Cholesterol Sulfate Alters Substrate Preference of Matrix Metalloproteinase-7 and Promotes Degradations of Pericellular Laminin-332 and Fibronectin

Abstract

Localization of secreted matrix metalloproteinases (MMPs) on the cell surface is required not only for processing of cell surface proteins, but also for controlled degradation of the extracellular matrix (ECM). Our previous study demonstrated that binding of MMP-7 (matrilysin) to cell surface cholesterol sulfate (CS) is essential for the cell membrane-associated proteolytic action of this MMP. In this study, we investigated the role of CS in the MMP-7-catalyzed degradation of protein components of ECM. We found that the degradation of laminin-332 (laminin-5) catalyzed by MMP-7 was accelerated dramatically in the presence of CS, whereas the sulfated lipid inhibited the degradation of casein catalyzed by the protease. The MMP-7-catalyzed degradation of fibronectin was partially inhibited in the presence of low concentrations of CS, whereas it was accelerated significantly at high concentrations of the lipid. Therefore, it is likely that CS alters the substrate preference of MMP-7. We also found that the proteins of which MMP-7-catalyzed degradation were accelerated by CS also had affinities for CS, suggesting that CS facilitates the proteolyses by cross-linking MMP-7 to its substrates. Moreover, MMP-7 tethered to cancer cell surface via CS degraded fibronectin and laminin-332 coated on a culture plate. The degradations of the adhesive proteins led to significant detachment of the cells from the plate. Taken together, our findings provide a novel mechanism in which cell surface CS promotes the proteolytic activities of MMP-7 toward selective substrates in the pericellular ECM, thereby contributing to cancer cell migration and metastasis.

Introduction

The matrix metalloproteinases (MMPs)² comprise a family of zinc-dependent endopeptidases that degrade components of extracellular matrix (ECM) and play pivotal roles in tissue remodeling under physiological and pathological conditions such as morphogenesis, angiogenesis, tissue repair, and tumor invasion (1–4). It has been also suggested that several MMPs proteolytically modulate the biological functions of various cell surface proteins, including growth factor precursors, growth factor receptors, and cell adhesion molecules (4); such regulation as well as MMP-catalyzed degradation of ECM is important for tumor growth, invasion, metastasis, and progression (5, 6). It is believed that the localization of secreted MMPs on the cell surface is required not only for processing of cell surface molecules, but also for controlled degradation of the pericellular ECM (7–9). Controlled ECM degradation is important for directed cellular invasion, whereas extensive uncontrolled ECM degradation rather impedes the invasive process, because cells require specific ECM components for their migration and survival (10, 11).

MMP-7 (matrilysin) is the smallest member of the MMP family, of which the pro-form consists of only a propeptide and a catalytic domain and lacks the C-terminal hemopexin-like domain that appears commonly in other MMPs. Although various MMPs are overexpressed both in stromal and tumor cells in cancer tissues, MMP-7 has been detected specifically in tumor cells but not in stromal cells (12). Expression of MMP-7 is correlated well with malignancy and metastasis of cancers, especially in liver metastasis of colon cancer (13). MMP-7 also has physiological roles besides the pathological roles; the data from the MMP-7-deficient mice suggest that this protease is responsible for the activation of prodefensins and thereby participates in innate host defense (14). It has been reported that MMP-7 processes several cell surface proteins, such as Fas-ligand (15), pro-tumor necrosis factor (16), syndecan-1 (7), and E-cadherin (17). This MMP also cleaves Notch on the cell surface and promotes the dedifferentiation of pancreatic acinar cells by activating the Notch signaling pathway (18); the dedifferentiation of the pancreatic cells is associated with an increased risk for tumorigenesis. We previously reported that active MMP-7 efficiently binds to the surface of colon cancer cells and induces E-cadherin-mediated cell aggregation by processing a cell surface protein(s). The aggregated cells showed dramatically enhanced metastatic potential in the nude mouse model (19). Therefore, it seems likely that MMP-7-catalyzed processing of cell surface proteins associates preferentially with pathological stages.

Recently, we identified cholesterol sulfate (CS) as a major cell surface substance to which active MMP-7 binds and demonstrated that binding of MMP-7 to CS is essential for its membrane-associated proteolytic action and induction of the cell aggregation (20). More recently, we also identified the amino acid residues of MMP-7 essential for binding to CS (21). In the three-dimensional structure, the residues constituting the CS-binding site are located on the molecular surface in the opposite side of the catalytic cleft of the protease. Therefore, it is assumed that the active site of MMP-7 bound to the cell surface is directed outside. The direction of the cell-bound MMP-7 likely makes it feasible for the protease to cleave its protein substrates on cell surface as well as those in the pericellular ECM. Our previous study (21) demonstrated that the catalytic activities of MMP-7 toward synthetic peptidyl substrates are reduced upon its complex formation with CS. Kinetic studies suggested that CS binds to the protease and allosterically destabilizes the Michaelis complex, thereby reducing the affinities between the enzyme and the peptidyl substrates (21). In contrast, the binding of MMP-7 to cell surface CS potentiates the ability of the protease to cleave specific cell surface proteins (20). These data led us to speculate that CS modifies the substrate preference of MMP-7. In this study, we investigated the effects of CS on the MMP-7-catalyzed degradations of protein substrates, including components of ECM. We found that MMP-7-catalyzed degradations of laminin-332 (Lm332 and laminin-5) and fibronectin (FN) were accelerated dramatically in the presence of CS, whereas that of casein was inhibited. Moreover, we demonstrated that affinities between CS and the protein substrates are critical for the CS promotion of the MMP-7-catalyzed proteolyses. Our findings provide a molecular basis to explain how cell surface CS regulates the substrate preference of MMP-7. Clarification of the mechanism also provides the potential to develop MMP-7-targetted novel anti-cancer drugs that specifically block the proteolytic action of the MMP-7·CS complex formed on cancer cell surface.

EXPERIMENTAL PROCEDURES

Antibodies and Other Reagents

The sources of materials used were as follows: CS, cholesterol, dehydroepiandrosterone sulfate (DHEAS), human plasma FN, κ-casein, and β-cyclodextrin (β-CD) from Sigma; human recombinant MMP-7, heparin, and protease inhibitor mixture (containing 100 mm 4-(2-aminoethyl)benzenesulfonylfluoride hydrochloride, 80 mm aprotinin, 5 mm bestatin, 1.5 mm E-64 protease inhibitor, 2 mm leupeptin, and 1 mm pepstatin) from Wako Pure Chemical Industries (Osaka, Japan); the synthetic MMP inhibitor TAPI-1 (N-(R)-(2-(hydroxyaminocarbonyl)methyl)-4-methylpentanoyl-l-naphtyalanyl-l-alanine-2-aminoethyl amide) from Peptides Institute, Inc. (Osaka, Japan); the anti-human FN monoclonal antibody N-219 from Calbiochem (La Jolla, CA); the anti-human laminin β3 chain monoclonal antibody Kalinin B1 from BD Bioscience (San Jose, CA); the anti-human MMP-7 monoclonal antibody 11B4G was a kind gift from Dr. Y. Matsuo (Nagahama Institute of Oriental Yeast Co., Shiga, Japan). A recombinant human Lm332 was prepared as reported previously (22). The monoclonal antibodies against human laminin α3 chain (LSα3c3 and BG5) and γ2 chain (D4B5) were established and characterized previously (23, 24). The MMP-7 mutant having low CS-binding affinity and inactive MMP-7 mutant, which corresponding to MMP-7 (29, 33, 51, 55/M2) ΔC3 (21) and E215A mutant (20) in the previous reports, respectively, were prepared as described previously. All other chemicals used for the experiments were of analytical grade or the highest quality commercially available.

