Reversion of the Invasive Phenotype of Transformed Human Fibroblasts by Anti-Messenger Oligonucleotide Inhibition of Urokinase Receptor Gene Expression Alessandro Quattrone, Gabriella Fibbi, Enrica Anichini, et al. Cancer Res 1995;55:90-95. Published online January 1, 1995.
Updated Version
Citing Articles
E-mail alerts Reprints and Subscriptions Permissions
Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/55/1/90
This article has been cited by 9 HighWire-hosted articles. Access the articles at: http://cancerres.aacrjournals.org/content/55/1/90#related-urls
Sign up to receive free email-alerts related to this article or journal. To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at
[email protected]. To request permission to re-use all or part of this article, contact the AACR Publications Department at
[email protected].
Downloaded from cancerres.aacrjournals.org on July 13, 2011 Copyright © 1995 American Association for Cancer Research
[CANCER RESEARCH 55, 90-95, January 1. 1995]
Reversion of the Invasive Phenotype of Transformed Human Fibroblasts by AntiMessenger Oligonucleotide Inhibition of Urokinase Receptor Gene Expression1 Alessandro Quattrone, (¡abridla Fibbi, Enrica Anichini, Marco Pucci, Alessandra Zamperini, Sergio Capaccioli, and Mario Del Rosso Istillilo di Patologia Generale. Università di Firenze. Viale Morgagni, 50. 50134 Florence. Italy
role in the modulation of cell-substratum adhesion and extracellular matrix destruction (2, 5, 7, 8). We have shown that it is possible to obtain the adhesion contacts of the cells by detaching cell monolayers with the calcium chelator EDTA (5), which leaves on the substrate the so-called SAM (9). The adhesion sites specifically concentrate u-PARs. Here, we have investigated u-PAR distribution on human fibroblasts and their SV40-transformed counterpart and have evalu ated the role of u-PAR versus released u-PA in cell invasivity. We have switched off u-PAR gene expression by using the aODN strat egy. Various experimental models have demonstrated that aODNs are specific inhibitors of gene expression and function using a variety of mechanisms all of which involve aODN base-pairing to the target mRNA (10). AODNs may therefore represent new tools in anticancer and antiviral drug design, inasmuch as they are able to selectively inhibit activated oncogenes (11) and viral genes (12). An 18-mer phosphodiester aODN containing a double substitution with phosphorothioate residues at both 3'- and 5'-ends has been designed which
ABSTRACT The receptors for urokinase plasminogen activator were studied in both normal human fibroblasts (WI-38 cells) and their SV40-transformed counterpart (VA-13 cells). We have shown that transformed cells expose 10 times more urokinase plasminogen activator receptors (u-PAR) than normal cells. By cross-linking aliquots of cell lysates with the aniinoterminal fragment of the A chain of u-PA, containing the receptor-binding sequence, we have observed a u-PAR concentration at focal contacts in both cell lines. Only transformed cells were able to efficiently invade the basement membrane Matrigel. Switching off the receptor gene expression by the anti-messenger oligodeoxynucleotides strategy abolished the inva sive properties of transformed cells. The anti-messenger oligodeoxynucleotide sequence we have designed inhibited the u-PAR gene expression, lowering both the receptor and the receptor mRNA. This indicates that overexpression of u-PAR gene is itself responsible for invasivity of trans formed fibroblasts in our cell model system and that antisense compound therapy may prove to be of clinical interest in the control of cancer spreading.
selectively targets the u-PAR gene mRNA and has been applied to VA-13 cell cultures in combination with a cationic lipid as a vector (13). The aODN sequence inhibited u-PAR gene expression of transformed cells, dramatically changing their invasive properties.
