134 578 _THS_ A .' A ,'f ' \f‘ /‘ P . 0, IllllllHllHl||||IlllllllillllllllllllllilllHINlIHIlIllHllI 3 1293 02080 6075 LIBRARY MiChigan State University PLACE IN REFURN BOX to remove this checkout'from your record. TO AVOID FINE return on or before date due. MAY BE RECALLED with earlier due date if requested. DATE DUE DATE DUE DATE DUE 11/00 mumpss-pu Chemotherapeutic and chem0preventive roles of sphingolipids in human breast cancer By Hong Yang A THESIS Submitted to Michigan State University In partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Food Science and Human Nutrition 1999 Dr. Joseph J. Schroeder ABSTRACT CHEMOTHERAPEUTIC AND CHEMOPREVENTIVE ROLES OF SPHINGOLIPIDS IN HUMAN BREAST CANCER By Hong Yang Complex sphingolipids are significant constituents of food (Ahn and Schroeder, 1998) and are digested and absorbed as the bioactive metabolites, sphingosine and cerarnide (Schmelz er al., 1994). Sphingosine and the cell-permeable, ceramide-analog Cz-ceramide have been shown to inhibit proliferation and cause death of estrogen receptor-negative, MDA-MB-231 hmnan breast cancer cells (Zhang and Schroeder, 1998) and induce differentiation of HL—60 leukemia cells (Okazaki et al., 1990). The objectives of this study were to assess the chemotherapeutic potential of sphingosine and Cz—ceramide by comparing their efl‘ects on proliferation and death of breast cancer cells to conventional normal human breast epithelial cells (HBEC) which have a basal cell phenotype (Type II HBEC) and to assess the chemopreventive potential of these sphingolipids by examining their effects on proliferation, difi'erentiation and apoptosis of HBEC which lmve stem cell characteristics and are susceptible to carcinogenesis (Type I HBEC). The results show that both sphingosine and Cz-ceramide inhibit proliferation and cause death of tumorigenic breast cells (in vitro neoplastically transformed Type I HBEC and MCF-7 and MDA-MB-231 breast cancer cells). The inhibition of tumorigenic cell grth by sphingosine was accompanied by dephosphorylation of retinoblastoma protein and inhibition of telomerase activity. Death of transformed Type I HBEC and MDA—MB-231 cells caused by sphingosine and Cz-ceramide had characteristics of apoptosis (i.e. morphological changes and formation of a DNA ladder on agarose gel sir so 5? inc TCC electrophoresis). The sensitivities of the tumorigenic breast cell lines to Cz-ceramide are similar to that of the Type II HBEC. However, the tumorigenic breast cells are more sensitive to inhibition of proliferation by sphingosine than Type II HBEC. This suggests that sphingosine is a potential chemotherapeutic agent for breast cancer; whereas, Cz-ceramide might be toxic for normal human breast tissue at concentrations that are anti-proliferative for cancer cells. The sensitivities of Type I HBEC to growth inhibition by sphingosine and C2- ceramide were similar to that of tumorigenic breast cells. Cz-ceramide appeared to cause death of Type I HBEC by inducing apoptosis. At non-cytotoxic concentrations (1-3 uM), sphingosine induced differentiation of Type I HBEC; whereas, Cz-cerarnide did not affect differentiation. Therefore, sphingosine might also function as a chemopreventive agent by inducing the differentiation of Type I HBEC with stem cell characteristics and, thereby, reducing the targets for neoplastic transformation. Tr en C W} ACKNOWLEDGMENTS My deepest thanks go to my major advisor, Dr. Joseph J. Schroeder, who guided me through all stages of my graduate program with his profound insight about science, and about life. His belief in me challenges me to excel. I have been most fortunate to work in the laboratory of Dr. C. C. Chang and Dr. J. Trosko. My special thanks go to Dr. Chia-Cheng Chang, whose door is always open for an encouraging and problem-solving discussion. His exceptional intelligence, warmth and calming nature has great influence on me. My sincere gratitude goes to Drs. Maurice Bennink, James Trosko and Louis King, who ofl‘ered me great support and excellent suggestions. Enormous thanks go to Drs. Ching-Yi Hsieh, Wei Sun, Melinda Wilson, Brad Upham, Hye-Kyung Na, Gang Chen, Mei-Hui Tai, and Margo Holland for their thoughtfirl, constructive advise, generous support and genuine fi'iendship. Thanks for my fellow graduate students Nestor DeoCampo, Chi Zhang, Eun-Hyun Ahn, Min Sun Kim, Jason “fresinger, Angela Cruz and Maki Saitoh for their consistent support, valued comments and friendship. Finally, my deepest gratitude goes to my parents, my sister and brother-in-law, and all my fiiends for their endless love and support. They believe in me before I believe in myself. I will never have enough words to express how important they are to me. TABLE OF CONTENTS Page ACKNOWLEDGMENTS .................................................................. iv LIST OF TABLES ............................................................................... vii LIST OF FIGURES ........................................................................... viii I. INTRODUCTION ........................................................................... 1 II. LITERATURE REVIEW ................................................................ 3 A Breast cancer .................................................................................. 3 1. Breast cancer epidemiology ..................................................... 3 2. Breast cancer etiology .............................................................. 3 3. Breast cancer and diet ............................................................ 3 4. Role of stem cell differentiation in breast cancer prevention... 5 5. Breast cancer chemotherapy and apoptosis ............................ 5 6. Telomerase activity, a biomarker of cell cycle progression and difierentiation ....................................................................... 6 B. Sphingosine and ceramide regulate cell behavior ........................... 7 1. Sphingolipid metabolism ........................................................ 7 2. Cerarnide inhibits cell proliferation, causes differentiation, and induces apoptosis ................................................................... 8 3. Sphingosine inhibits cell proliferation, causes differentiation, and induce apoptosis ............................................................ 