1...: «v. :5: :, ... ~ : :r a . . H . ...a......£ .. 2 .02... .. 2.7.. . Q "II I! .P ' 11' rlq III III I- I I II | | I: ' LIBRARY Michigan State University Molecular Basis of Estrogenic Endocrine Disruptor-Estrogen Receptor Interactions: A Comparison Among Species By Jason Bruce Matthews A DISSERTATION Submitted to Michigan State University In partial fulfillment of the requirements For the degree of DOCTOR OF PHILOSOPHY Department of Biochemistry and Molecular Biology 2001 Abstract Molecular Basis of Estrogenic Endocrine Disruptor-Estrogen Receptor Interactions: A Comparison Among Species Jason Bruce Matthews In recent years, there has been heightened concern that environmental exposure to estrogen mimicking chemicals, known as estrogenic endocrine disruptors (EEDs) may cause adverse health effects in humans and wildlife. Many of the effects of EEDs are mediated through the estrogen receptor (ER). Although the physiological actions of the ER are conserved among species, variation within the amino acid sequence of ligand binding domains suggests that species may exhibit different responses and sensitivities to EEDs. Species-specific responses to EEDs were first examined in competitive binding assays using glutathione-S-transferase (GST)-ER fusion proteins from several different species. Fusion proteins consisted of the ER D, E, and F domains of human alpha (GST- hERadef), mouse alpha (GST-mERadef), chicken (GST-cERdef), green anole (GST- aERdef) and rainbow trout ERs (GST-rtERadef). Although, the fusion proteins exhibited similar binding preferences for many EEDs, several differences were observed. The GST-rtERadef, which has the greatest amino acid sequence variability in its ligand binding domain compared to hERadef, exhibited the most striking differences compared to the other GST-ERs. The ability of several of these EEDs to induce gene expression mediated by the various ERs was then examined in MCF—7 cells transiently transfected with Gal4-ERdef chimeric receptors. Overall, the data in the gene expression assay correlated with the competitive binding results. However, there were examples where EEDs bound to GST-ERs but were unable to significantly induce ER-mediated gene expression. Intriguingly, the E2-induced response mediated by Gal4-rtERadef was 2 orders of magnitude lower compared to the other receptors examined. Much of this effect was due to temperature, since when compared to hERa the 280-fold difference at 37°C was reduced to only 9—fold at 20°C. A comparison of their ligand binding pockets identified two conservative amino acid substitutions in rtERa (M317, 1496) and hEROL ‘ (L349, M528). The effect of these substitutions on ligand binding and transactivation was examined by constructing reciprocal mutants. The rtERadef M317L:I496M mutant exhibited a hERa phenotype with increased E2 binding affinity and transactivation ability at higher temperatures. The hERa L349M:M5281 mutant also exhibited a modest trend towards adopting the rtERa phenotype. The lack of a complete exchange of phenotypes indicates that factors outside of the ligand binding pocket are also involved. Taken together these results demonstrate that ERs from different vertebrate species exhibit different affinities and transactivation responses to EEDs. Since few differences were observed, these data do not preclude the use of a single surrogate ER to examine estrogenic responses for all vertebrate species. The present report also highlights the impact of temperature when comparing functional characteristics of proteins from poikilothermic species, such as rainbow trout, and humans. Acknowledgements I would like to thank my thesis advisor, Dr. Timothy Zacharewski, for all his help and valuable support throughout my tenure in his laboratory. The countless opportunities he has given me have been greatly appreciated. Several individuals contributed to the results presented thoughout my dissertation and I would to acknowledge them for their help. Cora Fong, Mark Wilke, Kirsten Fertuck, and Drs. Trine Celius and Yue-Wem Huang made contributions to data presented in Chapters 3, 5, and 8. Dr. Janine Clemons made significant contributions to the results presented in Chapter 6. I also thank other members of the laboratory for their support and assistance, with special thanks to Mark Fielden for his continued friendship and for making graduate school bearable and at times quite an enjoyable experience. I thank my committee members Drs. Don Jump, John Wilson, Leslie Kuhn, Larry Fischer, and John LaPres for their guidance and valuable comments. I would also like to thank my family for their support and for always believing in me throughout all my endeavours. Last but not least, I thank Trine Celius for her love, support, and companionship. iv TABLE OF CONTENTS LIST OF TABLES ........................................................................................................... viii LIST OF FIGURES ............................................................................................................. x LIST OF ABBREVIATIONS .......................................................................................... xiv CHAPTER 1 ........................................................................................................................ l OBJECTIVES ..................................................................................................................... 1 Rationale ......................................................................................................................... 2 Hypothesis ....................................................................................................................... 3 Aims ................................................................................................................................ 3 References ....................................................................................................................... 4 CHAPTER 2 ........................................................................................................................ 6 LITERATURE REVIEW .................................................................................................... 6 Hormones and Cellular Regulation ................................................................................. 7 Estrogen Receptor Agonist and Antagonist Ligands ...................................................... 8 Environmental Endocrine Disruptors ............................................................................ 10 Nuclear Receptors ......................................................................................................... 14 Estrogen Receptor 0t, [3, and y ....................................................................................... 22 Phylogenetic Comparison of ERs ................................................................................. 28 Coregulators .................................................................................................................. 32 Estrogen-Responsive Genes.................‘ ......................................................................... 36 Species-Specific Binding Preferences and Transactivation Responses to EEDs .......... 36 CHAPTER 3 ...................................................................................................................... 52 DIFFERENTIAL BINDING AFFINITIES OF PCBS, HO-PCBS AND AROCLORS WITH RECOMBINANT HUMAN, RAINBOW TROUT (ONCORHYNCHUS MYKISS) AND GREEN ANOLE (ANOLIS CAROLINENSIS) ESTROGEN RECEPTORS USING A SEMI-HIGH THROUGHPUT COMPETITIVE BINDING ASSAY .............................................................................................................................. 52 Abstract ......................................................................................................................... 53 Introduction ................................................................................................................... 53 Materials and Methods .................................................................................................. 54 Results ........................................................................................................................... 56 References ..................................................................................................................... 65 CHAPTER 4 ...................................................................................................................... 67 DIFFERENTIAL ESTROGEN RECEPTOR BINDING OF ESTROGEN IC SUBSTANCES: A SPECIES COMPARISON ................................................................. 67 Abstract ......................................................................................................................... 68 Introduction ................................................................................................................... 68 Materials and Methods .................................................................................................. 69 Results ........................................................................................................................... 71 Discussion ..................................................................................................................... 76 References ..................................................................................................................... 77 CHAPTER 5 ...................................................................................................................... 80 ABILITY OF SEVERAL NATURAL AND SYNTHETIC ESTROGENIC CHEMICALS TO INDUCE GENE EXPRESSION MEDIATED BY ESTROGEN RECEPTORS FROM SEVERAL VERTEBRATE SPECIES ......................................... 80 Abstract ......................................................................................................................... 81 Introduction ................................................................................................................... 83 Materials and methods .................................................................................................. 85 Results ........................................................................................................................... 91 Discussion ................................................................................................................... 102 CHAPTER 6 .................................................................................................................... 109 RECIPROCAL MUTAGENESIS BETWEEN HUMAN (1(L349, M528) AND RAINBOW TROUT 0t (M317, 1496) ESTROGEN RECEPTOR RESIDUES DEMONSTRATES THEIR IMPORTANCE IN LIGAND BINDING AND GENE EXPRESSION AT DIFFERENT TEMPERATURES ................................................... 109 Abstract ....................................................................................................................... 1 10 Introduction ................................................................................................................. 1 1 1 Materials and Methods ................................................................................................ 114 Results ......................................................................................................................... 120 Discussion ................................................................................................................... 136 References ................................................................................................................... 142 CHAPTER 7 .................................................................................................................... 145 IN VIT R0 AND IN VIVO INTERACTIONS OF BISPHENOL A AND ITS METABOLITE, BISPHENOL A GLUCURONIDE, WITH ESTROGEN RECEPTORS 0: AND B ......................................................................................................................... 145 Abstract ....................................................................................................................... 146 Introduction ................................................................................................................. 146 Experimental Procedures ............................................................................................ 147 Results ......................................................................................................................... 148 Discussion ................................................................................................................... 151 References ................................................................................................................... l 52 CHAPTER 8 .................................................................................................................... 155 PURIFICATION AND CHARACTERIZATION OF THE RTEROL D, E, AND F DOMAINS ...................................................................................................................... 155 Abstract ....................................................................................................................... 156 Materials and Methods ................................................................................................ 159 Results ......................................................................................................................... 164 Discussion ................................................................................................................... 172 References ................................................................................................................... 1 75 CHAPTER 9 .................................................................................................................... 177 CONCLUSIONS AND FUTURE DIRECTIONS .......................................................... 177 vi Conclusions ................................................................................................................. 178 Future Directions ......................................................................................................... 179 References ................................................................................................................... 1 8 1 vii Chapter 2 1. Chapter 3 1. Chapter 4 1. Chapter 5 LIST OF TABLES Comparison of the affinity for 17B-estradiol among different vertebrate ERs. PCR primers used in the cloning and recloning of the green anole and rainbow trout estrogen receptor (ER) D, E, and F domains and in the construction of the GST-ERdef containing plasmids. Comparison of the amino acid sequences from the D, E, and F domains of estrogen receptor from rainbow trout taken from four different sources. Summary of the ability of 44 polychlorinated biphenyls to compete with [3H] l7B-estradiol for binding to recombinant fusion proteins containing the human, green anole and rainbow trout estrogen receptor D, E and F domains. Receptor binding activity of 9 hydroxylated polychlorinated biphenyls to each the recombinant fusion proteins containing the human, green anole and rainbow trout estrogen receptor D, E, and F domains Effects of dimethyl sulphoxide on the ability of select polychlorinated biphenyls to competitively displace [3H]17B- estradiol from recombinant fusion proteins containing the human, green anole and rainbow trout estrogen receptor D, E and F domains. Comparison of the dissociation constants (Kd) of the GST-ERdef fusion proteins with reported values. IC50 values and relative binding affinities (RBAs) of natural and synthetic estrogenic chemicals for GST-ERdef fusion proteins following a 24 h incubation at 4°C. viii 38 55 57 60 61 62 72 73 Chapter 6 1 Chapter 7 1. ECso values of luciferase reporter gene activity mediated by the ERs from several different species following a 24 h exposure to natural and synthetic estrogenic chemicals 37°C. Hormone binding of wild-type and mutant ERs at different temperatures. GST-ERdef fusion proteins were incubated with increasing concentrations of [3H]17B-estradiol at 4, 20, 30, or 37°C for 24 h, 2 h, l h, or 30 min, respectively. Competitive binding and transactivation abilities of wild-type and mutant receptors in response to estrogenic chemicals. Summary of the ICso values of the competitive binding of 17B- estradiol, bisphenol A and bisphenol A glucuronide to GST- hERadef, recombinant hERB and mouse uterine cytosol. Summary of the ability of 17l3-estradiol, bisphenol A and bisphenol A glucuronide to induce estrogen receptor-mediated gene expression. Effect of oral gavage of bisphenol A on blotted and dry uterine weight in rats. Microscopic changes in the uterus following oral gavage of bisphenol A. Effect of subcutaneous administration of bisphenol A on blotted and dry weight in rats. Microscopic changes in the uterus following subcutaneous administration of bisphenol A. ix 98 123 133 149 I49 150 150 151 151 -' r- vvr. I‘ll!" Chapter 2 1. 10. 11. 12. 13. Chapter 3 LIST OF FIGURES Negative and positive feedback regulation of estrogens and progesterone on the hypothalamic-pituitary-ovarian axis. Estrogenic (agonist) and antiestrogenic (antagonist) ligands of the ER. Suspected estrogenic endocrine disruptors (EEDs) and their sources. Proposed machanism of action of (A) class I and (B) class 11 nuclear receptors. Conserved domain structure of nuclear receptors. Hypothetical "zinc fingers" in the DNA binding domain of nuclear receptors. Comparison of the conserved canonical fold of nuclear receptor ligand binding domains. The ligand induced conformational change of human retinoid X receptor (RXR). Interactions made by 17B-estradiol in the human estrogen receptor 0t ligand binding cavity. Position of helix l2 (AF-2) in (A) raloxifene, (B) genistein complexes with rat and human estrogen receptor [3, respectively. (C) Position of helix 12 in the human estrogen receptor a-17B- estradiol complex. Phylogentic analysis of estrogen receptor (ER) sequences. Structural based alignment of selected vertebrate estrogen receptor D, E and F domains. Proposed model for transactivation by ER including interactions with coactivators and corepressors. 11 13 16 18 20 21 23 25 27 30 31 35 Chapter 4 l. 4. Chapter 5 1. Partial nucleotide sequence of the green anole ER. A comparison of the amino acid sequence of human, green anole and recloned rainbow trout ER D, E and F domains. (A) A schematic representation the GST-ER fusion proteins expressed in bacteria. (B) SDS- PAGE analysis of the GST- ERdef fusion proteins purified using GSH affinity chromatography. Saturation analysis of (A) GST-hERadef, (B) GST-aERdef, and (C) GST-rtERadef fusion proteins. Classification of PCB interactions with GST-ERdef fusion proteins. Representative competitive binding curves for E2 to GST- hEROLdef fusion protein and unlabeled PCB 47 with GST- rtERadef, GST-hERadef and GST-aERdef fusion proteins. (A) Competitive binding of E2 and PCB 184 at 4%, 10% and 20% DMSO in assay buffer. (B) The effect of increasing DMSO concentration on the binding of [3H]17B-estradiol to the GST-ER DEF fusion protein. Alignment of the estrogen receptor D, E and F domains from human (hERadef), mouse (mERoedef), chicken (cERdef), green anole (aERdef) and rainbow trout (rtERozdef). (A) A schematic representation the GST-ERdef fusion proteins expressed in bacteria. (B) SDS-PAGE analysis of the GST- ERdef fusion proteins purified using GSH affinity chromatography. Representative competitive binding curves of selected test chemicals to GST-hERocdef GST-mERocdef, GST-cERdef, GST- aERdef, and GST-rtERordef fusion proteins. Dendogram generated from the aligned ERdef sequences. Schematic representation of Gal4-ER chimeric receptors transiently transfected into MCF—7 human breast cancer cells. xi 56 57 58 59 59 60 63 71 72 74 75 92 Chapter 6 1. Comparison of the ER-induced transactivation ability mediated by Gal4-ERs. Differential expression levels of SRC coactivators in MCF—7 cells. The rtERa DEF binds human SRC-2 and SRC-3. Mycotoxins and phytoestrogens differ in their ability to induce gene expression mediated by both ER isofroms and ERs of different vertebrate species. Xenobiotics differ in their ability to induce gene expression mediated by both ER isofonns and ERs of different vertebrate species. Comparison of ICso and ECSO values among different vertebrate ERs. Alignment of the estrogen receptor D, E and F domains from human (hERocdef; GenBank accession no. M11457) and rainbow trout (rtERadef; GenBank accession no. AJ242741). The affinity of wild-type and mutant receptors for 17B-estradiol at different temperatures. Temperature effects on wild-type human and rainbow trout estrogen receptor-mediated induction of reporter gene activity in the presence of 17B-estradiol. Temperature effects on wild-type and mutant estrogen receptor- mediated 17l3-estradiol induced reporter gene activity. Temperature effects on wild-type and mutant estrogen receptor- mediated 17B-estradiol induced reporter gene activity. Western blot analysis of wild-type and mutant Gal4-ER expression level. Effect of reciprocal mutagenesis on the competitive binding ability of coumestrol. xii 93 95 96 100 101 104 121 122 125 127 128 130 134 Chapter 7 1. Chapter 8 1 . 2. Transactivation abilities of wild-type and mutant receptors in response to coumestrol. Competitive binding of BPA and BPA glucuronide to (A) the mouse uterine ER, (B) GST-hERadef and (C) recombinant hERB. Effect of BPA and BPA glucuronide on reporter gene expression. Effect of BPA and BPA-G cotreatment on E2-induction of reporter gene activity. SDS-PAGE analysis of GST-rtERa purification. GroEL co-purifies with GST-ERdef fusion proteins. Comparison of the activity of the GST-rtERa and his_rtEROt proteins. SDS-PAGE analysis of his_rtEROt purification. Influence of the D domain on the solubility of his_rtER01. xiii 135 149 149 150 165 166 167 169 170 AF-l AF-2 AR ACTH CBP COUP-TF DBD DDT DES D—box DR DRIP EDC EDSTAC E2 ER ECSO ERE EED E 1 E3 ERR FSH GnRH GR GST HAT HDAC HO-PCB 1Cso LBD LH MAPK NGFI-B NCoR NR PBP pCAF PCB PPAR PAS P-box LIST OF ABBREVIATIONS Ligand-independent activation function 1 Ligand-dependent activation function 2 Androgen receptor Adrenocorticoid hormone CAMP responsive element-binding (CREB)- binding protein Chicken ovalbumin upstream promoter DNA binding domain 1 ,1 ,1-trichloro-2,2-bis(p-chlorophenyl)ethane Diethylstibestrol Distal-box Direct repeat VDR interacting protein Endocrine disrupting chemical Endocrine disruptor screening and testing advisory committee 17B-estradiol Estrogen receptor 50% maximal response concentration Estrogen response element Estrogenic endocrine disruptor Estrone Estriol Estrogen‘related receptor Follicle-stimulating hormone Gonadotropin-releasing hormone Glucocorticoid receptor Glutathione-S-transferase Histone acetylase transferase Histone deacetylase Hydroxylated polychlorinated biphenyl 50% inhibitory concentration Ligand binding domain Leutenizing hormone Mitogen-activating protein kinase Nerve growth factor-inducible gene B Nuclear receptor corepressor Nuclear receptor PPAR-binding protein p300/CBP associated factor Polychlorinated biphenyl Peroxisome proliferator activated receptor Per ARNT Sim Proximal-box xiv SRC SERM SMRT SHP- 1 TRAP VDR WT Retinoid X receptor Retinoic acid receptor Steroid receptor coactivator Selective estrogen receptor modulator Silencing mediator for retinoid and thyroid- hormone receptor Small heterodimer partner Thyroid hormone receptor-associated protein Thyroid receptor Vitamin D receptor Wild-type XV CHAPTER 1 Objectives Rationale Estrogens regulate several diverse physiological responses including reproductive development, cardiovascular function, and bone metabolism. Estrogens have also been implicated in a number of human disease states, including breast and endometrial cancers (1), cardiovascular disease (2), osteoporosis (3), and Alzheimer’s disease (4). In recent years, it has been suggested that exposure to chemicals in the environment with estrogen-mimicking activities may cause adverse health effects in humans and wildlife (5). These chemicals, known as estrogenic endocrine disruptors (EEDs), encompass a wide range of structurally diverse chemicals including natural products, industrial chemicals, pharmaceuticals, and environmental pollutants. The actions of EEDs are mediated through their binding to the estrogen receptor (ER), though additional mechanisms of action cannot be discounted (6-13). The ER is a member of the nuclear receptor superfamily, a family of ligand- and non-ligand-regulated transcription factors that modulate target gene expression. Even though the physiological actions of the ER are conserved among different species, amino acid sequences of ligand binding domains are variable. A limited number of studies have shown significant differences in ligand preference, binding affinity, and transactivation activity in response to EED exposure among vertebrate species (14-19). Hypothesis ERs among different classes of vertebrate species exhibit differences in ligand preference, binding affinities, and transactivation responses to EED exposure due to variation in the amino acid sequence within their ligand binding domains. Aims 1. Examine and compare differences in ligand preferences and binding affinities of representative ERs from different classes of vertebrate species for EEDs using a competitive binding assay. 2. Investigate the ability of EEDs to induce gene expression mediated by representative ERs from different classes of vertebrate species using transiently transfected MCF-7 human breast cancer cells. 3. Identify amino acid residues putatively involved in ligand binding and perform site- directed mutagenesis to verify the role of these residues in the differential respones observed in Aims 1 and 2. 4. Purification of rainbow trout ERa for further biophysical characterization. 10. 11. 12. 13. 14. 15. References Bernstein, L., Ross, R. K., and Henderson, B. E. (1992) J Natl Cancer Inst Monogr 12, 137-47 Chae, C. U., Ridker, P. M., and Manson, J. E. (1997) Thromb Haemost 78(1), 770-80. Battistini, M. ( 1997) Rev Rhum Engl Ed 64(6 Suppl), 448-49S. Birkhauser, M. H., Stmad, J ., Kampf, C., and Bahro, M. (2000) Int J Geriatr Psychiatry 15(7), 600-9. Colborn, T., vom Saal, F.S., Soto, A.M. (1993) Environ. Health Perspect. 101, 378-384 Adlercreutz, H., Hockerstedt, K., Bannwart, C., Bloigu, S., Hamalainen, E., Fotsis, T., and Ollus, A. (1987) J Steroid Biochem 27(4-6), 1135-44 Brouwer, A., and van den Berg, K. J. (1986) Toxicol Appl Pharmacol 85(3), 301- 12. Hodgert Jury, H., Zacharewski, T. R., and Hammond, G. L. (2000) J Steroid Biochem Mol Biol 75(2-3), 167-76. Loomis, A. K., and Thomas, P. (2000) Biol Reprod 62(4), 995-1004. Revelli, A., Massobrio, M., and Tesarik, J. (1998) Endocr Rev 19(1), 3-17. Blumberg, B., Sabbagh, W., Jr., Juguilon, H., Bolado, J ., Jr., van Meter, C. M., Ong, E. S., and Evans, R. M. (1998) Genes Dev 12(20), 3195-205. Masuyama, H., Hiramatsu, Y., Kunitomi, M., Kudo, T., and MacDonald, P. N. (2000) Mol Endocrinol 14(3), 421-8. Moore, J. T., and Kliewer, S. A. (2000) Toxicology 153(1-3), 1-10 Le Drean, Y., Kern, L., Pakdel, F., and Valotaire, Y. (1995) Mol. Cell. Endocrinol. 109, 27-35 Connor, K., Ramamoorthy, K., Moore, M., Mustain, M., Chen, 1., Safe, S., Zacharewski, T., Gillesby, B., Joyeux, A., and Balaguer, P. (1997) Toxicol Appl Pharmacol 145(1), 111-23 16. 17. 18. 19. Vonier, P. M., Guillette, J ., L.J., McLachlan, J. A., and Arnold, S. F. (1997) Biochem Biophys Res Comm 232(2), 308-312 Fitzpatrik, D. W., Picken, C. A., Murphy, L. C., and Buhr, M. M. (1989) Comp Biochem Physiol 94C(2), 691-694 Lutz, I., and Kloas, W. (1999) Sci Total Environ 225(1-2), 49-57 Petit, F., Valotaire, Y., and Pakdel, F. (1995) Eur J Biochem 233(2), 584-92 CHAPTER 2 Literature Review Hormones and Cellular Regulation Hormones are non-nutritive substances that are released into the intercellular environment and act as information carriers or messengers causing changes in cellular stimulation. There are three types of cellular stimulation mechanisms: endocrine, paracrine, and autocrine. Endocrine hormones are released into the bloodstream and carried to distant target sites. In addition to long-range signaling, most cells release chemical signals into the extracellular fluid that influence neighboring cells (paracrine stimulation), though many cells also release chemical signals that modulate their own function, a process known as autocrine stimulation. Frequently, the same hormone can be involved in all three types of cellular stimulation paradigms. The initial step in which hormones exert a response involves their interaction with specific receptors. Hormones like FSH (follicle-stimulating hormone), LH (luteinizing hormone), insulin, somatotrophin and others bind to transmembrane receptors on the cell surface that initiate a signal transduction cascade resulting in changes a number of intracellular events, including the release of second messengers as well as the activation of kinases (1). This leads to a very rapid response that is quickly attenuated once the cellular equilibrium is re-established. Steroid hormones such as estrogens, androgens, and glucocorticoids and the thyroid hormone diffuse into the cell and bind with high affinity to their cognate receptors. Li ganded receptors either translocate to the nucleus or are already present in the nucleus where they bind DNA, causing dose-dependent changes in transcription. Since changes at the transcriptional level are relatively slow, these compounds produce more long-terrn cellular effects compared to peptide hormones that mediate their effects through a signal transduction pathway. There is increasing evidence that in addition to leading to changes in gene expression, estrogens are also able to bind to membrane receptors leading to induction of a signal transduction cascade (2). The actions of estrogens are regulated through negative and positive feedback mechanisms on the hypothalamic-pituitary-gonadal axis (Figure 1). The synthesis and release of estrogen from the ovaries is synchronized by the pituitary gonadotropins FSH and LH, which are regulated by the release of hypothalamic gonadotropin-releasing hormone (GnRH). GnRH concentrations are influenced through negative and positive feedback by steroid hormones. Adrenocorticoid hormone (ACTH) released from the pituitary acts on the adrenal glands to cause the synthesis and release of androgens, which become converted to estrogen by the enzyme aromatase. Estrogens as well as progesterone can also act directly on the pituitary to decrease FSH and LH. Moreover, the ovarian protein, inhibin, also down regulates FSH synthesis (1). Estrogen Receptor Agonist and Antagonist Ligands Estrogens play a critical role in growth, development, and maintenance of a number of cellular functions in all vertebrate species (3). The three natural estrogens are l7B-estradiol (E2), estrone (E1), and estriol (E3), though there are a number of metabolic derivatives (Figure 2). E2 is the most potent of the three and the primary estrogen in mammals and non-mammalian vertebrates (3). In addition to the role estrogens play in reproductive physiology, estrogens can also induce the growth and proliferation of cancer cells. As a result considerable effort has been put forth in developing antiestrogen drugs for the treatment and prevention of breast cancer (4). Ligands such as E2 and the Figure l I Hypothalamus 1 I () l GnRH Inhibin (-) ® .(-) J r fl ACTH ACTH * * * , Estrogen + I Adrenal I FSH LHI Adrenal I—P Ad1pose Progesterone Androgens Figure 1. Negative and positive feedback regulation of estrogens and progesterone on the hypothalamic-pituitary-ovarian axis. Arrowheads denote positive regulation. GnRH, gonadotropin-releasing hormone, ACTH, adrenocorticotropic hormone, LH, luteinizing hormone, FSH, follicle stimulating hormone. synthetic nonsteroidal estrogen diethylstilbestrol (DES) function as pure agonists, whereas others such as ICI 182,780 function as pure antagonists (Figure 2), though it has been reported to function as a partial agonist in the sheep uterus (5). The synthetic drugs, raloxifene and tamoxifen (6), are routinely used in the treatment of breast cancer, and belong to a class of compounds known as selective estrogen receptor modulators (SERMs) (Figure 2). The antagonistic and agonistic activities of these compounds are tissue- and promoter-specific (7,8). Environmental Endocrine Disruptors In recent years, there has been heightened concern that exposure to hormonally active chemicals, known as endocrine disrupting chemicals (EDCs) may cause adverse health effects in humans and wildlife (9). EDCs have been defined as any exogenous substances that can cause adverse health effects in an intact organism or its progeny, secondary to changes in endocrine function (10). Suspected EDCs encompass a wide range of compounds including natural products (e. g. phytoestrogens, mycotoxins), environmental pollutants (e. g. polychlorinated biphenyls, dioxin), pharmaceuticals, pesticides and industrial chemicals. Some effects attributed to EDC exposure include: increased susceptibility to hormone-dependent cancers (11,12), reproductive tract abnormalities (13), compromised reproductive fitness (14), and neurological (15) and developmental abnormalities (16,17). In response, new legislation was implemented in the form of the Food Quality Protection Act (1996 - Bill number PL. 104-170) as well as amendements to the Safe Drinking Water Act (05 1996 - Bill number S. 1316) that require the United States Environmental 10 Agonists Natural 17B-estradiol estrone estriol Synthetic moxestrol “0 diethylstilbestrol Antagonists SERMs O\/\N O 0 O 0 (WWW O \ 0H _ ,0 o. o o raloxifene W0 OH tamoxith O O o o. D. O ONO Iasofene EM-652 OH, on .. co ,0 0. 01421930012130st ICI 182780 '0' 154354 Pure antagonists CH2),°OON(CH,)CN(CH:,)C.H. Figure 2. Estrogenic (agonist) and antiestrogenic (antagonist) ligands of the ER. SERMs, selective estrogen receptor modulators. 11 Protection Agency (US EPA) to screen and test chemicals for estrogenic, androgenic and thyroid-like activities. The Endocrine Disruptor Screening and Testing Advisory Committee (EDSTAC) was established and charged with the development and implementation of a screening and testing strategy (18). EDSTAC deliberations resulted in recommendations for tiered strategy consisting of prioritization followed by screening and testing approaches involving a combination of in vitro and in vivo assays. It has been estimated that more than 30,000 chemicals require testing. Much attention has focused on estrogenic endocrine disruptors (EEDs) (16,19- 21). EEDs encompass a wide range of compounds including natural products, environmental pollutants, pharmaceuticals and industrial chemicals (Figure 3). Many of these agents do not share any obvious structural similarity to the prototypical estrogen, E2, which makes identification based solely on molecular structure difficult (22). Although controversial, there are reports of a decrease in human sperm production and seminal volume during the past half-century (23-25). However, reproductive abnormalities have been observed in mammals (26), reptiles (17,27), birds (28,29) and several fish species (30) following exposure to environmental contaminants. The feminization of gull embryos (31) and gonadal sex reversal of turtles (32,33) have also been reported following exposure to l,l,l—trichloro-2,2-bis(p-chlorophenyl)ethane (DDT) and polychlorinated biphenyls (PCBs), respectively. Induction of vitellogenin, a gene normally induced in egg laying females, has been detected in male fish located downstream of sewage treatment effluent (34). It has been proposed that the activity of estrogenic, as well as androgenic and thyroid-like chemicals is mediated by their respective nuclear receptors (NRs) (35), 12 Natural 0H 0 CH3 HO C ' HO O u o zearalenone genistein Coumestrol Pesticides CI CI Cl | “H ° ' WHO“! col, C, 0' cc:a o p‘-DDT Methoxychlor Dieldrin Industrial Chemicals/Contaminants H3 03H17 C'x / “my Bisphenol A octylphenol Polychlonntated biphenyl o /C4H9 O O O 851$ O\ C4H9 O 0 ° 0 O Dibutyl phthalate benz[a]anthrecene HO OH Phenol Red impurity Figure 3. Suspected estrogenic endocrine disruptors (EEDs) and their classifications. 13 though other mechanisms such as interactions with binding globulins (36,37), inhibition of aromatase, the enzyme that converts testosterone to estrogen (38,39), and binding to membrane receptors (40,41) or other NRs (42,43) cannot be excluded. Nuclear Receptors NRs are a superfamily of ligand- and non-ligand-regulated transcription factors that play critical roles in development, differentiation, and homeostasis. They regulate gene expression through binding small hydrophobic ligands, such as steroid and thyroid hormones, vitamin D, and retinoids. These proteins represent the largest family of transcription factors known; yet relatively low amounts of these receptors (103-104 copies/cell) are present in a cell (35). This superfamily can be divided into four broad categories based on their ligand binding, dimerization abilities, cellular distribution, and DNA binding properties. The type I NRs include receptors for steroids such as estrogens, androgens, progestins, glucocorticoids, and mineralcorticoids. These receptors are bound to heat shock proteins and sequestered in the cytoplasm in the absence of ligand. Ligand binding causes the dissociation of heat shock proteins and translocation to the nucleus where they bind their cognate palindromic response elements as homodimers (Figure 4A). The ER, unlike other type I NRs, is found exclusively in the nucleus, a behavior that has been confirmed in the presence and absence of ligand (44,45). The type H NRs include among others, receptors for thyroid (TR), vitamin D (VDR), and all-trans retinoic acid (RAR). These receptors are found solely in the nucleus and bind to direct repeat response elements as heterodimers with 9-cis retinoid X receptor (RXR) (45) (Figure 4B). The type III and type IV NRs contain a large number of putative receptors for 14 Figure 4. Proposed mechanism of action of (A) class I and (B) class 11 nuclear receptors. See text for details. NR, nuclear receptor; hsp90, heat shock protein 90; hsp70, heat shock protein 70; HRE, hormone response element; RXR, retinoid X receptor. This figure is presented in color. 15 \ Cellular Response ‘\ Q} ammu- Protein which ligands have not been identified or may not exist, and are referred to as orphan receptors (46,47). These putative receptors have been identified based on their sequence homology to type I and type II NRs. Type 111 members include: peroxisome proliferator activated receptor (PPAR) (48), chicken ovalbumin upstream promoter transcription factor (COUP-TF) (49), hepatic nuclear factor-4 (50), and several others (47,51). These putative receptors bind to DNA as homodimers, and/or heterodimers with RXR. Type IV NRs, which are also classified as orphan receptors include: estrogen-related receptor (ER) (52), nerve growth factor-inducible gene B (NGFI-B) (53), SF-l (54), and many others (47,51). Type IV members bind DNA exclusively as monomers, a feature that distinguishes them from Other NRs. Moreover, the orphan receptor DAX-l contains a unique three repeat domain in the N-terminus, which represents a novel type of single stranded DNA/RNA-binding domain (55). Short heterodimerization partner (SHP) is an unusual orphan receptor that lacks a conventional DNA binding domain (DBD) but contains a putative central ligand binding domain (56). SHP interacts with a variety of NRs including RXR and several of its heterodimerization partners and acts as a negative regulatory of ER function (57). Orphan receptors have attracted considerable interest since they could help uncover novel endocrine regulatory systems. Most NRs display a modular structure with five to six distinct regions, termed A- F domains and contain two separate activation functions (Figure 5). The N-terminal AlB regions contain a ligand-independent activation function (AF-1), which can constitutively activate transcription. The C region contains the DBD, which is characterized by two 17 AF’I <— Ligand Binding _, PH Domain A/B C D E F ® DNA Binding Q9 ®HJ Domain AF'2 Figure 5. Conserved domain structure of nuclear receptors. A schematic representation of the human estrogen receptor 01 is shown. AF-l and AF-2 refer to the ligand independent and ligand dependent transactivation domains, respectively. The encircled P refers to known sites of phosphorylation, while the encircled Ac identifies sites of acetylation. 18 zinc fingers and recognizes cognate cis elements in the upstream regulatory region of target genes, referred to as response elements (Figure 6A). The proximal-box (P-box) located in the first zinc finger interacts with DNA and is involved in binding site discrimination (58,59), whereas the distal-box (D-box) located in the second zinc finger supports receptor dimerization (59-61). The nature of the response element varies among the different types of NRs. NR response elements for type I receptors consist of inverted palindromic sequences, which are separated by a three nucleotide spacer (Figure 6B). Type 11 NR response elements are generally direct repeats (DR) composed of a minimal hexanucleotide core spaced by 1 to 5 nucleotides (Figure 6B). A flexible hinge or D region separates the DBD from the E region, which contains the ligand binding domain (LBD), dimerization interface, and the ligand-dependent activation function, AF-2. Recent studies have demonstrated that the hinge region may serve as a docking site for corepressor binding (62,63). Acetylation of lysine residues within the hinge region of the ER has recently been shown to influence ligand sensitivity as well as recruitment of corepressor complexes (64). The activities of AF-1 and AF-2 vary depending on the promoter (65) and cell type, and in some cases both are required for full transcriptional activation (66). Several receptors also contain a variable C-terrninal F domain, which has been shown to influence interactions with coactivators (67,68). X-ray structures of the LBD reveal that NR LBDs exhibit a similar overall architecture with the LBD folded into a three-layered antiparallel a-helical sandwich comprising a central core layer of helices (H5, H6, H9, and H10) found between two additional layers of helices (H1-4 and H7, H8, and H11) (69-76) (Figure 7). Biochemical 19 o G H . R 0 Y O o A 9 K o . _ . D C D 9 I o o O O o C C Dbox C Zn Zn : G / \ G PW / \ G C C L KoFFoRooooooooYc C '0'GM Direct Repeat Inverted Repeat Palindromic Sequences q nxfi fl nxh flnx<== TR, RAR, VDR, FXR, RAR, TR, ER, GR, MR, PPAR, RXR, PXR, PPAR, RXR PR, AR FXR, CAR, v-ErbA Everted Repeat Monomeric RNA Hairpins No DNA interaction h 11x q fi 5 ? RAR, TR, PXR NGFIB, SF-l, [)qu SHP-1 ERR1,ERR2, ROR Figure 6. (A) Hypothetical "zinc fingers" in the DNA binding domain of nuclear receptors. Conserved amino acid residues are shown, whereas amino acid residues for which no consensus exists, are represented by a closed circle. Residues which make up the P (proximal) box and D (distal) box are boxed. (B) Different types of hormone response elements and DNA interactions among nuclear receptors. 20 .880 E 02888 mm 2:?» m3... .8888 88.5380 928.580 :5qu 8m 2% wee—000 ca 8.8% .88: use 823:8 388805.: 8m 228208 a me 00888 :00; 3: 2m .8 882:8.“ 282a $5. .538 9:qu EB»: 05 88 vocation mm 50:80 8838.» NJ? 05 .8 080 05 8 mun—80280 £033 Afimv Q 30$ 65%: .3 0085000 2 $50 3 {can 38 :03? .8 .038 98:5 0588.». a 8320 20% £5. 808658 30:28 8:22:23 0208— -025 a 85 320% 2a 8888 950:3 gum: 88002 82032 .83 20a 08:32 8.233 .53 0288880 88002 28 28:2 as ea .35 0853383632. 5:». 38358 8882 .8965 as .83 segment 58. 85358 a 88002 885mm a @5280 mama—E 05%: 88002 820:: .8 28m 30:80 8888 05 .8 823800 .5 083m 21 and structural studies have demonstrated that a substantial conformational change occurs after ligand binding (77,78). Comparison of the apo-RXR and 9-cis retinoic acid bound RXR crystal structures confirms the existence of a ligand-induced conformational change, which triggers the receptor to form an activated complex and expose amino acid residues within AF-2 (helix 12), which are necessary for interaction with coactivators (78). The main conformational change occurs in the N-terminal regions of helices 3, 11, and 12 (Figure 8). This movement causes hydrophobic residues previously exposed to solvent to be directed towards the ligand binding cavity, while at the same time allows other hydrophobic residues to form a protein surface necessary for coregulator interaction (78). Estrogen Receptor 0t, [3, and 7 Many of the effects of estrogen are mediated through its binding to the ER and subsequent alterations in gene expression, though non-genomic effects have also been reported (2). Two ER isofonns exist, EROt and ERB, each one is encoded from a distinct gene and displays tissue-specific distribution and differential ligand preference (79,80). A comparison of the amino acid sequence identity between the two isoforms reveals that the most conserved region is the DBD (97%), whereas the NHz-terminal A/B domains and LBD are only 20% and 60% identical in amino acid sequence (81). Although both isoforms exhibit similar affinity for E2, a comparison of their tissue distributions suggests that they may have distinct biological roles (81). Several groups have shown that EROt and ERB form functional heterodimers in vitro and in vivo (82-84). Interestingly the isoforms, the heterodimers would predominate (81,83). The physiological role of each 22 B holo-RXRoc Figure 8. The ligand induced conformational change of human retinoid X receptor (RXR). Comparison of the (A) unliganded (apo-RXROL, shown in blue) and liganded (hold-RXRa shown in orange). The main conformational differences were observed in helices 3 (H3) and H12 following ligand binding. The arrows emphasize the ligand-induced conformational change. The ligand, 9-cis retinoic acid, is shown in light blue. This figure is presented in color 23 isoform is being intensely studied through the generation of isoform-specific knockout mice (85-87). The DBDs of EROt and ERB, like other NR5, allow the receptors to bind to their cognate DNA target sequence, referred to as an estrogen responsive element (ERE). It has been demonstrated that both ER isoforms do not signal only through binding EREs but can also signal through AP-l and Sp-l elements (88-90). Intriguingly, ERB exhibited distinct properties from ERa on both AP-l and Sp-l elements. Antiestrogen binding to ERB caused an increase in reporter gene expression, whereas binding to EROL resulted in an opposite effect examined under the same conditions (88,90) The recent discovery of a third unique type of ER, ERy, isolated from Atlantic croaker and zebra fish, has amplified the complexity of estrogen signaling (91). Phylogenetic analysis revealed that ERy arose from a relatively recent gene duplication of ERB (91). Teleost (bony fish) genomes have been shown to support significantly more gene duplications than their mammalian counterparts (92). Therefore, ER-y may not represent a third unique ER isoform and may simply represent a genetic duplication of ERB. In addition, ERa and ERB have been described in several vertebrate species, ERy has thus far only been reported in teleosts (91). X-ray structures of several NR LBD complexes with various ligands have provided insight into the molecular basis of agonism and antagonism. Ligand binding is achieved by a combination of specific hydrogen bonding interactions and the hydrophobic nature of the binding pocket. In the ERa—EZ complex, E2 makes direct 24 Figure 9. Interactions made by HIS-estradiol in the human estrogen receptor a ligand-binding cavity. This figure was generated in InsightII using the Brookhaven Protein Data Bank entry lERE (73). The three amino acids residues (Glu353, Arg394, and Hi5524) that participate in direct hydrogen bonds with the ligand and the distances of the interactions are shown. This figure is presented in color. 