ANTICANCER RESEARCH 24: 139-144 (2004)

‚-Galactosidase and ·- Inhibit Formation of Multicellular Nodules in Breast Cancer Cell Cultures

KATHLEEN F. ARCARO1, JOANN WANG1, CHRISTOPHER N. OTIS2 and BERT M. ZUCKERMAN3

1Department of Veterinary and Animal Sciences, University of Massachusetts-Amherst, Amherst, MA; 2Director of Surgical Pathology, Baystate Medical Center, Springfield, MA; 3Department of Microbiology, University of Massachusetts-Amherst, Amherst, ª∞, U.S.A.

Abstract. In response to an estrogen, confluent monolayers hydrocarbon receptor (AhR) agonists (3). It is proposed of MCF-7 cell cultures develop multi-cellular nodules, termed that AhR agonists inhibit foci by either 1) decreasing the foci. Post-confluent development of foci occurs with available E2, due to increased of E2 by physiologic levels of 17‚-estradiol and are inhibited by various cytochrome P450s, or 2) cross-talk between the ER and anti-estrogens acting through either the estrogen or aryl AhR (3). Many breast cancers, however, are not ER- hydrocarbon receptors. In the present paper we report that positive and therefore not treatable with ER antagonists or disruption of the terminal sugars on membrane receptors other types of hormonal intervention. results in inhibition of foci. Treatment with 0.013-0.05 units/ml To evaluate alternative mechanisms of inhibiting the of ‚-galactosidase completely inhibited the development of foci formation of foci in breast cancer cell cultures, we attempted while leaving the monolayer of cells intact. Trials with ·- to disrupt cell-cell signaling by interfering with membrane mannosidase resulted in a similar but less potent inhibition of surface receptors. The sugar moieties of cell-surface and cell- foci. -fluorescent conjugates, RCA (Ricinus communis cell are thought to be involved in recognition agglutinin), and ConA (Canavalia ensiformis agglutinin) were events during cancer metastasis and invasion. are used to identify membrane surface carbohydrates on MCF-7 of a non-immune origin that have the capacity to cells. Binding of the RCA-fluorescent conjugate was inhibited bind sugars with a high degree of specificity. Over the past by co-treatment with galactose or lactose. Binding of ConA- twenty years there has been considerable interest in using fluorescent conjugate was significantly inhibited by mannose lectins to investigate the distribution of glycoproteins in and n-acetyl-glucosamine. This is the first report of inhibition breast cancer cell lines (4, 5) and numerous studies have of foci development in MCF-7 cell cultures by disruption of demonstrated changes in the content, composition, and/or surface carbohydrates on membrane receptors. distribution of oligosaccharides of glycoproteins in malignant tissue (6-12). Treatment with lectins in vitro has resulted in The MCF-7 breast cancer cell line forms a confluent both increased cell growth (13, 14) and decreased cell growth monolayer in plastic culture dishes. In the presence of the due to toxicity (14). In the present study, lectin-fluorescent steroidal hormone, 17‚-estradiol (E2), cells in confluent probes were used to identify surface carbohydrates, and mild cultures can continue to grow and divide, forming discrete enzymatic treatment with ‚-galactosidase and ·-mannosidase multi-cellular nodules termed foci (1). The formation of foci to target these carbohydrates on MCF-7 breast cancer cells occurs at physiologic concentrations of E2 and is estrogen in an effort to disrupt foci formation, while leaving the receptor (ER)-dependent. Thus, the post-confluent monolayer of cells intact. development of foci has been proposed as a model for the study of breast cancer (1). Formation of foci is inhibited by Materials and Methods ER antagonists such as LY156 758 (2), as well as by aryl Cell culture. MCF-7 cells (a gift from J. Gierthy) were maintained in T-75 flasks (Falcon Brand; Beckton Dickinson Laboratories; Franklin Lakes, NJ, USA) in phenol red-free Dulbecco’s modified Correspondence to: Dr. Kathleen Arcaro, Environmental Sciences Eagle’s medium, (DMEM, Sigma Chemicals; St.Louis, MO, USA) Program, N235 Morrill 1, University of Massachusetts, 639 North supplemented with 5% Bovine Calf Serum (BCS) (Hyclone; Logan, Pleasant Street, Amherst, MA 01003-9298 U.S.A. Tel: (413) 577- UT, USA), 10 ng/ml insulin, 0.1mM nonessential amino acids, and 1823, Fax: (413) 545-5731, e-mail: [email protected] 2 mM L-glutamine. Cells were split when just confluent by treating the cultures with 0.5% trypsin and 0.025% EDTA in phosphate- Key Words: Breast cancer, MCF-7, ‚-galactosidase, ·-mannosidase, buffered saline (PBS) and gently triturating until a single-cell fluorescent-lectin conjugates, membrane receptors. suspension was obtained. Cells were seeded and test media were