10 4. Cerarnide and sphingosine have the potential to prevent and treat breast cancer ................................................................ 11 C. Normal human breast epithelial and cancer cells in culture as experimental models .................................................................. 12 III. MATERIALS AND METHODS ................................................ 16 IV. RESULTS ..................................................................................... 21 A Sphingosine and ceramide inhibit cell proliferation ........................ 21 B. Sphingosine, but not ceramide, induces differentiation of Type I HBEC ........................................................................................... 33 C. Sphingosine and ceramide induce apoptosis ................................. 37 D. Sphingosine decreases telomerase activity in transformed Type I HBEC ........................................................................................... 44 E. Sphingosine causes dephosphorylation of retinoblastoma protein in MCF-7 breast cancer cells but not Type II HBEC ....................... V. DISCUSSION .............................................................................. VI. SUMMARY ................................................................................ VII. REFERENCES ........................................................................... vi 49 51 55 56 LIST OF TABLES Table l. Sphingosine and ceramide afl‘ect cell-cycle distribution of Type I HBEC ................................................................................... Table 2. Sphingosine and ceramide afi‘ect cell-cycle distribution of transformed Type I HBEC ................................................................ Table 3. Sphingosine and ceramide affect cell-cycle distribution of MCF-7 breast cancer cell lines .......................................................... Table 4. Sphingosine causes differentiation of Type I HBEC ............. Table 5. Cerarnide does not cause differentiation of Type I HBEC. . .. Table 6. Effects of sphingosine and ceramide on apoptosis of transformed Type I HBEC ............................................................... Page 27 28 29 34 36 47 LIST OF FIGURES Figure 1. Structures of sphingosine, ceramide and sphingomyelin ........ Figure 2. Sphingomyelin turnover pathway ........................................ Figure 3. Morphology of two types of normal human breast epithelial cells (HBEC) (photographs) ............................................................... Figure 4. Type I HBEC derived from reduction mammoplasty have the ability to differentiate to Type II HBEC and are susceptible to neoplastic transformation .................................................................... Figure 5. Sphingosine inhibits proliferation and causes death of Type Ibut not Type II HBEC ..................................................................... Figure 6. Sphingosine inhibits proliferation and causes death of tumorigenic breast cells ...................................................................... Figure 7. Cerarnide inhibits proliferation and causes death of Type I and Type II HBEC ............................................................................. Figure 8. Ceramide inhibits proliferation and causes death of tumorigenic breast cells .................................................................... Figure 9. Sphingosine stereoisomers inhibit proliferation and cause death of tumorigenic breast cells ....................................................... Figure 10. Flow cytometric analysis of transformed Type I HBEC adtured with sphingosine ...................................................................... Figure 11. Flow cytometric analysis of transformed Type I HBEC cultured with ceramide ...................................................................... Figure 12. Sphingosine induces differentiation of Type I HBEC (photographs) ..................................................................................... Figure 13. Sphingosine and ceramide cause morphological changes in Type I HBEC indicative of apoptosis (photographs) ...................... Figure 14. Sphingosine and ceramide cause morphological changes in transformed Type I HBEC indicative of apoptosis (photographs). Figure 15. Sphingosine and ceramide cause morphological changes in MDA-MB-231 cells indicative of apoptosis (photographs) ........... Page 14 15 22 23 25 26 30 31 32 35 38 39 40 Figure 16. Cerarnide causes intemucleosomal DNA fragmentation in Type I HBEC ................................................................................... Figure 17. Sphingosine and ceramide cause intemucleosomal DNA fragmentation in transformed Type I HBEC ......................................... Figure 18. Sphingosine and ceramide cause time-dependent increase in intemucleosomal DNA fragmentation in transformed Type I I-IBEC.. Figure 19. Sphingosine causes intemucleosomal DNA fragmentation in MDA-MB-23l breast cancer cells ................................................... Figure 20. Cerarnide causes intemucleosomal DNA fragmentation in MDA—MB-231 breast cancer cells ....................................................... Figure 21. Sphingosine decreases telomerase activity in transformed Type I HBEC ..................................................................................... Figure 22. Sphingosine alters the expression of retinoblastoma protein (Rb) in MCF-7 breast cancer cells but not Type II HBEC ................. 