Regulation of Membrane-type-1 Matrix Metalloproteinase Activity by Its Cytoplasmic Domain

Membrane-type-1 matrix metalloproteinase (MT1-MMP) has transmembrane and cytoplasmic domains, which target it to invasive fronts. We analyzed the role of the cytoplasmic tail by expressing wild type MT1-MMP and three mutants with progressively truncated C termini in human Bowes melanoma cells. We examined gelatinase A activation and the localization and processing of recombinant proteins in stable cell clones using gelatin zymography, immunoblotting, and immunofluorescence. Cell invasion was analyzedin vitro by Matrigel invasion assays. Gelatinase A was activated in all cell clones. However, the localization of MT1-MMP to the leading edge of migrating cells and cell invasion through Matrigel were strongly enhanced only in cells expressing either wild type or truncated MT1-MMP lacking 6 C-terminal amino acid residues (Δ577). Truncations of 10 or 16 amino acid residues in the cytoplasmic domain (Δ567 and Δ573, respectively) disturbed MT1-MMP localization. The expression of wild type and Δ577 MT1-MMPs induced also their cleavage to 43-kDa cell surface forms and the release of soluble, ∼20-kDa N-terminal fragments containing the catalytic center. A synthetic MMP inhibitor but not a gelatinase inhibitor prevented the processing, suggesting that autocatalytic cleavage occurs. Purified soluble MT1-MMP was also autoproteolytically processed to 43- and 20-kDa forms in vitro. Our results indicate that the cytoplasmic domain has an important role in cell invasion by controlling both the targeting and degradation/turnover of MT1-MMP. Membrane-type-1 matrix metalloproteinase (MT1-MMP) has transmembrane and cytoplasmic domains, which target it to invasive fronts. We analyzed the role of the cytoplasmic tail by expressing wild type MT1-MMP and three mutants with progressively truncated C termini in human Bowes melanoma cells. We examined gelatinase A activation and the localization and processing of recombinant proteins in stable cell clones using gelatin zymography, immunoblotting, and immunofluorescence. Cell invasion was analyzedin vitro by Matrigel invasion assays. Gelatinase A was activated in all cell clones. However, the localization of MT1-MMP to the leading edge of migrating cells and cell invasion through Matrigel were strongly enhanced only in cells expressing either wild type or truncated MT1-MMP lacking 6 C-terminal amino acid residues (Δ577). Truncations of 10 or 16 amino acid residues in the cytoplasmic domain (Δ567 and Δ573, respectively) disturbed MT1-MMP localization. The expression of wild type and Δ577 MT1-MMPs induced also their cleavage to 43-kDa cell surface forms and the release of soluble, ∼20-kDa N-terminal fragments containing the catalytic center. A synthetic MMP inhibitor but not a gelatinase inhibitor prevented the processing, suggesting that autocatalytic cleavage occurs. Purified soluble MT1-MMP was also autoproteolytically processed to 43- and 20-kDa forms in vitro. Our results indicate that the cytoplasmic domain has an important role in cell invasion by controlling both the targeting and degradation/turnover of MT1-MMP. extracellular matrix matrix metalloproteinase membrane-type-1 matrix metalloproteinase tissue inhibitor of metalloproteinases, types 1 and 2 Eagle's minimal essential medium polymerase chain reaction base pair(s) polyacrylamide gel electrophoresis phosphate-buffered saline wild type The three-step model for tumor cell invasion describes the critical process of cancer metastasis in terms of cell attachment to the extracellular matrix (ECM),1 localized degradation of the ECM, and cell migration through the digested barrier (1.Liotta L.A. Stetler-Stevenson W.G. Cancer