Detachment of Transformed Cells

Transcription

1 by Humana Press Inc. All rights of any nature whatsoever reserved /95/ /$7.80 Detachment of Transformed Cells Role of CD44 Variants CARLOS SANTOS, 1 KAREN CHANDLER, 2 STEPHEN ZIMMER, 3 PAUL B. FISHER, 4 URSULA GUNTHERT, 5 AND KIMBERLY WARD ANDERSON *,1 I Department of Chemical Engineering and Center of Membrane Sciences, University of Kentucky, Lexington, KY 40506; 2Department of Chemical Engineering, Georgia Institute of Technology, Atlanta, GA; 3Department of Microbiology and Immunology, University of Kentucky, Lexington, KY; 4Departments of Pathology and Urology, Comprehensive Cancer Center~Institute of Cancer Research, Columbia University, College of Physicians and Surgeons, New York; 5Basel Institute for Immunology, Basel, Switzerland Received April 2, 1994; Accepted May 2, 1994 ABSTRACT A parallel-plate flow chamber was used to quantify the detachment of normal cloned rat embryo fibroblasts (CREF) fibroblasts, ras-transformed CREF fibroblasts (CREF T24), and CREF T24 fibroblasts transfected with a Krev/RAPIA suppressor gene (HK B1) from a confluent monolayer of normal CREF fibroblasts to determine if the expression patterns of CD44 variants (tool *Author to whom all correspondence and reprint requests should be addressed. Cell Biophysics 1 Volume 26, 1995

2 2 Santos et al. wt 110 and 140 kda) corresponded with detachment properties and metastatic potential. In the detachment assay, known shear stresses ranging from dyn/cm 2 were applied to the adherent cells and the number of cells detached from the monolayer after 180 s was determined. Results showed that cellular expression of CD44 variants correlated with the metastatic potential of the cells and with the cells' ability to detach from a monolayer of normal cells. Western blot analysis showed a low level of expression of the CD44 variants in the normal cell line, CREF, and the lowly metastatic cell line, HK B1. Detachment studies showed a low percentage of detachment of both of these cell lines from a normal cell monolayer. Tumor-derived (HK B1-T) and lung nodule-derived (HK B1-M) cell lines were established and both formed tumors and metastasis with reduced latency periods as compared to HK B1, but still showed a markedly delayed latency period compared to the highly metastatic cell line, CREF T24. Both of these cell lines showed a higher expression of the CD44 variants as compared to CREF and HK B1, and detached easier than CREF and HK B1. CREF T24 showed a much higher level of expression of the variants and had a higher percentage detachment than all other cell lines. To further test the role of the CD44 variants in the ability of the cells to detach from the normal monolayer, CREF cells were transfected with a DNA construct that constitutively expresses the CD44 variants and the detachment properties of three randomly selected clones were studied. Clones 2 and 3 showed a low level of expression of the CD44 variants after transfection and detached from the normal monolayer similar to CREF. Clone 1 showed a high level of expression of the CD44 variants and the detachment of these cells was significantly higher than CREF. From these results, it is concluded that in the five cell lines studied, expression of the CD44 variants play a significant role in the ability of the cells to detach from a monolayer of normal cells. It is hypothesized that this detachment may be an important component of a cell's ability to metastasize. Index Entries: Cancer; metastasis; adhesion; flow chamber; CD44. INTRODUCTION Metastasis, or the spread of cancer throughout the body, is a major contributing factor to patient morbidity. This complex process involves a number of distinct steps: detachment of tumor cells

3 Detachment of Transformed Cells 3 from the primary tumor, invasion through normal tissue and intravasation into blood vessels or lymphatics; dissemination of the tumor cells through the vasculature; and arrest, extravasation, and proliferation at second sites. Although biochemical factors have been shown to play a major role in the ability of a cell to metastasize, several physical attributes have been hypothesized to be important in metastasis formation. Two such physical properties include adhesion and detachment. The majority of the previous studies attempting to relate adhesion and detachment properties of tumor cells with tumorigenic or metastatic potential have focused on either the detachment of tumor cells from the primary tumor (1,2) or adhesion of tumor cells to vascular endothelium and the accompanying extracellular matrix (3,4). Previously, we have shown that detachment of tumor cells from normal cells may also reflect an important cellular function associated with the process of metastasis (5). It was shown that following transfection of normal cloned rat embryo fibroblasts (CREF) (6) with a ras-t24 oncogene to form CREF T24 cells, the cells became highly tumorigenic and metastatic (7,8) and were more easily detached from a monolayer of normal cells (5). Transfection of the highly metastatic CREF T24 cells with a Krev/RAPIA suppressor gene to form HK B1 cells resulted in a more normal phenotype, but the doubly transfected cells formed tumors and metastases following subcutaneous implantation only after a long latency period and did not form metastasis in the tail vein assay (8). These cells returned to levels of detachment of the normal cells (5). Recent biochemical investigations of metastasis formation have focused on the transmembrane glycoprotein, CD44. CD44 exists in a variety of forms, the most prevalent being mol wt kda (9). This protein has been found in many types of cells including lymphocytes, monocytes, granulocytes, erythrocytes, fibroblasts, and epithelial cells (10). CD44 is involved in cellular adhesion to high endothelial venules (11), collagen, fibronectin (12,13), and hyaluronic acid (14). Various isoforms of CD44 are generated by alternative splicing of the nuclear RNA, resulting in variation in the extracellular portion of the molecule near the membrane proximal region (15). The genomic DNA contains 20 exons, 10 of which are alternatively spliced (16). Although the standard CD44 molecule is implicated mainly in cell-cell and cell-matrix contacts during migration and lymphocyte circulation (17), recent investigations have demonstrated that the variant isoforms may play a significant role in metastatic and invasive capabilities of tumor cells (9,18-28) and lymphocyte activation (28,29). The variant forms of the molecule have