25 hydrogen bonds with the side chains of glutamate 353, arginine 394 and histidine 524. Hydrophobic interactions further stabilize the ligand in the binding cavity (Figure 9). Interestingly, ligands with agonist and antagonist activity stabilize the LBD in distinct conformations, most notably in the orientation of helix 12. In the E2-liganded complexes, helix 12 is positioned over the ligand binding cavity and forms a hydrophobic groove with helices 3, 4, and 5, as well as a turn between helices 3 and 4 (Figure 10). This hydrophobic groove has been proposed as the AF-2 surface, which is capable of interacting with coactivators (73,93). In the raloxifene-liganded structure, helix 12 is prevented from forming part of the AF-2 interaction surface and instead lies in a groove between helices 3 and 5, in effect blocking the coactivator recognition site and mimicking the interactions formed between the NR box motif and the LBD (73,74,93) (Figure 10). The position adopted by helix 12 in the partial agonist genistein-bound structure only partially mimics the NR-box binding mode (74) (Figure 10). In the pure antagonist bound structure the bulky side chain of ICI 164,384 binds along the NR-box binding site and physically prevents helix 12 from adopting either agonist or antagonist orientation (94), further illustrating the importance of the orientation of helix 12 in ER activity. In addition to the ligand-induced activation of the ER, phosphorylation has been shown to play a key role in modulating ER activation (95-97). Treatment of cells with growth factors was shown to increase the activity of the ER as well as its phosphorylation state. Subsequent studies demonstrated the serine 118 was phosphorylated and required for full ligand-independent receptor activation (95). Serine 118 was later shown to be a substrate for mitogen-activating protein kinase (MAPK), linking ER activity to the RAS- MAPK pathway (96). ERa is also phosphorylated at tyrosine 537, which when mutated 26 .830 E @2585 mm 98 $3 89a :83 33 05mm mi... .5388 953 cacao—E assuage cowobmo a8 ADV v5 £0388 c.52— Eoummeomla SERVE comebmo SE5: Rs 5 Amt 53:80 952— Emwabmoé: -6 8532 comebmo 58:: A3 05 E 2 “£2— .«o :oEmom .2 £0: mo cassava woos—cam 93m: 2: mo scam—8:80 .3 chew?— 27 causes a ligand—independent increase ER activity (97). Phylogenetic Comparison of ERs Phylogenetic analysis reveals that there are distinct ER types. The teleost EROts and ERBs form distinct groups compared to all other vertebrate ERs (Figure 11). The interspecies homology is higher than the similarity between the 0t and B isoforms. Interestingly, the two ERys in the database, cluster with the other teleost ERBs and do not form a separate and distinct branch. A more detailed analysis of the sequence differences among ER D, E, and F domains from several different species, including representative species from 5 different vertebrate classes; mammalian, avian, amphibian, reptilian, and actinopterygii (ray-finned fish) is shown in figure 12. These results demonstrate that significant variability exists within the ligand-binding portion of the different receptors and between 0t and [3 isoforms. Despite the differences in degree of sequence identity among species, ERs from all species harbor the same three equivalent amino acids to hEROL (glutamate 353, arginine394 and histidine524) that participate in direct hydrogen bonds and stabilize E2 in the binding pocket. (73,74,98). However, differential binding of several natural and synthetic chemicals to hEROt and hERB, as well as to ERs from different species have been reported (see below), suggesting that additional amino acid residues may also play a role in determining ligand preference and relative binding affinity. Despite significant differences in amino acid sequence within ER LBDs a comparison of the amino acid sequences that line the ligand-binding cavity and interact with E2 reveals that few amino acid differences exist among different species compared to hERa. For the aER and xER only single amino acid substitution exist, phenylalanine 28 Figure 11. Phylogenetic analysis of estrogen receptor (ER) sequences. The distance between nodes reflects the degree of sequence identity when doing pair wise alignments. The value of 0.] corresponds to a difference of 10% among sequences. This figure was generated using the ClustalW alignment function in MacVector 6.5 (Oxford Molecular Ltd.). GenBank accession numbers: anolis EROL AFO9591], African catfish a (a. catfish 0t) X84743, Atlantic croaker ERa (a. croaker a) AF298183, Atlantic croaker B (a. croaker B) AF29818 1, Atlantic croaker ER y (a. croaker 7), Atlantic salmon at (a. salmon 0t) X89959, blue tilapia 0t (b. tilapia 0t) P50240, catfish a AFO61275, catfish B AF 185568, chicken 0t X03805, chicken B AB036415, cow B Yl8017, gilthead seabream 0t (g. seabr. 0t) A1006039, gilthead seabream (g. seabr. B) AF 136980), goldfish B AF06l269, hamster 0t AF 181077, horse 0t AF 124093, human 0t X03635, human B AB006590, Japanese eel B (i. eel B) AB003356, marmoset B Y09372, medaka B ABO33491, mouse on M38651, mouse B U8l451, Nile tilapia 0t (11. tilapia 0t) U75604, Nile tilapia B (n. tilapia B) U75605, pig (1 AF035775, pig B AF267736, quail B AF045149, rainbow trout (11 (r. trout al) A] 242740, rainbow trout a2 (r. trout 0L2) AJ 242741, rainbow trout B (r. trout B) A] 289883, rat 0t Y00102, rat B A10026OZ, red seabream on (r. seabr. 0t) AB007453), rhesus monkey B (rh monkey B) AF 1 19229, sheep B AF 177936, Starling B AF113513, whiptail 0t $79923, wrasse 0t AF326201, xenopus 0t L20738, zebrafinch 0t L79911, zebrafish 0t AF349412, zebrafish B AF349414, zebrafish ‘y AF349413. 29 human 0t horse on ‘ rat a mouse on . hamster 0t — pig a h' k a __ C re en i - . anolis 0t whiptail 0t xenbogttis ab 0 ze r inc —[ r-—-catfish on a. catfish 0t wrasse at r. seabr. a :e _. ——E g. seabr. 0t menu. a. croaker 0t ‘e was r—n. tilapia a ' . tilapia 0t cular medaka Oé .. -. r. trout 0t ,dtllsh Er. trout a1 . a. salmon 0t ;almon catfish B zebrafish y . croaker B . ee lbrcam . r—J—— zegraftisflh B r. on 5 LC: goldfish [3 n. tilapia % g. seabr. 304, r . a. croaker 'y f—Z‘ starlmg B quail B 1) chicken B {M - cow sheep plg B a ‘ marmoset B 1P” *_‘l Erh. monkey B arafishl hutnban B - ra {mouse B Scale 0.100 30 ha --MLKHKRQRDDGEGRGEVGSAGDMRAANLWPSPLMIKRSKK ----- NBLALSLTADQIVEALLDIIPPILYBIY m -- .......... L...N.M.AS ............. V..HT.. ----- ..P...................II.... C -E. M. Q. .EEQDS. NGEA. STEL. .PT. .T. .VV. HN.. ----- ..P......I.......l.....V.... a -- ........ EENDS. N- A. ALTEA. STA ......... H. . ----- .P..............I....VV.... x --L ...... KEEQ. QKN- DVDPSEI. T. SI. VN. SVK--. M. ----- L. PV.....I.LI...II..A..V...B rt RVLR.D..YCGPAGD.EKPYGDLEH.T.PPQDGGRNSSS.LNGGGGIRGPRITIPPI.VLFL.QC....A.C.RQ hB --LVR-~...SAD.QLHCA.K.KRSGGHAPRVRE.LLD -------------- ALBPI.DVL@..I....EVLIBR ‘) H3 W) H5/H6 Hm- a to a. . fit at e ‘ e DPTRPPBIASIIGLLTNLADRIEVEIINIIRRVPGIVDLTLHDQVHLLICANLIILIIGLVWRSNIHPGKLLIAP 0.80.00.00.00000...O0.00.0.00...OOOOOOOUCOOOOOOOOOOOOOOOOOO00.00.000.000... ..N...N.....T.............................................................. ..8...fl...V.T..........................A.................V................. .8.R.L......T............................................V..I...V......8... KVI~.YT.VT..T...SI..K......A...K....QI.8.....Q...88..{V.....I...IHC....I..Q -P8&..T.....I8..K...K......8...KI....I.8.P...R...8C.I{V}.M..M...ID.....I... ‘2?) 110-6) H8 H) H9 H mo 3 * NLLLDRNOGXCVIGIVIIFDILLAIBBRIRIINLQGIIPVCLKSIILLNSGVYTPLBSTLRSLIIKDHIERVLDK .........................AA......................................R.Y....... ......88..V...’..................VR..............P.I..Y.......V..R......... .........R....L.......VT.A$.....R.R....I....................I...DT.L..II... D.I...SI.D.....A.........V}....LK.KP.......A.......A’8.C.N8VI..HNBBIVISI..N D.V}..nl......IL.........T....ILR..EK.YL.V}AM.....8M.PLV!A.QDID88R-KLABL.HA O H) H11 )—QE— ee .9 o ITDTLIBLIIKAGDTLQQQBQRLAQLLLILSHIRHIBHIGIIELYSIRCKNVVPLYDLLLIILDAERLEAPTSRG O....000......OOOOOOROOOOOOOOOOOOOOOOOOOOOOOOuOIOOOOOOOOOOOOOOOOOOOOOOOAQ0M ...........B..8.....R........................N.........................A.RS .....M.....‘..8.....R..................................................AAK. .I...V.F...8..8....QR..............................................I.T.KDKT ...A...HISHB.ABV...PR.Q.......................I....R............G...QS.GKVA V..A.VWVI..8.188...8fl...H..IL...V}.A........Lfl...R...PV.......LN..V.RGCK.SI GASVEETDQSHLATAGSTSSHSLQKYYITGEAEG-FPATV ------------- 595 .VPP..PS.TQ...TS...A....T...PP....-..N.I ------------- S99 A.PM..ENRNQ.T..-PA ...... SF..NSKE.ESMQN.I ------------- 589 SPPS.DDPLNQ..VP-.PSM...LPC.VNKQE..NEQEAI ------------- 349 TTQE.DSRSPPTT.VNGA.PCLQPY.TNTE.VSLQSTV --------------- 586 Q.GEQTEGP.TTT.TSTG..IGPMRGSQDTH.RSPGS.VLQYGSPSSDQMPIP 577 TG.ECSPAEDSKSKEG.QNPQ.Q ------------------------------ 530 Figure 12. Structural based alignment of the estrogen receptor D, E and F domains of human 0t (ha), human B (hB), mouse (m), chicken (c), anolis (a), xenopus (x), and rainbow trout (rta). Identical amino acid residues are represented as dots while missing residues are shown as dashes. The ligand binding domains (E domains) are shown in bold. Residues that line the hormone binding pocket and that interact with 17B-estradiol are marked with a * and #, respectively. Boundaries of the helices and beta sheets were taken from PDB lERE (73). 31 175 in the aER and leucine 413 are both substituted to methionine 421 in hERa (Figure 12). Two amino acid substitutions exist between hERa and both hERB and rtERa. In hERB, methionine 336 and isoleucine 373 are substituted for leucine 384 and methionine 421, respectively, whereas in rtERa methionine 317 and isoleucine 496 are substituted for leucine 349 and methionine 528, respectively (Figure 12). This suggests that these residues may contribute to the observed differences in ligand preferences, binding affinities, and transactivation responses to EEDs among different ERs (see below). Coregulators Biochemical and expression cloning studies have identified a number of proteins that interact with N Rs in a ligand-dependent and/or ligand-independent manner (99). Many of these proteins have been shown to potentiate NR activity in cell based cotransfection assays, suggesting that they act as coregulators of NR function (99). Coregulators can be classified as coactivators and corepressors. Coactivators lead to an increase in NR activity, whereas corepressor attenuate or inhibit NR activity. Several structurally distinct classes of nuclear receptor coregulators have been identified, some of which are subunits of large multiprotein complexes. For example, the steroid receptor coactivator (SRC) family of coactivators (100) is a family of 160 kDa molecular weight proteins that were the first coregulators shown to interact with NRs. Members of this family include, SRC-1 (also known as p160/NCoA-1/ERAP-l60) (100,101), SRC-2 (also known as TIF2/GRIP-l/NCoA-2) (102,103), and SRC-3 (also known as AIB1/ACTR/RAC-3/TRAM-1/pCIP/NCoA-3) (104-108). Once recruited by NRs, these coactivators form a mutiprotein complex with CBP/p300 (105,107) leading to increases 32 in (Ir-3 in transcription. SRCs are members of the PAS domain family, which contain a N- terminal basic helix-loop-helix region, followed by PAS A and PAS B domains, and a C- terminal glutamine rich region (109,110). Functional and structural studies have shown that coactivators interact with the AF-2 region via short NR-boxes or LxxLL motifs, where L refers to leucine and x refers to any other amino acid residues, to transduce the ligand signal to the transcriptional machinery (93,111,112). Recent studies have shown that nuclear receptors have distinct affinities for coactivators; for example PPARy has a greater affinity for CBP than for SRC-1, whereas the ER preferentially binds SRC-1 but exhibits very weak affinity for CBP (113). In addition, the NR affinities of different coactivators and various NR boxes present within coactivators have been demonstrated to exhibit ligand dependent variability (114,1 15). Several lines of evidence suggest that the SRC family of coactivators are important in NR function and disease (104,116,117). However, studies reveal potential redundancy in NR coactivator function since only subtle effects in SRC-1 knockout mice were observed, which may have been the result of a compensatory increase in SRC-2 levels (118). In contrast, SRC-3 is required for normal development and female reproductive function, suggesting that it plays a different physiological role than SRC-1 and provides evidence of diversity within the SRC family (119). Coregulators and NRs have been shown to interact with the general transcription factors CBP/p300 and pCAF. SRC-1, SRC-3, CBP/p300 and pCAF possess histone acetyltransferase (HAT) activity, which disrupts nucleosomes and leads to transcriptional activation (105,120). In the absence of ligand some NRs associate with nuclear receptor 33 corepressors SMRT (63) and NCoR (62). The association of NCoR with TR is crucial for the ligand independent nuclear retention of TR (121). Both SMRT and NCoR recruit SIN3 and HDACs to form a large corepressor complex that contains histone deacetylase activity (122,123), implicating histone deacetylation in transcriptional repression. Three models, sequential, combinatorial, and parallel, have been proposed to explain how the extraordinary numbers of coregulators modulate NR transcriptional activation (99). In the sequential model, one of the complexes is recruited to the promoter, which leads to the recruitment of additional proteins. Alternatively, a combinatorial model has been put forth where multiple complexes are present and their combined activity is required for transcriptional activation in vivo. Finally, transcriptional activation may occur in a parallel fashion, with different complexes forming on the same promoter, allowing for tissue or cell specific responses. Recent reports favor the sequential model where different cofactors cycle during the onset and course of transcription (Figure 13) (124). The binding of E2 to the ER causes the exchange of the corepressor complex with a coactivator complex. This is followed by the recruitment of coactivators with HAT activity such as p300/SRC complex and PBP, the protein that anchors the DRIP/T RAP complex to NRs (125). The p300 complex modifies chromatin, but is present only during the initial cycle of transcription. Once the RNA polymerase II C-terminal tail is phosphorylated, CBP replaces the p300 complex causing the recruitment of pCAF. This leads to the concomitant release of SRC and the ER. CBP and pCAF disassemble and the cycle is set to repeat ( 124). 34 Repression AF -1 mediated tanscription l 7B—estradiol e '\‘ as S IE3; / re-initiation acetylation , no ligand Figure 13. Proposed model for transactivation by ER including interactions with coactivators and corepressors. This model was modified from Shang et_al. (124), which demonstrated the sequential and cyclical formation of complexes leading to transcriptional activation in the presence of an agonist such as l7B-estradiol. This figure is presented in color. 35 Estrogen-Responsive Genes Several estrogen responsive genes have been characterized in a variety of species. Analysis of the upstream regulatory regions of many estrogen-responsive genes revealed a common sequence motif referred to as an estrogen responsive element (ERE), which was determined to have the following consensus sequence 5’-GGTCAnnnTGACC-3' In. (126,127). Electromobility gel electrophoresis assays demonstrated that the ER binds with high affinity to consensus EREs as a homodimer (56). The sequences flanking the ERE have also been shown to influence stability of the ER-ERE interaction (128,129). Genes such as Xenopus vitellogenin A2 (126) and human oxytocin (130) contain a single consensus ERE, whereas others like Xenopus vitellogenin B1 ( 126,131), the human PR (132), human c-fos (133) and human p82 (134) contain single or multiple copies of EREs, which contain one or more nucleotide differences from the consensus ERE. Some estrogen-responsive genes such as ovalbumin (135) contain multiple copies of ERE—half sites, but not the palindromic sequence. These genes also contain response elements for RXR, RAR, and TR (136,137), suggesting that the cellular concentrations of NR5 influence the regulation of target genes. In addition, the identification and study of estrogen-responsive genes is further complicated by the observation that the ER can regulate gene expression through interaction with other transcription factors (88). Species-Specific Binding Preferences and Transactivation Responses to EEDs Although the physiological actions of the ER are conserved among different species, the amino acid sequences of the regions involved in ligand binding are quite 36 variable. This suggests that species may exhibit different responses and sensitivities to EEDs, and that one species may not be an appropriate surrogate for use in identifying and predicting responses in other species. Significant variability exists in the reported binding affinity for E2 among vertebrate ERs (Table 1). Receptor binding studies have shown that E2 exhibits a 5-fold lower binding affinity for the rainbow trout ERa (rtERoc) than for the hERa (138,139). When expressed in yeast the rtEROt exhibits a 10-fold lower responsiveness to E2 and a significantly reduced affinity for E2 at elevated temperature when compared to hERa (140). Weaker E2 responsiveness of rtEROt has also been observed in embryonic salmonid cells (STE-137) transfected with rtERa (139). Similarly, the transactivation ability of another piscine EROt, the Oreochromis gureus (OaERa), also exhibits reduced activity at temperatures above its normal physiological range (141). The molecular basis for the reduced rtERa function at elevated temperature is unclear, though other fish proteins have been shown to exhibit optimal activity at lower temperatures than their mammalian counterparts (142-144). Interestingly, exchanging a 35 amino acid stretch, between the OaEROL and the chicken ER (cER) partially rescued the thermal deficient transactivation of the OaERa (141), which suggests that additional factors also contribute to the temperature sensitive phenotype of piscine ERs. Domain interchange and the generation of chimeric receptors revealed that the rtER (hEROt C domain) chimeric receptor exhibited greater DNA binding affinity, while the rtER (hEROL E domain) chimeric receptor displayed an increased E2 transactivational response compared to wild-type rtERa (145). These results suggest that some of the functional differences between hERon and rtERa reside within the DNA-binding domain. However, competitive binding studies using GST-hERa and GST-rtERa fusion proteins 37 Table 1. Comparison of the affinity for 17B-estradiol among different vertebrate ERs. ‘ Vertebrate Smies Source Kdrin nM Reference Amphibians Urodele Amphibian Testis 0.1 (146) Xenopus Liver cytosol 0.5 (147) Hepatocytes 0.4 ( 148) Avian Chicken Oviduct cytosol 0.23 (149) Oviduct cytosol 3.7 (150) Liver cytosol 0.7 (151) Japanese quail Liver cytosol 0.2 ( 152) Fish Atlantic croaker Testis 033-04 (153) B In vitro transcribed 1.38 (91) In vitro transcribed 1.16 (91) Atlantic Salmon Liver cytosolic 2-4 (154) Liver nuclear 5-6 (154) Catfish Liver cytosolic 1-1.3 (155) Rainbow trout a In vitro transcribed 5 (138) Mammals Human (1 Liver cytosol 0.47 (156) ER mammalian cells 0.2 (157) Full length in yeast 0.35 (158) Truncated in yeast 1.0 (158) ER—protein A fusion 1.49 (158) Mouse Uterine cytosol 1.4 (159) or In vitro transcribed 0.2 (160) B In vitro transcribed 0.5 (160) Pig Uterine cytosol 0.53 (149) Rat Uterine cytosol 0.7 (161) a Expressed in COS-1 0.8 (161) cells In vitro transcribed 0.1 (162) B In vitro transcribed 0.4 (162) Reptiles Alligator Oviduct 0.5 (163) Turtle Liver - cytosolic 17 (164) Liver - nuclear 17 (164) Testis - cytosol 0.7 (165) 38 consisting of only the D, E and F domains have identified several differences in relative binding affinities of estrogenic chemicals, supporting the hypothesis that the LBD significantly contributes to the reported differences between hERa and rtERa (166,167) (Chapter 3 & 4). Atrazine and simazine, two chloro-S-triazine derived pesticides were unable to induce expression of a hEROL regulated reporter gene transiently transfected in MCF-7 cells (168), while atrazine and other chloro—S-tn'azines were able to compete with tritiated E2 for binding to the American alligator ER (163). The pig ER exhibits significantly greater affinity for at zearalenol, a metabolite of the mycotoxin zearalenone, than does the ER from the Leghorn chicken (149). Genistein has been shown to bind with 30-fold greater affinity to hERB than hERa (162). The recent report of the crystal structure of ERB in complex with genistein has suggested that this ligand preference may be attributed to two conservative mutations: leucine 384 and methionine 421 in ERa are replaced by methionine 336 and isoleucine 373 in ERB (74). Comparison of the amino acid differences within the ligand binding pockets of ER from several different species reveals that methionine 421 in hERa is replaced by phenylalanine 175 in the anolis ER and leucine 413 in xenopus ER, whereas leucine 349 and methionine 528 in hERa are substituted with methionine 317 and isoleucine 496 in the rtERa (Figure 12). Although, these residue differences are considered to be conservative substitutions, they affect the volume and hydrophobic character of the binding pocket (74,169). Taken together, these observations suggest that these residue substitutions may contribute to the observed differences in the relative binding affinities of select EEDs. Recent crystallography data has demonstrated that both 39 the volume and hydrophobic character of the binding pocket can influence ligand binding. For example, the volume of the probe-occupied ligand pocket of EROt-EZ crystal complex has been determined to be 490 A while that of ERB-genistein is 390 A with the reduction being primarily due to the replacement of the leucine 384 in hERa with a bulkier methionine 336 residue in hERB (74). This allows the residues that line the pocket to pack more tightly around genistein, stabilizing the ligand in the binding pocket in ERB (74), and possibly resulting in greater binding affinity. 40 References 1. 10. ll. 12. l3. 14. 15. 16. Human, P. (1994) Human Pharmacology, Second Edition Ed. (Brody, T. M., Larner, J ., Minneman, K. P., and Neu, H. C., Eds), Mosby, St. Louis Razandi, M., Pedram, A., and Levin, E. R. (2000) Mol Endocrinol 14(9), 1434- 47. Comp, E. (1987) Fundamentals of comparative vertebrate endocrinology (Chester-Jones, I., Ingelton, P. M., and Phillips, J. G., Eds), Plenum Press, New York and London Dutertre, M., and Smith, C. L. (2000) J Pharmacol Exp Ther 295(2), 431-7. Robertson, J. A., Zhang, Y., and Ing, N. H. (2001) J Steroid Biochem Mol Biol 77(4-5), 281-7. Grainger, D. J ., and Metcalfe, J. C. (1996) Nat Med 2(4), 381-5. Grese, T. A., and Dodge, J. A. (1998) Curr Pharm Des 4(1), 71-92. Berry, M., Metzger, D., and Chambon, P. (1990) EMBO J 9(9), 2811—2818. McLachlan, J. A., and Korach, K. S. (1995) Environ Health Perspect 103 Suppl 7, 3—4. European Workshop on the Impact of Endocrine Disruptors on Human Health and Wildlife (1996) Vol. 1452, pp. 103-120., Weybridge, U.K. Davis, D. L., Bradlow, H. L., Wolff, M., Woodruff, T., Hoel, D. G., and Anton- Culver, H. (1993) Environ Health Perspect 101(5), 372-377. Wolff, M. S., Toniolo, P.G., Lee, E.W., Rivera, M., Dubin, N. (1993) J Natl Cancer Inst 85(8), 648-652. Giwercman, A., Carlsen, E., Keiding, N., and Skakkebaek, N. E. (1993) Environ Health Perspect 101 Suppl 2, 65-71. Carlsen, E., Giwercman, A., Keiding, N., and Skakkebaek, N. E. (1992) ij 305(6854), 609-613. Tilson, H. A., and Kodavanti, P. R. (1997) Neurotoxicology 18(3), 727-743. Colbom, T., vorn Saal, F.S., Soto, A.M. (1993) Environ. Health Perspect. 101, 378-384. 41 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. Guillette, L. J. J ., Gross, T.S., Masson, G.R., Matter, J .M., Percival, H.F., Woodward, AR. (1994) Environ Health Perspect 102(8), 680-688. EDSTAC, F. R. (1998), pp. 71541-71568, Federal Register Notice. McLachlan, J. A. (1993) Environ Health Perspect 101(5), 386-387. Safe, S. H. (1995) Environ. Health Perspect. 103, 346-351. Harrison, P. T. C., Holmes, P. Humfrey, C.D.N. (1997) Sci Total Environ 205, 97-106. Katzenellenbogen, J. A. (1995) Environ Health Perspect 103 Suppl 7, 99-101. Carlsen, E., Giwercman, A., Keiding, N., Skakkebaek, NE. (1995) Environ Health Perspect 103(suppl 7), 137-139. Sharpe, R. M., Skakkebaek, NE. (1993) Lancet 341(8857), 1392—1395 Irvine, D. s. (1994) Br Med J 309(6952), 476. Reijnders, P. J. (1986) Nature 324(6096), 456-457. Guillette, L. J. J ., Pickford, D.B., Crain, D.A., Rooney, A.A., Percival, HF. (1996) Gen Comp Endocrinol 101(1), 32-42. Fry, D. M. (1995) Environ Health Perspect103(suppl 7), 165-171. O'Shea, T. J ., Fleming, W.J., Cromartie, E. (1980) Science 209(4455), 509-560. Von Westemhagen, H., Rosenthal, H., Dethlefsen, V., Ernst, W., Harms, U.L., Hansen, PD. (1981) Aquatic Tax 1, 85-99. Fry, D. M., Toone, CK. (1981) Science 213, 922-924. Crews, D., Bergeron, J. M., and McLachlan, J. A. (1995) Environ Health Perspect 103(Suppl 7), 73-77. Crews, D., Bergeron, J .M., Bull, J .J ., Flores, D., Tousignant, A., Skipper, J .K., Wibbels, T. (1994) Dev Genet 15(3), 297-312. Sumpter, J. P., and Jobling, S. (1995) Environ Health Perspect 103 Suppl 7, 173- 178. Evans, R. (1988) Science 240, 889-895. 42 36. 37. 38. 39. 40. 41. 42. 43. 45. 46. 47. 48. 49. 50. 51. 52. 53. Adlercreutz, H., Hockerstedt, K., Bannwart, C., Bloigu, S., Hamalainen, E., Fotsis, T., and Ollus, A. (1987) J Steroid Biochem 27(4-6), 1135-1144. Hodgert Jury, H., Zacharewski, T. R., and Hammond, G. L. (2000) J Steroid Biochem Mol Biol 75(2-3), 167-176. Chen, S., Kao, Y. C., and Laughton, C. A. (1997) J Steroid Biochem Mol Biol 61(3-6), 107-115. Kao, Y. C., Zhou, C., Sherman, M., Laughton, C. A., and Chen, S. (1998) Environ Health Perspect 106(2), 85-92. Loomis, A. K., and Thomas, P. (2000) Biol Reprod 62(4), 995-1004. Revelli, A., Massobrio, M., and Tesarik, J. (1998) Endocr Rev 19(1), 3-17. Blumberg, B., Sabbagh, W., Jr., Juguilon, H., Bolado, J ., Jr., van Meter, C. M., Ong, E. S., and Evans, R. M. (1998) Genes Dev 12(20), 3195-205. Moore, J. T., and Kliewer, S. A. (2000) Toxicology 153(1-3), 1-10 Htun, H., Holth, L. T., Walker, D., Davie, J. R., and Hager, G. L. (1999) Mol Biol Cell 10(2), 471-486. Hager, G. L., Lim, C. S., Elbi, C., and Baumann, C. T. (2000) J Steroid Biochem Mol Biol 74(5), 249-254. Mangelsdorf, D. J ., Evans, RM. (1995) Cell 83(6), 841-850. Enmark, E., and Gustafsson, J. A. (1996) Mol Endocrinol 10(11), 1293-307. Issemann, I., and Green, S. (1990) Nature 347(6294), 645-650. Wang, L. H., Tsai, S. Y., Cook, R. G., Beattie, W. G., Tsai, M. J ., and O'Malley, B. W. (1989) Nature 340(6229), 163-166. Sladek, F. M., Zhong, W. M., Lai, E., and Darnell, J. E., Jr. (1990) Genes Dev 4(12B), 2353-2365. Giguere, V. (1999) Endocr Rev 20(5), 689-725. Giguere, V., Yang, N., Segui, P., and Evans, R. M. (1988) Nature 331(6151), 91- 94. Song, C., Kokontis, J. M., Hiipakka, R. A., and Liao, S. (1994) Proc Natl Acad Sci U SA 91(23), 10809-10813. 43 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 65. 66. 67. 68. Lala, D. 8., Rice, D. A., and Parker, K. L. (1992) Mol Endocrinol 6(8), 1249- 1258. Zanaria, E., Muscatelli, F., Bardoni, B., Strom, T. M., Guioli, S., Guo, W., Lalli, E., Moser, C., Walker, A. P., McCabe, E. R., and et a1. (1994) Nature 372(6507), 635-641. Seol, W., Choi, H. S., and Moore, D. D. (1996) Science 272(5266), 1336-1339. Johansson, L., Thomsen, J. S., Damdimopoulos, A. E., Spyrou, G., Gustafsson, J. A., and Treuter, E. (1999) J Biol Chem 274(1), 345—353. F Mader, S., Kumar, V., dc Vemeuil, H., and Chambon, P. (1989) Nature 338(6212), 271-274. Umesono, K., and Evans, R. M. (1989) Cell 57(7), 1139-1146. Kumar, V., and Chambon, P. (1988) Cell 55(1), 145-156. * Dahlman-Wright, K., Wright, A., Gustafsson, J. A., and Carlstedt-Duke, J. (1991) J Biol Chem 266(5), 3107-3112. Horlein, A. J ., Naar, A. M., Heinzel, T., Torchia, J ., Gloss, B., Kurokawa, R., Ryan, A., Kamei, Y., Soderstrom, M., Glass, C. K., and et a1. (1995) Nature 377(6548), 397-404. Chen, J. D., and Evans, R. M. (1995) Nature 377(6548), 454-457. Wang, C., Fu, M., Angeletti, R. H., Siconolfi-Baez, L., Reutens, A. T., Albanese, C., Lisanti, M. P., Katzenellenbogen, B. S., Kato, S., Hopp, T., Fuqua, S. A., Lopez, G. N., Kushner, P. J ., and Pestell, R. G. (2001) J Biol Chem 276(21), 18375-18383. Metzger, D., Losson, R., Bornert, J. M., Lemoine, Y., and Chambon, P. (1992) Nucleic Acids Res 20(11), 2813-2817. Tora, L., White, J ., Brou, C., Tasset, D., Webster, N., Scheer, E., and Chambon, P. (1989) Cell 59, 477-487. Montano, M. M., Muller, V., Trobaugh, A., and Katzenellenbogen, B. S. (1995) Molecular Endocrinology 9(7), 814-825. Warnmark, A., Almlof, T., Leers, J ., Gustafsson, J. A., and Treuter, E. (2001) J Biol Chem 276, 23397-23404. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82. 83. Bourguet, W., Ruff, M., Chambon, P., Gronemeyer, H., and Moras, D. ( 1995) Nature 375(6530), 377-382. Renaud, J. P., Roche], N., Ruff, M., Vivat, V., Chambon, P., Gronemeyer, H., and Moras, D. (1995) Nature 378(6558), 681-689. Wagner, R. L., Apriletti, J. W., McGrath, M. E., West, B. L., Baxter, J. D., and Fletterick, R. J. (1995) Nature 378(6558), 690-697. Wurtz, J. M., Bourguet, W., Renaud, J. P., Vivat, V., Chambon, P., Moras, D., and Gronemeyer, H. (1996) Nat Struct Biol 3(1), 87-94. E Brzozowski, A. M., Pike, A. C., Dauter, Z., Hubbard, R. E., Bonn, T., Engstrtim, O., Ohman, L., Greene, G. L., Gustafsson, J. A., and Carlquist, M. (1997) Nature 389(6652), 753-758. Pike, A. C., Brzozowski, A. M., Hubbard, R. E., Bonn, T., Thorsell, A. G., L Engstrom, O., Ljunggren, J ., Gustafsson, J ., and Carlquist, M. (1999) Embo J 18(17), 4608-4618. Matias, P. M., Donner, P., Coelho, R., Thomaz, M., Peixoto, C., Macedo, S., Otto, N., Joschko, S., Scholz, P., Wegg, A., Basler, S., Schafer, M., Egner, U., and Carrondo, M. A. (2000) J Biol Chem 275(34), 26164-26171. Sack, J. S., Kish, K. P., Wang, C., Attar, R. M., Kiefer, S. E., An, Y., Wu, G. Y., Scheffler, J. E., Salvati, M. E., Krystek, S. R., Jr., Weinmann, R., and Einspahr, H. M. (2001) Proc Natl Acad Sci U S A 98(9), 4904-4909. Gee, A. C., and Katzenellenbogen, J. A. (2001) Mol Endocrinol 15(3), 421-428. Egea, P. F., Mitschler, A., Roche], N., Ruff, M., Chambon, P., and Moras, D. (2000) Embo J 19(11), 2592-25601. Mosselman, S., Polman, J ., and Dijkema, R. (1996) FEBS Lett 392(1), 49-53. Kuiper, G. G., Enmark, E., Pelto-Huikko, M., Nilsson, S., Gustafsson, J .A. (1996) Proc Natl Acad Sci U S A 93(12), 5925-5930. Giguere, V., Tremblay, A., and Tremblay, G. B. (1998) Steroids 63(5-6), 335- 339. Cowley, S., Hoare, S., Mosselman, S., and Parker, M. (1997) J Biol Chem 272(32), 19858-19862. Pettersson, K., Grandien, K., Kuiper, G. G., and Gustafsson, J. A. (1997) Mol Endocrinol 11(10), 1486- 1496. 45 84. 85. 86. 87. 88. 89. 90. 91. 92. 93. 94. 95. 96. 97. 98. Pace, P., Taylor, J ., Suntharalingarn, S., Coombes, R. C., and Ali, S. (1997) J Biol Chem 272(41), 25832-25838. Krege, J. H., Hodgin, J. B., Couse, J. F., Enmark, E., Warner, M., Mahler, J. F., Sar, M., Korach, K. S., Gustafsson, J. A., and Smithies, O. (1998) Proc Natl Acad Sci U S A 95(26), 15677-15682. Couse, J. F., Hewitt, S. C., Bunch, D. O., Sar, M., Walker, V. R., Davis, B. J., and Korach, K. S. (1999) Science 286(5448), 2328-2331. Lubahn, D. B., Moyer, J. S., Golding, T. S., Couse, J. F., Korach, K. S., and Smithies, O. (1993) Proc Natl Acad Sci U S A 90(23), 11162-11166. Paech, K., Webb, P., Kuiper, G. G., Nilsson, S., Gustafsson, J ., Kushner, P. J ., and Scanlan, T. S. (1997) Science 277(5331), 1508-1510. Webb, P., Nguyen, P., Valentine, C., Lopez, G. N., Kwok, G. R., McInemey, E., Katzenellenbogen, B. S., Enmark, E., Gustafsson, J. A., Nilsson, S., and Kushner, P. J. (1999) Mol Endocrinol 13(10), 1672-1685. Zou, A., Marschke, K. 8., Arnold, K. E., Berger, E. M., Fitzgerald, P., Mais, D. E., and Allegretto, E. A. (1999) Mol Endocrinol 13(3), 418-430. Hawkins, M. B., Thornton, J. W., Crews, D., Skipper, J. K., Dotte, A., and Thomas, P. (2000) Proc Natl Acad Sci U S A 97(20), 10751-10756. Robinson-Rechavi M, Marcth 0, Escriva H, Bardet PL, Zelus D, Hughes S, and V, L. (2001) Genome Res 11(5), 781-788. Shiau, A. K., Barstad, D., Loria, P.M., Cheng, L., Kushner, P.J., Agard, D.A., Greene, G.L. (1998) Cell 95(7), 927-937. Pike, A. C., Brzozowski, A. M., Walton, J ., Hubbard, R. E., Thorsell, A., Li, Y., Gustafsson, J ., and Carlquist, M. (2001) Structure 9(2), 145-153. Ali, S., Metzger, D., Bornert, J. M., and Chambon, P. (1993) Embo J 12(3), 1153- 1160. Kato, S., Endoh, H., Masuhiro, Y., Kitamoto, T., Uchiyama, 8., Sasaki, H., Masushige, S., Gotoh, Y., Nishida, E., Kawashima, H., and et a1. (1995) Science 270(5241), 1491-1494. White, R., Sjoberg, M., Kalkhoven, E., and Parker, M. G. (1997) Embo J 16(6), 1427-1435. Tanenbaum, D. M. (1998) in Graduate School Vol. 1, pp. 149, Yale univeristy. 46 99. 100. 101. 102. 103. 104. 105. 106. 107. 108. 109. 110. 111. 112. 113. Glass, C. K., and Rosenfeld, M. G. (2000) Genes Dev 14(2), 121-141. Onate, S. A., Tsai, S.Y., Tsai, M.J., O'Malley, B.W. (1995) Science 1995 270(5240), 1354-1357. Halachmi, S., Marden, E., Martin, G., MacKay, H., Abbondanza, C., Brown, M. (1994) Science 264(5164), 1455-1458. Hong, H., Kohli, K., Garabedian, M., and Stallcup, M. (1997) Mol Cell Biol 17(5), 2735-2744. Voegel, J. J ., Heine, M. J. S., Zechel, C., Chambon, P., and Gronemeyer, H. (1996) EMBO J. 1996, 3667-3675. Anzick, S. L., Kononen, J ., Walker, R. L., Azorsa, D. O., Tanner, M. M., Guan, X. Y., Sauter, G., Kallioniemi, O. P., Trent, J. M., and Meltzer, P. S. (1997) Science 277(5328), 965-968. Chen, H., Lin, R. J ., Schiltz, R. L., Chakravarti, D., Nash, A., Nagy, L., Privalsky, M. L., Nakatani, Y., and Evans, R. M. (1997) Cell 90(3), 569-580. Takeshita, A., Cardona, G. R., Koibuchi, N., Suen, C. S., and Chin, W. W. (1997) J Biol Chem 272(44), 27629-27634. Torchia, J ., Rose, D. W., Inostroza, J ., Kamei, Y., Westin, S., Glass, C. K., and Rosenfeld, M. G. (1997) Nature 387(6634), 677-684. Li, H., Gomes, P. J ., and Chen, J. D. (1997) Proc Natl Acad Sci U S A 94(16), 8479-8484. Hankinson, O. (1995) Annu Rev Pharmacol Toxicol 35, 307-340. Gu, Y. Z., Hogenesch, J. B., and Bradfield, C. A. (2000) Annu Rev Pharmacol Toxicol 40, 519-561. Heery, D. M., Kalkhoven, E., Hoare, S., Parker, MG. (1997) Nature 387(6634), 733-736. Darimont, B. D., Wagner, R. L., Apriletti, J. W., Stallcup, M. R., Kushner, P. J ., Baxter, J. D., Fletterick, R. J ., and Yamamoto, K. R. (1998) Genes Dev 12(21), 3343-3356. Zhou, G., Cummings, R., Li, Y., Mitra, S., Wilkinson, H. A., Elbrecht, A., Hermes, J. D., Schaeffer, J. M., Smith, R. G., and Moller, D. E. (1998) Mol Endocrinol 12( 10), 1594-1604. 47 114. 115. 116. 117. 118. 119. 120. 121. 122. 123. 124. 125. 126. 127. Bramlett, K. 8., Wu, Y., and Burris, T. P. (2001) Mol Endocrinol 15(6), 909-922. Routledge, E. J ., White, R., Parker, M. G., and Sumpter, J. P. (2000) J Biol Chem 275(46), 35986-35993. Lee, S. K., Anzick, S. L., Choi, J. E., Bubendorf, L., Guan, X. Y., Jung, Y. K., Kallioniemi, O. P., Kononen, J ., Trent, J. M., Azorsa, D., Jhun, B. H., Cheong, J. H., Lee, Y. C., Meltzer, P. S., and Lee, J. W. (1999) J Biol Chem 274(48), 34283- 34293. Tai, H., Kubota, N ., and Kato, S. (2000) Biochem Biophys Res Commun 267(1), 31 1-316. Xu, J., Qiu, Y., DeMayo, F. J., Tsai, S. Y., Tsai, M. J., and O'Malley, B. W. (1998) Science 279(5358), 1922-1925. Xu, J ., Liao, L., Ning, G., Yoshida—Komiya, H., Deng, C., and O'Malley, B. W. (2000) Proc Natl Acad Sci U S A 97(12), 6379-6384. Spencer, T. E., Jenster, G., Burcin, M. M., Allis, C. D., Zhou, J ., Mizzen, C. A., McKenna, N. J ., Onate, S. A., Tsai, S. Y., Tsai, M. J., and O'Malley, B. W. (1997) Nature 389(6647), 194-198. Baumann, C. T., Maruvada, P., Hager, G. L., and Yen, P. M. (2001) J Biol Chem 276(14), 11237-11245. Alland, L., Muhle, R., Hou, H., Jr., Potes, J ., Chin, L., Schreiber-Agus, N., and DePinho, R. A. (1997) Nature 387(6628), 49-55. Heinzel, T., Lavinsky, R. M., Mullen, T. M., Soderstrom, M., Laherty, C. D., Torchia, J ., Yang, W. M., Brard, G., Ngo, S. D., Davie, J. R., Seto, E., Eisenman, R. N., Rose, D. W., Glass, C. K., and Rosenfeld, M. G. (1997) Nature 387(6628), 43-48. Shang, Y., Hu, X., DiRenzo, J ., Lazar, M. A., and Brown, M. (2000) Cell 103(6), 843-852. Rachez, C., Gamble, M., Chang, C. P., Atkins, G. B., Lazar, M. A., and Freedman, L. P. (2000) Mol Cell Biol 20(8), 2718-2726. Walker, P., Germond, J. E., Brown-Luedi, M., Givel, F., and Wahli, W. (1984) Nucleic Acids Res 12(22), 8611-8626. Klein-Hitpass, L., Schorpp, M., Wagner, U., and Ryffel, G. U. (1986) Cell 46(7), 1053-1061. 48 128. 129. 130. 131. 132. 133. 134. 135. 136. 137. 138. 139. 140. 141. 142. 143. 144. Driscoll, M. D., Sathya, G., Muyan, M., Klinge, C. M., Hilf, R., and Bambara, R. A. (1998) J Biol Chem 273(45), 29321-29330. Hyder, S., Nawaz, Z., Chiappetta, C., Yokoyama, K., and Stance], G. (1995) J Biol Chem 270( 15), 8506-8513. Richard, S., and Zingg, H. H. (1990) J Biol Chem 265(11), 6098-6103. Martinez, E., Give], F., and Wahli, W. (1987) Embo J 6(12), 3719-3727. Kastner, P., Krust, A., Turcotte, B., Stropp, U., Tora, L., Gronemeyer, H., and Chambon, P. (1990) Embo J 9(5), 1603-1614. Weisz, A., and Rosales, R. (1990) Nucleic Acids Res 18(17), 5097-5106. Berry, M., Nunez, A. M., and Chambon, P. (1989) Proc Natl Acad Sci U S A 86(4), 1218-1222. Kato, S., Tora, L., Yamauchi, J ., Masushige, S., Bellard, M., and Chambon, P. (1992) Cell 68(4), 73 1-742. Lee, M. 0., Liu, Y., and Zhang, X. K. (1995) Mol Cell Biol 15(8), 4194-4207. Klinge, C. M., Bodenner, D. L., Desai, D., Niles, R. M., and Traish, A. M. (1997) Nucleic Acids Res 25(10), 1903-1912. Pakdel, F., Le Gac, F., Le Goff, P., and Valotaire, Y. (1990) Mol Cell Endocrinol 71(3), 195-204. Le Drean, Y., Kern, L., Pakdel, F., and Valotaire, Y. (1995) Mol. Cell. Endocrinol. 109, 27-35. Petit, F., Valotaire, Y., and Pakdel, F. (1995) Eur J Biochem 233(2), 584-592. Tan, N. S., Frecer, V., Lam, T. J ., and Ding, J. L. (1999) Biochim Biophys Acta 1452(2), 103-120. Klaunig, J. E., Ruch, R. J ., and Goldblatt, P. J. (1985) In Vitro Cell Dev Biol 21(4), 221-228. de Toledo, F. G., Albuquerque, M. C., Goulart, B. H., and Chini, E. N. (1995) Comp Biochem Physiol C Pharmacol Toxicol Endocrinol 111(1), 93-98. Pohl, R. J ., Bend, J. R., Guarino, A. M., and Fouts, J. R. (1974) Drug Metab Dispos 2(6), 545-555. 49 145. 146. 147. 148. 149. 150. 151. 152. 153. 154. 155. 156. 157. 158. 159. 160. 161. 162. Petit, F. G., Valotaire, Y., and Pakdel, F. (2000) Nucleic Acids Res 28(14), 2634- 2642. Mak, P., Callard, I.P., Callard, G.V. (1983) Biol Reprod 28(2), 261-270. Westley, B., Knowland, J. (1978) Cell 15(2), 367-374. Perlman, A. J ., Wolffe, A.P., Champion, J ., Tata, J .R. (1984) Mol Cell Endocrinol 38(2-3),151-161. Fitzpatrik, D. W., Picken, C. A., Murphy, L. C., and Buhr, M. M. (1989) Comp Biochem Physiol 94C(2), 691-694. Mulvihill, E. R., Palmiter, RD. (1977) J Biol Chem 252(6), 2060-2068. Lazier, C. B. (1987) J Steroid Biochem 27(4-6), 877-882. Turner, R. T. (1984) Gen Comp Endocrinol 54(3), 457-461. Loomis, A. K., and Thomas, P. (1999) Biol Reprod 61(1), 51-60. Lazier, C. B., Lonergan, K., Mommsen, TR (1985) Gen Comp Endocrinol 57(2), 234-245. Nimrod, A. C., and Benson, W. H. (1997) Toxicol Appl Pharmacol 147(2), 38]- 390. Rossini, G. P., Baldini, G.M., Villa, E., Manenti, F. (1989) Gastroenterology 96(4), 1102-1109. Eriksson, H. A. (1982) J Steroid Biochem 17(5), 471-477. Wooge, C., H,. Nilsson, G,M,. Heierson, A., McDonnell, D.P., Katzenellenbogen, BS. (1992) Mol Endocrinol 6(6), 861-869. Horigome, T., Golding, T.S., Quarmby, V.E., Lubahn, D.B., McCarty, K. Sr., Korach, KS. (1987) Endocrinology 121(6), 2099-2111. Tremblay, G. B., Tremblay, A., Copeland, N.G., Gilbert, D.J., Jenkins, N.A., Labrie, F., Giguere, V. (1997) Mol Endocrinol 11(3), 353-365. Koike, S. 8., M. Muramatsum, M. (1987) Nucleic Acids Res 15(6), 2499-2513. Kuiper, G. G., Carlsson, B., Grandien, K., Enmark, E., Haggblad, J ., Nilsson, S., Gustafsson, J .A. (1997) Endocrinology 138(3), 863-870. 50 163. 164. 165. 166. 167. 168. 169. Vonier, P. M., Guillette, J ., L.J., McLachlan, J. A., and Arnold, S. F. (1997) Biochem Biophys Res Comm 232(2), 308-312. Ho, S. M., Fehrer, S., Yu, M., Liang, L.C., Press, D. (1988) Gen Comp Endocrinol 70(3), 382-394. Mak, P., H0, S.M., Callard, LP. (1983) Gen Comp Endocrinol 52(2), 182-189. Matthews, J. B., Celius, T., Halgren, R., and Zacharewski, T. R. (2000) J Steroid Biochem Mol Bio 74(4), 223-234. Matthews, J. B., and Zacharewski, T. R. (2000) Toxicol Sci 53(2), 326-339. Connor, K., Howell, J ., Chen, 1., Liu, H., Berhane, K., Sciarretta, C., Safe, S., and Zacharewski, T. (1995) Fund. Appl. Toxicol. 30, 93-101. Stickle, D. F., Presta, L. G., Dill, K. A., and Rose, G. D. (1992) J Mol Biol 226(4), 1143-1159. 51 CHAPTER 3 Differential binding afl‘mities of PCBs, HO-PCBs and aroclors with recombinant human, rainbow trout (Oncorhynchus mykiss) and green anole (Anolis carolinensis) estrogen receptors using a semi-high throughput competitive binding assay These data have been published in the following article: Matthews, J .B. and Zacharewski, T.R. 2000 Differential binding affinities PCBs, HO- PCBs and Aroclors with recombinant human, rainbow trout (Oncorhvnchus mvkiss) and reptilian (Anolis cafllinensis) estrogen receptors using a semi-high throughput competitive binding assay. Toxicol Sci, 53, 326-339. 52 NariMW“In.“"Eut.aiupfiuiaisg.n..‘q'IElia n‘ k1. "first—7....4 TOXICOLOGICAL SCIENCES 53, 326 -339 (2000) Copyright © 2000 by the Socrety of Toxrcology Differential Binding Affinities of PCBs, HO-PCBs, and Aroclors with Recombinant Human, Rainbow Trout (Onchorhynkiss mykiss), and Green Anole (Anolis carolinensis) Estrogen Receptors, Using a Semi-High Throughput Competitive Binding Assay Jason Matthews and Tim Zacharewski' Department of Biochemistry and National Food Safety and Toxicology Center. Michigan State University, East Lansing. Michigan 48824 —-l3 [9 Received September 8. 1999; accepted October 15. 1999 Acomparativestudy wasnndertakentoassesstheabilityof“ polychlorinated biphenyls (PCBs), 9 hydroxylated PCBs (HO-PCBs), and 8 aroclors at concentrations ranging from 1 MA to 10 M to compete with [’H]l7B—eetradiol (E2) for binding to bacterially ex- prwed fusion proteins using a semi-high throughput competitive- lindingassay.ThefusionproteinsconsistedoftheD,E,andF domains of human (a), cloned reptilian (Anolis carolinensis) and recloned rainbow trout (Onchorhynkiss mykiss) estrogen receptors (ER) linked to the glutathione S-h'ansferase (GST) protein. GST- hERadef (human), GST-aERdef (reptile) and GST-rtERdef (rain- bow trout) fusion proteins exhibited high affinity for EZ with disso- ciation constants (K.,) of 0.4 :t 0.1 nM, 0.7 t 0.2 nM, and 0.6 :1: 0.1 nM, respectively. Of the 44 PCBs examined, only PCBs 104, 184, and 188 effectively competed with [’H1E2 for binding to the GST-rtERdef protein with 1C,o values ranging from 0.4-1.3 M. In contrast, these same congeners only caused a 30% displacement of [’HIEZ in GST- hERaadef and GST-aERdef proteins. Several additional congeners were found to bind to the GST-rtERdef fusion protein, although the degree of interaction varied among congeners. Among the HO-PCBs, 2',3',4',5'-tetrachloro-4-biphenylol and 2,6,2',6'-tetrachloro-4-biphe- nylol bound to all three fusion proteins with IC,« values ranging from 0.1—0.3 MM. Dimethyl sulphoxide (DMSO) concentrations of 20% significantly increased the ability of PCBs 104. 184, and 188 to compete with [’HIEZ for binding to the GST-ERdef fusion proteins. whereas at 20% DMSO, 3 significant reduction in saturable binding was observed. These results demonstrate that ERs from different speciesexhibitdifierentialligandpreferencesandrelativebinding afiinities for PCBs, which can be dramatically affected by DMSO concentration. Key Words: polychlorinated biphenyls; estrogenic endocrine dis- ruptors; estrogen receptor; comparative; competitive binding. In recent years, there has been heightened concern regarding exposure to chemicals in the environment that may interfere ' To whom correspondence should be addressed at Michigan State Univer- sity. Department of Biochemistry, 419 Biochemistry Building, Wilson Road, East Lansing, Michigan, 48824-1319. Fax: (517) 353—9334. E-mail: tzachare®pilot.rnsu.edu. Web site: http://www.bch.msu.edu/~zacharetl tzlabhttn. with the endocrine system and adversely affect normal repro- ductive development and fitness of humans and wildlife (McLachlan and Korach, 1995). These substances, known as endocrine disrupting chemicals (EDCs). are commonly used throughout modern society. In response, the Safe Drinking Water Act and the Food Quality Protection Act were amended to require the United States Environmental Protection Agency (US. EPA) to screen and test for chemicals that mimic or inhibit the activities of estrogen, androgen, and thyroid hor- mones. The Endocrine Disruptor Screening and Testing Advi- sory Committee (EDSTAC) was established and charged with the development and implementation of a screening and testing strategy (EDSTAC, 1998). EDSTAC deliberations resulted in a tiered strategy consisting of prioritization followed by screen- ing and testing approaches that involve a combination of in vitro and in vivo assays. The US. EPA is currently exploring the possibility of using high-throughput receptor binding and reporter gene assays to assist in prioritizing chemicals that require testing. Much attention has been focused on estrogenic endocrine disruptors (EEDs). These chemicals encompass a wide range of substances including natural products, environmental pol- lutants, pharmaceuticals, and industrial chemicals (Colbom, 1993; Katzenellenbogen, 1995). Many of these chemicals do not share any obvious structural similarity to the endogenous ligand for the estrogen receptor (ER), 17B-estradiol (E2), which makes identification based solely on molecular structure difficult (Katzenellenbogen, 1995). It has been hypothesized that many of the effects elicited by estrogenic substances are the result of ER-mediated modulation of gene expression (McLachlan, 1993), although additional modes of action can- not be discounted. The ER is a member of the nuclear receptor superfamily, a family of nuclear proteins that function as transcription factors to modulate gene expression in a ligand-dependent manner (Evans, 1988). As with other members of this family, the ER has a modular structure consisting of six domains (A-F) (Evans, 1988; Tenbaum and Baniahmad, 1997). The highly 326 53 DIFFERENTIAL BINDING OF PCBS TO GST-ERdef FUSION PROTEINS conserved DNA-binding domain (C domain) separates a highly variable NHz-termina] region (A/B domains) and a COOH- terminal 1igand-binding domain (D, E, and F domains). The ER and other steroid hormone receptors are activated by interac- tion with specific ligands that bind with high affinity to the ligand-binding domain. Ligand-occupied ERs undergo ho- modimerization, and the resulting complex binds to its cognate DNA target site. These sites are referred to as estrogen respon- sive elements (EREs) and are located in the regulatory region of estrogen-inducible genes. Once bound to the ERE, the ER-homodimer complex may induce or inhibit gene transcrip- tion. thereby altering the levels of proteins that are important for development, cell proliferation, and the maintenance of homeostasis. Consequently, the ER acts as the primary gate- keeper for initiation of a number of estrogenic responses. Even though the physiological actions of the ER are con- served among different species, amino acid sequences of ii- gand binding domains are variable. This suggests that species may exhibit different responses and sensitivities to EEDs, and that one species may not be an appropriate surrogate for use in identifying and predicting responses in other species. A num- ber of competitive-binding studies have shown that EEDs exhibit differential binding preferences and relative binding affinities for the ER of different species (Connor et al., 1997: Fitzpatrik et al., 1989; Le Drean et al., 1995; Vonier et al., 1997). Polychlorinated biphenyls (PCBS) are a class of halogenated aromatic industrial compounds that are ubiquitous, persistent environmental contaminants detected in almost every ecosys- tem (Bellschmiter et al., 1981). They were commercially man- ufactured as mixtures (e.g. Aroclors) containing varying de- grees of chlorination made up of 140-150 of the 209 possible congeners (Mullin et al., 1984; Safe, 1993; Schulz er al., 1989). PCBS evoke or elicit a number of in vitro and in vivo responses. PCBS. and hydroxylated (HO)-PCBs have been identified in wildlife and humans, and they represent a class of EEDs that differs significantly from the endogenous ER ligand, E2. Co- planar congeners and their planar, mono-ortho substituted de- rivatives induce responses that correlate with their binding affinity for the aryl hydrocarbon receptor (AhR) and evoke 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD)-1ike responses (Safe, 1993). Non-coplanar (di-, tri- and tetra-ortho substi- tuted) PCBS and some HO-PCBs induce multiple responses (Narasimhan er al., 1991: Schuur et al., 1998; Van den Berg et al., 1991) including in vitro and in vivo estrogenic activities independent of the AhR (Bitman and Cecil. 