutrophils contribute to mammary adenocarcinoma metastasis ( 40, 41). The lack of MMP-9 did not alter neutrophil influx, because we observed equivalent numbers of neutrophils in both MMP-9-null and wild-type mice after tumor cell inoculation. We speculate that as metastatic tumor cells enter the lung, neutrophils are recruited and perform protumorigenic functions by releasing MMP-9 and allowing tumor cells to establish and initiate growth.

In exploring cellular mechanisms by which MMP-9 affects lung tumor establishment, we observed significantly more apoptosis in the tumor cells in MMP-9 null lungs (8%) compared with control lungs (2%), 6 hours after injection, providing an explanation for the decrease in tumor cells in MMP-9 null mice at all subsequent times. It has been shown that blocking MMP-9 in vitro using an anti-MMP-9 monoclonal antibody induces apoptosis in B-cell chronic lymphocytic leukemia cells but only when these cells are in contact with bone marrow stromal cells ( 42). This data is interesting in light of our finding that MMP-9 from the bone marrow is contributing to tumor cell survival in the lung. Interestingly, at 20 hours after injection, the effect was reversed, and significantly more tumor cells in the control mice were undergoing apoptosis compared with MMP-9 null mice. The reason for this difference is unclear but could be explained if one considers that apoptosis of cells when they reach an ectopic environment is probably a normal event, but the timing is accelerated in the absence of MMP-9. A proportion of the tumor cells are capable of surviving long enough to reach the next stage of initiating proliferation, but the number of surviving cells capable of initiating proliferation is reduced in MMP-9 null mice as a result of this premature apoptosis. Sustained growth of the colonies that manage to initiate proliferation is unaffected by MMP-9 because we observed similar growth rates and tumor sizes in control and MMP-9 null mice.

How MMP-9 contributes to survival or the absence of MMP-9 contributes to cell death remains unknown. MMP-9 may contribute to the apoptotic process by modifying cell surface molecules or processing matrix to release soluble factors. MMP-9 has been shown to process latent transforming growth factor β (TGFβ) to its active form at the cell surface by complexing with CD44 ( 43). When this complex of MMP-9 and CD44 is disrupted, tumor cells injected into the lung undergo apoptosis. It is therefore possible that the absence of MMP-9 is inducing apoptosis by allowing less active TGFβ to be present to induce tumor cell survival. In addition, apoptosis of tumor cells in MMP-9 null mice may be related to tumor cell attachment to the lung endothelium. It has been shown that rat lung endothelium has exposed areas of basement membrane, where laminin-5 in these areas is important in binding to integrin α3β1 on tumor cells to facilitate tumor cell arrest in the lung vasculature ( 44). Therefore, an absence of MMP-9 in the lung may prevent tumor cells from attaching to the lung vasculature because MMP-9 is not available to expose basement membrane on the lung endothelium to facilitate adhesion.

Additional evidence for MMP-9 contributing to tumor establishment is observed from the results in the orthotopic model. Fifty percent fewer MMP-9 null mice had primary tumors in the lung compared with control mice after orthotopic injection of A549 cells. However, of the tumors that did develop in the lung, tumors in MMP-9 null mice were able to convert to macroscopic size in numbers equivalent to control mice. MMP-9 affected early tumor take in the lung and did not contribute to subsequent growth of tumors in the lung. Therefore, in the orthotopic model, as was observed in the experimental metastasis assay, MMP-9 contributed to tumor cell survival, and the absence of MMP-9 caused fewer primary tumors to form in the lung.

MMPs have multiple roles in different stages of tumor progression, and we have shown that host MMP-9 can contribute to the early stages of metastasis in the lung as well as establishment of transplanted primary lung tumors. Treatment of patients with late-stage lung cancer with several synthetic MMPIs, all of which targeted MMP-9, was not successful ( 10). Although our data indicate a contribution of MMP-9 to metastasis, it is only for a narrow window of time following the arrival of the tumor cells in the lung microenvironment, after which point, ablation of MMP-9 has no effect on subsequent tumor growth. Based on this information, MMP-9 inhibition with MMPIs would be expected to have little or no effect on patients in which metastasis had already occurred, such as those enrolled in the MMPI clinical trials. Effective treatment based on MMP-9 inhibition would require identifying patients before metastatic seeding occurs and maintaining constant levels of MMP inhibition. These results show the importance of understanding the specific stage(s) of tumor progression recapitulated in animal models and the necessity of a thorough understanding of mechanism of action of targeted therapies.