4 4 Santos et al. larger apparent molecular weights ( kda) and have been detected in cutaneous lymphomas, colon and vulnar carcinomas, neutrophils, and keratinocytes (19). These isoforms are likely to contact further ligands in addition to or instead of the aforementioned ligands for the standard molecule (15,19). Gunthert and coworkers (18) recently isolated cdna clones that encode several splicing variants of the glycoprotein CD44. These variants carry additional extracellular domains and are found to be expressed in metastatic cells but not in nonmetastatic cells. Rudy and coworkers (30) showed that the two most abundant variants found in a pancreas carcinoma cell line have 162 and 85 extra amino acid sequences, respectively. Seiter and coworkers (31) showed that a monoclonal antibody (MAb), 1.1ASML, which recognizes CD44 variants containing the 6v exon, efficiently prevents formation of metastases in rat carcinoma cells. In the present study, the normal (CREF) and transformed cloned rat embryo fibroblasts (CREF T24 and HK B1) were tested for CD44 variant expression using the MAb 1.1ASML and a parallel-plate flow chamber was used to determine if the expression pattern correlated with the detachment properties of these cells from a monolayer of normal cells. The flow chamber was used to apply a controlled shear stress to the cells. In addition, the detachment properties of normal CREF transfected with the DNA expressing CD44 variants were investigated to further test the role of the CD44 variant in detachment. MATERIALS AND METHODS Normal and Transformed Cells The normal and transformed cells used in this study were previously described (7,8). Briefly, the parental cell line CREF (6) serves as the normal cell control in this series of cell lines. This cell line has a normal cell morphology, a near diploid DNA content, and is contact inhibited. The CREF cell line was transfected with a T24 ras oncogene (PT24 C3 and psv2neo) to produce a transformed cell line CREF T24 (7,8). The CREF T24 cell line was further transfected with a Krev/RAPIA suppressor gene and prsvi.1 (containing a hygromycin resistance gene) (32) to produce the HK B1 cell line (8). The Krev/RAPIA gene has been shown to reverse the transformed phenotype of ras transfected cells (8,33). The HK B1 cells showed a flattened revertant phenotype. HK B1-T cells

5 Detachment of Transformed Cells 5 were obtained from a tumor following subcutaneous implantation of HK B1 cells into the flank of rats. HK B1-M cells were obtained from the lung nodules that formed following subcutaneous implantation of HK B1 cells. The properties of these lines have been extensively characterized (8). These normal and transformed cells were grown in 75 and 150 cm 2 Corning tissue-culture flasks in minimum essential media supplemented with 10% fetal calf serum, 10,000 U penicillin/ml, 5 mg streptomycin/ml, 0.3% sodium bicarbonate, and 1 mm glutamine. Cells were maintained in a humidified incubator at 37~ in an atmosphere of 5% CO2 and 95% air. The normal CREF cells were seeded onto Permanox slides with polystyrene chambers (21.6 x 45.7 mm, Model , Nunc, Naperville, IL) and allowed to grow until confluent. Cell suspensions of all the cells were prepared by detaching the cells from the culture flask by trypsinization (0.25% trypsin for 5 rain) and resuspending the cells in culture medium at a density of 1.25 x 106 cells/ml. The viability of all cell lines as measured by trypan blue exclusion was greater than 98%. CD44 Transfected Cells CREF was transfected to produce cell lines expressing CD44 variants. Approximately 2 x 105 cells/5 ml of the culture media described in the previous section were seeded onto 60 mm Costar cell-culture dishes. The cells were incubated at 37~ in an atmosphere of 5% CO2 and 95% air until they reached 40-60% confluency. Five micrograms of cdna that encodes the CD44 variant exons 5v and 6v, 1 ~g of hygromycin drug resistant plasmid, and enough serum-free media to bring the total volume to 100 ~L was added directly to the cells. Twenty-five microliters of Lipofectin Reagent (No. 8292SA, Gibco BRL, Life Technologies, Gaithersburg, MD) in solution with 75/~L of serum-free media was then added. After a 5-h incubation period at 37~ the medium was removed and the cells were incubated in the culture media described in the previous section. Hygromycin-resistant cells were selected in medium containing 400 #g of hygromycin/ml. CREF was also transfected to produce a cell line expressing only hygromycin resistance (CREF/Drug Resistant). The procedure is identical to that just described for transfecting CREF with the CD44 variants with the exception that no cdna encoding the CD44 variant exons 5v and 6v was added.