41 42 43 45 46 48 50 I. INTRODUCTION Breast cancer is the most common cancer and the second leading cause of cancer- related deaths in women in the United States (Cancer Facts and Figures, 1998). The disease arises as a result of the accumulation of mutations of critical genes that regulate cell proliferation, difl‘erentiation and apoptosis in breast cells (Russo et al., 1990). The majority of breast cancers are adenocarcinomas originating from epithelial cells (Osteen et al., 1986). Terminal end buds, which contain highly proliferating mammary epithelial stem cells, are considered to be the target of mammary neoplastic transformation (Russo and Russo, 1987). Chemotherapeutic agents are being investigated with the goal of treating this disease without side effects or the development of drug resistance. Recent research also has focused on identifying specific dietary components that may prevent the development of breast cancer (Love, 1994). Complex sphingolipids are significant constituents of food (Vesper, et al., 1999, Ahn and Schroeder, 1998) and are digested and absorbed as the bioactive metabolites ceramide and sphingosine (Schemelz et al., 1994). Ceramide and sphingosine regulate cell behavior including cell proliferation, differentiation and apoptosis (Merrill et al., 1995). Cerarnide may mediate the effects of ionizing radiation and the breast cancer chemotherapeutic agents vincristine and doxorubicinn which have been shown to cause cellular accumulation of ceramide (Hannun, 1997). Since ceramide can be deacylated by ceramidase to form sphingosine, some cellular effects originally attributed to ceramide might be mediated by sphingosine (Ohta et al., 1994). Because of their abilities to inhibit proliferation and induce apoptosis in breast cancer cell lines (Gill et al., 1997; Cai et al., 1997; Zhang and Schroeder, 1998), ceramide and sphingosine may be useful agents to treat breast cancer. However, no study has examined the efiects of ceramide and sphingosine on the proliferation of normal breast epithelial cells or evaluated the chemopreventive potential of these sphingolipids. Recently, two types of morphologically distinguishable normal human breast epithelial cells (HBEC) were derived from reduction marmnoplasty (Kao et al., 1995). Type I HBEC express estrogen receptors (Kang et al., 1997) and have stem cell characteristics (1'. e. deficiency in gap junctional intercellular communication, ability to difl‘erentiate to Type II HBEC and to form budding/ductal structures on Matrigel) (Kao et al., 1995; Sun et al., 1999). Significantly, Type I HBEC are susceptible to neoplastic transformation and SV40 large T-antigen transformed Type I HBEC were capable of anchorage independent growth and were more susceptible to telomerase activation and immortalization (Kao et al., 1995; Sun et al., 1999). Therefore, Type I HBEC appear to be the target cells for neoplastic transformation. In contrast, Type II HBEC express basal epithelial cell markers (i.e. rat-6 integrin and cytokeratin-l4) and rarely become immortal afier SV40 transformation. The specific aims of the proposed studies were to use the unique HBEC culture system as well as MCF-7 and MDA-MB—231 breast cancer cells to determine if ceramide and sphingosine have: 1) Chemopreventive activities-Are ceramide and sphingosine capable of inhibiting the proliferation and/or inducing the difi'erentiation of Type I HBEC, thereby reducing the targets for breast carcinogenesis?; and 2) Chemotheraputic activities-Do sphingosine and ceramide inhibit proliferation and induce apoptosis of tumorigenic breast cells more potently than for normal Type H HBEC? [1. LITERATURE REVIEW A. Breast cancer 1. Breast cancer epidemiology-Breast cancer is the most common cancer and the second leading cause of cancer-related deaths in women in the United States. About 178,700 new invasive cases were estimated to occur in 1998 in the United States and the incidence of breast cancer is approximately 110 cases per 100,000 women (Cancer Facts and Figures, 1998). Although the mortality rate has stabilized in recent years mainly due to earlier detection and improved treatment, there were still an estimated 43,500 deaths during 1998. 2. Breast cancer etiology-Breast cancer arises as a result of the accumulation of mutations of critical genes that regulate cell proliferation, differentiation and death in breast cells (Osteen et al., 1986). The majority of breast cancers are adenocarcinomas originating from epithelial cells. Terminal end buds, which contain highly proliferating mammary epithelial stem cells, are considered to be the target of mammary neoplastic transformation (Russo and Russo, 1987). Many risk factors for breast cancer in females have been documented (Marshall et al., 1993). The most significant ones include old age, early menarche, late first full-term pregnancy, late menopause, obesity after menopause, ovariectomy, history of fibrocystic disease and history of primary cancer in ovary or endometrium, family history of premenopausal bilateral breast cancer and place of birth (North America, Europe >Asia, Africa). 3. Breast cancer and diet-The diet is a complex mixture containing both pro- carcinogenic and anti-carcinogenic agents. The correlation of higher breast cancer incidence with early menarche and taller stature may be reflective of the influence of nutrition during childhood and adolescence (Pollner, 1993). Animal studies show that caloric restriction reduces breast cancer incidence (Klurfeld et al., 1991). Obese postmenopausal women have a higher risk of breast cancer (Cleary and Maihle, 1997) and overweight breast cancer patients have a poorer prognosis (Bastarrachea et al., 1994). High alcohol consumption increases the risk of breast cancer (Davis etal., 1993) while dietary carotenoids (Verhoeven er al., 1997) and vitamins A (Sankaranarayanan and Mathew, 1996), C (Verhoeven et al., 1997) and E (Kimmick et al., 1997) may help to prevent breast cancer mainly by acting as antioxidants. Blood 1,25-dihydroxyvitamin D3 levels are associated with low risk of breast cancer (Janowsky et al., 1997), probably through regulating proliferation and difl'erentiation of breast cells (Reichel et al., 1989). Epidemiology studies suggest a reduced risk of breast cancer in women who consume fermented milk products (Veer etal., 1989) and in vitro studies show antiproliferative activity of fermented milk in MCF-7 breast cancer cells (Bifli er al., 1997). Milk fat components including conjugated linoleic acid, butyric acid, ether lipids, and sphingomyelin show anticarcinogenic effects in animal experiments (Parodi, 1997). Soy intake was found to be inversely correlated with the risk of breast cancer among 142,857 women in Japan over a period of 17 years (Messina etal., 1994) and diets containing soybeans reduced mammary tumor occurrence induced by irradiation (Troll er al., 1980) and the chemical carcinogen, N—methyl-N-nitrosourea WU) in animal models (Barnes er al., 1988). Phytoestrogens present in soybeans appear to help prevent mammary cancer (Messina et a1 ., 1997). 