Assay of MMP-7-catalyzed Cleavages of FN, κ-Casein, and Lm332

FN (2.5 μm), κ-casein (20 μm), or Lm332 (0.3 μm) were incubated with 0, 20, or 25 nm MMP-7 in the presence of various concentrations of CS in 50 μl of 50 mm Tris-HCl (pH 7.5) containing 150 mm NaCl, 10 mm CaCl₂, and 0.05% polyethylene glycol (PEG) 4000 at 37 °C for various lengths of time. After incubation, samples were mixed with SDS-sampling buffer to make the final concentrations to be 50 mm Tris-HCl (pH 6.8), 2% 2-mercaptoethanol, 2% SDS, and 10% glycerol and subjected to SDS-PAGE. The separated proteins on the gel were visualized by using CBB R-250 staining or silver staining. For Western blot analysis, proteins separated by SDS-PAGE were transferred onto nitrocellulose membranes, and the membranes were incubated with specific antibodies. The immunoreactive protein bands were then detected using the biotin-conjugated secondary antibody, the streptavidin-conjugated alkali phosphatase, and the reagents for color development that contain 5-bromo-4-chloro-3-indolyl phosphate and nitro blue tetrazolium as described previously (20).

Amino Acid Sequence Analyses of Fragments of FN Produced by MMP-7 Digestion

Fragments of FN produced by MMP-7 digestion were separated on 4–20% polyacrylamide gels by SDS-PAGE. Proteins on the gels were transferred to polyvinylidene difluoride membranes (Millipore, MA). The proteins on the membrane were visualized by CBB R-250 staining, and the N-terminal amino acid sequences of individual fragments were analyzed on a 492 Procise protein sequencer (Applied Biosystems).

Cell Lines and Culture Conditions

Human colon cancer cell lines Colo201 and WiDr, and human bladder cancer cell line EJ-1, were obtained from the Japanese Cancer Resources Bank. These cell lines were maintained in Dulbecco's modified Eagle's medium and Ham's F-12 medium (DMEM/F-12, Invitrogen) supplemented with 10% fetal bovine serum.

β-CD Treatment of Cells and Replenishment of Cells with CS or Cholesterol

β-CD treatment of cells and their replenishment with CS or cholesterol was carried out as described previously (20). Briefly, Colo201 cells (1 × 10⁶) were incubated without or with 10 mm β-CD in 2 ml of serum-free medium on 35-mm dishes (Celltight X, Sumitomo Bakelite, Tokyo, Japan) at 37 °C for 30 min. To replenish the cells with CS or cholesterol, the β-CD-treated cells were washed three times with serum-free medium, and then incubated in 2 ml of serum-free medium in the presence of 250 μm β-CD saturated with CS or cholesterol at 37 °C for 1 h, respectively. The cells treated as described above were washed three times with serum-free medium and used for assays.

Preparation of Membrane Fraction of Colo201 Cells

The membrane fraction of cells was prepared as described previously (20). Briefly, Colo201 cells collected by centrifugation were washed three times with serum-free medium, and then homogenized in 1 ml of 20 mm HEPES buffer (pH 7.5) containing 250 mm sucrose by a Potter-Elvehjem-type homogenizer. The homogenates were centrifuged at 800 × g for 7 min to remove nuclei and cellular debris. The postnuclear supernatant was centrifuged at 21,000 × g for 30 min, and the resultant precipitate was used as the cell membrane fraction.

Assay of Binding of Proteins to CS

FN (1 μm), κ-casein (20 μm), Lm332 (0.3 μm), or MMP-7 (50 nm) was incubated with various concentrations of CS in 40 μl of 50 mm Tris-HCl (pH 7.5) containing 150 mm NaCl, 10 mm CaCl₂, and 0.05% PEG 4000 at 25 °C for 1 h. The samples were then centrifuged at 21,000 × g for 15 min. The resultant supernatants and precipitates were separately mixed with the SDS-sampling buffer and subjected to SDS-PAGE.

Assay of Binding of Proteins to Colo201 Cells

Colo201 cells (2 × 10⁶) treated with β-CD or those replenished with CS or cholesterol as described above were incubated with 5 μm FN, 1 μm Lm332, or 50 nm MMP-7 in 1 ml of serum-free DMEM/F-12 medium in the 35-mm dishes at 37 °C for 6 h. After incubation, the supernatants and the membrane fractions of the cells were prepared as described above. The proteins in the supernatants or membrane fractions were analyzed by Western blot analysis with their specific antibodies.

Preparation of FITC-conjugated FN

FN (4 μm) was incubated with 50 μm fluorescein isothiocyanate (FITC) in 50 mm carbonate buffer (pH 9.5) containing 150 mm NaCl at 25 °C for 1 h. After incubation, FITC-conjugated FN was separated from free fluorescein by gel-filtration chromatography on Sephadex G-25 column equilibrated with Ca²⁺/Mg²⁺-free phosphate-buffered saline (PBS).

Assay of Degradations of Proteins Catalyzed by Cell-bound MMP-7

Colo201 cells (2 × 10⁶) were incubated with 0 or 50 nm MMP-7 in 2 ml of serum-free medium on the 35-mm dishes at 37 °C for 1 h. After incubation, cells were washed three times with serum-free medium, and then the cells were suspended in 1 ml of serum-free medium containing the protease inhibitor mixture of which the final concentrations of the constituents were adjusted to be 100 μm 4-(2-aminoethyl)benzenesulfonylfluoride hydrochloride, 80 μm aprotinin, 5 μm bestatin, 1.5 μm E-64 protease inhibitor, 2 μm leupeptin, and 1 μm pepstatin, respectively, were incubated further with 2.5 μm FN or 0.3 μm Lm332 in the 35-mm dishes in the presence of 0 or 20 μm TAPI-1 at 37 °C for 24 h. After incubation, the cells were removed by centrifugation. The proteins in the resultant supernatants were analyzed by Western blot analysis with their specific antibodies.

