Cancer Cell Biology
Calcitonin stimulates the secretion of urokinase-type plasminogen activator from prostate cancer cells: Its possible implications on tumor cell invasion
Version of Record online: 27 DEC 2005
Copyright © 2005 Wiley-Liss, Inc.
International Journal of Cancer
Volume 118, Issue 11, pages 2694–2702, 1 June 2006
How to Cite
Sabbisetti, V., Chigurupati, S., Thomas, S. and Shah, G. (2006), Calcitonin stimulates the secretion of urokinase-type plasminogen activator from prostate cancer cells: Its possible implications on tumor cell invasion. Int. J. Cancer, 118: 2694–2702. doi: 10.1002/ijc.21625
- Issue online: 14 MAR 2006
- Version of Record online: 27 DEC 2005
- Manuscript Accepted: 7 SEP 2005
- Manuscript Received: 2 APR 2005
- NIH. Grant Number: CA96534
- prostate cancer;
Calcitonin (CT) is synthesized and secreted in prostate epithelium, and its secretion from malignant prostates is several folds higher than that in benign prostates. CT receptor (CTR) is expressed in malignant prostate epithelium, and its activation increases invasiveness of prostate cancer (PC) cells via activation of protein kinase A. Since the role of urokinase-type plasminogen activator (uPA) in invasion of PC has been established, we tested the hypothesis that CT increases invasion of PC cells by stimulating uPA secretion from PC cells. Exogenously added CT stimulated the secretion of uPA from PC-3M cells in a dose-dependent manner, which was blocked by Rp.cAMP, a competitive inhibitor of protein kinase A. CT stimulated the secretion of MMP-2 and MMP-9 from PC-3M cells, and also increased their invasiveness. Both these actions of CT were blocked by uPA-neutralizing antibodies. Immunofluorescence studies with PC-3M cells suggest that CT stimulated redistribution of cellular uPA to focal adhesion sites, which was further confirmed by co-immunoprecipitation of uPA with focal adhesion kinase (FAK) in response to CT. These results suggest that CT increases invasiveness of PC cells by stimulating PKA-mediated uPA secretion and by redirecting the secreted uPA to focal adhesion sites. The results also suggest that uPA may, at least in part, mediate proinvasive actions of CT on PC cells by stimulating the secretion of gelatinases and degradation of focal adhesion sites. © 2005 Wiley-Liss, Inc.
Prostate cancer (PC) is the most commonly diagnosed form of cancer among men in the United States and is second only to lung cancer as a cause of cancer-related deaths.1, 2, 3 Despite recent progress in PC detection and treatment, most of the deaths from PC occur in patients with androgen-independent metastatic disease.1, 4 Cancer metastasis is a complex multistep process involving tumor cell proliferation, loss of adhesion, cell migration, invasion, angiogenesis and cell implantation in distant organs.5, 6, 7 Once PC becomes metastatic, its treatment options are limited to androgen withdrawal. In this environment, the cells often develop an androgen-independent state resulting in patient demise.8, 9
There is evidence of high correlation between tumor growth, increasing tumor grade, neuroendocrine differentiation and microvessel density.7, 10, 11, 12 Primary PC cells have been shown to secrete several neuropeptides, including bombesin, neurotensin, vasoactive intestinal peptide and calcitonin (CT).13 The role of neuropeptides, growth factors and proteases in the acquisition of hormonal independence and the underlying molecular mechanisms involved in this process remain poorly understood.13, 14, 15, 16, 17 Previous studies from this laboratory have shown that CT and its receptors are expressed in primary prostate tumors, and the level of their expression positively correlates with tumor grade.18 CT and CT receptors (CTRs) have also been identified in several established PC cell lines such as LNCaP, PC-3, DU-145 and PC-3M.19, 20, 21, 22 Exogenously added CT stimulates the proliferation of LNCaP cells by stimulating cAMP accumulation and increasing cytoplasmic Ca2+ transients, and stimulates invasion by activating protein kinase A-mediated mechanisms.19, 23
Urokinase-type plasminogen activator (uPA), a member of the serine protease family, is strongly implicated as a promoter of tumor progression in various human malignancies, including breast cancer and PC.24 The ability of uPA to break down various components of the extracellular matrix, including laminin, fibronectin and collagen, through the activation of plasminogen to plasmin may facilitate the invasion of tumor cells.16, 25, 26 Although uPA is produced by normal and benign hyperplastic prostate, elevated levels of uPA are observed in patients with PC.27 The expression of human uPA gene has been shown to be under the regulation of various growth factors and cytokines.28, 29 However, the role of neuroendocrine peptides such as CT in the regulation of uPA production in PC has not been examined.
