2...... .I. .. s -3. a... a. .d! 7.3.: r3. 6. .2... 55...... 1. x .2 :5: l... 5.. .x- f. U ‘m .56. . Ila-.3 3.53.44. J . Iii-g.» 1 hm.“ 2.5. "In .5 hr “flunk"... ”Ta I333: .: .5. I£l5l&_¢-:: O 1;.2‘11 .151: . i9 .20 : t . . 1...: . a. My! .2 :0... 1. , :\ $1.»..c =th v o . any," va‘naan‘ti.’r ’at-t“ 11“}: B. , ah... rt... I37 1.113%...»2.‘ i .9 .2 .0.-. All, It...“ ‘OAF 1 , ii skunfl 5... A... .. . z..- , . . . L: .53.... 2.52“... .3 . .11 fifififixuuwu : “mach... ‘ . :01... hand... $3.... |. 2...! v.3...) .zinryhlliz.-. .1 1:19:51; ,itlls. 721.. . ..‘.!..riv .3. 3‘ .I». .. 2‘1»?! 7... 1.. tl \. lHtfilS I 2.050 This is to certify that the thesis entitled INFLUENCE OF SOYBEAN PRODUCTS AND THE NON_STEROIDAL ANTI:INFLAMMATORY DRUG SULINDAC ON COLORECTAL CANCER presented by BING WANG has been accepted towards fulfillment ' of the requirements for MS , HUMAN NUTRITION degree in Major professor Date A‘f‘M‘RII‘QwO 0-7639 MS U is an Affirmative Action/Equal Opportunity Institution LIBRARY Michigan State University PLACE IN RETURN BOX to remove this checkout from your record. To AVOID FINES return on or before date due. MAY BE RECALLED with earlier due date if requested. DATE DUE DATE DUE DATE DUE 8/01 chlFlC/DatoDuepGS—p. 15 INFLUENCE OF SOYBEAN PRODUCTS AND THE NON-STEROIDAL ANTI-INFLAMMATORY DRUG SULINDAC ON COLORECTAL CANCER By BING WANG 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 2000 ABSTRACT INFLUENCE OF SOYBEAN PRODUCTS AND THE NON-STEROIDAL ANTI-INFLAMMATORY DRUG SULINDAC ON COLORECTAL CANCER BY BING WANG A study was conducted to investigate the chemopreventive effects of soybean products and a non-steroidal anti-inflammatory drug, sulindac, on colorectal cancer. Five dietary treatments were used in two experiments. The protein sources for the diets were: 1) casein, 2) soy concentrate, 3) an extruded 30% soy product, and 4) a non-extruded soy product identical to 3. A fifih treatment group was fed the casein diet and given sulindac (200 ppm) in the drinking water. In experiment one, azoxymethane-induced aberrant crypt foci (ACF) in rat colon were used as the endpoint to study colon cancer initiation. No effects of soy protein or sulindac were found in terms of the number of total ACF, large ACF, and average size of ACF. In experiment two, adenoma development in the intestines of Min mice was used as the endpoint to study colon cancer promotion. No protective effects of soy protein were found. Mice fed the extruded soy had larger (P < 0.05) small intestinal tumors than mice fed soy concentrate or non-extruded soy, and also tended to have higher tumor incidence in the large intestine compared to mice in the other treatment groups. However, the intestinal tumori genesis of mice on the three soy diets was not statistically different from that observed in mice fed the casein diet. Sulindac significantly (P < 0.05) reduced the size and number of small intestinal tumors. Sulindac did not significantly influence tumor formation in the large intestine. ACKNOWLEDGMENTS I would like to sincerely thank my major professor, Dr. Leslie Bourquin for his guidance and constant support. Thanks are also extended to members of my committee, Dr. Dale Romsos and Dr. Maurice Bennink for their guidance and encouragement throughout the program. I would like to thank all my friends for the help they have given to me. I would also like to express my gratefulness to my dear wife and my parents for their love and understanding. iii TABLE OF CONTENTS page LIST OF TABLES vi LIST OF FIGURES vii I. INTRODUCTION 1 11. REVIEW OF LITERATURE 3 A. Epidemiological studies 3 1. Colorectal cancer incidence and distribution 2. Overview about diet and colorectal cancer B. Genetic study of colorectal cancer 4 1. Multi-step nature of colorectal tumorigenesis 2. Genetic basis of colorectal cancer C. Colorectal cancer chemoprevention 6 D. Soy consumption and colorectal cancer 7 1. Epidemiological evidence 2. Influence of processing process on soybean products 3. Animal studies with soy products 4. Proposed anti-carcinogenic compounds in soybeans 5. Human clinical trials with soybean products E. Models and biomarkers for chemoprevention studies 15 1. Animal models vs. cell culture 2. Carcinogen-induced colorectal carcinogenesis model 3. ACF as early precancerous lesions 4. Min mouse model F. Anti-colorectal cancer effect of sulindac 22 III. JUSTIFICATION 25 IV. OBJECTIVES 26 iv V. MATERIALS AND METHODS A. Diets B. Experiment one 1. Animals and housing 2. Experimental design 3. Determination of aberrant crypts 4. Statistical analysis C. Experiment two 1. Min mouse breeding 2. Min mutation genotyping 3. Experimental design 4. Tumor analysis 5. Statistical analysis VI. RESULTS AND DISCUSSIONS VII. SUMMARY AND CONCLUSIONS VIII.RECOMMENDATION AND FUTURE STUDIES IX. REFEREENCES 27 27 3O 32 37 62 64 65 LIST OF TABLES Table 1. Compostion of diets (g/ 100g). Table 2. Composition of the 30% soy product (added to diets at 81.4 g/ 1 00g diet) Table 3. Effect of diet and sulindac on adenoma distribution in the small intestine of Min mice. Table 4. Effects of diet and sulindac on tumor incidence, number of tumors/mouse, volume of tumors (mm3)/mouse in the cecum and large intestine of Min mice. vi Page 28 29 53 55 LIST OF FIGURES Page . Effect of diet, sulindac, and AOM-inj ection on body weight of rats. 42 . Effect of diet and sulindac on total ACF number in AOM-inj ected rats. 43 . Effect of diet and sulindac on average size of ACF inAOM-injected rats. 44 . Effect of diet and sulindac on number of large ACF (AC>3) in AOM-inj ected rats. 45 . Effect of diet and sulindac on body weight of male and female Min mice. 49 . Effect of diet and sulindac on body weight of male Min mice. 50 . Effect of diet and sulindac on body weight of female Min mice. 51 . Effect of diet and sulindac on small intestinal adenoma number in Min mice. 52 . Effect of diet and sulindac on small intestinal adenoma size in Min mice. 54 vii I. INTRODUCTION Colorectal cancer claims nearly 50,000 lives every year in the United States (Parker et al., 1996; AICR/WCRF,1997). Researchers have revealed that colorectal cancer development is a multi-step process, which typically develops over decades and appears to require at least seven genetic events for completion of a malignant tumor (Fearon et al., 1990; Kinzler et al., 1996). The multi-step nature of colorectal cancer makes it possible - through administering natural or synthetic substances - to interrupt the carcinogenesis process, and ultimately reduce the age-specific colorectal cancer incidence. This strategy of cancer prevention is referred as chemoprevention (Lippman et a1, 1994; Owen et al., 1998) Significant evidence has demonstrated that dietary factors strongly modify colorectal cancer incidence (AICR/WCRF,1997). There is increasing interest in identifying dietary ingredients for colorectal cancer chemoprevention. Epidemiological evidence has suggested that consumption of soybean products contributes to the lower cancer incidences observed in some Asian countries (ACS, 1994). Many compounds in soy products - such as the isoflavone genistein, the Bowman-Birk protease inhibitor, phytic acid and saponins - have been proposed to suppress carcinogenesis (Messina et al., 1994; Kennedy et al., 1995). Amounting evidence has also suggested that sulindac, a non-steroid anti-inflammatory drug, protects against colorectal cancer (Giardiello et al., 1993; Taketo et al., 1998). Many studies have been conducted to evaluate the protective effects of some specific components in soybeans against cancer, but only relatively little research has focused on 1 the potential of soybean protein to prevent colorectal cancer (Reddy et al., 1976; Clinton et al., 1979; Thiagarajan et al., 1998). Until now, the soybean protective effect against cancer and which compounds may contribute to any protection offered by soy are still unclear. Few studies have been conducted to evaluate the influence of processing on the cancer preventive effects of soy products. This is important, because processing of soybeans is necessary to produce food for human consumption. In Western society, extrusion processing is used to incorporate soy products into foods such as breakfast cereals or certain snack items to increase soy consumption. The current study was conducted to investigate the chemopreventive effects of different soy proteins and sulindac on colorectal cancer. The first objective was to test the protective effects of soybean products on colorectal cancer in two animal models Protective effects in the initiation stage were tested with the AOM-induced colon carcinogenic rat model and were tested in the promotion stage with the Min mouse model. The second objective was to test the influence of soybean processing on colorectal carcinogenesis. The third objective was to determine the potential of sulindac to protect against colon carcinogenesis, as sulindac was also utilized as a positive control for studies on the potential of soy products to reduce colon carcinogenesis. II. REVIEW OF LITERATURE A. Epidemiological Studies 1. Colorectal Cancer Incidence and Distribution With the improvement of living standards and medicine during the past hundred years, there has been a dramatic decline of deaths caused by infectious disease and life expectancy has increased in most parts of the world. Concurrent with the reduction in deaths due to infectious diseases, some aging related chronic diseases such as cancer and cardiovascular disease have emerged as major causes of death. Colorectal cancer is the fourth most common cancer in the world. In 1996, an estimated 875,000 new cases were diagnosed worldwide. In the United States, colorectal cancer is the second most common cause of cancer mortality. In 1997, there was an estimated 94,100 new cases and 46,600 deaths (Parker et al., 1996; AICR/WCRF, 1997). Colorectal cancer incidence varies about twenty-fold across countries of the world. The highest rates of incidence are found in the developed world — North America, Western Europe and Australia - with age adjusted incidence rates of 25-35 cases per 100,000 population. In contrast, much lower incidence occurs in countries of Asia, Afiica and southern Europe (ACS, 1992). 2. Overview about Diet and Colorectal Cancer Cancer as a disease is multi-factorial in its causality. Many factors such as hereditary defects, diet, smoking, and exposure to environmental carcinogenic chemicals may all contribute to the generation of cancers. In 1969, Burkitt first suggested that a low intake of dietary fiber might be involved in the causation of colorectal cancer (Burkitt, 1969). Since 3 that time, abundant epidemiological evidence has revealed that dietary factors strongly influence colorectal cancer incidence. Higher risk for developing colon cancer is generally related with higher meat and fat consumptions (Stemmerrnan et al. 1984; Enstrom, 1975), whereas lower risk is associated with higher consumptions of vegetables and fruits (Howe et a1. 1992). Dietary components may be one of the most important factors contributing to cross-cultural differences in colorectal cancer incidence. The American Institute for Cancer Research estimated that about 66% to 75% colorectal cancers may be prevented by adequate diets (AICR/WCRF, 1997). B. Genetic Study of Colorectal Cancer 1. Multi-Step Nature of Colorectal Tumorigenesis Studies on experimental carcinogenesis and clinical data indicate that colorectal cancer development is a multi-step process consisting of at least three different stages - initiation, promotion and progression (F oulds, 1958; Weiberg, 1989). Initiation refers to alterations at the genomic level caused by hereditary defects or carcinogenic factors, which confers the cell with some selective grth advantage. Promotion involves clonal expansion of the initiated cells, which is generally associated with altered phenotypic changes and accumulation of genetic mutations. Progression involves further genotypic and phenotypic changes associated with malignancy and metastasis. Each of the steps involves multiple events. 2. Genetic Basis of Colorectal Cancer Some of the molecular events involved in the process of colorectal carcinogenesis have been identified in recent years. Fearon and Vogelstein have proposed a genetic model for colorectal tumorigenesis (Fearon et al., 1990). Tumorigenesis proceeds through 4 a series of genetic alterations involving mutation of proto-oncogenes such as the ras gene, and inactivation of tumor suppressor genes such as those on chromosomes 5q, 17p and 18q (the APC, p53, and DCC genes, respectively). Adenomatous polyposis coli (APC) gene mutations on 5q chromosome are thought mainly responsible for epithelial cell hyperproliferation. An inherited mutation of the APC gene results in the formation of numerous colorectal polyps and ultimately colon cancer in PAP (familial adenomatous polyposis) patients. APC gene mutations also are found in colon tumors arising in patients without polyposis. It has been demonstrated that the APC gene is frequently lost or mutated at a relatively early stage of tumorigenesis. Ras mutations (usually K-ras) on chromosome 12p appear to occur in cells of a small adenoma, which then becomes a larger and more dysplastic tumor through clonal expansion. Loss of the DCC gene (Deleted in Colon Cancer) on 18q, and the p53 gene on 17p usually occur at later stages of colorectal tumorigenesis. There are other unidentified gene mutations also involved in the process. Each of the mutations confers the affected cell with some growth advantage, allowing it to outgrow other neoplastic cells within the tumor and to become the predominant cell type constituting the neoplasm (clonal expansion). A large body of evidence supports the idea that accumulated genetic changes underlie the development of colorectal cancer, which typically develops over decades and appears to require at least seven genetic events for completion of a malignant tumor (Kinzler et al., 1996). Fewer mutations suffice for benign tumorigenesis. The total accumulation of mutations rather than their order with respect to one another is more important in determining the biological properties of the tumors (F earon et al., 1990; Kinzler et al., 1996). C. Colorectal Cancer Chemoprevention Most colorectal cancers develop initially from hyperproliferative lesions, progress through different stages of adenomas, and ultimately, to large metastatic carcinomas (Fearon et al., 1990). In humans, the process of development from a benign to a malignant colorectal tumor averages from 10 to 15 years (Marshall, 1992). The multi-step nature of colorectal cancer makes early detection difficult. Current treatments like surgery, chemotherapy and radiation may be ineffective on the later stage tumors. On the other hand, the multi-step carcinogenic process makes it possible - through administering natural or synthetic substances - to interrupt the carcinogenic process, retard cancer development, and ultimately reduce the age-specific colorectal cancer incidence. This strategy of cancer prevention is referred as chemoprevention (Lippman et al, 1994; Owen et al., 1998; Mettlin et al., 1997). Considering that nearly all colon cancers are believed to originate from previously benign adenomas (Sugarbaker et a1. 