4. Role of stem cell differentiation in breast cancer prevention-One important histologic feature of malignancy is anaplasia (ie. a loss of differentiation). As a result, cancer has been described as a disease of differentiation (Markert, 1968) or oncogeny as blocked or partially blocked ontogeny (Potter, 1978 and 1987). Stem cells are undifferentiated cells that are capable of proliferation and self-maintenance, and which produce a large number of differentiated progeny cells (Loefller, 1997). The relatively undifferentiated nature of tumor cells could be due to the de-differentiation of differentiated cells or blocked difi'erentiation in stem cells which give rise to cancer cells (V armur and Weinberg, 1993). As mentioned before, early full-term pregnancy decreases risk for breast cancer. This might be related to stem cell multiplication that occurs begimring at the time of puperty and during each ovarian cycle until, but not after, the first pregnancy (Cairns, 197 5). Alternatively, pregnancy may induce firll differentiation of the mammary gland (Russo et al., 1990), thus decreasing the susceptibility for carcinogenesis by reducing the number of stem cells. Similarly, dietary components that reduce the incidence of breast cancer may act by reducing the number of stem cells. 5. Breast cancer chemotherapy and apoptosis-Apoptosis is a type of programmed cell death with distinctive morphological and biochemical changes which allows deletion of unwanted cells from an organism (Vaux et al., 1996). In apoptosis, the nuclear chromatin is ' condensed and aggregates under the nuclear membrane. Then, activation of endonucleases causes fiagrnentation of DNA into multiples of 200 base pair nucleosome—sized pieces. Cells shrink, exhibit cytoplasmic budding, and fragment into membrance-bound vesicles of cytosol and organelles termed apoptotic bodies. Normally, apoptotic bodies are phagocytosed and degraded without eliciting an inflammatory response in the surrounding tissue (Walker et al., 1988). The occurrence of apoptosis can be determined by the appearance of morphologic alterations and endonucleosomal DNA fragments. Apoptosis-inducing agents which selectively kill cancer cells and/or have less effects on neighboring normal cells would be ideal Chemotheraputic agents. In fact, many chemotherapeutic agents cause cancer cell death mainly by inducing apoptosis (Hannun, 1997). Difi‘erent types of cells vary in their susceptibility to apoptotic induction; thus, treatments may induce apoptosis in tumor cells while arresting the cell cycle in normal cell counterparts (Fisher, 1994). The apoptosis pathway may be disrupted in tumor cells, which leads to both a survival advantage and resistance to treatment (Fisher, 1994). Alternatively, improper stimulation of proliferation may lead to apoptosis since prom-oncogenes such as c—myc (Green, 1997) as well as c-fos and c-jun (Preston er al., 1996) which stimulate cell proliferation can also induce apoptosis. 6. Telomerase activity, a biomarker of cell cyde progression and differentiation- Telomeres are repetitive TTAGGG sequences at the ends of eukaryotic chromosomes that shorten by 50-200 base pairs after each cell division because of the incomplete replication of the 5’ ends of DNA molecules (Shay et al., 1993). Telomeres are required for proper chromosome segregation during mitosis by preventing nuclease degradation and end-to-end firsion of chromosomes during replication (Kirk et al., 1997). Telomerase is a nbonucleoprotein enzyme that synthesizes telomeric DNA, thereby preventing the replication- dependent shortening of DNA Telomerase activity is detected in 85% to 95% of immortal and tumor cells including breast cancer cells but rarely in normal cells except germ cells and stem cells (Shay et al., 1993, Belair er al., 1997). In human breast cancer, telomerase activity is associated with cell cycle regulatory defects such as overexpression of cyclin D1 and/or cyclin B (Landberg eta1., 1997). Also, absence of telomerase activity was reported in G2/M- synchronized MCF-7 and MBA-435 breast cancer cells (Zhu et al., 1996). Recently, telomerase also has been used as a biomarker of cell differentiation because of the correlation between telomerase activity and classification of colorectal carcinomas (Okayasu et al., 1998), esophageal cancer (Asai et al., 1998), prostate cancer (Uemura et al., 1998) and leukemia (Zhang et al., 1996). Telomerase activity also is correlated with tumor aggressiveness and therapeutic effects (Hoos et al., 1998). Inactivation of telomerase activity and differentiation therapy are new therapeutic approaches in prostate cancer (Schalken, 1998). The regulation of telomerase in cancer cells is not clear; however, protein kinase C (PKC) inhibitors such as sphingosine were shown to specifically inhibit telomerase activity in human nasopharyngeal cancer cells (Ku etal., 1997). B. Sphingosine and ceramide regulate cell behavior 1. Sphingolipid metabolism-Sphingolipids are bioactive molecules with a sphingoid base backbone. Complex sphingolipids, such as sphingomyelin, are major constituents of all eukaryotic and some prokaryotic cells and account for about 20% of plasma membrane lipid (Kolesnick et al., 1991). Sphingomyelin has a sphingoid base backbone primarily composed of sphingosine, an amide-linked fatty acid and a phosphorylcholine polar head group (Figure 1). Upon stimulation with agonists such as 10:, 25-dihydroxyvitamin D3 (Okazak et al., 1989 and Nikolova et al., 1997), tumor necrosis factor-a, y-interferon (Kim et al., 1991) and interleukin-113 (Ballou et al., 1992 and Nikolova-Karakashian et al., 1997), sphingomyelin is hydrolyzed to ceramide and/or sphingosine (N ikolova-Karakashian et al., 1997). Both ceramide and sphingosine act as second messengers that mediate diverse cellular behaviors (Figure 2) including cell proliferation, difi‘erentiation and apoptosis (Merrill et al., 1997). Because of their ability to inhibit proliferation and induce apoptosis in certain tumor cell lines, sphingolipids may be useful agents in treating various cancers. As components of food, complex sphingolipids are found in relatively high concentration in soybeans, eggs and dairy products (Vesper et al., 1996, Ahn and Schroeder, 1998). About 90% of complex sphingolipids are digested and absorbed throughout the small intestine to ceramide, sphingosine and other metabolites; however, about 10% reaches the colon (Schemelz et al., 1994). Feeding milk sphingomyelin was shown to suppress the appearance of both the dysplasia lesions and the more advanced malignant colon tumors in rats treated with the colon carcinogen 1,2-dirnethylhydrazine (Schemelz et al., 1996). Dietary sphingomyelin can increase serum sphingomyelin in a dose-dependent manner (Irnaizurni er al., 1992), which suggests that dietary sphingolipids can also reach other organs via the circulation. 