CD44 Modulates Hs578T Human Breast Cancer Cell Adhesion, Migration, and Invasiveness

Experimental and Molecular Pathology 66, 99–108 (1999) Article ID exmp.1999.2236, available online at http://www.idealibrary.com on

REVIEW CD44 Modulates Hs578T Human Breast Cancer Cell Adhesion, Migration, and Invasiveness1

Andrea Herrera-Gayol* and Serge Jothy†,2 *Department of Pathology, McGill University, Montreal, Quebec, Canada; and †Department of Laboratory Medicine and Pathobiology, University of Toronto, and Sunnybrook and Women’s College Health Sciences Centre, Sunnybrook Site, Toronto, Ontario, Canada

Received December 22, 1998; accepted January 13, 1999

CD44 is an adhesion molecule that has been implicated in tumor progression of epithelial and nonepithelial tumors. One of its variants, CD44v6, is involved in the production of experimental metastasis. Previous reports have indicated that in human breast cancer the overexpression of CD44, and moreover the presence of CD44v6, correlated with poor prognosis. This study focuses on the role of these molecules in in vitro invasion of breast cancer cells. The effect of antibodies against all CD44 isoforms and CD44v6 was evaluated in different in vitro experimental assays that are closely related to tumor cell invasion in vivo: adhesion to hyaluronan and purified extracellular matrix components; cell motility; haptotaxis; and invasion of purified extracellular matrix components. The highly metastatic human breast cancer cell line Hs578T was used in all assays. Our results show that both antibodies have a blocking effect on cell migration, on haptotatic migration, on in vitro invasion, and on adhesion to hyaluronan and purified extracellular matrix components. In conclusion, our data show that, in addition to its participation in adhesion to components of the extracellular matrix, CD44v6 is involved in the motility and in invasion of tumoral cells. q 1999 Academic Press Key Words: adhesion molecules; breast cancer; adhesion; motility; invasion.

INTRODUCTION

Tumor cell invasion is composed of a minimum of four steps: cell adhesion to the extracellular matrix, degradation of the extracellular matrix components, and tumor cell motil¨ ity followed by cell detachment (Babaı, 1976; Fidler et al., 1978; Liotta, 1986). Homotypic cell–cell interactions and cell–matrix interactions are involved in each of those steps (reviewed in Hart and Saini, 1992; and in Behrens, 1993). A restricted number of cell-associated proteins can fulfill this dual role. One of them is CD44, which encompasses a family of cell adhesion glycoprotein isoforms (reviewed in Rudzki and Jothy, 1997; Naor et al. 1997). The variant CD44 isoforms differ from the standard form CD44s (also called “hematopoietic,” CD44H) in their proximal extracellular region. In that region, 10 exons, designated v1 to v10, can be inserted by alternative splicing (Screaton et al., 1992). Hyaluronan, a negatively charged high-molecular-weight glycosaminoglycan present at high concentrations in most carcinomas, is the main ligand for CD44. In addition, CD44s can bind to collagen and fibronectin (Lesley et al., 1993). The intracytoplasmic tail of CD44 is connected to the cytoskeleton by ankyrin (Bourguignon et al. 1992) and by proteins of the ezrin, radixin, and moesin complex (Tsukita et al., 1994). CD44 is involved in leukocyte adhesion to endothelial cells and T cell activation (Haynes et al., 1989;

1

Supported by operating grants from the Medical Research Council and the National Cancer Institute of Canada to S.J. A.H-G. is supported ´ by a Fellowship from the FRSQ-FCAR Sante, Quebec, Canada. 2 To whom reprint requests should be addressed at Division of Anatomic Pathology, Sunnybrook and Women’s College Health Sciences Centre, Sunnybrook Site, 2075 Bayview Avenue, Room E441e, Toronto, ON, Canada, M4N 3M5. E-mail: serge.jothy@ sunnybrook.on.ca.

0014-4800/99 $30.00 Copyright q 1999 by Academic Press All rights of reproduction in any form reserved.

99

100 Arch et al., 1992), recirculation of lymphocytes (Stamenkovic et al., 1991), organogenesis (Wheatley et al., ¨ 1993), and tumor progression (Gunthert et al., 1991; Bartolazzi et al., 1995; Koochekpour et al., 1995). A CD44 isoform with a sequence encoded by exon v6 has been shown to induce metastatic capacity to nonmetastatic derivatives ¨ (Gunthert et al., 1991). Motility is a required step in the acquisition and maintenance of tumor cell invasion and metastasis, a process already emphasized in 1863 by Virchow. Since then, several groups have shown that the in vitro capacity of cells to move correlates with their invasive behavior in vivo (Mohler et al., 1988). Cell motility is a complex process requiring the coordination of several changes: polarization, formation of protrusive structures at the leading edge, cell–substratum attachments, multiple force-generating mechanisms, and cell detachment at the trailing edge (reviewed in Lauffenburger and Horwitz, 1996). Adhesion molecules and antiadhesion molecules play important functions in those steps (Jothy et al., 1995; reviewed in Lauffenburger and Horwitz, 1996; Gumbiner, 1996). A number of studies with melanoma and glioma cell lines have shown that cells overexpressing CD44 are more motile, more invasive, and more adherent to hyaluronan (Birch et al., 1991; Koochekpour et al., 1996; Thomas et al., 1992, 1993; Goebeler et al., 1996). The role of CD44 in cell motility is less well defined for epithelial cells, in particular breast cancer cells. In this study we tested the hypothesis that CD44, previously described mostly as an adhesion protein, also acts as a determinant in the motility of human breast cancer cells and therefore is involved in their invasive properties.

