———_—————_—_ 7L.- by: 531 I9 4., ‘ flu wnheu * News SITYLIB RRA RIES lllllllllllll!llllHlllllllllllllIlUIHIHHIllllllllll 3 1293 00914 0991 This is to certify that the dissertation entitled HORMONAL REGULATION W THE 814 GENE IN 3T3-F442A ADIPCXIYTEB presented by Gerald Joseph lepar has been accepted towards fulfillment of the requirements for Pho D degree in PhYSj—Olwy iggm Major professor Date A/Oy. 3; /790 MS U is an Affirmative Action/Equal Opportunity Institution 0-12771 ..——-—— -q—-—-—————-.——__—i..- 4—— _..._. —_ FL. “ KARI ”kWh State l University A ‘— PLACE IN RETURN BOX to remove this checkout from your record. TO AVOID FINES return on or before date due. DATE DUE DATE DUE DATE DUE MSU Is An Affirmative ActlorVEqual Opportunity Institution cidrchrna-pd HORMONAL REGULATION OF THE 814 GENE IN 3T3-F442A ADIPOCYTES By Gerald Joseph Leper B.S., University of Illinois, 1979 M.S., University of Illinois, 1982 A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Physiology 1990 ABSTRACT HORMONRL REGULATION OF ADIPOCYTE GENE EXPRESSION BY Gerald Joseph Lepar Expression of the 814 gene is regulated by thyroid hormone (T3) and glucocorticoids (DEX) in adult mouse liver and white adipose tissue (WAT). The primary goal of this dissertation was to determine the utility of the adipocyte-forming 3T3—F442A. cell line as an ip vitro model to define regulation and function of the 814 protein (17,000 Mr; 4.9 pI). Fibroblast (preadipocyte) and adipocyte cultures of 3T3-F442A cells lack basal 814 expression. Fibroblasts also lack mRNA314 induction by T3, DEX, and retinoic acid (RA). In contrast, mRNASl4 is induced by DEX and RA, but not T3, in 3T3-F442A adipocytes. The lack of T3-stimulated accumulation of mRNASl4 is paradoxical since other gene products are responsive to T3 in 3T3-F442A adipocytes. This apparent T3 resistance may be due to: l) 70% reduction of T3 receptor (TR) levels in 3T3-F442A adipocytes compared to rat liver and WAT; 2) high levels of expression of the c-erbAa2 (TR) isoform, a non-T3 binding isoform; or 3) other cellular factors involved in T3 control of $14 expression. Retinoic acid and DEX stimulate mRNASl4 (1.12 kilobases) 37-fold and 266-fold, respectively, while DEX plus RA induce a synergistic 3200-fold increase after 72 h. Glucocorticoid effects on mRNASl4 are detectable 30 min after stimulation and increase to le-fold induction by 4 h, while RA plus DEX induce a zSO-fold increase in mRNASl4 after 4 h. Kinetic, dose-response, and analog studies indicate both DEX and RA effects on mRNASl4 are receptor mediated. lg yitrg nuclear transcription run-on assays show these effects are directed primarily at the transcriptional level. The presence of RA receptor alpha transcripts implicates this isoform in mediating retinoic acid effects. Transfection analysis shows that developmentally and glucocorticoid-activated trans-acting factors interact with DNA sequences within -4316 kilobases of the start site for $14 gene transcription, while RA actions are mediated by different sequences. Epinephrine and the cAMP analog, 8-CPT CAMP, rapidly inhibit glucocorticoid-stimulated increases of mRNASl4 (t1/2=90 min) in 3T3-F442A cells. Preliminary studies to correlate function of the 814 protein with lipogenesis or lipolysis in 3T3-F442A cells were inconclusive. These results demonstrate for the first time, that in 3T3-F442A cells: 1) 814 gene expression is dependent on the differentiation state; 2) the 814 protein is unnecessary for basal lipid metabolism; 3) only the glucococorticoid 814 regulatory mechanism is operative compared with DEX and T3 in 2122; 4) T3 regulation is selective; 5) RA influences lipid metabolism; and 6) a-kinase stimulated effects are dominant over glucocorticoid effects on 814 gene expression. DEDICATION To Becky, for whose love, patience, and knowledge this would not have been possible or meaningful. To Danielle, Katherine, and Achilles, who helped foster a healthy perspective throughout this effort. To Mom and Dad, for your love, support, and patience, particularly during holiday meals when you would dutifully ask "so, tell us again, what exactly is it that you do iv ACKNOWLEDGEMENTS The amount of support I received from my wife, Dr. Rebecca R. Sandborg, probably should entitle her to an honorary degree. Thank-you, Becky, for all the personal and professional sacrifices you have made for me over the past several years. I would like to thank Dr. Donald B. Jump for his guidance, and the opportunity to complete this work in his laboratory. I would also like to thank the other members of my committee: Dr. Stephen Heidemann requires a special thank-you for the generous use of many essential pieces of equipment and his open and honest discussions; Dr. Dale R. Romsos provided many useful insights and suggestions on many aspects of this dissertation; and Drs. William S. Spielman and Birgit Zipser for their assistance during the completion of this work. There were several other members in the laboratory during my "tour of duty." Each one of these people contributed in their own way to enriching my life and making this completed work possible. These people include Aneka Bell, Larry Herberholz, Mary Peterson, Vivian Santiago, and Andrew Veit. I would also like to thank Ormond MacDougald for providing some relief under the magnifying glass as well as for all the stimulating and enjoyable conversations. I wish all these people the success and happiness they deserve. Of course, it is never possible to acknowledge all the people that contribute to an individuals success and happiness, but, I would like to mention several other people in the Department of Physiology who I was fortunate enough to know. These individuals included Dr. Gregory Adams, Dr. Mary Lynn Bajt, Esther Brenke, Robert Cole, Dr. James G. Cunningham, Dr. Tim Dennerll, Dr. Cheryl Killingsworth, Mary Jane Rice, Vivian Steele, and Jing Zheng. I would also like to acknowledge Dr. William Helferich, from the Department of Food Science and Human Nutrition, for his interest in many aspects of my career at Michigan State University. I would like to thank Dr. B. Spiegelman for providing us with the 3T3-F442A cells and Drs. Chambon, Cleveland, Kedes, Lomedico, Pfahl, Spiegelman, and Towle for the recombinant probes used in this study. Dr. Donald Jump also performed the 13 vitro nuclear transcription run-on assay performed in Chapter 5. This work was supported by the Department of Physiology (MSU), College of Natural Sciences (MSU), and NIH Grant GM 36851. EASE: OF CONTENTS List of Tables.........................................vii List of Figures........................................viii List of Symbols, Abbreviations, and Nomenclature.......xi Introduction...........................................1 Chapter 1. Literature Review A. Initiation of Gene Transcription ...... 5 B. Hormonal Control of Gene Expression...10 C. 314 Background ........................ 21 1. Tissue-specific regulation ......... 25 2. Developmental regulation...........29 3. Hormonal regulation................30 4. Circadian expression...............34 D. Significance..........................35 Chapter 2. Expression of the 814 Gene and Functional Analysis of the 814 Protein in 3T3-F442A Adipocytes A. Introduction..........................37 B. Materials and Methods.................40 C. Results...............................46 D. Discussion ..... ....... ..... ...........63 Chapter 3. Glucocorticoid Regulation of 814 Gene Expression in 3T3-F442A Cells A. Introduction... ..... .... ............ ..69 B. Materials and Methods.................70 C. Results...............................78 D. Discussion............................93 Chapter 4. Lack of Thyroid Hormone (T3) Control of 814 Gene Expression in 3T3-F442A Cells A. Introduction..........................101 B. Materials and Methods....... ....... ...102 C. Results....... ..... ...................106 D. Discussion................. ........ ...119 Chapter 5. Retinoic Acid and Dexamethasone Interact to Regulate 814 Gene Transcription in 3T3-F442A Adipocytes A. Introduction..........................124 B. Materials and Methods.................126 C. Results...............................128 D. Discussion............................144 Chapter 6. Summary and Conclusions....................152 Bibliography...........................................163 vi Table 1. IST OF TABLES Relative Abundance of mRNAs in 3T3-F442A Fibroblast and Adipocytes vii 50 Figure 1. Figure 2. Figure 3. Figure 4. Figure 5. Figure 6. Figure 7. Figure 8. Figure 9. Figure 10. LIST OF FIGURES General Diagram Illustrating Hormonal Control of Physiological Functions Control of Gene Expression and Protein Synthesis General Structural and Functional Diagram of the Steroid and Thyroid Hormone Receptors Structural Organization and Characterization of the Rat 814 Gene, mRNA, and Protein Multifactorial Control of the 814 Gene Relative Expression of mRNA514 in Mouse Liver cAMP-Mediated Inhibition of mRNA314 and mRNAB-actin in Adipocytes Northern Blot Analysis of mRNA314 from Mouse and Rat Liver and from 3T3-F442A Cells [3H120-Incorporation as a Measure of Lipogenesis in Adipocytes Glycerol Release as a Measure of Lipolysis in Adipocytes viii 18 24 27 48 54 57 60 62 Figure 11. Structure of the Plasmid Vector 72 pSl4-CAT-4.3 Figure 12. Kinetics of Dexamethasone Induction of 80 mRNA514 in Adipocytes Figure 13. Decline in mRNA514 Following 83 Dexamethasone Removal in Adipocytes Figure 14. Dose-Response Relationship for 85 Dexamethasone Induction of mRNA314 in Adipocytes Figure 15. Steroid Analog Specificity of mRNA314 88 Induction in Adipocytes Figure 16. In Vitro Transcriptional run-on 91 activity of the 814 Gene in Adipocytes Figure 1?. Developmental and Dexamethasone Control 95 of mRNA514 Abundance and the CAT-Fusion Gene Containing 5'-Flanking Sequences of the 814 Gene in Adipocytes Figure 18. Activation of $14 Promoter Activity by 98 Development and Dexamethasone Figure 19. Scatchard Analysis of T3-Binding in 108 Mouse Liver and 3T3-F442A Adipocytes Figure 20. T3 and Dexamethasone Effects on 111 Protein Biosynthesis in Adipocytes Figure 21. Northern Blot Analysis of Thyroid 114 Hormone Receptors Isoforms in Rat Liver and 3T3-F442A Adipocytes ix Figure Figure Figure Figure Figure Figure Figure Figure 22. 23. 24. 25. 26. 27. 28. 29. T3 and Dexamethasone Control of mRNA514 Abundance and the CAT-Fusion Gene Containing 5’-F1anking Sequences of the 814 Gene in Adipocytes Retinoic Acid and Dexamethasone Interact to Regulate mRNA314 Abundance in Adipocytes Dose-Response Relationship for Retinoic Acid Induction of mRNA514 in Adipocytes Kinetics of Retinoic Acid and Dexamethasone Induction of mRNA514 in Adipocytes Dexamethasone and Retinoic Acid Interact to Regulate 814 Gene Transcription in Adipocytes Northern Blot Analysis of Retinoic Acid Receptor Isoforms in Rat Liver and 3T3-F442A Adipocytes Retinoic Acid and Dexamethasone Control of mRNA314 and the Cat-Fusion Gene Containing 5’-Flanking Sequences of the 814 Gene in Adipocytes Developmental and Hormonal Control of The 814 Gene in 3T3-F442A Cells 118 130 134 136 139 142 146 162 LIST OF SYMBOLS. ABBREVIATIONS; OR NOMENCLATURE A250 A280 ATP bp CAT pCAT(AN) c-erbAal c-erbABl c-erbABz CTP absorbance at wavelength 260 nanometers absorbance at wavelength 280 nanometers adenosine triphosphate base-pair degrees centigrade calcium chloride cyclic adenosine 3',5’-monophosphate chloramphenicol acetyltransferase promoterless CAT reporter plasmid cellular alpha one isoform of the thyroid hormone receptor homologous to v-erbA cellular alpha two isoform of the thyroid hormone receptor homologous to v—erbA cellular beta one isoform of the thyroid hormone receptor homologous to v-erbA cellular beta two isoform of the thyroid hormone receptor homologous to v-erbA cytidine triphosphate xi 8-CPT CAMP DEX DM DMEM DOTMA DNA DNase I DTT EDso EDTA EGTA EtOH ER FAS GPD GR GRE GTP HEPES Hss i.p. 8-(4-chloropheny1thio)-cyclic adenosine 3',5'-monophosphate dexamethasone differentiation medium Dulbecco’s modified eagles medium N-[l-(2,3-dioleylosy)propy1]-N,N,N- trimethylammonium chloride deoxyribonucleic acid deoxyribonuclease I (DNA endonuclease) dithiothreitol 50% effective dose ethylenediamine-tetraacetic acid [ethylenebis(oxyethylenenitrilo)]tetra- acetic acid ethanol estrogen receptor fatty acid synthase gram glycerol phosphate dehydrogenase glucocorticoid receptor ' glucocorticoid response element guanosine triphosphate hour 4-(2-hydroxyethyl)-1-piperazine ethane sulfonic acid hypersensitive site intraperitoneal xii KCl kD mCi MgClz mg min m1 MOPS mRNA mRNA514 MW n NaCl NaOH pEMBLSl4-l3 pI kilobase potassium chloride kilodalton dissociation constant liter molar millicurie magnesium chloride milligram minute milliliter millimeter millimolar maintenance medium 3-(N-morpholino)propanesulfonic acid relative molecular weight messenger ribonucleic acid mature mRNA transcribed from the 814 gene molecular weight number of samples sodium chloride sodium hydroxide plasmid containing genomic 814 DNA extending from +1.7 to -11.3 kb relative to the 814 start site isoelectric point xiii PPI PR pRCII pSl4-Cl pSl4-CAT-2.1 pSl4-CAT-4.3 pSl4exoPEII-8 pSl4-IVS pSV2-Neo PMSF RA preproinsulin II progesterone receptor plasmid containing preproinsulin II cDNA genomic $14 probe containing 3’exon sequences plasmid containing from +19 to -2111 bp of $14 promoter sequences fused to CAT reporter gene plasmid containing from +19 to -4316 bp of $14 promoter sequences fused to CAT reporter gene genomic Sl4 probe containing sequences representing +23 to +483 in the S'exon of the rat 814 gene genomic $14 probe containing intervening sequences plasmid conferring neomycin resistance phenylmethylsulfonyl fluoride retinoic acid retinoic acid receptor alpha isoform of the retinoic acid receptor beta isoform of the retinoic acid receptor retinoic acid response element gamma isoform of the retinoic acid xiv $14 SDS SGC t1/2 TBS TLC TR TTP uCi ug UTP v-erbA WAT X 9 receptor ribonucleic acid revolutions per minute Spot 14 gene sodium dodecyl sulfate seconds half-life 3,5,3'-triiodothyronine tris-buffered saline thin-layer chromatography thyroid hormone receptor c-erbAal c-erbAaz c-erbABl c-erbABZ thyroid hormone response element thymidine triphosphate unit microcurie microgram micromolar uridine triphosphate viral erythroblastosis oncogene white adipose tissue (factor) times gravity INTRODUCTION The ability of animals to integrate discrete functions of organs and tissues is largely dependent on two regulatory mechanisms: 1) the endocrine system: and 2) the nervous system. This was first demonstrated by Starling in 1905 when he described separate endocrine and neurogenic control of gastric function (Starling, 1905). Hormones control four general areas of cell physiology, these include: reproduction: growth and development; maintenance of the internal environment; and energy production, utilization, and storage (Figure 1). This dissertation examines hormonal control of gene expression in the context of energy production, utilization, and storage. Regulation of gene expression represents a major target for hormonal control of cell function. Steroid and thyroid hormones acting at the transcriptional and post- transcriptional levels control small networks of genes essential for signalling the appropriate change for a cell. Because of the tremendous effects of steroid and thyroid hormones on animal physiology, considerable efforts have been directed towards understanding the initial events in hormonal regulation of gene expression. Increasing our FIGURE 1. GENERAL DIAGRAM ILLUSTRATING HORMONAL CONTROL OF PHYSIOLOGICAL FUNCTIONS. The coordination of biological activities within the body requires a communication network allowing for both local and systemic control. [From Wilson, J.D. Wilson and D.W. Foster, 1985, In "Williams Textbook of Endocrinology" (eds. J.D. Wilson and D.W. Foster) W.B. Saunders Co., Philadelphia, Pennsylvania] Growth and Reproduction Development Hormones Maintenance of Energy Production, Internal Environment Utlllzatlon, and Storage 4 understanding of the mechanisms regulating gene expression will enhance our view of normal cell function as well as contributing to our understanding of aberrant cell function. Elucidating the molecular interactions responsible for hormonal control of gene expression requires the use of model gene systems. The $14 gene is a model gene that was originally identified for studying the primary events associated with T3 control of gene expression. Transcriptional activity of the hepatic $14 gene is stimulated within 5 min of T3 administration to hypothyroid animals, making it one of the most rapidly T3 responding genes reported to date. This observation makes the 814 gene an excellent candidate for understanding the early events in T3 action on a cell. Since its initial discovery, expression of the 814 gene has been shown affected by developmental, tissue-specific, nutritional, and other hormonal factors (see Literature Review C). Therefore, the $14 gene provides a good model for understanding specific aspects of multifactorial control of gene expression. Since the majority of studies on the $14 gene have been carried out in the rat, the primary objective of this study was to establish a cell line to examine hormonal regulation of $14 gene expression. An in yitzg system would provide a less complicated and more easily Chapter 1. Literature Review A. Initiation 9; Gene Transcription Gene transcription requires the interaction of proteins (trans-acting factors) with appropriate regulatory DNA sequences (cis-acting sequences). The basic features of this have been well described (for reviews see Lewin, 1985: Lillie and Green, 1989: Saltzman and Weinman, 1989). The short length cis-acting DNA sequences important for gene transcription are categorized as either promoter or enhancer/silencer elements (Figure 2). Promoter elements are found immediately upstream from the 5’-end of the gene. Promoters function in the binding of RNA polymerase II and associated transcription factors (e.g. TFIIA-F) to allow formation of a transcription preinitiation complex (Buratowski et al., 1989: Gallagher , 1989: O’Malley, 1990). This usually requires the presence of a TATA binding sequence and a CAAT binding sequence. The rate- limiting step in gene transcription is formation of the transcription preinitiation complex. Once formed, RNA Pol II initiates mRNA synthesis at a specific nucletide (start site) and transcribes the gene to yield the full length primary transcript. FIGURE 2. CONTROL OF GENE EXPRESSION AND PROTEIN SYNTHESIS. Transcription initiation requires the formation of a trancriptional complex (black circles) on the DNA promoter elements located immediately 5’-to the start site (arrow on exon 1). The interaction of various other regulatory proteins (trans-acting regulatory factors: shaded circles) on separate and distinct DNA enhancer sequences (cis-linked regulatory sequences) alters the rate of transcription initiation from the promoter. A pre-mRNA is first synthesized in the nucleus containing all exons and introns which is then methylated (CH3 on exon 1) and poly-adenylated (AAA on exon 2). Splicing together of exons and removal of introns occurs in the nucleus before the mature mRNA is transported out of the nucleus. Cytoplasmically located mRNA can then direct translation of a protein on the rough endoplasmic reticulum. The protein may then be post-translationally modified in the golgi complex. Regulation can occur at any of these levels and the stability of both the mRNA as well as the protein will also determine the amount of either of these that is present. eee fl CAAY TATA. l J L__J Enhancer Promoter l CHI Dre-mRNA H“ on: 1 mRNA Exon 1 [Exon 2)“ Protein 8 Enhancer/silencer elements can be located within a gene or flanking either the 5'- or 3'-ends of the gene. Enhancers function to induce, while silencers suppress gene transcription, by directly affecting the rate-limiting step in gene transcription. Promoter/silencer enhancer elements are also short segments of DNA typically only 13-18 nucleotides in length that may respond to single or multiple trans—acting factors. These stimulus-specific enhancer/silencer elements are used to coordinate regulation of a network of genes by the same effector. Enhancer/silencer elements for the same trans-acting factor located on different genes show a high degree of sequence homology. Slight differences in enhancer sequences may contribute to the varying magnitude of responses of different genes to the same stimuli (Glass et al., 1989). Several enhancer elements have been identified which function in developmental and tissue-specific control of gene expression (Distel et al., 1987: Higuchi et al., 1988: Herrera et al., 1989: Hunt et al., 1989; Lassar et al., 1989). Enhancer sequences have also been identified that interact with steroid hormone receptors, thyroid hormone receptors, retinoic acid receptors (for review see Evans, 1988), and other transcription factors whose functions are not all completely understood (Wingender, 1988). 9 The ability of promoter and enhancer/silencer sequences to control gene expression requires their interaction with appropriate trans-acting regulatory factors. While many trans-acting factors have been identified, their precise function is not always fully understood (Wingender, 1988: Meyer et al., 1989). The spectrum of trans-acting factors expressed, and the accesibility of promoter and enhancer/silencer elements to these factors in a cell, determines the regulatory pattern for a specific gene. These protein-DNA interactions can acutely regulate gene transcription, as seen during energy production, utilization, and storage (for review see Goodridge, 1987), or effect long-term regulation e.g. developmental and tissue-specific expression (Distel et al., 1987: Higuchi et al., 1988; Herrera et al., 1989: Hunt et al., 1986: Lassar et al., 1989). Once transcription begins at a start site, RNA polymerase II transcribes the entire gene including both protein coding and non-coding exons, and introns (see Figure 2: Lewin, 1985). This primary transcript is either degraded or capped by methylation at the 5’-end of the transcript. The transcript is usually polyadenylated at the 3’-end to form a pre-mRNA. The pre-mRNA is then either degraded or processed by removal of introns to form a mature mRNA that is transported out of the nucleus. In the cytoplasm, the mRNA will be translated into protein at 10 polysomes or rough endoplasmic reticulum. The stability to degradation of the different mature mRNAs can influence how many proteins are translated from that transcript. Once translated, the protein may be modified in the golgi apparatus and either remain within the cell or be secreted. This multi-step pathway in protein synthesis provides many sites for control of expression. Normal cell function is, in part, the result of properly coordinated gene regulation while abnormal cell function can be the result of deviant gene regulation. Therefore, understanding the molecular interactions responsible for gene control is of great interest. 