6 6 Santos et al. Characterization of CD44 Expression patterns of the CD44 variants were obtained by Westem Blotting Analysis. Total cellular proteins (80 #g) were separated by gel electrophoresis of cellular extracts on a 10% polyacrylamide gel with a 4% stacking gel according to the method of Laemmli (34). The proteins were transferred to nitrocellulose filters (Schleicher and 5chuell, Keene, NH) overnight at 18 V constant voltage in a transfer buffer consisting of 15 mm glycine, 20 mm Tris-base, and 20% methanol. The nitrocellulose filter was then placed in blocking buffer (10 g of dry skim milk in 200 ml of phosphate-buffered saline [PBS]) at room temperature for 2.5 h. The primary antibody (MAb 1.1ASML) directed against the CD44 variant region (18) was applied in 10 ml of blocking buffer and incubated overnight at 4~ The nitrocellulose filter was then washed in PBS and developed using the Vectastain immunoperoxidase kit (Vector Labs, Burlingame, CA) as recommended by the manufacturer. Detachment Studies Detachment of cells from the normal cell (CREF) monolayer was studied using a parallel-plate flow chamber (Biomedical Engineering Laboratory, Rice University, Houston, TX) as previously described (5). In the parallel-plate flow chamber, a known and consistent shear stress is applied to the attached cells. A schematic of the experimental system is shown in Fig. 1. The bottom plate of the flow chamber consisted of the Permanox slide seeded with CREF cells. The top plate consisted of a polycarbonate plate with slits for inflow and outflow of cell suspension. Pressure ports located in the top plate were connected to a variable reluctance differential pressure transducer (Validyne Engineering, Northridge, CA). The top and bottom plates were separated by a gasket of silastic medical grade sheeting (Dow Coming, Midland, MI) and the entire assembly was held together by a vacuum. To perform the detachment studies, the flow chamber was placed on the stage of an inverted phase contrast microscope (Model IM35, Zeiss Instruments, Batavia, IL) equipped with a SIT video camera (Hamamatsu C2400, Photonic Microscopy, Oak Brook, IL). For each cell line studied, individual cells were allowed to settle onto the CREF monolayer for 30 rain. The monolayer with adherent cells was then perfused for 3 rain with culture medium at flow rates ( ml/min) which produced a shear stress in the range of dyn/cm 2. Flow rates through the chamber were controlled by a syringe pump (Model 911, Harvard Instruments,

7 Detachment of Transformed Cells FLOW CHAMBER ASSEM MONITOR C I APPARATUS CHART RECORDER ~ PRESSURE A \ TRANSDUCER VACUUM PUMP MICROSCOPE SYRINGE PUMP Fig. 1. Schematic of experimental system used for detachment studies. The parallel-plate flow chamber is placed on the stage of an inverted phase contrast microscope equipped with a SIT video camera. Boston, MA). Experiments were viewed using a video monitor (Model PVM-122, Sony, Teaneck, NJ) and recorded using a video cassette recorder (Model AG-6500, Panasonic, Secaucus, NJ) so that the number of adherent cells which detached from the monolayer could be quantified. Data Analysis The wall shear stress (J was calculated using the following relation:

8 8 Santos et al. r = APd / 2L where AP is the pressure drop across the monolayer, d is the height of the flow in the chamber, and L is the length between pressure ports (2 cm). The height of flow in the chamber was calculated using: d 3 = 12/~QL / bap where Q is the flow rate of the perfusate, # is the viscosity of the perfusate, and b is the width of flow in the chamber (1.45 cm). For the given shear stress range, the percent of detached cells was determined by counting the cells remaining on the monolayer at 180 s after the start of the flow. This time was chosen based on previous studies which showed that the percent of cells detached reached a maximum at 180 s after the start of flow. The percent of cells detached from the monolayer for all the cell types was compared with the percent detachment of CREF from the monolayer using one-way analysis of variance. RESULTS Cell Transformation and Tumorigenic Properties A major indicator of cellular transformation in rodent fibroblasts is the acquisition of anchorage-independent growth. Hence, each cell line in the series was assayed for its capacity to grow in a semi-solid medium (soft agar). The results of the growth test for the normal and transformed cell lines are shown in Table 1. Compared to the transformed and highly malignant CREF T24 cells, the other cells showed either no growth or reduced growth in soft agar. Approximately 60% of the CREF T24 cells formed colonies in soft agar while CREF and HK B1 showed no growth at all. The HK B1-T and HK B1-M cell lines have been reported to show increased growth in soft agar in comparison with HK B1 cells (8). Tumorigenicity was measured in nude mice by implanting 5 x 104 cells in an alginate bead subcutaneously into the flank. Metastatic capability of the cells was determined by assaying for metastatic nodules following tumor formation in animals or by an experimental tail vein injection. The experimental metastasis assay measures the formation of tumor nodules after injection of 5 x 104 cells into the tail vein of nude mice. The experimental metastasis Celt Biophysics Volume 26, 1995

