The synthesis and study of novel selenazolidines: potential cancer chemopreventive agents.

Selenium-containing compounds have been studied extensively for their cancer chemo-preventive properties. Although many of these drugs, such as sodium selenite or 1,4-phenylenebis(methylene)selenocyanate(p-XSC), have shown promise as anticancer agents, the toxicity of such compounds limits their usefulness. A selenocysteine prodrug is proposed to supply selenium in a less toxic manner by utilizing selenocysteine beta-lyase to provide selenium in a biochemically available form. This prodrug approach would overcome the instability and potential toxicity associated with selenocysteine alone, but allows slow release of the active component to proved chemoprevention. A novel prodrug of selenocysteine, 2-oxoselenazolidine-4-carboxylic acid (OSCA), was synthesized and preliminary biological data collected utilizing Chinese hamster lung fibroblast cells. Toxicity data collected on both the L- and D- isomers of OSCA showed a significant reduction in cytotoxicity as compared with the parent compound, selenocysteine. OSCA was also less toxic than other selenium-containing compounds such as sodium selenite, sodium selenite, selenomethionine and p-XSC. Studies were conduced to determine if OSCA released the active component, selenocysteine. Enzymatic hydrolysis studies of OSCA with 5-oxoprolinase were attempted, but were not successful. Additionally, non-enzymatic studies were carried out and compared with the sulfur-containing compounds, 2-thiazolidine-4-carboxylic acid (OTCA). As expected OSCA showed no significant hydrolysis under physiological condition, much like OTCA. Finally, glutathione peroxidase induction (GSH-Px) was studied to determine if OSCA provided selenium in a biochemically available form. The results were compared to sodium selenite and p-XSC. Neither L- or D-OSCA showed induction of GSH-Px in preliminary studies. In conclusion, these studies provide evidence that prodrug forms of selenocysteine can reduce toxicity commonly associated with selenium-containing compounds. However, further investigation is warranted to determine their effectiveness at providing selenium for protein synthesis or other metabolic pathways that may provide cancer chemo-preventive effects.

THE SYNTHESIS AND STUDY OF NOVEL SELENAZOLIDINES: POTENTIAL CANCER CHEMOPREVENTIVE AGENTS by Megan D. Short A thesis submitted to the faculty of The University of Utah in partial fulfillment of the requirements for the degree of Master of Science Department of Medicinal Chemistry The University of Utah May 2001 Copyright © Megan D. Short 2001 All Rights Reserved THE UNIVERSITY OF UTAH GRADUATE SCHOOL SUPERVISORY COMMITTEE APPROVAL of a thesis submitted by Megan D. Short This thesis has been read by each member of the following supervisory committee and by majority vote has been found to be satisfactory. Chair: JeaOftte Roberts ! ' ~?i~ Michael ... "..,."..lrl ..... THE UNIVERSITY OF UTAH GRADUATE SCHOOL FINAL READING APPROVAL To the Graduate Council of the University of Utah: I have read the thesis of Megan D. Short in its final form and have found that (1) its format, citations, and bibliographic style are consistent and acceptable~ (2) its illustrative materials including figures, tables, and charts are in place; and (3) the final manuscript is satisfactory to the supervisory committee and is ready for submission to The Graduate School. Approved for the Major Department Approved for the Graduate Council :b~~ S.c.Q~-- . David S. Chapman Dean of The Graduate School ABSTRACT Selenium-containing compounds have been studied extensively for their cancer chemopreventive properties. Although many of these drugs, such as sodium selenite or l,4-phenylenebis(methylene)selenocyanate (p-XSC), have shown promise as anticancer agents, the toxicity of such compounds limits their usefulness. A selenocysteine prodrug is proposed to supply selenium in a less toxic manner by utilizing selenocysteine ~-lyase to provide selenium in a biochemically available form. This prodrug approach would overcome the instability and potential toxicity associated with selenocysteine alone, but allow slow release of the active component to provide chemoprevention. A novel prodrug of selenocysteine, 2-oxoselenazolidine-4-carboxylic acid (OSCA), was synthesized and preliminary biological data collected utilizing Chinese hamster lung fibroblast cells. Toxicity data collected on both the L- and D- isomers of OSCA showed a significant reduction in cytotoxicity as compared with the parent compound, selenocystine. OSCA was also less toxic than other selenium-containing compounds such as sodium selenite, sodium selenate, selenomethionine and p-XSC. Studies were conducted to determine if OSCA released the active component, selenocysteine. Enzymatic hydrolysis studies of OSCA with 5-oxoprolinase were attempted, but were not successful. Additionally, nonenzymatic studies were carried out and compared with the sulfur-containing compound, 2-thiazolidine-4-carboxylic acid (OTCA). As expected OSCA showed no significant hydrolysis under physiological conditions, much like OTCA. Finally, glutathione peroxidase induction (GSH-Px) was studied to determine ifOSCA provided selenium in a biochemically available form. The results were compared to sodium selenite and p-XSC. Neither L- or D-OSCA showed induction of GSH-Px in prelin1inary studies. In conclusion, these studies provide evidence that prodrug forms of selenocysteine can reduce toxicity commonly associated with selenium-containing compounds. However, further investigation is warranted to determine their effectiveness at providing seleniun1 for protein synthesis or other metabolic pathways that may provide cancer chemopreventive effects. v To my family and friends, for their infinite love and support TABLE OF CONTENTS ABSTRACT ....................................................................................................................... iv LIST OF FIGU"RES ........................................................................................................... ix LIST OF TABLES ............................................................................................................ xi ACKNOWLEDGEMENTS ............................................................................................. xii rnTRODUCTION ............................................................................................................... 1 History ............................................................................................................................ 1 Distribution atld Metabolism ......................................................................................... 6 Selenocysteine Incorporation into Protein ..................................................................... 8 Cancer Chemoprevention ............................................................................................. 13 Selenium atld HIV ........................................................................................................ 14 Selenium Supplements ................................................................................................. 15 RATIONALE ..................................................................................................................... 18 Selenocysteine Prodrug ................................................................................................ 18 Evaluation of Biological Activity ................................................................................ 20 RESULTS AND DISCUSSION ........................................................................................ 24 Target Compound S)'Ilthesis ........................................................................................ 24 Analysis of Drug Toxicity ........................................................................................... 33 Enzymatic Hydrolysis Studies of L-OSCA (2) and D-OSCA (3) ............................... 39 Nonenzymatic Hydrolysis Studies ofL-OSCA (2) atld D-OSCA (3) ........................ .54 Evaluation of Glutathione Peroxidase Induction ........................................................ 57 CONCLUSIONS ................................................................................................................ 62 EXPERIMENTAL ............................................................................................................. 65 S)'Ilthesis ...................................................................................................................... 65 Cell Culture Procedures ............................................................................................... 69 Drug Toxicity Assay .................................................................................................... 72 General Procedures for Hydrolysis Studies ................................................................. 73 Evaluation of Enzymatic Hydrolysis ........................................................................... 74 Evaluation of Nonenzymatic Hydrolysis ..................................................................... 78 Glutathione Peroxidase Assay ..................................................................................... 78 APPENDIX A: NMR SPECTRA ..................................................................................................... 80 B: MASS SPECTRA ................................................................................................... 87 C: HPLC CHROMATOGRAMS ................................................................................ 92 REFERENCES .................................................................................................................. 96 viii LIST OF FIGURES Figure 1.1. Whole blood selenium levels for adults in various regions of the world ................. 2 1.2. Schematic diagram showing metabolic pathways of selenium ................................ 7 1.3. Schematic representation of the process of selenocysteine incorporation into protein ............................................................................................................. 1 0 1.4. Chemical structure of 1 ,4-phenylenebis( methylene )selenocyanate (p-XSC) ........ 17 1.5 Structure of selenazolidine prodrugs ..................................................................... 19 1.6. Reaction scheme of 5-xxoprolinase with 5-oxoproline and OTCA ....................... 20 1.7. Reaction scheme of the bioreduction ofMTS Tetrazolium to Fom1azan .............. 21 1.8. Chemical structure of dinitrofluorobenzene (DNFB) and dinitrophenyl derivatives of selenocystine and selenocysteine .................................................... 22 1.9. Coupled reaction of glutathione peroxidase (GSH-Px) and glutathione reductase (GSSG-Rx) ............................................................................................. 23 2.1. Target prodrugs of selenocysteine ......................................................................... 25 2.2 Chemical structure of L- and D-selenocystine ....................................................... 25 2.3. Synthetic scheme of selenocystine ......................................................................... 26 2.4. OSCA synthetic scheme ........................................................................................ 27 2.5. Reaction of selenocystine with various carbonyl donors ....................................... 28 2.6. Selenocysteine synthesis and OSCA reaction attempts ......................................... 31 2.7. Reaction attempt to form OSCA ........................................................................... .32 2.8. Drug toxicity assay in V79-4 cells ........................................................................ .35 2.9. Drug toxicity assay in V79-4 cells ........................................................................ .36 2.10. Drug toxicity assay in V79-4 cells ......................................................................... 37 2.11. Drug toxicity assay in V79-4 cells ......................................................................... 38 2.12. Drug toxicity assay in V79-4 cells ........................................................................ .40 2.13. Drug toxicity assay in V79-4 cells ........................................................................ .41 2.14. Drug toxicity assay in hepa1c1c7 cells ................................................................. .42 2.15. Drug toxicity assay in hepa1c1c7 cells ................................................................. .43 2.16. Drug toxicity assay in hepa1c1c7 cells ................................................................. .44 2.17. Drug toxicity assay in hepa1c1c7 cells .................................................................. 45 2.18. Drug toxicity assay in hepa1c1c7 cells ................................................................. .46 2.19. Drug toxicity assay in hepa1c1c7 cells ................................................................. .47 2.20. Coupled assay of pyruvate kinase (Pyr Kin) and lactate dehydrogenase (Lac DeH) used for 5-oxoprolinase enzyme activity assay .................................. .49 2.21. Analysis of 5-oxoprolinase purification by SDS-PAGE ....................................... 51 2.22. Analysis of 5-oxoprolinase purification by SDS-PAGE ...................................... 52 2.23. Selenocystine standard curve ................................................................................. 56 2.24. GSH-Px activity comparison between the hepa1c1c7 and V79 cellline ............... 59 2.25. Induction of GSH-Px activity in V79-4 cells ......................................................... 61 x LIST OF TABLES Table 2.1. Purification of 5-0xoprolinase from Rat Kidney .................................................. 49 2.2. Comparison of 5-0xoprolinase Purification Attempts .......................................... 52 2.3. Comparison of 5-0xoprolinase Purification Attempts .......................................... 54 2.4. Hydrolysis of OTCA Over Time at Varying pH ................................................... 