Analysis of MMP-7-catalyzed Degradation of Lm332 or FN Coated on the Plastic Plate

Each well of eight-well Lab-Tek chamber slides (Nunc, Naperville, IL) was incubated with 150 μl of 3 μm FITC-conjugated FN or 0.1 μm Lm332 in PBS at 4 °C for 24 h to coat the protein. EJ-1 cells (1 × 10⁶) were incubated with 0 or 50 nm MMP-7 in 2 ml of serum-free medium on 35-mm dishes at 37 °C for 1 h. After incubation, the cells were washed three times with serum-free medium and then suspended in serum-free medium containing the protease inhibitor mixture at a density of 5 × 10⁵ cells/ml. 250 μl of the cell suspension was inoculated per well of the protein-coated plates and incubated at 37 °C for 24 h in the presence of 0 or 20 μm TAPI-1. After incubation, cells were lysed with 20 mm NH₄OH, and the plates were washed with PBS. Fluorescence image of the FITC-labeled FN remained on the plastic plates was analyzed directly using a fluorescence microscope (model BZ-8000, Keyence, Osaka, Japan). The image of remaining Lm332 was also obtained using the FITC-labeled antibody as follows: each washed well was blocked with 1.2% BSA in PBS at 25 °C for 1 h, and then incubated with 3 μg/ml of the anti-laminin α3 chain monoclonal antibody (BG5) in PBS containing 1.2% BSA at 25 °C for 1 h. The well was washed and incubated further for 1 h with the FITC-conjugated secondary antibody (Vector Laboratories, Burlingame, CA) diluted 200-fold with PBS containing 1.2% BSA.

Cell Adhesion Assay

The cell adhesion assay was performed as described previously (25). Briefly, 96-well microtiter plates were incubated with 50 μl of 2.5 μm FN or 0.1 μm Lm332 in PBS at 4 °C for 24 h to coat the protein, and then blocked with 1.2% BSA in PBS at 25 °C for 1 h. Cells incubated with or without MMP-7 as described above were suspended in serum-free medium containing the protease inhibitor mixture at a density of 5 × 10⁵ cells/ml, and a 100-μl aliquot was inoculated per well of the plates in the presence of 0 or 20 μm TAPI-1 and incubated at 37 °C for various lengths of time. After incubation, non-adherent cells were removed after gentle agitation, and adherent cells were fixed with 2.5% glutaraldehyde and stained by incubating with 0.0005% Hoechst 33342 dissolved in 0.001% Triton X-100 at 25 °C for 1.5 h. The fluorescent intensity of each well of the plates was measured using the spectrofluorometer Plate Chameleon (Hidex, Turku, Finland) with 355 nm excitation and 460 nm emission filters.

RESULTS

Effects of CS on MMP-7-catalyzed Degradations of FN, κ-Casein, and Lm332

Because FN and Lm332 are components of ECM accessible to cell-bound MMP-7, these proteins might be the physiological substrates of the CS-bound MMP-7. FN is indeed known as a substrate of free MMP-7 (26). It is also known that casein is cleaved efficiently by MMP-7, and this protein has been used as a substrate to assay the activity of MMP-7, although the protein probably is not a physiological substrate of the protease.

To examine the effect of CS on the proteolytic activity of MMP-7, FN, κ-casein, and Lm332 were incubated with MMP-7 in the presence of various concentrations of CS, and degradation of these proteins was analyzed by SDS-PAGE. As shown in Fig. 1A, when FN was incubated with MMP-7, this protein (250 kDa) was degraded, and several lower molecular weight fragments were yielded after 8-h incubation. The degradation of FN was partially inhibited in the presence of 50 μm CS, whereas it was significantly accelerated at 500 μm CS. We also examined the effect of CS on the time course of the degradation of FN. As shown in Fig. 1 (B and C), 41% of the original FN was converted to the smaller fragments after 3-h incubation in the absence of CS, whereas intact FN remained after the incubation in the presence of 500 μm CS was ∼28% of the original one. Therefore, the MMP-7-catalyzed degradation of FN was accelerated ∼1.8-fold by the concentration of CS. When FN was incubated with MMP-7 alone, 190- and 150-kDa fragments appeared as the major products, whereas these fragments only transiently appeared, and instead, a 110-kDa fragment appeared when FN was incubated with MMP-7 in the presence of 500 μm CS. Therefore, it is likely that the sites of FN preferentially cleaved by MMP-7 were altered by CS. We performed the N-terminal sequence analyses of the individual fragments produced by MMP-7 digestion of FN in the absence or presence of CS. As shown in Fig. 1D, all of the 190-, 150-, and 110-kDa fragments had LVATS sequence corresponding to 721–725 of FN in their N termini. We also found that the fragments having N-terminal LVQTA and LNQPT sequences corresponding to 1624–1628 and 2197–2201 of FN, respectively, were produced only in the presence of CS. Therefore, the MMP-7-catalyzed cleavages of the peptide bond between Pro¹⁶²³ and Leu¹⁶²⁴, and that between Gly²¹⁹⁶ and Leu²¹⁹⁷ of FN are likely accelerated in the presence of CS. It is possible that CS accelerates the cleavage of the Pro¹⁶²³–Leu¹⁶²⁴ bond in the 190- and 150-kDa fragments to yield the 110-kDa one corresponding to 721–1623 of FN. In contrast to the case of FN, MMP-7-catalyzed degradation of κ-casein was simply inhibited with increasing concentrations of CS, and it was inhibited almost completely in the presence of 500 μm CS (Fig. 1E).

FIGURE 1.