9 Detachment of Transformed Cells Table 1 Soft Agar Growth Test, Tumorigenicity, and Metastatic Potential Cell line Tumorigenicity and metastasis b (Average latency period, days) Transfected Soft agar Experimental Spontaneous Average latency gene growth a metastasis metastasis period, days - - CREF None No CREF T24 ras T24 Yes HK B1 ras T24 No 136 Krev-la HK B1-T ras T24 Reduced Krev-la HK B1-M ras T24 Reduced Krev-la a Soft agar growth was determined by counting colonies of > 0.1 mm after 14 days in 0.4% Noble agar as previously described (35). Yes = growth at high efficiency; No =no macroscopic colonies; Reduced = lower than CREF T24 cells. btumorigenicity was determined by implanting 5 x 104 cells in an alginate bead subcutaneously in nude mice. At visible signs of metastasis (shallow breathing, lethargy, and weight loss), animals were sacrificed and all major organs were examined for overt metastasis. Experimental metastasis was determined by injecting 5 x 104 cells into the tail vein of nude mice. Latency period is defined as the time for visible signs of tumor formation and metastasis to appear. assay only assesses part of the metastatic process, i.e., the ability of a cell to extravasate and form tumor nodules. The spontaneous metastasis assay scores for the formation of tumors at secondary sites in animals following cell implantation. In order for a cell to form metastatic foci after subcutaneous implantation, the cell must successfully complete the entire metastatic process. In both metastasis assays, each animal is subjected to a complete necropsy with screening for overt metastases using a dissecting microscope. Briefly, CREF cells produced no tumors and showed no signs of metastasis in nude mice. CREF T24 cells formed tumors following subcutaneous implantation in nude mice within 10 d and animals died within 30 d. The lungs were filled with numerous overt metastatic foci. In the experimental tail vein assay, the lungs were filled with numerous lung foci and animals died within 15 d. HK B1 cells produced tumors in nude mice with an average latency period of 136 d and eventually formed metastatic foci within 180 d.

10 i0 Santos et al. A B C D E kd Fig. 2. Western blot analysis of CD44 variant in CREF cell lines. The variant CD44 specific MAb 1.1ASML was used to stain the protein. Position of the molecular weight in kilodaltons is indicated. A, CREF; B, CREF T24; C, HK B1; D, HK B1-T; E, HK B1-M. These cells did not form metastases in the experimental assay. Tumor-derived and lung nodule-derived HK B1 cells also formed tumors but with a shorter latency periods of d, and the animals died of metastasis around 90 d. The lung nodule-derived cells tested positive for experimental metastasis formation although it took d. CD44 Characterization Results from the Western blot analysis for CREF, CREF T24, HK B1, HK B1-T, and HK B1-M cell lines are shown in Fig. 2. The variants had apparent molecular weights of 110 and 140 kda as detected with MAb 1.1ASML, which is specific for CD44-6v isoforms. The CREF cells showed a low level of expression of the CD44 variants. However, CREF T24 cells displayed a significant increase in expression. This is consistent with the finding that transient expression of ras induces increased expression of CD44 variants (36). Transfection of CREF T24 with the Krev/RAPIA suppressor gene resulted in the production of a cell line with a level of expression similar to that observed in CREF. Both the tumorderived and lung nodule-derived HK B1 cells showed levels of expression slightly higher than CREF but lower than CREF T24. Following transfection of CREF with the CD44 variants and selection for hygromycin resistance, three randomly chosen clones were obtained with differing levels of expression. Results from the Western blot analysis for these three clones are shown in Fig. 3. Blots of CREF and CREF T24 are also shown for comparison. Clone I showed a high level of expression but not as high as CREF T24, whereas Clones 2 and 3 showed levels of expression similar to CREF. CelI Biophysics Volume26, 1995