55 ACKNOWLEDGEMENTS I would like to express my deepest appreciation to Dr. Jeanette Roberts for her guidance, support and wisdom. In addition, I would like to thank my labmates for their advice, direction and priceless friendship. I would like to say a special thanks to Pam Cassidy for serving as a great mentor to me. I appreciate all her helpful advice and suggestions. Thanks also to Britta Wilmore and Pam Dominick for all their help and support throughout my project. I would also like to express my gratitude to my committee members, Dr. Arthur Broom and Dr. Michael Franklin, for their help and time. Thanks also to the department of Medicinal Chemistry faculty and staff for all of their efforts and to the department of Pharmacology and Toxicology for the use of their instruments. Special thanks to nly mother, Sarah Harvey, for her guidance, unconditional love and support. Thanks also to my wonderful husband, Reggy for his constant companionship, understanding, and support. Additionally, I extend my appreciation to Sheryl Verbitski for her generosity in providing a home to me over the last 18 months and for her invaluable friendship. Finally, I would like to recognize the charitable financial support I have received from the Huntsman Cancer Institute, the Cancer Research Foundation of America, the University of Utah Research Committee, and also the underlying support of the instrumental facilities by NIH grant 1 S10 RR00626 and NCI grant 5 P30 CA42014. INTRODUCTION History lons lackob Berzelius, a Swedish chemist, first discovered selenium in 1817. Selenium is found in the VIA group of the periodic table and exists in many oxidation states, +4 (SeOl), +6 (Se042-), 0 (elemental Se) and -2 (diselenides, etc). Selenium is considered to be a metalloid meaning it has properties of both metals and nonmetals. It has six stable isotopes that exist in nature, 74Se (0.82%), 76Se (8.660/0), 77Se (7.31 %), 78Se (23.21 %), 80Se (50.65%), and 82Se (8.350/0), giving it a unique mass spectrometry pattern. The 77 Se isotope can also be observed by nuclear magnetic resonance spectroscopy (NMR). Selenium exists as a trace element found primarily at the earth's surface in varying concentrations. High soil concentrations of selenium have been reported in regions of the United States such as Wyoming and South Dakota, while areas of New Zealand and Finland report a much lower selenium content (Allaway, et aI., 1968; McKenzie, et aI., 1978). As a result human blood concentrations of selenium vary significantly among people in different regions of the world (Figure 1.1). For example, whole blood selenium levels in healthy adults of New Zealand are approximately 68.3 f.lg/mL, whereas individuals in the United States average 194 f.lg/mL (Allaway, et aI., 1968; McKenzie, et aI., 1978). Selenium-containing compounds were first widely recognized for their toxicity. E -~- I: -(1) 300 (1) en ~ 200 -cE 0--'" 00) _I: m~ -o(1) J: ~ 100 o 230 254 26 210 20 10 Sample Size Figure 1.1. Whole blood selenium levels for adults in various regions of the world. SD South Dakota, WY = Wyoming. Allaway, et aI., 1968; McKenzie, et aI., 1978; Dickson, et aI., 1967; Westermarck, et aI., 1977. N 3 Reports of selenium toxicity to animals can be traced back as far as the 13th century. Marco Polo wrote of a poisonous plant that caused the hooves of his horses to fall off, and modern researchers believe these plants contained toxic levels of selenium (Combs and Combs, 1986; Reilly, 1996). Such diseases among livestock became known as alkali or blind staggers disease. Animals suffering from this condition experienced loss of hair from the mane and tail as well as a crippling necrotic hoof disease. The recognition of selenium compounds as the causative factor in alkali and blind staggers disease came in the early 1930s when Kurte Franke at the South Dakota State Experimental Station assayed for trace elements in grain samples and found selenium to be present in toxic grains (Franke and Painter, 1935; Franke, et aI., 1936). Documented cases of acute selenium toxicity in humans are due mainly to occupational exposure of workers in copper smelting or selenium rectifier plants (Clinton, 1947; Wilson, 1962). The most common exposure results from inhalation of selenium fumes, and symptoms include irritation of the mucous membranes and upper respiratory tract leading to coughing, sneezing, headaches, dizziness and nausea. In severe cases pulmonary edema may occur. Chronic exposure to high levels of selenium is documented among populations in seleniferous geographic locations. Symptoms of chronic exposure include fatigue, loss of hair, scaly dermatitis and garlic breath (Diskin, et aI., 1979; Yang, et aI., 1983). Despite the toxic effects commonly associated with selenium, it was also found to have biological significance. The biological importance of selenium was first recognized in the early 1950s when scientists replaced vitamin E with selenium in the diets of chicks 4 and prevented exudative diathesis, a condition in which fluid is exuded from tissues or its capillaries (Patterson, et aI., 1957). Both vitamin E and selenium have similar functions as antioxidants. Vitamin E acts as a lipid antioxidant by scavenging free radicals. Selenium prevents lipid oxidation through the enzyme glutathione peroxidase by converting hydrogen peroxide to water. Furthermore, scientists recognized the nutritional importance of selenium when Chinese researchers identified a human disease linked to selenium deficiency. Named for the region in which it was discovered, Keshan disease is a condition in humans causing cardion1yopathy (Xia, et aI., 1994; Reilly, 1996). Chinese veterinarians recognized selenium deficiency as the causative factor in Keshan disease due to their experience with animals suffering from white muscle disease. (White muscle disease was already established as a condition occurring in selenium-deficient ruminants.) Researchers found endemic areas to be selenium-deficient and later evidence was provided that selenium supplementation could prevent Keshan disease (Xia, et aI., 1994). These studies served as the first evidence that seleniun1 was an important micronutrient. Not only was selenium an essential micronutrient, but further investigation uncovered selenium-containing compounds as cancer chemopreventive agents. Studies showed inhibition of tumorigenesis occurred at nontoxic levels of selenium (Medina and Morrison, 1988). In the mouse mammary tumor model, inhibition of tumor formation occurred at selenium levels at least five fold greater than nutritional requirements (Medina and Morrison, 1988). It is this pharmacological property of selenium that is of significant interest. Early work in this area by Shamberger and Frost suggested an 5 inverse correlation between environmental levels of selenium and the incidence of some forms of cancers in hunlans (Shamberger and Frost, 1969). Qualitative studies throughout the 1970s demonstrated the ability of selenium-containing compounds such as sodium selenite or selenium dioxide to inhibit chemically-induced carcinogenesis. For example, studies done by Schrauzer and Ismael in 1974 showed that selenium dioxide administered through the drinking water inhibited mammary tumor formation in virgin female mice by more than 700/0 (Schrauzer and Ishmael, 1974). Further studies in 1977 by Jacobs showed that sodium selenite administered in the drinking water inhibited 1,2- dimethylhydrazine-induced colon carcinogenesis in Sprague-Dawley rats by more than 50% (Jacobs, et aI., 1977). More recent studies have also shown the ability of selenium to prevent UVB-induced skin cancer in mice at levels above nutritional values (Pence, et aI., 1994). In addition, promising studies have been carried out in human subjects (Clark, et aI., 1996). Although this study showed no significant effect of selenium supplementation on the incidence of basal cell or squamous cell skin cancer, the study did show a significant decrease in overall cancer incidence. The selenium group had fewer prostate cancers, colorectal cancers, and lung cancers as compared to control groups. Although this study is only a preliminary step in determining the ability of selenium compounds to act as anticancer agents, the data support the potential for selenium supplements to prevent human cancers. Distribution and Metabolism Naturally occurring sources of selenium are plants, seafoods and, to a lesser extent, other meats. Daily selenium requirements are estimated according to examination of areas with and without selenium deficiency. Overall the adult recommended dietary selenium allowance is approximately 0.87 Jlg/kg or 70 and 55 Jlg/day for the average North American male and female, respectively (Recommended Dietary Allowances, 1989). 6 Selenium is distributed throughout the body, but during adequate selenium intake the highest concentrations are found in the liver and kidney (Lindberg and Jacobsson, 1970; Oster, et ai., 1988). Although the metabolism of selenium in the human body is not fully understood, data gathered over the last half century have been focused on answering such questions. The two main forms of selenium in food are selenomethionine, which is found in plant products, and selenocysteine, found mainly in animal products (Olson, et ai., 1970; Combs and Combs, 1984). The inorganic forms of selenium such as selenate or selenite occur in soil, but are not naturally found in foods. Even so, these compounds are most commonly utilized as dietary supplements. Figure 1.2 shows the metabolic pathways of selenium-containing compounds in higher animals. Selenomethionine and the inorganic forms of selenium (selenates and selenites) utilize glutathione to provide selenium to a selenium "pool." These compounds react with glutathione forming a selenodiglutathione intermediate, which is further nletabolized to selenopersulfide and then H2Se by reaction with glutathione reductase and NADPH SeMeth Selenites Selenates 1 (GS-Se-SG) 1 (GSSeH) H2Se IIselenium pool" Selenocysteine ! selenocysteine beta lyase SeD methY7 ~ Excretion Products HSe(P03)2- ! Selenocysteine synthase Seryl tRNA Selenoproteins Figure 1.2. Schematic diagram showing metabolic pathways of selenium. 7 8 (Reilly, 1996). Selenocysteine requires a specific enzyme, selenocysteine beta lyase, to release elemental selenium (Esaki, et aI., 1982; Tanaka, et aI., 1985). This enzyme has been isolated from porcine and human tissues. Additionally, selenocysteine has been found to be the only substrate with a Km = 0.5 mM (Daher and Van Lente, 1992). Since the Km is higher than cellular levels, the enzyme probably turns over the substrate rather slowly. Following enzymatic cleavage, elemental selenium is non-enzymatically reduced to provide for the selenium "pool." The resulting hydrogen selenide can either be detoxified yielding methylated excretion products or phosphorlyated to an activated form used in selenoprotein synthesis (Ganther, 1986; Reilly, 1996). Selenocysteine Incorporation into Protein Selenium exists in two main forms in the body, selenomethionine and selenocysteine. Animals do not distinguish between selenomethionine and methionine, but incorporate both interchangeably during protein synthesis via the same enzymatic pathway (Reilly, 1996). Thus, selenomethionine-containing proteins do not have a selenium-specific function. Rather relative amounts of the S-amino acid verses the Se-amino acid depend upon nutritional intake (Burk, 1986). Only selenocysteine is specifically incorporated during protein synthesis and regulated physiologically. Selenocysteine has been identified as the 21 5t amino acid in ribosome­directed protein synthesis (Stadtman, 1996). During protein synthesis, selenocysteine does not become directly esterified to a tRNA, but instead is formed in situ. The selenocysteyl-tRNA is formed from a unique serine-charged tRNA (Hatfield, et aI., 1982; Lee, et aI., 1996). As shown in Figure 1.3, a pyridoxal phosphate-dependent 9 selenocysteine synthase reacts with the serine residue yielding an aminoacrylate intermediate (Forchhammer, et aI., 1991). This intermediate then reacts with an activated selenium donor, monoselenophosphate, to form the selenocysteine-charged tRNA (Forchhammer and Bock, 1991). Selenocysteine is then incorporated into protein by recognition of a UGA codon, commonly believed to be a stop codon (Leinfelder, et aI., 1988). Further studies have been conducted to determine what recognition elements are present to distinguish a stop codon from selenocysteine incorporation. This mechanism has been thoroughly investigated by mutagenesis studies of both prokaryotic and eukaryotic genes (Zinoni, et aI., 1990). Using E. coli formate dehydrogenase, researchers found that gene deletion from the 3' side of TGA 140 identified a 40-base sequence directly downstream from the UGA 140 position in the message that must be present for efficient recognition as selenocysteine (Zinoni, et aI., 1990). The putative secondary structure of this 40-base region appears to be a stem loop structure (Heider, et aI., 1992; Chen, et aI., 1993). This was also found to be true in eukaryotes. A cloned type 1 5'­deiodinase (D 1), from both rat and human, required a specific 3' untranslated region (approximately 200 nucleotides) for selenocysteine incorporation (Berry, et aI., 1991). The stem-loop structure required for decoding UGA is located in the 3' untranslated regions of the mRNA and is known as selenocysteine insertion sequences (SECIS)(Stadtman, 1996). These sequences were first recognized in the message of type 1 5' -deiodinase and glutathione peroxidase (Berry, et aI., 1991). Later studies also identified a similar SECrS in mRNA of rat selenoprotein P (Berry, et aI., 1993). o I ° i-°tlASE tRNA 0 OH H2N-r° CH2 I OH Seryl-tRNA Selenocysteine Synthase .. o o I ° i-°tlASE tRNA 0 OH H2Ni:: H I O=i-O~ASE tRNA 0 OH H2 N-,>=O CH2 I SeH HSe(P03).2.- Aminoacrylate intermediate Selenocysteyl-tRNA Selenoproteins Figure 1.3. Schematic representation of the process of selenocysteine incorporation into protein. o 11 The major outcome of selenium uptake is incorporation into proteins. There are many known selenium-containing proteins and more are being discovered (Gladyshev, et aI., 1998). A few known selenium-containing proteins include glutathione peroxidase (the first known selenoenzyme) (Harris, 1992), selenoprotein P (Yang, et aI., 1987), selenoprotein W (Vendeland, et aI., 1993), type 1 thyroid hormone deiodinase (Behne, et aI., 1990) and the more recently discovered thioredoxin reductase (Ganther, 1999). The cellular or classical glutathione peroxidase (cGSH-Px) was discovered as a selenoenzyme in 1973 (Harris, 1992). It contains four identical subunits each containing a selenocysteine residue (Flohe, 1982; Ren, et aI., 1997). Later studies identified two more seleniunl-dependent GSH-Px enzymes, phospholipid hydroperoxide glutathione peroxidase (phGSH-Px) (Ursini, et aI., 1985) and plasma glutathione peroxidase (pIGSH­Px) (Takahashi, et aI., 1987). In general, these enzymes function as antioxidants by reducing hydroperoxides (Flohe, 1982). The crystal structure of the human pIGSH-Px has been resolved at 2.9A (Ren, et aI., 1997). Selenoprotein P is a 51 kD glycoprotein isolated from rat plasma (Yang, et aI., 1987; Read, et aI., 1990). The cDNA indicates selenoprotein P contains 10 selenocysteine residues. However, analysis of the purified protein shows selenium content, on average, to be 7.5 selenium atoms per molecule (Hill, et aI., 1991). It is the first and only selenoprotein identified with more than one selenium atom per polypeptide chain (Read, et aI., 1990). The function of this selenoprotein is still unknown. However, in addition to the selenocysteine residues present, this selenoprotein also contains numerous cysteine (17) residues indicating it may playa role as an antioxidant. 