0250-7005/2004 $2.00+.40 139 ANTICANCER RESEARCH 24: 139-144 (2004) prepared in supplemented DMEM in which the BCS was stripped Results with a charcoal-dextran slurry to remove all trace amounts of E2. Figure 1 shows results from a typical focus assay. After 9 Focus assay. MCF-7 cells were seeded in 48-well tissue culture plates days of treatment with E , a dose-dependent increase in the at a density of 1x105 cells/ml in 0.5 ml medium/well. After 24 hours, 2 number of foci is detected as an increase in relative the medium in each well was removed and cells were refed with 0.8 ml of test medium containing various concentrations of the vehicle fluorescence (bottom panel). The foci are multi-layered structures (top panel C and D), which retain more control, E2, ‚-galactosidase, ·-mannosidase, or E2 plus an . The refeed was repeated on day 5 and the cultures fixed with 10% Rhodamine B than does the surrounding monolayer. formalin on day 9. After drying, the cells were stained with 1% To determine whether treatment with , which target Rhodamine B in distilled water (0.5 ml/well) and rinsed with tap specific sugars on cell surface receptors, increase or decrease water. Rinses were repeated until most of the stain was removed the formation of foci, MCF-7 cultures were treated with from the monolayer. The stained foci are easily discriminated from increasing concentrations of ‚-galactosidase or ·-mannosidase the monolayer background and can be quantified directly using an automated colony counter, or indirectly by measuring the in both the absence and presence of 0.1 nM E2. Neither ‚- fluorescence in a plate reader. In the present study, cells were galactosidase nor ·-mannosidase increased the formation of quantified indirectly using a Fusion Universal Microplate Analyzer foci (Figure 2: filled symbols), however, both enzymes (Packard Instrument Co.; Meriden, CT, USA), set at a bottom read significantly inhibited the foci induced by 0.1 nM E2 (Figure 2: with 550/10 excitation and 590/20 emission filters. Results are open symbols). ‚-galactosidase was a more potent inhibitor of reported as relative fluorescent units (RFU). Data were normalized foci than was ·-mannosidase (IC =0.017 and 0.026 Units/ml, to maximum response obtained with E . Due to evaporation of 50s 2 respectively). ‚-galactosidase also was more efficacious than medium from the outer wells, data were collected from the inner 24 wells only. was ·-mannosidase; it inhibited 100% of the foci, whereas the highest concentration of ·-mannosidase inhibited only 60% of Preconfluent growth assay. The effect of ‚-galactosidase and ·- the foci induced by E2. To determine whether enzyme mannosidase on preconfluent growth of MCF-7 cells was treatment resulted in a permanent decrease in foci ® determined using the CellTiter 96 AQueous Non-Radioactive Cell development, a focus assay was run in which cells were treated Proliferation Assay (Promega; Madison WI, USA), in which with enzymes for 9 days and then treated with control medium metabolically active cells reduce a tetrazolium compound into a for an additional 5 days. Removal of enzyme followed by formazan product that is soluble in tissue culture medium. MCF-7 treatment with medium 0.1 nM E for an additional 5 days cells were seeded in 96-well plates at a density of 1x103 cells/100 2 Ìl/well. The day following seeding, cells were treated with ‚- resulted in a 30% increase in foci as compared to control. galactosidase and ·-mannosidase at concentrations of 0, 0.01, 0.02, During the course of the 11-day focus assay, the cultures 0.03, 0.04 and 0.05 units/ml of enzyme activity at 100 Ìl/well. On were observed to determine whether the enzyme treatment day 4 after seeding, 20 Ìl of CellTiter solution was added to each altered preconfluent growth. Visual inspections indicated that well and the absorbance of the formazan at 490nm was measured the concentrations of enzymes used in these experiments did directly in the 96-well plate. not kill or retard the growth of the cells. Cells in all wells, from the DMSO control to the highest test concentration, Labeling of surface carbohydrates. Lectins conjugated to fluorescent appeared to reach confluence at nearly the same time. probes, Ricinus communis agglutinin-TRITC (RCA-I) and Canavalia ensiformis agglutinin-TRITC (ConA), and inhibitory However, since subtle effects on preconfluent growth might carbohydrates (galactose and lactose for RCA-I; and ·-D-mannose, be missed by visual inspection, an enzymatic cell proliferation ·-D-glucose and ·-N-acetyl-D-glucosamine for ConA (EY assay was used. Treatment with either ‚-galactosidase or ·- Laboratories Inc.; San Mateo, CA, USA) were used to identify mannosidase did not alter preconfluent cell growth (Figure 3). surface carbohydrates on MCF-7 cells. MCF-7 cells were seeded in The presence of specific carbohydrate residues on the 5 96-well tissue culture plates at a density of 5x10 cells/ml. One day membrane of MCF-7 cells was investigated using lectin- after seeding, solutions of inhibitory carbohydrates were prepared fluorescent conjugates and appropriate inhibitory sugars. in 0.1M PBS and incubated at a 1:1 (v/v) solution with the appropriate lectin at 100 Ìg lectin/ml PBS. The solutions were kept Figure 4 shows significant fluorescence in MCF-7 cell at room temperature in the dark for 45 minutes. Cells were then cultures treated with either RCA-TRITC or ConA-TRITC. treated with either the lectin solution or the lectin/inhibitory The binding of the RCA was significantly inhibited by both carbohydrate mixture at 100 Ìl/well and incubated at 37ÆC for 15 galactose and lactose. The binding of ConA was significantly minutes. After incubation, cells were rinsed 3 times with PBS and inhibited by ·-D-mannose and ·-N-acetyl-D-glucosamine, the fluorescence quantified using a plate reader with 550/10 but not by ·-D-glucose. Fluorescent microscopy (not excitation and 590/20 emission filters. shown) confirmed that the lectin-fluorescent conjugates bound to membrane surface residues. Statistical analysis. Data were analyzed (ANOVA and Student’s t- tests) using Microsoft Excel 2000 and graphed using SigmaPlot After determining the presence of specific carbohydrates graphics package (SPSS Inc.; Chicago, IL, USA). All tests were on the surface of MCF-7 cells, we ran a series of focus two-tailed and a significant p value was set at > 0.05. assays with lectins in an attempt to inhibit foci development

