Cellular Membrane Type-1 Matrix Metalloproteinase (MT1-MMP) Cleaves C3b, an Essential Component of the Complement System*

Neoplasms have developed numerous strategies to protect themselves against the host immune system. Membrane type-1 matrix metalloproteinase (MT1-MMP) is strongly associated with many cancer types and is up-regulated in the aggressive, metastatic neoplasms. During the past few years, there has been an increasing appreciation of the important, albeit incompletely understood, role of MT1-MMP in cancer. We have discovered, using cell-free and cell-based assays in vitro, that MT1-MMP proteolysis specifically targets C3b, an essential component of the complement propagation pathway. MT1-MMP proteolysis liberates the deposited C3 activation fragments from the cell surface. The shedding of these cell-deposited opsonins by MT1-MMP inhibits the complement cascade and protects breast carcinoma MCF7 cells from direct complement-mediated injury in the in vitro tests. The functional link associating MT1-MMP with the host immune system, heretofore unrecognized, may empower tumors with an escape mechanism that contributes to the protection against the host anti-tumor immunity as well as to the survival of invading and metastatic malignant cells in the bloodstream.

It is well established that the progression of metastatic cancer involves the interplay of the host environment with the malignant cells (1). Neoplasms employ multiple means to sustain themselves and to proliferate in vivo. Evidence has emerged that tumor immunity is an important defense mechanism protecting malignant cells from the host immune surveillance. The host immune system is an apparatus directed against foreign invading organisms and tumor cells (2)(3)(4). Controlled activation of the complement system is a critical component of host immunity. Complement activation products stimulate a localized protective inflammation and are involved in both the inductive and effector phases of an immune response (5). In many cases malignant cells exhibit antigens that are not typically associated with normal cells. These antigens can be identified by the complement system via antibody recognition and, as a result, the recognized cells are attacked by the immune system (6).
The complement system is comprised of soluble proteins that interact in a stepwise manner. Complement can be activated via three different pathways: the classical pathway that is usually antibody-dependent, and the alternative and lectin pathways. In the classical pathway, immunoglobulin-coated targets bind and subsequently activate the complement component C1. This event starts the complement cascade, the propagation of which results in the generation of anaphylatoxins (C3a, C4a, and C5a) and ultimately the cytolytic C5b-9 membrane attack complex (MAC) 1 (7).
In the process of complement propagation, proteolytic cleavage of serum C3 and C4 creates transient soluble C3b and C4b products with an exposed reactive thioester group. Once exposed, the thioester group forms amide and ester bonds with the target cell surface molecules. This binding of C3b and C4b is critical for amplification of the cascade and for MAC formation. Subsequent proteolytic cleavage transforms the cell-bound C3b into the cleavage fragments iC3b, C3dg, and C3d, which remain covalently attached to the cell surface. These C3 fragments serve as ligands for receptors on phagocytic and NK cells. Opsonization of target cells with these C3 fragments promotes and enhances both antibodydependent and complement-dependent cell cytotoxicity, two additional effector systems that play an important role in the elimination of neoplastic cells (8,9). The deposition of C3b and C4b and the follow-up amplification of the complement cascade also results in the generation of soluble bioactive C3a and C5a peptides that may potentiate anti-tumor responses via their chemoattractant and proinflammatory activities. Thus, the inactivation of C3b and C4b and their removal from a cell surface represents an important immune evasion mechanism of tumor cells.
An important role for complement resistance of tumor cells is indicated by the fact that many tumor cells overexpress one or more of the cell surface-associated complement regulatory proteins: CD46/MCP, CD55/DAF, and CD59/protectin. These regulatory proteins act at different stages of the complement propagation (10 -13). In addition to the cell-associated regulatory proteins, some tumor cells bind serum complement inhibitory proteins (14,15), express proteases to clear the C3b protein from the cell surface, and inhibit the complement cascade (16 -18).
Membrane type-1 matrix metalloproteinase (MT1-MMP) is the most common protease from the membrane-tethered enzyme subfamily of MMPs (19). MT1-MMP plays an important, albeit insufficiently characterized, role in tissue remodeling and cell motility, and especially in tumor progression, metas-tasis, and angiogenesis (20,21). MT1-MMP is a cell surface activator of soluble pro-MMP-2 and pro-MMP-13 and has also been implicated in matrix proteolysis and turnover as well as in the proteolytic processing of cell surface-associated adhesion and signaling receptors (19,22,23). Although MT1-MMP is detectable in normal tissue, the expression of this protease is strongly associated with aggressive, invasive malignant cells (19,24).