Acknowledgments

Grant support: NIH/National Cancer Institute grant P50 CA90949 and predoctoral trainee grant T32 CA009385 (H.B. Acuff).

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.

We thank Drs. David Johnson and David Carbone for their continual support and encouragement and Oliver McIntyre for his help with the in vivo imaging studies.

Footnotes

    • Received July 18, 2005.
    • Revision received October 20, 2005.
    • Accepted October 24, 2005.

    References

    1. ↵
      Fidler IJ. The organ microenvironment and cancer metastasis. Differentiation 2002; 70: 498–505.
      OpenUrlCrossRefPubMed
    2. ↵
      Egeblad M, Werb Z. New functions for the matrix metalloproteinases in cancer progression. Nat Rev Cancer 2002; 2: 161–74.
      OpenUrlCrossRefPubMed
    3. ↵
      Stamenkovic I. Matrix metalloproteinases in tumor invasion and metastasis. Semin Cancer Biol 2000; 10: 415–33.
      OpenUrlCrossRefPubMed
    4. ↵
      Lynch CC, Matrisian LM. Matrix metalloproteinases in tumor-host cell communication. Differentiation 2002; 70: 561–73.
      OpenUrlCrossRefPubMed
    5. ↵
      DeClerk YA. Interactions between tumour cells and stromal cells and proteolytic modification of the extracellular matrix by metalloproteases in cancer. Eur J Cancer 2000; 36: 1258–68.
      OpenUrlCrossRefPubMed
    6. ↵
      Coussens LM, Tinkle CL, Hanahan D, Werb Z. MMP9 supplied by bone marrow-derived cells contributes to skin carcinogenesis. Cell 2000; 103: 481–90.
      OpenUrlCrossRefPubMed
    7. ↵
      Bergers G, Brekken R, McMahon G, et al. Matrix metalloproteinase-9 triggers the angiogenic switch during carcinogenesis. Nat Cell Bio 2000; 2: 737–44.
      OpenUrlCrossRefPubMed
    8. ↵
      Masson R, Lefebvre O, Noël A, et al. In vivo evidence that stromelysin-3 metalloproteinase contributes in a paracrine manner to epithelial cell malignancy. J Cell Biol 1998; 140: 1535–41.
      OpenUrlAbstract/FREE Full Text
    9. ↵
      Huang S, Van Arsdall M, Tedjarati S, et al. Contributions of stromal metalloproteinase-9 to angiogenesis and growth of human ovarian carcinoma in mice. J Natl Cancer Inst 2002; 94: 1134–42.
      OpenUrlAbstract/FREE Full Text
    10. ↵
      Coussens LM, Fingleton B, Matrisian LM. Matrix metalloproteinase inhibitors and cancer: trials and tribulations. Science 2002; 295: 2387–92.
      OpenUrlAbstract/FREE Full Text
    11. ↵
      Vu TH, Shipley JM, Bergers G, et al. MMP9/gelatinase B is a key regulator of growth plate angiogenesis and apoptosis of hypertrophic chrondrocytes. Cell 1998; 93: 411–22.
      OpenUrlCrossRefPubMed
    12. ↵
      Wilson CL, Heppner KJ, Labosky PA, Hogan BLM, Matrisian LM. Intestinal tumorigenesis is suppressed in mice lacking the metalloproteinase matrilysin. Proc Natl Acad Sci U S A 1997; 94: 1402–7.
      OpenUrlAbstract/FREE Full Text
    13. ↵
      Itoh T, Ikeda T, Gomi H, Nakao S, Suzuki T, Itohara S. Unaltered secretion of β-amyloid precursor protein in gelatinase A (matrix metalloproteinase 2)-deficient mice. J Biol Chem 1997; 272: 22389–92.
      OpenUrlAbstract/FREE Full Text
    14. ↵
      McLemore T, Liu M, Blacker P, et al. Novel intrapulmonary model for orthotopic propagation of human lung cancers in athymic nude mice. Cancer Res 1987; 47: 5132–40.
      OpenUrlAbstract/FREE Full Text
    15. ↵
      Koay MA, Gao X, Washington MK, et al. Macrophages are necessary for maximal nuclear factor-κB activation in response to endotoxin. Am J Respir Cell Mol Biol 2002; 26: 572–8.
      OpenUrlCrossRefPubMed