2. Ceramide inhibits cell proliferation, causes differentiation, and induces apoptosis-Cell-penneable ceramide analogues, such as Cz-ceramide and C‘s-ceramide can mimic the effects of exogenous stimuli, which supports a role of ceramide in mediating multiple cellular functions triggered by agonists. Among them, the most distinctive actions are inhibition of proliferation and induction of apoptosis (Kolesnick and Kronke, 1998). In addition, ceramide may mediate the effects of ionizing radiation and the breast cancer chemotherapeutic agents vincristine and doxorubicinn which have been shown to cause accumulation of ceramide (Hannun, 1997). Cerarnide inhibits the proliferation of normal fibroblasts (Harmun er al., 1994) and induces apoptosis in fibroblasts (Cifone et al., 1994) in a number of other cell lines including human myeloid leukemia cells (Jarvis et al., 1994), human pancreatic cancer cells (Yamada etal., 1997), prostate cancer cells (Herrmann et al., 1997), oligodendrocytes (Larocca er al., 1997) and rat neonatal cardiomyocytes (Bielawska et al., 1997). In HL-60 leukemia cells, Cz-ceramide inhibited cell proliferation at concentrations as low as 1-10 uM and induced cell differentiation (Okazaki et al., 1989). Exogenous C6-cerarnide (15 uM) induced cell cycle arrest at the G,,/G1 phase in Molt-4 T H OH Sphingosine MMM/Vtr' Ceramide W0“ 0" Sphingomyelin WMWWWWM W 0 Figure 1. Structures of sphingosine, ceramide and sphingomyelin Agonists Sphingomyelin \ / Plasma membrane Sphingomyelinase l Cerarnide "'”’ Growth arrest Ceramidase l Differentiation . . m... Apoptosis Sphingosrne \\ // Figure 2. Sphingomyelin turnover pathway. leukemia cells and Wi 38 human fibroblast cells (Jayadev et al., 1995). 3. Sphingosine inhibits cell proliferation, causes differentiation, and induces apoptosis-Ceramide can be deacylated by ceramidase to form sphingosine. Therefore, some cellular effects originally attributed to ceramide might be mediated by sphingosine (Ohta et al., 1994). One piece of evidence that supports this hypothesis is that exogenously added sphingosine induced apoptosis much earlier than ceramide in human neutrophils (Ohta et al., 1995). Sphingosine induced apoptosis in human myeloid leukemia cells, human prostatic carcinoma cells (Shiraharna et al., 1997), and other solid tumor cell lines (Sakakura et al., 1996, Sweeney et al., 1996). An interesting observation is sphingosine induces apoptosis in SV-40 transformed epithelial cells such as HUVECS and rat mesangial cells, but not in their primary culture counterparts (Sweeney et al., 1996), which implies sphingosine may be an excellent candidate as an anti-cancer agent. Sphingosine inhibited the proliferation of Chinese hamster ovary cells (Merrill, er al., 1989) and human T-lymphocytes (Borchaidt et al., 1994), and caused cell-cycle arrest at GOIG, in HL-60 cells (Chao et al., 1992). Like the induction of apoptosis, the efl‘ects of ceramide on cell proliferation and differentiation could also be mediated by sphingosine. In neuroblastrna Neuro 2a cells, there was a concomitant, early and sustained increase in the ceramide concentration during retinoic acid-induced differentiation; whereas, supplying either sphingosine or ceramide induced neurite formation and inhibited thymidine incorporation into DNA (Riboni et al., 1995). The mechanisms whereby sphingosine inhibits tumor cell proliferation and induces apoptosis are not firlly understood. Sphingosine inhibits protein kinase C isoenzyme family members and has anti-proliferative properties (Hannun et al., 1986). Sphingosine is also a 10 potent inhibitor of a mammalian RNA primase in vitro (Simbulan et al., 1994) and the suppression of proliferation of human leukemic I-IL—60 cells was correlated with DNA primase inhibition (Tamiya et al., 1997). Sphingosine (15 uM) induced apoptosis in androgen- independent human prostatic carcinoma DU-145 cells by down-regulation of either bcl-2 or bcl--XL gene expression (Sakakure et al., 1996”). Sphingosine was also shown to down- regulate c-myc gene expression and caused retinoblastoma protein (Rb) dephosphorylation (Hannun et al., 1993, Merrill et al., 1991, 1993). In hernatopoietic cells, sphingosine-induced dephosphorylation of retinoblastoma protein preceded inhibition of proliferation and cell cycle arrest (Chao et al., 1992). 4. Ceramide and sphingosine have the potential to prevent and treat breast cancer-The potent inhibition of proliferation and induction of difl‘erentiation and apoptosis in certain tumor cell lines by ceramide and sphingosine suggests that they could be useful agents in the treatment of cancer. Previous studies showed that ceramide and sphingosine induced cell death in estrogen receptor-negative MDA-MB-231 human breast cancer cells (Zhang and Schroeder, 1998); however, only sphingosine, but not ceramide, caused a DNA ladder in agarose gel electrophoresis and the formation of a pre-GOIG, peak in flow cytometric analysis, both of which are indications of apoptosis. Other studies suggest that ceramide can induce apoptosis in human breast cancer cell lines. Cz-ceramide induced a dose-dependent increase in apoptosis in HS 578T and there was a significant increase in the percentage of cells in the pre-Gl phase of the cycle in cells treated with as low as 4 uM Cz-ceramide for 24 hours (Gill et al., 1997). Cé-ceramide caused death of MCF-7 cells in a dose-dependent manner at 48 hours by inducing apoptosis (Cai et al., 1997). In the present study, we have investigated the potential of ceramide and sphingosine to prevent and treat human breast 11 cancer using breast epithelial and cancer cell models. C. Normal human breast epithelial and cancer cells in culture as experimental models. Recently, two types of morphologically distinguishable normal human breast epithelial