UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Characterization of a human prothrombin gene enhancer Chow, Billy Kowk-Chong 1991

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


831-UBC_1991_A1 C56.pdf [ 8.05MB ]
JSON: 831-1.0100492.json
JSON-LD: 831-1.0100492-ld.json
RDF/XML (Pretty): 831-1.0100492-rdf.xml
RDF/JSON: 831-1.0100492-rdf.json
Turtle: 831-1.0100492-turtle.txt
N-Triples: 831-1.0100492-rdf-ntriples.txt
Original Record: 831-1.0100492-source.json
Full Text

Full Text

Characterization of a Human Prothrombin Gene Enhancer by Billy Kwok-Chong Chow B. Sc (Hon.), The University of British Columbia, 1986 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF Doctor of Philosophy in THE F A C U L T Y OF GRADUATE STUDIES Department of Biochemistry We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA May 1991 QD Billy Kwok-Chong Chow, 1991 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia/ I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of RloCpVEM 1 VT£,y The University of British Columbia Vancouver, Canada Date / V W ^ V \ Y \\<\) DE-6 (2/88) Abstract ABSTRACT The 5' flanking sequence of the human prothrombin gene was isolated by screening a human liver phage library with a human prothrombin cDNA as a hybridization probe. A phage was identified that contained 3 kilobasepairs of D N A upstream of the initiator methionine codon. Primer extension studies showed that the major transcription initiation sites were located 23 and 36 basepairs upstream of the initiator codon. D N A sequences in the 5' flanking region of the human prothrombin gene were then analyzed for cis-activating transcriptional activity by a transient expression system using the human growth hormone gene as the reporter gene. The chimeric expression vector was introduced into HepG2 cells, and secreted human growth hormone was monitored by using a radioimmunoassay. These studies showed that the 3 kbp fragment contained sequences that were sufficient for the initiation of transcription in HepG2 cells. Subsequent deletion studies showed that the 3 kbp fragment contained two elements: a weak promoter in the region immediately upstream of the mRNA coding sequence, and an enhancer located between nucleotides -860 and -940. The enhancer element was active at a distance and in either orientation. In addition, the enhancer was liver cell specific, and acted on heterologous promoters including the herpes simplex virus thymidine kinase promoter and the mouse metallothionein I promoter. Comparison of the nucleotide sequence of the enhancer with a D N A sequence data base showed the enhancer sequence to be unique. The enhancer sequence is flanked by an inverted repeat, 5' CCTCCC 3', and contains a putative binding site for hepatic nuclear factor 1 (HNF-1). Deoxyribonuclease I footprint analysis and linker scanning mutagenesis showed that the enhancer contains multiple protein binding motifs. Mutagenesis of the 3' boundary CCTCCC sequence eliminated the enhancer activity. Comparison with other liver genes showed the presence of the C C T C C C ii Abstract sequence in the hepatitis B virus enhancer, the a-1 antitrypsin promoter, and the fibrinogen (5-chain promoter, suggesting a functional role for this motif. Using the concatenated human prothrombin enhancer as a probe to screen a HepG2 expression library, a cDNA encoding for the Y-box binding protein was identified. A putative Y-box was also found in the enhancer region, suggesting that the protein factor may be partially responsible for the human prothrombin gene expression. Northern blot analysis using the Y-box binding protein cDNA as a probe indicated that the Y-box binding protein mRNA is expressed in all of the tested tissues. This protein may be one of the constitutively expressed transcription factors responsible for the regulation of a number of genes. iii Table of Contents T A B L E OF C O N T E N T S ABSTRACT ii T A B L E OF C O N T E N T S iv LIST OF TABLES viii LIST OF FIGURES ix L I S T OF A B B R E V I A T I O N S