Advances in Brief Inhibition of Furin-mediated Processing Results in Suppression of Astrocytoma Cell Growth and Invasiveness

Purpose: Astrocytoma arises in the central nervous system as a tumor of great lethality, in part because of the invasive potential of the neoplastic cells that are able to release extracellular matrix-degrading enzymes. Furin convertase activates several precursor matrix metalloproteases involved in the breakdown of the extracellular matrix. In the present study inhibition of furin was achieved by gene transfer of 1-antitrypsin Portland (PDX) cDNA. Experimental Design: This furin inhibitor was transfected into two tumorigenic astrocytoma cell lines. The inhibitory effect was evaluated using in vivo tumorigenicity, invasion, and proliferation assays, as well as by investigating impairment of furin substrate processing. Results: Expression of PDX prevented the s.c. growth of the transfected cells. Invasion assays demonstrated that PDX-transfected cells exhibited a reduced invasive ability in vitro and in vivo. Furthermore, s.c. growth of PDX transfectant xenotransplants showed a significant reduction in size that coincided with a significant decrease of the in vitro doubling time and of the in vivo cell proliferation ability. Additional studies showed that the furin substrates insulinlike growth factor IR, transforming growth factor and membrane type 1-matrix metalloprotease were not activated in PDX-expressing astrocytoma cells. Conclusions: PDX expression in astrocytoma cells demonstrated a direct mechanistic link between furin inhibition, and decreased astrocytoma proliferation and invasive ability. Because furin inhibition inhibits both invasiveness and cell growth in astrocytoma, furin should be considered a promising target for glioblastoma therapy. Introduction Limited proteolysis aimed at removing the initial propeptide fragment in the NH2-terminal end of proproteins is a molecular event required by many latent protein precursors to acquire biological activity. A family of Ca -dependent serine endoproteases named PCs, which participate in processing latent substrates to fully active enzymes, has been identified during the last decade. The PCs share a common phylogenetic origin with bacterial subtilisin and the yeast protease kexin, and their catalytic domains exhibit a high proportion (50–75%) of conserved residues (1, 2). To date, several members have been characterized in mammals, mostly in nervous and endocrine tissues. Among these, furin, PACE4, PC6B, and PC7 are ubiquitous enzymes expressed in many different tissues. Proteolytic cleavage by PCs occurs behind motifs of basic pairs KR2 or RR2, although basic residues at 4 and 6 upstream positions also contribute to substrate recognition (2). The up-regulated expression of PACE4 and furin in some types of cancer supports a possible functional role in tumorigenesis (reviewed in Ref. 3). Enhanced expression of the PC PACE4 was found in invasive mouse skin squamous cell carcinomas, whereas levels remained low in the less aggressive tumors (4). Among the most attractive substrates of PCs that might link proprotein processing with progression of cancer disease are MMPs, synthesized as inactive precursors, because of their role in the proteolytic digestion of the ECM that anticipate tumor invasion. Stromysin-3 is processed into its active form by furin (5, 6), and the group of MT-MMPs also contains insertions between the propeptide and the catalytic domains that include the paired basic amino acid recognition site for furin processing (7). Recently, a variant 1-antitrypsin was constructed, which contains in its reactive site Arg-X-X-Arg-, the minimal sequence required for efficient processing by furin (8). This furin-specific inhibitor called PDX, unlike other related serine protease inhibitors, does not inhibit either elastase or thrombin. PDX is a potent competitive inhibitor of furin (IC50 0.6 nM), and when expressed in cells (either by stable or transient transfection), blocks the processing of HIV-1 gp160 and measles virus-Fo inhibiting virus spread. PDX is also 10-fold more effective than currently used antiherpetic agents in cellculture models. The requirement of furin for the processing of envelope glycoproteins from many pathogenic viruses and for the activation of several bacterial toxins suggests that selective inhibitors of furin have potential as broad-based antipathogens (8). Furthermore, furin substrate processing was inhibited by PDX in four (furin-overexpressing) epithelial-derived Received 10/8/01; revised 2/26/02; accepted 3/11/02. 