    16. ↵
      Itoh T, Tanioka M, Yoshida H, Yoshioka T, Nishimoto H, Itohara S. Reduced angiogenesis and tumor progression in gelatinase A-deficient mice. Cancer Res 1998; 58: 1048–51.
      OpenUrlAbstract/FREE Full Text
    17. ↵
      Itoh T, Tanioka M, Matsuda H, et al. Experimental metastasis is suppressed in MMP9 deficient mice. Clin Exp Metastasis 1999; 17: 177–81.
      OpenUrlCrossRefPubMed
    18. ↵
      Cox G, Jones JL, O'Byrne KJ. Matrix metalloproteinase 9 and epidermal growth factor signal in operable non-small cell lung cancer. Clin Cancer Res 2000; 6: 2349–55.
      OpenUrlAbstract/FREE Full Text
    19. ↵
      Cox G, Jones JL, Andi A, Walker DA, O'Byrne KJ. A biological staging model for operable non-small cell lung cancer. Thorax 2001; 56: 561–6.
      OpenUrlAbstract/FREE Full Text
    20. ↵
      Yang L, Debusk LM, Fukada K, et al. Expansion of myeloid immune suppressor Gr+CD11b+ cells in tumor-bearing host directly promotes tumor angiogenesis. Cancer Cell 2004; 6: 409–21.
      OpenUrlCrossRefPubMed
    21. ↵
      Giraudo E, Inoue M, Hanahan D. An amino-bisphosphonate targets MMP9 expressing macrophages and angiogenesis to impair cervical carcinogenesis. J Clin Invest 2004; 114: 623–33.
      OpenUrlCrossRefPubMed
    22. ↵
      Atkinson JJ, Senior RM. Translational Review: matrix metalloproteinase-9 in lung remodeling. Am J Respir Cell Mol Biol 2003; 28: 12–24.
      OpenUrlCrossRefPubMed
    23. ↵
      McCawley LJ, Crawford HC, King LE, Mudgett J, Matrisian LM. A protective role for matrix metalloproteinase-3 in squamous cell carcinoma. Cancer Res 2004; 64: 6965–72.
      OpenUrlAbstract/FREE Full Text
    24. ↵
      Balbin M, Fueyo A, Tester AM, et al. Loss of collagenase-2 confers increased skin tumor susceptibility to male mice. Nat Genet 2003; 35: 252–7.
      OpenUrlCrossRefPubMed
    25. ↵
      Krüger A, Soeltl R, Sopov I, et al. Hydroxamate-type matrix metalloproteinase inhibitor batimastat promotes liver metastasis. Cancer Res 2001; 61: 1272–5.
      OpenUrlAbstract/FREE Full Text
    26. ↵
      Chen X, Su Y, Fingleton B, et al. Increased plasma MMP9 in integrin α1-null mice enhances lung metastasis of colon carcinoma cells. Int J Cancer 2005; 116: 52–61.
      OpenUrlCrossRefPubMed
    27. ↵
      Moore MJ, Hamm J, Eisenberg M, et al. A comparison between gemcitabine (GEM) and the matrix metalloproteinase (MMP) inhibitor BAY12-9566 in patients with advanced pancreatic cancer. Proc Am Soc Clin Oncol 2000; 19: 240a.
      OpenUrl
    28. ↵
      Krüger A, Arlt M, Gerg M, et al. Antimetastatic activity of a novel mechanism-based gelatinase inhibitor. Cancer Res 2005; 65: 3523–6.
      OpenUrlAbstract/FREE Full Text
    29. ↵
      Yu Q, Stamenkovix I. Localization of matrix metalloproteinase 9 to the cell surface provides a mechanism for CD44-mediated tumor invasion. Genes Dev 1999; 13: 35–48.
      OpenUrlAbstract/FREE Full Text
    30. ↵
      Koslowski R, Fichtner F, Barth K, Roehlecke C, Seidel D, Kasper M. Apoptosis and release of CD44s in bleomycin-treated L132 cells. J Cell Biochem 2005; 95: 1146–56.
      OpenUrlPubMed
    31. ↵
      McIntyre JO, Fingleton B, Wells KS. Development of a novel fluorogenic proteolytic beacon for in vivo detection and imaging of tumour-associated matrix metalloproteinase-7 activity. Biochem J 2004; 377: 617–28.
      OpenUrlPubMed
    32. ↵
      Pozzi A, Lefine WF, Gardner HA. Low plasma levels of matrix metalloproteinase 9 permit increased tumor angiogenesis. Oncogene 2002; 21: 272–81.