INTRODUCTION Plasminogen activation is considered a central process in the reg ulation of pericellular proteolysis which occurs under both normal and pathological conditions, including cancer invasion, when tissue de struction and cell migration are required (1, 2). That both tissue type plasminogen activator and u-PA2 are responsible for the fibrinolytic
MATERIALS
AND METHODS
Cell Cultures. The primary cell line WI-38 (human embryonic lung fibro blasts), and their SV40-transformed counterpart, VA-13 cells, were purchased
activity observed in both in vivo neoplasias and in vitro transformed cells remains to be fully investigated; however, it is now well estab lished that u-PA is a hallmark of neoplastic cell transformation (3). After secretion as a zymogen (pro-u-PA), the active form is obtained by cleavage at lysine 158 into two chains connected by a disulfide bridge. As a result of their interaction with a specific receptor (u-PAR) (2) which belongs to the glycosylphosphatidylinositol-anchored pro tein family (4, 5), pro-u-PA and the active enzyme equilibrate be tween solution and the cell surface. Pro-u-PA binds to the receptor by a sequence enframed within the amino-terminal fragment of the A chain (6), thus exposing the B chain, which contains the catalytic site, to the extracellular milieu. On some cell lines, growing as monolayers in vitro, u-PARs are localized at specific sites on the ventral side of the cell: the so-called focal contacts (2, 5, 7). Other components of the plasminogen activation system have been identified on the extracel lular side of cell contacts and have been suggested to play a regulatory
from Flow Laboratories (Milan, Italy) and cultured in DMEM supplemented with 10% fetal calf serum, in an atmosphere of 5% CO2-95% air at 37°C. VA-13 cells were obtained by infection of WI-38 fibroblasts with the trans forming virus in the Wistar Institute (Philadelphia, PA) and were selected on the basis of fibrosarcoma production upon autologous and homologous im plantation (14). lodination of u-PA and ATF and Binding on Cell Monolayers. Human u-PA was prepared as described previously by commercial sources (5). Human ATF (which is the amino-terminal 140 amino acid residues stretch of human u-PA) was kindly provided by Dr. Jack Henkin (Abbott Laboratories). Both molecules were subjected to radioiodination with Na'25I (Amersham Italia SRL, Milan, Italy) using lodo-Gen (Pierce Eurochemie BV, Amsterdam,
the
Netherlands) as described previously (5). The specific activities of the prepa rations used in this study were 21.2 and 69.5 /J.Ci/fj.g for u-PA and ATF, respectively. Binding of u-PA was performed on adherent cell monolayers both as such and pretreated with acidic wash (50 mM glycine-HCl buffer, pH 3.0-0.1 M NaCI) to uncouple endogenous u-PA from the receptor, as described (15). NaOH-soluble radioactivity was measured at the end of incubation and
Received 8/2/94; accepted 10/29/94. 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. ' Work supported by grants from CNR (P. F. ACRO, Attività di Comitato and Progetto Strategico "Nucleotidi Antisenso come Strumenti di Nuove Strategie Terapeutiche"),
specific binding was calculated as described (5). Cross-Linking Assay. Aliquots of the medium of both cell lines, grown to confluence in 25-cm2 Falcon flasks, were used directly in cross-linking exper
MURST (40% and 60%). Ente Cassa di Risparmio di Firenze, and in part by Associazione Italiana per la Ricerca sul Cancro (AIRC). E. A. is the recipient of a fellowship from Accademia Nazionale dei Lincei. A. O. is the recipient of a fellowship from AIRC. 2 The abbreviations used are: t-PA, tissue type plasminogen activator; u-PA. urokinase-type activator; u-PAR, u-PA receptor; pro-u-PA, u-PA zymogen; SAM, surfaceattached material; aODN, anti-messenger oligodeoxynucleotides; ATF, amino-terminal fragment; PIPLC, phosphatidylinositol-specific phospholipase C; PAI-I, plasminogen activator inhibitor 1; PA, plasminogen activator; CM. conditioned medium; ODN. oligodeoxynucleotide; DOTAP, /V-[l-(2,3-dioleyloxy)propyl]-MA',A'-trimethyammonium
After neutralization and further washing in PBS, cells were lysed in the flasks with 500 fil lysis buffer (16) and cleared by centrifugation. In parallel cultures, cells were detached wit 500 fil EDTA (0.2 g/liter in PBS) and centrifuged. The pellet was lysed in 500 ju,llysis buffer and an aliquot was used for cross-linking with I25I-ATF, using disuccinimidyl suberate, as described (5, 16). The mate
iments. Cell monolayers were washed twice with PBS and subjected to acid wash as described above to uncouple endogenous u-PA from the receptor.
rial eventually remaining on the flask after EDTA detachment (SAM) was washed 5 times with 10 ml PBS and extracted with 500 ju.1lysis buffer, and an aliquot was subjected to cross-linking. Phosphatidylinositol-specific phospho-
methylsulfate.
90
Downloaded from cancerres.aacrjournals.org on July 13, 2011 Copyright © 1995 American Association for Cancer Research
aODN EFFECT ON INVASIVE
PHENOTYPE
OF TRANSFORMED
lipase C (PIPLC) release of the u-PA receptor was obtained by incubating cell monolayers at 37°Cfor l h in 0.5 ml DMEM-HEPES containing 10 ^1 PIPLC
u-PAR Gene mRNA Quantitation by Reverse Transcriptase-PCR. m-RNA levels of the u-PAR gene both under basal conditions and after ODN treatment were determined by an internal standard-based quantitative reverse transcriptase-PCR assay.1 ß-Actinwas chosen as internal standard gene be
from Bacillus cereus (final concentration, 0.4 /xg/ml). At the end of incubation the medium was collected and cleared by centrifugation, and an aliquot was subjected to cross-linking. Samples were then analyzed by 12.5% SDS-PAGE
cause of its growth-independent expression pattern. Total cellular RNA was extracted from 10'' cells according to the method of Chomczynski and Sacchi
under nonreducing conditions and autoradiographed. Evaluation of Secreted PA and PAI-1. The enzyme activity present in the culture medium of WI-38 and VA-13 fibroblasts was quantitated by measuring the plasminogen-dependent release of soluble, radioactive fibrin degradation peptides from an insoluble substrate of 125I-labeled fibrin adhering to the walls
(23), treated with RQ1 RNase-free DNase (Promcga, Madison, WI) to elimi nate possible contaminating DNA, and reversely transcribed to cDNA. Both u-PAR and ß-actingene segments were separately amplified using serial dilutions of cDNA. Amplimers used for amplification of the u-PAR gene were GAG AAG ACG TGC AGO GAC CCC (bases 2-21 of the cDNA
of plastic wells, as described by Strickland and Beers (17) and reported by this laboratory (15). PAI-1 and u-PA antigen were measured in aliquots of the harvest medium fluid, using u-PA and PAI-1 ELISA kits (Monozyme ApS,
sequence; sense strand) and GCC AGA GTT GCC CTG GTT GCA (bases 338—358of the cDNA sequence; antisense strand) (22). Amplimers for ß-actin gene were as described by Ng et al. (24). Amplimers for u-PAR gene yielded a segment of 356 base pairs encompassing the anti-messenger ODN-targeted region, and amplimers for ß-actin yielded a segment of 234 base pairs.