140 Arcaro et al: Nodule Inhibition in MCF-7 Cell Cultures

Figure 1. The top panel shows a confluent monolayer of MCF-7 cells grown in the absence of 17‚-estradiol (A and B) and the development of foci in cultures treated with 0.1 nM 17‚-estradiol (C and D); phase contrast optics (A and C); side view with Nomarski optics (B and D); reprinted from Arcaro and Gierthy (28) with permission of the authors. The bottom panel shows a concentration-dependent increase of foci in MCF-7 cell cultures treated with 17‚-estradiol. A 17‚- estradiol control curve was conducted with each experiment as described in the Materials and Methods. by capping terminal sugars. In contrast to the enzyme first demonstration of a non-hormonal treatment assays, all concentrations of lectins that inhibited foci inhibiting the formation of foci in breast cancer cell development also were toxic to the cells, as they significantly cultures. Since the formation of foci in MCF-7 cell inhibited preconfluent cell growth. cultures serves as a model for breast cancer, the inhibition of these solid, multi-cellular, tumor-like structures may be Discussion relevant to therapy. ·-mannosidase also inhibited the E2- induced formation of foci, but since ‚-galactosidase was The most important observation of these studies is that the more potent and efficacious inhibitor of foci, it is a mild enzymatic treatment with ‚-galactosidase completely stronger candidate for future investigations. Results from inhibits the E2-induced formation of foci in MCF-7 breast an earlier study by Sargent and colleagues (15) showed cancer cell cultures while leaving the confluent monolayer that mild enzymatic treatment with trypsin resulted in a of cells completely intact. To our knowledge this is the slight but consistent decrease in the capability of cells

