Glioblastoma is a severe type of primary brain tumor and its invasion is strongly correlated with the secretion of matrix metalloproteinases (MMPs). To investigate a role of PTEN, a tumor suppressor gene, in the regulation of hyaluronic acid (HA)-induced invasion of glioma cells, we examined the secretion of MMP-9 in various glioma cells with or without a functional PTEN gene. The secretion of MMP-9 in glioma cells lacking functional PTEN (U87MG, U251MG, and U373MG) was induced by HA, although not in wildtype (wt)-PTEN-harboring cells (LN229, LN18, and LN428). In addition, stable expression of wt-PTEN into U87MG cells significantly decreased the secretion of HA-induced MMP-9 and basal levels of MMP-2, inhibiting the activation of focal adhesion kinase and extracellular signal-regulated kinase 1/2, whereas the secretion levels of the tissue inhibitor of metalloproteinase-1 and -2 were increased, finally resulting in the inhibition of invasion by HA in vitro. Ectopic expressions of adenoviral (Ad)-wt-PTEN and -lipid phosphatase-deficient (G129E)-PTEN, but not both protein and -lipid phosphatase-deficient (C124S)-PTEN, reduced MMP-9 secretion and invasion by HA. These results were also confirmed by expressions of Ad-wt-PTEN and Ad-G129E-PTEN in other glioblastoma cells lacking functional PTEN, U251MG, and U373MG. These findings strongly suggest the possibility that PTEN may block HA-induced MMP-9 secretion and invasion through its protein phosphatase activity.
Glioblastoma, a severe type of primary brain tumor, is lethal because of local invasion into brain parenchyma. Glioma invasion is strongly correlated with the secretion of MMPs3 (1, 2). MMPs are a large family of zinc-dependent neutral endopeptidases, and are involved in the degradation of many different components of the extracellular matrix. Among the MMPs, MMP-9 specifically targets type IV collagen, a major component of the basement membrane, and appears to play a crucial role in glioma invasion across this barrier (3).
HA is the principal glycosaminoglycan found in extracellular matrix of brain. It binds to cell-surface receptors such as CD44 and appears to be involved in cell adhesion, migration, proliferation, and tumor progression (4, 5). In human brain, HA is distributed in white matter fiber tracts, which form the most frequent route of glioma dissemination (6). There are several studies that suggest an important role of HA in glioma cell invasion in vitro (7, 8). However, there have been only few studies on the molecular mechanism of HA-mediated invasion of glioma cells (9). Furthermore, the specific role of HA on the secretion of MMPs and the invasion by glioma cells is not well understood.
PTEN (also called MMAC1) is a tumor-suppressor gene located on human chromosome 10q23.3 (10, 11), and it regulates cell growth, apoptosis (12), and interaction with the extracellular matrix, and inhibits cell migration, spreading, and focal adhesion (13). PTEN protein exhibits dual specificity protein phosphatase activity in vitro (14) and its potential cellular target of FAK (15). However, it can also dephosphorylate the lipid signal transduction molecules phosphatidylinositol 3,4,5-triphosphate and phosphatidylinositol 3,4-bisphosphate, which are both involved in the PI3K pathway (14). It has been suggested that dephosphorylation of FAK by the PTEN protein is correlated with cell spreading, migration, and invasion (13, 15). In addition, Koul et al. (16) recently reported that PTEN suppresses MMP-2 gene expression, and invasion of glioma cells and phosphatase activity is essential in these events. On the other hand, it has also been reported that the lipid phosphatase activity of PTEN is not required in the invasive potential of glioma cells (17).
In this report, we describe the effect of HA on the secretion of MMP-9 and invasion of human glioma cell lines, and the role of PTEN in these events. Our findings suggest that PTEN suppresses HA-induced secretion of MMP-9, possibly through FAK dephosphorylation; thus, PTEN inhibits HA-induced invasion via control of MMP-9 secretion.
MATERIALS AND METHODS
Reagents and Cell Culture.
HA was obtained from Sigma Chemical Co. (St. Louis, MO) and reconstituted in DMEM (Life Technologies, Inc., Grand Island, NY). The following inhibitors purchased from Calbiochem (La Jolla, CA) were used in this study: Ras-specific inhibitor (damnacanthal), PKC inhibitor (G06983 and GF109203X), MAPK inhibitor (PD98059) and ERK-1/2 inhibitor (SB203580). Human glioma cell lines [U87MG, U251MG, and U373MG (obtained from American Type Culture Collection) and LN18, LN428, and LN229 (a generous gift from Dr. Frank Funari, Ludwig Institute for Cancer Research, La Jolla, CA)] were maintained in DMEM supplemented with 10% heat-inactivated FBS, penicillin (100 units/ml) and streptomycin (100 μg/ml) at 37°C in a humidified atmosphere containing 5% CO2.
