Abstract
The invasive properties of melanoma cells correlate with the expression of matrix metalloproteinases (MMPs) and their physiological modulators (tissue inhibitors of metalloproteinase and membrane-type MMPs) and with that of the αVβ3 integrin. We investigated the effect of anterior lens capsule type IV collagen and of the α3(IV) collagen chain on the invasive properties of various tumor cell lines (HT-144 melanoma cells, HT-1080 fibrosarcoma cells). We demonstrated that anterior lens capsule type IV collagen or specifically the synthetic peptide α3(IV) 185–203 inhibited both the migration of melanoma or fibrosarcoma cells as well as the activation of membrane-bound MMP-2 by decreasing the expressions of MT1-MMP and the β3 integrin subunit.
INTRODUCTION
Tumor cell invasion is characterized by interdependent steps of interactions between tumor cells and the extracellular matrix, involving sequential proteolytic degradation of basement membranes, migration through blood vessels, and adhesion to extracellular matrix proteins (1) . The degradation of basement membranes involves various proteolytic enzymes, mainly MMPs3 ,(2 , 3) . MMPs are a large family of at least 18 members of zinc-containing proteinases that degrade extracellular matrix proteins, such as collagens, proteoglycans, laminins, and fibronectin (4) . Among them, the 72-kDa gelatinase (MMP-2) and stromelysin (MMP-3) and the 92 kDa-gelatinase (MMP-9) have been shown to play an important role in the degradation of basement membranes and in tumor progression. MMPs are secreted as inactive zymogens, and their activation occurs during basement membrane crossing as a result of an imbalance between levels of TIMPs and physiological activators of MMPs, such as urokinase, plasmin, or MT-MMPs (5 , 6) . MT-MMPs constitute a new subgroup of MMPs containing an additional transmembrane domain, and they have been shown to activate latent MMP-2 (7) .
Integrins are αβ heterodimeric cell surface glycoproteins that interact with extracellular matrix proteins and mediate tumor cell adhesion to basement membrane components during tumor progression (8 , 9) . A particular role has been demonstrated for theα Vβ3 integrin in melanoma cell migration and invasion (10 , 11) . Up-regulated levels of the expression of αVβ3 integrin are induced in invasive melanoma cells in an in vitro model in nude mice (12 , 13) . Recent studies have also demonstrated the role of the αVβ3 integrin that colocalizes with MMP-2 in a functionally active form on the melanoma cell surface (14 , 15) .
Type IV collagen is a major component of basement membranes (16) , and one of its important functions is the ability to promote the adhesion and motility of various normal or transformed cells. It is a heterotrimer formed from any of six α chains. The most prominent molecular species, as well as type IV collagen from EHS tumor, is composed of two α1(IV) and one α2(IV) chain (17) . The genes for the additional minor type IV collagen chains, α3(IV), α4(IV), α5(IV), and α6(IV), have been cloned, and their tissue distribution has been examined (18, 19, 20) . The α chains of type IV collagen contain a long collagenous domain of about 1400 amino acids and a COOH-terminal NC1 domain of about 230 amino acids. Several studies have ascribed diverse biological activities to the various domains of the type IV collagen molecule and have demonstrated the role of specific peptide sequences from both the helical and the NC1 domains on melanoma cell adhesion and spreading (21, 22, 23) .
In recent studies, we have demonstrated that ALC type IV collagen and a specific sequence comprising residues 185–203 of the NC1 domain of theα 3(IV) chain were able to prevent oxygen free radical (O2−) production and granule exocytosis in polymorphonuclear leukocytes in response to various stimuli (24 , 25) . In addition, the α3(IV) 185–203 peptide was shown to promote adhesion of melanoma cells and to inhibit tumor cell proliferation (26) . In melanoma cells, theα 3(IV) collagen chain binds to αVβ3 integrin and CD47/integrin-associated protein, which serve as membrane receptors (27) , and triggers an intracellular transduction pathway involving an increase of cytoplasmic cAMP and cAMP-dependent protein kinases (28) . Because of the crucial role of MMPs in tumor cell invasion, we investigated the effect of the α3(IV) 185–203 peptide on the activities of MMPs in tumor cells. In this paper, we present evidence that the α3(IV) 185–203 peptide is also able to inhibit tumor cell migration in an in vitro model. The amount of the inactive form of MMP-2 secreted into the medium was not altered by the α3(IV) 185–203 peptide. In contrast, the fraction of MMP-2 bound to the plasma membrane of the tumor cells was markedly decreased. The activation of this MMP-2 fraction was strongly inhibited, and this inhibition coincided with the inhibition of the expression of both MT1-MMP and β3 integrin subunit.