xi ACKNOWLEDGEMENTS xiv I. I N T R O D U C T I O N 1 A . Hemostasis - an overview 1 B. Blood coagulation 4 C. Prothrombin 8 D. Expression of the human prothrombin gene 14 E . Eukaryotic gene regulation 16 1. Active chromatin 17 2. The cis-acting regulatory elements 18 3. The protein factors 19 4. The leucine zipper 20 5. DNA-binding motifs 21 F. The promoter 23 1. The T A T A box 23 2. The C C A A T box 26 3. Distal regulatory elements 27 G. The enhancer 28 H . Liver specific Expression 31 iv Table of Contents 1. CCAAT/enhancer-binding protein (C/EBP) 32 2. D binding protein (DBP) 32 3. Hepatocyte-specific nuclear factor-1 (HNF-1 or LF-B1) 33 4. Hepatocyte nuclear factor-3A (HNF-3A) 34 I. Transcriptional regulation of blood clotting factors 35 J . The present study 37 M A T E R I A L S A N D M E T H O D S 38 A . Strains, vectors and media 38 1. Bacterial strains 38 2. Vectors 38 3. Media 39 B. Gel electrophoresis 39 1. Non-denaturing agarose gels and Southern blot analysis 39 2. Formaldehyde agarose gels and Northern blot analysis 40 3. Non-denaturing polyacrylamide gels 41 4. Denaturing polyacrylamide gels 41 5. SDS-denaturing polyacrylamide gels and Western blot analysis 42 C . Isolation of D N A 42 1. Isolation of plasmid D N A 42 2. Isolation of lambda phage D N A 43 D. D N A subcloning 44 1. Isolation and preparation of D N A fragments for subcloning 44 2. Ligation and transformation of D N A into bacteria 44 E. Radioactive labeling of D N A 45 Table of Contents 1. Klenow labeling 45 2. Labeling with T4 polynucleotide kinase 46 3. Labeling by nick translation 46 F. Oligodeoxyribonucleotides used in this study 47 G. Site directed mutagenesis 49 1. Site-directed mutagenesis in M13 bacteriophage 49 2. PCR-mutagenesis: production of linker-scanning mutants..!.'. 49 H . Primer extension analysis of a messenger RNA transcript 52 I. Tissue culture techniques 53 1. Maintaining the cell-lines 53 2. Transfection and transient expression of hGH 54 J. Characterization of the protein factors binding to the human prothrombin gene enhancer 55 1. Preparation of rat liver nuclear extract 55 2. DNase I footprinting analysis and gel retention assays 56 K . Screening of a lambda gtl 1 expression library 57 III. RESULTS 59 A . Identification of the human prothrombin gene 5' flanking region 59 B. Characterization of the human prothrombin gene 5' flanking region 61 C. The human growth hormone transient expression system 68 D. The effects of prothrombin fragment 1 and vitamin K on the expression of prothrombin at the transcriptional level 73 E. Initial characterization of the human prothrombin gene 5' flanking region 75 F. Fine mapping of the human prothrombin 5' region 78 vi Table of Contents G. Characterization of the upstream element 81 H . PCR-linker scanning mutation analysis of the prothrombin enhancer 85 I. The human prothrombin gene enhancer contains multiple protein binding modules 85 J. Characterization of a lambda g t l l expression clone 91 IV. DISCUSSION 101 A . Characterization of the human prothrombin gene 5' flanking region 101 B. Characterization of the human growth hormone transient expression system. 102 C. The effects of prothrombin fragment 1 and vitamin K on the expression of prothrombin at the transcriptional level 105 D. Functional characterization of the 3 Kbp 5' flanking region 106 E. Characterization of the upstream enhancer element 107 F. PCR-linker scanning mutation analysis of the prothrombin enhancer 108 G. DNase I footprint analysis of the human prothrombin gene enhancer 109 H . Analysis of the enhancer motifs 110 I. Characterization of the Y-box binding protein 114 J. Putative model for human prothrombin gene expression in the liver 120 K . Future studies 121 V . REFERENCES 124 vii List of tables LIST O F T A B L E S Table 1. Table of oligodeoxyribonucleotides used in the study of human prothrombin gene regulation 48 Table 2. Transfection results using the plasmids pOGH, pXGH5, p T K G H and pFIIGH1.3 70 Table 3. Effects of vitamin K on the human prothrombin gene expression 76 viii List of Figures LIST O F F I G U R E S Fig . 