anticoagulant activity (196 USP U/mg), and chemically ... sodium salt with less than 5 USP U/mg each were a generous gift ..... Laboratorio de HistologÃa Animal.
Journal of Biological Regulators and Homeostatic Agents
Heparin increases the adhesion of murine mammary adenocarcinoma cells (LM3). Correlation with the presence of heparin receptors on cell surface G. BERTOLESI, A.M. EIJAN, J.C. CALVO, L. LAURIA DE CIDRE Facultad de Ciencias Exactas y Naturales, Departamento de Biodiversidad y Biología Experimental, Universidad de Buenos Aires, Ciudad Universitaria, Buenos Aires, Argentina
ABSTRACT: Mastocytosis is a common feature around solid tumors. Due to mast cell (MC) degranulation, heparin and other chemical mediators are released to surrounding tissues. The aim of this paper is to investigate the role of heparin and chemically modified heparins, on a murine mammary adenocarcinoma cell line adhesion properties, and the relationship with the presence of heparin binding sites in tumor cells. We show that heparin increases tumor cell adhesion in a dose-dependent manner. When the number of heparin binding sites was regulated, by culturing the cells with different FCS concentration for 24 hours, a correlation between binding capacity and heparin effect on cell adhesion was observed. The increment on cell adhesion by heparin was lower on cells with less heparin binding sites. Moreover, only heparin and a chemically modified heparin (partially N-desulfated N-acetylated), which bound to heparin-receptor, retained the ability to stimulate cell adhesion, while other modified heparins lost both effects. The increase in cell adhesion was observed on plastic dishes, albumin, as well as on fibronectin pre-coated ones suggesting that heparin effect is substratum independent. Our results show a direct relation between heparin binding to specific cell receptors and increase in cell attachment. (J Biol Regul Homeost Agents 2005; 19: ) KEY WORDS: Heparin, adhesion, cell receptor, fibronectin, tumor Received: Revised: Accepted:
It is well known that mast cells (MC) are abundant around solid tumors, but their functional significance is not entirely clear. Tumor infiltrating MCs have been correlated with an increase in growth and invasion in several types of experimental and human cancers. Other researchers have demonstrated that MC act as a host defense against tumor (see for review 1). Particularly, our investigation on the role of MCs in relation to the tumorigenicity of two murine mammary adenocarcinomas has shown that they are able to modulate several steps in tumor development (2). As a consequence of tumor activation MCs release heparin (3, 4) and heparin can bind to specific cellsurface receptors inhibiting cell proliferation in vitro (5) and experimental metastases in vivo (6). Heparin has been widely used for almost 60 years in clinical treatments for its anticoagulant and antithrombotic properties. However, during the last 15 years several evidences on clinical and animal models have shown that heparin could also contribute to the inhibiti on of cancer growth and metastatic dissemination (see for review 7). Al though its mechanism of action is not completely understood
some, specific properties of heparin have been suggested to be responsible for its antimetastatic a c t i v i t y. Indeed, a lower number of experimental metastases has been associated with the effect of heparin in blocking heparanases and the activity of other enzymes necessary for migration and transvasation of tumor cells (8). A lower survival of tumor cells on the vasculature as a consequence of the antithrombotic and profibrinolytic activities of heparin has also been suggested as mechanism of heparin antimetastatic activity (6, 9). Adhesion is also an important feature related to migration and invasion on metastasis development. Molecular events involved in such cell-substrate adhesi on have been elucidated by studies on adhesion molecules and their cell surface receptors (10,11). Heparin possesses several properties by which it could modulate cell adhesion. First, it is a molecule closely related to heparan sulfate, a proteoglycan abundant in the extracelular matrix. The role of glycosaminoglycans on cell adhesion has been examined and, particularly, the heparan sulfate proteoglycan has been implicated as a ligand for a cell surface receptor that transducts signals via protein
0393-974X/000-08 $15.00
© Wichtig Editore, 2005
INTORODUCTION
Heparin increases cell adhesion.