MATERIALS AND METHODS Cell Culture The human breast cancer cell lines Hs578T and MCF-7 were obtained from the American Type Culture Collection (ATCC, Rockville, MD). Cells were cultured in DMEM (Gibco, Burlington, Ontario, Canada) supplemented with 5% fetal bovine serum (Gibco), 2 mM glutamine, and 100 IU/ml penicillin/streptomycin. Monoclonal Antibodies Various antibodies were used in the assays. The antiCD44s antibody against all CD44 isoforms, clone F.10.44.2,

HERRERA-GAYOL AND JOTHY

was purchased from Novocastra Laboratories (Newcastle, UK). The anti-CD44v6 antibody, which recognized CD44 isoforms containing variant v6, clone 2F10, was obtained from R & D Systems (Minneapolis, MN). For the different experiments, the anti-CD44s and the anti-CD44v6 antibodies were used at 1/30 to 1/50 and 1/250 to 1/500 dilutions, respectively. An irrelevant antibody directed against chromogranin A (Boeheringer-Mannheim, Laval, Quebec, Canada) was used as negative control. Migration Assays Migration assays were based on the repopulation of wounded cultures and performed as described (Birch et al., 1991) with modifications. The technique used for coating glass coverslips with hyaluronan was adapted from Thomas et al. (1992). Round glass coverslips were placed in 6well plates. Hyaluronan extracted from rooster comb at a concentration of 5 mg/ml (Sigma, St. Louis, MO) was applied to the coverslips, left at 48C overnight, dried at room temperature for 1 h, washed in PBS, and incubated with 0.2% bovine serum albumin (Sigma) for 2 h at 378C. The coverslips were washed again in PBS before 1 3 105 cells were plated in 100 ml of medium per coverslip. After reaching confluence, cells were “wounded” with a 1-mm tissue scraper along the entire diameter of the coverslip. The medium was removed, and the CD44 antibodies were added to the cells and incubated for 1 h at 378C. An irrelevant monoclonal antibody against chromogranin A (Boeheringer Mannheim) was used as a negative control. Afterward, 2 ml of complete culture medium was added to each well. The cells were allowed to migrate into the wound area for 24 h. Then they were washed in PBS fixed in cold methanol and stained with Gills’ hematoxylin. The wounded areas were photographed on diapositive slides and projected on reading sheets, and all cells were counted within a 9-mm2 area of the coverslip. Experiments were done in triplicate and repeated four times. As an internal control for this assay, we compared the natural migratory capacity of the cell line Hs578T to that of the cell line MCF-7. Adhesion Assays The adhesion assays were performed in 96-well plates (Costar, Cambridge, MA) coated with hyaluronan (500 mg/ ml or 5 mg/ml) or 2% BSA as a negative control as described above for the coverslips. Afterward, all wells were incubated with 0.2% BSA for 2 h at 378C. Cells were trypsinized and

101

CD44 AND Hs578T BREAST CANCER CELLS

allowed to recover for 30 min to 1 h in culture medium. Trypsin/EDTA proved to be the best method for obtaining viable cells. Cells were then resuspended in serum-free medium (for adhesion assays performed for 1 h) or regular culture medium (for assays performed for 24 h) at a density of 2 3 103 cells per 100 ml/well. Before being plated, the cell suspensions were incubated with antibodies against CD44s, CD44v6, or chromogranin A for 1 h at 48C. Adhesion was performed at 48C for 1 h or at 378C for 24 h. The wells were washed in PBS, fixed in cold methanol for 10 min, and stained with Gill’s hematoxylin. Adherent cells were counted under an inverted microscope. Adhesion to Matrigel was performed in 96-well plates coated with 5 mg of Matrigel (Becton Dickinson, Bedford, MA) in a final volume of 35 ml. The wells were allowed to dry at room temperature overnight and reconstituted with 50 ml of serum-free medium. The cells were harvested as described before and resuspended in serum-free medium at a density of 2 3 103 cells per 100 ml/well. Before being plated, the cell suspensions were incubated with antibodies against CD44s, CD44v6, or chromogranin A for 1 h at 48C. Adhesion was performed at room temperature for 2 h. The wells were washed and fixed as described above. Adherent cells were counted in an inverted microscope. Experiments were repeated three to six times.