MATERIALS AND METHODS
Reagents.
All of the reagents, unless specifically indicated, were obtained from Sigma (St. Louis, MO). The reagents for molecular biology were from Life Technologies (Cergy Pontoise, France). Primers for PCR were synthesized by Eurogentec (Seraing, Belgium) or by Genome Express (Paris, France). Monoclonal antibody to αVβ3 integrin (clone 23C6) was from PharMingen (San Diego, CA).
Cell Cultures.
The human metastatic melanoma cell line HT-144 obtained from Dr. P. Braquet (Bioinova, France) was grown in McCoy’s 5A medium (Life Technologies, France) containing 10% FBS. The human fibrosarcoma cell line HT-1080 was obtained from the American Type Culture Collection (Rockville, MD), and human normal dermal fibroblasts, used as controls, were explanted in the laboratory. Both these cell lines were grown in DMEM supplemented with 10% FBS. All cultures were maintained at 37°C in a humidified atmosphere containing 95% air and 5% CO2. Cell proliferation was evaluated by nuclei staining with crystal violet (29) .
Preparation of Collagen Substrates.
Native EHS type IV collagen was obtained from Sigma. Native and pepsin-treated ALC type IV collagen and the NC1 domain from ALC type IV collagen were prepared as previously described (25) . Peptides corresponding to the residues 185–203 of the human NC1 domains of the α1(IV) chain (CNYYANAYSFWLATIERSE) and α3(IV) chain (CNYYSNSYSFWLASLNPER), as well as a scrambled peptide, corresponding to the α3(IV), (YAPLWNRSSFENSLNYSCY), were prepared by solid-phase synthesis (26) .
Cell Migration Assays.
Cell migration assays were performed using modified Boyden chambers containing polyvinylpyrrolidone-free polycarbonate membranes (tissue culture-treated, 6.5-mm diameter, 8-μm pore; Transwell, Costar, Cambridge, MA). Membranes were coated with different collagen IV forms (native or pepsinized ALC type IV collagen, native EHS type IV collagen, NC1 domain of ALC type IV collagen; all at 20 μg/membrane). Sterile type IV collagen solubilized in 18 mm acetic acid was deposited onto the membranes and dried under a laminar hood. Cells (105 cells/well) suspended in medium containing 0.2% BSA were deposited onto the upper surface of the membrane. The lower compartment was filled with medium supplemented with 2% BSA and 10% FBS. After a 3-h incubation period, >90% of the cells adhered to the membrane, and the medium of the upper compartment was replaced by fresh medium containing 0.2% BSA without FBS. Migration was measured at 37°C in a humidified atmosphere of 95% air and 5% CO2 for 6 or 72 h depending on the cell lines (HT-1080 or HT-144, respectively). At the end of the incubation period, the cells remaining at the upper surface of the membrane were removed with a cotton swab. The membranes were fixed with methanol and stained with crystal violet. The number of cells that migrated into the lower compartment of the chamber was estimated by measuring the absorbance at 560 nm.
Collagen IV Degradation.
Collagen IV degradation by melanoma cells was evaluated according to two different procedures to check the involvement of the adhesion process in the induction of collagenase activity: (a) pepsinized type IV collagen (25 μg/dish), which was used as a degradation substrate, was added under a soluble form into the incubation medium of HT-144 melanoma cells cultured on plastic. In addition, the different collagen IV forms used as effectors were added in the incubation medium, all at 25 μg/dish (native or pepsinized ALC type IV collagen, native EHS type IV collagen, NC1 domain of ALC type IV collagen, α3(IV) 185–203 peptide); (b) cells adhered to and were grown on pepsinized type IV collagen-coated dishes, and the different forms of type IV collagen, used as effectors, were added into the incubation medium. In both cases, melanoma cells and dermal fibroblasts were grown for 48 h in MEM containing 0.5% FBS. At the end of the incubation period, collagen IV degradation was evaluated by measuring 4-hyp-containing peptides liberated into the culture medium. The degradation products of type IV collagen were separated from the nondigested molecules by precipitation with 80% ethanol and quantified in the supernatants obtained by centrifugation at 10,000 g for 30 min at 4°C. 4-hyp was measured by a fluorometric technique after NBD-Cl derivatization as described elsewhere (30) .
Collagen IV degradation was also measured under the same experimental conditions with [3H]-labeled collagen IV as degradation substrate and small radioactive peptides liberated into the incubation medium were quantified after 80% ethanol precipitation.