Charlottenlund, Denmark), according to the instructions of the manufacturer. Zymography of Released PAs. Aliquots of culture medium were sub jected to SDS-polyacrylamide slab gel electrophoresis (12.5%) under nonre
Samples obtained by omitting the Moloney murine leukemia virus enzyme in the reverse transcription mixture were used as reference controls. PCR ampli fication, according to the hot start procedure (25), was carried out for 1 min at 92°C,1 min at 48°C,and 1 min at 72°Cfor 30 cycles. PCR products were
ducing conditions (18), and migrated proteins were transferred onto a 0.45 jim pore size nitrocellulose filter (Bio-Rad Laboratories, Richmond, CA) as de scribed (15). The nitrocellulose sheet was removed and placed on an indicating layer containing casein and plasminogen, as described (15). After incubation at 37°C clear bands of lysis were visible in the cloudy casein background,
identified by direct sequence analysis using a cycle sequencing method (26). Samples were electrophoresed on 5% nondenaturing polyacrylamide gel, au toradiographed, excised, and counted in a liquid scintillation counter. At least three values of u-PAR gene PCR products, falling within the linear range of
corresponding to the position of PA bands in the polyacrylamide gel. Basement Membrane Matrigel and Chemoinvasion Assay. The base ment membrane Matrigel (Collaborative Research) was applied to 13-mm-
amplification, were normalized to the starting cDNA volumes and referred to ß-actinfor relative quantitation.
diameter filters (8 firn pore size; Nuclepore), essentially as described by Albini et al. (19). The coated filters were placed in Boyden chambers. Cells (40 X IO3), suspended in serum-free DMEM, were added to the upper chamber. CM (obtained from WI-38 human fibroblasts maintained in culture in serum-free conditions for 24 h after confluence) or DMEM containing u-PA
RESULTS Measurement of u-PAR of WI-38 and VA-13 Cell Lines by Saturation Binding Assay with l2SI-u-PA and Evaluation of Re
was placed in the lower compartment (20). Cell migration in the absence of chemoattractants (DMEM alone), was taken as background migration for each cell line over a 5-h period. In some experiments, cells in the upper compart ment of the migration chamber were incubated with mAbs. Anti-human u-PA A»chain mAb (namely, the 5B4 mAb) was a gift from Dr. M. L. Molli (Dow-Lepetit, Várese, Italy). Mouse mAb R3, against u-PAR, was kindly provided by Professor F. Blasi. Assays were carried out at 37°Cin 5% CO2 for
ceptor Occupancy by Endogenous Plasminogen Activators. The Scatchard analysis of binding data (27) has indicated that nonacidified WI-38 normal human fibroblasts expose about 23.3 ±3 (SE) X IO3 u-PAR/cell, with a Kd of about 1.5 nM, while the values of nonacidi fied VA-13-transformed cells are 175.0 ±35 X IO3 and 4.0 nM,
5 h. Filters were then removed and fixed in methanol for 2 h. The cells on the upper part of the filters were completely removed by wiping with a rubber swab. The filters were stained with acid fuchsin and all the cells attached to the lower surface, i.e., the invasive cells, were counted. Each assay was performed in quadruplicate. Oligodeoxynucleotides. To inhibit u-PAR gene expression an 18-mer ODN phosphodiester has been used. Such anti-messenger ODN was protected against exonuclease cleavage by a double substitution at 3'-end and 5'-end with phosphorothioate residues (Genosys). The anti-u-PAR ODN (5'-CGG CGG GTG ACC CAT GTC-3'), prepared and HPLC purified, was designed to encompass to residues bled ODN GGG TCA
respectively. Following acidification, the binding assay revealed an increase of u-PAR in WI-38 normal fibroblasts of about 36% (31.7 ±4 X IO3 receptors/cell). No increase was observed in VA-13 cells. Cell-associated and soluble PA activities were also measured. Con fluent monolayers were incubated for 24 h in serum-free DMEM at 37°Cand 5% CO2, and aliquots of conditioned medium were then harvested for the fibrinolytic assay. Monolayers were acid washed, and aliquots of the washing solution were collected. The two samples were used in the solid-phase 125I-fibrin assay described above. While
the translation start site of the targeted mRNA (21), corresponding 44-61 of the human u-PAR gene cDNA sequence (22). A scram sequence with the same phosphorothioate substitutions (5'-CCT GCG CGC TAG-3') was used as a control. Both ODN uptake and
the plasminogen activator detectable in the culture medium of WI-38 cells was 554.0 mIU/106 cells, the cell-associated plasminogen acti vator accounted for 258.0 mIU/106 cells. In the VA-13 system, the values were 24.5 mIU/106 cells for the culture medium, while the cell-associated enzyme resulted undetectable, thus confirming the
stability were enhanced by combining ODNs with a cationic liposome, namely DOTAP (Boehringer Mannheim, Mannheim, Germany), as described previ ously (13). Anti-messenger or scrambled ODNs were preincubated at 37°Cfor