Haptotaxis Hyaluronan coated inserts were prepared following a slight modification of the procedure of Goebeler et al. (1996). Three inserts containing tissue culture treated polyethylene terephthalate membranes of 8-mm pore size (Falcon, Becton Dickinson Labware, Bedford, MA) were floated overnight at 378C, on a solution of 1 mg/ml hyaluronan. Inserts were allowed to dry at room temperature for 1 h, washed carefully with PBS, and placed inside their companion 24-well plates (Falcon). Hyaluronan at 100 mg/ml in the regular culture medium was pipetted in the lower chamber. Suspensions containing 5 3 104 cells in 100 ml of medium per insert were prepared. The cell suspensions were preincubated with the antibodies for 1 h at 48C before being plated. The chamber units were incubated for 4 h at 378C in a 5% CO2 atmosphere. The cells in the inside face of the membrane were wiped off. The inserts were washed with PBS and fixed in cold methanol. The cells that migrated and reached the underside of the membrane were stained with Gill’s hematoxylin and counted under an inverted microscope. Five microscope fields per insert were counted. Experiments were done in triplicate and repeated three times.

The natural haptotactic migratory capacities of Hs578T and MCF-7 were compared.

In Vitro Invasion Assays Invasion was determined using a modified Boyden chamber system, as described before for the haptotactic motility assays. Cell suspensions containing 1 3 105 cells in 100 ml of medium without serum were incubated with the CD44 or control antibodies for 1 h at 48C and plated. The lower chambers were filled with 800 ml of a 100 mg/ml solution of hyaluronan in DMEM and the cells were incubated for 48 h at 378C. The cells that had invaded the Matrigel layer were counted on the underside of the membrane as described above. Experiments were done in triplicate and repeated three times.

Flow Cytometry Analysis Viable cells (250 3 103 cells) were incubated with the anti-CD44s and anti-CD44v6 antibodies diluted in 0.05% NaN3 in PBS, for 45 min at 48C. Cells were washed in NaN3/PBS and incubated with the secondary FITC-labeled goat anti-mouse IgG antibody (Cedarlane, Hornby, Ontario, Canada) for 45 min at 48C. After washing, fluorescence analysis was performed on a EPICS scan (Coulter Corp., Hialeah, FL). Background controls were established by omitting the primary antibodies.

Statistical Analysis Comparisons of the experiments were conducted using one-way analysis of variance (ANOVA) followed by posttest methods.

RESULTS

The highly invasive and metastatic cell line Hs578T was chosen for this study based on previously published data describing its aggressive biological behavior (Thompson et al., 1992; Bae et al., 1993; Sommers et al., 1994) and our

102

HERRERA-GAYOL AND JOTHY

own data (Table I) Hs578T cells are derived from a primary human ductal breast carcinoma, estrogen receptor negative and vimentin positive, and grow in cultures as “fibroblastlike” cells. We found that Hs578T express a high level of CD44 as shown by flow cytometry (Table I). By contrast MCF-7 cells express a low level of CD44 (measured by flow cytometry and immunohistochemistry; data not shown, and Culty et al., 1994) have a more differentiated morphology, and consequently, were used as low CD44 controls in motility and haptotaxis assays. Contribution of CD44s and CD44v6 in Cell Migration The wound assay was employed to evaluate the migratory capacity of the Hs578T cell line. This assay has been used to evaluate and quantify the motility of different cell types, on a 2-dimensional (2D) system (Allardyce et al., 1979; Birch et al., 1991; Thomas et al., 1992; Goupille et al., 1997; Mathew et al., 1997). The method consists of plating the cells and letting them reach confluence before creating a wound. Cells that are actively motile will move into the cell-free wounded area. The cells were visually monitored at 8, 24, 48, and 72 h. Eight hours after wounding, cellular protrusions were seen projecting into the wound area. At 24 h, tumor cell migration formed partial areas of healing. Healing with total closure of the wound was achieved between 48 and 72 h. A comparable time course of migration into the wound area was observed when the cells were plated on hyaluronan-coated glass or uncoated glass coverslips. Cells were allowed to migrate into the wound area in the presence of serum, which was required for maintenance of cell attachment. MCF-7 were unable to close the wound area, even at 72 h. Using the migration kinetics as baseline of Hs578T motility values, the participation of CD44 was investigated. Antibodies against an epitope common to all CD44 isoforms (antiCD44s), against CD44 variant isoforms containing exon v6 (anti-CD44v6), or against an unrelated control protein

TABLE I CD44s and CD44v6 Expression by Flow Cytometry

Mean fluorescence 6 SD % of positive cells 6 SD P valueb

a

a b

Negative control

CD44s

CD44v6

0.4 6 0.1 3.13 6 2.3

10.6 6 10.2 89.5 6 8.9 ,0.05

1.6 6 0.4 93.3 6 5.6 ,0.05

Data from three different experiments. Posttest performed after ANOVA.

were added immediately after the culture was wounded. The rationale for using an antibody against the CD44v6 isoform is based on the notion that it plays an important role in ¨ tumor progression (Gunthert et al., 1991). The migration assays were terminated at 24 h. The anti-CD44s antibody had a mean inhibitory effect of 40% on cell migration with no significant difference between hyaluronan coated and uncoated surfaces (Figs. 1 and 2). Mean inhibitory effects on cell migration of 41 and 53% were achieved with the antibody directed against the v6 epitope of CD44 when the cells were plated on uncoated or hyaluronan-coated glass coverslips, respectively (Fig. 1). As a compounding factor, it is possible that the antibodies have an inhibitory effect on cellular proliferation, reducing the number of cells seen in the wound area. We checked that possibility by performing cell counts. No inhibitory effect on cell proliferation was seen when the cells were cultured in the presence of the antibodies (data not shown).