kinase C pathways (12,13). Second, various components of the ECM are already known to contain heparin-binding sites, including fibronectin, laminin and collagen (14). The effect of heparin on cell adhesion to surfaces precoated with these proteins varied according to the substrate, the cell type and the specific experimental conditions (15). Finally, receptors for heparin/heparan sulfate have been identified and could also modulate cell adhesion. The presence of heparin-binding sites has been documented in several normal cells (16) as well as tumor cells with different metastatic ability (5, 17, 18). The precise role of these receptors is not well defined. It is possible that binding of heparin to its receptor inhibits cel l growth. Our previous results on adenocarcinoma cells, are in agreement with this hypothesis. Even more, heparin and chemically modified heparins which retained the binding capacity, had a tumor cell antiproliferative activity (4, 5). Other researches (16) have also demonstrated an antiproliferative activity on smooth muscle cells. The antiproliferative activity of heparin is associated with inhibition of oncogenes involved in cell proliferation, such as c-fos and c-myc by the protein kinase C (PKC) pathway (19, 20). In the present work, we studied the effect of heparin and chemically modified heparins without anticoagulant activi ty on the adhesion of an adenocarcinoma cell line (LM3) obtained from a murine mammary adenocarcinoma that expresses a high number of heparin-receptors (5) and in which experimental lung metastasis has been inhibited by heparin (6). We analyzed and correlated cell adhesion with the binding of heparin to cell surface binding sites.
MATERIALS AND METHODS Heparins Porcine intestine heparin sodium salt wi th anticoagulant acti vity (196 USP U/mg), and chemically modified heparins: 1) O-desulfated heparin (O-des) sodium salt; 2) O/N-desulfated N -acetylated heparin (O/N-des N-Ac) sodium salt; 3) N-desulfated heparin (N-des) sodium salt and 4) N-desulfated N-acetylated heparin (N-des N-Ac) sodium salt with less than 5 USP U/mg each were a generous gift by Syntex S.A., Buenos A i r e s , Argentina. The composition of heparin and heparin derivatives used, was obtained by NMR analysis of 13C at 50 MHz accordi ng to manufacturer ’s specifications; all data in relative % (moles/100 moles of compound) was described previously (21). Tumor cell line A murine mammary adenocarcinoma cell line (LM3) that was obtained in our laboratory from primary 2
cultures of a BALB/c transplantable mammary adenocarcinoma M3 was used (22). LM3 cells were maintained in tissue culture flasks (75 cm2; Falcon; Lincoln Park; NJ) in MEM (Gibco BRL, Gaithersburg, MD) plus glutamine (2 mM) and gentamycin (80 µg/ml) supplemented with 5% fetal calf serum (FCS) (GEN, Buenos Aires; Argentina). The cells were kept in a humidified 5% CO2-air atmosphere at 37 °C. For each assay, subconfluent monolayer cultures (about 70-80% confluent) growing in MEM supplemented with 5% FCS and, for the last 24 hours, with different concentrations of FCS (see below), were detached by exposure to trypsin-EDTA ( 0 . 2 5 % ; 0.075% respectively). Cell suspensions were washed twice with serum-free medium and maintained in the same medium for 2 hours to allow replenishment of cell-surface components. Cell viability was over 95 % as assayed by the trypan blue exclusion test. Cell aggregation assay Aliquots from LM3 cell suspensions (2 x 105 cells/ml) without (control) or with heparin (250 (µg/ml) were taken at different times (0 to 90 minutes). The percentage of cell aggregates was determined by scoring more than 200 cells under phase contrast microscope (NikonJapan). Monodispersed cells or cells forming aggregates from 2 to 7 cells per group were registered. Data were collected at random in triplicate. Adhesion assay LM3 cells were plated on 24-well plastic trays, at 1x105 cells/well in 0.5 ml of serum-free MEM without (control) or with different concentration of heparin or chemically modified heparins (0 to 500 µg), and incubated for different times (0 to 120 min). T h e heparin concentration and incubation times for each experimental condition are indicated in the Results secti on for each particular case. Cells were maintained in MEM without FCS for the last 24 h previous to the experiment unless otherwise stated. At the end of the assay, cells were carefully washed with ice-cold phosphate-buffered saline (PBS) and the amount of attached cells was eval uated by measurement of DNA content with Hoechst 33258 (23). The rate of adhesion was expressed as a percentage of the total number of cells seeded. Binding of 3H-heparin Subconfluent cell monolayers, growing for the last 24 hrs at different FCS concentrations (0%, 2% and 5%), were used for 3 H-heparin ( 3 H-hep) binding assay. Heparin binding was determined as previously described (5). Briefly, cells were washed three times with cold PBS and then, 120 µl/well of 3H-hep (2x10 5 cpm/well) (0.57 mCi/mg) (New England Nuclear; Boston MA) were added to the cells for 30 min. This is the time needed to reach equilibrium between heparin
Bertolesi et al Fig. 1. Effect of heparin on LM3 cells adh es ion as a f unct ion of time . Subconfluent monolayers of LM3 cells grown the last 24 h without FCS, we re d eta che d a nd t heir adh es ion at diffe rent time wa s determined in media in the absence (❏) or presence (■) of heparin (250 µg/ml). The percent adhesion was calculated as the DNA content of the attached cells divided by the total DNA added to the well. Data show mean ± SD (n=4) of one representative experiment out of three performed independently (*= p< 0.05).
