Cell Biology International 29 (2005) 959e967www.elsevier.com/locate/cellbi

Invasive growth and topoisomerase-switch induced by tumorousextracellular matrix in osteosarcoma cell culture

Revekka Harisi a,*, Jozsef Dudas a, Ferenc Timar b, Gabor Pogany b, Jozsef Timar a,Ilona Kovalszky a, Miklos SzendrTi c, Andras Jeney a

a 1st Institute of Pathology and Experimental Cancer Research, Faculty of Medicine, Semmelweis University,

Budapest, H-1085, UllTi ut 26, Hungaryb Joint Research Organization of the Hungarian Academy of Sciences and Semmelweis University,

Department of Molecular Pathology, Budapest, Hungaryc Department of Orthopaedics, Faculty of Medicine, Semmelweis University, Budapest, Hungary

Received 21 June 2005; revised 26 July 2005; accepted 10 August 2005

Abstract

Osteosarcoma cells are capable of extracellular matrix (ECM) synthesis. The ability of ECM to trigger the proliferation of a novel osteosar-coma cell line (OSCORT) was tested in this study in relation to a known tumor ECM, isolated from EngelbretheHolmeSwarm (EHS) sarcoma(EHS-ECM). OSCORT was grown in monolayer, in EHS-ECM and in ECM deposited by the cells (OSCORT-ECM). Both EHS-ECM andOSCORT-ECM increased the proliferation and migration of OSCORT cells. Among the ECM biopolymers, heparan sulfate proteoglycan(HSPG) and fibronectin enhanced invasive growth, collagen type IV reduced it, while laminin had no effect. Among the ECM componentsHSPG and collagen IV increased both the synthesis and activation of collagenase type IV, and all the ECM components substantially increasedb1 integrin levels in the cells. The majority of ECM biopolymers decreased the level of topoisomerase I (except laminin) and elevated topoi-somerase II (except fibronectin) in OSCORT. The switch in the ratio between the activities of topoisomerases I and II was mainly due to HSPG.The HSPG synthesized by OSCORT cells is described as agrin, which is a novel finding. The present study showed that HSPG (agrin) showedthe most remarkable stimulatory action on the growth and migration of OSCORT cells. HSPG-induced topoisomerase II-induction deservesfurther experimentation, to discover its relevance to tumor progression.� 2005 International Federation for Cell Biology. Published by Elsevier Ltd. All rights reserved.

Keywords: Extracellular matrix; Osteosarcoma; Invasion; Matrix metalloproteinase; b1 integrin; Topoisomerase; Heparan sulfate proteoglycan; Fibronectin

1. Introduction

Human osteosarcoma is a malignant bone tumor with poorprognosis due to its propensity for metastasis preferentially tothe lungs (Bacci et al., 1988). The majority of the cell lines derived

Abbreviations: ECM, extracellular matrix; EHS-ECM, extracellular matrix

isolated from EngelbretheHolmeSwarm sarcoma; OSCORT-ECM, extracel-

lular matrix prepared from OSCORT cell culture; HSPG, heparan sulfate pro-

teoglycan; MMP, matrix metalloproteinase; Topo, topoisomerase.

* Corresponding author. Tel.: C36 1 457 6518/C36 20 553 4298; fax: C36

1 457 6600.

E-mail address: revekka.harisi@novartis.com (R. Harisi).

1065-6995/$ - see front matter � 2005 International Federation for Cell Biology

doi:10.1016/j.cellbi.2005.08.010

from human osteosarcoma do not form metastasis in immune-suppressed mice (Jia et al., 1999), but their involvement in boneinduction was suspected in the case of the Hu09 cell line (Haraet al., 1996). Correspondingly, an osteosarcoma cell line wasestablished in our laboratory, i.e. OSCORT, from a primaryosteosarcoma of the humerus of a male adolescent.

Several authors indicated the critical role of extracellularmatrix (ECM) in tumor metastasis (Fedarko et al., 1989;Tovari et al., 1997; Brenner et al., 2000; Liotta and Kohn,2001; Bissell and Radinsky, 2001; Bodey et al., 2001; Coxand O’Byrne, 2001; Sasisekharan et al., 2002), based partlyon the expression of ECM-receptors (integrins) on the surfaceof tumor cells (van der Pluijm et al., 1997), and on the ability

. Published by Elsevier Ltd. All rights reserved.

960 R. Harisi et al. / Cell Biology International 29 (2005) 959e967

of the tumor cells to degrade extracellular matrix with matrixmetalloproteinases (Nielsen et al., 1997; Kawashima et al.,2000; Thorns et al., 2003). Accordingly, we decided toexamine the proliferation, invasive growth and migration ofOSCORT cell culture exposed to mouse EngelbretheHolmeSwarm (EHS) sarcoma-derived ECM (Pogany et al., 2001;Harisi et al., 2005), as well as to ECM deposited by OSCORTcells. Besides MTT-measurement and cell counting, DNA top-oisomerase expression and activity were measured in relationto cell proliferation, i.e. DNA synthesis and mitosis (Hsianget al., 1988). In order to get a deeper insight into the compo-nents responsible for the induced effects of ECM on osteosar-coma cells, treatment with ECM components, such as heparansulfate proteoglycan (HSPG), fibronectin, laminin, or collagentype IV, was performed. It was found that ECM supported themigratory and invasive potential of the OSCORT cells, whileHSPG and collagen IV induced a switch in topoisomeraseactivity.

2. Material and methods

2.1. Chemicals

Cell culture media reagents and chemicals of analytical grade were pur-

chased from Sigma (St. Louis, MO, USA), Merck (Darmstadt, Germany),

Boehringer Ingelheim (Heidelberg, Germany) and Boehringer Mannheim

(Mannheim, Germany). The antibodies are specified in the appropriate sections.