FIG. 1. Participation of CD44s and CD44v6 in cell migration. Cells were plated on hyaluronan coated or uncoated glass coverslips, “wounded,” and incubated with anti-CD44 and anti-CD44v6 antibodies, as described under Materials and Methods. Cells were allowed to migrate for 24 h. Results are shown as the mean 6 SD of percentage of the control. The effect of the anti-CD44s and anti-CD44v6 antibodies on cell migration was tested on uncoated (A) or hyaluronan coated coverslips (B). Difference in mean values among groups by ANOVA (A) P 5 0.0002, (B) P 5 0.003. (A) Posttest* P , 0.01; ** P , 0.01; (B)*P , 0.05, ** P , 0.01.

CD44 AND Hs578T BREAST CANCER CELLS

103 epithelium, the breast cancer cells were exposed to various concentrations of this extracellular-matrix component. As shown in Fig. 3A, cells adhere more to hyaluronan coated wells than to BSA coated wells and the level of attachment increases with increasing hyaluronan concentration. Adhesion of Hs578T was markedly reduced by preincubating the cells with an antibody against all isoforms of CD44. A mean inhibitory effect of 50% was observed at 1 h (data not shown) and a mean inhibitory effect of up to 63% was seen when CD44 antibodies were present during the 24-h adhesion assay to hyaluronan (Fig. 3B). Inhibition of CD44 resulted in comparable inhibition of adhesion to the low and high concentrations of hyaluronan. Blocking the v6 epitope of CD44 inhibited cell adhesion to hyaluronan at 24 h with an inhibitory effect comparable to that achieved by anti-CD44s (Fig. 3). To gain more information on the binding properties of the

FIG. 2. Participation of CD44s in cell migration. Cells were plated on uncoated surfaces. (A) Cells left to migrate for 24 h in the presence of negative control antibody have moved into the wound area. (B) The motility of cells incubated in the presence of anti-CD44s antibody is markedly inhibited. Bar, 100 mm.

Therefore, CD44 including its v6 isoform plays a significant role in cellular migration independent of cell growth. Adhesive Properties of CD44 and Variant CD44v6 Tumor cell attachment to the extracellular environment is necessary for motility. Consequently, it was relevant to test the role of CD44 in adhesion to hyaluronan and to purified extracellular matrix components present in Matrigel. Adhesion assays were performed for 1 to 2 h, a time during which no significant cell migration occurs but during which cells bind to the extracellular substratum. In addition, because cell migration might depend on CD44-mediated adhesion, we also measured adhesion to hyaluronan, 24 h after cell seeding. Because carcinomas grow in vivo in a microenvironment that contains a markedly increased concentration of hyaluronan (De la Torre et al., 1993), compared to normal

FIG. 3. Participation of CD44 in cell adhesion to immobilized hyaluronan. (A) Cells were plated on BSA or hyaluronan (HA) at 500 mg/ml or 5 mg/ml and allowed to adhere for 24 h. Results are shown as the mean 6 SD of percentage of control. ANOVA P 5 0.0013; posttest *P , 0.01. (B) Cells were preincubated with the anti-CD44s and anti-CD44v6 antibodies and plated into the wells coated with 500 mg/ml or 5 mg/ml of hyaluronan or BSA for 24 h as described under Materials and Methods. Results are shown as the mean 6 SD of percentage of control. C, control antibody; S, antibody against all isoforms of CD44; v6, antibody against Cd44 v6 isoform. Cells plated on 500 g/ml HA; ANOVA P 5 0.007; cells 5 mg/ml; ANOVA P 5 0.0001. Posttest *P , 0.05.

104

HERRERA-GAYOL AND JOTHY

breast cancer cells to a more complex extracellular substrate, the same type of adhesion experiments were performed using Matrigel. CD44 is also involved in this interaction. A significant inhibition of binding was observed when antibodies to standard and variant CD44 were used. The mean inhibitory effect of the anti-CD44s and anti-CD44v6 antibodies was 45 and 37%, respectively (Fig. 4). Haptotaxis and in Vitro Invasion Haptotaxis is the capacity of cells to migrate toward molecular species bound to a solid support. This type of assay is consistent with the migratory signals sensed by invading tumor cells in vivo (Aznavoorian et al., 1990). In the blocking experiments with the cell line Hs578T, the results show that the antibodies against all CD44 isoforms or those against isoforms containing the v6 epitope have a mean inhibitory effect of 41 or 43%, respectively (Fig. 5). The Hs578T cells migrated through the pores of the membranes and bound to their inferior surface where hyaluronan was previously adsorbed. In contrast, MCF-7 cells were unable to migrate to the underside of the membranes. Therefore, this cell line was not used for blocking experiments. The fact that a low CD44 expressor was unable to migrate toward hyaluronan was used to strengthen the relationship between the expression of CD44 and the haptotactic capacity of the cell line Hs578T. To assess the role of CD44 in invasion into and through a complex biological extracellular matrix substrate, the Hs578T cells were exposed for 48 h to Matrigel before those

FIG. 4. Participation of CD44 in cell adhesion to Matrigel. Cells were harvested and incubated with the anti-CD44s and anti-CD44v6 antibodies prior to seeding on wells coated with diluted Matrigel as described under Materials and Methods. Cells were allowed to adhere for 2 h. Results are shown as the mean 6 SD of percentage of control. Differences in mean values among groups by ANOVA P , 0.001, Posttest *P , 0.05, **P , 0.05.