Gelatinase Activity.
Gelatinase activity was determined in conditioned media or cell layers by gelatin zymography. Tumor cells were grown on type IV collagen-coated wells in FBS-free medium containing 0.1% BSA for 48 h. The preparation of conditioned media and cell extracts and the determination of their gelatinase activity were done as previously described (31) .
Evaluation of Expression of Metalloproteinase Genes: RNA Extraction and RT.
Cells were grown on collagen IV-coated dishes (native or pepsinized ALC type IV collagen, native EHS type IV collagen, NC1 domain of ALC type IV collagen; all at 25 μg/ml) or in the presence of the α1(IV) orα 3(IV) peptide (5 μg/ml) for 48 h. RNA was extracted with guanidinium/phenol/chloroform as previously described (32) . Total RNA content was measured by an A260-nm measurement, and its integrity was checked by 1.5% agarose electrophoresis.
cDNA was prepared from 5 μg of total cellular RNA by RT at 42°C for 45 min. The 100-μl reaction volume contained 200 units of murine-Moloney leukemia virus reverse transcriptase (Life Technologies, France), 2.5 μm random hexamers, 0.8 mm dATP, dCTP, dGTP, and dTTP, 2 units of RNase inhibitor (RNAsin; Promega, France), 10 mm DTT, 5 mm MgCl2, and 50 mm KCl in 20 mm Tris-HCl buffer (pH 8.4). The RT reaction product (2.5μ l) was amplified in a 25-μl PCR mixture containing 0.2μ m forward and 0.2 μm reverse primers, 200μ m dATP, dCTP, dGTP, and dTTP, 0.5 units of Taq DNA polymerase (Life Technologies, France), 1.5 mm MgCl2, and 50 mm KCl in the same buffer as above. The PCR reaction was performed in a Hybaid Omnigen thermocycler (Teddington, Middx, United Kingdom), with 32 cycles consisting each of denaturation at 95°C for 20 s, primer annealing at 55°C for 30 s, and extension at 72°C for 30 s.
Competitive PCR.
Internal standards of DNA fragments for MMPs, TIMPs, MT1-MMP, β3 integrin subunit, and GAPDH were prepared in the laboratory by generating slightly smaller products than the fragment amplified from extracted RNAs. For that purpose, composite primers were constructed as shown in Table 1⇓ . These DNA fragments were purified by a Geneclean kit (Bio 101, La Jolla, CA), quantified by A260 nm, and used in PCR experiments. The nature of the amplified fragments were confirmed by restriction enzyme digestion.
Primer pairs of internal standards (competitor) and amplification of GAPDH, MMP-2, TIMP-2, MT1-MMP and β3 integrin subunit mRNAs (target)
As a rule, series of seven dilutions of the internal standards (78 fg to 5 pg for GAPDH, 7.8 fg to 0.5 pg for MMP-2, 31 fg to 2 pg for TIMP-2, 6 fg to 0.4 pg for MT1-MMP, 6 fg to 0.4 pg for β3 integrin subunit) were performed, and these amounts of internal standards were added as competitors to a constant amount of cDNA prepared from cellular extracted RNAs. PCR products were separated by agarose gel electrophoresis and quantified by fluorometric scanning. The steady-state levels of mRNA were calculated as previously described (32) .
Evaluation of αVβ3 Integrin Expression.
The expression of the αVβ3 integrin was investigated by Western blot analysis on HT-144 melanoma cell extracts. Cells were grown on culture dishes coated with the different collagen IV substrates (native or pepsinized ALC type IV collagen, native EHS type IV collagen, NC1 domain of ALC type IV collagen; all at 25 μg/dish) or in the presence of the α1(IV) or α3(IV) peptide (5 μg/ml) for 48 h. Cell layers were washed three times with cold saline solution and lysed with 5 ml of hypotonic solution. Cell ghosts were solubilized with 100 μl of electrophoresis sample buffer and submitted to SDS-PAGE through a 10% polyacrylamide gel under reducing conditions. Proteins were transferred onto an Immobilon membrane (Millipore, Bedford, MA) and revealed with a monoclonal antibody to αVβ3 integrin (clone 23C6) and with a second antibody to mouse IgG coupled to alkaline phosphatase.
Expression of the β3 integrin subunit gene was also evaluated by competitive RT-PCR as described above.
Measurement of in Vitro Binding of MMP-2 to Plasma Membrane.