2.2. Establishment of OSCORT cell line

The OSCORT cell line is based on a 17-year-old male patient, whose left

humerus diaphysis possessed primary osteosarcoma with fibroblastic differen-

tiation. Fragments of the first biopsy material were subcutaneously implanted

into immuno-suppressed CBA/CA mice. The tumor was grown and histolog-

ically stabilised in the mice, its osteoid-producing capacity remaining in the

xenograft for several passages. The resulting xenograft osteosarcoma retained

the original histological structure, and a cell line, OSCORT, was started from

it, in tissue culture conditions described previously (Harisi et al., 2005). Both

the xenograft and the cell line showed human specific lactate-dehydrogenase-

isoenzyme patterns. The human origin of OSCORT was also confirmed by

karyotype analysis. Immunohistochemical staining confirmed the synthesis

of osteopontin, decorin, osteocalcin and collagen Ic peptide synthesis by the

OSCORT cells (not shown).

The xenograft was established with the written consent of the patient, in

accordance with the Institutes’ Ethical Commitee Guidelines.

The maintenance, care and all operative procedures were performed ac-

cording to the Guidelines issued by Semmelweis University (Rules for Animal

Care No. 100/1999).

The experimental study in oncology at the Institute of Pathology and

Experimental Cancer Research has been approved by the Veterinary Control

Station at Budapest (25e20/2001).

2.3. Preparation of extracellular matrix and isolation of itsbiopolymers

ECM was isolated from EngelbretheHolmeSwarm (EHS) sarcoma (EHS-

ECM) as previously described (Kramer et al., 1986). ECM was also prepared

from OSCORT cell culture (OSCORT-ECM) in the following way: 106

OSCORT cells were plated on T-25 Sarstedt culture dishes (Sarstedt, Germany)

and cultured in 5 ml RPMI-1640 medium for 96 h. After that the medium was

removed and the cells were lysed in phosphate-buffered saline (PBS) contain-

ing 0.5% Triton X-100, 20 mM NH4OH (pH 7.4) for 3 min. The deposited

cell-free ECM layer was washed 4 times with PBS, solubilized with 10 mM

EDTA in PBS and dialyzed against PBS (Fuks et al., 1992).

The ECM biopolymers were prepared from EHS-ECM crude extract as re-

ported (Kleinman et al., 1982; Bissell et al., 1987; Lyon and Gallagher, 1991).

Fibronectin from human plasma was purchased from Sigma (St. Louis, MO,

USA). Composition of EHS-ECM isolated in our laboratory was reported pre-

viously (Pogany et al., 2001). The HSPG component of the OSCORT-ECM

was characterized with western or dot blots using monoclonar antibodies against

agrin [mouse monoclonal (clone 7E12) 1:500 dilution (donated by Eva Kemeny)

(Kemeny et al., 1988)] and perlecan [mouse monoclonal (clone A7L6) 1:500

dilution (Chemicon, Temecula, USA)]. The comparison of EHS-ECM and

OSCORT-ECM showed the presence of both agrin and perlecan in the EHS-

ECM, whereas in OSCORT-ECM only agrin could be detected (Fig. 1). Perlecan

was isolated from EHS-ECM and described in a previous study (Castillo et al.,

1996). The perlecan content of EHS-ECM, was shown in the dot blot reaction in

Fig. 1, as a positive control for the negative reaction of OSCORT-ECM.

2.4. Measurements on cell culture growth

About 7 ! 104 OSCORT cells/cm2 were cultured on 24-well plates

(Greiner, Nurtingen, Germany) for 24e72 h in RPMI-1640 medium containing

10% fetal calf serum and supplemented with penicillin (100 Units/ml medium)

and streptomycin (100 mg/ml medium), at 37 �C, in a humidified 5% CO2

atmosphere. The OSCORT cells were free from contaminants, mycoplasma

contamination was regularly tested for, and treated. OSCORT cells growing as

monolayer culture for 48 h were overlaid by either EHS-ECM or OSCORT-

ECM then allowed to grow for an additional 72 h. Alternatively, ECM or the

various ECM biopolymers (HSPG, fibronectin, laminin, collagen type IV)

were added in soluble form at a protein concentration of 10, 50 or 100 mg/ml.

Cells both from the monolayer and the three-dimensional ECM-gel were

obtained by 0.02% EDTA and by separating the migrated cells in ECM-gel

from the cells remaining in the monolayer. Alternatively, the cells isolated

from the ECM-gel were plated on plastic dishes to characterize the repopu-

lated cells (cell number, viability and topoisomerase activity).

Proliferation of OSCORT cells growing in monolayer or in three-dimen-

sional gel was measured by cell counting in a Buerker haemocytometer, or

using the succinate dehydrogenase test (MTT-assay) (Mosmann, 1983).

Flow-cytometry was used to measure cell population kinetics (Babo et al.,

1999). Cell cycle parameters of 106 suspended cells/ml in 0.1% citrate, 0.1%

Triton X-100, 0.05 mg/ml RN-ase, pH 7.3 were measured after staining with

50 mg/ml propidium iodide for 20 min. Fluorescence measurements were per-

formed on a FACScan flow cytometer (Becton-Dickinson Immunocytometry

Systems, San Jose, CA, USA). The percent of cells in various phases of cell

cycle was then determined by applying Robinovitch’s Multicycle software

(Phoenix Flow Systems Inc., San Diego, CA, USA).

Fig. 1. Detection of heparan sulfate proteoglycans in EHS-ECM and

OSCORT-ECM. (A) Proteoglycan (10 mg) isolated from EHS-ECM or

OSCORT-ECM was analyzed on western or dot blots using anti-HSPG

monoclonal antibody. Lane 1, OSCORT-ECM; Lane 2, EHS-ECM. (B) Proteo-

glycan isolated from EHS-ECM and OSCORT-ECM was dot blotted onto

nitrocellulose membrane, which was reacted with anti-perlecan monoclonal

antibody. Anti-HSPG antibody recognized agrin, the sizes of the glycanated

and non-glycanated proteins (w400 and w200 kDa) are also characteristic

of agrin. Agrin was detected both in EHS-ECM and OSCORT-ECM. Oppo-

sitely to the EHS-ECM used as positive control, perlecan was not detected

in the OSCORT-ECM (B).