1985), and that at least 50% of the persons in Western populations may develop benign colorectal tumors by the age of 70 (Parker et al., 1996), effective chemoprevention should greatly reduce age- related colorectal cancer mortality. Ideal chemopreventive agents should be potent, inexpensive, widely applicable, and free of serious side effects. Given the epidemiological evidence relating diet to colorectal cancer incidence, there has been increasing interest in finding natural dietary components to be used as chemopreventive agents in recent years. D. Soy Consumption and Colorectal Cancer 1. Epidemiological Evidence In the United States, the incidence and mortality of colorectal, breast and prostate cancers are higher than that observed in some Asian countries such as Japan and China (ACS, 1994). Japanese migrating to the US develop the higher risk rates for colon cancer more similar to. American whites rather than their counterparts in Japan. Adoption of the diets in their new country was thought to play an important role in the development of colorectal cancer (Haenszel et a1. 1968). Although Western and Eastern diets vary widely in many aspects such as fat and fiber content, one striking distinction between them is the major source of protein. In Eastern diets, soybeans are an important protein source. In general, Japanese and Chinese pe0ple consume 20-80 grams of soy foods per day, whereas Americans consume only 1-3 grams daily (Barnes et a1. 1995). As people in these Asian countries consume more soybean products than Americans, consuming large amounts of soybean products has been suggested to be associated with overall low cancer incidence and mortality rates, particularly for cancers of the colon, breast and prostate (Messina et al., 1994; Kennedy et al., 1995). Several epidemiological studies related soy consumption with colorectal cancer incidence. In a case-control study in China, Hu et a1. (1991) found a protective effect of soybean products against rectal cancer risk in man. Another study in Japan also found that consumption of beans and bean curd was associated with lower rectal cancer risk (Watanabe et al., 1984). However in a cross-cultural study of 38 countries, no association between soybean intake and colon cancer risk was found (McKeown-Eyssen et al., 1984). 2. Influence of Processing Process on Soybean Products In some Asian countries such as China and Japan, soybeans are usually used for making soymilk, tofu and fermented soy foods such as miso, soybean paste. However, more emphases are focused on isolation of protein and oil fractions of soybeans in the Western countries. Soybean oil is prepared by extracting crushed soybeans with hexane. The defatted soybeans are pulverized to form defatted soy flour. Soybeans can also be ground directly to make full-fat soy flour or full-fat soy flakes. Soy flour comes in several grades based on the extent to which it is heat-treated. Heating is used to inactivate phytohemagglutinins and protease inhibitors such as trypsin inhibitor and Bowman-Birk inhibitor in order to reduce the growth-depressing effect of the unheated soy products. Defatted soy flour is further processed to generate products with higher protein content such as soy concentrate and isolated soy protein. Aqueous washing of soy flour is commonly used to remove the soluble carbohydrate and increase the percentage of protein. Sometimes, in order to get taste-free and color-free soy concentrate, soy flour is washed with a mixture of water and alcohol. Soybean processing significantly influences the composition of soy products. Heating inactivates the protease inhibitors. A number of soy-containing foods were found to have no more than 1% of the quantity of protease inhibitors present in raw soy flours (Anderson et al., 1995). Soybeans contain high concentrations of isoflavones (average from 1200 to 4200 ug/g). However, the concentration of isoflavones in soy isolate is reduced (average from 600 to 1000 ug/ g), and a soy concentrate made by alcohol extraction is virtually devoid of isoflavones (Wang et al. 1994). Soybeans typically contain 0.1-0.5 % saponins. After alcohol extraction almost no saponins were detected in soy concentrate (Anderson et al., 1995). Recently, soy extrusion commonly has been used to incorporate soy products into breakfast cereals and some snack items. However, extrusion conditions such as high temperature and pressure may also significantly change the composition of soybean products. For example, the protease inhibitor will be largely inactivated. As recent studies have suggested that these minor components in soybeans may have beneficial biological effects such as lowering blood cholesterol or preventing cancers, further study of the influence of soybean processing on its potential anti-cancer effect is necessary. 3. Animal Studies with Soy Products Animal studies related to soy consumption and colorectal cancer incidence are inconsistent. Reddy et a1. (1976) and Clinton et a1. (1979) compared soy protein with beef protein in 1,2-dimethylhydrazine (DMH)-induced colonic carcinogenesis in rats, and found no protective effect of soy protein on the colon tumor incidence and multiplicity. In these two studies, the soy protein sources were not specified. Thiagarajan et a1. (1998) found that feeding defatted soy flour and fiill-fat soy flakes significantly reduced the number of colonic aberrant crypt foci (ACF) in azoxymethane (AOM)-treated rats. Conversely, feeding ethanol-washed soy concentrate, which is virtually devoid of genistein, did not influence ACF number (Thiagarajan et al., 1998). 4. Proposed Anti-Carcinogenic Compounds in Soybeans Numerous studies have been conducted to examine the effects of specific components in soybeans on colon cancer risk. Several compounds in soy products have 9 been shown to suppress carcinogenesis in vitro and in vivo. These compounds include the isoflavone genistein, the Bowman-Birk protease inhibitor, inositol hexaphosphate (phytic acid), saponins, and phytosterols such as B-sitosterol (Messina et al., 1994; Kennedy et al., 1995). Isoflavones are a group of natural heterocyclic phenols which are abundant in soybeans. Genistin one of the major isoflavones found in soybeans, is the B-glucoside form of genistein (4’, 5, 7- trihydroxyisoflavone) (Kudou et al., 1991). Soybeans are the major dietary source of genistein for humans. On the basis of average annual consumption of soybeans, daily intakes of genistein by the Japanese are estimated to be 1.5-4.4 mg/person (F ukutake et al., 1996). American diets, which usually do not include appreciable amounts of soy products, are almost completely lacking in genistein. Genistein has been hypothesized to be the most potent compound in soy products to reduce tumor incidence (Messina et al., 1994). Most of the evidence for cancer prevention by genistein comes from in vitro experiments. Genistein can suppress the growth of a wide range of cancer cells in vitro (Akiyama et al., 1991). Genistein also can induce differentiation and apoptosis in many cancer cell lines (Constantinou et al., 1995; Spinozzi et al., 1994). While precise mechanisms are still unknown, genistein is known to be a potent inhibitor of protein tyrosine kinases. Enzymes in this kinase family phosphorylate the tyrosine residues on key proteins involved in signal transduction events in normal and tumor cells (Akiyama et al., 1987; Makishima et al., 1991). Genistein also inhibits DNA topoisomerases I and II (Okura et al., 1988; Markovitz et al., 1989), which are involved in DNA replication. Other anticarcinogenic properties of genistein include suppression of 10 angiogenesis (F otsis et al., 1993) and scavenging of exogenous hydrogen peroxide free radicals (Wei et al., 1995). Most cell culture studies indicate that the inhibitory effect of genistein on tumor cell grth can only be achieved at relatively high concentrations. The IC50 (concentration required to cause a 50% inhibition of cell growth) is typically greater than 13.2 umol genistein/L (Barnes et al., 1995). By comparison, the plasma level of genistein in persons consuming soy-rich diets is only 1-4 umol/L (Adlerceutz et al., 1993). This discrepancy suggests that genistein may not be protective in vivo because plasma concentrations are too low. Genistein shares structural similarity with naturally occurring estrogens and can bind to estrogen receptors, although with weaker activity compared with 17 B-estradiol (Martin et al., 1978). Because of its phytoestrogen activity, genistein may influence the incidence of some honnone-dependent breast and prostate cancers (Setchell et al., 1984; 1998). Although colon cancer occurs with approximately equal frequency in men and women, the subsites of colon cancer vary by sex. There is a female excess of right—sided colon cancers and a male excess of left-sided cancers (McMichael et al., 1980). Further, it is suggested that differences in gut metabolism, colonic bacterial populations and fermentation rates exist between the sexes. These differences may be mediated ultimately by hormones (Potter 1995). Studies also have shown that in the 19703 there was a modest and brief decline in colon cancer incidence among women compared with men in countries in which use of the original high-dose oral contraceptives had been common. It has been suggested that oral contraceptive use was associated with lower risk for colon cancer in women 11 (McMichael et al., 1980). Thus, it is possible that exogenous hormone-like compounds such as genistein can also influence the normal function of the colon as well as colon carcinogenesis. Animal studies on the potential of genistein to influence colon cancer are not conclusive. Several experiments reported that genistein reduced ACF, putative precancerous lesions in ADM-induced colon carcinogenesis in rats (Pereira et al., 1994; Steele et al., 1995; Thiagarajan et al., 1998). Conversely, Rao et a1. (1997) reported that genistein (250 ppm) in the diet significantly increased noninvasive and total adenocarcinoma multiplicity in AOM-treated rats when compared with casein control diet. Bennink et a1. (1999) also found that rats fed genistin (500 ppm genistin added to a soy concentrate diet) had higher colon tumor incidence compared with rats fed the soy concentrate diet alone. Another study employed the Min mouse model and found that soy isoflavones (475 ppm in the diet; about 280 ppm genistein) had no effect on adenoma formation (Sarensen et al., 1998). Soybean is a plentiful source of protease inhibitors. About 6% of the protein of soybeans consists of protease inhibitors (Birk, 1993). Soy-derived protease inhibitors can suppress tumor grth both in vivo and in vitro (Kennedy, 1993). Among the protease inhibitors in soy, the Bowman-Birk protease inhibitor (BBI) has been the most extensively studied. BBI is a polypeptide with seven disulfide bridges and two homologous sites with inhibitory effects on trypsin and chymotrypsin (Birk, 1993). Several studies have shown BBI is a potent anticarcinogen. BBI suppressed DMH- induced colonic carcinogenesis (Weed et al., 1985; Billings et al., 1990) and reduced adenoma formation in the Min mouse model (Kennedy et al., 1996). The underlying 12 mechanisms by which BBI suppresses carcinogenesis are unknown. One of the contributing mechanisms may be the selective toxicity of BBI for pre-malignant and certain malignant cells. Kennedy and associates suggested that protease inhibitors suppress carcinogenesis via inhibition of expression of some specific oncogenes and proteases thought to be involved in the transformation of a cell from normal to malignant (Kennedy, 1993 and 1998). Yavelow et a1. (1983) reported that although dietary BBI reaches the colon in its active form, very little is taken up in the blood stream and delivered to other internal organs. Intakes of large dietary quantities of BBI may result in high levels in the colonic lumen and feces without a simultaneous increase in internal organ levels (Yavelow et al., 1983; Troll etal., 1984). Thus, BBI may be able to influence colon cancer development only via its actions in the colonic lumen - not through systemic action - because its absorption is minimal. Soybeans are also a major dietary source of saponins, containing about 0.5% saponins by weight (Anderson et al., 1995). Saponins are glycosides composed of a lipid- soluble aglycone (consisting of either a sterol or triterpenoid) linked to one or more water-soluble sugar residues (Koratkar et al., 1997; Rao and Snug, 1995). Hence, saponin molecules are arnpiphilic with the triterpene portion being hydrophobic and the sugar portion hydrophilic, giving them some characteristic surface activity (Oakenfull et al., 1990). Saponin concentrations ranging from 150-600 ppm have been shown to inhibit proliferation of human carcinoma cells (HCT 15) in vitro (Rao and Snug, 1995). Koratkar (1997) also reported that 3% soybean saponins in the diet significantly reduced AOM- induced preneoplastic lesions ACF on the colons of mice compared with control AIN-76 diet. There are several proposed mechanisms to explain the effect of saponins on colon 13 cancer risk. These proposed mechanisms include selective toxicity toward cancer cells, immune modulation, and regulation of cell proliferation (Rao and Snug, 1995). Inositol hexaphosphate (phytic acid), has demonstrated anticancer potential in both in vivo and in vitro experiments (Shamsuddin, 1995). Several mechanisms have been suggested. Since inositol-1,4,5 -triphosphate acts as a second messenger in cellular proliferation and differentiation pathways, it has been hypothesized that inositol hexaphosphate may exert its antineoplastic and antiproliferative effects by increasing intercellular levels of inositol triphosphate (Shamsuddin et al., 1995). Phytic acid can decrease metal pro-oxidant reactivity and reduce radical generation by binding to ions like iron and zinc (Graf et al., 1990). In doing so, it may also reduce colon cancer risk (Messina et al, 1991). Other compounds in soybean also have been shown to have-anticancer effects. Raicht et a1. (1980) reported that 0.2% dietary B-sitosterol reduced colon cancer in rats. At this time, it is unclear whether one of these compounds attributes to the potential of soy to reduce colon cancer risk, or the cumulative effects of these compounds attribute to the protective effect of soy. 5. Human Clinical Trials with Soybean Products Few human clinical trials with soy products have been conducted to evaluate their potential to protect against colon cancer risk. One double blind prospective study with forty-seven subjects found that consumption of 39 g per day of isolated soy protein for one year significantly reduced the proliferative capacity in colonic crypts compared with consumption of casein (Thiagarajan et al., 1999). 