In the current study, we examined the regulation of uPA production by CT in androgen-resistant and highly aggressive PC-3M PC cell line, which displays several characteristics of aggressive PCs such as highly metastatic behavior, androgen-independence and co-expression of CT, CTR and VEGF. The potential actions of uPA in stimulating invasiveness of PC-3M cells were also investigated.
Material and methods
Fibrinogen and plasminogen were purchased from EMD Biosciences, San Diego, CA (catalogue no. 341576 and 528175, respectively). Monoclonal antibody against B chain of human uPA was purchased from American Diagnostica, Stanford, CT (product no. 39401). Purified mouse antifocal adhesion kinase (FAK) and rabbit antiphospho paxillin antibodies were purchased from BD Biosciences Pharmingen, San Diego, CA (catalogue no. 611806) and Cell Signaling, Beverly, MA, respectively. Amido black was purchased from Bio-Rad, Hercules, CA (catalogue no. 161-0402).
PC-3M PC cell line was kindly provided by Dr. Isiah Fidler (Anderson Cancer Center, Houston, TX). The cells were maintained in the complete medium (RPMI 1640 medium supplemented with 5% fetal calf serum and 15% horse serum, 100 IU/ml penicillin G and 100 μg/ml streptomycin) under standard culture conditions.
Plasmids and generation of PC-3M variant cell lines
Overexpression of CT
A full-length human CT cDNA was cloned from total mRNA extracts of primary prostate tumor specimens by reverse transcription-PCR, as previously described.18 The cloned cDNA was inserted into mammalian expression vector pcDNA3.1 (Research Genetics) in sense direction. The sense orientation of the insert in the cloned plasmid was confirmed by sequencing at the institutional DNA sequencing facility.
CT cDNA-pcDNA3.1 construct was transfected into PC-3M and LNCaP cell lines using FuGene 6 transfection reagent (Roche, Indianapolis, IN). G418-resistant colonies were selected after 3 weeks. Cells derived from individual colonies after transfection and drug selection were obtained by cloning cylinders and further expanded. Total RNA was prepared from transfectants derived from the control vector-transfected group (PC-3M-v) as well as the CT expression vector group (PC-3M-CT+) and was screened for CT mRNA abundance by S1-nuclease assay as previously described.18 Conditioned media were collected from the cultures of both control vector-transfected and CT expression vector-transfected groups, and analyzed for secretory CT levels, using a specific CT RIA.30
Silencing of CT
Two strands of 5 templates encoding each ribozyme against CT cDNA having a BamH1 and HindIII site at the 3′ and 5′ ends, respectively, were synthesized as described below at the Genemed Biosynthesis, (San Francisco, CA).
R4: 5′-GAAGATCTTC/CAAGGGCAG/TTTCGTCCTCACGCACTCATCAG/ATCTGGCT/CCGCTCGAGCGG-3′; 5′-CTTCTAGAAG/GTTCCCGTCA/AAAGCAGGATGCCGTGAGAGTC/TAGACCGA/GGCGAGCUCGCC-3′. R5: 5′-GAAGATCTTC/AGCTTCTAG/TTTCGTCCTCACGCACTCATCAG/ATCTGGCT/CCGCTCGAGCGG-3′; 5′-CTTCTAGAAG/TCGAAGATC/AAAGCAGGATGCCGTGAGAGTC/TAGACCGA/GGCGAGCUCGCC-3′.
The oligonucleotides were purified by polyacrylamide gel electrophoresis, annealed and cloned into BamH1 and HindIII sites of ptv5 vector.31 The vector is a pUC-based plasmid carrying a U6 polIII promoter driving the transcription of the ribozyme. The transcription is terminated by a polIII polymerase termination signal (4 thymidine residues) to generate ribozymes with very little extraneous sequences. Each ribozyme was co-transfected with plasmid pcDNA3.1-zeocin into PC-3M cells, and zeocin-resistant colonies were selected after 4 weeks. As a control, vector ptv5 lacking the ribozyme template was transfected to generate PC-3M-v2 cell line. The clones were then screened for CT mRNA expression and CT secretion by S1-nuclease assay and specific RIA, respectively.18
Secretion of uPA by PC-3M cells: uPA antigen assay
PC-3M cells were plated in 6-well culture dishes at a density of 1 × 105 cells/well in the complete RPMI 1640 medium (supplemented with 5% fetal calf serum and 15% horse serum, 100 IU/ml penicillin G and 100 μg/ml streptomycin) for 24 hr. The medium was then replaced with serum-free basal RPMI1640 medium (containing no serum but 0.3% BSA, 10 mM HEPES, 10 μM bacitracin, 100 IU/ml penicillin G and 100 μg/ml streptomycin). The cells were then treated with various concentrations of CT or other agents for 24 hr as described in the Results section. The aliquots of the conditioned media were then collected and assayed for antigenic uPA. A commercial kit (IMUBIND uPA ELISA kit, American Diagnostica) was used and the manufacturer's suggested protocol was followed. The unconditioned serum-free basal medium was used as the blank. Results were expressed as mean ± SEM in ng/106 cells of triplicate experiments.