1970; Fielden et al., 1997; Li et al., 1994). In order to investigate the ability of PCBs to compete with E2 for binding to the ER and to identify potential differences in ER binding among species, a comparative study was under- taken in which a semi-high throughput competitive-binding assay, using bacterially expressed GST-ER fusion proteins, was developed. In this study. 44 PCB congeners, 8 commercial Aroclor mixtures, and 9 HO-PCBs, 7 of which have been 327 detected in human serum (Bergman et al., 1994; Moore er al., 1997) were examined for their ability to compete with [3H]E‘2 for binding to the recombinant ERs from human, green anole (Anolis carolinensis) and rainbow trout (Onchorhynkiss mykiss). MATERIALS AND METHODS Chemicals. Polychlorinated biphenyls (PCB) lntemational Union of Pure and Applied Chemistry (lUPAC) it 41, 51, 58, 60, 68, 78, 91. 99, 104, 112. 115, 126. 143. 153. 169. 173, 184. 188, 190, and 193 were purchased from AccuStandard (New Haven. CT) and provided by M. Tysklind (Umea Uni- versity, Umeé. Sweden). PCBS lUPAC it 18, 44, 45, 47, 54, 70, 74, 84. 87. 10]. 128, 138, 151, 158. 168, 177, 178, 183, 187, and 194; Aroclors 1016. 1221. 1232, 1242, 1248, 1254, 1260, and 1268 were synthesized or provided by 8. Safe (Texas A&M University, College Station, TX). Hydroxylated PCBs 2.3.5.3',4’-pentachloro—4-biphenylol: 2.3.5.2'.4',5'-hexachloro-4-biphenylol; 2.4.5.2'.3',4’,5"-heptachloro-3-biphenylol: 3.5,2’,3',4'-pentachloro—4-biphe- nylol; 2,3,5,2',3',4'-hexach|oro-4-bipheny|ol; 2.3,5,2'.3’,4'.5'-heptachloro-4- blphenylol; 2.3.5.6.2'.4',5'.-heptachloro-4-biphcnylol: 2.6.2'.6’-tetrachloro-4- biphenylol; and 2'.3',4’,5'-tetrachloro-4-biphenylol were synthesized by 8. Safe (Connor et al., 1997; Safe et al., 1995). Dimethyl sulphoxide (DMSO) and FIB-estradiol (E2) were obtained from Sigma (St. Louis. MO). [2,4,6,7,16.17-’H] 17B-estradiol ([’H)E2; 123 Ci/mmol) was purchased from New England Nuclear (Boston, MA). MicroScint 20 was obtained from Packard Instruments (Meriden, CT). lsopropylthio-B—D-galactoside (lPTG) and bovine serum albumin (BSA) were obtained from Fisher Scientific (Pitts- burgh, PA). The glutathione S-transferase (GST) expression vector, pGEX6p3, glutathione (GSH) sepharose and XK16 columns were purchased from Am- ersham/Pharrnacia (Piscataway, NJ). SuperScn'pt 1] reverse transcriptase and Trim] Reagent were purchased from Life Technologies (Gaithesburg. MD). Vent DNA polymerase was purchased from New England Biolabs (Beverly, MA), and restriction enzymes and Taq DNA polymerase were obtained from Roche/Boehringer Mannheim (Indianapolis, IN). All other chemicals and biochemicals were of the highest quality available from commercial sources. RNA isolation. Total RNA from a 1 cm’ liver section from a female green anole (Anolis carolinensis: kindly provided by J. Wade, Departments of Psychology and Zoology, Michigan State University) was isolated using Trizol Reagent. The Trizol Reagent procedure is a modification of the single step- RN A isolation method developed by Cbomczynski and Sacchi, (1987). Green anole liver sections were homogenized in the presence of Trizol Reagent. using a Brinkman Polytron homogenizer. Following a 5-min incubation at ambient temperature, chloroform was added and the mixture was separated by centrif- ugation at 12.000 X g for 15 min at 4°C. The aqueous layer containing the isolated RNA was removed and the RNA was precipitated using isopropano]. The RNA was pelleted by centrifugation at 12,000 X g for 10 min at 4°C, and the resulting pellet was washed with 75% ethanol diluted with diethyl pyro- carbonate (DEPC)-treated sterile water. The pellet was then air dried and resuspended in DEPC-treated sterile water. RNA was stored at —80°C until use. Cloning of green anole estrogen receptor DEF domain. Total RNA (5 ug) was incubated for 10 min at 70°C with 500 nM oligo dT primer (PR1r). Following a 5 min incubation on ice, the mRNA was reverse transcribed in a 20 a] reaction mixture containing PCR buffer (20 mM Tris-HCl (pH 8.4), 50 mM KC]. 2.5 mM MgCll), 10 mM dithiothreitol (DTI‘), 500 [AM dNTPs and 200 units of SuperScript 1] reverse transcriptase at 42°C for 50 min. The reaction was terminated with a 15 min incubation at 70°C. The reverse transcription (RT) reaction was then incubated with 1 unit of RNase H for 30 min at 37°C. One tenth of the RT reaction was used in the subsequent PCR reactions. RACE (rapid amplification of cDNA ends) PCR reactions were performed according to the manufacturer's instructions (Life Technologies). Three de- 54 328 MATTHEWS AND ZACHAREWSKI TABLE 1 PCR Primers Used in the Cloning and Recloning of the Green Anole and Rainbow Trout Estrogen Receptor (ER) D. E and F Domains and in the Construction of the GST—ERdef Containing Plasmids . Restriction Primer Sequence enzyme Description Per' 5'—GGCCACGCGTCGACTAGTACT.7-3' Sall” OligodT primer‘ PRZr 5'—CUACUACUACUAGGCCACGCGTCGACTAGTAC-3' Sall Universal amplification primer‘ PR3f“ 5'—TA/CfT CAAGGA/G/T C AC/I' AAC/T GACT ACfT ATGTGC/T C—3' - Degenerate primer based on ER DNA binding domain (human ERa a.a. 213-222) PR4f 5' AAAAGAATTCGGATCCGC/I'I‘AC/TGAAGTAJGGGA/C/ EcoRl Degenerate primer based on ER DNA binding GATGATGAAAGG—3' BamHl domain, upstream of PR3f (human ERa a.a. 244—253) PRSr 5'—GCGGCCGCCTCGAGGA/G/TGAA/GAGA/GATGAGGA- Notl Degenerate primer based on ER hormone GGAGGAGCT-3' Xhol binding domain (human ERa a.a. 517—506) PR6f 5'—CAAAGAATT@AT$CATGTTGCTGGCCGCFICITCTC—3' EcoRI Anole ER specific primer BamHI PR7f 5'—AAAAQGATQQATGTI'GAAACACAAGCGCCAGAGAG—3' BamHI Human ERa DEF forward primer (pGEX) PR8r 5'—AAAACTCGAGTCAGACTGTGGCAGGGAAACCCT—3' Xhol Human ERa DEF reverse primer (pGEX) PR9f 5'-AAAAGGATCCATGCTGAAACACAAACGTCAAAGAG—3’ BamHI Anole ER DEF forward primer (pGEX) PR10r 5'—AAAAGGATCCCTCGAGTCAAATI‘GC'ITCCTGCTCAT- BamHl Anole ER DEF reverse primer (pGEX) TTCCC—3' Xhol PR1 1f 5'-AAAAGGATCQCI‘CGAGGGCGGGTTCTCAGGGATAAGCG—3' Xhol Rainbow trout ER DEF reverse primer (pGEX) PR12r 5'—AAAACTCGAGTCACGGAATGGGCATCTGGTCTG—3’ Xhol Rainbow trout ER DEF reverse primer (PGEX) ‘ “r" denotes reverse primer. "' Restriction enzyme sites are underlined. ‘ Commercially available primers provided in the 3'RACE kit (Life Technologies). ' “i" denotes forward primer. generate primers were used in the cloning strategy. The oligonucleotides were identified from a consensus sequence derived by a multiple sequence align- ment of the ERs from 10 different species. Two of the primers (PR3f and PR4f; see Table l) were based on the highly conserved ER DNA binding domain and the third primer. PR5r. was derived from a highly conserved region in the ligand binding domain (LBD). Optimal PCR reaction conditions were deter- mined to be 20 mM Tris-HCI (pH 8.4), 50 mM KC]. 3 mM MgC12. 200 nM of each primer. 200 uM dNTPs and 2.5 units of Taq polymerase. Following the addition of template. the samples were incubated at 95°C for 2 min and amplified for 35 cycles. Each cycle included: 1 min denaturation at 95°C, 1 min annealing at 62°C and 2 min elongation at 72°C. Ten percent of the first strand synthesis reaction was PCR amplified using primers PR3f and PR2r. Using a nested PCR strategy. a 2 ul aliquot of the initial PCR reaction was used as a template for a subsequent PCR amplification using primers PR4f and PR2r. The products from the second round of PCR were used as a template in a third PCR reaction using primers PR4f and PR5r. This reaction produced a fragment of approximately 800 bp, which was digested with BamHI and Xhol. cloned into the eukaryotic expression vector, p111 and sequenced using ABI/Prism automated sequencing (Perkin Elmer Applied Biosystems; Foster City. CA). Based on this sequence, a green anole ER-specific primer (PR6f) was designed and used with PR2r to amplify the 3’ end of the ER using a 2 id aliquot of the PR4f/PR2r PCR reaction as template. The resulting llOO-bp product was cloned into pGEM plasmid (Promega) and sequenced using ABl/Prism automated sequencing. The boundaries of the green anole ER D. E and F domains were determined by sequence alignment to the human ERa. Sequence analysis was performed using MacVector 6.5 and the GCG Wiscon- sin Package (Oxford Molecular Ltd.. Beaverton OR). Rec-Inning of rainbow trout ER DEF domains. Total RNA (1 #3) from the liver of a female rainbow trout (Onchorhynkiss myki'ss; kindly provided by S. Wagner. Department of Physiology. University of Western Ontario) was reverse transcribed as previously described (Gillesby and Zacharewski, 1999) in the presence of first-strand buffer (50 mM Tris-HCI (pH 8.3). 37.5 mM KCI, 1.5 mM MgCl;). 10 mM DTT. 500 nM reverse primer (PR14r). 1 mM dNTPs, and 40 units of SuperScript II. The RT reaction was incubated for 10 min at 25°C . followed by 50 min at 42°C. 15 min at 70°C and 5 min at 4°C . The entire cDNA product produced in the RT reaction was used in subsequent PCR reactions. The PCR reaction mixture. containing Thermopol buffer (20 mM Tris—HCI (pH 8.8). 10 mM KCl. 10 mM (NH.),SO.. 2 mM MgSO.. and 0.1% Triton X-lOO). 200 uM dNTPs, 1 mM MgSO.. 2 uM primers (PRl3f/PR14r). and 1.25 units of Vent DNA polymerase was amplified for 30 cycles using the following conditions: 94°C for l min. 60°C for l min, and 72°C for 2 min. Construction of GST-ERdef fusion proteins. The plasmids pGEXhERadef. pGEXaERdef. and pGEXrtERdef were constructed by PCR amplification of the human ERa (kindly provided by P. Chambon, INSERM U184. Strasbourg. France). green anole. and rainbow trout ER DEF domains using primers PR7f/PR8r, PR9f/PR10r, and PR1 lf/PR12r, respectively. The fragments were digested with the appropriate restriction enzymes (see Table l) 55 .n .a. . la FIJ .Lf... t n.1— mfl no. DIFFERENTIAL BINDING OF PCBS TO GST-ERdef FUSION PROTEINS and ligated into the GST fusion protein expression vector, pGEX6p3. The PCR amplification was performed using Vent DNA polymerase (New England Biolabs) as described above. The PCR reaction mixture containing Thermopol buffer. 200 [AM dNTPs, 1 mM MgSO., 500 nM primer. and 1.25 units of polymerase was heated to 94°C for 5 min followed by 35 rounds of 94°C for 45 s. 60°C for 45 s. and 72°C for 1 min 45 s. The sequence of each construct was confirmed with restriction enzyme digest and ARI/Prism automated se- quencing. Expression and purification of GST ER fusion proteins. Overnight cul- tures of E. coli strain BL21 (AmershamlPharmacia) containing pGEX-ERdef constructs were diluted 1:100 in 500 ml of LB broth (1% (w/v) tryptone. 0.5% (wlv) yeast extract. and 1% (w/v) NaCl. pH 7.5) containing 100 ug ampicil- lin/ml and incubated at 37°C with constant shaking. The cells were grown to an optical density of approximately 1.0 at 600 nm. and induced with IPT G at a final concentration of 1 mM. The induced cultures were incubated for 4 h at 37°C. then pelleted by centrifugation at 1000 X g for 10 min at 4°C. Cell pellets were resuspended in 25 ml of buffer A (50 mM HEPES. 3 mM EDTA. 5 mM DTT. 50 mM NaCl. and 10% (v/v) glycerol. pH 7.5) containing 0.1 mg/ml lysozyme. 100 ug/ml phenylmethylsulfonyl fluoride (PMSF). 10 ug/ml 1eupeptin.and 10 rig/ml pepstatin A. Cells were then lysed by sonication on ice for 3 X 15 s. separated by 10 5 intervals. Tween20, to a final concentration of 0.1%, was added to the cellular debris and incubated for 30 min at 4°C under constant shaking. Cell debris was pelleted by centrifugation at 20.000 X g for 40 min at 4°C. Supematants were stored at -80°C until further use. The supematants containing the GST fusion proteins were applied to an XK16 column containing GSH sepharose pre-equilibrated with buffer A at a constant flow rate of 0.5 mllmin at 4°C. After adsorption of the protein. the GSH sepharose was washed with 100 ml of buffer B (50 mM HEPES. 3 mM EDTA. 5 mM DTT, 150 mM NaCl and 10% (v/v) glycerol. pH 7.5). Bound proteins were eluted in 25 ml of buffer C (50 mM HEPES. 3 mM EDTA. 5 mM DTT. 150 mM NaCl and 10% (v/v) glycerol. pH 8.0) containing 10 mM GSH. The partially purified protein was concentrated to a l-ml final volume using Millipore Ultrafree-15 filter columns with a 50-kDa molecular weight cutoff (Millipore Corp. Bedford MA). Protein concentration was determined using the Bradford (1976) method. Protein was diluted to 0.5 mg/ml and stored at -80°C until further use. Partially purified fusion proteins were separated by SDS-PAGE according to Laemmli (1970), using a 4% stacking and 10% separating gel. Proteins were visualized by Coomassie brilliant blue R250 staining. Receptor binding assays. Partially purified GST-ERdef fusion proteins were diluted in TEGD buffer (10 mM Tris pH 7.6, 1.5 mM EDTA. 1 mM DTI‘ and 10% (v/v) glycerol) containing 1 mg/ml bovine serum albumin (BSA) as a carrier protein. and incubated at 4°C for 2 h with 0.1—3.5 nM [’H]E2 in 1 ml glass tubes arranged in a 96-well format (Marsh Scientific. Rochester. NY). Fusion protein preparations were diluted to ensure 10.000 dpms of total binding (dilutions varied from 750-3000-fold). Binding assays were initiated by adding 240 pl of protein preparation to glass tubes containing 5 ul of DMSO and 5 ul of [’HlEZ; thus. the solvent concentration did not exceed 4%. unless stated otherwise. The amount of nonspecific binding was determined in the presence of a 400-fold excess of unlabeled EZ. Bound [’H]E2 was sepa- rated from free using a 96-well filter plate and vacuum pump harvester (Packard Instruments). Filter plates containing the protein were washed with 3 X 50 ml of T50 (10 mM Tris buffer (pH 7.6). 1.5 mM EDTA. and 10% (v/v) glycerol) and allowed to dry under continued suction for 30 5. After drying. the underside of the filter plates were sealed and 50 ul of MicroScint 20 scintil- lation cocktail was added to each well. Bound [’H]E2 was measured using a TopCount luminescence and scintillation counter (Packard Instruments). Competitive-ligand-binding assays were performed essentially as described above with the following modifications. Partially purified GST-ERdef fusion protein was diluted in TEGD containing 1 mg/ml BSA and was incubated with 2.5 nM [’HJEZ (5 ul aliquot) and increasing concentrations of unlabeled competitor. PCB (1.0 nM-IO MM. 5 id aliquots) at 4°C for 2 h. Bound [’HJEZ was separated from free as described above. Nonspecific binding was deter- mined in the presence of a 400ofold excess of unlabeled E2. Each treatment I GATACCAAGTGGGGATGATGAAAGGTGGAATTCGGAAAGACCGCAGASGAGGCCGTATGC 20 Y F V G M H K G G I R K D R R G G P A 1 61 TGAAACACAAACGTCAAAGAGAGGAAAATGATAGCAGGAATGCASGGGCTTTAACAGASG 40 K h K R O R E E N D S R N A G A L T E A 2! CGAGGAGCACGGCTCTCTGGCCAAGTCCCCTGATGATCAAACATAGCAAAAASAACAGCC 6C R S T A L N P S P L H i K H S K K N S P 181 CAGCCTTSTCTCTCACTCCACATCACATCGTCACTGCTTTGCTGGAGGCAGAGCCACCTG 8C A L S L T A D O M V S A L L E A E P P V TSTCTACTCSGMTATUAC."CTALK'AGA‘ '(IT'I‘TCAATUAAGCT ".‘CACTAAI GACAC If.“ 1': {7 V Y S E Y C P S R P F N E A S V M T l L 301 rAACCAACCTTGCTCACAGASAACTGGTGCACATGATCAACTGGGCAAAAAGAGTTSCAG 123 T N L A U R L L V H H i N u A K R V P G GATTTSTGGATTTAGCACTCCATGATCAGGTCCATCTGCTGGAATGTGCCTQUTTAGACA F V 3 L A L H D O V H L L E C A I L E I TACTGATGGTTGGCTTGGIUTGGCGATCAATGGAGCATCCAGGAAAACTGTTCTTTCCCC L M V G L V H R S M E H P G K L L F A P CTAACTTACTATTCCACAGGAGCCATGGGAAAGTTCTTGAGGGTTTTGTGSAAATATTTG 183 N L L L D R S H G K V V E G F V E l F J ACATGLTbLiubLLULTIL.TL.LGL117LhAATCATCAATCTACCCCCACAAGAATTTG M L L A A S S R F R M M N v R G E E F V 63; TGTGCCTTAAATCCATCATCTTACTCAATCCCGGAATCTATACATATCTTTFCAGTACC? 273 CLKSIXLLNPGIYTYLSSTL TAAAATCAGTGGAGGAAAGGGATCATATCCACCSTCTCCTAGATAAAATCACASACACGT K 5 V E E R D M I H R V L 3 K I T D I L TGATGCATTTGATGSCCAAATCGGGTCTCTCTCTACACCAGCAGCATAGGCGACTCGCTC 560 H L M A K S G L S L C Q 0 H R R L A Q AGCTTCTTCTCATTCTTTCCCATATCAGGCACATGAGCAACAAGGGAATGGAGCA:CTTT 282 L L L I L S H I R H M S N K 5 M L H Y ATACCATGAAGTGCAAGAACGTGGTCCCTCTATATGATCTGTTACTCGAGATGCTAGATG S M K C K N V V P L Y D L L L E H L 3 A 30( CTCACCGCTTGCATGCCCCTGCAGCAAAAGGCAGCCCTCCAAGTGAAGATGATCCTCTAA 320 NHLnAF‘AAKGSPPSEUDPI.N 961 ATCAGTTGGCCGTCCCATCCCCTTCARTGCATTCCTTGCTACCTTGTTATGTGAACAAA: 14c Q L A V P S P 5 M N 5 L L F C r v N K D AGGAAGAGGGAAATGACCAGSAAGCAATTTGAGASTCTACATTTLAGTSAGCCCAGAYGT 369 E E G N E U E A I end 1C2! TTCTGAGAACCCCATCAGAAGAVCAPCAACTTCAGCSTTACCTGAACTTCTACTATGTGT AAGCACTCCAGTCCTCACCAATCGCAATTGCCAAGCTGATTGCTTAAATCCTTAAACGAA 10:4:- 114'. FIG. 1. Partial nucleotide sequence of the green anole ER. The first amino acid of the D domain is underlined. The boundaries of the domains were determined from an alignment to human ERa. The numbers on the left indicate the nucleotide position, and the values on the right refer to the amino acids (GenBank accession number AFO95911). was performed in quadruplicate and results are expressed as percent specific binding of ["HlEZ versus log of competitor concentration. 1C” values were determined from nonlinear regression for single site competitive binding analysis. using Equation 1. Y = bottom + (top — bottom)/(l + 10"‘4°"°‘°’) (l) The reported 1C“, values represent the concentration of test compound required to displace 50% [’H]E2 from the GST-ER fusion proteins as compared to the 50% displacement of [’HlEZ achieved by unlabeled E2. These analyses were performed using GraphPad Prism 3.0 software (GraphPad Software Inc. San Diego. CA). RESULTS Isolation of Green Anole ER DEF cDNA Figure 1 shows the partial nucleotide sequence of the green anole ER (aER). The region spanning the D. E and F domains 56 This is to certify that the dissertation entitled Molecular Basis of Estrogenic Endocrine Disruptor—Estrogen Receptor Interactions: A Comparison Among Species presented by Jason Bruce Matthews has been accepted towards fulfillment of the requirements for Ph.D. degree in Biochemistry and Molecular Biology 9//€ / "till...- 1“- .a' A - '- ll'L . , 1 rr TABLE 2 Major professor 042771 D m! to the human ERa e D. E and F domains terms of amino acid as the most conserved Comparison of the Amino Acid Sequences from the D. E. and F Domains of Estrogen Receptor from Rainbow Trout Taken from Four Different Sources. Boundaries of the D. E and F Domains Were Determined from a Sequence Alignment to the Human ERa Type of amino-acid sequence difference Source of Rainbow trout ER sequence Frame shift Single amino acid differences Pakdel er al. (1990) 24l-KI’I’VTWSTGQRPQ-253‘ 287-QGQTPA-292 399-RFGML-403 419-LNPGA-423 514-LNAPG-518 GenBank Accession number AF099079 18- KPYGDLEHRTAEPQJI“ 64-QGAEI_’PA-70 178-RFRML-182 l98-LNSGA—202 293-LQSPG-297 GenBank Accession number N242740' 287~KPYGDLEHRTA£PQ~300 333-QGAEEPA-339 447-RFRML—451 467-LNSGA471 562-LQSPG-566 GenBank Accession number A124274ll 242-KPYGDLEHRTAEPQ-255 288-QGAE_PPA-294 402-RFRML-406 422~LNSGA~426 5|7-LQSPG-521 ‘ The numbers correspond to the amino acid residues in the sequence. which only includes the D. E and F domains. ’ Amino acid residues found to be different among the sequences are shown in bold. ' Inserted amino acid residues are shown underlined and in bold. ‘ Recently deposited in GenBank by the author who originally published the rainbow trout ER sequence. 57 DIFFERENTIAL BINDING OF PCBS TO GST-ERdef FUSION PROTEINS 331 a o---- awn Mama-we B . u m "L , w‘ I“ hDu-l-‘I' .“ m M m 9" M D . i w . l ”3;. :1? a; (be-Mud l.$| I. at” |“ 'g‘ll‘ “eh-f “6‘“!!“ 4s — i :33 .‘f 3% ' I .2‘. W” “M 31 in I at and on“); :: i‘ a I .~ " mm «m» FIG. 3. 31.5 (A) A schematic representation of the GST-ER fusion proteins expressed in bacteria. Amino acid residues that link the 2 proteins are provided above the transition point. The residues shown in italics represent glutathione S-transferase (CST). The first 3 residues of the ER ligand binding domain (DEF domains) are shown in regular text. Numbers provided above identify the amino acids used in the construction. The numbers within the domains represent % identity while those in parentheses represent % similarity to hERa. ' Only a portion of the green anole ER sequence was cloned. and the start of the D domain corresponds to amino acrd residue 19 in the cloned sequence. ' Refers to the first amino acid of the recloned rainbow trout ER partial sequence (GenBank accession number AF099079). (B) SDS-PAGE analysts of the GST-ERdef fusion proteins purified using GSH affinity chromatography. Lane A: 5 pg of partially purified GST-hERadef (predicted mw = 64.2 kDa. Lane 8: 5 ug of partially purified GST-aERdef (predicted mw = 64.3 kDa. Lane C: 5 pg of partially purified GST-nERdef (predicted mw = 65.5 kDa). Proteins were analyzed using a 4% stacking and a 12% separating gel stained with Coomassie brilliant blue R250. (89% sequence identity and 92% sequence similarity) while the D (57% sequence identity) and F (32% sequence identity) domains were less conserved. Recloning of Rainbow Trout ER DEF cDNA The rainbow trout ER (rtER) D. E, and F domains were recloned using primers based on the previously published sequence (Pakdel et al., 1990). Upon sequencing the frag- ment, significant differences with the published sequence were observed (Table 2). and these differences were con- firmed by sequencing 7 independent clones from 3 different animals (data not shown). The recloned sequence (GenBank accession AF099079) is in agreement with 2 unpublished rtER sequences recently deposited in the GenBank database (GenBank accession AF242740 and AF242741) by the au- thors, who originally published the rtER sequence (Table 2). Combined. the D, E. and F domains are 40% identical and 47% similar in terms of amino acid composition. to the hERa. The E domain was the most conserved (60% se- quence identity and 66% sequence similarity) while the D and F domains were less conserved (18% and 19% sequence identity. respectively). Purification of GST-ERdef Fusion Proteins The purity of the GST-ERdef fusion proteins was qualita- tively determined to be approximately 85%. based on a Coo. massie-stained SDS—PAGE analysis (Fig. 38). All 3 fusion proteins migrated according to their predicted molecular weights (mw): GST-hERadef (64.2 kDa). GST-aERdef protein (64.3 kDa), and GST-rtERdef (65.5 kDa). The recovery of the fusion proteins varied among species. with yields ranging from 1-3 mg/l. As shown in Figure 3B. lane 1. the GST-hERadef fusion migrated as a doublet. The higher MW band most likely represents the full length product. whereas the lower band may result from proteolytic cleavage. Although the GST-aERdef and GST-rtERdef fusions did not appear to migrate as dou- blets, additional higher and lower MW proteins co-purified with the fusion proteins. Characterization of the Purified GST-ERdef Fusion Proteins Binding affinities of the partially purified GST-ERdef fusion proteins for 132 were determined by saturation analysis (inset, Fig. 4) and linear transformation of the data (Scatchard, 1949) (Fig. 4). Differences in the amounts of receptor required to attain the desired 10.000 dpm at saturation were species- dependent and may be due, in part. to differences in protein purity, functionality. and level of expression between prepara- tions. All GST-ERdef fusion proteins exhibited high binding affinity for E2. with dissociation constants (K.,) of 0.4 t 0.1 nM, 0.7 .+_ 0.2 nM, and 0.6 t 0.1 nM for GST-hERadef. GST-aERdef and GST-rtERdef, respectively. These values are means 1‘ standard deviations from 4 independent experiments. Ligand Binding Analysis A set of 44 PCBS. 8 Aroclors, and 9 HO-PCBs were exam- ined for their ability to compete with [’H]E2 for binding to the 58 332 0.30 ' 0.20 T 0.10 " 0.10 "' Bound/ Free 0.00 " K, - 0.4 10.: 0M 0.000 0.070 0.100 Bound (nM) 00‘ I I 0.020 0.000 w 0.10 W g 0.10 J E 0.10 - 0 CD 0.054 K, = 0.7 1: 0.2 nM - o.” I f I I V 0.000 0.020 0.000 0070 0.100 0.125 Bound (nM) 0.20 1 0 010 - 1 E E 0.10 - ca 0.00 - K., - 0.6 t 0.1 nM 0.00 . fl . . 0.000 0.020 0.000 0.070 0.100 Bound (nM) FIG. 4. Saturation analysis of (A) GST-hERadef. (B) GST-aERdef. and (C) GST-rtERdef fusion proteins. Various concentrations of [’HlEZ were incubated with known amounts of partially purified fusion protein for 2 h at 4°C as described in Materials and Methods. Saturation data shown in the inset were plotted by the method of Scatchard. The reported K. values were averaged from 4 independent experiments. MATTHEWS AND ZACHAREWSKI 140 120- f3: 1001 3 8 80- 3 a so- ..= — 4o- 20- O I I I r I I j 42 -11 -10 -0 -0 -7 -0 -5 4 Log concentration competitor (M) FIG. 5. Classification of PCB interactions with GST-ERdef fusion pro- teins. An aliquot of partially purified GST-rtERdef was incubated with 2.5 nM [’HlE, and increasing concentrations of unlabeled competitor [filled circle] E2 (positive contro|)—exhibits full displacement curve with lC,o 2.4 t 1.0 nM. [empty square] PCB 77—non-binder (nb). did not compete with [’HIEZ for binding to GST-ERdef fusion protein. [filled square] PCB 153—weak binder (wb). only lO-30‘Ir displacement of [’H]E2 from GST-ERdef fusion protein was observed at the highest dose examined. X PCB 9l—congener capable of displacing at least 50% [’HJEZ at the highest examined dose examined from GST-ERdef fusion proteins however characteristic displacement was not achieved 1C” > 10 nM. and [empty circle] PCB 184 exhibits full displace- ment curve with EC” = 3.1 I 0.6 nM. The results (mean and standard deviation) are from a representative experiment that was repeated 3 times. GST—ERdef fusion proteins using a semi-high throughput com- petitive-binding assay. Protein preparations were incubated with 2.5 nM [’HlEZ and increasing concentrations of PCBs (1.0 nM- 10 MM) for 2 h at 4°C. Kinetic studies with E2 showed that a 2-h incubation was sufficient to reach full saturation (data not shown). Figure 5 shows binding profiles of 4 representative PCB congeners to the GST-rtERdef fusion protein. and illustrates the criteria used to classify competitive binding. The binding patterns observed for PCB 77 and PCB 153 represent PCB congeners that are classified as non-binders (ab) and weak binders (wb), respectively. A PCB congener was classified as a non-binder if less than 10% competitive binding was ob- served; similarly. a PCB congener was classified as a weak binder if only 10%-50% of [3H]E2 was displaced at the highest concentration of competitor examined (10 nM). PCB 91 cf- fectively displaced 50%-70% [’H]E2 from the GST-rtERdef, however, a characteristic one-site competitive displacement curve was not achieved. Consequently, an 1C,o greater than the highest concentration of test compound was ascribed. PCB 184 effectively competed with [’HlEZ. displacing more than 80% [’HlEZ from the fusion protein and an IC,o value was calcu- lated using Graphpad Prism 3.0. Concentrations greater than 10 59 DIFFERENTIAL BINDING OF PCBS TO GST-ERdef FUSION PROTEINS TABLE 3 Summary of the Ability of 44 Polychlorinated Biphenyls to Compete with [’Hll7B-Estradiol for Binding to Recombinant Fu- sion Proteins Containing the Human, Green Anole and Rainbow Trout Estrogen Receptor D, E, and F Domains PCB‘ Substitution GST- GST- GST- (IUPAC) pattern hERadef aERdef nERdef Mono—ortho 58 2.3.3',5'- nb" nb wb‘ 60' 2.3.4.4'- nb nb wb 68 2,4.3'.5'- nb nb wb 70‘ 2.5.3'.4'.- nb nb wb 74‘ 2.4.5.4’- nb nb wb Di-urtho l8‘ 2.5.2'- nb nb wb 41 2.3.4.2’- nb nb >10 uM‘ 44‘ 2.3.2'.3'- nb nb wb 47' 2.4.2'.4'- nb nb >10 itM 49‘ 2.4.2'.5'- nb nb wb 87‘ 2.3,4,2’.5'- nb nb nb 99‘ 2,4,4,2',4'~ nb nb wb 101‘ 2,4,5.2’.5'- uh uh wb 112 2,3.5,6.3’- nb nb wb 115 2.3,4.6.4'- nb nb >10 uM 128' 2.3.4.2'.3'4’- nb nb wb 138‘ 2,3,4,2'.4',5'- nb nb wb l53' 2.4.5.2'.4'.5'- nb nb wb 158 2.3,4.6.3'.4'- nb nb nb 168 2.4.6.3'.4’,5'- nb nb nb 190 2.3.4.5,6.3'.4’- nb nb nb 193 2,3.5,6.3'4',5'- nb nb nb I94” 2,3.4.5.2',3'.4'.5'- nb nb nb Tri-orrho 45' 2.3.6.2’- nb nb >10 uM 51 2.4.2'.6'- nb nb >10 piM 84‘ 2.3.6.2'3’- nb nb wb 9i” 2.3.6.2',4'- nb nb >10 uM 95‘ 2.3.6.2'5'- uh uh wb 143 2.3.4.5.2'.6’- nb nb >10 p.M 149' 2.6.2'.4'5- nb nb wb 151‘ 2.3.5.6.2'.5'- nb nb wb 173 2.3.4.5.6.2’.3'— nb nb >10 uM 177‘ 2.3,4.2',3'.5',6'- nb nb >10 MM 178 2.3.5,6.2'.3'.5'- nb nb wb 183‘ 2.3,4.6.2',4'.5'- nb nb wb 187' 2.3.5.6,2'.4'.5'— uh uh wb Tetra-orrlw 54 2,6.2'.6'— nb wb >10 “M 104 2.4.6.2'.6'- >10 “M >10 uM 1.3 t 0.6 M’ 184 2.3.4.6,2'.4',6'- >10 uM >10 M 0.4 I 0.! uM l88 2,3.5.6.2'.4'.6' >10 uM >0 FM 1.3 t 1.2 uM Co—planar 77‘ 3,4.3'.4'- nb nb nb 78 3.4.5.3'- nb nb nb 126‘ 3 .4.5.3' .4' - nb nb nb 169 3.4.5.3'.4'.5' nb nb wb ' PCBs were examined at l nM—lO uM concentrations. ’Denotes non binder (nb) since no significant displace of radiolabeled l7B—estradiol was observed at the highest dose examined (10 uM). following replicate experiments. 333 120 G 100- B 8 8 30- «*3 a toe E _' 40" 20- o- I I I I I I I -12 -11 -10 -0 -0 -7 -0 -5 4 Log concentration of competitor (M) FIG. 6. Representative competitive binding curves for [filled circle] 52 to GST-hERadef fusion protein and unlabeled PCB 47 with [empty circle] GST-nERdef. [filled square] GST-hERadef and [empty square] GST-aERdef fusion proteins. An aliquot of partially purified GST-ERdef fusion proteins was incubated with 2.5 nM [’H]EZ and increasing concentrations of unlabeled E2 or PCB 47 as described in Materials and Methods. The results (mean and standard deviation) are from a representative experiment that was repeated 3 times. ptM were not examined, due to potential solubility limitations of the test compounds. Table 3 summarizes the competitive binding ability of 44 PCB congeners to GST-hERadef, GST-aERdef and GST- rtERdef fusion proteins. Only PCBS 104. 184, and 188 com- peted with [’HlEZ for binding to GST-hERadef and GST- aERdef fusion proteins. The remaining compounds did not significantly bind to either fusion protein. In contrast, several PCBs competitively bound to GST- rtERdef. All 5 mono-ortho-substituted PCBS (PCB 58, 60, 68, 70. and 74) weakly interacted with GST-rtERdef fusion protein and displaced less than 30% of the radioligand. Nine of the 18 di-ortho—substituted congeners (PCB 18, 44, 49, 99. 101, 112, 128, 138. and 153) exhibited weak competitive binding pro- ‘ Denotes weak binder (wb) lO-50% displacement of radiolabeled 17]?- estradiol was observed at the highest examined dose (10 M) in two separate experiments. ‘ Environmentally relevant congeners as described in McFarland and Clarke (1989) and Hansen (1998). ‘ Uncharacteristic single site competition curve with greater than 50% displacement observed at the highest dose examined from three separate experiments. thus an IC” greater than 10 pM was assigned. ’ Characteristic single site competition curve with greater than 80% dis- placement of radiolabeled INS-estradiol observed at the highest dose examined (10 uM). The 1C,o was determined as described in Materials and Methods and represents the averages and standard deviations from 3 independent experi. ments. 6O 334 MATTHEWS AND ZACHAREWSKI TABLE 4 Receptor Binding Activity of 9 Hydroxylated Polychlorinated Biphenyls to Each of the Recombinant Fusion Proteins Containing the Human, Green Anole, and Rainbow Trout Estrogen Receptor D, E, and F Domains HO-PC B Substitution pattern GST-hERadef GST-aERdef GST-rtERdef 1‘ 2,3,5,3 ' ,4'-pentachloro—4-biphenylol rib" rib nb 2‘ 2.3.5.2',4',5’-hexachloro—4—biphenylol rib rib nb 3‘ 2,4.5,2’,3’,4',5'-heptachloro-3-biphenylol nb rib rib 4‘ 3,5,2',3',4'-pentachloro-4-biphenylol nb rib rib 5' 2.3.5.2'.3'.4’-hexachloro—4—biphenylol rib rib nb 6‘ 2.3.5.2'.3',4',5'-heptachloro-4-biphenylol nb rib nb 7' 2.3.5.6.2'.4'.5’-heptachloro-4-biphenylol nb rib nb X' 2',3',4'.5'-tetrachloro—4-biphenylol 0.1 $0.02uM‘ 0.25 IO.O9;LM 0.27:0.02uM 54‘ 2,6.2',6'-tetrachloro—4-biphenylol 0530.02 pM 0.5 :02 11M 0.3 2:0.ltiM ‘ Nomenclature taken from Moore er al., 1997. ' Denotes non binder (rib) since no significant displace of radiolabeled FIB-estradiol was observed at the highest examined dose (10 MM), following replicate experiments. ‘ Greater than 80% displacement of radiolabeled l7B—estradiol observed at the highest dose examined (10 MM). The [Cu was determined as described in Materials and Methods and represents the average and standard deviation from 3 independent experiments. ’ A hydroxylated metabolite of the polychlorinated biphenyl IUPAC it 54. files. while PCB 41, 47. and 115 displaced between 50 and 70% [’H]E2 from the GST-rtERdef fusion protein. Of the 13 tri—orrho-substituted PCBS examined, 7 congeners (PCB 84, 95, 149, 151, 178, 183, and 187) weakly competed with [’H]E2, and 6 congeners (PCB 45, 51, 91, 143, 173, and 177) displaced between 50 and 70% of [’HlEZ from the GST- rtERdef fusion protein. The most active congeners contained 4 orrho-chloro substituents; PCB 104, 184, and 188 were strong binders, while PCB 54 only displaced 50-70% [’HlEZ from the fusion protein. None of the 4 co-planar (non-ortho substi- tuted) PCBS examined effectively competed with [’HlEZ for binding, with the exception of PCB 169, which demonstrated a weak interaction with the GST-rtERdef fusion protein. The differential binding of PCB 47 to the GST-ER fusion proteins is shown in Figure 6. PCBS 41, 45, 51, and 91 showed similar binding curves across the species (data not shown). Results in Table 4 summarize the competitive binding of 9 hydroxylated PCB congeners to GST-ERdef fusion proteins. HO-PCB X and HO-PCB 54 competitively bound to all 3 fusion proteins with similar affinity (1C,0 values 0.1-0.3 uM). None of the 8 Aroclor mixtures competitively bound to any of the GST-ER fusion proteins at the highest concentration tested (10 W). Higher concentrations were not examined due to possible solubility limitations. Table 5 summarizes effects of DMSO on competitive bind- ing of E2 and PCB to GST-ERdef fusion proteins. ng, values for E2 were unchanged at DMSO concentrations up to 20% v/v in the assay buffer. In contrast, DMSO increased the binding affinity of select PCB congeners for the fusion proteins. Figure 7A shows the effect of DMSO on the competitive binding of E2 and PCB 184 to GST-hERadef. At 4% DMSO, PCB 184 displayed characteristics of a weak binder; however, as the DMSO concentration increased to 20%, PCB 184 was a more effective competitor for binding to GST-hERadef. DISCUSSION Recombinantly expressed human, green anole, and rainbow trout ERs were used in this study to systematically investigate the potential differences in the ligand-binding preference of structurally diverse PCBs and HO-PCBs among ERs from different species. The human ERa (hERa). considered to be the prototypical ER, was selected as the basis for all compar- isons. due to the information available on its ligand-binding characteristics and the structure of the ligand-binding pocket (Brzozowski et al., 1997; Tanenbaum et al., 1998). The rain- bow trout (Onchorhynkus mykiss) ER (rtER) was included in the study. since it represents an environmentally relevant spe- cies, and the rtER has the most divergent amino acid sequence within its ligand-binding domains (LBD, domains D, E, and F) of any cloned ER with percent identity and similarity of 40 and 47%. respectively, when compared to the hERa LBD. The ER from the green anole (Anolis carolinensis), a lizard commonly found throughout the southeastern United States, was also included to further investigate claims of unexpected interac- tions between reptilian ERs and xenobiotics (Crain et al., 1998; Crews et al., 1995; Vonier et al., 1996). The green anole ER (aER) represents the first reported complete LBD sequence for a reptile, although partial sequences have been reported (Bergeron et al., 1998; Young et al., 1995). The LBD of aER is intermediate in amino acid sequence divergence relative to hERa and rtER LBD, with percent identity and similarity of 79 and 83%, respectively, compared to hERa. These divergent sequences were selected in order to assess the appropriateness of using a single surrogate species to prioritize, identify. and assess potential EEDs. Moreover, analysis of the data and comparison of ER sequence information may identify amino acid residues that might contribute to differences in ligand preference and relative binding affinities among species. 61 DIFFERENTIAL BINDING OF PCBS TO GST-ERdef FUSION PROTEINS 335 TABLE 5 Effects of Dimethyl Sulphoxide on the Ability of Select Polychlorinated Biphenyls to Competitively Displace [’Hll7BoEstradiol from Recombinant Fusion Proteins Containing the Human, Green Anole, and Rainbow Trout Estrogen Receptor D, E, and F Domains Compound DMSO % GST—hERardef GST-aERdef GST-rtERdef E2' 4 2.4 I 1.0 nM 2.5 I 1.3 nM 3.1 I 0.6 nM 10 2.8 I 0.8 nM 2.3 I 0.7 nM 3.1 I 0.6 nM 20 3.1 I 1.1 nM 27 I 0.7 nM 4.3 I 0.4 nM PCB 41 4 nb" rib >10 pM‘ 10 nb rib 6.5 I 0.7 nM' 20 wb' >10 uM 4.9 I 0.1 uM PCB 45 4 nb rib >10 11M 10 rib rib 5.6 I 1.4 “M 20 wb wb 2.5 I 0.7 1.4M PCB 47 4 rib rib >10 MM 10 rib nb 5.1 I 0.9 MM 20 wb wb 6.2 I 1.9 uM PCB 51 4 rib rib >10 11M 10 rib ii!) 2.4 I 0.6 1.1M 20 wb >10 “M 1.4 I 0.3 uM PCB 54 4 rib wb >10 MM 10 uh uh 1.9 I 1.2 ”M 20 wb wb 1.6 I 0.3 uM PCB 91 4 nb nb >10 uM IO rib rib 3.8 I 0.9 11M 20 wb >10 1.5M 2.0 I 0.3 uM PCB 104 4 >10 pM >10 itM 1.3 I 0.6 p.M 10 3.9 I 1.1 MM 4.1 I 1.0 pM 0.48 I 0.02 pM 20 1.1 I0.2 uM 1.1 I03 uM 0.4I0.1uM PCB 184 4 >10 MM >10 uM 0.4 I 0.1uM 10 4.0 I 2.2 uM 1.8 I 0.6 M 0.28 I 0.02 itM 20 0.5 I 0.2 p.M 0.7 I 0.1 M 0.14 I 0.02 M PCB 188 4 >10p.M >10uM 1.3 I 1.2M 10 4.9 I 0.9 M 2.9 I 1.1 uM 0.72 I 0.01 11M 20 1.9 I 0.9 M 1.4 I 0.7 “M 0.8 I 0.2 uM ' The 1C” was determined as described in the Materials and Methods and represents the average and standard deviation from at least 10 independent experiments. ’ Denotes non binder (rib) since no significant displace of radioloabeled HIS-estradiol was observed at the highest examined dose (10 uM), following replicate experiments. ‘ Uncharacteristic single site competition curve with greater than 50% displacement observed at the highest dose examined from three separate experiments, thus an 1C,o greater than 10 nM was assigned. ' Characteristic single site competition curve with greater than 80% displacement of radioloabeled 17B—estradiol observed at the highest dose examined (10 M). The lCm was determined as described in the Materials and Methods and represents the average and standard deviation from three independent experiments. ' Denotes weak binder (wb) since 10-50% displacement of radioloabeled l7Bestradiol was observed at the highest examined dose (10 M) in two separate experiments. The D, E, and F domains of the rtER were recloned and contained several differences when compared to the originally published sequence (Table 2). A single cytosine inserted at nucleotide position 726 of the originally published sequence (Pakdel er al., 1990) resulted in a shift in the reading frame. This change was later corrected by a second insertion of a pair of cytosines at nucleotide positions 756 and 757 (Pakdel et al., 1990). The insertion caused a change of 9 amino acids and resulted in the addition of a single amino acid residue within the insertion sites (Table 2). Moreover, single amino acid changes were found at various sites within the D, E, and F sequences (Table 2). It is doubtful that our sequence represents a polymorphism since the same changes were reported in two recently submitted rtER sequences in the GenBank database (GenBank accession AF242740 and AF242741). Potential species-specific sensitivities to PCBs were inves- tigated using the ER D, E, and F domains from mammalian (human), reptilian (green anole),and fish (rainbow trout) spe- cies expressed in bacteria as GST fusion proteins, using a semi-high throughput competitive binding assay. In vitro ER competitive binding assays have been well established and extensively used to investigate ER-ligand interactions. All competitive binding assays involve the displacement of a re- ceptor-bound probe molecule by a test compound. The probe is usually [’H]E2; however, fluorescently labeled high-affinity ER ligands have also been used (Bolger er al., 1998). Separa- 62 336 A . i 100- - ammo 1 ' Flu-10%“ 3 70‘ ' 524% V . gnu-mouse % ’°‘ cent is - is- r . . 1 r u - 1 41 40 -0 4 .1 a .0 B Luge-mumb- olewm) I» 0000 g ‘5 none . i e 4000 . 2000 0 Species II x In it x in n A II FIIIIMSO 4 4 4 II to lo 20 I II warmth-(V0) FIG. 7. (A) Competitive binding of E2 and PCB 184 at4%. 10%, and 20% DMSO in assay buffer. (B) The effect of increasing DMSO concentration on the binding of [‘HIEZ to the GST—ER DEF fusion protein. A M0 iLL aliquot of the GST-hERadef (H), GST-aERchA) or GST-rtERdef (RI) fusion proteins were dissolved in TEG buffer. pH 7.6 containing I ngml BSA as a carrier protein. and incubated with 2.5 nM [‘H1E2 and the indicated final concentra- tion of DMSO and unlabeled competitor as indicated. for 2 h at 4°C. The amount of bound radionucleotide was compared within each species and among different concentrations of DMSO. The results (mean and standard devration) are from a representative experiment that was repeated two times. ‘ Indicates DMSO treatment significantly different compared to other concen- trations of DMSO for each fusion protein (17 < 0 05) using a paired Student's r-Iest. tion of receptor-bound from free ligand can be done using dextran-coated charcoal (DCC) (Stoessel and Leclercq, I986), hydroxyapatite (HAP) (Laws er al., 1996) or by protein binding to a glass fiber membrane (Coleman er al., 1997). Shortcom- ings of the assay include the inability to distinguish between receptor agonists and antagonists and the possibility that high concentrations of competing ligand may lead to an increase in non-specific binding (Zacharewski, 1997). However, the assay is amenable to a high throughput format and can be used to investigate direct ligandzER interactions, which are the initial steps in many estrogenic responses. Heterologous expression systems have been used to purify and characterize several proteins, including steroid hormone receptors (Metzger er al., 1988; Wooge, 1992). Many fusion proteins exhibit activity comparable to that of their native forms (Jaglaguier er al., 1996; Wittliff er al., 1990). Expressing proteins as fusions facilitates the production of significant quantities of the desired regions or mutations of interest and their purification. In addition, it allows for precise control of assay conditions (protein concentration. metabolism. and back- MA'ITHEWS AND ZACHAREWSKI ground proteins) making direct comparisons among different species possible. The affinity of the bacterially expressed GST- ERdef fusion proteins for E2 was in agreement with the Kd values reported for full length ERs from human and other species (Nimrod and Benson, 1997; Pakdel er al., 1990; Vonier el al., 1997: Wooge, 1992). However. the affinity of the GST— rtERdef for E2 was approximately lO-fold higher than that reported for full length rainbow trout ER (Le Drean er al., 1995; Pakdel el al., 1990). This discrepancy may be attributed to differences in protein purity. assay conditions and the lack of accessory proteins or differences in post-translational modifi- cations. In addition, a wide range of K" values have been reported for some species, for example the K, determined from Xenopus liver cytosol ER has been reponed to vary from 0.5 to 15 nM (Lutz and Kloas, 1999; Westley. 1978). This suggests that differences in protein preparation and assay conditions may also contribute to the variability in the reported Kd values. It has been demonstrated that the degree of chlorination and the substitution pattern of PCB congeners can significantly influence their estrogenic properties (Korach er al., 1988'. Moore er al., 1997). X-ray crystallography studies have dem- onstrated that (who-substitution causes severe conformational restriction about the inter-ring bond, and conformationally restricted hydroxylated PCBs have been shown to be effective ligands for the ER (Korach er al., 1988). Quantitative structure activity relationships (QSARs) have also suggested that PCBS containing ortho- and para-chlorinated substituents are capable of binding to the ER (McKinney and Waller, I994; Waller er al., 1995). In this study, the only PCBS found to interact with GST-hERadef and GST-aERdef receptors were three tetra- orrho—substituted PCBS. a penta-chlorinated PCB (PCB 104). and two hepta—chlorinated PCB (PCB 184, and PCB 188) congeners. This is in agreement with reports showing that PCB 104 was able to compete with [’H]E2 for binding to mouse uterine ER and induce ER-mediated gene expression (Fielden er al., 1997) and that PCB 188 and 104 induce MCF—7 cell proliferation (Andersson er al., 1999). PCB 54, the fourth tetra-ortho substituted PCB examined, exhibited a weak inter- action with the GST-aERdef protein and did not bind to the GST-hERadef. which agrees with results reported by Arcaro el al. (1999) using a recombinant hERa preparation. Conversely, several PCB congeners, including PCB 104. 184, and 188. bound to the GST—rtERdcf fusion protein, with the degree of interaction increasing as the number of chlorinated substituents increased. Many of the congeners that displaced at least 50% [’HlEZ from the GST-rtERdef fusion protein also contained at least one para—chlorinated substituent in addition to the ortho substitutions. Of the environmentally relevant PCBS. only PCB 45, 47, 91, and 177 competitively displaced at least 50% [’HlEZ from the GST—rtERdef fusion protein, however none of these congeners bound to the GST-hERadef or GST-aERdef fusion proteins. The differences in PCB interaction between the GST- hERadef. GST-aERdef and GST-rtERdef fusion proteins may 63 Ian 9. Lulu ”w;- ..' ..%ill UH” E in.» Mt.