Here, we report a novel, unexpected and highly significant function of MT1-MMP in the proteolysis of the opsonic complement components C3b and C4b. This proteolysis efficiently inhibits complement activation in cell-based models. We suspect that MT1-MMP is likely to make malignant cells more resistant to complement-mediated cytotoxicity in vivo.
Our data indicate that this novel function of MT1-MMP is involved in the release of C3b from the tumor cell surface, and is likely to contribute to the survival and propagation of malignant cells. We believe that the proteolysis of the complement components by MT1-MMP is a powerful and efficient mechanism employed by aggressive malignant cells to protect themselves against host complement, immune surveillance, and destruction.

MATERIALS AND METHODS
Antibodies and Reagents-All reagents were purchased from Sigma unless otherwise indicated. Purified human C3b, iC3b, and C4b, goat anti-human C3 and C4 antibodies, normal human serum, and C5depleted human serum were from Advanced Research Technologies. Monoclonal murine antibody H206 against human C3b ␣ chain was purchased from Research Diagnostics. Fluorescein isothiocyanate-conjugated goat anti-mouse IgG and HRP-conjugated F(abЈ) 2 fragment goat anti-mouse antibodies were obtained from Jackson ImmunoResearch Laboratories. Rabbit antibody AB815 against the hinge domain of MT1-MMP, HRP-conjugated rabbit anti-goat antibodies, the TMB/M and TMB/E substrates, and GM6001 (a broad-range hydroxamate inhibitor of MMPs) were from Chemicon. Murine anti-human CD59 monoclonal antibody BRIC229 and murine anti-human CD55 monoclonal antibody 1A10 are as described earlier (25,26). Murine antihuman CD46 monoclonal antibody M75 was obtained from BD Pharmingen. The recombinant catalytic domain of human MT1-MMP (MT1-CAT) was expressed in Escherichia coli, purified from the inclusion bodies, and refolded to restore its native conformation as previously described (27). Rabbit antiserum against the breast carcinoma MCF7 cell membranes was prepared by routine techniques and, where indicated, was used to sensitize MCF7 cells to human complement. Rabbit antibody against the recombinant catalytic domain of MT1-MMP was generated in our laboratory.
Cell Lines-Human breast carcinoma MCF7 cells (ATCC) stably transfected with the empty pcDNA3-zeo vector (Invitrogen) (control MCF-zeo cells) and the full-length wild type MT1-MMP (MCF-MT cells) were constructed and extensively characterized in our prior work (28). Human fibrosarcoma HT1080 cells, which synthesize MT1-MMP and MMP-2 naturally, were obtained from ATCC. Transfected cells were routinely grown in Dulbecco's modified Eagle's medium (DMEM) (Irvine Scientific) supplemented with 10% fetal bovine serum (FBS) (Tissue Culture Biologicals) and 0.2 mg/ml zeocin. Human breast carcinoma BT549, MCF7, MDA-MB-231, T47D, and SK-Br-3 cells (ATCC) were propagated in DMEM supplemented with 5% FBS and penicillin-streptomycin (100 IU/ml and 100 g/ml). Immunoprecipitation and Western Blotting-HT1080, MCF-zeo, and MCF-MT cells were grown in DMEM/FCS. Where indicated, PMA (5 ng/ml) was added to the cells. After incubation for 12 h, cells were washed with PBS and surface biotinylated with sulfo-NHS-LC-biotin (Pierce) according to the manufacturer's instructions. Next, cells were washed with ice-cold PBS and lysed with 50 mM N-octyl-␤-D-glucopy-ranoside (Amresco) in PBS supplemented with 1 mM CaCl 2 , 1 mM MgCl 2 , and protease inhibitor mixture containing 1 mM phenylmethylsulfonyl fluoride and 1 g/ml each of aprotinin, pepstatin, and leupeptin. The lysates were pre-cleared with Protein G-agarose beads (Calbiochem). The samples of cell lysates each containing 1.0 mg of protein were mixed with an MT1-MMP antibody (1 g) and Protein G-agarose, and incubated at 4°C overnight. After extensive washings, immune complexes were released by boiling the beads for 5 min in 2ϫ SDS sample buffer containing 50 mM dithiothreitol. Solubilized proteins were subjected to SDS-PAGE and Western blotting (28). For flow cytometry analyses the cells were detached by an enzymefree buffer (Specialty Media), washed, and stained for 1 h at 4°C with the murine monoclonal antibody H206 against human C3b ␣ chain (5 g/ml) followed by incubation for 30 min with an fluorescein isothiocyanate-conjugated goat anti-mouse antibody. All incubation steps were performed in PBS supplemented with 1% BSA and 0.01% NaN 3 . After removal of unbound antibodies, cells were re-suspended in PBS supplemented with 3 g/ml propidium iodide (Molecular Probes), 1% BSA, and 0.01% NaN 3 . Viable cells were analyzed on a FACScan flow cytometer (BD Biosciences). Where indicated, GM6001 (50 M) was co-incubated with the cells for 24 h.