8. Hormonal Control p; pppp Expression One of the functions of the endocrine system is to coordinate energy production, utilization, and storage within various tissues (Berne and Levy, 1983: Martin, 1985; Wilson and Foster, 1985). All hormones released by endocrine glands are typically at femtomolar to nanomolar concentrations in the circulation. Changes in the circulating levels of hormones are responsible for signalling changes in physiologic responses of specific target cells. In addition to systemic distribution of hormones via the circulation, several additional mechanisms exist for directing hormones to specific target tissues to coordinate their function (Genuth, 1983: Martin, 1985: Wilson and 11 Foster, 1985). For example, insulin action on liver function is an example of delivery of a hormone within a specific circulation. Insulin is very effective in the liver because the pancreas releases insulin directly into the hepatic circulation, and not because of unique hepatic insulin receptor concentrations. Hormones can also have local effects independent of the circulation. Paracrine and autocrine effects rely on simple diffussion. Paracrine effectors are released by cells and direct their effects to adjacently located cells. Autocrine effectors influence the cells that make them after they are released. Finally, intracrine effectors may be synthesized and retained within a cell to control cell function, as suggested for ligand binding to some of the "orphan receptor" (see below) transcription factors (O'Malley, 1989, 1990). In target cells, hormones are typically recognized by specific protein receptor molecules capable of binding the ligand in a reversible low capacity-high affinity manner. The different hormone receptors are usually in highest concentrations in target cells. However, receptors can also be present in non-responding tissues, such as insulin receptors in red blood cells, which do not respond to insulin (Genuth, 1983). This observation demonstrates that target tissues need appropriate signal transduction mechanisms in addition to receptors for proper hormonal signalling. After signal transduction, different networks 12 of responses are elicited dependent on the hormone and cell type. Traditionally, hormones are grouped into two classes, those which diffuse through the cell membrane and bind intracellular receptors and those which interact with receptors within the plasma membrane (Axelrod and Reisine, 1984: Evans, 1988: Beato, 1989: O’Malley, 1990: Saltiel, 1990). However, this classification may require modification since a growing body of evidence indicates steroid hormones also bind plasma membrane receptors (for review see Touchette, 1990). Oxytocin, growth hormone, and insulin are examples of peptide or protein hormones which bind to plasma membrane receptors. Activated plasma membrane receptors can influence many cell functions. For example, ligand activated insulin receptors can lead to changes in DNA, RNA, and protein synthesis, increasing and decreasing enzyme activity, and can also lead to changes in cellular localization of proteins (for reviews see: Granner, 1987: Saltiel, 1990). The mechanisms by which activated insulin receptors cause these changes are not well understood. However, they are initiated by ligand- induced tyrosine-kinase activity of the insulin receptor (Kasuga et al., 1983: Kahn, 1985). Secondary to increased tyrosine-kinase activity are biological responses at the plasma membrane (Klip et al., 1988), within the cytoplasm (Goldfine, 1987; Klip et al., 1988), and at the nuclear 13 level (Sasaki et al., 1984; Colosia et al., 1988: Pape et al., 1988). These appear related to changes in protein phosphorylation states, concentrations of diacylglycerol, inositol triphosphate, and calcium (Granner, 1987: Saltiel, 1990). This demonstrates that activated plasma membrane receptors transduce their hormonal signals through a diversity of pathways. Alternatively, steroid hormones, thyroid hormone (T3), vitamin D, and the vitamin A derived retinoic acid (RA) are all lipid soluble and bind to intracellular receptors. The receptors for each of these ligands is similar in structure and function and belong to a steroid/thyroid superfamily of genes related to the proto-oncogene v-erbA (for review see: Evans, 1988: Beato, 1990; O'Malley, 1990). The principal mechanism of action for these hormone activated receptors i.e. ligand-activated transcription factors, is to alter gene expression. Some examples include glucocorticoid control of tyrosine aminotransferase (Jantzen et al., 1987) and phosphenolpyruvate carboxykinase gene activity (Sasaki et al., 1984: Sasaki and Granner, 1988) and T3 control of growth hormone (Glass et al., 1987), S14 (Jump, 1989a), and myosin heavy chain gene activity (Izumo and Mahdavi, 1988). Ligand binding presumably induces a conformational change in the receptor molecule yielding an activated receptor (Evans, 1988: Beato, 1989: O’Malley, 1990). For all but the nuclear located estrogen, progesterone, and 14 thyroid hormone receptors (TR), this is followed by translocation of the receptor-ligand complex into the nucleus. These activated, nuclear located, steroid receptors (trans-acting factors) show increased binding affinity to specific DNA enhancer sequences critical for altering gene expression. These DNA enhancer sequences are referred to as hormone response elements (HRE). In contrast to steroid receptors, TR (and possibly RAR) binding to an HRE is independent of the ligand. Thyroid hormone receptors are always bound to their respective TRE, however, without T3 binding, this receptor is unable to activate gene expression. Ligand binding somehow alters the transcription activating potential of TRs (Jump et al., 1981: Glass et al., 1987: Dalman et al., 1990). The various HREs demonstrate some specificity in their- interaction with different receptors from this superfamily. For example, the glucocorticoid response element (GRE) interacts with the progesterone receptor (PR) and the glucocorticoid receptor (GR) and is called a GRE/PRE (progesterone response element: Tsai et al., 1988). However, a GRE/PRE only weakly binds a TR (Glass et al., 1988), and is not responsive to estrogen receptor (ER: Mader et al., 1989). Thyroid hormone receptors can bind thyroid hormone response elements (TREs: Glass et al., 1987) and estrogen response elements (EREs: Glass et al., 1988), while RARs can bind TREs and RA response elements 15 (RAREs: Umesono et al., 1988: Graupner et al., 1989: de The et al., 1990: Sucov et al., 1990: Vasios et al., 1990) and vitamin D responsive elements (VDREs: Schule et al., 1990). Slight differences in HRE sequences between genes for a receptor, such as for TREs between the growth hormone gene and the myosin heavy chain gene, may account for some of the difference in the magnitude of response of these two genes to T3 (Glass et al., 1989). Activated receptors on a HRE typically induce nucleosomal displacement that allows for binding of general and specific transcription factors to sites near the HRE (Beato, 1989: Baniahmad et al., 1990). The receptor and additional transcription factors are thought to act cooperatively to increase transcription from the cis-linked gene. In addition to transcriptional control, the steroid/thyroid hormone receptor superfamily also regulates cell function at the post-transcriptional level. Examples of genes controlled in this manner include phosphoenolpyruvate carboxykinase (Petersen et al., 1989,), 3-hydroxy-3-methylglutaryl-coenzyme A reductase (Chen and Shapiro, 1990), and growth hormone (Diamond and Goodman, 1985: Bedo et al., 1989). Regulation can occur at one or any combination of the following levels including pre-mRNA processing, mRNA export from the nucleus, stability of mature mRNA, transport of mRNA to the endoplasmic reticulum, translation rate, post-translational l6 modifications of proteins, and enzyme activity (for reviews see: Berk, 1989: Bernstein and Ross, 1989; Klausner and Harford, 1989). Control by transcriptional or post- transcriptional mechanisms can be either primary or secondary (Hess et al., 1990). This depends on whether the receptors directly effect the cis-linked gene or they influence the action of other trans-acting factors on the controlled gene. The similarity in structure and function of the steroid/thyroid hormone receptor superfamily is shown in Figure 3 (Evans, 1988). Region E is responsible for recognizing and binding the correct ligand. A hinge region of the receptor is encoded by region D which appears to be active in nuclear translocation and trans-activation of the receptor. The ability of these receptors to bind with the appropriate HRE is determined by region C, and region A/B is sometimes important in correctly transducing the hormone signal. This superfamily of genes codes for one androgen receptor (AR), one GR, and one vitamin D receptor (VDR: Evans, 1988). However, multiple isoforms of the PR, ER, TR, and RAR are synthesized. Presently, there are 2 ERs and PRs (Evans, 1988: Kastner et al., 1990), at least 4 TRs have been identified (Sap et al., 1986: Weinberger et al., 1986: Thompson et al., 1987: Izumo and Mahdavi, 1988: Koenig et al., 1988: Lazar et al., 1988, 1989: Hodin et al., 1989: Miyajima et al., 1989), and 10 RARs (Giguere et 17 FIGURE 3. GENERAL STRUCTURAL AND FUNCTIONAL DIAGRAM OF THE STEROID AND THYROID HORMONE RECEPTORS. This general structure and functional arrangement is shared by all steroid and thyroid hormone receptors. The function of region A/B is not completely understood, but sometimes is effective in modulating the efficiency of signal transduction of the activated receptor. Region C is responsible for binding the appropriate DNA response element. Region D is a hinge region in the receptor , molecule. Finally, region B confers ligand specificity on the receptor molecule since this is the site of ligand binding. [M. Beato, 1990, Cell 56:335-344] NH- AIS 18 Modulator Ligand ]— coon 19 al., 1987: Petkovich et al., 1987: Benbrook et al., 1988: Brand et al., 1988: Zelent et al., 1989: Kastner et al., 1990: Mangelsdorf et al., 1990). Of special note for studies described in Chapter 4, is the fact that of the four TR isoforms described, TRalr TRaz, TRBlr TRBZI all but the TRaz isoform is capable of binding T3. The inability of the TRaz isoform to bind T3 makes it incapable of transducing the T3 signal in tissues it is highly expressed, thereby, blocking T3 action (Izumo and Mahdavi, 1988: Lazar et al., 1988). The multiplicity of receptors for some of these hormones presumably increases the ability of an organism to discretely regulate gene expression. Receptors from the steroid/thyroid hormone receptor superfamily form dimers in order to regulate gene expression from HREs. This requirement presumably increases the diversity of control by these receptors (Evans, 1988: Beato, 1989: O’Malley, 1989b). Particularly for receptors with multiple isoforms or those that are able to form heterodimers with different receptors. This latter case includes ERs which are able to form heterodimers with GRs and PRs (Cato and Ponta, 1989), and TRs that can associate with RARs. The interaction of TRs and RARs appears to exploit the potential diversity of control afforded by receptor dimer formation when one considers there are at least 4 TR isoforms (Sap et al., 1986: Weinberger et al., 1986: Thompson et al., 1987: Izumo and 20 Mahdavi, 1988: Koenig et al., 1988: Lazar et al., 1988, 1989: Hodin et al., 1989: Miyajima et al., 1989) and 10 RAR isoforms (Giguere et al., 1987: Petkovich et al., 1987: Benbrook et al., 1988: Brand et al., 1988: Zelent et al., 1989: Kastner et al., 1990: Mangelsdorf et al., 1990), apparently all capable of forming homodimers or heterodimers (Umesono et al., 1988: Forman et al., 1989: Glass et al., 1989: Graupner et al., 1989). Additionally, some proto-oncogenes appear to recognize and cross-couple to the same HREs that some of the steroid/thyroid hormone receptors bind (Schule et al., 1990). These interactions suggest gene regulation by steroid/thyroid hormone receptors is complex. It should also be mentioned that there is a growing list of structurally related proteins to the steroid/thyroid hormone receptor superfamily called "orphan receptors" whose ligand and function have not yet been identified (Evans, 1988: O’Malley, 1990). The importance of these putative transcription factors remains to be discovered. The explosion of information on this superfamily of receptors and their DNA binding sites in the last five years presents a daunting but exciting challenge for the future. One can begin to envision the mechanisms necessary to achieve the diversity of control for normal growth and development. However, to fully understand the complexity of steroid hormone, thyroid hormone, and vitamin A and D 21 control in response to developmental and environmental signals, we must identify all the proteins capable of influencing their action and identify all the permissible interactions. Studies directed towards this goal require the use of model gene systems to test hypotheses. These models are used to understand both the specifics of that particular system and possibly elucidate more general mechanisms of control that are applied by other genes. A major objective of my dissertation was to gain a better understanding of eukaryotic hormonal control of gene expression utilizing the model 814 gene (see below). C. Si; Background The early observation by Oppenheimer et a1. (1972) that thyroid hormone receptors were localized in the nucleus of cells supported the hypothesis that ligand- activated thyroid hormone receptors probably had nuclear effects. To better understand these nuclear T3 actions, Seelig et al.(1981, 1982) and Liaw et al. (1983) identified a group of mRNAs from rat liver that rapidly responded to altered thyroid hormone concentrations. "Spot 14," or $14, was the designation given to one of these mRNAs. $14 was originally identified by two-dimensional gel analysis of the ip yippp translational products from mRNAs isolated from rat livers of animals exposed to varying concentrations of T3 (Seelig et al., 1981, 1982: Liaw et al., 1983: Jump et al., 1984). 22 The S14 gene is a single copy, 4.4 kb gene in rat liver, consisting of two exons (Fig. 4). The presence of two polyadenylation signals in the 3’-exon of the $14 gene accounts for the presence of two mRNA species (mRNA514: 1.2 and 1.37 kb) detected by northern analysis. The protein coded by the 814 gene is small (17 kD), acidic (pI 4.9), and located within the cytosol (Jump et al., 1984: Liaw & Towle, 1984: Kinlaw et al., 1989). Western analysis of 514 protein levels with a specific antibody (Kinlaw et al., 1989, Strait et al., 1989) demonstrates the protein parallels changes in mRNA514 abundance, supporting and extending previous studies that 814 regulation occurs primarily at the transcriptional level (Jump et al., 1988: Jump, 1989b). Despite the lack of definition of a precise biochemical function of the 814 protein, a body of circumstantial evidence suggests it is involved in lipid metabolism. This is based on tissue-specific (Jump et al., 1984: Jump and Oppenheimer, 1985), developmental (Jump et al., 1986: Perez-Castillo et al., 1987: Jump et al., 1988: Clarke et al., 1989a), hormonal (Seelig et al., 1981, 1982: Carr et al., 1984: Jump et al., 1984, 1990a, 1990b: Beyer et al., 1985: Jump and Oppenheimer, 1985: Kinlaw et al., 1986, 1987a, 1988, 1989; Jump, 1989a, 1989b: Strait, 1989) and nutritional (Liaw et al., 1983: Topliss et al., 1983: Mariash et al., 1986: Freake and Oppenheimer, 1987: Kinlaw 23 FIGURE 4. STRUCTURAL ORGANIZATION AND CHARACTERIZATION OF THE RAT S14 GENE, mRNA, AND PROTEIN. The rat 814 gene is a single copy/cell gene. It is approximately 4400 bases in length consisting of two exons and one intron. Two poly-adenylation signals within the second exon allows for the synthesis of two different mRNA species. The open reading frame for the protein is located entirely within exon one. The protein is 150 amino acids long and is cytosolically located. 24 RAT S14 Gene ,. l l 700 me $14 mRNA —_AAA 1200 nts AAA 1320 nts S14 Protein 17.000 pl 4.9 25 et al., 1987a: Clarke et al., 1989a, 1989b: Hamblin et al., 1989: Jacoby et al., 1989: Mariash, 1989: Wong et al., 1989a: Jump et al., 1990a, 1990b,), regulation of expression of mRNASl4 coding for the 814 protein (Figure 5). It is the dynamic regulation of the 814 gene ip zixp that makes it an attractive model to understand the molecular basis of multifactorial control of gene expression. The questions addressed by this dissertation focus on understanding hormonal regulation of 814 gene expression. 1. Tissue-specific regplation p; 8;; gppp pgppession The $14 gene is expressed only in those tissues associated with the synthesis and storage of mobilizable lipid, i.e. triglycerides (Jump et al., 1984: Jump and Oppenheimer, 1985). While levels of mRNASl4 expressed in liver approximate 1200 copies/cell (intermediate abundance class: Jump et al., 1988), levels of expression in lactating mammary gland and white adipose tissue are 4-fold and 10-fold higher, respectively (Jump and Oppenheimer, 1985: Freake and Oppenheimer, 1987). Gene expression in the appropriate tissue requires the interaction of tissue-specific trans-acting factors with DNA regulatory sequences. Identification of these sequences is facilitated by DNase I hypersensitive site (Hss) analysis. DNase I Hss are typically found flanking the 5’-end of most active genes and reflect local 26 FIGURE 5. MULTIFACTORIAL CONTROL OF THE 814 SYSTEM. See text for description. 27 DEVELOPMENT TISSUE SPECIFIC V HORMONAL S 1 4 REGULATION / \ NUTRITION CIRCADIAN 28 disruptions in nucleosomal periodicity due to the interactions of trans-acting factors (e.g. developmental factors, thyroid hormone receptors) with underlying DNA sequences (e.g. enhancers, HREs: Gross and Garrard, 1988). Because of the altered chromatin structure in Hss, they are more sensitive to DNase I digestion than other locally arranged chromatin. The 5’-f1anking region of the 814 gene from adult euthyroid rat liver contains six major DNase I Hss located at -65 to -265 bp (Hss-l), -1.2 kb (Hss-Z), - 1.8 kb (Hss-3'), -2.67 kb (Hss-3), -5.3 kb (Hss-4), and — 6.2 kb (Hss-S) upstream from the $14 transcriptional start site (Jump et al., 1988, 1989a: 1990a). However, in tissues where the $14 gene is not expressed, Hss-l, 2, 3’, and Hes-3 are not found implicating these DNA sequences and associated proteins in the tissue-specific regulation of the 814 gene. - In the three tissues mRNASl4 is expressed at high levels (liver, white adipose tissue, and lactating mammary gland) T3 regulates mRNASl4 levels (Jump et al., 1984: Jump and Oppenheimer, 1985: Freake and Oppenheimer, 1987). Although mRNASl4 is detected in brain, lung, kidney, heart, spleen, testes, and pituitary, the level of expression is less than 6% of the liver value, and T3 regulation is absent. The differential chromatin organization of the $14 gene and the variation in 814 gene transcription accounts for the tissue-specific variation in mRNASl4 abundance 29 (Jump et al., 1988: Jump, 1989b). In order to define the molecular basis for tissue-specific expression of the 814 gene, an ip vitro cell culture model that mimics the ip gigp state is required. 2. Developmental Contgol While rat liver mRNASl4 is expressed at approximately 1200 copies/hepatocyte in the adult animal, expression in the 5 or 15 day old rat is <1% of the adult value (Jump et al., 1988). The major increase in hepatic mRNASl4 to adult levels occurs at the time of weaning. Hepatic mRNASl4 increases >100-fold between 18 and 22 days post-partum (Jump et al., 1986, 1988). This increase in mRNASl4 is due to a >40-fold induction of $14 gene transcription. Prior to activation of S14 gene transcription, the Hss-l, 3, and 3' sites are absent. The formation of Hss-l and Hss-3 at weaning correlates with activation of the S14 gene (Jump et al., 1988: Jump, 1989b).‘ Future studies to determine the significance of these DNA sequences and trans-acting factors also requires the establishment of an ip zippp model system. The induction of mRNASl4 during post-natal development can be influenced by dietary manipulation. Weaning pups onto rat chow at day 18 post-partum accelerates the induction of mRNASl4 (Jump et al., 1988: Jump, 1989b), while weaning rat pups onto rat chow containing high fat will delay the normal rise in mRNASl4 (Clarke et al., 1989b: Jump, 1989b). These results suggest that the 30 dietary switch at weaning from a high fat milk diet to rat chow represents the principal activator of 814 gene expression. The developmental profile for S14 expression during this period is similar to the increase in activity reported for other lipogenic enzymes such as fatty acid synthase (Volpe & Kishimoto, 1972: Clarke et al., 1989b) and acetyl-CoA carboxylase (Lockwood et al., 1970). The increases in activity for fatty acid synthase and acetyl-CoA carboxylase is due primarily to increased protein synthesis (Volpe & Vagelos, 1976). For fatty acid synthase this corresponds to an increase in mRNA production that is paralleled by similar changes in mRNASl4 (Clarke et al., 1989b). These changes may represent a coordinated adaptive response of the liver to significant changes in dietary composition. Understanding the developmental control of 814 expression may provide important clues as to how a class of specific enzymes, i.e., lipogenic enzymes, may be involved in metabolic adaptation. 3. Hopponal Regplation pf 81; Gppp Egpression a. THYROID HORMONE Since its initial discovery, the 814 gene has proven to be an excellent model for studying the primary effects of T3 on gene expression. Methimazole induced hypothyroidism protects Hss-3’ and Hss-3 regions of the $14 promoter from DNase I digestion, so that they are either absent or present at diminished levels. However, within 5 31 min of T3 injection to hypothyroid animals, the sensitivity of Hss-3 to DNase I was increased. Hyperthyroid animals show a Hss pattern similar to euthyroid rats, suggesting Hss-3 is important for rapid T3 control of the $14 gene (Jump, 1989a: 1990a). Finding a DNA sequence within the Hss-3 region showing high homology to the growth hormone TRE further supports the View that the Hss-3 region is critical for T3 control of 814 gene expression. Zilz et a1. (1990), using a DNA transfection approach, reported that TREs responsible for T3 control of $14 gene transcription were found between -2.6 and -2.7 kb upstream from the 5'-end of the gene. This region corresponds to the T3-inducible Hss-3. Therefore, T3, like steroid hormones, regulates gene transcription by modification of chromatin structure (Jump, 1989a). Thyroid hormone control of 814 protein expression can be defined in steps: 1) T3 enters nucleus and binds to its DNA-associated receptor (Jump et al., 1981): 2) TR induces a DNase I Hss at -2.6 to -2.7 kb in front of the 814 gene (Jump, 1989a): 3) S14 gene transcription increases within 15 minutes (Jump, 1989a): 4) mRNASl4 increases within 20-30 min (Jump et al., 1984: Jump, 1989): 5) 814 protein increases after 2 h (Kinlaw et al., 1989). Varying the thyroidal status of animals also changes the profile of additional trans-acting factors that bind to cis-linked regulatory elements important for $14 gene 32 expression (Wong et al., 1989a, 1990). These results demonstrate T3 has profound effects on $14 gene transcription, as well as, some post-transcriptional effects on hepatic mRNASl4 abundance (Jump, 1989a). While other genes are also transcriptionally and post- transcriptionally regulated by T3, such as pituitary growth hormone (Yaffe and Samuels, 1984), malic enzyme (Dozin et al., 1986), and HMG-CoA reductase (Simonet and Ness, 1988), 814 is one of the most rapidly responding genes to T3 stimulation. b. GLUCOCORTICOIDS Although glucocorticoids appear to influence Sl4 gene expression in rat liver, their effects are not well characterized or understood. In adrenalectomized animals and in animals receiving the glucocorticoid agonist, dexamethasone (DEX), hepatic mRNASl4 levels were similarly higher than in intact animals (Beyer et al., 1985). These confounding results suggest either that glucocorticoid regulation of mRNASl4 expression in rat liver is very complex or that one or more of the studies were in error. Our studies in the mouse (Lepar and Jump, 1989: Chapter 2) clearly show that glucocorticoids induce mRNASl4 in liver and WAT. In fact, T3 and glucocorticoids interact additively to regulate mRNASl4. c. INSULIN AND GLUCAGON 33 Since the original observation that insulin to glucagon ratios might contribute in regulating 814 gene expression (Carr et al., 1984), Jump et al. (1990b) have demonstrated that insulin rapidly stimulates rat hepatic 814 gene transcription. The carbohydrate, sucrose, and its constituents, glucose and fructose, also appear capable of individually regulating the $14 gene. However, maximal stimulation requires carbohydrate, as well as, insulin and thyroid hormone (Mariash et al., 1986: Kinlaw et al., 1987; Hamblin et al., 1989: Jump et al., 1990a, 1990b). Transfection analysis of $14 promoter elements suggests sequences located as far as -4316 from the start site of transcription may be important for carbohydrate control (Jacoby et al., 1989). More specifically, DNase I Hss site analysis shows changes in chromatin structure around Hss—1 and Hss-2 may be significant for carbohydrate control (Jump et al., 1990a). Regulation of $14 gene expression by carbohydrates may require the synthesis of metabolic intermediates to properly control expression of the $14 gene (Mariash and Schwartz, 1986: Mariash et al., 1986: Hamblin et al., 1990: Jump et al., 1990a). In contrast to these positive effects on S14 gene expression, hormones or compounds which elevate intracellular hepatic levels of cAMP, like glucagon, epinephrine, and theophylline, or the membrane permeable cAMP analog dibutyrl-cAMP, attenuate hepatic abundance of 34 mRNASl4. Administration of glucagon to rats at 2000 hr leads to a fall in hepatic mRNASl4 with a t1/2 of 90 min (Kinlaw et al., 1986). Kinlaw et al. (1986, 1987a) have provided evidence to suggest that glucagon activates an A-kinase second messenger system which leads to the inhibition of 814 gene transcription. Thyroid hormone administration to glucagon-treated animals can reverse the inhibitory effect and stimulate 814 gene transcription, as well as, the accumulation of mRNASl4 (Kinlaw et al., 1986, 1987a). However, insulin is unable to overcome the inhibitory effects of cAMP on the $14 gene (Jump et al., 1990b). Hepatic mRNASl4 levels and $14 gene transcription are also decreased in states of streptozotocin-induced diabetes and in starvation (Carr et al., 1984: Jump, 1989b: Jump et al., 1990a, 1990b). Thus, several hormones, including: T3, glucocorticoids, insulin, glucagon: and epinephrine may contribute to the multihormonal regulation of 814 gene expression in murine liver and WAT. 4. circadian Expression The diurnal circadian rhythym in $14 gene expression follows the feeding pattern of the rat. This corresponds to highest mRNASl4 levels at night when animals are eating (2200 h) and lowest in the morning when animals are restful (0800 h). Despite the apparent correlation between nutritional status and mRNASl4 levels, circadian mRNASl4 expression appears dependent on some undetermined non- 35 nutritional regulatory factor (Kinlaw et al., 1986: Kinlaw et al., 1987a, 1987b: Wong et al., 1989). Additional studies are necessary to better understand circadian regulation of the $14 gene. D. Sippificance Understanding how hormones that interact with nuclear receptors, and hormones that activate plasma membrane-generated second messenger systems, function to regulate gene transcription is a central issue in physiology. Efforts by this laboratory to define these processes (see above) required the development of an ip yipgp cell model to examine specific aspects of gene expression. This dissertation describes for the first time 814 gene regulation in an immortalized non-neoplastic cell line. Biochemical and molecular genetic techniques were used in conjunction with the adipocyte forming 3T3-F442A cell line to examine thyroid hormone, glucocorticoid hormone, retinoic acid, and A-kinase mediated aspects of 814 gene expression. This cell line was also used in preliminary studies to begin to understand the unique function of the 814 protein. This dissertation successfully illustrates the usefulness of 3T3-F442A cells as a model system for performing defined experiments to understand the molecular aspects of hormonal control of gene expression. Although not specifically addressed in this dissertation, 3T3-F442A 36 cells will also provide a useful model for further understanding developmental and tissue-specific $14 gene control. Chapter x. Expression pg mRNASl4 and Functional Comparison with Lipogenesis and Lipolysis ip 3T3-F442A Adipocytes Iptrodpction The 814 gene encodes a small acidic protein (MW 17 kd, pI 4.9) found within hepatic cytosol (Seelig et al., 1981, 1982: Liaw et al., 1983: Liaw and Towle, 1984). Regulation of $14 gene expression has been extensively studied ix yiyp by a number of laboratories and demonstrates expression of this gene is under complex regulation. Although originally identified to study early events in T3 stimulation of rat liver gene expression, the 514 gene is also regulated by developmental, tissue-specific, nutritional, and other hormonal signals (see Literature Review C). In the rat, stimulation of the $14 gene is associated with lipogenesis (see Literature Review C). However, the precise function of the 814 protein remains unknown and the 814 protein sequence does not demonstrate close homology with any sequences recorded in national computer sequence libraries. This suggests the 814 protein is a novel protein. Efforts to identify molecular events controlled by specific effectors of 814 gene expression, as well as, 37 38 analysis of 814 protein function have been hampered by the lack of a good in yipxp model system. Primary cultures of isolated hepatocytes have been used for $14 studies (Topliss et al., 1983: Mariash et al., 1984: Jacoby et al., 1989), but the difficulties associated with isolation and cultivation of primary cultures of hepatocytes (Oppenheimer et al., 1987), as well as the differing response of the 814 gene in these cells when compared to ix yiyp, raises some concerns about the efficacy of the hepatocyte model. Recently, the neoplastic kidney cell line, COS-1 has been used for transfection studies on 814 gene regulation (Zilz et al., 1990), but kidney only minimally expresses the 814 gene in 2129 (Jump and Oppenheimer, 1985). Therefore, this laboratory was interested in establishing a well characterized, non-neoplastic, immortalized cell line for conducting regulation and functional studies of the S14 protein. Because the 814 gene is highly expressed in WAT (Jump and Oppenheimer, 1985), I examined the adipose forming 3T3-F442A cell line for regulatable 814 gene expression. 