— .Hh .cfl AU .Nh n), 3.: up... this LPG, #5.: l. . A .I .rV . II. I... V.,: pal til . I . l ..... DIFFERENTIAL BINDING OF PCBS TO GST-ERdef FUSION PROTEINS be due to amino acid-sequence differences among the recep- tors. particularly, the amino acids that form the binding pocket. Indeed, the promiscuity of the ER has been partially attributed to the size of the ligand-binding pocket, which is approxi- mately 2 times the volume of E2 (Brzozowski er al., 1997). Amino acid sequence alignments of the ERs from different species reveal that the region of the receptor involved in ligand binding is variable. For example, the LBD of the hERa shares 90% amino acid sequence identity with the mouse ERa, 82% with the chicken ER, 79% with the green anole ER, 70% with the xenopus ER, and only 40% with the rainbow trout ER. However, identification of critical amino acid residues or mo- tifs within the LBDs that contribute to observed differences in ligand preference and relative binding affinity through simple amino acid sequence alignment may be difficult. Despite dif- ferences in sequence identity of these ERs, the sequences from all species harbor the same 3 equivalent amino acid residues, Glu 353, Arg 394, and His 524, which participate in direct hydrogen bonds with 132 to stabilize the agonist in the binding pocket (Brzozowski et al., 1997). Rodents treated with Aroclors experience a variety of estro- genic responses, including increases in uterine glycogen con- tent and uterine wet weight (Ecobichon and MacKenzie, 1974). However, there have been few studies examining the ER binding affinities of these mixtures. Aroclor 1221 and 1254 have been shown to weakly bind the rat uterine ER (Nelson, 1974) while Aroclors 1221, 1248, and 1268 are capable of displacing [’H]E2 from the rainbow trout ER expressed in yeast (Petit er al., 1997). Aroclors 1221 and 1248 (10 and 100 nM. respectively) have also been reported to induce vitelloge- nin synthesis in rainbow trout hepatocytes (Petit et al., 1997). However, none of the Aroclors examined in this study were found to bind to any of the GST-ERdef fusion proteins. Com- plete congener analysis of 8 Aroclor mixtures (1221, 1232, 1242, 1016, 1248, 1254, 1260, and 1268), using capillary gas chromatography, demonstrated that PCBS 104, 184, and 188 are not detectable or are present at concentrations less than 0.05% wt (Schulz er al., 1989). In addition, none of the PCB congeners found to preferentially bind to GST-rtERdef were observed to exceed 2.6% wt (Schulz er al., 1989), resulting in a concentration that is unable to bind to the ER. The discrep- ancies between our results and those reported in the literature may be due to different assay conditions, measured endpoints. and differences in metabolic activity within the assays. It is well known that hydroxylation of select PCB congeners sig- nificantly increases their affinity for the ER (Fielden er al., 1997; Korach er al., 1988). thus suggesting that hydroxylation of PCB congeners plays an important role in the in vivo estrogenicity of Aroclor mixtures. Although, the major HO-PCBs identified in human serum (HO-PCB 1—7) examined in this assay have been shown to significantly inhibit ER-mediated gene expression in tran- siently transfected MCF-7 cells (Moore et al., 1997), none of the congeners were found to compete with [’HJEZ for binding 337 to the GST-ERdef fusion proteins. These results are similar to those reported by Kuiper er al. (1998) using baculovirus ex- pressed hERa and rat ERB preparations. HO-PCB 7 is a para-hydroxylated metabolite of PCB 187; however, the hy- droxylation of this congener did not increase its affinity for any of the GST-ERdef fusion proteins. This was in contrast to the para~hydroxylation of PCB 54 and 104, which significantly increased the affinity of the HO-PCBs for the ER of all 3 species. HO-PCB X, which binds both mouse and rat uterine ER, also bound to all 3 GST-ERdef fusion proteins. Unlike the fully ortho-chloro-substituted HO-PCB congener HO-PCB 54 (2,6,2',6'-tetrachloro-4-biphenylol), HO-PCB X (2',3',4’,5’- tetrachloro-4-biphenylol) contains a single ortho substitution, but was found to bind to all 3 GST-ERdef fusion proteins with a slightly lower affinity than HO-PCB 54. In addition. nonphe- nolic chloro-substituted HO-PCBs have been shown to effec- tively compete for binding to the ER (Kuiper er al., 1998), although HO-PCB 54 consists of both phenolic and nonphe- nolic chloro-substitutions and competes for binding to all 3 fusion proteins. This suggests that in addition to the degree of ortho substitution, the chlorination pattern and position of the hydroxyl are important determinants of ER binding as previ- ously described (Connor er al., 1997; Korach er al., 1988). Increasing DMSO concentrations to 10% was found to effect ligand preference and relative binding affinity of PCBS. In contrast, it had little effect on E2 interactions with GST-ERdef fusion proteins, suggesting that DMSO may increase the sol- ubility of PCB congeners and their availability for receptor interaction. For example, PCBS that bound weakly to the GST-ERdef fusion proteins with 4% DMSO in the assay mix- ture exhibited a significant increase in binding affinity in so- lutions containing up to 20% DMSO (Fig. 7A). However, at a final concentration of 20% DMSO, a significant decrease in total binding was observed, indicating direct effects on protein function. This observation has important implications for as- sessment of relative ligand-binding affinities for the ER, since organic solvent concentration may markedly influence the binding of some substances. These results demonstrate that ERs from different species exhibit differential ligand preferences and relative binding affinities for PCBS, which can be dramatically affected by solvent concentration. Although many of the environmentally relevant PCBS did not effectively compete with [’H]132 for binding to the GST-ERdef fusion proteins, the data generated from this study can be used for further development of ER QSARs (Waller er al., 1995) and also help in the derivation of species-specific QSARs (Tong er al., 1997). In summary, we report the cloning of the first complete reptilian ER DEF sequence, which has been used in a study comparing the differential binding of PCBS and HO-PCBs to the ERs from human, green anole, and rainbow trout using a semi-high throughput, competitive binding assay. Surprisingly, several examples of differences in the absolute and relative binding affinity of a number of structurally-related PCBS 64 338 among the GST-hERadef, GST-aERdef, and GST-rtERdef proteins were observed. The lack of differences between bind- ing affinities for the human and green anole proteins is most likely due to the higher degree of amino acid sequence identity throughout their ligand binding domains. The most notable differences were observed between the GST-rtERdef and either of the other two GST-ER fusion proteins. This may have implications for risk assessment when extrapolating data be- tween two such divergent species as humans and rainbow trout. Studies are currently underway that examine more structurally diverse substances, including pharmaceuticals, natural prod- ucts, environmental pollutants, and industrial chemicals, for potential differences in ER-binding affinity across species. ACKNOWLEDGMENTS We thank Stephen Safe for providing several of the PCB congeners. HO- PCB. and the commercial Aroclor mixtures. We also wish to thank Matts Tysklind and Patrik Andersson for their helpful comments and for providing some of the PCB: used in this study. and Juli Wade for providing the green anole liver tissue. The authors are also grateful to Stephen Safe. Trine Celius. Rob Halgren, Kirsten Fertuck, and Mark Fielden for their critical reading of this manuscript. Funding for this research was provided by the United States Environmental Protection Agency (R 82630I—01—0) and the Chemical Man- ufacturers Association. REFERENCES Andersson. P. I... Blom, A., Johannisson, A., Pesonen, M., Tysklind. M., Berg, A. H.. Olsson, P., and Norrgren. L. (1999). Assessment of PCBs and hydroxylated PCBs as potential xenoestrogens: In vitro studies based on MCF—7 cell proliferation and induction of vitellogenin in primary culture of rainbow trout hepatocytes. Arch. Environ. Contam. Toxicol. 37, I45~150. Arcaro. K. F.. Yi, L. Seegal, R. F.. Vakharia, D. D., Yang, Y., Spink. D. C.. Brosch, K., and Gierthy. .I. F. (1999) 2.2’.6,6’-Tetrachlorobiphenyl is estro- genic in vitro and in viva. J. Cell. Biochem. 72, 94-102. Bellschmiter, K., Buchert. S. H., Zell, M., Figge, K., Polzhofer. K., and Hoerschelrnan. H. (1981). Studies of the global baseline production. V: Monitoring the baseline pollution of the subantarctic by penguins as bioindi- caters. Anal. Chem. 309, 1-7. Bergeron, .l. M., Gahr, M., Horan, K., Wibbels, T., and Crews, D. (1998) Cloning and in situ hybridization analysis of estrogen receptor in the developing gonad of the red-cared slider turtle. a species with temperature- dependent sex determination. Dev. Growth Differ. 40, 243-254. Bergman, A., Klasson-Wehler. E. and Kuroki, H. (1994). Selective retention of hydroxylated PCB metabolites in blood. Environ. Health Perspect. 102, 464—469. Bitman. J., and Cecil. H. C. (I970). Estrogenic activity of DDT analogs and polychlorinated biphenyls. J. Agric. Food Chem. 18, 1108—1112. Bolger, R., Wiese, T. E.. Ervin, K., Nestich, 8., and Checovich, W. (1998). Rapid screening of environmental chemicals for estrogen receptor binding capacity. Environ. Health Perspect. 106, 55I-557. Bradford, M. M. (1976). A rapid and sensitive method for the quantitation of rnicrogram quantities of protein utilizing the principle of protein-dye bind- ing. Anal. Biochem. 72, 248 —254. Brzozowski, A. M., Pike. A. C.. Dauter, Z., Hubbard. R. E.. Bonn, T., Engstrom, 0.. Ohman, L, Greene. G. L.. Gustafsson, I. A., and Carlquist, M. (1997). Molecular basis of agonism and antagonism in the oestrogen receptor. Nature 389, 753—758. MATTHEWS AND ZACHAREWSKI Chomczynski, P., and Sacchi. N. (1987). Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction. Anal. Bio- chem. 162, 156-159. Colborn. T., vorn Saal, F. 8., and Soto, A. M. (1993). Developmental effects of endocrine—disrupting chemicals in wildlife and humans. Environ. Health Perspect. 101, 378 -384. Coleman, K. G., Wautlet, B. 8., Morrissey. D., Mulheron. J., Sedman. S. A., Brinkley, P., Price. 8., and Webster. K. R. (1997). Identification of CDK4 sequences involved in cyclin D1 and p16 binding. J. Biol. Chem. 272. 18869-18874. Connor, K., Ramamoorthy. K., Moore, M., Mustain, M., Chen, 1.. Safe. 8., Zacharewski, T., Gillesby, 8.. Joyeux, A., and Balaguer, P. (I997). Hy- droxylated polychlorinated biphenyls (PCBs) as estrogens and antiestro‘ gens: Structure-activity relationships. Toxicol. Appl. Pharmacol. 145, 111— 123. Crain. D. A., Noriega, N.. Vonier, P. M., Arnold. S. F., McLachlan, J. A., and Guillette, L. J., Jr. (1998). Cellular bioavailability of natural hormones and environmental contaminants as a function of serum and cytosolic binding factors. Toxicol. Ind. Health 14, 261-273. Crews. D., Bergeron, J. M., and McLachlan, J. A. (1995). The role of estrogen in turtle sex-determination and the effect of PCBs. Environ. Health Per- spect. 103, 73-77. Ecobichon, D. J., and MacKenzie. D. O. (1974). The uterotropic activity of commercial and isomerically-pure chlorobiphenyls in the rat. Res. Commun. Chem. Pathol. Pharmacol. 9, 85—95. EDSTAC, FR. (1998). Endocrine Disruptor Screening and Testing Advisory Committee Final Report: Vol. 1. Evans, R. (1988) The steroid and thyroid hormone receptor superfamily. Science 240, 889—895. Fielden. M. R., Chen. 1., Chittim, 3., Safe, S. H., and Zacharewski, T. R. (1997). Examination of the estrogenicity of 2.4.6.2’,6’-pentachlorobi- phenyl (PCB 104). its hydroxylated metabolite 2.4,6.2'.6'-pentachloro- 4-biphenylol (HO-PCB l04). and a further chlorinated derivative. 2.4.6.2’.4',6’-hexachlorobiphenyl. Environ. Health Perspect. 105, 1238—1248. Fitzpatrick, D. W.. Picken. C. A., Murphy. L. C.. and Buhr. M. M. (1989). Measurement of the relative binding affinity of zearalenone, a-zearalenol. and b-zearalenol for uterine and oviduct estrogen receptors in swine, rats. and chickens: An indicator of estrogenic potencies. Comp. Biochem. Physiol 94C, 69l—694. Gillesby, B., and Zacharewski, T. R. (1999). p82 (TFF2) mRNA levels in human breast tumor samples: Correlation with clinical and histological prognostic markers. Breast Cancer Res. Treat. 1466, l—l3. Hansen. L. G. (1998). Stepping backward to improve assessment of PCB congener toxicities. Environ. Health Perspect. 106(Suppl I). 171—189. Jaglaguier. S., Mesnier. D., Lager. I. L., and Auzou, G. ( I 996). Putative steroid binding domain of the human mineralocorticoid receptor. expressed in E. coli in the presence of heat shock proteins shows typical native receptor characteristics. J. Steroid Biochem. Mol. Biol. 57, 43—50. Katzenellenbogen, I. A. (1995) The structural pervasiveness of estrogenic activity. Environ Health Perspect. 103(Suppl 7). 99-101. Korach, K. 8.. Sarver, P., Chae. K., McLachlan, l. A., and McKinney. .I. D. (1988). Estrogen receptor-binding activity of polychlorinated hydroxybiphe- nyls: Conformationally restricted structural probes. Mal. Pharmacol. 33, 120-126. ~ Kuiper, G. 0.. Lemmen. I. G.. Carlsson, 3., Conan. J. C., Safe, S. H., van der Saag, P. T., van der Burg, B.. and Gustafsson, J. A. (1998). Interaction of estrogenic chemicals and phytoestrogens with estrogen receptor beta. En- docrinology 139, 4252—4263. Laemmli, U. K. (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227, 680-685. 65 DIFFERENTIAL BINDING OF PCBS TO GST-ERdef FUSION PROTEINS Laws, 8. C.. Carey. S. A.. Kelce, W. R., Cooper, R. I... and Gray, L. E. (1996). Vinclozolin does not alter progesterone receptor (PR) function in viva despite inhibition of PR binding by its metabolites in vitro. Toxicology 110, 1-1 1. be Drean. Y., Kern, L., Pakdel. F., and Valotaire, Y. (1995). Rainbow trout estrogen receptor presents an equal specificity but a differential sensitivity for estrogens than human estrogen receptor. Mal. Cell. Endocrinol. 109, 27-35. Li, M. H., Zhao, Y. D., and Hansen, L. G. (1994). Multiple dose toxicokinetic influence on the estrogenicity of 2.2',4.4'.5,5'-hexachlorobiphenyl. Bull. Environ. Contam. Toxicol. 53, 583-590. Lutz, I., and Kloas, W. (1999). Amphibians as a model to study endocrine disruptors: 1. Environmental pollution and estrogen receptor binding. Sci. Total Environ. 225, 49-57. McFarland, V. A., and Clarke. J.,U. (1989). Environmental occurrence. abun- dance, and potential toxicity of polychlorinated biphenyl congeners: Con- siderations for a congener-specific analysis. Environ. Health Perspect. 81, 225-239. McKinney. J. D.. and Waller, C. L. (1994). Polychlorinated biphenyls as hormonally active structural analogues. Environ. Health Perspect. 102, 290-297. McLachlan, l. A. (1993). Functional toxicology: A new approach to detect biologically active xenobiotics. Environ. Health Perspect. 101, 386-387. McLachlan, .I. A., and Korach, K. S. (1995). Symposium on estrogens in the environment: III. Environ. Health Perspect. 103(Suppl. 7), 3-4. Metzger. D.. White. J. H.. and Chambon, P. (1988'). The human oestrogen receptor functions in yeast. Nature 334, 31—36. Moore. M., Mustain, M., Daniel. K., Chen. 1.. Safe. 8., Zacharewski, T., Gillesby, B.. oneux. A., and Balaguer, P. (1997). Antiestrogenic activity of hydroxylated polychlorinated biphenyl congeners identified in human se- rum. Taxicol. Appl. Pharmacol. 142, 160-168. Mullin. M. D.. Pochini. C. M., McCrindle. 8., Romkes. M., Safe. S. H., and Safe, L. M. (1984). High-resolution PCB analysis: Synthesis and chromato- graphic properties of all 209 PCB congeners. Environ. Sci. Technol. 18, 468-476. Narasimhan. T. R., Kim. H. L.. and Safe. S. H. (1991). Effects of hydroxylated polychlorinated biphenyls on mouse liver mitochondrial oxidative phos- phorylation. J. Biochem. Toxicol. 6, 229-236. Nelson, J. A. (1974). Effects of dichluu "r“ (DDT) ana- logs and polychlorinated biphenyl (PCB) mixtures on l7-beta-[3H]estradiol binding to rat uterine receptor. Biochem. Pharmacol. 23, 447-451. Nimrod, A. C.. and Benson. W. H. (1997). Xenobiotic interaction with and alteration of channel catfish estrogen receptor. Toxicol. Appl. Pharmacol. 147, 381—390. Pakdel, F.. Le Gac, F.. be Goff, P., and Valotaire, Y. (I990). Full-length sequence and in vitro expression of rainbow trout estrogen receptor cDNA. Mal. Cell. Endocrinol. 71, 195-204. Petit. F., Le Goff, P., Cravedi, J., Valotaire, Y., and Pakdel. F. (1997). Two complementary bioassays for screening the estrogenic potency of xenobi- otics: recombinant yeast for trout estrogen receptor and trout hepatocyte cultures. J. Mol. Endocrinol. 19, 321-335. Safe. S. (1993). Toxicology. structure-function relationship, and human and environmental health impacts of polychlorinated biphenyls: Progress and problems. Environ. Health Perspect. 100, 259—268. Safe. 8., Washbum. K., Zacharewski, T., and Phillips, T. (1995). Synthesis and I. ' LI ,IuILIIIUI 339 characterization of hydroxylated polychlorinated biphenyls (PCBs) identi- fied in human serum. Chemosphere 31, 3017—3023. Scatchard. G. (1949). The attractions of proteins for small molecules. Ann. N Y Acad. Sci. 51, 660-672. Schulz. D. E.. Petrick. G.. and Duinker. .I. C. (I989). Complete characterization of polychlorinated biphenyl congeners in commercial Aroclor and Clophcn mixtures by multidimensional gas chromatography-electron capture detec- tion. Environ. Sci. Technol. 23, 852-859. Schuur. A. G.. Brouwer, A.. Bergman. A.. Coughtrie, M. W., and Visser. T. J (1998). Inhibition of thyroid hormone sulfation by hydroxylated metabolites of polychlorinated biphenyls. Chem. Biol. Interact. 109, 293-297. Stoessel. 8.. and Leclercq. G. (1986). Competitive binding assay for estrogen receptor in monolayer culture: Measure of receptor activation potency. J. Steroid Biochem. 25, 677—682. Tanenbaum, D. M., Wang, Y., Williams, S. P., and Sigler, P. B. (1998) Crystallographic comparison of the estrogen and progesterone receptor's ligand binding domains. Proc. Natl. Acad. Sci. U S A 95, 5998—6003. Tenbaum. 8.. and Baniahmad. A. (1997). Nuclear receptors: Structure. func- tion. and involvement in disease. Int. J. Biochem. Cell. Biol. 29, 1325-1341. Tong. W., Perkins, R.. Strelitz. R., Collantes. E. R., Keenan. 8.. Welsh, W. J., Branham, W. 8., and Sheehan. D. M. (1997). Quantitative structure-activity relationships (QSARs) for estrogen binding to the estrogen receptor: Pre— dictions across species. Environ. Health Perspect. 105, 1116—1124. Van den Berg. K. J., van Raaij. I. A., Bragt, P. C., and Notten, W. R. (1991). Interactions of halogenated industrial chemicals with transthyretin and ef- fects on thyroid hormone levels in viva. Arch. Toxicol. 65, 15-19. Vonier. P. M., Crain. D. A., McLachlan, I. A., Guillette. L. J., Jr., and Arnold. S. F. (1996). Interaction of environmental chemicals with the estrogen and progesterone receptors from the oviduct of the American alligator. Environ. Health Perspect. 104, 1318-1322. Vonier. P. M., Guillette, J., L. .l.. McLachlan, .I. A., and Arnold. S. F. (1997). Identification and characterization of estrogen and progesterone receptors from the oviduct of the American alligator (Alligator mississippiensis). Biochem. Biophys. Res. Comm. 232, 308 -312. Waller, C. L., Minor. D. L.. and McKinney. .l. D. (1995). Using three- dimensional quantitative structure-activity relationships to examine estrogen receptor binding affinities of polychlorinated hydroxybiphenyls. Environ. Health Perspect. 103, 702-707. Westley, B.. and Knowland, .I. (1978). An estrogen receptor from Xenopus laevis liver possibly connected with vitellogenin synthesis. Cell 15, 367— 374. Wittliff, I. L.. Wenz, L. L.. Dong, J., Nawaz. Z., and Butt. T. R. (1990). Expression and characterization of an active human estrogen receptor as a ubiquitin fusion protein from Escherichia coli. J. Biol. Chem. 265, 220l6- 22022. Wooge. C. H., Nilsson, G. M., Heierson, A., McDonnell, D. P., and Katzenel- lenbogen, B. 8. (1992). Structural requirements for high affinity ligand binding by estrogen receptors: a comparative analysis of truncated and full length estrogen receptors expressed in bacteria. yeast. and mammalian cells. Mol Endocrinol. 6, 861—869. Young. L. 1.. Godwin, J., Crammer. M., Gahr. M., and Crews. D. (1995). Reptilian sex steroid receptors: Amplification, sequence. and expression analysis. J. Steroid Biochem. Molec. Biol. 55, 261—269. Zacharewski, T. (1997). In vitro bioassays for assessing estrogenic substances. Environ. Sci. Tech. 31, 613—623. 66 CHAPTER 4 Differential estrogen receptor binding of estrogenic substances: a species comparison These data have been pubilished in the following article: Matthews, J .B., Celius, T, Halgren, R, and Zacharewski, T.R. (2000) Differential estrogen receptor binding of estrogenic substances: a species comparison. J Steroid Biochem Mol Biol. 74, 223-234. 67 twp-add Steroid Biochemistry & Molecular Biology PERGAMON Journal of Steroid Biochemistry & Molecular Biology 74 (2000) 223—234 www.clsevier.com/locate/jsbmb Differential estrogen receptor binding of estrogenic substances: a species comparison Jason Matthews, Trine Celius, Robert Halgren, Tim Zacharewski * Department of Biochemistry, Institute for Environmental Toxicology and National Food Safety and Toxicology Center, Michigan State University, 223 Biochemistry Building, Wilson Road, East Lansing, MI 48824-1319, USA Accepted 6 July 2000 Abstract The study investigated the ability of 34 natural and synthetic chemicals to compete with [’H]l7B-estradiol (E2) for binding to bacterially expressed glutathione-S-transferase (GST)-estrogen receptors (ER) fusion proteins from five different species. Fusion proteins consisted of the ER D, E and F domains of human alpha (GST-hERadef), mouse alpha (GST-mERadef), chicken (GST-cERdef), green anole (GST-aERdef) and rainbow trout ERs (GST-rtERdel). All five fusion proteins displayed high affinity for E2 with dissociation constants (K.,) ranging from 0.3 to 0.9 nM. Although, the fusion proteins exhibited similar binding preferences and binding affinities for many of the chemicals, several differences were observed. For example, a-zearalenol bound with greater affinity to GST-rtERdef than E2, which was in contrast to other GST-ERdef fusion proteins examined. Coumestrol, genistein and naringenin bound with higher affinity to the GST~aERdef, than to the other GST-ERdef fusion proteins. Many of the industrial chemicals examined preferentially bound to GST-rtERdef. Bisphenol A, 4-t-octylphenol and a,p' DDT bound with approximately a ten-fold greater affinity to GST-rtERdef than to other GST-ERdefs. Methoxychlor, p,p’-DDT, op'-DDE, p,p'-DDE, a-endosulfan and dieldrin weakly bound to the ERs from the human, mouse, chicken and green anole. In contrast, these compounds completely displaced [3H]E2 from GST-rtERdef. These results demonstrate that ERs from different species exhibit differential ligand preferences and relative binding affinities for estrogenic compounds and that these dilTerences may be due to the variability in the amino acid sequence within their respective ER ligand binding domains. © 2000 Elsevier Science Ltd. All rights reserved. Keywords: Estrogen receptor; Phytoestrogens; Competitive binding; Estrogenic 1, Introduction are modular in structure and consist of six distinct domains (A-F) [2]. The DNA-binding domain (C do- Estrogen influences the growth, development, behav- main) separates .the NHz-terrninal ligand-independent ior and regulation of reproductive tissues in all verte- activationdomam. (All; domains) and the 900134” ' brates. Many of the effects of estrogens are mediated mmal {3810‘}, Which Includes a hinge region (D do- through binding to the estrogen receptor (ER). Follow- main), the ligand binding domain (E domain) and a variable F domain. It has been suggested that exposure to natural and synthetic estrogenic chemicals may adversely affect wildlife and human health [3]. There have been contro- versial reports of decreases in sperm production and seminal volume in humans during the past half-century [4] and increases in reproductive abnormalities in mam- mals [5], reptiles [6], birds [7] and several fish species [8] ing estrogen binding, the ER undergoes a conforrna- tional change, which facilitates chromatin binding and the modulation of estrogen responsive gene expression. The ER exists as two subtypes, ERct and ERB, which are distinct genes that differ in their tissue distri- bution, and ligand preference [1]. Both receptors ‘Corresponding author. Tel: +l-517-3551607; fax: +1-517- followmg _exPosure to env'ronmefital contammant?‘ 3539334. However, it has also been argued that weak estrogenic E-mailaddress: tzacharc@pilot.msu.edu (T. Zacharewski). chemicals do not possess sufficient potency to elicit 0960-0760/‘00/‘3 - see front matter © 2000 Elsevier Science Ltd. All rights reserved. PII: 80960-0760(00)00126-6 68 224 J. Matthews et al. ,‘Journal of Steroid Biochemistry & Molecular Biology 74 (2000) 223—234 these adverse health effects when compared to the intake and potency of natural estrogenic chemicals, such as phytoestrogens [9]. The diets of humans and other species consist of several natural, non-steroidal estrogenic compounds ei- ther produced by plants (phytoestrogens) or fungal molds (mycotoxins) [10]. There are three chemically distinct classes of phytoestrogens: flavonoids (e.g. genis- tein and naringenin), coumestans (e.g. coumesterol) and lignans (e.g. enterdiol and enterolactone), which have been associated with reproductive abnormalities in grazing animals. However, the presence of high levels of phytoestrogens in Asian diets, combined with compara- tively lower rates of hormone-induced cancers prevalent in western populations, suggests that they may have a chemoprotective effect [1 1]. Of the mycotoxins, zear- alenone and its metabolites a- and B-zearalenol are the most commonly studied and have been shown to cause reproductive problems in swine and cattle fed contami- nated grain [12]. Results from several studies suggest that estrogenic compounds may exhibit differential binding preferences and relative binding alfinities for both ER subtypes [l3] and for ERs from different species [14]. 17B-estradiol (E2) exhibits a ten-fold lower affinity for the rainbow trout (Onchorhynkus mykt'ss) ER (rtER) than for the human ERa (hERet) [15]. Moreover, the pig ER ex- hibits a significantly greater affinity for a-zearalenol than does the ER from the Leghorn chicken [16]. Although, these differences may be due to the variabil- ity in the amino acid sequence within the ER ligand binding domain among species [14,17], many of these studies used different assay conditions and examined a limited set of test chemicals, making overall compari- sons difficult. Under controlled conditions potential differences in ER binding among species were further investigated using bacterially expressed glutathione-S-transferase (GSH—ERdef fusion proteins consisting of the D, E and F domains of human alpha, mouse alpha, chicken, green anole and rainbow trout ERs. The ability of several endogenous, synthetic and natural compounds to compete with E2 for binding to GST-ERdef fusion proteins was examined using a semi-high throughput competitive binding assay. The hERor, considered to be the prototypical ER, was selected as the basis for all comparisons due to the information available on its ligand binding characteristics and structure of its ligand binding pocket [18,19]. The mouse (Mus musculus) ERG: and chicken (Gallus gallus) ER were included as repre- sentative rodent and avian ERs, respectively. The ER from the green anole (Anolis carolinensis; aER), a lizard commonly found throughout the southeastern United States, was also included to investigate interactions with a representative reptilian ER. In addition the aER represents the only reported complete ligand binding domain sequence for a reptile [14], although partial sequences have been previously reported [20,21]. The rtER was also examined due to its environmental rele- vance and because it has a highly divergent amino acid sequence within its ligand binding domain, with percent identity and similarity of 60% and 67%, respectively, when compared to the hERa. 2. Materials and methods The steroids l7B-estradiol (l,3,5[10]-estratriene-3,l7 B-diol), l7et-ethynyl estradiol (l7a-ethynyl-1,3,5[lO]-es- tratiene-3, 1 7B-diol), estrone ( l ,3,5[10]-estratrien-3-ol- 17-one), estriol (l ,3,5[lO]-estratriene-3, 16a, 1 7B-triol), B-estradiol benzoate (l,3,5[lO]-estratriene-3,l7B-diol 3- benzoate), DHT (dihydrotestosterone, Set-androstan- l7B-ol-3-one) and DHEA (dehydroisoandrosterone, 5-androsten-3B-ol-l7-one) were purchased from Sigma (St. Louis, MO) Synthetic estrogens tamoxifen ([Z]-l-[podimethy- laminoethoxyphenyll-l ,2-diphenyl-1-butene), 4-hydrox- ytamoxifen ([Z]- l -[p-dimethylaminoethoxyphenyll- l ,2- diphenyl- l -buten-4-ol), diethylstilbestrol (4,4’-(1 ,2-di- ethyl-l,2-ethene-diyl)-bisphenol) were from Sigma. The antiestrogen ICI 164,384 (n,n -buty1-l l-(3, l 7B-dihydrox- estra - l,3,5(10)trien - 7a - yl) - N - methyl - undecan- amide) was a gift from Alan Wakeling of AstraZeneca (Alderley Park, UK). The mycotoxin zearalenone (2,4-dihydroxy-6-[lO-di- hydroxy-6-oxo-undecyl]benzoic acid u-lactone), its metabolites a-zearalenol (2,4-dihydroxy-6-[6a,lO-dihy- droxy-undecyllbenzoic acid u-lactone), and B-zearalenol (2,4-dihydroxy-6-[6B, 1 O-dihydroxy-undecyl]benzoic acid u-lactone), and the fiavonoids genistein (4’,5,7-trihy- droxyisoflavone), naringenin (4’,5,7-trihydroxyfia- vanone 7-rhamnoglucoside) and quercitin (3,3’,4’,5,7- pentahydroxyflavone), and B-sitosterol (22,23-dihydros- timasterol) were from Sigma. Coumestrol (2-(2,4-dihy- droxyphenyl)-6-hydroxy~3-benzofurancarboxylic acid lactone) was obtained from Arcos Organics (Pittsburgh, PA). The pesticide methoxychlor (1,1 , l -trichloro-2,2-bis(p- methoxphenylkthane) and its bis-hydroxylated metabo- lite HPTE (2,2-bis(p—hydroxyphenyl)-1,1,l,-trichloro- ethane) were provided by William Kelce (Monsanto, St. Louis, MO). o,p'-DDT (1,1,1-trichloro-2-[2-chloro- phenyl]-2-[4-chlorophenyl]ethane), p,p ’-DDT ( l ,1 ,1- trichloro-Z,2-bis[4-chlorophenyl]ethane), o,p’-DDE 1,1- dichloro-Z-[Z-chlorophenyl]-2-[4-chlorophenyl]ethylene) and p,p’-DDE (1,1- dichloro-bis[4-chlorophenyl] ethylene) were purchased from AccuStandard (New Haven, CT). The alkyl phenolic compound 4-t- octylphenol and bisphenol A (4,4'-isopropylidenediphe- nol) were obtained from Aldrich (Milwaukee, WI). Atrazine, simazine and chlordecone (kepone) (de- cachloro - octahydro - 1,3,4 - metheno - 2H - cyclobuta 69 .1. Matthews at al. / Journal of Steroid Biochemistry & Molecular Biology 74 (2000) 223—234 225 (cd)pentalene) were from Chem-Service (West Chester, PA). Dieldrin (1,4:5,8-dimethanonaphthalene), ct-endo- sulfan (hexachlorohexahydromethano-Z,4,3-benzo-di- oxathiepin-3-oxide), B-endosulfan (hexachlorohexa- hydromethano-2,4,3-benzo-dioxathiepin-3-oxide) were provided by S. Safe (Texas A&M University, College Station, TX). Monsanto (St. Louis, MO) and Eastman Chemical Company (Kingsport, TN) supplied butyl- benzylphthalate and dibutylbenzylphthlate, res- pectively. Radiolabeled [2,4,6,7,16,17-3H] l7B-estradiol ([3H]E2; 123 Ci/mmol) was purchased from New England Nu- clear (Boston, MA). SuperScript II reverse transcriptase and Trizol Reagent were purchased from Life Tech- nologies (Gaithesburg, MD). Vent DNA polymerase was purchased from New England Biolabs (Beverly, MA), and restriction enzymes and Taq DNA poly- merase were obtained from Roche/Boehringer Mannheim (Indianapolis, IN). All other chemicals and biochemicals were of the highest quality available from commercial sources 2.]. Recloning of chicken ER DEF domains Total RNA from a l-cm3 liver section from a G. gallus (chicken) was isolated using Trizol Reagent ac- cording to manufacturer’s instructions. One microgram of RNA was then reverse transcribed using primer PRcr-5’-aaaactcgagttatattgtattctgcatactctcctc-3’ as pre- viously described [14]. The entire cDNA products from the reverse transcription (RT) reaction were used in a PCR mixture containing 200 uM dNTPs, 2 uM primers (PRcf - 5' - aaaagaattccgaaatgatgaaacagaaacgtcaaag - 3’ and PRcr) and 1.25 units of Vent DNA polymerase was amplified for 30 cycles using the following conditions: 94°C for 1 min, 62"C for 1 min and 72°C for 2 min. Sequence analysis was performed using MacVector 6.5 and the GCG Wisconsin Package (Oxford Molecular Ltd., Beaverton OR). 2.2. Construction of GS T-ER DEF fusion proteins The construction of pGEX-hERadef, pGEX-aERdef and pGEX-rtERdef vectors has already been described [14]. The pGEX-mERadef plasmid (mERa a.a. 268— 599) was generated by PCR amplification of the plas- mid pJ3MOR containing the complete mouse ER cDNA (provided by M.G. Parker; Molecular En- docrinology Research Laboratory, London, UK) using primers PRmf-5’-aaaaggatccatgttgaagcacaagcgtcaga- gag-3’ and PRmr-5’-aaaagaattcgcgccgctcagatcgtgttgggg- aagccctc-3’. The pGEX-cERdef plasmid (cER a.a. 258—589) was prepared using the products of the RT- PCR reaction described above. The mERadef and cERdef PCR fragments were digested with the BamHI/ Natl and EcoRl/Xhol restriction enzymes and ligated into the appropriately digested GST fusion protein expression vector, pGEX6p3 (Amersham/Pharmacia; Piscataway, NJ). The PCR amplification was per- formed using Vent DNA polymerase (New England Biolabs) as described above. The sequence of each construct was confirmed with restriction enzyme digest and ABI/Prism automated sequencing (Perkin Elmer Applied Biosystems; Foster City, CA). 2.3. Expression and purification of GS T-ERdef fusion proteins Expression and purification of GST-ERdef fusion proteins was done as previously described [14]. Partially purified fusion proteins were separated by SDS—PAGE using a 4% stacking and 10% separating gel. Proteins were visualized by coomassie brilliant blue R250 staining. 2.4. Receptor binding assays Receptor binding assays were performed as previ- ously described [14]. Briefly, GST-ERdef fusion proteins were diluted in TEGD buffer (10 mM Tris pH 7.6, 1.5 mM EDTA, 1 mM DTT and 10% (v/v) glyc- erol) containing 1 mg/ml bovine serum albumin (BSA), and incubated at 4°C for 2 h with 0.1—3.5 nM [3H]E2 in 1 ml glass tubes arranged in a 96-well format (Marsh Scientific, Rochester, NY). Fusion protein preparations were diluted to ensure 10 000 dpms of total binding (varied from 750—2000-fold). Binding assays were ini- tiated by adding 240 pl of protein preparation to glass tubes containing 5 pl of DMSO and 5 ul [3H]E2, thus the concentration of solvent did not exceed 4%. Bound [3H]E2 was separated from free using a 96-well filter plate and vacuum pump harvester (Packard Instru- ments). After drying, the filter plates were sealed and 50 ul of MicroScint 20 scintillation cocktail (Packard ln- struments) was added to each well. Bound [3H]E2 was measured using a TopCount luminescense and scintilla- tion counter (Packard Instruments). Competitive ligand binding assays were performed as described above except diluted GST-ERdef fusion protein preparations were incubated with a final con- centration of 2.5 nM [3H]E2 (5 ul aliquot) and increas- ing final concentrations of unlabeled competitor (0.1 nM-IOO uM, 5 ul aliquots) at 4°C for 24 h. Each treatment was performed in quadruplicate and results are expressed as percent specific binding of [3H]E2 versus log of competitor concentration. ICso values were determined from non-linear regression for single site competitive binding analysis. The reported 1Cso values represent the concentration of test compound required to displace 50% [3H]E2 from the GST-ERdef fusion proteins as compared to the 50% displacement of [3H]E2 achieved by unlabeled E2. Analyses were per- 70 226 J. Matthews et al. / Journal of Steroid Biochemistry & Molecular Biology 74 (2000) 223—234 formed using GraphPad Prism software (GraphPad Software Inc., San Diego, CA). 3. Results 3.]. Recloning of chicken ERdef sequence The recloned chicken (G. gallus) ERdef sequence was identical to the previously published sequence [22]. A comparison of the amino acid sequences of the cERdef with those of human, mouse, green anole and rainbow trout is shown in Fig. 1. Amino acids found to interact with E2 and/or line the ligand binding pocket are boxed. Though these residues are highly conserved among the species, there are differences. For example, Met421 in hERadef is substituted for Phel75 in aERdef. In addition, hERadef differs in two locations from rtERdef. in hERadef Leu349 and Met528 replace Met93 and Ile272, in rtERdef. 3.2. Expression and saturation analysis of GS T-ERdef fusion proteins The amino acids used in the construction of the GST-ERdef fusion proteins for each species and their hsaadet 264 mERadef 268 cERdef 257 cERdef 19‘ -- ........ BENDS .N-A.ALTI:A.STA. . . C.DVDOBTOOOO ----- ’0...’... sequence identity compared to that of the human ERa are shown in Fig. 2. All five fusion proteins migrated according to their predicted molecular weights (MWs): GST-hERadef (64.2 kDa), GST-mERadef (64.4 kDa), GST-cERdef (65.2 kDa), GST-aERdef protein (64.3 kDa) and GST-rtERdef (65.5 kDa), although each appears to migrate as a doublet (Fig. 2(A)). The higher MW bands most likely represent the full-length product, whereas the lower bands may result from proteolytic cleavage [23]. In addition, higher and lower MW proteins co-purified with the proteins. The purity of the GST-ERdef fusion proteins varied among protein preparations, with yields ranging from 1 to 4 mg/l. This was evident with the GST-mERadef prepa- ration (Fig. 2(B) lane m) which contained lower amounts of the fusion protein, when compared to the other fusion protein preparations. However the GST- mERadef preparations resulted in sufficient recombi- nant receptor to investigate the competitive binding of approximately 500 compounds per liter culture. This value varied among protein preparations, ranging from 100 to 600 compounds per liter culture. Binding affinities of the partially purified GST-ERdef fusion proteins for E2 were determined by saturation analysis and linear transformation of the data [24] (Table 1). Differences in the amount of receptor re- - -MLKHKRQRDDGEGRGEVGSAGDKWIMPSPLMI KRSKHI ----- ”monumental! 3 3 1 -- .......... L...N.fl.h3 .......... -B.H.Q....EEQDS.NGEA.STEL..PT..T...VV.HN.... ----- 2......I.......l.....V.... 325 ...... H.....‘----P..............I....VV.... 85 COCO-OOOIOOnCIII 335 rtERdet 1. RVLR.D..YCGPAGD.EKPYGDLBH.T.PPQDGGRNSSS.LNGGGGIIG'IIIIDII.VLIL.96....A»C.IO 75 332 406 336 410 326 400 86 160 76 150 407 481 £11 485 401 475 161 235 151 225 482 $55 486 560 476 550 236 310 226 300 557 GASVEETDQSRLATAGSTSSHSLQK----YYITGEAEG-PPATV --------- 595 561 .VPP..PS.TQ...T3...3....T----...PP....-..N.I --------- 599 551 A.PM..ENRNQ.T..-PA......S----F..NSKE.ESMQN.I --------- 589 311 SPPS.DDPLNQ..VP-.PSH...LP----C.VNKQE..NEQEAI --------- 349 301 Q.GEQTEGP.TTT.TSTG..IGPHRGSQDTH.RSPGS.VLQYGSPSSDOHPIP 353 Fig. 1. Alignment of the estrogen receptor D, E and F domains from human (hERadel). mouse (mERadel), chicken (cERdef), green anole (aERdef) and rainbow trout (rtERdef). Numbers refer to amino acid position in the full-length sequence. Identical amino acid residues are represented as dots while missing residues are shown as dashes. The E domains are shown in bold. Residues that line the hormone binding pocket and/or interact with bound E2 are boxed. ‘Only a portion of the green anole ER sequence was cloned and the start of the D domain corresponds to amino acid residue 19 in the cloned sequence [14]. ”Refers to the first amino acid of a recloned rainbow trout ER partial sequence [14]. This figure was modified from Pike et al. (1999) [42]. 71 J. Matthews et al. / Journal of Steroid Biochemistry (1 Molecular Biology 74 (2000) 223—234 227 WWW“ A walls-IQ when... assume" hmcs“ '90k-— Fig. 2. (A) A schematic representation the GST-ERdef fusion proteins expressed in bacteria. Amino acid residues that link the two proteins are provided above the transition point. The residues shown in reduced font and italics represent amino acids from the linker region downstream of GST. The first three residues of the ER D domains are shown in regular text. Numbers provided above identify the amino acids used in the construction. The numbers within the domains represent the percent identity while those in parentheses resent rcent similarity to Mike, 'Only a portion of the green anole and rainbow trout ER sequences were cloned as previously described [14]. (B) SDS—PAGE analysis of the GST-ERdef fusion proteins purified using GSH affinity chromatography. Lane h: GST-hERadef (predicted molecular weight (MW) = 64.2 kDa). Lane in: GST-mERadef (predicted MW=64.4 kDa). Lane c: GST-cERdef (predicted MW=65.2 kDa). Lane a: GST-aERdef (predicted MW = 64.3 kDa). Lane rt: GST-nERdef (predicted MW = 65.5 kDa). Each lane was loaded with 5 ug of partially purified GST-ERdef fusion protein. Proteins were analyzed using a 4% stacking and a 10% separating gel stained with coomassie brilliant blue R250. quired to attain the desired 10 000 dpm at saturation Table 2 and Fig. 3 summarize the ability of several were species-dependent and may be due, in part, to natural and synthetic compounds to compete with differences in protein purity, functionality and level of [3H]EZ for binding to GST-ERdef fusion proteins. eXpression between preparations All GS T-ERdef fu- Non-binders and weak-binders were tested in a single 510“ proteins exhibited hlgh binding affinity for E2, experiment in which each treatment was performed in with dissociation constants (Kd) similar to other reports quadruplicate. ICso values were determined for com- (see Table 1 and references therein). pounds that displaced at greater than 50% of the [3H]E2 from the GST-ERdef fusion proteins and are shown as averages from at least two experiments. The 3.3. R l t' b' d’ '1' nd i 'd , ea we m ing affinl 1e: ofe ogenous S eroi s relative binding affinities were determined for each and antiestrogens Table l The classification of the competitive binding ability Companson or me dissociation commits (K4) or me 0512“,” of the test compounds followed the same criteria as fusion prom“, Wm, reported value previously discussed [14]. Briefly, compounds were classified as non-binders (nb) if less than 10% displace- Protein Ki (nM) Kr reponed (nM)‘ References ment was observed or as weak binders (wb) if only . . GST-hERadef 0.4 i 0.1 01—15 [27,451 10~.50% of [3H]E2 was displaced at the highest concen- GST-mERordef 0.6 10.2 0.1—1.4 [46.47] tratlon (100 uM) of competitor exarnrned. For com- GsTigkdef 09 i OJ 0.2-31 [16.48] pounds that were capable of displacing greater than GST-aERdet 0.7 1 0.2 0.5—l7 [33,35] 50% of the [3H]E2 from the GST-ERdef fusion proteins GST-“ERdef 0-6 i 0-1 09-6 [29.301 an IC” vahle was calcmated usmg Graphpad Pnsm.3'0' ' IQl values were derived from a variety of different ER sources Concentrations great" th_a_n '09 “M were "0" examined including in vitro translated proteins, recombinant proteins expressed due to potential solubility limitations of the test in [mum yea“ and SF9 “Us, and cytosol prepared from mm compounds. testis and liver tissue. 72 A21 oo: 83. 35833 30:»... 0.... .a 92,830 mm? 3.253.?— 33.3062 .0 595829.... Eda—him: on no...» 35 3.2.5-5: 38.83 H. .2: 8: 98v v2.3.8... 7.2.»... o... .a 328.... .33 36.2.8.9. vo_ons_o_v8 .8 32:829.... $3.2 8.2m 33. 3.2.5 233 3.9.09 9 .8. x .x 3:888 30:28.3: 22.1%.... 22.... 8.2.... 8.3.3. .. 8:08.398 E3599: 0...: .30. :w :8... 5.3.3.6 23:2». can owaeozw o... 2:383. c5. N 528% a. coatomov 3 Use .a cognac... $3 a 9.30:8 3:25.098 wives 3.52.58 80.... “coinage“. 893 8:3.» RU~ . I a: I n: I n: I a: I a: _Ouo~n0~_m.n M i e. i 2. - e. i e. - 2. Sega H i a. i 2. i e. - e. - e. 2.3.2 3 I 93 i a: i n: n c: I A... cab—aaoccma )_ .3 .7 2 x e¢ H on - es . a... i e. i a... 8.8.350: m ~.¢¢ T2 x v. H .2 . es - h.._ - as i a... 52385-... M ¢~¢ .. 2 x 2 H S i a... - e3 i 2.. . ea u8.2.2....En....b.....a 7 i 93 a n3 : a? i Q... 1 n3 o.m.u;EQ_>N.—on_b=m m .72 xe¢ H 3 i 2.. . .3 i a... - 2... 59...... w 98¢ .. 2 x ¢¢ H 3. i “.3 i H... - a... i .3 39...... .w :¢ .72 XS H N... - as i as - .3 i a... moon? 1w $¢ 72 x .¢ H w... - a... 28¢ .72 x N. H 2 28¢ n.2 x 2 H e. - a... 59...... m 98¢ .72 x¢.~H¢¢ 23 T2 v.N¢H¢.. 28¢ T2 x2 H N... i e3 i 2., 5.225 m w~¢¢ .2 x ~¢ H a. - es . .3 - as i .3 <83 & 3.2. .2 x n. H a... 83 .72 x ..¢ H 3 88¢ .2 x v¢ H 3 - e... i as 25.....32 w, .8... .72 Xe¢H~¢ ..¢¢ T21¢HZ ..¢¢ ni2 x .¢H¢¢ 28¢ t2 xm¢He¢ 88¢ 72 x... Hm... 2.2.8. m .3 #2 x 2 H e. :¢ .72 x... H .3 38¢ .72 xi H 3 88¢ .2 x 2. H 2.. 88¢ .2 x... H em < .822... m x¢¢ . -2 x 3 H ¢.. .2 72 x N. H N... 88¢ T2 x e¢ H mm 28¢ .72 x I H e... $¢¢ .72 x ¢¢ H 3 2228:32825 hm NH 72 x ~¢ H... 83 f2 3.. H 3 $¢ 72 x .¢H 3 :¢ T2 x .¢ H e. ~.¢ .72 x; H E 358.55.... a $¢ 72 x ..¢ H 2 n. 72 x .¢ H 3 HS 72 X: H 3 HQ 72 xv¢ H 3. 2.¢ T2 xH¢ H M... 22250 m e~¢ .72 x .¢ H 3 2 72 x e¢ H S 2¢ 72 3.. H 3. a... 72 x2 H ¢... ...¢ 72 x¢¢He¢ 328.58 M. 2 72 x .¢Hv.~ .3. 721.33. m... 72 x~.~H..¢ N. 72x~¢H- N. T2xw¢Hn~ E: N a. .2 x .8 H 3. N. 72 x 3H 2 2 .72 x I H 3 N. 72 xm¢H 2 m... 72 x...¢H .H 2583.3 m ¢¢ .72 x 3H 2. m. 72 x S H X 2 72 x .¢ H S N. .72 x ~¢H 2 2 T2 xm¢H em 2822. .2256 m a 72 x 2. H a. 2 72 x H¢H ¢¢ 2 72 x S H ..N 2 72 x .¢ H e... : T2 xe¢H 3 523:5 w a .-2 x2H S N... 72 x... H 2 a 72 x .¢H.... : T2 xe. H 2 m. 72 xm¢H 2 65.23.. H 2 72 xe¢ H S. 8 72 x _¢H¢.. .. 72 x .¢H 3 2 .2 x3 H ..~ 8 72 xm¢H¢.. .25.". a. a. .12 x S H ¢.. 2 .72 x ~¢ H... 8 .72 x... H S 3 .72 x2 H on He .12 xm¢H ¢¢ 3.3.2 .0. a z 72 x ~¢ H 2 8 .72 x .¢ H .H 8 72 x S H 3 8 .72 x .8 H 2. 9 T2 xnoH 3 2.23 m SN .72 x .¢Hm.. 8 $22.13. 2 .12 x¢.¢H¢.e a .72 xH¢H .H .3 T2 x~¢H .e 38.288... w. .2 .72 x .¢H¢.~ .2 s2 x¢.. H3 8. .12 xeoHnm a. T2 xn¢H~¢ a T2 x.¢H~.m 2.3.2.255 m 8. 72 x 3 H 2 8. 12.. 2 H .H 8. .12 x 3 H 2 8. 12.38 H 2 8. 22 x3 H 2 36834.: . 8. .72 x¢.¢ H .H a. .72 x .¢H 2 E .12 x .¢ H o. 2. .12 x ..¢H 2 R. .72 x .¢Hm.~ 2.5.8 .22... J E .12 x¢¢ H N. 3 .72 x .¢ H 2 8. .12 xm¢ H m... N: .72 x e¢ H N. 2. .72 x .¢ H H... 523825.82... <5. 9.2 <5. .2. so. <5. .2. so. <3. .2. so. .25. .2. .32 83.8-50 822-50 835950 882550 .8528.-me 8:858 .3 228 00¢ .a c0383... .73 m worse—.8 3.205 :23. dovmmFmO 5.. 2.8.82.0 0.58:8 2.05:? 9:. .832. .0 .355 music? wiper. 95.2». new 83g SU— N 95:. 73 J. Mflfl’lt’H‘X e! ul Journal 0/ Slcnml Brmlmnmn & Molecular Bio/ngy 74 (.7000) .733 234 129 v - u v v T v u v -I 1 4° d v. -7 4 -. O -1 4 4 4 17B-Estradiol Coumestrol .rrri 9!§¥j iii? ‘1 v a :1 4 ~s 4—r-octylphenol l? R a 1 v u v 4 -1 4 -I ICI l64,384 Genistein Bisphenol A e in b b b 4: b d- L ‘1 L4 u n a % [3H]17B-Estradiol Bound v v v v u u 4 -1 O -l Dihydrotestosterone Naringenin o.p'-DDT ‘a 9!!§§ 919a? erraé L 4a b in 54 5. 6 a. b d- ‘1 ersL %: «irri‘ sari" ‘2 -H -10 4 J -7 -I :0 -. :7 -I u-Zearalenol HPI‘E Dibutylbenzylphthalate Log Competitor (M) Fig. 3. Representative competitive binding curves of selected test chemicals to O GST-hERadef, C) GST-mERadef, I GST~cERdef, U GST-aERdef. and O GST-rtERdef fusion proteins An aliquot of partially purified GST-ERdef fusion proteins was incubated with 2.5 nM [‘H]l7[}-estradiol and increasing concentrations of unlabeled test chemical and incubated for 24 h at 4°C as described in Section 2. Hie results are from a representative experiment that was repeated at least two times. Standard deviations for points on graph ranged between S and ISM. of the mean ‘1 b b 5 L‘ compound as compared to the [C50 value of E2 for depending on the competitive binding ability of the test each GST-ERdef fusion protein. The compounds have compound. been arranged in order of potency in comparison to Overall the steroidal and antiestrogenic compounds GST-hERotdef. Chemicals were tested at concentrations exhibited similar binding preferences and relative bind- ranging from 0.1 nM to 100 pM. This range varied ing affinities for GST-ERdef fusion proteins. The 52 74 230 J. Matthews et al. / Journal of Steroid Biochemistry & Molecular Biology 74 (2000) 223—234 binding was similar across species with 1C“, values ranging from 2.7 to 3.3 nM (Fig. 3). 4-Hydroxytamox- ifen, the hydroxylated metabolite of tamoxifen, bound with greatest affinity to all 5 of the fusion proteins, with a ten- to 25-fold greater affinity than the parent com- pound. Diethylstilbestrol and ethynyl estradiol bound with similar affinities, 1.9—3.1 nM and 2.0—3.2 nM, respectively, to the five fusion proteins. Though the rank order of these chemicals was similar among species, some notable differences were observed. The pure ER antagonist ICI 164,384 bound to GST- hERadef, ~mERozdef, -cERdef and -aERdef proteins with a five- to ten-fold lower affinity than to GST- rtERdef (Fig. 3). In addition, ICI 164,384 bound with higher affinity than E2 to GST-rtERdef, which was not observed with the other four fusion proteins. DHT bound to GST-aERdef with a seven- to lZ-fold greater affinity than to GST-hERadef, ~mERadef and -rtER- def, but bound with a 42-fold greater affinity than to GST-cERdef (Fig. 3), which shares 91% amino acid sequence identity to aERdef within their respective ER B, ligand binding, domains. DHEA, a precursor in the endogenous synthesis of estrogens and androgens, weakly bound to the ERs of human, mouse, chicken and green anole; however it exhibited an ICso value of 12 i 2 pM with GST-rtERdef. 