Expression of Chicken Ovalbumin Upstream Promoter-Transcription Factor II Enhances Invasiveness of Human Lung Carcinoma Cells

Degradation of the extracellular matrix (ECM) and components of the basement membrane by proteases facilitates the detachment of tumor cells, their crossing of tissue boundaries, and invasion into adjacent tissue compartments. Several of these phenotypic alterations result from the dysregulation of gene expression, particularly at the level of transcription (1). Analysis of mRNA expression in vitro suggested that alterations in expression of nuclear transcription factors such as COUP-TFs might be involved in this dysregulation (2).

COUP-TFs are the best characterized “orphan” (ligands not yet identified) nuclear receptors (2). They belong to the steroid/thyroid hormone receptor superfamily of nuclear receptor proteins and are required for regulation of gene expression (3), development, differentiation, and homeostasis (4). Angiogenesis in COUP-TFII mutant mice is largely impaired (5), and defects mimic the phenotypes exhibited by mice lacking angiopoietin-1 or its receptor, TIE2. It has also been suggested that COUP-TFII plays an important role in mesenchymal-endothelial interactions (5). COUP-TFs are expressed in some tumor cell lines (6), including human endometrial cancer cells (7), lung cancer cells (8), and in adrenal tumors (9), but are not expressed in terminally differentiated epithelial cells. The relationship between COUP-TFII expression and cancer development is not known.

Several types of proteases contribute to the degradation of the extracellular matrix: serine proteases (e.g., plasmin and urokinase-type plasminogen activator; uPA; Ref. 10), cysteine proteases (e.g., cathepsins B and L; Ref. 11), and matrix metalloproteinases (MMP; Ref. 12). MMPs are implicated in tumor invasion and metastasis (13). Other features of tumor evolution, including survival, growth, and angiogenesis, may also be dependent on MMPs (14). MMP members are classified into subgroups on the basis of their structure and substrate preference (15). Among these, the gelatinases (MMP-2 and MMP-9) are closely associated with tumor invasiveness and metastasis because of their potent ability to degrade type IV collagen present in the basement membrane that surrounds blood vessels. Elevated levels of gelatinases are found in many types of human cancers (16).

In the extracellular milieu, the activity of MMPs is controlled by tissue inhibitors of MMPs (TIMP; Ref. 17). Although these inhibitors have similar inhibitory activities against most MMPs, they differ in many aspects including their interactions with pro-MMPs, solubility, transcriptional regulation, and tissue specificity (17). TIMP-1 forms complexes with pro-MMP-9, and TIMP-2 and TIMP-4 with pro-MMP-2 (18).

MMPs are produced by cells as proenzymes, and they require additional processing to generate the active enzyme. The activation of pro-MMPs is facilitated by active forms of other MMPs. For example, the activation of pro-MMP-2 requires a membrane-type MMP at the cell surface (16). The activation of pro-MMPs can also be mediated by other groups of proteinases (19). For example, a serine proteinase, plasmin, can activate gelatinases without the action of other metallo- or acidic proteinases (20). It has been shown that increased expression of uPA and its membrane-bound receptor (uPAR; CD87) is closely correlated with an increase in disease recurrence and with early death of lung and other cancer patients (21). Cell membrane-associated uPAR is a key molecule for the induction of pericellular proteolysis, because plasminogen is efficiently activated to plasmin by cell surface-associated interactions with uPAR-bound uPA (22).

Interference with the membrane-associated function of uPAR should result in a reduction in plasminogen activation, which would decrease tumor cell proliferation, invasion, and metastasis (23). In this context, it is important to note that the membrane anchoring of uPAR is not a prerequisite for uPA activation but is necessary for plasminogen activation (22). In addition to the proteolytic function of uPA, the uPA/uPAR interaction induces downstream signaling, resulting in the induction of cell proliferation, adherence, migration, and chemotaxis (24).