14 E. Models and Biomarkers for Chemoprevention Studies 1. Animal Models vs. Cell Culture For chemoprevention studies, one of the most important priorities is to choose the proper animal model and valid biomarkers to evaluate anti-cancer effects. Although in vitro studies with transformed cancer cell lines can help to elucidate the anticancer mechanisms of specific compounds in the diet, such studies do not offer conclusive evidence that a compound may be anticarcinogenic in vivo. There are several reasons. First of all, the cancer cells cultured in vitro likely behave very differently compared to normal cells or precancerous cells in vivo. Secondly, the concentrations of specific compounds used in cell culture studies are usually much higher than what is physiologically achievable through dietary consumption. Thirdly, it is virtually impossible to study the metabolism or interactions between different diet ingredients in cell culture studies. Based on these reasons, animal models are usually employed to study the potential of diet to prevent colorectal cancer. Despite the fact that animal models have numerous advantages over in vitro models, it still is difficult to extrapolate the results from animals to humans because of numerous species differences. 2. Carcinogen-Induced Colorectal Carcinogenesis Model Early studies of human colorectal cancer have been hampered by the lack of experimental animal models. In 1963, Laqueur fed rats with a diet containing cycasin, one of the major toxic glycosides derived from the cycad nut, and demonstrated that intestinal tumors could be induced (Laqueur et al., 1963,1965 and 1968). Since then many related compounds have been synthesized. One such compound - 1,2- dimethylhydrazine (DMH) - has been shown to be extremely selective in the production 15 of colon cancer (Druckrey, 1967 and 1968; Martin et al., 1973). The mechanism of induction of colon cancer by DMH is not totally understood. It is postulated that after DMH is taken into the body, it is converted into azomethane, then azoxymethane (AOM), and finally into methyl-azoxy-methanol (MAM) in the liver by some detoxification enzymes. MAM is an activated methylating agent which can cause the methylation of guanine residues in both RNA and DNA. Guanine methylation results in mutations in the DNA sequence during replication. In addition, DMH is cytotoxic. One to three days after administration, hyperproliferation occurs in the colonic epithelium to replace the dead cells. Induction of these changes in the colon leads to genetic and epigenetic alterations, and eventually the formation of tumors. Early studies used colonic tumors as the endpoint for protective evaluation. Because of the relatively low colonic tumor production, a large number of animals were needed to detect diet effects. In addition, these studies usually employed relatively large doses of carcinogen that were administered repeatedly over a long period of time (Thumherr et al., 1973; Deschner, 1977). Carcinogen-induced colon cancer studies have several disadvantages. Administering large doses of carcinogen has serious side effects on the animals. Repeated administrations of carcinogens also bring some risk to the researchers. In addition, maintaining large numbers of animals for a long period of time is extremely expensive. For these reasons, researchers worked to identify biomarkers other than colonic tumors to evaluate the protective effect of different dietary factors. Deschner described colonic mucosal proliferation patterns in DMH-treated mice (Deschner et al., 1974). Isolated crypts of DMH-treated mice had higher labeling indices 16 and showed an upward shifi in the major zone of DNA synthesis. Compared with untreated controls, this hyperproliferation pattern in the colon mucosa has been correlated with colonic carcinogenesis. Some biomarkers of ‘hyperproliferation have been used to predict the risk for colorectal cancer in humans (Lipkin et al., 1974; 1985). However, the concept that increased levels of cell proliferation in a tissue are predictive of tumor formation has been challenged recently. Farber ( 1995) argued that cell proliferation per se has not been directly associated with tumorigenesis. For example, tumors in the small intestine are extremely rare despite the higher rates of proliferation in the small intestine compared to the colon (Farber, 1995). Other factors such as DNA damage repairing capability and apoptosis have also been suggested as being involved in the colorectal carcinogenesis (Bedi et al., 1995; Garewall et al., 1996). Chang et a1. (1997) conducted an experiment to determine the prognostic significance of proliferation, differentiation and apoptosis as intermediate markers for colon tumor development in AOM-induced rat colon cancer model. They found measurements of differentiation and apoptosis had greater prognostic value to detect dietary effects on tumor incidence than did measurements of cell proliferation (Chang et al., 1997). Given the complexity and inconsistency to measure these biomarkers for predicting colorectal cancer risk, researchers are still searching for more valid early predictive biomarkers. 3. ACF as Early Precancerous Lesions In 1987, Bird and associates identified aberrant crypts (AC) in the colon mucosa of rodents treated with AOM. After staining fixed colonic tissue with methylene blue, AC are distinguished fi'om normal crypts by the bluer color, enlarged size, thickened epithelial lining and increased pericryptal zone (Bird, 1987, 1995). It was hypothesized that AC 17 represented preneoplastic lesions. ACF are numerous in carcinogen-treated rodent colons, but are rarely detected in rodents that are not treated with carcinogens. ACF are also found in the colonic mucosa of human patients with colorectal tumors (Roncucci et a1. 1991; Pretlow et al., 1991). Some ACF induced by DMH in rat colon epithelium exhibit dysplasia (McLellan et al., 1991). Epithelial cells in aberrant crypts often carry K-ras mutations. Losi et a1. (1996) found that 57 percent of ACF examined from human colorectal cancer patients carry K-ras mutations. Jen et al. (1994) examined the gene alterations of ACF in patients with colorectal cancer, and found K-ras mutations were present in 19 of 20 ACF, and the only ACF without K-ras mutation carried on APC mutation. An aberrant crypt foci (ACF) may contain only a single aberrant crypt, or a cluster of aberrant crypts. The number of ACF and the multiplicity of ACF (AC number per ACF) are proposed to be correlated with colorectal carcinogenesis (Mclellan et al., 1991; Bird, 1995). Aberrant crypts can be induced within a few weeks after one or two AOM injections at relatively small doses, and have been shown to be sensitive biomarkers to detect different diet effects (Bird, 1995). Many studies have used ACF as biomarkers for colonic carcinogenesis, however, there has been little consistency in the protocols used and what kind of ACF parameters (such as the total ACF or large ACF) that are used as the endpoint for predicting colon cancer risk. Different strain of rodents have different sensitivities to the carcinogen, and the carcinogen doses, the number of injections and the injection time of the day will influence the number of ACF induced (Pereira et al., 1994). Some researchers have used the total number of ACF to assess anticancer effect of diets without considering the ACF multiplicity (Pereira l8 et al., 1994; Cassand et al., 1997). However, other researchers have suggested that only the number of large ACF is correlated with the final tumor incidence (Pretlow et al., 1992; Magnuson et al., 1993; Caderni et al., 1995). Different definitions of large ACF have been used in different studies and more researchers have defined the ACF with 4 or more than 4 aberrant crypts as the large ACF (Pretlow et al., 1992; Magnuson et al., 1993; Davies et al., 1999). Although many studies have used ACF as precursor lesions of colorectal cancer, there are still some discrepancies and questions. Hardman et a1. (1991) found that the number of ACF and the incidence of adenocarcinomas were significantly different in rats treated with DMH. All attempts to show a significant correlation between the mean number of ACF per rat and the incidence of adenocarcinomas failed (Hardman et al., 1991). As questions exist about using ACF to predict final colon cancer risk, the use of other more reliable animal models may be helpful to predict dietary chemopreventive effects. 4. Min Mouse Model Recently, the Min (multiple intestinal neoplasia) mouse model has been used extensively for colorectal cancer research. Min is an ethylnitrosourea (ENU)-induced gemline mutation in the murine APC (adenomatous polyposis coli) gene (Moser et al., 1990; Su et al., 1992), which contains a mutation (Leu to stop) at codon 850 resulting in premature truncation of the polypeptide. This is similar to the germline mutation of APC gene in humans with F AP, which predisposes the patient to develop hundreds of colonic adenomas (Joslyn et al., 1991; Nishisho et al., 1991). The reason why APC mutations primarily initiate tumori genesis in the intestine and not in other organs is unclear. APC gene mutations are thought to be mainly responsible 19 for the epithelial hyperproliferation and numerous colorectal polyps observed in FAP patients. Kinzler and Vogelstein (1996) have proposed a “gatekeeper” function for APC in human colorectal tissue in which APC is believed to be responsible for maintaining a constant census in renewing cell populations. A mutation of the “gatekeeper” gene leads to a permanent imbalance of cell division over cell death, and leads to colorectal tumorigenesis (Kinzler, 1996). Studies have indicated that APC gene mutation is an early event in colorectal cancer development (Fearon et al., 1990; Powell et al., 1992). Although all cells in the intestine of a Min mouse carry a mutant copy of Apc, only a relatively small number of tumors form. This indicates that further somatic events are required before tumors can develop. It has been suggested that the somatic mutation of the wild-type APC allele inherited from the unaffected parent is the rate-limiting step in the tumor initiation (Kinzler 1996). Loss or mutation of the second (wild-type) allele of APC has been detected in the earliest neopleastic lesions as well as the majority of tumors (Jen et al., 1994; Moser et al., 1992; Nagase et al., 1993). Inactivation of both alleles of the murine APC gene occurs very early in mouse colon tumor development (Levy, 1994). The mechanisms underlying APC gatekeeper function are not totally clear. The APC protein is apparently located at the basolateral membrane in colorectal epithelial cells, with expression more pronounced as cells migrate up through the crypt column (Smith et al., 1993). Expression of wild-type APC in colorectal epithelial cells with APC mutations results in apoptosis, suggesting that APC is involved in controlling of the cell death process (Morin et al., 1996). Loss of such a “death signal” could alter the precise homeostatic balance required in renewing cell populations. 20 The APC gene encodes a protein with 2843 amino acids which contains several potentially important functional domains. The middle third of APC protein contains two classes of B-catenin binding repeats. When APC protein binds B-catenin, it promotes the degradation of B-catenin by a ubiquitin/proteosome mediated process. APC mutation generally results in the loss of at least one of the B-catenin binding sites. Consequently, the cytoplasmic B-catenin concentration increases due to reduced B-catenin degradation. This promotes the association of B-catenin with some transcription factors such as T-cell factor (ch) and lymphoid enhancer factor (Lef) (Behrens et al., 1996). Complexes of B-catenin with ch/Lef will increase the transcription of some target genes in the nucleus. These genes are largely unknown, although a recent study identified c-MYC as a possible target (Behrens et al., 1998; Stappenbeck et al., 1998). The B-catenin association also links APC to cellular adhesion. B-catenin is necessary for E-cadherin-mediated cell adhesion between epithelial cells. Because the binding of B-catenin to E-cadherin and APC is mutually exclusive, APC could modulate such adhesion by regulating B-catenin concentration in the cytosol (Su et al., 1993; Hulsken et al., 1994). The carboxyl terminal region of the APC protein involves interactions with microtubules and two other proteins. One is EB-l , a highly -conserved 30 KDa protein of unknown function (Su et al., 1995). Another is the human homologue of the Drosophila tumor suppressor gene discs large (DLG) (Matsumine et al., 1996). As mutated APC protein loses its carboxyl terminus, it also loses these interactions with DLG and EB 1, which may be important for APC’s growth-controlling function. 21 For unknown reasons, most tumors which develop in Min mice are located in the small intestine, with relatively few tumors in the cecum and colon. This is in contrast to human FAP patients, where the colon is the primary site of tumor development. Min mice may also develop tumors in other tissues including desmoid tumors, epidermoid cysts and mammary tumors. Although APC mutations first were identified in F AP patients, somatic APC mutations are also found in the great majority of sporadic colorectal tumors (Nishisho et al., 1991; Powell et al., 1992). It has been suggested that nearly all colon cancers originate from benign adenomas (Sugarbaker et a1. 1985). Hence, the adenomas which develop in Min mice can be used as a valid endpoint to predict the final cancer risk. Since the Min mouse model closely mimics the mechanism of APC gene inactivation in F AP and most sporadic human colon cancers, it should be an ideal animal model for colorectal cancer chemoprevention studies. F. Anti-Colorectal Cancer Effect of Sulindac Recently, several epidemiological studies suggested that non-steroidal antiflammatory drugs (N SAIDs) such as aspirin, indomethacin, piroxicarn and sulindac may lower the risk of colorectal cancer (Thun et al., 1991; Giovanucci et al., 1994). Sulindac has been shown to cause regression of colonic polyps in the treatment of patients with FAP (Giardiello et al., 1993; Taketo et al., 1998). In cell culture studies the sulindac metabolites, sulindac sulfide and sulindac sulfone, inhibited colon cancer cell proliferation, blocked the cell cycle at G1 phase, and induced apoptosis in some colon cancer cell lines (Shiff et al., 1995; Goldberg et al., 1996). 22 The underlying mechanism whereby sulindac causes polyp regression is still unknown. Evidence for involvement of cyclooxygenases (COX) in cancer was suggested from the findings that various animal and human tumor tissues, including human colon cancer, had increased concentrations of prostaglandins (Kargman et al., 1995). Cyclooxygenase is one of the key enzymes in arachidonate metabolism and catalyzes the conversion of arachidonic acid to prostaglandin H2, the precursor for the prostaglandins and thromboxanes (Smith et al., 1996). There are two isoforms of COX: COX-1 and COX-2. COX-1 is constitutively expressed in most tissues and is involved in cellular homeostasis, synthesizing prostaglandins. COX-2 is inducible in many inflammatory reactions, and inflammatory cells usually have more abundant COX-2 activity (Smith et al., 1996). The levels of COX-2 in colorectal adenomas and carcinomas are significantly increased compared with that found in normal colonic mucusa (Eberhart et al., 1994). Additional evidence implicating COX-2 in carcinogenesis is the finding that a null mutation for COX-2 markedly reduced the number and size of intestinal tumors in APCA716 knockout mice, another murine model of F AP (Oshima et a1. 