3.4. Differential binding of phytoestrogens and mycotoxin: Differences in ligand preferences and binding affini- ties were also seen with some mycotoxins and phy- toestrogens. a-Zearalenol, a hydroxylated metabolite of . .051 hERudet 0.000 mERadef 0.049 (:8ka _ 0.073 0.01 - aERdef 0.112 I I 0.417 '_ ' tEltdet Fig. 4. Dendogram generated from the aligned ERdef sequences. The distance between nodes reflects the degree of sequence identity when doing pairwise alignments. The value of 0.] corresponds to a differ- ence of l0°/o between two sequences. This figure was generated using the ClustalW alignment function in MacVector 6.5 (Oxford Molecu- lar Ltd). the mycotoxin zearalenone, consistently bound with greater affinity to the GST-rtERdef than to any other fusion proteins (Fig. 3). Similarly, the parent compound zearalenone and another metabolite, B-zearalenol, bound with greater affinity to the GST-rtERdef than to ERs from other species (Table 2). Interestingly, a-zear- alenol bound with greater affinity than E2 to GST- rtERdef (2.6-fold) but bound to the other 4 fusion proteins with approximately half the affinity of £2. Overall the phytoestrogens displayed higher affinity for GST—aERdef than the other GST-ERdef fusion proteins. However, quercitin bound with slightly greater affinity to GST-rtERdef and B-sitosterol was unable to displace [3H]EZ from any of the GST-ERdef fusion proteins at the highest concentration examined (100 pM; Table 2). Coumestrol bound with greatest affinity to GST-aERdef (1C,0 = 0.10 i 0.04 pM). It ex- hibited similar affinity to the human, mouse, chicken and ERdef proteins (IC50 values ranging from 0.36 to 0.80 uM) and bound with a l4-fold lower affinity to the rainbow trout ER (IC,0 = 1.4 i 0.1 uM; Fig. 4). Genis- tein exhibited similar binding affinities for all five GST- ERdef fusion proteins, but consistently bound with higher affinity to GST-aERdef (Fig. 4). Moreover, naringenin bound with highest affinity to GST-aERdef (IC,0=4.7i-0.8 uM) and slightly lower affinity to GST-rtERdef (8.7 $1.3 uM). However it bound with approximately a ten-fold lower affinity to GST-cERdef and bound weakly to GST-hERadef and GST- mERadef (Fig. 4). 3.5. Differential binding of synthetic chemicals Overall, this class of compounds bound with greater affinity to GST-rtERdef than to any other fusion protein. Bisphenol A, 4—t-octylphenol and o,p'-DDT bound with approximately a lO—fold greater affinity to GST-rtERdef than to the ERs of the other species (Fig. 4). Complete displacement of [3H]E2 by these com- pounds was only observed with GST-rtERdef at the highest concentration examined (100 uM). Methoxy- chlor, p,p’-DDT, o,p'-DDE, p,p'-DDE, at-endosulfan and dieldrin were found to bind weakly to the ERs from the human, mouse, chicken and green anole. In contrast, these compounds completely displaced [3H]EZ from GST-rtERdef (Table 2). Although, ci-endosulfan effectively displaced [3H]E2 from GST-rtERdef, its iso- mer B-endosulfan did not displace greater than 30% [3H]E2 and was therefore classified as a weak binder. Butylbenzylphthalate and dibutylbenzylphthalate bound weakly to the ERs of the different species. However, dibutylbenzylphthalate was found to displace of 75% [3H]EZ from GST-rtERdef (IC,0=I.7i2.3 uM) at the highest concentration (1 uM) examined (Fig. 3). A visible precipitate was observed at concen- trations greater than 10 uM and 1 nM, for butylben- 75 Ii: 1‘ .I. Matthews et al. / Journal of Steroid Biochemistry & Molecular Biology 74 (2000) 223-234 231 zylphthalate and dibutylbenzylphthalate, respectively, thus higher concentrations were not examined. Finally, two chloro-S-triazines (Atrazine and Simazine) were unable to displace [3H]E2 from any of the GST-ERdef fusion proteins at the highest concentration examined (100 uM). The time to equilibration for steroid recep- torzcompetitor complexes has been reported to be greater than 16 h at lower temperatures [25]. Since the competitive ligand-binding assay used a 24-h incuba- tion time it is unlikely that the observed differences in binding affinities were due to a lack of equilibration. However when this data set was examined following a 2-h incubation at 4°C, the binding profiles of most chemicals and differences in ligand preferences, and binding affinities were similar to that of the 24 h incubation time, though there were some exceptions (unpublished data). 4. Discmsion Previous studies using a variety of protein prepara- tions and assay conditions suggest that there may be differences in the absolute and relative binding affinities of structurally diverse estrogenic chemicals to ERs from different species. In order to investigate species-specific ligand preferences, and differences in relative and abso- lute binding affinities, GST-ERdef fusion proteins con~ sisting of mammalian (human and mouse), avian (chicken), reptilian (green anole) and fish (rainbow trout) ERs were constructed. It has been reported that truncated forms of the glucocorticoid receptor can effect protein stability and receptor function [26]. However, truncated forms of nuclear receptors overexpressed in heterologous expres- sion systems have been previously shown to exhibit comparable affinities and ligand preferences relative to their native forms [27,28]. Heterologous expression of GST-ERdef facilitates purification of these fusion proteins and allows for precise control of the competitive binding assay conditions (e.g. protein con- centration, metabolism, non-specific binding, back- ground/accessory proteins), making direct comparisons possible. The affinity of the bacterially expressed GST-ERdef fusion proteins for E2 was similar to the Kd values reported for full length ERs (Table 2). However, the affinity of the GST-rtERdef for £2 was approximately ten-fold higher than that reported for full length rain- bow trout ER [29], but was in agreement with reports using [3H]moxestrol [30]. In general, the reported Kd values for some species vary considerably, which may be due to the use of different assays such as dextran- coated charcoal and hydroxylapatite methods. For ex- ample, the Kd value determined from Xenopus liver cytosol ER has been reported to vary from 0.5 to 15 nM [31,32]. Similarly, the Kd value determined from turtle ER using two different receptor sources, hepatic and testis cytosol, varies from 0.7 to 17 nM, respec- tively [33,34]. This suggests that differences in protein preparation, assay conditions and assay methods may contribute to the variability in the reported [(6 values. The ER has been shown to bind several structurally diverse chemicals. This property appears to be unique among nuclear receptors and is also true for mam- malian ER: and ERB subtypes [13] as well as for ERs from non-mammalian species [15,35]. Crystal structures of hERa E domain in complex with E2 [18,19] support ER-EZ interaction models generated from binding stud- ies, structure activity relationships and three-dimen- sional homology models using crystallographic data from other nuclear receptors [36-38]. EZ binding is achieved by a combination of specific hydrogen bond- ing interactions and the hydrophobic nature of the binding pocket. The promiscuity of the ER has been partially attributed to the size of the ligand binding pocket, which is almost twice the volume of E2 [19]. Despite the differences in sequence identity among spe- cies [14,17], ERs from all species harbor the same three equivalent amino acids to hERa (Glu353, Arg394 and Hi5524) that participate in direct hydrogen bonds and stabilize E2 in the binding pocket [18,19]. However, differential binding of several natural and synthetic chemicals to hERa and hERB, as well as to ERs from different species have been reported [l4—l6,35,39-4l]. This suggests that additional amino acid residues may also play a role in determining ligand preference and relative binding affinity. Genistein has been shown to preferentially bind with 30-fold greater affinity to hERB than hERa [13]. The recent report of the crystal structure of ERB in complex with genistein has suggested that this ligand preference may be attributed to two conservative mutations within the binding pocket that may be responsible for further stabilizing the hERB-genistein complex [42]. Although many of the compounds examined in this study, including E2, bound with similar affinity to all five GST-ERdef fusion proteins, some notable differ- ences were reported. The most striking differences in relative binding affinities were seen with GST-aERdef and GST-rtERdef. Comparison of the amino acid dif- ferences within their respective ligand binding domains suggests that aERdeszhel75, rtERdeszet93 and rtERdelele272 may contribute to the observed differ- ences in ligand preference and relative binding affinities. Preliminary mutagenesis studies indicate that these residues influence relative and absolute binding affini- ties of a subset of estrogenic compounds (Matthews, J.B. et al. manuscript in preparation). These residues may change the hydrophobicity and volume of the binding pocket as well as result in unique ligand-residue 76 H‘ '1 if?! E. T! C'.‘ 232 J. Matthews et al. /’Journal of Steroid Biochemistry & Molecular Biology 74 (2000) 223-234 interactions. Recent crystallography data has demon- strated that each of these parameters can influence ligand binding. For example, the volume of the probe- occupied ligand pocket of ERa-EZ crystal complex has been determined to be 490 A while that of ERB-genis- tein is 390 A with the reduction being primarily due to the replacement of the Leu384 in hERa with a bulkier Met336 residue in hERB [42]. This allows the residues that line the pocket to pack more tightly around genis- tein, stabilizing the ligand in the binding pocket in ERB [42]. There were no obvious relationships between se- quence identity and binding affinity. GST-rtERdef has the lowest sequence identity compared to GST-hERdef, but this was not predictive of the binding affinity of a compound for GST-rtERdef. For example, 4-t- octylphenol bound with greater affinity to GST-rtER- def (le = 0.11 :t 0.02 uM) compared to GST-hERadef (ICSO=2.4;+_0.7 uM) while the rank order binding affinities were reversed for coumestrol (IC,0 values of 0.36:0.03 uM vs. 1.4i0.l uM for GST-hERozdef and GST-rtERdef, respectively). How- ever, some patterns in the relative binding affinity data were observed. Cluster analysis based on amino acid sequence identity suggested that hERadef and mERadef shared greater similarity than cERdef and aERdef, with rtERdef being the most divergent when compared to the other ER sequences (Fig. 4). In gen- eral, ERdef proteins with greater similarity exhibited similar relative binding affinities as illustrated in Fig. 3, with some notable exceptions. Although cERdef and aERdef shared the greatest similarity, DHT exhibited a 42-fold difference in relative binding affinity between the two species. The difference was only seven-, eight- and 12-fold for GST-hERdef, GST-mERdef and GST- rtERdef, respectively. GST-rtERdef, which has the most divergent amino acid sequence according to the cluster analysis, exhibited the greatest promiscuity in its ligand preference, further supporting the hypothesis that structural differences within the ligand binding domains among ERs of different species influences ligand preference and relative binding affinity. The results demonstrate that ERs from human, mouse, chicken, green anole and rainbow trout exhibit differential ligand preferences and relative binding affinities for a number of natural and synthetic com- pounds. This data can be used to further develop ER quantitative structure activity relationships (QSARs) [43] and to evaluate the feasibility of species-specific ER QSARs [44]. Although the majority of substances ex- amined in this study exhibited comparable relative binding affinities across ERs, a significant number of differences were observed. The relative binding affinities of the GST-rtERdef, which has the greatest amino acid variation in its E domain relative to the other species examined, exhibited the most striking differences. The rtER also had the greatest ligand promiscuity, binding a significantly greater number of structurally diverse estrogenic compounds. However, pharmacokinetic and pharrnacodynamic differences between species make it unlikely that differences in binding affinities for individ- ual estrogenic compounds would be observed in vivo. Nevertheless, in the absence of structural data for natu- ral and synthetic ligands, this cross species comparison provides valuable insights into potentially important residues that may play critical roles in the interaction between structurally diverse ligands and the ER binding pocket. Acknowledgements The authors would like to thank the United States Environmental Protection Agency (R 826301-01-0) and the Chemical Manufacturers Association for providing funding for this research. References [1] S. Mosselman, J. Polman, R. Dijkema, ER beta: identification and characterization of a novel human estrogen receptor, FEBS Lett. 392 (1996) 49-53. [2] R. Evans, The steroid and thyroid hormone receptor superfam- ily, Science 240 (1988) 889—895. [3] T. Colbom, F.S. vom Saal, A.M. Soto, Developmental effects of endocrine-disrupting chemicals in wildlife and humans, Environ. Health Perspect. 101 (l993) 378—384. [4] E. Carlsen, A. Giwercman, N. Keiding, N.E. Skakkebaek, De- clining semen quality and increasing incidence of testicular can- cer: is there a common cause?, Environ. Health Perspect. 103 (1995) 137-139. [5] PJ. Reijnders, Reproductive failure in common seals feeding on fish from polluted coastal waters, Nature 324 (1986) 456—457. [6] L.J.J. Guillette, T.S. Gross, GR. Masson, J.M. Matter, HF. Percival, A.R. Woodward, Developmental abnormalities of the gonad and abnormal sex hormone concentrations in juvenile alligators from contaminated and control lakes in Florida, Envi« ron. Health Perspect. 102 (1994) 680-688. [7] D.M. Fry, Reproductive effects in birds exposed to pesticides and industrial chemicals, Environ. Health Perspect. 103 (1995) 165—171. [8] H. Von Westemhagen, H. Rosenthal, V. Dethlefsen, W. Ernst, U.L. Harms, P.D. Hansen, Bioaccumulating substances and reproductive success in the Baltic flounder, Platichthys flesus, Aquatic Tox. l (1981) 85-99. [9] SH. Safe, Environmental and dietary estrogens and human health: is there a problem? Environ. Health Perspect. 103 (1995) 346—351. [10] KS. Korach, S. Migliaccio, V.L. Davis, in: P.L. Munson (Ed), Principles of Pharmacology —— Basic Concepts and Clinical Applications, Chapman and Hall, New York, 1994, pp. 809- 825. [11] MJ. Messina. V. Persky, K.D. Setchell, S. Barnes, Soy intake and cancer risk: a review of the in vitro and in vivo data, Nutr. Cancer 21 (1994) ll3-l3l. [12] CJ. Mirocha, S.V. Pathre. C.M. Christensen, in: Y. Pomeranz (Ed), Advances in Cereal Chemistry. American Association of Cereal Chemists, St. Paul, MI, 1980. pp. 159-225. 77 .1. Matthews at al. /’ Journal of Steroid Biochemistry & Molecular Biology 74 (2000) 223—234 233 [13] GO. Kuiper, J.G. Lemmen, B. Carlsson, J.C. Corton, S.H. Safe, P.T. van der Saag, B. van der Burg, J .A. Gustafsson, Interaction of estrogenic chemicals and phytoestrogens with estrogen recep- tor beta, Endocrinology 139 (I998) 4252—4263. [14] J.B. Matthews, T.R. Zacharewski, Differential binding affinities PCBS, HO-PCBs and Aroclors with recombinant human, rain- bow trout (Onchorhynlciss mykiss) and reptilian (Anolis caroli- nensis) estrogen receptors using a semi-high throughput competitive binding assay, Toxicol. Sci. 53 (2000) 326-339. [15] Y. Le Drean, L. Kern, F. Pakdel, Y. Valotaire, Rainbow trout estrogen receptor presents an equal specificity but a differential sensitivity for estrogens than human estrogen receptor, Mol. Cell. Endocrinol. 109 (1995) 27-35. [16] D.W. Fitzpatrik, CA. Picken, LC. Murphy, M.M. Buhr, Mea- surement of the relative binding affinity of zearalenone, zit-zear- alenol and B-zearalenol for uterine and oviduct estrogen receptors in swine, rats and chickens: an indicator of estrogenic potencies. Comp. Biochem. Physiol. 94C (1989) 69l —694. [17] MR. Fielden, 1. Chen, B. Chittim, S.H. Safe, T.R. Zacharewski, Examination of the estrogenicity of 2,4,6,2,6-pentachloro— biphenyl (PCB 104), its hydroxylated metabolite 2,4,6,2,6-pen- tachloro—4-biphenylol (HO-PCB 104), and a further chlorinated derivative,2,4,6,2,4,6-hexachlorobiphenyl, Environ. Health Per- spcct. 105 (1997) 1238-1248. [18] D.M. Tanenbaum, Y. Wang, S.P. Williams, P.B. Sigler, Crystal- lographic comparison of the estrogen and progesterone recep- tor’s ligand binding domains, Proc. Natl. Acad. Sci. USA 95 (1998) 5998-6003. [19] A.M. Brzozowski, A.C. Pike, Z. Dauter, R.E. Hubbard, T. Bonn, O. Engstrom, L. Ohman, G.L. Greene, J.A. Gustafsson, M. Carlquist, Molecular basis of agonism and antagonism in the oestrogen receptor, Nature 389 (1997) 753—758. [20] L.J. Young, J. Godwin, M. Grammer, M. Gahr, D. Crews, Reptilian sex steroid receptors: amplification, sequence and ex- pression analysis, J. Steroid Biochem. Mol. Biol. 55 (1995) 261-269. [21] J.M. Bergeron, M. Gahr, K. Horan, T. Wibbels, D. Crews, Cloning and in situ hybridization analysis of estrogen receptor in the developing gonad of the red-cared slider turtle, a species with temperature-dependent sex determination, Dev. Growth Differ. 40 (1998) 243~254. [22] A. Krust, S. Green, P. Argos, V. Kumar, P. Walter, J. Bornert, P. Chambon, The chicken oestrogen receptor sequence: homol- ogy with vverbA and the human oestrogen and glucocorticoid receptors, EMBO J. 5 (1986) 891—897. [23] DA. Seielstad. K.E. Carlson, J.A. Katzenellenbogen, P.J. Kush- ner, G.L. Greene, Molecular characterization by mass spec- trometry of the human estrogen receptor ligand-binding domain expressed in Escherichia coli, Mol. Endocrinol. 9 (1995) 647- 658. [24] G. Scatchard, The attractions of proteins for small molecules, Ann. New York Acad. Sci. 51 (1949) 660—672. [25] P. Aranyi, Kinetics of the hormone-receptor interaction. Compe- tition experiments with slowly equilibrating ligands, Biochim. Biophys. Acta 628 (1980) 220—227. [26] M. Xu, P.K. Chakraborti, MJ. Garabedian, K.R. Yamamoto, S.S. Simons, Modular structure of glucocorticoid receptor do- mains is not equivalent to functional independence. Stability and activity of the steroid binding domain are controlled by se- quences in separate domains, J. Biol. Chem. 271 (1996) 21430— 21438. [27] CH. Wooge. G.M. Nilsson, A. Heierson, D.P. McDonnell, B.S. Katzenellenbogen, Structural requirements for high affinity lig- and binding by estrogen receptors: a comparative analysis of truncated and full length estrogen receptors expressed in bacte- ria, yeast, and mammalian cells, Mol. Endocrinol. 6 (1992) 861 —- 869. [28] S. Jaglaguier, D. Mesnier, J .L. Lager, G. Auzou, Putative steroid binding domain of the human mineralocorticoid receptor, ex- pressed in E. coli in the presence of heat shock proteins shows typical native receptor characteristics, J. Steroid Biochem. Mol. Biol. 57 (1996) 43-50. [29] F. Pakdel. F. Le Gac, P. Le Goff, Y. Valotaire, Full-length sequence and in vitro expression of rainbow trout estrogen receptor cDNA, Mol. Cell Endocrinol. 71 (1990) 195-204. [30] RM. Donohoe, LR. Curtis, Estrogenic activity of chlordecone, o,p'-DDT and op'-DDE in juvenile rainbow trout: indication of vitellogenesis and interaction with hepatic estrogen binding sites, Aquatic Toxicol. 36 (1996) 31-52. [31] B. Westley, J. Knowland, An estrogen receptor from Xenopus laevis liver possibly connected with vitellogenin synthesis, Cell 15 (1978) 367-374. [32] l. Lutz, W. Kloas, Amphibians as a model to study endocrine disruptors: l. Environmental pollution and estrogen receptor binding, Sci. Total Environ. 225 (1999) 49—57. [33] SM. Ho, S. Fehrer, M. Yu, L.C. Liang, D. Press, High-affinity binding of [’H]estradiol-l7 beta by an estrogen receptor in the liver of the turtle, Gen. Comp. Endocrinol. 70 (1988) 382—394. [34] P. Mak, S.M. Ho, l.P. Callard, Characterization of an estrogen receptor in the turtle testis, Gen. Comp. Endocrinol. 52 (1983) 182—189. [35] RM. Vonier, J.L.J. Guillette, J.A. McLachlan, S.F. Arnold, Identification and characterization of estrogen and progesterone receptors from the oviduct of the American alligator (Alligator mississippiensis), Biochem. Biophys. Res. Commun. 232 (1997) 308—312. [36] GM. Anstead, K.E. Carlson, J.A. Katzenellenbogen, The estra- diol phannacophore: ligand structure—estrogen receptor binding affinity relationships and a model for the receptor binding site, Steroids 62 (1997) 268—303. [37] J.M. Wurtz, U. Egner, N. Heinrich, D. Moras, A. Mueller- Fahrnow, Three-dimensional models of estrogen receptor ligand binding domain complexes, based on related crystal structures and mutational and structure-activity relationship data, J. Med. Chem. 41 (1998) 1803—1814. [38] D. Lewis, M. Parker, R. King, Molecular modelling of the human estrogen receptor and ligand interactions based on site- directed mutagenesis and amino acid sequence homology, Steroid Biochem. Mol. Biol. 52 (1995) 55-65. [39] GO. Kuiper, B. Carlsson, K. Grandien, E. Enmark, J. Hagg- blad, S. Nilsson, J.A. Gustafsson, Comparison of the ligand binding specificity and transcript tissue distribution of estrogen receptors alpha and beta, Endocrinology 138 (1997) 863—870. [40] T. Barkhem, B. Carlsson, Y. Nilsson, E. Enmark, J. Gustafsson, S. Nilsson, Differential response of estrogen receptor alpha and estrogen receptor beta to partial estrogen agonists/antagonists, Mol. Pharmacol. 54 (1998) 105—112. [41] K. Connor, J. Howell, 1. Chen, H. Liu, K. Berhane, C. Sciar- retta, S. Safe, T. Zacharewski, Failure of chloro~S-triazine- derived compounds to induce estrogen receptor-mediated responses in vivo and in vitro, Fund. Appl. Toxicol. 30 (1995) 93—101. [42] AC. Pike, A.M. Brzozowski, R.E. Hubbard, T. Bonn, A.G. Thorsell, 0. Engstrom, J. Ljunggren, J. Gustafsson, M. Car- lquist, Structure of the ligand-binding domain of oestrogen receptor beta in the presence of a partial agonist and a full antagonist, EMBO 1. I8 (1999) 4608—4618. [43] CL. Waller, D.L. Minor, J.D. McKinney, Using three—dimen- sional quantitative structure—activity relationships to examine estrogen receptor binding affinities of polychlorinated hydroxyo biphenyls, Environ. Health Perspect. 103 (1995) 702—707. [44] W. Tong, R. Perkins, R. Strelitz, E.R. Collantes, S. Keenan. W.J. Welsh, W.S. Branham, D.M. Sheehan. Quantitative struc- tureactivity relationships (QSARs) for estrogen binding to the 78 234 J. Matthews at al. ,/ Journal of Steroid Biochemistry & Molecular Biology 74 (2000) 22.3—234 estrogen receptor: predictions across species, Environ. Health Perspect. 105 (1997) 1116-1124. [45] R.L. Eckert, B.S. Katzenellenbogen, Physical properties of estro- gen receptor complexes in MCF-7 human breast cancer cells. Differences with anti-estrogen and estrogen, J. Biol. Chem. 257 (1982) 8840—8846. [46] GB. Tremblay, A. Tremblay, N.G. Copeland, D.J. Gilbert. N.A. Jenkins, F. Labrie, V. Giguere, Cloning, chromo- somal localization, and functional analysis of the murine estrogen receptor beta, Mol. Endocrinol. 11 (1997) 353— 365. [47] T. Horigome, T.S. Golding, V.E. Quarmby, D.B. Lubahn, K. McCarty, Sr, K.S. Korach, Purification and characterization of mouse uterine estrogen receptor under conditions of varying hormonal status, Endocrinology 121 (1987) 2099-2111. [48] ER. Mulvihill, R.D. Palmiter, Relationship of nuclear estrogen receptor levels to induction of ovalbumin and conalbumin mRNA in chick oviduct, J. Biol. Chem. 252 (1977) 2060-2068. 79 CHAPTER 5 Ability of several natural and synthetic estrogenic chemicals to induce gene expression mediated by estrogen receptors from several vertebrate species 80 Abstract The ability of 15 natural and synthetic estrogenic compounds to induce gene expression mediated through EROt and B isoforms and ERs from different vertebrate species was examined. MCF—7 cells were transiently transfected with a Gal4-ER chimeric receptor consisting of the D, E and F domains of the human 0t (Gal4-hEROtdef), mouse 0t (Gal4-mEROLdef), mouse B (Gal4-mERBdef), chicken (Gal4-cERdef), green anole (Gal4-aERdef), xenopus (Gal4-xERdef) or rainbow trout 0t ERs (Gal4—rtEROtdef). The 17B-estradiol (E2) induced reporter gene expression was similar among the different constructs with ECso values ranging from 0.05 to 0.7 nM. Intriguingly, the E2 induced response mediated by Gal4-rtEROtdef was 2 orders of magnitude lower, with an ECSO value of 28 nM at 37°C. However, at 20°C only a 9-fold difference in ECso values was observed. As a result, the ability of test compounds to induced Gal4-rtEROtdef-mediated gene expression was examined at both 37°C and 20°C. Although, the response of E2 was similar among the ERs, many differences were observed. Coumestrol induced Gal4- mERBdef— and Gal4-aERdef-mediated reporter gene expression 164-fold and 8-fold greater compared to the other Gal4-ERs, respectively. In contrast to other Gal4-ERs, 0t- zearalenol induced Gal4-rtEROtdef-mediated reporter gene expression at lower concentrations than E2. In general, the gene expression data correlated well with the competitive binding results presented in chapter 4, with a Pearson r value of 0.86. Overall the results show that certain estrogenic compounds exhibit a differential ability to induce reporter gene activity mediated by ERs from different vertebrate species. In 81 addition, these data also highlight the importance of incubation temperature when examining rtEROL-mediated activity. 82 Introduction Several studies have reported differences in relative binding affinity (RBA) of estrogenic endocrine disruptors (BEDS) for different ER isoforms and ERs from different species (1,2). For example, in chapter 3 we have demonstrated that the rtERa exhibits greater affinity for several PCB congeners than either the human or anole ERs. In . chapter 4, several differences in RBA of EEDs for ERs from different classes of vertebrate species were observed. However, in other studies the RBAs of several EEDs did not correlate with their ability to induce the expression of estrogen responsive complement C3 mRNA in treated rat endometrial adenocarcinoma cells (3), suggesting that RBA is not a good predictor of estrogenic activity. This may not be surprising since ligand binding to the ER is the first event in a complex mechanism leading to changes in gene expression. The transcriptional activity of the ER is mediated by two separate activation functions, an NHz-terminal ligand- independent activation function (AF-1) and the ligand-dependent activation function (AF-2) located in the ligand binding domain. Although the mechanism by which the AF- 2 region transmits ligand signals to the basal transcriptional machinery is poorly understood, several proteins that interact with the AF-2 region in a ligand-dependent manner have been identified (4,5). These proteins, collectively termed cofactors, function as coactivators or corepressors to induce or inhibit gene expression, respectively. After estrogen binding and subsequent dimerization, the ER undergoes a conformational change, which allows it to bind to its cognate DNA target site, referred to as estrogen responsive elements (EREs) located in the regulatory region of estrogen-inducible promoters resulting in the transcriptional regulation of target genes (6). In addition, the 83 ligand-induced conformational change exposes critical residues within the AF-2 region that are essential for cofactor recruitment. Functional and structural studies have shown that coactivators interact with the AF-2 region via short leucine motifs (ie. LxxLL or NR- box) to transduce the ligand signal to the basal transcriptional machinery. Several structurally distinct classes of coactivator proteins have been identified. For example, the steroid receptor coactivator (SRC) family of coactivators (7) is a family of 160 kDa molecular weight proteins that were the first coregulators shown to interact with nuclear receptors. Members of this family include, SRC-1 (7,8), SRC-2 (9,10), and SRC-3 (11- 15) and they form a mutiprotein complex with CBP/p300 (12,14). Recent studies have shown that the ability of the ER to recruit SRC coactivators is dependent on the structural characteristics of the ligand (16). There are reports in which RBAs of EEDs do not correlated with transactivation ability (1,17). For example, despite the apparent greater affinity of genistein for ERB, genistein exhibits a similar ability to induce reporter gene expression mediated by either ERB or ERa (1). In addition, we have observed that bisphenol A also binds to ERB with greater affinity (IO-fold) than EROL, but only exhibited a 2-fold greater ability to induce gene expression mediated by either isoform (chapter 7). Since we have examined the ability of several EEDs to compete for binding to ERs from different species, it was of interest to examine whether differences in RBA correlated with the ability of EEDs to induce gene expression mediated by EROts from five different vertebrate classes and a representative mammalian ERB. 84 Materials and methods Chemicals and biochemicals 175-estradiol (1 ,3,5[10]-estratriene-3,17B-diol), DHT (dihydrotestosterone, 50t- androstan-17B-ol-3-one), diethylstilbestrol (4,4 ’-( l ,2-diethyl- l ,2-ethene-diyl)-bisphenol), a—zearalenol (2,4-dihydroxy-6-[60t,lO-dihydroxy-undecyl]benzoic acid [.t—lactone), B- E zearalenol (2,4-dihydroxy-6-[6B,10-dihydroxy-undecyl]benzoic acid [t-lactone), genistein (4’,5,7-trihydroxyisoflavone), naringenin (4’,5,7—trihydroxyflavanone 7- rhamnoglucoside) and B-sitosterol (22,23-dihydrostimasterol) were purchased from L Sigma (St. Louis, MO). Coumestrol (2-(2,4-dihydroxyphenyl)-6-hydroxy—3- benzofurancarboxylic acid lactone) was obtained from Arcos Organics (Pittsburgh, PA). Methoxychlor (1,1,1-trichloro-2,2-bis(p-methoxphenyl)ethane) was provided by William Kelce (Monsanto, St. Louis, MO). o,p’-DDT (1,1,1-trichloro-2-[2-chlorophenyl]-2-[4- chlorophenyl]ethane), and p,p’-DDT (1,1,1-trichloro-2,2-bis[4-chlorophenyl]ethane), were purchased from AccuStandard (New Haven, CT). The alkyl phenolic compound 4- t-octylphenol and bisphenol A (4,4’-isopropy1idenediphenol) were obtained from Aldrich (Milwaukee, WI). Atrazine, was from Chem-Service (West Chester, PA). Vent DNA polymerase was purchased from New England Biolabs (Beverly, MA), and restriction enzymes were obtained from Roche/Boehringer Mannheim (Indianapolis, IN). Phenol red-free Dulbecco’s modified Eagle’s medium (DMEM) and medium supplements were from Life Technologies (Gaithesburg, MD). Fetal bovine serum (FBS) and D-luciferin were purchased from Intergen (Purchase, NY) and Molecular Probes (Eugene, OR), respectively. All other chemicals and biochemicals were of the highest quality available from commercial sources. 85 Construction of plasmids The plasmid pG4M-hEROidef (Gal4-HEGO) was a gift from Dr. P. Chambon (IGBMC CNRS-LGME, Illkirch Cedex C.U. de Strasbourg, France). The construction of pG4M-mERBdef has been described elsewhere (18). The plasmid pG4M-rtERadef was constructed by PCR amplifying amino acid residues 214-576 of the rainbow trout EROL as described previously (Matthews M in press). The plasmid pG4M-mEROtdef (a.a. 268- 599) was generated by PCR amplification of the plasmid pJ3MOR containing the complete mouse ER cDN A (provided by M.G. Parker, Molecular Endocrinology Research Laboratory, London, United Kingdom) using primers 5’- caaagaaattcatcgattggcggcatacggaaagaccgc-3’ (forward) and 5 ’- aaaagaattcgcggccgctcagatcgtgttggggaagccctc-3’ (reverse). The plasmid pG4M-cERdef (a.a. 257-599) was constructed by PCR amplification using primers 5'- aaaactcgagccaaaggtggaatccggaaagac-3’ (forward) and 5’- aaaaagatctttatattgtattctgcatactctcctct-3' (reverse). The plasmid pG4M-aERdef (a.a. 8-349) was generated by PCR amplification using primers 5'-aaaa ggatccctcgagccggtggaattcggaaagaccgcag-3' (forward) and 5'- aaaaggatccctcgagtcaaattgcttcctgctcatttccc-3' (reverse), whereas the plasmid pG4M- xERdef (a.a. 248-586) was constructed by PCR amplification of the plasmid CMVSxERl containing the complete Xenopus ER cDNA (provided by Dr. D. Shapiro, University of Chicago, Urbana, IL) using primers 5’-aaaactcgagccgggggcattcgaaaggatcgca-3' (forward) and 5’-aaaaggtaccgagctctcatactgtgctttgtaagctcact-3' (reverse). PCR fragments were digested with the appropriate restriction enzymes and ligated into the eukaryotic 86 expression vector containing the DNA binding domain of the yeast transcription factor Gal4, pG4MpolyII (Dr. P. Chambon). The plasmid pGEXKG-ACTR (SRC-3) (a.a. 615 to 768) was a gift from Dr. L. Freedman (Memorial Sloan-Kittering Cancer Center, New York, NY), whereas the plasmid pGEX-TIF2 (SRC—2) (a.a. 594 to 766) was constructed by PCR amplification of the plasmid pSGS-TIF2 (Dr. P. Chambon) using primers 5'- caaaggatccgaaggtacaactggacaagcagag 3' (forward) and 5'- caaactcgagtcaatctgcttactgtccagtctctc 3' (reverse). PCR fragments were digested with the appropriate restriction enzymes and ligated into pGEX6p3, a GST expression vector (Amersham Pharmacia, Piscataway, NJ). All PCR amplification was performed as previously described (19). The sequence of each construct was confirmed with restriction enzyme digest and ABI/Prism ' automated sequencing (Perkin Elmer Applied Biosystems, Foster City, CA). Cell culture and transient transfection assays MCF-7 human breast cancer estrogen receptor positive cells (obtained from Dr. L. Murphy, University of Manitoba) were maintained with phenol red-free DMEM supplemented with 3.7 g/l NaHCO3, 2 mM L-glutamine, 10% FBS, 10 mM HEPES, 500 [Lg/ml gentamicin, 2.5 ttg/ml amphotericin B, 100 IU/ml penicillin G, and 100 [lg/ml streptomycin. Cells were cultured in a humidified environment at 37°C with 5% C02. Transient transfections and gene transcription assays were performed essentially as previously described (20,21). Briefly, MCF—7 cells were seeded at approximately 50% confluency in 6-well tissue culture plates in medium supplemented with 5% dextran- coated charcoal-treated fetal bovine serum (DCC-FBS) and allowed to settle for 7 h. 87 MCF-7 cells were transiently transfected by the calcium phosphate precipitation method (22) with 1.5 pg of l7m5-G-Luc (provided by Dr. P Chambon), 0.2 pg of either pG4M- hERadef, pG4M -mEROLdef, pG4M -mERBdef, pG4M -rtEROtdef or 0.5 pg of pG4M - cERdef, pG4M -cERdef, or pG4M -xERdef, and 0.05 pg of pCMV-lacZ (B-galactosidase expression vector). Transiently transfected cells were washed 16 h later with sterile phosphate buffered saline and fresh medium was added to each well. Transiently transfected cells were exposed to final concentrations ranging from 10'12 to 10'5 M of test compound. Final concentrations were obtained by adding 2 pl of test chemical to 2 ml of medium. After a 24 h incubation with test compound, cells were harvested and assayed for luciferase activity according to standard methods (23). Each treatment was done in duplicate and two samples were taken from each replicate. Each experiment was repeated three times. Values are reported as a percentage relative to the maximum induction observed with E2. Reverse transcriptase polymerase chain reaction Total RNA from MCF-7 cells was isolated using the Trizol reagent according to manufacturer’s instructions (Life Technologies). In brief, cells were lysed with two successive washes of 325 pl aliquots of Trizol. The lysates were pooled and incubated with 300 pl of chloroform for 10 min at 23°C. The mixture was separated by centrifugation at 12,000xg for 15 min at 4°C and the aqueous layer (1 ml) containing the RNA was removed and precipitated with 500 pl of isopropanol and 2 pl of linear acrylamide (10 mg/ml). The RNA was pelleted by centrifugation at 12,000xg for 10 min at 4°C and the resulting pellet was washed with 75% ethanol. The pellet was then air 88 dried and resuspended in lOOuL of 3M sodium citrate. The RNA was stored at —20°C until use. Total RNA (5 pg) was incubated for 10 min at 70°C with 0.5pg oligo dT primer (5'-ttttttttttttttttttttvn-3'). The reaction was chilled on ice for 2 min and the mRNA was reverse transcribed in a 20 pl reaction mixture containing PCR buffer (20 mM Tris-HCI (pH 8.4), 50 mM KCl, 2.5 mM MgClg), 10 mM D'I'I‘, 500 pM dNTPs and 200 IU of SuperScript II reverse transcriptase for 60 min at 42°C. The reaction was terminated with a 15 min incubation at 70°C. One pl of the reverse transcription reaction was used in the subsequent PCR reactions using the following primers for B-actin: 5'- aaaagcatccaagcttctgaagtaccccattgaacatggca—3' (forward) (5'- aaaactcgaggcggccgctgtcacgcacgatttccctctcag-3’ (reverse), SRC-1: 5'- caaaccatggatccagacagtaaatactctcaaaccagtc-3' (forward) (5'- caaactcgagtcaatcaggctcgacagacaaagtgg-3' (reverse), SRC-2: 5'- caaaggatccagaaggtacaactggagaagcagag-3' (forward) and 5'- caaagaattctcagtgatggtgatggtgatgatctgtcttactgtccagtctctc-3' (reverse), and SRC-3: 5'- gaaagtaaggagagcagtgttgag-3' (forward) and 5'-gtcagaactagtcagatcaccaag 3' (reverse). After the addition of template, the samples were incubated at 94°C for 3 min and amplified for 25 cycles. Each cycle included: 45 sec. denaturation at 94°C, 45 sec annealing at 60°C and 1 min elongation at 72°C. PCR products were separated by 2% agarose electrophoresis and visualized with ethidium bromide staining. GST pull-down assays 89 The expression and purification of GST fusion proteins was performed as described previously (24,25). Glutathione-Sepharose beads (Amersham Pharmacia) were prewashed in NETN buffer (0.5% Nonidet P-40, 1 mM EDTA, 20 mM Tris, pH 7.4, 100 mM NaCl, 10% glycerol). Crude bacterial extracts containing the fusion proteins (GST- SRC-l, GST-SRC-2, and GST-SRC—3) were purified onto the prewashed beads (10 pl beads + 20 pl crude in 500 pl total reaction) by incubation at 4°C for 1.5 h on a rotary mixer. The beads (loaded with fusion protein) were collected by centrifugation and washed three times with NETN buffer. GST fusion proteins or GST alone were incubated overnight in 1.5 ml centrifuge tubes containing 490 pl of NETN buffer containing 1% bovine serum albumin, 5 pl of the in vitro translated 3SS-labeled receptor, in the presence of 5 p1 of vehicle or test compound. Beads were washed four times with NETN, and dried under vacuum for 10 min. The dried beads were resuspended in 25 pl of 3x protein loading buffer (reducing), incubated for 5 min at 95°C, and the entire sample was separated by 12% SDS-PAGE. The gels were fixed, dried, and the radiolabelled rtER was visualized by fluorography. The amount of bound 35S-labelled protein was quantified with a PhosphorImager (Molecular Dynamics, Sunnyvale, CA). 90 Results Ability of 17B-estradiol to induce gene expression mediated by ERs from several different species E2 and 14 EEDs, several of which have been shown to exhibit preferential affinity for different vertebrate ERs (Chapter 4 and reference (24)) were examined for their ability to induce ER-mediated gene expression by measuring luciferase activity using MCF-7 cells co-transfected with the Gal4-ERdef and a Gal4-regulated luciferase reporter gene, l7m5-G-Luc. The amino acid residues of the ER D, E, and F domains used to generate the Gal4-ERdef fusion proteins Gal4-hEROtdef (human or), GaI4-mEROtdef (mouse or), Gal4-mERBdef (mouse B), Gal4-cERdef (chicken), Gal4-aERdef (anole),- Gal4-xERdef (xenopus), and Gal4-rtEROtdef (rainbow trout) and their sequence identity compared to hERa are shown in figure 1. The results in figure 2 show that E2 treatment of transiently transfected MCF-7 cells with Gal4-ERdef caused a concentration- dependent increase in luciferase activity. The ECso value for this response was similar among the different chimeric receptors ranging from 0.05 to 0.7 nM. However, the E2 induced response mediated by Gal4-rtEROtdef was 2 orders of magnitude lower, with an ECSO value of 28 nM when transfected cells were maintained at 37°C. When, MCF-7 cells transiently transfected with Gal4-rtEROtdef were incubated at 20°C, a more physiological relevant temperature for rainbow trout, the ECso value was reduced to 1 nM. The molecular basis for this temperature sensitive phenotype of the rtERa is more thoroughly investigated in chapter 6. Surprisingly, neither the xenOpus nor 91 Tl , 595 I ‘ Gal4-hERa 599 T 13313-] 93'4‘mER0‘ 485 I '17 I Gal4-mERI3 9 . “58 ‘. ”([19] Gal4-cER .. 349 I32] Gal4-aER 587 , I 37 I Gal4-xER . 576 I19I Gal4-rtERa Figure 1. Schematic representation of Gal4-ER chimeric receptors transiently transfected into MCF-7 human breast cancer cells. The chimeric receptors consisted of the DNA binding domain of the yeast transcription factor, Gal4, linked upstream of the last 11 amino acid residues of the C domain and the entire D, E, and F domains of ERs from several vertebrate species. Numbers provided above identify the amino acids used in the construction of the expression vectors. The numbers wthin the domains indicate the % amino acid sequence identity compared to hERa. 92 > w 125 125 O .2 38100- as... >,o .t: 3 '58 75- O 32""; 50! .7?» .53.: '8 o :3 25‘ 45 5° 0- Log [concentration] Figure 2. Comparison of the ER-induced transactivation ability mediated by Gal4-ERs. MCF—7 human breast cancer cells transiently transfected with 0.2 pg Gal4-hEROt (o), Gal4-mERB (I), Gal4-rtEROt ( O) or 0.5 pg of Gal4-cER (A), Gal4—aER (O), Gal4-xER ( El), 1.5 pg of Gal4 regulated-luciferase reporter gene, 17m5-G-Luc, and pCMV as described in the methods section. Following dosing with E2, cells were incubated at the 37°C (panel A), with the exception of cells transfected with Gal4-rtEROt which were incubated at 20°C (panel B). 93 the anole ERs exhibited temperature sensitive phenotypes as judged by their ability to induce reporter gene expression at 37°C (Figure 2). The maximal E2 induction ranged from 20- to lOO-fold among the Gal4-ERs. Due to the variability in fold induction the data are reported as percent luciferase activity relative to the maximal activity achieved by E2 with each receptor, which was arbitrarily set to 100% for comparative purposes. RtERa recruits SRC coactivators GST pull-down assays were performed to verify that the reduced transactivation activity of the rtEROt was not due to an inability of the rtERa to interact with human coactivators. Figure 3 shows a comparison of the expression levels of three different classes of the SRC family of coactivators. RT-PCR analysis of total RNA isolated from MCF-7 cells revealed that the expression level of SRC-3 (ACT R) was approximately 5- fold higher than either SRC-l or SRC-2 (TIF2). Figure 4 shows a typical GST pull-down assay, where the dose-dependent recruitment of 35s-rtERot by GST-SRC-2 and GST- SRC-3 with E2 is demonstrated. Similar results were seen with SRC-1 (data not shown). These data suggest that differences in transactivation are not a result of the rtEROt failing to recruit human coactivators and are rather due to functional differences between the hERa and rtERor that may be the result of the sequence variability within their respective ligand binding domains. 94 f "t \ Figure 3. Differential expression levels of SRC coactivators in MCF-7 cells. Total RNA from MCF-7 cells was reverse transcribed with an oligo dT primer, followed by a 25 cycle PCR amplification with gene specific primers. The samples were separted on 2% agarose gel and visualized with ethidium bromide staining. Lane 1: 100 bp molecular marker, lane 2: B—actin , lane 3: SRC-l, lane 4: SRC-2, and lane 5: SRC-3. 95 r'r- 10% input Beads E-' 8 GST-SRC—2 GST-SRC-3 E2 (M) - -8 —6 - -8 -6 rtERorDEF—> t... 5000- 3000- 2000- s 9 ’~ e a to e a re ”>09 ea" 0” (9° to a ¢~° e2 «a e " GST-SRC-2 GST-SRC—3 Figure 4. The rtERor DEF binds human SRC-2 and SRC-3. (A) GST fusion proteins previously coupled to Sepharose beads were incubated with in vitro translated, [3SS]methionine-labelled rtERa DEF, in the presence or absence of l7B-estradiol (E2) After extensive washing, samples were separated by 12% SDS- polyacrylamide gels. Gels were fixed, dried and visualized by fluorography. (B) Graph of the quantification of the pixel intensity using a phosphorimager. 96 Ability of BEDS to induce gene expression mediated by ERs from several different species A summary of the half maximal activation (ECso) values for the 14 EEDs examined is presented in Table 1. For EEDs unable to achieve 50% of the maximal activation of E2, the percent activation at the highest dose (10 pM) was reported. EEDs unable to achieve 20% of the maximal E2 response were classified as weak inducers (wi), while EEDs that induced less than 10% of the maximal E2 response were considered non—inducers (ni). Since the Gal4-rtEROtdef displayed improved E2 induced transactivation ability at 20°C compared to 37°C, both temperatures were used to examine rtERor mediated gene expression. In general, the ability of several compounds to induced Gal4-rtERdef-mediated reporter gene expression was increased at 20°C (Table l) and therefore this was considered to be a more suitable temperature to examine rtERa mediated transactivation activity. Overall the E2 induced reporter gene expression was similar among the ERs, though several differences in the rank order and transactivation response of EEDs were observed. DES and E2 exhibited a similar ability to induce reporter gene expression mediated by the different ERs, whereas or-zearalenol induced Gal4-rtERordef-mediated reporter gene expression at lower concentrations than E2 or DES, which was not observed in other Gal4-ERs. The rank order of the phytoestrogens was similar among the ERs (coumestrol >/= genistein>>naringenin), though differences in potency among the receptors were observed. Coumestrol induced reported gene expression 164- and 8-fold greater mediated through Gal4-mERBdef and Gal4-aERdef, respectively, than through 97 :21 o: 8% poems—axe “mop—wE 05 E @9830 83 >658 28w cocoa“: we 53.65 58$:me e: 85m 95 8265 :o: messen— b .921 o: poem—=98 omen “monwE 05 E .3358 038E:— Eoocoa 8 E83— ... .335“ 05% cocoa»: E ommocoE eaom a 52: v.8— 5 3:32 EoEEob 85m 95 .5265 x33 8.280 e .o.m EEEEO E conga omaoamBoEe EEoEwB ofi wEms coca—e83 203 $2? 80m e E E E E E E v: E o=§h< E E E E E E e: eE .80395& E E E E E E on: 5 bank: a e a e L: x we a 3 a re x I seen Ea .3 a a a a a s, s, .5 82:5st e-o_ x mN H N6 83 E eb— x 0.. H .N 1: x wd H 0; #2 x wd H 3. as x QN «.2 x Nd H 9m chwEEZ c2 x v; H md 92 x md H m.m we on a.m H ma co— x fie H m4 we x 3 H ore. eo— x 0.0 H me #2 x m; #8 x Ne H 0% EBmEoO no. x .N H _.n maom co. x QN H as 92 x ed H n6 1: x mm H m6 ea: x «A H Em we x fin 3.2 x fin H mp 358830 co. x w.m H 2. $8 oaom 92 x mg H we L: x a._ H N._ L: x to H ad 1: x ms cc. x mg H 9.. Becca—booed c2 x 9N H m.m meow flRoom econ *3 c2 x N._ H a.m c2 x _.N e2 x Nd H mm (am we: fioHo; $2 co— x : Hvé 9o. xadH .N we :2 He; 92 xadHa._ ub— xod L: .56de FIG 23 x _.o H ma ”-2 x a.~ H 90 we x Q». H 2. so. x Wm H arm 93 x no H m.m eb— x o._ H Nd an: x m._ H M: ea: x 0N H as Eco—«Si 2.2 x We H _.~ 92 x N. H md 12 x _.N H rem eb— x mM. H Om 2.2 x m.m H 9a 2.3 x 06 H v6 2.2 x _.o H A: 2-2 x Nd H as Eco—«SA. 2.2 x me H ma ”.2 x m6 H m._ 2.0— x wd H w; :b. x Wm H ad 2-2 x A: H a._ :b. x a; H 90 :.o_ x ON H NM 2-2 x 6c H mg amn— e.o_ x We H o._ “.3 x wd H wd 2.2 x Wm H ad 2.9: cd H We Sb— x wd H ad .72 x EN H as :-o_ x Wm H No :b. x Na H ma EEabmotn: Deon Us?“ ammutiao azure—a0 «more—«O «wee—a0 mug—m0 anEéEO dam—ciao amuse—a0 be sum .oflacofio BE; mmoE: .Ueam E mEoEono oEomobmo 0:059? new ESE: 9 8:398 a em a mike—Ea $6on :85ng 3.38 :8: mam 2: 3 3362: 53:8 ocow cotoaoc 38863 no v.25; 80m A 0.3—ah 98 the other receptors. Genistein and naringenin also preferentially induced reporter gene expression mediated by GaI4-mERBdef at concentrations 70 to 95 fold lower than the other receptors (Figure 5). Although these phytoestrogens are less potent than E2, coumestrol and genistein were able to generate a response mediated through ERB at concentrations similar to physiological hormone levels (IO-100 nM). B-Sitosterol did not induce any significant reporter gene expression through any of the ERs examined. Of the industrial chemicals examined, 4-t—octy1phenol was the most potent among the different ERs with ECSO values ranging from 0.01 to 0.1 pM (Figure 6). The EC50 values for bisphenol A (BPA) ranged from 0.3 to 3 pM, with BPA exhibiting a preferential ability to induce reporter gene expression mediated by Gal4-mERBdef and Gal4-rtEROtdef. BPA was only able to induce 25-30% of maximal activity of E2 through Gal4-cERdef, -aERdef, and -xERdef. The pesticide o,p’-DDT exhibited the most variability of all the compounds examined. o,p’-DDT induced reporter gene expression with ECso values of 0.8 :t 0.1 and 2.9 i 2.5 pM for Gal4-cERdef and Gal4-mEROtdef, respectively. However, only 40% maximal response was observed following incubation with cells transiently transfected with Gal4-hEROtdef and o,p’-DDT was classified as a non-inducer for the other receptors (Figure 6). Methoxychlor, p,p’-DDT, and atrazine failed to significantly induce reporter gene expression through any of the ERs. These data show that certain estrogenic compounds exhibit a differential ability to induce reporter gene activity mediated by both ER isoforms and ERs from different vertebrate species. Although differences in the ability of EEDs to induce gene expression were observed among the different vertebrate ERs, the most dramatic differences were observed with the phytoestrogens interacting with ERB. 99 -\\ 200 125 Gal4-hEROtdef Gal4-mERBdef 150-I 100‘ 100d 50- 0i ‘ d an. O ‘D fn-t V m. m. C l 125 150 Gal4-cERdef ° W _Gal4-aERdef ° 125 125 ‘°°i Gal4-xERdef 1001 Ga14-"ERadef . 