- Email: [email protected]
Effect of Hyaluronan on Xenotransplanted Breast Cancer
Experimental and Molecular Pathology 72, 179 –185 (2002) doi:10.1006/exmp.2002.2437, available online at http://www.idealibrary.com on
Effect of Hyaluronan on Xenotransplanted Breast Cancer
Andrea Herrera-Gayol* and Serge Jothy† ,1 *Department of Pathology, McGill University, Montreal, Quebec, Canada; †Department of Laboratory Medicine and Department of Pathobiology, St. Michael’s Hospital and University of Toronto, Toronto, Ontario M5B 1W8, Canada Received February 1, 2002
Hyaluronan is a major glycosaminoglycan component of the extracellular matrix and CD44 is its principal ligand. In previous in vitro studies we have shown that CD44 and hyaluronan are involved in the invasive properties of the human breast cancer cell line Hs578T. The aim of this study was to test whether experimental therapy with hyaluronan interferes with tumor invasion and has an inhibitory effect on tumor growth in vivo. The Hs578T cell line was xenotransplanted orthotopically into the mammary fat pad of nu/nu mice. After tumor growth reached a maximum size of 5⫻5 mm, 50 g of hyaluronan was injected intratumorally. The tumors of control nu/nu mice were injected with PBS. Four of 12 tumors from the hyaluronan-treated group regressed completely. This effect could be due to a saturation of the hyaluronan-binding sites on tumor cells or to an acceleration of tumor rejection by a non-T-cell-dependent mechanism. This study gives a rationale for future work on the antineoplastic effects of hyaluronan. © 2002 Elsevier Science (USA)
the tumor (Bertrand et al., 1992). High tissular levels of HA are also present in pathological conditions such as inflammation, wound healing, and graft rejection (reviewed in Laurent and Fraser, 1992; Gerdin and Hallgren, 1997). HA binds to different cellular receptors such as CD44 and RHAMM (reviewed in Entwistle et al., 1996). CD44 encompasses a family of membrane glycoproteins that differ in properties such as size, glycosylation, and function (reviewed in Rudzki and Jothy, 1997). Tumor cells receive both positive and negative proliferation and motility signals from other cells as well as ECM components such as HA present in the tumor microenvironment. Breast cancer cells, from established cell lines or from human tumor samples, can be characterized on the basis of selective expression of CD44 variant isoforms (Herrera-Gayol et al., 1999a). CD44 has been identified as a key determinant in the progression to a metastatic phenotype in an experimental rat pancreatic cancer model (Gunthert et al., 1991). HA alters melanoma and breast cancer cell motility and increases their CD44 expression (Yoshinari et al., 1999; Herrera-Gayol et al., 2001). In addition, CD44 and HA play a role in angiogenesis induced by nonepithelial tumor cells (West and Kumar, 1991; Trochon et al., 1996). We have previously reported that CD44 is involved in the invasive properties of the human breast cancer Hs578T cell line in vitro (HerreraGayol et al., 1999b). Soluble HA and immobilized HA also affect the adhesive and in vitro motility capacities of this cell line (Herrera-Gayol et al., 2001). Both forms of the polymer as well as hyaluronidase treatment could increase the expression of membrane CD44. Whether these effects can facilitate or inhibit tumor growth in vivo needs to be determined. The effect of HA on the growth of nonepithelial malignant tumors in vivo has also been investigated. In a melanoma xenotransplant model, small HA oligomers had an inhibitory effect on the growth of the transplanted tumors
INTRODUCTION Breast cancer is one of the major causes of death from cancer in developed and developing areas of the world (Pisani et al., 1999). As with other malignant tumors, lethality is due to a basic feature of cancer cells, their ability to invade and metastasize. Invasion is a complex series of events where tumor cell adhesion to the extracellular matrix (ECM), degradation of the ECM, and cell migration throughout altered stroma facilitating invasion are the principal events. Lack of immunological rejection of tumor cells and tumor-initiated angiogenesis are additional mechanisms allowing the growth of the neoplastic foci. Hyaluronan (HA) is a glycosaminoglycan present in the ECM and is found at much higher concentrations at the tumor invasion front of breast cancer than in central areas of 1 To whom correspondence and reprint requests should be addressed at Department of Laboratory Medicine and Pathobiology, St. Michael’s Hospital, 30 Bond Street, Toronto, ON, M5B 1W8, Canada.
0014-4800/02 $35.00 © 2002 Elsevier Science (USA) All rights reserved.
179
180
HERRERA-GAYOL AND JOTHY
MATERIALS AND METHODS
dimension, at the timed end-point of the study (62– 64 days from the date of intratumoral injection), or in the presence of any sign of distress (skin ulcers, weight loss). Samples from tumors, lymph nodes, and internal organs were processed for histological examination and immunohistochemistry. Histological evaluation of muscle invasion was performed as described previously (Herrera-Gayol et al., 1995).
Cell Culture and Reagents
Evaluation of Treatment Response
The Hs578T human breast cancer cell line was obtained from the American Type Culture Collection (ATCC, Rockville, MD). Cells were cultured as monolayers in regular culture medium (RCM) composed of DMEM (Gibco, Burlington, Ontario, Canada) supplemented with 5% fetal bovine serum (FBS, Gibco), 2 mM glutamine, and gentamycin (50 g/ml, Gibco). A highly purified HA preparation (Suplasyn, Bioniche Life Science Inc., Belleville, Ontario, Canada) supplied as a sterile solution in PBS was used for intratumoral injections. Sterile PBS was used as the vehicle control.
Following intratumoral injections of HA or PBS, a positive response to treatment was considered to be present when: (1) the tumor did not grow further and then regressed completely or (2) the tumor initially continued to grow but then regressed to an undetectable mass or smaller than the initial volume or had become 60% smaller than the maximum volume reached.
(Zeng et al., 1998). Based on these data and our previous in vitro observations, we have investigated whether HA affects the in vivo growth and/or invasiveness of tumors derived from the Hs578T human breast cancer cell line.