12 Selenoprotein W was isolated from rat muscle and contains only one selenocysteine residue (Vendeland, et aI., 1993). Western blot analysis of rat tissues showed a distribution of selenoprotein W throughout muscle, spleen, testis and brain with the highest concentrations in muscle and brain (Yeh, et aI., 1995). However, the function of this protein is still unknown. Mammalian type 1 5' -deiodinase (D 1) is an integral membrane protein of microsomes (Behne, et aI., 1990). This enzyme consists of two identical subunits each with a selenocysteine residue and is most prominent in the thyroid, liver, kidney and pituitary (Stadtman, 1996). In the rat model, D 1 requires selenocysteine to function normally (Berry, et aI., 1991). This enzyme catalyzes the conversion of the inactive prohormone, thyroxine, into the active 3,3' ,5-triiodothyronine. Recently, thioredoxin reductase (TRx) was recognized as a mammalian selenoprotein isolated from a hun1an lung adenocarcinoma cell line (Tamura and Stadtman, 1996). The mammalian enzyme differs from the E. coli enzyme in that it is larger and it contains an additional redox center with selenocysteine present (Ganther, 1999). This protein is a homodimer with a subunit mass of 57 kD. Using NADPH as an electron donor, it catalyzes thioredoxin-dependent reduction of insulin. Extensive studies on TRx are being carried out to determine the role of selenium in the enzyme mechanism as well as its role in cancer chemoprevention (Mustacich and Powis, 2000). ]3 Cancer Chemoprevention As the metabolic processes of selenium become clear, more can be understood about the function of selenium as a cancer chemopreventive agent. Although a plethora of information exists supporting the ability of selenium to prevent tumor formation, the mechanism by which prevention occurs is still unclear. Experimental data suggests selenium may function at multiple levels to accomplish chemoprevention. Possible mechanisms of cancer chemoprevention include: 1) the induction of specific selenoproteins that function to inhibit tumor formation, 2) the induction of apoptosis, 3) the alteration of carcinogen metabolism, 4) the inhibition of DNA-adduct formation and 5) the inhibition of angiogenesis. Thioredoxin reductase (TRx) is one example of a newly discovered selenoprotein with a possible role in cancer prevention. Some studies have shown an increase in activity with selenium supplementation while others have shown an inhibitory effect of selenium on thioredoxin leading to growth inhibition (Ganther, 1999; Mustacich and Powis, 2000). These interesting findings have lead to further investigation of the role of TRx in cancer prevention. Additionally, more selenoproteins continue to be discovered (Lescure, et aI., 1999). More recent studies have shown that selenium agents playa role in apoptosis, and thus may prevent tumorigenesis (Sinha, et aI., 1999; Shen, et ai., 1999). Selenium-containing compounds have been found to inhibit P450 enzymes from activating carcinogen metabolites (Shimada, et ai., 1997). Investigators have discovered the ability of sodium selenite to reduce the formation of aflatoxin B\-DNA adducts in the chick n10del (Chen, et ai., 1982; 14 Shi, et ai., 1994). Finally, studies of sodium selenite in human umbilical vein endothelial cells have shown evidence of selenium acting as a cancer chemopreventive agent by inhibiting angiogenesis (Jiang, et aI., 1999). Overall, the most reasonable explanation for the role of selenium in cancer chemoprevention incorporates all these mechanisms of action as well as some that still lay undiscovered. Selenium and HIV Not only does selenium playa important role in cancer chemoprevention, but selenium is also important to the overall function of the immune system. Selenium deficiency has been linked to such immune diseases as asthma (Horvathova, et aI., 1999), multiple sclerosis (Syburra and Passi, 1999) and viruses such as the human immunodeficiency virus (HIV). The HIV virus is known to produce selenoproteins which deplete the host's selenium levels. These selenoproteins appear to regulate viral growth according to the redox environment and selenium concentration of the host (Taylor, et aI., 1999). Studies in this area have led to the investigation of selenium supplementation and its effects on HIV. Nuclear factor KB (NF-KB) is a transcription factor known to induce HIV replication in infected T-lymphocytes. In one study selenium-supplen1ented cells increased GSH-Px activity which protected the cells against the activation ofNF-KB by H20 2 (Sappey, et aI., 1994). Further investigation showed selenium supplementation prior to exposure of chronically infected T -lymphocytic and monocytic cell lines to tumor necrosis factor a (TNF-a) decreased the induction of HI V type I replication (Hori, et aI., 1997). In 15 conjunction, selenium supplementation showed an increase in selenoprotein activity, such as GSH-Px and TRx, which contributes to antioxidant defense mechanisms. The therapeutic value of antioxidants in treating HIV has become a subject of interest. This has led to studies of glutathione repleting agents such as N-acetylcysteine (NAC) or 2-oxothiazolidine-4-carboxylic acid (OTCA). Such compounds or their derivatives, exhibited anti-HIV-I effects (airy, et aI., 1999). Selenium Supplements Studies have shown that the chemical form of selenium administered to the system has a dramatic effect on its overall availability and distribution (Lane, et aI., 1991). The most common selenium supplement used today for studies of cancer prevention is an inorganic form of selenium, sodium selenite. Although this form of selenium has been useful in providing cancer chemoprevention, the utilization of sodium selenite is limited by its toxicity. Furthermore, sodium selenite and other inorganic forms of selenium are bioreduced by reaction with glutathione, thus depletion of the system's natural defense mechanisms occurs (Reilly, 1996). Other approaches to selenium supplementation have focused more on organoselenium compounds. Synthetic phenylated or methylated selenium derivatives have displayed some success in preventing tumor formation in mammary and colon cell models (Ip, et aI., 1994; Reddy, et aI., 1996; Ip, et aI., 1997). Such compounds as methylselenocysteine, selenobetaine and dimethylselenide have been tested for their efficacy as anticancer agents (Ip, et aI., 1994). Research of triphenylselenonium and diphenylselenide has shown a reduced number of tumors in the DMBA-induced mammary tumor model (Ip, et aI., 1997). Triphenylselenonium chloride showed a reduction in total tumor number by 60-65% while diphenylselenide reduced total tumor number by about 40%. 16 Additionally, 1 ,4-phenylenebis(methylene )selenocyanate (p-XSC) represents another organoselenium compound of interest (Figure 1.4). This selenium agent and its isomers have been studied extensively as cancer chemopreventive agents in many in vitro and in vivo model systems. Studies have shown that p-XSC inhibited azoxymethane-induced colon carcinogenesis in rats at both the initiation and post initiation phases of carcinogenesis (Reddy, et ai., 1992). Studies ofp-XSC were also conducted on rats with DMBA-induced mammary carcinogenesis. Rats treated with p-XSC showed a decrease in total mammary tumors by 80%) at the initiation phase and 52% at post initiation (Jp, et aI., 1994). Evidence also shows p-XSC inhibits lung tumor formation induced by 4- (methylnitrosamino )-1-(3-pyridyl)-1-butanone (NNK) in All mice (EI-Bayouomy, et aI., 1996). Further investigation of p-XSC as a cancer chemopreventive agent has led to its evaluation for use in clinical trials (Kellof, et ai., 1996). Additionally studies of the XSC isomers have been done to determine their effectiveness at inhibiting xenobiotic metabolizing enzymes (Shimada, et aI., 1997; Sohn, et ai., 1999). To varying extents, the isomers did show inhibition of procarcinogenic substrate metabolism (Shimada, et ai., 1997). Animal studies showed that the XSC isomers played a role in inducing enzymes responsible for detoxifying chemical carcinogens (Sohn, et aI., 1999). Studies of naturally occurring selenoamino acids have also been conducted. Due to evidence that selenomethionine is misincorporated in the place of methionine during 17 CH2SeCN Figure 1.4. Chemical structure of 1 ,4-phenylenebis(methylene )selenocyanate (p-XSC). protein synthesis, it may not provide adequate therapeutic value. Selenocysteine has not been studied for its cancer chemopreventive properties due to its inherent toxicity and instability, but it's oxidized form, selenocystine, has been evaluated to a limited extent. Selenocystine provides a biochemically available form of selenium in that it elevates glutathione peroxidase levels in various tissues such as kidney, liver and muscle (Deagen, et aI., 1987). However, one study has shown no effect of selenocystine on mammary tumor incidence (Lane, et aI., 1990). So the investigation for finding an effective and less toxic selenium supplement continues. RATIONALE Selenocysteine Prodrug Selenocysteine represents the most efficiently processed form of selenium to H2Se by utilizing selenocysteine p-lyase instead of glutathione, which is important in other detoxification pathways (Figure 1.2). This metabolic pathway may provide a superior route to selenoprotein biosynthesis, and appears to be a reasonable target for selenium supplementation. However, the high degree of instability and potential toxicity associated with selenocysteine precludes it from being an effective choice as a selenium supplement. In order to provide selenocysteine in a more useful form, a prodrug form was designed. A prodrug is a chemically modified form of a drug which must be broken down to give the active drug form (Albert, 1958). The purpose of a prodrug is to implement chemical modification which diminishes any undesirable properties associated with the drug (Sinkula and Yalkowsky, 1975). Undesirable properties can be anything from taste and odor to limited absorption. A selenocysteine prodrug will function to supply selenium in a biochemically available form by increasing overall stability of selenocysteine, but still providing a form of selenium that can undergo efficient metabolic processing. The proposed selenocysteine prodrug is a condensation product of selenocysteine and a carbonyl donor. This class of compounds is known as selenazolidines (Figure 1.5a). One selenazolidine, selenazolidine-4-carboxylic acid or selenaproline (Figure 1.5b), has 19 Figure 1.5. Structure of selenazolidine prodrugs: general structure (a), selenaproline (b), 2-oxoselenazolidine-4-carboxylic acid or OSCA ( c). already been synthesized and preliminary studies have shown it inhibits protein synthesis in E. coli, rabbit reticulocytes and rat liver (De Marco, et aI., 1977; Antonucci, et aI., 1977). The target compounds for this project were both D- and L- isomers of OSCA (Figure 1. 5c). Both isomers were studied to potentially gain a better understanding of the mechanisms by which these compounds may provide selenium or function as a cancer chemopreventive agent. The design and development of selenazolidines was based on the sulfur-containing thiazolidines. 2-0xothiazolidine-4-carboxylic acid (OTCA or Procyst®) was originally synthesized in 1964 but a modified synthetic scheme was developed in 1984 (Kaneko, et aI., 1964; Boettcher and Meister, 1984). Biological studies with OTCA found it to be a substrate for the purified enzyme 5-oxoprolinase and an excellent source of cysteine for glutathione repletion (Willian1son and Meister, 1981; Williamson, et aI., 1982). 5- Oxoprolinase is an enzyme of the y-glutamyl cycle in which 5-oxoproline is hydrolyzed 20 to yield L-glutamate (Figure 1.6). OTCA is also a substrate for 5-oxoprolinase, producing cysteine (Figure 1.6). Sulfur and selenium share a number of chemical similarities. Both sulfur and selenium exist in the same oxidation states, +4, +6, 0, and -2. In addition, they also have similar atomic radii (S, 100 and Se, 115pm), electronegativities (S, 2.58 and Se, 2.55; Pauling type), and ground state electron configurations (S, _3p4 and Se, _4p4) (Pauling, 1964; Web Elements, 2000). Thus, it was reasoned the chemical synthesis of OSCA would be possible. Evaluation of Biological Activity In order to determine the effectiveness of OSCA as a prodrug of selenocysteine, biological studies were conducted in Chinese hamster lung fibroblast cells (V79-4). 9HC02H+ o 5-0xoproline ATP Enz .. Enz • ADP + Pi Cysteine Figure 1.6. Reaction scheme of 5-oxoprolinase with 5-oxoproline and OTCA. 21 These studies included 1) investigation of the toxicity of OSCA as compared to other selenium-containing anticancer agents, 2) evaluation of the ability of OSCA to release the active component, selenocysteine and 3) examination of OSCA to provide a biochemically available form of selenium as measured by GSH-Px induction. Toxicity Toxicity assays were performed to determine if OSCA was less toxic than other known selenium-containing compounds. Toxicity was measured by using an MTS tetrazolium dye which becomes bioreduced by viable cells to give formazan (Figure 1.7). These studies were conducted at various concentrations and exposure times. This information was also useful in conducting future experiments at subtoxic levels. MTS Tetrazolium Formazan Figure 1.7. Reaction scheme of the bioreduction ofMTS tetrazolium to formazan. 22 Hydrolysis OSCA was also investigated for its ability to release the active constituent, selenocysteine. Both enzymatic and non-enzymatic studies were attempted to gain a better understanding of the overall stability of OSCA and to determine if it acted as a substrate for 5-oxoprolinase as compared to OTCA. Hydrolysis was monitored by derivatizing experimental samples with dinitrofluorobenzene (DNFB) (Figure 1.8), separating the products by high performance liquid chromatography (HPLC) on a C18 column and observing at 365 nm (Dominick, et aI., 2000). r(YF oNMNO 2 2 Oinitrofluorobenzene i se -. o N-Q-~ NH:: COO 2 _ H N02 2 N,N-Oi-ONP-Selenocystine N ,Se-Oi-ON P-Selenocysteine Figure 1.8. Chemical structure of dinitrofluorobenzene (DNFB) and the dinitrophenyl derivative of selenocystine and selenocysteine. 23 GSH-Px Induction To conclude the biological evaluation of OSCA, glutathione peroxidase induction experiments were conducted to determine the ability of OSCA to provide biochemically available selenium. Comparisons were made with Na2Se03 (a known GSH-Px inducer) and p-XSC. Glutathione peroxidase activity was determined by using a coupled assay of GSH-Px and glutathione reductase (GSSG-Rx) in which NADPH depletion was nlonitored at 340 run (Figure 1.9). Summary To summarize, this project focused on developing a novel selenium-containing cancer chemopreventive agent. This approach included synthesizing a selenocysteine prodrug, OSCA, and evaluating its effectiveness through preliminary biological testing in cell culture. It was hypothesized that OSCA will (l) reduce toxicity commonly associated with selenium-containing compounds, (2) hydrolyze to release the active constituent, selenocysteine, and (3) provide a biochemically available form of selenium. 2GSH + R-OOH NADPH N}.\DP+ GSH-Px,. ROH + H20 + GSSG \.),. 2GSH GSSG-Rx Figure 1.9. Coupled reaction of glutathione peroxidase (GSH-Px) and glutathione reductase (GSSG-Rx). RESUL TS AND DISCUSSION Target Compound Synthesis In an effort to further explore the role of selenium-containing compounds as cancer chemopreventive agents, a selenocysteine prodrug, 2-oxoselenazolidine-4-carboxylic acid (OSCA), was synthesized. Both the D- and L- isomers were synthesized to better understand the mechanism of action of such conlpounds. This prodrug was designed to lower toxicity commonly associated with selenium agents and to provide a biochemically available form of selenium for use in protein synthesis or other important pathways. 2-0xoselenazolidine-4-carboxylic acid (L-OSCA and D-OSCA) The target compounds for this study were isomeric prodrugs of selenocysteine (1), L­OSCA (2) and D-OSCA (3) as shown in Figure 2.1. This drug delivery system was designed to provide controlled release of the active constituent through metabolic action. This should serve to lower overall toxicity associated with selenium, but still maintain adequate selenium levels for cancer chemoprevention. The overall synthesis of L-OSCA (2) and D-OSCA (3) proved to be quite difficult. The final synthetic procedure was a modified approach of the thiol-derived synthesis for OTCA (Boettcher and Meister, 1984). However, the yields from this synthetic scheme ("-'4-6%) make it impossible to synthesize this compound in a practical manner. Selenocysteine (1) C02H r+H SeyNH o L-OSCA (2) Figure 2.1. Target prodrugs of selenocysteine. C02H ,+"H SeyNH o D-OSCA (3) Accumulation of enough compound for use in biological studies was only accomplished through multiple reactions that required an immense amount of time and resources. OSCA was synthesized from the oxidized fom1, L-selenocystine (4) or D-selenocystine (5) (Figure 2.2). These intermediates were synthesized because the commercially available L-isomer was hard to obtain and the D-isomer could not be purchased. This was accomplished using previously published procedures (Tanaka and Soda, 1987; Klayman and Griffin, 1973). L-Selenocystine (4) O-Selenocystine (5) Figure 2.2. Chemical structure of L- and D-selenocystine. 