Effects of CS on MMP-7-catalyzed degradations of FN, κ-casein and Lm332. A, FN (2.5 μm) was incubated without (-MMP-7) or with 25 nm MMP-7 in the presence of indicated concentrations of CS at 37 °C for 8 h. B, FN (2.5 μm) was incubated without (-MMP-7) or with 25 nm MMP-7 in the absence (−) or presence (+) of 500 μm CS at 37 °C for indicated length of time. The arrowheads at 190, 150, and 110 kDa indicate the fragments of FN differentially produced by MMP-7 depending on the presence or absence of 500 μm CS. C, the intensities of the band of 250-kDa FN incubated without (open circles) or with (closed circles) 500 μm CS were measured using ImageJ (National Institutes of Health). The intensity of the band of 250-kDa FN before incubation was taken as 100%, and the relative intensity is shown on the ordinate. D, amino acid sequence analyses of fragments of FN produced by MMP-7 digestion. FN (2.5 μm) was incubated with 20 nm MMP-7 in the absence or presence of 500 μm CS at 37 °C for 9 h. The N-terminal amino acid sequences of the resultant fragments were analyzed as described under “Experimental Procedures.” The identified sequences (underlined letters) of the fragments corresponding to the bands indicated by a–e, and their locations in the primary structure of FN are shown on the right. The open arrowheads indicate the fragments at 190, 150, and 110 kDa, and the closed arrowheads in the sequences represent the deduced sites of MMP-7 cleavage in the sequence of FN. E, κ-casein (20 μm) was incubated without (-MMP-7) or with 25 nm MMP-7 in the presence of indicated concentrations of CS at 37 °C for 8 h. F, Lm332 (0.3 μm) was incubated with MMP-7 in the indicated enzyme to substrate molar ratio at 37 °C for 1 h. G–J, Lm332 (0.3 μm) was incubated without (-MMP-7) or with 20 nm MMP-7 in the presence of indicated concentrations of CS at 37 °C for 1 h. The arrowheads at 45, 29, and 26 kDa in I, and those at 65 and 26 kDa in J, indicate the fragments derived from the β3 and the γ2 chains, respectively. All of the reaction mixtures contained 50 mm Tris-HCl (pH 7.5), 150 mm NaCl, 10 mm CaCl₂, and 0.05% PEG 4000. The resultant digests were subjected to SDS-PAGE. The protein bands of fragments derived from FN (A, B, and left panel in D) or κ-casein (E) were visualized by CBB R-250 staining, and those from Lm332 (F and G) were visualized by silver staining. The degradation of each chain of Lm332 was further analyzed by Western blot analysis with the monoclonal antibodies against the α3 (H), β3 (I), and γ2 (J) chains, respectively.

As shown in Fig. 1F, when Lm332 was incubated with MMP-7 in an enzyme to substrate molar ratio of 1: 15, almost no cleavage of each chain was observed. Only the γ2 chain was degraded when Lm332 was incubated with MMP-7 in the ratio that the enzyme was 1.7-fold molar excess to the substrate, suggesting that Lm332 is relatively resistant to MMP-7 cleavage. On the other hand, the MMP-7-catalyzed degradation of Lm332 was accelerated dramatically by CS, and in the presence of 1 μm or higher concentrations of CS, all of the three chains were degraded even when Lm332 was incubated with MMP-7 in an enzyme to substrate molar ratio of 1:15 (Fig. 1G). When the products of the degradation of Lm332 were analyzed with specific antibodies against the α3, β3, and γ2 chains, 45-, 29-, and 26-kDa fragments derived from the β3 chain (Fig. 1I) and 65- and 26-kDa fragments from the γ2 chain (Fig. 1J) were detected. Although the α3 chain (160 kDa) was decreased significantly after the incubation, no smaller fragments were detected (Fig. 1H), suggesting that α3 chain is cleaved at the epitope of the antibody.

Effects of Cholesterol, DHEAS, and Heparin on MMP-7-catalyzed Cleavages of FN and Lm332

To examine whether the compounds structurally similar to CS or those having sulfate group affect the proteolytic activity of MMP-7, we tested the effects of cholesterol, DHEAS, and heparin on MMP-7-catalyzed degradations of FN and Lm332. As shown in Fig. 2 (A and B), these compounds did not affect MMP-7-catalyzed cleavages of FN and Lm332. We also found these compounds tested here had no effect on the peptidolytic activity of MMP-7 toward a synthetic substrate (7-methoxycoumarin-4-yl)-acetyl-Pro-Leu-Gly-Leu-[N^β-(2,4-dinitrophenyl)-l-2,3-diaminopropionyl]-Ala-Arg amide (data not shown). It should be noted that DHEAS does not have the ability to alter the MMP-7 activity, although this compound and CS share the common rings structure and sulfate group (Fig. 2C).

FIGURE 2.

Effects of cholesterol, DHEAS, and heparin on MMP-7-catalyzed degradations of FN and Lm332. A, FN (2.5 μm) was incubated without (-MMP-7) or with 25 nm MMP-7 in the absence (None) or presence of 500 μm CS, 500 μm cholesterol, 500 μm DHEAS, or indicated concentrations of heparin at 37 °C for 8 h. B, Lm332 (0.3 μm) was incubated without (-MMP-7) or with 20 nm MMP-7 for 1 h under the same conditions as described above. The resultant digests were subjected to SDS-PAGE. The protein bands of fragments derived from FN (A) or those of Lm332 (B) were visualized by CBB R-250 staining or silver staining, respectively. C, chemical structure of CS with carbon numbers and that of DHEAS are shown.

Effects of CS on Degradations of FN and Lm332 Catalyzed by the MMP-7 Mutant Having Essentially no Affinity for CS

We reported previously that an MMP-7 mutant, which had the residues essential for binding to CS modified, has essentially no affinity for CS (21). This mutant and wild-type MMP-7 have similar proteolytic activities toward FN, whereas the mutant lacks the ability to induce the colon cancer cell aggregation. To examine whether the interaction between MMP-7 and CS is necessary for the CS alteration of the proteolytic activities of MMP-7 toward FN and Lm332, these proteins were incubated with the MMP-7 mutant in the presence of various concentrations of CS, and the degradations of the proteins were analyzed by SDS-PAGE. As shown in Fig. 3A, in the presence of 50 μm CS, the degradation of FN by wild-type MMP-7 was significantly inhibited as compared with that in the absence of CS, whereas the MMP-7 mutant-catalyzed degradation of FN was not affected by CS. On the other hand, the mutant-catalyzed degradations of the 190- and 150-kDa fragments were accelerated significantly in the presence of 500 μm CS, and the 110-kDa fragment, which is the specific product of the degradation of FN catalyzed by wild-type MMP-7 in the presence of CS, was also produced. In our previous study (21), the MMP-7 mutant, which had the residues essential for binding to CS modified, showed much lower affinity for CS as compared with wild-type MMP-7. However, the peptidolytic activity of the mutant was reduced by 10–20% in the presence of 50 μm CS, suggesting that a small part of the mutant still bound to CS. Therefore, it is possible that an un-negligible part of this mutant binds to CS in the presence of 500 μm CS. This might be the reason why the mutant-catalyzed degradations of the 190- and 150-kDa fragments were significantly accelerated, and the 110-kDa fragment was produced in the presence of 500 μm CS. When Lm332 was incubated with the MMP-7 mutant in the presence of CS, no acceleration of its degradation by CS was observed, whereas CS dramatically accelerated the wild-type MMP-7-catalyzed degradation of Lm332 (Fig. 3B). These results suggest that the interaction between MMP-7 and CS is necessary for CS alteration of the proteolytic activity of MMP-7 toward FN and Lm332.

FIGURE 3.