and its receptor. Free 3 H-hep was eliminated by washing two times (200 µl/well) with PBS at 4 ºC. The cells were solubilized in 0.2 N NaOH (100 µl) and then neutralized with 0.2 N HCl. Non-specific binding was measured in the presence of 100-fold excess of unlabeled heparin. Radioactivity was determined using a liquid scintillation counter. Wells without cells were also exposed to 3H-hep and binding to plastic (< 5%) was considered background and subtracted. Fibronectin substrata Twenty four well plates were precoated with fibronectin: 0.33 ml of PBS containing 12 µg/ml of human plasma fibronectin. Sigma; St. Louis, MO, were added to each well and incubated for 2 h, at 37 °C. Then, wells were washed with PBS and allowed to dry in a laminar flow. Non-specific binding sites were blocked by incubation with bovine serum albumin (BSA, 1 mg/ml) for 30 minutes. Control wells were treated similarly with PBS without fibronectin and, subsequently with BSA. The RGDS cell-binding site was tested by incubating an aliquot of cell suspension with RGDS polypeptide (100 µg/ml), for 30 minutes at 37 °C before starting the cell adhesion assay. Statistic analysis All the experiments were repeated at least three times with four replicates for each point in cell adhesion assays and triplicate in binding assays. D i fferences between groups were analyzed by ANOVA using the SPSS program version 5.0 (SPSS Inc, Chicago Illinois, USA) and considered statistically significant when p < 0.05.
RESULTS Heparin effect on cell adhesion A time course adhesion analysis of LM3 cells grown in serum-free medium for 24 h, was performed in the absence or presence of heparin (250 µg/ml) (Fig. 1). Between 15 and 60 minutes, heparin significantly increased LM3 cell adhesion, with a value near to 75% at the time point where 50% of control cells were added. At 120 min, about 90% of the total treated and untreated cells were attached. Heparin stimulation of cell adhesion was dosedependent, as shown on Fig. 2. In 45 minutes and in the presence of heparin (100 µg/ml), 70% of the heparin-treated cells adhered, while only 45% of control cells attached. Heparin effect on LM3 cell aggregation Knowing that heparin has the capacity to aggregate platelets and other cell types, we tested its ability to do so on LM3 cell suspensions, and thus influence the adhesion evaluation test. The microscope examination of adhered cells in both control and experimental cultures showed that they were principall y monodispersed, w ithout cl usters. Furthermore, when cell suspensions containing heparin (250 µg/ml) were examined at different times (0 to 90 minutes), 70-80 % of the cells did not form aggregates (Fig. 3). Only 20 % of the cells were on clusters of two cells without significant differences compared to controls, suggesting that the effect on cell adhesion is not mediated by cell-cell aggregates. 3
Heparin increases cell adhesion. Fig. 2. Dose dependence of heparin effect on LM3 cells adhesion. LM3 cells grown the last 24 h in medium without FCS, were detached and their adhesion was determined in the absence (control) or presence of hep arin (0 .1 -5 00 µg/m l) for 4 5 minutes. The DNA content of the att ac hed a nd t ota l ce lls wa s mea sur ed a nd is ex pres se d a s percent of the total DNA added to the well. Data show mean ± SD (n=4) of one representative experiment out of three performed independently (*= p< 0.05).