The inhibitory effect of the α3(IV) 185–203 peptide on the binding and activation of pro-MMP-2 on the melanoma cell membrane was also investigated in an in vitro model. HT-144 melanoma cells were cultured on Biosilon beads (Nunc, Copenhagen, Denmark) in McCoy’s 5A medium without FBS and in the presence of 10 μg/ml of theα 3(IV) 185–203 peptide or of its homologous α1(IV) 185–203 peptide and were used as a negative control. Membrane extracts of HT-144 cells covering the culture beads were obtained by hypo-osmotic lysis, suspended in a 50 mm Tris-HCl buffer containing 0.15 m NaCl and 5 mm CaCl2, and incubated with purified pro-MMP-2 (Calbiochem, Meudon, France; 20 ng/ml) for 2 h at 37°C. The beads were then centrifuged at 350 g for 10 min and rinsed twice with cold PBS. The MMP-2 fraction bound to the cell membrane was desorbed with electrophoresis buffer and analyzed by gelatin zymography as described above. In several assays, the membrane extracts were previously incubated for 30 min with an anti-αVβ3 integrin antibody (clone 23C6; 6 μg IgG/ml) before the addition of MMP-2.
Statistical Analyses.
Statistical significances were calculated using the Student’s t test. All experiments were done in triplicate, and data represent the mean ± 1 SD of three different series.
RESULTS
We have previously demonstrated that a peptide sequence comprising residues 185–203 of the NC1 domain of the α3(IV) chain of ALC type IV collagen supported the attachment of various melanoma cell lines and inhibited their proliferation (26) .
Inhibition of Tumor Cell Migration.
Here, using an in vitro model, we investigated the effect of ALC type IV collagen on HT-144 melanoma and HT-1080 fibrosarcoma cell migration. ALC type IV collagen induced a 50% inhibition of HT-144 melanoma cell migration, whereas EHS type IV collagen, which does not contain the α3(IV) chain, had no effect (Fig. 1⇓ ). A similar inhibition was noted with the NC1 domain. Similar results were obtained with HT-1080 cells. When tumor cells were preincubated with a synthetic α1(IV) 185–203 peptide arising from a region of theα 1 chain, similar to that of the α3(IV) chain, tumor cell migration was not inhibited. On the other hand, preincubation with the α3(IV) 185–203 peptide resulted in 50% and 30% inhibition of HT-144 and HT-1080 cell migration, respectively.
Inhibitory effect of ALC type IV collagen or of theα 3(IV) 185–203 peptide on HT-144 melanoma cells (hatched bars) or HT-1080 fibrosarcoma cell (gray bars) migration. Tumor cell migration was measured on Transwell membranes. Polyvinyl pyrrolidone-free polycarbonate membranes (8 μm porosity) were coated with the different collagen IV forms (20 μg/membrane). To test the inhibitory effect of the synthetic peptides, tumor cells were preincubated for 24 h with the α1(IV) or α3(IV) synthetic peptides (5 μg/ml) and then deposited onto membranes coated with pepsinized ALC collagen IV (20 μg/membrane). 1, control; 2, native EHS type IV collagen; 3, native ALC type IV collagen; 4, Pepsinized ALC type IV collagen; 5, NC1 domain from ALC type IV collagen; 6, α1(IV) 185–203 peptide; 7, α3(IV) 185–203 peptide. Differences from control: NS, not significant; ∗, P < 0.01; ∗∗, P < 0.001.
Inhibition of Collagen IV Degradation.
When HT-144 melanoma cells were grown on noncoated plastic dishes, they failed to degrade pepsinized type IV collagen added under a soluble form as a degradation substrate (0.012 ± 0.003 nmol 4-hyp/ml versus 0.011 ± 0.005 for a control without cells). We did not find significant differences when cells were incubated in the presence of the various collagen IV forms, added as effectors in the incubation medium. We obtained similar results by using [3H]-labeled collagen IV as a substrate (data not shown).
We further investigated the degradative potential of HT-144 adhering to pepsinized collagen IV-coated dishes in the absence (control) or in the presence of the various collagen IV effectors added in the medium. Under these experimental conditions, HT-144 cells degraded pepsinized type IV collagen (5.8 ± 0.1 nmol 4-hyp/ml versus 0.020 ± 0.005 for the control without cells). The addition of native ALC type IV collagen or its NC1 domain or the α3(IV) peptide in the medium elicited a 50% inhibition of the degradative potential of melanoma cells but had no influence on that of normal dermal fibroblasts (Fig. 2⇓ ). Similar results were obtained using radiolabeled pepsinized type IV collagen as substratum (data not shown). These results show that a close contact between collagen IV and melanoma cells is needed to induce collagen IV degradation, suggesting the involvement of a collagenolytic enzyme associated with the plasma membrane in the degradation process.