3T3-F442A cells are a subclone of a fibroblast cell line originally isolated from Swiss mouse embryos (Todaro and Green, 1963: Green and Kehinde, 1976). The differentiation of 3T3-F442A fibroblasts into adipocytes is accelerated in media containing fetal bovine serum plus insulin (Kuri-Harcuch and Green, 1978: Smith et al., 1988). 39 3T3-F442A cells and other 3T3 related cell clones demonstrate many features characteristic of adipose tissue, including: changes in morphology (Todaro and Green, 1963), induction of several proteins involved in carbohydrate, fatty acid, and triacylglycerol synthesis (Mackall et al., 1976: Mackall and Lane, 1977: Coleman et al., 1978: Spooner et al., 1979: Wise & Green, 1979: Spiegelman and Green, 1980: Student et al., 1980: Weiss et al., 1980: Kawamura et al., 1981: Spiegelman and Farmer, 1982: Spiegelman et al., 1983: Bernlohr et al., 1984, 1985: Chapman et al., 1984, 1985: Wise et al., 1984: Cook et al., 1985: Paulauskis and Sul, 1988: Ntambi et al., 1988: de Herreros and Birnbaum, 1989: Kaestner et al., 1989, 1989) and the acquisition of responsiveness to both lipogenic and lipolytic hormones (Rubin et al., 1977, 1978: Reed et al., 1977: Lai et al., 1982: Grunfeld et al., 1985: Schwartz and Carter-Sn, 1988). These observations have led to the widespread use of 3T3- F442A cells for conducting studies on adipose-like tissue. Previous 1 vivo studies on $14 gene regulation have dealt only with the rat. Since 3T3-F442A cells were derived from the Swiss mouse, hormonal regulation of the $14 gene was initially examined in the liver of Swiss mice. I next examined expression of the $14 gene in both the 3T3- F442A fibroblast and adipocyte phenotypes. Effects of T3, the glucocorticoid analog dexamethasone (DEX), cAMP stimulating agents (epinephrine and forskolin), and a cAMP 40 analog (8-CPT cAMP) were examined for their effects on S14 gene expression in 3T3-F442A cells. Control studies examined the expression of mRNA’s from fatty acid synthase (FAS), glycerol phosphate dehydrogenase (GPD), and B-actin. It has previously been shown that the mRNAs for fatty-acid synthase and glycerophosphate dehydrogenase increase dramatically, while B-actin slightly decreases with adipocyte differentiation of 3T3-F442A cells (Student et al., 1980: Spiegelman and Green, 1980: Spiegelman and Farmer, 1982: Spiegelman et al., 1983: Cook et al., 1985). I also examined the association of mRNASl4 expression with changes in lipogenesis and lipolysis in an effort to gain some insight into the function of the 514 protein. MATERIALS AND METHODS Animals Intact male Swiss mice (25-35 g) were obtained from Charles River (Kalamazoo, MI). Mice were maintained on a stock diet (Teklad) and water ad libitum. Animals were administered T3 dissolved in 0.1 N NaOH at 200 ug T3/100 g body weight (I.P.), DEX dissolved in 100% EtOH at 200 ug DEX/100 g body weight (I.P.), or a combination of both hormones. Control animals received vehicle (0.01% EtOH and 0.001 N NaOH). Animals were killed between 0900 and 1300 h by ether anesthesia and exsanguination. 41 gpll Cultures 3T3-F442A fibroblasts were obtained from B. Spiegelman, Dana Farber Cancer Center, Harvard Medical School. 3T3-F442A preadipocytes and adipocytes were cultured as described by Cook et al. (1985). Preadipocytes were seeded at 105 cells/100 mm plate and grown in Maintenance Media (MM) containing Dulbecco’s Modified Eagles Medium (DMEM), 4.5 mg/ml glucose, 10% calf serum, 100 U/ml penicillin, and 100 U/ml streptomycin. Adipocyte cultures were obtained by transferring fibroblasts at <70% confluence to new culture dishes and growing cells in Differentiation Media (DM) containing MM plus 5 ug/ml bovine insulin and substituting 10% fetal calf serum for 10% calf serum. Adipocyte cultures attained 280% differentiation within 10 days of reaching confluence. Preadipocyte and adipocyte cultures were treated with hormones only after reaching confluency or full differentiation, respectively. DEX was dissolved in 100% EtOH and administered at 1 uM. The non-metabolizable cAMP analog, 8-CPT cAMP (Sasaki et al., 1984), and the adenylate cyclase stimulating compound, forskolin (Seamon & Daly, 1981), were dissolved in EtOH and administered at 0.1 mM and 50 uM concentrations, respectively. Epinephrine was dissolved in 0.1 N NaOH and administered at a final concentration of 1 uM for the indicated times. All media were changed every 48 h. 42 Complementaxy pg; Probes The p814 exoPEII-8 genomic probe used to measure $14 mRNA levels contains sequences extending from +23 to +483 bp from the 5'-end of the 814 gene. This region represents the 5'-exon of the rat liver $14 gene (Jump et al., 1988). The GPD cDNA was obtained from B. Spiegelman, Dana Farber Cancer Center, Harvard Medical School (Spiegelman et al., 1983). The PAS-17 cDNA was obtained from S. Clark, Upjohn Co., Kalamazoo, MI, and represents the FAS-17 cDNA cloned originally by Nepokroeff et al. (1984). The B-actin cDNA was obtained from L. Kedes, Stanford University, Palo Alto, CA. The method of plasmid isolation has been previously described (Jump, 1989). Mpasuxement pf mRNA Levels Total RNA was extracted from mouse liver or 3T3-F442A cells by the guanidinium thiocyanate procedure of Chirgwin et al. (1979). The levels of the indicated mRNA were measured by dot blot hybridization using a nick translated insert derived from the appropriate probe (Jump et al., 1988: Jump, 1989a). Following hybridization, blots were washed, dried, and exposed to X-ray film. Relative levels of hybridization were quantified by videodensitometry and related to an internal hybridization standard from an adult (60 day old) male rat (Jump et al., 1989a). This standard has been quantitated for abundance of mRNASl4. One unit of 43 mRNASl4 = 4.8 x 107 copies of mRNASl4/ug total RNA or 1200 copies of mRNASl4/liver cell. Noxthern Apalysig pg 3335814 Total RNA was denatured in 1X-MAE (40 mM MOPS, pH 7.0: 10 mM Na-acetate: 1 mM EDTA, pH 8.0), 2.2 M formaldehyde, and 50% formamide, and then electrophoretically separated in a 1.2% agarose gel containing 2.2 M formaldehyde (Maniatis et al., 1981). RNA was transferred to Zetabind, hybridized, and visualized as described above for dot blots. A separate lane of rat liver total RNA was co- electrophoresed and stained with ethidium bromide to determine the relative mobility of 18S and 288 ribosomal RNA. Fatty-acid Synthesis pp p Determinant p; Lipogenesis Lipogenesis in 3T3-F442A adipocytes was measured as described by Eadara et al. (1989) with the following modifications. Confluent 3T3-F442A adipocytes in 35 mm plates were washed twice with 1X-PBS, followed by addition of 2 ml Krebs-Henseleit Buffer (KHB: Krebs and Henseleit, 1932), pH 7.0, containing 10 mM glucose, 5 ug/ml insulin, 30 uCi/ml [3H]ZO, plus vehicle, 1 uM DEX, 1 uM epinephrine (5 mM stock dissolved in 0.1 N NaOH), or a combination DEX and epinephrine. Cultures were incubated for 2 h and incorporation of [3H120 into fatty-acids (FA) was determined by extraction with propylene oxide, and B- scintillation counting of labeled FA. Incorporation of 44 [3H120 into FA was linear for up to 3 h. After 2 h of labeling, a 10 ul media sample was removed from each plate and counted to determine the specific activity of [3H]20 in each plate. The remaining media was decanted, cells were washed twice with 1X-PBS, trypsinized, and collected in 1 ml glass distilled H20 (GD-HOH). A 50 ul aliquot was removed and frozen for protein determination by the Lowry method (1951). The remaining sample had 3 ml 30% KOH added , vortexed, followed by addition of 7 ml 95% EtOH, and vortexed again. Samples were saponified at 75°C for 90 min, chilled on ice, prior to adding 9 ml GD-HOH and vortexing. Samples were extracted with 5 ml petroleum ether and the lower aqueous phase was retained. The aqueous sample was acidified with 1.5 ml 11.7 N HCl and re- extracted with 5 ml of petroleum ether. The upper organic layer was removed and evaporated in a scintillation vial. Organic extraction with petroleum ether was repeated twice. Evaporated samples were quantitated by B-scintillation counting. Results are expressed as nmol [3H]20 incorporated in 2 h, and expressed relative to protein levels. Glycerol Detprmination gpx Lipolysis Lipolysis in 3T3-F442A adipocytes was measured as described by Coutino et al. (1990), Michigan State University, East Lansing, MI. Briefly, 3T3-F442A adipocytes grown in 100 mm plates were washed twice with 45 lX-PBS followed by addition of KHB, pH 7.0, containing 3% bovine serum albumin (Intergen/Armour Biochemical), 5 mM glucose, with either vehicle, 1 uM DEX, 1 uM epinephrine, or a combination treatment of DEX and epinephrine. Samples (1 ml) were removed at 30 min intervals. The 1 ml aliquot was replaced with fresh test media to maintain a constant buffer volume. Enzymatic activity was terminated by heating samples to 80°C for 15 min followed by freezing and thawing. Debris was removed by sedimentation at 1500 x g, and supernatants were transferred to 1.5 ml microfuge tubes for glycerol analysis. Glycerol in the supernatant was measured using a kit (Boehringer Mannheim). Glycerol release was linear for at least 2 hours. Cells remaining on the plate were trypsinized and collected for cell counting, protein analysis (Lowry et al., 1951), and mRNA analysis by dot blot hybridization. Statistical Analysis Statistical comparisons of samples first utilized a completely random analysis of variance (ANOVA). The a priori Dunnett’s test was used as indicated to compare all means with a control when there was significant difference in variance. The a posteriori Student-Newman-Keuls test (SNK) was used as indicated to look at all possible comparisons of means when there was a significant difference in variance (Steel & Torrie, 1980). 46 RESULTS Hoxxonal Regxlation pi EENA314 ix xpxgp xiypx While mRNASl4 is regulated by T3 in rat liver, similar studies in mouse liver have not been reported. Accordingly, intact male mice were administered a receptor- saturating dose of T3 (200 ug/lOO g body weight), DEX, or a combination of both hormones. Four hours after hormone treatment, mice were killed, livers removed and RNA extracted for measurement of mRNASl4 levels. Both T3 and DEX significantly induced mouse hepatic mRNASl4 (Figure 6). T3 induced mRNASl4 3-fold while DEX induced mRNASl4 4- fold. The combination of T3 and DEX induced hepatic mRNASl4 7-fold, indicating that each hormone acted independently to regulate hepatic mRNASl4 abundance. Similar results were obtained when mRNASl4 expression was examined in mouse white adipose tissue (data not shown). Hormonal Regxlation pi 814, FAB, GPD, and B-actin mRNA ix 3T3-F412A zibroblasts and Adipocytes We next examined the expression of mRNASl4 in 3T3- F442A preadipocytes and adipocytes. 3T3-F442A cells were >80% differentiated within 10 days of treatment with DM as measured by Oil Red 0 staining of triacylglycerides and by phase contrast microscopy. Virtually no expression of mRNASl4 was detected in either 3T3-F442A preadipocytes or adipocytes indicating that induction of mRNASl4 does not accompany differentiation of preadipocytes to the adipocyte 47 FIGURE 6. RELATIVE EXPRESSION OF mRNAS14 IN MOUSE LIVER. Intact male mice (25 g) were injected (i.p.) with a receptor saturating dose (200 ug/100 g body weight) of either T3, DEX, or a combination of both hormones. Four animals were in each group. Control animals were injected with vehicle (0.01% ethanol and 0.001 N NaOH, respectively). All animals were killed 4 h after hormone administration and livers were removed and extracted for total RNA. Hepatic mRNASl4 abundance was examined by RNA dot blot analysis as described in Materials and Methods. Results are expressed as mean + SE. One mRNASl4 Unit = 4.8 x 107 copies of mRNASl4/mg total RNA. aP<0.01: T3, DEX, and T3 plus DEX treated mice versus vehicle treated mice (ANOVA and Dunnett’s test). bP<0.001: T3 plus DEX treated mice versus T3 or DEX treated mice (ANOVA and SNK test). mRNAs1 4 Abundance 48 mRNAS1 4 EXPRESSION IN SWISS MOUSE LIVER (Units) 3.6- 3.01 2.41 1.81 1.22 0.64 o.b L—«o 0.0 99!qu x30 x30 + 91 49 phenotype (Table 1). While T3 treatment alone failed to induce mRNASl4 in either phenotype, DEX significantly induced mRNASl4 only in adipocytes. Following a 24 h exposure to DEX or DEX plus T3, mRNASl4 was induced 266- fold and 340-fold, respectively. Since we cannot detect S14 expression in untreated adipocytes or fibroblasts, we underestimate the fold induction of mRNASl4. In contrast to the 1 vivo studies in liver and WAT of rats and mice (Figure 6: Jump & Oppenheimer, 1985), mRNASl4 was found to be unresponsive to T3 in both 3T3-F442A fibroblasts and adipocytes. Dexamethasone induction of mRNASl4 in adipocytes was consistent with the DEX induction of this sequence in mouse liver (Figure 6) and WAT. The DEX-mediated induction of mRNASl4 in 3T3-F442A adipocytes is dependent on the differentiation state of the cells since DEX failed to induce mRNASl4 in preadipocytes. The mRNA expression from the lipogenic enzymes FAS and GPD were measured to confirm the differentiation state of 3T3-F442A cells. While mRNAFAS was induced 5.9-fold, mRNAGpD was induced 30.9-fold in the preadipocyte to adipocyte transition (Table 1). The induction of FAS and GPD during adipocyte differentiation has been previously reported (Spiegelman et al., 1983: Cook et al., 1985: Paulauskis & Sul, 1988). In contrast, the relative abundance of mRNAB-actin changed little with differentiation (Table 1). Others have reported changes in 50 TABLE 1 Relative Abundance of sillsa in 3T3-P442A Fibroblasts and Adipocytesb Adipocytesc T3 + Fibroblastsc Veh DEX T3 DEX 514 Nod ND 0.66 ND 0.40 (4) (4) + 0.12 (4)° (4) + 0.08 (4)e FAS 0.48 2.84 4.37 -- -- + 0.04 (3) + 0.63 (7) + 0.88 (6) CPD 0.49 15.16 19.30 -- -- + 0.07 (4) + 4.44 (3) + 5.18 (5) B-actin 0.89 1.83 1.05 -- -- + 0.05 (4) + 0.43 (5) + 0.12 (8) a Results are expressed as units of mRNA relative to the amount of mRNA detected in the liver of a male adult rat (250 9). One unit of mRNAs 4 represents 1200 copies of mRNASl4/cell, where a cell contains 25 pg of RNA (1 unit - 4.8 x 107 copies of mRNASl4/mg total RNA). This RNA represents an internal standard used in all hybridization analysis and facilitates normalization of results. Results are expressed as mean 3 8E. The number of samples are contained within the parentheses. b 3T3-F442A fibroblasts and adipocytes were grown in Dulbecco's Modified Eagles Medium containing calf serum or fetal calf serum plus insulin, respectively. Adipocytes were > 80% differentiated (see Materials and Methods). C Fibroblast and control cultures received no hormone or vehicle treatment. Vehicle for T3 was 0.001 N NaOH and the vehicle for dexamethasone was 0.01% ethanol. Hormones were administered at 1 uM for 24 h. d N.D.: not detectable. e p.m .05: a 0.0 0 5 J . q q q o 0 0 o 4 3 2 1 __oo\35 5305 I ouoi>mEv £035 .33 D Dexamethasone Epinephrine 63 epinephrine and DEX stimulated lipolysis. The combination treatment of DEX plus epinephrine induced an effect significantly greater than either agent alone. However, the effect was not additive, suggesting that lipolysis was maximally stimulated and that both agents acted through similar pathways. The effect of epinephrine on lipolysis reported here is similar to previous studies on 3T3-L1 cells (Kawamura et al., 1981). These treatments caused no significant changes in protein content per cell or plate. Although DEX stimulates both lipolysis and mRNASl4 accumulation, the combination treatment of DEX and epinephrine further augments the lipolytic response while dramatically inhibiting mRNASl4 accumulation. These confounding results do not indicate a simple association of mRNASl4 levels with lipolysis as measured by this assay in 3T3-F442A cells. DISCUSSION We examined the regulation of expression of the 814 gene in 3T3-F442A preadipocytes and adipocytes with four goals in mind: 1) to determine whether the $14 gene was expressed in the 3T3-F442A adipocyte-forming cell line; 2) to determine whether expression of the $14 gene followed the same differentiation-dependent expression as many other enzymes involved in lipid metabolism in 3T3-F442A cells (Mackall et al., 1976; Mackall and Lane, 1977; Coleman et al., 1978; Spooner et al., 1979; Wise and Green, 1979; 64 Spiegelman and Green, 1980; Student at al., 1980; Weiss et al., 1980; Kawamura et al., 1981; Spiegelman and Farmer, 1982; Spiegelman et al., 1983; Bernlohr et al., 1984, 1985; Chapman et al., 1984; Wise et al., 1984; Chapman et al., 1985; Cook et al., 1985; Paulauskis and Sul, 1988; Ntamni et al., 1988; de Herreros and Birnbaum, 1989; Kaestner et al., 1989a, 1989b); 3) to determine whether the 3T3-F442A cells would be a useful model to define the tissue-specific and hormonal regulation of 814 gene expression (Seelig et al., 1981; Topliss et al., 1983; Carr et al., 1984; Jump et al., 1984, 1987, 1990a, 1990b; Mariash et al., 1984, 1986: Jump and Oppenheimer, 1985; Kinlaw et al., 1986, 1987a, 1988; Jump, 1989a, 1989b; Zilz et al., 1990); and 4) to look for basal or regulatable expression of the 314 gene, and presumably 814 protein, suggesting this cell line could be used for functional studies of the S14 protein. My studies show the 814 gene is expressed only in 3T3- F442A adipocytes treated with the glucocorticoid agonist DEX (Table 1). The differentiation dependent hormonal regulation of the S14 gene is similar to the glucocorticoid regulated expression of the clone 10 gene expressed in TA-l adipocytes (Chapman et al., 1985). The dependence of mRNASl4 accumulation in 3T3-F442A cells on the differentiation state of the cells suggests the 814 protein is a member of a family of proteins involved in lipid metabolism. Messenger RNA levels for GPD 65 (Table 1; Spiegelman et al., 1983; Cook et al., 1985), adipsin (Cook et al., 1985; Djian et al., 1985), FAS (Table 1; Paulauskis and Sul, 1988), malic enzyme, ATP-citrate lyase (Wise et al., 1984), stearoyl-CoA desaturase 1 and 2 (Ntambi et al., 1988; Kaestner et al., 1989a), insulin responsive glucose transporters GT2 and GT3 (Kaestner et al., 1989a; de Herreros and Birnbaum, 1989), fructose-1,6- bisphosphate aldolase, and the adipocyte P2 protein (Bernlohr et al., 1984, 1985; Cook et al., 1985) are induced during differentiation of either 3T3-L1 or 3T3- F442A cells. However, the $14 gene is expressed only in fully differentiated 3T3-F442A adipocytes following glucocorticoid treatment. Although this finding reinforces the notion that the 814 protein functions in lipid metabolism, it also shows that the S14 protein does not play a role in either differentiation or in the basal accumulation of triacylglycerides in 3T3-F442A cells. Because the $14 gene is not basally expressed while these cells differentiate or accumulate lipid, S14 protein function is not required for these biological functions in 3T3-F442A cells. This is contrary to the view obtained from in 2129 tissue-specific, hormonal, and nutritional studies, consistent with the S14 protein functioning in lipogenesis (Jump and Oppenheimer, 1985; Freake and Oppenheimer, 1987; Perez-Castillo et al., 1987; Jump, 1989b). 66 Because of the purported role of the 814 protein in lipid metabolism, effects of A-kinase stimulation on the S14 gene were examined by Kinlaw et al. (1984) and Mariash et al. (1986, 1987, 1988). These investigators demonstrated mRNASl4 declined due to transcriptional inhibition by glucagon, epinephrine, and db-cAMP, similar to declines in mRNA for several other genes involved in lipogenesis including malic enzyme (Goodridge, 1987) FAS (Goodridge, 1987; Paulauskis and Sul, 1989), L-type pyruvate kinase (Decaux et al., 1989), aldolase B (Munnich et al., 1985), and glucokinase (Iynedjian et al., 1989). The inhibitory effects of epinephrine and 8-CPT cAMP on mRNASl4 abundance in DEX stimulated 3T3-F442A cells (Figure 7. A) indicates this inhibitory effect is mediated via the A-kinase pathway (Lefkowitz and Caron, 1987). Future studies will determine the contribution of transcriptional and post-transcriptional mechanisms and whether cAMP is directly acting on DNA enhancer sequences of the 814 gene. To better understand the function of the 814 protein, I examined the effects of glucocorticoids on lipogenesis and lipolysis in 3T3-F442A adipocytes to determine if a correlation existed with either of these functions and mRNASl4 levels. I made the assumption that changes in mRNASl4 parallel changes in $14 protein during these manipulations, as has been shown 13 2129 during T3 stimulation of rats (Kinlaw et al., 1989). No correlation 67 was demonstrated between the expression of the S14 gene and lipogenesis (Figure 9). In contrast, DEX enhanced lipolysis while stimulating mRNASl4 (Figure 10). However, epinephrine treatment also enhanced lipolysis, a treatment that inhibits $14 expression in adipocytes (Figure 7. A). The failure to detect changes in cellular lipogenesis or lipolysis that consistently correlate with 814 expression, suggests the role of the 814 protein in cell metabolism may be subtle. More definitive studies directed at either over-expressing the 814 protein or fully attenuating the 814 protein by antisense RNA methods may aid in defining the function of the 814 protein. Perhaps the most surprising observation made in this study was the absence of T3 regulation of $14 gene expression in 3T3-F442A adipocytes. Thyroid hormone regulates mRNASl4 levels in both rat and mouse liver and WAT (Figure 6; Jump et al., 1984; Jump and Oppenheimer, 1985; Jump, 1989a), and primary hepatocytes (Topliss et al., 1983a; Mariash et al., 1984a). The rapid induction of S14 gene transcription and a single DNase I hypersensitive site (Hss-3) upstream from the $14 gene supports the concept that T3 acts directly on the $14 gene. Subsequent transfection analysis of 5’-flanking sequences of the $14 gene confirms this (Zilz et al., 1990), and supports Jump’s previous suggestion that the cis-linked T3 regulated DNase I hypersensitive site (Hss-3) harbors a putative thyroid 68 hormone response element (TRE) controlling $14 gene transcription (Jump, 1989a, 1989b). Additional studies are required to define the mechanism accounting for the lack of T3 regulation of the S14 gene in 3T3-F442A cells (Chapter 3), and may provide a novel model for understanding certain aspects of T3 regulation of 314 gene expression. In conclusion, 3T3-F442A cells demonstrate positive control by glucocorticoids and negative control by a-kinase effects, similar to what has been reported i vivo. These cells provide a useful i3 vitro model to better understand regulation by these factors of the 814 protein. 