1. The blood coagulation cascade 5 Fig. 2. The organization of the human prothromin gene, mRNA and protein 10 Fig. 3. Organization of a eukaryotic polymerase II promoter 24 Fig. 4. Construction strategy of PCR-linker scanning mutants 50 Fig. 5. Southern blot analysis of the phage clone 1H51A 60 Fig. 6. Plasmid construction strategies of pFIIGH3.0, pFIIGH1.3, pFIIGH1.3R and pFIIGH0.4 62 Fig. 7. D N A sequence analysis of 1258 bp of the 5' flanking region of the human prothrombin gene 64 Fig. 8. Primer extension analysis to determine the transcription start site of the human prothrombin gene 66 Fig. 9. Characterization of the human growth hormone transient expression system 70 Fig. 10. Effects of the activation peptide (fragment-1) on the expression of human prothrombin gene 73 Fig. 11. Transfection of HepG2 cells with expression constructs containing different regions of the human prothrombin 5' flanking sequence 76 Fig. 12. Fine mapping of the upstream regulatory sequence (URS) 78 Fig. 13. Characterization of the human prothrombin gene URS 81 Fig. 14. Linker scanning mutation analysis of the region -859 to -959 bp from the translation start site 85 Fig. 15. DNase I footprint analysis of the human prothrombin gene enhancer 88 Fig. 16. Tissue specificity of rat prothrombin gene expression 91 ix List of Figures Fig. 17. Competitive gel-shift assays 94 Fig. 18. D N A and amino acid sequence analysis of the Y-box binding protein 96 Fig. 19. Northern blot analysis of the Y-box binding protein 99 Fig. 20. Summary of the protected sequences of the enhancer I l l Fig. 21. Analysis of the Y-box binding motif in the human prothrombin gene enhancer 115 Fig. 22. Model of human prothrombin gene expression in the liver 121 x List of Abbreviations LIST O F A B B R E V I A T I O N S A absorbance Amp ampicillin ATP adenosine triphosphate B S A bovine serum albumin B H K baby hamster kidney bis N.N'-methylenebisacrylamide bp(s) basepair(s) C A P transcription start site CAT chloramphenicol acetyl transferase cDNA complementary deoxyribonucleic acid C/EBP C C A A T enhancer binding protein Ci Curie cpm counts per minute DBP D-binding protein DEAE diethylaminoethyl D M E M Dulbecco's Modified Eagle Medium DMSO dimethyl sulfoxide DNA deoxyribonucleic acid DNase I deoxyribonuclease I dNTPs deoxyribonucleotide triphosphates DTT dithiothreitol EDTA ethylenediaminetetraacetic acid EtBr ethidium bromide xi List of Abbreviations Gla y-carboxyl glutamic acid h hour(s) HBS hepes buffer saline HEPES N-[2-hydroxyethyl]piperazine-N'-[2-ethane-sulfonic acid] h G H human growth hormone H N F hepatocyte-specific nuclear factor H S V herpes simplex virus IPTG isopropyl- p-D-thiogalactopyranoside L B Luria broth min minutes(s) M M L V - R T Moloney Murine Leukemia Virus reverse transcriptase mRNA messenger ribonucleic acid OAc acetate C D optical density P A G E polyacrylamide gel electrophoresis PBS phosphate buffer saline PIPES Piperazine-N,N'-bis[2-ethane-sulfonic acid] PMSF phenylmethylsulfonyl fluoride POU Pit-l,Oct-l,Unc-86 P V A polyvinylalcohol RNA ribonucleic acid RNase ribonuclease rpm revolutions per minute RSV Rous Sarcoma Virus xii List of Abbreviations sec second(s) SDS sodium dodecyl sulphate TEMED N,N,N\N'-tetramethylethylenediamine T K thymidine kinase Tris tris(hydroxymethyl)aminomethane tRNA transfer ribonucleic acid U V ultra-violet V Volts W Watts X-gal 5-bromo-4-chloro-3-indolyl-P-D-galactopyranoside xiii Acknowledgement A C K N O W L E D G E M E N T S I would like to give special thanks to my neighbor, Bruce, for his help and encouragement in the past four years. Also, Helene for her great company in times of trouble. Life will be so boring without Dave in the laboratory. When you are sick of looking at the meaningless footprinting gels, you can always go to the coffee room where Dave hosts a talk-show once every hour in the day-time. I have to thank Ross for his guidance and also for being such a money