961R. Harisi et al. / Cell Biology International 29 (2005) 959e967

2.5. Assay for invasive growth in vitro

To measure the action of ECM biopolymers on invasive growth: osteosar-

coma cells were cultured as monolayers for 48 h in the presence of 50 mg/ml

of ECM biopolymers then 0.75% agarose gel was overlaid and the tumor cells

were left to migrate into the agarose gel for 24 h. Then the agarose gel with the

migrated cells was separated from the monolayer mechanically. To compare

the size of the cell population in the non-migrated and migrated compartments,

the number of cells was counted using a Buerker haemocytometer both in the

monolayer and in the agarose (Harisi et al., 2005).

2.6. Preparation of cells for biochemical studies

For biochemical studies, OSCORT cells were separated from the EHS-ECM-

gel by digesting with 2.5 mg/ml dispase (Boehringer Ingelheim, Heidelberg,

Germany) for 20 min, digestion was stopped with 53.7 mmol/l EDTA. The

removal of the gel fragments was monitored under microscopic examination.

2.7. Immunoblot analysis

Cell lysates, cytoplasmic or cell nuclear extracts were analyzed on sodium

dodecyl sulfate gel electrophoresis (SDS-PAGE) on 10% polyacrylamide gels,

corresponding to 105 cells, and electrotransferred to nitrocellulose membranes

(BioRad, Hercules, CA, USA).

Antibodies were used at the following dilutions: anti-topoisomerase I (scl-

70): 1:5000 human polyclonal serum (TopoGen, Columbus, Ohio, USA), anti-

topoisomerase II-alpha: 1:200 rabbit polyclonal (Novocastra Laboratories

Ltd., Newcastle upon Tyne, UK), anti-b1 integrin (clone JB1a): 1:300 mouse

monoclonal (Chemicon, Temecula, CA, USA). Immunblots were developed

with chemiluminescence (Sigma) and densitometry was measured with an

Eagle-Eye II still video system (Stratagene, La Jolla, CA, USA).

2.8. Detection of metalloproteinase (gelatinase) activity

Metalloproteinase activity was detected from the serum-free conditioned

medium corresponding to 105 OSCORT cells per sample. Unheated aliquots

of conditioned media were concentrated through Centricon 31 filters (Amicon,

Millipore, Bedford, MA, USA), and mixed with Laemmli’s sample buffer

without 2-mercaptoethanol. Measurement of metalloproteinase activity was

performed as described previously (Heussen and Dowdle, 1980; Yamagata

et al., 1988; Babo et al., 1999). Using gelatinase zymography, MMP-2 (gelat-

inase A, 72 kDa collagenase type IV) and MMP-9 (gelatinase B, 92 kDa col-

lagenase type IV) activities were detected in the serum-free conditioned

medium of OSCORT cells grown either on plastic substrate or in the presence

of ECM biopolymers.

2.9. Topoisomerase activity measurements

Cell nuclear extracts were prepared from cultured cells, according to the

method reported by Duguet et al. (1983).

Topoisomerase I activity was measured using both the plasmid relaxation

assay (Taudou et al., 1984; Larsen et al., 1993) and the pBR322 cleavage tech-

nique (Larsen et al., 1993). The former was performed in 20 ml assay buffer

{50 mM TriseHCl (pH 7.5), 150 mM KCl, 10 mM MgCl2, 0.5 mM dithio-

threitol, 0.5 mM EDTA and 0.03 mg/ml bovine serum albumin} containing

cell nuclear extracts of 2.5 ! 104 OSCORT cells and 500 ng of supercoiled

plasmid DNA. The reaction after 30 min incubation at 37 �C was terminated

by the addition of 3% SDS and 1 mg/ml proteinase K (final concentration),

and reaction mixtures were further incubated at 50 �C for 30 min. After addi-

tion of 0.25% w/v bromophenol-blueeTBEeglycerol (30%) (6 ! gel loading

buffer) to the relaxation reaction samples, they were electrophoresed on 1%

agarose gels in TBE buffer {89 mM Triseborate, 2 mM EDTA (pH 8.3)} at

2.4 V/cm for 16 h. Gels were stained with ethidium bromide and imaged as

described above.

In the case of topoisomerase I mediated cleavage the specific inhibitor, top-

otecan, was added at 100 mmol/l to the same cleavage reaction as described for

topoisomerase II mediated cleavage and the electrophoresis was performed in

the presence of 0.45 mol/l NaOH to provide alkaline conditions for the sepa-

ration of single stranded breaks.

The catalytic activity of topoisomerase II was assayed using the decatena-

tion assay as described previously (Marini et al., 1980). The standard reaction

mixture for this assay was 50 mM TriseHCl (pH 7.5), 85 mM KCl, 10 mM

MgCl2, 0.5 mM dithiothreitol, 0.5 mM EDTA, 0.03 mg/ml bovine serum albu-

min and 1 mM ATP. The nuclear extract of 2.5 ! 104 OSCORT cells was

added to this buffer in a volume of 20 ml and was incubated with kDNA for

30 min at 37 �C. The reaction was terminated by incubation for 30 min at

50 �C by adding 3% SDS e 1 mg/ml proteinase K (final concentration) and

0.25% w/v bromophenol-blueeTBEeglycerol (30%) (6 ! gel loading buffer)

was subsequently added to the samples prior to electrophoresis. The samples

were then electrophoresed in 1% agarose gel using TBE buffer at 35 V for 4 h.

The gels were stained with ethidium bromide (1 mg/ml) and differentiated in

H2O. The DNA bands were visualized and analyzed in a Stratagene Eagle-

Eye II video densitometer (Stratagene, Gebouw, CA, USA).

In another assay the topoisomerase II mediated cleavage of end-labeled

pBR322 plasmid (Eco RI linearized pBR322 plasmid was end-labeled with

a32P-dATP using Klenow enzyme (Promega, Madison, WI, USA), followed

by the subsequent cleavage of one of the labeled ends with BamH1 (Promega))

was studied using the same reaction mixture as for the decatenation assay, but

with 20 times more nuclear protein and with 100 mmol/l etoposide as specific

inhibitor of topoisomerase II (Larsen et al., 1993). Double stranded breaks

were detected using gel electrophoresis in 1% agarose at 20 V for 16 h, fol-

lowed by overnight autoradiography.