FIG. 5. Participation of CD44 in haptotatic migration toward hyaluronan. (A) The haptotatic migration of the cell line Hs578T using membranes coated with 1 mg/ml of hyaluronan was evaluated as described under Materials and Methods. Results are shown as mean 6 SD of percentage of control. Differences in mean values among groups by ANOVA P , 0.04. Posttest *P , 0.05, **P , 0.05.

cells that have fully penetrated through the substrate were counted. This process was also found to depend on CD44 as mean inhibitory effects of 56% and 65% were observed with the v6 and standard antibodies, respectively (Fig. 6).

DISCUSSION AND CONCLUSIONS

In this study we asked whether CD44, a protein thus far characterized mostly as a cell adhesion protein, could also have a role in the migration of breast cancer cells. We also investigated whether in this system CD44 is involved in two functions related to tumor cell migration: haptotaxis and invasion. Various studies have shown that CD44 is implicated in migration and invasion of different types of mesenchymal cells (Birch et al., 1991; Koochekpour et al., 1996; Thomas et al., 1992; Goebeler et al., 1996). The role of CD44 in the motility and invasion of epithelial cell is less clear. In a rat pancreatic cancer cell model, a nonmetastatic tumor cell line acquired metastaic behavior after being ¨ transfected with the CD44v6 isoform (Gunthert et al. 1991). Moreover, lymph node and lung metastasis could be prevented by the injection of a monoclonal antibody against CD44v6 (Seiter et al., 1993). The highly tumorigenic Hs578T human breast cancer cell line was selected as it exemplifies the behavior of a clinically

CD44 AND Hs578T BREAST CANCER CELLS

105

FIG. 6. Participation of CD44 in Matrigel invasion. (A) Invasion assays were performed as described under Materials and Methods. Results are shown as mean 6 SD of percentage of control. Differences in mean values among groups by ANOVA P 5 0.002. Posttest *P , 0.01; **P , 0.01. (B, C and D) The effect of the antibodies against CD44s (C) and against CD44v6 (D) compared to control antibody (B). Arrowheads show cells adherent to the membrane and arrows show 8-mm pores. Bar, 50 mm.

aggressive tumor. Its pathobiological characteristics; expression of vimentin, loss of estrogen receptor expression, fibroblastic-type morphology, and invasive and metastatic potentials, are well documented (Thompson et al., 1992; Bae

et al., 1993; Sommers et al., 1994). Compared to other human breast cancer cell lines like MCF-7, the Hs578T cells express more CD44 (Culty et al., 1994) and are able to synthesize hyaluronan (Heldin et al., 1996). The adhesion

106 assays showed that Hs578 cells adhere to hyaluronan, a process that could be inhibited by CD44 antibodies. Tumor cell motility in a 2D system was evaluated by the wound assay on hyaluronan-coated and uncoated glass surfaces to investigate whether CD44 and its v6 variantcontaining isoforms participate in this process. The results of the wound assay indicate that the display of CD44 on the surface of tumor cells is an important contributor to the various mechanisms of motility. Interestingly, pretreatment of the migrating surface with hyaluronan did not modify the extent of motility inhibition by the CD44 antibodies. This might be consistent with the notion that cellular motility could be activated by a soluble factor present in the culture medium. When cell migration was done without added serum in the medium, solid-phase-attached hyaluronan had to be present for an increase in the cells migratory capacity to be observed (data not shown). Alternatively, it is possible that during the culture period the cells secrete enough hyaluronan to use this endogenously produced ligand in the motility process. The haptotatic migratory capacity of the tumor cells toward substratum-bound hyaluronan was also inhibited by preincubating the cells with the CD44 antibodies. Taking into account the results obtained in the migration and haptotaxis assays, it can be concluded that the CD44 isoforms containing the v6 epitope are involved in the motility of the Hs578T breast cancer cells in both 2D and 3D systems. To evaluate the possibility that the antibodies could have an effect on cellular proliferation or cell death in the previous assays, the cell line was cultured in the absence and in the presence of the antibodies. No effect was observed, confirming that the results obtained in both in vitro motility assays are related to an inhibition of cell migration and not a decrease in cell density. The inhibitory effect of CD44 antibodies on tumor cell migration could be due to the binding of the antibodies to the extracellular portion of cell-bound CD44. In this case, antibody binding might have several consequences: preventing further interaction between CD44 and ligands of the extracellular matrix or alternatively blocking the internal signal that CD44 uses to trigger cytoskeletal changes related to increased motility. A third possibility involves CD44 shed from the membrane, a natural process documented in several ¨ cell types (Goebeler et al., 1996; Ristamaki et al., 1997). In this case, the CD44 antibodies, by complexing shed CD44 molecules, would prevent interaction of the tumor cells with the extracellular ligands. The blocking effect of CD44 antibodies could be explained as follows. Normally, Hs578T cells express CD44 at a high level and also shed it from the cell membrane at a high rate. Shed CD44 is equivalent