We examined the relationship between COUP-TFII and ECM proteinases. We demonstrate that COUP-TFII-transfected A549 cells (a human lung cancer cell line in which COUP-TFII is not normally expressed) acquired invasive ability, had increased in vitro tumorigenicity and migratory ability, and displayed enhanced expression of collagenase type IV (MMP-2). These results were confirmed by transduction of different human lung cancer cell lines and a human breast carcinoma cell line (MDA-MB231) with purified COUP-TFII protein.

Interestingly, transfection of a very invasive human lung cancer cell line (H460SM) with COUP-TFII in the antisense orientation decreased COUP-TFII expression level, suppressed their invasive and migratory ability, and profoundly reduced the synthesis of both pro- and active forms of MMP-2. To analyze more specifically the contribution of COUP-TFII to the process of invasion, the regulations of human (h)uPA and/or huPAR were studied. Our study suggests a critical role for COUP-TFII in human lung cancer invasion.

A549, HeLa, and NCI-H460 (American Type Culture Collection, Rockville, MD) were maintained as monolayer culture in DMEM (A549 and NCI-H460) and Eagle’s MEM (HeLa) supplemented with 10% heat-inactivated fetal bovine serum (Life Technologies, Inc., Burlington, Ontario, Canada). H661, H520, H441, (American Type Culture Collection), and MDA-MB231 (kindly provided by Dr. Wolfgang Vogel from University of Toronto) were maintained as a monolayer culture in RPMI 1640 supplemented with 10% fetal bovine serum. H460SM (25) was grown in RPMI 1640 containing 10% fetal bovine serum.

A549 cells were transfected with a pCR 3.1 expression cassette containing a 1.5-Kb mouse COUP-TFII cDNA fragment (kindly provided by Dr. Ming-Jer Tsai, Baylor College of Medicine, Houston, TX) or with the empty vector (mock). To suppress the expression of COUP-TFII, H460SM cells were transfected with a plasmid containing the COUP-TFII coding region in the antisense orientation. This plasmid was made by ligating an EcoRI-XhoI fragment corresponding to the full-length COUP-TFII cDNA into the pEGFP.C1 expression vector (Clontech, Palo Alto, CA), in the antisense orientation.

Transfections were carried out using coprecipitation with calcium phosphate (26), and stable G418-resistant clones were isolated.

To generate a transducible TAT fusion COUP-TFII protein, we used a bacterial expression vector, pTAT-HA (27), which contains an NH₂-terminal 6-histidine leader followed by the 11-amino acid TAT protein transduction domain and a hemagglutinin (HA) tag. The murine COUP-TFII coding region, which has >90% homology to its human analog, was inserted into the polylinker region of pTAT-HA. To facilitate the cloning, restriction sites NcoI and EcoRI were included in the 5′(GCACCATGGCAATGGTAGTCAGCACGTGGC) and 3′(GGAATTCTTATTGAATTGCCATATATGGCCAG) primers, respectively. To express a control protein, we cloned a 1.1-kb Cre cDNA sequence into XhoI-EcoR1 sites of pTAT-HA.

The pTAT-HA-COUP-TFII and pTAT-HA-Cre plasmids were used to transform Escheria coli DH5α and, after sequencing confirmation, put into E. coli BL21(DE3). The transformed E. coli BL21(DE3) were cultured as described (28) and the fusion proteins purified under denaturing conditions (27). The proteins were refolded by gradually decreasing the concentration of guanidine hydrochloride (Gu-HCl) from 6 M to 0 M, by dialysis against a buffer containing 100 mm potassium phosphate (pH 6.0), 500 mm KCl, 10% glycerol, EDTA-free protease inhibitor mixture tablets (Roche), 8 mm βmercaptoethanol, and 0.1% Tween 20, and then dialysis against PBS containing 10% glycerol, 2 mm DTT, and 50 μm 4-(2-aminoethyl) benzenesulfonylfluoride.