Plasmid Construction and Cell Transfection.
wt-PTEN cDNA was obtained from Dr. Hong Sun (Yale University, New Haven, CT) and cloned into pcDNA3 vector to generate pcDNA3-PTEN. The entire open reading frame was sequenced to confirm the correct sequence. U87MG cells were transfected with pcDNA3-PTEN or pcDNA3 (no insert) in triplicate dishes and in three independent experiments by using the Effectene reagent (Qiagen, Valencia, CA) according to the manufacturer’s protocol. After 24 h, cells were split at a 1:5 dilution and exposed for 2–3 weeks in G418 (Boehringer Mannheim, Indianapolis, IN)-containing medium (800 μg/ml) and colonies were picked for their resistance to G418. Expression of PTEN was confirmed by Western blot analysis using a monoclonal antibody against PTEN.
Construction and Infection of Recombinant Adenovirus.
To construct Ad-PTEN vectors, 1.2-kb fragments of PTEN cDNAs [wt- and mutant type-PTEN (C124S and G129E, a generous gift from Dr. Young E. Whang, University of North Carolina School of Medicine, Chapel Hill, NC)] were cloned into the KpnI and XhoI sites of pShuttle-CMV vector (AdEasy Adenoviral Vector System; Stratagene, La Jolla, CA) and cotransformed into Escherichia coli BJ5183 cells with the pAdEasy-1 vector, using the electroporation method and selected colonies harboring homologous recombinant plasmid in agar plates containing kanamycin, as described in the manufacturer’s protocol. Recombinant adenoviruses were produced by transfection of homologous recombinant plasmid into the HEK cell line 293. As a control, Ad-LacZ was made from pSuttle-CMV-lacZ encoding the β-gal gene in the manner described previously. Viruses were propagated in the HEK293 cell line and purified by two rounds of CsCl density centrifugation; viral titers were measured in a limiting-dilution bioassay using the HEK293 cells. The recombinant adenoviruses were infected into glioma cells with 50 plaque-forming units per cell in serum-free medium containing 2.5 mg of Polybrene (Sigma) per ml and incubated for 90 min at 37°C.
Transfection of Antisense Oligonucleotides of FAK.
FAK sense (5′-AGTTCCATTCGTCGACGGTA-3′) and antisense (5′-AAGCAGCTGCCATTATTTTG-3′) phosphorothioate oligonucleotides were obtained from Bioneer Co. (Cheongwon, Chungbuk, Korea) and each olignucleotide was transfected into U87MG cells using Effectene reagent following the supplier’s instructions.
Glioma cells in subconfluent culture (∼70–80% cell density of confluent culture) were washed and refreshed with serum-free DMEM and were incubated with or without HA for 18 h. In some experiments, cells were preincubated for 30 min with various kinase inhibitors before the addition of HA. The enzymatic activity and molecular weight of electrophoretically separated gelatinolytic enzymes in the conditioned medium of glioma cells were determined by SDS-PAGE as follows. Twenty μl of serum-free culture medium per sample were prepared in nondenaturing loading buffer (0.5 M Tris-HCl, pH 6.8, 10% SDS, 0.1% bromophenol blue, and 10% glycerol) and were size-fractionated in 10% SDS-polyacrylamide gel impregnated with 0.1% gelatin. The gels were then washed with 2.5% Triton X-100 for 1 h at room temperature to remove SDS, rinsed twice with water, and then incubated in a developing buffer [50 mm Tris-HCl buffer (pH 7.4), 20 mm NaCl, 10 mm CaCl2, and 0.1 NaN3] for 18 h at 37°C. Subsequently, gels were fixed and stained with 10% 2-propanol and 10% acetic acid containing 0.5% Coomassie Blue R250. Gelatinase activity was visualized as clear bands within the stained gel.
Western Blot Analysis.