absence of receptor occupancy by endogenous activator formed cell line. Analysis of u-PAR Distribution by Cross-Linking 12SI-ATF. Fig. 1 shows the results of cross-linking in 125I-ATF is cross-linked by disuccinimidyl suberate
5 min in the presence of DOTAP, sterilized through a 0.2 firn Millex filter (Millipore, Bedford, MA), and applied to the culture medium containing 10% heat-inactivated PCS (GIBCO, Gaithersburg, MD) final concentrations of 10 /AM and 13 JJ.M ODN and DOTAP, respectively. DOTAP-combined ODN half-life and cellular toxicity were determined previously for each batch of heat-inactivated FCS. ODN Treatment of Cell Monolayers. Cells were seeded at 20 X lO'/well in a 24-well Petri dish in DMEM containing
10% heat-inactivated
HUMAN FIBROBLASTS
in the trans Assay with VA-13 cells. to the u-PA
receptor (5, 16) using the cell lysate of the whole monolayer, leading to the formation of a MT70,000 complex (Fig. 1, Lane 2). Lysates of cell monolayers pretreated with PIPLC show a strong decrease of complex formation (Fig. 1, Lane 7). After preincubation of the cell lysate with 100 nM unlabeled ATF, complex formation is impaired
FCS. An
exposure time to ODNs of 4 days was chosen on the basis of preliminary experiments indicating a steady reduction of u-PAR number after 3 days of aODN treatment. Since the half-life of DOTAP-combined ODNs in the culture medium was about 48 h, the initial treatment with 10 /IM DOTAP-combined ODNs was followed by a second addition of 5 JAMDOTAP-combined ODNs
1 A. Morelli, cultured method,
after 48 h in order to restore the initial concentration. On the fourth day, cells were collected and used for phenotypic and molecular analysis.
Ouattrone, L. Papucci, N. Schiarane, V. Saniini, A. Calastrelli. E. Copreni, S. A. Nicolin, and S. Capaccioli. Quanlitation of blc-2 oncogene expression in follicular lymphoma and primary leukemia B-cells by highly sensitive RT-PCR submitted for publication.
91
Downloaded from cancerres.aacrjournals.org on July 13, 2011 Copyright © 1995 American Association for Cancer Research
aODN EFFECT ON INVASIVE
PHENOTYPE
OF TRANSFORMED
HUMAN FIBROBLASTS
transformed cells is almost twice that of normal cells (Fig. 4). To characterize the released PA, we have used a zymographic underlay (Fig. 5). On the basis of recognition by specific anti-u-PA mAb used in the ELISA assay, of comigration with the Mr 54,000 standard u-PA band, and of the absence of other areas of caseinolysis at both higher and lower molecular weights, we conclude that the normal and trans formed human fibroblasts used in this study secrete exclusively u-PA.
mw Fig. 1. Cross-linking of human ATF to VA-13 cells. Lane a, 125I-ATF; Lane b, cross-linking with an aliquot of pooled mediums from mouse LB6 cells transfected with a soluble, recombinant human u-PA receptor and referred to as 4.19 medium (37), as positive control showing cross-linking with a u-PAR with a glycosylation degree that is lower than that observed in our samples; Lane I, culture medium of untreated cells; Lane 2, lysate of whole cell monolayer; Lane 3, lysate of SAM after EDTA detachment of VA-13 monolayer; Lane 4, lysate of SAM after PIPLC treatment; Lane 5, aliquot of the incubation medium of PIPLC—treated SAM; Lane 6, culture medium after PIPLC treatment of VA-13 cells; Lane 7, lysate of the cell monolayer (Lane 2) pretreated with PIPLC; Lane 8, cell lysate in the presence of 100 nM unlabeled ATF. mw, molecular weights reference standard migration. Molecular weights are reported on the left and are expressed in thousands. Arrow on the right, migration of the complex (about M, 70,000).
(Fig. 1, Lane 8). Aliquots of culture medium from PIPLC-untreated cells do not show any complex (Fig. 1, Lane 1), while the complex becomes evident when using medium of PIPLC-treated cells (Fig. 1, Lane 6). These data indicate that VA-13 cells have receptors for human ATF; such receptors are not spontaneously released into the culture medium but are cleavable from the cell surface by treatment with bacterial PIPLC. Cross-linking was also performed on the lysate of SAM which remained on the culture dish after EDTA cell detach ment (Fig. 1, Lane 3). PIPLC treatment of SAM can appreciably reduce 125I-ATF cross-linking (Fig. 1, Lane 4). At the same time the receptor appears in the incubation medium of PIPLC-treated SAM
14Fig. 2. Distribution of u-PAR on the cell body and SAM-associated cell contacts. Lane I, EDTA-detached WI-38 cells; Lane 2, SAM of WI-38 cells; Lane 3, EDTA-detached VA-13 cells; Lane 4, SAM of VA-13 cells. Molecular weights (mw) are reported on the left and are expressed in thousands. The amount of lysate subjected to cross-linking in Lanes I and 2 (WI-38) was 3 times higher in terms of protein (75 jig) than in Lanes 3 and 4 (25 /ig). The electrophoretic conditions were the same as in Fig. 1.