1996). Selective inhibitors of COX-2 cause a reduction in the number and size of polyps similar to that caused by the COX-2 gene knockout mutations (Oshima et al., 1996). The precise mechanism of how APC interacts with COX-2 is not clear. Recently, a study reported that introduction of full-length APC in HT-29 colon cancer cells down-regulates COX-2 expression at the translational level (Hsi et al., 1999). Inhibition of COX-2 activity may be one of the major mechanisms whereby sulindac and other NSAIDs cause polyp regression. However, some studies demonstrated that 23 NSAIDs inhibit the proliferation of colon cancer cell lines independent of their ability to inhibit prostaglandin synthesis (Hanif et al., 1996). Chiu et al. (1997) reported that administration of sulindac to Min mice reduced the tumor number by 95% but did not alter the levels of prostaglandin E2 (PGE2) and leukotriene B4 (LTB4) in intestinal tissues. Dietary arachidonic acid supplementation increased PGE2 and LTB4 levels by 44% but had no effect on tumor number or size. When sulindac was added to the arachidonate-supplementation diet, tumor number was reduced by 82%, whereas eicosanoid levels remained elevated (Chiu et al., 1997). Sulindac is a prodrug that is metabolized to a pharmacologically active sulfide derivative that potently inhibits prostaglandin synthesis. However, another derivative of sulindac which essentially lacks COX inhibitory effects, sulidac sulfone, was also able to inhibit HT—29 colon cancer cell growth by inducing apoptosis (Piazza et al., 1995). 24 III. JUSTIFICATION Several lines of evidence have suggested that soybean consumption reduces colorectal cancer risk. A number of compounds in soybeans have been proposed to explain its anticarcinogic effect. However, it is unclear whether one or a few specific compounds in soy are responsible for the protective effect, or if the cumulative effects of a large number of compounds in soy contributes to its protective potential. Studies on the potential of whole soy products to reduce colon cancer risk are warranted. Few studies have been conducted to evaluate the influence of soybean processing on the cancer preventive effect of soy products. In Western societies, extrusion processing is used to incorporate soy products into foods such as breakfast cereals or certain snack items to increase soy product consumption. Processing changes the chemical composition of soy. Studying the influence of processing on the biological effects of soy is important for the food industry and may help to elucidate the mechanisms whereby soybeans and their constituents may act as anticarcinogens. Little research has been conducted to evaluate the potential of soybean products to prevent the grth of intestinal tumors in Min mice. Kennedy et al. (1996) reported that supplementing the diet of Min mice with 0.5% Bowman-Birk Inhibitor Concentrate (BBIC) resulted in a 41% reduction in colonic tumori genesis and a similar reduction in small intestinal tumors. Sorensen et al.(1998) used the Min mouse model and found that soy isoflavones did not influence intestinal tumor numbers. Further research on the potential of soybean products to reduce tumor incidence in Min mice is warranted. 25 IV. OBJECTIVES The first objective of this research was to test the potential of soybean products to protect against colorectal cancer development in the two animal models. Specifically, we tested the potential of soy-containing diets to protect against colon cancer initiation using the AOM-induced colon carcinogenesis rat model and to delay colon cancer promotion using the Min mouse model. The second objective was to test the influence of soy processing on the colorectal carcinogenesis using these animal models. The third objective was to determine the potential of sulindac to reduce colonic carcinogenesis using the same two animal models, as sulindac was also utilized as a positive control for studies on the potential of soy products to reduce colon carcinogenesis. 26 V. MATERIALS AND METHODS Two experiments were conducted to determine the effects of soybean products and sulindac on colon cancer initiation and promotion. A. Diets Five dietary treatments were used in the two experiments. All the diets were based on AIN-93G diets (Reeves et al., 1993), and were modified to contain 19% protein and 15% fat (from soy oil). The energy density was also increased about 10% compared to AIN- 93G diets and the increased energy percentage from fat and protein were at the cost of a proportionate reduction of carbohydrate. All five diets were adjusted to have the same energy density. Dietary concentrations of other essential nutrients such as minerals and vitamins were also elevated with the increased diet energy density relative to standard AIN-93G. The major distinction between the different diets was their protein source. The protein sources were: 1) casein, 2) soy concentrate, 3) an extruded 30% soy product, and 4) a non-extruded soy product identical to 3. The composition of the diets is listed in Table 1. The composition of the 30% soy product used in the diets is listed in Table 2. The extruded and non-extruded 30% soy products were provided by the Kellogg Company. The extrusion conditions were not provided in this manuscript because the process was proprietary. A fifth treatment group was fed the casein-containing diet and given sulindac (200 ppm) in the drinking water. This level of sulindac has been shown to be effective against tumor development in Min mice (Boolbol et al., 1996). 27 Table 1. Composition of diets (g/ 100g) Soy Extruded Non-extruded Ingredient Casein Concentrate Soy Product Soy Product Casein 22.1 - - - Soy Concentrate - 27.3 - - 30% Soy Product" - - 81.4 81.4 Soybean Oil 15 14.8 9.4 9.4 Cellulose 5 - - - AIN-93G-MX 3.9 3.9 3.9 3.9 AIN-93-VX 1.1 1.1 1.1 1.1 L-Cystine 0.3 0.3 0.3 0.3 Methionine - 0.2 0.2 0.2 Choline Bitartrate 0.3 0.3 0.3 0.3 Cornstarch 31.8 31.6 2.5 2.5 Dextrinized Cornstarch 10.5 10.5 0.8 0.8 Sucrose 10 10 - - " Composition of the 30% soy product is presented in Table 2. 28 Table 2. Composition of the 30% soy product (added to diets at 81.4 g/100g diet) Ingredient g/ 100g total diet Whole Wheat Flour 30.5 Oat Flour 16.4 Corn Flour 3.9 Salt 0.8 Sodium Bicarbonate 0.4 Citric Acid 0.1 Full fat Soy Flour 24.4 Soy Protein Isolate 3.7 Soy Lifea 1.2 ' Soy Life: soybean hypocotyl, rich in isoflavones. 29 B. Experiment One Experiment one was conducted to test the effects of soy and sulindac on the initiation and early pomotion stages of colon cancer in azoxymethane (AOM)-induced colon carcinogenesis in rats. 1. Animals and Housing Four-week-old male Fisher 344 rats were obtained from Harlan Sprague-Dawley (Indianapolis, IN), and randomly divided into the different treatment groups. Each of the five diet treatment groups had 15 rats. A sixth group with 9 rats was given the casein diet and was injected with saline to serve as a non-AOM-injected control. Rats were ear- notched for individual identification and were housed with three per cage. All rats were housed in a room with constant temperature and humidity (21-24°C, 60% humidity) and a 12: 12 hour light-dark cycle. Rats had free access to the diets and water. Health condition, food and water supply were checked daily and rat weights were measured once a week. This study was approved by the Michigan State University All-University Committee on Animal Use and Care. 2. Experimental Design After 1 week of dietary treatment, AOM was injected subcutaneously (15 mg/kg body weight) in the morning between 8:00 am to 10:00 am. The same dose of AOM was given again one week later. The sixth group (9 rats) was injected with the same volume of saline. Rats were kept on the dietary treatments throughout the experiment. Six weeks after the second AOM injection, rats were euthanized by C02 inhalation. The colons were removed, opened longitudinally, flushed with water, pinned on cardboard, and fixed with 10% neutral buffered formalin overnight. All the tissues were stored in 1% neutral buffered 30 formalin solution and the number of aberrant crypt foci were determined. The experimental design is shown schematically as follows: Fisher 344 rats AOM injection 4 weeks old 15mg/kg BW Sacrifice l l l ,1 0 l 2 3 4 5 6 7 8 < Weeks on dietary treatment b 3. Determination of Aberrant Crypts The fixed colon tissues were stained with 0.2% methylene blue in phosphorate buffered solution (PBS) for 3-5 minutes. The aberrant crypts (AC) were determined with the aid of stereo microscope at a magnification of about 40 times. An AC is distinguished from normal crypts by the bluer color, enlarged size, thick epithlial lining and increased pericryptal zone. ACF (aberrant crypt foci) may have one or more than one aberrant crypts. ACF consisting of more than one crypt were characterized as follows: there are no norrnal-appearing crypts separating the crypts within the ACF, and the total area occupied by the aberrant crypts composing the ACF is greater than the area occupied by an equivalent number of surrounding morphologically normal crypts (Bird, 1987, 1995; McLellan et al. 1991 ). The number of ACF observed in each colon and the number of AC in each focus were recorded. 4. Statistical Analysis Rat weight data were analyzed by repeated measures analysis of variance. The total ACF number, ACF mutiplicity (AC/ACF), and the number of ACF which had more than 3 AC were evaluated by one-way analysis of variance. When significant differences were 31 detected (P < 0.05), treatment means were compared using the Least Significant Difference method. C. Experiment Two Experiment two was conducted to test the effect of soy products and sulindac on the promotion stage of colon cancer in Min mice. 1. Min Mouse Breeding C57BL/6J male Min mice (Min/+) were mated with normal C57BL/6J female mice (+/+) (obtained from the Jackson Laboratory) in a breeding colony maintained in a facility operated by MSU University Laboratory Animal Research. Mice were housed in a room with constant temperature and humidity (21-24 °C, 60% humidity) and a 12:12 hour light- dark cycle. Commercial rodent diet (Teklad 8640) was given to the mice in the breeding colony. All the male breeding Min mice were given sulindac (200 ppm in drinking water) when separated from female mice after mating, and female mice were given tap water. The health condition was checked every day. 2. Min Mutation Genotyping Progeny were genotyped to identify if they were heterozygous for the Min allele at 3- 4 weeks of age as described below. 1) Blood collection: At 3-4 weeks of age, 30-50 ul of blood was obtained from each mouse by dorsal pedal vein bleeding with a heparinized microcapillary tube. Blood was expelled into a 1.5 ml microfuge tube containing 10 pl of 10 mM EDTA and mixed to prevent clot formation immediately. All the samples were stored on ice until processing. 32 2) DNA extraction for PCR (Polymerase Chain Reaction): The blood drawn from the mice was used for DNA extraction. The procedures were as follows: A) Centrifuge samples 5 minutes at 1000x g to pellet cells. B) Remove and discard supernatant. C) Add 200 pl Lysis Buffer to each tube and vortex to suspend evenly. D) Microfuge 25 seconds at 16000x g to pellet nuclei. E) Remove and discard supernatant and repeat steps C to D two more times, or until no hemoglobin remains. F) Resuspend nuclear pellet in 100 pl PBND with 1 p1 of 60 pg / ml proteinase K and incubate at 55°C over night. G) Heat samples to 97°C for 10 minutes to inactivate proteinase K. H) DNA extractions were stored at —20°C in freezer for PCR. * Composition of reagents: (1) Lysis Buffer: 0.32 M Sucrose, 10 mM Tris-HCl (pH 7.5), 5 mM MgC12, 1% Triton X- 100 (V/V). (2) PBND (PCR Buffer with Nonionic Detergents): 50 mM KCl, 10 mM Tris-HCI (pH 8.3), 2.5 mM MgC12, 0.1 mg/ml gelatin, 0.45% (V/V) Nonidet P40, 0.45% (V/V) Tween 20. PBND was autoclaved to sterilize and dissolve gelatin and stored frozen. Proteinase K was added into PBND (60 pg/ml) immediately prior to using. 3) PCR assay and detection of Min Apc mutations: 33 The presence of the Apc mutant allele was detected by PCR assay for the known Apc Min mutation (Dietrich et al. 1993; J acoby et al., 1996). Three primers are used in the PCR. The primer oligonucleotide sequences are as follows: wild type Apc forward primer, 5’-GCC ATC CCT TCA CGT TAG-3’; mutant Apc forward primer, 5’- TTC TGA GAA AGA CAG AAG TTA-3’; and reverse Apc primer 5’-TTC CAC TTT GGC ATA AGG-3’. Each DNA sample was amplified in a 50 p1 reaction containing: 3 p1 DNA sample; 5.0 p1 10x Perkin Elmer buffer, 1.5 mM MgC12; 0.4 pM wild type Apc forward primer; 0.4 pM mutant Apc forward primer; 0.4 pM reverse primer; 200 pM concentrations (each) of dCTP, dGTP, dATP, and dTTP; and 3.75 units of Taq DNA polymerase. Distilled water was added to make the total volume 50 pl. PCR reactions were run in a Perkin Elmer Thermal Cycler under following conditions: 1 cycle at 94°C for 3 min followed by 35 cycles at 94°C for 30 seconds, 55°C for 30 seconds, and 72°C for 1 minute followed by 1 cycle at 72°C for 7 min. Samples were then stored at 4°C until electrophoresis. After amplified through PCR, the wild type Apc forward and reverse primer generated a product of 619 bp, and Apc mutant primer and reverse primer generated a product of 313 bp. The two different products were identified by agarose gel electrophoresis. PCR products 10 p1 were mixed with loading buffer and electrophoresed through 1% agarose gels at a constant voltage of 40-50 V for about 4 hours. After staining with ethidium bromide solution (0.5 pg per 100 ml distilled water) for about 10-20 minutes, the bands were observed with the aid of transilluminator. Mice that expressed the two different Apc gene products (both the 619 and 313 hp products) were identified as 34 Min carriers, whereas mice that expressed only the 619 bp product were identified as normal. 3. Experimental Design Fifty Min mice were randomly assigned to each of the following five dietary groups after weaning (4-6 weeks old). Each group had almost the same numbers of males and females (6 males and 4 females in the casein and sulindac group; 5 males and 5 females in the soy concentrate, extruded soy and non-extruded soy group, respectively). Four dietary treatments had different protein sources, which were: 1) casein, 2) soy concentrate, 3) an extruded 30% soy product, and 4) a non-extruded soy product identical to 3. The composition of the diets is listed in Table 1 and Table 2. A fifth treatment group was fed the casein-containing diet and given sulindac (200 ppm) in the drinking water. All mice were housed in a room with constant temperature and humidity (21-24°C , 60% humidity) and a 12: 12 hour light-dark cycle. Mice had free access to the diets and water. Health condition, food and water supply were checked daily and weight was measured once a week. This study was approved by the Michigan State University All-University Committee on Animal Use and Care. Mice were kept on the dietary treatment for 10 weeks. The experimental design is shown as follows: Min mice 4-6 weeks old Sacrifice 0 1 2 3 4 5 6 7 8 9 10 ‘ Weeks on dietary treatment b 35 4. Tumor Analysis After 10 weeks of dieary treatment, mice were euthanized by C02 inhalation. The colon, cecum, and small intestine were removed, opened longitudinally, rinsed with water, pinned on cardboard, and fixed with 10% neutral buffered formalin overnight. Tissue samples were stored in 1% neutral buffered formalin solution. The number of adenomas in each section was determined with the aid of a dissecting microscope. The diameters of adenomas were also measured using a transparent grid placed under the specimen. The diameter of each adenoma was estimated to the nearest 0.5 mm. For flat tumors, the diameters were measured in two dimensions, and the average value was used to represent the size. For solid tumors, the diameters were measured in three dimensions (the three values are: (1,, (12 and d,), and the spheric volumes were caculated to represent the tumor burden by using the formula: spheric volume = (1r * dl * d2 * d3) / 6. 