11 Detachment of Transformed Cells 11 A B C D E kd II j_,40 i~ ~ 110 Fig. 3. Western blot analysis of CREF cells transfected with the DNA expressing the CD44 variant. The CD44 specific MAb 1.1ASML was used to stain the protein. Position of the molecular weight in kilodaltons is indicated. A, CREF; B, CREF T24; C, Clone 1; D, Clone 2; E, Clone 3. 'lo c- O [] CREF [] CREF T24 [] HK BI [] HK BI-T [] HK B1- M dyn/cm 3 Shear Stress Fig. 4. Percent of transformed cells detached from a CREF monolayer after 180 s of flow at shear stresses of dyn/cmt The percent of cell detached for each cell line was compared with the percent of CREF cells that detached using a one-way analysis of variance (*p < 0.02; +p < 0.054). N = 10. Detachment Studies The percent of adherent cells (mean + SEM) which detached from the CREF monolayer after 180 s of flow when subjected to a shear stress range of dyn/cm 2 is shown in Fig. 4 for CREF, CREF T24, HK B1, HK B1-T, and HK B1-M cell lines. The percent detachment of CREF cells from the monolayer was %. When treating these cells with the T24 ras oncogene (CREF T24), percent detachment significantly increased to %. Analysis of HK B1 cells revealed that the percent detachment under identical experimental conditions was similar to CREF ( %). The percent detachments of HK B1-T and HK B1-M cells were significantly higher than CREF but not as high as the percent detachment of CREF T24 ( and , respectively).

12 12 Santos et al. * 9 CREF "0 0 5O 4O 3O 2ol ~J Clone 1 Clone 2 Clone 3 CREF-Drug Res dyn/cm 2 Shear Stress Fig. 5. Percent of cells detached from a CREF monolayer after 180 s of flow at shear stresses ranging from dyn/cml The percent of cells detached for each cell line was compared with the percent of CREF cells that detached using a one-way analysis of variance (*p < 0.02). N = 10. i The percent of adherent cells (mean + SEM) which detached from the CREF monolayer after 180 s of flow is shown in Fig. 5 for the CREF cell lines transfected with the DNA expressing the CD44 variant. The results for CREF (control cell lines in transfection) and CREF T24 are also shown for comparison. The percent detachment of Clone 1 ( ) was significantly higher than the detachment of CREF but not as high as the percent detachment of CREF T24. Clones 2 and 3 both showed percent detachments that were not significantly different from CREF ( and , respectively). The percent detachment of the CREF/Drug Resistant cell line ( ) was not significantly different from CREF. DISCUSSION Adhesion and detachment of tumor cells are important phenomena associated with the ability of the cells to metastasize. The importance of these phenomena have been discussed in detail by Weiss and Ward (37). The majority of the previous studies have focused on adhesion of tumor cells to endothelial cells prior to extravasation and detachment of tumor cells from each other during invasion. It is hypothesized that detachment of tumor cells from normal cells reflects an important cellular property associated with Cell Biophysics Volume26, 1995

13 Detachment of Transformed Cells i3 the metastatic cascade (5). In a previous study, we have shown that when transfecting normal CREF with a ras gene to form a highly tumorigenic and metastatic cell line (CREF T24), the cells become less adherent to normal cells. Furthermore, expression of a ras revertant gene (Krev/RaplA) that reverses the transformed phenotype of ras-transfected cells (the cells exhibit an extended latency period of 136 d, but eventually exhibit metastasis), results in detachment levels to that of the normal cells (5). Birch et al. (9) noted that the process of metastatic dissemination of solid tumors is comparable to that of lymphocyte migration (38,39) and consequently, a cell surface molecule such as CD44, which plays a major role in determining the trafficking of normal lymphocytes (40,41), may play a part in regulating the dissemination of cells from solid tumors. Indeed, several studies have reinforced this assumption by showing that malignant cells express high levels of CD44 (9,18-28). In addition, the level of expression of CD44 has been correlated with invasive potential of human bladder carcinoma cells (42). The role, however, of the standard CD44 (tool wt kda) vs variants of CD44 (mol wt kda) in metastasis formation have not been clear in the literature. Birch and coworkers (9) reported the formation of more and larger lung nodules when intravenously injecting human melanoma cells expressing high levels of CD44 (tool wt 90 kda) as opposed to injecting cells with low CD44 expression. Metastatic behavior was also reported following transfection of a CD44 variant (mol wt kda) in a nonmetastatic rat carcinoma cell line (18). Hofmann and coworkers (19) reported that homologous sequences of this CD44 variant are expressed in human tumor cell lines. Two forms of CD44, CD44H and CD44E, have been recently isolated by Sy et al. (20). CD44H was kda in size and had a high affinity for hyaluronate. CD44E was 150 kda in size and had a low affinity for hyaluronate. Hyaluronate is a glycosaminoglycan molecule and participates in the formation of the extracellular matrix. It is believed to play a major role in wound healing and inflammation. Increased production of this molecule has been associated with tumor growth. It was found in their investigation that transfecting human lymphoma cells with cdna clones encoding CD44H enhanced local tumor formation and metastatic potential. Transfecting with CD44E did not affect potential. They also reported that tissue distribution of the metastasis was not affected by CD44H, suggesting that this glycoprotein does not