2.10. Statistical analysis

Statistical significance ( p! 0.05) of the measured effects and differences

was determined using Student’s t-test.

3. Results

3.1. Effect of ECM biopolymers on the growth ofOSCORT cells

The ECM derived from EHS and OSCORT cell cultureenhanced the growth of the cells in a concentration-dependentmanner (Fig. 2). Among the ECM components, fibronectinand HSPG significantly enhanced the growth of the cellseven more than the EHS-ECM ( p! 0.001), while collagentype IV, in proportion to its concentration, reduced the sizeof the cell culture even at 10 mg/ml ( p! 0.001). Laminindid not influence cell growth. The effect of ECM componentson the growth of cells was detected using succinate dehydro-genase (MTT-assay). The MTT results of cells cultured in con-trol conditions (on plastic surface) were considered to be100%, and the SD was 3.77%.

In the presence of ECM, HSPG or fibronectin the percentof cells in S-phase was twice as much as in the untreated cul-tures, while the cell cycle was not affected by laminin. Incontrast, collagen type IV induced a fourfold increase in theG2-phase, with a concomitant decrease in the proportion ofboth G1- and S-phase cell populations.

3.2. Effect of ECM biopolymers on the migration ofOSCORT cells

To elucidate the individual action of ECM components,agarose gel was used instead of EHS-ECM gels. Cells in themonolayer cultures were treated with fibronectin, HSPG,

962 R. Harisi et al. / Cell Biology International 29 (2005) 959e967

laminin or collagen type IV for 48 h and then the monolayercultures were overlaid with agarose, which offered an upwardmigration for the cells. As shown in Table 1, the ECM compo-nents elicited rather modest changes in the total size of theOSCORT cell population, i.e. 20% and 31% elevation in thecase of fibronectin and HSPG, respectively; collagen type IVreduced the total number of cells by 20% and laminin causedno changes. Interestingly, the distribution of the OSCORT cellpopulation between the monolayer and the agarose phase wasshowing the migratory preference of OSCORT cells in thepresence of ECM proteins (Table 1). The migration was mark-edly stimulated after treatment with fibronectin (203%) andHSPG (356%).

3.3. Matrix metalloproteinase (MMPs) activity inducedby ECM and its biopolymers

There was no sign of constitutive synthesis of MMPs in theOSCORT culture, whereas in the presence of ECM, both

Fig. 2. Proliferation of OSCORT cells in the presence of ECM and its biopol-

ymers. Cells were attached to 24-well plates in medium supplemented with

serum for 48 h. After that they were further cultured on plastic. Solubilized

EHS-ECM, OSCORT-ECM and ECM components were added to the medium

at a concentration range of 10e100 mg/ml. Cells were incubated for an addi-

tional 72 h, then an MTT-assay was performed. The figure shows optical den-

sities after MTT-assay related to control cells cultured on plastic substrate.

Solubilized EHS-ECM, fibronectin, HSPG and OSCORT-ECM significantly

increased cell proliferation above 50 mg/ml ( p! 0.001, with Student’s

t-test). Laminin had no significant effect on proliferation ( pO 0.05), while

collagen type IV significantly decreased cell proliferation even at 10 mg/ml

( p! 0.001). Bars represent the results of 3 independent experiments in all

cases. The MTT results of cells cultured in control conditions (on a plastic sur-

face) were considered to be 100%, its SD was: 3.77%, which is demonstrated

by the horizontal line in the figure, to show the inductive and inhibiting effects

of the ECM components.

92 kDa collagenase type IV and its active form (82 kDa) couldbe recognized (Fig. 3) (Harisi et al., 2005). Among the ECMcomponents, fibronectin increased the synthesis of 92 kDa col-lagenase type IV, but not its activation, whereas HSPG andcollagen IV increased both the 92 and the 82 kDa forms. Aswith the EHS-ECM, the OSCORT-ECM increased both thesynthesis and activation of collagenase type IV, although itseffect was more moderate.

3.4. b1 integrin synthesis in OSCORT cells in thepresence of ECM and its biopolymers

ECM-gel increased the protein synthesis of b1 integrin, asdescribed in our previous study (Harisi et al., 2005). We cannow confirm that all the ECM components substantially in-creased the b1 integrin levels of OSCORT (Fig. 4).

3.5. Topoisomerase-switch in the migrating OSCORTcells

Comparing topoisomerases I and II activities in the prolif-erating and migrating compartments of OSCORT cultures,a marked difference could be observed. Topoisomerase I activ-ity predominated over topoisomerase II in the cells originatingfrom the monolayer, while in the cells collected from theECM-gel (i.e. migrating compartment), topoisomerase II ac-tivity was markedly higher (Fig. 5; lanes a-3,4 and b-3,4). Aquestion had been raised as to whether the shift in the domi-nant topoisomerase activity was stable, so the migrating andnon-migrating cells were re-cultured and re-examined after 3division cycles. As Fig. 5 (lanes a-5,6 and b-5,6) shows, trans-fer of OSCORT cells from the non-migrating or migratingcompartments, and culture in the absence of ECM, resultedin the reactivation of topoisomerase I dominance.

It was investigated whether the topoisomerase-switch is eli-cited by ECM only in its gel form or in its soluble form aswell. Furthermore, measurements were done on the effectsof the individual ECM biopolymers, to give insight into thetype of molecule that could be responsible for the changesin topoisomerase activity during tumor cell migration. Figs. 6and 7 show that the lower level of topoisomerase I and ele-vation of topoisomerase II in OSCORT cells was observed inthe presence of the majority of ECM biopolymers. Interestingly,however, laminin did not reduce either the expression or the

Table 1

ECM biopolymers’ action on migration of OSCORT cells in agarose gel

ECM treatment (50 mg/ml) Number of cells (!104) Total growth

changes (%)

Changes in the migratory

compartment (%)Non-migrated Migrated

Control 38.57 G 1.27 9.29 G 1.50 100 G 4.70 100 G 16.14

Fibronectin 39.14 G 2.40 18.86 G 2.60a 121 G 12.96a 203 G 13.95a

HSPG 29.57 G 0.98a 33.14 G 2.48a 131 G 8.97a 356 G 26.69a

Laminin 38.70 G 1.50 9.57 G 1.51 100 G 7.80 103 G 16.25

Collagen type IV 26.40 G 2.23a 12.00 G 1.83 80 G 10.52a 129 G 19.69a

Data are given as mean G SD. All numbers represent 3 independent experiments.a Significant difference from the value measured for plastic ( p! 0.05), using Student’s t-test.