75d % Luciferase activity relative the highest dose of E2 50‘ I 251 Log [concentration] Figure 5. Mycotoxins and phytoestrogens differ in their ability to induce gene expression mediated by both ER isofonns and ERs from different vertebrate species. Comparison of the ability of a—zearalenol (O ), coumestrol ( I), genistein (O), and naringenin ( CI) to induce gene expression mediated by Gal4-hEROtdef, Gal4-mERBdef, Gal4-cERdef, Gal4-aERdef, Gal4-xERdef, or Ga14-rtEROtdef. MCF—7 human breast cancer cells were transiently transfected with 0.2 to 0.5 pg of Gal4-ER chimeras, 1.5 pg of l7m5-G-Luc, and 0.1 pg of pCMV as described in the methods section. Cells transfected with Gal4-rtER were incubated for 24 h at 20°C after dosing with test compound. Cells transfected with the other Gal4- ER chimeric receptors were incubated for 24 h at 37°C. 100 % Luciferase activity relative the highest dose of E2 125 100‘ 75- 50‘ 25. Gal4-hEROtdef OJ 125 100- 75- Gal4-cERdef 125 -1-10 -9 -a -7 -6'-5 -4 Gal4-mERBdef Gal4-aERdef 1O 9 8 125 100‘ Gal4-xERdef 100- Gal4-rtEROtdef 75-l 75‘ 50d 50‘ 25-1 25. (pm 9 , , , T . , -1O -9 -8 -7 -6 -5 ~10 -9 8 -7 6 -5 Log [concentration] Figure 6. Xenobiotics differ in their ability to induce gene expression mediated by both ER isoforms and ERs from different vertebrate species. Comparison of the ability of 4-t-octylphenol (O), Bisphenol A (l), o, p ’ -DDT (O ), and rmthoxychlor ( ) to induce gene expression mediated by Gal4-hEROLdef, Gal4- mERBdef, Gal4-cERdef, Gal4-aERdef, Gal4-xERdef, or Gal4-rtEROtdef. MCF—7 human breast cancer cells transiently transfected with Gal4-ER chimeras described in the methods section. Cells transfected with Gal4-rtEROt were incubated for 24 h at 20°C after dosing with test compound. Cells transfected with the other Gal4-ER chimeric receptors were incubated for 24 h at 37°C. 101 Discussion In chapter 4, ligand binding assays were used to examine the ability of several EEDs to bind to hERa, mEROt, cER, aER, and rtERor. Several differences in relative and absolute binding affinities were observed, with the rtEROL, which has the greatest amino acid variation in its ligand binding domain relative to the other species examined, exhibiting the most striking differences. In radioligand binding assays only compounds capable of competing with radioligand are detected. Also, these assays do not distinguish between agonists and antagonists. Cell based transcription assays using a reporter gene under the transcriptional control of a specific receptor offer an alternative to radioligand binding assays and provide valuable information about the ability of test compounds to induce transcriptional responses. In order to focus on differences in transactivation ability among different vertebrate ERs that may reside within their respective ligand binding domains, Gal4-ERdef fusion proteins consisting of the D, E, and F domains of each receptor were constructed. This ensured that effects imposed by other domains, such as the DNA binding domain would not influence the results (26) and also allowed for comparison with the competitive binding data presented in chapter 4 where similar GST-ERdef fusion proteins were used. The Gal4-ERdef fusion proteins exhibited similar E2 transactivation ability when compared to full-length ERs (27,28). Although the rtERa exhibited reduced transactivation ability at 37°C, once the incubation temperature was reduced to 20°C, the E2-induced transactivation response agreed with previous studies in yeast transiently transfected with full-length rtERor (29). This is consistent with another study investigating the transactivation ability of the Oreochromis gens 102 Tl EROL (OaEROL), which also exhibits reduced activity at temperatures above its normal physiological range (30). There were instances where the ligand binding data did not support the gene expression results. For example, 4-t-octylphenol preferentially bound to GST-rtEROtdef with lO-fold greater affinity than to the GST-aERdef, but was able to induce reporter gene expression at approximately 10-fold lower concentrations through aER than through rtERor. In addition, some compounds found to compete for binding to a specific ER were unable to induced gene expression mediated by the ER from the same species. For example, o,p’-DDT bound to GST-rtEROtdef but did not elicit any reporter gene activity mediated by Gal4-rtEROtdef. In many other instances the gene expression data support the competitive binding studies presented in chapter 4. The mycotoxin, (it-zearalenol, which bound with greater affinity than E2 to GST-rtERocdef, also induced reporter gene expression mediated by Gal4-rtEROtdef at lower concentrations than E2. Similarly, coumestrol which bound to GST-aERdef with higher affinity than to the other ERs, also induced gene expression mediated by Gal4-aERdef at lower concentrations compared to the Gal4-hEROtdef, - mERadef, -cERdef, and -rtEROLdef. A plot of the log transformed ICso values (Chapter 4, table 2, page 77) vs ECso values (Table 1 page 101) for hERa, mERor, cER, aER, and rtERor shown in figure 7 reveals a linear correlation with a Pearson r value of 0.86. Linear regression analysis revealed an overall slope of 0.82, indicating a strong correlation between ligand binding and transactivation ability. The half-maximal responses of E2 and DES in the gene expression assay were approximately IO—fold lower than that of the ligand binding assay. In contrast the half- 103 Log IC50 (M) -10.. '11 I I I I I I -11 -1o -9 -8 -7 -6 -5 -4 Log EC50 (M) Figure 7. Comparison of IC50 and EC50 values among different vertebrate ERs. Correlation analysis of the log transformed ICso values (Chapter 4, Table 2, page 76) vs EC50 values (Table 1, page 101) for hER(1( o) , mERor (O) , cER ( l), aER (El), and rtERa ( 0) reveals an overall linear correlation with a Pearson r value of 0.86. 104 maximal responses to ligand for the progesterone receptor (PR) and the glucocorticoid receptor (GR) reflect their affinity for that ligand (31,32). Unlike the GR and PR, the ER is capable of binding DNA in the absence of ligand (33) which may be a contributing factor to the hormone-independent activity of the receptor and as been proposed has an explanation for the discrepancy in half-maximal response between gene expression and ligand binding assays (27). DNA binding and subsequent recruitment of coactivators E may represent a rate limiting process in the activation of NRs. Since the ER can exist in a “prebound state” to DNA, essentially overcoming the rate-limiting step, maximal transactivational activity may be achieved at lower concentrations (27). Gel shift assays L using the Gal4-ER constructs have not been performed to support this hypothesis. The phytoestrogens exhibited an increased ability to induce gene expression mediated by Gal4-mERBdef, which was consistent with previous competitive binding studies where genistein exhibited a 30-fold greater affinity for ERB compared to ERor (1). These previous studies reported only a 3-fold lower ECso value for ERB compared to ERor (1). This is far lower than the 95-fold (genistein) and 164-fold (coumestrol) lower EC50 values for mERB compared to hERor reported in the present study (Table 1). These discrepancies may be due to differences in the cell—based assay conditions, since the previous studies were done in human embryonic kidney 293 cells transiently transfected with full length hERa and hERB and an ERE driven reporter gene (1). BEDS have also been shown to differentially recruit the coactivators, SRC-l and SRC-2 to EROt and ERB, in which both full-length and ligand-binding domain truncations were used (16). The HO-PCB, 2’,3’,4’,5-tetrachlorobiphenyl, exhibited a similar RBA for both EROL and ERB. Treatment of this compound, however, failed to recruit SRC-1 or 105 SRC-2 to EROt but resulted in complete recruitment of SRC-l and submaximal recruitment of SRC-2 to ERB. The 30-fold greater binding affinity of genistein for ERB resulted in a 12 000- and 33-fold greater ability of ERB to recruit SRC-l and SRC-2, respectively (16), which supports the increased ability of genistein to induce ERB- mediated reporter gene activity (Table I, Figure 5). However, x-ray crystallographic data of the genistein-ERB complex reveals that the critical helix 12, which contains the AF-2 region is positioned intermediate between agonist and antagonist conformations suggesting that the complex may not adequately interact with coactivators. The differential expression levels of coactivators in cell lines combined with the observation that SRC family of coactivators interact with NRs via both the AF-l and AF-2 regions may also account for some of the discrepancies observed among different studies. Take together these results demonstrate that ERs from different vertebrate species exhibit differential transactivation profiles in response to a number of EEDs. The ability of E2 and other estrogenic chemicals to induce gene expression mediated by the rtEROtdef is influenced by temperature. In general the results from the ligand binding studies presented in chapter 4 support the transactivation data present here, though there were instances where the two assays did not correlate. Pharmacokinetic and pharmacodynamic differences among species make it difficult to predict whether the differences in functional activities among ERs would be observed in vivo. 106 10. 11. 12. 13. 14. 15. References Kuiper, G. G., Lemmen, J. G., Carlsson, B., Corton, J. C., Safe, S. H., van der Saag, P. T., van der Burg, B., and Gustafsson, J. A. (1998) Endocrinology 139, 4252-4263. Fitzpatrik, D. W., Picken, C. A., Murphy, L. C., and Buhr, M. M. (1989) Comp Biochem Physiol 94C, 691-694. Strunck, E., Stemmann, N., Hopert, A., Wunsche, W., Frank, K., and Vollmer, G. (2000) J Steroid Biochem Mol Biol 74, 73-81. Glass, C. K., and Rosenfeld, M. G. (2000) Genes Dev 14, 121-141. Glass, C. K., Rose, D. W., and Rosenfeld, M. G. (1997) Curr Opin Cell Biol 9, 222-232. Katzenellenbogen, B. S. (1996) Biol Reprod 54, 287-293. Onate, S. A., Boonyaratanakomkit, V., Spencer, T.E., Tsai, S.Y., Tsai, M.J., Edwards, D.P., O'Malley, B.W. (1998) J Biol Chem 273, 12101-12108. Halachmi, S., Marden, E., Martin, G., MacKay, H., Abbondanza, C., Brown, M. (1994) Science 264, 1455-1458. Hong, H., Kohli, K., Garabedian, M., and Stallcup, M. (1997) Mol Cell Biol 17, 2735-2744. Voegel, J. J., Heine, M. J. S., Zechel, C., Chambon, P., and Gronemeyer, H. (1996) EMBO J. 1996, 3667-3675. Anzick, S. L., Kononen, J., Walker, R. L., Azorsa, D. O., Tanner, M. M., Guan, X. Y., Sauter, G., Kallioniemi, O. P., Trent, J. M., and Meltzer, P. S. (1997) Science 277, 965-968. Chen, H., Lin, R. J., Schiltz, R. L., Chakravarti, D., Nash, A., Nagy, L., Privalsky, M. L., Nakatani, Y., and Evans, R. M. (1997) Cell 90, 569-580. Takeshita, A., Cardona, G. R., Koibuchi, N., Suen, C. 8., and Chin, W. W. (1997) J Biol Chem 272, 27629-27634. Torchia, J ., Rose, D. W., Inostroza, J., Kamei, Y., Westin, 8., Glass, C. K., and Rosenfeld, M. G. (1997) Nature 387, 677-684. Li, H., Gomes, P. J ., and Chen, J. D. (1997) Proc Natl Acad Sci U S A 94, 8479- 8484. 107 16. I7. l8. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. Routledge, E. J ., White, R., Parker, M. G., and Sumpter, J. P. (2000) J Biol Chem 275, 35986-35993. Pike, A. C., Brzozowski, A. M., Hubbard, R. E., Bonn, T., Thorsell, A. G., Engstrom, O., Ljunggren, J., Gustafsson, J ., and Carlquist, M. (1999) Embo J 18, 4608-4618. Fertuck, K. C., Matthews, J. B., and Zacharewski, T. R. (2001) Toxicol Sci 59, 231-240. Gillesby, B., and Zacharewski, T. R. (1999) Breast Cancer Res Treat 56, 253-265 Balaguer, P., Joyeux, A., Denison, M.S., Vincent,R., Gillesby, B.E., Zacharewski, T. (1996) Can J Physiol Pharmacol 74, 216-222. Zacharewski, T. R., Meek, M.D., Clemons, J .H., Wu, Z.F., Fielden, M.F., Matthews, J .B. (1998) Toxicol Sci 46, 282-293. Sambrook, J ., Fristch, E.F., Maniatis, T. (1989) Molecular Cloning. A Laboratory Manual, Cold Spring Harbor Laboratory Press, New York. Brasier, A. R., Tate, J .E., Habener, J.F. (1989) Biotechniques 7, 1116-1122 Matthews, J. B., Celius, T., Halgren, R., and Zacharewski, T. R. (2000) J Steroid Biochem Mol Bio 74, 223-234. Matthews, J. B., and Zacharewski, T. R. (2000) Toxicol Sci 53, 326-339. Petit, F. G., Valotaire, Y., and Pakdel, F. (2000) Nucleic Acids Res 28, 2634-2642. Reese, J. C., and Katzenellenbogen, B. S. (1992) J Biol Chem 267, 9868-9873. Pakdel, F., Le Gac, F., Le Goff, P., and Valotaire, Y. (1990) Mol Cell Endocrinol 71, 195-204. Petit, F., Valotaire, Y., and Pakdel, F. (1995) Eur J Biochem 233, 584-592. Tan, N. S., Frecer, V., Lam, T. J ., and Ding, J. L. (1999) Biochim Biophys Acta 1452, 103-120. Giguere, V., Hollenberg, S. M., Rosenfeld, M. G., and Evans, R. M. (1986) Cell 46, 645-652. el-Ashry, D., Onate, S. A., Nordeen, S. K., and Edwards, D. P. (1989) Mol Endocrinol 3, 1545-1558. Tora, L., Mullick, A., Metzger, D., Ponglikitmongkol, M., Park, I., and Chambon, P. (1989) EMBO J 8, 1981-1986. 108 fibril CHAPTER 6 Reciprocal mutagenesis between human or (L349, M528) and rainbow trout or (M317, 1496) estrogen receptor residues demonstrates their importance in ligand binding and gene expression at different temperatures These data have been published in the following article: Matthews, J.B., Clemons, J.H., and Zacharewski, T.R. (2001) Reciprocal mutagenesis between human or (L349, M528) and rainbow trout (M317, 1496) estrogen receptor residues demonstrates their importance in ligand binding and gene expression at different temperatures. Mol Cell Endocrinol in press 109 Abstract Several fish proteins exhibit compromised function at temperatures outside of their normal physiological range. In this study, the effect of temperature on the ligand binding and the transactivation abilities of the rainbow trout estrogen receptor (rtER) and human estrogen receptor alpha (hEROt) were examined. Saturation analysis and gene expression assays, using GST-ER and Gal4-ER fusion proteins consisting of the D, E and F domains of human (hERadef) and rainbow trout (rtERdef) receptors, show that GST- rtERdef E2 binding affinity and transactivation ability decrease with increasing temperature. A comparison of the amino acid sequence differences between their ligand binding pockets identified two conservative amino acid residue substitutions in rtER (M317, 1496) and hERor (L349, M528). The effect of these substitutions on ligand binding and transactivation were examined by constructing reciprocal mutants, which effectively exchanged the binding pockets between rtER and hERa. The rtERdef M317L:I496M double mutant exhibited increased E2 binding affinity and transactivation ability at higher temperatures, and displayed hERor phenotypic behavior for the phytoestrogen, coumestrol. The hERordef L349M:M528I double mutant also exhibited a modest trend towards adopting the rtER phenotype. These studies demonstrate that conservative changes in residue hydrophobicity and volume can significantly affect ER ligand binding and transactivation ability in a temperature-dependent manner. The lack of a complete exchange of phenotypes between rtER and hERa indicates that factors outside of the ligand-binding pocket are also involved. 110 Introduction The estrogen receptor (ER) is a member of the nuclear receptor superfamily, a family of ligand-regulated transcription factors (1-3). As with other members of the nuclear receptor superfamily, the ER consists of several functional domains (A-F), with the DNA-binding domain (C domain) separating the NHz-terminal ligand-independent activation domain (A/B domains) and the COOH-terrninal region, which includes a hinge region (D domain), the ligand binding domain (E domain) and a variable F domain (1,4). Two ER subtypes exist, ERa and ERB, each is encoded from a distinct gene and display tissue-specific distribution and ligand preference (5,6). Although the mechanism of action of estrogen and the domain organization of the ER is well conserved among species, several studies have reported significant differences in transactivation ability and ligand preference, especially between human and rainbow trout (Oncorhynchus mykiss) ERs (7-10). A comparison of the human and rainbow trout ERs indicates that amino acid sequence identity within domains can vary significantly between the two species. The most conserved region is the DNA-binding domain (92%). The NHz-terminal A/B domains are only 20% identical, while the E domain, which contains the horrnone- binding pocket, is 60% identical in amino acid sequence (11,12). Despite the significant sequence identity in some domains, functional differences in response to 17B-estradiol (E2) and other estrogenic chemicals have been reported. For example, when expressed in yeast the rtER exhibits a lO-fold lower responsiveness to E2 and a significantly reduced affinity for E2 at elevated temperature when compared to hEROL (10). Weaker E2 responsiveness of rtER has also been observed in embryonic salmonid cells (STE-137) lll transfected with rtER (9). In addition, functional differences in response to other hormones and antihormones have been reported (10). The molecular basis for the reduced rtER function at elevated temperature is unclear, although other fish proteins have been-shown to exhibit optimal activity at lower temperatures when compared to their mammalian orthologs (13-15). Domain interchange and the generation of chimeric receptors demonstrated that the rtER (hERa C domain) chimeric receptor had greater DNA binding affinity, while the rtER (hER E domain) chimeric receptor exhibited an increased E2 responsive compared to wild-type rtER (16). These results suggest that some of the functional differences between hEROt and rtER reside within the DNA-binding domain. However, competitive binding studies using GST-hERor and GST-rtER fusion proteins consisting of only the D, E and F domains have also identified several differences in relative binding affinities of estrogenic chemicals, supporting the hypothesis that the ligand binding domain significantly contributes to the reported differences between hERor and rtER (8). Sequence alignment using crystallographic data of the hERor-E2 complex identified two residues (17), L349 and M528 in hERa and the corresponding residues, M317 and 1496 in rtER, that differ between their respective ligand binding pockets. In the present study we examined the role these residues play in ligand binding and transactivation by preparing reciprocal mutants, i.e. hERa (L349M, M528I) and rtER (M317L, I496M), which effectively exchanged the ligand binding pockets of hEROL and rtER. Our results show that conservative amino acid residue substitutions within the ligand-binding pocket significantly affect ligand binding and transactivation ability. In addition, these residues also contribute to the reduced binding affinity and transactivation 112 of the rtER at higher temperatures. Interestingly, reciprocal mutagenesis did not result in a complete exchange of phenotypes, suggesting that residues outside of the ligand- binding pocket also influence ligand binding and transactivation ability. 113 Materials and Methods Chemicals 17B-estradiol (l ,3,5[ lO]-estratriene—3, l7B-diol) and a-zearalenol (2,4-dihydroxy- 6-[60t,10-dihydroxy-undecyl]benzoic acid p-lactone) were purchased from Sigma (St. Louis, MO). Coumestrol (2-(2,4-dihydroxyphenyl)-6-hydroxy-3-benzofurancarboxylic acid lactone) was obtained from Acros Organics (Pittsburgh, PA). 4-t-octy1phenol, was purchased from Aldrich (Milwaukee, WI). Radiolabeled [2,4,6,7,16,17-3H] l7B-estradiol ([3H]E2; 123 Ci/mmol) was from New England Nuclear (Boston, MA). pfu DNA polymerase was purchased from Stratagene (La Jolla, CA), and restriction enzymes were obtained from Roche/Boehringer Mannheim (Indianapolis, IN). Phenol red-free Dulbecco’s modified Eagle’s medium (DMEM) was from Life Technologies (Grand Island, NY). Fetal bovine serum (FBS) and D-luciferin were purchased from Intergen (Purchase, NY) and Molecular Probes (Eugene, OR), respectively. All other chemicals and biochemicals were of the highest quality available from commercial sources. Plasmid construction The plasmids pGEX-hERadef, and pGEX-rtERdef harboring the human and rainbow trout ER D, E, and F domains were constructed as previously described (7). The pG4M-hEROtdef (Gal4-HEGO) was a gift from Dr. P Chambon (IGBMC CNRS-LGME, Illkirch Cedex C.U. de Strasbourg, France). The pG4M-rtERdef plasmid was constructed by PCR amplifying amino acid residues 214-576 of the rainbow trout ER. The fragment was digested with HindIII and ClaI and ligated into the similarly digested eukaryotic expression vector containing the DNA binding domain of the yeast transcription factor 114 Gal4, pG4MpolyII. PCR amplification was performed essentially as previously described (18). All mutants were prepared by PCR based mutagenesis strategy, using the QuikChange Site-Directed Mutagenesis Kit (Stratagene, La Jolla, CA). The oligonucleotides used in the mutagenesis were: 5-tactgaccaaca_tggcagacagg—3and 5- cctgtctgccatgttggtcagta-3 for hERor L349M (changes Leu349 to Met), 5- catctgtacagcgtgaagtgcaagaac-3 and 5-gttcttgcactttz;tgctgtacagatg-3 for hERor M5281 (changes Met528 to Ile), 5-gctcaccagcc_tggctgacaagg-3 and 5-ccttgtcagcg1ggctggtgagc-3 for rtER M317L (changes Met317 to Leu), and 5- cacctttacagca_tgaaatgtaagaac-3 and 5- ttcttacatttc_atgctgtaaaggtg-3 for rtER I496M (changes Ile496 to Met). The mutated residues are underlined. The following PCR conditions were used for the mutagenesis: 95°C for 5 min followed by 25 cycles of 95°C for 1 min, 50°C for 1 min and 72°C for 10 min. The sequence of each construct was confirmed by restriction enzyme digest and ABI/Prism automated sequencing (Perkin Elmer Applied Biosystems; Foster City, CA). Expression and purification of GST ER fusion protein Expression and purification of GST-ERdef fusion proteins was performed as previously described (7,8). Briefly, E. coli strain BL21 (Amersham/Pharmacia) containing pGEX-ERdef constructs were incubated at 37°C with constant shaking. Cells were induced with 1 mM IPTG at an optical density of 0.8 at 600 nm and grown for 4 h at 30°C. After centrifugation, cell pellets were resuspended in 25 ml of buffer A (50 mM HEPES, 3 mM EDTA, 5 mM DTT, 50 mM NaCl and 10% (v/v) glycerol, pH 7.5) containing 0.1 mg/ml lysozyme, and protease inhibitors. Cells were then lysed by 115 sonication on ice and cell debris was pelleted by centrifugation at 20,000xg for 40 min at 4°C. Supematants were stored at -80°C until further use. The supematants containing the GST fusion proteins were applied to GSH Sepharose pre-equilibrated with buffer A at 4°C. After adsorption of the protein, the GSH Sepharose was washed with 100 ml of buffer B (50 mM HEPES, 3 mM EDTA, 5 mM DTT, 150 mM NaCl and 10% (v/v) glycerol, pH 7.5). Bound proteins were eluted in 25 ml of buffer C (50 mM HEPES, 3 mM EDTA, 5 mM DTT, 150 mM NaCl and 10% (v/v) glycerol, pH 8.0) containing 10 mM GSH. The partially purified protein was concentrated to a 1 m1 final volume and the protein concentration was determined using the Bradford method (19). Protein was diluted to 0.5 mg/ml and stored at —80°C until further use. Receptor binding assays Partially purified GST-hERordef and GST-rtERdef fusion protein were diluted in TEGD buffer (10 mM Tris pH 7.6, 1.5 mM EDTA, 1 mM D'I'I‘ and 10% (v/v) glycerol) containing 1 mg/ml bovine serum albumin as a carrier protein. Receptor binding and competitive binding assays were performed as previously described (8). Briefly, GST- ERdef fusion proteins were diluted in TEGD buffer (10 mM Tris pH 7.6, 1.5 mM EDTA, 1 mM DTT and 10% (v/v) glycerol) containing 1 mg/ml bovine serum albumin (BSA), and incubated at 37°C for 30 min, 30°C for 1 hr, 20°C for 2 h, and 4°C for 24 h with 0.1- 3.5 nM [3H]E2 in 1 ml glass tubes arranged in a 96-well format (Marsh Scientific, Rochester, NY). Fusion protein preparations were diluted to ensure 10,000 dpms of total binding (varied from 1000-2000 fold). Binding assays were initiated by adding 240 pl of 116 protein preparation to glass tubes containing 5 pl of DMSO and 5 p1 [3H]E2, thus the concentration of solvent did not exceed 4 %. Bound [3H]E2 was separated from free using a 96-well filter plate and vacuum pump harvester (Packard Instruments). After drying, the filter plates were sealed and 50 pl of MicroScint 20 scintillation cocktail (Packard Instruments) was added to each well. Bound [3H]E2 was measured using a TopCount luminescense and scintillation counter (Packard Instruments). Competitive ligand binding assays were performed as described above except diluted GST-ERdef fusion protein preparations were incubated with a final concentration of 2.5 nM [3H]E2 (5 pl aliquot) and increasing final concentrations of unlabeled competitor (0.1 nM — 100 pM, 5 pl aliquots) at 4°C for 24 h. Each treatment was performed in quadruplicate and results are expressed as percent specific binding of [3H]E2 versus log of competitor concentration. ICso values were determined from nonlinear regression for single site competitive binding analysis. The reported ICso values represent the concentration of test compound required to displace 50% [3HIE2 from the GST-ERdef fusion proteins as compared to the 50% displacement of [3H]E2 achieved by unlabeled E2. Analyses were performed using GraphPad Prism software (GraphPad Software Inc., San Diego, CA). Cell Culture and transient transfection Estrogen receptor positive MCF-7 human breast cancer cells (obtained from Dr. L. Murphy, University of Manitoba, at passage 32-35) were maintained with phenol red- free DMEM supplemented with 10% FBS, 2 mM L—glutamine, 10 mM HEPES, 500 pg/ml gentamicin, 2.5 pg/ml amphotericin B, 100 IU/ml penicillin G, and 100 pg/ml 117 streptomycin. Cells were cultured in a humidified environment at 37°C in the presence of 5% C02 atmosphere. Transient transfections and gene transcription assays were performed essentially as previously described (20,21). Briefly, MCF-7 cells were seeded at approximately 50% confluency in 6-well tissue culture plates in medium supplemented with 5% dextran- coated charcoal-treated fetal bovine serum (DCC-FBS) and allowed to settle for 7 h. For the preparation of whole cell extracts used in western blot analysis, MCF-7 cells were transiently transfected by the calcium phosphate precipitation method (22) with 5 pg of Gal4-ER expression vector and 0.05 pg of pCMV-lacZ, while for the reporter gene assay, cells were transiently transfected with 1.5 pg of l7m5-G-Luc (provided by Dr. P Chambon), and 0.2 pg of Gal4-hEROtdef or 0.2 pg Gal4-rtERdef along with 0.05 pg of pCMV-lacZ. Cells were washed 16 h later with sterile phosphate buffered saline and fresh medium was added to each well. Cells were then treated with E2, or test compound and allowed to incubate for 24 h. Following incubation cells were harvested and assayed for luciferase activity as described previously (23). Each treatment was done in duplicate and two samples were taken from each replicate and each experiment was repeated three times. The concentration required to induce half maximal luciferase activity (ECSO value) was calculated using GraphPad 3.0. Western Blotting Whole cell extracts from transfected MCF—7 cells were fractionated on a 10% polyacrylamide/SDS gel and transferred to nitrocellulose (Amersham/Pharmacia, Piscataway, NJ). Blots were incubated with anti-Gal4 DNA binding antibody (Santa 118 Cruz, Santa Cruz, CA) for 2 h at room temperature in the presence of 2% fat-free milk powder and then incubated for 2 h with the secondary conjugated goat anti—(mouse IgG)- horseradish peroxidase (Santa Cruz). Proteins were visualized using ECL detection system (Amersham/Pharmacia) and subsequent exposure to film. Statistical Analysis Comparisons of ECso and ICso values between wild type and mutant proteins (both Gal4 ER chimeras and GST fusion proteins) were performed with a one-way ANOVA using the GLM procedure of SAS version 7 (SAS Inc, Cary, NC). Multiple i comparisons between wild type (control) and mutant forms of the ER within species were adjusted by Dunnett's method. Comparisons between species were adjusted by the Tukey method. The level of significance was 5 %. 119 Results Comparison of the hormone binding properties of GST-hERocdef and GST-rtERdef wild- type and mutant receptors at different temperatures Sequence alignment identified amino acid residues that line the ligand binding pockets of hERoc and rtER and differ between the two receptors i.e. L349 and M528 in hERa and the corresponding residues, M317 and I496 in rtER (Figure 1). Reciprocal mutants were prepared which effectively exchanged the ligand binding pockets of hERor and rtER. GST-ERdef fusion proteins used in this study exhibit similar purity as previously described (7 ,8). Mutagenesis did not affect the level of protein expression (data not shown). GST-ERdef fusion proteins were incubated with increasing concentrations of [3HIE2 and the binding affinity was measured in vitro at 4, 20, 30, and 37°C using the method of Scatchard (24) (Figure 2 and Table 1). The K, value of GST-hERordef wild- type (WT) for E2 was similar to published values (25-27) and was unaffected by temperature. Similarly, the K, values of the GST-hEROtdef mutants differed less than 2- fold when compared to GST-hERadef WT at the different temperatures examined. However, the affinities of the GST-hERordef L349M:M528I double mutant and the GST- hERordef M5281 single mutant were significantly lower than GST-hERocdef WT at 4°C. The K; value of GST-rtERdef WT was 2-fold lower compared to GST-hERordef WT at 4°C (Table 1), but comparable to published values (12,28). All three GST-rtERdef mutants displayed significantly higher affinity for 120 Ti 0088:. b80898: 083 85 80.0.80 28 05:8 0... 8.00000. .00. 8w8. 2.... i 8 5.3 .80....000. 08 38.000 0300800.... 8.88 £0.23 .8028: 0.0.3 .* 8 5.3 00.7.3500. o8 .0..80.8-n\t. 5.3 w:..0:00 :owoceb. :. 00202.. 800.8”. .200 :. 08.8.23: 08 88:80 m 2.... .8980 8 0300.0 08 80.0.8: $08.8 0.0....» 80.0 8 08:08:00: 08 800.8. 0.8 00.08 80:80. 00:00.08 530...... o... :. 00.0.80 0.08 0:08 0. 8.0.. 800087. évamvmz .0: 00.8008 0.5586 ”.8355 .000 300:8: E8 €3.34 .0: 5.0808 0.8m000 ”.0335... .882. 60:. 88800 .0 .08 m. d 00.88.. 0.8008 2.. ..0 825%.? .. 00%.". a 0m mHngmmmmmwog. mgmm . SSOWOMZQOH . . 09mg. . BBB . QONBOWO . 0 m N m m a m IIIIIIIII EdmhiOgOBH;I I i ixoammmmfimmfimfiémgfimgw h m m «mm «EOOQOCOOOOICOCOO...OOK.O.CH...O0.0.C.0......0.....a.:...§.ag§000‘000 omv m m m DamsaufigdgflgflgggsaggHgH mm v e e at 00 mvv ICCEgawEOCEBOU. u:.......‘.......a. §OCCOO>COCOCCCCC‘CCOOCGC BCCCHCG mbm Hmv .. . hov fibm GOOHOOOOOIHOICHOOOOO>000qu00°...00“."00O00K.O.‘OOOOOOKCCEOOOBOOEOQOOE oom mov gagHgsgggghgfl Nmm 0 Coco co fleece e mam 9.6.4. . . .oo.§.umgnx.mo§020 . 802000000890000088 00.300». 0:002. mam 0000000 0 m m Euqnmggaang ttttt 050880200m0302<<§0000000000mom§x§t t e 0 m 0 00000: 121 0.20‘ 0.3‘ 0.15' Bound/Free 0-2‘ 0.104 0.1 -‘ 0.05 0.0 W i I I 0.” 3 y t 0.00 0.05 0.10 0.15 0.20 0-25 0.00 0.05 0.10 0.15 0.20 0.25 Bound (nM) Figure 2. The affinity of wild-type and mutant receptors for 173- estradiol at different temperatures. Partially purified (A) GST-hERadef wild-type or (B) GST-rtERdef wild-type were incubated with increasing concentrations of [3H]17B-estradiol at 4°C for 24 h (0), 20°C for 2 h (0), 30°C for l h (I) or 37°C for 30 min ([3). Affinities of the receptors for E2 were determined by the method of Scatchard {Scatchard, 1949 #209}. 122 Table 1. Hormone binding of wild-type and mutant ERs at different temperatures. GST- ERdef fusion proteins were incubated with increasing concentrations of [3H]17B-estradiol at 4, 20, 30, or 37°C for 24 h, 2 h, 1 h, or 30 min, respectively. Affinities are expressed as equilibrium dissociation constants (K.,), and represent the mean and standard deviation of three determinations. Average K. :t S.D. (nM) Protein Preparation 4°C 20°C 30°C 37°C GST-hERouief wild-type 0.50 :t 0.15 0.50 :l: 0.10 0.74 :l: 0.12 0.57 3; 0.02 L349M 0.52 :t 0.04 0.49 :t 0.07 0.60 i 0.10 0.64 i 0.09 M5281 0.92 :t 0.08‘ 0.49 i 0.13 0.66 i 0.04b 0.76 :t 0.07b L349MzM528I 0.91 1 0.26‘ 0.92 :l: 0.15" 0.98 i 0.058 0.90.: 0.203 GST-rtERdef wild-type 1.03 :t 0.14 1.82 :l: 0.21 4.58 i 0.64d 5.37 :t 1.25d M3171. 0.56 i 004° 1.39 :1: 0.63 2.27 :1: 0.208d 3.31 a: 069°d I496M 0.82 i 0.11° 1.00 i 005° 1.47 i 0.13°-d 1.60 :l: 0.1 1°-d M317L:I496M 0.76 i 008° 0.87 :t 018° 1.53 i 0.01°~cl 1.82 3: 003°°| a P < 0.05 vs GST-hERocdef wild-type within each temperature b P < 0.05 vs corresponding GST-hERadef wild-type or mutant protein at 4°C c P < 0.05 vs GST-rtERdef wild-type within each temperature d P < 0.05 vs corresponding GST-rtERdef wild-type or mutant protein at 4°C All data analyses were performed using SAS version 7 (SAS Inc, Cary, NC). Kd values were analyzed by a one-way analysis of variance (ANOVA), using the GLM procedure of SAS, with either mutation or temperature as a fixed effect. Pair-wise comparisons between groups and the control (wild-type protein or 4°C) were performed using the Dunnett's test. The level of significance was P < 0.05. 123 E2 than GST-rtERdef WT. The K., values of GST-rtERdef WT and mutant receptors significantly increased with temperature (Table 1). 'For GST-rtERdef WT, the K., value was 5.3-fold higher at 37°C compared to 4°C. The K., values increased 3.4-fold with the I496M mutation, 1.6-fold with the M317L mutant and less than 1.8-fold for the GST- rtERdef M317L:I496M double mutant at 37°C compared to WT at 4°C. The number of binding sites in GST-hERocdef WT remained constant among the different temperatures (Figure 2). In contrast, GST-rtER WT exhibited dramatic reductions in the number of binding sites at elevated temperatures (Figure 2). A reduction in the number of binding sites at elevated temperature was also seen with the rtER mutants, while only slight reductions were observed with the hERa mutants (data not shown). Gal4-hERozdef and Gal4-rtERdef wild-type and mutant receptor mediated reporter gene activity at different temperatures WT and mutant Gal4-ERdef mediated luciferase reporter gene induction was examined at 20, 30, and 37°C for 24 h following treatment with E2. The morphology of the cells was similar at the various temperatures examined. At 37°C cells transfected with Gal4-rtERdef required a 280—fold greater concentration of E2 to achieve half- maximal response compared to Gal4-hEROtdef at the same temperature (Figure 3). This difference in transactivation was reduced to 33-fold and to less than lO-fold when transfected MCF-7 cells were maintained at 30 and 20°C, respectively. This was caused by a shift in the dose-response curve for the rtER, since the E2-induced half-maximal response for hERa changed less than 2.5-fold among the three different temperatures examined. 124 37°C 100% q, 75- VJ t: a 50~ (I) 32 25° N m 0‘ I I I Tés‘ -12 -11 -10 -9 -8 -7 -s g 125 O 5 100- 3°C a .“a’ 75‘ 8 "as 50‘ 04 5‘ 25' “E OJ I I I I I I I <1: -12 -11 -10 -9 -a -7 -0 Q) U) <3 8 5:: Q 3 .4 0° 32 -1'1 30 :9 :8 -7 -6 Log [concentration] Figure 3. Temperature effects on wild-type human and rainbow trout estrogen receptor-mediated induction of reporter gene activity in the presence of 173- estradiol. MCF—7 cells were cotransfected with the reporter plasmid 17m5-G- Luc and the Gal4-ERdef expression vector for hEROt (O) or rtER (O) for 16 h at 37°C. Following the incubation, cells were rinsed with phosphate buffered saline, fresh media was added and cells were incubated at 20, 30 or 37°C for 24 h in the presence of increasing concentrations of 17 B-estradiol. The results shown are from a representative experiment that was repeated two times. 125 The maximal E2 induction ranged from 20- to 30-fold for Gal4-hERocdef and 50- to 100-fold for Gal4-rtERdef at both 30 and 37°C, but decreased to 4- to 7-fold at 20°C for both receptors. Background luciferase activity for both receptors increased at 20°C, and therefore the data are reported as percent luciferase activity relative to the maximal activity achieved by E2 with each receptor, which was set to 100% for comparative purposes. No changes in induction of luciferase activity were evident with the Gal4- hEROLdef WT and single mutations at 20, 30 or 37°C (Figure 4a). In contrast, the ECso value of the Gal4-hERocdef L349M:M528I double mutant was significantly (p<0.0001) increased 20-fold at 37°C (Figure 4a and 5a), but was only slightly increased at 20 and 30°C, indicating that induction of reporter gene activity by the double mutant was dramatically affected by temperature. At 37°C, Gal4-rtERdef M317L:I496M-mediated reporter gene expression increased S-fold compared to Gal4-rtERdef WT and was only 60-fold lower than that of Gal4-hERadef WT (Figure 4b and 5b). The difference in transactivation of Gal4-rtERdef M317L:I496M compared to Gal4-hERoodef WT was reduced to 8- and 3-fold at 30 and 20°C, respectively. Interestingly, the transactivation profiles of Gal4-rtERdef M317L:I496M at 20°C and Gal4-hEROtdef WT at 37°C were superimposable (Figure 5b), and their respective ECso values were not significantly different (p>0.05). Both Gal4-rtERdef M317L and I496M mutations contributed to the decreased transcriptional activity, with the M317L mutation having a greater influence than the I496M mutation (Figure 4b). This is in contrast to the effect on GST-rtERdef binding where the I496M mutation had a greater affinity for E2 compared to the M317L mutant at elevated temperatures (Table 1). 126 E. "'3’ 4c 2'0“ -0-hWT -0-rtWT E 1““ +M528l -I-|496M v -t:1-hL349M:M528I -D-rtM317L:l496M 9, 1.0- 20- 8 0.5- 0.0 . g 20 30 37 20 30 37 Temperature (°C) Figure 4. Temperature effects on wild—type and mutant estrogen receptor- mediated 17B-estradiol induced reporter gene activity on EC50 values. MCF—7 cells were cotransfected with (A) Gal4-hEROtdef wild type (0), L349M (O), M5281(I), or L349M2M5281 (III) or (B) Gal4-rtERdef wild type (0), M317L (O), I496M (I), or M31721496M ([3). Following transfection and dosing, cells were incubated for 24 h at 37°C, 30°C or 20°C. Cells were harvested and luciferase and B-galactosidase activity were determined. Though a complete reverse of phenotypes was not seen, the Gal4-hER01def L349M2M528I double mutant adopted a temperature sensitive phenotype, similar to Gal4—rtERdef WT. Also Gal4-rtERdef M317L:I496M double mutant displayed significantly reduced temperature sensitivity, which was primarily due to the M317L mutation. The results are from a representative experiment that was repeated 3 times. 127 1% m 37°C 8' 95$ T 42 41 4'0 -9 -a -7 3 4'2 41 40 :9 -s -7 -6 30°C 53‘ § % Luciferase Activity Relative to Maximal E2 Response 9 Q 3 4' 4'2 31 4'0 5 :8 :7 :5 42 4'1 40 39 :8 51 :6 125 125 100- 20°C 100- 20°C a 75- 75- 50- so- 35‘ 254 0‘ 0- 0-4" 42 4'1 40 :9 30 51 :6 42 4'1 4'0 39 :8 51 3 Log [concentration] Figure 5. Temperature effects on wild-type and mutant estrogen receptor-mediated 17B-estradiol induced reporter gene activity. MCF-7 cells were cotransfected with Gal4-hERadef wild-type (O), or Gal4-hER01def L349M:M528I (I) in panel A, and Gal4-rtERdef wild-type (O) or Gal4-rtERdef M317L:I496M (CI) in panel B, as described in Experimental Procedures. The Gal4-hERocdef (O) mediated-reporter gene activity at 37°C (dashed line) in panel B illustrates the improved transactivation ability of wild-type and mutant rtER at the three different temperatures. The results are from a representative experiment that was repeated two times. 128 The Gal4-rtERdef M317L21496M double mutant also displayed an increased transcriptional activity of approximately 4-fold compared to Gal4-rtERdef WT at each of the temperatures examined (Figure 4b and 5b). In contrast, the Gal4-hEROLdef L349M2M5281 displayed a decreased transcriptional activity at the three different temperatures (Figure 4a and Figure 5a), as would be expected from the reciprocal mutagenesis. These data suggest that residues L349/M317 and M528/I496 play important roles in the observed differences in transactivation between hERa and rtER, as well as in influencing the decreased transcriptional activity of the rtER at different temperatures. Western blot analysis Western blot analysis was performed to assess the expression level of the WT and mutant Gal4-ERdef proteins. At the three temperatures examined, both Gal4-hEROLdef and Gal4-rtERdef WT and mutant proteins were expressed at comparable levels in MCF- 7 cells (Figure 6). Therefore, the differences in transactivation observed between Gal4- rtERdef WT and Gal4-hERocdef WT were not due to differences in protein expression levels. However, expression levels for both constructs were always lower at 20°C than at 30 or 37°C, which was consistent with the decreased induction of luciferase activity. Interaction of estrogenic chemicals with wild-type and mutant receptors Previous competitive binding studies have shown that oc-zearalenol, coumestrol, and 4-t—octylphenol exhibit different relative binding affinities for GST-hERocdef WT compared to GST-rtERdef WT (8). Therefore, the ability of these 129 .025 330%.. 33 35 EoEcomxo 3:8:829: a Bob 8a 3:58 05. .3335 430-35“ 53, @8088 203 332.3 .3 35:03 2:805 wocmmté—nw we. hm can “08355430 «9. mm of. .Uocm “a @8385 £00 .8“. 8:32 we? Swan: 38: u>c 23.35 SW .5532 65: com 393— 203 Am: 0: £88m :39 mo 3:588 Raym— .m2=voooi RacoEtoaxm E engrave mm 35$-QO B @3833 use taxman 0.83 flombxm .EEQE =23“;me Mm mo w: m 5:» 3889.5“: 230 5452 Bob 3398a 803 £8be :8 32$» .32: 56835 yam—-430 E538 93 256:? Mo 39:98 83 5883 .c 2sz . , .. assuage 99.9 .9 29 929 a... er 9% er as; aw 9w 9 9... 9.9.. a 99., .9. 99, 99.9 9.. 99, 9% 2% 2%, 9 9.8 9.8 0% x V A 4 ..... E39330 ? 9 «I x/ 8 AV p. $7 p. AV 8 a7 ? A? 9 47 .05 er :2 as er .05 v. an. r .0». { 9V»? Aw AV 0?»? la AV 2721.0 a? 4&0 as». are @an. 130 compounds to compete with [3H]E2 for binding to the mutant receptors was examined in order to investigate the effect of the mutagenesis on the differential ligand binding behaviors of hERa and rtER. Although reciprocal mutagenesis had little effect on the relative binding affinity of on—zearalenol and 4—t-octylphenol (data not shown), GST- hERadef and GST-rtERdef double mutants experienced a nearly complete exchange of binding phenotypes for coumestrol (Table 2 and Figure 7). The ICso value for coumestrol was significantly (p<0.0001) lower (IO-fold) for GST-hERadef WT than for GST- rtERdef WT, while the GST-hERadef WT was not significantly different (p=0.1554) from GST—rtERdef M317L:I496M. Similarly, the ICso value for coumestrol was significantly (p<0.0003) increased (IO-fold) for the GST-hERadef L349M2M5281 double mutant compared to that of GST-hERadef WT, but was not significantly different compared to that of GST-rtERdef WT. The ability of coumestrol to induce reporter gene expression mediated by Gal4- ERdef WT and mutant receptors was investigated in order to assess whether the effects observed in competitive binding studies correlated with transactivation activities (Table 2 and Figure 8). Transcriptional activity was examined at the optimal temperatures for Gal4-hEROt (37°C) and Gal4-rtERdef (20°C). Similar to the competitive binding data, coumestrol-induced transcriptional activity significantly decreased for the Gal4-hEROtdef L349MzM5281 double mutant. An increase for the Gal4-rtERdef M317L:I496M double mutant was also observed (Table 2), but was not statistically significant (p=0.079). When the transactivation activity of coumestrol mediated by Gal4-rtERdef WT and mutant receptors was examined at 37°C, the same trends seen at 20°C were observed, except that 131 there was approximately a 10-fold reduction in ECso values for both WT and mutant proteins. 132 Table 2. Competitive binding and transactivation abilities of wild-type and mutant receptors in response to estrogenic chemicals. The ICso and ECSO values represent the mean and standard deviation of 3 and 2 independent experiments, respectively. E2 Coumestrol Protein Preparation ICso (M) ECso (M) ICso (M) ECso (M) GST-hERa wild-type 2.2 i 0.4 x 10“ 9.8 i 9.2 x 10'“ 1.2 i 0.1 x 10'7 6.1 i 3.8 x 10'7 L349M 2.4 i 0.2 x 10'9 4.1 :l: 2.5 x 10-10 7.9 :l: 1.5 x 10'“ 6.9 :l: 3.8 x 10'7 M5281 2.3 i 0.1 x 10‘9 3.4 :l: 0.3 x 10'10 3.0 :1: 0.7 x 10'7 7.1 :l: 3.4 x 10'7 L349M2M5281 3.4 i- 1.0 x 10'9‘ 1.9 :t 0.5 x109" 1.2 i 0.3 x 10'6“ 2.9 :t 0.5 x 10'6b GST-rtER wild-type 2.7 :l: 0.2 x 10'9 1.0 i 0.4 x 10'9 1.1 :t 0.1 x 10'6 5.1 i 2.1 x 10'7 M317L 2.5 1 0.1 x 10'9 2.8 :l: 0.4 x 10"“ 8.7 :l: 1.1 x 107C 8.8 :t 1.6 x 10'8 I496M 2.8 i 0.5 x 10'9 5.6 :1: 0.2 x 1o"° 4.0 21.5 x 107° 1.4 :1: 0.3 x 10'7 M317L:I496M 3.2 :l: 0.6 x 10'9 3.6 i 0.8 x 10'10“ 4.6 i 1.5 x 107C 7.2 i 2.2 x 10'8 “ P < 0.05 vs GST-hERadef wild-type b P < 0.05 vs Gal4-hERocdef wild-type C P < 0.05 vs GST-rtERdef wild-type d P < 0.05 vs cGal4-rtERdef wild—type All data analyses were performed using SAS version 7 (SAS Inc, Cary, NC). ICso or ECSO values were analyzed by a one-way analysis of variance (ANOVA), using the GLM procedure of SAS, with either mutation or temperature as a fixed effect. Pair-wise comparisons between groups and the control (wild-type protein) were performed using the Dunnett's test. The level of significance was P < 0.05. 133 125 125 E 3100- 1004 o m 75. 75. N a 50. 504 m 21.. g0 25- 25- o. M I 1 I -9 -8 -7 -6 -5 -4 -9 -8 -7 -6 -5 -4 Log [concentration] Figure 7. Effect of reciprocal mutagenesis on the competitive binding ability of coumestrol. Representative competitive binding curves of coumestrol with (A) GST-hERocdef wild type (0), L349M (O), M5281 (I), and L349M:M5281(D), or (B) GST-rtERdef wild type (0), M317L (O), I496M (U), and M317:I496M (D). The results are from a representative experiment that was repeated at least two times. 134 175T 150- 1254 moi 754 50-1 to Maximal E2 Response 25‘ % Luciferase Activity Relative Log [concentration] Figure 8. Transactivation abilities of wild-type and mutant receptors in response to coumestrol. MCF-7 cells were cotransfected with (A) Gal4-hERocdef wild-type (O) Gal4-hEROLdef L349M2M5281 (I), or (B) Gal4—rtERdef wild-type (O) Gal4- rtERdef M317lzl496M (D), as described in the Materials and Methods. Following transfection and dosing, cells were transfected with Gal4-hEROtdef were incubated for 24 h at 37°C, while cells transfected with rtERdef were incubated for 24 h at 20°C. The results are from a representative experiment that was repeated two times. 