Relationship between cell adhesion and heparin binding receptor To test the possibility that the heparin-mediated increase on cell adhesion was associated to specific binding to cell surface receptors, we employed two experimental approaches: a) we analyzed the effect of heparin on cells expressing different number of heparin-binding sites; and b) we compared the effect of chemically modified heparins on their heparinreceptor binding capacity and their effect on cell adhesion. a) Cell adhesion and number of heparin binding sites The primary tumor M3 has heparin receptors that diminish when the cells are incubated at higher FCS concentration (6.9 x 105 and 3.5 x 105 sites/cell on cells grown without or with 2% of FCS, respectively) (5). Thus, we tested whether LM3 cells had the same behavior and we performed a 3H-heparin binding assay to LM3 cells grown in different concentrations of FCS. The cells were kept in 0; 2 and 5% of FCS for the last 24 h previous to the experiment. The specific binding was highest for cells grown without FCS and diminished to 43% and 17%, for cells maintained in MEM plus 2% and 5% of FCS respectively (Fig. 4A). This decrease was not consequence of binding sites blocked by heparin-like molecules from FCS, since pre-incubation with FCS before binding assays showed similar results (data not shown). Then cells were assayed for their cell attachment properties in the presence of heparin. Interestingly, while LM3 grown without FCS required approximately 45 min to reach 50% of cell adhesion (fig. 1), this 4
time diminished in cells grown the last 24 hours in presence of 2 and 5% of FCS to 37 and 26 minutes respectively. In order to analyze the effect of heparin, the kinetics of cell adhesion was analyzed. Fig. 4B shows heparin stimulation of cell adhesion, measured when 50% of control cells were adhered. The highest adhesion was observed in cells grown in MEM without FC S and showed a good correlation wi th the presence of heparin-binding sites. b) Effect of chemically modified heparins on cell adhesion to plastic It has been shown that chemical modifications of the heparin molecule modulate its binding affinity to heparin receptors. Fig. 5 shows the displacement of 3 H-heparin binding in the presence of 100 molar excess of heparin and chemically modified heparins. We observed that, while N-desulfated N-acetylated heparin partially retained heparin-receptor binding c a p a c i t y, other chemical modifications such as Odesulfation and N-desulfation rendered heparins without binding properties, as was previously shown (5,16). Next, we studied the effect of these heparins on the adhesion capacity of LM3 cells. These experiments were performed on plastic and on albumin-coated dishes to diminish possible interactions of heparin charges to plastic. Similar effects were observed in both conditions, although the number of attached cells was greater on plastic than on albumin coated dishes (data not shown). We observed that O-des, N-des and O/N-des N-Ac heparins lost the capacity to increase cell adhesion while N-des N-Ac retained it (Fig. 6).
Bertolesi et al Fig. 3. Effect of heparin on LM3 cell aggregation. Aliquots from a LM3 cells suspension (2 x 105 cells/ ml) without (❏ ) or with 2 50 µg/ml of heparin (■) were taken at different times and the number of monodispersed cells (1) or cell clusters (2 to 7 cells per cluster) were counted under light field microscopy. A minimum of 200 cells were counted. Results are expressed as percentage of total cells.
These results show the correlation between binding capacity to their receptor and modulation of cell adhesion for both, heparin and the chemically modified heparin N-des N-Ac. Effect of heparin and chemically modified heparins on cell adhesion to fibronectin Besides binding to its receptor, heparin could also interact with extracellular matrix proteins, such as FN, known to contain heparin-binding domains (14, 24). FN (4 µg/well) induced a slow but significant increase on LM3 cell line adhesion (52 ± 3 and 64 ± 3% on albumin and FN substrata, respectively) (Fig. 6). We also tested whether heparin and chemically modified heparins could modify LM3 cells attachment to FN. We observed that heparin, as well as N des NAc heparin, increased cell adhesion to FN substrate. All other chemically modified heparins, had no activity (Fig. 6). FN has been shown to interact with integrins, many of which can recognize the specific peptide sequence RGD in cell attachment domains (10). When LM3 cells were treated with an excess of RGDS (100 µg/ml), the adhesion to fibronectin-substrate diminished. H o w e v e r, when cells were incubated simultaneously with RGDS and heparin, adhesion was restored, and was higher than that observed in non treated cells (Fig. 6). This shows that heparin increases LM3 cell adhesion to FN independently of the cell-FN binding site. Moreover, heparin effect on cell adhesion to FN substrata was similar to that observed for plastic and albumin substrata.
DISCUSSION Proteoglycans and gl ycosami noglycans are components of the extracellular matrix. They can regulate tumor development by their interaction with cell membrane proteins (25). Heparin is a glycosaminoglycan especially present in neoplastic matrices due to the high concentration of MCs in active degranulation elicited by tumor cells and cells from tumor stroma. Cell adhesion to extracellular matrix is a key event in the metastatic cascade and heparin could modulate it. This paper demonstrates that LM3 is a good model to study the influence of heparin and heparin binding sites on cell adhesion. We show that heparin increases LM3 cell adhesion in a dose-dependent manner. When the number of heparin binding sites was regulated by culturing the cells with different FCS concentration for 24 hours, a correlation between binding capacity and heparin effect on cell adhesion was observed. Moreover, when chemically modified heparins without anticoagulant activity were tested, we observed that only heparins with affinity to its receptor (native heparin and N-des N-Ac heparin), increased cell adhesion. The correlation between adhesion and the binding of heparin to its receptors was observed in other cell types. For example, Castellot et al (16) showed that the binding and internalization of heparin by vascular smooth muscle cells, modified cell adhesion. A significant increase in cell adhesion was detected at heparin concentration of 50 µg/ml or higher (approximately 10 USP/ml). Although these concentrations could be higher than those used on clinical treatments, it must be considered that not all 5
Heparin increases cell adhesion.