963R. Harisi et al. / Cell Biology International 29 (2005) 959e967

Fig. 3. MMP-9 induction by ECM and its biopolymers. After attachment, OSCORT cells were further cultured for 72 h under different treatment conditions. EHS-

ECM, OSCORT-ECM and ECM components were added to the culture. Metalloproteinase activities were detected from the conditioned medium of the cells.

Plastic: plastic cultured cells; EHS-ECM: EHS-ECM cultured cells; Fibronectin (FN), Heparan sulfate proteoglycan (HSPG), Laminin (LN), Collagen type IV

(C IV): ECM components added to the culture medium at 100 mg/ml; OSCORT-ECM: OSCORT-ECM added in excess at 100 mg/ml. Metalloproteinase activity

was measured by densitometry of the white bands on blue gels, and indicated in the table below the image. The image represents a typical zymogram from 3

repeated experiments.

catalytic activity of topoisomerase I; in addition, topoisomer-ase II was not elevated by fibronectin.

4. Discussion

The invasion of cancer cells is mainly regulated by theirproteolytic remodeling of the surrounding extracellular matrixand by their integrin expression (Giannelli et al., 2001; Hoodand Cheresh, 2002). Following malignant transformation,ECMeintegrin interactions are still involved during a varietyof cell functions, such as motility, proliferation and viability(Senoo et al., 1996; Miranti and Brugge, 2002).

Sarcoma cells, such as those of the EHS tumor or the oste-osarcoma cells themselves, synthesize extracellular matrixcomponents (Kleinman et al., 1982; Pogany et al., 2001; Harisiet al., 2005). The data presented here, in addition to our pre-vious study (Harisi et al., 2005), confirmed the positive effectof the tumorous ECM (derived from EHS tumor) on the pro-liferation, migration, MMP-9 activity and b1 integrin synthe-sis of OSCORT cells. It is the first time that cell-freeextracellular matrix deposited by osteosarcoma cells hasbeen prepared and used for treatment. Interestingly, theOSCORT-derived ECM exhibited effects comparable with

EHS-ECM. HSPG and fibronectin were found to be the tumor-matrix components responsible for the stimulatory effects. Incontrast, reduced proliferative capacity was observed in thepresence of collagen type IV, and laminin showed no signifi-cant effect. HSPG induced b1 integrin synthesis, MMP-9gelatinase activity, topoisomerase II expression and enzymeactivity. The effects of fibronectin were more modest at themolecular level.

It is important to note that tumorous ECM is cruciallyimportant to support the metastatic ability of osteosarcoma.Several previous studies failed to model the metastatic potentialof osteosarcomas in vitro, when compared to the clinical ex-perience (Bacci et al., 1988; Hara et al., 1996; Jia et al.,1999). From the current study, it can be seen that tumor ECMneeds to be considered in the investigation of invasive growthof osteosarcomas. The ECM component that presumablyundergoes the most tumor-specific changes is HSPG, depend-ing on the variabilities in heparan-sulfate synthesis (Kemenyet al., 1988; Lyon and Gallagher, 1991; Dudas et al., 2000;Sasisekharan et al., 2002; Yamauchi et al., 2005). HSPG mightcontribute to the malignant phenotype and tumor invasion, aswell as effect against it (Serra et al., 2005), depending on theoligosaccharide sequence and whether it is present in the

Fig. 4. b1 integrin expression in the presence of ECM and its biopolymers. After attachment OSCORT cells were further cultured for 72 h under different treatment

conditions. b1 integrin was detected from the cytoplasm extract of 105 cells. Plastic: plastic cultured cells; EHS-ECM: EHS-ECM cultured cells; Fibronectin (FN),

Heparan sulfate proteoglycan (HSPG), Laminin (LN), Collagen type IV (C IV): ECM components added to the culture medium at 100 mg/ml; OSCORT-ECM:

OSCORT-ECM added in excess at 100 mg/ml. Immunoblot results were analyzed with densitometry, the protein expression is indicated in the table below the

image. The image represents a typical immunoblot from 3 repeated experiments.

964 R. Harisi et al. / Cell Biology International 29 (2005) 959e967

Fig. 5. Extracellular matrix action on topoisomerases I and II activity. (A) Topoisomerase I activity detected by the relaxation of supercoiled pBR322 plasmid

DNA. (B) Topoisomerase II activity detected by the ATP-dependent decatenation of kDNA. 1: Negative control, i.e. 500 ng plasmid DNA without nuclear extract.

2: Positive control, i.e. 500 ng plasmid DNA with 2 units of topoisomerase I or II (TopoGen, Columbus, Ohio, USA). 3: Topo I (A), II (B) activities in 72 h mono-

layer OSCORT culture. 4: Topo I (A), II (B) activities in 72 h three-dimensional OSCORT culture. 5: Topo I (A), II (B) activities in 7 day monolayer OSCORT

culture transferred from the 72 h monolayer OSCORT culture. 6: Topo I (A), II (B) activities in monolayer OSCORT cells transferred from the 72 h three-dimen-

sional OSCORT culture. MW: molecular weight. The figure represents typical gel images from 3 repeated experiments.

ECM in a soluble form or anchored at the cell surface (Fedarkoet al., 1989; Battaglia et al., 1993; Tovari et al., 1997;Sasisekharan et al., 2002; Merz et al., 2003). The HSPG ofthe EHS-ECM used in our experiments contained both perlecan(Merz et al., 2003; Castillo et al., 1996) and agrin (Wang et al.,2003). Agrin was found in the OSCORT-ECM (Fig. 1). Agrinsynthesis by sarcoma cells is a novel finding of this study.