      OpenUrlCrossRefPubMed
    33. ↵
      Hamano Y, Zeisberg M, Sugimoto H, et al. Physiological levels of tumstatin, a fragment of collagen IV α3 chain, are generated by MMP9 proteolysis and suppress angiogenesis via αVβ3 integrin. Cancer Cell 2003; 3: 589–601.
      OpenUrlCrossRefPubMed
    34. ↵
      Folkman J. The role of angiogenesis in tumor growth. Semin Cancer Biol 1992; 3: 65–71.
      OpenUrlPubMed
    35. ↵
      Jones JL, Glynn P, Walker RA. Expression of MMP-2 and MMP-9, their inhibitors, and the activator MT1-MMP in primary breast carcinomas. J Pathology 1999; 189: 161–8.
      OpenUrlCrossRefPubMed
    36. ↵
      Saito K, Takeha S, Shiiba K, et al. Clinicopathologic significance of urokinase receptor- and MMP-9-positive stromal cells in human colorectal cancer: functional multiplicity of matrix degradation on hematogenous metastasis. Int J Cancer 2000; 86: 24–9.
      OpenUrlPubMed
    37. ↵
      Weeks BS, Schnaper W, Handy M, Holloway E, Kleinman HK. Human T lymphocytes synthesize the 92 kDa type IV collagenase (gelatinase B). J Cell Physiol 1993; 157: 644–9.
      OpenUrlCrossRefPubMed
    38. ↵
      Leppert D, Waubant E, Galardy R, Bunnett NW, Hauser SL. T cell gelatinases mediate basement membrane transmigration in vitro. J Immunol 1995; 154: 4379–89.
      OpenUrlAbstract
    39. ↵
      Ferrari-Lacraz S, Nicod LP, Chicheportiche R, Welgus HG, Dayer J. Human lung tissue macrophages, but not alveolar macrophages, express matrix metalloproteinase after direct contact with activated T lymphocytes. Am J Respir Cell Mol Biol 2001; 24: 442–51.
      OpenUrlCrossRefPubMed
    40. ↵
      Aeed PA, Nakajima M, Welch DR. The role of polymorphonuclear leukocytes (PMN) on the growth and metastatic potential of 13762NF mammary adenocarcinoma cells. Int J Cancer 1988; 42: 748–59.
      OpenUrlCrossRefPubMed
    41. ↵
      Welch DR, Schissel DJ, Howery RP, Aeed PA. Tumor-elicited polymorphonuclear cells, in contrast to “normal” circulating polymorphonuclear cells, stimulate invasive and metastatic potentials of rat mammary adenocarcinoma cells. Proc Natl Acad Sci U S A 1989; 86: 5859–63.
      OpenUrlAbstract/FREE Full Text
    42. ↵
      Ringshausen I, Dechow T, Schneller F, et al. Constitutive activation of the MAPkinase p38 is critical for MMP-9 production and survival of B-CLL cell on bone marrow cells. Leukemia 2004; 18: 1964–70.
      OpenUrlCrossRefPubMed
    43. ↵
      Yu Q, Stamenkovic I. Transforming growth factor-β facilitates breast carcinoma metastasis by promoting tumor cell survival. Clin Exp Metastasis 2004; 21: 235–42.
      OpenUrlCrossRefPubMed
    44. ↵
      Wang H, Fu W, Im J, et al. Tumor cell α3β1 integrin and vascular laminin-5 mediate pulmonary arrest and metastasis. J Cell Biol 2004; 164: 935–41.
      OpenUrlAbstract/FREE Full Text
    Back to top
    Cancer Research: 66 (1)
    January 2006
    Volume 66, Issue 1

    Sign up for alerts

    View this article with LENS

    Open full page PDF
    Article Alerts
    Email Article
    Citation Tools
    Matrix Metalloproteinase-9 from Bone Marrow–Derived Cells Contributes to Survival but not Growth of Tumor Cells in the Lung Microenvironment
    Heath B. Acuff, Kathy J. Carter, Barbara Fingleton, D. Lee Gorden and Lynn M. Matrisian
    Cancer Res January 1 2006 (66) (1) 259-266; DOI: 10.1158/0008-5472.CAN-05-2502
    del.icio.us logo Digg logo Reddit logo Twitter logo CiteULike logo Facebook logo Google logo Mendeley logo

    Jump to section

    Advertisement