making machine so graduate students like us can buy expensive reagents and go to international conferences. I also would like to thank John (Dr. Brown) for having so much confidence in an ordinary student like me. I want to thank Shiho who gives me the initiative to apply for fellowships, and finally, I have to thank my family especially my parents for their support so I could survive through and hence experience those memorable graduate student days. xiv Introduction I. I N T R O D U C T I O N A. HEMOSTASIS - AN OVERVIEW Hemostasis is a self-defence mechanism to arrest bleeding from damaged blood vessels. Physiologically, there are four interrelated processes: vasoconstriction; formation of a platelet plug; blood coagulation and finally, as a secondary event, fibrinolysis (Colman et al., 1987). The end result of the first three processes is the formation of a stable blood clot which mechanically impedes the flow of blood from the injured vessel to reduce blood loss. On the other hand, fibrinolysis is important in clot remodelling during healing. Stimulation of vascular smooth muscle cells upon injury results in an immediate intense vasoconstriction to reduce blood loss from the site of injury. Vasoconstriction in response to direct stimulation is generally transient. Additional vasoconstrictors such as platelet-derived thromboxane A2 secreted from the activated platelets contribute to a more prolonged effect (Colman et al., 1987). Vascular endothelial cells are largely responsible for the prevention of abnormal blood clot formation in blood vessels (Vasiliev and Gelfand, 1978). Disruption of this protective layer exposes the subendothelium which interacts with the integral glycoproteins of the plasma membrane of platelets, and this process initiates the formation of the platelet plug. The platelet can bind either directly to subendothelial collagen or through a multimeric plasma protein, von Willebrand factor, which interacts with both collagen and surface receptor glycoproteins on the platelet plasma membrane (Packham and Musturd, 1984). Binding of platelets to the vascular basement membrane proteins laminin and fibronectin also helps to localize platelets to the injury site at the early stage of hemostasis. Platelets are subsequently activated following contact with collagen in the presence of trace 1 Introduction amounts of the coagulation factor thrombin. The activated platelets release thromboxane A2 and A D P which act synergistically to recruit more platelets to form a loosely aggregated platelet plug (Colman et al., 1987). Platelet activation is a complicated process consisting of a series of progressive overlapping events. Immediately after their exposure to activating agents, the platelets change shape from flattened disks to spheres with multiple projecting pseudopods. This change in shape is a result of the polymerization of actin and myosin monomers to form actin and myosin microfilaments (White, 1979). Platelet activation also causes calcium ions to flow from the extracellular fluid and intracellular dense tubular system into the cytosol. The rise in intracellular calcium concentration initiates a number of cellular responses. Arachidonic acid is liberated and oxidized to produce thromboxane A2, a potent mediator of platelet aggregation and vasoconstriction. The calcium flux also induces fusion of the secretion granules (oc-granules and dense bodies) with the open cannulicular system (Holmsen and Weiss, 1979). As a result, the contents of the granules are released into the plasma and surrounding tissues. The cc-granules contain a variety of proteins involved in hemostasis. The dense granules contain ADP and calcium ions (Holmsen and Weiss, 1979). A D P plays an important role in the amplification of platelet activation and in recruiting new platelets into the growing mass. Finally, the calcium flux also initiates contraction of the actomyosin microfilaments responsible for the contraction of the fibrin clot (Holmsen and Weiss, 1979). Simultaneous