40
(Fig. 1, Lane 5). Cell contacts contained in the SAM concentrate almost all u-PARs present on the surface membrane of VA 13 cells, as evident from a comparison of Lanes 2 and 3 of Fig. 1. Fig. 2 shows the comparison of 125I-ATF-u-PAR complex forma
o
WI-38
30
20
tion between EDTA-detached cells and SAM from normal and trans formed fibroblasts. The lanes within both cell lines are comparable on a per /ug protein basis. While the u-PAR of EDTA-detached cells does not show appreciable differences between the two cell lines (Fig. 2, Lanes 1 and 3), complex formation with SAM indicates that all the difference between normal and transformed cells (more than 6-fold, as indicated by saturation binding experiments) is accounted for by cell contact-associated u-PARs (Fig. 2, Lanes 2 and 4 for WI-38 and VA
• •= total u-PA O—O= active u-PA
£13^Q) (10 J(432.//DVA-"
15 (hours)I time
U-
13 cells, respectively). Vinculin was present in the SAM, as indicated by immunoprobing of SAM and EDTA lysates with anti-vinculin mAb (not shown), thus suggesting that the SAM-associated cell membrane regions belong to the so-called focal adhesions (28). Determination of Secreted Plasminogen Activator and PAI-1 in Normal and Transformed Cells. Confluent monolayers of WI-38 and VA-13 cells (in 6-cm-diameter plates) were incubated under standard culture conditions in 3 ml of serum-free medium. Aliquots of the culture medium were collected at regular intervals and assayed for u-PA and PAI-1 as antigens and for PA activity on 125I-fibrin-coated
OBoX10A
(— -•
.
/Î-o5-13Î~0n • O-
•= total u-PAb20• O= active u-PA
Õ I 0«= 10
15
20
time (hours)
plates. Fig. 3 indicates that normal cells show a secretion of enzyme (measured as both antigen and activity) exceeding by about 1 order of magnitude that of transformed cells, suggesting that secreted inhibi tors are able to buffer the PA activity leaving about 15 to 20% of uninhibited enzyme in both cell lines, even if PAI-1 secretion of
Fig. 3. U-PA in the culture medium of WI-38 and VA-13 cells. Aliquots of the harvest fluid were subjected to determination of u-PA antigen with an ELISA kit, as described. Parallel aliquots were subjected to PA determination by the fibrinolytic method described. In this case the PA activity was expressed as ng/106 cells, assuming a specific PA activity of 100 X IO5 lU/mg. Points, mean of three independent experiments each performed in triplicate (n = 9); bars, SE.
92
Downloaded from cancerres.aacrjournals.org on July 13, 2011 Copyright © 1995 American Association for Cancer Research
aODN EFFECT ON INVASIVE
250
2
Fig. 5. Zymographic assay. Samples of culture medium from WI-38 (Lane 1) and VA-13 (Lane 2) cells were separated by SDS-PAGE, blotted on nitrocellulose, and subjected to zymography for 8 h at 37°C.u-PA, zymogram of standard u-PA (10 IU), with the lysis bands corresponding to high molecular weight u-PA (M, 54,000) and low molecular weight u-PA (M, 33,000). Lane I, 20 ;il of culture medium from WI-38 fibroblasts; Lane 2, 200 ¡JL\ of culture medium from VA-13 fibroblasts.
Chemoinvasion Assay. In a previous paper we have shown that human fibroblasts undergo u-PA-dependent chemotaxis, reaching a peak at the concentration of 50 ng/ml of u-PA in the lower compart ment of the Boyden chamber (20). Thus, the invasion assay was performed either in the presence of DMEM alone in the lower cham ber of the blind well (conditions of random migration) or in the presence of chemoattractant (conditions of chemotaxis, or directed cell migration). The normal fibroblasts showed little or no ability to spontaneously penetrate the barrier. These cells did not display any invasive behavior, even when stimulated with native u-PA or CM (not shown). On the contrary, transformed cells became invasive (Fig. 6), either spontaneously or under chemoinvasion conditions; u-PA stim ulated a good increase of Matrigel invasion, while the CM was exceedingly effective. Both u-PA-dependent and CM-dependent VA-13 invasion can be suppressed in the presence of a mAb against u-PAR (mAB R3, able to impair u-PA/u-PAR interaction). Another mAb against the A chain of u-PA, which impairs the interaction
14—
Fig. 7. Cross-linking of lysates from aODN-treated VA-13 cells with I25I-ATF.Lane 1, aliquot of 4.19 medium (37); Lane 2, SAM of untreatedcells; Lane 3, EDTA-detached untreated cells; Lane 4, SAM of aODN-treated cells; Lane 5, EDTA-detached aODNtreated cells; Lane 6, SAM of untreated cells incubated with 100 nM ATF; Lane 7, standard ATF. Molecular weights (mw) are reported on the left and arc expressed in thousands.