To transduce the tumor cells with TAT-HA-COUP-TFII and TAT-HA-Cre, we followed the method described by Vives et al. (29). Cells were seeded at a concentration of 2 × 10⁵ per 10-cm dish (for Matrigel invasion and migration assays) or 3 × 10⁴ cells/well on coverslips (for immunostaining).

The purified proteins were dissolved in serum-free medium at a concentration of 1 nm, incubated at 37°C for 15 min, then incubated with the cells at 37°C for 2–4 h. The cells were then rinsed three times with PBS (pH 7.3) at room temperature and prepared for Matrigel invasion and migration assays or immunostaining. Immunostaining was performed using monoclonal anti-HA antibody (BAbCO, Richmond CA) as described (28).

Invasive and migratory ability of tumor cells was measured using the Matrigel invasion assay (30). The soft agar cloning assay was performed as described (31). Briefly, tumor cells (10⁴) were suspended in 0.8% low melting point agarose (Difco Laboratories Inc., Detroit, MI) at room temperature, mixed with an equal volume of 2× concentrated culture medium, and plated onto an agarose bed consisting of 2% low melting point agarose and the same medium. After 12 days, colonies were stained with neutral red and those exceeding 250 μm in diameter were enumerated using an inverted microscope (Leica; DMIRB).

Total RNA was prepared using Trizol reagent (Life Technologies, Inc., Rockville, MD) according to the manufacturer’s protocol. RNA blots on Hybond nylon (Amersham, Oakville, Ontario, Canada) were probed with ³²P-labeled 1.5 kb mouse COUP-TFII cDNA, 905-bp human GAPDH cDNA (Ambion Inc., Austin, TX), or full-length cDNA of human uPA or uPAR (provided by Dr. Shafaat A. Rabbani, The Royal Victoria Hospital, Montreal, Quebec, Canada). DNA probes were prepared by random primer extension (30). Relative amounts of the mRNA transcripts visualized by autoradiography were analyzed using NIH Image software and normalized to the internal GAPDH control.

COUP-TFII-transfected and mock-transfected or control A549 cells were seeded on coverslips, washed twice with PBS, and fixed with 3% paraformaldehyde.

Cells were permeabilized with 0.25% Triton X-100. Actin was stained with rhodaminephalloidin for 30 min.

COUP-TFII-transfected and mock (vector) -transfected or control A549 cells were serum-starved for 24 h, washed twice with PBS, and collected as a single cell suspension in DMEM with or without 400 μg/ml G418 plus 0.25 mg/ml BSA. Cells were kept in suspension for 60 min at 37°C and then seeded on fibronectin-coated or noncoated culture plates. Cells were harvested as described (32), and the lysate was then centrifuged at 100,000 × g for 30 min at 4°C. The resulting supernatant represented the cytoplasmic fraction and was subjected to immunoprecipitation.

Immunoprecipitation was performed as described (31). Ten μg of polyclonal rabbit anti-FAK antibodies (kindly provided by Dr. Jun-Lin Guan, Cornell University, Ithaca, NY) was added to the cytoplasmic fraction for 2 h at 4°C. Immune complexes were precipitated by incubation with protein A Sepharose beads for 60 min at 4°C. The beads were washed four times with the hypotonic buffer (31), resuspended in Laemmli SDS sample buffer (33), and boiled for 10 min. The eluted proteins were separated by electrophoresis on 8% SDS- polyacrylamide gels under reducing conditions.

Proteins (30 μg) were separated by electrophoresis on 8% SDS-polyacrylamide gels and transferred onto nitrocellulose filters (Bio-Rad, Hercules, CA). The blots were incubated in TNT buffer [0.15 M NaCl (pH 7.5) containing 0.05% Tween 20 and 10 mm Tris] containing 5% skimmed milk and 2% BSA, or 3% skimmed milk (for FAK phosphorylation detection), and probed with MMP-2 antiserum or an anti-TIMP-2 antibody (Oncogene Research Products, Boston, MA). The blots were also probed with anti-FAK or anti-FAK phosphotyrosine (Tyr 397; Upstate, Lake Placid, NY) antibodies at a 1:1000 dilution. To visualize the bands, blots were incubated with horseradish peroxidase-conjugated antimouse or rabbit IgG antibodies, and detected by enhanced chemiluminescence (Amersham, Baie d’Urfé, Quebec, Canada).