The activation of ERK 1/2 and FAK was determined by Western blotting using antibodies specific for phosphorylated forms of the corresponding to ERK 1/2 and FAK. Glioma cells were stimulated with HA at 100 μg/ml and lysed in lysis buffer [20 mm Tris (pH 7.4), 150 mm NaCl, 1 mm EDTA, 1 mm EGTA, 1% Triton, 2.5 mm sodium PPI, 1 mm β-glycerolphosphate, 1 mm Na3VO4, 1 μg/ml leupeptin, and 1 mm phenolmethylsulfonyl fluoride]. After a brief sonication, the lysates were clarified by centrifugation at 12,000 × g for 15 min at 4°C, and protein content was measured by a Bradford’s method. An aliquot (30 μg protein/lane) of the total protein was separated by 10 or 12% SDS-PAGE and blotted to nitrocellulose transfer membrane (0.2 μm; Amersham, Arlington Heights, IL). The membrane was blocked with 5% nonfat skim milk in TBST [20 mm Tris-HCl (pH 7.6), 137 mm NaCl, and 0.01% Tween 20] for 1 h at room temperature, followed by incubation with the primary antibodies. After extensive washing with TBST, the membrane was reprobed with the secondary antibody of horseradish peroxidase-linked antirabbit immunoglobulin at 1:3,000 in TBST for 40 min at room temperature. Immunoblots were visualized by enhanced chemiluminescence (Amersham), according to the manufacturer’s protocol.
Conditioned medium from each sample was also subjected to protein analysis of MMP-2 and -9 and TIMP-1 and -2. For this purpose, culture medium in each tissue culture dish was collected and concentrated using a Centricon 10 microconcentrator (Amicon, Beverly, MA), and 5-fold concentrated conditioned medium (20 μl) was then used for SDS-PAGE analysis.
In Vitro Invasion Assay.
Invasion assays were performed using modified Boyden chambers with polycarbonate Nucleopore membrane (Corning, Corning, NY). Precoated filters (6.5 mm in diameter, 8 μm pore-size, Matrigel 100 μg/cm2) were rehydrated with 100 μl of medium, and 2 × 105 cells in 100 μl of medium, with or without HA, were seeded in triplicate into the upper part of each chamber, and the lower compartment was filled with 1 ml of serum-free DMEM supplemented with 0.1% BSA. After incubation for 18 h at 37°C, noninvaded cells on the upper surface of the filter were wiped with a cotton swab, and migrated cells on the lower surface of the filter were fixed and stained with Diff-Quick kit. Invasiveness was determined by counting cells in five microscopic fields per well, and the extent of invasion was expressed as an average number of cells per microscopic field.
HA Induces MMP-9 Secretion in Glioma Cell Lines Lacking Functional PTEN.
To investigate the effect of HA on the secretion of MMPs, we examined MMP activities in various glioma cells stimulated with HA. Interestingly, HA markedly induced secretion of MMP-9 in functional-PTEN-deficient glioma cell lines (U87MG, U251MG, and U373MG; Fig. 1,A), whereas it did not do so in glioma cell lines harboring wt-PTEN (LN18, LN229, and LN428; Fig. 1 B). These results prompted us to examine PTEN function in HA-induced MMP-9 secretion.
PTEN Inhibits HA-induced MMP-9 Secretion and Invasion in U87MG Cells.
To examine the role of PTEN in HA-induced MMP-9 secretion, the wt-PTEN gene was transfected into U87MG cells to get a stable clone highly expressing wt-PTEN. Expression of wt-PTEN was verified by Western blot analysis (Fig. 2,A). Overexpression of wt-PTEN substantially reduced the phosphorylation of protein kinase B/Akt and enhanced the expression of p27KIP1 cyclin-dependent kinase inhibitor compared with controls in these cells, whereas p21WAF1/CIP1 and p16INK4a were unaffected (Fig. 2 A).
A gelatin zymogram showed that the secretion of HA-induced MMP-9 and basal levels of MMP-2 were reduced compared with U87MG cells (vector only) in wt-PTEN-transfected cells (Fig. 2,B). Expression levels of MMP-2 and -9 were also confirmed by Western blot analysis in concentrated conditioned medium (Fig. 2,C). In addition, the secretion levels of TIMP-1 and -2, endogenous inhibitors of MMPs, were increased in conditioned medium from PTEN-transfected cells (Fig. 2,C). Based on these results, we examined the effect of PTEN in invasion of the glioma cells. As shown in Figure 2D, the transfection of PTEN reduced the basal and the HA-induced in vitro invasion in these cells (Fig. 2 D). These results suggest that PTEN plays an important role in controlling MMP and TIMP secretion and in the in vitro invasion in U87MG cells whether or not stimulated with HA.
Ras, FAK, and ERK 1/2 Signaling Plays a Role in HA-induced MMP-9 Secretion in U87MG Cells and PTEN Inhibits These Signaling Pathways.