Res. 1991; 51: 5054s-5059sPubMed Google Scholar). A critical balance of protease activation and inhibition at specific cell surface sites is essential for coordinated proteolysis required for cell attachment and invasion. Matrix metalloproteinases (MMPs) and serine proteases are involved in the proteolytic cascades leading to ECM degradation. Recent evidence suggests that MMPs are important also in creating an environment that supports the growth of tumors (2.Chambers A.F. Matrisian L.M. J. Natl. Cancer Inst. 1997; 89: 1260-1270Crossref PubMed Scopus (1427) Google Scholar). MMPs are Zn2+-dependent endopeptidases, which share a common domain structure. They are produced as inactive proenzymes, and their activation by the proteolytic removal of the N-terminal prodomain is an important control step in ECM degradation. Activated MMPs are inhibited by tissue inhibitors of metalloproteinases (TIMPs). MMPs are expressed in many tumor tissues at higher levels than in normal tissues, and their expression is tightly regulated by growth factors and cytokines (2.Chambers A.F. Matrisian L.M. J. Natl. Cancer Inst. 1997; 89: 1260-1270Crossref PubMed Scopus (1427) Google Scholar). Membrane-type-1 matrix metalloproteinase (MT1-MMP) is one of the six currently known MT-MMPs (3.Sato H. Takino T. Okada Y. Cao J. Shinagawa A. Yamamoto E. Seiki M. Nature. 1994; 370: 61-65Crossref PubMed Scopus (2367) Google Scholar, 4.Will H. Hinzmann B. Eur. J. Biochem. 1995; 231: 602-608Crossref PubMed Scopus (317) Google Scholar, 5.Takino T. Sato H. Shinagawa A. Seiki M. J. Biol. Chem. 1995; 270: 23013-23020Abstract Full Text Full Text PDF PubMed Scopus (447) Google Scholar, 6.Puente X.S. Pendas A.M. Llano E. Velasco G. Lopez-Otin C. Cancer Res. 1996; 56: 944-949PubMed Google Scholar, 7.Pei D. J. Biol. Chem. 1999; 274: 8925-8932Abstract Full Text Full Text PDF PubMed Scopus (228) Google Scholar, 8.Pei D. Cell Res. 1999; 9: 291-303Crossref PubMed Scopus (169) Google Scholar). It is expressed in association with gelatinase A (MMP-2) in various human malignant tumors, including colon, breast, gastric, head and neck, lung, and cervical carcinomas (9.Seiki M. APMIS. 1999; 107: 137-143Crossref PubMed Scopus (274) Google Scholar). MT-MMPs differ from all other MMPs in that they have hydrophophic transmembrane domains and short cytoplasmic tails. In addition, they contain recognition sites for the furin-like proprotein convertases between their pro and catalytic domains. Therefore, they appear to be constitutively activated by these proteases (10.Pei D. Weiss S.J. J. Biol. Chem. 1996; 271: 9135-9140Abstract Full Text Full Text PDF PubMed Scopus (364) Google Scholar). The structure of the human MT1-MMP/MMP-14 gene differs from the otherMMP genes in exon organization, suggesting a relatively early divergence in evolution (11.Lohi J. Lehti K. Valtanen H. Parks W.C. Keski-Oja J. Gene. 2000; 242: 75-86Crossref PubMed Scopus (115) Google Scholar). Until recently, the primary function assigned to MT-MMPs has been the binding and processing of secreted progelatinase A to its active form. The activation depends on the formation of a ternary complex between MT1-MMP, TIMP-2, and gelatinase A and initial cleavage of gelatinase A prodomain by another MT1-MMP molecule (12.Atkinson S.J. Crabbe T. Cowell S. Ward R.V. Butler M.J. Sato H. Seiki M. Reynolds J.J. Murphy G. J. Biol. Chem. 1995; 270: 30479-30485Abstract Full Text Full Text PDF PubMed Scopus (227) Google Scholar, 13.Imai K. Ohuchi E. Aoki T. Nomura H. Fujii Y. Sato H. Seiki M. Okada Y. Cancer Res. 1996; 56: 2707-2710PubMed Google Scholar, 14.Strongin A.Y. Collier I. Bannikov G. Marmer B.L. Grant G.A. Goldberg G.I. J. Biol. Chem. 1995; 270: 5331-5338Abstract Full Text Full Text PDF PubMed