COUP-TFII-transfected A549 cells and COUP-TFII-antisense-transfected H460SM cells, together with the control and mock-transfected cells, were incubated in serum-free medium at 37°C for 48 h. The conditioned medium was then collected, dialyzed, and concentrated by freeze-drying at −50°C. Proteins in concentrated conditioned medium (30 μg/sample) were diluted in nonreduced SDS sample buffer and separated by electrophoresis in 10% SDS-polyacrylamide gels copolymerized with 1 mg/ml of gelatin or casein (for MMP-2 or plasmin activity detection, respectively). Gels were washed with 2.5% Triton X-100 for 1 h and then twice in Tris-HCl (pH 8.0) for 15 min at room temperature. The gels were incubated with substrate buffer [50 mm Tris-HCl (pH 8.0) and 10 mm CaCl₂] for 18 h at 37°C. The gels were then stained with Coomassie brilliant blue and destained until the clear bands of lysis appeared. To confirm the lytic bands, gels were treated with 20 mm EDTA (a metalloproteinase inhibitor) or 500 μg/ml amino-n-caproic acid (a plasmin inhibitor) in the substrate buffer for 18 h at 37°C.

A549 and H460SM cells were cultured in DMEM and RPMI 1640, respectively, without serum. The COUP-TFII-transfected A549 and COUP-TFII antisense-transfected H460SM cells were cultured in serum-free medium containing 400 μg/ml G418 for 24 h, dispersed, and seeded in 96-well plates (Falcon, Lincoln Park, NJ) at a density of 10⁵ cells/well. The cells were washed three times with serum-free medium, incubated at 37°C for 30 min, and then incubated for 1 h on ice with 5 μg/ml of monoclonal antibody against huPAR (CD87; American Diagnostica Inc., Greenwich, CT). After extensive washing with cold medium, the cells were incubated with FITC-labeled goat antimouse IgG antibody (1:200) for 1 h on ice, washed, then fixed in PBS containing 1% paraformaldehyde. The labeled cells were analyzed by flow cytofluorometry using a FACScan System (BD Bioscience).

The two-tailed t test was used to analyze differences in the invasiveness and migration of tumor cells.

It is reported that COUP-TFs are expressed in some human lung cancer cell lines, such as H460 and H661, but not expressed in others, such as A549, H520, and H441(34). To test the role of COUP-TFII in tumor progression, we stably transfected A549 cells with a plasmid vector (pCR3.1) expressing COUPTFII cDNA or with pCR3.1 as a mock control. As shown by Northern blot analysis (Fig. 1,A), an isolated stable clone (clone 1) and a total population of 20–30 clones (T.P.3) expressed higher levels of COUP-TFII mRNA than their parental cell line A549 or the mock controls (mock 3 from a single clone and mock T.P.1 from a mixture of ∼20 clones). COUP-TFII expression was higher in clone 1 than in T.P.3. (Fig. 1 A).

As a complementary approach, we generated stable cell lines of H460SM cells expressing COUP-TFII antisense RNA. According to Northern blot analysis, cells stably transfected with antisense COUP-TFII (T.P.2 from total population of ∼20 colonies and clone 20 from a single clone) showed 59–68% reduction in COUP-TFII mRNA expression compared with H460SM or mock T.P.5 (a mixture of ∼20 clones; Fig. 1 B). NCI-H460, the parental cell line of H460SM, also exhibited a lower level (48%) of COUP-TFII expression compared with H460SM.