Enzymatic activity of secreted uPA: Fibrin zymography
Biological activity of uPA secreted by PC-3M cells in the conditioned medium was determined by zymography, as previously described.32, 33 Conditioned media were loaded in equal amounts of protein (1 μg/well) onto 10% SDS polyacrylamide gels containing fibrinogen and plasminogen (10 mg/ml and 50 μg/ml, respectively) as substrates.34 Following electrophoresis, the gel was washed with 2.5% (v/v) Triton X-100 for 2 times of 30 min each, and was incubated in glycine buffer (0.1 M, pH 8.0) for 16–22 hr at 37°C. The gel was first stained with 0.1% Amido black (in 25% v/v methanol, 10% v/v acetic acid in water), and then destained with (25% v/v methanol, 10% v/v acetic acid in water without Amido black). After destaining, the gel displayed a uniform background except in regions where uPA had migrated and cleaved its substrates. The uPA enzymatic activity was determined by densitometric analysis of fibrin zymograms. Two sets of controls that include basal incubation medium without incubation with PC-3M cells and purified uPA (American Diagnostica) were run in parallel with the samples. In addition, a separate but concurrent gel was run in the absence of plasminogen substrate.
Cellular uPA: Western blot analysis
For determining the content of cell-bound and intracellular uPA immunoreactivity, PC-3M cells (2 × 106 per 100 mm dish) were washed in PBS and lysed on ice under nonreducing condition in a 50 mM Tris buffer (pH 7.4, containing 1% NP-40, 0.25% Na-deoxycholate, 1 mM EDTA, and freshly supplemented with 1 mM PMSF, 1 μg/ml leupeptin, 1 μg/ml aprotinin, 1 μg/ml pepstatin, 1 mM Na3VO4 and 1 mM NaF) for 10 min. Equal amounts of proteins (∼5 mg/ml) were separated on a 12% sodium dodecyl sulfate-polyacrylamide gel (SDS-PAGE) and blotted onto a polyvinyl difluoride membrane (Bio-Rad Laboratories). Membranes were blocked in a Tris-buffered saline solution with 5% nonfat dry milk and incubated with monoclonal uPA antibody (10 μg/ml, American Diagnostica) overnight at 4°C. Immunoreactive signals were detected by incubation with horseradish peroxidase-conjugated secondary antibodies (Santa Cruz Biotechnology, Santacruz, CA) followed by chemiluminescent detection (ECL detection system, Radiochemical Center, Amersham). The antibodies were stripped from the membrane, and the blot was reprobed with mouse α-tubulin antibodies to determine α-tubulin content.
uPA and invasion of PC-3M cells: In vitro invasion assay
These experiments were conducted in 24-well, 2-compartmented, MatrigelTM invasion chambers (Becton Dickinson, Bedford, MA). Exponentially growing PC-3M cells were serum-starved for 24 hr with basal RPMI medium containing no serum or growth factors (but containing 0.1% BSA, 10 mM HEPES, 4 mM L-glutamine, 100 IU/ml penicillin G and 100 μg/ml streptomycin). The cells were then harvested, and seeded at a density of 25 × 103 cells/well in the upper insert of the Matrigel invasion chamber. The lower chamber received the chemoattractant medium, which consisted of 90% complete medium and 10% conditioned medium from the cultures of PC-3 M cells expressing constitutively active Gsα protein.35 The cells also received CT and uPA antibodies as described in the Results section, which were added to the insert. The incubations were carried out for 24 hr, after which the Matrigel (along with noninvading cells) was scraped off with cotton swabs, and outer side of the insert was fixed and stained using Diff Quik staining kit (Dade Behring Diagnostics, Aguada, PR). The number of cells migrated on the outer bottom side of the insert was counted under the microscope in 6 or more randomly selected fields (magnification: 100×). The final results were expressed as mean number of invaded cells ± SEM per 100× field. Each experiment was done in triplicates, and the experiment was repeated twice.