141 ANTICANCER RESEARCH 24: 139-144 (2004)

Figure 2. Effects of enzyme treatment on foci development in MCF-7 cell cultures. The focus assay was conducted with increasing concentrations of ‚- galactosidase or ·-mannosidase in the absence or presence of 0.1 nM 17‚-estradiol as described in the Materials and Methods. Means±S.D. from a single experiment are presented; n=4 for each concentration; the experiment was repeated four times with similar results.

Figure 3. Effects of enzyme treatment on preconfluent growth in MCF-7 cell cultures. Cells were seeded in 96-well plates, treated with either vehicle control, 0.1 nM E2, 0.05 units of enzyme, or E2 plus enzyme. The CellTiter assay was performed on day 3 when the cultures were roughly 80% confluent, as described in the Materials and Methods. Means±S.D. from a single experiment are shown; n=6 for each test compound; t-tests with a Bonferoni correction for multiple comparisons showed no differences between enzyme treatment and control; the experiment was repeated twice with similar results.

142 Arcaro et al: Nodule Inhibition in MCF-7 Cell Cultures

Figure 4. Labeling of carbohydrate residues on membrane receptors with lectin-fluorescent conjugates. MCF-7 cells in 96-well plates were treated with 10 mg of the lectin-fluorescent conjugate (solid bars), or with the conjugate plus 0.1 mM of the inhibitory carbohydrate (hatched bars). Means±S.D. of relative fluorescent units (RFU) are presented; n=6 for each test compound; *=significantly different from control using t-tests with Bonferoni correction for multiple comparisons; the experiment was repeated twice with similar results. from a murine mammary tumor to colonize the lung, from mammary tumors that are strong colonizers of lung where as treatment with either or tissues as compared with tumor cells that did not colonize did not diminish colonizing potential of lung tissue (15). tumor cells. Inhibition of foci with no observable changes in the Results from our experiments with lectin-fluorescent monolayer suggests that either the surface carbohydrates conjugates and inhibitory carbohydrates identified galactose differ between the monolayer and foci, or that the cells in and mannose/glucosamine residues as important targets of the foci are sensitive to enzymatic treatment whereas cells the enzyme treatments. This is in agreement with other in the monolayer are resistant. These differences between studies that have demonstrated the presence of these the monolayer and foci could be investigated using lectin- carbohydrate residues on MCF-7 cell cultures (15, 16). In histochemistry and confocal microscopy. Determining the contrast to the enzyme treatment, inhibiting foci by capping molecular differences between cells in the monolayer and the terminal sugars using specific lectins resulted in foci may lead to the understanding of the etiology of breast significant cytotoxicity of the monolayer at all concentrations cancer and suggest avenues for treatment and prevention. that inhibited foci formation. Lectin treatment has resulted Cell-to-cell interactions appear to be primary importance in various effects on cell growth, ranging from increased to metastasis of tumor cells. These interactions occur in proliferation to toxicity (14, 17, 18). Indeed, many of the many cases through integral membrane proteins and lectins routinely used to identify surface carbohydrates are glycoproteins. Mucins, such as Muc1, are secreted by breast extremely toxic, e.g. RCA-II also is known as ricin. Since adenocarcinoma cells (25). Members of the adhesion and lectins are present in many foods, are resistant to digestion, integrin families have been shown to regulate growth of cross the intestinal barrier in appreciable amounts and breast cancer in vitro. Integrins have been implicated in potentially reach many organs in the body, the role of lectins affecting cell adhesion, signaling, metastasis, apoptosis and as both a causative and therapeutic agent for cancer has cell motility (26, 27). Our next series of experiments are received considerable attention (19-24). Lectin adsorption designed to explore the connection of the integrin studies did not reveal significant differences between cells superfamily of proteins to the phenomena described here.