Inhibitory effect of ALC type IV collagen on collagen IV degradation by HT-144 melanoma cells (hatched bars) or normal dermal fibroblasts (gray bars). Cells were grown on pepsinized type IV collagen-coated dishes (25 μg/dish) for 48 h in MEM without FBS. Nondigested type IV collagen was removed by precipitation with 80% ethanol. The amount of 4-hyp contained in the supernatant was evaluated by a fluorometric technique after derivatization with NBD-Cl. 1, control; 2, native EHS type IV collagen; 3, native ALC type IV collagen; 4, Pepsinized ALC type IV collagen; 5, NC1 domain from ALC type IV collagen; 6, α3(IV) 185–203 peptide (all at 25 μg/dish). Differences from controls: NS, not significant; ∗, P < 0.01; ∗∗, P < 0.001.
Inhibition of the Gelatinolytic Potential of Tumor Cells.
To correlate the results obtained on collagen IV degradation by tumor cells with MMP activity, we studied the gelatinolytic activities of HT-144 melanoma cells and HT-1080 fibrosarcoma cells by gelatin zymography. Tumor cells were cultured for 48 h in the absence of FBS on the various collagen IV forms as described above. Under these various experimental culture conditions, HT-144 melanoma cells only secreted gelatinase A (MMP-2) in a latent form (Fig. 3Aa⇓ ), and MMP-2 secretion was only slightly affected by the presence of ALC collagen IV or by the α3(IV) 185–203 peptide (Fig. 3Ba⇓ ). No activation of MMP-2 was found in the incubation medium. Identical results were obtained with HT-1080 fibrosarcoma cells (Fig. 3, Ab and Bb⇓ ).
Inhibitory effect of ALC type IV collagen and of theα
3(IV) 185–203 peptide on gelatinase activity of tumor cells. HT-144
melanoma cells (a and c) and HT-1080
fibrosarcoma cells (b and d) were grown
for 48 h on type IV collagen-coated dishes (25 μg/dish) in the
absence of FBS. A, gelatinolytic activities were
evaluated by zymography of incubation media or of membrane extracts
prepared from cell layers. B, the quantification of the
zymograms expressed as a percentage of the control. 1,
plastic; 2, native EHS type IV collagen;
3, native ALC type IV collagen; 4,
pepsinized ALC type IV collagen; 5, NC1 domain from ALC
type IV collagen; 6, α3(IV) 185–203 peptide (5μ
g/ml). Differences from control significant at: ∗,
P < 0.05; ∗∗,
P < 0.01. ▪, latent form of MMP-2 (72
kDa). , active forms of MMP-2 (64 or 62 kDa).
In contrast, we found a strong decrease of the amount of MMP-2 present in the plasma membrane extracts of HT-144 cells cultured in the presence of ALC type IV collagen, its NC1 domain, or the α3(IV) 185–203 peptide (−71.2, −73.3, and −81.5% from the control, respectively; Fig. 3Ac⇓ , Lanes 3, 5, and 6). This decrease was not observed in the presence of EHS type IV collagen or pepsinized ALC type IV collagen (+2.7 and− 15.4% from the control, respectively; Fig. 3Ac,⇓ Lanes 2 and 4), as well as when cells have been incubated in the presence of the α1(IV) 185–203 peptide or the scrambled peptide corresponding to the α3(IV) 185–203 peptide (−3.7% and + 2.7%, respectively; data not shown).
In the membrane extracts from HT-144 cells, MMP-2 was present as the latent form of 72 kDa and also as the active form of 64 kDa. The activation of MMP-2 was largely decreased when the cells were cultured in the presence of either ALC type IV collagen (−79.1%; Fig. 3Ac,⇓ Lane 3), NC1 domain from ALC type IV collagen (−77.1%; Fig. 3A⇓ , Lane 5), or theα 3(IV) 185–203 peptide (−69.6%; Fig. 3Ac,⇓ Lane 6). In contrast, the presence in the medium of EHS type IV collagen or pepsinized ALC type IV collagen increased the activation of MMP-2 bound to the plasma membrane (+29.5 and +25.5%, respectively; Fig. 3Ac,⇓ Lanes 2 and 4, respectively). Similar results were obtained with HT-1080 cells, and the inhibitory effect affected both the 64-kDa and the 62-kDa active forms (Fig. 3, Ad and Bd⇓ ).