Effects of CS on degradations of FN and Lm332 catalyzed by MMP-7 mutant having essentially no affinity for CS. A, FN (2.5 μm) was incubated without (-Enzyme) or with 25 nm MMP-7 (WT) or 25 nm the MMP-7 mutant having essentially no affinity with CS (Mutant) in the presence of indicated concentrations of CS at 37 °C for 9 h. The arrowheads indicate the 190-, 150-, and 110-kDa fragments corresponding to those in Fig. 1B. B, Lm332 (0.3 μm) was incubated without (-Enzyme) or with 20 nm MMP-7 (WT) or 20 nm the MMP-7 mutant (Mutant) in the presence of the indicated concentrations of CS at 37 °C for 1 h. The resultant digests were subjected to SDS-PAGE. The protein bands of the fragments derived from FN (A) or those from Lm332 (B) were visualized by CBB R-250 staining or silver staining, respectively.

Abilities of Protein Substrates of MMP-7 to Bind to CS

It is well known that phospholipids dramatically accelerate the cascade reactions of the blood coagulation system, in which the upstream active forms of serine proteases proteolytically activate the downstream protease zymogens, by interacting with both the enzyme and its substrate (27). Therefore, it is possible that CS accelerates the MMP-7-catalyzed degradation of protein substrates by interacting with both the enzyme and substrate. To test this possibility, we examined whether CS has affinities for the protein substrates of MMP-7. As we previously found that free and CS-bound forms of MMP-7 can be separated by centrifugation, and the precipitate fraction contains the CS-bound form of the enzyme (21), FN, κ-casein, and Lm332 as well as MMP-7 were first incubated with CS, respectively, and then separated by centrifugation. The proteins in the resultant supernatant and precipitate fractions were analyzed by Western blot analysis. As shown in Fig. 4A, MMP-7 in the precipitate fraction was increased with increasing concentrations of CS, and a concomitant decrease of the MMP in the supernatant was observed, as expected. When FN and Lm332 were incubated with the various concentrations of CS, these proteins in the precipitate fraction were also increased with increasing concentrations of CS, and almost all proteins were precipitated when they were incubated with 500 μm CS (Fig. 4, B and D). In contrast the cases of FN and Lm332, κ-casein was not precipitated even when it was incubated with 500 μm CS, suggesting that κ-casein does not have an affinity for CS (Fig. 4C).

FIGURE 4.

Abilities of proteins to bind to CS. MMP-7 (50 nm), FN (1 μm), κ-casein (20 μm), and Lm332 (0.3 μm) were incubated with the indicated concentrations of CS at 25 °C for 1 h. Each sample was centrifuged, and then the resultant supernatant (Sup.) and precipitate (Ppt.) were subjected to SDS-PAGE as described under “Experimental Procedures.” MMP-7, FN, and Lm332 partitioned to each the fractions were detected by Western blot analysis with monoclonal antibodies against MMP-7 (A), FN (B), or anti-laminin α3 chain (D). The band of the partitioned κ-casein was visualized by CBB R-250 staining (C).

Abilities of Protein Substrates of MMP-7 to Bind to Cell Surface CS

We previously demonstrated that the binding of MMP-7 to colon cancer cell line Colo201 is mainly mediated by cell surface CS (20). Moreover, we found that the amount of cell surface CS is significantly decreased by β-CD treatment, and that it can be replenished by adding CS exogenously (20). To examine whether cell surface CS has the ability to bind to FN and Lm332, Colo201 cells without the β-CD treatment, cells treated with β-CD, or those replenished with CS or cholesterol after the treatment with β-CD were incubated with FN and Lm332 in the culture medium. The membrane fractions prepared from the incubated cells were then analyzed by Western blot analysis. We confirmed that all of the cells tested here expressed negligible levels of FN and Lm332. As shown in Fig. 5, when the Colo201 cells without the β-CD treatment were incubated with FN and Lm332, these proteins were detected in the membrane fractions. The amounts of FN and Lm332 in the membrane fractions from the β-CD-treated cells were much smaller than those from the non-treated cells. Replenishment of the β-CD-treated cells with CS, but not cholesterol, led to recovery of the proteins in the membrane fraction, suggesting that cell surface CS has abilities to bind to FN and Lm332.

FIGURE 5.

Abilities of FN and Lm332 to bind to Colo201 cells. Colo201 cells suspended in serum-free medium were incubated without (−) or with (+) 10 mm β-CD at 37 °C for 30 min. To replenish the cells with cholesterol or CS, the cells incubated with β-CD were washed and further incubated without (−) or with (+) 250 μm β-CD saturated with cholesterol or CS at 37 °C for 1 h. After incubation, the cells washed and re-suspended were incubated further without (−) or with (+) 5 μm FN or 1 μm Lm332 at 37 °C for 6 h. The incubated cells were washed, homogenized, and then fractionated by centrifugation as described under “Experimental Procedures.” The proteins partitioned in the supernatant (Sup.) and the membrane fraction (Mem.) were detected by Western blot analysis with the monoclonal antibodies against FN (A) and the laminin β3 chain (C). The intensities of the bands of the proteins in the membrane fraction were measured using the NIH image software. The intensity of the band of each protein in the membrane fraction of the cells without the β-CD-treatment was taken as 100%. The amounts of FN (B) or Lm332 (D) in the membrane fraction of the cells treated as indicated at the bottom of each column are shown as the relative intensity of the band. Each bar represents the mean ± S.D. for triplicate assays. Statistical significance was determined by an unpaired t test. *, p < 0.002; **, p > 0.1.

Effects of CS on MMP-7-catalyzed Degradations of Chymotryptic Fragments of FN

The data in Figs. 1 and 4 suggest that the proteins of which MMP-7-catalyzed degradation are accelerated by CS also have affinities for CS. To further test this possibility, we examined effects of CS on MMP-7-catalyzed cleavages of chymotryptic fragments of FN; the mixture of the chymotryptic fragments was used as a peptide library that contains peptides with or without the CS-binding affinity. As shown Fig. 6A, the limited chymotryptic digestion of FN yielded several lower molecular weight fragments. When the mixture of the fragments were incubated with CS and then separated by centrifugation, some of the fragments showed the ability to bind to CS (Fig. 6A). On the other hand, when the chymotryptic fragments were incubated with MMP-7 in the presence of CS, some of the fragments were gradually degraded, whereas MMP-7 alone failed to degrade any of the fragments. As shown in Fig. 6B, all of the fragments cleaved by MMP-7 in the presence of CS also showed ability to bind to CS, suggesting again that the binding of CS to the substrates is critical for the CS promotion of the MMP-7-catalyzed proteolyses.

FIGURE 6.