5. Statistical Analysis The weight data were analyzed by repeated measures. The tumor number and tumor size of Min mice were analyzed by two-way analysis of variance (sex and treatment). Data on solid tumor incidence were analyzed by F isher's exact test. When significant differences were detected (P < 0.05), treatment means were compared using the Least Significant Difference method. One male mouse in the sulindac group was found missing an intestinal tissue. Another male in the non-extruded soy group had no intestinal adenomas and was identified as non-Min carrier. Both were excluded from the statistical analysis. 36 VI. RESULTS AND DISCUSSIONS In these studies, five different dietary treatments were employed. All diets were modified from the AIN-93G diet, which is formulated to support the growth, pregnancy and lactation phases of rodents (Reeves et al., 1993). The rationale for increasing the fat content of diets using soybean oil from 7% to 15% was to more closely mimic the fat intakes characteristic of Western diets. All the diets were balanced to have the same energy density and the same amount of essential nutrients such as vitamins and minerals. The major distinction between the different diets was their protein source. There were little differences in carbohydrate or fiber compositions among the diets. Thus, carbohydrate composition was unlikely to influence colorectal carcinogenesis in these experiments. The casein diet was used as the negative control for the soy products in the studies. Casein diet plus sulindac (200 ppm) in drinking water was employed as a positive control, as it has been shown by other researchers to protect against tumorigenesis in the Min mouse model (Boolbol et al., 1996). Three different soybean diets were used in the experiments. The soy concentrate (Arcon F, Archer Daniels Midland, Decatur, IL) was prepared from soy flour via an alcohol extraction process. Hence, the soy concentrate was expected to be virtually devoid of some phytochemicals such as isoflavones and saponins. These compounds recently have been suggested to have benefits against some cancers and cardiovascular diseases (Messina et al., 1994; Kennedy et al., 1995). On the other hand, the extruded and non-extruded soy diets, in which the major protein source was full fat soy flour and did not go through the alcohol-extraction process, should be abundant in 37 isoflavones and saponins compared with soy concentrate. So, comparing the different effects between the soy concentrate and the extruded or non-extruded soy product diets will help to understand the influence of alcohol extraction on the soy products, and may also help to explain the roles of isoflavones and saponins on colon carcinogenesis. The extruded soy product had the identical ingredient composition as the non-extruded soy product. During the extrusion process the product is subjected to conditions, such as high temperature and pressure, which may alter its chemical composition. For example, protease inhibitors such as the Bowman-Birk Inhibitor will be largely inactivated, and other soy components such as isoflavones and saponins will likely stay unchanged (Anderson et al., 1995). So, comparing the different effects observed when feeding non- extruded or extruded soy products on colorectal carcinogenesis will allow us to identify the influence of extrusion on soy products, and may also help us to elucidate the role of protease inhibitors on colorectal cancer. In experiment one, the AOM-induced colon carcinogenesis rat model was used, and ACF (aberrant crypt foci) were used as the endpoint to study colon cancer initiation. ACF are putative precancerous lesions that have the potential to progress into colon tumors. Observations of ACF development allow us to investigate dietary effects on the initiation and early promotion stages of colorectal carcinogenesis. Data on rat body weight are shown in figure 1. Body weights were similar for rats in all treatment groups during the first two weeks of experiment. After this time rats fed the casein (control) diet and injected with saline had greater (P < 0.05) body weights than rats that were injected with AOM. AOM injection significantly delayed the body weight gain of the rats for about one week compared with saline injection. Among the AOM-injected 38 groups, the body weights of rats in the three soy product groups were not significantly different from rats fed the control (casein) diet. Rats fed the casein diet were heavier (P < 0.05) than rats fed the same diet and given sulindac (200 ppm) in drinking water from weeks 4-10 of the experiment. As diet intakes and digestibilities were not measured in this experiment, it was not clear why differences in body weight gain were observed. Drinking water containing sulindac might influence food intake or digestibility of rats. ACF are widely used as surrogate markers of colon cancer risk in many studies. However, there is no clear consensus about how to quantify and interpret ACF data. Different indices have been used by different researchers to evaluate the dietary treatment effects. These indices include the total number of ACF, the total number of aberrant crypts, the ACF size (average AC per ACF), and the number of large ACF (ACF with 4 or more than 4 AC per ACF). In the AOM-induced colon carcinogenesis rat model, the dose of AOM, number of injections, and even the time of the day when AOM is injected will influence the number of ACF (Pereira et al., 1994; Bird, 1995). In a study by McLellan et al. (1991), the grth and morphological characteristics of ACF were sequentially analyzed. It was found that 6 weeks after injecting a single dose of DMH (125 mg/kg BW), the total ACF number was close to its maximum level. Hence, 6 weeks after the second AOM injection was deemed sufficient for the ACF to develop and to detect the dietary effects on initiation and early promotion of colon cancer. The total number of ACF detected in rats administered the different treatments are presented in figure 2. No ACF were found in the saline-injected control group, and all five AOM-injected groups carried ACF, confirming that ACF were specifically induced in the colons of carcinogen-injected animals. Most of the ACF were located in the middle and 39 distal thirds of colon. In AOM-inj ected groups, the total number of ACF per rat ranged between 121-139 for rats in the different treatment groups. Total ACF number/rat was not significantly different for the five different dietary treatment groups. Data on ACF size are presented in figure 3. ACF size or ACF multiplicity, which is the average number of aberrant crypts in each ACF, ranged from 2.2 to 2.4 AC/ACF for rats in the five dietary treatment groups. ACF size was not significantly different for the dietary treatment groups. Some researchers have observed that only the number of large ACF is correlated with the final colorectal tumor incidence (Pretlow et al., 1992; Caderi et al., 1995). In this experiment, the number of large ACF was also compared between different dietary treatment groups. Different definitions of large ACF have been used in different studies, but the majority of studies have defined the ACF with more than three AC as large ACF. The number of large ACF has been shown to correlate better with final tumor incidence than total ACF (Pretlow et al., 1992; Magnuson et al., 1993; Davies et al., 1999). As shown in figure 4, no significant differences in the number of large ACF (AC > 3 / ACF) were observed among the five dietary treatment groups. The number of large ACF ranged from 15 to 23 per rat for different treatment groups. In this experiment, no differences in ACF development were found among the five treatment groups. Soy-containing diets had no influence on colon cancer initiation in this model. One study (Thiagarajan et al., 1998) found that feeding diets containing soy flakes and soy flour significantly reduced the ACF number compared with a diet containing alcohol-extracted soy concentrate. However, the calcium concentration in the diets used by Thiagarajan et al. (1998) was reduced to 0.1%, whereas it is 0.5% in standard AFN-93G 40 diets. Because high calcium diets have been shown to protect against colorectal cancer, it was suggested that 0.5% calcium in the AIN-93G diet might mask protective effects of other components in the diet (Thiagarajan et al., 1998). Boolbol and associates (1996) first reported that sulindac (200 ppm in drinking water) significantly reduced adenomas in Min mice. In our study, sulindac (200 ppm in drinking water) did not prevent ACF development in rats. Large dose intakes of sulindac may cause some serious side effects such as intestinal ulcers and bleeding (Ishikawa et al., 1997). However, the minimal effective dose of sulindac to reduce colonic carcinogenesis is still not clear. Several studies have been conducted to test the protective effects of sulindac using the AOM-induced carcinogenesis model. Pereira et al. (1994) evaluated the effect of different doses of sulindac on ACF induction. They found that 400 ppm sulindac in the diet significantly reduced the ACF number in the colon, whereas 200 ppm sulindac did not show any effect (Pereira et al., 1994). However, Rao et al. reported that sulindac (160 ppm in the diet) significantly reduced the tumor incidence and multiplicity in AOM- induced colon cancer model (Rao et al., 1995). The sulindac dose (200 ppm in drinking water) used in the current study was estimated to be roughly equal to the dose of 160 ppm in the diet used by Boolbol et al. (1996). It is possible that sulindac may not effectively prevent the initiation of colon cancer, but may be effective to prevent the promotion of colonic carcinogenesis. We conducted another experiment using the Min mouse model to further study the influence of the same dietary treatments on intestinal cancer promotion. Min mice carry a germline mutation in APC gene, which predisposes them to develop numerous adenomas in the intestine. Unlike ACF induced by AOM-inj ection, the endpoint used for Min mice feeding studies is adenomas, which have been demonstrated to be 41 Body weight (grams) 350 . 300 - 250 - 200 — 150 a Figure 1. Effect of Diet, Sulindac, and AOM Injection on Body Weight of Rats 1... /° injection 1 1 1 I o 1 2 3 . 4 Weeks of dietary treatment 1 1 ' l I 5 6 7 8 Saline-injection (F9) —A— Extruded Soy (n=15) — -D- — Soy Concentrate (n=15) —o— Casein (n=15) ——<>— Non-extruded Soy (n=15) _ .+- _ Sulindac + Casein (n=15) 42 . 1204 ACF number per rat Figure 2. Effect of Diet and Sulindac on Total ACF Nunlrer in AOM-injected Rats 160 1 140 _ 100 - 80- 604 40. 20- I l I l 0 _ Dietary treatment E Casein (n=15) Soy Concentrate (n=15) I Extruded Soy (n=15) I Non-extruded Soy (n=15) Cl Sulindac + Casein (n=15) 43 Average size of ACF (crypts/ACF) Figure 3. Effect of Diet and Sulindac on Average Size of ACF in AOM-injected Rats Dietary treatment E Casein (n=15) I Soy Concentrate (n=15) II Extruded Soy (n=15) ! Non-extruded Soy (n=15) El Sulindac + Casein (n=15) Number of large ACF per rat Figure 4. Effect of Diet and Sulindac on Number of Large ACF (AC > 3) in AOM-injected Rats I I I I I Dietary treatment 30- 25- N G I Casein (n=15) I Soy Concentrate (n=15) I Extruded Soy (n=15) I Non-extruded Soy (n=15) El Sulindac + Casein (n=15) 45 validated precursors for most human colorectal cancers (Sugarbaker et al. 1985). It is not clear why most adenomas which develop in Min mice are located in the small intestine, whereas F AP patients express most of their adenomas in the colon. With that exception, the phenotypic and genetic characteristics of adenomas developed in Min mice are thought similar to those observed in F AP patients. It is suggested that Min mice offer a powerful model for studying complicated gene-environment interactions, and is a good model for colorectal cancer studies (Su et al., 1992). The onset of intestinal tumor development in Min mice occurs early in life (Shoemaker et al., 1995). In our experimental design, 4-6 week-old Min mice were exposed to different dietary treatments for ten weeks to test the effect of soy and sulindac on the promotion stage of colorectal carcinogenesis. Body weights of Min mice are shown in figure 5. Male mice were significantly larger than females throughout the experiment. Body weights of male and female mice are shown in figure 6 and figure 7, respectively. Overall, there was no treatment effect on the body weights of mice. Averaged across time points, no sex by treatment interaction was detected. However, a significant treatment by time interaction was observed. During the experiment period, some differences on the body weight gain were observed for mice on different diet treatments. Mice in the sulindac, casein and extruded soy groups were heavier than mice in the non-extruded soy or soy concentrate groups. Because the diet intakes and digestibilities were not measured in this experiment, it is not clear why this difference occurred. During the last two weeks of experiment, mice in the sulindac group continued to gain weight, while mice in other four groups stopped gaining, or started to lose weight. This observation was probably due to the smaller tumor burden in mice 46 treated with sulindac. Generation of adenomas in the mouse intestine causes bleeding, intestinal obstruction and reduced food intake. Four mice were sacrificed before the experiment ended due to becoming moribund. One male fed the soy concentrate was sacrificed after 7 weeks on the treatment. Another male fed the non-extruded soy was sacrificed after 8 weeks on the trearntent. One female fed the extruded soy and another female fed the non-extruded soy were sacrificed after 9 weeks on their treatments. All the remaining Min mice were sacrificed after ten weeks of treatment. All the Min mice had multiple adenomas in the small intestine. Adenoma multiplicity was not statistically different between the female and male mice. Data on adenoma multiplicity in the whole small intestine are summarized in figure 8. The sulindac group had significantly fewer adenomas (P < 0.05) than any other diet group, averaging only 5.6 adenomas/mouse. No significant differences in adenoma number in the small intestine were detected among the other four treatment groups, averaging between 31 and 42 adenomas/mouse. Data on adenoma distribution in the small intestine are shown in Table 3. Most of the small intestinal adenomas were located in the middle and distal sections. In the proximal small intestine, the