HSPGs have been reported as essential regulators of theWnt/b-catenin signaling pathway (Merz et al., 2003; Wanget al., 2003; Schambony et al., 2004). This pathway is involvedin the regulation of cell migration. Mutations in the unc-52gene of Claenorhabditis elegans, which encodes a homologueof the basement membrane heparan sulfate proteoglycan,

perlecan, led to migratory defects in the distal tip cells, whichwere partially related to the Wnt/b-catenin signaling pathway(Merz et al., 2003). A potential of the Wnt/b-catenin pathwayto crosstalk with the agrin signaling cascade was also reported(Wang et al., 2003). Interestingly, one of the targets of thetranscription factors induced by the Wnt/b-catenin signalingpathway is MMP-9, which at the same time takes an importantactive role in the signaling pathway, since MMP-9 knockdownconcomitantly results in increased levels of surface E-cadherin, redistribution at the plasma membrane of b-catenin,and its physical association with E-cadherin (Sanceau et al.,2003). This relationship could have been observed in our studyas well, since MMP-9 was significantly induced by the

Fig. 6. Changes in protein synthesis of topoisomerases in the presence of ECM and its biopolymers. After attachment, OSCORT cells were further cultured for 72 h

under different conditions. Topoisomerase I and II protein levels were measured using the cell nuclear extracts of 105 OSCORT cells. Immunoblots were developed

with chemiluminescence and were analyzed with densitometry. 1 (Plastic): plastic cultured cells; 2 (EHS-ECM ): EHS-ECM cultured cells; 3 (FN ): fibronectin; 4

(HSPG): heparan sulfate proteoglycan; 5 (LN ): laminin; 6 (C IV): collagen type IV: ECM components added to the culture medium at 100 mg/ml. The figure

represents typical immunoblots from 3 repeated experiments.

965R. Harisi et al. / Cell Biology International 29 (2005) 959e967

Fig. 7. Changes in enzyme activity of topoisomerases in the presence of ECM and its biopolymers. After attachment, OSCORT cells were further cultured for 72 h

under different treatment conditions, followed by cell nuclear protein extraction. Topoisomerase I activity was detected by the relaxation of the supercoiled pBR322

plasmid DNA in the presence of nuclear extract from 2.5 ! 104 OSCORT cells. R: relaxed DNA, S: supercoiled DNA. NC, negative control, i.e. 500 ng plasmid

DNA without nuclear extract. PC, positive control, i.e. 500 ng plasmid DNA with 2 units of topoisomerase I (TopoGen, Columbus, Ohio, USA). 1e6, 500 ng

plasmid DNA with nuclear extract from OSCORT cells cultured under different conditions. Topoisomerase II activity was detected by the ATP-dependent deca-

tenation of kDNA in the presence of nuclear extract from 2.5 ! 104 OSCORT cells. During the reaction, interlocking kDNA network is decatenated to individual

minicircles. NW: high molecular weight kDNA network. Mc: minicircles. NC, negative control, i.e. 500 ng kDNA without nuclear extract. PC, positive control, i.e.

500 ng kDNA with 2 units of calf thymus topoisomerase II (TopoGen, Columbus, Ohio, USA). 1e6, 500 ng kDNA with nuclear extract from 2.5 ! 104 OSCORT

cells cultured under different conditions: 1 (Plastic): plastic cultured cells; 2 (EHS-ECM ): EHS-ECM cultured cells. 3e6: ECM components were added to the

culture medium at 100 mg/ml: 3 (FN ): fibronectin, 4 (HSPG): heparan sulfate proteoglycan, 5 (LN ): laminin, 6 (C IV): collagen type IV. The relationship of relaxed

bands to total plasmid DNA, as well as decatenated bands to total kDNA, were analyzed with densitometry and indicated in the tables under the images. The figure

represents typical gel images from 3 repeated experiments.

OSCORT-ECM and by the isolated HSPG (agrin). The otherpotential target of the Wnt/b-catenin signaling is fibronectin(FN). The fibronectin promoter contains LEF/TCF-bindingsites, which are a target of the Wnt-signaling pathway(Schambony et al., 2004).

A novel issue described in this study is the so-called ‘‘top-oisomerase-switch’’. An increase in topo II and decrease intopo I activities was detected in relation to tumor ECM-inducedinvasive growth, and was also induced by HSPG. The topoi-somerase II-induction of HSPG might be related to the integ-rin-mediated signaling to the Wnt-signaling pathway(Schambony et al., 2004), which might lead to the activationof focal adhesion kinase, with a potential to stimulate extra-cellular signal-related kinase 2 (ERK2), and to achieve theactivation of topoisomerase II (Shapiro et al., 1999).

Several reports have shown that stimulation of cell prolifer-ation accompanies increased synthesis and activity of topoiso-merase II (Hsiang et al., 1988; Ludvikova et al., 2005). It is,however, odd that collagen type IV was showing inhibitoryaction on the invasive growth of OSCORT cells and alsoincreased topoisomerase II activity, which was related to theaccumulation of OSCORT cells in G2/M phase. This issuemight suggest distinct signal transduction pathways for thetopoisomerase-switch and induction of cell migration. Aspreviously reported in vitro, diphosphorylated active ERK2phosphorylated topoisomerase IIa and enhanced its specificactivity by sevenfold, as measured by DNA relaxation assays,whereas unphosphorylated ERK2 had no effect (Shapiro et al.,1999). ERK2 is regulated by mitogen activated protein kinase1/2 (MAPK 1/2), which might be a target of collagen IV

mediated signaling (Anbalagan and Rao, 2004). On the otherhand, the collagen environment was reported to control theactivity of the Wnt-pathway, and also Wnt-expression. MCF-10A cells down-regulate the expression of Wnt-5A, but notWnt-7B, when grown on collagen (Bui et al., 1997; Schambonyet al., 2004), which might be related to the detected anti-migratory effect of collagen IV, independent of the inductionof topoisomerase II via a different signal transductionpathway.