143 ANTICANCER RESEARCH 24: 139-144 (2004)

Acknowledgements 15 Sargent NSE, Price JE, Darling DL, Flynn MP and Tarin T: Effects of altering surface composition on This research was supported by NIEHS grant R01 ES09795, and a metastatic colonization potential of murine mammary-tumor grant from the Baystate/UMASS Collaborative Research Center, cells. Br J Cancer 55: 21-28, 1987. USA. The authors thank Drs. John F. Gierthy and Jonathan B. 16 Kolar Z and Negrini R: Lectin histochemistry in the estradiol Zuckerman for helpful comments on an earlier version of this report. stimulated and nonstimulated MCF- 7 cells. Acta Univ Palacki Olomuc Fac Med 122: 121-126, 1989. References 17 Valentine U, Fabian S, Schumacher U and Leathem AJ: The influence of dietary lectins on the cell proliferation of human breast cancer cell lines in vitro. Anticancer Res 23: 1197-1206, 2003. 1 Gierthy JF, Lincoln DWI, Roth KE, Bowser SS, Bennett JA, 18 Lorea P, Goldschmidt D, Darro F, Salmon I, Bovin N, Gabius Bradley L and Dickerman HW: Estrogen-stimulation of HG, Kiss R and Danguy A: In vitro characterization of lectin- postconfluent cell accumulation and foci formation of human induced alterations on the proliferative activity of three human MCF-7 breast cancer cells. J Cell Biochem 45: 177-187, 1991. melanoma cell lines. Melanoma Res 7: 353-363, 1997. 2 Arcaro KF, Yi LD, Seegal RD, Vakharia DD, Yang Y, Spink 19 Jordinson M, El-Hariry I, Calam D and Pgnatelli M: Vicia faba DC, Brosch K and Gierthy JF: 2,2',6,6'-tetrachlorobiphenyl is agglutinin, the lectin present in broad beans, stimulates estrogenic in vitro and in vivo. J Cell Biochem 72: 94, 1999. differentiation of undifferentiated colon cancer cells. 3 Arcaro KF, O'Keefe PW, Yang Y, Clayton W and Gierthy JF: Gastroenterol 44: 709-714, 1999. Antiestrogenicity of environmental polycyclic aromatic 20 Steck PA and Nicolson GL: Cell surface glycoproteins of hydrocarbons in human breast cancer cells. Toxicol 133: 115, 1999. 13762NF mammary adenocarcinoma clones of differing 4 Kolar Z, Negrini R and Lisato L: Effect of some sugars on the metastic potentials. Exp Cell Res 147: 255-267, 1983. growth and differentiation of MCF-7 cells: Detection of 21 Klefstrom P, Nuortio L and Taskinen E: Postoperative radiation glycosylative changes using lectin histochemistry. Acta Univ therapy and adjuvant chemoimmunotherapy in breast cancer - Palacki Olomuc Fac Med 122: 113-119, 1989. aspects of timing and immune competence. Acta Radiologica 5 Konska G, Vissac C, Zagla K, Chezet F, Vasson MP, Bernard- Oncol 25: 161-166, 1986. Gallon D and Guillot J: Ultrastructural localization of binding 22 Taeda Y, Nose M, Hiraizumi S and Ohuchi N: Expression of L- sites for PNA and VVA- B-4 lectins in human breast cancer cell PHA-binding proteins in breast cancer: reconstitution and lines detected by confocal fluorescence microscopy. Int J Oncol molecular characterization of beta 1-6 branched oligosaccharides 21: 1009-1019, 2002. in three-dimensional cell culture. Breast Cancer Res Treatment 6 Ohuchi N, Nose M, Abe R and Kyogoku M: Lectin-binding 38: 313-324, 