Animals Female athymic CD1 nude mice (nu/nu), 6 to 8 weeks old, were obtained from Charles River Laboratories (St. Constant, Quebec, Canada). Animals were housed under specific pathogen-free conditions according to the Canadian Council on Animal Care guidelines. Orthotopic Tumor Cell Injection, Tumor Development, Injection of HA, and Autopsies Tumor cells in culture were trypsinized and washed in PBS by centrifugation, and viable cells were counted by trypan blue exclusion. For each mouse, 1 ⫻ 10 6 viable cells were resuspended in 50 l of sterile PBS. Cells were injected orthotopically into one of the lower abdominal mammary fat pads to induce tumor formation (one tumor per mouse). When tumors reached 5 ⫻ 5 mm (length ⫻ breadth), each of 12 tumors in 12 mice was injected once with 50 l of 1 mg/ml of Suplasyn (treatment group) and tumors in 12 other mice were injected with 50 l of sterile PBS (control group). Animals were monitored twice weekly, when the weight of the animals and their tumor dimensions were recorded. Macroscopic tumor volume was calculated with the formula: 21 (tumor width 2 ⫻ tumor length) (Paine-Murrieta et al., 1997). Animals were sacrificed when tumors reached a maximum of 10 mm in either
Immunohistochemistry Slides were processed for immunohistochemistry with monoclonal antibodies using the MOM (mouse on mouse) kit from Vector Laboratories (Burlington, Ontario, Canada) according to the manufacturer’s instructions. Primary antibodies were as follows: anti-CD44s clone F10.44.2 (Novocastra, Newcastle upon Tyne, UK); anti-CD44 v6, clone 2F10 (R & D Systems, Minneapolis, MN); anti-vimentin (Novocastra), anti-muscle-specific antigen (Dako, Carpinteria, CA). All primary antibodies were incubated on tissue sections overnight at 4°C. Biotinylated secondary antibody incubations were followed by an avidin– biotin–peroxidase complex final step. Reactions were developed using the NovaRed chromogen (Vector Laboratories) and the slides were counterstained with Gill’s hematoxylin. Cell Cycle and Apoptosis Analysis by Flow Cytometry Cell cycle analysis and apoptosis were studied by flow cytometry of cell suspensions prepared from paraffin-embedded material. Four to five 50-m sections were obtained from each tumor sample, or from an area where the tumor initially grew and then regressed, or from normal mammary fat pad areas where tumor cells were not injected. Sections were deparaffinized in xylene followed by decreasing concentrations of ethanol. The sections were finely minced, kept in water for 5 days, followed by incubation with 5 mg/ml of pepsin (Sigma, St. Louis, MO) in 0.9% NaCl, pH 1.5, for 30 min at 37°C, and then washed twice in cold PBS by centrifugation. Pellets were incubated overnight at 4°C
181
HYALURONAN AND TUMOR REGRESSION
TABLE I Mean Initial and Final Volumes in HA-Treated and Control Groups
Control group HA-treated group
Initial volume in mm 3 (mean ⫾ SD)
Final volume in mm 3 (mean ⫾ SD)
55 ⫾ 14
526 ⫾ 184
57 ⫾ 12
379 ⫾ 345
TABLE II Effect of Treatment on Tumor Growth
Final volume control vs final volume treated P value
No. of mice in the group Control group HA-treated group
12 12
No. of mice with tumor regression (%) P value 0 (0%) 4 (33%)
0.047
0.09
RESULTS in a propidium iodide/citrate solution containing RNase and NP-40. Cell suspensions were passed through a 53-mesh nylon filter. Flow cytometry was carried out using a Coulter EPICS XL analyzer (Beckman Coulter, Mississauga, Ontario, Canada). Morphological Characteristics of Tumor Regression: Evaluation of Inflammation Score, Fibrosis Score, and Mitotic Activity Index Histological parameters relating to immunological regression of tumors were determined using hematoxylin and eosin (H&E) stained sections. Tumors were stained with Alcian blue, Alcian blue followed by hyaluronidase treatment, or Masson trichrome to evaluate the composition of the extracellular environment. Inflammation was graded as 0 to 2 according to the density of inflammatory cells at the tumor injection site. Fibrosis was graded as 0 to 3 according to the amount of extracellular matrix at the injection site. The mitotic activity index was established by counting the number of mitotic figures in 10 high-power fields. Mitotic activity index was classified as grade 1 (ⱕ10 mitoses), grade 2 (11–20), and grade 3 (ⱖ20).
Tumor Development HA treatment resulted in tumor regression in 4 of 12 mice (33%) compared to none (0%) from the control group (Fisher’s exact test, P ⬍ 0.05; Tables I and II). The growth pattern of the four tumors that had regressed is depicted in Fig. 1. As shown in Fig. 1, the tumor in mouse 2 increased in volume and then regressed to a small palpable lesion consisting of fibrosis. The absence of tumor cells in the area where tumors regressed after HA treatment in these mice was confirmed by histological examination and by cell cycle analysis. The mean initial volume of the four tumors that regressed was 48 ⫾ 11 mm 3, the mean maximum volume achieved was 104 ⫾ 69 mm 3, and the mean final volume was 4.5 ⫾ 9 mm 3 (initial vs final volume, t test, P ⫽ 0.015). Although HA injection improved intermediate time-point survival, the difference observed in overall survival between the HA-treated mice and the control mice did not reach statistical significance (Fig. 2).
Evaluation of Histological Changes in Lymph Nodes Inguinal and para-aortic lymph nodes were fixed and processed for regular histology (H&E sections). Evaluation of the histopathological changes in inguinal and para-aortic lymph nodes was based on description protocols developed for lymph nodes in nude mice with and without xenotransplanted tumor cell lines (Sainte-Marie and Peng, 1983). Statistics Comparisons between HA-treated animals and the control group were performed using t tests, Fisher’s exact test, Log rank test, or 2 test depending on the assays.
FIG. 1.
Growth pattern of four regressing HA-treated tumors.
182
HERRERA-GAYOL AND JOTHY
percent of mice from the HA-treated group had normal lymph nodes and the other 50% had morphological changes compatible with a reaction to xenotransplanted tumors. No lymph node metastasis was observed in either group. Macroscopic and microscopic studies of various organs showed no distant metastasis in either group. Cell Cycle and Apoptosis
FIG. 2.