25 26 Sodium borohydride was reacted with elemental selenium in water to produce Na2Se2 (Figure 2.3). The resulting Na2Se2 was immediately reacted with J3-chloro-alanine (pH 9) over 2 h. The product was then isolated and utilized in future reactions. OSCA was synthesized by reducing selenocystine with sodium borohydride, in an aqueous system, and then introducing phenylchloroformate as a carbonyl donor (Figure 2.4). The product from this reaction was isolated and combined with the products of several reactions before taking it to the final step of washing with methanol to remove boric acid. A scaled-up procedure was attempted, but was not successful. Many attempts were made to find a synthetic scheme that afforded a better product yield, but to no avail. Some of these approaches included changing the order of synthetic steps, using various reducing agents and carbonyl donors as well as trying to introduce a carboxylate protecting group to increase organic solubility. Initial investigation began with trying to react selenocystine with a carbonyl donor 1. H;P, RT, NH2 20 min r+ 1. pH 9, 2 h 2Se + 2NaBH4 c· Na2Se2 + H • 2. 100° , 2 370C 2 min CI COOH· , overnight p-Chloro-D-or L-alanine Figure 2.3. Synthetic scheme of selenocystine. Selenocystine D49% L 58% -tse~~20Hl 4 or 5 2. Acetic Acid j pH 6-7 1. NaBH41 0.05 N NaOH [ eOOH] Hse~~2 g,0I(CI j 1. 1.& ° I Toluene p 7 2. pH 8-9 (6 N NaOH) 2 or 3 Figure 2.4. OSCA synthetic scheme. 27 28 first, and then introducing a reducing agent to form product. Such carbonyl donors included phenylchloroformate, ethylchloroformate, and 1, l-carbonyldiimidazole. Reduction was attempted with ZnJAcOH or NaBH4• A reaction of selenocystine dissolved in 2 M NaOH with phenylchloroformate over a period of 30 min was attempted at O°C (Figure 2.5). The reaction continued stirring for an additional 30 min at room tenlperature, the pH was adjusted to 7, and the mixture was concentrated by rotary evaporation. The resulting reaction mixture became dark red suggesting the presence of elemental selenium. This reaction was attempted again, but the product was extracted into ethyl acetate and then observed by nuclear magnetic resonance (NMR) spectroscopy. The spectrum showed no evidence of a or p Anhydride by-product no significant product 40rS no significant product Dark red product Figure 2.5. Reaction of selenocystine with various carbonyl donors. 29 protons corresponding to product. During the course of the attempted reactions, a noticeable white precipitate formed upon addition of pheny1chloroformate. TLC analysis suggested the possibility that the white precipitate corresponded to the phenylchloroformate hydrolysis product, phenol. Due to the rapid formation of this compound and the lack of significant OSCA formation, it was reasoned that pheny1chloroformate was too reactive with solvent. One final reaction attempt incorporated excess phenylchloroformate to increase the chances of reaction with selenocystine. Although product formation was evident, this approach did not increase the overall yield of the reaction. So the next step entailed finding a different carbonyl donor. Carbonyldiimidazole was chosen for two reasons. First, it is less reactive than phenylchloroformate and second, it is soluble in H20 providing a monophasic system. Therefore, selenocystine was reacted with 1, 1-carbonyldiimidazole in 0.5 M NaOH for 2 h (Figure 2.5). The pH was then adjusted to 5, and Zn metal was added. Unfortunately, the reduction step did not appear to succeed with Zn so N aBH4 was added. Analysis by NMR showed potential product formation so another attempt at the reaction was made. The reaction was carried out similarly except NaBH4 was used as the reducing agent. NMR analysis showed little evidence of product formation. The evidence pointed to the conclusion that the 1,1- carbonyldiimidazole was not reactive enough with the free selenol of selenocysteine to form product before the diselenide, selenocystine, was reformed. An attempt was made with yet another carbonyl donor. Ethylchloroformate was chosen under the premise it would be more reactive than 1, 1-carbonyldiimidazole, but not as reactive as phenylchloroformate. Thus, ethylchloroformate was reacted with selenocystine in 50% dioxane/H20 at pH 8-9 with 4 M KOH (Figure 2.5). Attempts at reacting selenocystine with ethylchloroformate revealed no product formation by NMR analysis. Preliminary analysis of the NMR data suggested that the carboxylic acid may have reacted with the carbonyl donor to yield an anhydride side product. 30 The discovery of a published procedure for the reduction of diselenides led to an investigation of reducing selenocystine and isolating the free selenol, selenocysteine (Gunther, 1966). Selenocystine was suspended in absolute ethanol and hypophosphorous acid was added (Figure 2.6). The reaction was refluxed until the yellow solution became colorless (30-45 min). As the reaction cooled to room temperature, a white precipitate began to form. The white crystals, selenocysteine, were recovered. This reaction worked well the first time, but all subsequent reactions failed for unknown reasons. With the selenocysteine retrieved from the first and only successful reaction, a few attempts were made to synthesize OSCA. The isolated selenocysteine was reacted with both phenylchloroformate and 1,1- carbonyldiimidazole, but phenylchloroformate turned out to be the best carbonyl donor in these experiments. A reaction of selenocysteine with phenylchloroformate (1 : 1 molar ratio) in dioxane yielded promising preliminary results. Analysis by NMR showed definite product formation with some minor impurities present. Unfortunately, further investigation of this synthetic approach was halted because the selenocysteine synthesis failed to work. 31 -f ,r--(COOH] Se NH2 2 4 or 5 Selenocysteine Product + Impurities No product Figure 2.6. Selenocysteine synthesis (Gunther, 1966) and OSCA reaction attempts. Since working in an aqueous system caused so many competing side reactions to take place, attempts were made to make selenocystine more organic soluble. This approach included trying to protect the carboxylate of the diselenide. Model studies were first conducted using cystine (disulfide counterpart of cysteine) as starting material in order to conserve the more precious selenocystine. However, studies in this area were 32 unsuccessful due to the limited solubility of cystine. Attempted reaction of cystine with 2-(p-tolylsulfonyl) ethanol only produced unreacted starting material. A final attempt to improve the synthesis of OSCA was investigated. This synthetic scheme involved reacting selenocystine with NaBH4 in methanol, isolating the selenocysteine intermediate by repetitive dissolution in methanol and concentration by rotary evaporation to remove boric acid (Figure 2.7). The resulting solid was then reacted with phenylchloroformate in pyridine. Problems arose immediately. The selenocystine would not dissolve sufficiently in methanol to allow reduction to take place. Thus this method was modified to reducing se1enocystine in a 70-80% methanol/H20 solution. In addition, the se1enocysteine intermediate was sensitive to extensive manipulation, so the methanol washes to remove boric acid were unsuccessful. Reaction of crude selenocysteine with phenylchloroformate showed little product formation by NMR. During the reaction, the mixture became more and more yellow in color suggesting formation of the diselenide. In the interest of time and due to exhaustion of new ideas the search for a better synthetic scheme for OSCA was terminated. Much time and effort was put into slowly NaBH4 .. MeOH (),O!CI .. pyridine No product 4 or 5 Figure 2.7. Reaction attempt to form OS CA. accumulating each of the isomers for later use in biological studies. Although this adopted synthesis does not present a practical approach for more intensive studies with OSCA (such as animal studies), essential preliminary data were obtained from this project. Analysis of Drug Toxicity 33 Although selenium-containing compounds have shown much success as cancer chemopreventive agents, it is also known that such compounds can be unacceptably toxic. One objective of this project focused on identifying a less toxic selenium­containing agent for potential use as a cancer chemopreventive. The toxicity of L-OSCA (2) and D-OSCA (3) was investigated to determine if the selenazolidines were less toxic than the parent drug (selenocystine) and known selenium-containing anticancer agents. In addition, evaluation of drug toxicity also paved the way for other biological testing at sub-toxic levels. Toxicity assays were carried out in two cell lines, V79-4 and hepa1c1c7 cells. The V79-4 is a Chinese hamster lung fibroblast cell line and the hepa1c1c7 is a mouse hepatoma cell line. The preferred cell line for these studies was the mouse hepatoma cell line since this cell line has been used previously for investigation of cancer chemopreventive agents, but it was unclear whether future studies, such as investigation of GSH-Px induction, could be conducted in these cells (Uda, et aI., 1997; Song, et aI., 1999). However, analysis of GSH-Px activity in V79-4 cells had been done previously (Ochi and Miyaura, 1989; Ochi, 1990; Teixeira and Meneghini, 1996). Thus, studies were conducted in both cell lines. Drug toxicity was evaluated using an MTS tetrazolium assay. This assay monitors cell viability by producing a color change when MTS tetrazolium dye is bioreduced to form formazan (Berridge and Tan, 1993). Both the V79-4 and hepal clc7 cells were exposed to varying concentrations of drug for 1, 2, and 3 days. Drug solutions were removed, MTS tetrazolium dye was immediately added, and absorbance measurements were made at 490 nm. Surviving fraction was calculated by comparison of drug treated cells to untreated cells. 34 Figure 2.8 shows a comparison between the isomers of selenocystine and OSCA in the V79-4 cells after being exposed to drug for 1 day. L-OSCA (2), D-OSCA (3) and D­selenocystine (5) showed little toxicity, while the LD50 of L-selenocystine was ,-.., 150 J,lM. After exposing cells to drug for 2 days, L-OSCA (2) showed an LD50 of 900 J,lM, but D­OSCA (3) demonstrated no toxicity up to 1 mM (Figure 2.9). However, the LD50 ofL­and D-selenocystine was 50 J,lM and 75 J,lM, respectively. At 3 days exposure, D-OSCA (3) still showed no toxicity (Figure 2.10). The L- isomer had an LD50 of 600 J,lM and both and D-selenocystine showed an LD50 of 20 J,lM. A toxicity comparison was also made between OSCA and other known selenium­containing agents such as sodium selenite, sodium selenate, 1,4-phenylene­bis( methylene)selenocyanate (p-XSC) and selenomethionine. As shown in Figure 2.11, L-OSCA (2), D-OSCA (3), and selenomethionine were noticeably less toxic than any of the other agents tested. Both L- and D-OSCA showed no toxicity up to 1 mM. p-XSC 1.5 _ --O-OSCA c: ..0-. , • L-OSCA ~ 1.of* ~aj * • D-Selenocystine • L-Selenocystine 0) .c-: > .~ 0.5 ~ en O.O~I--~--~------~------~ 'I r-. o 25 50 75100125150300550800 Drug Concentration (~M) Figure 2.8. Drug toxicity assay in V79-4 cells. Comparison of L- and D-OSCA to the isomers of selenocystine. Each point represents 12 replicates and error bars correspond to the standard error of the mean of each data point. Cells were exposed to drug for 1 day. w VI 0.0 I I I I I I I I I I I a 100 200 300 400 500 600 700 800 900 1000 Drug Concentration (J-lM) Figure 2.9. Drug toxicity assay in V79-4 cells. Comparison of L- and D-OSCA to the isomers of selenocystine. Each point represents 12 replicates and error bars correspond to the standard error of the Inean of each data point. Cells were exposed to drug for 2 days. w 0\ 1.5 -ll-O-OSCA c .0- --..- L-OSCA U ~ ---4L- D-Selenocystine ~ 1.0 LL --+- L-Selenocystine C) .c- > I- t: 0.5 ::s tJ) 0.0 I I I II --. ... _ .. _ ... I o 10 20 30 40 50 200 2200 4200 Drug Concentration (f.LM) Figure 2.10. Drug toxicity assay in V79-4 cells. Comparison of L- and D-OSCA to the isomers of selenocystine. Each point represents 12 replicates and error bars correspond to the standard error of the mean of each data point. Cells were exposed to drug for 3 days. (",) ...J c: ...0- .. , U .tG.. LL C) .->c--: .>.. ::s U) 1.5 • • O-OSCA L-OSCA 1.0 , ! --v- SeMeth Selenate 0.5 a . a -1-1----.----.---.---,---; I 1- 1 1 a 10 20 30 40 50 250 500 750 1000 Drug Concentration (Jl M) 8- Selenite -o-p-XSC Figure 2.11. Drug toxicity assay in V79-4 cells. Comparison of L- and D-OSCA to other selenium-containing compounds. Each point represents 12 replicates and error bars correspond to the standard error of the mean of each data point. Cells were exposed to drug for 1 day. w Q() 39 was the most toxic reagent with an LDso of 10 J.lM. Sodium selenite showed similar results with an LDso of 15 J.lM. Cells were exposed to drug for 1 day. Cells that had been exposed to drug for 2 days showed basically the same results (Figure 2.12). However, the LDso of L-OSCA (2) and selenomethionine were 1 mM. At 3-day drug incubations, selenomethionine and L-OSCA (2) showed an LDso of 500 J.lM and 600 J.lM, respectively (Figure 2.13). However, D-OSCA (3), sodium selenite and p-XSC showed no change. Figures 2.14 to 2.19 show the same comparison studies, but in the hepal c 1 c7 cell line. The same general trends apply. Both L-OSCA (2) and D-OSCA (3) are less toxic than either selenocystine isomer at all 3 days of exposure (Figure 2.14, 2.15, and 2.16) and are also less toxic than any of the other selenium-containing compounds tested (Figure 2.17, 2.18, and 2.19). However, in this cell line sodium selenite was the most toxic reagent with an LDso of <5 J.lM (Figure 2.18). p-XSC also showed a high degree of toxicity with an LDso of 15 flM (Figure 2.18). D-OSCA (3) showed no toxicity up to 5 mM at all three time points and the LDso ofL-OSCA (2) was 4 mM at I-day, 3.5 mM at 2-days, and 2 mM at 3-days exposure (Figure 2.17,2.18, and 2.19). Enzymatic Hydrolysis Studies of L-OSCA (2) and D-OSCA (3) In an effort to determine if OSCA releases the active constituent, selenocysteine, enzymatic hydrolysis studies were carried out. These studies were based on previous evidence that the sulfur-containing compound, OTCA requires an enzyme for hydrolysis. OTCA was found to be a substrate for 5-oxoprolinase, an enzyme of the y- glutamyl 1.5 c ...-0. .. , CJ .C.'G. 1.0 u. en .c- .>- .>.. 0.5 ::J U) O. 0 I 'd --~, I • , ·~r' --, o 1 0 20 30 40 50 500 1 000 1500 Drug Concentration (f..1M) • • O-OSCA L-OSCA --sv- SeMeth ----.-.-!' .. ~-.-.. -.-- Selenate -- Selenite -0- p-xSC Figure 2.12. Drug toxicity assay in V79-4 cells. Comparison of L-and D-OSCA to other selenium-containing compounds. Each point represents 12 replicates and error bars correspond to the standard error of the mean of each data point. Cells were exposed to drug for 2 days. ..j;:>. o 1.5 s:: ..0-. . (.) CU 1.0 l- LL C) .s-:: .c>-: ,0.5 ::::s U) o .0 I LJrnIyI ')\,.'";-"~~::, t· I'-~ 2, I o 1 0 20 30 40 50 500 1 000 1500 Drug Concentration (JlM) • • O-OSCA L-OSCA ----9- SeMeth _. __ . -" . _-_ .. _.. S elenate ...... Selenite --0-- p-xSC Figure 2.13. Drug toxicity assay in V79·4 cells. Comparison of L- and D-OSCA to other selenium-containing compounds. Each point represents 12 replicates and error bars correspond to the standard error of the mean of each data point. Cells were exposed to drug for 3 days. -~ 1.5 _ .- D-OSCA -c0-: • L-OSCA D-Selenocysti ne ....., (.) • m 1.0 ~ s... • L- Selenocysti ne u. C) -->cc--::: 0.5 :::J C/) a . a -+I---,.--~-or------r----i I I I I o 40 80 120 160 200 1 000 3000 5000 Drug Concentration (f.lM) Figure 2.14. Drug toxicity assay in hepa1c1c7 cells. Comparison ofL- and D-OSCA to the isomers ofselenocystine. Each point represents 6 replicates and error bars correspond to the standard error of the mean of each data point. Cells were exposed to drug for 1 day. +:>. N 1.5 'Y- D-OSCA (c): • L-OSCA ta ;; D-Selenocys me CJ lr. 1.0""=,t ... • L-Selenocystm. e en .t-: > .~ 0.5 ::s CIJ 0.0 -+I---r--.....---~---.,..