C3b liberated from the cell surface by proteolytic shedding and released into the extracellular milieu was identified by ELISA of the medium samples. For these purposes, MCF-zeo and -MT cells (10 5 cells each) were incubated for 1 h at 37°C in DMEM supplemented with 20% C5-depleted human serum to induce deposition of C3b on cell surfaces. Next, unbound material was removed by washing cells with PBS. A mixture of DMEM/FBS (200 l) was added to the cells. Following incubation for 30 -120 min to release the cell-bound C3b, the aliquots of medium were withdrawn for a subsequent analysis of the liberated C3b. Where indicated, GM6001 (50 M) was co-incubated with the cells to block activity of the cellular MT1-MMP.
ELISA of Soluble C3b-Wells of a 96-well plate (Corning) were coated with goat anti-human C3 antibody (5 g/ml) and then blocked with 1% BSA. Medium aliquots (100 l) were allowed to bind for 1 h at 37°C with the antibody-coated wells. The bound C3b was detected with biotin-labeled goat anti-human C3 (1 g/ml) followed by streptavidin-HRP and the TMB/E substrate. The absorbance of the samples was measured at 450 nm.
Flow Cytometry of the Complement Regulatory Proteins-Flow cytometry was used to assess the cell surface expression of complement regulatory proteins CD46, CD55, and CD59 in MCF-zeo and -MT cells. For these purposes, cells were detached by the enzyme-free buffer (Specialty Media) and co-incubated with the respective primary anti-body (5 g/ml each) followed by incubation with fluorescein isothiocyanate-labeled secondary antibody (1:500 dilution). Population gates were set by using cells incubated with normal murine IgG. Cells were analyzed on a FACStar flow cytometer (BD Biosciences).
Cytotoxicity Assay-MCF-zeo and -MT cells (1 ϫ 10 5 each) were grown in DMEM/FBS in wells of a 48-well plate. To sensitize cells, 20% heat-inactivated rabbit MCF7 antiserum was co-incubated with the cells for 30 min at 37°C. The sensitized cells were then placed for 1 h at 37°C in 20% normal human serum to induce the activation of the complement pathway and lysis of the cells by the resulting membrane attack complex. The treatment with rabbit anti-MCF7 serum was omitted in control samples. The efficiency of the cell lysis was determined by using a Vybrant Cytotoxicity Assay Kit V-23111 (Molecular Probes) in accordance with the manufacturer's instructions.
The cytotoxic effect of the complement was also assessed by microscopy. For this purpose, MCF-zeo and -MT cells (1 ϫ 10 5 cells each) were grown in DMEM/FBS in wells of an 8-well Lab-Tek TM chamber glass slide (Nalge Nunc). Heat-inactivated 20% rabbit anti-MCF7 serum was co-incubated with the cells for 30 min at 37°C. The sensitized cells were further incubated for 1 h at 37°C in 20% normal human serum and then fixed with 4% glutaraldehyde for 1 h at ambient temperature and photographed. Dead cells were made visible with propidium iodide using a LIVE/DEAD Reduced Biohazard Viability/Cytotoxicity L-7013 Kit (Molecular Probes) and the images were taken by a Nikon Eclipse TE300 fluorescence microscope equipped with a SPOT Real-Time SP402-115 digital camera (Diagnostic Instruments).