Since PC-3M cells exhibit high proliferation rate, it is likely that the cells migrated during early part of the 24 hr incubation period could proliferate during the remaining period of incubation, leading to a slight overestimation of the final results. To correct this probability, we determined the growth rate of cell lines under identical culture conditions. About 25 × 103 cells were plated at hourly intervals in 6-well dishes and cultured with/without CT (50 nM) for1–24 hr. Mean percent increase in the cell number was determined at the end of the incubation period by counting the net increase in the number of cells. The relative CT-induced increase of the pooled results of all time points was found to be 1.19 (vehicle control = 1). This correction was applied to the results of invasion assays.
Effect of uPA on MMP activation: Gelatin zymography
Conditioned media of PC-3M cell cultures were concentrated approximately 8 times with YM-30 centricon ultrafilters (Millipore, Bedford, MA). The concentrated conditioned media of PC-3M cultures treated with or without CT and/or uPA antibodies were analyzed for MMP gelatinolytic activities by gelatin zymography as described previously.36 Briefly, the concentrated conditioned media (4 μg protein per sample) were fractionated by SDS-PAGE on an 8% gel containing 1.0 mg/ml gelatin (Sigma, St. Louis, MO) under nonreducing conditions. After 2 washes in Tris buffer (50 mM Tris, 200 mM NaCl, 10 mM CaCl2, 1 mM ZnCl2, 1% Triton X-100, pH 7.5), the gel was incubated in the same buffer in the absence of Triton X-100 for 18 hr at 37°C. After being stained with Coomassie Brilliant Blue R-250, the gel was destained with 10% acetic acid/25% MeOH (v/v), and clear bands resulting from the digestion of the substrate by gelatinase enzymes were visualized.
Localization of uPA in PC-3M cells: Immunofluorescence
PC-3M cells were plated in multichamber cell culture slides in the complete medium for 20 hr. The medium was then replaced with serum-free medium for additional 24 hr. The cells were then treated with or without 50 nM CT for several incubation periods up to 60 min. Then cells were fixed with −20°C methanol for 20 min. After 3 washes with PBS, the cells were incubated at 25°C with monoclonal mouse anti-human urokinase IgG (American Diagnostica) and rabbit antiphospho paxillin IgG (Cell Signaling) in blocking solution (PBS, 0.3% Triton X-100, 10% goat serum) overnight at 4°C. The cells were then washed 3 times with PBS, and incubated with anti-mouse IgG-Alexa 488TM conjugate (for uPA) and TRITC-conjugated anti-rabbit IgG (for phospho paxillin) for 1 hr in dark at room temperature. Controls with either nonimmune goat IgG or no primary antisera were used in all studies. After 3 final washes in PBS, the slides were mounted in PBS containing 3% n-propyl-gallate (Sigma) and 25% glycerol. Digital photographs were taken with Retiga 1300 camera connected to a Nikon Optiphot-2 microscope equipped for epifluorescence. Co-localization of individual images of the same field at different wavelengths was performed using IP LabTM image analysis program.
Immunoprecipitation and immunoblotting
Serum-starved PC-3M cells on 100 mm dishes (2 × 106 per 100 mm dish) were stimulated with 50 nM CT for the indicated times as described in the Results section. The cells were then washed with cold PBS, lysed at 4°C for 10 min in 500 μl modified lysis buffer (50 mM Tris buffer (pH 7.4), containing 1% NP-40, 0.25% Na-deoxycholate, 1 mM EDTA, and freshly supplemented with 1 mM PMSF, 1 μg/ml leupeptin, 1 μg/ml aprotinin, 1 μg/ml pepstatin, 1 mM Na3VO4 and 1 mM NaF), scraped and collected, diluted with 500 μl of HNTG buffer (50 mM HEPES (pH 7.4), 150 mM NaCl, 0.1% Triton X-100, 10% glycerol), sheared by a passage through 22-gauge needle, and precleared by incubation with agarose beads.
The cell lysates for the IP contained ∼300 μg of total cell protein. Antibodies to FAK (2 μg of IgG) were incubated with lysates for 1 hr at 4°C and immunoprecipitates were collected with 10 μl of Protein A-agarose bead slurry (Pharmacia).
Antibody-complexed proteins were washed at 4°C with Triton X-100 lysis buffer, followed by washes with HNTG buffer and were separated by SDS-PAGE. Equal amounts of protein (∼5 mg/ml) were separated on a nonreducing 12% SDS-PAGE and blotted onto a polyvinyl difluoride membrane (Bio-Rad Laboratories). Membranes were blocked in a Tris-buffered saline solution with 5% nonfat dry milk and incubated with monoclonal uPA antibody (10 μg/ml) overnight at 4°C. Immunoreactive signals were detected by incubation with horseradish peroxidase-conjugated secondary antibodies (Santa Cruz Biotechnology) followed by chemiluminescent detection (ECL detection system, Radiochemical Center, Amersham). The antibodies were stripped from the membrane, and the blots were reprobed with mouse anti-FAK, and total FAK contents were determined.