3T3-F442A cells also provide an 13 2i229 system to facilitate identification of the function of the 814 protein. The unexpected lack of T3 control of the $14 gene in 3T3-F442A cells also appears to provide some very unique opportunities to understand gene control by thyroid hormone receptor isoforms (see Chapter 4). Chapters 3, 4, and 5, further characterize certain aspects of S14 gene regulation in 3T3-F442A cells, and with Chapter 6, suggest the implications of this research and suggest additional experiments on hormonal control of $14 gene expression in 3T3-F442A cells. Cheats; ;. Glucocogticoid gggglation g; 814 Gene Egpresgiop 12 3T3-F442A Adipocztes IEIRODUCTION In chapter 2, I demonstrated that the $14 gene was responsive to glucocorticoids in 3T3-F442A cells in a differentiation-dependent manner. In this chapter, I will characterize the glucocorticoid-mediated regulation of $14 gene expression in 3T3-F442A adipocytes. These studies include kinetics of induction, dose response, and glucocorticoid analog analysis. I will use Lg 2i§gg transcriptional run-on and transfection analysis to evaluate the mechanism of glucocorticoid regulation of 814 gene expression in 3T3-F442A adipocytes, and provide preliminary data on the localization of putative glucocorticoid and developmental cis-acting elements controlling $14 gene expression. These studies suggest glucocorticoids directly induce 814 gene transcription by an activated glucocorticoid receptor (GR) to increase mRNASl4 abundance. The DNA sequences responsible for developmental and glucocorticoid control of $14 gene expression are located within -4.3 kilobases (kb) of the start site for 814 gene transcription. 69 70 MATERIALS 532 METHODS 92;; Cultures All cultures were maintained as previously described (Lepar and Jump, 1989; Chapter 2). Complementagy 9!; Probes The pSl4exoPEII-8 genomic and B-actin cDNA probes used were described in Chapter 2. A genomic probe containing the preproinsulin II gene (pRCII) was obtained from P. Lomedico, LaRoche Institute, Nutley, NJ (Lomedico et al., 1979). Complementary DNA probes for a- and B-tubulin were obtained from D. Cleveland, the Johns Hopkins University, Baltimore, MD (Sullivan & Cleveland, 1986). The method of plasmid isolation was as described (Jump, 1988). Construction 9; 8;4-Chloramphenicol Acetyl Transferase Fusion Genes The plasmid used in the DNA transfection studies was pSl4-CAT-4.3 (Figure 11). This plasmid contains portions of the rat liver 814 genomic DNA fused to the chloramphenicol acetyltransferase reporter gene. The $14 genomic DNA extends from +19 to -4316 bp relative to the S'-end of the rat liver 814 gene. The pSl4-CAT-4.3 plasmid was constructed by ligating a 2203 bp BamHI-HindIII fragment isolated from the pEMBLSl4-13 plasmid into the pSl4-CAT-2.1 plasmid generously provided by Dr. H. Towle, Biochemistry Department, University of Minnesota (Zilz et al., 1990). The pEMBLSl4-13 plasmid was originally cloned 71 FIGURE 11. STRUCTURE OF THE PLASMID VECTOR pSll-CAT-4.3. Promoter sequences of the $14 gene that begin at +19 relative to the start site of transcription and extending upstream to -4316, were inserted into the multicloning site of vector pCAT. This fusion plasmid was named pSl4-CAT-4.3. 72 +19 pS14-CAT-4.3 73 from a size-selected rat liver genomic library and contains 314 genomic DNA extending from +1.7 to -11.3 kb relative to the 5’-end of the S14 gene (Jump, 1989). The pSl4-CAT-2.1 plasmid contains 814 genomic DNA extending from +19 to -2111 bp relative to the 5' end of the rat liver $14 gene. The 2203 bp BamHI-HindIII fragment from pEMBLSl4-13 was cloned into the BamHI-HindIII sites of the pSl4-CAT-2.1 plasmid. The HindIII site at 2111 and the BamHI just 5’- to this site place the insert in the native orientation. These plasmids were isolated and purified through CsCl, and sterilized with ethanol prior to transfection. Measurement 2; 335814 Levels Total RNA was extracted, blotted, and quantified as described in Chapter 2. 122152122 2: 122:21125 £22121 Nuclei were isolated from confluent 3T3-F442A adipocytes as described by Chapman et al (1985). For each treatment, confluent cells from three 100 mm plates were suspended in 5 ml Hypotonic Buffer (20 mM Tris-Cl, pH 8, 4 mM MgC12, 6 mM CaClz, 0.5 mM DTT) transferred to a homogenizing flask containing 5 ml Lysis Buffer (0.6 M sucrose, 0.2% Triton X-100, 0.5 mM DTT) and 2 mM PMSF. Cells were lysed and nuclei sedimented at 2000 RPM for 10 min. Nuclei were resuspended in Resuspension Buffer (0.25 M sucrose, 10 mM Tris-Cl, pH 8, 10 mM MgC12, 1 mM DTT), nonidet NP-40 and PHSF were added to 0.1% and 1 mM, 74 respectively. Nuclei were sedimented as described above. The nuclear pellets were washed with Resuspension Buffer, centrifuged, and resuspended in 3T3-F442A Nuclear Storage Buffer (50% glycerol, 10 mM Tris-Cl, pH 8, 10 mM MgC12, 1 mM DTT). Isolated nuclei were adjusted to 60 A26O/ml and stored at -80°C for transcriptional analysis. I; 21222 Transggiptional "Run-On" Analysis In 21229 transcriptional run-on analysis of the $14 gene was previously described (Jump, 1989a). The composition of the transcription buffer was: 25% glycerol; 75 mM HEPES, pH 7.5; 3 mM MgC12: 100 mM KCl; 0.1 mM EDTA; 0.05 mM EGTA; 1.0 mM spermidine; 1 mM DTT; 16 ug/ml creatine kinase; 100 ug/ml creatine phosphate; 1 mM ATP; 0.5 mM CTP; 0.5 mM GTP; so uci 32P-UTP at 0.25 uM; 120 units/ml RNasin; 10 A250 units of nuclei in a final volume of 300 ul. Transcriptional assays were run at 26°C for 10 min followed by DNase I treatment (5 min). RNA was purified by proteinase K treatment followed by phenol:chloroform (1:1) extraction and chloroform:isoamyl alcohol (49:1) extraction. DNA was removed by additional DNase I treatment. Radiolabeled 32P-RNA was hybridized to cDNAs affixed to Zetabind. Hybridization was quantified by videodensitometry. zgepargtiog g; ggggl2 Egagggected gggzgggg; 9;;12 3T3-F442A fibroblasts were stably transfected with the pSl4-CAT-4.3 plasmid in the presence of the selection 75 plasmid, pSV2-Neo (generously provided by Dr. S. Conrad, Microbiology Department, Michigan State University). Actively growing fibroblasts at 50% confluence were transfected by the lipofection technique described by Felgner, et al (1987). Fibroblasts in 60 mm petri dishes were washed 3 times with Tris-buffered saline (TBS: 20 mM Tris-Cl, pH 7.5) and once with DMEM without serum. pSV2-Neo (2 ug) was mixed with pSl4-CAT-4.3 (20 ug) in 2 ml Opti-I MEM medium, without serum (Life Technologies, Bethesda MD). The lipofection reagent (DOTMA:(N-[l-(z,3-dioleylosy)propyl] -N,N,N-trimethylammonium chloride}) at 1 mg/ml (from Life Technologies, Inc, Bethesda, MD) was diluted by adding 50 ul of reagent to 1 ml of Opti-I MEM to give 50 ug/ml. The 1 ml diluted lipofection reagent was added to the 2 m1 of diluted DNA, mixed and applied directly to fibroblast monolayers in 60 mm petri dishes. After 24 h of treatment with the transfection reagent in a humidified OZ/COZ culture incubator at 37°C, medium was decanted and cells fed DMEM + 10% calf serum. After 48 h, cells were trypsinized and passed from a 60 mm plate to a 100 mm plate. Cells were switched to selection medium, i.e. DMEM + 10% calf serum + 400 ug/ml geneticin. Cells were maintained on selection medium through the experiment. While >90% of the cells were dead within 1 week of trypsinization, small foci of geneticin-resistant colonies 76 were apparent. Typically 20-30 colonies of resistant cells were present on each plate. When the colony size reached 3 mm in diameter, cells were trypsinized and reseeded onto the same plate. When the plate was 70% confluent cells were trypsinized and passed 1:8. When these cells were 70% confluent, cells were trypsinized and frozen at -135°C in DMEM + 10% calf serum + 400 ug/ml geneticin + 10% DMSO. Experiments with transfected cells involved removing a vial of frozen cells from the freezer, thawing and growing cells to 70% confluence in DMEM + 10% calf serum + 400 ug/ml geneticin. Typical experiments involved growing cells in 60 mm plates to confluence, treating cells with differentiation medium (DMEM + 10% fetal calf serum + 5 ug/ml insulin + 250 ng/ml dexamethasone + 500 ug/ml isomethylbutylxanthine) for 48 h, followed by trypsinization, replating 1:1, and maintaining cells on DMEM + 10% fetal calf serum + 5 ug/ml insulin for 7 to 14 days. Cells were fully differentiated within 7-10 days. Hormone treatments were for 96 h at 1 uM DEX. 9;: Assays Chloramphenicol acetyl transferase (CAT) activity was assayed as described by Gormon (1985) and as modified by Jacoby, et a1 (1989). Briefly, transfected 3T3-F442A fibroblast or adipocyte monolayers were washed once with TBS and scraped in a volume of 500 ul of TBS into a 77 microfuge tube. Cells were sedimented at 1500 x g (10 min), the supernatant decanted and the cells were resuspended in 0.25 M Tris-Cl, pH 7.5. Cells were lysed by 4 cycles of freezing and thawing. Cell debris was removed by sedimentation at 5000 x g (10 min) and the supernatant frozed at -20°C until use. Protein concentration was determined using a kit from Bio—Rad. Cell extracts were heat denatured at 65°C for 15 min to inactivate proteases and deacetylases. The CAT assay included: 100 mM Tris-Cl, pH 7.5; 100 uM acetyl coenzyme A, 10 uM [14C1-chloramphenicol (0.1 uCi) and 50 ug cellular extracted protein. Reactions were incubated for 4 h at 37°C and then terminated with the addition of 1 ml ethyl acetate. Samples were vigorously vortexed for 30 sec, centrifuged at 12,000 x g for 5 min and the upper organic phase transferred to a microfuge tube. The ethyl acetate was dried and the residual chloramphenicol (and acetylated forms) was resuspended in 12 ul of ethyl acetate and applied directly to a silica G gel thin layer chromatogram. Chloramphenicol and the acetylated forms were separated on the TLC in chloroform:methanol (95:5). The developed TLC was dried and exposed to X-ray film overnight. The fraction of chloramphenicol acetylated was quantified by cutting and B-scintillation counting. 3T3-F442A fibroblast and adipocytes contain no activity that acetylates chloramphenicol. In addition, we 78 detected virtually no CAT activity in cells stably transfected with the promoter-less pCAT(AN) plasmid. Transfection of cells with the Rous sarcoma virus (RSV)-CAT showed high levels of CAT activity in all fibroblast and adipocytes examined. However, the activity was not affected by any of the hormonal treatments. Thus, the CAT activity measured in pSl4-CAT transfected cells represents changes in gene transcription and not changes in CAT enzymatic activity or stability induced by the specific treatments. RESULTS Kinetics g; 2335814 Induction 32g Deciine i2 3T3-F442A Adipocytes I The kinetics of DEX induction of mRNASl4 were examined in 3T3-F442A adipocytes. Cultures of differentiated adipocytes were exposed to either DEX (1 uM) or vehicle. Cultures were harvested after the indicated times of exposure and the relative abundance of mRNASl4 was determined by RNA dot blot hybridization. Following DEX administration to 3T3-F442A adipocytes, mRNASl4 was induced in a linear fashion (r=0.925) up to 72 h (Figure 12). A significant (p<0.05: ANOVA and Dunnett’s test) induction of mRNASl4 was detected 30 min after hormone treatment (see insert, Figure 12). Maximal mRNASl4 induction in 3T3-F442A adipocytes required 3 days of hormone exposure and these levels were maintained at least 7 days. 79 FIGURE 12. KINETICS OF DEXAMETHASONE INDUCTION OF mRNASl4 IN ADIPOCYTES. Fully differentiated 3T3-F442A cells were exposed to either 1 uM DEX (open circles) or vehicle (closed circles) containing DM. Medias were changed every 48 h. Cells were harvested after the indicated times and the relative abundance of mRNASl4 was determined by RNA dot blot hybridization. Values are expressed as Units of mRNASl4 (see Table 1 for description), mean i SE, n=4. INSERT: Results obtained over the initial 4 h induction period are replotted to illustrate the rapidity of the effect of DEX on 814 gene expression. 8O DEXAMETHASONE KINETICS OF mRNAS14 INDUCTION 1.00- ‘ o_° «Ha-2 0.104 .—. M Q 0.80" . 0 one c 1 O E i 3 A 0060‘ .'=0 3 5 5 ‘3 \ .D I! < 9: , R wt 5 / '- V 0.404 ° 5’ J / E 0.204 ) 0.60Lm¢=2—/—:—\\—o——¢-.-.-I._. 72 120 168 216 264 O 2 4 24 Hours 81 We next determined whether DEX was required to maintain mRNASl4 expression (Figure 13). Cultures were treated with DEX (1 uM) for 72 hr to maximally induce mRNASl4. Media was changed to differentiation medium containing no DEX and mRNASl4 abundance was measured at the indicated intervals. A second set of cultures were maintained with DEX for an additional 7 days. While cultures maintained with DEX for 10 days showed no significant decline in mRNASl4 abundance, cultures from which DEX was removed showed a significant loss of mRNASl4. Within 5 days of hormone removal, mRNASl4 returned to undetectable levels. These results indicate that DEX is required both to induce and to maintain mRNASl4 expression in 3T3-F442A adipocytes. The slow decline in mRNASl4 (t1/2 = 26 h) probably represents incomplete removal of the lipophilic steroid hormone. Dose-Dependent Induction g; 3335814 i; Adipocytes Fully differentiated cultures of adipocytes were exposed to various concentrations of DEX for 24 h. Cells were subsequently harvested and analyzed for relative abundance of mRNASl4 as determined by RNA dot blot analysis (Figure 14). While cultures without DEX expressed no detectable mRNASl4, cultures exposed to DEX concentrations as low as 10'10M induced mRNASl4 significantly. A linear increase in mRNASl4 was observed between 10'10 to 10'8M and maximal 82 FIGURE 13. DECLINE IN mRNAsl‘ FOLLOWING DEXAMETHASONE REMOVAL IN ADIPOCYTES. Fully differentiated 3T3-F442A cells were exposed to 1 uM DEX containing DM for 72 h to maximally induce mRNASl4. Cultures were switched to ON containing no DEX (open diamonds) or maintained in DM containing DEX (open circles). A third set of cultures were grown in the absence of DEX for the duration of the experiment (closed circles). Medias were changed every 48 h. Cells were harvested at the indicated times for analysis of relative mRNASl4 abundance by dot blot hybridization. Results are expressed as a percent of mRNASl4 abundance after 72 h of DEX treatment, mean 1 SE, n=4. % Maximal Response 83 DECLINE IN mRNAs1 4 FOLLOWING DEXAMETHASONE REMOVAL 125q 0—0 mm . O—O (-)Deneme0\eeone O—O Vehicle +:n-2 100 84 FIGURE 14. DOSE-RESPONSE RELATIONSHIP FOR DEXAMETHASONE INDUCTION OP mRNAsl‘ IN ADIPOCYTES. Fully differentiated 3T3-F442A cells were exposed to DEX at concentrations ranging from 10'10 to 10’5M for 24 h. Cells were harvested and relative mRNASl4 abundance was measured by dot blot hybridization. Max1mal response was attained at 10'5M DEX. The ED50 = 4 x 10’10M. Results are expressed as Units of mRNASl4 (see Table 1 for description), mean i SE, n=4-8, two replicate experiments. 85 24 HOUR DEXAMETHASONE DOSE-RESPONSE IN 3T3—F442A ADIPOCYTES 120- 100; 80$ 604 4o- % Maximal Response 20~ o 1o"° 1o“ 10" 10"7 1o" [Dexamethasone] (M) 86 levels of expression were achieved with 10’8M DEX. The half-maximal response of 4 x 10'10M correlated well with the Kd of DEX binding to the glucocorticoid receptor (Durant et al., 1986). Analog Specificity g; m§§;314 Induction i2 3T3-F442A Adipocytes Fully differentiated 3T3-F442A adipocytes were exposed to various steroid hormones (1 uM) for 24 h (Figure 15). Cells were harvested and analyzed for relative abundance of mRNASl4 as determined by RNA dot blot analysis. Cultures of adipocytes treated with corticosterone and hydrocortisone induced mRNASl4 to levels representing 88.7 and 88.1%, respectively of the value obtained after DEX exposure. Thus, glucocorticoid agonists stimulated mRNASl4 expression in adipocytes. The mineralocorticoid, aldosterone, also stimulated mRNASl4 accumulation to 78.3% of the maximal DEX response. Since aldosterone displays glucocorticoid activity ig 2i2g (Bondy, 1985), aldosterone may be stimulating mRNASl4 expression through the glucocorticoid receptor. To test this possibility, we treated cultures with the mineralocorticoid agonist deoxycorticosterone acetate, a steroid which displays low glucocorticoid activity ig 212g (Bondy, 1985). Deoxycorticosterone acetate at 1 uM induced mRNASl4 to only 14.9% of the maximal DEX response. This observation is 87 FIGURE 15. STEROID ANALOG SPECIFICITY OF mRNASl4 INDUCTION IN ADIPOCYTES. Fully differentiated 3T3-F442A cells were exposed to various steroid hormones (1 uM) for 24 h. Cells were harvested and analyzed for relative abundance of mRNASl4 by dot blot hybridization. Results are expressed as percentage of DEX stimulated mRNASl4 response, mean 1 SE, n=4. 88 ANALOG SPECIFICITY OF mRNAs1 4 INDUCTION IN 3T3—F442A ADIPOCYTES Ezémogoaozm “Imago—I pacemaozm _ _ I LIII 238332;? II— €86.on v|._ IéflOOOfldMOZm le_ ooflzoomdmeZM 1 4 d u 4 O O 0 O O 8 6 4. 2 100- omcoamom xmo & 89 consistent with interaction of aldosterone with glucocorticoid receptors in adipocytes. The sex steroids progesterone, B-estradiol and dihydrotestosterone failed to induce mRNASl4. The specificity of induction of mRNASl4 by glucocorticoid agonists is consistent with the view that mRNASl4 abundance in 3T3-F442A adipocytes is regulated through the glucocorticoid receptor. Dexamethasone induced 814 Gene Transcription i2 3T3-F442A Adi oc tes We next determined whether DEX regulated mRNASl4 expression at the transcriptional level using an i3 2i§gg run-on assay. ACcordingly, we compared the change in mRNASl4 abundance with the change in $14 transcriptional run-on activity in fully differentiated adipocyte cultures receiving either vehicle or DEX (1 uM) for 24 h. 814 transcriptional run-on activity and mRNASl4 abundance were measured as described in Materials and Methods and Figure 16. Untreated 3T3-F442A adipocyte cultures expressed no measurable mRNASl4 or 814 transcriptional activity, indicating that the inability to detect mRNASl4 in either fibroblasts or control adipocytes is not due to rapid turnover of the transcript. However, cultures receiving DEX for 24 h induced mRNASl4 from undetectable levels to 0.67 1 0.05 units and $14 transcription activity from undetectable levels to 10.0 + 4.2 units. In our kinetic 90 FIGURE 16. IN VITRO TRANSCRIPTIONAL RUN-ON ACTIVITY OF THE 814 GENE IN ADIPOCYTES. Fully differentiated 3T3-F442A cells were treated with vehicle or DEX (1 uM) for 24 h. Cells were isolated and extracted for RNA and nuclei. mRNASl abundance (open bars) was determined by dot blot hybridization and expressed as units (see Table 1 for description). 814 transcriptional run-on activity (cross-hatched bars) was measured as described in Materials and Methods. Autoradiograms of the transcription dot blots were quantified by videodensitometry. $14 run-on activity is expressed in arbitrary units, n=7-8, two replicate experiments, mean i SE. Transcriptional run-on B-actin, a- and B-tubulin were found to be unaffected by DEX treatment of adipocytes. 91 DEXAMETHASONE INDUCED sI4 GENE TRANSCRIPTIONAL ACTIVITY IN 3T3-F442A ADIPOCYTES E 3:5 pages 5322 :OIcsm .ocozatoecoth ccoo Im O O m 2 .U “823 0. 00:09.32 v 5(sz 92 analysis, we sometimes detected an induction of $14 gene transcription as early as 30 min after DEX treatment of adipocytes (data not shown). This rapid effect suggests induction of $14 gene transcription may represent a primary effect of glucocorticoids on 3T3-F442A adipocyte gene expression. Since we cannot accurately measure mRNASl4 or $14 gene transcription rates in untreated adipocytes, we cannot determine the extent of post-transcriptional control affected by DEX. Transcriptional activity of B-actin, a-tubulin and B-tubulin, was not affected by DEX treatment indicating that the rapid induction of $14 gene transcription is a selective response of 3T3-F442A adipocytes to glucocorticoids. And the negative control, preproinsulin, showed no detectable expression under control or DEX stimulated conditions. 9;: Analysis 91 gig giomoter Eigmgggg Involved i2 Deveiopmentai 22g Dexamethasone Contiol To determine if 5'-flanking elements were necessary for developmental and DEX control of $14 gene expression, 3T3-F442A fibroblasts were stably transfected with -4316 bp of 5’-Sl4 DNA sequences joined to the CAT reporter gene (pSl4-CAT-4.3) . Cultures of these transfected cells were then analyzed during differentiation and after DEX stimulation for 814 promoter-driven CAT activity. 93 There was no detectable mRNASl4 expression in fibroblast, fibroblast cultures exposed to Differentiation Medium (DM) for 24 h, 50% differentiated cultures, or vehicle treated adipocyte cultures of 3T3-F442A cells transfected with pSl4-CAT-4.3 (Figure 17). In contrast, cells transfected with the pSl4-CAT-4.3 plasmid showed an increase in $14 promoter driven CAT activity as early as 24 h after cultures of fibroblasts were transferred to DM (7.6-fold). This increased to a maximum of 23.7-fold at 50% differentiation and declined to 13.2-fold above fibroblast levels in the fully differentiated adipocyte. Cultures of adipocytes exposed to either 10'10M or 10'5M DEX show a 5.4- and 62.1-fold increase in mRNASl4 after 24 h over vehicle treated adipocytes (Figure 17). However, CAT activity stimulated by 10"10 and 10"6 DEX was only 1.2-fold and 3.0-fold above vehicle treated adipocytes. DISCUSSION In chapter 2, I demonstrated how the 814 gene is regulated by glucocorticoids in a differentiation-dependent manner. This observation was examined in greater detail here showing DEX induces mRNASl4 accumulation and 814 gene transcription within 30 min of hormone administration, suggesting a primary action of glucocorticoids on the $14 gene. Results of the CAT assays also support the ig 2iggg transcription results. In addition, the CAT results 94 FIGURE 17. DEVELOPMENTAL AND DEXAMETHASONE CONTROL OF mRNASl4 ABUNDANCE AND THE CAT FUSION GENE CONTAINING 5’-FLANRING SEQUENCES OF THE 814 GENE IN ADIPOCYTES. Cultures of 3T3-F442A fibroblasts stably transfected with pSl4-CAT-4.3 were grown up and differentiated in MM and DM as described in Materials and Methods. Cultures of confluent fibroblasts, fibroblast cells 24 h after initial DM exposure, and cultures 50% differentiated in DM were harvested and analyzed for mRNASl4 and ability to acetylate chloramphenicol. Fully differentiated cultures of transfected adipocytes were exposed for 72 h to DM containing vehicle, 0.1 nM DEX, or 1 uM DEX and analyzed. for mRNASl4 and ability to acetylate chloramphenicol. Relative mRNASl4 abundance was measured by dot blot hybridization and presented as Units (see Table l for description). Chloramphenicol and its acetylated products were cut from TLC plates, counted by liquid scintillation counting, and expressed as percent acetylation of chloramphenicol. n=2. 95 EFFECTS OF DEVELOPMENT AND DEXAMETHASONE ON mRNA514 ABUNDANCE AND p51 4—CAT-4.3 ACTIVITY mRNA51 4 Abundance C3 (Units) .0 .0 7‘ 7‘ N N O U! Q U" Q U‘ F'Ibl'obiofll 24 hr Poet - DH 50! Differentiated Venicie °°‘ ”-10 “ j— -5 J °°‘ ‘° " — .o - f '5 o N .U‘ O" 0 CAT Activity - (% Conversion) 96 suggest a direct interaction between a DEX stimulated trans-acting factor and sequences located within -4316 bp of the start site for $14 gene transcription (Figure 18. A). It is interesting to note that there are two palindromic sequences within this region (Figure 18. B) that show significant homology with the consensus glucocorticoid response element (GRE) that binds the activated glucocorticoid hormone receptor and activates gene transcription (Jantzen et al., 1987). The potential 314 GRE located at -671 has homology with 7 of 12 bases of the consensus GRE (-AATTTTNNNTGTTCT-; bold letters show homology with consensus) and the potential 814 GRE located at -2643 bp has homology with 9 of 12 bases (- TTCACANNNTGTTCT-). Preliminary studies (data not shown) with the pSl4-CAT-2.1 plasmid, which includes 814 5’- flanking sequences from the start site to only -2.1 kb in front of the $14 gene, gave similar results to pSl4-CAT- 4.3. This indicates developmental factors and glucocorticoids act within -2.1 kb on the 814 gene. Additional studies will determine the specific DNA sequences responsible for control. Kinetic analysis of mRNASl4 induction showed maximal accumulation was achieved after 72 h of hormone treatment. Removal of DEX from cultures deinduced mRNASl4 indicating the glucocorticoid was required to maintain 814 expression. The dose-response and analog study indicated mRNASl4 97 FIGURE 18. ACTIVATION OF 814 PROMOTER ACTIVITY BY DEVELOPMENT AND DEXAMETHASONE. A. Regulatory sequences responsible for differentiation-dependent and glucocorticoid stimulated expression of the $14 gene appear to be located within - 4316 bp of the start site for $14 gene transcription. B. Three putative $14 developmental regulatory sequences are located -1153, -1347, and -1956, relative to the start site for 814 gene transcription. The first two sequences show significant homology with ESE-2 elements and the last with a FSE-l element (Hunt et al., 1986; Distel et al., 1987; Herrera et al., 1989). Two putative GREs responsible for glucocorticoid effects on the $14 gene are located -674 and -2643 relative to the start site for $14 gene transcription. They show significant homology with the GRE consensus sequence (Jantzen et al., 1987). 98 DEVELOPMENTAL AND GLUCOCORTICOID CONTROL / \ S14 1’ i f ~4a1e 0 S14 I> 3U ‘38:!" 3233 380-. 