Purification and Iodination of TIMP-2-Human TIMP-2 was produced by Chinese hamster ovary cells stably transfected with the Cterminal His-tagged human TIMP-2. TIMP-2 was purified from conditioned medium by chelating chromatography and ion-exchange chromatography using the slightly modified protocol described in Ref. 30. TIMP-2 (10 g) was labeled with Na 125 I (Amersham Biosciences) using the IODO-GEN iodination reagent (Pierce) and separated from unincorporated radioactivity by gel-filtration. Normally, the specific radioactivity of labeled TIMP-2 was 4 -5 Ci/g.
Cell Surface Binding of Radiolabeled TIMP-2-To remove cell-bound endogenous TIMP-2, cells (1.5 ϫ 10 5 cells per well of a 12-well plate) were washed prior to assays with 50 mM glycine-HCl buffer, pH 3.0, containing 100 mM NaCl, then neutralized with 0.5 M HEPES buffer, pH 7.5, containing 100 mM NaCl and, finally, equilibrated with DMEM supplemented with 20 mM HEPES, pH 7.5, and 0.2% BSA. Increasing concentrations of 125 I-TIMP-2 (0.03-14 nM) were added to the cells. Following a 3-h incubation at 4°C, cells were washed twice with ice-cold DMEM supplemented with 20 mM HEPES, pH 7.5, and 0.2% BSA, and lysed in 0.5 M NaOH containing 0.1% SDS; the radioactivity was then counted. The nonspecific background binding was determined using a 200-fold excess of unlabeled over highest concentration of 125 I ligand. One well of each cluster was used to count the number of cells. Each experimental point was measured in duplicate. GraphPad Prism software was used to calculate the Scatchard plot and the K d value.
Isolation of RNA and Northern Blotting Analysis-The total RNA was isolated from the cells with a RNAzol B reagent (Tel-Tech). RNA (15 g) was separated by electrophoresis on a 1% agarose/formaldehyde gel and then transferred to a Hybond-N membrane (Amersham Biosciences). The membrane was hybridized with the 32 P-labeled MT1-MMP cDNA and with the 32 P-labeled 28 S rRNA oligonucleotide probe (Clontech). Gelatin zymography was employed to visualize the levels of MMP-2 activation. MCF7 cells were supplemented in these assays with external pro-MMP-2 in amounts that were similar to those naturally synthesized by HT1080 cells. To promote the activation of MMP-2, HT1080 cells were stimulated with PMA (5 ng/ml) (31,32). Our data indicate that the efficiency of MMP-2 activation by the transfected MCF-MT cells is highly similar to that of HT1080 cells (Fig. 1A), suggesting similar levels of cell surface-associated mature MT1-MMP in these two cell types.

Transfected Breast Carcinoma Cells Express Physiologically
To confirm these observations, we directly identified the levels of cell surface-associated MT1-MMP in MCF-MT and HT1080 cells. For these purposes, cells were surface-labeled with membrane-impermeable biotin and then analyzed. Biotinlabeled, cell surface-associated MT1-MMP was immunoprecipitated from cell lysates and detected by Western blotting with streptavidin-HRP (Fig. 1B). These studies demonstrated that cell surface-associated, full-length 60-kDa MT1-MMP was equally represented in HT1080 and MCF-MT cells. In addition, MCF-MT cells exhibit significant amounts of the 43-45-kDa catalytically inactive ectodomain forms of the protease. PMA treatment increased the levels of the degraded MT1-MMP in HT1080 cells. The existence of these full-length and degraded species of MT1-MMP is in agreement with our earlier findings (28) and the results published by others (32). Importantly, these data, which agree well with those of the MMP-2 activation studies, indicate similar levels of the full-length, catalytically potent MT1-MMP in HT1080 and MCF-MT cells.
To extend these comparison studies further, we compared by Northern blotting the levels of the MT1-MMP mRNA in fibrosarcoma HT1080 cells with those in breast carcinoma BT549, MDA-MD-231, and MCF7 cells (Fig. 1C). In agreement with the data of Western blotting that did not detect the expression of the MT1-MMP protein in MCF7-zeo cells (Fig. 1B), Northern blotting did not identify any detectable amounts of the MT1-MMP mRNA in this cell type. In turn, the levels of the MT1-MMP mRNA were relatively low in MDA-MD-231 cells but significantly higher in BT549 cells and comparable with the levels of the messenger identified in HT1080 cells.