Scopus (1435) Google Scholar, 15.Butler G.S. Butler M.J. Atkinson S.J. Will H. Tamura T. van Westrum S.S. Crabbe T. Clements J. d'Ortho M.P. Murphy G. J. Biol. Chem. 1998; 273: 871-880Abstract Full Text Full Text PDF PubMed Scopus (539) Google Scholar). MT1-MMP can also activate procollagenase-3 (MMP-13) (16.Knauper V. Will H. Lopez-Otin C. Smith B. Atkinson S.J. Stanton H. Hembry R.M. Murphy G. J. Biol. Chem. 1996; 271: 17124-17131Abstract Full Text Full Text PDF PubMed Scopus (618) Google Scholar). MT1-MMP expression levels in tumor tissues correlate with gelatinase A activation levels, and the malignancy and invasiveness of the tumor, supporting the proposed function as gelatinase A activator (17.Ueno H. Nakamura H. Inoue M. Imai K. Noguchi M. Sato H. Seiki M. Okada Y. Cancer Res. 1997; 57: 2055-2060PubMed Google Scholar). MT1-MMP is abundantly expressed by highly invasive melanoma cells in vivo, whereas gelatinase A and TIMP-2 are detected in stromal cells bordering tumors (18.Airola K. Karonen T. Vaalamo M. Lehti K. Lohi J. Kariniemi A.L. Keski-Oja J. Saarialho-Kere U.K. Br. J. Cancer. 1999; 80: 733-743Crossref PubMed Scopus (205) Google Scholar). MT1-MMP itself is also a potent matrix-degrading protease with broad collagenolytic, glycoproteolytic, and gelatinolytic activities (19.Ohuchi E. Imai K. Fujii Y. Sato H. Seiki M. Okada Y. J. Biol. Chem. 1997; 272: 2446-2451Abstract Full Text Full Text PDF PubMed Scopus (829) Google Scholar, 20.d'Ortho M.P. Will H. Atkinson S. Butler G. Messent A. Gavrilovic J. Smith B. Timpl R. Zardi L. Murphy G. Eur. J. Biochem. 1997; 250: 751-757Crossref PubMed Scopus (383) Google Scholar). In addition, it is a very efficient fibrinolytic protease (21.Hiraoka N. Allen E. Apel I.J. Gyetko M.R. Weiss S.J. Cell. 1998; 95: 365-377Abstract Full Text Full Text PDF PubMed Scopus (642) Google Scholar), and collagen turnover in MT1-MMP-deficient mice was severely compromised (22.Holmbeck K. Bianco P. Caterina J. Yamada S. Kromer M. Kuznetsov S.A. Mankani M. Robey P.G. Poole A.R. Pidoux I. Ward J.M. Birkedal-Hansen H. Cell. 1999; 99: 81-92Abstract Full Text Full Text PDF PubMed Scopus (1102) Google Scholar). MT1-MMP appears to have an important role in promoting cell migration through the ECM. MT1-MMP overexpression in HT-1080 cells enhances the rate of cell migration through a Matrigel (Becton Dickinson, Mountain View, CA) barrier (3.Sato H. Takino T. Okada Y. Cao J. Shinagawa A. Yamamoto E. Seiki M. Nature. 1994; 370: 61-65Crossref PubMed Scopus (2367) Google Scholar). In addition, MT1-MMP overexpression in melanoma cells increases the degradation rate of fibronectin-coated, cross-linked gelatin films (23.Nakahara H. Howard L. Thompson E.W. Sato H. Seiki M. Yeh Y. Chen W.T. Proc. Natl. Acad. Sci. U. S. A. 1997; 94: 7959-7964Crossref PubMed Scopus (361) Google Scholar). The cytoplasmic and transmembrane domains of MT1-MMP are important for the localization of MT1-MMP to the sites of matrix degradation and for the cellular invasion. Active gelatinase A is partially localized to the same sites, thus generating a restricted pericellular environment for ECM degradation (23.Nakahara H. Howard L. Thompson E.W. Sato H. Seiki M. Yeh Y. Chen W.T. Proc. Natl. Acad. Sci. U. S. A. 1997; 94: 7959-7964Crossref PubMed Scopus (361) Google Scholar). During gelatinase A activation MT1-MMP is processed by an MMP-dependent mechanism to its 43-kDa form, which lacks most of the catalytic domain (24.Lehti K. Lohi J. Valtanen H. Keski-Oja J. Biochem. J. 1998; 334: 345-353Crossref PubMed Scopus (201) Google Scholar, 25.Lohi J. Lehti K. Westermarck J. Kähäri V.M. Keski-Oja J. Eur. J. Biochem. 