14 14 Santos et al. distinguish specific endothelial cell receptors in different organs. Hence, they suggested that CD44H may enhance metastasis by facilitating interactions between tumor cells and host tissues. Thomas et al. (43) recently showed that CD44H expression in human melanoma lines increases motility of these cells on hyaluronate coated surfaces, whereas CD44E did not. They suggested that CD44 may play a role in in vivo aggressiveness of tumors through hyaluronan-rich stroma (44). A recent study by Faasen and coworkers (21) showed that a cell surface chondroitin sulfate proteglycan (CSPG), which is immunologically related to CD44, increases cell migration and invasion behavior of a metastatic mouse melanoma cell line. In this investigation, it was shown that two CD44 variants containing exon 6v were expressed in the highly tumorigenic and metastatic fibroblast cells (CREF T24). These variants are 110 and 140 kda in size. These results agree with the recent investigation of Hofmann and coworkers (36) which showed increased expression of a CD44 variant induced in cells transiently transfected with ras. Following transfection with the Krev/RAPIA suppressor gene, the cells expressed a low level of CD44 similar to the CREF cells. Our results suggest that expression of the CD44 variants in CREF cells can be correlated with the detachment of cells from a CREF monolayer, and that this property may play a role in enhancing metastatic potential. CREF T24, the cell line which had the highest expression of the CD44 variants, had the highest percentage detachment from the CREF monolayer as compared to the normal cell line, CREF, or any of the other cell lines. HK B1 showed both a low expression level of the CD44 variants and a percent detachment similar to CREF. HK B-T and HK B1-M cells showed levels of expression of the CD44 variants that were slightly higher than CREF and the percent detachments of these cells from the CREF monolayer were higher than CREF, but not as high as the CREF T24 cells. The three cell lines which resulted from transfecting CREF with the CD44 variants also showed a positive correlation between the expression of the CD44 variants and detachment from normal tissue. These three clones were randomly chosen following the transfection. Clone 1, which had a higher expression of the CD44 variants than CREF, showed a significantly higher percent detachment than CREF. Both Clones 2 and 3, which showed an expression of variants similar to that of CREF, showed detachment similar to CREF. The fact that Clones 2 and 3 resulted from the same transfection procedure as Clone 1, but did not have a relatively

15 Detachment of Transformed Cells 15 high expression of the CD44 variants, provides stronger support that the modification in detachment of Clone 1 was a result of the relatively high expression of the variant and not a result of some other modification of the cell line during the transfection procedure. Additionally, a clone of CREF containing only the drug resistant exhibited detachment properties similar to the parental CREF line; thus supporting the notion that the hygromycin drug resistant used in the transfection process did not influence the cell detachment observed in Clone 1. The mechanisms by which the CD44 variant alters cell adhesion in the CREF cell line is currently unknown. It is possible that normal CD44 plays a role in adhesion of these cells to the normal monolayer, and modification of the amino acid sequence during formation of the variant disrupts the normal CD44 function and hence, weakens adhesion. St. John et al. (45) showed that normal CD44 (mol wt 88 kda) promotes aggregation of fibroblast cells. It is also possible that the CD44 variant competes with other cell-surface receptors for binding sites and forms a weaker bond of adhesion. Seiter et al. (31) suggested that an MAb to a CD44 variant molecule interferes with the invasion of the metastatic cells into the lungs of a rat pancreatic adenocarcinoma by blocking a ligand interaction. Although these postulated mechanisms assume that the bond of adhesion is altered by the variant, it should be noted that detachment, not adhesion, was measured in this investigation. It has been well documented by Weiss and coworkers (37,46-50) that cell adhesion and detachment are different processes. They showed that when cells are detached from glass, they do not detach clearly from the surface. Instead, rupturing of the cells occur leaving cellular or extracellular material behind (46,47). These studies suggest that the observed changes in cell detachment may not be simply due to weakening of bonds at the cell surface. Investigations of the point of cohesive failure in the cell-cell detachment studies are clearly warranted to pinpoint the exact mechanisms of CD44 variantinduced altered cell detachment. In conclusion, we have shown that high levels of expression of CD44 variants in CREF studied correlated with greater metastatic potential and tumorigenicity. In addition, results show that in the five cell lines studied, expression of the variants increased the ability of the cells to detach from a monolayer of normal cells and hence, may play a role in the invasion stage of the metastatic process. Current investigations are concentrating on mechanisms of CD44 variantinduced changes in cell detachment.