Overall survival of HA- and PBS-injected mice.
Histologic Evaluation The histology of the tumor which regressed to a small palpable lesion is illustrated in Fig. 3, as compared to one of the control tumors. The area of tumor regression was composed of fibrosis with small foci of inflammatory cell infiltration. Collagen deposition was confirmed with the Masson trichrome stain. Samples of the three other mammary fat pad regions where tumors developed and totally regressed after HA treatment showed small areas of ECM deposition, small patches of inflammatory infiltration, or a normal mammary fat pad histology. Morphological changes were compared to the 8 nonregressing tumors of the HA-treated mice and the 12 tumors of the control mice. Two types of malignant cells were present: polygonal cells and “fibroblast-like” cells. All tumor cells were vimentin positive. A high proportion of tumor cells expressed CD44s and were seen in different areas of the tumor. Tumor cells expressed CD44v6 weakly. CD44v6-positive cells were seen mostly at the periphery of tumors, in invasive areas, and in tumor cell aggregates. No difference in the expression of these markers was observed between the nonregressing tumors from the HA-treated group and those from the control group. Malignant cells were observed invading the skeletal muscle, the skin, and the mammary ducts in both groups. Aberrant mitotic figures were seen in both groups and no difference in mitotic score was observed. A mononuclear inflammatory infiltrate was present in 9/12 (75%) tumors of the control group and in all tumors from the HA-treated group (8/8). Globally, the tumors that had not regressed following HA injection had histological features comparable to those of the untreated tumor group. Histologic examination of lymph nodes showed that 58% of control mice had normal lymph nodes and 42% presented signs compatible with a reaction to xenotransplanted tumors: expansion of germinal centers and histiocytosis. Fifty
Using flow cytometry, samples from tissues of the treated group where tumors developed and then regressed showed only a diploid cell cycle population, confirming the absence of tumor cells. An aneuploid population was observed in tumor samples from control and nonregressing HA-treated groups. No differences in the S-phase of the aneuploid population or in the proportion of apoptotic cells were observed between the last two groups.
DISCUSSION AND CONCLUSIONS The human breast cancer cell line Hs578T used in this study has an aggressive phenotype (Thompson et al., 1992; Bae et al., 1993) and a high level of CD44 expression (Herrera-Gayol et al., 1999b; Culty et al., 1994). CD44 is the main ligand for HA, and this system provides a means of evaluating the effect of exogenous HA on tumor growth and metastasis. The central observation in the present study is that 33% of the HA-treated tumors showed a positive response to treatment and regressed completely. This regression might result from several different mechanisms. One possible mechanism is a saturation of HA-binding sites, such as CD44, on the membrane of tumor cells by the large amount of HA injected within the tumors. This saturation could mask membrane-binding sites required for the attachment of tumor cells to the extracellular matrix. Consequently the lack of tumor cell attachment would lead to anoikis, a form of apoptosis, triggered by disruption of cell– extracellular matrix adhesion (reviewed in Shanmugathasan and Jothy, 2001). Alternatively, the high-molecular-weight HA injected intratumorally might have been cleaved into HA oligomers, which in a melanoma study were shown to inhibit tumor growth in vivo (Zeng et al., 1998). Another possible mechanism of tumor regression induced by HA might involve shedding of CD44 which commonly accompanies CD44 overexpression. We have shown that HA increases CD44 expression in Hs578T cells in vitro
HYALURONAN AND TUMOR REGRESSION
183
FIG. 3. Histology of regressive changes in an orthotopic HA-treated breast tumor compared to a PBS-treated control tumor. (A) Histology of a regressed HA-treated tumor site in the mammary fat pad. The short arrow shows a mouse mammary duct. The previous site of the tumor is replaced by a fibrotic area (F). The long arrow shows the limit of the fibrotic area. A thin inflammatory cell infiltrate (arrowhead) is present at the periphery. H & E stain; 135⫻. (B) Histology of a tumor in a control mouse. A large tumor (T) is present. The long arrows show the tumor margin, growing into the mammary fat pad. The short arrow shows a mouse mammary duct. There is a central area of necrosis (N) and small patches of inflammatory infiltrate (arrowhead). H&E stain; 135⫻.
(Herrera-Gayol et al., 2001). Shed CD44 disrupts the interaction between cell membrane CD44 and the extracellular matrix, thus altering signals necessary for proliferation (Sy
et al., 1992; Bartolazzi et al., 1995) or creating an environment that favors apoptosis (Yu et al., 1996). Overall, high expression levels of CD44 could result in its increased
184 shedding locally within the tumor and consequently favor anoikis. An additional mechanism whereby HA may be implicated in tumor regression is by affecting host resistance. Despite an absence of functional T-lymphocytes, nude mice are able to generate an immune response against xenotransplanted tumors through antibody production, macrophage infiltration, and NK activity (Kim et al., 1982). In vitro studies have shown that HA increases membrane CD44 expression in lymphocytes while the in vivo injection of HA triggers B cell proliferation (Rafi et al., 1997). It is also possible that the injection of HA into the tumors stimulated a CD44-mediated NK response. CD44 plays a stimulatory role in a number of NK cell functions, such as TNF-␣ secretion and expression of CD69 through a protein tyrosine kinase pathway (Galandrini et al., 1996). Whatever the mechanisms of tumor involution are in the present study, the four tumors that regressed showed histological signs similar to those reported in the spontaneous regression of human seminomas and melanomas (McGovern, 1975; Holmes et al., 1986). In conclusion, we have shown in this study that a significant number of Hs578T-derived breast tumors in nude mice regressed completely following intratumoral injection of HA. Further studies are required to understand the mechanisms of regression induced by HA in this and other tumor models.