To substantiate these observations even further, we identified the number of the TIMP-2-binding MT1-MMP cell surface sites in HT1080 and BT549 cells. For these purposes, we incubated the cells with increasing concentrations of radioactively labeled TIMP-2. An excess of unlabeled TIMP-2 fully blocked the binding of the labeled inhibitor with the cells. TIMP-1 (a poor inhibitor of MT1-MMP) was incapable of interfering with the labeled TIMP-2 ligand (data not shown). Following incubation with labeled TIMP-2 and washing to remove the unbound radioactivity, cells were lysed and the radioactivity was counted in cell lysates. Scatchard plot analysis of binding data demonstrated the existence of high affinity binding sites with K D ϭ 3.9 nM and 235,000 sites per HT1080 cell (Fig. 1D). Similarly, BT549 cells exhibited 136,000 MT1-MMP sites/cell and K D ϭ 1.34 nM. These data correlate well with the concentrations of the MT1-MMP mRNA identified by Northern blotting in HT1080 and BT549 cells (Fig. 1C). The data from literature also are in agreement with our results (33,34). Thus, real-time PCR studies identified that the levels of MT1-MMP expression in HT1080 cells are highly similar to those found in human umbilical vein endothelial cells, prostate carcinoma PC3, breast carcinoma BT549, and melanoma G361 cells (33).
We conclude from these findings and the data available from the literature that the levels of MT1-MMP in the MT1-MMPtransfected breast carcinoma MCF7 cell model are comparable with those existing naturally in fibrosarcoma HT1080 cells as well as in numerous other cell types including breast carcinoma BT549 cells, human umbilical vein endothelial cells, and melanoma G361 cells. Therefore, the effects of MT1-MMP on the complement system reported below with the transfected MCF7 cells are directly applicable to many other cell systems in which this membrane protease is synthesized and activated naturally.

MT1-MMP Efficiently Cleaves C3b in Both Cell-free and Cellbased Systems-Previous
studies have demonstrated the susceptibility of cell surface-bound opsonic C3b to several distinct proteases (16 -18). To extend these observations and to examine if the complement components C3b, iC3b, and C4b are susceptible to MT1-MMP proteolysis, we co-incubated these purified human components of complement with increasing concentrations of the proteolytically potent catalytic domain of MT1-MMP (MT1-CAT) and analyzed the digest samples by Western blotting. C4b and, especially C3b and iC3b, were cleaved by catalytic amounts of MT1-CAT. Multiple 30 -50-kDa cleavage fragments were observed in the C3b/iC3b cleavage reactions. The ␣ chain of both C3 and C4 was especially sensitive to MT1-MMP proteolysis in vitro (Fig. 2). A potent broadrange hydroxamate inhibitor of MMP activity, GM6001 (5 M), blocked MT1-MMP proteolysis of C3b, iC3b, and C4b.
To investigate if MT1-MMP cleaves the complement components when deposited on a tumor cell surface, we performed proteolytic assays using complement opsonized MCF-zeo and MCF-MT cells. To activate complement and deposit C3b and C4b on the cell surface, the cells were preincubated with anti-MCF7 cell-specific antibodies and then incubated for 1 h in human C5-depleted (to prevent MAC-mediated cell lysis) serum. Cellular MT1-MMP was then allowed to cleave the deposited C3b and C4b. The residual levels of cell-associated C3b and C4b were next determined by Western blotting and flow cytometry.
Western blotting showed that MCF-MT cells but not MCF-zeo cells reduced the levels of cell surface-associated C4b and, especially, C3b (Fig. 3). In agreement, the staining of the cells with an antibody specific to the C3b ␣ subunit followed by the    (Fig. 4B). Because treatment of MCF-MT cells with GM6001 resulted in a lower clearance of C3 from the cell surfaces than was observed for untreated MCFzeo cells, our data also indicate that other proteinases (most probably, ADAM and ADAMTS enzymes, which are sensitive to inhibition by hydroxamate inhibitors) are potentially involved in shedding C3.
To analyze the relationships between MT1-MMP and C3b in more detail, we determined the rate of deposition of C3b on MCF-zeo and -MT cells. For these purposes, antibody-sensitized cells were incubated with human C5-depleted serum for different periods of time and the deposition of C3b was determined by flow cytometry. FACS analysis showed that in 15-60 min the resultant deposition of C3b was significantly lower on MCF-MT cells than on MCF-zeo (Fig. 5). These data agree well with the observations of others (35,36) who evaluated the effect of the distinct but functionally similar cathepsin L protease on the deposition of C3b on cell surfaces. Our findings show that MT1-MMP inhibits complement activation by clearing cell surface-deposited C3b and C4b from a tumor cell with the resultant release of their respective inactive soluble forms into the extracellular milieu.