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 NIH National Cancer Institute Grants CA 75028, CA 06927, and by an appropriation of the Commonwealth of Pennsylvania. 2 To whom requests for reprints should be addressed, at Department of Pathology, Fox Chase Cancer Center, Philadelphia, PA 19111. Phone: (215) 728-3154; Fax: (215) 728-2899; E-mail: [email protected]. 3 The abbreviations used are: PC, proprotein convertase; PDX, 1antitrypsin Portland; ECM, extracellular matrix; IGF-IR, insulin-like growth factor receptor 1; LI, labeling indices; MMP, matrix metalloproteinase; MT-MMP, membrane type-MMP; TGF, tumor growth factor ; NHA, normal human astrocyte. 1740 Vol. 8, 1740–1746, June 2002 Clinical Cancer Research Research. on April 28, 2017. © 2002 American Association for Cancer clincancerres.aacrjournals.org Downloaded from malignant cell lines resulting in reduced tumorigenicity and invasiveness (9, 10). Although to our knowledge, furin expression levels have not been investigated in brain tumors, low levels of expression have been detected in neurons and glial cells of murine and human normal brain by in situ hybridization (11–13). High grade astrocytomas are characterized by enhanced proliferation and invasiveness, and their lethality is because of local brain tissue destruction rather than distant metastases. Invasion of glioma cells in the brain is preceded by a process of proteolysis and solubilization of the ECM that enables tumor cells to spread (14). Because local invasiveness is the major cause of glioma mortality and because furin is expressed in the nervous tissue together with several MMPs that contain sequences for cleavage and activation by furin, astrocytoma cells are excellent candidates to investigate the effects of furin inhibition as a possible new approach to astrocytoma therapy. Herein we demonstrate a very significant reduction in cell proliferation, tumorigenicity, and invasiveness after introduction of a cDNA plasmid for expression of the furin inhibitor PDX into furin-expressing anaplastic astrocytoma cell lines U87MG and U118MG. Materials and Methods Expression Plasmids. The EcoRI/SalI fragment encompassing the full length cDNA of PDX was directionally cloned into the pCI.neo vector (Promega) to yield pCIN.PDX. Cell Culture and Transfection. Parental U87MG and U118MG astrocytoma cells (American Type Culture Collection, Manassas, VA) were maintained in MEM supplemented with 2 mM L-glutamine, 1.5 g/liter sodium bicarbonate, 0.1 mM nonessential amino acids, 1 mM sodium pyruvate, and DMEM, respectively, with 10% FBS, 100 units/ml penicillin, and 100 g/ml streptomycin. Both lines were stably transfected by lipofection (LipofectAMINE PLUS, Life Technologies, Inc., Gaithersburg, MD) with either pCIN.PDX or pCI.neo vector alone (pCIN). G418 (1 mg/ml)-resistant colonies were picked, cloned, and screened for high 1-PDX expression by Western analysis. Transfectants were propagated in medium supplemented with 400 g/ml G418. Subconfluent cells were serum-starved 1 day before harvesting. Primary cultures of NHAs from Clonetics (San Diego, CA) were used as control cells. Protein Sample Preparation and Western Blot Analysis. For cell lysates, cell monolayers were harvested, washed, and disrupted with radioimmunoprecipitation assay (PBS containing 0.1% SDS, 0.5% sodium deoxycholate, 1% Triton X-100, 1% NP40, plus freshly added protease inhibitors: 2 mM phenylmethylsulfonyl fluoride, 0.2 mg/ml aprotinin, and 1 mM Na3V04) for 30 min at 4°C. The soluble part was used as protein extract. Furin was first immunoprecipitated from lysates with MON-148 (Alexis Biochemicals, San Diego, CA), a monoclonal antibody raised against an epitope in the catalytic domain. Total protein (100 g) was cleared with normal mouse IgG/protein A/G PLUS agarose (Santa Cruz Biotechnology, Santa Cruz, CA), and the cleared rest was then incubated with MON-148 and protein A/G PLUS agarose. Extracellular proteins were assessed on 24 h-conditioned medium concentrated 50-fold with Millipore centrifugal filter devices. For expression analysis, proteins were subjected to SDSPAGE using Tris-glycine gels, electrotransferred overnight onto nitrocellulose membranes, and probed with antibody against either furin (MON-152; Alexis Biochemicals), PDX ( 1-antitrypsin antibody; Sigma, St. Louis, MO), TGFaffinity-purified goat antihuman latent-associated peptide IgG (R&D, Minneapolis, MN), MT1-MMP (AB815; Chemicon