---t I o 40 80 120 160 200 1000 3000 5000 Drug Concentration (J.lM) Figure 2.15. Drug toxicity assay in hepa1c1c7 cells. Comparison ofL- and D-OSCA with the isomers of seienocystine. Each point represents 6 replicates and error bars correspond to the standard error of the mean of each data point. Cells were exposed to drug for 2 days. ~ w 1.5 II- O-OSCA c -0- I .... . ~ • L-OSCA ~ 100M ~~ " D-Selenocy sti ne • • L-Selenocystine C) --cc>--: 0.5 ::::s en 0.0 I .... I , r -I I 1 o 40 80 120 160 2001000 3000 5000 Drug Concentration (J-lM) Figure 2.16. Drug toxicity assay in hepalclc7 cells. Comparison of L- and D-OSCA to the isomers of selenocystine. Each point represents 6 replicates and error bars correspond to the standard error of the mean of each data point. Cells were exposed to drug for 3 days. .,t.. .,t.. 1.5 --DOSCA c -.0.-. -----0 , -o-LOSCA (.) ~ 1.0 b...... .. \ SeMeth .. • LL ~ v Selenate C) --cc>--: 0.5 • pSe-xleSnCit e :::J en 0.0 I l' , I I I .--. (---I I o 40 80 400 800 200035005000 Drug Concentration (JlM) Figure 2.17. Drug toxicity assay in hepalclc7 cells. Comparison ofL- and D-OSCA to other selenium-containing compounds. Each point represents 6 replicates and error bars correspond to the standard error of the mean of each data point. Cells were exposed to drug for 1 day. .j::.. Vl 1.5 c ...0.-. . CJ .c.a. 1.0 u. 0) .c- .>- E= 0.5 :::s U) a a I "'="~·C?"~ o --DOSCA -o-LOSCA ------ 0 • Selenate ~ - • SeMeth • p-xSC --------- ~ Selenite • I I I I I I T 40 80 400 800 200035005000 Drug Concentration (JJ,M) Figure 2.18. Drug toxicity assay in hepalclc7 cells. Comparison ofL- and D-OSCA to other selenium-containing compounds. Each point represents 6 replicates and error bars correspond to the standard error of the mean of each data point. Cells were exposed to drug for 2 days. ~ 0\ 1.5 II- D-OSCA c -.0.-. ~ • L-OSCA CJ ~ 1.0! '" ~ --f.r- Selenate u. C) ~ ------------. c -s;}- SeMeth --c>--: 0.5 ~ -o-p-xSC • Selenite ::::s tn 0.0 I - ~-""f""'l! I I·· - I I I o 40 80 200 600 1 000 2000 4000 Drug Concentration (JlM) Figure 2.19. Drug toxicity assay in hepa1clc7 cells. Comparison ofL- and D-OSCA to other selenium-containing compounds. Each point represents 6 replicates and error bars correspond to the standard error of the mean of each data point. Cells were exposed to drug for 3 days. ..j::;.. -l cycle (Williamson and Meister, 1981). Thus, it was reasoned that OSCA would potentially utilize the same enzyme activation. 48 To begin the hydrolysis studies of OSCA with 5-oxoprolinase, the enzyme was isolated from rat kidney, according to a published procedure (Van Der Werf, et aI., 1975). Briefly, the rat kidneys were homogenized and centrifuged to remove cellular debris. Ammonium sulfate was added to the resulting solution, to precipitate protein, and centrifuged again. The protein pellet was dissolved in buffer and dialyzed overnight. The dialysate was applied to the top of a Sephadex G-200 sizing column. Active fractions were combined and applied to the top of a DEAE cellulose anion exchange column. Again active fractions were pooled and protein was ammonium sulfate precipitated. The protein pellet was collected by centrifugation and dissolved in buffer. This solution was applied to a final Sephadex G-200 column. 5-0xoprolinase activity was followed using a coupled enzyme assay of pyruvate kinase and lactate dehydrogenase (Figure 2.20). The depletion ofNADH absorbance was monitored at 340 nm. The enzyme was isolated in the presence of its natural substrate, 5-oxoproline, for stability reasons. Just prior to using the isolated enzyme in experiments, the substrate was dialyzed away from the enzyme against buffer containing no 5-oxoproline. Initial purification of the enzyme achieved a sevenfold purification level (Table 2.1). To verify the enzyme of the correct molecular weight was isolated, sodium dodecyl sulfate (SDS) polyacrylamide gel electrophoresis (PAGE) was perfom1ed on enzyme samples after each purification step. The gel was stained with coomassie blue. The gel showed that outer lying band intensity diminished while one band became more intense 49 CH2 0 HOOCA OP02- Pyr Kin HoocAcH3 .. Lac DeH... OH 3 n n HOOC~CH3 ADP ATP NADH NAD+ Phosphoenol Pyruvate (PEP) Pyruvate L-Lactate Figure 2.20. Coupled assay of pyruvate kinase (Pyr Kin) and lactate dehydrogenase (Lac DeH) used for 5-oxoprolinase enzyme activity assay. Table 2.1. Purification of 5-oxoprolinase from rat kidney. Purification Volume Protein Total Specific Purification 010 Step (mL) (mg) Units Activity Factor Yield (J.lmollhr) (U/mg) Dialysate 121 338.8 700 2.0 1.0 100 SephadexO- 70 45.5 585 12.8 6.4 84 200 DEAE- 49 10.8 368 31.1 17.0 53 Cellulose (NH4hS04 25 19.0 289 15.2 7.6 41 Precipitation Sephadex 0- 8 0.7 10 15.1 7.5 2 200 50 (Figure 2.21a). This band corresponded to a molecular weight of --150,000 Da. This corresponds to the subunit molecular weight of 5-oxoprolinase. A calibration curve was plotted to determine the molecular weight of the unknown band (Figure 2.21 b). In addition, the enzyme assay showed no activity present when the natural substrate was dialyzed away from the enzyme, but enzyme activity could be regained when 5- oxoproline was added back to the enzyme solution. This result suggested that the desired enzyme, 5-oxoprolinase, was the enzyme isolated during this process. To further verify that this protein is 5-oxoprolinase, SDS-PAGE could be performed with the cloned protein to see if they migrate similarly (Guo-jie, 1996). Once OSCA was synthesized, 5-oxoprolinase was purified again to obtain material for use in hydrolysis studies. Several attempts were made to purify the enzyme, but to no avail. Purification yields at each step were lower in this experiment as compared to the first experiment (Table 2.2). Most importantly, active fractions could not be retrieved from the DEAE-cellulose anion exchange column. Every attempt with this column resulted in loss of enzyme activity. It was reasoned that the enzyme was either not eluting from the column or it was eluting inactive. Thus SDS-PAGE was run on various fractions from the DEAE-cellulose to determine whether the enzyme was eluting from the column. These results showed the enzyme was eluting from the column, but with no enzyme activity. As shown in Figure 2.22a, it appears the protein may be eluting in fraction 94 (",0.15 M NaCl) from the DEAE-cellulose column, although co-migration with the cloned enzyme would be useful in verifying this protein (Guo-jie, 1996). A calibration curve determined the approximate molecular weight of the unknown band to a. b. 6 -5 E ~ "'CJ 4 ~ .~ 3 :i CI) g 2 oJJ 1 200,000 116,250 97,400 66200 • o +---------.--~-----~------~-------~-------, o 50 100 150 200 250 Molecular Weight (kDa) 51 Figure 2.21. Analysis of 5-oxoprolinase purification. (a) SDS-PAGE of 5-oxoprolinase purification. Lane (1) molecular weight markers, (2) dialysate, (3) sephadex G-200, (4) DEAE-cellulose, (5) sephadex G-200; (b) Calibration curve to determine molecular weight of unknown band. N=l, r2=O.93. 52 Table 2.2. Comparison of 5-oxoprolinase purification attempts. Exp. Purification Volume Protein Total Units Specific Purif. 0/0 # Step (mL) (mg) (J.1mollhr) Activity Factor Yield I (U/m~) 1 Dialysate 121 338.8 700 2.0 1.0 100 SephadexG- 70 45.5 585 12.8 6.4 84 200 2 Dialysate 52 239.2 1003 4.2 1.0 100 Sephadex G- Il 27.0 169 6.3 1.5 17 200 a. b. • 200,000 116,250 97,400 66,200 50 100 150 200 250 Molecular Weight (kOal Figure 2.22. Analysis of 5-oxoprolinase purification. (a) SDS-PAGE of various fractions from the DEAE-cellulose column. Lane (1) molecular weight markers, (2) dialysate, (3) sephadex G-200, (4) DEAE-cellulose, fraction 73, (5) DEAE-cellulose, fraction 94. Arrow represents band of interest. (b) Calibration curve to determine molecular weight of unknown band. N=l, r2=0.95. 53 be 170,000 Da (Figure 2.22b). The same band is present in lanes (2) and (3), although poor reproduction of the gel makes this difficult to see. Further attempts were made to avoid enzyme degradation. First, new DEAE-cellulose resin was purchased from Sigma Chemical Company (St. Louis, MO). Second, the resin was thoroughly washed with ethanol to destroy any potential microbes present. Even in light of these precautions no active enzyme was recovered fron1 this column. In a last attempt to purify 5-oxoprolinase, a modified approach to another purification procedure was investigated (Meister, et aI., 1985). The first step involved applying the dialysate to a DEAE-cellulose column. Since no success was obtained from the DEAE­cellulose column, this step was avoided and the next step was attempted. This step involved a size exclusion column, much like the previous procedure, except an ultrogel AcA 34 resin was used instead of sephadex G-200. Although purification yields were higher than those from the sephadex G-200 column, the purification factor did not change considerably «2-fold difference) (Table 2.3). Although attempts were made to run hydrolysis studies with OTCA at this enzyme purification level, no hydrolysis product was observed. Hydrolysis products were analyzed by derivatizing samples with dinitrofluorbenzene (DNFB), separating the DNP derivatives by HPLC on a C18 column and monitoring UV absorbance at 365 nm (Dominick, et aI., 2000). It was not discovered until the conclusion of this project that the cloned version of 5-oxoprolinase is available (Guo-jie, 1996). Due to the difficulties encountered with purifying this enzyme future studies may include obtaining the cloned enzyme for use in hydrolysis studies as well as potentially answering some questions about the purification process. 54 Table 2.3. Comparison of 5-oxoprolinase purification using sephadex 0-200 and ultrogel AcA 34 resins. Method Purification Volume Protein Total Specific Purif. 0/0 Step (mL) (mg) Units Activity Factor Yield (J.Lmollhr) (U/mg) 1 Dialysate 52 239.2 1003 4.2 1 100 SephadexO- 11 27 169 6.3 1.5 16.8 200 2 Dialysate 50 475 2100 4.4 1 100 Ultrogel AcA 27 ! 111 925 8.3 1.9 44 34 Ultrogel AcA 12 18.4 185 10.1 2.3 8.8 34 Since the isolated enzyme hydrolysis studies were unsuccessful, attempts were also made at whole cell lysate studies. V79-4 cells were incubated with selenocystine or OSCA and compared with controls. Since the HPLC retention times of selenocystine overlapped with that of cysteine, no information could be obtained from these studies due to the high levels of endogenous cysteine in the cells. Furthermore, later studies showed that this HPLC method was incompatible with detecting OSCA hydrolysis products, so no attempts were made at modifying this procedure. The reasons for not using the HPLC for detecting OSCA hydrolysis products will be described in more detail in the following section. Nonenzymatic Hydrolysis Studies of L-OSCA (2) and D-OSCA (3) Nonenzymatic hydrolysis of OSCA was studied by HPLC and compared with OTCA. This involved incubating samples at 37°C, pH 7 for 1, 2, and 3 days. The sulfur analog, 55 OTCA, showed no hydrolysis, even at 3 days (Table 2.4). Additionally, <1 % hydrolysis was observed when OTCA was incubated at 37°C for 3 days at pH 4 or 9. However, hydrolysis studies with OSCA could not be followed by HPLC. An unidentifiable peak consistently appeared in the chromatogram, but no peak corresponding to N,N-bis-dinitrophenyl selenocystine was observed. Several attempts were made to identify this peak. First, a sample of OSCA was spiked with selenocystine to determine if the peak was shifting in the matrix. Instead of eluting at 30 min, selenocystine eluted on top of the unidentified peak at 31 min (Appendix C). Second, the peak was collected and submitted for mass spectrometry analysis and NMR. Unfortunately, the data conflicted. The mass spectrum showed a prominent peak at 333.1 m/z, but no selenium isotope pattern was present (Appendix B). In addition the NMR showed peaks consistent with the a and ~ protons of OSCA, but no aromatic peaks Table 2.4. Hydrolysis of OTCA over time at varying pH. ND Not Detected. Limit of detection for cysteine = < 0.10/0. pH Day 0/0 Hydrolysis I I 4 1 0.2 2 0.3 3 0.4 7 1 ND 2 ND 3 ND 9 1 0.3 2 0.3 3 0.3 56 were observed suggesting no dinitrophenyl functionality was present (Appendix A). Since the peak could not be identified, further analysis of hydrolysis by HPLC was terminated. It appeared the derivatization assay was incompatible with monitoring OSCA hydrolysis for unknown reasons. In a final attempt to study the hydrolysis of OSCA, a colorimetric assay was used to detect the presence of selenocysteine (Esaki and Soda, 1987). This assay was a modified procedure of the Gaitonde method which detects cysteine (Gaitonde, 1967). Samples of both L-OSCA (2) and D-OSCA (3) were incubated at 37°C at pH 7 for varying amounts of time. The samples were then reduced with sodium borohydride and reacted with a ninhydrin solution. The absorbance was measured at 570 nm. A selenocystine standard curve was developed to quantitate hydrolysis in experimental samples (Figure 2.23). The absorbance values for L-OSCA (2) were 0.02 (0 h), 0.01 (1 h), 0.02 (12 h), 0.03 (24 h), 2.0 E s:: -~Bs:: 11..50 cu ..c.. o tI) 0.5 ~ O.O-+---.,...---r------r-----,-----, 0.0 0.5 1.0 1.5 2.0 2.5 Concentration (mM) Figure 2.23. Selenocystine standard curve (N=2, r2=0.99). 57 and 0.01 (48 h). The D- isomer showed similar results with absorbance values of 0.01 (0 h), 0.01 (1 h), 0.01 (12 h), 0.04 (24 h), and 0.02 (48 h). The absorbance values were very low and the incubated samples did not change from the control samples that were not incubated. Thus L- and D-OSCA showed no hydrolysis suggesting it exhibits similar stability as OTCA. Enzyme studies will be necessary to determine if release of selenocysteine is possible from these prodrugs. Evaluation of Glutathione Peroxidase Induction Finally, the prodrugs (L- and D-OSCA) were evaluated for their ability to provide selenium in a biochemically available form. Glutathione peroxidase (GSH-Px) is a well­known selenium-dependent enzyme that is commonly used for monitoring available selenium levels (Lane, et aI., 1991; Christensen and Burgener, 1992; Ricetti, et aI., 1994; Zbikowska, et aI., 1997; Yoshida, et aI., 1999). This is done through measuring GSH-Px induction at either the gene level (transcription) or protein level (translation). Although it's hypothesized that OSCA will induce GSH-Px activity through gene up-regulation, studies were only done to monitor GSH-Px activity at the protein level. Both L-OSCA (2) and D-OSCA (3) were tested for their ability to provide selenium to GSH-Px. Ideally, these studies were to be carried out in a hepatoma cell line to provide a common model system among cancer chemopreventive studies. However, it was unclear whether GSH-Px activity could be detected in this cell line so the V79-4 cell line was also considered since some studies of GSH-Px induction had been investigated in this cell line (Ochi and Miyaura, 1989; Ochi, 1990; Teixeira and Meneghini, 1996). 58 Therefore, initial studies of GSH-Px activity focused on comparing the hepalclc7 cell line to the V79-4 cells. Studies were carried out by exposing both cell lines to increasing concentrations ofNa2Se03 (positive control) for 2 days and then assaying for GSH-Px activity. The cells were exposed to selenium for 2 days to allow sufficient time for selenocysteine incorporation during protein synthesis to occur. This was accomplished by monitoring the depletion ofNADPH (340 nm) through a coupled assay of GSH-Px and glutathione reductase (Figure 1.9). Protein content of celllysates was measured using a Bradford assay (Bradford, 1976). GSH-Px activity was expressed as Jlmol NADPH oxidized per min per mg protein. Figure 2.24 shows a comparison of GSH-Px induction (by Na2Se03) between the hepalclc7 and V79-4 cells. Both cell lines were exposed to varying drug concentrations for 48 h and then GSH-Px activity was assessed. Although both cell lines showed some GSH-Px induction, the V79-4 cell line showed a much greater response with increasing levels ofNa2Se03. However, as drug concentrations approached toxic levels GSH-Px activity began to drop off in the V79-4 cell line. Induction of GSH -Px activity in the hepalclc7 cell line was much lower. Further investigation was not carried out in the hepalclc7 cell line. However, due to the slow growth of the hepalclc7 cells, as compared to the V79-4 cells, future studies could entail modifying the procedure to see if the hepalclc7 cells give a better response. Since studies ofGSH-Px activity were already published using the V79-4 cell line, these cells were used for further experimentation. Both selenazolidine isomers were tested for their ability to induce GSH-Px activity 150 -c 125 cu""-'" N C) :E ~ 100 >< .