In summary, osteosarcoma-derived ECM supported the in-vasive growth of a human osteosarcoma cell line. This supportwas in association with the induction of b1 integrin and MMP-9 enzymatic activity, and was related to the suppression of top-oisomerase I and enhancement of topoisomerase II activities.It could be concluded that, among the ECM components,HSPG was the most effective inducer of tumor cell migration.Therapeutic intervention in order to interfere with the stimula-tory effects of HSPG might have potential in the treatment ofthe metastatical processes of osteosarcoma.

Acknowledgements

This study was supported by the Hungarian National Scien-tific Research Foundation (Grant OTKA T32751), by the Hun-garian Ministry of Health (Grants 224/2000, 145/2000) and bythe Hungarian Ministry of Education (Grant NKFB 1/482000). The authors wish to thank Janos Fuzesi for his helpwith the artwork.

966 R. Harisi et al. / Cell Biology International 29 (2005) 959e967

References

Anbalagan M, Rao AJ. Collagen IV-mediated signaling is involved in progen-

itor Leydig cell proliferation. Reprod Biomed Online 2004;9:391e403.

Babo I, Bocsi J, Jeney A. The site-dependent growth characteristics of a human

xenotransplanted basaloid squamous cell carcinoma. J Cancer Res Clin

Oncol 1999;125:35e41.

Bacci G, Avella M, Picci P, Briccoli A, Dallari D, Campanacci. Metastasis

pattern in osteosarcoma. Tumori 1988;74:421e7.

Battaglia C, Aumailley M, Mann K, Mayer U, Timpl R. Structural basis of b1

integrin-mediated cell adhesion to a large heparan sulfate proteoglycan

from basement membranes. Eur J Cell Biol 1993;61:92e9.

Bissell DM, Arenson DM, Maher JJ, Roll FJ. J Clin Invest 1987;79:801e12.

Bissell MJ, Radinsky D. Putting tumours in context. Nat Rev Cancer 2001;

1:46e54.

Bodey B, Bodey Jr B, Siegel SE, Kaiser HE. Matrix metalloproteinases in

neoplasm-induced extracellular matrix remodeling in breast carcinomas.

Anticancer Res 2001;21:2021e8.

Brenner W, Gross S, Steinbach F, Horn S, Hohenfellner R, Thuroff JW. Dif-

ferential inhibition of renal cancer cell invasion mediated by fibronectin,

collagen IV and laminin. Cancer Lett 2000;155:199e205.

Bui TD, Tortora G, Ciardiello F, Harris AL. Expression of Wnt5a is downre-

gulated by extracellular matrix and mutated c-Ha-ras in the human mam-

mary epithelial cell line MCF-10A. Biochem Biophys Res Commun

1997;239:911e7.

Castillo GM, Cummings JA, Ngo C, Yang W, Snow AD. Novel purification

and detailed characterization of perlecan isolated from the EngelbretheHolmeSwarm tumor for use in an animal model of fibrillar A beta amyloid

persistence in brain. J Biochem (Tokyo) 1996;120:433e44.

Cox G, O’Byrne KJ. Matrix metalloproteinases and cancer. Anticancer Res

2001;21:4207e19.

Dudas J, Ramadori G, Knittel T, Neubauer K, Raddatz D, Egedy K, et al. Effect

of heparin and liver heparan sulphate on interaction of HepG2-derived tran-

scription factors and their cis-acting elements: altered potential of hepato-

cellular carcinoma heparan sulphate. Biochem J 2000;15:245e51.

Duguet M, Lavenot C, Harper F, Mirambeau G, De Recondo AM. DNA top-

oisomerases from rat liver: physiological variants. Nucleic Acids Res

1983;11:1059e75.

Fedarko NS, Ishihara M, Conrad HE. Control of cell division in hepatoma cells by

exogenous heparan sulfate proteoglycan. J Cell Physiol 1989;139:287e94.

Fuks Z, Vlodavsky I, Andreeff M, Mcloughlin M, Haimovitz-Friedman A. Ef-

fects of extracellular matrix on the response of endothelial cells to radia-

tion in vitro. Eur J Cancer 1992;28A:725e31.

Giannelli G, Bergamini C, Fransvea E, Marinosci F, Quaranta V, Antonaci S.

Human hepatocellular carcinoma (HCC) cells require both alpha3beta1 in-

tegrin and matrix metalloproteinases activity for migration and invasion.

Lab Invest 2001;81:613e27.

Hara A, Ikeda T, Nomura S, Yagita H, Okumura K, Yamauchi Y, et al. In vivo

implantation of human osteosarcoma cells in nude mice induces bones

with human-derived osteoblasts and mouse-derived osteocytes. Lab Invest

1996;75:707e17.

Harisi R, Dudas J, Timar F, Pogany G, Paku S, Timar J, et al. Antiproliferative

and antimigratory effects of doxorubicin in human osteosarcoma cells ex-

posed to extracellular matrix. Anticancer Res 2005;25:805e14.

Heussen C, Dowdle EB. Electrophoretic analysis of plasminogen activators in

polyacrylamide gels containing sodium dodecyl sulfate and copolymerized

substrates. Anal Biochem 1980;102:196e202.

Hood JD, Cheresh DA. Role of integrins in cell invasion and migration. Nat

Rev Cancer 2002;2:91e100.

Hsiang YH, Wu HY, Liu LF. Proliferation-dependent regulation of DNA top-

oisomerase II in cultured human cells. Cancer Res 1988;48:3230e5.

Jia SF, Worth LL, Kleinerman ES. A nude mouse model of human osteosarco-

ma lung metastases for evaluating new therapeutic strategies. Clin Exp

Metastasis 1999;17:501e6.

Kawashima A, Kawahara E, Tokuda R, Nakanishi I. Tumour necrosis factor-

a provokes upregulation of a2b1 and a5b1 integrins, and cell migration in

OST osteosarcoma cells. Cell Biol Int 2000;25:319e29.