The inhibitory effect of ALC type IV collagen or of the α3(IV) 185–203 peptide seemed not to depend on variations of TIMP secretion, as measured by reverse zymography (data not shown).
Effect of ALC Type IV Collagen and α3(IV) 185–203 Peptide on MMP Gene Expression.
Fig. 4, A and B⇓ shows that native ALC type IV collagen or the α3(IV) 185–203 peptide did not elicit any significant changes in MMP-2 and TIMP-2 gene expression, respectively, as evaluated by competitive RT-PCR.
Effect of ALC type IV collagen or of the α3(IV) 185–203 peptide on the expression of MMP-2 (A), TIMP-2 (B), and MT1-MMP (C) genes. HT-144 melanoma cells were grown for 48 h on type IV collagen-coated dishes (25 μg/dish) or in the presence of the α3(IV) 185–203 peptide (5 μg/ml). MMP-2, TIMP-2, and MT1-MMP mRNAs were evaluated by competitive RT-PCR and expressed as a ratio to GAPDH mRNA. 1, control; 2, native EHS type IV collagen; 3, native ALC type IV collagen; 4, Pepsinized ALC type IV collagen; 5, NC1 domain from ALC type IV collagen; 6, Peptideα 3(IV) 185–203. Differences from control significant at: ∗, P < 0.05; ∗∗, P < 0.01.
In contrast, ALC type IV collagen or its NC1 domain or the α3(IV) 185–203 peptide triggered a strong inhibition of up to 80% in the expression of the MT1-MMP (Fig. 4C⇓ ), whereas EHS type IV collagen or pepsinized ALC type IV collagen had no significant effect compared to the control, as well as the α1(IV) 185–203 peptide (0.38 ± 0.08 versus 0.40 ± 0.07 for the control). The inhibition of the MT1-MMP gene correlated well with the decrease observed in the activation of the MMP-2 fraction bound to the plasma membrane.
Effect of ALC Type IV Collagen and α3(IV) 185–203 Peptide onβ 3 Integrin Gene Expression.
To investigate the role of the αVβ3 integrin in the binding of MMP-2 to the plasma membrane and its activation as well as its role in cell migration, we measured the effect of ALC type IV collagen and theα 3(IV) 185–203 peptide on the expression of the β3 integrin subunit gene by competitive RT-PCR with HT-144 melanoma cells (Fig. 5⇓ ). We found a significant inhibition (50%) of the expression of theβ 3 integrin subunit gene induced by ALC type IV collagen or its NC1 domain and the α3(IV) 185–203 peptide, whereas EHS type IV collagen or pepsinized ALC type IV collagen were without effect. Similar results were obtained by Northern blot analysis of mRNAs and by evaluating the expression of the β3 integrin subunit by Western blot analyses in membrane extracts of HT-144 melanoma cells (data not shown).
Inhibitory effect of ALC type IV collagen or of α3(IV) 185–203 peptide on β3 integrin subunit expression. HT-144 melanoma cells were grown for 48 h on type IV collagen-coated dishes (25μ g/ml) or in the presence of the α1 or α3(IV) 185–203 peptide (5μ g/ml). β3 subunit integrin mRNA was evaluated by competitive RT-PCR and expressed as a ratio to GAPDH mRNA. 1, control; 2, native EHS type IV collagen; 3, native ALC type IV collagen; 4, pepsinized ALC type IV collagen; 5, NC1 domain from ALC type IV collagen; 6, α3(IV) 185–203 peptide; 7, α1(IV) 185–203 peptide. Differences from control significant at: ∗, P < 0.05; ∗∗, P < 0.01.
The inhibitory effect of the α3(IV) 185–203 peptide on the binding and activation of pro-MMP-2 on the plasma membrane was also investigated in an in vitro model with membrane extracts prepared from HT-144 melanoma cells cultured on Biosilon beads. When HT-144 cells have been cultured in the presence of the α1(IV) 185–203 peptide, purified pro-MMP-2 bound to membrane extract-coated beads (Fig. 6⇓ , Lane 5). In contrast, this binding of pro-MMP-2 was decreased (−36%) when cells were cultured in the presence of theα 3(IV) 185–203 peptide (Lane 6), suggesting a decrease of potential membrane receptors (MT1-MMP or αVβ3 integrin) for MMP-2 in these extracts. It is likely that a preincubation of membrane extracts from HT-144 cells with a monoclonal antibody to αVβ3 integrin (Lane 8) also induced a large decrease in the binding of purified pro-MMP-2 (−50.7%), reproducing the result obtained with HT-144 cells cultured in the presence of the α3(IV) 185–203 peptide.