Effect of CS on MMP-7-catalyzed degradation of the chymotryptic fragments of FN. A, FN (5 μm) was incubated with 40 nm α-chymotrypsin in 80 μl of 50 mm Tris-HCl (pH 7.5), 150 mm NaCl, 10 mm CaCl₂, and 0.05% PEG 4000 at 37 °C for 3 h. After incubation, the sample was mixed with 8 μl of 100 mm PMSF, dissolved in methanol, and further incubated at 25 °C for 15 min to inactivate α-chymotrypsin. The chymotrypsin digest of FN was further incubated with 25 nm MMP-7 in the absence (−CS) or presence (+CS) of 500 μm CS at 37 °C for the indicated length of time. After incubation, each sample was subjected to SDS-PAGE. The chymotrypsin digest of FN was also incubated with 500 μm CS at 25 °C for 30 min. After incubation, the sample was centrifuged and the resultant supernatant (CS-unbound) and precipitate (CS-bound) were subjected to SDS-PAGE. The bands of fragments derived from FN were visualized by CBB R-250 staining. The white, gray, and black arrowheads at 41, 32, and 15 kDa, respectively, indicate the fragments that were degraded by MMP-7 only in the presence of CS. B, the enlarged pictures of A focusing on the 41-, 32-, and 15-kDa fragments in the CS-bound (Bound) and CS-unbound (Unbound) fractions or those after 20-h incubation with MMP-7 (+MMP-7, 20 h) in the absence (−CS) or presence (+CS) of 500 μm CS.

Degradations of FN and Lm332 in Culture Medium by Cell-bound MMP-7

We demonstrated previously that MMP-7 binds to surface of several cancer cell lines, including Colo201 cells, when the cells are incubated with active MMP-7 (20). To examine whether cell-bound MMP-7 has an ability to degrade FN or Lm332, Colo201 cells carrying MMP-7 on their surface were incubated with these proteins in culture medium, and degradations of the proteins were analyzed by Western blot analysis. As shown in Fig. 7A, MMP-7 was detected in the membrane fractions when Colo201 cells were incubated with MMP-7. As MMP-7 was not detected in the supernatant, it is likely that almost all of the added MMP-7 had bound to the cells after the incubation. On the other hand, the hydroxamate-based metalloproteinase inhibitor TAPI-1 significantly inhibited the binding of MMP-7 to the cells as reported previously (20). As shown in Fig. 7 (B and C), when the Colo201 cells carrying MMP-7 were incubated with FN, ∼80% of 250-kDa FN was degraded after 24-h incubation. The degradation of FN was significantly inhibited by TAPI-1, suggesting that the metalloproteinase is responsible for the degradation of FN. As shown in Fig. 7 (D, F, and H), when the cells carrying MMP-7 were incubated with Lm332, ∼80, 65, and 62% of the α3 (Fig. 7E), β3 (Fig. 7G), and γ2 (Fig. 7I) chains were degraded, respectively, after 24-h incubation. The degradation of each chain of Lm332 was also inhibited by TAPI-1.

FIGURE 7.

Degradations of FN and Lm332 in culture medium by cell-bound MMP-7. Colo201 cells suspended in serum-free medium were incubated without (−) or with (+) 50 nm MMP-7 at 37 °C for 1 h. After incubation, the cells were washed and then further incubated without (A) or with (B) 2.5 μm FN or 0.3 μm Lm332 (D, F, and H) in the absence (−) or presence (+) of 20 μm TAPI-1 at 37 °C for 24 h. The incubated cells were washed, homogenized, and fractionated by centrifugation as described under “Experimental Procedures.” A, MMP-7 in the supernatant (Sup.) and the membrane fraction (Mem.) were detected by Western blot analysis with the anti-MMP-7 monoclonal antibody. B, FN and its fragments in the supernatant were detected with the anti-FN monoclonal antibody. D, F, and H, Lm332 or its fragments in the supernatant were detected with the monoclonal antibodies against the α3 (D), β3 (F), and γ2 (H) chains. The same amount of each protein without the incubation (Before incu.) was also subjected to Western blot analysis. The intensities of the bands of 250 kDa FN (C), 160 kDa α3 chain (E), 135 kDa β3 chain (G), and 105 kDa γ2 chain (I) were measured using NIH ImageJ. The percentages of degraded proteins were calculated by using the equation, (1 − (I_a/I_b)) × 100, where I_a and I_b represent the intensities of the bands of the proteins after and before incubation, respectively. Each bar represents the mean ± S.D. for triplicate assays. Statistical significance was determined by an unpaired t test. *, p < 0.001; **, p > 0.05.

Degradations of FN and Lm332 Coated on Culture Plate by Cell-bound MMP-7

We next examined whether cell-bound MMP-7 has the ability to degrade FN and Lm332 coated on plastic culture plates. As shown in Fig. 8A, when EJ-1 cells carrying MMP-7 were plated on the FITC-conjugated FN-coated plate and then incubated, several dark spots, of which sizes are similar to those of EJ-1 cells, were observed in the fluorescence microscopic analysis. On the other hand, no spot was observed when the cells carrying MMP-7 were plated and incubated in the presence of TAPI-1, or EJ-1 cells without MMP-7 were incubated in the absence of the inhibitor, suggesting that MMP-7 on cell surface is responsible for the degradation of FITC-conjugated FN coated on the plate. Similar results were obtained when EJ-1cells carrying MMP-7 were incubated with the plate coated with Lm332, and then degradation of the coated Lm332 was analyzed using the FITC-labeled antibody (Fig. 8B).

FIGURE 8.

Degradations of FN and Lm332 coated on the plate by cell-bound MMP-7. FITC-conjugated FN (A) and Lm332 (B) were coated on 8-well plastic plates as described under “Experimental Procedures.” EJ-1 cells suspended in serum medium were incubated without (-MMP-7) or with 50 nm MMP-7 at 37 °C for 1 h. The cells were washed and then plated onto the plates coated with FN or Lm332 and incubated in the absence or presence of 20 μm TAPI-1 at 37 °C for 24 h. After incubation, the cells were removed, and the FITC-conjugated FN or Lm332 remaining on the plates was visualized as described under “Experimental Procedures.” The arrowheads indicate the dark spots where the coated proteins were degraded. Scale bar, 10 μm.

Detachment of Cells Carrying MMP-7 from Culture Plate Coated with FN or Lm332

The data in Fig. 8 suggest that cell-bound MMP-7 degrades the pericellular ECM components such as FN an Lm332. If FN and Lm332 lose their cell adhesion activities upon the MMP-7-catalyzed degradation, the cells carrying MMP-7 may detach from the culture plate coated with FN or Lm332. We examined this possibility.