CT stimulates the secretion of uPA by PC-3M cells: Rp.cAMP attenuates CT action
CT significantly stimulated the secretion of immunoreactive uPA by PC-3M cells in a dose-dependent fashion (Fig. 1a). The increase was at least 3-fold in response to 10 nM CT (the lowest tested concentration), and increased to 7-fold when treated with 1 μM CT (the highest tested concentration). Since our previous studies have shown that CT increases invasiveness of various PC cell lines through the activation of protein kinase A,37 we tested a possibility that CT may stimulate uPA secretion by the same mechanism. We examined the effect of Rp.cAMP (100 μM), a membrane-permeable competitive inhibitor of cyclic AMP-dependent protein kinase A, on CT-induced uPA secretion by PC-3M cells. The results of Figure 1a have shown that Rp.cAMP significantly inhibited basal uPA secretion, and also abolished CT-induced increases at all concentrations.
8-Bromo-cAMP mimics the effect of CT on uPA secretion
To further confirm the role of cAMP signaling pathway in the regulation of uPA secretion by PC-3M cells, we examined the effect of 8-bromo-cAMP, a membrane-permeable analog of cAMP, on uPA secretion. 8-Bromo-cAMP also stimulated PC-3M cells to increase uPA secretion in a dose-dependent manner. The increase caused by 1 mM of 8-bromocAMP was comparable to that by 100 nM CT (Fig. 1b).
Measurement of uPA activity by fibrin zymography
Since uPA is secreted in multiple isoforms, including proenzymes, we wanted to test whether CT-stimulated uPA is biologically active. We examined this by analyzing proteolytic activity of secreted uPA by fibrin zymography. The basal incubation medium was used as the blank (lane 0 of Fig. 1c), and standard uPA solution (10 ng/lane; lane 1 of Fig. 1c) was used as the positive control. A parallel gel that lacked only plasminogen was also run under similar conditions, and did not display any uPA bands (data not shown). The results of Figure 1c show that biologically active uPA was present in 3 major molecular sizes in the conditioned media of vehicle-treated PC-3M cells (lane 2 of Fig. 1c), and the intensity of all 3 bands increased in response to increasing CT concentrations (lanes 3–5 of Fig. 1c). Densitometric scanning of the low-molecular-weight band (52 kDa, as suggested by the standard uPA in lane 1) showed a significant and dose-dependent increase in response to CT at the concentrations of 50 and 100 nM (p < 0.001). Higher molecular weight bands also displayed a similar profile, suggesting these bands may represent bound forms of uPA.
Modulation of CT expression in PC-3M cells alters uPA secretion
Considering that tumor cells of primary PCs secrete CT and uPA and their amounts may vary with their tumor grade,18 we modulated endogenous CT expression of PC-3M cells by stably transfecting either constitutively active CT vector (PC-3M-CT+) or anti-CT ribozyme (PC-3M-CT−) as described recently.38 The controls for these 2 cell lines expressed either the empty vector pcDNA3.1 (PC-3M-v) or inactive ribozyme (R4). The effect of these constructs on CT expression was analyzed by testing (i) CT mRNA abundance by S1-nuclease assay; and (ii) the secretion of CT in the conditioned media by a specific CT RIA. The results of Figure 2a show that PC-3M-CT+ cells displayed markedly higher abundance of CT mRNA than PC-3M-v. In contrast, PC-3M-CT− cells displayed a marked decline. As expected, control cell lines expressing either empty pcDNA3.1 vector (PC-3M-v) or inactive ribozyme (R4) contained the equivalent levels of CT mRNA. Since all samples contained equivalent amounts of β-actin mRNA, the changes in CT mRNA abundance are specific.
The changes in CT expression of PC-3M sublines were further confirmed by analyzing the secretion of CT. PC-3M-CT+ cells secreted over 3-fold greater levels of CT as compared to PC-3M-v (Fig. 2b). In contrast, CT levels were undetectable in the conditioned media of PC-3M-CT− cells. As expected, the secretion of CT by R4 cells was similar to that of PC-3M-v.
To examine the effect of modulation of CT expression on uPA secretion, we analyzed the conditioned media of these cell lines for uPA activity by fibrin zymography. The results demonstrate that PC-3M-v and R4 cells secreted closely similar amounts of active uPA (lanes 2 and 3 of Fig. 2c). However, the uPA activity in the conditioned medium of PC-3M-CT+ cells was greater than that of PC-3M-v or R4 media by ∼2-fold (lane 4 of Fig. 2c). In contrast, the conditioned medium of PC-3M-CT− cells contained almost undetectable uPA activity (lane 5 of Fig. 2c). Basal incubation medium displayed no uPA activity and standard uPA displayed 1 band of 52 kDa size (lanes 1 and 2 of Fig. 2c).