ACKNOWLEDGMENTS Our thanks go to Mrs. S. Schiller for her technical assistance with the flow cytometric analysis; Drs. G. Prud’homme and N. C. Phillips for critically reviewing the manuscript; Dr. F. Halwani, Mrs. E. Fragiskatos, Mrs. C. Loiselle, and H. Khoury for their technical advice; Mrs. H. St-Croix for performing the orthotopic injections; Mr. F. Rouah for his technical assistance with the statistical analysis; Mrs. D. Avery for editing the manuscript; and Bioniche Life Sciences Inc. for providing Suplasyn. This work was supported by an operating grant (10315) from the Canadian Institutes of Health Research. A.H-G. was supported by a Fellowship from the Royal Victoria Hospital Research Institute, Montreal, Quebec, Canada.
REFERENCES
HERRERA-GAYOL AND JOTHY
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. Bertrand, P., Girard, N., Delpech, B., Duval, C., d’Anjou, J., and Dauce, J. P. (1992). Hyaluronan (hyaluronic acid) and hyaluronectin in the extracellular matrix of human breast carcinomas: Comparison between invasive and non-invasive areas. Int. J. Cancer 52, 1– 6. Culty, M., Shizari, M., Thompson, E. W., and Underhill, C. B. (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. Entwistle, J., Hall, C. L., and Turley, E. A. (1996). HA receptors: Regulators of signalling to the cytoskeleton. J. Cell Biochem. 61, 569 –577. Galandrini, R., Piccoli, M., Frati, L., and Santoni, A. (1996). Tyrosine kinase-dependent activation of human NK cell functions upon triggering through CD44 receptor. Eur. J. Immunol. 26, 2807–2811. Gerdin, B., and Hallgren, R. (1997). Dynamic role of hyaluronan (HYA) in connective tissue activation and inflammation. J. Intern. Med. 242, 49 –55. 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. Herrera-Gayol, A., and Jothy, S. (1999a). Adhesion proteins in the biology of breast cancer: Contribution of CD44. Exp. Mol. Pathol. 66, 149 –156. Herrera-Gayol, A., and Jothy, S. (1999b). CD44 modulates Hs578T human breast cancer cell adhesion, migration, and invasiveness. Exp. Mol. Pathol. 66, 99 –108. Herrera-Gayol, A., Royal, A., and Babaı¨, F. (1995). Correlation between cell differentiation stage, types of invasion, and hematogenous metastasis in experimental rhabdomyosarcomas. Exp. Mol. Pathol. 63, 1–15. Herrera-Gayol, A., Royal, A., and Babaı¨, F. (2001). Effect of hyaluronan on the invasive properties of human breast cancer cells in vitro. Int. J. Exp. Pathol. 82, 193–200. Holmes, A. S., Klimberg, I. W., Stonesifer, K. J., Kramer, B. S., and Wajsman, Z. (1986). Spontaneous regression of testicular seminoma: Case report. J. Urol. 135, 795–796. Kim, U., Han, T., Ghosh, S., Freedman, V. H., and Shin, S-I. (1982). Immunologic mechanisms of selective graft resistance to certain malignant tumors and prevention of metastases by athymic nude mice. In “Proceedings of the Third International Workshop on Nude Mice,” pp. 365–377, Gustav Fischer, New York. Laurent, T. C., and Fraser, J. R. (1992). Hyaluronan. FASEB J. 6, 2397– 2404. McGovern, V. J. (1975). Spontaneous regression of melanoma. Pathology 7, 91–99. Paine-Murrieta, G. D., Taylor, C. W., Curtis, R. A., Lopez, M. H., Dorr, R. T., Johnson, C. S., Funk, C. Y., Thompson, F., and Hersh, E. M. (1997). Human tumor models in the severe combined immune deficient (SCID) mouse. Cancer Chemother. Pharmacol. 40, 209 –214. Pisani, P., Parkin, D. M., Bray, F., and Ferlay, J. (1999). Estimates of the worldwide mortality from 25 cancers in 1990. Int. J. Cancer 83, 18 –29.
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.
Rafi, A., Nagarkatti, M., and Nagarkatti, P. S. (1997). Hyaluronate–CD44 interactions can induce murine B-cell activation. Blood 89, 2901–2908.
Bartolazzi, A., Jackson, D., Bennett, K., Aruffo, A., Dickinson, R., Shields,
Sainte-Marie, G., and Peng, F. S. (1983). Structural and cell population
Rudzki, Z., and Jothy, S. (1997). CD44 and the adhesion of neoplastic cells. Mol. Pathol. 50, 57–71.
HYALURONAN AND TUMOR REGRESSION
changes in the lymph nodes of the athymic nude mouse. Lab. Invest. 49, 420 – 429. Shanmugathasan, M., and Jothy, S. (2001). Apoptosis, anoikis and their relevance to the pathobiology of colon cancer. Pathol. Int. 50, 273–279. Sy, M. S., Guo, Y. J., and Stamenkovic, I. (1992). Inhibition of tumor growth in vivo with a soluble CD44 –immunoglobulin fusion protein. J. Exp. Med. 176, 623– 627. Thompson, E. W., Paik, S., Brunner, N., Sommers, C. L., Zugmaier, G., Clarke, R., Shima, T. B., Torri, J., Donahue, S., and Lippman, M. E. (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. Trochon, V., Mabilat, C., Bertrand, P., Legrand, Y., Smadja-Joffe, F., Soria, C., Delpech, B., and Lu, H. (1996). Evidence of involvement of
185 CD44 in endothelial cell proliferation, migration and angiogenesis in vitro. Int. J. Cancer 66, 664 – 668. West, D. C., and Kumar, S. (1991). Tumour-associated hyaluronan: A potential regulator of tumour angiogenesis. Int. J. Rad. Biol. 60, 55– 60. Yoshinari, C., Mizusawa, N., Byers, H. R., and Akasaka, T. (1999). CD44 variant isoform CD44v10 expression of human melanoma cell lines is upregulated by hyaluronate and correlates with migration. Melanoma Res. 9, 223–231. Yu, Q., Toole, B. P., and Stamenkovic, I. (1997). Induction of apoptosis of metastatic mammary carcinoma cells in vivo by disruption of tumor cell surface CD44 function. J. Exp. Med. 186, 1985–1996. Zeng, C., Toole, B. P., Kinney, S. D., Kuo, J. W., and Stamenkovic, I. (1998). Inhibition of tumor growth in vivo by hyaluronan oligomers. Int. J. Cancer 77, 396 – 401.