Fig. 4A) Heparin binding to LM3 cells, grown 24 h without or with different FCS concentrations (2% and 5%). 3H-Heparin ( 1 0 5 cpm/well) was added for 30 min at 4 ºC to washed monolayers of LM3 cells. Non specific binding with 100-fold excess unlabeled heparin was subtracted. Specific binding is exp res se d a s mea n ± SD ( n=3 ) of one rep res ent at iv e experiment out of three performed independently.
Fig. 4B) Heparin effect on the adhesion of LM3 cells grown the last 24 h on medium without or with different FCS content. Adhesion of control and heparin treated cells (200 µg/ml), measured as shown in Fig. 1 was determined. Data show the time (min) for the adhesion of 50% of control cells and the percentage of heparin treated cells adhered at this time point.
the effects observed closely correlate with the anticoagulant properties of heparin and, that this could be due to the fact that heparin is a highly heterogeneous molecule, which binds to its receptor in a very small percentage (approximately 1%), as we have previously shown (5). Thus, as is the case with antithrombin-heparin interaction (26), heparin binding to its receptor may require a specific sequence, as suggested by displacement results observed using modified heparins. Taking into account that heparin is a molecule structurall y rel ated to heparan sulfate, and considering that both act similarly in several biological or biochemical responses such as bindi ng to receptors, antiproliferative activity or migration inhibition (4, 16, 27), it is likely that the latter may be the natural ligand for heparin receptors. Heparan sulfate proteoglycan, as well as FN, is found in the attachment portion of the cell while the detachment sites contain chondroitin sulphate and hyaluronic acid (28). It is well accepted that the cell adhesion process starts with the attachment of cells to specific sites of adhesive matrix glycoproteins, like FN, which is recognized by cell surface receptors of the integrin family (10). Although cell surface receptors to matrix proteoglycans such as heparan sulfate are less known, it was recently observed that sulfated glycolipids promoted adhesion, acting as receptors for brevican, other abundant proteoglycan from the
extracellular-matrix (29). Moreover, although the heparin/heparan sulfate receptor has not been characterized yet, monoclonal agonist antibodies against this molecule induced the same signaling pathway as heparin on its antiproliferative activity (15). Thus, heparin interaction with its receptor may induce two different biological responses: inhibition of cell proliferation and increase in cell adhesion. FN-coated wells increased LM3 cell adhesion when compared to albumin pre-coated ones. Although FN has several domains containing cell-binding sites, RGDS peptide was able to abolish the increase in cell adhesion mediated by FN, indicating the involvement of RGD binding sites in cell attachment to this substrate as previously shown (10, 30). Interestingly, our results show that heparin was able to increase cell binding to FN, even in the presence of RGDS, suggesting that other FN-binding sites may be involved in heparin-mediated cell adhesion. T h i s result is in apparent contradiction with observations on smooth muscle cells where heparin and heparan sulfate inhibited cell adhesion (16). H o w e v e r, inhibition of cell adhesion was only observed when heparin was added as a coating with FN, possibly as a consequence of conformational changes in the FN molecules, while no change was detected when heparin was added as a second coat to a preformed FN matrix and at lower doses (24). In this study, on the contrary, the increase on LM3 cell adhesion by heparin, appears to be independent of
6
Bertolesi et al
Fig. 5. Displacement of 3H-heparin binding by heparin and chemically modified heparins. Subconfluent cultures of LM3 cells grown the last 24 h without FCS were incubated with 3Hheparin (2x105 cpm/well) without (0%) or in presence of 100 molar excess of heparin (hep) and chemically modified heparins partially N-desulfated N-acetylated (N-d N-Ac); Odesulfated (O-d); O/N-desulfated N-acetylated (O/N-d N-Ac) or N-desulfated (N-d).
Fig. 6. Effect of heparin and chemically modified heparins on LM3 cell adhesion to albumin (■) or fibronectin (❏) substrate. LM3 cells (1x105 cell/well) were added for 45 min on 24 well plas tic dis hes wit hout (c ontrol -C-) or with heparin or chemically modified heparins (250 µg/ml). Other cells were incubated with RGDS peptide (100 µg/ml) for 30 minutes bef ore s tart ing c ell ad hes ion a ss ay s. Adhe sion wa s determined by DNA content and expressed as percent of total added. Data shows mean ± SD (n=4) of one representative experiment out of three performed independently. °p