135 Discussion In order to focus on differences in ligand binding and transactivation ability between hEROL and rtER that may reside within their respective ligand binding domains, GST-ER and Gal4-ER fusion proteins consisting of the D, E, and F domains of each receptor were constructed. This ensured that effects imposed by other domains, such as the DNA binding domain would not influence the results (16). The GST-ERdef and Gal4-ERdef fusion proteins used in this study exhibited similar E2 affinity and transactivation ability when compared to full-length ERs (25-27). Saturation analysis confirmed that GST-rtERdef exhibited reduced binding affinity for E2 at elevated temperatures. Gal4-rtERdef transactivation ability was also significantly compromised, as greater E2 concentrations were required to induce a half- maximal response in transfected cells maintained at 37°C compared to 20°C. The results demonstrate that the rtERdef exhibits reduced E2 binding affinity and transactivation ability at elevated temperatures, both of which reflect the loss of protein secondary structure. In contrast, the binding affinity and the transactivation ability of hEROLdef were not affected by temperature, indicating that the hERadef exhibits greater stability across a wider range of temperatures. . The enhanced activity of rtER at lower temperatures has been previously observed in transfected fish and human cell lines (9) and, to a lesser extent, in yeast grown at 30°C (10). These results are not surprising since the optimal temperature for maintaining rainbow trout is 16 :1 °C. Similarly, the transactivation ability of another piscine ER, the Oreochromis aureus (OaER), also exhibits reduced activity at temperatures above its normal physiological range (29). Moreover, other fish proteins 136 Ti have also been shown to function more efficiently at lower temperatures compared to their mammalian counterparts (13-15). For example, incubation temperature was shown to influence cytochrome P450 activity in rainbow trout hepatocytes, with lower temperatures producing greater activity (13). Rapid inactivation of dogfish shark P450 enzyme activity in liver microsomes was observed at 38°C, while in rabbit microsomes rapid inactivation occurs at 48°C and the reaction is still linear at 43°C (15). Rainbow trout ATPase activity has also been shown to be more heat-sensitive than that of rabbit ATPase (14). Although a substantial amount of experimental evidence has been accumulated on protein thermal stability (sequence, structure, mutagenesis, and E thermodynamics), no single mechanism or motif has been identified that is responsible for protein thermal stabilization (30). Thermal stabilization is a combination of intrinsic (hydrogen bonds, salt bridges, and hydrophobic effects) and extrinsic characteristics (31). The hERa and rtER E domains share 60% sequence identity with only two residue differences, L349 and M528 in hERa and M317 and 1496 in rtER, in their respective ligand binding pockets (8). Results obtained with the rtERdef reciprocal mutants demonstrate that M317 and 1496 affect ligand binding and transactivation ability at different temperatures. In general, the introduction of human residues, M317L and I496M, resulted in the rtERdef mutant displaying the phenotype of hERadef. The rtER M317L:I496M double mutant exhibited increased E2 affinity and transactivation compared to rtERdef WT at all temperatures examined. In addition, increasing the incubation temperature had a less pronounced effect on reducing the E2 affinity of the rtERdef M317L:I496M double mutant. This increase in thermal stability may be a result of ligand—stabilized protein secondary structure since the rtERdef M317L:I496M double 137 mutant displayed a significantly higher affinity for E2 compared to WT. The hERadef L349M:M528I double mutant showed modest rtER phenotypic behavior and also exhibited significantly decreased E2 binding affinity and E2-induced transactivation ability at 37°C. Decreasing temperatures increased the transactivation ability of the hERadef L349M:M5281 double mutant, as was seen with rtERdef WT. Interestingly, exchanging a 35 amino acid stretch, (B3 region) between the OaER and the chicken ER (cER) partially rescued the thermal deficient transactivation of the OaER (29), confirming that reciprocal mutagenesis of residues L349, M528 in hERoz and M317, 1496 in rtER is important but not sufficient for phenotypic reciprocity. The B3 region, which encompasses amino acid residues Glu444 to Val478 and corresponds to helix 9 and half of helix 10 of hERa, is highly conserved among the different EROts, with only two amino acid residue differences between hERa and cER. However, only 46% of the residues are conserved in the OaER B3 region. Similarly, the rtER B3 region also shares low identity (49%) relative to hERa and cER, compared to 69% identity with the OaER B3 region (29). Although hERa and hERB isoforms only share 31% identity within the B3 region, there have been no reports of hERB displaying any temperature deficient phenotypes. This suggests that variability within the B3 region may not be sufficient for temperature sensitivity, though in certain contexts it can contribute to thermal stabilization (29). The OaER also contains Met304, analogous to Met317 in rtER, suggesting that it may also contribute to thermal instability. It is interesting to speculate that the combination of reciprocal mutagenesis with B3 region exchange as described by Tan and coworkers (29), could lead to complete reciprocity of the temperature deficient transactivation phenotype. 138 Two amino acid substitutions also exist within the hERa and hERB binding pockets, Leu384 and Met421 in ERa are replaced by Met336 and Ile373 in ERB (17). Both receptors also share a similar degree (i.e. 58%) of sequence identity in their E domains (32) compared to hEROt and rtER. The residues used in the reciprocal mutants for the present study and the residue substitutions in the ERa and ERB binding pockets are considered to be conservative, but will affect the volume and hydrophobic character of the binding pocket (17,33). Moreover, the greater relative binding affinity of genistein for ERB compared to EROL has been attributed to these two residue differences (17,34). Although phenotype exchange was most evident in the double mutants, examination of single mutant data suggests that rtER I496M and hERa M5281 made greater contributions to the adopted binding phenotype while rtER M317L and hERa ’ L349M appeared to contribute more to the adopted transactivation abilities. This is consistent with the location of residue M528 in hERa, which occurs in a region (amino acids 515-535) that is important in the recognition of structurally diverse estrogens and antiestrogens (35). Existing crystal structures for the hERa E domain bound with E2 reveal that M528 interacts with the D-ring while L349 makes contact near the A-ring of the steroid (36,37). Other non-conservative mutations (i.e. C447A, K449E, M528A) in this region also resulted in ER phenotypes that exhibited compromised transactivation ability (38) and instability at elevated temperatures (27,39,40). Since the effect of the reciprocal mutagenesis was significantly more pronounced in rtER compared to hERa, factors outside of the binding pocket also influence E2 binding affinity and transactivation ability at different temperatures. 139 Mutations in hERadef that affected the relative binding affinity of coumestrol, also resulted in a corresponding change in its transactivation ability (Table 2). However, for selected ligands, the large volume of the ER binding pocket appears to allow the mutations to exert differential effects that alter transactivation ability without corresponding changes to binding affinity. For example, a mutation of M532 in mouse ERa, analogous to M528 in hERa, diminished the transactivation response of F diethylstilbestrol 7.5-fold but enhanced indenestrol A response 40-fold, with negligible changes in binding affinity (41). Single amino acid differences within nuclear receptors from different species have been previously identified (42). Reciprocal mutagenesis in the human and chicken progesterone receptor has been shown to affect sensitivity to the abortifacient, RU486 (43). The human and chicken peroxisome proliferator-activated receptors also undergo a reversal of phenotypes following reciprocal mutagenesis of a single amino acid residue (44). Although reciprocal mutagenesis of the rtER and hERa in this study did not result in a complete reciprocity of phenotypes, significant alterations in behavior were observed. In the absence of structural data, the key ligand-protein interactions and the consequences of these mutations on protein structure, as well as subsequent ligand binding and transactivation ability at different temperatures, will be difficult to ascertain. Nevertheless, results from this study demonstrate that L349 and M528, and M317 and 1496 play important roles in the differences observed between hERa and rtER, respectively, at different temperatures. Other binding pocket residue differences among species may reveal additional ligand-protein interactions that may provide further insights into species-specific ligand-induced receptor phenotypes. 140 \ This report identifies two amino acid residues that contribute not only to differences in binding and transactivation between hERoc and rtER but also play significant roles in the reduced activity of the rtER at elevated temperatures. The biological relevance of the rtER exhibiting reduced function at elevated temperature is unclear, but may involve increased flexibility in the binding pocket of the rtER compared to hERa, that causes the rtER to become unstable at elevated temperatures. The present f report also highlights the impact of temperature when comparing functional characteristics of proteins from poikilothermic species, such as rainbow trout, and humans. 141 References Evans, R. (1988) Science 240, 889-895. Carson-Jurica, M. A., Schrader, W.T., O'Malley, B.W. (1990) Endocr Rev 11(2), 201-220. Beato, M. (1991) FASEB J 5(7), 2044-2051. Tenbaum, S., and Baniahmad, A. (1997) Int J Biochem Cell Biol 29(12), 1325- 1341. Mosselman, S., Polman, J ., and Dijkema, R. (1996) FEBS Lett 392(1), 49-53. Kuiper, G. G., Enmark, E., Pelto-Huikko, M., Nilsson, S., Gustafsson, J .A. (1996) Proc Natl Acad Sci U S A 93(12), 5925-5930. Matthews, J. B., and Zacharewski, T. R. (2000) Toxicol Sci 53(2), 326-339. Matthews, J. B., Celius, T., Halgren, R., and Zacharewski, T. R. (2000) J Steroid Biochem Mol Bio 74(4), 223-234. Le Drean, Y., Kern, L., Pakdel, F., and Valotaire, Y. (1995) Mol. Cell. Endocrinol. 109, 27-35. Petit, F., Valotaire, Y., and Pakdel, F. (1995) Eur J Biochem 233(2), 584-592. Pakdel, F., Le Guellec, C., Vaillant, C., Le Roux, M., and Valotaire, Y. (1989) Mol Endocrinol 3(1), 44-5 1 . Pakdel, F., Le Gac, F., Le Goff, P., and Valotaire, Y. (1990) Mol Cell Endocrinol 71(3), 195-204. Klaunig, J. E, Ruch, R. J ., and Goldblatt, P. J. (1985) In Vitro Cell Dev Biol 21(4), 221-228. de Toledo, F. 6., Albuquerque, M. C., Goulart, B. H., and Chini, E. N. (1995) Comp Biochem Physiol C Pharmacol Toxicol Endocrinol 111(1), 93-98. Pohl, R. J ., Bend, J. R., Guarino, A. M., and Fouts, J. R. (1974) Drug Metab Dispos 2(6), 545-555. Petit, F. G., Valotaire, Y., and Pakdel, F. (2000) Nucleic Acids Res 28(14), 2634- 42. 142 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. Pike, A. C., Brzozowski, A. M., Hubbard, R. E., Bonn, T., Thorsell, A. G., Engstrom, 0., Ljunggren, J., Gustafsson, J ., and Carlquist, M. (1999) Embo J 18(17), 4608-4618. Gillesby, B., and Zacharewski, T. R. (1999) Breast Cancer Res Treat 56, 253- 265. Bradford, M. M. (1976) Anal Biochem 72(2), 248-254. Balaguer, P., Joyeux, A., Denison, M.S., Vincent,R., Gillesby, B.E., Zacharewski, T. (1996) Can J Physiol Pharmacol 74(2), 216-222. D Zacharewski, T. R., Meek, M.D., Clemons, J .H., Wu, Z.F., Fielden, M.F., Matthews, J .B. (1998) Toxicol Sci 46, 282-293. Sambrook, J ., Fristch, E.F., Maniatis, T. (1989) Molecular Cloning. A Laboratory Manual, Cold Spring Harbor Laboratory Press, New York. L Brasier, A. R., Tate, J .E., Habener, J .F. (1989) Biotechniques 7(10), 11 16-1122 Scatchard, G. (1949) Ann NY Acad Sci 51, 660-672. Wooge, C., H,. Nilsson, G,M,. Heierson, A., McDonnell, D.P., Katzenellenbogen, BS. (1992) Mol Endocrinol 6(6), 861-869. J aglaguier, S., Mesnier, D., Lager, J. L., and Auzou, G. (1996) J Steroid Biochem Mol Biol 57(1-2), 43-50. Reese, J. C., and Katzenellenbogen, B. S. (1992) J Biol Chem 267(14), 9868- 9873. Donohoe, R. M., and Curtis, L. R. (1996) Aquatic Toxicol 36, 31-52. Tan, N. S., Frecer, V., Lam, T. J ., and Ding, J. L. (1999) Biochim Biophys Acta 1452(2), 103-120. Kumar, S., Tsai, C. J ., and Nussinov, R. (2000) Protein Eng 13(3), 179-191. Vieille, C., Burdettc, D. S., and Zeikus, J. G. (1996) Biotechnol Annu Rev 2(3), 1- 83. Pace, P., Taylor, J ., Suntharalingam, S., Coombes, R. C., and Ali, S. (1997) J Biol Chem 272(41), 25832-25838. Stickle, D. F., Presta, L. G., Dill, K. A., and Rose, G. D. (1992) J Mol Biol 226(4), 1143-1159. 143 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. Kuiper, G. G., Lemmen, J. G., Carlsson, B., Corton, J. C., Safe, S. H., van der Saag, P. T., van der Burg, B., and Gustafsson, J. A. (1998) Endocrinology 139(10), 4252-4263. Ekena, K., Weis, K. B., Katzenellenbogen, J. A., and Katzenellenbogen, B. S. (1997) J Biol Chem 272(8), 5069-5075. Brzozowski, A. M., Pike, A. C., Dauter, Z., Hubbard, R. E., Bonn, T., Engstrom, 0., Ohman, L., Greene, G. L., Gustafsson, J. A., and Carlquist, M. (1997) Nature 389(6652), 753-758. Tanenbaum, D. M., Wang, Y., Williams, S. P., and Sigler, P. B. (1998) Proc Natl Acad Sci U S A 95(11), 5998-6003. Ekena, K., Weis, K. E., Katzenellenbogen, J ., A.,, and Katzenellenbogen, B. S. (1996) J Biol Chem 27(33), 20053-20059. Reese, J. C., and Katzenellenbogen, B. S. (1991) J Biol Chem 266(17), 10880- 10887. Pakdel, F., Reese, J .C., Katzenellenbogen, BS. (1993) Mol Endocrinol 7(1 1), 1408-1417. Kohno, H., Bocchinfuso, W. P., Gandini, 0., Curtis, S. W., and Korach, K. S. (1996) J Mol Endocrinol 16(3), 277-285. Govindan, M. V. (1990) Mol Endocrinol 4(3), 417-427. Benhamou, B., Garcia, T., Lerouge, T., Vergezac, A., Gofflo, D., Bigogne, C., Chambon, P., and Gronemeyer, H. (1992) Science 255(5041), 206-209. Takada, 1., Yu, R. T., Xu, H. E., Lambert, M. H., Montana, V. G., Kliewer, S. A., Evans, R. M., and Umesono, K. (2000) Mol Endocrinol 14(5), 733-740 144 CHAPTER 7 In vitro and in vivo interactions of bisphenol A and its metabolite, bisphenol A glucuronide, with estrogen receptors (1 and B These data were published in the following article: Matthews, J .B., Twomey, K, and Zacharewski, T.R. (2001) In vitro and in vivo interactions of bisphenol A and its metabolite, bisphenol A glucuronide, with estrogen receptors 0t and B. Chem Res Toxicol 14, 149-157. 145 FEBRUARY 2001 VOLUME 14, NUMBER 2 0 Copyright 2001 by the American Chemical Society Articles In Vitro and in Vivo Interactions of Bisphenol A and Its Metabolite, Bisphenol A Glucuronide, with Estrogen Receptors a and fl Jason B. Matthews,’r Ken Twomey,t and Timothy R. Zacharewski’” Department of Biochemistry & Molecular Biology, National Food Safety & Toxicology Center, and Institute for Environmental Toxicology, Michigan State University, East Lansing, Michigan 48824-1319, and Zeneca Central Toxicology Laboratory, Macclesfield, Cheshire SK10 4TJ, UK. Received August 22, 2000 The estrogenic activities of bisphenol A (BPA) and its major metabolite BPA glucuronide (BPA- G) were assessed in a number of in vitro and in vivo assays. BPA competed with [3Hl-l7fi-estradiol (E2) for binding to mouse uterine cytosol ER, a glutathione S-transferase (GST)—-human ER D, E, and F domain fusion protein (GST-hERadeO and full-length recombinant hERB. The ICso values for E2 were similar for all three receptor preparations, whereas BPA competed more effectively for binding to hERB (0.96 pM) than to either mouse uterine cytosol ER (26 uM) or GST-hERadef (36 pM). In contrast, BPA-G did not competitively displace [3H1E2 from any of the ER preparations. In MCF-7 cells transiently transfected with Gal4-hERndef or Gal4-hERfidef, BPA induced reporter gene activity with comparable E050 values (71 and 39 ”M, respectively). N 0 significant induction of reporter gene activity was seen for BPA-G. Cotreatment studies showed that concentrations of ( 10 pM) BPA and BPA-G did not antagonize E2-induced luciferase mediated through either Gal4-hERadef or Gal4-hERfidef. In vivo, the uterotropic effect of gavage or subcutaneous (sc) administration of 0.002-800 mg of BPA/kg of body weight/day for three consecutive days was examined in immature rats. Dose-related estrogenic effects on the rat uterus were observed at oral doses of 200 and 800 mg/kg and at sc doses of 10, 100, and 800 mg/kg. These results demonstrate that BPA competes more effectively for binding to ERB, but induces ERa- and ERB-mediated gene expression with comparable efficacy. In contrast, BPA-G did not exhibit any in vitro estrogenic activity. In addition, there was a clear route dependency on the ability of BPA to induce estrogenic responses in vivo. Introduction Accumulating evidence suggests that exposure to natu- ral and synthetic chemicals that mimic the activity of 17B-estradiol (E2) may adversely affect wildlife and human health. Although controversial (1 ), there have been reports of decreases in sperm production and semi- nal volume in humans during the past half-century (2-4). In wildlife, field studies indicate there are in- ‘ To whom correspondence should be addressed: Department of Biochemistry & Molecular Biology, 223 Biochemistry Building, Wilson Road, Michigan State University, East Lansing, MI 48842-1319. Telephone: (517) 355-1607. Fax: (517) 353-9334. E-mail: tzschareQ pilot.msu.edu. ' Michigan State University. ‘ Zeneca Central Toxicology Laboratory. creases in reproductive abnormalities in mammals (5), reptiles (6), birds (7), and several fish species (8) following exposure to estrogenic environmental contaminants. Many of these effects are thought to occur through an estrogen receptor (ER)-mediated mechanism of action. 10 1021/tx0001833 CCC: $20 00 O 2001 American Chemical Society Published on Web 12/30/2000 146 150 Chem. Res. Toxicol, Vol. 14, No. 2, 2001 The ER is a member of the nuclear receptor superfam- ily, a family of nuclear proteins that function as tran- scription factors to modulate gene expression in a ligand- dependent manner (9). The two subtypes, EKG and ERB, are products of distinct genes and differ in their tissue distribution, and in their ligand preferences (10, 11). For example, in rat, ERa. is expressed at higher levels in the uterus, kidney, and epididymis, while ERB levels are higher in the prostate, ovaries, and lung (12). As with other members of the nuclear receptor superfamily, the ER has a modular structure consisting of six domains (A-F) (9, 13). The highly conserved DNA-binding domain (C domain) separates a highly variable NHg-terminal region (A and B domains), which contains a ligand-inde- pendent activation region (AF-l), and a COOH-terminal region, which includes a hinge region (D domain), the ligand binding domain (E domain), and a variable F domain. The ER E domain also harbors a nuclear locali- zation signal, a dimerization region, and a ligand-de- pendent activation region (AF-2) as well as residues which interact with heat shock protein 90 (hsp90) (14- 16). When estrogen binds, the ER undergoes a confor- mational change, which allows it to bind to its cognate DNA target site, termed estrogen responsive elements (EREs), located in the regulatory region of estrogen-in- ducible genes, thereby modulating the expression of es- trogen responsive genes. Bisphenol A l4,4’-isopropylidene-2-diphenol (BPA)], a monomer used in the production of polycarbonate and epoxy resins, has been shown to elicit ER-mediated activity in a number of in vitro and in vivo assays. BPA has been shown to compete with [3H]E2 for binding to ERa and ERB (11), and induce a number of in vitro effects, including ERa- and ERB-mediated reporter gene activity (11), MCF-7 human breast cancer cell prolifera- tion, progesterone receptor expression (17), vitellogenin expression in carp hepatocytes (18), and prolactin release in a pituitary tumor cell line (19). High doses of EPA have been reported to elicit repro- ductive toxicity and abnormal cellular development in rodents (20). However, unlike E2, BPA had no effect on uterine weight in exposed rats at doses as high as 150 mg/kg of body weight, but induced peroxidase activity and elevated progesterone receptor (PR) levels similar to that of E2 (21 ). Furthermore, in cotreatment studies, BPA antagonized the E2 stimulatory effects on both peroxi- dase activity and PR levels but did not inhibit E2-induced increases in uterine weight, suggesting that BPA may use a mechanism of action distinct from that of E2 (21 ). BPA is readily metabolized in vivo through glucu- ronidation to BPA glucuronide (BPA-G), and subse- quently excreted in the feces and urine (22, 23). Phar- macokinetic studies have shown that there is a clear route dependency in the bioavailability of parent BPA, with considerably higher systemic blood concentrations of BPA after sc and intraperitoneal administration than afier oral administration (23). There has been increasing concern regarding the estrogenic activities of BPA as a result of reports that trace levels can leach from the lining of food cans, polycarbonate plastic ware, and dental resins (24, 25). Consequently, the estrogenic activities of BPA and BPA-G were investigated for their ability to compete with [3H1E2 for binding to the recombinantly expressed glu- tathione S-tranferase (GST)-estrogen receptor a. (ERa) fusion protein, consisting of the D-F domains of hERa. Matthews et al. linked to GST, and to full-length ERB proteins, to induce and antagonize ERa- and ERB-mediated reporter gene expression in MCF-7 cells transfected with Gal4-hERndef or Gal-hERBdef and the Gabi-regulated luciferase re- porter gene (17m5-G-Luc) in vitro. To account for phar- macodynamic and pharmacokinetic interactions that may affect in vivo estrogenic activity, the ability of BPA to induce a variety of uterine responses was investigated in immature Sprague-Dawley rats using two different routes of exposure. Experimental Procedures Chemicals. Bisphenol A (99.9% pure, Bayer AG) and bisphe- nol A glucuronide were provided by the BPA Global Industry Group. Dimethyl sulfoxide (DMSO) and 17fi-estradiol(E2) used in the in vitro studies were obtained from Sigma (St. Louis, MO), and E2 used in the in vivo studies was from Sigma (Poole, Dorset, UK). [2,4,6,7.16,17-3H1E2 (123 Ci/mmol) was purchased from New England Nuclear (Boston, MA). MicroScint 20 was obtained from Packard Instruments (Meriden, CT). Vent DNA polymerase was purchased from New England Biolabs (Beverly, MA), and restriction enzymes were obtained from Roche! Boehringer Mannheim (Indianapolis, IN). Phenol red-free Dul- becco's modified Eagle's medium (DMEM) and medium supple- ments were from Life Technologies (Gaithesburg, MD). Fetal bovine serum (FBS) and D-luciferin were purchased from Intergen (Purchase, NY) and Molecular Probes (Eugene, OR), respectively. All other chemicals and biochemicals were of the highest quality available from commercial sources. Animals. Immature female AlpkrAPf'SD rats (21-22 days old) with body weights of 38-53 g were obtained from the breeding unit at AstraZeneca (Alderley Park, U.K.). Animals were housed in multiple rat racks suitable for animals of this strain and the weight range expected during the course of the study. Humidity was controlled, and a 12 h light/dark cycle was maintained. Animals were weaned on Rat and Mouse No. 3 (RM3) breeding diet (Special Diets Services Ltd., Witham, Essex, UK) until they were transferred to AstraZeneca Central Toxicology Laboratory (CTL) and were then maintained on Rat and Mouse No. 1 (RMl) maintenance diet, as previously recommended (26). All animals were acclimatized to the labora- tory for at least 18 h before being treated. Construction of Plasmids. The plasmid pGEX-hERadef was constructed as previously described (27). The plasmid pGal4-hERfldef (amino acids 204-530) was constructed by PCR amplification of the hERfi (kindly provided by L. Murphy, University of Manitoba, Winnipeg, MB), using the primers pr- hf' (5'-caaactcgagcctgccgacttcggaagtgttacga-S') and pr-hr (5’- caaaggatcctcactgagactgtgggttctgg-3’). The fragment was di- gested with final and BamHI and ligated into the similarly digested eukaryotic expression vector containing the DNA binding domain of the yeast transcription factor Gal4, pG4MpolyII (kindy provided by P. Chambon, IGBMC CNRS-LGME, Illkirch Cedex C. U. de Strasbourg, France). PCR amplification was performed essentially as previously described (28) using Vent DNA polymerase in a reaction mixture containing Thermopol buffer, 200 uM dNTPs, 1 mM MgSO;, 500 nM primer, and 1.25 units of polymerase, which was heated to 94 °C for 5 min followed by 35 rounds at 94 °C for 45 s, 60 °C for 45 s, and 72 'C for 105 s. The sequence of each construct was confirmed with restriction enzyme digestion and ABl/Prism automated sequenc- ing (Perkin-Elmer Applied Biosystems, Foster City, CA). Expression and Purification of the GST—ER Fusion Protein. Expression, purification, and characterization of the GST-hERadef fusion protein were carried out as previously described (27). Meptor Binding Assays. The partially purified GST— hERadef fusion protein or recombinant hERB (PanVera, Madi- son, WI) was diluted in TEGD buffer [10 mM Tris (pl-I 7.6), 1.5 mM EDTA, 1 mM DTT, and 10% (v/v) glycerol] containing 1 mg/mL bovine serum albumin (BSA) as a carrier protein. For 147 Ti Estrogenic Activity of BPA and BPA-G mouse uterine cytosol preparation, immature female CD-1 mice were obtained from Charles River Laboratories (Wilmington, MA) and maintained in a controlled environment (40-60% humidity; 20-22 “C ambient temperature) on a 12 h light/dark cycle. Harlen Teklad 22/5 Rodent Diet 8640 (Madison, WI) and tap water were provided ad libitum. Uterine tissue from 21- day-old CD-1 mice was excised, trimmed of excess fat and connective tissue, weighed, and homogenized in 1.0 mL of TEGD [10 mM Tris base, 1.5 mM EDTA, 10% glycerol, and 1.0 mM DTT (pH 7.6)] per 50 mg of uterine tissue with 3 x 20 s bursts using a Brinkman Polytron homogenizer at 50% output. Samples were centrifuged at 3000g for 10 min at 4 'C. The supernatant (cytosol) was centrifuged at 105000g for l h at 4 °C. The protein concentration of the cytosol was adjusted to 2.0 mg/mL and the protein stored at -80 °C. Protein preparations and the mouse uterine cytosol were incubated at 4 'C for 24 h and at 30 °C for 2 h, respectively, with 2.5 nM [3HIE2 and increasing concentrations of unlabeled competitor in 1 mL glass tubes arranged in a 96-well format (Marsh Scientific, Rochester, NY). Protein preparations were diluted to ensure 10 000 dpm of total binding (dilutions were 500-fold hERB and lSOO-fold GST-hERadeD. Binding assays were initiated by adding 240 ,uL of the protein preparation to glass tubes containing 5 pL of DMSO and 5 uL of [3H1E2; thus, the solvent concentration did not exceed 4%. The amount of nonspecific binding was determined in the presence of a 400- fold excess of unlabeled E2. Bound [’HlE2 was separated from free [3H1E2 using a 96-well filter plate and vacuum pump harvester (Packard Instruments). Filter plates containing the protein were washed with 3 x 50 mL of TEG [10 mM Tris buffer (pH 7.6), 1.5 mM EDTA, and 10% (v/v) glycerol] and allowed to dry under continued suction for 30 3. After drying, the under- sides of the filter plates were sealed and 50 uL of MicroScint 20 scintillation cocktail was added to each well. The amount of bound [3H1E2 was measured using a TopCount luminescense and scintillation counter (Packard Instruments). Each treatment was performed in quadruplicate, and results are expressed as the percent specific binding of [3HlE2 versus the log of the competitor concentration. The reported 1050 values represent the concentration of test compound required to displace 50% of the [3H] E2 from the ER preparation as compared to the 50% displacement of [3H1E2 achieved by unlabeled E2. These analyses were performed using GraphPad Prism 3.0 software (GraphPad Software Inc., San Diego, CA). Cell Culture. MCF-7 human breast cancer estrogen receptor positive cells (obtained from L. Murphy, University of Manitoba, at passage 32-35) were maintained with phenol red-free D-MEM supplemented with 10% FBS, 10 mM HEPES, 2 mM L—glutamine, 500 ug/mL gentamicin, 2.5 ug/mL amphotericin B, 100 IU/mL penicillin G, and 100 ug/mL streptomycin. Cells were cultured in a humidified environment at 37 °C with 5% C03. Transfection and Reporter Gene Assays. Transient trans- fections and gene transcription assays were performed es- sentially as previously described (29, 30). Briefly, MCF-7 cells were seeded at approximately 50% confluency in six-well tissue culture plates in medium supplemented with 5% dextran-coated charcoal-treated fetal bovine serum (DCC-FBS) and allowed to settle for 7 h. Cells were transiently transfected by the calcium phosphate coprecipitation method (31) with 2.5 pg of 17m5-G- Luc, 0.2 ,ug of Gal4-hERadef, or 0.2 ug Gal4-hER/idef, along with 0.01 pg of pCMV-lacZ (B-galactosidase expression vector). Cells were washed 16 h later with sterile phosphate-buffered saline (PBS), and then fresh medium was added to each well. Transiently transfected MOE-7 cells were exposed to final concentrations ranging (i) from 1 nM to 10 uM for BPA or BPA-G and (ii) from 1 pM to 10 nM for E2 or (iii) DMSO (solvent) alone. Final concentrations were obtained by adding 2 uL of the test chemical to 2 mL of the medium. Following incubation with the sample for 24 h, cells were harvested and assayed for luciferase activity according to the method described previously (32). In the transiently transfected MCF-7 cells, the reference plasmid pCMV-lacZ was cotransfected as an internal control to Chem. Res. Toxicol, Vol. 14, No. 2, 2001 151 correct for differences in transfection efficiencies and extraction differences. B-Galactosidase activity was measured according to standard protocols (31). Each treatment was carried out in duplicate, and two samples were taken from each replicate. Each experiment was repeated three times. Values are reported as a percentage relative to the maximum induction observed with 10 nM E2. Uterotrophic Assay. The protocol for the uterotrophic assay has been previously described (26, 33). Animals were treated by either gavage or so injection with BPA suspended in arachis (peanut) oil. Clinical observations and body weights were recorded daily. The dosing volume used in both studies was 5 mng of body weight, calculated on the weight of the animal immediately prior to administration. Animals were treated with BPA for three consecutive days and were killed by an overdose of halothane Ph Eur vapor approximately 24 h after the final dose. The dose levels for the oral gavage study were 0.002, 0.02, 0.2, 1.0, 10, 100, 200, and 800 mg of BPA/kg/day with 10 animals per treatment group. The dose levels for the sc injection study were 0.002, 0.02, 0.2, 1.0, 10, 100, and 800 mg of BPA/kg/day with 10 animals per treatment group. The high dose levels for both studies were based on preliminary work which established that oral and subcutaneous doses of up to 800 mg of BPA/kg] day did not induce more than mild toxicity in the experimental animals. The remaining dose levels were chosen by the BPA Global Industry Group to cover a wide range of exposure to the experimental animal. E2 was used as the positive control for both studies, using a dose level of 0.4 mg/kg of body weight} day, as recommended previously (26). At necropsy, the uteri were excised, trimmed free of fat and any adhering nonuterine tissue, pierced, and blotted to remove excess fluid. The body of the uterus was cut just above its junction with the cervix and at the junction of the uterine horn and the ovaries. The uterus was then weighed (wet weight). prior to bisection of the uterus at the junction of the uterine horns. One uterine horn was reweighed (wet) and then dried in a preweighed glass vial for a minimum of 24 h in an oven at a temperature of 70 ’C. After being dried, the sample was removed and left to cool in a desiccator for a minimum of 1 h and the uterine dry weight recorded. The other section of the uterus was preserved in 10% neutral buffered formol saline and processed for histopathological examination. Liver weight and plasma alanine aminotransferase and aspartate aminotrans- ferase levels were recorded, but the data are not presented in this paper. Statistical analysis was conducted using the SAS (1989) package as follows: body weights, by analysis of covariance on initial body weight; and uterus weight, by analysis of variance and analysis of covariance on terminal body weights. Results Competitive Binding Ability of BPA and BPA-G. The ability of BPA and BPA-G to compete with [3H1E2 for binding to mouse uterine ER, GST-hERudef, and commercially available baculovirus-expressed full-length hERB (Figure 1) preparations was investigated in vitro using a semi-high-throughput competitive binding assay. BPA competed for binding to each of the ER preparations; however, the 1050 values varied among preparations (Table l). BPA bound to hERB approximately 27- and 38- fold better than GST-hERadef and mouse uterine cyto- sol, respectively (Table l). BPA-G did not competitively displace [3HlE2 from any of the ER preparations (Figure l) at the concentrations that were examined (1 nM to 100 pM). Concentrations of > 100 uM were not examined due to possible artifactual results from competitor ligand precipitation. Ability of BPA and BPA-G To Induce ERa- and ERB-Mediated Gene Expression. BPA and BPA-G 148 152 Chem. Res. Toxicol., Vol. 14, No. 2, 2001 Matthews et al. Table 1. Summary of the [Cu Values of the Competitive Binding of l7p-Estradiol, Bisphenol A, and Bisphenol A Glucuronide to GST—hERadef, Recombinant hERfl, and Mouse Uterine Cytosol ICso (MyI chemical GST-hERadef hER/i mouse cytosol l7/3-estradiol (2.9 :i: 0.5) x 10’9 (3.6 :i: 0.4) x 10'9 (3.0 :i: 0.2) x 10‘9 bisphenol A (3.6 :1: 1.6) x 10“ (9.6 :1: 2.3) x 10'7 (2.6 :t 1.1) x 10"“ bisphenol A glucuronide nb" nb nb " ICgo values represent the means and standard deviations from three independent experiments. " nb denotes nonbinder since this chemical did not displace more than 10% of the [3H1E2 at the highest concentration that was examined (100 uM). A 125- v. (1mm Bound $ El 31.173.333.334 v. [1mm Bound to effi’ij I I 1 41.1730 3 .7 s -s I “‘1 “”1 75-i % [3H)E2 Bound O 4'1-1'0-3 -'s .1 -s .5 .4 Log [Competitor] Figure 1. Competitive binding of BPA and BPA glucuronide to (A) the mouse uterine ER, (B) GST-hERadef, and (C) recombinant hERB. Protein preparations were incubated with 2.5 nM [3H]E2 and increasing concentrations of unlabeled (0) E2, (0) BPA, or (I) BPA-G. The competitive displacement of radioligand was assessed using a vacuum pump filteration method as described in Experimental Procedures. Displacement curves were obtained for E2 and BPA. No significant displace- ment was observed with BPA-G. The data are the results from a representative experiment that was repeated three times. induction of ER-mediated gene expression was assessed by measuring luciferase activity using MCF-7 cells cotransfected with Gal4-hERadef or Gal4-hER,6def and the Gabi-regulated luciferase reporter gene, 17m5-G-Luc. The results in Figure 2 show that E2 treatment of transiently transfected MCF-7 cells with either Gal4- hERndef or Gal4-hERBdef caused a concentration-de- pendent increase in luciferase activity. The E050 values for this response were similar for both chimeric receptors (Table 2). MCF-7 cells transiently transfected with Gal4- hERadef and Gal4-hER,6def exhibited a maximum induc- tion response of approximately 30- and 20-fold, respec- tively, following treatment with 10 nM E2. > 125- 1”- %l.ucifcrxsc Activity 5' 5 if 0‘ 4241-10 4 -I -7 -C 4 3 % Luciferase Activity w a??? 2.21" 412-1130 :9 -'s 37 -'s .3 Log [Concentration] Figure 2. Effect of BPA and BPA glucuronide on reporter gene expression. MCF-7 cells transiently transfected with Gal4- human estrogen receptor a (Gal4-hERade0 or Gabi-human estrogen receptor fl (Gal4-hERfidef) and the Gal4-regulated reporter gene (17m-5-G-Luc) were treated with (0) E2, (0) BPA, or (I) BPA-G. Luciferase activity was measured 24 h after treatment as described in Experimental Procedures. The data are the results from a representative experiment that was repeated three times. Table 2. Summary of the Ability of I'm-estradiol, Bisphenol A. and Bisphenol A Glucuronide to Induce Estrogen Receptor-mediated Gene Expression ECso (M)ll chemical Gal4-hERadef Gal4-hERJidef 17/3-estradiol (5.3 :t 2.1) x 10‘11 (8.3 :i: 2.9) x 10’11 bisphenolA (7.1 :l: 2.9) x 10‘7 (4.5 :i: 1.8) x 10'? bisphenol A glucuronide ni" ni ‘ ECso values represent the means and standard deviations from three independent experiments. 5 ni denotes no induction. BPA caused a dose-dependent increase in luciferase activity in MCF-7 cells transiently transfected with either Gal4-hERadef or Gal4-hERBdef (Figure 2). The ECso values for the response were similar for both ERa and ERB (Table 2). Concentrations of >10 uM were not examined due to visible precipitation in the medium. BPA-G did not significantly induce ERa- or ERB-medi- ated luciferase activity at the highest concentration that was examined (10 uM) (Figure 2). Similar results were observed using a Gal4-mERBdef consutruct (data not shown). Induction of luciferase activity by BPA was absent in transiently transfected MCF-7 cells if the chimeric construct, Gal4-hERadef or Gal4-hER,8def, was not cotransfected (data not shown). 149 Estrogenic Activity of BPA and BPA-G 125 5‘ €100 < {75 in I! 0 case +:-- +--- E: -+++ -+++ on -—+- --+- BPA-(1 ---+ ---+ GsM-hERudef m! Figure 3. Effect of BPA and BPA-G cotreatment on E2 induction of reporter gene activity. MCF-7 ce transiently transfected with Gabi-human estrogen receptor a (Gal4- hERadel) or Gal4- human estrogen receptor [3 (Gal4-hERfldef) and the Gal4- -ureg urlated reporter gene (17m- 5—-GLuc) were cotreated with10 nM E210uM EPA or 10 FM BPA- G. The data are the results from a representative experiment that was repeated two times Figure 3 illustrates that cotreatnnent of MCF-7 cells with 10 uM BPA or BPA-G did not result in significant changes in the E2 induction of luciferase activity in MCF-7 cells transiently transfected with either Gal4- hERudef or Gal4-hER45def. Uterine Effects following Oral Administration of BPA. There was a significant increase in mean uterine wet and dry weight in the groups treated with 200 and 800 mg of BPA/kg/day in comparison with the vehicle Chem. Res. Toxicol., Vol. 14, No. 2, 2001 153 control group (Table 3). As expected, the uterine wet and dry weights in the positive control group were signifi- cantly increased compared with those of the vehicle control group. Minimal to moderate endometrial hypertrophy/hyper- plasia was observed in eight of the females treated with 800 mg of BPA/kg/day (Table 4). This efi‘ect was similar to, but less pronounced than, that of the positive control group (treated with 0.4 mg of EWkg/day), where all the females exhibited moderate or marked change. Dose- related increases in severity of endometrial glandular epithelial apoptosis and the number of stromsl neutro- phils were observed in females treated with 200 or 800 mg of BPA/kg/day and also in the positive control group animals, where the increases were greater than those observed in the BPA-treated animals. An increased incidence and severity of lumenal epithelial apoptosis was also observed in females treated with 800 mg of BPA! kg/day, although this was not observed in the positive control group. There were no treatment-related micro- scopic changes observed in the uteri of the animals from the remaining groups. One female in the group treated with 800 mg of BPA/ kg/day was killed humanely due to adverse clinical signs, although there were no abnormalities observed at necrop- sy that could explain these signs. There was no efi'ect of treatment on body weight gain. Uterine Effects following Subcutaneous Iniection of BPA. There was a slight but significant reduction in normalized body weight over the treatment period in the Table 3. Effect of Oral Gavage of Bisphenol A on Blotted and Dry Uterine Weight in Rats 1/2 uterus terminal body uterus l/z uterus V2 uterust percent loss dose level no. of weight (3) weight (mg) blotted weight on ' (mg/kg/day) animals (mean :1: SD) (mean :1: SD) (mg) (mean i SD) (mg) (me:nz :1: tSD) (mean 3: SD) arachis oil 10 61.7 i 4.5 29 :1: 9 13 d: 3 3.4 :i: 0.04 73.3 :i: 3.8 bisphenol A 0.002 10 61.2 i: 6.1 34 :t 8 16 i 3 3 8 :1: 0.09 75.9 i: 2.2 bisphenol A 0.02 10 59.8 i 6.3 31 :i: 9 15 :1: 3 3 8 :i: 0.06 73.4 i 3.6 bisphenol A 0 2 10 60.5 :1: 8.4 31 i: 9 15 :t 4 3.7 :1: 0.12 75.4 :1: 3.1 bisphenol A l 10 61.2 :h 6.6 33 d: 9 16 2t 4‘ 4 3" 2t 0.11 73.7 i 5.2 bisphenol A 10 10 61.8 :1: 6.3 30 i 11 13 :1: 3 3 7 :i: 0.09 71.8 :t 5.5 bisphenol A 100 10 60.6 :t 6.9 29 2t 6 14 :i: 3 3 6 i 0.08 74 0 i 3 0 bisphenol A 200 10 61.8 i 5.5 36 d: 8" 17 :i: 4" 4 3 :i: 0.07'I 74 4 d: 2 2 bisphenol A 800 9 60.7 i 6.7 58 i 14" 28 i 6" 6.0 i O. 08" 78 1 i 2. 6" ”fl-estradiol 0.4 10 61.4 d: 5.1 109 i 19” 52 :i: 11” 10 3 :i: 0.18" 80.0 d: 1.1” " p < 0.05 with a Student’s t test (two-sided). " p < 0.01 with a Student's t test (two-sided). Table 4. Microscopic Changes in the Uterus following Oral Gavage of Bisphenol A 0'I 0.002“ 0.02“ 0.2‘I 1" 10“ 100‘l 200' 800“ 0.4” no oaf male 10 10 10 10 10 10 10 10 9 10 endometrial hypertrophy/hyperplasia minimal 0 0 0 0 0 0 0 2 4 0 slight 0 0 0 0 0 0 0 0 3 0 moderate 0 0 0 0 O 0 0 0 I 8 Red 0 0 0 0 0 0 0 0 0 2 endometrial epithelial apoptosis (lumen) minimal 1 0 0 1 0 0 l O 3 3 slight 0 0 0 0 0 0 0 1 2 0 0 0 0 0 0 0 0 0 l 0 endometrial epithelial apoptosis (glands) minimal 7 5 4 8 4 5 6 6 3 0 slight 0 0 0 0 0 0 0 2 1 1 moderate 0 0 0 0 0 0 0 0 3 8 increased stromsl neutrophils ‘nim 0 0 0 0 0 0 0 2 4 0 slight 0 0 0 0 0 O 0 0 4 2 moderate 0 0 0 0 0 0 0 O 0 8 ° Milligrams of BPA per kilogram of body weight per day. " Milligram of E2 per kilogram of body weight per day. 150 154 Chem. Res. Toxicol., Vol. 14, No. 2, 2001 Matthews et al. Table 5. Effect of Subcutaneous Administration of Bisphenol A on Blotted and Dry Uterine Weight in Rats l/2 uterus terminal body uterus ll: uterus V2 uterus percent loss dose level no. of weight (g) weight (mg) blotted weight dry weight on drying (mg/kg/day) animals (mean :1: SD) (mean :l: SD) (mg) (mean 1 SD) (mg) (mean :i: SD) (mean :1: SD) arachis oil 0 10 63.3 d: 2.4 42 i 8 19 :t 5.1 2.9 :t 1.7 85.3 :i: 6.7 bisphenol A 0.002 10 61.9 :t 4.5 41 :i: 14 19 :l: 5.6 3.0 :i: 2.7 85.2 :t 13.4 bisphenol A 0.02 10 61.4 :i: 3.4 48 :i: 21 22 :i: 11 3.3 i: 4.7 88.7 i 10.5 bisphenol A 0.2 10 63.4 :i: 6.4 37 :i: 8 16 :i: 3.3 1.6 :i: 1.4 90.1 :1: 7.8 bisphenol A 1 10 66.5 :i: 3.5 43 :b 10 21 :i: 4.2 3.0 :i: 2.4 86.7 :t 9.1 bisphenol A 10 10 66.4 i: 1.7 44 :t 10 19 :i: 4.9 2.0 i: 2.1 89.5 :i: 12.8 bisphenol A 100 10 64.4 :h 2.4 51 i 5 22 :i: 3.7 1.2 :h 1.2 94.9 :1: 4.7“ bisphenol A 800 10 62.6 :i: 4.0 91 :i: 33" 42 i 14 4.3 :i: 3.7 90.7 i: 6.0 17fl-estradiol 0.4 10 62.4 :i: 5.8 135 :i: 14" 60 :i: 6.45 7.7 i 2.4“ 87.5 :i; 3.4 ‘ p < 0.05 with a Student’s t test (two-sided). " p < 0.01 with a Student’s t test (two-sided). Table 6. Microscopic Changes in the Uterus following Subcutaneous Administration of Bisphenol A no. of animals On H O 0.002“ 10 0.02“ 10 0.2‘ll 10 la H O 10‘I g.- Q p—n O 100“ 800‘I 0.4" 10 H O endometrial hypertrophy/hyperplasia minimal slight moderate marked endometrial epithelial apoptosis (lumen) minimal slight moderate endometrial epithelial apoptosis (glands) minimal slight moderate increased stromal neutrophils minimal slight moderate vacuolar degeneration of the endometrial epithelium minimal slight mononuclear cell infiltration (total) minimal 0000 GOO GOO COO 0000 0000 GOO GOO COO 0000 ooooo Han-Io OtDr-a no.5 H0000 COO QNOO no 0000 GOO GOO COO 0000 HHOO COO OQDH GOOD GOO GOO 0°C 00°C COCO GOO GOO COO 0000 GOOD GOO COO) 000) 0000 0000 OWN QWH CONDO! (Du-IO ' Milligrams of BPA per kilogram of body weight per day. " Milligram of E2 per kilogram of body weight per day. group treated with 800 mg of BPA/kg/day and in the positive control group in comparison with the control group (data not shown). There was a significant increase in mean uterine wet and dry weights in the group treated with 800 mg of BPA/ kg/day, compared with those in the vehicle control group (Table 5). The mean uterine wet weight for the group treated with 100 mg of BPA/kg/day was also increased, but was not statistically significance. As expected, the uterine wet and dry weights in the positive control group were significantly increased compared with those of the vehicle control group. Slight to marked endometrial hypertrophy/hyperplasia was observed in all females treated with 800 mg of BPA/ kg/day (Table 6). This effect was similar to, but less pronounced than, that of the positive control group (treated with 0.4 mg of E2/kg/day). At 100 mg of BPA/ kg/day, eight females were observed with minimal change. Increases in the severity of endometrial glandular epi- thelial apoptosis were observed in females treated with 10, 100, or 800 mg of BPA/kg/day. These findings were also observed in the positive control group animals, where the increase and severity were lower than those observed in the animals treated with 800 mg of BPA. Increases in the number of stromal neutrophils were observed in females treated with 100 or 800 mg of BPA/kg/day. These findings were also observed in the positive control group animals, where the increases were higher than those observed in the animals treated with 800 mg of BPA. Luminal epithelial apoptosis was observed in females treated with 10, 100, and 800 mg of BPA/kg/day, with the highest incidence occurring in the group treated with 100 mg of BPA, although this was not observed in the positive control group. There were no treatment-related microscopic changes observed in the uteri of the animals from the remaining groups. There were no clinical signs observed during the study that were considered to be related to treatment. Discussion The ligand binding affinities of ERG. and E125 proteins for physiological ligands, such as E2, are quite similar (12). However, differences in ligand preference and relative binding affinity between subtypes for other estrogenic chemicals and phytoestrogens have been demonstrated (11, 34, 35) and were observed in this study. BPA exhibited a 38- and 27 -fold greater ability to compete for binding to ER]? than to GST-hERudef and mouse uterine cytosol, respectively. This differential binding is comparable to previously reported leo values (12), although lower values have been reported for other ERfi preparations (11). The reported differences may be due to the use of a baculovirus/SF9-expressed ERB preparation in the current study that included an ad- 151 Estrogenic Activity of BPA and BPA-G ditional 53 amino acid residues at the N-terminus (36), which was absent from earlier clones (10, 37) and not present in the ERB protein preparations used in other binding studies (11, 12). Nevertheless, the amino acid sequences within the respective ligand binding domains are identical in all of the competitive binding studies. Differences in reported ERfl ICso values for the same ligand among studies may also be due to variation in assay protocols. In contrast, BPA-G did not compete with [3HlE2 for binding to either protein preparation. Other studies have also reported differences in the relative binding affinity between ERa and ERfi for the same compound. For example, an aryl-substituted pyra- zole exhibited higher binding affinity for EKG. and 120- fold greater potency in transactivational activity than ERfl (35). In addition, genistein has been shown to preferentially bind with 30-fold greater potency to hERfi than to hERa (12). The overall structure of the ERB ligand binding pocket is similar to that of ERu with ligand binding being achieved by a combination of specific hydrogen bonding interactions and the hydrophobic nature of the binding pocket (38-40). It has been suggested that the preference for genistein may be attributed to two conservative mutations within the binding pocket that further stabilize the hERfi-genistein complex (38). BPA has been shown to exhibit mixed agonist and antagonistic effects through ERa while behaving solely as an agonist through ERB (41 ). However, in the current study, no antagonistic effects were observed. Despite the preferential binding for hERB, BPA exhibited an activity in Gal4-hERfldef reporter gene induction that was <2- fold greater than the activity in Gal4-hERadef. The lack of correlation between binding affinity and transactiva- tional activity may be due, in part, to alterations in the ability of the liganded ER complex to bind DNA (42, 43). Differential interactions between EEO. and EM with cellular proteins, such as coactivators, may also be a contributing factor (44). Crystal structure data for ER/B complexed with genistein indicate that helix 12, which interacts with coactivators, is positioned along a cleft rather than over the binding cavity as seen ERa-EZ and Elia—diethylstilbestrol structures (39, 45). Therefore, the ligand-dependent positioning of helix 12 may account for the difference in the ability of genistein to induce ERa— and ERB-mediated gene expression, despite the greater affinity for ERB. To date, no ERa-genistein structure has been reported to support this hypothesis. In vivo effects of BPA on uterine weight and uterine microscopic changes were examined in immature rats treated by gavage to mimic the primary route of exposure, and by sc to bypass hepatic first-pass elimination and subsequent metabolism to BPA-G. Increases in wet and dry uterine weight were observed at doses of 800 mg of BPA/kg/day for administration by both routes of exposure and also at a dose of 200 mg of BPA/kg/day for oral administration. The effects on uterine weight are con- sistent with other recent reports (46, 47). However, rat strain differences in vaginal responses to BPA have been reported (48). In addition, sc administration of BPA has been shown to increase uterine wet weight in BGC3F1 mice at doses ranging from 0.02 to 8 mg/day (49). E050 values of 0.72 mg of BPA/day and 19.4 ng of E2/day have been estimated for this response (49), although it has been argued that these values do not represent true potency comparisons, since BPA-induced a <2-fold in- Chem. Res. Toxicol., Vol. 14, No. 2, 2001 155 crease in uterine weight compared to a >5-f'old increase induced by E2 (50). Tinwell and co-workers (51) were unable to demonstrate that BPA is reproducibly active in the mouse uterotrophic assay in which AP mice were treated with either by sc or gavage with doses ranging from 0.02 pg/kg to 300 mg/kg of BPA. This is also consistent with the work of Coldham et al. (52), where sc injection of BPA failed to increase uterine weight at dose levels of 3, 33, and 330 mg/kg. Increases in the incidence of endometrial hypertrophy, endometrial epithelial apoptosis in the lumen, and en- dometrial and stromal neutrophils were observed in gavaged rats treated with 200 and 800 mg of BPA/kg] day. Subcutaneous injections of 10, 100, and 800 mg of BPA/kg/day also increased the incidence of changes in the microscopic structure of the uterus. These microscopic findings are similar to effects induced by other weak estrogen agonists (53) and are consistent with vaginal cornification results (47). BPA has also been shown to induce molecular, cellular, and tissue level effects in rats at doses that did not induce uterine weight. BPA induced progesterone receptor ex- pression in rats treated with 5-150 mg/kg (21 ). Similarly, BPA and the phytoestrogen, diadzin, were weak stimula- tors of uterine growth in rat, but were able to alter the expression of the androgen receptor, the estrogen recep- tor, and complement 3 (54). Therefore, weak estrogenic chemicals can induce changes at the gene expression and cellular levels without affecting uterine weight. However, the toxicological significance of these changes is unclear. At low doses that do not induce uterine wet weight, gestational exposure of mice to BPA has been reported to affect accessory reproductive organ weights and daily sperm production in male offspring (55, 56). In contrast, other rodent studies indicate that 0.2-200 pg of BPA/ kg/day in drinking water does not affect reproductive organ development or sperm counts (57, 58). These contradictory results observed in similar assays make the assessment of BPA and other xenoestrogens a conten- tious issue that needs to be resolved (50). The results of this study demonstrate that BPA exhibits a greater ability to compete for binding to ERB than to ERu. However, the ability of BPA to induce reporter gene expression mediated by either isotype is comparable. In addition, these results demonstrate that there is a clear route of administration dependency on the ability of BPA to induce uterine responses, as observed in Long Evans rats (59). This route dependency supports results from competitive binding and reporter gene assays that demonstrate the BPA-G does not exhibit significant estrogenic activity in the uterus, since BPA is more rapidly metabolized to BPA-G when administered by gavage than when administered by sc injection (23). Acknowledgment. We thank Mark Fielden and other members of the lab for their helpful discussions during the course of these studies. This work was supported by the BPA Global Industry Group. References (1) Safe, S. H. (2000) Endocrine Disruptors and Human Health-Is There a Problem? An Update. Environ. Health Perspect. 