Effect of Hyaluronan on Xenotransplanted Breast Cancer
Andrea Herrera-Gayol* and Serge Jothy† ,1 *Department of Pathology, McGill University, Montreal, Quebec, Canada; †Department of Laboratory Medicine and Department of Pathobiology, St. Michael’s Hospital and University of Toronto, Toronto, Ontario M5B 1W8, Canada Received February 1, 2002
Hyaluronan is a major glycosaminoglycan component of the extracellular matrix and CD44 is its principal ligand. In previous in vitro studies we have shown that CD44 and hyaluronan are involved in the invasive properties of the human breast cancer cell line Hs578T. The aim of this study was to test whether experimental therapy with hyaluronan interferes with tumor invasion and has an inhibitory effect on tumor growth in vivo. The Hs578T cell line was xenotransplanted orthotopically into the mammary fat pad of nu/nu mice. After tumor growth reached a maximum size of 5⫻5 mm, 50 g of hyaluronan was injected intratumorally. The tumors of control nu/nu mice were injected with PBS. Four of 12 tumors from the hyaluronan-treated group regressed completely. This effect could be due to a saturation of the hyaluronan-binding sites on tumor cells or to an acceleration of tumor rejection by a non-T-cell-dependent mechanism. This study gives a rationale for future work on the antineoplastic effects of hyaluronan. © 2002 Elsevier Science (USA)
the tumor (Bertrand et al., 1992). High tissular levels of HA are also present in pathological conditions such as inflammation, wound healing, and graft rejection (reviewed in Laurent and Fraser, 1992; Gerdin and Hallgren, 1997). HA binds to different cellular receptors such as CD44 and RHAMM (reviewed in Entwistle et al., 1996). CD44 encompasses a family of membrane glycoproteins that differ in properties such as size, glycosylation, and function (reviewed in Rudzki and Jothy, 1997). Tumor cells receive both positive and negative proliferation and motility signals from other cells as well as ECM components such as HA present in the tumor microenvironment. Breast cancer cells, from established cell lines or from human tumor samples, can be characterized on the basis of selective expression of CD44 variant isoforms (Herrera-Gayol et al., 1999a). CD44 has been identified as a key determinant in the progression to a metastatic phenotype in an experimental rat pancreatic cancer model (Gunthert et al., 1991). HA alters melanoma and breast cancer cell motility and increases their CD44 expression (Yoshinari et al., 1999; Herrera-Gayol et al., 2001). In addition, CD44 and HA play a role in angiogenesis induced by nonepithelial tumor cells (West and Kumar, 1991; Trochon et al., 1996). We have previously reported that CD44 is involved in the invasive properties of the human breast cancer Hs578T cell line in vitro (HerreraGayol et al., 1999b). Soluble HA and immobilized HA also affect the adhesive and in vitro motility capacities of this cell line (Herrera-Gayol et al., 2001). Both forms of the polymer as well as hyaluronidase treatment could increase the expression of membrane CD44. Whether these effects can facilitate or inhibit tumor growth in vivo needs to be determined. The effect of HA on the growth of nonepithelial malignant tumors in vivo has also been investigated. In a melanoma xenotransplant model, small HA oligomers had an inhibitory effect on the growth of the transplanted tumors
INTRODUCTION Breast cancer is one of the major causes of death from cancer in developed and developing areas of the world (Pisani et al., 1999). As with other malignant tumors, lethality is due to a basic feature of cancer cells, their ability to invade and metastasize. Invasion is a complex series of events where tumor cell adhesion to the extracellular matrix (ECM), degradation of the ECM, and cell migration throughout altered stroma facilitating invasion are the principal events. Lack of immunological rejection of tumor cells and tumor-initiated angiogenesis are additional mechanisms allowing the growth of the neoplastic foci. Hyaluronan (HA) is a glycosaminoglycan present in the ECM and is found at much higher concentrations at the tumor invasion front of breast cancer than in central areas of 1 To whom correspondence and reprint requests should be addressed at Department of Laboratory Medicine and Pathobiology, St. Michael’s Hospital, 30 Bond Street, Toronto, ON, M5B 1W8, Canada.
0014-4800/02 $35.00 © 2002 Elsevier Science (USA) All rights reserved.
179
180
HERRERA-GAYOL AND JOTHY
MATERIALS AND METHODS
dimension, at the timed end-point of the study (62– 64 days from the date of intratumoral injection), or in the presence of any sign of distress (skin ulcers, weight loss). Samples from tumors, lymph nodes, and internal organs were processed for histological examination and immunohistochemistry. Histological evaluation of muscle invasion was performed as described previously (Herrera-Gayol et al., 1995).
Cell Culture and Reagents
Evaluation of Treatment Response
The Hs578T human breast cancer cell line was obtained from the American Type Culture Collection (ATCC, Rockville, MD). Cells were cultured as monolayers in regular culture medium (RCM) composed of DMEM (Gibco, Burlington, Ontario, Canada) supplemented with 5% fetal bovine serum (FBS, Gibco), 2 mM glutamine, and gentamycin (50 g/ml, Gibco). A highly purified HA preparation (Suplasyn, Bioniche Life Science Inc., Belleville, Ontario, Canada) supplied as a sterile solution in PBS was used for intratumoral injections. Sterile PBS was used as the vehicle control.
Following intratumoral injections of HA or PBS, a positive response to treatment was considered to be present when: (1) the tumor did not grow further and then regressed completely or (2) the tumor initially continued to grow but then regressed to an undetectable mass or smaller than the initial volume or had become 60% smaller than the maximum volume reached.
(Zeng et al., 1998). Based on these data and our previous in vitro observations, we have investigated whether HA affects the in vivo growth and/or invasiveness of tumors derived from the Hs578T human breast cancer cell line.