16 16 Santos et al. ACKNOWLEDGMENTS This work was supported by a National Institutes of Health Shannon Award, The National Science Foundation Research Experience for Undergraduates Program, and The Department of Education. REFERENCES 1. Coman, D. R. (1944) Decreased mutual adhesiveness, a property of cells from squamos cell carcinomas. Cancer Res. 4, Volk, T., Geiger, B., and Raz, A. (1984) Motility and adhesive properties of high- and low-metastatic murine neoplastic cells. Cancer Res. 44, Nicolson, G. L. (1988) Organ specificity of tumor metastasis: role of preferential adhesion, invasion and growth of malignant cells at specific secondary sites. Cancer Met. Rev. 7, Nicolson, G. L. (1989) Metastatic tumor cell interactions with endothelium, basement membrane and tissue. Curr. Opinion Cell Biol. 1, Cezeaux, J. L., Austin, V., Hosseinipour, M. C., Ward (Anderson), K. A., and Zimmer, S. (1991) The effcts of shear stress and metastatic phenotype on the detachment of transformed cells. Biorheology 28, Fisher, P. B., Babiss, L. F., Weinstein, I. B., and Ginsberg, H. S. (1982) Analysis of type 5 adenovirus transformation with a cloned rat embryo cell line (CREF). Proc. Natl. Acad. Sci. USA 79, Boylan, J. F., Jackson, J., Steiner, M. R., Shih, T. Y., Duigou, G. J., Roszman, T., Fisher, P. B., and Zimmer, S. G. (1990) Role of the Haras (Ras H) oncogene in mediating progression of the tumor cell phenotype. Anticancer Res. 10, Su, Z.-Z., Austin, V. N., Zimmer, S. G., and Fisher, P. B. (1993) Defining the critical gene expression changes associated with the expression and suppression of the tumorigenic and metastatic phenotype in Ha-ras-transformed cloned rat embryo fibroblast cells. Oncogene 8, 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, Bosworth, B. T., St. John, T., Gallatin, W. M., and Harp, J. A. (1991) Sequence of the bovine CD44 cdna: comparison with human and mouse sequences. Mol. Immunol. 28, Jalkanen, S., Jalkanen, M., Bargatze, R., Tammi, M., and Butcher, E. C. (1988) Biochemical properties of glycoproteins involved in lym-

17 Detachment of ~ansforrned Ceils 17 phocyte recognition of high endothelial venules in man. J. Immunol. 141, Carter, W. G. and Wayner, E. A. (1988) Characterization of the class III collagen receptor, a phosphorylated, transmembrane glycoprotein expressed in nucleated human cells. J. Biol. Chem. 263, Jalkanen, S. and Jalkanen, M. (1992) Lymphocyte CD44 binds the COOH-terminal heparin-binding domain of fibronectin. J. Cell. Biol. 116, Culty, M., Miyake, K., Kincade, P. W., Silorski, E., Butcher, E. C., and Underhill, C. (1990) The hyaluronate receptor is a member of the CD44 (H-CAM) family of cell surface glycoproteins. J. Cell Biol. 111, Lesley, J., Hyman, R., and Kincade, P. W. (1993) CD44 and its interaction with the extracellular matrix. Adv. Immunol. 54, 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, 12,160-12, 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, 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, Hofmann, M., Rudy, W., Zoller, M., Tolg, C., Ponta, H., Herrlich, P., and Gunthert, U. (1991) CD44 splice variants confer metastatic behavior in rats: homologous sequences are expressed in human tumor cell lines. Cancer Res. 51, Sy, B. M. S., Guo, Y., and Stamenkovic, I. (1991) Distinct effects of two CD44 isoforms on tumor growth in vivo. J. Exp. Med. 174, Faassen, A. E., Schrager, J. A., Klein, D. J., Oegema, T. R., Couchman, J. R., and McCarthy, J. B. (1992) A cell surface chondroitin sulfate proteoglycan, immunologically related to CD44, is involved in type I collagen-mediated melanoma cell motility and invasion. J. Cell Biol. 116, Kennel, S. J. (1992) Effects of target antigen competition on distribution of monoclonal antibody to solid tumors. Cancer Res. 52, Kuppner, M. C., Van Meir, E., Gauthier, Th., Hamou, M.-F., and Tribolet, N. (1992) Differential expression of the CD44 molecule in human brain tumors. Int. J. Cancer 50, Pandolfi, F., Trentin, L., Boyle, L. A., Stamenkovic, I., Byers, H. R., Colvin, R. B., and Kurnick, J. T. (1992) Expression of cell adhesion