HERRERA-GAYOL AND JOTHY

to cell-free CD44 that would still have the ability to bind hyaluronan in the extracellular matrix since the binding domain is located in the distal part of the molecule. This would result in the masking of hyaluronan by cell-free CD44 and consequently result in an inability of the tumor cell to attach to the substrate through its cell-bound CD44. Eventually the tumor cells are more loosely attached to the extracellular matrix and consequently more able to move. In this model, antibodies added to the culture medium would preferably bind to cell-free CD44, and hyaluronan would then be more available for cellular attachment. Overall, the main feature of this model is that CD44 has a dual and opposite role depending on whether it is cellbound or cell-free. This is consistent with a cell adhesion model that we have proposed (Jothy et al., 1995). Experiments showed that under certain conditions CD44 can behave as an anti-adhesion protein (Koochekpour et al., 1996), a process particularly applicable to CD44 isoforms containing the CD44v6 epitope (Santos et al., 1995). One important factor in the detachment–attachment properties of CD44 would be related to the concentration of hyaluronan in the extracellular environment and shedding of soluble CD44 from the tumor cell membrane. Both antibodies blocked cell adhesion and invasion into Matrigel. Matrigel is a solubilized basement membrane preparation extracted from the Engelbreth–Holm–Swarm sarcoma, rich in laminin, collagen, and other extracellular matrix proteins (Kleinman et al., 1982; Timpl et al., 1984). However, its hyaluronan concentration is unknown. Therefore, both antibodies could have a blocking effect on the binding of the cells to hyaluronan or other ligands present in this matrix such as collagen or fibronectin. The inhibition of migration through the Matrigel could be related to an inhibition of the cellular contacts to the extracellular environment that are necessary to transduce motility signals. Several molecules such as RHAMM (reviewed in Turley, 1992), integrins, and cell surface proteoglycans (reviewed in Lester and McCarthy, 1992) are involved in motility; CD44 acts as a part in the total context. Our data showed that in a highly tumorigenic human breast cancer cell line, CD44 and, in particular, CD44v6 are involved in motility and in invasion. The role of CD44 in motility and invasion has been reported by others in different experimental models using mesenchymal cells (Birch et al., 1991; Koochekpour et al., 1996; Goebeler et al., 1996). The expression of CD44v6 in basal cells of different tissues (Wirth et al., 1993; Ruiz et al., 1995; Fasano et al., 1997) leads to the suggestion of a possible role for this variant in cellular motility. Moreover, CD44v6 has been directly implicated in the motility of Langerhans cells (Weiss et al.,

CD44 AND Hs578T BREAST CANCER CELLS

1997) and indirectly in the increased migration capacity of colon cancer cells overexpressing H-blood groups (Goupille et al., 1997). Our contribution to the understanding of the role of the CD44 family in tumorigenesis is based on the description of a function for CD44v6. We suggest that the underlying mechanism is related to the adhesive properties of the molecules in response to the extracellular environment, where hyaluronan plays a key role. This study indicates that in malignant epithelial cells, CD44 and its v6 variant participate in invasion by changes in the adhesion to extracellular ligands that are necessary to create a bridge between signals from the extracellular matrix to the intracellular motility machinery.

ACKNOWLEDGMENTS ¨ We thank Dr. Z. Rudzki, Dr. F. Halwani, Dr. F. Babaı, and Mr. L. LeDuy for their critical comments; Mrs. S. Schiller for her technical assistance with the flow cytometry analysis; and Mrs. L. Fallavollita and Ms. E. Fragiskatos who contributed technical advice.

REFERENCES

Allardyce, R. A., Hunt, J. S., and Stewart, R. J. (1979). A micro macrophage migration inhibition test for the detection of cellular immunity in vitro. J. Immunol. Methods 27, 9–18. Arch, R., Wirth, K., Hofmann, M., Ponta, H., Matzku, S., Herrlich, P., ¨ and Zoller, M. (1992). Participation in normal immune responses of a metastasis-inducing splice variant of CD44. Science 257, 682–685. Aznavoorian, S., Stracke, M. L., Krutzsch, H., Schiffmann, E., and Liotta, L. A. (1990). Signal transduction for chemotaxis and haptotaxis by matrix molecules in tumor cells. J. Cell Biol. 110, 1427– 1438. ¨ ´ Babaı, F. (1976). Etude ultrastructurale sur la pathogenie de l’invasion ´ du muscle strie par des tumeurs transplantables. J. Ultrastruct. Res. 56, 287–303. Bae, S-N., Arand, G., Azzam, H., Pavasant, P., Torri, J., Frandsen, T. L., and Thompson, E. W. (1993). Molecular and cellular analysis of basement membrane invasion by human breast cancer cells in Matrigel-based in vitro assays. Breast Cancer Res. Treat. 24, 241–255. Bartolazzi, A., Jackson, D., Bennett, K., Aruffo, A., Dickinson, R., Shields, J., Whittle, N., and Stamenkovic, I. (1995). Regulation of growth and dissemination of a human lymphoma by CD44 splice variants. J. Cell Sci. 108, 1723–1733. Behrens, J. (1993). The role of cell adhesion molecules in cancer invasion and metastasis. Breast Cancer Res. Treat. 24, 175–184.