Cell-bound uPA in PC-3M cell lysates
The measurement of uPA activity in the conditioned media provides the estimate of secreted uPA activity. However, only cell-bound or intracellular uPA can act on the cells to alter their characteristics such as motility or adhesion properties. Therefore, we examined uPA activity in PC-3M cell lysates by Western blot analysis, after treating the cells with CT. Serum-starved PC-3M cells were treated with various concentrations of CT for 24 hr, lysed, and the lysates were processed for Western blot analysis using monoclonal antibody against human uPA B-chain. It can be seen from Figure 3 that the antibody recognized 3 major bands of approximately 140, 52 and 50 kDa. According to the present evidence, 52-kDa bands correspond to high-molecular-weight or active form of uPA, which has two (A and B) chains and 3 linked domains, such as the aminoterminal domain, the kringle domain and aminoterminal domain.39, 40 The results suggest that CT increased the immunoreactivity of this band in a biphasic manner with a maximal increase at 50 nM CT. The profile of these results is almost identical to that of CT-induced increase in invasion, as described previously.23 A second major high-molecular-weight band (140 kDa) may correspond to PAI-1-uPA-uPA receptor complex as suggested by a recent report.40
Effect of CT on invasion of PC-3M cells: Mediatoryrole of uPA in CT action
We have shown that CT increases the invasiveness of several PC cell lines.23 Since CT also stimulates the secretion of active uPA in PC-3M cells, we tested whether proinvasive actions of CT are mediated by uPA. We tested this by examining the effect of CT (50 nM, maximal effective dose23) on the invasion of PC-3M cells in Matrigel in the presence or absence of uPA neutralizing antibodies (50 ng/ml). It can be seen from Figure 4a that the tested dose of uPA antibody did not affect the invasiveness of untreated PC-3M cells, but abolished CT-induced increase.
The effect of modulation of CT expression and the subsequent modulation of uPA secretion on invasiveness was tested in PC-3M sublines. R4 and PC-3M-v (2 carrier control cell lines) exhibited similar invasiveness in Matrigel (Fig. 4b). Overexpression of CT in PC-3M-CT+ cells led to an increase in uPA secretion (Fig. 2c) and a 2-fold increase in invasiveness (Fig. 4b). Again, PC-3M-CT− cells, which did not secrete CT or uPA, were almost noninvasive.
Role of uPA in CT-induced activation of gelatinolytic activity
To examine the potential role of uPA in CT-induced increase in gelatinolytic activity, we treated the cells with CT in the presence or absence of uPA-neutralizing antibody for 24 hr. Zymography of the conditioned media demonstrate that uPA antibody by itself did not have any significant effect on basal activity of MMP-2 (72-kDa band in lane 2 of Fig. 5). CT (50 nM) markedly increased MMP2 activity (72-kDa band in lane 3 of Fig. 5). However, CT could not increase MMP-2 activity in the presence of uPA antibodies (72-kDa band in lane 4 of Fig. 5).
The profile of MMP-9 was more intriguing. uPA antibody had a stimulatory effect on MMP-9 activity (90-kDa band in lanes 1–2 of Fig. 5). CT had a potent stimulatory effect on MMP-9 activity (90-kDa band in lane 3 of Fig. 5). However, CT action was completely reversed by the presence of uPA antibodies (lane 4 of Fig. 5).
Effect of CT on redistribution of uPA in focal adhesion sites
To examine the potential role of uPA in migratory invasive action, we examined the presence of uPA and its potential co-localization with paxillin, a component of focal adhesion complex.41, 42 PC-3M cells were grown in chamber slides, serum-starved and treated with 50 nM CT for various times (0–60 min). The cells were then processed for uPA-paxillin double immunofluorescence. Figure 6 shows vehicle- and CT-treated cells after 15 min of treatment. Vehicle-treated cells tightly adhered to the basement membrane as characterized by the stretched cytoplasm and elongated shape of the cell (left panels of Fig. 6). Moreover, paxillin, a component of focal adhesion sites, was randomly distributed throughout the stretched cytoplasm, with higher concentrations around nuclear core and edges of the cell (left paxillin panel of Fig. 6). uPA was also uniformly diffused throughout the stretched cytoplasm (left uPA panel of Fig. 6). Merging of uPA and paxillin images suggest little co-localization of these 2 components except perhaps a narrow stretch surrounding the nuclear core (left merge panel of Fig. 6). The effect of CT on redistribution of focal adhesion sites was further examined by FAK staining. The results of left FAK panel show that vehicle-treated PC-3M cells displayed stretched morphology with numerous FAK-positive sites concentrated throughout the stretched cytoplasm.