- o E :t::::: 75 D.. 0 ~} 50 25 O~I--------------------~---------- o 10 20 30 Sodium Selenite (f.1M) • V79 -'~Hepa Figure 2.24. GSH-Px activity comparison bet\veen the hepalclc7 and V79-4 cell line. Induction of GSH-Px was measured by exposing cells to varying concentrations ofNa2Se03 for 48 h. Each point represents 3 replicates and error bars correspond to the standard error of the mean for each data point. VI '>0 60 along with p-XSC and Na2Se03 as a positive control (Figure 2.25). Na2Se03 showed the most GSH-Px induction over any of the other drugs tested. The greatest GSH-Px induction occurred at 25 J.lM where values were approximately fivefold greater than control values. Beyond 25 J.lM Na2Se03, GSH-Px activity began to decrease probably due to a lack of cell growth at higher concentrations. Ochi and Miyaura reported the greatest GSH-Px induction at 60 nM where values were approximately 1.5- to 2-fold greater than control values (Ochi and Miyaura, 1989). p-XSC showed some induction of the enzyme, but both L-OSCA (2) and D-OSCA (3) showed no significant GSH-Px induction even at concentrations of 400 J.lM. It is unclear at this point whether the prodrug is not being hydrolyzed to yield the active component utilized in GSH-Px synthesis or whether there is simply no increase in protein synthesis in the presence of the agent. The V79-4 cells do contain 5-oxoprolinase, but it is unclear whether enough time was allotted to allow hydrolysis of OSCA (Russo, et aI., 1985). Future studies may include examining GSH-Px activity of cells that have been exposed to drug at various time points to determine if sufficient time is allotted for hydrolysis of OSCA to occur. 500 ...-.. ~ 400 ...c..-.. . -..Eo..... 300 E 200 ::t ----- lE 100~ • ~ • • ii , Z 0 -1 00 -+1 --r----r---,------y----1 r - -, . - I I o 10 20 30 40 50 150 250 350 450 Concentration (f.lM) --O-OSCA '-L-OSCA .- p-xSC Selenite Figure 2.25. Induction of GSH-Px activity in V79-4 cells. Cells were exposed to each drug for 48 h. Each point represents 9 replicates from three experiments. Error bars correspond to the standard error of the mean for each data point. .Q...\. CONCLUSIONS Considerable evidence has been collected which supports the ability of selenium­containing compounds to provide cancer chemopreventive activity. Although some selenium compounds (Na2Se03 or p-XSC) have shown promise as anticancer agents, the utilization of these compounds is limited by their toxicity. This might present a significant problem with using these agents as pharmaceuticals. In order to overcome the toxicity that can be associated with selenium-containing compounds, a prodrug approach was investigated. A novel selenocysteine prodrug, 2- oxoselenazolidine-4-carboxylic acid (OSCA), was synthesized. In addition, biological studies were performed to (1) determine if OSCA was less toxic than other known selenium-containing anticancer agents, (2) evaluate if OSCA released the active constituent, selenocysteine, and (3) investigate whether OSCA provided a biochemically available form of selenium. Both the L- and D- isomers of OSCA were synthesized to gain a better understanding of the mechanism by which these compounds may provide selenium for anticarcinogenic activity. If the isomers show similar biological effects this may clarify whether enzyme­mediated (stereospecific) processes are being utilized. Although the synthesis of the compounds proved to be difficult, important preliminary data were obtained from this project. For future experin1entation with OSCA, such as animal studies, a more effective synthetic scheme needs to be developed. 63 One of the goals of this project was to develop a selenium-containing compound with less toxicity than other related compounds. A Chinese lung fibroblast (V79-4) and a mouse hepatoma (hepalclc7) cell line were used to test OSCA toxicity. Comparisons were made between the L- and D- isomers of OSCA and L- and D-selenocystine. In both cell lines, OSCA was significantly less toxic than the isomers of selenocystine. Cell survivability for OSCA was greater than 50% up to 200 11M while both selenocystine isomers showed survival rates less than 10% at the same concentration. The D- isomer of OSCA showed the least toxicity with about 1000/0 survival up to 1 mM while L-OSCA survival dropped to <20% (1 mM) in the V79-4 cells. This result suggests that there may be an enzymatic or other stereospecific nlechanism of action for the OSCA compounds. OSCA was evaluated for its ability to release selenocysteine. Initially, enzymatic studies with 5-oxoprolinase were proposed. However, the enzyme purification procedure was unsuccessful so other avenues of hydrolysis were investigated. Both whole cell lysate and nonenzymatic studies were pursued. However, some difficulties arose with monitoring the hydrolysis product, selenocystine, by HPLC. Studies of the synthesized and in vitro N,N-bis-dinitrophenyl- selenocystine (DNP-selenocystine) showed a retention time of 30 min by HPLC. But experimental samples of OSCA always produced a peak at 31 min. This peak was collected for identification. Unfortunately, the mass spectrometry and NMR data were inconsistent and the identity of this peak was not elucidated. It appeared the HPLC procedure was not compatible with monitoring OSCA hydrolysis. Therefore, OSCA hydrolysis was monitored using a selenocysteine colorimetric assay. Both L- and D-OSCA showed no hydrolysis after being incubated at 64 37°C, pH 7 for up to 48 h. These data suggest OSCA has similar stability to OTCA. Lastly, OSCA was studied for its ability to provide a biochemically available form of selenium. This was accomplished by measuring glutathione peroxidase (GSH-Px) activity in V79-4 and hepalclc7 cells. Since the V79-4 cells showed a greater response to GSH-Px induction, these cells were used to study OSCA. The results of OSCA induction were compared to Na2Se03 (a known GSH-Px inducer) and p-XSC. Na2Se03 exhibited the greatest GSH-Px induction while p-XSC only showed limited induction. Both L-OSCA (2) and D-OSCA (3) showed no GSH-Px induction. It is unclear at this point whether OSCA does not provide selenium in a biochemically available form or if OSCA does not hydrolyze to provide the active component, selenocysteine. In summary, OSCA was a difficult compound to synthesize. However, sufficient product was obtained to perform some preliminary biological testing. It appears that both L- and D-OSCA are significantly less toxic than other selenium-containing compounds. The D- isomer, however, was less toxic than the L- isomer. The OSCA isomers appear to be as stable as OTCA, under physiological conditions, and release no detectable selenocysteine. Finally, neither L- or D-OSCA induced GSH-Px activity under the conditions tested. Due to the inherent stability of OSCA under physiological conditions, it appears that enzyme activation may be necessary for selenocysteine release. However, enzyn1e studies would be necessary to confirm this hypothesis. EXPERIMENTAL Synthesis The following instruments were used for characterization of synthesized compounds. Nuclear magnetic resonance (NMR) spectra were obtained from either an IBM instruments 200 MHz or a Varian Unity 500 MHz FT-NMR spectrometer. Deuterated solvents were purchased from Aldrich Chemical Company (Milwaukee, WI). Dimethyl selenide (Aldrich) was used as a reference in 77Se NMR studies. Infrared (lR) spectra were obtained using a Perkin-Elmer FT-IR Spectrophotometer (1600 series). Mass spectrometry analysis was conducted at the University of Utah Department of Chemistry using a Finnegan MAT 95. Elemental analyses were performed at Galbraith Laboratories (Knoxville, TN). Polarimetry data were collected using a JASCO Digital Polarimeter (Model DIP-370). Melting points were determined using a Laboratory Devices USA Mel-Temp II nlelting point instrument and are uncorrected. Thin layer chromatography (TLC) was carried out using Whatman flexible back 60 A silica gel plates with a layer thickness of 0.25 mm (Clifton, NJ). The chemicals used for synthesis were purchased from Aldrich (Milwaukee, WI) or Sigma Chemical Company (St. Louis, MO). All other materials were acquired from Fisher Scientific (Pittsburgh, PA). Prior to beginning synthetic procedures, precautions were taken to avoid reaction with oxygen in the air. This was accomplished by degassing all solvents and constantly bubbling argon through solvents during synthesis. Nanopure H20 was obtained from a Barnstead E-pure (series 582) water dispenser and was used in all reactions. L-Selenocystine (4) and D-Selenocystine (5) 66 Either the D- or L- isomer of selenocystine was synthesized by slight modification to known procedures (Klayman and Griffin, 1973; Tanaka and Soda, 1987). At room temperature, selenium powder (3.0 g, 38 mmol) was suspended in H20 (19 mL). Sodium borohydride (3.0 g, 79 mmol) was dissolved in H20 (19 mL) and slowly added to the selenium suspension with magnetic stirring. Another equivalent of selenium powder (3.0 g, 38 mmol) was added to the reaction and the mixture was stirred for 15 min. The reaction mixture was placed briefly on a steam bath (1 min) to drive the reaction to completion. Beta-chloro-D (or L)-alanine (3.2 g, 26 mmol) was dissolved in H20 at pH 9 and added dropwise to the selenium solution over 2 h. The reaction was incubated overnight at 37°C. The reaction mixture was acidified to pH 2 and hydroxylamine hydrochloride (218 mg, 3.1 nlmol) was added. The exhaust was blown through two lead acetate traps for 2 h. Vacuum filtration was performed to remove elemental selenium. The filtrate was adjusted to pH 6-6.5 and left to crystallize at 4°C for 3-5 days. The yellowish-orange crystals were collected by vacuum filtration and redissolved in HCI (1 M). Any remaining particulates were removed by vacuum filtration. The filtrate was adjusted to pH 6-6.5 and left to crystallize at 4°C for 3-5 days. The yellow crystals were collected by vacuum filtration and dried by vacuum overnight. Product yields were approximately 54% (2.3 g, 6.9 mmol), mp = 174-176°C (d), (reported, 184-185°C) (Tanaka and Soda, 1987). TLC n-BuOH:H20:Acetic Acid (3:2:1), Rf= 0.32. tH NMR (D20lNaOD, 500 MHz) 8 3.7 (dd, J= 5, 7 Hz, 1H), 3.3 (dd, J 5, 12 Hz, 1H), (dd, J= 7,12 Hz, 1 H) ppm; 77Se NMR (D20lNaOD, 500 MHz,) 8288 (s, 2Se) ppm; IR (KBr) Vmax 3500, 3000 cm-l; FABMS [M - Hr mlz 335.0; [af5o -30.9° (L), +27.3° (D); Anal. (C6H1204N2Se2): Calc% C 21.4 Obs% C 21.2 H 3.6 H3.5 N 8.3 N 8.3 Se 47.6 Se 47.9 67 NOTE: Added safety precautions were implemented when running this reaction due to the production of hydrogen selenide gas during the work-up of the reaction. Reaction vessels were sealed tightly to prevent release of H2Se into the air and the exhaust was forced through two lead acetate traps. As an additional precaution, an H2S-rated respirator was worn for escape only purposes. 2-0xo-selenazolidine-4-carboxylic acid (L-OSCA (2) and D-OSCA (3) D (or L)-Selenocystine (0.5 g, 1.5 mmol) was suspended in a flask containing 0.05 N NaOH (5 mL) and to that solution sodium borohydride (0.22 g, 5.8 mmol) was added slowly over about 10 min. The reaction mixture was stirred for an additional 10 min and then placed on an ice bath. The pH was adjusted to 6 with concentrated HCl. A 1: 1 (v/v) solution of phenyl chloroform ate (0.5 mL, 4.0 mmol) and toluene (0.5 mL) was added in 68 200 f.lL aliquots over 30 min. The pH was maintained at 7 during phenylchloroformate addition. The pH was adjusted to 8-9 (6 N NaOH) and the reaction stirred for 30 min. To isolate the product the pH of the reaction was adjusted to 7 and the solution was washed with diethyl ether. The pH of the aqueous layer was adjusted to 2 and extracted with ethyl acetate. The ethyl acetate was dried over MgS04 and concentrated. The resulting oily solid was partitioned between CHCb and H20 to dissolve all material present. The CHCb layer was extracted with H20, and the H20 fraction was concentrated and dried under vacuum. Finally, the white solid was washed with methanol, concentrated and dried under vacuum. Product yields were approximately 5% (0.025 g, 0.13 mmol). TLC n-BuOH:H20:Acetic Acid (3:2:1), Rf= 0.65. IH NMR (D20, 500 MHz) 84.5 (dd, J 6,8 Hz, IH), 3.8 (dd, J= 8,10 Hz, IH), 3.6 (dd, J= 6,10 Hz, IH) ppm; 77Se NMR (D20, 500 MHz,) 8 1353 (s, ISe) ppm; IR (KBr) Vrnax 3300,3000, 1700 em-I; FABMS [M-Hr mlz 193.9; [a]2sD -67.9° (L), +61.2° (D); Anal. (C4Hs03NSe): Calc% C 24.7 Obs% C 24.4 H2.6 H2.7 N7.2 N6.9 Se 40.7 Se 38.9 Cell Culture Procedures V79-4 Chinese hamster lung fibroblasts and hepa1c1c7 mouse hepatoma cells were purchased from American Type Culture Collection (Manassas, V A). The following items were acquired from Sigma Chemical Company (St. Louis, MO): powdered Dulbecco's modified minimum essential media (D-MEM), Eagle's modified minimum essential media (E-MEM), alpha modified minimum essential media (a-MEM), Hanks' balanced salt solution, trypsin (1 :250), phosphate buffered saline (PBS), antibiotic/antimycotic solution (1 OOX), glutathione, nicotinamide adenine dinucleotide phosphate (~-NADPH) reduced form, glutathione reductase Type III from Bakers yeast and glutathione peroxidase from bovine erythrocytes. Sodium dodecyl sulfate was purchased from BioRad Laboratories (Richmond, CA). Hydrogen peroxide (20%) and Nalgene sterile filter units containing a nylon membrane (0.2 J-lm) were purchased from Fisher Scientific (Pittsburgh, P A). Fetal Clone I serum was purchased from Hyclone Laboratories (Logan, UT). Cell Titer 96 AQueous One solution was obtained from Promega (Madison, WI) for use in MTS tetrazolium cell proliferation assays. 69 The cells were maintained at 37°C, 5% CO2-air atmosphere in a Baxter Scientific Ultra-Tech (Model WJ 301 D) incubator. Sterile handling procedures were carried out in a Forma Scientific Class II Type AB3 (Model 1186) biological safety cabinet (Marietta, OH). Cell counting for experiments was done on a Beckman Coulter particle counter (Fullerton, CA). Data analysis was done by GraphPad Prism software (version 2.0). A Hewlett Packard UV diode array spectrophotometer (Model 8452A) equipped with 70 a Peltier temperature control (Model HP8909A) was used for analysis of glutathione peroxidase induction. In addition, a Dynatech Laboratories Inc. MR 300 microplate reader (Chantilly, VA) was used, in conjunction with the Bradford assay, to measure UV absorbance for determining protein concentrations and also for measuring UV absorbance for monitoring bioreduction of MTS dye in cell toxicity experiments. Solutions The following solutions were used for nlaintenance and experimentation with cells in culture. D-MEM: To Dulbecco's modified minimum essential media was added 3.7 giL sodium bicarbonate. The pH was adjusted to 7.1 and then the solution was sterile filtered. Finally, 100/0 Fetal Clone I and 0.1 % antibioticlantimycotic was added and the solution was stored at 4°C. E-MEM: To Eagle's modified minimum essential media was added 2.2 giL sodium bicarbonate and 110 mglL sodium pyruvate. The pH was adjusted to 7.1 and then the solution was sterile filtered. Finally, 10% Fetal Clone I and 0.1 % antibioticlantinlycotic was added and the solution was stored at 4°C. ex.