107 Birch, M., Mitchell, S., and Hart, I. R. (1991). Isolation and characterization of human melanoma cell variants expressing high and low levels of CD44. Cancer Res. 51, 6660–6667. Bourguignon, L. Y. W., Lokeshwar, V. B., He, J., Chen, X., and Bourguignon, G. J. (1992). A CD44-like endothelial cell transmembrane glycoprotein (GP116) interacts with extracellular matrix and ankyrin. Mol. Cell. Biol. 12, 4464–4471. Culty, M., Shizari, M., Thompson, E. W., and Underhill, C. (1994). Binding and degradation of hyaluronan by human breast cancer cell lines expressing different forms of CD44: Correlation with invasive potential. J. Cell Physiol. 160, 275–286. De la Torre, M., Wells, A. F., Bergh, J., and Lindgren, A. (1993). Localization of hyaluronan in normal breast tissue, radial scar, and tubular breast carcinoma. Hum. Pathol. 24, 1294–1297. Fasano, M., Sabatini, M. T., Wieczorek, R., Sidhu, G., Goswami, S., and Jagirdar, J. (1997). CD44 and its v6 spliced variant in lung tumors: A role in histogenesis? Cancer 80, 34–41. Fidler, I. J, Gersten, D. M., and Hart, I. R. (1978). The biology of cancer invasion and metastasis. Adv. Cancer Res. 28, 149–250. ¨ Goebeler, M., Kaufmann, D., Brocker, E-B., and Klein, C. E. (1996). Migration of highly aggressive melanoma cells on hyaluronic acid is associated with functional changes, increased turnover and shedding of CD44 receptors. J. Cell Sci. 109, 1957–1964. Goupille, C., Hallouin, F., Meflah, K., and Le Pendu, J. (1997). Increase of rat colon carcinoma cells tumorigenicity by a-(1– 2)fucosyltransferase gene transfection. Glycobiology 7, 221–229. Gumbiner, B. M. (1996). Cell adhesion: The molecular basis of tissue architecture and morphogenesis. Cell 84, 345–357. ¨ ¨ Gunthert, U., Hofmann, M., Rudy, W., Reber, S., Zoller, M., Haußmann, I., Matzku, S., Wenzel, A., Ponta, H., and Herrlich, P. (1991). A new variant of glycoprotein CD44 confers metastatic potential to rat carcinoma cells. Cell 65, 13–24. Hart, I. R., and Saini, A. (1992) Biology of tumour metastasis. Lancet 339, 1453–1457. Haynes, B. F., Telen, M. J., Hale, L. P., and Denning, S. M. (1989). CD44—A molecule involved in leukocyte adherence and T-cell activation. Immunol. Today 10, 423–428. Heldin, P., De la Torre, M., Ytterberg, D., and Bergh, J. (1996). Differential synthesis and binding of hyaluronan by human breast cancer cell lines. Relationship to hormone receptor status. Oncol. Rep. 3, 1011–1016. Johnson, M. D., Torri, J. A., Lippman, M. E., and Dickson, R. B. (1993). The role of cathepsin D in the invasiveness of human breast cancer cells. Cancer Res. 53, 873–877. Jothy, S., Munro, S. B., LeDuy, L., McClure, D., and Blaschuk, O. W. (1995). Adhesion or anti-adhesion in cancer: What matters more? Cancer Metastasis Rev. 14, 363–376. Kleinman, H. K., McGarvey, M. L., Liotta, L. A., Robey, P. G., Tryggvason, K., and Martin, G. R. (1982). Isolation and characterization of type IV procollagen, laminin, and heparan sulfate proteoglycan from EHS sarcoma. Biochemistry 21, 6188–6193. Koochekpour, S., Pilkington, G. J., and Merzak, A. (1996). Hyaluronic acid/CD44H interaction induces cell detachment and stimulates migration and invasion of human glioma cells in vitro. Int. J. Cancer 63, 450–454. Lauffenburger, D. A., and Horwitz, A. F. (1996). Cell migration: A physically integrated molecular process. Cell 84, 359–369.