1996; 239: 239-247Crossref PubMed Scopus (180) Google Scholar). The present study was carried out to characterize the role of the cytoplasmic domain for targeting MT1-MMP activities using human Bowes melanoma cells and site-directed mutagenesis. We found that the middle part of the MT1-MMP cytoplasmic domain was essential for the correct localization of MT1-MMP to the leading edges of migrating cells. Focusing of MT1-MMP activity enhanced the invasive potential of Bowes melanoma cells and caused an autocatalytic cleavage of MT1-MMP to cell-bound 43-kDa and soluble 20-kDa fragments. Thermostable AmpliTag polymerase was from Perkin-Elmer (Branschburg, NJ), and Pfu polymerase was from Stratagene (La Jolla, CA). All the DNA-modifying enzymes were purchased from New England Biolabs (Beverly, MA). pcDNA3 eukaryotic expression vector was from InVitrogen (Oxford, UK). Reagents for dideoxy sequencing were from Amersham Pharmacia Biotech. Cyclic peptide inhibitor CTTHWGFTLC, which inhibits specifically gelatinases A and B (26.Koivunen E. Arap W. Valtanen H. Rainisalo A. Medina O.P. Heikkilä P. Kantor C. Gahmberg C.G. Salo T. Konttinen Y.T. Sorsa T. Ruoslahti E. Pasqualini R. Nat. Biotechnol. 1999; 17: 768-774Crossref PubMed Scopus (513) Google Scholar), was obtained from Dr. E. Koivunen (Department of Biochemistry, University of Helsinki). The general inhibitor of metalloproteinase activity, BB-3103, was a generous gift from British Biotech Pharmaceuticals Ltd. (Oxford, UK). Human Bowes melanoma cells were selected for the invasion and activity studies in the MT1-MMP overexpression system, because they have low endogenous expression levels of MT1-MMP, TIMPs, and other MMPs as compared with several other melanoma cells and they produce low levels of gelatinase A, which remains latent even in the presence of phorbol 12-myristate 13-acetate (18.Airola K. Karonen T. Vaalamo M. Lehti K. Lohi J. Kariniemi A.L. Keski-Oja J. Saarialho-Kere U.K. Br. J. Cancer. 1999; 80: 733-743Crossref PubMed Scopus (205) Google Scholar). Cells were cultivated in Eagle's minimal essential medium (MEM) containing 10% heat-inactivated fetal calf serum (Life Technologies, Inc., Gaithersburg, MD), 100 IU/ml penicillin, and 50 μg/ml streptomycin. The cultures were incubated at 37 °C in humidified 5% CO2 atmosphere until confluency. All experiments were carried out in serum-free conditions. For the collection of conditioned medium and cell lysates, the cultures were rinsed twice with serum-free medium, incubated under serum-free conditions for 6 h, and rinsed again with serum-free medium. The conditioned medium was then collected after 24 h. The cells were lysed on ice with Triton lysis buffer (50 mm Tris-HCl buffer, pH 8.0, containing 150 mm NaCl, 1% Triton X-100, 0.02% sodium azide, 10 mm EDTA, 10 μg/ml aprotinin, 1 μg/ml pepstatin A, and 1 μg/ml aminoethylbenzene sulfonyl fluoride; Calbiochem, San Diego, CA). The lysates were clarified by centrifugation. The cloning of bases 1–2369 of MT-MMP-1 cDNA as an EcoRI fragment to pcDNA3 eukaryotic expression vector (MTpc3SE) has been described (25.Lohi J. Lehti K. Westermarck J. Kähäri V.M. Keski-Oja J. Eur. J. Biochem. 1996; 239: 239-247Crossref PubMed Scopus (180) Google Scholar). Segments of MT1-MMP cDNA (bases 1418–1820, 1418–1835, and 1418–1848) were amplified using polymerase chain reaction (PCR) with oligonucleotides containing a TGA-stop codon and the XbaI restriction endonuclease recognition site at the 3′-end. Subsequently, the fragments were cloned as BsmBI-XbaI fragments to MTpc3SE, thus generating the constructsΔ577, Δ573, and Δ567. Thewt construct containing the whole open reading frame of MT1-MMP was made by amplifying by PCR a segment containing bases 1418–1906 of MT1-MMP cDNA and a XbaI recognition site at the 3′-end and cloning it as a BsmBI-XbaI fragment to MTpc3SE. The