: The α3(IV) 185–203 peptide or an anti-β3 integrin subunit monoclonal antibody inhibits the binding of pro-MMP-2 on the HT-144 cell membrane. HT-144 melanoma cells were cultured on Biosilon beads as described in “Materials and Methods,” and gelatinase activities were evaluated by gelatin zymography. 1, cell membrane extract of the HT-144 cell cultured in the presence of theα 1(IV) 185–203 peptide (5 μg/ml); 2, cell membrane extract of HT-144 cells cultured in the presence of the α3(IV) 185–203 peptide (5 μg/ml); 3, pro-MMP-2 (control); 4, pro-MMP-2 incubated with the α3(IV) 185–203 peptide; 5, cell membrane extract of HT-144 cells cultured in the presence of the α1(IV) 185–203 peptide and incubated with pro-MMP-2; 6, cell membrane extract of HT-144 cells cultured in the presence of the α3(IV) 185–203 peptide (5 μg/ml) and incubated with pro-MMP-2; 7, pro-MMP-2 activated with APMA; 8, cell membrane extract of HT-144 cells cultured in the presence of the α1(IV) 185–203 peptide, preincubated with a monoclonal antibody anti-αVβ3 integrin, and then incubated with pro-MMP-2; 9, anti-αVβ3 integrin monoclonal antibody alone.
DISCUSSION
Modulation of the interactions between the cells and extracellular matrix involves the action of proteolytic enzymatic systems responsible for hydrolysis of various extracellular matrix components. The regulation of the integrity and composition of the extracellular matrix structures by these enzymatic systems controls the signals elicited by matrix molecules. In the case of tumor cells, matrix molecules modulate cell adhesion, migration, and invasion, as well as their expression of various proteinases. Basement membranes particularly regulate adhesion or migration of tumor cells and their invasive properties (21) . For example, laminin-1 promotes adhesion, spreading, and migration of tumor cells through many short peptide sequences located along its α1 chain (33 , 34) . Laminin-5 also induces adhesion of different cell types by binding α3β1 integrin (35 , 36) . Furthermore, the cleavage of the laminin-5 molecule by MMP-2 reveals a cryptic site corresponding to residues 582–593 of the γ2 chain and induces cell migration (3) . This feature appears to be specific to laminin-5, which was not known to be a substrate for MMP-2 and is not shared by type IV collagen, another basement membrane component, which also modulates cell-matrix interactions. Melanoma cells interact with different specific peptide sequences located in the triple helix or in the globular NC1 domain of the α1(IV) chain (21 , 37 , 38) . These sequences promote adhesion of tumor cells in a conformation-dependent manner or depending on a RGDT motif (21 , 23 , 39) . The contact between tumor cells and type IV collagen involves several integrin receptors,α 3β1 or αVβ3, and leads to alterations of the invasive properties of these cells, usually correlated with changes in the expression of various proteinases and MMPs, such as MMP-1 or MMP-2 (39, 40, 41) .
We have demonstrated that the peptide sequence corresponding to residues 185–203 of the α3(IV) chain of ALC type IV collagen was able to inhibit the proliferation of different melanoma, fibrosarcoma, or osteosarcoma cells. This peptide sequence contains an SNS triplet in position 189–191, which is unique to the α3(IV) chain, and the replacement of a serine residue by alanine abolishes the inhibitory activity of this peptide (26) . The intact molecule of ALC type IV collagen shares the same inhibitory activity as the α3(IV) 185–203 peptide, whereas EHS type IV collagen, which does not contain the α3(IV) chain, has no effect. The fact that the whole molecule shows the same activity as the synthetic peptide clearly suggests that a cleavage of the type IV collagen molecule by gelatinases or other proteinases is not required to reveal the inhibitory activity as it was described elsewhere for laminin-5 (3) . Now, we demonstrate that the inhibitory activity of native ALC type IV collagen also affects the invasive potential of HT-144 melanoma cells or HT-1080 fibrosarcoma cells. The α3(IV) 185–203 peptide reproduces this inhibition, whereas the α1(IV) 185–203 peptide, which does not contain the SNS 189–191 triplet, has no effect.