99 expression in adipocytes is regulated in a glucocorticoid-dependent fashion through the glucocorticoid receptor. These results are consistent with direct action of glucocorticoid receptors on 814 gene transcription in 3T3-F442A adipocytes. From these i2 vitro studies, we can infer that the glucocorticoid mediated control of mRNASl4 in mouse liver and white adipose tissue may be mediated through interaction of glucocorticoid receptors with the 814 gene. The differentiation-dependent expression of the $14 gene seen in 3T3-F442A cells also appears to involve sequences located within 4316 bp from the start site of 814 gene transcription (Figure 18. A). Although no specific sequences have yet been identified within this region for developmental control of 814 gene expression, it is interesting to note two areas of sequence homology with fat specific enhancer elements (FSE) 1 and 2 (Figure 18. B). These elements have been implicated in the developmental expression of the aP2 gene, a lipid binding protein (Bernlohr et al., 1984) expressed in 3T3-F442A cells (Hunt et al., 1986; Distel et al., 1987; de Herrera et al., 1989). The 814 gene shows a FSE 1 like site at -1956, with homology in 9 of 13 bases of the consensus (-CATCCTGGTCAGG-; bold letters indicate homology with consensus). An FSE 2 like element is located at -1153 and -1347 from the start site of transcription of the $14 gene. 100 The more proximal FSE 2 element demonstrates homology with 8 of 14 bases (-TCACNAGAGGACCCT-) and the more distal element demonstrates 7 of 14 bases homology (-GCTGGAGAGGACTCC-). Studies in progress will determine if these are the important sites for developmental control or whether additional sites are required. It will also be interesting to follow the progress on the clone 10 gene from TA1 cells since this is the only other reported gene that demonstrates glucocorticoid control after adipocyte differentiation (Chapman et al., 1985). It is not known why the transfected pSl4-CAT-4.3 activity increases during differentiation while native mRNASl4 levels do not change. Two possible explanations might be that: 1) we are transfecting a rat Sl4 promoter into a mouse cell line, and this may result in some unusual aspects of transfected gene expression; or 2) the transfected sequences may not contain all the essential sequences for proper control of $14 gene control i.e. additional sequences may be needed from further upstream, downstream, or within the gene for proper $14 expression. These results indicate that 3T3-F442A cells are a viable model to examine aspects of glucocorticoid and developmental control of $14 gene expression as seen in vivo. M5..- mmmwgm gen; figpigggign i; 3T3-E453; Ldipogyges INTRODUCTION Thyroid hormone (T3) is known to stimulate accumulation of mRNASl4 in murine liver and white adipose tissue (see Literature Review C. 3; Lepar and Jump, 1989). Suprisingly, however, adipocyte—forming 3T3-F442A cells lack T3 control of $14 gene expression (Lepar and Jump, 1989; Chapter 2). The inability of T3 to regulate 814 gene expression in 3T3-F442A cells may be due to: 1) lack of T3 receptors (TR); 2) inappropriate expression of one or more of the four TR isoforms (TRalr TRaz, TRBlr and TRBz; see Literature Review B); 3) lack of appropriate ancillary factors; 4) presence of inhibitory factors for T3 action; or 5) some combination of the above. Experiments were conducted to assess the molecular basis for the deficiency of T3 response of the $14 gene in 3T3-F442A cells. Initially, I was interested in determining whether specific T3 binding was present in 3T3-F442A cells. Adipocytes were analyzed by a TR radio-ligand binding assay indicating reduced TR level compared to murine liver and rat white adipose tissue 101 102 (WAT). To assess general biological responsiveness of 3T3-F442A cells to T3, two-dimensional gel analysis of metabolically labeled proteins from cells exposed to T3 were performed. This data suggests other cellular components of 3T3-F442A cells are effected by T3, while 314 is resistant. These results prompted me to identify TR isoform transcripts present in 3T3-F442A cells which might be responsible for translating the TR(s) involved in T3 binding that lead to biological responses. Of the four TR isoforms examined, transcripts for only the TRaz and TRal isoforms were detected (TR32>>TRa1) by Northern analysis. I also examined a known functional 814 T3 response element (pSl4-CAT-4.3) by transfection analysis in 3T3—F442A cells and found it also lacked any T3 control. This study provides information on a novel model system to understand differential T3 control of gene expression. MATERIALS AND METEODS Animals Intact male Swiss mice (25-30 g) were grown, maintained, and killed as described (Lepar and Jump, 1989; Chapter 2). ggii Cultures All cultures were maintained as described (Lepar and Jump, 1989; Chapter 2). 103 Complementapy pg; Probes The pSl4exoPEII-8 genomic probe used was already described in chapter 2. Complementary DNA probes for rat c-erb Aa-1 (TRa-1) and c-erb AB-l (TRB-2) isoforms of the TR were obtained from H. Towle, Minneapolis, MN (Murray et al., 1988). Construction 92 814-Chloramphenicol Acetyl Transgegase Fusion Genes The plasmid used in the DNA transfection studies was pSl4-CAT-4.3. This plasmid contains portions of rat liver $14 genomic DNA fused to the chloramphenicol acetyltransferase (CAT) reporter gene (Chapter 3). Bait-Extracted :3 Receptor Binding Aggpy Nuclei were isolated from confluent 3T3-F442A adipocytes as described (Lepar and Jump, 1989; Chapter 3). Nuclei were prepared from fresh mouse liver as described by Hewish and Burgoyne (1973) with modifications by Jump et a1 (1937). Nuclear T3 receptors were extracted from nuclei with 0.4 M KCl at 4°C and quantitated as previously described by Silva et al. (1977), and Jump et al. (1981), with the following modifications. Salt-extracted nuclear receptors were incubated with [12511T3 [(0.5-20) x 10'10M] in 0.4 ml of Buffer C (15 mM Tris, pH 7.5, 60 mM KCl, 15 mM NaCl, 0.15 mM spermidine, 0.1 mM DTT, 0.1 mM EDTA, 0.5 mM EGTA, 0.25 M sucrose) , 1 mM DTT, 3 mM MgCl2, 1 mM EDTA, 20 mM 104 Na-bisulfite, pH 7.0, and 0.4 M KCl for 2 h at 25°C, then at 0°C for 18 h. Bound [12511T3 was separated from free ligand by treatment with an equal volume of an ion-exchange slurry (Bio-Rad AG-1X8: 160 mg/ml; Buffer C, 0.4 M KCl, 3 mM MgC12, 1 mM EDTA, 1 MM DTT) for 15 min at 4°C with intermitant mixing. [1251]T3 remaining resistant to resin uptake was considered bound to protein. Nonspecific binding of [12511T3 was determined by including 10 uM T3 (nonradioactive) in a tube containing 2 nM [125IJT3. Data were analyzed by the method of Scatchard (1949) to obtain maximum binding capacity and dissociation constants. Metaboiic Labeling Cultures pi 3T3-F442A Adipocytep Metabolic labeling was carried out as described by Spiegelman and Farmer (1982) with the following modifications. Fully differentiated cultures of adipocytes exposed to either vehicle, 10’6M DEX, or 10'5M T3, for 72 h were rinsed twice with 1X-PBS, 4°C, and 10 ml methionine free DMEM was added containing 5 uCi/ml 35S-methionine plus either vehicle, 10’6M DEX, 10’6M T3. .Cells were labeled for 2 h at 37°C. Cultures were harvested in 1X-PBS, spun at 1500 X g (10 min), supernatants decanted, and frozen at -80°C. Samples were then thawed and resuspended in 100 ul Hypotonic buffer (20 mM Tris-Cl, pH 8, 4 mM MgC12, 6 mM CaC12, 0.5 mM DTT) plus 2% NP-40. Samples had an aliquot TCA precipitated and counted by liquid scintillation counting. 105 Two-Dimensional fig; Analysis Aliquots of samples containing 5 x 105 total counts of TCA-insoluble 35S-labelled proteins were separated by two dimensional gel analysis as described by O’Farrell (1975) and O'Farrell et al. (1977) with modifications by Jump et al. (1984). Separation in the first dimension was by equilibrium isoelectric focusing in polyacrylamide gels containing 2% ampholine 3-10 and 3% ampholine 5-7. Separation by molecular weight in the second dimension was in a 15% SDS-polyacrylamide gel. Molecular weight standards were also run and visualized by Coomasie blue staining. Gels containing 35S—labeled proteins were exposed to X-Omat AR-S film to visualize 35S-labeled proteins. Meagurement g; 2335814 Levels Total RNA was extracted, blotted, and quantitated as already described (Lepar and Jump, 1989; Chapter 3). uggppggp Lpalysis p; 13 geceppo; Igoggrmg Poly(A+)mRNA was prepared using Sephadex bound Poly(U) columns. Total RNA at a concentration of 1-2 mg/ml in 1 mM EDTA, pH 7.0, was heat denatured at 70°C for 1 min followed by chilling on ice. RNA sample was added to an equal volume of 2X-Binding Buffer (IX-Binding Buffer: 0.2 M NaCl, 10 mM Tris, pH 7.4, 1.0 mM EDTA, 0.2% SDS) and applied to column. Column was then washed with 1X-Binding Buffer to remove residual protein. Bound RNA is eluted with Elution 106 Buffer (90% formamide, 10 mM Tris, pH 7.4, 1.0 mM EDTA, 0.2% SDS) and recovered by ethanol precipitation (1 ml 4 M NaCl, 1.5 m1 sterile distilled H20, and 20 ml 95% ethanol are added to 7.5 ml column eluate and stored at -20°C for 16 h) and centrifugation. Electrophoresis, capillary transfer to Zeta-bind, hybridization, and autoradiography of mRNA samples were as described previously (Lepar and Jump, 1989; Chapter 2). Preparation 92 stably Transfected 3T3-F442; Cells and CAT Assays 3T3-F442A fibroblasts were stably transfected with the pSl4-CAT-4.3 plasmid in the presence of the selection plasmid, pSV2-Neo, and assayed for CAT activity as already described (Chapter 3). RESULTS :3 Receptor Binding Studies 1 3T3-F442A Adipocytes Cultures of fully differentiated 3T3—F442A adipocytes were examined for TRs, and compared in their affinity and number to those found in rat liver. Scatchard analysis of T3 binding to salt extracted nuclear proteins from 3T3-F442A adipocytes demonstrated low abundance high affinity binding (Figure 19). These TRs bind T3 with a Kd of 156 pH which is similar to the affinity of TRs found in rat liver and WAT (Oppenheimer 1983; Buergi and Abbuehl, 1984). Because salt extracted nuclear receptors were analyzed, it is not possible to determine absolute numbers 107 FIGURE 19. SCATCKARD ANALYSIS OF T3-BINDING IN MOUSE LIVER AND 3T3-F442A ADIPOCYTES. Solubilized nuclear receptors from mouse liver (closed circles) and 3T3-F442A (open circles) adipocytes were analyzed for the level and affinity of TR. Protein extracts from 3 A250 equivalents of nuclei/tube were assayed for TR with increasing concentrations of 5I-T3. B/F, molar ratio of bound to free ligand. B/F 108 SALT EXTRACTED T3 RECEPTORS O NIH Swiss Mouse Liver O NIH 3T3-F442A Adipocytee K0875.6 pM O 5.89 15.3 1\ / / ' I 19‘ Fr 10' I Y ‘ I ' Y ' t I I ' l O 2 4 6 8 10 12 14 16 18 20 T3 Bound (fmoles) 109 of receptors in these tissues. However, the relative concentration of receptors in 3T3-F442A adipocytes was only 30% of that found in mouse liver. Effects pg 23 ppg 9;; pp Protein Biosypthesis ip 313-?442A 21115 To determine if the lack of T3 control was specific for the $14 gene or was a more general phenomenon, fully differentiated cultures of 3T3-F442A cells treated with either vehicle, T3, or DEX for 72 h, were metabolically labeled with 35S-methionine and analyzed for effects at the protein level. Two-dimensional gel analysis of labeled proteins suggests that of 219 identifiable spots, T3 influenced approximately 11.8% of the proteins and DEX influenced approximately 18.3% of the proteins. T3 stimulated accumulation of approximately 12 proteins and inhibited accumulation of approximately 14 proteins. DEX stimulated accumulation of approximately 17 proteins and inhibited accumulation of approximately 23 proteins. Figure 20 (arrow) illustrates an example of one protein that appears unaffected by DEX but is stimulated by T3 when compared to control cultures. These results represent an n=2 for each treatment and need to be confirmed by additional studies. However, they suggest that not all gene products are T3 resistant in 3T3-F442A cells, and therefore, these cells either lack some factor(s) responsible for $14 gene activation or contain inhibitory 110 FIGURE 20. T3 AND DEXAMETHASONE EFFECTS ON PROTEIN BIOSYNTBESIS IN ADIPOCYTES. Fully differentiated 3T3-F442A cells were exposed to vehicle, 1 uM DEX, or 1 uM T3 containing DM for 72 h. Medias were changed to methionine-free DMEM, containing 5 uCi/ml 35S-methionine, plus either vehicle, 1 uM DEX, or 1 uM T3 for 2 h to label proteins. Proteins were extracted and 5 x 10'5 total 358 counts/sample were separated by 2-dimensional gel electrophoresis. These figures represent one of several regions where changes in expression of specific proteins was detected. pH range is across the top and molecular weights are along the sides. n=2. DH 46 5.5 I I 37 kd- o _ 39 kd- V9" V 37 kd— . , . 39 kd- ,. 99" 37 kd- .— .. T 39 kd- v. , 3 112 factors preventing T3 from transducing its signal to control 814 gene expression. Interestingly, the 814 protein was not identifiable in DEX treated adipocytes, indicating it is either absent or present at undetectable levels in 3T3-F442A adipocytes, utilizing this technique. 13 Receptor Isoform Characterization ip 3T3-F442A Adipocytes Because 3T3-F442A cells specifically bind T3, I wanted to determine the TR isoform(s) expressed in adipocytes. Northern analysis (Figure 21) of poly(A+)mRNA from 3T3-F442A adipocytes (lane 1) and rat liver (lane 2) was performed. 3T3-F442A adipocytes and rat liver were examined for expression of the four TR transcript isoforms: c-erb Aal (TRa1)r c-erb Aaz (TRaz), c-erb A31 (TRBl)r and c-erb A32 (TRBZ) (Weinberger et al., 1986; Sap et al., 1986; Lazar et al., 1988; Thompson et al., 1987; Izumo and Mahdavi, 1988; Hodin et al., 1989; Mitsuhashi et al., 1988; Koenig et al., 1988). The full length rat TRal probe was used to identify both the 5.0 kb TRal mRNA and the 2.5 kb TRaz mRNA (Lazar et al., 1988). This probe cross-hybridizes to the TRal and TRaz transcript allowing for simultaneous identification of both of these TR isoforms. 3T3-F442A adipocytes show a much lower abundance of TRa1 mRNA than TRaz mRNA (Figure 21, alpha side, lane 2). Rat liver poly(A+)mRNA was also 113 FIGURE 21. NORTHERN SLOT ANALYSIS OF THYROID HORMONE RECEPTOR ISOFORMS IN RAT LIVER AND 3T3-F442A ADIPOCYTES. Total RNA was extracted from rat liver and 3T3-F442A adipocytes and was enriched for poly(A+)mRNA. Duplicate samples of 20 ug of poly(A+)mRNA from rat liver (lane 1) and 3T3-F442A adipocytes (lane 2) were concurrently electrophoretically separated on the same denaturing 1.2% agarose gel, transferred to Zetabind, and either hybridized with a TRal (Alpha) or TRBl probe (Beta). The Alpha blot was exposed to x-ray film for 48 h and the Beta blot was exposed for 72 h. Rat liver 18S and 288 ribosomal RNAs were visualized by ethidium bromide staining. The size of the TRal is 5.0 kb, TRaz is 2.6 kb, and TRBl is 6.2 kb (Lazar et al., 1988). 114 Alpha Beta 2.5 kb - --18 S- 12 12 VI 115 analyzed for TRal and TRaZ distribution (Figure 21, alpha side, lane 1). Rat liver also demonstrates minimal expression of either TRal or TRaz, and agrees with previously published results (Murray et al., 1988). The level of expression of both alpha isoforms is lower in rat liver when compared to 3T3—F442A adipocytes. Identical poly(A+)mRNA samples as those analyzed for TRal and TRaZ mRNA analysis were analyzed for expression of TRBl and TRBZ transcript isoforms. A full length cDNA probe for the rat TRBl isoform was used that is capable of cross-hybridizing to both the TRBl and TRBZ transcript isoforms. However, since the TRB2 isoform has only been found in the anterior pituitary (Hodin et al., 1989), I did not expect nor detect any cross-hybridization to any mRNA species corresponding to it in 3T3-F442A adipocytes or rat liver (Figure 21). 3T3-F442A adipocytes also lacked expression of the TRbl isoform (6.2 kb; Lazar et al., 1988) (Figure 21, beta side, lane 2). Rat liver had higher levels of TRBl expression than 3T3-F442A adipocytes (Figure 21, beta side, lane 1). These rat liver results also agree with previous observations (Murray et al., 1988). 9;! Analysis 2; Functional Thygoid ggipgpp gggpppgg Elements Transfected ipig 3T3-F142A ggipggyigg To assess whether T3 could regulate an authentic gene harboring a TRE in 3T3-F442A cells, I transfected cells with a plasmid containing 5’-genomic sequences from the rat 116 $14 gene. This genomic fragment includes -4316 bp from the $14 start site and functional thyroid hormone response elements (TREs) at -2.6 to -2.7 kb (Zilz et al., 1990). This sequence was fused to a CAT reporter gene (pSl4-CAT-4.3, Chapter 3) to determine whether T3 stimulation of the rat $14 TREs could drive expression of the pSl4-CAT-4.3 fusion plasmid. Cells stably transfected with pSl4-CAT-4.3 were differentiated and stimulated with T3 (10'5M) or T3 plus DEX (10‘6M) for 24 h to determine whether an exogenous Sl4 promoter could drive expression of the CAT reporter gene. The abundance of mRNASl4 in transfected cultures during these stimulations was examined for comparison. CAT activity and mRNASl4 abundance for vehicle and DEX treated cultures (10'6M, 24 h) from Figure 17 (Chapter 3) are reproduced here for reference. Thyroid hormone did not induce CAT activity in cells transfected with pSl4-CAT-4.3 (Figure 22). The 4-fold increase in mRNASl4 from T3-treated versus vehicle-treated pSl4-CAT-4.3 transfected cells represents a variable response sometimes seen in 3T3-F442A cells. While DEX induced mRNASl4 and CAT activity 62.1-fold and 3-fold, respectively, T3 plus DEX induced mRNASl4 32-fold, and CAT activity 2.3-fold. T3 failed to induce significantly either the endogenous $14 gene (mRNASl4) or the transfected gene (CAT activity). 117 FIGURE 22. T3 AND DEXAMETHASONE CONTROL OF mRNASl4 ABUNDANCE AND THE CAT FUSION GENE CONTAINING 5'-FLANHING SEQUENCES OF THE 814 GENE IN ADIPOCYTES. Cultures of 3T3-F442A fibroblasts stably transfected with pSl4-CAT-4.3 were grown up and differentiated in MM and DM as described in Materials and Methods. Fully differentiated cultures of transfected adipocytes were exposed for 72 h to DM containing vehicle, 1 uM DEX, 1 uM T3, or the combination treatment and analyzed for mRNASl4 and ability to acetylate chloramphenicol. Relative mRNASl4 abundance was measured by dot blot hybridization and presented as units (see Table l for description). Chloramphenicol and its acetylated products were out from TLC plates, counted by liquid scintillation counting, and expressed as percent acetylation of chloramphenicol. n=2. 118 EFFECTS OF T3 AND DEXAMETHASONE ON mRNAs14 ABUNDANCE AND p51 4—CAT-4.3 ACTIVITY mRNAS1 4 Abundance [:3 (Units) 9 .0 . 7‘ I9 N o 01 c an o 01 Vehicle -5 7 DEX 10 M -6 T3 16 M 7‘ ”-6” i “E" ”-6“ E.— .N 01 CAT Activity - (3 Conversion) 0'0 0'93 J 119 DISCUSSION In 3T3-F442A adipocytes, the $14 gene fails to respond to T3 (Lepar and Jump, 1989; Chapter 1). It has also been shown that the $14 gene (Hausdorf et al., 1988) and malic enzyme (Goodridge et al., 1984) are only minimally responsive to T3 in the related 3T3-L1 cell line. These observations are contrary to what is observed in 2129 in murine liver and WAT where the 814 gene and malic enzyme are controlled by T3 (Literature Review C. 3; Wise and Ball, 1964; Silpananta and Goodridge, 1971; Towle et al., 1980). The $14 gene is also regulated by T3 in primary hepatocyte cultures indicating that T3 control of $14 gene expression in hepatocytes is through direct interaction (Mariash et al., 1984). The rapid induction of $14 gene transcription seen 19 2129, as well as, a single DNase I hypersensitive site (Hss-3) upstream from the 814 gene also suggests T3 acts directly on the $14 gene. Jump (1989a: Jump et al., 1990a) suggested the cis-linked T3-regulated DNase I hypersensitive site (Hss-3) harbors a TRE controlling 814 gene transcription, which has since been supported by transfection analysis (Zilz et al., 1990). Thyroid hormone resistance of the $14 gene in 3T3-F442A cells cannot be explained by a total absence of TR capable of transducing the hormone signal. Adipocytes demonstrate low capacity high affinity T3 binding with Kd values (0.156 x 10’9M) similar to those reported in murine 120 liver and WAT (Figure 19). However, the actual number of TRs in 3T3-F442A adipocytes is approximately only one-third of the 4000 TRs found in murine liver and WAT (Oppenheimer, 1983; Buerghi and Abbuehl, 1984). This diminished TR number may be, in part , responsible for the lack of T3 control of the 814 gene in 3T3-F442A cells. Vanderbilt et al. (1987), have reported for the glucocorticoid receptors (GR), that changes in biological responsiveness to glucocorticoids directly relates to the level of GR present. Interestingly, 2-dimensional gel analysis suggests that T3 may regulate other gene products in 3T3-F442A adipocytes (Figure 20). This implies that T3 can differentially regulate specific genes in 3T3-F442A adipocytes. A similar type of selective T3 response is found in 3T3-L1 adipocytes. In 3T3-L1 cells, T3 only minimally stimulates the $14 gene (Hausdorf et al., 1988) and malic enzyme (Goodridge and Fisch, 1984), while significantly increasing basal and isoproterenol stimulated lipolysis and inhibiting cAMP phosphodiesterase activity (Elks and Manganiello, 1985). Although TR number in 3T3-L1 cells was not determined in these other studies, reduced TR levels may also contribute to their discordant T3 control Goodridge and Fisch, 1984). Recent identification of four TR isoforms, TRal, TRaz, TRBl, and TRBZ (Weinberger et al., 1986; Sap et al., 1986; 121 Lazar et al., 1988; Thompson et al., 1987; Izumo and Mahdavi, 1988; Hodin, 1989; Mitsuhashi et al., 1988; Koenig et al., 1988), raised the question of which TR isoform(s) was responsible for T3 binding in 3T3-F442A cells (Figure 19) leading to biological responses (Figure 20). Thyroid hormone receptor identification is possible only at the transcript level, and inferences can only be made to the abundance of actual receptors synthesized from the TR transcripts. In the rat, while TRs are found in all tissues (Oppenheimer, 1983) transcripts of the TRal are most abundant in skeletal muscle and brown fat, TRaz in brain and hypothalamus (Lazar et al., 1988), TRBl in kidney and liver (Koenig et al., 1988), and TRBZ only in the anterior pituitary (Hodin et al., 1989). Northern analysis of these TR transcript isoforms in 3T3-F442A adipocytes (Figure 21) showed TRa2 predominantly expressed over TRalr with no detectable expression of TRBl or TRsz- If TR transcript levels are any indication of TR protein levels, this would suggest the majority of TR translated in 3T3-F442A adipocytes are TRaz. Interestingly, only TRal, TRBl, and TRBZ' are capable of T3 binding (Weinberger et al., 1986; Sap et al., 1986, Thompson et al., 1987; Hodin et al., 1989; Koenig et al., 1988; Murray et al., 1988) and presumably transducing the hormone signal after ligand binding (Glass et al., 1987; Forman et al., 1989; Thompson et al., 1989). Since the only ligand binding TR transcript 122 found in 3T3-F442A cells is TRalr this TR is probably involved in T3 binding and mediating T3 responses. Predominant expression of TRaz transcripts in 3T3-F442A adipocytes may also be significant in blocking T3 control of the $14 gene in 3T3-F442A adipocytes. The TRaz appears capable of binding a TRE, but unable to bind T3 (Izumo and Mahdavi, 1988; Lazar et al., 1988), thereby, effectively making the TRa2 a silencer of T3 action (Koenig et al., 1989). A combination of reduced TR number and predominant expression of an inactive TR isoform may account for lack of T3 control of the $14 gene in 3T3-F442A adipocytes. A similar observation of high levels of TRaZ mRNA expression in 3T3-L1 adipocytes is noticed when data of Murray et al. (1988) is examined, possibly contributing to the T3 resistance of the $14 gene and malic enzyme in these cells. The mechanism(s) by which the $14 gene fails to respond to T3 in 3T3-F442A cells remains unclear. The lack of T3 control of transfected rat $14 promoter elements in 3T3-F442A adipocytes (Figure 22) suggests that while levels of TRal present in 3T3-F442A cells are sufficient to regulate some gene products (Figure 20), TRal either: 1) does not control 814 gene expression; 2) is not expressed in high enough abundance; 3) is blocked in its action by high levels of TRaz; 4) lacks some ancillary factor; 5) is inhibited by some factor expressed in 3T3-F442A cells; or 123 6) some combination of the above. Future experiments where functional TR isoforms are transfected into 3T3-F442A cells will assess these possibilities. It should also be remembered that 3T3-F442A cells were originally isolated from embryonic tissue (Todaro and Green, 1963; Green and Kehinde, 1976), which might explain some of the reasons 3T3-F442A cells respond differently from adult mouse tissue. Interestingly, TR isoforms are regulated during chick development, and this coincides with T3 responsiveness of certain tissues during this period (Forrest et al., 1990). Possibly TR receptor distribution in 3T3-F442A cells is not representative of what is seen 19 2129 in adult mouse white adipose tissue. Examination of T3 resistance of the $14 gene in 3T3-F442A cells may provide a unique example for understanding how T3 can differentially regulate specific genes in a tissue selective manner during development. Chapteg 9. Retinoic 99id and Dexp9ethasone Iptgract 99 Reggiate 814 Gene Transcription 19 3T3- F442A Adipocytes INTROQUCIION The retinoids are a group of natural and synthetic compounds that have profound effects on differentiation, and growth and development in a wide variety of systems (for reviews see Mandel, 1985; Sporn et al., 1984). These compounds have been reported to suppress carcinogenesis 19 2129 (Sporn and Roberts, 1983; Roberts and Sporn, 1984) and effect pattern formation in developing and regenerating limbs (Maden, 1985). 