1996. patterns of breast-carcinoma – significance on structural 23 Schumacher U and Adam E: Lectin histochemical HPA-binding atypism. Tohoku J Exp Med 143: 491-499, 1984. pattern of human breast and colon cancers is associated with 7 Ng RCY, Roberts AN, Wilson RG, Latner AL and Turner GA: metastases formation in severe combined immunodeficient Analyses of extracts of human-breast cancers – changes mice. Histochem J 29: 677-684, 1997. in glycoprotein content linked to the malignant phenotype. Br J 24 Pae HO, Seo WG, Oh GS, Shin MK, Lee HS, Lee HS, Kim SB Cancer 55: 249-254, 1987. and Chung HT: Potentiation of tumor necrosis factor-alpha- 8 Dansey R, Murray J, Ninin D and Bezwoda WR: Lectin binding induced apoptosis by mistletoe lectin, Immunopharmacol in human breast cancer – clinical and pathologic correlations Immunotoxicol 22: 697-709, 2000. with fluorescein-conjugated peanut, wheat germ and 25 Devine P, Clark B, Birrell G, Layton G, Ward B, Alewood P concanavalin-a binding. Oncol 45: 300-302, 1988. and Mckenzie I: The breast tumor-associated defined 9 Skutelsky E, Hoenig S, Griffel B and Alroy J: The distribution by monoclonal antibody 3EI2 is an O-linked mucin of lectin receptor sites in human breast lesions. Pathol Res carbohydrate containing N-glyconeuraminic acid. Cancer Res Practice 183: 469-475, 1988. 51: 5826-5836, 1991. 10 Karuna V: Lectin binding patterns in benign and malignant 26 Kirshner J, Chen C, Loi P, Huang J and Shikeley J: CEACGMI- lesions of the breast. Indian J Path Micro 35: 289-297, 1992. 45, a cell-cell adhesion molecule mediates apoptosis and reverts 11 Ulloa F and Real FX: Differential distribution of in mammary carcinoma cells to normal morphogenic phenotype in alpha2,3 and alpha2,6 linkages in the apical membrane of a 3D culture. Proc Nat Acad Sci USA 100: 521-526, 2003. cultured epithelial cells and tissues. J Histochem Cytochem 49: 27 Mercurio AM, Bachelder RE, Rabinovitz I, O’Conner KL, Tani 501-509, 2001. T and Shaw LM: The metastatic odyssey, the integrin connection. 12 Dwek MV, Ross HA, Streets AJ, Brooks SA, Adam E, Titcomb Cancer Metastasis, Biol Clin Aspects 10: 313-328, 2001. A, Woodside JW, Schumacher U and Leathem AJ: Helix 28 Arcaro KF and Gierthy JF: Assessing modulation of estrogenic pomatia agglutinin lectin-binding oligosaccharides of aggressive activity of environmental and pharmaceutical compounds using breast cancer. Int J Cancer 95: 79-85, 2001. MCF-7 focus assay. In: Methods in Molecular Biology. Vol. 176: 13 Fik E, Wolun-Cholewa M, Kistowska M, Warchol JB and Steroid Receptor Methods: Protocols and Assays, ed. Lieberman Gozdzicka-Jozefiak A: Effect of lectin from Chelidonium majus BA, pp. 341-351. Totowa, N.J: Humana Press Inc. 2001 L. on normal and cancer cells in culture. Folia Histochemica Cytobiologica 39: 215, 2001. 14 Camby I, Janssen T, De Decker R, Petein M, Raviv G, Pasteels J-L, Kiss R and Danguy A: Lectin-induced alterations on the proliferation of three human prostatic cancer cell lines. In Vitro Received September 18, 2003 Cell Dev Biol 32: 633-639, 1996. Accepted October 16, 2003

144