To examine whether COUP-TFII expression correlates with the invasiveness of the cancer cell lines, we determined in vitro the cell invasiveness using reconstituted basement membrane (Matrigel) invasion assays. As shown in Fig. 2,A and B, human lung cancer cell lines A549, H520, and H441 and a human breast carcinoma cell line (MDA-MB231), which do not express COUP-TFII (34), had poor invasive and migratory abilities. In contrast, human lung cancer cell lines NCI-H460, its highly invasive variant H460SM, H661, and HeLa cells exhibited higher levels of invasiveness (∼2-, 14-, 13-, and 27-fold increases in invasiveness compared with A549, H520, H441, and MDA-MB231, respectively) and migratory ability (∼7-, 16-, 14-, and 19-fold increases in migration compared with A549, H520, H441, and MDA-MB231, respectively; Fig. 2, A and B) in agreement with their levels of COUP-TFII expression (34).

Using COUP-TFII-transfected cells, we observed that in both T.P.3 and clone 1, invasion was significantly (P < 0.05) increased (∼5- and 17-fold) relative to nontransfected and mock-transfected A549 cells (Fig. 2,C). In addition, we discovered a significant increase (P < 0.05; approximately 20–30-fold) in the migratory ability of the COUP-TFII-transfected A549 cells compared with controls (Fig. 2,D), measured on uncoated or type IV collagen-coated filters. Accordingly, we found that the level of invasion and migration were significantly (P < 0.05) decreased (approximately 14–27-fold in invasion and approximately 2–4-fold in migration) in H460SM cells expressing COUP-TFII-antisense mRNA (Fig. 2, E and F).

To additionally confirm the role of COUP-TFII in lung cancer invasiveness, we transduced A549, H520, and H441 with purified TAT-HA-COUP-TFII and TAT-HA-Cre (as a control) fusion proteins. We also used a human breast carcinoma cell line (MDA-MB231) to represent another type of tumor cell to determine whether the effect of COUP-TFII was restricted to lung cancer. The HIV-1-derived TAT protein is used in this system to mediate cell entry of the fusion protein (35). Immunocytochemistry analysis was performed with an anti-HA antibody, and both fusion proteins were found in the nucleus (data not shown; Supplementary Fig. 1). We found a significant (P < 0.05) increase in the invasive and migratory ability of all of the tumor cells, A549, H520, H441, and MDA-MB231 (∼5-, 22-, 18-, and 3-fold in invasion; ∼3-, 4-, 12-, and 2-fold in migration; Fig. 2, G and H) transduced with the purified COUP-TFII protein, compared with controls (nontransduced or transduced with TAT-HA-Cre).

To investigate the role of COUP-TFII in regulation of human lung cancer tumorigenicity, the clonogenicity of the cells was measured in semi-solid agarose plates. We found an increase of >100% in the size of the colonies formed by COUP-TFII-transfected A549 cells (Fig. 3, A and B). In agreement with this finding, we also detected an 89% reduction in the size of the colonies formed by COUP-TFII antisense-transfected H460SM cells (Fig. 3, C and D). For both cell types the number of colonies relative to control cells changed slightly (Fig. 3, B and D).

An increase in anchorage-independent growth and tumorigenicity requires changes in the organization of actin filaments (36). Our analysis of actin filaments showed a change in distribution of actin filaments; we observed more stress fibers in the COUP-TFII-transfected A549 cells than controls (Fig. 4,A). Because phosphorylation of p125-FAK is linked to the regulation of cytoskeletal reorganization, cell spreading, and migration, it was of interest to examine changes in FAK phosphorylation. We found that FAK phosphorylation was increased in COUP-TFII-transfected A549 cells (T.P.3 and clone 1) when they were cultured on fibronectin-coated plates (Fig. 4,B), whereas the total FAK levels were comparable with that in nontransfected and COUP-TFII-transfected A549 cells (Fig. 4 C). Interestingly, in H460SM cells expressing COUP-TFII antisense RNA, no changes were observed in the distribution of actin filaments. However, the cells acquired a more rounded morphology, which was particularly apparent when they were seeded on Matrigel-coated culture plates (data not shown; Supplementary Fig. 2).