Downloaded from cancerres.aacrjournals.org on July 13, 2011 Copyright © 1995 American Association for Cancer Research
aODN EFFECT ON INVASIVE
PHENOTYPE
OF TRANSFORMED
in nude mice (35). The use of the antimessenger strategy alternatively against each gene has shown that u-PA and cathepsin L can substitute for each other in the promotion of metastasis. In the present study we have considered various molecules of the fibrinolytic pathway. We have observed that normal WI-38 fibroblasts display about 20 X IO3 u-PAR/cell, against about 170 X 103/cell in
IH «control ~
5000
=5
4000
CZl=aODN ES=nsODN R3 = nsODN + mAb 5B4
•S3000
SV-40 transformed VA-13 fibroblasts. In both cell lines the majority of u-PAR is concentrated at the focal contacts. Such location renders the u-PAR a key determinant in the generation of directed cell surface-bound plasmin activity, required for extracellular matrix de struction at the sites of cell progression within tissues. While PAI production does not differ significantly between the two cell lines, u-PA secretion is about 10-fold higher for normal cells. Matrigel invasion assays have shown that both spontaneous and chemotactic invasivity (stimulated by u-PA and/or by CM) of transformed fibro blasts largely exceeds that displayed by normal cells under the same conditions. Monoclonal antibodies against the u-PA A chain and u-PAR were efficient in impairing cell invasivity of transformed cells in Matrigel assay. Taken together, these results indicate that cell invasivity of human fibroblasts correlates with the number of u-PARs
2000 1000
VA—13 / cond.
HUMAN FIBROBLASTS
medium
Fig. 8. Effect of ODN treatment of VA-13 cells on cell invasivity. Cells were subjected to WI-38-conditioned medium chemoinvasivity before and after ODN treatment. The difference between control and ns-ODN-treated cells was not significant. Monoclonal antibodies (R3 and 5B4, 5 /¿g/ml)were used to show that ns-ODN-treated VA-13 cells still retained the possibility to be inhibited by mAbs, as shown previously for untreated VA-13 cells. Points, mean of four independent experiments each performed in duplicate (n = 8); bars, SE.
localized at the cell focal contacts, while it is inversely related with the amount of released u-PA. Moreover, taking into consideration the higher u-PAR occupancy observed in normal cells, one is tempted to speculate that a low u-PAR occupancy in cells displaying many u-PARs, matched with a low u-PA production (transformed cells), is more efficient in promoting cell invasivity than a high u-PA secretion matched with a low number of u-PARs completely occupied by endogenous u-PA (normal fibroblasts). These data are in agreement with data of Tsuboi and Rifkin (36), indicating that high levels of plasmin, generated by secreted PAs, may cause uncontrolled degra dation of the extracellular matrix, thus interrupting the cell-matrix interactions necessary for cell migration and tissue invasion. These observations further strengthen the importance of receptor-bound u-PA as the key determinant of cell invasivity. We have sown that invasivity of transformed fibroblasts is impaired by aODNs against u-PAR. Since the u-PAR gene antisense-targeted sequence did not reveal significant homology with other known human expressed genes, the aODN-induced decrease of VA-13 fibro blasts invasiveness provides direct evidence that increased cell contact association of u-PARs is largely responsible for the invasive proper ties of these cells. The target specificity of the designed aODN is evidenced by the ineffectiveness of scrambled sequence ODN both in lowering u-PAR and u-PAR mRNA and in inhibiting the invasivity of VA-13 cells. Our model, which involves the substitution of 5'- and 3'-ends with phosphorothioate residues and the cationic lipid delivery
VA-13 monolayers indicated a remarkable decrease of u-PAR follow ing aODN treatment, 37.5 ±6.3 X IO3 receptors/cell, while cells treated with nsODN did not show significant variations of u-PAR number with respect to control untreated VA-13 cells (136.0 ±21 X IO3 receptors/cell). Fig. 8 shows that aODN treatment resulted in a considerable decrease of WI-38-CM-dependent invasiv ity of VA-13 cells. The number of cells on the underside of the migration filter was even lower than under conditions of spontaneous random migration, approaching that of WI-38 cells. VA-13 cells treated with nsODN did not show significant differences when com pared with untreated cells, remaining sensitive to the suppression of their invasivity by the mAbs 5B4 and R3. Effect of aODN on u-PAR mRNA Levels. A segment of mRNA of the u-PAR gene, including the aODN-targeted sequence, was reversely transcribed and amplified. For quantitation, PCR product values of the target gene, falling within the linearity range, were normalized to that of ß-actingene used as an internal standard. The 4-day treatment of VA-13 cells with 10 /LIMaODN vehiculated by 13 /J.MDOTAP resulted in a dramatic decrease of u-PAR mRNA levels. In fact, the u-PAR/ß-actin ratio resulted 0.731 ± 0.102 and 0.706 ±0.084 for untreated and nsODN-treated VA-13 cells, respec tively, while the same ratio in aODN-treated cells was 0.105 ±0.11, approaching the value obtained in untreated WI-38 cells
of aODNs in order to inhibit u-PAR gene expression, offers oppor tunities for further studies on the relevance of cell-associated plasminogen activation in the invasive process. Furthermore, assuming that our in vitro model is applicable in vivo, antisense compound therapy may well prove to be of clinical interest in the control of cancer growth.