difference between constructE240A and wt is that the codon GAG coding for E240 was mutated to the alanine codon GCG by PCR-mediated mutagenesis using two partially overlapping oligonucleotides. This point mutation inactivates the enzyme (27.Valtanen, H., Lehti, K., Lohi, J., and Keski-Oja, J. (2000) Protein Expr. Purif., in pressGoogle Scholar). For the ΔTM construct, a segment containing bases 1482–1733 was amplified by PCR with oligonucleotides containing a TGA-stop codon, six additional histidine codons (CAC), and the EcoRI restriction endonuclease recognition site at the 3′-end. The PCR segment was digested withBamHI and ligated with a 1490-base pair (bp)EcoRI-BamHI fragment of MTpc3SE. The generated EcoRI fragment was then cloned to theMTpc3SE. The constructed deletion and point mutations are illustrated in Fig. 1. All the constructs were verified by automated sequencing using ABI 310 (Perkin-Elmer, Norwalk, CT). Cells were cultured in 6-well plates to reach 50–80% confluence and transfected using 2 μg of plasmid DNA and 5 μl of FuGENE 6 transfection reagent for each transfection according to the manufacturer's instructions (Roche Molecular Biochemicals, Mannheim, Germany). Transfection mixture was applied onto cells in the presence of serum and antibiotics, and the incubation was carried out for 6 h. The cells were then washed twice with serum-free MEM. After 48 h the cells were fixed for immunofluorescence staining, or the selection for stable transfectants was started with 1.2 mg/ml G418 (Life Technologies, Inc.) in complete medium. Expression of the recombinant MT1-MMPs by transfected cell clones was analyzed by Northern hybridization and immunoblotting. Cell clones producing about 10-fold excess of wild type or mutated MT1-MMP as compared with endogenous expression level of MT1-MMP were selected for the experiments. Polyclonal antibodies to MT1-MMP were raised in rabbits. The immunogens used were (illustrated in Fig. 1): a bacterially generated recombinant fusion protein of Schistosoma japonicum glutathione S-transferase and amino acid residues 260–484 of MT1-MMP (Ab-1), a synthetic peptide of 26 amino acid residues corresponding to the C-terminal intracellular domain of MT1-MMP (Ab-2), and a synthetic peptide of 24 amino acid residues corresponding to amino acids 154–167 of MT1-MMP (Ab-3). The antibodies were affinity-purified using the respective immunogens coupled to Sepharose 4B (25.Lohi J. Lehti K. Westermarck J. Kähäri V.M. Keski-Oja J. Eur. J. Biochem. 1996; 239: 239-247Crossref PubMed Scopus (180) Google Scholar). SDS-polyacrylamide gel electrophoresis (PAGE) was carried out using 4–15% gradient or 10% standard Laemmli polyacrylamide gels (Bio-Rad, Richmond, VA). Proteins were electrophoretically transferred to nitrocellulose membranes (Gelman Sciences, Ann Arbor, MI). Immunodetection of the proteins was performed as described (25.Lohi J. Lehti K. Westermarck J. Kähäri V.M. Keski-Oja J. Eur. J. Biochem. 1996; 239: 239-247Crossref PubMed Scopus (180) Google Scholar). To analyze the gelatinolytic proteins, aliquots of cell-conditioned medium were analyzed by zymography essentially as described previously (28.Lohi J. Harvima I. Keski-Oja J. J. Cell. Biochem. 