108, 487—493. (2) Carlsen, E., Giwercman, A., Keiding, N., and Skakkebaek, N. E. (1995) Declining semen quality and increasing incidence of testicular cancer: is there a common cause? Environ. Health Perspect. 103, 137-139. 152 156 Chem. Res. Toxicol., Val. 14, No. 2, 2001 (3) Sharpe, R. M., and Skakkebaek, N. E. (1993) Are oestrogens involved in falling sperm counts and disorders of the male reproductive tract? Lancet 341, 1392- 1395. (4) Irvine, D. S. (1994) Falling sperm quality. Br. Med. J. 909, 476. (5) Reijnders, P. J. (1986) Reproductive failure in common seals feeding on fish from polluted coastal waters. Nature 324, 456— 457 (6) Guillette, L. J. J., Gross, T. S., Masson, G. R., Matter, J. M., Percival, H. P., and Woodward, A. R. (1994) Developmental abnormalities of the gonad and abnormal sex hormone concentra- tions in juvenile alligators from contaminated and control lakes in Florida. Environ. Health Perspect. 102, 680-688. (7) Fry, D. M. (1995) Reproductive effects in birds exposed to pesticides and industrial chemicals. Environ. Health Perspect. 103, 165—171. (8) Von Westemhagen, H., Rosenthal, H., Dethlefsen, V., Ernst, W., Harms, U. L., and Hansen, P. D. (1981) Bioaccumulating sub. stances and reproductive success in the Baltic flounder, Platich- thys flesus. Aquat. Toxicol. 1, 85-99. (9) Evans, R. (1988) The steroid and thyroid hormone receptor superfamily. Science 240, 889—895. (10) Mosselman, S., Polman, J., and Dijkema, R. (1996) ER beta: identification and characterization of a novel human estrogen receptor. FEBS Lett. 392, 49-53. (11) Kuiper, G. G., Lemmen, J. G., Carlsson, B., Carton, J. 0., Safe, S. H., van der Saag, P. T., van der Burg, B., and Gustafsson, J. A. (1998) Interaction of estrogenic chemicals and phytoestrogens with estrogen receptor beta. Endocrinology 139, 4252—4263. (12) Kuiper, G. G., Carlsson, B., Grandien, K., Enmark, E., Haggblad, J., Nilsson, S., and Gustafsson, J. A (1997) Comparison of the ligand binding specificity and transcript tissue distribution of estrogen receptors alpha and beta. Endocrinology 138, 863- 870. (13) Tenbaum, S., and Baniahmad, A (1997) Nuclear receptors: structure, function and involvement in disease. Int. J. Biochem. Cell Biol. 29, 1325-1341. (14) Webster, N. J., Green, S., Jin, J. R., and Chambon, P. (1988) The hormone-binding domains of the estrogen and glucocorticoid receptors contain an inducible transcription activation function. Cell 54, 199—207. (15) White, R., Fawell, S. E., and Parker, M. G. (1991) Analysis of oestrogen receptor dimerisation using chimeric proteins. J. Steroid Biochem. Mol. Biol. 40, 333-341. (16) Danielian, P. S., White, R., Lees, J. A., and Parker, M. G. (1992) Identification of a conserved region required for hormone depend- ent transcriptional activation by steroid hormone receptors. EMBO J. 11, 1025-1033. (17) Krishnan, A. V., Stathis, P., Permuth, S. F., Tokes, L., and Feldman, D. (1993) Bisphenol-A: an estrogenic substance is released from polycarbonate flasks during autoclaving. Endocri- nology 132, 2279-2286. (18) Smeets, J. M., van Holsteijn, I., Giesy, J. P., Seinen, W., and van den Berg, M. (1999) Estrogenic potencies of several environmental pollutants, as determined by vitellogenin induction in a carp hepatocyte assay. Toxicol. Sci. 50, 206—213. (19) Chun, T. Y., and Gorski, J. (2000) High concentrations of bisphenol A induce cell growth and prolactin secretion in an estrogen-responsive pituitary tumor cell line. Toxicol. Appl. Pharmacol. 182, 161-165. (20) Morrissey. R. E.. George, J. D., Price, C. J., Tyl, R. W., Marr. M. C., and Kimmel, C. A (1987) The developmental toxicity of bisphenol A in rats and mice. Fundam. Appl. Toxicol. 8. 571- 582. (21) Gould, J. C., Leonard, L. S., Maness, S. C., Wagner, B. L., Conner, K, Zacharewski, T., Safe, 8., McDonnell, D. P., and Gaido, K. W. (1998) Bisphenol A interacts with the estrogen receptor alpha in a distinct manner from estradiol. Mal. Cell. Endocrinol. 142, 203- 214. (22) Knaak, J. B., and Sullivan, L. J. (1966) Metabolism of bisphenol A in the rat. Toxicol. Appl. Pharmacol. 8, 175-184. (23) Pottenger, L. H., Domoradzki, J. Y., Markham, D. A., Hansen, S. C., Cagen, S. 2., and Waechter, J. M., Jr. (2000) The relative bioavailability and metabolism of bisphenol A in rats is dependent upon the route of administration. Toxicol. Sci. 54, 3-18. (24) Brotons, J. A., Clea-Serrano, M. E, Villalobos, M., Pedraza, V., and Olea, N. (1995) Xenoestrogens released from lacquer coatings in food cans. Environ. Health Perspect. 103, 608-612. (25) Olea, N ., Pulgar, R., Perez, P., Clea-Serrano, F., Rivas, A., Novillo- Fertrell, A., Pedraza, V., Soto, A. M., and Sonnenschein, C. (1996) Estrogenicity of resin-based composites and sealants used in dentistry. Environ. Health Perspect. 104, 298-305. (26) Odum, J., Lefevre, P. A., 'I‘ittensor, S., Paton. D., Routledge, E. J., Beresford, N. A., Sumpter, J. P., and Ashby, J. (1997) The Matthews et al. rodent uterotrophic assay: critical protocol features, studies with nonyl phenols, and comparison with a yeast estrogenicity assay. Regal. Toxicol. Pharmacol. 25, 176-188. (27) Matthews, J. B., and Zacharewski, T. R (2000) Differential binding affinities of PCBS, HO-PCBs and Aroclors with recom- binant human, rainbow trout (Onchorhynkiss mylziss) and rep- tilian (Anolis carolinensis) estrogen receptors using a semi- high throughput competitive binding assay. Toxicol. Sci. 53, 326- 339. (28) Gillesby, B., and Zacharewski, T. R. ( 1999) p82 (TFF2) mRNA levels in human breast tumor samples: correlation with clinical and histological prognostic markers. Breast Cancer Res. Treat. 58. 253—265. (29) Balaguer, P., Joyeux, A., Denison, M. 8., Vincent, R., Gillesby, B. E., and Zacharewski, T. (1996) Assessing the estrogenic and dioxin-like activities of chemicals and complex mixtures using in vitro recombinant receptor-reporter gene assays. Can. J. Physiol. Pharmacol. 74 (2), 216-222. (30) Zacharewski, T. R, Meek, M. D., Clemons, J. H., Wu, Z. P., Fielden, M. F., and Matthews, J. B. (1998) Examination of the in vitro and in viva estrogenic activities of eight commercial phtha- late esters. Toxicol. Sci. 48, 282-293. (31) Sambrook, J ., Fristch, E. R., and Maniatis, T. (1989) Molecular Cloning. A Laboratory Manual, Cold Spring Harbor Laboratory Press, Plainview, NY. (32) Brasier, A. R., Tate, J. E., and Habener, J. F. (1989) Optimized use of the firefly luciferase assay as a reporter gene in mammalian cell lines. BioTechniques 7, 1116—1122. (33) Wakeling, A. E., Dukes, M., and Bowler, J. (1991) A potent specific pure antiestrogen with clinical potential. Cancer Res. 51, 3867— 3873. (34) Barkhem, T., Carlsson, B., Nilsson, Y., Enmark, E., Gustafsson, J ., and Nilsson, S. (1998) Differential response of estrogen receptor alpha and estrogen receptor beta to partial estrogen agonists/antagonists. Mol. Pharmacol. 54, 105—112. (35) Sun, J., Meyers, M. J., Fink, B. E., Rajendran, R., Katzenellen- bogen, J. A., and Katzenellenbogen, B. S. ( 1999) Novel ligands that function as selective estrogens or antiestrogens for estrogen receptor-alpha or estrogen receptor-beta. Endocrinology 140, 800- 804. (36) Ogawa, S., Inoue, S., Watanabe, T., Hiroi, H., Orimo. A, Hosoi, T., Ouchi, Y., and Muramatsu, M. (1998) The complete primary structure of human estrogen receptor beta (hER beta) and its heterodimerization with ER alpha in vivo and in vitro. Biochem. Biophys. Res. Commun. 243, 122-126. (37) Kuiper, G. G., Enmark, E., Pelto-Huikko, M., Nilsson, S., and Gustafsson, J. A. (1996) Cloning of a novel receptor expressed in rat prostate and ovary. Proc. Natl. Acad. Sci. USA. 98, 5925- 5930. (38) Pike, A. C., Brzozowski, A. M., Hubbard, R. E., Bonn, T., Thorsell, A. G., Engstrom, 0., Ljunggren, J ., Gustafsson, J ., and Carlquist, M. (1999) Structure of the ligand-binding domain of oestrogen receptor beta in the presence of a partial agonist and a full antagonist. EMBO J. 18, 4608-4618. (39) Brzozowski, A. M., Pike, A. C., Dauter, Z., Hubbard, R. E.. Bonn, T., Engstrom, 0., Ohman, L., Greene, G. L., Gustafsson, J. A., and Carlquist, M. (1997) Molecular basis of agonism and antago- nism in the oestrogen receptor. Nature 389, 753-758. (40) Tanenbaum, D. M., Wang, Y., Williams, S. P., and Sigler, P. B. (1998) Crystallographic comparison of the estrogen and proges- terone receptor's ligand binding domains. Proc. Natl. Acad. Sci. USA. 95, 5998-6003. (41) Hiroi, H., Tsutsumi, 0., Momoeda, M., Takai, Y., Osuga, Y., and Taketani, Y. (1999) Differential interactions of bisphenol A and 175-estradiol with estrogen receptor a (ERa) and ERB. Endocr. J. 48, 773-778. (42) Christman, J. K., Nehls. 8., Palin, L., and Brooks, S. C. (1995) Relationship between estrogen structure and conformational changes in estrogen receptor/DNA complexes. J. Steroid Biochem. Mol. Biol. 54, 201—210. (43) Pennie, W. D., Aldridge, T. C., and Brooks, A N. (1998) Dif- ferential activation by xenoestrogens of ER alpha and ER beta when linked to different response elements. J. Endocrinol. 158, R11—R14. (44) Paige, L. A, Christensen, D. J., Gran, H., Norris, J. D., Gottlin, E. B., Padilla, K. M., Chang, C. Y., Ballas, L. M., Hamilton, P. T., McDonnell, D. P., and Fowlkes, D. M. (1999) Estrogen receptor (ER) modulators each induce distinct conformational changes in ER alpha and ER beta. Proc. Natl. Acad. Sci. U.S.A. 98, 3999- 4004. (45) Shiau, A. K., Barstad, D., Loris. P. M., Cheng, L., Kushner, P. J ., Agard, D. A., and Greene. G. L. (1998) The structural basis of 153 Estrogenic Activity of BPA and BPAoG (46) (47) (48) (49) (50) (51) (52) (53) estrogen receptor/coactivator recognition and the antagonism of this interaction by tamoxifen. Cell 95, 927-937. Ashby, J ., and Tinwell, H. (1998) Uterotrophic activity of bisphec nol A in the immature rat. Environ. Health Perspect. 106, 719- 720. Ashby, J., Odum, J., Paton, D., Lefevre, P. A., Beresford, N., and Sumpter, J. P. (2000) Re-evaluation of the first synthetic estrogen, 1-keto—1,2,3,4-tetrahydrophenanthrene, and bisphenol A, using both the ovariectomised rat model used in 1933 and additional assays. Toxicol. Lett. 115, 231-238. Long, X, Steinmetz, R., Ben-Jonathan, N., Caperell-Grant, A., Young, P. C., Nephew, K. P., and Bigsby, R. M. (2000) Strain differences in vaginal responses to the xenoestrogen bisphenol A. Environ. Health Perspect. 108, 243—247. Papaconstantinou, A. D., Umbreit, T. H., Fisher, B. R., Goering, P. L., Lappas, N. T., and Brown, K. M. (2000) Bisphenol A-induced increase in uterine weight and alterations in uterine morphology in ovariectomized BSC3F1 mice: role of the estrogen receptor. Toxicol. Sci. 56, 332—339. Safe, S. (2000) Bisphenol A and related endocrine disruptors. Toxicol. Sci. 56, 251-252. 'finwell, H., Joiner, R., Pate, 1., Soames, A., Foster, J., and Ashby, J. (2000) Uterotrophic activity of bisphenol A in the immature mouse. Regal. Toxicol. Pharmacol. 32, 118-126. Coldham, N. 0., Dave, M., Sivapathasundaram, 8., McDonnell, D. P., Connor, 0, and Sauer, M. J. (1997) Evaluation of a recombinant yeast cell estrogen screening assay. Environ. Health Perspect. 105, 734-742. Branham, W. S., Zehr, D. R., and Sheehan, D. M. (1993) Differential sensitivity of rat uterine growth and epithelium hypertrophy to estrogens and antiestrogens. Proc. Soc. Exp. Biol. Med. 203, 297-303. (54) (55 v (56) (57) (58) (59) Chem. Res. Toxicol., Val. 14, Na. 2, 2001 I57 Diel, P., Schulz, T., Smolnikar, K., Strunck, E., Vollmer, G., and Michna, H. (2000) Ability of xeno- and phytoestrogens to modulate expression of estrogen-sensitive genes in rat uterus: estrogenicity profiles and uterotropic activity. J. Steroid Biochem. Mol. Biol. 73, 1-10. ' Nagel, S. C., vom Saal, F. S., Thayer, K. A., Dhar, M. (3., Boechler, M., and Welshons, W. V. (1997) Relative binding affinity-serum modified access (RBA-SMA) assay predicts the relative in viva bioactivity of the xenoestrogens bisphenol A and octylphenol [see comments]. Environ. Health Perspect. 105, 70-76. vom Saal, F. S., Cooke, P. S., Buchanan, D. L., Palanza, P., Thayer, K. A., Nagel, S. C., Parmigiani, S., and Welshons, W. V. (1998) A physiologically based approach to the study of bisphenol A and other estrogenic chemicals on the size of reproductive organs, daily sperm production, and behavior. Toxicol. Ind. Health 14, 239—260. Cagen, S. 2., Waechter, J. M., Jr., Dimond, S. S., Breslin, W. J., Butala, J. H., Jekat, F. W., Joiner, R. L., Shiatsuka, R. N., Veenstra, G. E., and Harris, L. R. ( 1999) Normal reproductive organ development in Wistar rats exposed to bisphenol A in the drinking water. Regal. Toxicol. Pharmacol. 30, 130-139. Cagen. S. 2., Waechter, J. M., Jr., Dimond, S. 8., Breslin, W. J., Butala, J. H., Jekat, F. W., Joiner, R. L., Shiotsuka, R. N., Veenstra, G. E., and Harris, L. R. (1999) Normal reproductive organ development in CF-l mice following prenatal exposure to bisphenol A. Toxicol. Sci. 50, 36-44. Laws, S. 0., Carey, S. A., Ferrell, J. M., Badman, G. J., and Cooper, R. L. (2000) Estrogenic activity of octylphenol, nonylphe- nol, bisphenol A and methoxychlor in rats. Toxicol. Sci. 54, 154- 167. TX0001833 154 CHAPTER 8 Purification and characterization of the rtERa D, E, and F domains 155 Abstract The rainbow trout estrogen receptor 0t (rtERa) and human ERa (hERa) have different functional activities in response to WIS-estradiol (E2). Sequence alignment has identified two amino acid residues that line their respective ligand binding pockets and differ between the rtEROL (Met317, Ile496) and hERa (Leu349, Met528). In the rtERa, these residues have been previously shown to influence E2 binding and transactivation ability. In an effort to design an efficient expression and purification strategy to generate sufficient amounts of protein for further biophysical characterization of the rtERa, a series of recombith proteins consisting of combinations of the D, E, and F domains fused to either glutathione-S-transferase (GST) (GST-rtERa) or a 6x histidine tag (his_rtEROt) were examined. Although both expression systems produced protein with similar activity, the 6x histidine tagged approach resulted in increased yield and purity. In both expression systems the addition of the hydrophilic D domain improved protein solubility, whereas the F domain did not appear to have any effect. After consideration of yield, purity, and activity, the his_rtEROt DE protein was chosen for initial crystallographic screens. Unfortunately, two successive screens have not produced any crystals. A more thorough purification strategy using ion exchange and/or size exclusion chromatography maybe required. These data demonstrate that the D domain can significantly improve the solubility of the rtERa. In addition, we have developed a robust expression and purification strategy that can yield sufficient amounts of protein for biophysical characterization of the rtERa.‘ 156 Introduction Nuclear magnetic resonance and x-ray diffraction studies of proteins are often limited by the amount of purified protein. Performing such studies on native proteins is limited because many proteins of interest are produced at very low levels within the cell. Heterologous expression systems circumvent this problem since they generally provide sufficient levels of proteins for these types of studies. Heterologous systems using Escherichia coli provide several advantages for protein expression since they have relative short doubling times and extensive knowledge exists about their genetic make up, biochemistry, and physiology. All of which can be tailored to a specific expression problem (1),(2). However, many post-translation modifications do not occur in M, which may affect overall protein folding and consequently its function. A comparison of the human 0: (hERa) and rainbow trout 0t rtERa) estrogen receptors indicates that the E domain is 60% identical in amino acid sequence (Chapter 3, figure 3A page 61), whereas the D and F domains are only 18% and 19% identical, respectively. Previous studies and data presented in chapter 6 show that the rtEROL exhibits differential functional activities in response to E2 compared to hERa (3),(4). Sequence alignment has identified two amino acid substitutions within the ligand binding pocket that have been shown in chapter 6 to influence E2 mediated transactivation and the temperature sensitivity of the rtERa. However, in the absence of structural data of the rtERa-EZ complex, the role that these residues (M317 and I496) play in influencing the binding mode of E2 is unknown. It has been reported that in the expression of GST-rtERa (CD) and GST-rtERa (EF) in E. coli nearly 100% of both fusion proteins were insoluble (5). The authors used 157 denaturing conditions to solubilize the protein followed by renaturation using gradual dialysis to verify receptor activity (5). This is in contrast to the data presented in chapter 3 where we describe satisfactory yields of a GST-rtERa DEF fusion protein using native conditions. This suggests that the different domains included in the fusion protein may influence its expression in E_.<_:9_l_i. Several nuclear receptors have been overexpressed in bacteria, purified and their structures determined. Although previous ER proteins were purified and subsequently crystallized using a customized E2-affinity column (6,7), it has been recently reported that his tagged hERa E domain wild-type and mutant proteins were crystallized from metal affinity chromatography purified protein (8). The overall structure of the wild-type hERa E domain was similar to previous reports (8). To date the structure of only the E domains of human and rodent ERs have been solved and these structures do not contain any information regarding the role that the D and F domains may play in influencing ligand binding and/or coactivator interaction. Determining the position of the F domain in the ER crystal structure may be of significant interest since it this region has been shown to influence coactivator recruitment (9), and is also poorly conserved among ERs from several vertebrate species. To examine the potential role of the D and F domains on the solubility and functionality of the rtERa, as well as produce sufficiently pure protein to initiate crystallographic studies of the rtERa, a series of GST and histidine tagged fusion proteins were expressed in bacteria and purified using affinity chromatography in support of future crystallization trials. 158 Materials and Methods Plasmid construction The plasmid pGEX-rtERa DEF harboring the rainbow trout ERa D, E, and F domains (a.a. 225 to 577) was constructed as described in chapter 3. pGEX-rtERa E (a. a. 265 to 521), pGEX-rtERa EF (a. a. 265 to 577), and pGEX-rtERa DE (a. a. 225 to 521) were constructed by PCR amplification of the desired sequences of the rtERa using the following primer combinations rtlf (5'caaaggatccggatggcgtgggcccagaatc -3') and rt2r (5'-caaactcgatggggattggagccggtgacc -3'), rtlf and rt3r (5'- aaaactcgagtcacggaatgggcatctggtctg -3'), and rt4f (5'-aaaaggatcccgggttctcaggaaggataagcg- 3') and rt3r, respectively. The fragments were digested with the appropriate restriction enzymes and ligated into the similarly digested GST expression vector, pGEX6p3 (Amersham Pharmacia). pEThis-rtERa E, pEThis-rtERa EF, pEThis—rtERa DE, and pEThis-rtEROt DEF were constructed in a similar manner using the following PCR primer combinations rt5f and rt2r, rt5f (5'- caaaccatggatccacatcaccatcaccatcacctcaatggtggtggaggatggc -3') and rt3r, rt6f (5'- caaaccatggatccacatcaccatcachcgggttctcaggaaggataagcg -3') and rt2r, and rt6f and rt4r, respectively. For the construction of the pEThis-rtERa expression vectors, the 6x histidine sequence, which has been underlined above, was incorporated into the forward PCR primers creating an NHz-terminal 6x histidine protein. The PCR fragments were digested with the appropriate restriction enzymes and ligated into the similarly digested pET23d(+) expression vector (Invitrogen), which contains a MDP leader sequence as described previously (7). 159 Expression and purification of GST-rtERa fusion proteins L021 strain BL21 (Amersham/Pharmacia) containing pGEX-rtERa constructs were grown in 2 x Luria broth (LB) and incubated at 30°C with constant shaking. Cells were induced with 1 mM IPT G at an optical density of 0.8 at 600 nm and grown for 4 h at 25°C. After centrifugation, cell pellets were resuspended in 15 ml of buffer A (50 mM HEPES, 3 mM EDTA, 5 mM DTT, 50 mM NaCl and 10% (v/v) glycerol, pH 7.5) containing 0.1 mg/ml lysozyme, 200 U of Dnase I, and protease inhibitors. Cells were then lysed by sonication on ice and cell debris was pelleted by centrifugation at 20,000xg for 30 min at 4°C. The pellet was resuspended a second time and the supematants pooled. The supematants containing the GST fusion proteins were incubated in batch with GSH Sepharose pre-equilibrated with buffer A (1 ml/15 ml supernatant) for 1.5 h at 4°C. After adsorption of the protein, the GSH Sepharose was applied to a 20 ml disposable column and washed with 50x bed volume of buffer A (50 mM HEPES, 3 mM EDTA, 5 mM DTT, 200 mM NaCl and 10% (v/v) glycerol, pH 7.5), 50x bed volume of buffer B (50 mM HEPES, 3 mM EDTA, 5 mM DTT, 1 M NaCl and 10% (v/v) glycerol, pH 7.5), and finally 10x bed volume of buffer A. Bound proteins were eluted in 25 ml of buffer C (50 mM HEPES, 3 mM EDTA, 5 mM DTT, 200 mM NaCl and 10% (v/v) glycerol, pH 8.0) containing 20 mM GSH. The partially purified protein was concentrated to a 1 m1 final volume and the protein concentration was determined using the BioRad protein assay (BioRad). Expression and purification of 6x histagged rtERa fusion proteins 160 E._co_li strain BL21(DE3)pLysS (Invitrogen) containing pEThis-rtEROt constructs were grown in 2x LB and incubated at 30°C with constant shaking. Cells were induced with 0.5 mM IPT G at an optical density of 0.8 at 600 nm and grown for 3 h at 25°C. After centrifugation, the pellets were resuspended in 20 ml of lysis buffer (20 mM Tris, 300 mM NaCl, 2 mM MgC12, 5 mM imidazole, 100 ug/ml PMSF, 5 mM [3- mercaptoethanol, 200 U Dnase I, and 1 mg/ml lysozyme). Cells were then lysed by sonication on ice and cell debris was pelleted by centrifugation at 20,000xg for 30 min at 4°C. The pellet was resuspended a second time and the supematants pooled. The supematants containing the his_rtEROLs were incubated in batch with nickel nitrilotriacetic acid resin (Qiagen) pre-equilibrated with resuspension buffer (1 ml/4O ml supernatant) for 1.5 h at 4°C. After adsorption of the protein, the nickel affinity resin was applied to a 20 ml disposable column and washed with 50x bed volume of wash I (20 mM Tris, 500 mM NaCl, 5 mM imidazole, 5 mM B-mercaptoethanol), 50x bed volume of wash II (20 mM Tris, l M NaCl, 20 mM imidazole, 5 mM B-mercaptoethanol), and finally 25x bed volume of wash III (20 mM Tris, 300 mM NaCl, 20 mM imidazole, 5 mM B-mercaptoethanol). Bound proteins were eluted in elution buffer (20 mM Tris, 300 mM NaCl, 5 mM imidazole, 100 mM EDTA) and collected as 1 m1 fractions every 10 min. The protein concentration of each fraction was determined and fractions containing protein were pooled and the final concentration was determined using the BioRad protein assay (BioRad). For the purification of the his_rtEROt DE to be used in subsequent crystallization trials, 20 uM E2 was added to all buffers during the expression and purification steps. In addition, the protein was carboxymethylated in situ in the presence of 15 mM iodoacetic 161 acid in was IH, followed by a 5x bed volume wash with wash III. The protein was then eluted as described above. The protein was dialyzed overnight against (20 mM Tris pH 7.4, 150 mM NaCl, 3 mM EDTA, 20 M E2), concentrated to 7-10 mg/ml and stored at — 80°C. Receptor binding assays Aliquots of crude and partially purified GST-rtERa and his_rtEROt fusion proteins Were diluted in TEGD buffer (10 mM Tris pH 7.6, 1.5 mM EDTA, 1 mM DTT and 10% (v/v) glycerol) containing 1 mg/ml bovine serum albumin as a carrier protein. The reactions were incubated 4°C for 24 h with a saturating amount of [3H]E2 (3 nM) in 1 ml glass tubes arranged in a 96-well format (Marsh Scientific, Rochester, NY). Binding assays were initiated by adding 240 u] of protein preparation to glass tubes containing 5 pl of DMSO and 5 ul [3H]E2; thus the concentration of solvent did not exceed 4 %. Bound [3H]E2 was separated from free using a 96-well filter plate and vacuum pump harvester (Packard Instruments). After drying, the filter plates were sealed and 50 ul of MicroScint 20 scintillation cocktail (Packard Instruments) was added to each well. Bound [3H]E2 was measured using a TopCount scintillation and luminescence counter (Packard Instruments). Western Blotting Purified GST-hERadef samples (250 ng) were fractionated on a 10% polyacrylamide/SDS gel and transferred to nitrocellulose (Amersham/Pharmacia, Piscataway, NJ). Blots were incubated with anti-GST antibody (Santa Cruz, Santa Cruz, 162 CA) or anti-GroEL (Dr. Jon Kaguni, Department of Biochemistry and Molecular Biology, Michigan State University, East Lansing, Michigan) for 2 h at room temperature in the presence of 2% fat-free milk powder and then incubated for 2 h with the secondary conjugated goat anti-(mouse IgG) or mouse anti-(rabbit)-horseradish peroxidase (Santa Cruz). Proteins were visualized using ECL detection system (Amersham/Pharmacia) and subsequent exposure to film. 163 Results Expression and purification of the GST-rtERa fusion proteins The GST-rtERa fusion proteins mi grated according to their predicted molecular weights (MWs): GST-rtERa E (55.6 kDa), GST-rtERoc EF (61.3 kDa), GST-nERa DE (60.0 kDa), and GST-rtERa DEF protein (65.7 kDa) (Figure 1). In addition, higher and lower MW proteins co-purified with the fusion proteins. Western blot analysis of partially purified GST-hERocDEF revealed that the contaminating band at approximately 60 kDa represents the bacterial chaperon protein GroEL (figure 2). GroEL has been reported to be a common copurifying contaminant of GST fusion proteins (10). The purity of the GST-rtERa fusion proteins varied among protein preparations and was estimated to be 60%-80%. The yields also varied, with GST-rtERoc E, GST-rtERa EF, GST-rtERa DE, and GST-rtERa DEF yielding approximately 0.4, 2.5, 4.5, and 3.5 mg partially purified protein/L culture, respectively. In addition, only the GST-rtEROL DE and GST-rtERa DEF proteins were visible in the crude proteins extracts, which have been denoted by an asteriks in Figure 1. Receptor binding assays were performed on the crude and purified samples for each of the fusion proteins to evaluate the activity of the respective proteins. After normalization for protein concentration, an aliquot of the crude as well as the purified protein samples were diluted in TEGD buffer containing 1 mg/ml BSA and incubated in the presence of 3 nM [3H]E2. The results in figure 3 demonstrate that the inclusion of the D domain had relatively little effect on activity, though it increased protein yield. 164 kDa123456 “33123456 kDa 56— 'f :55 'V w 43— a!" 36— 3"” «'7 ‘ I.“ '0‘ «mg-per: a.” 27— g g; .. 27.: ii: Figure 1. SDS-PAGE analysis of GST-rtERoc purification. GST fusion proteins (A) GST-rtERa E, (B) GST—rtERa EF, (C) GST-rtERa DE, and (D) GST- rtEROL DEF were expressed and purified as described in the Materials and Methods section. Various fractions were separated by 12% SDS-PAGE and visualized by coomassie blue staining. Lane 1: 20 pg of crude lysate, lane 2: 20 pg of flow through, lane 3: 10 pl of 10x concentrated wash I, lane 4: 10 pl of 10x concentrated wash II, lane 5: 10 pl of 10x concentrated wash 1]], lane 6: 3 pg of purified GST-rtERa protein. The asteriks identify the GST-rtERa fusion protein in the crude lysate. 165 Coomassie Stain Western Blot Figure 2. GroEL co—purifies with GST-ERdef fusion proteins. GST-hERadef was extracted and purified as described in the Materials and Methods section of Chapter 3. Coomassie stain of purified GST-hERadef (left panel) indicated two distinct bands were separated by 10% SDS-PAGE. After subsquent western blot analysis (right panel) the upper band was identified as GST-hERoodef, while the lower band was determined to be GroEL, a common contaminant of GST fusion protein preparations. 166 109 .S O) E 108 a. 01) g 107 Q -o 30 106 .—1 1os¢zfi¢§ @QQIQ ©Q©Q QQQQ $¢°2°°¢°¢ §°e9¢239¢°® §6e°@«3‘°¢°¢ geéé‘oée we.» a 4' «615” a 9 48' a 0 c? 0904‘ 0 0" 0909’“ r e” $3“ 0 «yaw H4 Crude Purified Crude Purified Figure 3. Comparison of the activity of the GST-rtERa and his_rtEROt proteins. Aliquots of the crude and purified proteins were diluted in TEGD buffer containing 1 mg/ml bovine serum albumin and incubated for 16-18 h at 4°C as described in the Materials and Methods section. 167 One of the complications of using a GST fusion protein is the potential that the addition of the GST moiety may affect the activity of the protein and as a result it is generally removed by proteolytic cleavage. The GST-rtERoc fusion proteins contain an engineered Precision protease site upstream of the rtER; however after repeated attempts with the GST-rtEROLDEF and GST-hEROLDEF proteins only marginal cleavage was observed (data not shown). Expression and purification of his_rtER proteins As an alternative to GST tagged proteins, the rtERa was also expressed as a 6x histidine tagged (his) protein and purified with nickel affinity chromatography. This approach proved superior to the GST—rtEROL method in terms of overall protein expression, and purity, whereas similar levels of functional protein were observed. The results in figure 4 show that the his_rtEROt E (30.0 kDa), his_rtERa EF (35.6 kDa), his_rtERoc DE (34.4 kDa), and his_rtERa DEF (40.0 kDa) proteins migrated according to their predicted MWs with their purity ranging from 80-90%. However, there were significant contaminating bands in the his_rtEROt EF preparation. Lower bands were also apparent in the his_rtERa DEF preparations, which most likely represent products of proteolytic cleavage, though they are not easily seen on the scanned gel presented in figure 4. Similar to that observed with the GST-rtERa proteins, the inclusion of the D domain dramatically increased protein yields with the expression of his_rtERa E, his_rtEROt EF, his_rtERa DE, and his_rtERa DEF resulting in yields of 2.0, 2.0, 7.5, and 1 1 mg partially purified protein/L culture, respectively. A comparison of the crude 168 A B kDa 1 2 3 4 5 6 kDa 1 2 3 4 5 6 66— I. '1' . . 66 _> 43— 43 __,_ . 36— .A . 36 — ‘ kDa 1 2 3 4 5 6 kDa .._ w .._ 43— 43 _ 36— p q .. .- 3: _ .f f l ’l! we: Figure 4. SDS-PAGE analysis of his—rtER purification. The his_rtER proteins (A) his_rtER E, (B) his_rtER EF, (C) his_rtER DE, and (D) his_rtER DEF were expressed and purified as described in the Materials and Methods section. Various fractions were separated by 12% SDS-PAGE and visualized by coomassie blue staining. Lane 1: 50 pg of crude lysate, lane 2: 50 pg of flow through, lane 3: 10 pl of 10x concentrated wash 1, lane 4: 10 pl of 10x concentrated wash 11, lane 5: 10 pl of 10x concentrated wash III, lane 6: 3 pg of purified his_rtER protein. The asteriks identifies the his_rtER protein in the crude lysate. 169 Figure 5. Influence of the D domain on the solubility of his_rtERa. Extraction of his_rtERa proteins from BL21 (DE3) cell pellets was performed as described in the Materials and Methods section. Equal amounts of crude protein extracts (30 pg) were separated by 12% SDS-PAGE and visualized by coomassie blue staining. 170 extracts from each of the four his_rtERoc proteins shown in figure 5 clearly demonstrates the effect of the D domain on rtERa solubility. In figure 5, both the his_rtERoc DE and his_rtERa DEF are visible in the crude extracts, while neither his_rtERa E nor his_rtEROL EF were visible. Receptor binding assays were used to assess the activity of the his_rtERa proteins and despite the increased solubility of the his_rtEROt DE and the his_rtEROL DEF both proteins were approximately 33% as active as the his_rtERa EF, but exhibited similar activity to the his_rtERa E protein (Figure 3). Based on expression, purity, and functionality, the his_rtEch DE was chosen as a candidate for initial crystallization trials. To ensure that the rtERa DE was saturated with E2, 20 pM E2 was added to all buffers throughout the expression and purification steps outlined in the Materials and Methods section. Two different crystal screens done in collaboration with Dr. Jim Geiger (MSU Departement of Chemistry) using the Magic buffer set and an incubation temperature of 16°C have failed to produce any crystals. 171 Discussion Although the GST-rtERa and his_rtERa fusion proteins exhibit comparable ability to bind E2, there were marked differences in the purity and yield between both methods. The histidine tagged system was found to be superior to using GST as an affinity tag. Reasons for these differences may include the types of promoters that regulate the recombinant protein. The GST-rtERa fusion proteins are under the control of the ptac promoter, whereas the T7 promoter regulates his_rtERa protein expression. The GST moiety may inadvertently affect folding of the ER, thus leading to a larger proportion of insoluble products. Previous studies examining GST-rtERoc CD and GST- rtEROL EF constructs reported that nearly 100% of the expressed protein was insoluble (5). The inclusion of the D domain to both GST and histidine tagged proteins significantly improved the yield by increasing the population of soluble protein. This was clearly evident in a comparison among the crude extracts from the his_rtEROL proteins. Analysis of the amino acid sequence of the rtERCX. D domain (Chapter 2, figure 12 page 31) indicates that there are several charged hydrophilic residues that may influence solubility. The D domain is considered to be relatively flexible since it is thought to behave as a hinge, bridging the DNA binding and ligand binding domains ‘ together (1 1-13). This hydrophilic D domain may behave like a sequence tag that promotes ER solubility. In parallel studies in our laboratory a similar, albeit less dramatic effect, was observed with hEROL. 172 Several nuclear receptors have been crystallized after expression and purification as recombinant fusion proteins in ELoli (14,15). The androgen receptor has been crystallized following expression and purification as GST fusions as well as histidine tagged proteins ( 16,17). Both moieties were proteolytically cleaved prior to concentration and crystallization (16,17 ). Since previous structural studies of the ER focused on the E domain, the his_rtEROt E domain was initially selected for purification with the intent of x-ray crystallization. Subsequent analysis revealed that the his_rtEROL E domain formed a precipitate during prolonged incubation at 4°C and/or when the protein was concentrated above 3 mg/ml. This solubility problem was not encountered with the his_rtEROt DE construct and because of its increased expression and purity, the his_rtEROL DE was chosen for initial crystallographic screens. In general one step purifications with the intent of crystallization are relative uncommon; however, the retinoic X receptor (RXR) DE has been successfully crystallized in this manner (18). In this study, we describe a robust expression system that uses commercially available components and does not require specialized expression vectors (2). These data demonstrate that histidine tagged expression system is more suitable for the expression and purification of the rtERa than a GST based expression strategy. Overall protein yields and purity were best using the his_rtERa expression strategy. The hydrophilic D domain significantly improves solubility of the rtERa yielding 90-95% pure protein following a single step purification using metal affinity chromatography. Although this strategy has yet to produce any crystals, these screens are in their earlier stages and with a more thorough purification methodology may ultimate lead to the formation of crystals suitable for x-ray diffraction. In addition, an E2-affinity purification strategy, which has 173 been successfully used by several different groups to crystallize the hERa (19-21), could be used as an alternative to metal affinity chromatography. 174 10. 11. 12. 13. 14. 15. References Baneyx, F. (1999) Curr Opin Biotechnol 10, 411-421. Nygaard, F. B., and Harlow, K. W. (2001) Protein Expr Purif 21, 500-509. Le Drean, Y., Kern, L., Pakdel, F., and Valotaire, Y. (1995) Mol. Cell. Endocrinol. 109, 27-35. Pakdel, F., Le Gac, F., Le Goff, P., and Valotaire, Y. (1990) Mol Cell Endocrinol 71, 195-204. Pakdel, F., Petit, F., Anglade, I., Kah, 0., Delaunay, F., Bailhache, T., Valotaire, Y. (1994) Mol Cell Endocrinol 104, 81-93. Greene, G. L., Nolan, C., Engler, J.P., Jensen, E.V. (1980) Proc Natl Acad Sci U SA 77, 5115-5119. Seielstad, D. A., Carlson, K.E., Katzenellenbogen, J.A., Kushner, P.J., Greene, G.L. (1995) Molecular Endocrinology 9, 647-658. Eiler, S., Gangloff, M., Duclaud, S., Moras, D., and Ruff, M. (2001) Protein Expr Purif22, 165-173. Warnmark, A., Almlof, T., Leers, J., Gustafsson, J. A., and Treuter, E. (2001) J Biol Chem 276, 23397-23404. Rohman, M., and Harrison-Lavoie, K. J. (2000) Protein Expr Purif 20, 45—47. Giguere, V. (1999) Endocr Rev 20, 689-725. Zechel, C., Shen, X. Q., Chen, J. Y., Chen, Z. P., Chambon, P., and Gronemeyer, H. (1994) Embo J 13, 1425-1433. Zechel, C., Shen, X. Q., Chambon, P., and Gronemeyer, H. (1994) Embo J 13, 1414-1424. Watkins, R. E., Wisely, G. B., Moore, L. B., Collins, J. L., Lambert, M. H., Williams, S. P., Willson, T. M., Kliewer, S. A., and Redinbo, M. R. (2001) Science 292, 2329-2333. Bourguet, W., Ruff, M., Bonnier, D., Granger, F., Boeglin, M., Chambon, P., Moras, D., and Gronemeyer, H. (1995) Protein Expr Purif 6, 604-608. 175 l6. 17. 18. 19. 20. 21. Matias, P. M., Donner, P., Coelho, R., Thomaz, M., Peixoto, C., Macedo, S., Otto, N., Joschko, S., Scholz, P., Wegg, A., Basler, S., Schafer, M., Egner, U., and Carrondo, M. A. (2000) J Biol Chem 275, 26164-26171. Sack, J. S., Kish, K. R, Wang, C., Attar, R. M., Kiefer, S. E., An, Y., Wu, G. Y., Scheffler, J. E., Salvati, M. E., Krystek, S. R., Jr., Weinmann, R., and Einspahr, H. M. (2001) Proc Natl Acad Sci U S A 98, 4904-4909. Bourguet, W., Ruff, M., Chambon, P., Gronemeyer, H., and Moras, D. (1995) Nature 375, 377-382. Brzozowski, A. M., Pike, A. C., Dauter, Z., Hubbard, R. E., Bonn, T., Engstrom, O., Ohman, L., Greene, G. L., Gustafsson, J. A., and Carlquist, M. (1997) Nature 389, 753-758. Shiau, A. K., Barstad, D., Loria, P.M., Cheng, L., Kushner, P.J., Agard, D.A., Greene, G.L. (1998) Cell 95, 927-937. Goldstein, S. W., Bordner, J ., Hoth, L. R., and Geoghegan, K. F. (2001) Bioconjug Chem 12, 406-413. 176 CHAPTER 9 Conclusions and Future Directions 177 Conclusions The data described above support our hypothesis that ERs among different vertebrate species exhibit differential binding affinity and in vitro gene expression responses to EEDs due to sequence divergence within their ligand binding domains. Although there were no obvious relationships between sequence identity and binding affinity and transactivation ability, (low sequence identity was not predictive of binding affinity of a compound), ERs with greater sequence identity exhibited similar relative binding affinities and transactivation abilities. The rtEROt and mERB, which have relatively low identity (60%) compared to hERa, exhibited the greatest differences among the ERs examined. Multiple sequence alignment identified amino acid residues that line the hormone binding pockets and differ among the vertebrate ERs examined. Taken together, this cross species comparison provides valuable insights into potentially important residues that may play critical roles in the interaction between structurally diverse ligands and the ER binding pocket. Although this study has identified several examples of EEDs exhibiting preferential binding affinity or ability to induce gene expression mediated by a specific vertebrate ER, the differences were relatively subtle and were rarely greater than an order of magnitude. This suggests the use of a single vertebrate ER to examine or screen for suspected EEDs is feasible, since the species- specific differences in binding and transactivation response of EEDs were relatively small. 178 t 'l ‘ .‘-I'L\.I J t .,. "3.1! . Future Directions 1. Since, ERB exhibits significantly greater affinity for EEDs, especially the phytoestrogens when compared to EROL. A comparison of the binding and transactivation profiles of EEDs for ERBs from representative vertebrate ERs would be an exciting continuation of the data presented above. ERBs exhibit significant amino acid sequence variability among vertebrate species. It would be interesting to examine whether vertebrate ERBs respond equally well to phytoestrogens and not just the mammalian ERBs 2. During the course of the experiments described in chapters 3 & 4 we noted that the rtERa exhibits a temperature sensitive phenotype that was not seen in the other species examined. Although lowering the temperature significantly improved the function of the rtERa, an approximate lO-fold lower transactivation response compared to hERa was still observed. Using a reciprocal mutagenesis strategy, two amino acid residues (Met317, Ile496) were found to contribute not only to differences in binding and transactivation between hERa and rtEROL but also to the reduced activity of the rtERa at elevated temperatures. The biological relevance of the rtERa exhibiting reduced function at elevated temperature is unclear, but may involve increased flexibility in the binding pocket of the rtERa compared to hERa. that causes the rtERa to become unstable at elevated temperatures. In the absence of structural data the role these residues play influencing the position of E2 in the binding pocket is unclear. Similarly, the transactivation ability of another piscine ER, the Oreochromis aureus (OaER), also 179 exhibits reduced activity at temperatures above its normal physiological range. Interestingly, exchanging a 35 amino acid stretch ([3 region), between the OaER and the chicken ER (cER) partially rescued the thermal deficient transactivation of the OaER. Therefore, combining the reciprocal mutagenesis described above with the exchange of the [3 region between rtERoc and hEROL may result in complete phenotypic reversal between the two receptors. 3. A robust purification strategy was established to produce sufficient quantities of rtEROL for further biophysical characterization. After comparing a series of different rtEROt recombinant proteins his_rtEROt DE was selected for initial crystallographic screens. Although the crystal screens have not produced any crystals, more extensive screens with different buffer conditions may enhance our success. Proteolytic removal of the 6x histidine tag combined with a more extensive purification strategy including ion exchange and/or size exclusion chromatography may also increase the probability of success. In crystallization studies of histidine tagged hERa E domain, only protein subjected to ion exchange gave crystals suitable for diffraction, while protein purified without ion exchange did not (3). Solving the crystal structure of the rtERa represents an important research challenge and hopefully the data presented here will contribute towards this goal. 180 References Waller, C. L., Minor, D. L., and McKinney, J. D. (1995) Environ Health Perspect 103(7-8), 702-707. Tong, W., Perkins, R., Strelitz, R., Collantes, E.R., Keenan, 8., Welsh, W.J., Branham, W.S., Sheehan, D.M. (1997) Environ Health Perspect 105(10), 1116- 1 124. Eiler, S., Gangloff, M., Duclaud, S., Moras, D., and Ruff, M. (2001) Protein Expr Purif 22(2), 165-173. 181 lllllllllllllljlljllllllllll’1111111211111