-MEM: To ex. modified minimum essential media was added 2.2 giL sodium bicarbonate. The pH was adjusted to 7.1 and then the solution was sterile filtered. Finally, 10% Fetal Clone I and 0.1 % antibioticlantimycotic was added and the solution was stored at 4°C. Hanks': To Hank's balanced salt solution was added 2.2 giL sodium bicarbonate. The pH was adjusted to 7.1, sterile filtered and stored at 4°C. Trypsin: Trypsin was prepared by adding 2.5 giL trypsin powder to a 10 mM phosphate buffered saline solution. The pH 71 was adjusted to 7.1, sterile filtered and stored at 4°C. MTS tetrazolium dye: The dye was defrosted before each use and diluted in E-MEM (1 :5). Maintenance Subculture. V79-4 cells were subcultured in a 75 cm2 flask by removing media (D­MEM), washing with Hanks' (10 mL), adding 1.0 mL trypsin and incubating at 37°C for 3-5 min (or until cells had sufficiently detached from the flask). The trypsin was quenched and cells suspended by the addition of media (10 mL). The resulting cell suspension was used to seed new cultures and for experimentation. Approximately 1 x 106 cells were seeded into a new culture flask. The cell cultures were passaged every 3 days. Hepalclc7 cells were subcultured in a 75 cm2 flask by removing media (a-MEM), washing with Hanks' (10 mL), adding 1.0 mL trypsin and incubating at 37°C for 5-7 min. The trypsin was quenched and the cells suspended in media (10 mL). The resulting cell suspension was used to seed new cultures and for experimentation. Approximately 5 x 105 cells were seeded into a new culture flask. The cell cultures were passaged every 7 days. Cell counting. A sample of cell suspension was syringed through a 26-gauge disposable needle to adequately disrupt cells and a 0.050 mL aliquot of cell suspension was diluted in 20 mL Coulter diluent. The cells were then counted according to Coulter protocol. 72 Drug Toxicity Assay Drug toxicity was determined using a 96-well plate assay . Wells were seeded with cell suspension and appropriate media (D-MEM or a-MEM) to a total volume of 0.2 mL. Drug toxicity was monitored over a 3-day time period. In the V79-4 cells, plates incubated for 1 day received 1 x 104 cells, 2-day plates received 2.5 x 103 cells and 3-day plates received 1 x 103 cells in each well. In the hepa1 c 1 c7 cells, plates incubated for 1 day received 2 x 104 cells, 2-day plates received 1 x 104 cells and 3-day plates received 5 x 103 cells in each well. All drug concentrations were tested with six replicates and a control (no drug treatment) on every plate. The plates were incubated overnight to allow the cells to adhere to the bottom of the wells. The media was removed from each well and media-containing drug at varying concentrations was added (0.2 mL). Control lanes received 0.2 mL media. Again the plates were incubated for either 1, 2, or 3 days. The drug media was removed and cell proliferation was determined using MTS tetrazolium dye. The dye was combined with E-MEM (l :5) and added to each well (0.12 mL). The plates were incubated for 2 hand 10% sodium dodecyl sulfate (SDS, 0.025 mL) was then added to quench the reaction. The plates were stored at 4°C overnight. Prior to taking absorbance measurements, the plates were warmed slightly at 37°C (1 0-15 min). The absorbance readings were monitored at 490 nm and the value corresponding to a blank (well containing E-MEM and dye only) was subtracted from each sample measurement. The resulting sample values were expressed as a surviving fraction by dividing drug treated wells by the average of the control lanes. General Procedures for Hydrolysis Studies 73 The following chemicals and materials were used in the OSCA hydrolysis studies. 2,4-Dinitrofluorobenzene (DNFB) and 70% perchloric acid were obtained fronl Aldrich (Milwaukee, WI). Solvents and the Whatman PVDF syringe filters were purchased from Fisher Scientific (Pittsburgh, PA). HPLC analysis of samples (Dominick, et aI., 2000) Analysis of all samples was carried out using a Hitachi model L-6200A pump equipped with a Hitachi 4250 UV -VIS detector and a AS-2000 autosampler with a Rheodyne injection valve and a 0.1 mL sample loop. The system was directed by Hitachi model D6000 version 2 (revision 06) software. Sample separation was achieved on a Rainin Dynamax 5 j.lm, 4.6 x 250 cm, C18 column fitted with a C18 guard column. Samples were eluted with an acetonitrile (MeCN)/H20 gradient each containing 0.1 % trifluoroacetic acid. Sample volumes of 0.1 mL were injected and eluted at a flow rate of 1 mL/min. Samples were eluted for 5 min at 20% MeCNI H20, followed by a steady gradient, over 32 min, to 100% MeCN. Sample elution was monitored at 365 nm. 74 Synthesis of N,N-bis-dinitrophenyl-selenocystine Selenocystine (0.2 g, 0.6 mmol) was dissolved in 6 mL 0.5 N NaOH. At room temperature, 2,4-dinitrofluorobenzene (DNFB) (0.33 mL, 2.7 nlmol) in 2.1 IT1L methanol was added with magnetic stirring. The reaction was covered to prevent exposure to light and stirred overnight. The precipitate that formed during the course of the reaction was removed by vacuum filtration. The remaining filtrate was washed with diethyl ether. The aqueous layer was acidified to pH 2 with 1 M HCI and extracted with ethyl acetate. The ethyl acetate fractions were combined, concentrated by rotary evaporation, and dried under vacuum. The product yield was >90% pure by NMR and HPLC analysis. I H NMR (acetone-d6, 200 MHz): 83.7 (dd, J= 4, 8 Hz, 2H), 5.0 (dd, J= 4,8 Hz, 1H), 7.3 (d,J 7 Hz, 1H), 8.3 (dd, J= 2,7 Hz, 2H), 8.9 (d, J= 2 Hz, 1H) ppm. The HPLC retention time was 30 min. Sample derivatization for HPLC analysis (Dominick, et aI., 2000) The 0.6 mL aqueous samples were acidified with 0.1 mL 70% perchloric acid and a 0.5 mL aliquot was used in the derivatization reaction. To this 0.5 mL aliquot, a solution (0.48 mL) of potassium hydroxide (2 M) and potassium bicarbonate (2.4 M), followed by 1 mL 1 % DNFB in ethanol (100%) was added. The reaction mixture was agitated and stored at room temperature for 24 h. The samples were then acidified by the addition of 0.3 mL 70% perchloric acid and syringe filtered using 0.45 J.!m PVDF filters. 75 Evaluation of Enzymatic Hydrolysis Rat kidneys were purchased from Pel-Freez (Rogers, AK). The following materials were purchased fronl Sigma: pyruvate kinase Type II from rabbit muscle, lactate dehydrogenase Type XXXIX from rabbit muscle, and nicotinamide adenine dinucleotide (P-NADH) reduced form. Sephadex G-200 chromatography resin was purchased from Pharmacia Biotech (Piscataway, NJ). Ultrogel AcA 34 chromatography resin was purchased from Life Technologies (Rockville, MD). 5-0xoprolinase was evaluated for its ability to hydrolyze OSCA. This enzyme was chosen due to previous studies which have demonstrated its ability to hydrolyze the sulfur-containing compound, 2-oxothiazolidine-4-carboxylic acid (OTCA) (Williamson and Meister, 1981). 5-0xoprolinase was purified using a previously published procedure (Van Der Werf, et aI., 1975). In addition, OSCA hydrolysis was also investigated in whole celllysates using V79-4 cells. The techniques and materials used were described in the general cell culture section. 5-0xoprolinase purification and evaluation of hydrolysis Solutions. Buffer A: To 50 mM Tris-HCI buffer was added, 0.1 mM ethylenediaminetetraacetic acid disodium salt (EDTA) and 5 mM 5-oxoproline. The solution was adjusted to pH 7.4 and 2 mM dithiothreitol (DTT) was added. Buffer B: In addition to the elements present in buffer A, 0.3 M NaCI was added. Assay buffer: The assay buffer contained 100 mM NaHEPES, 150 mM potassium chloride, 8 mM magnesium chloride, 2 mM phospho( enol)pyruvate, 5 mM adenosine 5' -triphosphate, and 0.2 mM 5-oxoproline. All procedures were conducted at 4°C unless otherwise stated. 76 Activity assay. Enzyme activity was measured by a coupled reaction of pyruvate kinase and lactate dehydrogenase. A 0.2 mL enzyme sample was combined with 0.39 mL assay buffer, 0.4 mL pyruvate kinase (20 units), and 10 !-lL lactate dehydrogenase (0.2 units). The samples were then warmed on a heat block at 37°C for 5 min, then 5 !-lL NADH (0.2 mM) was added. The depletion ofNADH absorbance was measured at 340 nm. Protein concentrations were determined using a modified Bradford assay (Bradford, 1976). The solution volumes were reduced to accommodate the use of a 96-well plate. 5-0xoprolinase activity was expressed as !-lmol NADH per hour per mg protein. Purification procedure. Fifty rat kidneys were thawed in 100 mL buffer A for 30 min and then homogenized. The volume of the solution was adjusted to 200 mL and centrifuged at 16,000 g for 90 min to remove unbroken cells, connective tissue, etc. The supernatant was filtered through glass wool and the volume was adjusted to 200 mL. Ammonium sulfate (20.9 g/nlL) was slowly added to the solution with stirring over 15 min. The solution was then stirred for an additional 30 min and centrifuged at 16,000 g for 90 min. The resulting precipitate was dissolved in 100 mL buffer A and dialyzed (MWCO 12-14,000) overnight against two changes of buffer A. The dialysate was applied to the top of a sephadex G-200 column (3.0 x 65 cm) and eluted with buffer A at a flow rate of approximately 20 mL/h. Approximately 100 fractions (5 mL) were collected. Active fractions were combined and applied to the top of a DEAE-cellulose 77 anion exchange column (3.0 x 18 cm). The protein solution was loaded on the column by washing through one column volume buffer A. Protein fractions were eluted with a step gradient between buffer A and buffer B. Approximately one column volun1e of 0.75, 0.125,0.15,0.225, and 0.3 M NaCI was collected at a flow rate of 50 mLIh. Approximately 200 fractions (4 mL) were collected. Active fractions were combined and ammonium sulfate (0.35 g/mL) was added over 40-45 min with stirring. The solution was centrifuged at 16,000 g for 40 min and the resulting pellet was dissolved in 25 mL buffer A. The protein solution was applied to the top of a sephadex G-200 column (3.0 x 65 cm) which was eluted with buffer A at a flow rate of 20 mL/h. The most active fraction was used for study. Hydrolysis experiments were carried out by incubating (37°C) 0.2 mL enzyme solution with 0.1 mL 50 mM OTCA and 0.3 mL assay buffer for 24 h. Control experiments consisted of samples containing only enzyme and only OTCA. Samples were then analyzed by HPLC. Evaluation of hydrolysis by whole celllysates A 75 cm2 flask containing 25 mL D-MEM was seeded with 1 x 106 cells for each treatment and incubated overnight to allow cell adhesion. The media was removed and treated cells received 25 J.lM L-selenocystine (25 mL) or 100 J.lM L-OSCA (25 mL). The control flask received fresh media (25 mL). Flasks were incubated for 1, 2, and 3 days. The media was poured off and preserved for HPLC analysis. The cells were harvested and pelleted (100 g). The resulting media supernatant (10 mL) was preserved for HPLC analysis. The cell pellet was resuspended in 0.6 mL PBS containing 0.9% NaCl and 1 mM bathophenathroline disulfonic acid (BPDS). The resuspension was treated with 50 ~l 70% perchloric acid, sonicated for 2 min, and centrifuged at 3500 g for 5 min. The resulting supernatant was preserved for HPLC analysis. Evaluation of Nonenzymatic Hydrolysis Since data could not be obtained from the enzymatic hydrolysis studies, non­enzymatic studies were carried out. Comparisons were made between L-OSCA (2), D­OSCA (3), and OTCA. 78 HPLC sample analysis. An aqueous sample (0.6 mL, pH 7) of L-OSCA (2) (90 nmol), D-OSCA (3) (90 nmol), or OTCA (~940 nmol, 50 mM) was incubated at 37°C for 1, 2, or 3 days. San1ples were then acidified by addition of 0.1 mL 70% perchloric acid and a 0.5 mL aliquot was used for derivatization. Samples were derivatized according to the procedure outlined in the general procedures for hydrolysis studies section. Selenocysteine colorimetric assay. A 50 ~L sample of L- or D-OSCA (3 mM) was combined with 50 ~L NaBH4 (50 mM in 0.1 N NaOH) and incubated at room temperature for 15 min. One milliliter of a 1 % ninhydrin solution (acetic acid) was added and incubated at 70°C for 10 min. Samples were quickly cooled to 25°C and UV absorbance was measured at 570 nm. A selenocystine standard curve was developed by plotting absorbance verses concentration (0.125 to 2 mM). This curve was used for determining selenocystine concentrations in experimental samples. Both L- and D­OSCA were incubated at 37°C, pH 7 for 12, 24, and 48 h. 79 Glutathione Peroxidase Assay Cells were trypsinized as mentioned previously and the resulting cell suspension was used for experimentation. For the V79-4 cells, a 25 cm2 flask containing 10 mL media (D-MEM) was seeded with 5.0 x 105 cells. For the hepalclc7 cells, a 25 cm2 flask containing 10 mL media (a-MEM) was seeded with 1 x 106 cells. One flask was seeded for each drug concentration, including a control flask which received no drug treatment. The seeded flasks were then incubated at 37°C overnight to allow adhesion of cells to the plate. The old media was poured off and fresh media containing drug at varying concentrations was added (10 mL). The control flask received 10 mL fresh media. After incubating for 48 h with drug solution, the media was removed and fresh media was added (10 mL). The flasks were incubated for an additional 24 h. The media was poured off and the cells were washed twice with ice cold phosphate buffered saline. The cells were trypsinized with 0.5 mL trypsin and quenched with media (5 mL). Each cell suspension was collected and pelleted by centrifugation (100 g, 5 min) at 4°C. The cell pellet was resuspended in 0.1 mM potassium phosphate buffer (pH 7.4, mL) containing 2 mM EDT A and 2 mM NaN3. The cells were then sonicated for 2 min and centrifuged (l00 g, 20 min) at 4°C. The resulting supernatant (0.1 mL) was combined with 0.5 mL potassium phosphate buffer (as mentioned above), 0.1 mL glutathione (50 mM), and 0.1 mL glutathione reductase (18 U/mL). This solution was incubated in a water bath at 25-27°C for 10 min. A 4 mM nicotinamide adenine dinucleotide phosphate (NADPH, 0.1 mL) solution was added and UV absorbance (340 nrn) was monitored for 3 min. Finally, to initiate the enzymatic reaction 0.1 mL H202 (2.2 mM) was added and absorbance was monitored for an additional 2 min. Protein concentrations were determined using a Bradford assay. Glutathione peroxidase activity was expressed as !lmol NADPH oxidized per min per mg protein. 80 APPENDIX A NMRSPECTRA I' 700 600 500 400 300 200 100 o -100 Figure A.2. 77Se NMR of L-selenocystine. Sample was prepared in D20INaOD. Dimethyl selenide was used as an external reference. ppm 00 N .N.. . 1"\ <II <II N I I .--.- ,--, , , 12 11 10 9 8 7 6 5 4 3 2 1 -0 ppm Figure A.3. IH NMR of2-DL-oxoselenazolidine-4-carboxylic acid. The spectrum of the individual isomers looks identical. Sample was prepared in D20 containing 3-(trimethylsily)propionic-2,2,3,3-d4 acid, sodium salt Peaks are labeled in hertz. 00 w 1500 1400 1300 1200 1100 1000 900 800 100 600 Figure A.4. 77Se NMR ofDL-OSCA. Sample was prepared in D20. Dimethyl selenide was used as an external reference. ppm 00 ~ .~ ~ 2: ~l g,I) I 3.0 2.0 Figure A.5. IH NMR ofN,N-bis-dinitrophenyl-DL-selenocystine. Sample was prepared in acetone-d6. Peaks at 1.2, 2.0, and 4. ° ppm are a result of ethyl acetate. 00 VI ______. ..,.. .. ___. ,.",.ll., ...~ J PI 't I ")UUUlJ)U "". *" 11.., J ..l)~ 9 8 7 6 5 4 3 2 1 Figure A.6. IH NMR of the unidentified HPLC peak collected at 31 minutes. Sample was prepared in deuterated methanol. ppm 00 0\ APPENDIXB MASS SPECTRA 100 - 80- 60- 40- 20- 315 332.0 I 330.0 I 317.1 321.1 329~ I I 319.1 I I 320 327..:..,Q I 3P."0 I I I I I I 325 330 335~0 E+ 03 r- 6.95 336.0 344.0 r-- l 339.9 I 1 I 335 340 345 Figure B.l. Mass spectrum ofL-selenocystine. FABMS [M - Hr calculated for C6H12N20480Se2 (335.0). 