1992; 50: 337-349Crossref PubMed Scopus (150) Google Scholar). Bowes cells were cultured to ∼80% confluence and transfected with MT1-MMP constructs as described, or the cell clones were cultured to 100% or ∼20% confluence on glass coverslips. For the staining of migrating cells, a gel containing serum-free MEM and 0.5% low melting point agar was poured above the cells. Chemoattractant gradient was generated to the solid agar gel by adding 1 ng of fibroblast growth factor-2 or 5 μl of fetal calf serum to one side of the coverslips (29.Alanko T. Tienari J. Lehtonen E. Saksela O. Dev. Biol. 1994; 161: 141-153Crossref PubMed Scopus (9) Google Scholar). After 24 h the gel was removed by suction, and the cells were washed with serum-free MEM and fixed with 3% paraformaldehyde. For intracellular staining, the cells were permeabilized with 0.1% Triton X-100 in phosphate-buffered saline (PBS) for 20 min. Subsequently, the coverslips were treated with 5% bovine serum albumin in PBS for 30 min, washed twice with PBS containing 0.5% bovine serum albumin and incubated with affinity-purified polyclonal Ab-1 (see Fig. 1) for 1 h. The coverslips were then washed three times and incubated with fluorescein isothiocyanate-conjugated anti-rabbit antibodies (Jackson ImmunoResearch Laboratories, West Grove, PA) for 1 h. Next, the coverslips were washed four times and mounted on glass slides using Vectashield (Vector Laboratories, Burlingame, CA). The fluorescence images were obtained by using an epifluorescence microscope or a Bio-Rad MRC-1024 confocal microscope. The Boyden chamber cell invasion assay was carried out according to manufacturer's instructions (Becton Dickinson). Bowes control cells or stably transfected Bowes cells expressing wild type or mutated MT1-MMP proteins (1 × 105 cells/ml) were suspended in MEM containing 0.1% bovine serum albumin and seeded onto uncoated or Matrigel-coated Transwell filters (8-μm pore size) in Biocoat Matrigel invasion chambers. Proteinase inhibitors aprotinin (100 μg/ml), BB-3103 (10 μm), or peptide inhibitor CTTHWGFTLC (500 nm) were added to the suspension of control cells or cells expressing wt MT1-MMP, and the cells were seeded onto the filters after 0.5 h of incubation. Fetal calf serum (5%) was used as a chemoattractant in the lower chambers. After 18 h of incubation, noninvading cells were removed from the top surface of the filter, and the filters were stained with Diff-Quik (Dade AG, Dudingen, Switzerland). Cells migrated onto the lower surface of the filter were counted under a microscope at × 40 or × 200 magnification. Each assay was carried out in triplicate, and five fields were counted. Recombinant soluble forms of wild type and inactive MT1-MMP (ΔTM and E240AΔTM, respectively) were expressed in a baculovirus/insect cell system and purified by cation exchange and gel filtration chromatography (27.Valtanen, H., Lehti, K., Lohi, J., and Keski-Oja, J. (2000) Protein Expr. Purif., in pressGoogle Scholar). ΔTM and E240AΔTM (10 ng) were incubated alone or with proteinase inhibitors as indicated in 50 mm Tris-HCl buffer, pH 7.5, containing 0.15 m NaCl, 10 mmCaCl2, 0.05% Brij 35, and 0.02% NaN3 at 37 °C for 3 h. The reactions were terminated with reducing Laemmli sample buffer, and the MT1-MMP fragments were separated by SDS-PAGE and characterized by immunoblotting. To examine the effects of the cytoplasmic domain of MT1-MMP on its own protein expression, gelatinase A activation, and cell invasion, we prepared three MT1-MMP constructs encoding proteins with deletions in the cytoplasmic domain. The C-terminal fragments containing amino acids from Ser577, Tyr 573, or Thr567 to Val582 were eliminated from the mutated MT1-MMP proteins designated as Δ577, Δ573, and Δ567, respectively (Fig.1). Wild type MT1-MMP (wt); soluble truncated MT1-MMP, lacking the transmembrane and cytoplasmic domains (ΔTM); and inactive MT1-MMP, containing point mutation E240A, were used as controls. The constructs were transfected to human Bowes melanoma cells, and stable cell clones were generated by G418 selection and dilution cloning. Positive cell clones were selected by immunoblotting. Two separate cell clones expressing each of the deletion mutants Δ577, Δ573, and Δ567, and one cell clone expressing each