(0.155 ±0.014). DISCUSSION A substantial body of evidence indicates that urokinase receptor exposure is a major determinant of the invasive phenotype of tumor cells (29-32). Nonetheless, care must be taken in considering u-PAR as the main factor of invasivity among the members of the fibrinolytic system. In vivo tumors contain more PA activity than their corresponding normal counterpart (33), with a considerable variation among histologically similar tumors. Ossowski and Reich (34) have shown that antibodies against human u-PA prevented the formation of métastasesof human carcinoma cells in the chicken embryo lung. Moreover, transforma tion of NIH 3T3 by different ras oncogenes activate two discrete biochemical signal transduction pathways responsible for the expres sion of two alternative patterns of protease genes (either u-PA or cathepsin L), each able to promote the metastatic potential of the cells
REFERENCES 1. Saksela, O. Plasminogen activation and regulation of pericellular proteolysis. Biochim. Biophys. Acta, 823: 35-65, 1985. 2. Pollanen, J., Stephens, R. W., and Vaheri, A. Directed plasminogen activation at the surface of normal and malignant cells. Adv. Cancer Res., 57: 273-328, 1991. 3. Mignatti, P., and Rifkin, D. B. Biology and biochemistry of proteinases in tumor invasion. Physiol. Rev., 73: 161-195, 1993. 4. Ploug, M., Rönne,E., Behrendt, N., Jensen, A. L., Blasi, F., and Dan0, K. Cellular receptor for urokinase plasminogen activator. Carboxyl-terminal processing and membrane anchoring by glycosyl-phosphatidylinositol. J. Biol. Chem., 266: 1926-1933, 1991. 5. Del Rosso, M., Pedersen, N., Fibbi, G., Pucci, M., Dini, G., Anichini, E., and Blasi, F. Selective localization of receptors for urokinase amino-terminal fragment at
94
Downloaded from cancerres.aacrjournals.org on July 13, 2011 Copyright © 1995 American Association for Cancer Research
aODN EFFECT ON INVASIVE
6.
7. 8.
9. 10. 11.
12.
13.
14.
15.
16.
17.
18. 19.
20.
21.
PHENOTYPE
OF TRANSFORMED
substratum contact sites of an in v/rro-established line of human epidermal cells. Exp. Cell Res., 203: 427-434, 1992. Appella, E., Robinson, E. a., Ullrich, S. J., Stoppelli, M. P., Corti, A., Cassani, A., and Blasi, F. The receptor-binding sequence of urokinase. J. Biol. Chem., 262: 44374440, 1987. Pollanen, J., Hedman, K., Nielsen, L. s., Dan0, K., and Vaheri, A. Ultrastructural localisation of plasma membrane-associated urokinase-type plasminogen activator at focal contacts. J. Cell Biol., 106: 87-95, 1988. Emonard, H. P., Remacle, A. G., Noel, A. C, Grimaud, J. A., Stetler-Stevenson, C. A., and Foidart, J. M. Tumor surface-associated binding site for the Mr 72,000 type IV collagenase. Cancer Res., 52: 5845-5848, 1992. Rosen, J. J., and Gulp, L. A. Morphology and cellular origins of substrate-attached material from mouse fibroblasts. Exp. Cell Res., 107: 139-149, 1977. Uhlmann, E., and Peyman, A. Antisense oligonucleotides: a new therapeutic princi ple. Chem. Rev., 90: 543-583, 1990. Monia, B. P., Johnston, J. F., Ecker, D. J., Zounes, M. A., Lima, W. F., and Freier, S. M. Selective inhibition of mutant Ha-ras mRNA expression by antisense oligo nucleotides. J. Biol. Chem., 267: 19954-19962, 1992. Leiter, J. M. E., Agrawal, S., Palese, P., and Zamecnick, P. C. Inhibition of influenza virus replication by phosphorothioate oligodeoxynucleotides. Proc. Nati. Acad. Sci. USA, 87: 3430-3434, 1990. Capaccioli, S., Di Pasquale, G., Mini, E., Mazzei, T., and Quattrone, A. Cationic lipids improve antisense oligonucleotide uptake and prevent degradation in cultured cells and in human serum. Biochem. Biophys. Res. Commun., 107: 818-825, 1993. Jensen, F., Koprowski, H., Pagano, J. S., Ponten, J., and Ravdin, R. G. Autologous and homologous implantation of human cells transformed by simian virus 40. J. Nati. Cancer Inst., 32: 917-937, 1964. Fibbi, G., Magnelli, L., Pucci, M., and Del Rosso, M. Interaction of urokinase A chain with the receptor of human keratinocytes stimulates release of urokinase-like plas minogen activator. Exp. Cell Res., 187: 33-38, 1990. Nielsen, L. S., Kellerman, G. M., Behrendt, N., Picone, R., Dan0, K., and Blasi, F. A 55,000-66,000 M, receptor protein for urokinase-type plasminogen activator. Iden tification in human tumor cell lines and partial purification. J. Biol. Chem., 263: 2358-2363, 1988. Strickland, S., and Beers, W. H. Studies on the role of plasminogen activator in ovulation: in vitro response of granulosa cells to gonadotropins, cyclic nucteotides, and prostaglandins. J. Biol. Chem., 251: 5694-5702, 1976. Laemli, U. K. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature (Lond.), 227: 680-687, 1970. Albini, A., Iwamoto, Y., Kleinman, H. K., Martin, G. R., Aaronson, S. A., Kozlowski, J. M., and McEvan, R. N. A rapid in vitro assay for quantitating the invasive potential of tumor cells. Cancer Res., 47: 3239-3245, 1987. Anichini, E., Fibbi, G., Pucci, M., Caldini, R., Chevanne, M., and Del Rosso, M. Production of second messengers following chemotactic and mitogenic urokinasereceptor interaction in human fibroblasts and mouse fibroblasts transfected with human urokinase receptor. Exp. Cell Res., 213: 438-448, 1994. Toulmè, J. J. 1992. Artificial regulation of gene expression by complementary oligonucleotides. An overview. In: J. A. H. Murray (ed.), Antisense RNA and DNA. New York: Wiley-Liss, pp. 175-194.