average number of adenomas was not significantly influenced by treatment. In the middle small intestine, mice in the sulindac group had significantly fewer tumors than mice in any of the other groups, averaging only 1.1 adenomas/mouse compared with 9.3 to 11.5 adenomas/mouse for the other four groups. In the distal small intestine, mice in sulindac group also had significantly fewer tumors than mice in any of the other groups, averaging 1.7 adenomas/mouse. Mice in the non-extruded soy group had 21.9 adenomas/mouse in the distal section, significantly more than mice in the control 47 (13.5 adenomas/mouse) or soy concentrate (13.6 adenomas/mouse) groups. Mice fed the extruded soy averaged 16.5 tumors in the distal small intestine, which was not significantly different from mice in the control, soy concentrate and non-extruded soy groups. Data on small intestinal adenoma diameter are shown in Figure 9. All of the small intestinal adenomas were sessile morphologically. Most small intestinal adenomas were between 1 and 2 mm in diameter. The larger adenomas were usually ulcerative in the center. The mice in the sulindac group had significantly smaller adenomas in the small intestine than mice in any of the other groups, averaging 1.4 mm in diameter. Mice fed the extruded soy diet had significantly larger tumors (1.99 mm) than those fed soy concentrate (1.74 mm) or non-extruded soy (1.72 mm). Average small intestinal adenoma diameter in mice fed the casein control diet was intermediate in size (1.92 mm) when compared with mice fed the three soy-containing diets. Data on the large intestinal tumorigenesis of Min mice are shown in table 4. Tumors in the large intestine were usually protruded solid tumors, and most of them were located in the colon. Only two tumors were found in cecum. Male and female Min mice did not differ in the number of large intestinal tumors. The incidences of solid tumors in the large intestine were 20, 30, 70, 33 and 33% for mice in the casein, soy concentrate, extruded soy, non-extruded soy and sulindac groups, respectively. Mice fed extruded soy averaged about 1.0 tumors/mouse, and mice in the other groups had 0.3 to 0.5 tumors per mouse. Spherical solid tumor volume was used to represent tumor burden. The average tumor burdens per mouse were 1.6, 2.8, 5.0, 5.2, 1.4 mm3 for mice in the casein, soy concentrate, extruded soy, non-extruded soy and sulindac groups, respectively. No significant 48 Body weight (grams) Figure 5. Effect of Diet and Sulindac on Body Weight of Male and Female Min Mice 35. 303 254 203 15- 10 I I T I I I I 0 1 2 3 4 5 6 7 8 9 .1 —1 Weeks of dietary treatrrrent —O— Casein (n=10) --A-- Extruded Soy (n=10) --0-- N on-extruded Soy (n=9) —C}— Soy Concentrate (n=10) —-— Sulindac + Casein (n=9) 49 10 Body weight (grams) Figure 6. Effect of Diet and Sulindac on Body Weight of Male Min Mice 35 , 30 - ’. - 4’: A II I t 25 -‘ . 1' t ‘A ‘ in}! -" I "- t... _::§:‘~ . 4. 4 I 'C 20 ./ E/ 15 . lo I I I I I T I I I I 0 l 2 3 4 5 Weeks of dietary treatment —O— Casein (n=6) - -A- - Extruded Soy (n=5) — -<>- - Non-extruded Soy (n=4) —Cl— Soy Concentrate (n=5) —-— Sulindac + Casein (n=5) 50 Figure 7. Effect of Diet and Sulindac on Body Weight of Female Min Mice 30- 254 15 10 Bogy weight (grams) N C \\ O» ‘\ l e I I p ‘ V \ \O ' l I l l r l l l I I I I I I I I 0 1 2 3 4 5 6 7 8 9 10 Weeks of dietary treatment —-O— Casein (n=4) - -A- - Extruded Soy (n=5) - -<>- - Non—extruded Soy (n=5) —0— Soy Concentrate (n=5) —-— Sulindac + Casein (n=4) 51 figure 8. Effect of Diet and Sulindac on Small Intestinal Adenonm Nurrber in Min Mice 50. 40-1 a a 30- Number of adenomas 201 10- Dietary treatment “b Bars not sharing a common superscript are significantly different (P < 0.05). Casein (FIG) I Soy Concentrate (n=10) m Extruded Soy (n=10) I Non-extruded Soy (n=9) Cl Sulindac + Casein (n=9) 52 Table 3. Effect of diet and sulindac on adenoma distribution in the small intestine of Min mice Treatment Proximal SI Middle SI Distal SI Casein (n=10) 5.6 i 1.8 11.5 i 1.8 13.5 i 2.8 Soy Concentrate (n=10) 5.7 i 1.8 10.1i 1.8 13.6 a: 2.8 Extruded Soy (n=10) 5.9 :1: 1.8 10.6 i 1.8 16.5 i 2.8 Non-Extruded Soy (n=9) 10.6 3:1.9 9.3 :t 1.9 b21.9 i 3.0 Sulindac + Casein (n=9) 2.2 i 1.9 31.1 i 1.9 a1.7 i 3.0 " In middle SI and distal SI: Sulindac + Casein is significantly different from all other treatments (P < 0.01). b In distal SI: Non-extruded Soy is significantly different from Casein and Soy Concentrate treatments (P < 0.05). 53 Adenoma size (mm in diameter) Figure 9. Effect of Diet and Sulindac on Small Intestinal Adenoma Size in Min mice 2.5 - ab Dietary treatment m Bars not sharing a common superscript are significantly different (P < 0.05). :— Casein (n=10) .I Soy Concentrate (n=10) E Extruded Soy (n=10) I Non-extruded Soy (n=9) C! Sulindac + Casein (n=9) 54 Table 4. Effects of diet and sulindac on tumor incidence, number of tumors/mouse, and volume of tumors (mm3)/mouse in the cecum and large intestine of Min mice. Tumor Tumor volume Tumor number Treatment incidence /mouse /mouse Casein (n=10) 20% 1.6 :t 2.5 0.5 :t 0.3 Soy Concentrate (n=10) 30% 2.8 i 2.5 0.5 i 0.3 Extruded Soy (n=10) 70% 5.0 i 2.5 1.0 :t 0.3 Non-Extruded Soy (n=9) 33% 5.2 at 2.6 0.3 i 0.3 Sulindac + Casein (n=9) 33% 1.4 i 2.6 0.3 i 0.3 55 differences were found among any of the treatments in tumor incidence, average tumor numbers per mouse, or average tumor burden per mouse. This was probably due to insufficient statistical power in the experimental design to detect differences in large intestinal tumor development. In this study, Min mice were used to test for potential protective effects of soybean products against colorectal cancer. No protective effects of soy products were found in these experiments. Differences in soy processing resulted in some effects on adenoma formation in Min mice. Mice fed extruded soy had larger small intestinal adenomas (P < 0.05) than mice fed non-extruded soy or soy concentrate, but the small intestinal adenoma size of mice fed extruded soy was not significantly different from that observed in mice fed the casein control diet. Feeding soy products had no statistically significant effects on tumor development in the large intestine, although solid tumor incidence for mice in the extruded soy group was higher than all other treatment groups. Solid tumor incidence in the large intestine was 70% for the extruded soy group, compared with 20 to 33% for the other four groups. The reason why mice fed extruded soy had larger small intestinal adenomas and tended to have higher solid tumor incidence in the large intestine is not clear. The extruded soy product had identical ingredient composition as the non-extruded soy product. During the extrusion process, the product is subjected to high temperature and pressure, which may alter its chemical composition. For example, the protease inhibitors such as the Bowman-Birk Inhibitor (BBI) will be largely inactivated. It has been suggested that BBI is a potent anti-cancer compound (Kennedy et al. 1996). This may explain why mice fed extruded soy had larger adenomas than mice fed non-extruded soy or soy concentrate. However, it cannot explain why feeding non-extruded soy did not 56 significantly reduce tumor formation in Min mice compared with feeding the control diet. Gallaher et al. (1996) reported that feeding long-term stored soy protein isolate increased ACF in AOM-treated rats. Later, they found that Maillard reactions occurred between genistein and amino acids such as lysine in soy protein under standard storage conditions (room temperature and moisture). It was suggested that the products of this reaction may act as colon cancer promoters to increase ACF number in the AOM-treated rats (Davies et al., 1998). Maillard reactions may accelerate under conditions of high temperature and moisture occurring during extrusion processing, and the generated products may attribute to the tumor promotional effects of extruded soy products. As soy extrusion is important for incorporating soy products into cereals or snacks to increase the soy consumption in Western societies, more research on this topic is warranted. We are not aware of any published research studying the effects of whole soy products on intestinal tumorigenesis in the Min mouse model. Kennedy et al. (1996) reported that feeding Min mice with 0.5% soybean-derived BBI concentrate significantly reduced small intestinal tumors by 41 % and caused a similar reduction in the colon tumors. BBI is a chymotrypsin inhibitor present in soybeans and will be largely inactivated by high temperatures. Approximately 6% of soybean protein is protease inhibitors, and 20-25% of the total protease inhibitor content is BBI (Kennedy, 1995). Heat typically applied in soy processing for making soy flours is unlikely to inactivate protease inhibitors completely. Soy flours were found to contain BBI from <0.2 to 4.9 mg/g (Anderson et al., 1995). Kennedy (1995; 1998) also suggested that BBI has extremely potent anticancer effects. As little as 0.01% BBI in the diet can suppress liver carcinogenesis in mice and colon carcinogenesis in rats (Kennedy, 1995; 1998). In the 57 current study, no protective effects of soy concentrate or non-extruded soy diets were found when compared with the casein diet. Anti-cancer effects of BBI have been most extensively studied by the Kennedy group (Weed et al., 1985; Billings et al., 1990; Kennedy, 1998), and need to be confirmed by further studies in other laboratories. Genistein has been proposed to be the main protective component in soybeans against cancer (Messina et al., 1994). However, results observed when feeding soy in carcinogen- induced colorectal cancer models are inconsistent. Several studies have shown that genistein reduced ACF in rats (Pereira et al., 1994; Thiagarajan et al., 1998). However, several other studies found different results. One study compared a low-isoflavone soy protein diet with an isoflavone-containing soy protein diet on colon carcinogenesis in AOM-treated rats. It was reported that the isoflavone-containing soy protein did not reduce the ACF formation or tumor development (Davies et al., 1999). Rao et al. (1997) found that feeding 250 ppm genistein in the diet significantly increased noninvasive and total adenocarcinoma multiplicity in AOM treated rats when compared with a casein control diet. Bennink et al. (1999) found that when rats were fed genistin, AOM-induced colon tumor incidence was enhanced compared with rats fed soy flour. Gallaher et al. (1996) reported that feeding an isolated soy protein (ISP)-based diet reduced the ACF number in AOM-treated rats. However, the same ISP increased the number of ACF after storage at room temperature for more than two years. They ascribed the conflicting results to the development of Maillard reaction products between genistein and amino acids such as lysine during storage (Davies et al., 1998). Serensen et al. (1998) tested the effects of isoflavones on intestinal tumori genesis using the Min mouse model. They found that 475 ppm isoflavones in the diet not influence 58 adenoma formation in Min mice. This observation was partly confirmed by our studies, as we observed that feeding the relatively isoflavone—rich non-extruded and extruded soy diets did not reduce intestinal tumors relative to feeding isoflavone-poor soy concentrate. Boolbol et al. (1996) first reported that administration 200 ppm sulindac in the drinking water significantly reduced adenoma formation in Min mice. In the current study, sulindac (200 ppm in drinking water) significantly reduced the number and size of small intestinal adenomas in Min mice, but it did not significantly reduce tumor incidence in the large intestine. Our results are similar to those reported in other studies using sulindac or other NSAIDs (Jacoby et a1. 1996; Chiu et al., 1997; Serensen et al., 1998; Mahmound et al.,‘1998). Metabolism studies of sulindac have suggested that, after being absorbed in the upper intestine, sulindac is reduced to its sulfide form or oxidized to a sulfone form in the liver, and then excreted into the bile. Sulindac as well as its metabolites are also reabsorbed from the small intestine and undergo extensive enterohepatic circulation (Strong et al., 1985). The extensive enterohepatic circulation may expose the small intestinal epithelium to higher local sulindac concentrations when compared with the large intestine. The concentration of sulindac (200 ppm in drinking water) used in the current study was about equal to 160 ppm sulindac in the diet (Boolbol et al., 1996). One study has reported that feeding sulindac at levels of 160 ppm and 320 ppm in the diet reduced colon cancer incidence and multiplicity in AOM-treated rats in a dose-dependent manner (Rao et al., 1998). It may be possible that lower doses of sulindac can cause adenoma reduction in the small intestine, but higher doses are needed to prevent tumor development in the colon in Min mice. However, Serensen et al. (1998) found that feeding 300 ppm sulindac in diet 59 still did not reduce tumor development in the colon in Min mice. As the number of tumors generated in large intestine of Min mice is very low, larger numbers of mice than we used would be necessary to provide sufficient statistical power to detect treatment effects. Several researchers have suggested that the Min mouse model is a valid model for colorectal cancer studies (Dove, et al., 1995; Boolbol et al., 1996; Serensen et al., 1998). However, some researchers contend that because Min mice carry a germline APC mutation and tumor development begins very early in life, diet may not be able to protect against tumor development (Davies et al. 1999). Nevertheless, several studies have reported that adenoma development in Min mice can be modified by dietary factors (Kennedy et al., 1996; Paulsen et al., 1997; Wansan et al. 1997; Kranen et al.; 1998). Although tumors in Min mice express similar genetic defects and phenotype as that observed in F AP patients, the distribution of adenomas in the intestine is different for Min mice when compared with human FAP patients. Min mice develop most of their tumors in the small intestine, whereas FAP patients develop most of their tumors in colon. The small intestine and large intestine differ in terms of epithelial cell kinetics and luminal environment. Due to interactions between diet and microflora in the large intestine, dietary effects on small intestinal tumor development may not be the same as their effects on tumors in the colon. Under such circumstances, the effect of diets on small intestinal tumors and large intestinal tumors in Min mice should be considered separately. Carcinogen-induced colorectal cancer in rodents has been suggested as a valid model for human colorectal cancer (Maltzrnan et al., 1997). Results of the current study with soy products are in agreement with several published studies in which soy protein was compared with beef protein in 1,2-dimethylhydrazine-induced colon carcinogenesis. 