Kemeny E, Fillit HM, Damle S, Mahabir R, Kefalides NA, Gregory JD, et al.

Monoclonal antibodies to heparan sulfate proteoglycan: development and

application to the study of normal tissue and pathologic human kidney bi-

opsies. Connect Tissue Res 1988;18:9e25.

Kleinman HK, Mcgarvey ML, Liotta LA, Robey PG, Tryggvason K, Martin GR.

Isolation and characterization of type IV procollagen, laminin, and heparan

sulfate proteoglycan from EHS sarcoma. Biochemistry 1982;21:6188e893.

Kramer RH, Bensch KG, Wong J. Invasion of reconstituted basement mem-

brane matrix by metastatic human tumor cells. Cancer Res 1986;46:

1980e9.

Larsen AK, Grondard L, Couprie J, Desoize B, Comoe L, Jardillier JC, et al.

The antileukemic alkaloid fagaronine is an inhibitor of DNA topoiso-

merases I and II. Biochem Pharmacol 1993;46:1403e12.

Liotta LA, Kohn EC. The microenvironment of the tumorehost interface.

Nature 2001;411:375e9.

Ludvikova M, Holubec Jr L, Ryska A, Topolcan O. Proliferative markers in

diagnosis of thyroid tumors: a comparative study of MIB-1 and topoiso-

merase II-a immunostaining. Anticancer Res 2005;25:1835e40.

Lyon M, Gallagher JT. Purification and partial characterisation of the major

cell-associated heparan sulfate proteoglycan of rat liver. Biochem J 1991;

273:415e22.

Marini JC, Miller KG, Englund PT. Decatenation of kinetoplast DNA by top-

oisomerases. J Biol Chem 1980;255:4976e9.

Merz DC, Alves G, Kawano T, Zheng H, Culotti JG. UNC-52/perlecan affects

gonadal leader cell migrations in C. elegans hermaphrodites through alter-

ations in growth factor signaling. Dev Biol 2003;256:173e86.

Miranti CK, Brugge JS. Sensing the environment a historical perspective on

integrin signal transduction. Nat Cell Biol 2002;4:83e90.

Mosmann T. Rapid colorimetric assay for cellular growth and survival: appli-

cation to proliferation and cytotoxicity assays. J Immunol Methods 1983;

65:55e63.

Nielsen BS, Sehested M, Kjeldsen L, Borregaard N, Rygaard J, Dano K. Ex-

pression of matrix metalloprotease-9 in vascular pericytes in human breast

cancer. Lab Invest 1997;77:345e55.

Pogany G, Timar F, Olah J, Harisi R, Polony G, Paku S, et al. Role of

basement membrane in tumor cell dormancy and cytotoxic resistance.

Oncology 2001;60:274e81.

Sanceau J, Truchet S, Bauvois B. Matrix metalloproteinase-9 silencing by

RNA interference triggers the migratory-adhesive switch in Ewing’s sarco-

ma cells. J Biol Chem 2003;278:36537e46.

Sasisekharan R, Shriver Z, Venkataraman G, Narayanasami U. Roles of heparan-

sulphate glycosaminoglycans in cancer. Nat Rev Cancer 2002;2:521e8.

Schambony A, Kunz M, Gradl D. Cross-regulation of Wnt signaling and cell

adhesion. Differentiation 2004;72:307e18.

Senoo H, Imai K, Sato M, Kojima N, Miura M, Hata R. Three-dimensional

structure of extracellular matrix reversibly regulates morphology, prolifer-

ation and collagen metabolism of perisinusoidal stellate cells (vitamin

a-storing cells). Cell Biol Int 1996;20:501e12.

Serra M, Miquel L, Domenzain C, Docampo MJ, Fabra A, Wight TN, et al. V3

versican isoform expression alters the phenotype of melanoma cells and

their tumorigenic potential. Int J Cancer 2005;114:879e86.

Shapiro PS, Whalen AM, Tolwinski NS, Wilsbacher J, Froelich-Ammon SJ,

Garcia M, et al. Extracellular signal-regulated kinase activates topoisomer-

ase IIalpha through a mechanism independent of phosphorylation. Mol

Cell Biol 1999;19:3551e60.

Taudou G, Mirambeau G, Lavenot C, Der Garabedian A, Vermeersch J,

Duguet M. DNA topoisomerase activities in concavalin A stimulated lym-

phocytes. FEBS Lett 1984;176:431e5.

Thorns V, Walter GF, Thorns C. Expression of MMP-2, MMP-7, MMP-9,

MMP-10 and MMP-11 in human astrocytic and oligodendroglial gliomas.

Anticancer Res 2003;23:3937e44.

Tovari J, Paku S, Raso E, Pogany G, Kovalszky I, Ladanyi A, et al. Role of

sinusoidal heparan sulfate proteoglycan in liver metastasis formation. Int

J Cancer 1997;71:825e31.

van der Pluijm, Vloedgraven H, Papapoulos S, Lowick C, Grzesik W, Kerr J,

et al. Attachment characteristics and involvement of integrins in adhesion

of breast cancer cell lines to extracellular bone matrix components. Lab

Invest 1997;77:665e75.

967R. Harisi et al. / Cell Biology International 29 (2005) 959e967

Wang J, Jing Z, Zhang L, Zhou G, Braun J, Yao Y, et al. Regulation of acetylcholine

receptor clustering by the tumor suppressor APC. Nat Neurosci 2003;6:1017e8.

Yamagata S, Ito Y, Tanaka R, Shimizu S. Gelatinases of metastatic cell lines of

murine colonic carcinoma as detected by substrateegel electrophoresis.

Biochem Biophys Res Commun 1988;151:158e62.

Yamauchi N, Watanabe A, Hishinuma M, Ohashi KI, Midorikawa Y,

Morishita Y, et al. The glypican 3 oncofetal protein is a promising diagnos-

tic marker for hepatocellular carcinoma; heparan sulfate proteoglycan

binding promotes APRIL-induced tumor cell. Mod Pathol; Cell Death

Differ 2005;12:637e48.