Treatment with CT altered the shape of cells within 5 min (the earliest time point examined, data not shown). Right panels of Figure 6 present the changes induced by 50 nM CT in PC-3M cells after 15 min of incubation. The cells acquired a rounded shape (top right panel of Fig. 6). There was also redistribution of focal adhesion sites (right paxillin panel of Fig. 6). An increase in the intensity of uPA staining was observed throughout the cytoplasm along with the marked decrease in paxillin staining, particularly along the edges of the cell (right uPA panel of Fig. 6). Merging of these 2 images revealed increased co-localization of uPA with paxillin in the cytoplasm (right merge panel of Fig. 6). Treatment with CT resulted in a dramatic decline in the density FAK-positive sites in the cytoplasm (right FAK panel). Interestingly, the fewer remaining FAK sites were concentrated along the 2 edges of the cells, a typical characteristic of a migrating cell.
To confirm the finding of microscopy that CT induces redistribution of uPA to focal adhesion sites, we performed immunoprecipitation of lysates of CT-treated PC-3M cells with FAK antibodies. The immunoprecipitates were then separated by SDS-PAGE electrophoresis, transferred on the blot, and the blots were immunoprobed with either uPA or FAK antibodies. The results presented in Figure 7 show that FAK immunoprecipitates lacked uPA immunoreactivity in untreated PC-3M cells. However, CT treatment resulted in the co-immunoprecipitation of uPA with FAK within 5 min, and peaked at 10 min. However, a significant amount of uPA was still co-localized with FAK at 60 min, the longest time of incubation examined. Total FAK immunoreactivity at all time points were relatively constant, suggesting co-immunoprecipitation of uPA was a CT-mediated phenomenon (Fig. 7).
The uPA system (uPA-uPA-receptor-PAI) is linked to cellular invasion and migration through its capacity to promote pericellular proteolysis, regulate integrin function and mediate cell signaling in response to uPA binding to its receptor.43 uPA has been shown to break down various components of extracellular matrix, including laminin, fibronectin and collagen through the generation of plasmin.16 In general, tumor cells of several human malignancies express detectable cell surface-associated uPA and plasmin activity, and this positively correlates with their invasiveness.17, 44, 45, 46, 47, 48, 49 Since CT stimulates the invasion of several PC cell lines,23 we examined the actions of CT on the secretion of uPA by immunoassay and fibrin zymography. Present results have shown that CT markedly increased the secretion of uPA in the conditioned media. The results of ELISA as well as fibrin zymography have demonstrated that CT increased uPA secretion in a dose-dependent manner. This was further corroborated by the findings of Western blot analysis, which showed that CT increased the cellular levels of free as well as complexed uPA and both these forms were detected by the uPA antibody. Considering that aggressive PC display elevated expression of CT and its receptors, and persistently secrete several-fold greater amounts of CT,18, 30 we attempted to simulate this phenomenon by upregulating or downregulating CT expression in PC-3M cells using appropriate expression constructs. The results have demonstrated that the upregulation of CT expression in PC-3M cells led to a marked increase in the secretion of uPA activity. In contrast, the downregulation of CT expression caused a dramatic decrease. These results demonstrate that the modulation of CT expression in PC cells can significantly alter their uPA levels. Our recent results suggest that the action of CT on uPA expression may occur at the level of transcription as well as at its secretion. The upregulation of uPA expression by CT, although previously demonstrated in a porcine renal epithelial cell line,50, 51 is a new finding in invasive prostate carcinoma, and this action of CT may play an important role in the progression of PC to an invasive phenotype. In particular, we have shown that the expression of CT and its receptor is restricted to a population of cells in basal compartment of a developing benign epithelium. However, the distribution of CT- and CTR-positive cells in primary prostate tumors is throughout the whole malignant epithelium, and increases with tumor progression.18 Therefore, it is likely that the action of CT on uPA secretion may become pathophysiologically significant in advanced PCs where CT expression is markedly elevated.