cells (HBEC) were derived fi'om reduction mammoplasty (Kao er al., 1995, Figure 3 and 4). Type I HBEC express luminal epithelial cell markers (i. e. epithelial membrane antigen, cytokeratin-18, 19) and have stem cell characteristics (1'. e. deficiency in gap junctional intercellular communication, ability to differentiate to Type H HBEC and to form budding/ductal structures on Matrigel) (Kao et al., 1995; Sun et al., 1999). Type I HBEC express estrogen receptors (Kang et al., 1997) and are more susceptible to telomerase activation and irnortalization after SV40 transformation (Sun et al., 1999). Type I HBEC and these SV40 transformed cells were also capable of anchorage independent growth (Kao er al., 1995). Furthermore, neoplastically transformed Type I HBEC, similar to breast carcinomas, possess many phenotypes of Type I HBEC (i. e. expression of epithelial membranc antigen, cytokeratin 18 and estrogen receptors, and deficiency in gap junctional intercellular communication) (Kao et al., 1995; Kang et al., 1997). Therefore, Type I HBEC appear to be the target cells for neoplastic transformation. In contrast, Type II HBEC which express basal epithelial cell markers (at-6 integrin and cytokeratin-14) and gap junction genes (connexin 26 and 43) but not estrogen receptors, did not show anchorage independent growth and rarely became immortal after SV40 transformation (Kao et al., 1995, Sun et al., 1999, Figure 4). Type I HBEC are an excellent cell model to study the chemopreventive potential of sphingolipids for human breast cancer because they have stem cell characteristics and stem cells appear to be the targets of breast carcinogenesis. In addition, Type H HBEC and 12 Figure 3. Morphology of two types of normal human breast epithelial cells (HBEC) (photographs) tumorigenic breast cancer cells can be used to compare the effects of sphingolipids on the proliferation, difl‘erentiation and death of normal human breast epithelial cells and breast cancer cells to evaluate their chemotheraputic potential. We hypothesize that: (l) Sphingosine and ceramide may inhibit proliferation and induce differentiation of Type I HBEC. If so, they may be chemopreventive to human breast cancer by reducing the target cells for neoplastic transformation and inhibiting the proliferation of initiated (precancerous) cells; (2) Sphingosine and ceramide may also inhibit the proliferation or trigger the apoptotic death of the neoplastic transformed Type I HBEC and MCF-7 cells, and thus may be useful as chemotherapeutic drugs for human breast cancer. 14 Type I HBEC Type II HBEC Differentiation ’ $4— SV40 transfection a; Immortalized at Immortalized at high frequency low frequency I <— X-rays Weakly tumorigenic <— neu oncogene Highly tumorigenic Tumorigenic Transformed Type I HBEC Figure 4. Type I HBEC derived fiom reduction mammoplasty have the ability to difi‘erentiate to Type II I-IBEC and are susceptible to neoplastic transformation. 15 HI. MATERIALS AND METHODS Cell culture and sphingolipids treatment-Type I and Type H normal HBEC, an in vitro neoplastically transformed Type I HBEC (M13SV1R2N1) (Kang et al., 1998), MCF-7 and MDA-MB-231 breast cancer cells were kindly provided by Dr. C. C. Chang. Type I HBEC were cultured in MSU—l medium (Kao er al., 1995) with 5% fetal bovine serum (FBS). Type H HBEC were cultured in FBS-firee MSU-l medium supplemented with bovine pituitary extract (0.4 % V/V). A modified Eagle’s MEM (D medium) (Chang et al., 1981) with 5 % PBS was used to culture transformed Type I I-IBEC and breast cancer cell lines. Cells were cultured in incubators at 37°C and supplied with 5% CO2 and humidified air. Sphingolipids were obtained fi'om Matreya, Inc. Sphingosine was added as a 1:1 complex with bovine serum albumin (BSA) dissolved in phosphate bufi‘ered saline (PBS) while C2- ceramide was dissolved in 70% ethanol. Assessment of cell prolrferation-Cell proliferation was measured by quantitation of total nucleic acid extracted from the cultures (Li et al., 1990). Briefly, cells (6>< 10‘) were cultured in 6—well dishes in triplicate. The next day, various concentrations (1-10 uM) of C2- ceramide and sphingosine were added to FBS-free medium. Cells were incubated for 5 days with a change to fresh medium and treatments on day 3. The cells were harvested by rinsing twice with PBS followed by lysis with 1 mL of 0.1 N NaOH. Cell lysates were transfered to 2.2 mL Eppendorf tubes and centrifuged at 14,000 rpm for 2-3 min. The clear lysates were measured for absorbance at 260 nm using a Beckrnan DU-7400 spectrophotometer. Assasmart of T fire I HBEC rfiflerentiatr’m—Starting from single cell platings of pure Type I HBEC, the difl‘erentiation of Type I HBEC was measured by counting the number of Type H HBEC colonies and colonies of Type I surrounded by Type H HBEC. The 16 percentage of these colonies among total colonies indicates the difi‘erentiation potential of Type I HBEC under different treatments. Briefly, Type I HBEC (5X103) were cultured in MSU-l medium containing 5% FBS in triplicate in 60 mm dishes with grids. On the third day, various concentrations (1-3 pM) of Cz-ceramide and sphingosine were added to FBS- free MSU-l medium. Cells were incubated for 12 days with change of medium and renewed treatments every 2 days. Cholera toxin (CT) (1 ng/mL) was used as a positive control (Kao et al., 1995). Then, Type I, Type I surrounded by Type H and Type H colonies were visually identified under a microscope and quantitated. The identity of the treatment for each dish was unknown (blind) during counting to ensure objectivity. Analysis of DNA fragmentation by agarose gel electrophoresis-Cells were trypsinized and centrifirged at 2.2 K rpm for '10 min. Pellets were resuspended in 200 uL PBS and DNA was extracted with 400 uL phenol chloroform followed by centrifirgation at 14 K rpm for 5 min. After incubation with 0.2 mg/mL RNase A at 37° C for 1 hr, the extraction was repeated with 400 uL phenol/chloroform to inactivate RN ase A Twenty uL of 3 M NaOAc and 500 uL of 100% ethanol were added to the solution before storing at —-20°C overnight. After centrifirgation at 14 K rpm (4° C) for 30 min, DNA pellets were washed with 70% ethanol and dried en vacuo. The DNA pellets were then dissolved in Tris- EDTA (pH= 8) and the DNA was separated on a 2% agarose gel at 100 V for 1 hr. SDS-PA GE and Western blotting-Proteins were extracted from cells in 100 mm dishes by treatment with 20% SDS lysis solution containing protease and phosphatase inhibitors (1 mM phenylrnethylsulfonyl fluoride, 1 uM leupeptin, 1 uM antipain, 0.1 uM aprotinin, 0.1 M sodium orthovanadate, 5 mM sodium fluoride). After sonication via three 10-s pulses with a probe sonicator, the cell lysates were stored at —20°C until use. The 17 protein amounts were determined using the DC protein assay kit (Bio-Rad Co., Richmond, CA). Proteins were separated on 12.5% SDS polyacrylamide gels and transferred to PVDF membranes at 20 V for 16 hr. After blocking with 5% dried skim milk in PBS containing 0. 