As shown in Fig. 9A, when the cells carrying MMP-7 were plated onto the plates coated with FN and Lm332 and incubated, the majority of the cells attached to the plate and spread after 3-h incubation, whereas majority of them detached from the plate and formed aggregations after 24-h incubation. In contrast, EJ-1 cells without MMP-7 attached on the plate and spread throughout the 3- to 24-h incubation, and they did not detach from the plate. The detachment of the cells carrying MMP-7 after 24-h incubation was also inhibited by TAPI-1. When several cancer cell lines, including EJ-1, were examined quantitatively for their detachments induced by the cell surface MMP-7, 79% EJ-1 cells and 70% WiDr cells carrying MMP-7 detached from the FN-coated plates, whereas 62% EJ-1 cells and 79% Colo201 cells carrying the protease detached from the Lm332-coated plates after 24-h incubation (Fig. 9B). These results are consistent with the view that MMP-7 tethered to cell surface via CS degrades FN and Lm332 on the plate and abrogates their cell adhesion activities, thereby leading to cell detachment. However, considering that CS has the ability to bind to FN and Lm332, there is another possibility that cell surface CS directly supports the cell adhesion by binding to FN and Lm332, and competitive binding of MMP-7 to CS led to the cell detachment. To test the latter possibility, we examined the effect of the catalytically inactive MMP-7 mutant, of which affinity for CS is almost the same as that of wild-type MMP-7 (20), on the cell detachment. When EJ-1 cells carrying the inactive mutant of MMP-7 were plated onto the plate coated with FN and Lm332 and incubated, no detachment of the cells was observed even after 24-h incubation, suggesting that proteolytic activity of MMP-7 on the cell surface is necessary for the induction of cell detachment.

FIGURE 9.

Detachment of cancer cells carrying MMP-7 from culture plate coated with FN or Lm332. A, EJ-1 cells suspended in serum-free medium were incubated without (-MMP-7) or with 50 nm MMP-7 or 50 nm the catalytically inactive MMP-7 mutant (+E/A MMP-7) at 37 °C for 3 h. The incubated cells were washed, and then plated on each well of 8-well plates coated with FN or Lm332 and incubated in the absence or presence of 20 μm TAPI-1. After 3- or 24-h incubation, cells were photographed. The arrowheads represent the aggregations of the cells detached from the plates. Scale bar, 20 μm. B, EJ-1, WiDr, or Colo201 cells suspended in serum-free medium were incubated without (−) or with (+) 50 nm MMP-7 at 37 °C for 3 h. After incubation, the cells were washed, and then plated on each well of 96-well plates coated with the FN or the Lm332 and further incubated. After 24 h, quantities of adherent cells were determined by staining the cells with Hoechst 33342 followed by measurement of their fluorescent intensities. The quantity of adherent cells, which had been incubated without MMP-7, was taken as 100%; columns shown are relative quantities of the adherent cells that had been incubated under the indicated conditions. Each bar represents the mean ± S.D. for triplicate assays. Statistical significance was determined by an unpaired t test. *, p < 0.001. Each column represents the mean of triplicate assays. Bars, ±S.D. C and D, EJ-1 cells suspended in serum-free medium were incubated without (open circles) or with (closed circles) 50 nm MMP-7 at 37 °C for 3 h. The incubated cells were washed, plated on each well of 96-well plates coated with FN (C) or Lm332 (D), and incubated. After 24 h, the cells were trypsinized, plated on each well of the newly prepared 96-well plates that were coated with FN (C) or Lm332 (D), and further incubated. After incubation for indicated lengths of time, the quantities of adherent cells were determined by staining with Hoechst 33342 followed by measurement of their fluorescent intensities. Each point represents the mean ± S.D. for triplicate assays.

We previously demonstrated that the binding of MMP-7 to cell surface CS is essential for MMP-7-catalyzed cleavage of membrane proteins (20). To examine whether MMP-7-catalyzed cleavages of membrane proteins affect the ability of the cells to attach to the plate coated with the adhesive proteins, EJ-1 cells carrying MMP-7 or those without MMP-7 were first incubated for 24 h, and then the incubated cells were plated onto the newly prepared plate coated with FN or Lm332. As shown in Fig. 9C, when the EJ-1 cells carrying MMP-7 and those without MMP-7 were incubated on FN-coated plate for 24 h and then the incubated cells were plated on the newly prepared FN-coated plate and incubated, there was no significant difference in the rates of cell attachment between the cells carrying MMP-7 and those without MMP-7. Similar results were obtained in the experiment using the Lm332-coated plates (Fig. 9D), suggesting that the degradations of the adhesive proteins on the plate, but not the modification of cell surface proteins, are important for the induction of cell detachment.

DISCUSSION

We examined the effects of CS on the catalytic activities of MMP-7 toward protein substrates and found that CS dramatically accelerated MMP-7-catalyzed degradation of Lm332, whereas the lipid inhibited the degradation of casein catalyzed by the MMP. On the other hand, The MMP-7-catalyzed degradation of FN was partially inhibited in the presence of 50 μm CS, whereas it was accelerated significantly at 500 μm CS. Therefore, it is likely that CS alters the substrate preference of MMP-7. Because CS did not affect significantly the degradations of FN and Lm332 catalyzed by the MMP-7 mutant that has essentially no affinity for CS, the interaction between MMP-7 and CS is likely necessary for the alteration of the proteolytic activities of the protease. These data also rule out the possibility that CS promotes the denaturation of the protein substrates and thus makes them susceptible to the proteolytic cleavage. As we discussed in the previous report (21), CS molecules probably form clusters in aqueous solution, and several CS molecules in the cluster are involved in the interaction with one molecule of MMP-7. Although CS and DHEAS share the common rings structure and sulfate group, the latter compound did not affect the proteolytic activities of MMP-7 (Fig. 2). We also found that DHEAS failed to compete with CS to bind to MMP-7 (data not shown), suggesting that the sulfated steroid has essentially no affinity for MMP-7. We speculate that, unlike CS, DHEAS is unable to form clusters due to its polar carbonyl group at position 18 and lack of the hydrocarbon chain corresponding to the positions 20–27 of CS (see Fig. 2C), and MMP-7 does not have significant affinity for the monomer form of the sulfated steroid. We also found that FN and Lm332, but not casein, had affinities for CS, and all of the chymotryptic fragments of FN, of which degradation was accelerated by CS, also had affinities for CS. These data strongly suggest that the binding of CS to both MMP-7 and its substrate is critical for the CS promotion of the proteolytic activity of the enzyme. It is likely that the clustered CS molecules mediate the interaction between the enzyme and substrate and, thus, facilitate the enzyme reaction. On the other hand, we previously reported that CS reduces the catalytic activities of MMP-7 toward synthetic peptidyl substrates (21). Kinetic studies and other studies suggested that CS allosterically alters the substrate-binding site of MMP-7 and reduces the affinities between the active site of the enzyme and the peptides, thereby negatively regulating the catalytic activity (21). The CS inhibition of the MMP-7-catalyzed degradation of casein observed in Fig. 1E may reflect the adverse effect of CS on the active site of the enzyme. An explanation for the partial inhibition of the MMP-7-catalyzed degradation of FN in the presence of 50 μm CS might be that, at the low concentrations of CS, only MMP-7, but not FN, effectively interacts with CS; further evidence is needed to clarify this mechanism. Considering that CS accelerates the MMP-7-catalyzed degradation of the substrates, which have affinities for the lipid, the CS promotion of the interaction between MMP-7 and its substrate likely compensates and rather overcomes the adverse effect of CS on the active site. Because of the CS-mediated acceleration of the activities of MMP-7 toward selective substrates and its adverse effect on the activities toward others, CS may lead to drastic alteration of the substrate spectrum of MMP-7. For instance, during the MMP-7-catalyzed degradation of FN in the presence of CS, cleavages of the FN fragments likely proceed depending upon their affinities for CS; the fragments having affinities for CS are further cleaved but the cleavage of those without the affinity are inhibited. This is probably the reason why the products of the degradation of FN catalyzed by MMP-7 alone and those in the presence of CS observed in Fig. 1B were different. Physiologically, the proteolytic activity of MMP-7 tethered to cell surface via CS must be further restricted by the localization of CS. Our previous studies (20) suggest that CS is localized in membrane raft microdomains.