International, Inc., Temecula CA), and IGF-IR (sc-711; Santa Cruz Biotechnology) were used as primary antibodies. Immunocomplexes were revealed with enhanced chemiluminescence based on the use of peroxidase-labeled IgG (enhanced chemiluminescence; Amersham Pharmacia Biotech). Immunohistochemistry. Furin immunohistochemistry was performed using paraffin-embedded material. All of the paraffin sections were subjected to antigen retrieval for 10 min in distilled water. MON 152 (Alexis) was used as primary antibody at 1/100 dilutions. An avidin-biotin-peroxidase kit (Vectastain Elite, Vector, Burlingame, CA) was then used followed by the chromagen 3 3 -diaminobenzidine to develop the immunostain. Negative controls, not incubated with furin antibodies, were incubated with normal mouse IgG. All of the sections were counterstained with hematoxylin and mounted. Normal brain tissue (frontal cortex) from three individuals, eight high-grade astrocytomas, and tracheal xenotransplants of U87MGpCIN cells (see below) were used. Doubling Time. Cells were grown in 12-well plates and counted every other day. Log of cell numbers versus days was plotted to obtain the doubling time for each transfected cell line. Zymography. Cells (1 10) were grown overnight in serum-free S-MEM medium containing 2 mM L-glutamine and Pen-Strep. The conditioned media were concentrated down to 200 l using Amicon centripreps (Fisher, Springfield, NJ), and 20 l of each sample was loaded on a 10% Novex precast zymogram (gelatin) gel. The gel was run, renatured, and developed according to the manufacturer’s instruction. Gelatinase Zymography Standards were purchased from Chemicon International, Inc. In Vitro Invasion Assay. The ability of cells to degrade ECM in vitro was assessed using Biocoat Matrigel invasion chambers (Becton Dickinson, Bedford, MA) according to the manufacturer’s instructions (15). In Vivo Invasion Assay. Tracheal transplants were prepared as described previously (10, 15). Cells (5 10) from each transfected cell line were inoculated into de-epithelialized rat trachea (Zivic-Miller, Pittsburgh, PA). Six to ten tracheas were used for each cell line. After cell inoculation, the tracheas were sealed and transplanted into the dorsal s.c. tissues of Scid mice. Tracheal transplants were removed surgically at 8 weeks, sectioned into 3-mm thick rings, and fixed in 10% formalin. After H&E staining, the degree of invasion of the tracheal wall was determined by measuring the length of maximum penetration of the tumor cells into the tracheal wall (10, 15). All of the microscopic images of cross-sections of tracheal transplants were digitized at a magnification of 40. Using an adequate reference micrometric scale, the lengths were determined by measuring the distance between the luminal center and the most distant point of tumor invasion. If the lumen was obliterated, the 1741 Clinical Cancer Research Research. on April 28, 2017. © 2002 American Association for Cancer clincancerres.aacrjournals.org Downloaded from distance measured was between the geometric center of the tumor mass inside the tracheal lumen and the most distant point of tumor invasion either in or outside the tracheal wall. Each tracheal transplant was represented by two to six measurements corresponding to the number of available cross-sections containing tumor cells. A mean was calculated for each tracheal transplant and for each group of transfected cells. The results were expressed in m of penetration depth. Paraffin sections of tracheal transplants containing transfected tumor cell lines were used for the immunohistochemical detection of Ki-67 (Mib-1). A mouse monoclonal Mib-1 (Immunotech, Westbrook, ME) and an avidin-biotin-peroxidase kit (Vectastain Elite; Vector) were used. Negative controls were incubated with normal mouse IgG. Labeling indices (percentage of Mib-1-positive cells) were calculated by counting stained and unstained cells in tracheal transplants containing either vector aloneor 1-PDX transfected cells (at least 500 cells/tracheal transplant were counted). In Vivo Tumorigenicity. Either PDX-transfected or vector alone-transfected cells (5 10) were injected into the s.c. tissues of Scid mice. Tumors were measured twice a week after the appearance of the first tumor for each pair of PDXand vector alone-transfected cells using a Vernier caliper. Volumes (V) of the tumors were obtained using the following equation: V [(L1 L2)/2] L1 L2 0.526, where L1 and L2 are the length and width of the s.c. tumor.