Animals Female athymic CD1 nude mice (nu/nu), 6 to 8 weeks old, were obtained from Charles River Laboratories (St. Constant, Quebec, Canada). Animals were housed under specific pathogen-free conditions according to the Canadian Council on Animal Care guidelines. Orthotopic Tumor Cell Injection, Tumor Development, Injection of HA, and Autopsies Tumor cells in culture were trypsinized and washed in PBS by centrifugation, and viable cells were counted by trypan blue exclusion. For each mouse, 1 ⫻ 10 6 viable cells were resuspended in 50 l of sterile PBS. Cells were injected orthotopically into one of the lower abdominal mammary fat pads to induce tumor formation (one tumor per mouse). When tumors reached 5 ⫻ 5 mm (length ⫻ breadth), each of 12 tumors in 12 mice was injected once with 50 l of 1 mg/ml of Suplasyn (treatment group) and tumors in 12 other mice were injected with 50 l of sterile PBS (control group). Animals were monitored twice weekly, when the weight of the animals and their tumor dimensions were recorded. Macroscopic tumor volume was calculated with the formula: 21 (tumor width 2 ⫻ tumor length) (Paine-Murrieta et al., 1997). Animals were sacrificed when tumors reached a maximum of 10 mm in either
Immunohistochemistry Slides were processed for immunohistochemistry with monoclonal antibodies using the MOM (mouse on mouse) kit from Vector Laboratories (Burlington, Ontario, Canada) according to the manufacturer’s instructions. Primary antibodies were as follows: anti-CD44s clone F10.44.2 (Novocastra, Newcastle upon Tyne, UK); anti-CD44 v6, clone 2F10 (R & D Systems, Minneapolis, MN); anti-vimentin (Novocastra), anti-muscle-specific antigen (Dako, Carpinteria, CA). All primary antibodies were incubated on tissue sections overnight at 4°C. Biotinylated secondary antibody incubations were followed by an avidin– biotin–peroxidase complex final step. Reactions were developed using the NovaRed chromogen (Vector Laboratories) and the slides were counterstained with Gill’s hematoxylin. Cell Cycle and Apoptosis Analysis by Flow Cytometry Cell cycle analysis and apoptosis were studied by flow cytometry of cell suspensions prepared from paraffin-embedded material. Four to five 50-m sections were obtained from each tumor sample, or from an area where the tumor initially grew and then regressed, or from normal mammary fat pad areas where tumor cells were not injected. Sections were deparaffinized in xylene followed by decreasing concentrations of ethanol. The sections were finely minced, kept in water for 5 days, followed by incubation with 5 mg/ml of pepsin (Sigma, St. Louis, MO) in 0.9% NaCl, pH 1.5, for 30 min at 37°C, and then washed twice in cold PBS by centrifugation. Pellets were incubated overnight at 4°C
181
HYALURONAN AND TUMOR REGRESSION
TABLE I Mean Initial and Final Volumes in HA-Treated and Control Groups
Control group HA-treated group
Initial volume in mm 3 (mean ⫾ SD)
Final volume in mm 3 (mean ⫾ SD)
55 ⫾ 14
526 ⫾ 184
57 ⫾ 12
379 ⫾ 345
TABLE II Effect of Treatment on Tumor Growth
Final volume control vs final volume treated P value
No. of mice in the group Control group HA-treated group
12 12
No. of mice with tumor regression (%) P value 0 (0%) 4 (33%)
0.047
0.09
RESULTS in a propidium iodide/citrate solution containing RNase and NP-40. Cell suspensions were passed through a 53-mesh nylon filter. Flow cytometry was carried out using a Coulter EPICS XL analyzer (Beckman Coulter, Mississauga, Ontario, Canada). Morphological Characteristics of Tumor Regression: Evaluation of Inflammation Score, Fibrosis Score, and Mitotic Activity Index Histological parameters relating to immunological regression of tumors were determined using hematoxylin and eosin (H&E) stained sections. Tumors were stained with Alcian blue, Alcian blue followed by hyaluronidase treatment, or Masson trichrome to evaluate the composition of the extracellular environment. Inflammation was graded as 0 to 2 according to the density of inflammatory cells at the tumor injection site. Fibrosis was graded as 0 to 3 according to the amount of extracellular matrix at the injection site. The mitotic activity index was established by counting the number of mitotic figures in 10 high-power fields. Mitotic activity index was classified as grade 1 (ⱕ10 mitoses), grade 2 (11–20), and grade 3 (ⱖ20).
Tumor Development HA treatment resulted in tumor regression in 4 of 12 mice (33%) compared to none (0%) from the control group (Fisher’s exact test, P ⬍ 0.05; Tables I and II). The growth pattern of the four tumors that had regressed is depicted in Fig. 1. As shown in Fig. 1, the tumor in mouse 2 increased in volume and then regressed to a small palpable lesion consisting of fibrosis. The absence of tumor cells in the area where tumors regressed after HA treatment in these mice was confirmed by histological examination and by cell cycle analysis. The mean initial volume of the four tumors that regressed was 48 ⫾ 11 mm 3, the mean maximum volume achieved was 104 ⫾ 69 mm 3, and the mean final volume was 4.5 ⫾ 9 mm 3 (initial vs final volume, t test, P ⫽ 0.015). Although HA injection improved intermediate time-point survival, the difference observed in overall survival between the HA-treated mice and the control mice did not reach statistical significance (Fig. 2).
Evaluation of Histological Changes in Lymph Nodes Inguinal and para-aortic lymph nodes were fixed and processed for regular histology (H&E sections). Evaluation of the histopathological changes in inguinal and para-aortic lymph nodes was based on description protocols developed for lymph nodes in nude mice with and without xenotransplanted tumor cell lines (Sainte-Marie and Peng, 1983). Statistics Comparisons between HA-treated animals and the control group were performed using t tests, Fisher’s exact test, Log rank test, or 2 test depending on the assays.
FIG. 1.
Growth pattern of four regressing HA-treated tumors.
182
HERRERA-GAYOL AND JOTHY
percent of mice from the HA-treated group had normal lymph nodes and the other 50% had morphological changes compatible with a reaction to xenotransplanted tumors. No lymph node metastasis was observed in either group. Macroscopic and microscopic studies of various organs showed no distant metastasis in either group. Cell Cycle and Apoptosis
FIG. 2.