control protein, were used in the experiments. Northern blotting indicated that all clones had 8- to 10-fold overexpression ofwt or mutated MT1-MMP mRNA, as compared with endogenousMT1-MMP expression levels (data not shown). Northern hybridization was used also to detect possible changes in the expression levels of mRNAs for TIMP-2, αV and β3 integrins, urokinase-type plasminogen activator and urokinase-type plasminogen activator receptor by the cell clones. No significant differences, which could affect gelatinase A activity and cell invasion, were found. Bowes cells did not express detectable levels of mRNA for MMP-13, whose activation could also be affected by MT1-MMP (data not shown). By immunoblotting, using polyclonal Ab-3 against the N-terminal part of the MT1-MMP catalytic domain, MT1-MMP was detected as 60- and 63-kDa protein bands corresponding to its active and proenzyme forms, respectively (Fig. 2; characterized in detail in Ref. 24.Lehti K. Lohi J. Valtanen H. Keski-Oja J. Biochem. J. 1998; 334: 345-353Crossref PubMed Scopus (201) Google Scholar). Increased levels of the 60- and 63-kDa forms were detected in all transfected cells. Expression of Δ567 caused the appearance of an additional immunoreactive band of ∼58 kDa and occasionally an additional band of 35–40 kDa. The identities of these bands were not analyzed further. Overexpression of MT1-MMP induces gelatinase A activation (3.Sato H. Takino T. Okada Y. Cao J. Shinagawa A. Yamamoto E. Seiki M. Nature. 1994; 370: 61-65Crossref PubMed Scopus (2367) Google Scholar). To examine the effects of the cytoplasmic domain deletions of MT1-MMP on gelatinase A activation, aliquots of the conditioned media of Bowes cell clones were analyzed by gelatin zymography. Nontransfected cells as well as cloned cells expressing wt, E240A, and ΔTM were used as controls. Gelatinase A was processed to its activated 64- and 62-kDa forms by all cell clones expressing wild type MT1-MMP or mutants containing a deletion in the cytoplasmic domain (Δ577, Δ573, and Δ567) (Fig. 2). Cells expressing either soluble ΔTM or the inactive E240A were unable to activate gelatinase A. However, low levels of the intermediate forms were observable also in these cells. To confirm the specificities of the antibodies in immunofluorescence staining and to analyze the localization of the proteins encoded by the wt,ΔTM, E240A, Δ577,Δ573, and Δ567 constructs, we first carried out transient transfections of Bowes cells cultured on glass coverslips. The cells were fixed and subjected to immunofluorescence staining using affinity-purified Ab-1 against hinge/pexin-like domains of MT1-MMP. Comparison of the phase contrast and fluorescence images indicated that 5–10% of the cells were positive when stained with the MT1-MMP antibodies (Fig. 3 A). The endogenous expression of MT1-MMP was not sufficient to generate specific detectable staining in the rest of the cells. In cells transfected with either wt, E240A,Δ577, Δ573, or Δ567 plasmids, the protein products were localized to the cell surface as found inwt transfectants (Fig. 3 A), whereas in cells transfected with ΔTM only intracellular staining was observed (Fig. 3 A). wt, E240A, Δ577, Δ573, and Δ567 proteins were also accessible to cell surface biotinylation (data not shown) as observed earlier for wild type MT1-MMP in HT-1080 cells (24.Lehti K. Lohi J. Valtanen H. Keski-Oja J. Biochem. J. 1998; 334: 345-353Crossref PubMed Scopus (201) Google Scholar). The cell surface localization of MT1-MMP was examined next by immunofluorescence staining of cloned Bowes cells expressing the wt, E240A, Δ577, Δ573, and Δ567 forms of MT1-MMP. No specific staining was detected in nontransfected control cells (Fig.3 B). In confluent cell cultures