Several MMPs have been implicated in the invasive potential of melanoma cells and in the degradation of type IV collagen by these cells (42 , 43) . The inhibition of type IV collagen degradation by HT-144 melanoma cells under the influence of ALC type IV collagen suggests alterations in the secretion or the expression of MMPs by these cells. This degradative process depends on a cell membrane-bound fraction of MMP-2, and ALC type IV collagen induces a strong inhibition of both the amount and the activation of this fraction. Membrane-bound MMP-2 has been shown to play an important role in tumor invasion. Pro-MMP-2 binds to the cell membrane through a complex comprising TIMP-2 and MT1-MMP, in which TIMP-2, via its COOH-terminal domain, binds to the catalytic domain of MT1-MMP (44, 45, 46) . The activation of membrane-bound MMP-2 depends on MT1-MMP and is regulated by the level of TIMP-2 synthesis (47, 48, 49) . We did not find significant changes in MMP-2 and TIMP-2 secretion into the medium under the influence of ALC type IV collagen or the α3(IV) 185–203 peptide. In contrast, these effectors induced a large decrease of the expression of the MT1-MMP gene, whereas EHS type IV collagen had no effect. This result could explain the significant decrease in the amount of the membrane-bound MMP-2 and the decrease in its activation in the presence of the peptide. Another membrane receptor, the αVβ3 integrin, has been implicated in the binding of MMP-2 to the melanoma cell membrane (13) . This integrin is a membrane receptor for extracellular matrix proteins, such as vitronectin, fibronectin, or type IV collagen (50) , but might also serve as a MMP-2 receptor in parallel with the complex MT1-MMP-TIMP-2 (48) . ALC type IV collagen or the α3(IV) 185–203 peptide triggers a strong inhibition of the expression of the β3 integrin subunit, leading to a decrease in the MMP-2 amount bound to the cell membrane. In addition, the preincubation of a cell membrane extract from HT-144 melanoma cells with a monoclonal antibody to αVβ3 integrin also prevents the binding of pro-MMP-2 to cell membrane extract-coated beads and provides results very similar to that obtained with intact HT-144 melanoma cells incubated with the α3(IV) 185–203 peptide. Furthermore, the αVβ3 integrin has also been implicated in the tumor progression and cell migration of various melanoma cells and the tumorigenicity of melanoma cells correlated with elevated levels of αVβ3 integrin expression by these cells (10 , 12) . Our results suggest that ALC type IV collagen or the α3(IV) 185–203 peptide could regulate melanoma cell migration by decreasing the expression of the β3 integrin subunit.
For invading tissues, tumor cells could regulate the expression of MMPs and their activators as well as the expression of various integrins through cell-cell interactions (51) . In this paper, we provide evidence that tumor cells could also regulate the expression of these enzymatic complexes by cell-matrix interactions when they cross basement membranes and that the α3(IV) collagen chain may play a role in down-regulating the invasive properties of melanoma cells. In a previous paper, we have shown that newly synthesized α3(IV) chains were expressed in a large number of bronchogenic and alveolar tumors, whereas they were undetected in normal bronchi. The deposition of type IV collagen has always been associated with a protective role and correlates with a good prognosis in squamous cell carcinomas and in peripheral adenocarcinomas in the lung. These observations have suggested that the deposition of type IV collagen containing theα 3(IV) chain might reflect a potentially beneficial reaction of the host to the neoplasm (52) . Thus, the presence of theα 3(IV) chain within a basement membrane might increase its resistance against degradation by tumor cells and tumor invasion.
ACKNOWLEDGMENTS
We thank C. Perreau and M. Decarme for their technical assistance and Dr. W. Hornebeck and H. Emonard for helpful discussion.
Footnotes
-
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
-
↵1 Supported by grants from the University of Reims Champagne-Ardenne, CNRS (UPRESA 6021), the Ligue contre le Cancer, and the NIH (AR 20553, HL 29492, and AR 07490) and by a NATO Collaborative Research Grant.
-
↵2 To whom requests for reprints should be addressed, at Lab. Biochemistry, CNRS UPRESA 6021, UFR Medicine, 51 Rue Cognacq Jay, F-51095, Reims Cedex, France. Phone: (33)326913534; Fax: (33)326918055; E-mail: jc.monboisse{at}univ-reims.fr
-
↵3 The abbreviations used are: MMP, matrix metalloproteinase; MT-MMP, membrane-type MMPs; TIMP, tissue metalloproteinase inhibitor; NC1, noncollagenous; ALC, anterior lens capsule; FBS, fetal bovine serum; 4-hyp, 4-hydroxyproline; RT, reverse transcription; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; pg, picogram.
- Received June 7, 1999.
- Accepted November 12, 1999.
- ©2000 American Association for Cancer Research.