Because COUP-TFII is known to play an important role in angiogenesis (37), we investigated whether COUP-TFII can regulate the expression and activity of ECM-degrading proteinases involved in angiogenesis, such as MMP-2. We analyzed the activity of MMP-2 in COUP-TFII-transfected A549 cells (T.P.3 and clone 1) and found a marked increase in the level of MMP-2-mediated gelatinolytic activity, as assessed by zymography (Fig. 5,A). The increase in MMP-2 activity could be due to an increase in MMP-2 expression. Therefore, we measured MMP-2 levels in conditioned media of COUP-TFII-transfected A549 cells. Using Western blot analysis we found an increase in the production of pro-MMP-2 (72 kDa) in COUP-TFII-transfected A549 cells (Fig. 5,B). This was not due to differences in TIMP-2 or MT1-MMP synthesis in these cells, as revealed by Western blotting using the respective antibodies (data not shown). Moreover, in COUP-TFII-antisense expressing cells (T.P.2 and clone 20), a noticeable reduction in MMP-2 activity level relative to controls was observed (Fig. 5 C). In addition, H520 and H441 cells transduced with the recombinant TAT-HA-COUP-TFII protein showed higher pro-MMP-2 activity levels compared with control or nontransduced cells (data not shown; Supplementary Fig. 3).

To investigate the effect of COUP-TFII on plasmin production, casein zymography was performed on conditioned media derived from COUP-TFII-transfected A549 and COUP-TFII-antisenseexpressing H460SM cells. Two caseinolytic bands (81–83 kDa) were observed in COUP-TFII-transfected A549 cells (clone 1 and very weak bands observed for T.P.3; data not shown; Supplementary Fig. 4A). Conversely, in COUP-TFII antisense-transfected cells (T.P.2, clone 20), NCI-H460, the two caseinolytic bands could not be observed (data not shown; Supplementary Fig. 4C). These caseinolytic activities could be blocked by the plasmin inhibitor amino-n-caproic acid (data not shown; Supplementary Fig. 4, B and D). The levels of MT1-MMP and TIMP-2 protein were analyzed by Western blotting, and no difference was observed (data not shown).

To determine whether antisense COUP-TFII transfection altered uPAR function in H460SM cells, we analyzed cell surface receptor huPAR expression and synthesis by Northern blotting and flow cytometry. Northern blot analysis on H460SM cells revealed a 33% or 66% reduction in the steady-state mRNA levels of huPAR in T.P.2 and clone 20 cells, respectively (Fig. 6,A). The reduction in huPAR synthesis was additionally confirmed by immunocytofluorometry using a monoclonal antibody against huPAR (Fig. 6,B). There was no difference in the number of the human uPAR in the COUP-TFII-transfected A549 cells using immunocytofluorometry (data not shown). We found a 1.5-fold increase in the steady-state level of huPA mRNA in T.P.3 and a ∼2.5-fold increase in clone 1 (Fig. 6 C).

We demonstrated that cancer cell invasiveness correlates with COUP-TFII expression in several lung cancer cell lines as well as in a breast carcinoma line. Stable expression of COUPTFII in A549 cells increased the motility and invasiveness of the tumor cells. Transfection of a highly invasive human lung cancer cell line (H460SM) with a vector expressing COUP-TFII antisense cDNA reduced migration and invasion, providing negative as well as positive evidence for the involvement of COUP-TFII in invasion. In accordance with our analyses of stable cell lines, we found noticeable increases in the invasive and migratory ability of A549 cells and other human lung cancer cells transduced with TAT-COUP-TFII fusion protein.