1% Tween 20, the membrane were exposed to an anti-pr monoclonal antibody (Oncogene Science, NY). This was then followed by incubation with horseradish peroxidase-conjugated secondary antibody and detected with the ECL chemiluminescent detection reagent (Amersham Co., Arlington Heights, H.) X-ray film was exposed to membranes for 15 s to 1 min (Kang et al., 1996). Flow cytomeaic measurement and cell cycle analym's-A quantitative measurement of cell cycle distribution was obtained by flow cytometric analysis of DNA content-cell number frequency histograms, as described by Fraker et al. (1995). Briefly, cells were collected after 2 day treatment by trypsinization followed by centrifuging at 1200 rpm for 5 min Then the cells were resuspmded in 5 mL of PBS and transferred to a Falcon 2056 tube. The cells were pelleted by centrifuging at 1200 rpm for 5 min followed by aspiration of the PBS wash Cells were fixed in ice cold 70% ethanol with rapid but gentle mixing at a density of 1><106 cells/mL. After the cells were fixed in ethanol for 1 to 3 hr at 4°C, samples were stored at -20°C until analysis. For staining, the cells were centrifuged at 2500 rpm for 5 min to remove ethanol, washed one time with PBS and pelleted as above. The cells were resuspended in flow cytometric DNA staining reagent (0.1 mM EDTA [pH 7.4], 0.1% of Triton X-100, 0.05 mg/mL RNase A [50 units/mg], and SOag/mL propidium iodide [P1] in PBS [pH 7.4]) before incubating overnight in the dark at 4°C. Fluorescence was assessed on a FAC S Vantage (Beckton Dickinson) by excitation with an Argon laser at 488 nm and the emission was detected at 620 to 700 nm. Data were collected with Lysis H software and 18 the percentage of cells in each phase of the cell cycle was calculated with MPLUS software (Phoenix Flow). Apoptotic cells were determined as the percentage of cells with a DNA content of less than diploid. PCR-based telomerase assay-Cells were harvested by trypsinization. After cell counting, the cells were centrifuged to remove trypsin solution. Cell pellets were washed with 10 mL PBS and then centrifuged to remove PBS. Cells were then suspended at 1>< 10° cells/mL in PBS and aliquoted into Eppendoif tubes. After cells were centrifuged and the PBS was carefully removed, the cell pellets were stored at -85 °C. The telomerase assay was kindly performed by Dr. Wei Sun. The cell pellets were thawed and resuspended in 200 juL of 1x CHAPS lysis buffer/ 106 cells and left on ice for 30 min. The samples were centrifirged at 12,000 g for 20 min at 4°C. The cell lysates for each sample were aliquoted to several new tubes and stored at -85°C. The original lysates represent the concentration of 5000 cells/uL. Further dilution of cell lysates were adjusted based on the level of telomerase activity from individual cell lines. Telomerase activities were measured utilizing the TRAPezeTM Telomerase Detection Kit (Oncor, Gaithersburg, MD) which includes a primer of a 36 base pairs internal standard for amplification, thus providing a positive control for accurate quantitation of telomerase activity. Each analysis included a negative control (CHAPS-lysis bufl'er without sample), heat-inactivated control (sample incubated at 85 °C for 10 min prior to the assay) and positive cell line control (MCF-7 breast carcinoma cells). The products of the TRAP assay were resolved by electrophoresis in a non-denaturing 12% polyacrylamide gel electrophoresis (PAGE) in a buffer containing 54 mM Tris-HCL (pH 8.0), 54 mM boric acid and 1.2 mM EDTA. The gel was stained with Syber Green (Molecular Probes, Inc., Eugene, OR), and visualized using a 302 nm UV transilluminator. Images were captured and 19 analyzed by AlphalmagerTM (Alpha Innotech Corporation, SanLeandro, CA). Statistical analyses-Statistical analyses were conducted using the SPSS software. Data of cell proliferation were analyzed by Mo-way factorial analysis of variance (AN OVA). Differences in total nucleic acid content between control and treatment groups at specific culture periods were evaluated by multiple comparisons using Dunnett t-test. Difl‘erences were considered significant at p<0.05. 20 IV. RESULTS A. Sphingosine and ceramide inhibit cell proliferation To assess the efl‘ects of sphingosine and ceramide on the proliferation of normal and tumorigenic breast cells, subconfluent cells were cultured with sphingosine and Cz-ceramide. The total nucleic acid content was quantitated as an index of cell number. Sphingosine inhibits proliferation and causes death of Type I HBEC and tumorigenic breast cells, but not Type H HBEC at 8 uM-F or Type I HBEC (Figure 5A), control cultures grew slowly over the culture period with total nucleic acid increasing about 70% in 5 days. Sphingosine at 2 and 5 uM did not effect cell proliferation; however, sphingosine at 8 uM reduced the nucleic acid concentration to about 60% of the corresponding control cultures at day 1 and floating dead cells were visible in the medium. Thereafter, for cells treated with 8 uM sphingosine, the total nucleic acid concentration remained the same through day 5 suggesting that sphingosine blocked cell proliferation. For Type H HBEC (Figure 5B), the control cell number tripled in 2 days and remained in log-phase at 5 days of culture. Cell proliferation was not affected by sphingosine at concentrations as high as 8 uM, which was the highest concentration tested. The addition of sphingosine caused concentration- and time-dependent decreases in total nucleic acid concentration in all three tumorigenic breast cell lines (Figure 6). Sphingosine at 8 pM significantly inhibited the proliferation of and caused death of all of the tumorigenic breast cell lines. The transformed Type I HBEC line was the most sensitive, with 5 uM sphingosine reducing total nucleic acid content to about 50% compared with control at day 5 (Figure 6A). For MCF-7 cells and MDA-MB-231 cells, 5 uM sphingosine reduced total nucleic acid content to about 80% of controls at day 5 (Figure 6B and C). 