Overall survival of HA- and PBS-injected mice.
Histologic Evaluation The histology of the tumor which regressed to a small palpable lesion is illustrated in Fig. 3, as compared to one of the control tumors. The area of tumor regression was composed of fibrosis with small foci of inflammatory cell infiltration. Collagen deposition was confirmed with the Masson trichrome stain. Samples of the three other mammary fat pad regions where tumors developed and totally regressed after HA treatment showed small areas of ECM deposition, small patches of inflammatory infiltration, or a normal mammary fat pad histology. Morphological changes were compared to the 8 nonregressing tumors of the HA-treated mice and the 12 tumors of the control mice. Two types of malignant cells were present: polygonal cells and “fibroblast-like” cells. All tumor cells were vimentin positive. A high proportion of tumor cells expressed CD44s and were seen in different areas of the tumor. Tumor cells expressed CD44v6 weakly. CD44v6-positive cells were seen mostly at the periphery of tumors, in invasive areas, and in tumor cell aggregates. No difference in the expression of these markers was observed between the nonregressing tumors from the HA-treated group and those from the control group. Malignant cells were observed invading the skeletal muscle, the skin, and the mammary ducts in both groups. Aberrant mitotic figures were seen in both groups and no difference in mitotic score was observed. A mononuclear inflammatory infiltrate was present in 9/12 (75%) tumors of the control group and in all tumors from the HA-treated group (8/8). Globally, the tumors that had not regressed following HA injection had histological features comparable to those of the untreated tumor group. Histologic examination of lymph nodes showed that 58% of control mice had normal lymph nodes and 42% presented signs compatible with a reaction to xenotransplanted tumors: expansion of germinal centers and histiocytosis. Fifty
Using flow cytometry, samples from tissues of the treated group where tumors developed and then regressed showed only a diploid cell cycle population, confirming the absence of tumor cells. An aneuploid population was observed in tumor samples from control and nonregressing HA-treated groups. No differences in the S-phase of the aneuploid population or in the proportion of apoptotic cells were observed between the last two groups.
DISCUSSION AND CONCLUSIONS The human breast cancer cell line Hs578T used in this study has an aggressive phenotype (Thompson et al., 1992; Bae et al., 1993) and a high level of CD44 expression (Herrera-Gayol et al., 1999b; Culty et al., 1994). CD44 is the main ligand for HA, and this system provides a means of evaluating the effect of exogenous HA on tumor growth and metastasis. The central observation in the present study is that 33% of the HA-treated tumors showed a positive response to treatment and regressed completely. This regression might result from several different mechanisms. One possible mechanism is a saturation of HA-binding sites, such as CD44, on the membrane of tumor cells by the large amount of HA injected within the tumors. This saturation could mask membrane-binding sites required for the attachment of tumor cells to the extracellular matrix. Consequently the lack of tumor cell attachment would lead to anoikis, a form of apoptosis, triggered by disruption of cell– extracellular matrix adhesion (reviewed in Shanmugathasan and Jothy, 2001). Alternatively, the high-molecular-weight HA injected intratumorally might have been cleaved into HA oligomers, which in a melanoma study were shown to inhibit tumor growth in vivo (Zeng et al., 1998). Another possible mechanism of tumor regression induced by HA might involve shedding of CD44 which commonly accompanies CD44 overexpression. We have shown that HA increases CD44 expression in Hs578T cells in vitro
HYALURONAN AND TUMOR REGRESSION
183
FIG. 3. Histology of regressive changes in an orthotopic HA-treated breast tumor compared to a PBS-treated control tumor. (A) Histology of a regressed HA-treated tumor site in the mammary fat pad. The short arrow shows a mouse mammary duct. The previous site of the tumor is replaced by a fibrotic area (F). The long arrow shows the limit of the fibrotic area. A thin inflammatory cell infiltrate (arrowhead) is present at the periphery. H & E stain; 135⫻. (B) Histology of a tumor in a control mouse. A large tumor (T) is present. The long arrows show the tumor margin, growing into the mammary fat pad. The short arrow shows a mouse mammary duct. There is a central area of necrosis (N) and small patches of inflammatory infiltrate (arrowhead). H&E stain; 135⫻.
(Herrera-Gayol et al., 2001). Shed CD44 disrupts the interaction between cell membrane CD44 and the extracellular matrix, thus altering signals necessary for proliferation (Sy
et al., 1992; Bartolazzi et al., 1995) or creating an environment that favors apoptosis (Yu et al., 1996). Overall, high expression levels of CD44 could result in its increased
184 shedding locally within the tumor and consequently favor anoikis. An additional mechanism whereby HA may be implicated in tumor regression is by affecting host resistance. Despite an absence of functional T-lymphocytes, nude mice are able to generate an immune response against xenotransplanted tumors through antibody production, macrophage infiltration, and NK activity (Kim et al., 1982). In vitro studies have shown that HA increases membrane CD44 expression in lymphocytes while the in vivo injection of HA triggers B cell proliferation (Rafi et al., 1997). It is also possible that the injection of HA into the tumors stimulated a CD44-mediated NK response. CD44 plays a stimulatory role in a number of NK cell functions, such as TNF-␣ secretion and expression of CD69 through a protein tyrosine kinase pathway (Galandrini et al., 1996). Whatever the mechanisms of tumor involution are in the present study, the four tumors that regressed showed histological signs similar to those reported in the spontaneous regression of human seminomas and melanomas (McGovern, 1975; Holmes et al., 1986). In conclusion, we have shown in this study that a significant number of Hs578T-derived breast tumors in nude mice regressed completely following intratumoral injection of HA. Further studies are required to understand the mechanisms of regression induced by HA in this and other tumor models.
ACKNOWLEDGMENTS Our thanks go to Mrs. S. Schiller for her technical assistance with the flow cytometric analysis; Drs. G. Prud’homme and N. C. Phillips for critically reviewing the manuscript; Dr. F. Halwani, Mrs. E. Fragiskatos, Mrs. C. Loiselle, and H. Khoury for their technical advice; Mrs. H. St-Croix for performing the orthotopic injections; Mr. F. Rouah for his technical assistance with the statistical analysis; Mrs. D. Avery for editing the manuscript; and Bioniche Life Sciences Inc. for providing Suplasyn. This work was supported by an operating grant (10315) from the Canadian Institutes of Health Research. A.H-G. was supported by a Fellowship from the Royal Victoria Hospital Research Institute, Montreal, Quebec, Canada.