In addition to demonstrating stimulatory action of CT on uPA secretion, present results have also shown for the first time that uPA secretion by PC-3M cells is regulated by cAMP signaling cascade. 8-Bromo cAMP stimulated uPA secretion. In contrast, Rp.cAMP dramatically inhibited CT-induced uPA secretion and invasion of PC-3M cells.23 Since CT also produced similar effect on invasiveness of LNCaP cells, it is very likely that the stimulatory effect of CT on invasion of PC cells is not affected by the presence or absence of androgen receptor.22, 23 These results are consistent with several studies in renal cell models that have demonstrated that agents that increase cAMP such as cholera toxin, peptide hormones such as vasopressin and CT, as well as membrane-permeable cAMP analogues, can induce high levels of uPA mRNA expression in LLC-PK1 cell line without the need for new protein synthesis.52, 53, 54, 55, 56, 57
Since CT has been shown to increase the invasiveness of PC-3M cells,23 we wanted to test the role of uPA in CT-induced cell invasion. As expected, CT increased the invasion of PC-3M cells through Matrigel. However, the presence of uPA-neutralizing antibody attenuated this increase, raising a possibility that uPA may mediate some of proinvasive CT actions. This was further demonstrated by the findings that PC-3M subline (PC-3M-CT+) secreting uPA in higher amounts displayed greater invasiveness than PC-3M cells expressing the empty vector. In contrast, PC-3M subline secreting minimal uPA (PC-3M-CT−) lacked the ability to invade Matrigel. Matrix metalloproteinase-9 (MMP-9) and its isoform, MMP-2, can degrade type IV collagen, leading to the degradation of basement membrane and facilitation of the invasion of malignant cells. There is evidence to suggest that cell-bound uPA degrades extracellular matrix to generate and activate plasmin, which in turn, may activate proenzymes of gelatinases such as MMP-2 and MMP-9.40 Present results of gelatinase zymography have shown that CT increased MMP-2 and MMP-9 activities, and this increase was abolished in the presence of uPA-neutralizing antibody, which acts by binding to the B-chain of uPA.24 These results support the possibility that uPA is an important mediator of proinvasive CT actions.
The plasmin cascade driven by uPA has been shown to be spatially and temporally associated with cellular structures that regulate cell adhesion, migration and invasion.25, 26, 58, 59 The secretion of uPA and its binding to the uPA receptor leads to the co-localization of uPA receptors with integrins.25, 60, 61 The uPA receptor then focuses plasmin activation at or near the sites of focal contact between cell surface and extracellular matrix proteins.62, 63 Recent results have shown that vitronectin plays a crucial role in directing the redistribution of integrins and uPA to focal adhesion sites in several cancer cell lines, including PC cell lines.63 Once co-localized to focal points, uPA is thought to regulate cell migration as well as pericellular proteolysis by co-ordinating focal detachment.63 Present results have shown that uPA in CT-treated, but not untreated, cells was co-localized to the proteins of focal adhesion complex such as paxillin, suggesting that CT may direct the activity of uPA to focal adhesion sites. Moreover, the results that CT-treated cells displayed rounded, and not stretched, morphology and fewer focal adhesion sites are consistent with the apparent loss of adhesion, possibly mediated by uPA-induced disassembly of focal adhesion sites. Additional studies will be necessary to delineate the roles for vitronectin and integrins in uPA-mediated proinvasive actions of CT.
Amplification of uPA has been detected in some PCs as well as PC-3 cell line, and elevated uPA levels have been linked with increased aggressiveness and metastasizing ability of prostate tumors.64 Moreover, the inhibition of uPA has been shown to reduce metastasis and tumor growth in animal models.24, 65 Testosterone, which has been shown to reduce invasive ability of PC-3 cell line when transfected with androgen receptor cDNA, also reduced uPA expression.24 These results suggest a co-operative relationship between uPA secretion and invasive ability of PC cells, and the present findings may further explain a potential role of “CT System” in regulating uPA expression, secretion and activity, and consequently, the invasiveness and tumorigenicity of prostate tumors.
In summary, the present results have shown that CT increases the secretion of activated uPA from PC-3M PC cells, and this action of CT may be mediated by cAMP signaling. The results also suggest that uPA may mediate proinvasive actions of CT on PC cells by stimulating gelatinase activity and by decreasing the attachment of cells to the extracellular matrix, possibly by uPA-mediated disassembly of focal adhesion sites.
- 11Association between immunohistochemical expression of vascular endothelial growth factor (VEGF), VEGF-expressing neuroendocrine-differentiated tumor cells, and outcome in prostate cancer patients subjected to watchful waiting. Clin Cancer Res 2000; 6(5): 1882–90., , .
- 42Cytoskeleton-dependent tyrosine phosphorylation of the p130(Cas) family member HEF1 downstream of the G protein-coupled calcitonin receptor. Calcitonin induces the association of HEF1, paxillin, and focal adhesion kinase. J Biol Chem 1999; 274(35): 25093–8., , , .