It is known that rafts are specialized and dynamic membrane microdomains, which are enriched in cholesterol and glycosphingolipids (28). CS interacts with sphingolipids in a manner similar to that observed for cholesterol and has an ability to induce hydrocarbon ordering in lipid bilayers (29). Therefore, it is likely that rafts comprise specialized areas, in which CS controls proteolytic activity of MMP-7. It has been reported that MMP-7 proteolytically processes several cell surface proteins, such as Fas-ligand (15), pro-tumor necrosis factor (16), syndecan-1 (7), E-cadherin (17), Notch-1 (18), and β4 integrin (30). Interestingly, all cell surface proteins are constitutively or partially localized in rafts (31–36). We recently reported that MMP-7 cleaves annexin II bound to the tumor cell surface, and the cleavage leads to promotion of the binding of tissue-type plasminogen activator to the tumor cell surface (37). Because annexin II also localizes in rafts (38), the binding of MMP-7 to CS might play pivotal role in the MMP-catalyzed cleavages of specific cell surface proteins by localizing the enzyme in the lipids microdomains. On the other hand, it has been reported that rafts are located in the leading edge of migrating cells (39, 40). Therefore, the interaction between MMP-7 and cell surface CS likely brings the protease to the privileged interface where the pericellular ECM proteins in front of the invading cells are degraded. We demonstrated here that the MMP-7·CS complex degrades at least Lm332 and FN among the ECM proteins. It should be noted that Lm332 in solution and that coated on the plastic culture plate were degraded by MMP-7 bound to cell surface CS, although Lm332 was relatively resistant to the cleavage by MMP-7 alone. Although it is not feasible to define the densities of CS in rafts, the local concentrations of CS on cell surface may be >500 μm, because MMP-7 tethered to cells via CS effectively degraded FN in the culture medium (Fig. 7, B and C). Considering that 50 μm CS has an inhibitory effect on the degradation of FN, it is also possible that MMP-7 bound to rafts relatively poorly in CS has reduced proteolytic activity toward FN but an enhanced activity toward Lm332. In this case, ECM constituted by Lm332, such as basement membranes, will be degraded preferentially by the cell-bound protease. Cells might regulate the MMP-7-catalyzed degradation of ECM by controlling the density of CS on their surface.

Lm332 is a major laminin isoform in the epidermal basement membrane capable of promoting cell adhesion much more efficiently than other ECM proteins. The proteolytic processing of specific Lm332 subunits is known to modulate its cell adhesion and/or cell migration-promoting activities. For instance, proteolytic truncation of the N-terminal domains of γ2 chain of Lm332 suppresses its cell adhesion activity, whereas the truncation promotes its cell migration activity (25). It has been reported that MMP-7 and the γ2 chain of Lm332 are coexpressed in colorectal carcinomas and these molecules are suggested to have roles in the cancer invasion and metastasis. However, direct linkage between these two molecules has not been clarified. Recently, it was also reported that the β3 chain of Lm332 is N-terminally truncated by MMP-7, and the β3 chain (140 kDa) is converted to its 90-kDa form by the cleavage (41). Because the Lm332 having the truncated β3 chain shows an enhanced cell migration-promoting activity, it is suggested that the MMP-7-catalyzed cleavage of Lm332 is crucial for the regulation of cancer cell migration. In the report, however, the β3 chain of Lm332 is only partially cleaved even when the laminin is incubated with MMP-7 in the ratio that the enzyme is 11-fold molar excess to the substrate. Because an extremely high concentration of MMP-7 (1.2 μm) is required for the partial cleavage of the Lm332, physiological significance of the cleavage is questionable. In the present study, we found that CS dramatically enhanced the susceptibility of Lm332 to MMP-7 cleavage, and 20 nm MMP-7 was sufficient for the complete degradation of the laminin. Therefore, as compared with free MMP-7, the MMP-7·CS complex is a much more plausible candidate for the physiological enzyme that cleaves Lm332.

We demonstrated that MMP-7 tethered to cell surface via CS degrades FN and Lm332 coated on the plate and abrogates their cell adhesion activities, thereby leading to cell detachment. The detached cancer cells formed aggregates, suggesting that the membrane-bound MMP-7 also induced the homotypic cell adhesion by proteolytically modulating the cell surface proteins as observed in the previous studies (19–21). We speculate that the detachment of cells from ECM is required for the cancer cell migration and invasion, whereas the homotypic cell adhesion is important for their anchorage-independent survival in the bloodstream and/or their facilitated lodging in the microvascular vessels and subsequent extravasation. Therefore, the MMP-7·CS complex possibly supports several steps of cancer metastasis.

The association of MMPs with cancer metastasis has suggested that these proteases represent attractive targets for the development of novel anti-tumor therapies. However, no MMP inhibitor has been developed successfully as anti-tumor drugs, mainly because of deleterious side effects. The lack of selectivity of MMP inhibitors must be a stiff obstacle for developing safe and effective drugs. We have recently identified a β-amyloid precursor protein-derived decapeptide having the ISYGNDALMP sequence as an MMP-2-selective inhibitor, which interacts with the active site of MMP-2 (42, 43). It has also been reported that the reactive site-modified tissue inhibitor of metalloproteinase-2 specifically blocks the membrane type-1 MMP-catalyzed activation of pro-MMP-2 (44). However, no MMP-7-selective inhibitor has been identified, so far. In this study, we provide evidence that CS accelerates the proteolytic activity of MMP-7 toward selective substrates by cross-linking the enzyme to substrates. Considering that the selective substrates have affinities for CS, MMP inhibitors designed to bind to CS may have enhanced selectivity toward MMP-7 in the presence of CS; such inhibitors thus target the MMP-7·CS complex formed on cell surface. This possibility is currently under investigation in our laboratory. Taken together, our findings provide the potential to develop MMP-7-targeted novel anti-cancer drugs that block specifically the membrane-associated proteolytic action of this MMP, thereby having restricted side effects.