60 Feeding soy protein did not reduce the incidence of colon cancer in DMH-treated rats (Reddy et al., 1976; Clinton et al., 1979). Davies et al. (1999) also reported that soy protein did not reduce ACF formation or tumor development in AOM-treated rats. Unfortunately, these three studies did not specify the soy protein source. Recently, Bennink et al. (2000) reported that feeding a 21% full fat soy flour (FFSF) diet significantly reduced colon tumor incidence in the AOM-induced rat colon cancer model. They suggested that a threshold of FF SF in the diet must be exceeded before a substantial decrease in tumor incidence is observed, with the minimum threshold being between 14% and 21% of the diet (Bennink et al., 2000). In the current study, the FFSF content in the non-extruded soy diet was 24.4%, but no protective effect was observed. The lack of effect of FFSF on intestinal tumori genesis observed in our study may indicate that the Min mouse model and AOM-induced rat model differ in their sensitivity to dietary intervention. 61 VII. SUMMARY AND CONCLUSIONS Soybean consumption has been suggested to reduce colorectal cancer risk. Two experiments were conducted to investigate the chemopreventive effects of soy protein and sulindac against colorectal cancer. The influence of soy processing on colorectal carcinogenesis was also evaluated. Five dietary treatments were used in the two experiments. The major difference among different dietary treatments was their protein source. The protein sources used in the diets were: 1) casein, 2) soy concentrate, 3) an extruded 30% soy product, and 4) a non-extruded soy product identical to 3. A fifth treatment group was fed the casein- containing diet and given sulindac (200 ppm) in the drinking water. In experiment one, the AOM-induced colon carcinogenesis rat model was used. The putative precancerous lesions ACF (aberrant crypt foci) were used as the endpoint to study colon cancer initiation. No effects of soy protein or sulindac were found in terms of the number of total ACF, large ACF, and the average size of ACF when compared with the casein-containing diet. In experiment two, the Min mouse model was employed. Intestinal adenoma development was used to investigate dietary effects on the promotion stage of colorectal cancer. No protective effects of soy products were found. Soy processing had little influence on colorectal carcinogenesis. Mice fed the extruded soy had larger small intestinal adenomas (P < 0.05) than mice fed soy concentrate or non-extruded soy, but the sizes of adenomas in three soy product groups were not significantly different from those observed in mice in the control group. The solid tumor incidences in the large intestine 62 were 20, 30, 70, 33, 33% for mice on the casein, soy concentrate, extruded soy, non- extruded soy, and sulindac groups, respectively. Sulindac significantly (P < 0.05) reduced the size and number of tumors in the small intestine. However, sulindac did not significantly influence tumor formation in the large intestine. In conclusion, when soy protein was consumed as the major protein source in the diet, it did not protect against initiation and promotion of colorectal cancer compared with casein. Feeding extruded soy increased the size of small intestinal adenomas, and also tended to increase tumor incidence in the large intestine in the Min mouse model. Moderate intake of sulindac significantly reduced tumor number and size in the small intestine of Min mice, but did not protect against carcinogenesis in large intestine. Because few studies have been conducted to investigate the effects of soy products and soy processing on colorectal cancer, further studies are warranted. As moderate intake of sulindac alone did not protect against colorectal carcinogenesis, further research is needed to determine its protective threshold and the effects of its combined use with other dietary factors to prevent colorectal cancer. 63 VIII. RECOMMENDATION AND FUTURE STUDIES In the current studies, the Min mouse model was shown to be sensitive to detecting dietary effects on intestinal tumorigenesis. Since colon tumor multiplicity in Min mice is low, in future studies more Min mice should be used in each treatment group to allow greater power for detecting dietary effects on colonic carcinogenesis. Bennink et al. (2000) reported that feeding a 21% full fat soy flour (FF SF) diet significantly reduced colon tumor incidence in the AOM-induced rat colon cancer model. In the current study, feeding 24.4% FF SF diets had no effects on adenoma development in Min mice. This may indicate that the Min mouse model and AOM-induced rat model differ in their sensitivity to dietary intervention. Further research on the protective effects of soy products and the differences between the two colon cancer models to dietary intervention are needed. Sulindac has been shown to be an effective colon cancer chemopreventive agent. However, high doses of sulindac cause serious side effects. Further research on its protective threshold on colonic carcinogenesis is warranted. In the current study, feeding an extruded soy product resulted in some promotional effects on adenoma formation in Min mice. As soy extrusion is important for incorporating soy into food items, further research on this area is necessary. IX. REFERENCES Adlerceutz, H., Fotsis, T., Lampe, J ., Wahala, K., Makela, T., Brunow, G. and Hase, T. Quantitative determination of lignans and isoflavonoids in plasma of omnivorous and vegetarian women by isotope dilution gas chromatography-mass spectrometry. Scand. J. Clin. Lab. Invest, 53 (suppl. 215): 5-18, 1993. Akiyama, T., Ishida, J., Nakagawa, S., Ogawara, H., Watanabe, S., Itoh, N.M., Shibuya, M. and Fukami, Y. Genistein, a specific inhibitor of tyrosine-specific protein kinases. J. Biol. Chem., 262: 5592-5595, 1987. Akiyama, T. and Ogawara, H. Use and specificity of genistein as inhibitor of protein- tyrosine kinases. Methods Enzymol., 201: 362-370, 1991. American Cancer Society. Facts and Figures. Atlanta, GA., 1992. American Cancer Society. Facts and Figures. Atlanta, GA., 1994. American Institute of Cancer Research/World Cancer Research Fund (AICR/WCRF). Food, Nutrition and the Prevention of Cancer: A Global Perspective. 1997. Anderson, RI. and Wolf, W.J. Compositional changes in trypsin inhibitors, phytic acid, saponins and isoflavones related to soybean processing. J. Nutr., 125: 581S-588S, 1995. Barnes, S. Effect of genistein on in vitro and in vivo models of cancer, J. Nutr., 125: 7778-7898, 1995. Bedi, A., Pasricha, P.J., Akhtar, A.J., Barber, J.P., Bedi, G.C., Giardiello, F.M., Zehnbauer, B.A., Harrrilton, SR. and Jones, R]. Inhibition of apoptosis during development of colorectal cancer. Cancer Res., 55: 1811-1816, 1995. Behrens, J., J erchow, B-A., Wiirtele, M., Grimm, J ., Asbrand, C ., Wirtz, R., Ki’rhl, M., Wedllich, D. and Birchmeier, M. Functional interaction of an axin homolog, conductin, with B-catenin, APC and GSK313. Science, 280: 590-599, 1998. Behrens, J ., von Kries, J .P., Kiihl, M., Bruhn, L., Wedllich, D., Grosschedl, R. and Birchmeier, W. Functional interaction of B-catenin with transcription factor Lef-l. Nature, 382: 638-642, 1996. Bennink, M.R., Om, AS. and Miyagi, Y. Inhibition of colon cancer by soy flour but not by genistein or a mixture of isoflavones. FASEB J., 13: A50, 1999. 65 Bennink, M.R., Om, AS. and Miyagi, Y. Inhibition of colon cancer by wheat bran, soy flour and flax. FASEB J., 14: A217, 2000. Billings, P.C., Newbeme, P. and Kennedy A.R. Protease inhibitor suppression of colon and anal gland carcinogenesis induced by dimethylhydrazine. Carcinogenesis, 11: 1083- 1086, 1990. Bird, R.P. Observation and quantification of aberrant crypts in the murine colon treated with a colon carcinogen: Preliminary findings. Cancer Lett., 37: 147-151, 1987. Bird, R.P. Role of aberrant crypt foci in understanding the pathogenesis of colon cancer. Cancer Lett., 93: 55-71, 1995. Birk, Y. Protease inhibitors of plant origin and role of protease inhibitors in human nutrition. In: W. Troll and AR. Kennedy (eds.), Protease Inhibitors as Cancer Chemopreventive Agents, pp. 97 —106. New York: Plemun Publishing Corp., 1993. Boolbol, S.K., Dannenberg, S.J., Chadbum, A., Martucci, C., Guo, X., Ramonetti, J .T., Abreu-Goris, M., Nwewmark, H.L, Lipkin, M.L., Decosse, J .J . and Bertagnolli, M.M. Cyclooxygenase-2 overexpression and tumor formation are blocked by sulindac in a murine model of familial adenomatous polyposis. Cancer Res., 56: 2556-2560, 1996. Burkitt, D.P. Related disease-related cause? Lancet, 2: 1229-31, 1969. Caderni, C., Giannini, A., Lancioni, C., Luceri, C., Biggeri, A. and Dolara, P. Characterization of aberrant crypt foci in carcinogen-treated rats: association with intestinal carciongenesis. Br. J. Cancer, 71: 763-769, 1995. Cassand, P., Maziere, S., Champ, M., Meflah, K., Bomet, F. and Narbonne, J .F. Effects of resistant starch and vitamin A-supplemented diets on the promotion of precursor lesions of colon cancer in rats. Nutr. Cancer, 27: 53-59, 1997. Chang, W.C., Chapkin, RS. and Lupton, J.R. Predictive value of proliferation, differentiation and apoptosis as intermediate markers for colon tumorigenesis. Carcinogenesis, 18: 721-730, 1997. Chiu, C.H., McEntee, M.F. and Whelan, J. Sulindac causes rapid regression of preexisting tumors in Min mice independent of prostaglandin biosynthesis. Cancer Res., 57: 4267-4273, 1997. Clinton, S.K., Destree, R.J., Anderson, D.R., Turex, R.C., Imrey, PB. and Visek, W.J. 1,2-dimethylhydrazine induced intestinal cancer in rats fed beef or soybean protein. Nutr. Rep. Intl., 20: 335-342,1979. 66 Constantinou, A. and Huberrnan, E. Genistein as inducer of tumor cell differentiation: possible mechanisms of action. Proc. Soc. Exp. Biol. Med., 202: 109-115, 1995. Davies, C.G., Netto, F .M., Glassenap, N., Gallaher, C.M., Labuza, T.P. and Gallaher, D.D. Indication of the Maillard Reaction during storage of protein isolates. J. Agric. Food Chem., 46: 2485-2489, 1998. Davies, M.J., Bowey, B.A., Adlercrwutz, H., Rowland, LR. and Rurnsby, P.C. Effects of soy or rye supplementation of high fat diets on colon tumor development in azoxymethane-treated rats. Carcinogenesis, 20: 927-931, 1999. Deschner, E.E. Experimental induced cancer of the colon. Cancer, 34: 824-828, 1974. Deschner, E.E. and Long, F .C. Colonic neoplasms in mice produced with six injections of 1,2-dimethyhydrazine. Oncology, 34: 255-257, 1977. Dietrich, W.F., Lander, E.S., Smith, J .S., Mose, A.R., Gould, K.A., Luongo, C., Borenstein, N. and Dove, W. Genetic identification of Mom-1, a major modifier locus affecting Min-induced intestinal neoplasia in the mouse. Cell, 75: 631-639, 1993. Dove, W.F., Gould, K.A., Luongo, C., Moser, A. and Shoemaker, A.R. Emergent issues in the genetics of intestinal neoplasia. Cancer Surv., 25: 335-355, 1995. Druckrey, H. Organotropic carcinogenesis and mechanisms of action of symmetrical dialkylhydrazines and azo-and azoalkanes. Food Cosmet. Toxicol., 6: 578-579, 1968. Druckrey, H., Preussman, R., Matzkies, F . and Ivankovic, S. Erzeugung von Darmkrebs bei Ratten durch 1,2-dimethylhydrazin. Naturewissenschaften, 54, 285-286, 1967. Eberhart, C.E., Coffey, R.J., Radhika, A., Giardiello, F.M., Ferrenbach, s. and Dubois, R.N. Upregulation of cyclooxygenase-2 gene expression in human colorectal adenomas and adenocarcinomas. Gastroenterology, 107: 1183-1188, 1994. Enstrom, J .E. Colorectal cancer and the consumption of beef and fat. Br. J. Cancer, 32: 432, 1975. Faber, B. Cell proliferation as a major risk factor for cancer: A concept of doubtful validity. Cancer Res., 55: 3759—3762, 1995. Fearon, ER. and Vogelstein, B. A genetic model for colorectal tumorigenesis. Cell, 61: 759-767, 1990. Fotsis, T., Pepper, M., Adlercreutz, H., Fleischmann, G., Hase, T., Montesano, R. and Schweigerer, L. Genistein, a dietary-derived inhibitor of in vitro angiogenesis. Proc. Natl. Acad. Sci. USA., 90: 2690-2694, 1993. 67 F oulds, L. The natural history of cancer. J. Chronic Dis., 8: 2-37, 1958. Fukutake, M., Takahashi, M., Ishida, K., Kawamura, H., Sugimura, T. and Wakabayashi, K. Quantification of genistein and genistin in soybeans and soybean products. Food Chem. Toxicol., 34: 457-461, 1996. Gallaher, D.D., Gallaher, C.M., AN, 2., Hoffinan, RM. Soy protein isolate and genistein: Effects on initiation and progression of colon cancer. Presented at the Second lntemational Symposium on the Role of Soy in Preventing and Treating Chronic Disease, Brussels, Belgium, Sept 15-18, 1996. Garewall, H., Bernstein, H., Bernstein, C., Sampliner, R. and Payne, C. Reduced bile acid-induced apoptosis in “normal” colorectal mucosa: A potential biological marker for cancer risk. Cancer Res., 56: 1480-1483, 1996. Giardiello, F.M., Hamilton, S.R., Kush, A.J., Piantadosi, S., Hylind, L.M., Celano, P., Brooker, S.V., Robinson, CR. and Offerhaus, G.J. Treatment of colonic and rectal adenomas with sulindac in familial adenomatous polyposis. N. Engl. J. Med., 328: 1313- 1316, 1993. Giovanucci, E., Rimm, E.B., Starnpfer, M.J., Colditz, G.A., Ascherio, A. and Willett, W.C. Aspirin use and the risk of colorectal cancer and adenoma in male health professionals. Ann. Intern. Med., 121 :241-246, 1994. Goldberg, Y., Nassif, I.I., Pittas, A., Tasai, L.L., Dynlacht, B.D., Rigas, B. and Shiff, S]. The anti-proliferative effect of sulindac and sulidac sulfide on HT-29 colon cancer cells: Alterations in tumor suppressor and cell-regulatory proteins. Oncology,12: 893-901, 1996. Graf, E. and Eaton, J .W. Antioxidant functions of phytic acid. Free Radical Biol. Med., 8: 61-69, 1990. Haenszel, W., Berg, J.W., Segi, M., Kurihar, M. and Locke, F.B. Large-bowel cancer in Hawaiian Japanese. J .Natl.Cancer.Inst., 51: 1765-1779, 1973. Hanif, R., Pittas, A., Feng, Y., Koutsos, M.I., Qiao, L., Staiano-Coico, L., Shiff, 8.1. and Rigas, B. Effects of nonsteroidal anti-inflammatory drugs on proliferation and on induction of apoptosis in colon cancer cells by a prostaglandin-independent pathway. Biochem. Pharrnacol 52: 237-245, 1996. Hardman, W.E., Cameron, I.L., Heitrnan, D.W. and Contreras, E. Demonstration of the need for endpoint validation of putative biomarkers: Failure of aberrant crypt foci to predict colon cancer incidence. Cancer Res., 51, 6388-6392, 1991. 68 Howe, G.R. Dietary intake of fiber and decreased risk of cancer 5 of the colon and rectum: Evidence fi'om combined analysis of thirteen case control studies. J. Natl. Cancer Inst., 84: 1887-1895, 1992. Hsi, L., Angerman-Stewart, J. and Eling, T.E. Introduction of full-length APC modulates cyclooxygenase-2 expression in HT-29 human colorectal carcinoma cells at translational level. Carcinogenesis, 20: 2045-2049, 1999. Hu, J ., Liu, Y., Yu, Y., Zhao, T., Liu and S. Diet and cancer of the colon and rectum: A case-control study in China. Int. J. Epidemiol, 20: 362-7,1991. Hulsken, J., Birchmeier, W. and Behrens, J. E-cadherin and APC compete for the interaction with B-catenin and the cytoskeleton. J. Cell Biol., 127: 2061-2069, 1994. Ishikawa, H., Akedo, I., Suzuki, T., Narahara, H. and Otani, T. Adverse effects of sulindac used for prevention of colorectal cancer. J. Natl. Cancer Inst., 89: 1381, 1997. Jacoby, R.F., Marshall, D.J., Newton, M.A., Novakovic, K., Tutsch, K., Cole, C.E., Lubet, R.A., Kelloff, G.J., Verma, A., Moser, AR. and Dove, W.F. Chemoprevention of spontaneous intestinal adenomas in the Apc Min mouse model by the nosteroidal anti- inflammatory drug piroxicarn. Cancer Res., 56: 710-714, 1996. Jen, J ., Powell, S.M., Papadopoulos, N., Smith, K.J., Hamilton, S.R., Vogelstein, B. and Kinzler, K.W. Molecular determinants of dysplasia in colorectal lesions. Cancer Res., 54: 5523-5526, 1994. J oslyn, G., Carlson, M. and Thliveris, A. Identification of deletion mutations and three new genes at the familial polyposis locus. Cell, 66: 601-613, 1991. Kargman, S.K., O’Neill, G.P., Vickers, P.J., Evans, J .F., Mancini, J.A. and Jothy, S. Expression of prostaglandin G/H syntheses-1 and -2 protein in human colon cancer. Cancer Res., 55: 2556-2559, 1995. Kennedy, A.R. Anticarcinogenic activity of protease inhibitors. In: W. Troll and AR. Kennedy (eds.), Protease Inhibitors as Cancer Chemopreventive Agents, pp. 97 —106. New York: Plenum Publishing Corp., 1993. Kennedy, A.R. The evidence for soybean products as cancer preventive agents. J. Nutr., 125(Suppl): 7338-743S, 1995. Kennedy, A.R. Chemopreventive agents: Protease inhibitors. Pharrnacol. Ther., 78: 167- 209, 1998. 