18 18 Santos et al. molecules in human melanoma cell lines and their role in cytotoxicity mediated by tumor-infiltrating lymphocytes. Cancer 69, Joensuu, H., Ristamaki, R., Klemi, P. J., and Jalkanen, S. (1993) Lymphocyte homing receptor (CD44) expression is associated with poor prognosis in gastrointestinal lymphoma. Br. J. Cancer 68, De Rossi, G., Zarcone, D., Mauro, F., Ceruti, G., Tenca, C., Pucceti, A., Mandelli, F., and Grossi, C. E. (1993) Adhesion molecule expression on B-cell chronic lymphocyte leukemia cells: malignant cell phenotypes define distinct disease subsets. Blood 81, Heider, K.-H., Hofmann, M., Hors, E., van den Berg, F., Ponta, H., Herrlich, P., and Pals, S. T. (1993) A human homologue of the rat metastasis-associated variant of CD44 is expressed in colorectal carcinomas and adenomatous polyps. J. Cell Biol. 120, Koopman, G., Heider, K.-H., Horst, E., Adolf, G. R., van der Berg, F., Ponta, H., Herrlich, P., and Pals, S. T. (1993) Activated human lymphocytes and aggressive non-hodgkin's lymphomas express a homologue of the rat metastasis-associated variant of CD44. J. Exp. Med. 177, Mackay, C. R., Terpe, A., Stauder, R., Marston, W. L., Stark, H., and Gunthert, U. (1994) Expression and modulation of CD44 variant isoforms in humans. J. Cell Biol. 124, Rudy, W., Hofmann, M., Schware-Albiez, R., Zoller, M., Heider, K.-H., Ponta, H., and Herrlich, P. (1993) The two major CD44 proteins expressed on a metastatic rat tumor cell line are derived from different splice variants: each one individually suffices to confer metastatic behavior. Cancer Res. 53, 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, Su, Z.-Z., Duigou, G. J., and Fisher, P. B. (1991) Low-level beta 1 protein kinase C expression in cloned rat embryo fibroblast cells enhances transformation induced by the adenovirus type 5 EIA gene. Mol. Carcinog 4, Kitayama, H., Sugimoto, Y., Matsuzaki, T., Ikawa, Y., and Noda, M. A. (1989) Ras-related gene with transformation suppressor activity. Cell 56, Laemmli, U. K. (1970) Cleavage of structural proteins during assembly of the lead of bacteriophage T4. Nature (Lond.) 227, Babiss, L. E., Liaw, W. S., Zimmer, S. G., Godman, G. C., Ginsberg, H. S., and Fisher, P. B. (1986) Mutations in the Ela gene of adenovirus type 5 alter the tumorigenic properties of transformed cloned rat embryo fibroblast cells. Proc. Natl. Acad. Sci USA 83,

19 Detachment of Transformed Ceils Hofmann, M., Wolfgang, R., Gunthert, U., Zimmer, S. G., Zoller, M., Zawadzki, V., Lichtner, R., Herrlich, P., and Ponta, H. (1992) A link between ras and metastatic behavior of tumor cells: ras induces CD44 promoter activity and leads to low-level expression of metastasis-specific variants of CD44 in CREF cells. Cancer Res. 53, Weiss, L. and Ward, P. M. (1983) Cell detachment and metastasis. Cancer Metastasis Rev. 2, Sher, B. T., Bargatze, R., Holzmann, B., Gallatin, W. M., Mathews, D., Wu, N., Picker, L., Butcher, E. C., and Weissman, I. L. (1988) Homing receptors and metastasis. Adv. Cancer Res. 51, Gowans, J. L. and Knight, E. J. (1964) The route of recirculation of lymphocytes in the rat. Proc. R. Soc. London Biol. Sci. 159, Jalkanen, S., Richert, R. A., Gallatin, W. M., Bargatze, R. F., Weissman, I. L., and Bucher, E. C. (1986) Homing receptors and the control of lymphocyte migration. Immunol. Rev. 91, Gallatin, M., St. John, T., Siegelman, M., Reicher, R., Butcher, E. C., and Weissman, I. L. (1987) Lymphocyte homing receptors. Cell 44, Knudson, W., Biswas, C., Li, X. Q., Nemec, R. E., and Toole, B. P. (1989), The role and regulation of tumor-associated hyaluronan, in D. Evered and J. Whelan, eds., Ciba Foundation Symposium: The Biology of Hyaluronan, Vol. 142, Wiley, New York, pp Thomas, L., Byers, H. R., Vink, J., and Stamenkovic, I. (1992) CD44H regulates tumor cell migration on hyaluronate-coated substrate. J. Cell Biol. 118, 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. Inv. Derm. 100, St. John, T., Meyer, J., Idzerda, R., and Gallatin, W. M. (1990) Expression of CD44 confers a new adhesive phenotype on transfected cells. Cell 60, Weiss, L. (1961) Studies on cellular adhesion in tissue culture. IV. The alterations of substrata by cell surfaces. Exp. Cell Res. 25, Weiss, L. and Coombs, R. R. A. (1963) The demonstration of rupture of cell surfaces by an immunological technique. Exp. Cell Res. 30, Weiss, L. (1963) Studies on cellular adhesion in tissue-culture. V. Some effects of enzymes on cell-detachment. Exp. Cell Res. 30, Weiss, L. and Lachmann, P. J. (1964) The origin of an antigenic zone surrounding HeLa cells cultured on glass. Exp. Cell Res. 36, Weiss, L. (1964) Studies on cellular adhesion in tissue culture. VII. Surface activity and cell detachment. Exp. Cell Res. 33,