The increase in the anchorage-independent growth of COUP-TFII-transfected A549 cells coincided with a change in the distribution of actin filaments (36). Interestingly, no difference was seen in the distribution of actin filaments in COUP-TFII antisense-transfected H460SM cells (data not shown). However, their morphology was altered, as they appeared as typical polygonal, rounded epithelial cells and lost the more elongated, mesenchymal appearance of the parental cells. This pattern of morphology has been observed when tumor cells are invasive and they start to invade through junctional margins by extending pseudopodia-like cytoplasmic processes (38).

The changes observed in the actin filaments or the morphology of the cell prompted us to examine the phosphorylation of p125-FAK. Activation of FAK, overexpressed in several human cancers, induces survival, proliferation, and motility of cells in culture (32). In our study, FAK phosphorylation of Tyr³⁹⁷ in COUP-TFII-transfected A549 cells was increased, mainly in cells grown on fibronectin-coated plates. It is conceivable that, after integrin-mediated cell adhesion (there is similar levels of integrin β1 production in both COUP-TFII-transfected and nontransfected A549 cells; data not shown), FAK undergoes tyrosine phosphorylation that eventually leads to cytoskeletal disorganization. Whether distinct effector proteins (such as Rho GTPases) are involved in the reorganization of stress fibers in COUP-TFII-transfected cells is our future work.

The functional importance of FAK activation in human tumor growth in vivo has been elucidated to be dependent on uPAR-integrin β1 association (39). Also, binding of the uPA to its receptor (uPAR) is involved in the activation of MMP-2 (20). The importance of tumor-associated proteases in invasion and metastasis has been demonstrated for a variety of solid malignant tumors (40). Therefore, we examined changes in tumor-associated proteases in COUP-TFII-transfected cells. The increase in activity and production of MMP-2 in COUP-TFII-transfected A549 cells encouraged us to investigate whether non-MMP proteinases were involved in the activation of MMP-2. There is compelling evidence to indicate that cell migration and invasion depend on the coordinated enzymatic activities of metallo- and serine proteinases (40). The mechanism of pro-MMP-2 activation has been described in association with the uPA-plasmin cascade, whereby uPA binds uPAR, leading to the conversion of plasminogen to plasmin (41). The appearance of plasmin in the conditioned medium suggests that plasminogen/plasmin conversion can still occur on the surface of the cells, but the enzyme may be rapidly released (42). Alternatively, it is possible that plasminogen activation occurs in the conditioned medium after it is released from the cell surface (43). It should be noted that uPAR and the uPA/uPAR complex have also been identified as receptors for the ECM protein vitronectin (44), whereas plasmin can mediate cell detachment from the ECM (45). Changes in extracellular plasmin levels could be the result of changes in huPAR expression and plasminogen/plasmin conversion. We noticed a decrease in both the expression and synthesis of huPAR after transfection of H460SM cells with COUP-TFII antisense. Although the number of huPAR receptors in A549 cells is low, the high level of huPA expression in the COUP-TFII-transfected A549 cells may be involved in the regulation of factors downstream of the uPA/uPAR complex, leading to the elevated expression of plasmin (which, in turn, can activate MMP-2).

FAK is associated with integrin β1 regardless of its state of activation, but only in the presence of high levels of uPAR is integrin β1-associated FAK phosphorylated (39). It is conceivable that a high level of uPA/uPAR complex on the cell surface clusters integrins and that ligand-induced conformational change increases the proximity between FAK molecules leading to their trans-phosphorylation (46). Other reports also show that uPAR blocking can down-regulate the signaling pathway mediated by integrin and induce cancer dormancy in vivo (47). On the basis of the results of our study, we propose that the regulation of huPA/huPAR expression by COUP-TFII promotes the association of uPA/uPAR complex with integrin β1, leading to FAK phosphorylation as well as MMP-2 synthesis and activation.

We thank Anan Wang for technical assistance, and Dr. David R. Koehler and Deanna Yaniw for reviewing the manuscript.