MT1-MMP Protects Breast Carcinoma Cells from
Complement-mediated Injury-To investigate further the functional role of MT1-MMP in protecting MCF7 cells from the complement-mediated attack, we performed a kinetic assay. In this assay complement-mediated lysis was assessed by measuring the release of the cytosolic glucose-6-phosphate dehydrogenase from the cells. Our results indicated that the release of glucose-6-phosphate dehydrogenase reached a plateau following a 30min treatment of MCF-zeo and MCF-MT cells with complement and afterward remained fairly constant for an additional 30 min (Fig. 6A). The lysis by complement was significantly repressed in MCF-MT cells. These data correlate well with the other works where a maximal lytic effect of complement was observed following an ϳ30-min treatment (37,38).
To corroborate these findings, we evaluated complementtreated and untreated cells by light microscopy (Fig. 6B). Visual observation showed that MCF-zeo cells did not survive treatment with complement, whereas MCF-MT cells were largely resistant to complement-mediated lysis. We also stained cells with propidium iodide and employed fluorescence microscopy to identify the stained, dying cells (Fig. 6B). Following treatment with human complement, MCF-zeo cultures exhibited a large number of dead cells stained with propidium iodide. In contrast, only a few of the propidium iodide-positive cells were observed in MCF-MT cultures.
The release of glucose-6-phosphate dehydrogenase measured in these individual cell samples, directly confirmed the data of Fig. 6A and the observations gained by microscopy. Thus, the glucose-6-phosphate dehydrogenase release was significantly suppressed in MCF-MT cells relative to that in MCF-zeo cells. Approximately 50% of MCF-zeo cells were injured following co-incubation with complement, whereas only 8% of the MCF-MT cells did not survive the treatment. We conclude that in cell culture conditions MT1-MMP efficiently protects tumor cells against complement activation and lysis by the MAC.

MT1-MMP Does Not Affect the Expression of the Cellular
Complement Regulatory Proteins-Tumor cells are known to defend themselves against the complement by expressing the complement regulatory proteins CD35, CD46, CD55, and CD59 (10,11,39). CD35, however, is not expressed by MCF7 cells (40). We used flow cytometry to determine the expression levels of CD46, CD55, and CD59 in MCF-zeo and -MT cells. There was no correlation between the expression of MT1-MMP and expression levels of CD46, CD55, or CD59 (Table I). Thus, we exclude the possibility that differential expression of the complement regulatory proteins explains the resistance of MCF-MT cells to complement. Conversely, these data further support the unique role of MT1-MMP in defending malignant cells against complement. DISCUSSION We believe that MT1-MMP directly contributes to tumor cell immune resistance and is thus associated with the survival of metastatic cells in the bloodstream. The efficient survival of MCF-MT cells in cytotoxicity tests (Fig. 6) strongly supports this hypothesis.
Our in vitro and cell-based data show that MT1-MMP inhibits complement by efficiently releasing the C3b and C4b complement components from tumor cell surfaces. Very recently, Tam et al. (41) has also demonstrated, by using isotope-coded affinity tag mass spectrometry, that C3 is a target of MT1-MMP proteolysis, thereby directly supporting our data. Conversely, our results identify the temporal and causal link associating MT1-MMP with C3, and demonstrate the biochemical mechanism and the subsequent physiological effect of the MT1-MMP interactions with C3.
Evidently, the complement-suppressing function of MT1-MMP is a novel proteolytic defense mechanism. The multifunctional MT1-MMP enzyme plays a number of important roles in connection with malignant cells. There is a strong correlation of MT1-MMP expression with the tumorigenicity of many cancer cell types (19,20). In particular, MT1-MMP has been detected in all invasive breast tumor samples characterized by lymph node metastasis and lymph vessel invasion (42)(43)(44). In agreement, the recent work of many authors has demonstrated that MT1-MMP is a potent protease involved in the cleavage of the extracellular matrix, in the activation of several soluble MMPs, and in the processing of the cellular adhesion and signaling receptors such as integrins, CD44, tissue transglutaminase, lipoprotein receptor-related protein (LRP), and gC1q-R (45)(46)(47)(48)(49)(50)(51)(52). Reciprocally, inhibition of the MMP activity by natural and synthetic inhibitors was shown to reduce tumor growth and metastasis (53).