HUMAN FIBROBLASTS
22. Roldan, A. L., Cubellis, M. V., Masucci, M. T., Behrendt, N., Lund, L. R., Dane, K., Appella, E., and Blasi, F. Cloning and expression of the receptor for human urokinase plasminogen activator, a central molecule in cell surface, plasmin-dependent proteolysis. EMBO J., 9: 467-474, 1990. 23. Chomczynsky, P., and Sacchi, N. Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction. Anal. Biochem., 162: 156-159, 1987. 24. Ng, S-Y., Gunning, P., Eddy, R., Ponte, P., Leavitt, J., Shows, T., and Kedes, L. Evolution of the functional human /3-actin gene and its multipseudogene family: conservation of non-coding regions and chromosomal dispersion of pseudogenes. Mol. Cell. Biol., J: 2720-2732, 1985. 25. Chou, 0., Rüssel,M., Birch, D. E., Raymond, J., and Bloch, W. Prevention of pre-PCR mis-priming and primer dimerization improves low-copy-number amplifi cations. Nucleic Acids Res., 20: 1717-1723, 1992. 26. Murray, V. Improved double stranded sequencing using the polymerase chain reac tion. Nucleic Acids Res., 17: 8889, 1989. 27. Scatchard, G. The attraction of proteins for small molecules and ions. Ann. NY Acad. Sci., 51: 660-672, 1949. 28. Burridge, K., Path, K., Kelly, T., Nuckolls, G., and Turner, C. Focal adhesions: transmembrane junctions between the extracellular matrix and the cytoskeleton. Annu. Rev. Cell Biol., 4: 487-525, 1988. 29. Schlechte, W. M., Murano, G., and Boyd, D. Examination of the role of the urokinase receptor in human colon cancer mediated laminin degradation. Cancer Res., 49: 6064-6069, 1989. 30. Hollas, W., Blasi, F., and Boyd, D. Role of the urokinase receptor in facilitating extracellular matrix invasion by cultured colon cancer. Cancer Res., 51: 3690-3695, 1991. 31. Crowley, C. W., Cohen, R., Lucas, B. K., Liu, G., Shuman, M. A., and Levinson, A. D. Prevention of metastasis by inhibition of the urokinase receptor. Proc. Nati. Acad. Sci. USA, 90: 5021-5025, 1993. 32. Kobayashi, H., Ohi, H., Shinohara, H., Fujii, T., Schmitt, M., Goretzki, L., Chucholowski, N., Janicke. F., and Graeff, H. Saturation of tumor cell surface receptors for urokinase-type plasminogen activator by amino-terminal fragment and subsequent effecl on reconstituted basement membrane invasion. Br. J. Cancer, 67: 537-544, 1993. 33. Dane, K., Andreasen, P. A., Grendahl-Hansen, J., Kristensen, P., Nielsen, L. S., and Skriver, L. Plasminogen activators, tissue degradation, and cancer. Adv. Cancer Res., 44: 146-289, 1985. 34. Ossowski, L., and Reich, E. Antibodies to plasminogen activator inhibit human tumor metastasis. Cell, 35: 611-619, 1983. 35. Zhang, J-Y., and Schultz, R. M. Fibroblasts transformed by different ras oncogenes show dissimilar patterns of protease gene expression and regulation. Cancer Res., 52: 6682-6689, 1992. 36. Tsuboi, R., and Rifkin, D. B. Bimodal relationship between invasion of the amniotic membrane and plasminogen activator activity. Int. J. Cancer, 46: 56-60, 1990. 37. Masucci, M. T., Pedorsen, N., and Blasi, F. A soluble, ligand binding mutant of the human urokinase plasominogen activator receptor. J. Biol. Chem., 226: 8655-8658, 1991.
95
Downloaded from cancerres.aacrjournals.org on July 13, 2011 Copyright © 1995 American Association for Cancer Research