Because PTEN is a dual specificity phosphatase, we next examined the involvement of phosphorylation of various kinases in HA-induced MMP-9 secretion using specific kinase inhibitors. Gelatin zymogram analysis showed that the secretion of HA-induced MMP-9 and basal levels of MMP-2 were reduced by inhibitors of Ras (damnacanthal) and ERK 1/2 (PD98059) in concentration-dependent manners, but not by the inhibitors of PKC (Go6983 and GF109203X) and PI3K (wortmannin; Fig. 3,A). These results suggest that Ras and ERK 1/2 signaling play a critical role in HA-induced MMP-9 secretion. To confirm this observation, we next examined ERK 1/2 activation by HA in U87MG control cells and in wt-PTEN-transfected cells. ERK 1/2 activation was observed within 5 min after HA treatment, and there was a marked increase in 15 min in control cells. In contrast, ERK 1/2 activation was significantly reduced in wt-PTEN-transfected cells compared with control cells (Fig. 3 B).
FAK has been known as one of the targets of PTEN (15). As shown in Fig. 3,C, HA induced phosphorylation of FAK, and wt-PTEN transfection reduced this effect. To confirm the involvement of FAK in HA-induced MMP-9 secretion, we introduced antisense oligonucleotides to FAK into U87MG cells and found that antisense FAK reduced the expression of FAK and the secretion of MMP-9 that had been induced by HA, as well as levels of MMP-2 (Fig. 3,D). Furthermore, antisense FAK transfection reduced the HA-induced invasion of these cells (Fig. 3 E). These results suggest that FAK signaling, at least in part, is required for the augmented secretion of MMP-9 by HA.
Protein Phosphatase Activity of PTEN Is Critical for the Invasion of U87MG Cells Induced by HA.
To examine the role of phosphatase activity in HA-induced MMP-9 secretion, we introduced three kinds of Ad-PTEN to U87MG: wt-, protein-and-lipid-phosphatase-deficient (C124S)-, and lipid-phosphatase-alone-deficient (G129E)-PTEN. Introduction of Ad-wt-PTEN and Ad-G129E-PTEN showed suppression of HA-induced MMP-9 secretion. In contrast, Ad-C124S-PTEN infection had no effect on the secretion of MMP-9 induced by HA. This indicates that protein phosphatase activity is crucial in HA-induced MMP-9 secretion (Fig. 4 A).
We next examined the effect of PTEN on HA-induced cell invasion using modified Boyden chambers with polycarbonate Nucleopore membrane. Ad-wt-PTEN infection reduced the in vitro invasion of U87MG cells. Inhibition of invasion was also observed with Ad-C124S-PTEN and Ad-G129E-PTEN; however, the effect was less than with Ad-wt-PTEN (Fig. 4,B). After treatment of HA, we observed markedly increased invasiveness in control and in Ad-LacZ-infected cells. In contrast, infection with Ad-wt-PTEN and Ad-G129E-PTEN showed HA-induced invasion reduced by ∼52 and 45%, respectively. However, Ad-C124S-PTEN-infected cells had no effect (Fig. 4 B). These data suggest the possibility that protein phosphatase activity of PTEN is critical in mediating the invasion of U87MG cells induced by HA.
PTEN Also Inhibits the MMP-9 Secretion and Invasion of Other Functional-PTEN-deficient Glioma Cells, U251MG and U373MG.
To investigate whether the effect of PTEN on the reduction of MMP-9 secretion and in vitro invasion by HA was a general phenomenon of functional-PTEN-lacking glioma cells, we infected Ad-wt-PTEN, Ad-C124S-PTEN, and Ad-G129E-PTEN into U251MG and U373MG cells, stimulated with HA, and performed gelatin zymography, Western blotting, and in vitro invasion assays. Expression of PTENs in these cells were confirmed by Western blot analysis (Fig. 5,A). Gelatin zymography showed that MMP-9 secretion was increased in Ad-LacZ and Ad-C124S-PTEN-infected U251MG and U37MG cells upon treatment of HA, whereas Ad-wt- and Ad-G129E-PTEN-infected cells suppressed the MMP-9 secretion (Fig. 5,B). Also, FAK and ERK 1/2 activation, and in vitro invasion by HA were reduced in the Ad wt- and Ad-G129E-PTEN-infected cells compared with the cells infected by Ad-LacZ and Ad-C124-PTEN (Fig. 5, B and C). These data indicate that PTEN generally inhibits the MMP-9 secretion and the in vitro invasion of functional-PTEN-deficient glioma cells induced by HA.