108 Lesley, J., Hyman, R., and Kincade, P. W. (1993). CD44 and its interaction with extracellular matrix. Adv. Immunol. 54, 271–335. Lester, B. R., and McCarthy, J. B. (1992). Tumor cell adhesion to the extracellular matrix and signal transduction mechanisms implicated in tumor cell motility, invasion and metastasis. Cancer Met. Rev. 11, 31–44. Liotta, L. A. (1986). Tumor invasion and metastases—Role of the extracellular matrix: Rhodes Memorial Award lecture. Cancer Res. 46, 1–7. Mathew, A. C., Rajah, T. T., Hurt, G. M., Abbas Abidi, S. M., Dmytryk, J. J., and Pento, J. T. (1997). Influence of antiestrogens on the migration of breast cancer cells using an in vitro wound model. Clin. Exp. Metastasis 15, 393–399. Mohler, J. L., Partin, A. W., Isaacs, J. T., and Coffey, D. S. (1988). Metastatic potential prediction by a visual grading system of cell motility: Prospective validation in the Dunning R-3327 prostatic adenocarcinoma model. Cancer Res. 48, 4312–4317. Naor, D., Vogt Sionov, R., and Ish-Shalom, D. (1997). CD44: Structure, function, and association with the malignant process. Adv. Cancer Res. 71, 241–319. ¨ ¨ Ristamaki, R., Joensuu, H., Gron-Virta, K., Salmi, M., and Jalkanen, S. (1997). Origin and function of circulating CD44 in non-Hodgkin’s lymphoma. J. Immunol. 158, 3000–3008. Rudzki, Z., and Jothy, S. (1997). CD44 and the adhesion of neoplastic cells. J. Clin. Pathol. Mol Pathol. 50, 57–71. ¨ ¨ Ruiz, P., Schwarzler, C., and Gunthert, U. (1995). CD44 isoforms during differentiation and development. BioEssays 17, 17–24. ¨ Santos, C., Chandler, K., Zimmer, S., Fisher, P. B., Gunthert, U., and Anderson, K. W. (1995). Detachment of transformed cells. Role of CD44 variants. Cell Biophysics 26, 1–19. Screaton, G. R., Bell, M. V., Jackson, D. G., Cornelis, F. B., Gerth, U., and Bell, J. I. (1992). Genomic structure of DNA encoding the lymphocyte homing receptor CD44 reveals at least 12 alternatively spliced exons. Proc. Natl. Acad. Sci. USA 89, 12160–12164. Seiter, S., Arch, R., Reber, S., Komitowski, D., Hofmann, M., Ponta, ¨ H., Herrlich, P., Matzku, S., and Zoller, M. (1993). Prevention of tumor metastasis formation by anti-variant CD44. J. Exp. Med. 177, 443–455. Sommers, C. L., Byers, S. W., Thompson, E. W., Torri, J. A., and Gelmann, E. P. (1994). Differentiation state and invasiveness of human breast cancer cell lines. Breast Cancer Res. Treat. 31, 325–335.

HERRERA-GAYOL AND JOTHY

Stamenkovic, I., Aruffo, A., Amiot, M., and Seed, B. (1991). The hematopoietic and epithelial forms of CD44 are distinct polypeptides with different adhesion potentials for hyaluronate-bearing cells. EMBO J. 10, 343–348. Thomas, L., Byers, H. R., Vink, J., and Stamenkovic, I. (1992). CD44H regulates tumor cell migration on hyaluronate-coated substrate. J. Cell Biol. 118, 971–977. Thomas, L., Etoh, T., Stamenkovic, I., Mihm, M. C., Jr., and Byers, H. R. (1993). Migration of human melanoma cells on hyaluronate is related to CD44 expression. J. Invest. Dermatol. 100, 115–120. ¨ Thompson, E. W., Paik, S., Brunner, N., Sommers, C. L., Zugmaier, G., Clark, R., Shima, T. B., Torri, J., Donahue, S., Lippman, M. E., Martin, G. R., and Dickson, R. B. (1992). Association of increased basement membrane invasiveness with absence of estrogen receptor and expression of vimentin in human breast cancer cell lines. J. Cell Physiol. 150, 534–544. Timpl, R., Fujiwara, S., Dziadek, M., Aumailley, M., Weber, S., and Engel, J. (1984) Laminin, proteoglycan, nidogen and collagen IV: Structural models and molecular interactions. Ciba Found. Symp. 108, 25–43. Tsukita, S., Oishi, K., Sato, N., Sagara, J., Kawai, A., and Tsukita, S. (1994). ERM family members as molecular linkers between the cell surface glycoprotein CD44 and actin-based cytoskeletons. J. Cell Biol. 126, 391–401. Turley, E. (1992). Hyaluronan and cell locomotion. Cancer Met. Rev. 11, 21–30. Virchow, R. (1863). Ueber bewegliche thierische Zellen. Arch. Pathol. Anat. Physiol. Klin. Med. 28, 237–240. Weiss, J. M., Sleeman, J., Renkl, A. C., Dittmar, H., Termeer, C. C., ¨ ¨ Taxis, S., Howells, N., Hofmann, M., Kohler, G., Schopf, E., Ponta, H., Herrlich, P., and Simon, J. C. (1997). An essential role for CD44 variant isoforms in epidermal Langerhans cell and blood dendritic cell function. J. Cell Biol. 137, 1137–1147. Wheatley, S. C., Isacke, C. M., and Crossley, P. H. (1993). Restricted expression of hyaluronan receptor, CD44, during postimplantation mouse embryogenesis suggests key roles in tissue formation and patterning. Development 119, 295–306. Wirth, K., Arch, R., Somasundaram, C., Hofmann, M., Weber, B., ¨ Herrlich, P., Matzku, S., and Zoller, M. (1993). Expression of CD44 isoforms carrying metastasis-associated sequences in newborn and adult rats. Eur. J. Cancer 29A, 1172–1177.