00 00 100 80 60 40 2°l 91.0 I 114.0 I I i ' 1." • 100 193.9 183.1 I 195.9 III~ , h, i JUI' I I It I I ~ II "1"!JI1t I t! i 200 285.9 388.8 I I 342.9 I , ! i "1"11111 1' I 300 400 Figure B.2. Mass spectrum ofDL-OSCA. FABMS [M - H]- calculated for C4HsN0380Se (194.0). E+ 05 3.86 00 \0 650.9 666.8 100 - 80- 60- 40-; 648-.,9 20- J 650 664--.-8. 652 .. 9 I 662.8 654.9 660-.,8 658.9 I II I I 660 I I 668.8 I 682.9 670.9 672.9 680-.,9 II Jj 11111 • ---r--l 670 680 I I I E+ 03 r- 5.13 Figure B.3. Mass spectrum ofN,N-bis-dinitrophenyl-L-selenocystine. FABMS [M- H]- calculated for C18H16N601280Se2 (667), \0 o 100. I ! .. "S 0:: 25 20 1S 158.9 217.9 240.1 196.0 279.9 333.1 334.1 352.1 500 600 mIz 700 800 900 1000 Figure B.4. Mass spectrum of the unidentified HPLC peak collected at 31 minutes. The data was collected by ESI using a methanol/water (50:50) and 0.1 % trifluoracetic acid matrix. . \...0.... APPENDIXC HPLC CHROMATOGRAMS 8.288 AU 8.888 ... " 38~88 Minute. •" 48.88 52.88 Figure C.l. HPLC chromatogram of the synthesized N,N-bis-dinitrophenyl-L-selenocystine. Retention time is 30 minutes. \v0,J 8.zaa AU 8.8aa, I lilli, -;-. I "NtI 8.aa J.a.aa CD ~ Figure C.2. HPLC chromatogram ofL-OSCA. Retention time of unidentified peak at 31 minutes. '0 ~ a.28a AU ...... a.a88'" 11111 I I N CIt 38.8a Minutes 48~8a 52.8a Figure C.3. HPLC chromatogram ofL-OSCA spiked with L-selenocystine. The smaller peak at 30 minutes corresponds to N,N-bis-dinitrophenyl-selenocystine. The larger peak at 31 minutes represents a combination of the peak commonly found in the OSCA samples and the spiked selenocystine. \0 V1 REFERENCES Albert, A. Chenlical Aspects of Selective Toxicity. Nature 1958, 182,421-423. Allaway, W., Kubota, J., Lossee, M., Roth, M. Selenium, Molybdenum, and Vanadium in Human Blood. Arch. Environ. Health 1968, 16, 342-348. Antonucci, A., Foppoli, C., De Marco, C., Cavallini, D. Inhibition of Protein Synthesis in Rabbit Reticulocytes by Selenaproline. Bull. Molec. Bioi. Med. 1977, 2, 80-84. Behne, D., Kyriakopoulos, A., Meinhold, H., Kohrle, J. Identification of Type-1 Iodothyronine 5-Deiodinase as a Selenoenzyme. Biochem. Biophys. Res. Comm. 1990, 173, 1143-1149. Berry, M., Banu, L., Chen, Y., Mandel, S., Kieffer, J., Harney, J., Larsen, P. Recognition ofUGA as a Selenocysteine Codon in Type 1 Deiodinase Requires Sequences in the 3' Untranslated Region. Nature 1991,353,273-276. Berry, M., Banu, L., Harney, J., Larsen, P. Functional Characterization of the Eukaryotic SECIS Elements with Direct Selenocysteine Insertion at UGA Codons. EMBO 1993,12, 3315-3322. Boettcher, B., Meister, A. Synthesis ofL-2-0xothiazolidine-4-carboxylic Acid. Anal. Biochem. 1984, 138,449-450. Bradford, M. A Rapid and Sensitive Method for the Quantitation of Microgram Quantities of Protein Utilizing the Principle of Protein-Dye Binding. Anal. Biochem. 1976, 72, 248-254. Burk, R. Seleniunl and Cancer: Meaning of Serum Selenium Levels. J. Nutr. 1986, 116, 1584-1586. Chen, G., Fang, L., Inouye, M. Effects of the Relative Position of the UGA Codon to the Unique Secondary Structure in the fdhF mRNA on its Decoding by Selenocysteinyl tRNA in Escherichia Coli. J Bioi. Chem. 1993,268, 23128-2313 Christensen, M., Burgener, K. Dietary Selenium Stabilizes Glutathione Peroxidase mRNA in Rat Liver. J. Nutr. 1992,122, 1620-1626. 97 Clark, L., Combs, G., Turnbull, B., Slate, E., Chalker, D., Chow, J., Davis, L., Glover, R., Graham, G., Gross, E., Krongrad, A., Lesher, l, Park, H., Sanders, B., Smith, C., Taylor, J. Effects of Selenium Supplementation for Cancer Prevention in Patients with Carcinoma of the Skin. J. Am. Med. Assoc. 1996,276, 1957-1963. Clinton, M., Jr. Selenium Fume Exposure. J. Ind. Hyg. Toxicol. 1947, 29, 225-226. Combs, G., Combs, S. The Nutritional Biochemistry of Selenium. 1984, Palo Alto, Annual Reviews, Inc. Combs, G., Combs, S. The Role of Selenium in Nutrition. 1986, Orlando, Academic Press, Inc. Daher, R., Van Lente, F. Characterization of Selenocysteine Lyase in Human Tissues and its Relationship to Tissue Selenium Concentrations. J. Trace Elem. Electrolytes Health Dis. 1992,6, 189-194. Deagen, l, Butler, l, Beilstein, M., Whanger, P. Effects of Dietary Selenite, Selenocystine and Selenomethionine on Selenocysteine Lyase and Glutathione Peroxidase Activities and on Selenium Levels in Rat Tissues. J. Nutr. 1987, 117, 91-98. De Marco, C., Busiello, V., Di Girolamo, M., Cavallini, D. Selenaproline and Protein Synthesis. Biochim. Biophys. Acta 1977,478, 156-166. Dickson, R., Tomilinson, R. Selenium in Blood and Human Tissues. Clin. Chim. Acta 1967, 16, 311-321. Diskin, C., Tommasso, C., Alper, J., Glaser, M., Fliegel., S. Long-term Selenium Exposure. Arch. Intern. Med. 1979, 139, 824-826. Dominick, P., Cassidy, P., Roberts, l Improved Chromatographic Quantitation of Cysteine, Glutathione, and Other Thiolamines of Biological Importance. Manuscript in Preparation. EI-Bayoumy, K., Upadhyaya, R., Desai, D., Anlin, S., Hoffmann, D., Wynder, E. Effects of 1 ,4-Phenylenebis(methylene )selenocyanate, Phenethyl Isothiocyanate, Indole-3- carbinol, and d-Limonene Individually and in Combination on the Tumorigenicity of Tobacco-Specific Nitrosamine 4-(Methylnitrosamino )-I-(3-pyridyl)-I-butanone in AlJ Mouse Lung. Anticancer Res. 1996, 16,2709-2712. 98 Esaki, N., Nakamura, T., Tanaka, H., Soda, K. Selenocysteine Lyase, a Novel Enzyme that Specifically Acts on Selenocysteine. J. Bioi. Chern. 1982, 257, 4386-4391. Esaki, N., Soda, K. Selenocysteine. Meth. Enzyrnol. 1987, 143, 148-149. Flohe, L. (1982). Glutathione Peroxidase Brought into Focus. In: Free Radicals in Biology 1982, 5, 223-254. Academic Press. Forchhamnler, K., Leinfelder, W., Boesmiller, K., Veprek, B., Bock, A. Selenocysteine Synthase from Escherichia Coli. Nucleotide Sequence of the Gene (seIA) and Purification of the Protein. J. Bioi. Chern. 1991,266, 6318-6323. Forchhammer, K., Bock, A. Selenocysteine Synthase from Escherichia Coli. Analysis of the Reaction Sequence. J. BioI. Chern. 1991, 266, 6324-6328. Franke, K., Painter, E. Selenium in Proteins from Toxic Foodstuffs. J. Nutr. 1935, 10, 599-611. Franke, K., Moxon, A., Poley, W., Tully, W. Monstrosities Produced by the Injection of Selenium Salts into Hens' Eggs. Anat. Rec. 1936, 65, 15-22. Gaitonde, M. A Spectrophotometric Method for the Direct Determination of Cysteine in the Presence of Other Naturally Occurring Amino Acids. Biochern. J. 1967, 104,627- 633. Ganther, H. Pathways of Selenium Metabolism Including Respiratory Excretory Products. J. Arn. Coli. Toxico!. 1986,5, 1-5. Ganther, H. Selenium Metabolism, Selenoproteins and Mechanisms of Cancer Prevention: Complexities with Thioredoxin Reductase. Carcinogenesis 1999,20, 1657- 1666. Gladyshev, V., Jeang, K., Wootton, 1., Hatfield, D. A New Hunlan Selenium-Containing Protein. J. BioI. Chern. 1998, 273, 8910-8915. Gunther, W. Hypophosphorous Acid, a Novel Reagent for the Reduction ofDiselenides and the Selenol-Catalyzed Reduction of Disulfides. J. Org. Chern. 1966,31,1202-1205. Guo-jie, Breslow, E., Meister, A. The Amino Acid Sequence of Rat Kidney 5-0xo-L­prolinase Determined by cDNA Cloning. J. Bioi. Chern. 1996,271,32293-32300. Hatfield, D., Diamond, A., Dudock, B. Opal Suppressor Serine tRNAs from Bovine Liver Form Phosphoseryl-tRNA. Proc. Natl. Acad. Sci. USA 1982,79,6215-6219. 99 Heider, J., Baron, C., Bock, A. Coding from a Distance: Dissection of the mRNA Detenninants Required for the Incorporation of Selenocysteine into Protein. EMBO 1992, 11, 3759-3766. Hill, K., Lloyd, R., Yang, J., Read, R., Burk, R. The cDNA for Rat Selenoprotein P Contains 10 TGA Codons in the Open Reading Frame. J. Bio!. Chem. 1991,266,10050- 10053. Hori, K., Hatfield, D., Maldarelli, F., Lee, B., Clouse, K. Selenium Supplementation Suppresses Tumor Necrosis Factor alpha Induced HIV Virus Type 1 Replication in vitro. AIDS Res. Hum. Ret. 1997,13,1325-1331. Horvathova, M., Jahnova, E., Gazdik, F. Effect of Selenium Supplementation in Asthmatic Subjects on the Expression of Endothelial Cell Adhesion Molecules in Culture. Bioi. Trace Elem. Res. 1999, 69, 15-26. Ip, C., EI-Bayoumy, K., Upadhyaya, P., Ganther, H., Vadhanavikit, S., Thompson, H. Comparative Effect of Inorganic and Organic Selenocyanate Derivatives in Mammary Cancer Chemoprevention. Carcinogenesis 1994, 15, 187-192. Ip, C., Lisk, D., Ganther, H., Thompson, H. Triphenylselenonium and Diphenylselenide in Cancer Chemoprevention: Comparative Studies of Anticarcinogenic Efficacy, Tissue Selenium Levels and Excretion Profile. Anticancer Res. 1997, 17, 3195-3200. Jacobs, M., Jansson, B., Griffin, A. Inhibitory Effects of Selenium on 1,2- Dimethylhydrazine and Methylazoxymethanol Acetate Induction of Colon Tumors. Cancer Lett. 1977, 2, 133-138. Jiang, C., Jiang, W., Ip, C., Ganther, H., Lu, J. Selenium-Induced Inhibition of Angiogenesis in Mammary Cancer at Chemopreventive Levels of Intake. Mol. Carcinog. 1999, 26, 213-225. Kaneko, T., Shimokobe, T., Ota, Y., Toyokawa, E., Inui, T., Shiba, T. Syntheses and Properties of 2-0xothiazolidine-4-Carboxylic Acid and Its Derivatives. Chem. Soc. Jap. Bull. 1964, 37, 242-244. Kellof, G., Crowell, H., Hawk, E., Steele, V., Lubet, R., Boone, C., Covey, J., Doody, L., Omenn, G., Greenwald, P., Hong, W., Parkinson, D., Bagheri, D., Baxter, G., Blunden, M., Doeltz, M., Eisenhauer, K., Johnson, K., Knapp, G., Longfellow, D., Malone, W., Nayfield, S., Seifried, H., Swan, L., Sigman, C. Strategy and Planning for Chemopreventive Drug Development: Clinical Development Plans II. J. Cell. Biochem. Suppl. 1996, 26, 54-71. 100 Klayman, D., Griffin, T. Reaction of Selenium with Sodium Borohydride in Protic Solvents. A Facile Method for the Introduction of Selenium into Organic Molecules. J Am. Chem. Soc. 1973,95, 197-199. Lane, H., Teer, P., Dukes, J., Johnson, 1., White, M. The Effect of Four Chemical Forms of Selenium on Mammary Tumor Incidence in BALB/c Female Mice Treated with 7,12- Dimethylbenz[a] anthracene. Cancer Lett. 1990, 50, 39-44. Lane, H., Strength, R., Johnson, 1., White, M. Effect of Chemical Form of Seleniun1 on Tissue Glutathione Peroxidase Activity in Developing Rats. J Nutr. 1991, 121, 80-86. Lee, B., Park, S., Park, J., Chittum, H., Hatfield, D. Molecular Biology of Selenium and Its Role in Human Health. Mol. Cells 1996, 6, 509-520. Leinfelder, W., Zehelin, E., Mandrand-Berthelot, M., Bock, A. Gene for a Novel tRNA Species that Accepts L-Serine and Cotranslationally Inserts Selenocysteine. Nature 1988,331,723-725. Lescure, A., Gautheret, D., Carbon, P., Krol, A. Novel Selenoproteins Identified in silico and in vivo by using a Conserved RNA Structural Motif. J BioI. Chem. 1999,274, 38147-38154. Lindberg, P., Jacobsson, S. Relationship Between Selenium Content of Forage, Blood and Organs of Sheep, and Lamb Mortality Rates. Acta Vet. Scand. 1970, 11,49-58. McKenzie, R., Rea, H., Thomson, C., Robinson, M. Selenium Concentration and Glutathione Peroxidase Activity in Blood of New Zealand Infants and Children. Am. J Clin. Nutr. 1978,31,1413-1418. Medina, D., Morrison, D. Current Ideas on Selenium as a Chemopreventive Agent. Pathol. Immunopathol. Res. 1988, 7, 187-199. Meister, A., Griffith, 0., Williamson, J. 5-0xo-L-prolinase from Rat Kidney. Meth. Enzymol. 1985, 113,445-458. Mustacich, D., Powis, G. Thioredoxin Reductase. Biochem. J 2000, 346, 1-8. National Research Council (U.S.) Subcommittee on the 10th Edition of the RDA's Recommended Daily Allowances. 1989, Washington D.C., National Academy Press. Ochi, T., Miyaura, S. Cytotoxicity of an Organic Hydroperoxide and Cellular Antioxidant Defense System Against Hydroperoxides in Cultured Mammalian Cells. Toxicol. 1989, 55, 69-82. 101 Ochi, T. Effects of an Organic Hydroperoxide on the Activity of Antioxidant Enzymes in Cultured Mammalian Cells. Toxicoi. 1990,61,229-239. Oiry, l, Puy, J., Mialocq, P., Clayette, P., Fretier, P., Jaccard, P., Dereuddre-Bosquet, N., Dormont, D., Imbach, J. Synthesis and in Vitro Anti-HIV Activity in Human Monocyte- Derived Macrophages of 2-0xothiazolidine-4(R)-carboxylic Acid Derivatives. J. Med. Chern. 1999, 42, 4733-4740. Olson, 0., Novacek, E., Whitehead, E., Palmer, I. Investigations on Selenium Wheat. Phytochemistry 1970, 9, 1181-1188. Oster, 0., Schmiedel, G., Prellwitz, W. The Organ Distribution of Selenium in German Adults. Bioi. Trace Eiem. Res. 1988, 15,23-45. Patterson, Milstrey, R., Stokstad, E. Effect of Selenium in Preventing Exudative Diathesis in Chicks. Proc. Soc. Exp. Bioi. Med. 1957, 95, 617-624. Pauling, L. College Chemistry. 1964, San Francisco, W.H. Freeman and Company. Pence, B., Delver, E., Dunn, D. Effects of Dietary Selenium on UVB-Induced Skin Carcinogenesis and Epidermal Antioxidant Status. J. Invest. Dermatol. 1994, 102, 759- 761. Promega Technical Bulletin, # 245. 1996, http://www.promega.com Read, R., Bellew, T., Yang, l, Hill, K., Palmer, I., Burk, R. Selenium and Amino Acid Composition of Selenoprotein P, the Major Selenoprotein in Rat Serum. J. Bioi. Chern. 1990,265,17899-17905. Reddy, B., Rivenson, A., Kulkarni, N., Upadhyaya, P., EI-Bayoumy, K. Chemoprevention of Colon Carcinogenesis by the Synthetic Organoselenium Compound 1,4-Phenylenebis(methylene)selenocyanate. Cancer Res. 1992, 52, 5635-5640. Reddy, B., Wynn, T., EI-Bayoumy, K., Upadhyaya, P., Fiala, E., Rao, C. Evaluation of Organoselenium Compounds for Potential Chemopreventive Properties in Colon Cancer. Anticancer Res. 1996, 16, 1123-1128. Reilly, C. Selenium in Food and Health. 1996, London, Blackie Academic and Professional. Ren, B., Huang, W., Akesson, B., Ladenstein, R. The Crystal Structure of Seleno­Glutathione Peroxidase from Human Plasma at 2.9A. Resolution. J. Mol. BioI. 1997, 268, 869-885. 102 Ricetti, M., Guidi, G., Bellisola, G., Marrocchella, R., Rigo, A., Perona, G. Selenium Enhances Glutathione Peroxidase Activity and Prostacyclin Release in Cultured Human Endothelial Cells. Bioi. Trace Elem. Res. 1994, 46, 113-123. Sappey, C., Legrand-Poels, S., Best-Belpomme, M., Favier, A., Rentier, B., Piette, J. Stimulation of Glutathione Peroxidase Activity Decreases HIV Type 1 Activation after Oxidative Stress. AIDS Res. Hum. Ret. 1994, 10, 1451-1459. Schrauzer, G., Ishmael, D. Effects of Selenium and of Arsenic on the Genesis of Spontaneous Mammary Tumors in Inbred C3 Mice. Ann. Clin. Lab. Sci. 1974,4,411- 447. Shamberger, R., Frost, D. Possible Protective Effect of Selenium Against Human Cancer. Can. Med. Assoc. J. 1969, 100, 682. Shen, H., Yang, C., Ong, C. Sodium Selenite-Induced Oxidative Stress and Apoptosis in Human Hepatoma HepG2 Cells. Int. J. Cancer. 1999, 81, 820-828. Shi, C., Chua, S., Lee H., Ong, C. Inhibition of Aflatoxin B I-DNA Binding and Adduct Formation by Selenium in Rats. Cancer Lett. 1994, 82,203-208. Shimada, T., El-Bayoumy, K., Upadhyaya, P., Sutter, T., Guengerich, F., Yamazaki, H. Inhibition ofHunlan Cytochrome P450-catalyzed Oxidations of Xenobiotics and Procarcinogens by Synthetic Organoselenium Compounds. Cancer Res. 1997, 57,4757- 4764. Sinha, R., Kiley, S., Lu, J., Thompson, H., Moraes, R., Jaken, S., Medina, D. Effects of Methyselenocysteine on PKC Activity, cdk2 Phosphorylation and gadd Gene Expression in Synchronized Mouse Mammary Epithelial Tumor Cells. Cancer Lett. 1999,146, 135- 145. Sinkula, A., Yalkowsky, S. Rationale for Design of Biologically Reversible Drug Derivatives: Prodrugs. J. Pharm. Sci. 1975, 146, 181-210. Sohn, 0., Fiala, E., Upadhyaya, P., Chae, Y., EI-Bayoumy, K. Comparative Effects of Phenylenebis(methylene )selenocyanate Isomers on Xenobiotic Metabolizing Enzymes in Organs of Female CD Rats. Carcinogenesis 1999, 20, 615-621. Song, L., Kosmeder II, J., Lee, S., Gerhauser, C., Lantvit, D., Moon, R., Moriarty, R., Pezzuto, J. Cancer Chemopreventive Activity Mediated by 4'-Bromoflavone, a Potent Inducer of Phase II Detoxification Enzymes. Cancer Res. 1999, 59, 578-585. Stadtman, T. Selenocysteine. Ann. Rev. Biochem. 1996,65,83-100. 103 Syburra, C., Passi, S. Oxidative Stress in