In this study, we demonstrated that HA induced MMP-9 secretion in U87MG, U251MG, and U373MG cells through the FAK and ERK 1/2 pathways and increased the invasion of the cells. In addition, introduction of the wt-PTEN gene reduced the phosphorylation of FAK and ERK 1/2, secretion of MMP-9, and the invasion induced by HA. Protein phosphatase activity is critical in these events, which suggests a role for PTEN in the HA-induced invasion of these gliomas.
HA, a major extracellular matrix component of brain parenchyma, facilitates glioma migration and invasion through receptors such as CD44 (7, 9, 18). However, the mechanisms of migration and invasion of glioma cells induced by HA are not well understood. To study the mechanisms of stimulation of MMP-9 secretion and invasion of the cells by HA, we used damnacanthal and PD98059, specific inhibitors of Ras and MAPK kinase-1, respectively, and an antisense oligonucleotide of FAK. We found that the inhibitors and antisense oligonucleotide reversed HA-induced secretion of MMP-9 in U87MG cells. However, Go6983 and GF109203X, and wortmannin, specific inhibitors of PKC and PI3K, respectively, did not influence the secretion of MMP-9. These results suggest the possibility that HA-induced MMP-9 secretion and in vitro invasion of glioma cells might be achieved by the FAK-ERK 1/2 and/or the Ras-ERK 1/2 pathways, but not by the PKC or PI3K pathways, although the specific mechanism needs to be further elucidated.
One of the main substrates of the PTEN protein is FAK (13). FAK is required for the integrin-mediated Ras-ERK 1/2 signaling pathway (19) and modulates cellular adhesion, migration, invasion, and cytoskeletal formation (13, 15). Gliomas are typically highly invasive and express high levels of FAK; the inhibition of FAK reduces invasion (20). FAK is also involved in the activation of MMPs secretion induced by fibronectin or ConA (21, 22). In our study, HA induced MMP-9 secretion in U87MG cells and activated the phosphorylation of FAK. Antisense oligonucleotides to FAK reduced MMP-9 secretion and invasion in these cells. Therefore, our results suggest that FAK may be involved, at least in part, in the HA-induced MMP-9 secretion and invasion of the cells through the ERK 1/2 pathway.
Although inhibition of FAK reduces the invasion of glioma cells, PTEN status does not influence the phosphorylation of FAK (20). In addition, functional-PTEN expression in U87MG cells does not inhibit FAK phosphorylation induced by fibronectin, and the phosphatase domain of the PTEN protein is not required in the invasion of glioma cells (17). On the contrary, Tamura et al. (13, 15) showed that the suppression of FAK-mediated cell spreading, migration, invasion, and cytoskeletal formation in glioma cells with mutated PTEN alleles, by expression of exogenous PTEN, requires protein phosphatase activity. Furthermore, Koul et al. (16) recently showed that expression of wt-PTEN in U251MG glioma cells results in the inhibition of MMP-2 expression and invasion, and that the phosphatase activity of PTEN is essential for this function. In our study, HA-induced phosphorylation of FAK and ERK 1/2, secretion of MMP-9, and invasion were reduced by the stable transfection of functional PTEN in U87MG cells. Furthermore, infection of Ad-wt- and Ad-G129E-PTEN, but not of Ad-C124S-PTEN, reduced the phosphorylation of FAK, MMP-9 secretion, and invasion of U87MG cells. In addition, the infection of Ad-wt-PTEN and Ad-G129E-PTEN in U251MG and U373MG, other functional-PTEN-deficient cells, also reduced phosphorylation of FAK and ERK 1/2, secretion of MMP-9, and invasion by HA. Therefore, our results suggest that the protein phosphatase activity of PTEN is required in the inhibition of HA-induced activation of FAK and the invasion of these glioma cells.
In summary, we suggest that HA induces the invasion of glioma cells by the induction of MMP-9 through the FAK-ERK 1/2 signaling pathway. Introduction of the functional-PTEN gene decreases these effects and the protein phosphatase activity of the PTEN protein is critical in these events.
Supported by the National Cancer Control Program of the Ministry of Health and Welfare and the National Nuclear R & D program of the Ministry of Science and Technology, Seoul, Korea.
The abbreviations used are: MMP, matrix metalloproteinase; HA, hyaluronic acid; FAK, focal adhesion kinase; PI3K, phosphoinositide 3-kinase; PKC, protein kinase C; MAPK, mitogen-activated protein kinase; ERK 1/2, extracellular signal-regulated kinase 1/2; TIMP, tissue inhibitor of metalloproteinase; wt, wild type; Ad, adenoviral; HEK, human embryonic kidney.