19 21999 studies show that retinoids promote differentiation of several cell types (Breitman et al., 1980; Lotan and Lotan, 1980; Strickland et al., 1980; Kuri-Harcuch, 1982) while blocking differentiation of others (Pacifici et al., 1980; Kuri-Harcuch, 1982; Castro- Munozledo et al., 1987; Pairault and Lasnier, 1987). Retinoic acid (RA), a biologically active retinoid, initiates action at the cellular level by binding to specific receptors which belong to the steroid/thyroid hormone receptor supergene family (see Literature Review B). Several retinoic acid receptor (RAR) isoforms have 124 125 been identified through molecular cloning (Giguere et al., 1987; Petkovich et al., 1987; Benbrook et al., 1988; Brand et al., 1988; Zelent et al., 1989; Kastner et al., 1990; Mangelsdorf et al., 1990). As ligand-activated transcription factors, RARs initiate changes in gene expression by binding to cis-linked response elements (Umesono et al., 1988; Graupner et al., 1989; Vasios et al., 1989; de The et al., 1990). Umesono et al. (1988), reported that RA activated expression of an artificial gene through thyroid hormone response elements (TRE) and suggested that T3 and retinoids regulate overlapping gene networks. The induction of growth hormone in cultured pituitary cells by both T3 (Samuels et al., 1988) and RA (Bedo et al., 1989; Morita et al., 1989) is in agreement with the view that both receptor types might activate gene expression through a common TRE. Because of the overlap in RA and thyroid hormone (T3) action, I considered the possibility that other T3- responsive genes were affected by RA. Expression of the rat liver $14 gene is regulated by T3, and has been well characterized (Literature Review C.3.a). Thyroid hormone rapidly induces Sl4 gene transcription through induction of a DNase I hypersensitive site containing functional TREs located between 2.6 and 2.7 kb upstream from the transcription start site of the 814 gene (Jump, 1989a, 1989b; Jump et al., 1990a; Zilz et al., 1990). In order to 126 determine whether the $14 gene was responsive to RA, I examined the effects of RA on 814 gene expression in the 3T3-F442A adipocyte cell line. 3T3-F442A cells are both responsive to RA (Kuri-Harcuch, 1982; Castro-Munozledo et al., 1987; Pairault and Lasnier, 1987) and express the $14 gene (Lepar and Jump, 1989; Chapter 2). In this report, I characterize the RA regulation of 814 gene expression, and identify the RAR isoform expressed in 3T3-F442A adipocytes. Retinoic acid was found to have significant effects on $14 gene expression in 3T3-F442A adipocytes, suggesting that RA action may extend beyond growth and differentiation to include regulation of adipocyte lipid metabolism. MATERIALS 992 METHODS 9911 Cultures All cultures were maintained as previously described (Lepar and Jump, 1989; Chapter 2). Complementagy 999 Probes The genomic pSl4exoPEII-8 probe, and cDNA probes for fatty acid synthase, glycerophosphate dehydrogenase, B-actin (Chapter 2), and preproinsulin II (pRCII) (Chapter 3) were as described. A cDNA for phosphoenolpyruvate carboxykinase (PEPCK-IO) was generously provided by R. Hanson (Lamers et al., 1982). Complementary DNA probes for for mouse RARa and RARE were generously provided by P. Chambon (Zelent et al., 1989). The method of plasmid isolation was as described (Jump, 1988). 127 Construction 91 814-Chloramphenicol Acetyl Transferase Fusion Genes The plasmid used in the DNA transfection studies was pSl4-CAT-4.3. This plasmid contains portions of rat liver Sl4 genomic DNA fused to the chloramphenicol acetyltransferase (CAT) reporter gene (Chapter 3). Measurement 91 9999314 Levels Total RNA was extracted, blotted, and quantitated as described (Lepar and Jump, 1989; Chapter 3). Northern Analysis 91 RAR Isofog9s Poly(A+)mRNA was obtained from 3T3-F442A adipocytes and mouse liver, electrophoretically separated, hybridized, transferred, and visualized as described (Chapter 3). T1anscription Nuclei were isolated from confluent 3T3-F442A adipocytes (Lepar and Jump, 1989; Chapter 3), adjusted to 60 A260/m1 and stored at -80°C for transcriptional analysis. 19 vitro transcriptional run-on analysis of the Sl4 gene was as previously described (Jump, 1990a; Lepar and Jump, 1989), with the following modifications. The composition of the the transcription buffer was: 25% glycerol; 75 mM HEPES, pH 7.5; 3 mM MgC12; 100 mM KCl; 0.1 mM EDTA; 0.05 mM EGTA; 1.0 mM spermidine; 1 mM DTT; 16 ug/ml creatine kinase; 100 ug/ml creatine phosphate; 1 mM ATP; 0.5 mM CTP; 0.5 mM GTP; 50 uCi 32P-UTP at 0.25 uM; 120 128 units/ml RNasin; 12 A260 units of nuclei in a final volume of 300 ul. Transcriptional assays were run at 30°C for 45 min followed by DNase I treatment (15 min). Ribonucleic acid was purified by proteinase K treatment followed by phenol:chloroform (1:1) extraction and chloroform:isoamyl alcohol (49:1) extraction and isopropanol precipitation. 32P-RNA (6-10 x 106 cpm/blot) was hybidized to cDNAs affixed to nitrocellulose for 72 h at 42°C. Following hybridization, blots were washed and exposed to x-ray film. Levels of hybridization were quantified by B-scintillation counting and videodensitometry. Preparatio9 91 stably Transfected 3T3-F442A Cells and CAT Assays 3T3-F442A fibroblasts were stably transfected with the pSl4-CAT-4.3 plasmid in the presence of the selection plasmid, pSV2-Neo, and assayed for CAT activity as described (Chapter 3). RESULTS 9999814 19 3T3-F442A Adipocytes The effect of RA on $14 gene expression was examined by treating fibroblasts and adipocytes with either RA, DEX, or a combination of both treatments for 24 h and measuring changes in mRNASl4 levels. In fibroblasts, mRNASl4 was expressed at low levels (<0.01 Units) and remained unaffected by these treatments (Figure 23). The basal 129 FIGURE 23. RETINOIC ACID AND DEXAMETHASONE INTERACT TO REGULATE mRNASl4 ABUNDANCE IN ADIPOCYTES. Fully differentiated 3T3-F442A cells were exposed to DM containing either vehicle, 10'10M DEX, 10'6 M DEX, 10'5M RA, 10'10M DEX + 10'5M RA, or 10‘5M DEX + 10'5M RA for 24 h. Cells were harvested and analyzed for the expression of mRNASl4 by dot blot hybridization. Results are expressed as Units (see Table 1 for description), mean 1 SE, n=3-9. 130 EFFECTS OF RETINOIC ACID AND DEXAMETHASONE ON mRNAS14 ABUNDANCE IN 3T3-F442A ADIPOCYTES mRNAs1 4 Abundance 0: n-3 1 .2 - 0.: n-5 m: n-9 , _I_ ’5 'E 0 61 a e j _L, 1 0.0 .1. I L 1: VEH DEX (M): 10‘10 1o" 1o"° 10"“ RA (M): 10“ 1O“ 1o“5 131 mRNASl4 expression in 3T3-F442A adipocytes was 0.01 Units and treatment of adipocytes with DEX at 10"10 and 10'6M induced mRNASl4 5- and 35-fold, respectively, consistent with previous observations (Lepar and Jump, 1989; Chapter 2). Adipocytes treated with RA at 10'5M showed only a marginal 3-fold increase in mRNASl4. However, cells receiving RA (10’6M) in the presence of DEX at 10'10 or 10' 6M showed a 45- and 100-fold increase in mRNASl4 levels, respectively. When compared to the marginal induction by RA (10‘5M) or DEX (lO'lOM) alone, the z40-fold induction in mRNASl4 with the combination of these treatments illustrates these agents act synergistically to regulate Sl4 gene expression in 3T3-F442A adipocytes. While RA induces major effects on $14 gene expression, this induction is dependent on both expression of the adipocyte phenotype and presence of glucocorticoids. E11ects 91 Retinoic Acid 99 mRNAsl‘ Leveig 19 gzg-F142A Adipocytes 19 Dose-Dependent The analysis of RA effects on $14 gene expression required low levels of DEX (10'10M). The 3050 for DEX induction of mRNASl4 is 4 x 10'10M (Lepar and Jump, 1989; Chapter 3). Accordingly, adipocytes were treated for 24 h with DEX at 10’1°M in the presence of varying concentrations of RA, ranging from 10'”11 to 10'6M. While levels of mRNASl4 in adipocytes exposed to 10'11 or 10'1°M 132 RA (plus lO‘lOM DEX) were not different from 10'10M DEX alone, a linear increase in mRNASl4 abundance was detected between 10'9 and 10'7M RA and maximal levels of expression was obtained at 10’7M RA (Figure 24). The ED50 of 3.5 x 1079M RA correlates well with the reported biological effects of RA (Sporn et al., 1984; Castro-Munozledo et al., 1987) and binding of RA to RARs (Petkovich et al., 1987). Therefore, RA-stimulated increases in 814 gene transcription are consistent with a RAR-mediated response. 9999314 19 3T3-F442A Adipocytes The rapid effects of RA on adipocyte gene expression was compared to the effects of DEX by treating cells with either 10‘5M RA, 10'5M DEX, or 10‘10M DEX plus 10'5M RA. Treatment of cells with RA alone induced only a 7-fold increase after 72 h (Figure 25). Dexamethasone at 10'6M induced mRNASl4 18-fold within 4 h, representing 30% of the maximal response measured after 72 h. Since DEX at 10'10M (Figure 23) gives a small increase in mRNASl4, the major RA-mediated induction of mRNASl4 measured at 4 h indicates that RA acts directly to affect $14 gene expression in adipocytes. These results suggest that RA increases the magnitude but not the rate of mRNASl4 accumulation. 133 FIGURE 24. DOSE RESPONSE RELATIONSHIP FOR RETINOIC ACID INDUCTION or mRNASl, IN ADIPOCYTES. Fully differentiated 3T3-F442A cells were exposed to DM containing 10"10 DEX plus RA at concentrations ranging from 10’11 to 10’6M for 24 h. Cells were harvested and mRNASl4 abundance determined by dot blot hybridization. Results are expressed relative to the maximal response achieved at 10’7M RA; mean 1 SE, n=4. % Maximal Response 134 24 HOUR RETINOIC ACID DOSE RESPONSE IN .1 nM DEXAMETHASONE TREATED 3T3-F442A ADIPOCYTES 1206 100< ofl 80. / 60- 3 I ._ EDso-3.5x10—9M RA 404 20- 0 fl ' "r'wt ' ' """I ‘ ' """U f' """I ' ' ' ""'l r ' "fi'w 10"11 10‘” 10‘“ 10" 10" 10'7 10" [Retinoic Acid] (M) 135 FIGURE 25. KINETICS OF RETINOIC ACID AND DEXAMETHASONE INDUCTION OF mRNASl4 IN ADIPOCYTES. Fully differentiated 3T3-F442A cells were exposed to DM containing 10’5M RA (open circles), 10'5M DEX (closed circles), or 10’5M RA + 10'5M DEX (open squares). Cells were harvested after the indicated times of exposure and the relative abundance of mRNASl4 was determined by RNA dot blot hybridization. Values are expressed as Units (see Table 1 for description), mean 1 SE, n=4. 136 KINETICS OF RETINOIC ACID AND DEXAMETHASONE STIMULATION ON mRNAs14 ABUNDANCE IN 3T3-F442A ADIPOCYTES mRNA51 4 Abundance Hours 137 If.» Ret c apd Degamgthasone 1nte1act 19 19dUC9 911 Cape Transcription 1_ 3T3-F442A Adipocytes We previously reported that DEX induced $14 gene expression at the transcriptional level (Lepar and Jump, 1989; Chapter 3). Nuclear run-on analysis was used to determine whether RA affected 814 gene expression at the transcriptional or post-transcriptional level of control. Differentiatied adipocytes were incubated without or with DEX and RA for 72 h (Figure 26). Cells were harvested to measure mRNASl4 levels and transcriptional run-on activity. The transcriptional dot blot in Figure 26. A, shows that, in addition to 814, we measured the relative transcription rate of fatty acid synthase (FAS), B-actin (Actin), glycerol phosphate dehydrogenase (GPD), preproinsulin II (PPI), and phosphoenolpyruvate carboxykinase (PEPCK). Run-on transcription rates of FAS (42 i 5.6 ppm), B-actin (81 i 6.2 ppm), and PEPCK (4.2 i 0.3) were not significantly affected by any treatment indicating that neither RA nor DEX induced generalized changes in adipocyte gene transcription. Transcription of PPI was not detected indicating the specificity of the hybridization reaction. In contrast, GPD transcription was inhibited 60% in all treatments receiving RA. This decline in transcription paralleled a similar decline in mRNAGPD (not shown). Therefore, RA has inhibitory effects on the expression of this lipogenic enzyme. 138 FIGURE 26. DEXAMETHASONE AND RETINOIC ACID INTERACT TO REGULATE 814 GENE TRANSCRIPTION IN ADIPOCYTES. Fully differentiated adipocytes were treated with vehicle, 10'10M DEX, 10'6M DEX, 10'5M RA, 10’10M DEX + 10’ 6M RA, or 10'5M DEX + 10‘5M RA for 72 h. Cells were harvested and extracted for RNA and nuclei. Isolated nuclei were added to a run-on transcription reaction. Isolated 32P-RNA (6-10 x 106 CPM) was hybridized to blots containing cDNAs affixed to nitrocellulose. Following hybridization, blots were washed and exposed to x-ray film. A transcription dot blot illustrating the relative level of transcription of several genes is illustrated in 9. The cDNAs affixed to nitrocellulose at the 6 positions are: 1: $14; 2: fatty acid synthase (FAS); 3: B-actin; 4: glycerol phosphate dehydrogenase (GPD); 5: prepreinsulin II (PPI): and 6: phophoenolpyruvate carboxykinase (PEPCK). B. The transcription results were quantified by cutting individual dots and B-scintillation counting and videodensitometry. Results are expressed as ppm, mean of duplicate samples. mRNASl4 abundance was determined by dot blot hybridization and expressed as Units, mean of duplicate samples. The variation between samples was <10%. 139 9 10°10 10-6 8. 10" 10" 1 10 A. 1. {D ‘. I‘D :I: q. 0 e DEX. (N) - 10'10 10" - RIT.A.(N) - . ; 10-6 KEY: 1 $14 4 GP 2 FAS 5 PF? 3 ACT!" 6 PEPCK B. I, 1.5- ‘A 10 .. . r 1"! ii CLSIF IDA! Ben. (I) - 10'1' 10" .. 10". 10“ 11.7.1.1» - -- - n" u" a" $14 Nuclear Run—On Activity 01*?» (III!) 140 Figure 26.8, illustrates the effects of DEX and RA on $14 gene expression. Basal Sl4 run-on activity and mRNASl4 in untreated adipocytes was 0.1 ppm and 0.01 Units, respectively. Dexamethasone at 10"10 and 10’6M induced Sl4 gene transcription 4.8- and 32-fold, respectively, and mRNASl4 levels 2.5- and 45-fold, respectively. Retinoic acid at 10'6M induced 814 gene transcription 5-fold and mRNASl4 5.2-fold. The combination treatment of RA (10’6M) with DEX at either 10’10M or 10’6M induced 814 gene transcription 27- and 84-fold, respectively, and mRNASl4 39- and 131-fold, respectively. While DEX treatment alone significantly induces 814 gene transcription, RA required the presence of at least 10'10M DEX to induce major changes in $14 gene transcription. The nearly parallel induction of 814 gene transcription and mRNASl4 suggest that neither RA nor DEX induces major changes in 814 gene expression by activating post-transcriptional mechanisms. Therefore, the principal target for both DEX and RA action is at the transcriptional level. Retinoic Acid Receptor Isoform Analysis I next examined 3T3-F442A adipocytes for expression of RAR transcript isoforms to determine which RAR might be mediating RA effects (Figure 27). Mouse liver was also examined for comparison. Poly(A+)mRNA from mouse liver (Lane 1) and 3T3-F442A adipocytes (Lane 3) was size- separated, transferred to nylon membranes, and hybridized 141 FIGURE 27. NORTHERN SLOT ANALYSIS OF RETINOIC ACID RECEPTOR ISOFORMS IN RAT LIVER AND 3T3-F442A ADIPOCYTES. Poly(A+)mRNA isolated from mouse liver (7.5 ug) and 3T3-F442A adipocytes (15 ug) was electrophoretically separated in 1.2% agarose gels containing 2.2 M formaldehyde and 1X-MAE as buffer. Following electrophoresis, mouse liver (Lane 1) and 3T3-F442A adipocytes (Lane 3) RNA was transferred to Zetabind and hybridized with either 32P-labeled cDNARAR or 32P-labeled CDNARARB- Lane 2 is blank. The blot hybr1dized with cDNA a was exposed to x-ray film for 72 h, while the blot hybr1d1zed to CDNARARB was exposed to x-ray film for 10 days. Rat liver 188 and 288 ribosomal RNAs served as standards and were visualized by ethidium bromide staining. 142 Alpha Beta - 28 S - n 43.8 kb ‘3.2 kb 3.8 kb - 2.8 kb - -18$- 143 with either RARa or RARE 32P-labeled cDNA probes. The mouse RARa cDNA hybridized to two species of mRNA in both mouse liver and 3T3-F442A adipocytes (Figure 27). These mRNAs are 2.8 and 3.8 kb in length and correspond in size to RARa mRNAs reported by Zelent et a1 (1989). The high RARa hybridization signal in 3T3-F442A adipocytes suggests RARa may be expressed at higher levels in adipocytes than liver. In contrast, nucleic acid hybridization analysis using RARE cDNA showed detected mRNAs of 3.6 and 3.2 kb expressed in mouse liver with no detectable RARE expression in 3T3- F442A cells. Based on the hybridization analysis, the RAR isoform expressed in 3T3-F442A adipocytes is the a- phenotype. We did not determine whether RARg (Zelent et al., 1989; Kastner et al., 1990) or other RAR isoforms (e.g. RXRa; Mangelsdorf et al., 1990) were present in these cells. CAT Analysis 91 814 Gene SI-Flanking Regions for Retipoig 9919 Response Elements To assess whether RA effects on mRNASl4 abundance occurred within -4316 bp of the start site of 814 gene transcription, differentiated cultures of fibroblasts stably transfected with the Chimeric pSl4-CAT-4.3 plasmid (Chapter 2), were examined for RA responsiveness. This plasmid uses -4316 bp of rat $14 promoter elements to drive expression of the CAT reporter gene. Cultures were 144 differentiated and stimulated with 10'6M RA or 10'5M RA plus 10’5M DEX for 24 h. These treated cultures were also examined for mRNASl4 abundance by dot blot hybridization. Chloramphenicol acetyltransferase activity and mRNASl4 abundance for vehicle and DEX (lo-GM, 24 h) treated cultures from Figure 17 (Chapter 3) are reproduced here for comparison. Cultures of transfected adipocytes exposed to 10'5M RA or 10‘5M RA plus 10-10M DEX demonstrated no CAT stimulated activity (Figure 28). However, levels of mRNASl4 increased 6.58- and 33.58-fold relative to minimal expression detected in vehicle treated cultures. These results suggest that activated RARs are not acting within -4316 bp of the $14 transcriptional start site to stimulate $14 gene transcription. DISCUSSION Retinoic acid regulation of $14 gene expression in 3T3-F442A cells was dependent on the adipocyte phenotype and was potentiated by the glucocorticoid agonist, DEX. Nuclear run-on analysis showed that nearly all the increase in mRNASl4 induced by RA or DEX could be accounted for by activation of 814 gene transcription (Figure 26). Although glucocorticoids control 814 gene expression within -4316 bp of the start site of the $14 gene, RA effects are directed elsewhere (Figure 28). These results indicate that a reasonable indication of the effects of RA and DEX on 814 145 FIGURE 28. RETINOIC ACID AND DEXAMETHASONE CONTROL OF mRNASl4 AND THE CAT FUSION GENE CONTAINING 5'-FLANRING SEQUENCES OF THE 814 GENE IN ADIPOCYTES. Cultures of 3T3-F442A fibroblasts stably transfected with pSl4-CAT-4.3 were grown up and differentiated in MM and DM as described in Materials and Methods. Fully differentiated cultures of transfected adipocytes were exposed for 72 h to DM containing vehicle, 10'6M DEX, 10’ 6M RA, or 10'6M RA + 10'10M DEX. These cultures were harvested and analyzed for mRNASl4 and ability to acetylate chloramphenicol. Relative mRNASl4 abundance was measured by dot blot hybridization and presented as Units (see Table 1 for description). Chloramphenicol and its acetylated products were cut from TLC plates, counted by liquid scintillation counting, and expressed as percent acetylation of chloramphenicol. n=2. 146 EFFECTS OF RETINOIC ACID AND DEXAMETHASONE ON mRNA514 ABUNDANCE AND p514-CAT-4.3 ACTIVITY MRNA51 9, ”004006. E (Unite) .° .6 .-‘ .-'- N N 9 1 ‘1' 11 ‘3‘ 1 9 - 1“ Vehicle — w no: 10" u RA 10" u M 10411 + no: 10"°u ‘ N in CAT Activity - (I Conversion) 0'0 ‘ 0'93 ‘ 147 gene transcription can be obtained by measuring mRNASl4 directly. While RA treatment of adipocytes induced only marginal changes in 814 gene expression, DEX potentiated the RA response. For example, treatment of adipocytes with 1 uM RA or 0.1 nM DEX induced Sl4 gene expression 3- and 5-fold, respectively. However, when cells were treated with both RA and DEX at these concentrations, mRNASl4 was induced 45- fold, indicating that RA and DEX interact synergistically to regulate $14 gene expression in adipocytes (Figure 23). In order to characterize the RA effects on $14 gene expression, we took advantage of the finding that DEX potentiated RA action. In this experiment, RA (1 uM) plus Dex (0.1 nM) induced mRNASl4 50-fold within 4 hr indicating that RA acted rapidly on the $14 gene (Figure 25). The RA effect was dose-dependent with an ED50 = 5 nM (Figure 24), a value that correlates well with the biological effects of RA (Sporn et al., 1984) and the binding of RA to its receptor (Petkovich, 1987). Receptor isoform analysis indicated that RARa, but not RARB, was expressed in 3T3- F442A adipocytes (Figure 27). Based on these observations, RA regulation of $14 gene transcription in 3T3-F442A adipocytes appears mediated, at least in part, by RARa. We can not exclude the possibility that other RAR isoforms (Zelent et al., 1989; Mangelsdorf et al., 1990; Kastner et al., 1990) are present in 3T3-F442A cells. 148 Since RARs sometimes utilize TREs to regulate gene transcription (Umesono et al., 1988; Forman et al., 1989; Graupner et al., 1989), a prospective target for RA action on the 814 gene is the TRE. In rat liver, the TREs controlling Sl4 gene transcription are located between -2.6 and -2.7 kb upstream from the $14 transcription start site (Jump, 1990a; Zilz et al., 1990). Because the mouse hepatic 814 gene is T3 responsive (Lepar and Jump, 1989; Chapter 2), I speculate the TREs mediating T3 control of $14 gene transcription in the mouse are also located far upstream. However, in mouse derived 3T3-F442A cells, the 814 gene is refractory to T3 control, despite the apparent presence of at least some functional T3 receptors (Lepar and Jump, 1989; Chapter 4). To assess whether RA was controlling $14 gene expression through the $14 TREs, transfected 3T3-F442A cells containing the pSl4-CAT-4.3 chimeric rat 814-reporter plasmid were stimulated with RA (Figure 26). Since RA failed to stimulate expression of this reporter plasmid, those sequences of the $14 promoter contained within the pSl4-CAT-4.3 plasmid, do not contain the sequences necessary for mediating RA effects. These results suggest that RA does not utilize the $14 TREs to activate 814 expression, but may utilize distinct RAREs, such as those reported for the laminin B-l (Vasiois et al., 1990) and RARE gene (de The et al., 1990; Sucov et al., 1990). This regulatory sequence would appear to be located 149 either further upstream, downstream, or within the $14 gene. Alternatively, the contribution of trans-acting factors to $14 gene expression in transfected 3T3-F442A cells is not known, and may be involved in successful $14 expression. Additional transfection analysis experiments are being performed to understand this. A striking feature associated with RA regulation of $14 gene transcription in adipocytes was that low levels of DEX potentiated the RA response. Interaction of RA with DEX was not a generalized requirement for RA action in 3T3- F442A cells since GPD transcription was equally inhibited by RA in the presence and absence of DEX (Figure 26.A). The mechanism reponsible for this interaction is unclear. However, it appears that glucocorticoid receptors interact within -4316 bp of the start site of the $14 gene (Chapter 3; Figure 28), while RARs effects are mediated through different DNA sequences either located further upstream, downstream, or within the $14 gene (Figure 26). It is not known whether direct interaction between these sites is responsible for synergistic 814 control, or the role of ancillary trans-acting factors in regulating expression. Glucocorticoid receptors have been shown to interact with other trans-acting factors to promote synergistic regulation of gene transcription (Schule et al., 1988; Strahle et al., 1988; Ron et al., 1990). In order to understand the mechanism of RA and DEX control of $14 gene 150 expression in 3T3-F442A cells, additional transfection studies utilizing 814-CAT fusion genes are being conducted. Retinoic acid is known to inhibit conversion of 3T3- F442A preadipocytes to adipocytes implicating a role for RA in adipocyte differentiation (Kuri-Harcuch, 1982;Castro- Munozledo et al., 1987; Pairault and Lasnier, 1987). However, once these cells are fully differentiated to adipocytes, RA regulates the expression of at least two genes involved in adipocyte lipid metabolism. Although the 814 protein has been postulated to function in lipid metabolism (see Literature Review C), GPD is a known lipogenic enzyme that plays a key role in fat triglyceride synthesis (Dobson et al., 1987). Retinoic acid inhibits GPD gene transcription, while stimulating Sl4 gene transcription (Figure 26). The effects of RA on adipocyte metabolism are unknown. 19 2129 studies with rats suggest that RA effects on serum triglycerides are mediated through increased hepatic triglyceride synthesis (Gerber and Erdman, 1980). Studies using cultured hamster fibroblasts show that RA augments palmitic acid synthesis (Ringler et al., 1984). While these 19 2129 and 19 21119 studies support a role for RA in lipid metabolism, our studies provide new evidence showing that RA may regulate adipocyte lipid metabolism through direct action at the genomic level. Whether RA effects on adipocyte gene expression 151 parallel changes in lipid metabolism and whether these effects mimic 19 vivo events remains to be determined. Chgpteg 9. 