Our current work supports and extends these earlier studies and provides evidence that the complement proteins C3b and C4b are novel and important targets of MT1-MMP. MT1-MMP decreases the deposition of C3b and C4b at the cell surface, and inhibits amplification of the complement cascade. Notably, FIG. 6. MT1-MMP activity protects MCF7 cells from the lysis by MAC. Sensitized MCF-zeo and -MT cells were incubated with 20% normal human serum to induce cell lysis. A, the release of cytosolic glucose-6-phosphate dehydrogenase follows treatment of MCF7 cells with complement. Cells were incubated for the time indicated with complement, and the extent of cell lysis (killing rate) was determined by measuring the rate of the release of cytosolic glucose-6-phosphate dehydrogenase using the Vybrant Cytotoxicity Assay Kit. The data are expressed as a percentage relative to the killing rate of MCF-zeo cells after a 30-min incubation with complement. B, micrographs of the cells after treatment with complement. Upper panels, Nomarski images of the MCF-zeo and -MT untreated cells. Middle panels, Nomarski images of the sensitized MCF-zeo and -MT cells co-incubated with 20% normal human serum. Bottom panels, fluorescence images of the sensitized MCF-zeo and -MT cells incubated with 20% normal human serum and stained with propidium iodide. Bar, 100 m. In addition, the efficiency of lysis (killing rate) was expressed as a percentage (shown below the micrograph panels) relative to the fully lysed samples. The data from one representative experiment performed in triplicate are shown. iC3b, a ligand for complement receptors expressed on phagocytes and NK cells, is also susceptible to MT1-MMP. This fact is important because phagocytes and NK cells can be highly potent in destroying malignant cells. Proteolytic regulation of C3b and C4b presentation at the cell surface plays a significant role in negating complement activation. It has been shown that the degradation of C3b by the p65 proteinase makes human melanoma SK-MEL-170 cells resistant to the complement (38). Conversely, inhibition of the C3-cleaving cathepsin L-like cysteine proteinase p39 strongly decreased tumorigenicity and the metastatic potential of human melanoma xenografts in nude mice (17). In agreement, transfection with L-cathepsin transformed non-metastatic melanoma cells into highly tumorigenic and metastatic cells (35,36). In a similar fashion, expression of rat Crry (an inhibitor of complement activation) inhibited the in vivo deposition of complement C3 fragments and sharply enhanced the tumorigenicity of MCF7 cells. In turn, the expression of rat CD59 (an inhibitor of the cytolytic membrane attack complex) had no effect on the incidence and growth rate of MCF7 tumor xenografts in nude rats (10). The level of C3b deposition is important in tumor immunity because C3b and its cleavage fragments on the cell surface serve as receptors for immune effector cells. Thus, in a mouse tumor model, the deposition of C3b enhanced phagocytosis of mastocytoma P815 cells by macrophages (8,54). Our data agree well with these studies and indicate that the modulation of C3b and C4b deposition by MT1-MMP is likely to help malignant cells to evade host immune surveillance and thus to grow more efficiently as tumors and as metastases. There are extensive data suggesting that antibody-dependent cell-mediated cytotoxicity enhanced by complement is a powerful mechanism to limit tumor growth (55). MT1-MMP-dependent proteolysis of C3b and C4b is thus very likely an important factor in tumor immunity.
It is well established that normal cells synthesize a number of the complement regulatory proteins to control and avoid aberrant activation of the complement system. These soluble and plasma membrane-associated proteins are frequently overexpressed by malignant cells to escape complement attack (11,12). We failed to find any correlation between MT1-MMP expression with the levels of the complement regulatory CD46, CD56, and CD59 proteins. We conclude therefore that MT1-MMP directly inhibits complement. This hypothesis correlates well with the earlier observations that indirectly indicated the existence of the regulatory mechanisms that are used by tumor cells to escape immune attack. These mechanisms are additional to the mechanisms involving the CD46, CD55, and CD59 complement regulatory proteins (12,40).
We believe that our studies have identified a biochemical mechanism that involves MT1-MMP in tumor immunity and have shed light on the exceptionally important role that this protease plays in connection with invading metastatic malignant cells. Our findings warrant additional studies. The studies employing the expression of murine MT1-MMP in murine tumor cell lines followed by tumor xenografts in complementdeficient mice will clarify the role of MT1-MMP in the tumor defense mechanisms.