REFERENCES
HERRERA-GAYOL AND JOTHY
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. Bertrand, P., Girard, N., Delpech, B., Duval, C., d’Anjou, J., and Dauce, J. P. (1992). Hyaluronan (hyaluronic acid) and hyaluronectin in the extracellular matrix of human breast carcinomas: Comparison between invasive and non-invasive areas. Int. J. Cancer 52, 1– 6. Culty, M., Shizari, M., Thompson, E. W., and Underhill, C. B. (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. Entwistle, J., Hall, C. L., and Turley, E. A. (1996). HA receptors: Regulators of signalling to the cytoskeleton. J. Cell Biochem. 61, 569 –577. Galandrini, R., Piccoli, M., Frati, L., and Santoni, A. (1996). Tyrosine kinase-dependent activation of human NK cell functions upon triggering through CD44 receptor. Eur. J. Immunol. 26, 2807–2811. Gerdin, B., and Hallgren, R. (1997). Dynamic role of hyaluronan (HYA) in connective tissue activation and inflammation. J. Intern. Med. 242, 49 –55. 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. Herrera-Gayol, A., and Jothy, S. (1999a). Adhesion proteins in the biology of breast cancer: Contribution of CD44. Exp. Mol. Pathol. 66, 149 –156. Herrera-Gayol, A., and Jothy, S. (1999b). CD44 modulates Hs578T human breast cancer cell adhesion, migration, and invasiveness. Exp. Mol. Pathol. 66, 99 –108. Herrera-Gayol, A., Royal, A., and Babaı¨, F. (1995). Correlation between cell differentiation stage, types of invasion, and hematogenous metastasis in experimental rhabdomyosarcomas. Exp. Mol. Pathol. 63, 1–15. Herrera-Gayol, A., Royal, A., and Babaı¨, F. (2001). Effect of hyaluronan on the invasive properties of human breast cancer cells in vitro. Int. J. Exp. Pathol. 82, 193–200. Holmes, A. S., Klimberg, I. W., Stonesifer, K. J., Kramer, B. S., and Wajsman, Z. (1986). Spontaneous regression of testicular seminoma: Case report. J. Urol. 135, 795–796. Kim, U., Han, T., Ghosh, S., Freedman, V. H., and Shin, S-I. (1982). Immunologic mechanisms of selective graft resistance to certain malignant tumors and prevention of metastases by athymic nude mice. In “Proceedings of the Third International Workshop on Nude Mice,” pp. 365–377, Gustav Fischer, New York. Laurent, T. C., and Fraser, J. R. (1992). Hyaluronan. FASEB J. 6, 2397– 2404. McGovern, V. J. (1975). Spontaneous regression of melanoma. Pathology 7, 91–99. Paine-Murrieta, G. D., Taylor, C. W., Curtis, R. A., Lopez, M. H., Dorr, R. T., Johnson, C. S., Funk, C. Y., Thompson, F., and Hersh, E. M. (1997). Human tumor models in the severe combined immune deficient (SCID) mouse. Cancer Chemother. Pharmacol. 40, 209 –214. Pisani, P., Parkin, D. M., Bray, F., and Ferlay, J. (1999). Estimates of the worldwide mortality from 25 cancers in 1990. Int. J. Cancer 83, 18 –29.
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.
Rafi, A., Nagarkatti, M., and Nagarkatti, P. S. (1997). Hyaluronate–CD44 interactions can induce murine B-cell activation. Blood 89, 2901–2908.
Bartolazzi, A., Jackson, D., Bennett, K., Aruffo, A., Dickinson, R., Shields,
Sainte-Marie, G., and Peng, F. S. (1983). Structural and cell population
Rudzki, Z., and Jothy, S. (1997). CD44 and the adhesion of neoplastic cells. Mol. Pathol. 50, 57–71.
HYALURONAN AND TUMOR REGRESSION
changes in the lymph nodes of the athymic nude mouse. Lab. Invest. 49, 420 – 429. Shanmugathasan, M., and Jothy, S. (2001). Apoptosis, anoikis and their relevance to the pathobiology of colon cancer. Pathol. Int. 50, 273–279. Sy, M. S., Guo, Y. J., and Stamenkovic, I. (1992). Inhibition of tumor growth in vivo with a soluble CD44 –immunoglobulin fusion protein. J. Exp. Med. 176, 623– 627. Thompson, E. W., Paik, S., Brunner, N., Sommers, C. L., Zugmaier, G., Clarke, R., Shima, T. B., Torri, J., Donahue, S., and Lippman, M. E. (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. Trochon, V., Mabilat, C., Bertrand, P., Legrand, Y., Smadja-Joffe, F., Soria, C., Delpech, B., and Lu, H. (1996). Evidence of involvement of
185 CD44 in endothelial cell proliferation, migration and angiogenesis in vitro. Int. J. Cancer 66, 664 – 668. West, D. C., and Kumar, S. (1991). Tumour-associated hyaluronan: A potential regulator of tumour angiogenesis. Int. J. Rad. Biol. 60, 55– 60. Yoshinari, C., Mizusawa, N., Byers, H. R., and Akasaka, T. (1999). CD44 variant isoform CD44v10 expression of human melanoma cell lines is upregulated by hyaluronate and correlates with migration. Melanoma Res. 9, 223–231. Yu, Q., Toole, B. P., and Stamenkovic, I. (1997). Induction of apoptosis of metastatic mammary carcinoma cells in vivo by disruption of tumor cell surface CD44 function. J. Exp. Med. 186, 1985–1996. Zeng, C., Toole, B. P., Kinney, S. D., Kuo, J. W., and Stamenkovic, I. (1998). Inhibition of tumor growth in vivo by hyaluronan oligomers. Int. J. Cancer 77, 396 – 401.