9999912 and Conciusions The primary goal of this dissertation was to establish an 19 vitro cell culture system for conducting studies on hormonal regulation and function of the $14 gene. The importance of finding a cell culture line for performing controlled experiments on $14 gene expression is significant due to the complex interdependent regulation of the $14 gene found 1 vivo (see Literature Review C.). My dissertation demonstrates the utility of the adipocyte- forming 3T3-F442A cell line for conducting 814 regulation and functional studies. Specifically, I have demonstrated 3T3-F442A cells provide a controlled environment to study developmental, tissue-specific, positive glucocorticoid and RA, negative cAMP-mediated, and unique aspects of T3 control of 814 gene expression. The results of the lipogenesis and lipolysis assays in 3T3-F442A cells will form the basis of future studies to ascertain the unknown function of the 814 protein. This dissertation demonstrates for the first time a non-neoplastic clonal cell line for conducting more controlled studies on 814 function and certain aspects of 814 gene expression. 152 153 While the $14 gene is not expressed under basal conditions in either fibroblast or adipocyte 3T3-F442A cells, 814 is expressed after treating adipocytes with DEX or RA. Dexamethasone or RA had no effect on $14 expression in preadipocytes. Thus hormonal control of 814 expression requires that 3T3-F442A cells differentiate to adipocytes (Table 1). Because DEX and RA actions are directed primarily at the $14 transcriptional level, with only minor effects to increase stability of mRNASl4 (Chapters 3 and 5), differentiation of 3T3-F442A cells must include changes in the cellular components necessary for transcriptional activation of the $14 gene by these agents. 19 2129, the $14 gene also requires certain developmental changes in order to be properly expressed after weaning of young rats (Jump et al., 1986, 1988). Therefore, 3T3-F442A cells should provide a useful model to better understand developmental control of $14 gene expression seen 19 2129. Several DNA regions are tentatively identified by DNase I Hss-analysis to play a role in developmental control of 814 gene expression. Changes in chromatin structure are indicated by the appearance of DNase I Hypersensitive site (Hss)-1 and Hss-3, located at -65 to - 265 and -2670 bp, during weaning in the rat (Jump et al., 1988). Additionally, transfection analysis of the $14 promoter shows that elements within -4316 kb of the 5’-end of the $14 gene participate in developmental control of $14 154 gene expression in 3T3-F442A adipocytes (Figures 17 and 18). Fibroblast cultures transfected with a chimeric plasmid containing -4316 bp of $14 5'-flanking sequences cloned to the CAT reporter gene failed to demonstrate any response (Chapter 3). However, adipocyte cultures transfected with this plasmid show both developmental and DEX stimulation of the 814-CAT reporter gene (Chapters 3 and 5). These combined results indicate.that 814 DNA regulatory sequences for developmental and glucocorticoid activation are located within these sequences. This dissertation provides the first demonstration of significant glucocorticoid stimulation of $14 gene expression both 19 2129 (mouse) and 19 21119 (Chapter 2; Lepar and Jump, 1989). Interestingly, DNA sequence analysis of 5’-regulatory regions of the $14 gene shows two regions with high homology to the canonical glucocorticoid response element (GRE) responsible for mediating glucocorticoid effects on other genes (Jantzen et al., 1977). These sequences are located at -674 and -2643 bp upstream from the $14 gene. Further transfection analysis of 814 promoter elements containing these putative GREs in 3T3-F442A cells will determine their significance in glucocorticoid control of the $14 gene. Results of these studies do not indicate where RA is acting to stimulate $14 gene expression (Chapter 5). Stimulatory transcriptional effects of RA on other genes 155 can be mediated through a distinct retinoic acid response element (RARE) such as for the laminin B-1 and RARB gene (Vasios et al., 1989; de The et al., 1990; Sucov et al., 1990), or through interaction with TREs such as reported for the growth hormone gene (Forman et al., 1989; Glass et al., 1990). Initial transfection analysis of 814 promoter sequences in 3T3-F442A cells indicates RA is acting through different sites (Chapter 5) than the described 814 TREs (Jump et al., 1990a; Zilz et al., 1990). In fact, RA does not appear to be regulating Sl4 gene expression within 4316 bp 5’ to the $14 start site. Additional studies are required to determine if RA is acting further upstream, downstream, within the gene, or controlling other trans- acting factors to influence 814 gene expression. The mechanism by which RA synergistically augments the DEX effect on $14 gene expression are similarly not understood. A similar type of RA and DEX interaction is also seen with the growth hormone gene (Bedo et al., 1989). Dexamethasone and RA, may be acting independently, together through some interaction of receptors as seen between estrogen and glucocorticoid receptors or estrogen and progesterone receptors (Cato and Ponta, 1989), or through the interaction of additional trans-acting factors. Because of the increasing significance of RA in cell and lipid metabolism (Lakshmanan et al., 1969; Krause et al., 1972; W155 and Wiss, 1980; Sporn et al., 1984), and its 156 effects on 814 gene expression, mechanisms of RA control will continue to be studied in 3T3-F442A cells. In light of the ability of 3T3-F442A cells to differentiate from fibroblasts into adipocytes, they also provide the opportunity to examine features of tissue- specific expression of the $14 gene (see Literature Review C.1). The $14 gene is expressed only in liver, WAT, and lactating mammary gland, all tissues involved in lipid metabolism. Presumably, a combination of differences in chromatin structure and trans-acting factors between expressing and non-expressing tissues accounts for tissue- selective $14 expression. Deoxyribonuclease I Hss analysis implicates Hss-1, 2, 3’, and Hss-3 of the 814 promoter region as areas involved in tissue-specific expression of the $14 gene (Jump et al., 1988). Sequence analysis of Hss-2 and Hss-3' indicates they contain DNA sequences with significant homology to DNA enhancer elements responsible for the tissue-specific expression of other genes in adipose tissue (Distel et al., 1987). These DNA sequences are called fat-specific enhancer elements 1 and 2 (FSE 1 and FSE 2). Two apparent FSE 2 sites reside within Hss-2 (-1153 and -1347 bp from the $14 start site) and one apparent FSE 1 site in Hss-3' (-1956 bp from the start site). The importance of these DNA sequences and associated trans-acting factors in tissue-selective $14 157 gene expression will be examined in future experiments in 3T3-F442A cells. 3T3-F442A cells also provide an opportunity to study negative control of $14 gene expression as reported 1 vivo (Kinlaw et al., 1986, 1987a, 1988; Jump et al., 1990b). While A-kinase stimulation i vivo is sufficient to inhibit 814 gene expression, 3T3-F442A cells must first be stimulated with DEX and then treated with an A-kinase stimulator such as epinephrine (Chapter 2). After dexamethasone stimulates mRNASl4 accumulation, A-kinase activation results in a dominant and rapid inhibition of mRNASl4 abundance. Although experiments were not conducted to assess whether this inhibitory effect was directed at the transcriptional level in 3T3-F442A cells, A-kinase stimulation 19 2129, primarily inhibits Sl4 gene transcription (Kinlaw et al., 1988; Jump et al., 1990b). However, the molecular events responsible for dominant A- kinase inhibition of $14 gene expression are unknown both in vivo and 19 vitro. 3T3-F442A cells will allow for transfection studies to address this issue. Surprisingly, 3T3-F442A cells lack T3 control of the $14 gene that is well characterized 19 vivo (Chapters 2 and 4; see Literature Review C.3). However, 2-dimensional gel analysis showed other gene products in 3T3-F442A cells were affected by T3. This selective T3 regulation of adipocyte genes may be related to: 1) 70% decreased thyroid hormone 158 receptor (TR) levels in 3T3-F442A cells compared to rat liver and white adipose tissue (WAT); 2) expression of TR isoforms which may block TRal action; 3) lack of additional required trans-acting factors; 4) presence of factors which interfere in TR - 814 gene interaction; or 5) a combination of the above. In experiments with the glucocorticoid receptor, varying the concentration of glucocorticoid receptor does limit the magnitude of response of certain genes (Vanderbilt et al., 1987). Possibly, functional TR levels are high enough in 3T3-F442A cells to allow control of some genes, but are not high enough to stimulate expression of other genes like $14. This effect might be enhanced by the variability in the magnitude of response to TR binding, that may be partially mediated by slight alterations in sequences among TREs of different genes (Glass et al., 1989). Possibly more significant is the apparent dominant expression of the TRaz isoform of the TRs in 3T3-F442A adipocytes (Chapter 4). The TRaz isoform is capable of binding a TRE, but not T3, effectively making it a silencer of T3 action (Izumo and Mahdavi, 1988; Lazar et al., 1988). Thyroid hormone binding and signal transduction for biological responses in 3T3-F442A cells, is probably due in part to the TRal isoform, which is capable of T3-binding, and is present at significantly reduced levels in 3T3- F442A cells compared to TRaz (Chapter 4). 3T3-F442A cells 159 may express a different TR isoform spectrum than adult mouse WAT that could prevent T3 regulation of the $14 gene. This seems plausible since 3T3-F442A cells were isolated from embryonic tissue and therefore, never develop as they would have 19 911_. Future studies utilizing 3T3-F442A cells will try and restore T3 responsiveness of the 814 gene. If successful, this would be the first functional association of a TR isoform with regulation of a specific gene. It would also provide the first evidence that developmental expression of TR isoforms (Forrest et al., 1990) contributes to developmental control of T3-regulated genes like the $14 gene. If these experiments are unsuccessful they will suggest some additional trans-acting factors are either inhibiting T3 stimulation or are missing to prevent T3 stimulation of gene expression. Experiments in 3T3-F442A cells are in progress to explore these possibilities. Initial studies in 3T3-F442A cells to associate function of the 814 protein with either lipogenesis or lipolysis were inconclusive (Chapter 2). $14 expression is not necessary for basal lipid accumulation in 3T3-F442A cells. Dexamethasone alone stimulates mRNASl4 and lipolysis, however, DEX plus epinephrine stimulates both lipolysis and lipogenesis at a time when mRNASl4 levels are rapidly falling. This apparent converse relationship can only suggest some purely hypothetical functions for the $14 160 protein e.g. it is somehow involved in mediating second messenger cascade from the plasma membrane; it is involved in some other aspect of lipid metabolism not accurately measured by these techniques. However, what is of general interest from these results, is the fact that mRNASl4 levels rise in 3T3-F442A cells in association with lipolysis, while results of 1 vivo studies implicate a role in lipogenesis for the 814 protein. Over-expression of the 814 protein in 3T3-F442A cells by transfection with concurrent monitoring of metabolic pathways, as well as, experiments with RNA anti-sense methods to block 814 action, should allow for significant progress on determining 814 function. In summary, the $14 gene in 3T3-F442A cells shows aspects of multifactorial control as it does 19 2129. These are summarized in Figure 29, differentiation, glucocorticoids, and RA control S14 gene expression primarily at the transcriptional level. The effects of TR isoforms on $14 gene expression are unclear, but, they prevent transcription of the $14 gene. Negative control of the 814 gene is also found in 3T3—F442A cells, although the mechanism of action is currently unknown. In conclusion, 3T3-F442A cells should provide a good 19 21119 model system to define both the regulation and function of the 814 protein. 161 FIGURE 29. DEVELOPMENTAL AND HORMONAL CONTROL OF THE 814 GENE IN 3T3-F442A CELLS. mRNASl4 is only detectable after 3T3-F442A fibroblasts differentiate into adipocytes. In the adipocyte phenotype glucocorticoids and retinoic acid positively stimulate Sl4 gene transcription and increase the stability of the 814 transcript. The lack of T3 control may in part be due to a decreased functional TR level as well as high levels of expression of the inactive TRaz isoform. Finally, cAMP mediates a dominant inhibitory effect over DEX stimulation of the $14 gene in 3T3-F442A cells. The mechanism(s) responsible for cAMP control are currently under investigation. 162 S14 GENE REGULATION IN 3T3-F442A CELLS Gene T3 1.. mg...“ _Dmerentletlon m 5 (+1 . DEX, RA -. (+) H 7 ( ) W" mRNA -—'-> Degradation A-Klneee k.) (+) 7 S14 Proteln BIBLIOGRAPHY BIBLIOGRAPHY Axelrod, J and Reisine, TD 1984 Stress hormones: Their interaction and regulation. Science 224:452-459 Baniahmad, A, Steiner, C, Kohne, AC, and Renkawitz, R 1990 Modular structure of a chicken lysozyme silencer: involvement of an unusual thyroid hormone receptor binding site. Cell 61:505-514 Beato, M 1990 Gene regulation by steroid hormones. Cell 56:335-344 Bedo, G, Santistiban, P, and Aranda, A 1989 Retinoic acid regulates growth hormone gene expression. Nature 339: 231-234 Benbrook, D, Lenhardt, E, and Pfahl, M 1988 A new retinoic acid receptor identified from a hepatocellular carcinoma. Nature 333:669-672 Berk, AJ 1989 Regulation of eukaryotic transcription factors by post-translational modification. Biochimica et Biophysica Acta 1009:103-109 Berne, H and Levy, M (1983) Physiology. The Mosby Co., Philadelphia, Pennsylvania Bernlohr, DA, Angus, CW, Lane, MD, Bolanowski, MA, and Kelly, TJ, Jr 1984 Expression of specific mRNAs during adipose differentiation: identification of an mRNA encoding a homologue of myelin P2 protein. Proc Natl Acad Sci USA 81:5468-5472 Bernlohr, DA, Bolanowski, MA, Kelly, TJ, Jr, and Lane, MD 1985 Evidence for and increase in transcription of specific mRNAs during differentiation of 3T3-L1 preadipocytes. J Biol Chem 260:5563-5567 Bernstein, P and Ross, J 1989 Poly(A), Poly(A) binding protein and the regulation of mRNA stability. TIBS 14:373-377 163 164 Beyer, HS, Carr, FE, Mariash, CN, and Oppenheimer, JH 1985 Hepatic messenger ribonucleic acid activity profile of rats subjected to alterations in thyroidal and adrenocorticortical states: evidence for significant interaction. Endocrinology 116:2669-2676 Bondy, PK 1985 Disorders of the adrenal cortex. In: Williams, Textbook of Endocrinology (eds. Wilson, JD and Foster, WB Saunders, Co, Philadelphia, pp 816-890 Brand, N, Petkovich, M, Krust, A, Chambon, P, de The, H, Marchio, A, Tiollais, P, and Dejean, A 1988 Identification of a second human retinoic acid receptor. Nature 332: 850- -853 Breitman, TR, Selonick, SE, and Collins, SJ 1980 Induction of differentiation of the human promyelocytic leukemia cell line (HL-60) by retinoic acid. Proc Natl Acad Sci USA 77:2936-2940 Buerghi, U and Abbuehl, U 1984 Nuclear triiodothyronine binding sites in rat adipocytes. Metabolism 33:326-328 Buratowski, S, Hahn, S, Guarente, L, and Sharp, PA 1989 Five intermediate complexes in transcription initiation by RNA polymerease II. Cell 56:549-561 Carr, FE, Bingham, JH, Kistner, C, and Mariash, CN 1984 Quantitative investigation of hepatic genomic response to hormonal and pathophysiological stimuli by multivariate analysis of two-dimensional mRNA profiles. Proc Natl Acad Sci USA 81:974-978 Castro-Munozledo, F, Marsch-Moreno, M, Beltran-Langarica, A and Kuri-Harcuch, W 1987 Commitment of adipocyte differentiation in 3T3-F442A cells is inhibited by retinoic acid and the expression of lipogenic enzymes is modulated through cytoskeleton stabilization. Differentiation 36:211-219 Cato, ACE, and Ponta, H 1989 Different regions of the estrogen receptor are required for synergistic action with the glucocorticoid and progesterone receptors. Mole and Cell Biol9:5324-5330 Chapman, AB, Knight, DM, Kieckman, BS, and Ringold, GM 1984 Analysis of gene expression during differentiation of adipogenic cells in culture and hormonal control of the developmental program. J Biol Chem 259:15548-15555. 165 Chapman, AB, Knight, DM, and Ringold, GM 1985 Glucocorticoid regulation of adipocyte differentiation: hormone triggering of the developmental program and induction of a differentiation-dependent gene. J Cell Biol 101:1227-1235. Chen, H, and Shapiro, DJ 1990 Nucleotide sequence and estrogen induction of Xenopus laevis 3-hydroxy-3- methylglutaryl-coenzyme A reductase. J Biol Chem 265:4622-4629 Chirgwin, JM, Przybyla, AE, MacDonald RJ, and Rutter, WJ 1979 Isolation of biologically active ribonucleic acid from sources enriched in ribonuclease. Biochemistry 18:5294-5299 a Clarke, SD, Armstrong, MK, and Jump, DB 1989 Nutritional control of rat liver fatty acid synthase and $14 mRNA abundance. J Nutr 120:218-224 b Clarke, SD, Armstrong, MK, and Jump, DB 1989 Dietary polyunsaturated fats uniquely suppress rat liver fatty acid synthase and 814 mRNA content. J Nutr 120:225-231 Coleman, RA, Reed, BC, Mackall, JC, Student, AK, Lane, MD, and Bell, RM 1978 Selective changes in microsomal enzymes of triacylglycerol, phosphotidylcholine and phosphotidylethanolamine biosynthesis during differentiation of 3T3-L1 preadipocytes. J Biol Chem 253:7256-7261 Cook, KS, Hunt, CR, and Spiegelman, BM 1985 Developmentally regulated mRNA’s in 3T3 adipocytes: Analysis of transcriptional control. J Cell Biol 100:514-520 Coutino, LL, Merkel, RA, Romsos, DR, Smith II, CK, and Bergen, WG 1990 Lipolytic activities of phenethanolamines in isolated porcine adipocytes. Submitted for publication (part of Dissertation at Michigan State University, East Lansing, Michigan) de Herreros, AG and Birnbaum, MJ 1989 The acquisition of increased insulin-responsive hexose transport in 3T3-L1 adipocytes corelates with expression of a novel transporter gene. J Biol Chem 264:19994-19999 de The, H, del Mar Vivanco-Ruiz, M, Tiollais, P, Stunnenberg, H, and Dejean, A 1990 Identification of a retinoic acid response element in the retinoic acid receptor B gene. Nature 343:177-180 166 Decaux, JF, Antoine, B, Kahn, A 1989 Regulation of the expression of the L-type pyruvate kinase gene in adult rat adipocytes in primary culture. J Biol Chem 264:11584- 11590 Diamond, DJ and Goodman, HM 1985 Regulation of growth hormone messenger RNA synthesis by dexamethasone and triidothyronine: transcriptional rate and mRNA stability changes in pituitary tumor cells. J Mol Biol 181:41-62 Distel, RJ, Ro, HS, Rosen, BS, Groves, DL, Spiegelman, BM 1987 Nucleoprotein complexes that regulate gene expression in adipocyte differentiation: direct participation of c-fos. Cell 49:835-844 Djian, P, Phillips, M, and Green, H 1985 The activation of specific gene transcription in the adipose conversion of 3T3 cells J Cell Physiol 124:554-556 Dobson, DE, Groves, DL, and Spiegelman, BM 1987 Nucleotide sequence and hormonal regulation of mouse glycerophosphate dehydrogenase mRNA during adipocyte and muscle cell differentiation. J Biol Chem 262:1804-1809 Dozin, B., Magnuson, MA, and Nikodem, VM 1986 Thyroid hormone regulation of malic enzyme synthesis: dual tissue-specific control J Biol Chem 261:10290-10292 Durant, S, Duval, D, and Homo-DeLarch, F 1986 Factors involved in control of fibroblast proliferation by glucocorticoids. Endocr Rev 7:254-269 Eadara, JK, Dalrymple, RH, DeLay, RL, Ricks, CA, and Romsos 1989 Effects of cimaterol, a beta-adrenergic agonist, on lipid metabolism in rats. Metab 38:522-529 Elks, ML and Manganiello, VC 1985 Effects of thyroid hormone on regulation of lipolysis and adenosine 3’,5’-monophosphate metabolism in 3T3-L1 adipocytes. Endocrinology 117:947-953 Evans, RM 1988 The steroid and thyroid hormone receptor superfamily. Science 240:889-895 Felgner, PL, Gadek, TR, Holm, M, Roman, R, Chan, HW, Wenz, M, Northrop, JP,Ringold, GM, and Danielsen, M 1987 Lipofection: a highly efficient, lipid-mediated DNA-tranfection procedure. Proc Natl Acad Sci USA 84:7413-7417 167 Feve, B and Pairault, J 1987 Dexamethasone-dependent expression of Bl-24corticotropin stimulated adenylate cyclase during adipose conversion of 3T3-F442A cells. FEBS Lett 219:56-64 Forrest, D, Sjoberg, M, and Vennstrom, B 1990 Contrasting developmental and tissue-specific expression of alpha and beta thyroid hormone receptor genes. EMBO 9:1519-1528 Forman, BM, Yang, C, Au, M, Casanova, J, Ghysdael, J, and Samuels, RH 1989 A domain containing leucine-zipper-like motifs mediate novel i9 vivo interactions between the thyroid hormone and retinoic acid receptors. Mol Endocrinol 3:1610-1626 Freake, HC and Oppenheimer, JH 1987 Stimulation of 814 mRNA and lipogenesis in brown fat by hypothyroidism, cold exposure, and cafeteria feeding: evidence supporting a general role for $14 in lipogenesis and the maintenance of thermogenesis. Proc Natl Acad Sci USA 84:3070-3074 Gallagher, GL 1989 Upstream, downstream: the intricate affairs of busy transcription factors. J NIH Res 1:97-100 Gerber, LE, and Erdman, LW, Jr 1980 Comparative effects of all-trans and 13-cos retinoic acid administration on serum and liver lipids in rats. J Nutr 110:343-351 Giguere, V, Ong, ES, Segui, P, and Evans, RM 1987 Identification of a receptor for the morphogen retinoic acid. Nature 330:624-629 Glass, CK, Franco, R, Weinberger, C, Albert, VR, Evans, RM, and Rosenfeld, MG 1987 A c-erb-A binding site in rat growth hormone gene mediates trans-activation by thyroid hormone. Nature 329:738-741 Glass, CK, Hollowat, JM, Devary, 0V, and Rosenfeld, MG 1988 The thyroid hormone receptor binds with opposite transcriptional effects to a common sequence motif in thyroid hormone and estrogen response elements. Cell 54:313-323 Glass, CK, Lipkin, SM, Devary, OV, Rosenfeld, MG 1989 Positive and negative regulation of gene transcription by a retinoic acid-thyroid hormone receptor heterodimer. Cell 59:697-708 Goodridge AG, Fisch, JE, and Glynias, MJ 1984 Regulation of the activity and synthesis of malic enzyme in 3T3-L1 cells. Arch Biochem Biophys 228:54-63 168 Goodridge, AG 1987 Dietary regulation of gene expression: enzymes involved in carbohydrate and lipid metabolism. Ann Rev Nutr 7:157-185 Gorman, CM, Merlino, GT, Willingham, MC, Pastan, I, and Howard, BH 1982 The rous sarcoma virus long terminal repeat is a strong promoter when introduced into a variety of eukaryotic cells by DNA-mediated transfection. Proc Natl Acad Sci USA 79:6777-6781 Gorman, CM 1985 High efficiency gene transfer into mammalian cells. In: DNA Cloning, Vol. III, A Practical Approach (eds. DM Glover), IRL Press, CITY, pp 143-190 Granner, DK 1987 The molecular biology of insulin action on protein synthesis. Kidney International 32:8-82-8-93 Graupner, G, Wills, KN, Tzukerman, M, Zhang, XK, and Pfahl, M 1989 Dual regulatory role for thyroid hormone receptors allows control of retinoic acid receptor activity. Nature 340:653—656 Green, H and Kehinde, O 1976 Spontaneous heritable changes leading to increased adipose conversion in 3T3 cells. Cell 7:105-113 Gross, DS and Garrard, WT 1988 Nuclease hypersensitive sites in chromatin Annu Rev Biochem 57:159-197 Grunfeld, C, Hagman, J, Sabin, EA, Buckley, DI, Jones, DS, and Ramachandran, J 1985 Characterization of adrenocorticotropin receptors that appear when 3T3-L1 cells differentiate into adipocytes. Endocrinology 116:113-117 Hamblin, PS, Ozawa, Y, Jefferds, A, and Mariash, CN 1989 Interaction between fructose and glucose on the regulation of the nuclear precursor for mRNA-$14. J Biol Chem 264:21646-21651 Hausdorf, S, Clement, J, and Loos, U 1988 Expression of the 814 gene during differentiaiton of 3T3-L1 cells. Harm Metab Res 20:723-724 * Herrera, R, Ro, HS, Robinson, GS, Xanthopoulos, KG, Spiegelman, BM 1989 A direct role for C/EBP and the AP-I-Binding site in gene expression linked to adipocyte differentiation. Mole Cell Biol 9:5331-5339 169 Hess, P, Meenakshi, T, Chan, GCK, Carlstedt-Duke, J, Gustaffson, JA, and Payvar, F 1990 Purified glucocorticoid receptors bind selectively i9 vit19 to a cloned DNA fragment that mediates a delayed secondary response to glucocorticoids in v1V9. Proc Natl Acad Sci USA 87:2564-2568 Hewish, RD and Burgoyne, LA 1973 Chromatin substrycture: the digestion of chromatin DNA at regularly spaced sites by a nuclear deoxyribonuclease Biochem Biophys Res Comm 52:504-510 Higuchi, K, Law, SW, Hoeg, JM, Schumacher, UK, Meglin, N, and Brewer, HB, Jr 1988 Tissue-specific expression of apolipoprotein A-1 (Apo-l) is regulated by the 5'- flanking region of the human ApoAl-gene. J Biol Chem 263:18539-18536 Hodin RA, Lazar MA, Wintman BI, Darling DS, Koenig RJ, Larsen PR, Moore DD, and Chin WW 1989 Identification of a thyroid hormone receptor that is pituitary specific. Science 244:76-79 Hunt, CR, Ro, JHS, Dobson, DE, Min, HY, Spiegelman, BM 1986 Adipocyte P2 gene: developmental expression and homology of 5'-flanking sequences among fat cell-specific genes. Proc Natl Acad sci USA 83:3786-3790 ' Iynedjian, PB, Jotterand, D, Nouspikel, T, Asfari, M, and Pilot, PR 1989 Transcriptional induction of glucokinase gene by insulin in cultured liver cells and its repression by the glucagon-cAMP system. J Biol Chem 264:21824-21829 Izumo, SB and Mahdavi, V 1988 Thyroid hormone receptor alpha isoforms generated by alternative splicing differentially activate myosin heavy chain gene transcription. Nature 334:539-542 Jacoby, DB, Zilz, ND, and Towle, HC 1989 Sequences within the 5'-flanking region of the 814 gene confer responsiveness to glucose in