21 A 200 a: -I— SuM A 150 E —A— SuM 2 (5 m + 2uM E 3 5" -0— Control 0 a: o . o 1 3 4 5 Culture Period (day) B A a: 1200 -I- 8uM .: 1000 E 800 -A— SuM % 600 + 2uM .E 400 % 200 + Control M . . . . o o 1 2 3 4 5 Culture Period (day) Figure 5. Sphingosine at 8 uM inhibits proliferation and causes death of Type I HBEC but not Type H HBEC. Type I (A) and H (B) HBEC (6X10‘) were cultured in 6-well dishes in triplicate and treated with sphingosine at day 0 and day 3. Total nucleic acid was measured by spectrophotometry (1:260 nm) and used as an index of cell number. Results shown are mean i SD (n=3). Standard deviations which are not visible are hidden by the symbols. 22 “B‘IOuM A g m + M. .4: E 600 .. sa— SuM E (g 400 + 2uM g 200 -e— 0.5uM a o - - . 0 l 2 3 4 5 + Control Culture Period (day) g B "E 2 CD 0 .g % a: E a c E 2 CD 5% .9. £2 Culture Period (day) Figure 6. Sphingosine inhibits proliferation and causes death of tumorigenic breast cells. MCF-7 (A) and MDA-MB-231 (B) breast cancer cells and transformed Type I HBEC (C)(6><10‘) were cultured in 6-well dish in triplicate and treated with sphingosine. Cell proliferation was assessed by total nucleic acid content. Results shown are mean at SD (n=3). Standard deviations which are not visible are hidden by the symbols. 23 nuclr Cz-c and at 5 and C3-c prol Typ unna with mom Iran C310 Pro Hon 1].: Cerarnide inhibits cell proliferation and causes death-For Type I HBEC, the total nucleic acid content for vehicle (ethanol) control culture nearly tripled in 5 days (Figure 7A). Cz-ceramide at 1 uM did not effect cell proliferation over 1 day; whereas, Cz-ceramide at 2 and 5 uM significantly inhibited cell proliferation and caused death within 1 day. Cz-ceramide at 5 uM reduced nucleic acid content to about 60% of the corresponding control at day 1 and to about 25% of the control by day 5. Cz-ceramide at 8 pM killed all of the cells by day 3 (data not shown). For Type H HBEC (Figure 7B), Cz-ceramide at 5 pM significantly inlnbited cell proliferation and caused cell death within 1 day. For tumorigenic breast cells, Cz-ceramide was more potent than sphingosine as 5 uM Cz-ceramide completely inhibited cell proliferation in all three tested cell lines (Figure 8). Sphingosine stereoisomers inhibit proliferation and cause death of transformed Type I HBEC-To study the structural requirements for sphingosine to cause cell death, three urmatural stereoisomers, D-threo, L-threo and Irerythro-sphingosine were examined together with D-erythro-sphingosine (Figure 9). All three unnatural stereoisomers of sphingosine were more potent than D-erythro—sphingosine in inhibiting proliferation and causing death of transformed Type I HBEC. L-erythro-sphingosine was the most potent. Sphingosine and ceramide cause cell cycle arrest at G./Gl or Gle-Flow cytometric analysis was used to study the effects of sphingosine and Cz-ceramide on cell cycle progression in Type I HBEC (Table 1 and Figure 10-11), MCF-7 cells (Table 2) and transformed Type I HBEC (Table 3) . As shown in Table 1, control Type I HBEC exhibited normal homeostatic cell cycle distribution with 73.4 % of the cells in the G0/G, phase and 11.3 % in the S (DNA synthesis) phase. Sphingosine at 10 uM caused cell cycle arrest at the G(/G1 phase within 1 day as indicated by an increase of cells in GolG1 phase to 91.5% and a 24 Relative Growth (95) Culture Period (day) E 13" " -+5nM rue . E ,._ 4...... 93 “0 ” +1.»! E "°‘ _3 2.. +Control “ e e r z 3 4 s CulturePerlod(day) Figure 7. Q-ceramide at 5 uM inhibits proliferation and causes death of Type I and H HBEC. Type I (A) and II (B) HBEC (6X10‘) were cultured in 6-well plates in triplicate and treated with Cz-ceramide. Cell proliferation was assessed by total nucleic acid content. Results shown are mean 3: SD (n=3). Standard deviations which are not visible are hidden by the symbols. 25 "5‘th A 3!; 800- +8uM Emo- +5uM Lg 400~ +2nM > 3 200 -9-o.5uM 0 M o 0 l 2 3 4 5+Control CulturePerlod(day) As.- a? -A-SIM B “no. E *r—v—m *- a an» “-6-!“ gm "-O-Co-u-ol 9 a: . . . . . O l 2 3 4 5 Culturel’erlod(day) C 231...- ? no 5.... 0 . g are .2! no 0 a O Culture Period (day) Figure 8. Cz-ceramide at 5 uM or higher concentrations inhibits proliferation and causes death of tumorigenic breast cells. MCF-7 (A) and MDA-MB-231 (B) breast cancer cells and transformed Type I HBEC (C)(6><10‘) were cultured in 6-well plates in triplicate and treated with Cz-ceramide. Cell proliferation was assessed by total nucleic acid content. Results shown are mean :h SD (n=3). Standard deviations which are not visible are hidden by the symbols. 26 Table 1. Sphingosine and ceramide afl‘ect cell-cycle distribution of Type I HBEC. Data are shown as percentage of the cells in each phase of cell cycle. Treatments G0 /G1 phase S phase (2le phase (‘70) ('70) (‘70) Control 73.4 11.3 15.4 Sphingosine (10 uM) 8 h 75.5 9.5 15.0 Sphingosine (10 uM) 24 h 91.5 2.0 6.5 Cz-ceramide (5 uM) 8 h 78.5 8.4 13.1 Cz-ceramide (5 uM) 24 h 77.7 10.2 ' 12.0 27 Table 2. Sphingosine and ceramide affect cell-cycle distribution of MCF-7 breast cancer cell lines. Data are showed as percentage of the cells in each phase of cell cycle. Treatments G0 /G1 phase S phase Gle phase (70) (W (‘70) Sphingosine (control) 43.9 29.8 26.3 Sphingosine (10 uM) 51.0 37.8 11.2 Cz-ceramide (control) 63.5 26.2 10.2 Cz-ceramide (10 uM) 79.1 10.0 10.9 28 Table 3. Sphingosine and ceramide affect cell-cycle distribution of transformed Type I HBEC. Data are showed as percentage of the cells in each phase of cell cycle. Treatments G0 /G1 phase S phase G2 [M phase (”/0) (Va) (%) Sphingosine (control) 1d 35.8 24.5 39.7 Sphingosine (10 pM) 1d 39.5 25.3 35.2 Cz-ceramide (control) 1d 41.5 34.9 23.7 Cz-ceramide (10 uM) 1d 47.3 19.7 33.0 Sphingosine (control) 2d 47.3 29.1 23.6 Sphingosine (10 uM) 2d 46.6 22.8 30.5 CZ-ceramide (control) 2d 49.0 23.5 27.5 Cz-ceramide (10 pM) 2d 62.0 0.0 38.0 29 Control D-erythro-So D-threo-So L-threo-So L-erythro-So Relative Growth (%) ¢+E|>+