69 Kennedy, A.R., Beazer-Barclay, Y., Kinzler, K.W. and Newebeme, P.M. Suppression of carcinogenesis in the intestines of Min mice by the soybean-derived Bowman-Birk inhibitor. Cancer Res., 56: 679-682, 1996. Kinzler, K.W. and Vogelstein, B. Lessons from hereditary colorectal cancer. Cell, 87: 159-170, 1996. Koratkar, R. and Rao, A.V. Effect of soybean saponins on azoxymethane-induced preneoplastic lesions in the colon of mice. Nutr. Cancer, 27: 206-209, 1997. Kranen, H.J., Iersel, P.W.C., Rijnkels, J .M., Beems, D.B., Alink, GM and Kreijl, C.F. Effects of dietary fat and a vegetable-fruit mixture on the development of intestinal neoplasia in the ApcMm mouse. Carcinogenesis, 19: 1597-1601; 1998. Kudou, S., Fleury, Y., Welt, D., Magnolato, D., Uchida, T., Kitarnura, K. and Okubo, K. Malonyl isoflavone glycosides in soybean seeds (Glycine max Merrill). Agric. Biol. Chem, 55: 2227-2233, 1991. Laqueur, G.L. The induction of intestinal neoplasms in rats with the glucoside cycasin and its aglycone. Virchows Arch. [PathoL Anat.], 340: 151-163, 1965. Laqueur, G.L., Michaelsen, 0., Whiting, MG. and Kurland, L.T. Carcinogenic properties of nuts from Cycas circinalis. J. Natl. Cancer Inst., 31: 919-933, 1963. Laqueur, G.L. and Spatz, M. Toxicology of cycasin. Cancer Res., 28, 2262-2267, 1968. Levy, D.B., Smith, K.J., Beazer-Barclay, Y., Hamilton, S.R., Vogelstein, B. and Kinzler, K.W. Inactivation of both APC alleles in human and mouse tumors. Cancer Res., 54: 5953-5958, 1994. Lipkin, M. Phase I and phase II proliferative lesions of colonic epithelial cells in diseases leading to colonic cancer. Cancer, 34: 878-888, 1974. Lipkin, M., Venara, K. and Winawer, S.J. Seventh-day adventurist vegetarians have a quiescent proliferative activity in colonic mucosa. Cancer Lett., 26: 139-144, 1985. Lippman, S.M., Benner, SE. and Hong, W.K. Cancer chemoprevention. J. Clin. Oncol., 12: 851-873, 1994. Losi, L., Roncucci, L., Gregorio, C.D., Ponzde Leon, M. and Benhattar, J. K-ras and p53 mutations in human colorectal aberrant crypt foci. J. Pathol., 178: 259-63, 1996. Magnuson, B.A., Carr, 1. and Bird, R.P. Ability of aberrant crypt foci characteristics to predict colonic tumor incidence in rats fed cholic acid. Cancer Res., 53: 4499-4504, 1993. 70 Makishima, M., Honma, Y., Hozurni, M., Sampi, K., Hattori, M., Umezawa, K. and Motoyoshi, K. Effects of inhibitors of protein tyrosine kinase activity and/or phospatidylinositol turnover on differentiation of some human myelomonocytic leukemia cells. Leuk. Res., 15: 701-708, 1991. Maltzman, T., Whittington, J ., Driggers, L., Stephens, J. and Ahnen, D. AOM-induced mouse colon tumors do not express full-length APC protein. Carcinogenesis, 18: 2435- 2439, 1997. Markovitz, J ., Linassier, C., Fosse, P., Couprie, J ., Pierre, J ., Jacquemin-Sablon, A., Saucier, J .M., Le Pecq J .B. and Larson, A.K. Inhibitory effects of the tyrosine kinase inhibitor genistein on mammalian DNA topoisomerase II. Cancer Res., 49: 51 11-5117, 1989. Marshall, J .B. Polyps in the colon. Intestinal polyps, 92 (6): 53-65, 1992. Martin, M.S., Martin, F., Michaels, R., Bastien, H., Justrabo, E., Bordes, M. and Viry, B. An experimental model for cancer of the colon and rectum. Digestion, 8: 22-34, 1973. Martin, P.M., Horwitz, K.B., Ryan, D.S. and McGuire, W.L. Phytoestrogen interaction with estrogen receptors in human breast cancer cells. Endocrinology, 103: 1860-1867, 1978. Matsumine, A., Ogai, A., Senda, T., Okumura, N., Satoh, K., Baeg, G.H., Kawahara, T., Kobayashi, S., Okada, M., Toyoshima, K. and Akiyama, T. Binding of APC to the human homolog of the drosophila discs large tumor suppressor protein. Science, 272: 1020- 1023, 1996. McKeown-Eyssen, GE. and Bright-See, E. Dietary factors in colon cancer: International relationships. Nutr. Cancer, 6: 160-170,]984. McLellan, B.A., Medline, A. and Bird, R.P. Sequential analyses of the growth and morphological characteristics of aberrant crypt foci: Putative preneoplastic lesions. Cancer Res., 51: 5270-5274, 1991. McMichael A.J. and Potter, J .D. Reproduction, endogenous and exogenous sex hormones, and colon cancer: A review and hypothesis. J. Natl. Cancer Inst., 65: 1201- 1207, 1980. Messina, M. and Barnes, S. The role of soy products in reducing risk of cancer. J. Natl. Cancer Inst., 83: 541-546, 1991. Messina, M. J ., Persky, V., Setchell, K.D.R. and Barnes, S. Soy intake and cancer risk: A review of in vitro and in vivo data. Nutr. Cancer, 21 :1 13-131, 1994. Mettlin, C. Chemoprevention: Will it work? Int. J. Cancer, (Suppl. 10): 18-21, 1997. 71 Morin, P.J., Vogelstein, B. and Kinzler, K.W. Apoptosis and APC in colorectal tumorigenesis. Proc. Natl. Acad. Sci. USA, 93: 7950-7954, 1996. Moser, A.R., Dove, W.F., Roth, K.A. and Cordon, J .I. The Min mutation: Its effect on gut epithelial cell differentiation and interaction with a modifier system. J. Cell. Biol., 1 16: 1517-1526, 1992. Moser, A.R., Pitot, HQ and Dove, W.F. A dominant mutation that predisposes to multiple intestinal neoplasia in the mouse. Science, 247: 32-324, 1990. Nagase, H. and Nakamura, Y. Mutations of the APC gene. Human Mutat., 2: 425-434, 1993. Nishisho, I., Nakamura, Y.M, Miyoshi, Y., Miki, Y., Ando, H., Morii, A., Koyama, K., Utsunomiya, J ., Hedge, P., Markham, A., Krush, A.J., Peteren, G., Hamilton, S.R., Oakenfirll, D. and Sidhu, G.S. Could saponins be a useful treatment for hypercholesterolaemia? Eur. J. Clin. Nutr., 44: 79-88, 1990. Okura, A., Arakawa, H., Yoshinari, T. and Monden, Y. Effect of genistein on topoisomerase activity and on the growth of [Va112] Ha-ras-transfonned NIH 3T3 cells. Biochem. Biophys. Res. Commun., 157: 183-189, 1988. Oshima, M., Dinchuk, J .E., Kargman, S.L., Oshima, H., Hancock, 3., Kwong, E., Trzaskos, J .M., Evans, J .F. and Taketo, M.M. Supression of intestinal polyposis in APCA716 knockout mice by inhibition of cyclooxygenase-2. Cell, 87: 803-809. 1996. Owen, R.W. Dietary and chemopreventive strategies. Recent Results in Cancer Res., 146: 195-213, 1998. Parker, S.L., Tong, T., Bolden, S. and Wingo, P.A. Cancer statistics. Ca. Cancer J. Clin., 47: 5-27, 1996. Paulsen, J .E., Elvsaas, I.K., Steffensen, LL. and Alexander, J. A fish oil derived concentrate enriched in eicosapentaenoic and docosahexaenoic acid as ethyl ester suppresses the formation and growth of intestinal polyps in the Min mouse. Carcinogenesis, 18: 1905-1910, 1997. Pereira, M.A, Barnes, L.H, Rassman, V.L., Kelloff, G.J. and Steele, V.E. Use of azoxymethane-induced foci of aberrant crypts in rat colon to identify potential cancer chemopreventive agents. Carcinogenesis 15: 1049-1054, 1994. Piazza, G.A., Kulchak Rahm, A.L., Frutzsch, M., Sperl, G., Paranka, N.S., Gross, P.H., Brendel, K., Burt, R.W., Alberts, D.S., Pamukcu, RP. and Ahnen, D.J. Antineoplastic 72 drugs sulindac sulfide and sulfone inhibit cell growth by inducing apoptosis. Cancer Res., 55:3110—3116,1995. Potter, J .D. Risk factors for colon neoplasia - epidemiology and biology. Eur. J. Cancer, 31A: 1033-1038, 1995. Powell, S.M., Zilz, N., Beazer-Barclay, Y., Bryan, T.M., Hamilton, S.R., Thibodeau, S.N., Vogelstein, B. and Kinzler, K.W. APC mutations occur early during colorectal tumorigenesis. Nature, 359:235-237,1992. Pretlow, T.P., Barrow, BI. and Ashton, W.S. Aberrant crypts: Putative preneoplastic foci in human colonic mucosa. Cancer Res., 51: 1564-1567, 1991. Pretlow, T.P., O’Riordan, M.A., Somich, G.A., Amini, SB. and Pretlow, T.G. Aberrant crypts correlate with tumor incidence in F344 rats treated with azoxymethane and phytate. Carcinogenesis, 13: 1509-1512, 1992. Raicht, R.F., Cohen, B.I., F azzini, E.P., Sarwal, A.M. and Takahashi, M. Protective effect of plant sterols against chemically induced colon tumors in rats. Cancer Res., 40: 403-5, 1980. Rao, A.V. and Sung, M.K. Saponins as anticarcinogens. J. Nutr., 125: 7171 S-724S, 1995. Rao, C.V., Newmark, H and Reddy, B.S Chemopreventive effect of squalene on colon cancer. Carcinogenesis, 19: 287-290, 1998. Rao, C.V., Rivenson, A., Simi, B., Zang, E., Kelloff, G., Steele, V. and Reddy, B.S: Chemoprevention of colon carcinogenesis by sulindac, a nonsteroidal anti-inflammatory agent. Cancer Res., 55: 1464-1472, 1995. Rao, C.V., Wang, C.X, Simi, B., Lubet, R., Kelloff, G., Steele, V. and Reddy, B.S. Enhancement of experimental colon cancer by genistein. Cancer Res., 57: 373 7-3722, 1997. Reddy, B.S., Narisawa, T. and Weisburger, J.H. Effect of a diet with high levels of protein and fat on colon carcinogenesis in F344 rats treated with 1,2-dimethlhydrazine. J. Natl. Cancer Inst., 57: 567-569, 1976. Reeves, P.G., Nielsen, RH. and Fahey, G.C. AIN-93 purified diets for laboratory rodents: Final report of the American Institution of Nutrition Ad Hoc writing committee on the reformulation of the AIN-76A rodent diet. J. Nutr. 123: 1939-1951, 1993. Roncucci, L., Stamp, D., Medline, A., Cullen, J .B. and Bruce, W.R. Identification and quantification of aberrant crypt foci and microadenomas in the human colon. Hum. Pathol., 22: 287-294, 1991. 73 Setchell, K.D.R. Phytoestrogens: The biochemistry, physiology, and implications for human health of soy isoflavones. Am. J. Clin. Nutr., 68(suppl): 1333S-1346S, 1998. Setchell, K.D.R., Boniello, S.P., Hulme, P., Kirk, D.N. and Axelson, M. Nonsteroidal estrogens of dietary origin: possible roles in hormone dependent disease. Am. J. Clin. Nutr., 40: 569-78, 1984. Shamsuddin, A. Inositol phosphates have novel anticancer function. J. Nutr., 125: 7255- 7328, 1995. Shiff, S.J., Qiao, L., Tsai, LL. and Rigas, B. Sulindac sulfide, an asprin-like compound, inhibits proliferation, causes cell cycle quiescence, and induces apoptosis in HT-29 colon adenocarcinoma cells. J. Clin. Invest, 96:491-503, 1995. Shoemaker, A.R, Moser, A.R, Dove, E.F., N-Ethyl-N-nitrosurea treatment of multiple intestinal neoplasia (Min) mice: age-related effects on the formation of intestinal adenomas, cystic crypts, and epiderrnoid cysts, Cancer Res., 55: 4479-4485, 1995. Smith, K.J., Johnson, K.A., Bryan, T.M., Hill, D.E., Markowitz, S., Wilson, J .K., Paraskeva, C., Petersen, G.M., Hamilton, S.R., Vogelstein, B. and Kinzler, K.W. The APC gene product in normal and tumor cells. Proc. Natl. Acad. Sci. USA, 90: 2846-2850, 1993. Smith, W.L., Garavito, RM. and DeWitt, D.L. Prostagladin endoperoxide H synthases (cyclooxygenase)-1 and -2. J. Biol. Chem, 271: 33157-33160, 1996. Sorensen, I.K., Kristiansen, E., Mortensen, A.J., Nicolaisen, G.M., Wijnands, I.A.H., Kranen, H.J. and Kreijl, CF. The effect of soy isoflavones on the development of intestinal neoplasia in Apc Min mouse. Cancer Lett., 130: 217-225, 1998. Spinozzi, F., Pagiacci, M.C., Migliorati, G., Moraca, R., Crignani, F., Riccardi, C. and Nicoletti, I. The natural tyrosine kinase inhibitor genistein produces cell cycle arrest and apoptosis in Jurkat T-leukemia cells. Leuk. Res., 18: 431-439, 1994. Stappenbeck, T.S., Wong, M.H., Saam, J .R., Mysorekar, I.U. and Gordon, J .1. Notes from some crypt watchers: Regulation of renewal in the mouse intestinal epithelium. Curr. Opinion in Cell Biol., 10: 702-709, 1998. Steele, V.E., Pereira, M.S., Sigrnean, CC. and Kelloff, G.J. Cancer chemoprevention agent development strategies for genistein. J. Nutr., 125: 713S-716S, 1995. Stemmerman, G.N., Nomura, A.M.Y. and Heilbrun, L.K. Dietary fat and the risk of colorectal cancer. Cancer Res., 44: 4633-4637, 1984. 74 Strong, B.A., Warner, N.J. Renwick, A.G., George, C.F. Sulindac metabolism: The importance of an intact colon. Clin. Pharrnacol. Ther., 38: 387-393, 1985. Su, L.K., Burrell, M., Hill, D.E., Gyuris, J ., Brent, R., Wiltshire, R., Trent, J ., Vogelstein, B. and Kinzler, K.W. APC binds to the novel protein EBl. Cancer Res., 55: 2972-2977, 1995. Su, L.K., Kinzler, K.W., Vogelstein, B., Preisinger, A.C., Moser, A.R., Luongo, C., Gould, K.A. and Dove, W.F. Multiple intestinal neoplasia caused by a mutation in the murine homolog of the APC gene. Science, 256: 668-670, 1992. Su, L.K., Vogelstein, B., Kinzler, K.W. Association of the APC tumor suppressor protein with catenins. Science, 262: 1734-1737,1993. Sugarbaker, J.P., Gunderson, LL. and Wittes, R.E. Colorectal cancer. In Cancer: Principles and Practices of Oncology, V.T. Devita, S. Hellman, and SA. Rosenberg, (eds.) Philadelphia: J. B. Lippincott, pp. 800-803, 1985. Taketo, M.M. Cyclooxygenase-2 inhibitors in tumorigenesis. J. Natl. Cancer Inst., 90: 1609-1620, 1998. Thiagarajan, D., Bennink, M.R., Bourquin, L.D. and Kavas, F.A. Prevention of precancerous colonic lesions in rats by soy flakes, soy flour, genistein, and calcium. Am. J. Clin. Nutr. 68 (suppl): l394S-9S, 1998. Thiagarajan, D., Bennink, M.R., Bourquin, L.D., Mayle, J.E., Seymour, E.M., and Mridvika. Effect of soy protein consumption on colonic cell proliferation in humans. FASEB J., 13: A370, 1999. Thun, M.J., Narnboodiri, M.M. and Heath, C.W. Aspirin use and reduced risk of fatal colon cancer. N. Engl. J. Med., 325: 1593-1596, 1991. Thumherr, N., Deschner, E.E., Stonehill, E.H. and Lipkin, M. Induction of adenocarcinomas of the colon in mice by weekly injections of 1,2-dimethylhydrazine. Cancer Res., 33: 940-945. 1973. Troll, W., Frenkel, K. and Wiesner, R. Protease inhibitors as anticarcinogens. J. Natl. Cancer Inst., 73: 1245-1250, 1984. Wang, H. and Murphy, P.A. Isoflavone content in commercial soybean foods. Food Chem., 42: 1666-1673, 1994. Wasan, H.S., Novelli, M., Bee, J. and Bodmer, W.F. Dietary fat influences on polyp phenotype in multiple intestinal neoplasia mice. Proc. Natl. Acad. Sci. USA, 94: 3308- 3313, 1997. 75 Watanabe, Y., Tada, M., Kawamoto, K., Uozumi, G., Kajiwara, Y. A case-control study of cancer of the rectum and the colon. Nippon Shokakibyo Gakkai Zasshi, 81: 185-194, 1984. Weed, H., McGandy, RB. and Kennedy, A.R. Protection against dimethylhydrazine induced adenomatous tumors of the mouse colon by the dietary addition of an extract of soybeans containing the Bowman-Birk protease inhibitor. Carcinogenesis, 6: 1239-1241, 1985. Wei, H., Bowen, R., Cai, Q, Barnes, S. and Wang, Y. Antioxidant and antipromotional effects of the soybean isoflavone genistein. Proc. Soc. Exp. Biol. Med., 208: 124 —l30, 1995. Weinberg, R.A. Oncogenes, antioncogenes, and the molecular bases of multistep Carcinogenesis. Cancer Res., 49, 3713-21, 1989. Yavelow, J ., Finlay, T.H., Kennedy, AR. and Troll, W. Bowman-Birk soybean protease inhibitor as an anticarcinogen. Cancr Res., 43: 2454S-2459S, 1983. 76 IIIIIIIIIIIIIIIIIIIII 11le11111111111111171111