primary hepatocytes. J Biol Chem 264:17623-17626 Jantzen, HM, Strahle, U, Gloss, B, Stewart, F, Schmid, W, Boshart, M, Miksicek, R, Schutz, G 1987 Cooperativity of glucocorticoid response elements located far upstream of the tyrosine aminotransferase gene. Cell 49:29-38 Jump, DB, Seelig, S, Schwartz, HL, and Oppenheimer, JH 1981 Association of the thyroid hormone receptor with rat liver chromatin. Biochem 20:6782-6789 170 Jump, DB, Narayan, P, Towle, HC, and Oppenheimer, JH 1984 Rapid effects of triiodothyronine on hepatic gene expression. J Biol Chem 259:2789-2797 Jump, DB and Oppenheimer, JH 1985 High basal expression and 3,5,3'-triiodothyronine regulation of messenger ribonucleic acid 814 in lipogenic tissues. Endocrinology 117:2259-2266 Jump, DB, Tao, T-Y, Towle, HC, and Oppenheimer, JC 1986 Dissociation of hepatic messenger ribonucleic acidSl4 levels and nuclear transcriptional rates in suckling rats. Endocrinology 118:1892-1896 Jump, DB, Wong, NCW, and Oppenheimer, JH 1987 Chromatin structure and methylation state of a thyroid hormone-responsive gene in rat liver. J Biol Chem 262:778-784 Jump, DB, Veit, A, Santiago, V, Lepar, G, and Herberholz, L 1988 Transcriptional activiation of the rat liver Sl4 gene during postnatal development. J Biol Chem 263:7254-7260 a Jump, DB 1989 Rapid induction of rat liver Sl4 gene transcription by thyroid hormone. J Biol Chem 264:4698-4703 b Jump, DB 1989 Thyroid hormone regulation of rat liver Sl4 gene expression. In: Gene Regulation by Steroid Hormones IV (eds, AK Roy and JH Clark), Springer-Verlag, New York, New York, pp 144-160 a Jump, DB, Bell, A, and Santiago, V 1990 Thyroid hormone and dietary carbohydrate interact to regulate rat liver $14 gene transcription and chromatin structure. J Biol Chem 265:3474-3478 b Jump, DB, Bell, A, Lepar, SJ, and Hu, D 1990 Insulin rapidly induces rat liver Sl4 gene transcription. Mol Endocrinol (In press) a Kaestner, KH, Christy, RJ, McLenithan, JC, Braitman, LT, Cornelius, P, Pekala, PH, and Lane, MD 1989 Sequence, Tissue distribution, and differential expression of mRNA for a putative insulin-responsive glucose transporter in mouse 3T3-L1 adipocytes. Proc Natl Acad Sci USA 86:3150-3154 171 b Kaestner, KH, Ntambi, JM, Kelly, TJ, Jr, and Lane, MD 1989 Differentiation-induced gene expression in 3T3-L1 preadipocytes: a second differentially expressed gene encoding stearoyl-CoA desaturase. J Biol Chem 264:14755-14761 Kastner, Ph, Krust, A, Mendelsohn, C, Garnier, JM, Zelent, A, Leroy, P, Staub, A, and Chambon, P 1990 Murine isoforms of retinoic acid receptor gamma with specific patterns of expression. Proc Natl Acad Sci USA 87:2700- 2704 Kastner, P. Krust, A, Turcotte, B, Stropp, U, Tora, L, Gronemeyer, H, and Chambon, P 1990 Two distinct estrogen -regulated promoters generate transcripts encoding the two functionally different human progesterone receptor forms A and B. EMBO 9:1603-1614 Kawai, A and Kuzuya, N 1981 Effects of glucocoticoids on hormone-stimulated lipolysis and calcium uptake in the adipose cells. Horm Metab Res 13:224-228 Kawamura, M, Jensen, DF, Wancewicz, EV, Joy, LL, Khoo, JC, and Steinberg, D 1981 Hormone-sensitive lipase in differentiated 3T3-L1 cells and its activation by cAMP dependent protein kinase. Proc Natl Acad Sci USA 78:732-736 Kinlaw, WB, Schwartz, HL, Towle, HC, and Oppenheimer, JH 1986 Opposing effects of glucagon and triiodothyronine on the hepatic levels of messenger ribonucleic acid 814 and the dependence of such effects on Circadian factors. J Clin Invest 78:1091-1096 a Kinlaw, WB, Perez-Castillo, AM, Fish, LH, Mariash, CN, Schwartz, HL, and Oppenheimer, JH 1987 Interaction of dietary carbohydrate and glucagon in regulation of rat hepatic messenger ribonucleic acid 814 expression: role of circadian factors and 3’,5'-cyclic adenosine monophosphate. Mol Endocrinol 1:609-613 b Kinlaw, WB, Fish LH, Schwartz, HL, and Oppenheimer, JH 1987 Diurnal variation in hepatic expression of the rat $14 gene is synchronized by the photoperiod. Endocrinology 120:1563-1567 Kinlaw, WB, Schwartz, HL, Hamblin, PS, Mariash, CN, and Oppenheimer, JH 1988 Triiodothyronine rapidly reverses inhibition of 814 gene transcription by glucagon. Endocrinology 123:2255-2260 172 Kinlaw, WB, Ling, NC, and Oppenheimer, JH 1989 Identification of rat 814 protein and comparison of its regulation with that of mRNA $14 employing synthetic peptide antisera. J Biol Chem 264:19779-19783 Klausner, RD and Harford, JB 1989 Cis-trans models for post-transcriptional gene regulation. Science 246:870-872 Koenig, RJ, zar, MA, Hodin, RA, Brent, GA, Larsen, PR, Chin, WW, and Moore, DD 1989 Inhibition of thyroid hormone action by a non-hormone binding c-erbA protein generated by alternative mRNA splicing. Nature 337:659- 661 Koenig RJ, Warne RL, Brent GA, Harney JW, Larsen PR, and Moore DD 1988 Isolation of a cDNA clone encoding a biologically active thyroid hormone receptor. Proc Natl Acad Sci USA 85:5031-5035 Krause, RF, Beamer, KC, and Lotspeich, FJ 1972 Phosphonolipid metabolism in control and vitamin A deficient rats. Proc Soc Exp Biol Med 140:544-547 Krebs, HA and Henseleit, K 1982 Untersuchungen uber die harnstoffbildung in tierkorper. Hoppe Seyler's z Physiol Chem 210:33-36 Kuri-Harcuch, W and Green, H 1978 Adipose conversion of 3T3 cells depends on a serum factor. Proc Natl Acad Sci USA 75:6107-6109 Lai, E, Rosen, OM, and Rubin, CS 1982 Dexamethasone regulates the b-adrenergic receptor subtype expressed by 3T3-L1 preadipocytes and adipocytes. J Biol Chem 257:6691-6696 Lakshmanan, MR, Phillips, WEJ, and Brien, RL 1969 Vitamin A status and the metabolism of cholestrol in the rat. Br J Nutr 23:25-30 Lamers, WH, Hanson, RW, and Meisner, HM 1982 cAMP stimulates transcription of the genes for cytosolic phospoenolpyruvate carboxykinase in rat liver nuclei. Proc Natl Acad Sci USA 79:5137-5141 Lassar, AB, Buskin, JN, Lockshon, D, Davis, RL, Apone, S, Hauschka, SD, Weintraub, H 1989 MyoD is a sequence- specific DNA binding protein requiring a region of myc homology to the muscle creatine kinase enhancer. Cell 58:823-831 173 Lazar, MA, Hodin, RA, Darling, DS, and Chin, WW 1988 Identification of a rat c-erbAa-related protein which binds deoxyribonucleic acid but does not bind thyroid hormone. Mol Endorinol 2:893-901 Lazar, MA, Hodin, RA, Darling, DS, and Chin, WW 1989 A novel member of the thyroid/steroid hormone receptor family is encoded by the opposite strand of the rat c- erbAalpha transcriptional unit. Mol Cell Biol 9:1128-1136 Lefkowitz, RJ, and Caron, MC 1987 Molecular and regulatory properties of adrenergic receptors. Rec Prog Hor Res 43:469-497 ' Lepar, GJ, and Jump, DB 1989 Hormonal regulation of the $14 gene in 3T3-F442A cells. Mol Endocrinology 3:1207-1214 Lewin, Benjamin 1985 Genes, second edition. John Wiley and Sons, New York, New York Liaw, C, Seelig, S, Mariash, CN, Oppenheimer, JH, and Towle, HC 1983 Interactions of thyroid hormone, growth hormone, and high carbohydrate, fat-free diet in regulating several rat liver messenger ribonucleic acid species. Biochemistry 22:213-221 Liaw, CW and Towle, HC 1984 Characterization of a thyroid hormone-responsive gene from rat. J Biol Chem 259:7253-7260 Lillie, JW and Green MR 1989 Activator’s target in sight. Nature 341:279-280 Lockwood, EA, Bailey, E, and Taylor, CB 1970 Factors involved in changes in hepatic lipogenesis during development of the rat. Biochem J 118:155-162 Lomedico, P, Rosenthal, N, Efstratiadis, A, Gilbert, W, Kolodner, R, and Tizard, R 1979 The structure and evolution of the two nonallelic preproinsulin genes. Cell 18:545-558 Lotan, R, and Lotan, R 1980 Stimulation of melanogenesis in human melanoma cells by retinoids. Cancer Res 40: 3345- 3350 Lowry, OH, Rosebrough, NJ, Farr, AL, Randall, RJ 1951 Protein measurements with the folin reagent. J Biol Chem 193:265-275 174 Mackall, JC, Student, AK, Palakis, SE, and Lane, MD 1976 Induction of lipogenesis during differentiation in a preadipocyte cell line. J Biol Chem 251: 6462-6464 Mackall, JC and Lane, MD 1977 Role of pyruvate carboxylase in fatty acid synthesis: alteration during preadipocyte differentiation. Biochem Biophys Res Comm 79:720-725 Maden, M 1985 Retinoids and the control of pattern formation in the regenerating limb. CIBA Found Symp 113:132-155 Mandel, HG Fat-soluble vitamins, vitamins A, K, and E. In: The pharmacological basis of therapeutics (eds. Goodman, LS and Gilman, A) MacMillan Publishing Co, New York, New York, pp 1573-1591 Mangelsdorf, DJ, Ong, ES, Dyck, JA, and Evans, RM 1990 Nuclear receptor that identifies a novel retinoic acid response pathway. Nature 345:224-229 Maniatis, T, Sambrook, J, and Fritch, E 1981 Molecular cloning: a laboratory manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, New York, Hew York Mariash, CN, Jump, DB, and Oppenheimer, JH 1984 T3 stimulates the synthesis of a specific mRNA in primary hepatocyte culture. Biochem Biophys Res Comm 123:1122-1129 Mariash and Schwartz 1986 Effect of dichloroacetic acid on rat hepatic messenger RNA activity profiles. Metabolism 35:452-456 Mariash, CN, Seelig, S, Schwartz, HL, and Oppenheimer, JH 1986 Rapid synergistic interaction between thyroid hormone and carbohydrate on the mRNASl4 induction. J Biol Chem 261:9583-9586 Mariash, CN 1989 A novel hepatic nucleotide is correlated with the carbohydrate induction of messenger ribonucleic acid-$14. Endocrinology 124:212-217 Martin, CR 1985 Endocrine Physiology. Oxford University Press, New York, New York Meyer, ME, Gronemeyer,H, Turcotte, B, Bocquel, MT, Tasset, D, and Chambon, P 1989 Steroid hormone receptors compete for factors that mediate their enhancer function. Cell 57:433-442 175 Mitsuhashi, T, Tennyson, GE, and Nikodem, VM 1988 Alternative splicing generates messages encoding rat-c- erbA protein that do not bind thyroid hormone. Proc Natl Acad Sci USA 85:5804-5808 Miyajima, N, Horiuchi, R, Shibuya, Y, Fukushige, S, Matsubara, K, Toyoshima, K, and Yamamoto, T 1989 Two erbA homologs encoding proteins with different T3 binding capacities are transcribed from opposite DNA strands of the same genetic locus. Cell 57:31-39 Morita, S Fernadez-Mejia, C, and Melmed, S 1989 Retinoic acid selectively stimulates growth hormone secretion and messenger ribonucleic acid levels in rat pituitary cells. Endocrinology 124:2052-2056 Moustaid, N, Lasnier, F, Haingue, B, Quignard-Boulange, A, and Pairault, J 1990 Analysis of gene expression in 3T3-F442A preadipocytes: insulin and dexamethasone control. J Cell Biochem 42:243-254 Munnich, A, Besmond, C, Darquy, S, Reach, G, Vaulont, S, Dreyfus, JC, Kahn, A 1985 Dietary and hormonal regulation of aldolase B gene expression. J Clin Invest 75:1045-1052 Murray, MB, Zilz, ND, McCreary, NL, MacDonald, MJ, and Towle, HC 1988 Isolation and characterization of rat cDNA clones for two distinct thyroid hormone receptors. J Biol Chem 263:12770-12777 Murray, MB, and Towle, HC 1989 Identification of nuclear factors that enhance binding of the thyroid hormone receptor to a thyroid hormone response element. Mol Endocrinol 3:1434-1442 Narayan, P, Liaw, CW, and Towle, HC 1984 Rapid induction of specific nuclear mRNA presursor by thyroid hormone. Proc Natl Acad Sci USA 81:4687-4691 Nepokroeff, CM, Adachi, K, Yan, C, and Porter, JW 1984 Cloning of DNA complementary to rat liver fatty acid synthestase mRNA. Eur J Biochem 140:441-445 Ntambi, JM, Buhrow, SA, Kaestner, KH, Christy, RJ, Sibley, E, Kelly, TJ, Jr, and Lane, MD 1988 Differentiation-induced gene expression in 3T3-L1 preadipocytes: characterization of a differentially expressed gene encoding stearoyl-CoA desaturase. J Biol Chem 263:17291-17300 176 Nyborg, JK and Spindler, SR 1986 Alterations in local chromatin structure accompany thyroid hormone induction of growth hormone gene transcription. J Biol Chem 261:5685-5688 O’Farrell PH 1975 High resolution two-dimensional electrophoresis of proteins. J Biol Chem 250:4007-4021 O’Farrell, Pz, Goodman, HM, O’Farrell, PH 1977 High resolution two-dimensional electrophoresis of basic as well as acidic proteins. Cell 12:1133-1142 O’Malley, BW 1989 Editorial: Did eucaryotic steroid receptors evolve from intracrine gene regulators? Endocrinology 125:1110-1120 O’Malley, BW 1990 The steroid receptor superfamily: more excitement predicted for the future. Mol Endocrinology 4:363-369 Oppenheimer, JH, Koerner, D, Schwartz, HL, and Surks, MI 1972 Specific nuclear triiodothyronine binding studies in rat liver and kidney. J Clin Endocrinol Metab 35:330-333 Oppenheimer JH 1983 The nuclear receptor-triidothyronine complex: relationship to thyroid hormone distribution, metabolism, and biological action. In: Molecular Basis of Thyroid Hormone Action (eds. Oppenheimer, JH and Samuels, HH). Academic Press, New York, New York, pp 1-34 Oppenheimer, JH, Kinlaw, WB, Wong, NC, Schwartz, HL, and Mariash, CN 1987 Regulation of gene S-14 by triiodothyronine in liver. Hor Met Res Sup 17:1-5 Pacifici, M, Cossu, G, Molinaro, M, and Tato, F 1980 Vitamin A inhibits chondrogenesis but not myogenesis. Exp Cell Res 129:469-474 Pairault, J, and Lasnier, F 1987 Control of adipogenic differentiation of 3T3-F442A cells by retinoic acid, dexamethasone, and insulin: a topographic analysis. J Biol Chem 132:279-286 Paulauskis, JD and Sul, HS 1988 Cloning and expression of mouse fatty acid synthase and other specific mRNAs: developmental and hormonal regulation in 3T3-L1 cells. J Biol Chem 263:7049-7054 Perez-Castillo, A, Schwartz, HL, and Oppenheimer, JH 1987 Rat hepatic mRNA-$14 and lipogenic enzymes during weaning: role of 814 in lipogenesis. Am J Physiol 253:E536-E542 177 Petersen, DD, Koch, SR, and Granner, DK 1989 3’ noncoding region of phosphoenolpyruvate carboxykinase mRNA contains a glucocorticoid-responsive mRNA-stabilizing element. Proc Natl Acad Sci USA 86:7800-7804 Petkovich, M, Brand, NJ, Krust, A, and Chambon, P 1987 A human retinoic acid receptor which belongs to the family of nuclear receptors. Nature 330:444-450 Reed, BC, Kaufman, SH, Mackall, JC, Student, AK, and Lane, MD 1977 Alterations in insulin binding accompanying differentiation of 3T3-Ll preadipocytes. Proc Natl Acad Sci USA 74:4876-4880 Ringler, MB, Erbland, JF, Singh, BB and Schuster, G 1984 The effects of retinoic acid on [14C1acetate incorporation into lipids of normal and transformed hamster fibroblasts. Exp Cell Res 154:171-180 Roberts, AB, and Sporn, MB 1984 Cellular biology and biochemistry of the retinoids, In: The retinoids (eds. Sporn, MB, Roberts, AB, and Goodman, DS) Vol. 2, pp 210- 286 . Roesler, WJ, Vanderbark, GR, and Hanson, RW 1988 Cyclic AMP and the induction of eukaryotic gene transcription. J Biol Chem 263:9063-9066 Ron, D, Brasier, AR, Wright, KA, and Habener, J 1990 The permissive role of glucocorticoids on interleukin-1 stimulation of angiotensinogen gene transcription is mediated by an interaction between inducible enhancers. Mol Cell Biol 10:4389-4395 Roncari, DA, and Murthy, VK 1975 Effects of thyroid hormones on enzymes involved in fatty acid glycerolipid synthesis. J Biol Chem 250:4134-4138 Rubin, CS, Lai, E, and Rosen, OM 1977 Acquisition of increased hormone sensitivity during in vitro adipocyte development. J Biol Chem 252:3554-3557 Rubin, CS, Hirsch, A, Fung, C, and Rosen, OM 1978 Development of hormone receptors and hormone responsiveness in vitro: insulin receptors and insulin sensitivity in the preadipocyte and adipocyte forms of 3T3-Ll cells. J Biol Chem 253:7570-7578 Saltiel, AR 1990 Signal transduction in insulin action. J Nutr Biochem 1:180-189 178 Saltzman, AG and Weinman, R 1989 Promoter specificity and modulation of RNA polymerase II transcription. FASEB J 3:1723-1733 Samuels, HH, Forman, MB, Horowitz, ZD, and Ye, Z-S 1988 Regulationof gene expression by thyroid hormone. J Clin Invest 81:957-967 Sap, J, Munoz, A, Damm, K, Goldberg, Y, Ghysdael, J, Leutz, A, Beug, H, and Vennstrom, B 1986 The c-erb-A protein is a high-affinity receptor for thyroid hormone. Nature 324:635-640 Sasaki, K, Cripe, T, Koch, S, Andreone, T, Peterson, D, Beale, E, and Granner, DK 1984 Multihormonal regulation of the phosphoenolpyruvate carboxykinase gene: the dominant role of insulin. J Biol Chem 259:15242-15251 Sasaki, K and Granner, DK 1988 Regulation of phosphoenolpyruvate carboxykinase gene transcription by insulin and cAMP: reciprocal actions on initiation and elongation. Proc Natl Acad Sci USA 85:2954-2958 Scatchard, G 1949 The attractions of proteins for small molecules and ions. Ann NY Acad Sci 51:73-77 Schule, R, Miller, M, Kaltschmidt, C, and Renkawitz, R 1988 Many transcription factors interact synergistically with steroid receptors. Science 242:1419-1420 Schule, R, Umesono, K, Mangelsdorf, DJ, Bolado, J, Pike, JW, and Evans, RM 1990 Jun-fos and receptors for vitamins A and D recognize a common response element in the human osteocalcin gene. Cell 61:497-504 Schwartz, J and Carter-Sn, C 1988 Effects of growth hormone on metabolism and glucose transport in 3T3-F442A cells: dependence on cell differentiation. Endocrinology 122:2247-2256 Seamon, KB and Daly, JW 1981 Forskolin: a unique diterpine activator of cyclic AMP generating systems. J Cyclic Nucleotide Res 7:210-224 Seelig, S, Liaw, C, Towle, HC, and Oppenheimer, JH 1981 Thyroid hormone attenuates and augments hepatic gene expression at a pretranslational level. Proc Natl Acad Sci USA 78:4733-4737 179 Seelig, S, Jump, DB, Towle, HC, Liaw, C, Mariash, CN, Schwartz, HL, and Oppenheimer, JH 1982 Paradoxical effects of cyclohexamide on the ultra-rapid induction of two hepatic mRNA sequences by triiodothyronine (T3). Endocrinology 110:671-673 Silva, ES, Astier, H, Thakare, U, Schwartz, HL, and Oppenheimer, JH 1977 Partial purification of the triiodothyronine receptor from rat liver nuclei: differences in the chromatographic mobility of occupied and unoccupied sites. J Biol Chem 252:6799-6805 Simonet, WS and Ness, GC 1988 Transcriptional and posttranscriptional regulation of rat hepatic 3-hydroxy- 3-methylglutaryl-coenzyme A reductase by thyroid hormones. J Biol Chem 263:12448-12453 Smith, PJ, Wise, LS, Berkowitz, R, Wan, C, and Rubin, CS 1988 Insulin-like growth factor-I is an essential regulator of the differentiation of 3T3-L1 adipocytes. J Biol Chem 263:9402-9408 Spiegelman, BM and Green, H 1980 Control of specific protein biosynthesis during the adipose conversion of 3T3-cells. J Biol Chem 255:8811-8818 Spiegelman, EM and Farmer, SR 1982 Decrease in tubulin and actin gene expression prior to morphological differentiation of 3T3-adipocytes. Cell 29:53-60 Spiegelman, BM, Frank, M and Green, H 1983 Molecular cloning of mRNA from 3T3 adipocytes: regulation of mRNA content for glycerophosphate dehydrogenase and other differentiation-dependent proteins during adipocyte development. J Biol Chem 258:10083-10089 Spooner, PM, Chernick, SS, Garrison, MM, and Scow, R0 1979 Development of lipoprotein lipase activity and accumulation of triacylglycerol in differentiating 3T3-L1 adipocytes: effects of prostaglandin F2a, 1-methyl-3-isobutylxanthine, prolactin, and insulin. J Biol Chem 254:1305-1311 Sporn, MB, and Roberts, AB (1983) Role of retinoids in differentiation and carcigenesis. Cancer Res 43: 3034- 3040 Sporn, MB, Roberts, AB, and Goodman, DS 1984 The retinoids, vols 1-2 180 Starling, EH 1905 The Croonian Lectures on the chemical correlation of the functions of the body. Lancet 2:339- 341: 423-425; 501-503; 579-583 Steel, RGD, and Torrie, JH 1980 Principles and procedures of statistics: a biometrical approach. McGraw-Hill, Inc, New York, New York pp 137-194 Strahle, U, Scmid, W, and Scchutz, G 1988 Synergistic action of the glucocorticoid receptor with transcription factors. EMBO 7:3389-3395 Strait, KA, Kinlaw, WB, Mariash, CN, and Oppenheimer, JH 1989 Kinetics of induction by thyroid hormone of the two hepatic mRNAs coding for cytosolic malic enzyme in the hypothyroid and euthyroid states: evidence against an obligatory role of 814 protein in malic enzyme gene expression. J Biol Chem 264:19784-19789 Strickland, S, Smith, KK, Mariotti, KR 1980 Hormonal induction of differentiation in teratocarcinoma stem cells: generation of parietal ondoderm by retinoic acid and dibutyryl cAMP. Cell 21:347-355 Strobl, WS, Gorder, NL, Lin-Lee, YC, Gotto, AM, Jr, and Patsch, W 1990 Role of thyroid hormones in apolipoprotein A-l gene expression in rat liver. J Clin Invest 85:659- 667 Student, AK, Hsu, RY, and Lane, MD 1980 Induction of fatty acid synthetase synthesis in differentiating 3T3-L1 preadipocytes. J Biol Chem 255:4745-4750 Sucov, HM, Murakami, KK, and Evans, RM 1990 Characterization of an autoregulated response element in the mouse retinoic acid receptor type B gene. Proc Natl Acad Sci USA 87:5392-5396 Sullivan, KF and Cleveland, D 1986 Identification of conserved isotype-defining variable region sequences for four vertebrate b-tubulin polypeptide classes. Proc Natl Acad Sci USA 83:4327-4331 Thompson, CC, Weinberger, C, Lebo, R, and Evans, RM 1987 Identification of a novel thyroid hormone receptor expressed in the mammalian central nervous system. Science 237:1610-1614 Todaro, GJ and Green, H 1963 Quantitative studies of the growth of mouse embryo cells in culture and their development into established lines. J Cell Biol 17:299-313 181 Topliss, DJ, Mariash, CN, Seelig, S, Carr, FE, and Oppenheimer, JH 1983 Effects of triiodothyronine and glucose on cultured rat hepatocyte gene expression. Endocrinology 112:1868-1870 Touchette, N 1990 Man bites dogma: a new role for steroid hormones. J NIH Res 2:71-74 Towle, HC, Mariash, CN, and Oppenheimer, JH 1980 Changes in the hepatic levels of messenger ribonucleic acid for malic enzyme during induction by thyroid hormone or diet. Biochemistry 19:579-585 Towle, HC and Mariash, CN 1986 Regulation of hepatic gene expression by lipogenic diet and thyroid hormone. Federation Proc 45:2406-2411 Tsai, SY., Carlstedt-Duke, J, Weigel, NL, Dahlman, K, Gustafsson, J, Tsai, M, and O’Malley, BW 1988 Molecular interactions of steroid hormone receptor with its enhancer element: evidence for receptor dimer formation. Cell 55:361-369 Umesono, K, Giguere, V, Glass, CK, Rosenfeld, MG, and Evans, RM 1988 Retinoic acid and thyroid hormone induce gene expression through a common responsive element. Nature 336:262-265 Vanderbilt, JN, Miesfeld, R, Maler, BA, and Yamamoto, KR 1987 Intracellular receptor concentration limits glucocorticoid-dependent enhancer activity. Mol Endocrinol 1:68-74 Vasios, GW, Gold, JD, Petkovich, M, Chambon, P, and Gudas, LJ 1989 A retinoic acid response element is present in the 5'-flanking region of the laminin Bl gene. Proc Nat Acad Sci USA 86:9099-9103 Volpe, JJ and Vagelos, PR 1976 Mechanisms and regulation of biosynthesis of saturated fatty acids Physiol Rev 56:339- 417 Volpe, JJ and Kishimoto, Y 1982 Fatty acid synthetase of brain: development, influence of nutritional and hormonal factors and comparison with liver. J Neurochem 19:737- 753 Weinberger, C, Thompson, CC, Ong, ES, Lebo, R, Gruol, DJ, and Evans, RM 1986 The c-erbA gene encodes a thyroid hormone receptor. Nature 324:641-646 182 Weiss, GH, Rosen, OM, and Rubin, CS 1980 Regulation of fatty acid synthetase concentration and activity during adipocyte differentiation: studies on 3T3-L1 cells. J Biol Chem 255:4751-4757 Wilson, JD and Foster, DW 1985 Introduction. In:Williams Textbook of Endocrinology (eds, Wilson, JD and Foster, DW) W.B. Saunders Co., Philadelphia, Pennsylvania, pp 1-8 Wingender, E 1988 Compilation of transcription regulating proteins. Nucleic Acids Res 16:1879-1902 Wise, EM and Ball, EG 1964 Malic Enzyme and lipogenesis Proc Natl Acad Sci USA 52:1255-1263 Wise, LS and Green, H 1979 Participation of one isozyme of cytosolic glycerophosphate dehydrogenase in the adipose conversion of 3T3 cells. J Biol Chem 254:273-275 Wise, LS, Sul, HS, and Rubin, CS 1984 Coordinate regulation of the biosynthesis of ATP-citrate lyase and malic enzyme during adipocyte differentiation: studies on 3T3-L1 cells. J Biol Chem 259:4827-4832 Wiss, O, and Wiss, V 1980 Alterations of the lipid metabolism of rat liver as early symptoms of vitamin A deficiency. 1980 Int J Vit Nutr Res 50:233-237 Wong, NCW, Schwartz, HL, Santos, A, and Oppenheimer, JH 1987 Cycloheximide inhibits S-14 gene transcription and abolishes DNase I Hypersensitive S-14 sites in the livers of euthyroid but not hypothyroid rats. Mol Endocrinol 1:459-467 a Wong, NCW, Perez-Castillo, AM, Sanders, MM, Schwartz, HL, and Oppenheimer, JH 1989 Thyroid hormone and circadian regulation of the binding activity of a liver-specific protein associated with the 5'-flanking region of the $14 gene. J Biol Chem 264:4466-4470 b Wong, NCW, Schwartz, HL, Strait, K, and Oppenheimer, JH 1989 Thyroid hormone-, carbohydrate-, and age-dependent, regulation of a methylation site in the hepatic Sl4 gene. Mol Endocrinol 3:645-650 Wong, NCW, Deschamps, BJ, Yeomans, ZA, and Cannon, PD 1990 Binding of nuclear proteins to the rat liver $14 gene is influenced by thyroid state. J Biol Chem 265:8775-8781 183 Yaffe, BM and Samuels, HH Hormonal regulation of the growth hormone gene: relationship of the rate of transcription to the level of nuclear thyroid hormone-receptor complexes. J Biol Chem 259:6284-6291 Yamamoto, KK, Gonzalez, GA, Biggs, WH, and Montminy, MR 1988 Phosphorylation-induced binding and transcriptional efficacy of nuclear factor CREB. Nature 334:494-498 Zelent, A Krust, A, Petkovich, M, Kastner, P, and Chambon, P 1989 Cloning of murine alpha and beta retinoic acid receptors and a novel receptor gamma predominantly expressed in skin. Nature 339:714-717 Zilz, ND, Murray, MB, and Towle, HC 1990 Identification of multiple thyroid hormone response elements located far upstream from the rat $14 promoter. J Biol Chem 265:8136-8143