all the recombinant MT1-MMPs containing transmembrane domains were spread around the plasma membranes similarly with wt (Fig. 3 B). In sparse cultures, part of the cells had a random migrating phenotype, and in those cells the full length recombinant MT1-MMPs (wt and E240A) were concentrated to special membrane sites as those seen in Bowes cell clone expressing wt (Fig. 3 B). To characterize the effects of cytoplasmic deletions on the MT1-MMP localization in migrating cells, we cultured the cells in a chemotactic gradient and analyzed the MT1-MMP staining with confocal microscopy. Optical section images focused from the matrix contact to the top of the cells indicated that the wild type MT1-MMP as well as the Δ577 proteins were concentrated at the leading edges of the cells at the substratum contact level (Fig.4). On the contrary, the shorter forms Δ573 and Δ567 were distributed more equally around the plasma membranes from the bottom to the top of the cells. To analyze the effects of expression of wild type MT1-MMP or the mutants on the ability of Bowes melanoma cells to invade through basement membranes, we carried out Boyden chamber cell invasion assays. Cell invasionin vitro through basement membrane matrices was strongly enhanced by the expression of either the wild type MT1-MMP or Δ577 lacking 6 C-terminal amino acids (Fig.5 A). Invasion was also appreciably enhanced by the expression of truncated MT1-MMP lacking 10 C-terminal amino acids (Δ573). On the contrary, the expression of Δ567 lacking 16 C-terminal amino acids, soluble truncated (ΔTM), or the inactive MT1-MMP (E240A) did not significantly enhance cell invasion. We analyzed next the inhibition of invasion by three proteinase inhibitors: a broad spectrum MMP inhibitor BB-3103; the cyclic peptide CTTHWGFTLC (26.Koivunen E. Arap W. Valtanen H. Rainisalo A. Medina O.P. Heikkilä P. Kantor C. Gahmberg C.G. Salo T. Konttinen Y.T. Sorsa T. Ruoslahti E. Pasqualini R. Nat. Biotechnol. 1999; 17: 768-774Crossref PubMed Scopus (513) Google Scholar) that is inhibitory for gelatinase A and B activities, but does not inhibit MT1-MMP; and aprotinin. The invasion of control cells was only moderately affected by the inhibitors (Fig.5 B). In contrast, the invasion of cells expressing wt MT1-MMP was decreased to the level of control cells by BB-3103. The cyclic peptide CTTHWGFTLC inhibited the invasion of MT1-MMP overexpressing cells by about 56%. However, there was still a clear difference in the invasion rate as compared with the control cells. Aprotinin had only a moderate inhibitory effect. Gelatinase A activation in fibrosarcoma cells is accompanied by the N-terminal processing of MT1-MMP to its inactive 43-kDa form by an MMP-dependent mechanism (24.Lehti K. Lohi J. Valtanen H. Keski-Oja J. Biochem. J. 1998; 334: 345-353Crossref PubMed Scopus (201) Google Scholar, 25.Lohi J. Lehti K. Westermarck J. Kähäri V.M. Keski-Oja J. Eur. J. Biochem. 1996; 239: 239-247Crossref PubMed Scopus (180) Google Scholar). Accordingly, immunoblotting analyses with Ab-1 against hinge/pexin-like domains of MT1-MMP revealed the appearance of the 43-kDa form also in Bowes melanoma cells (Fig.6 A). The levels of the 60-kDa active and 63-kDa latent forms of MT1-MMP in cell lysates correlated with the detection of these forms with Ab-3 (Fig. 2). Expression of wt MT1-MMP and Δ577, and to some extent Δ573, also induced the appearance of 43-kDa form. In cells expressing Δ567 with the shortest cytoplasmic tail or the E240A inactive mutant, the 43-kDa form was not observed. The level of the 43-kDa form was notably lower in the cells expressing Δ573 than in cells expressing Δ577 or wt MT1-MMP. To further characterize the cleavage cascade and to detect the cleaved N-terminal f