with platelet plug formation, exposure of the injured vessel wall and activated platelets to the plasma clotting factors initiates blood coagulation. This process involves a sequential enzymatic activation of a family of serine protease proenzymes by limited proteolysis. Efficient coagulation requires the presence of a number of cofactors, 2 Introduction phospholipids and receptors on the activated platelet surface (Furie and Furie, 1988). The blood coagulation cascade can be subdivided into two steps: the generation of the coagulant enzyme thrombin and the reaction of thrombin with fibrinogen and XHIa leading to the deposition of cross-linked polymers of fibrin onto the platelet mass. The result of the coagulation process is the formation of a stable hemostatic seal consisting of a fused mass of platelets reinforced by the meshwork of an insoluble cross-linked fibrin clot (Doolittle, 1981). In addition to the initiation of coagulant forces responsible for the formation of a stable hemostatic seal, anticoagulant mechanisms are also set in motion to prevent the excessive growth of the platelet mass and fibrin clot. Endothelial cells in the vicinity of the injury site secrete prostacyclin and tissue plasminogen activator to inhibit platelet aggregation and to activate fibrinolysis respectively (Colman et al., 1987). The hemostatic seal is constantly being dissolved and reformed at the injury site. This remodelling process occurs simultaneously with the repair of the vessel wall. Tissue repair is stimulated by platelet-derived growth factor which is a potent mitogen of fibroblasts and smooth muscle cells on the vessel wall. While the vessel wall is being repaired, the fibrin clot is slowly degraded by plasmin. Tissue plasminogen activator (tPA) and urokinase are physiological activators of plasminogen secreted at low levels by vascular endothelial cells (Lijnen and Collen, 1982). Both tPA and urokinase activate plasminogen by limited proteolysis. Excessive fibrinolysis is prevented primarily by the potent inactivator of plasmin called oc2-antiplasmin (Lijnen and Collen, 1982). oc2-antiplasmin inhibits plasmin by binding to it within seconds; subsequently, the complex is cleared by the liver. a2-antiplasmin is also covalendy incorporated into the fibrin clot by the action of factor Xl l l a ; fibrinolysis will only occur when the local production of tPA exceeds the capability of the clot-bound oc2-3 Introduction antiplasmin to inhibit plasmin. Because of the potency of c£2-antiplasmin, active fibrinolysis requires a constant supply of new plasmin (Collen, 1980). Excessive fibrinolysis is also prevented by the rapid clearance of the tPA from the blood by the liver. The biological half-live of tPA in human plasma is 15 minutes (Colman et al., 1987). B. BLOOD COAGULATION Blood coagulation involves the sequential enzymatic activation of a number of serine protease proenzymes resulting in the formation of the cross-linked polymer of fibrin molecules. The process can be subdivided into two steps: the formation of the coagulant enzyme, thrombin, and the deposition of the insoluble fibrin clot on the platelet mass. Successive proteolytic activation of proenzymes at the beginning of coagulation has an amplification effect, therefore, a relatively high concentration of thrombin is formed at the end. Cofactors such as factor Va, V i l l a , phospholipids, surface receptors and calcium ions are required to confer specificity to the enzymatic reactions and to localize the reactions to the area of injury. Two pathways for the initiation of blood coagulation have been characterized: the intrinsic and the extrinsic pathway. Both pathways generate factor Xa as the physiological activator of prothrombin. The intrinsic pathway involves four plasma proteins: factor XI , X U , prekallikrein and high molecular weight kininogen (HMWK) (Fig. 1). The pathway is initiated by the reciprocal activation of factor XII and prekallikrein when blood is exposed to a negatively charged surface (Cochrane, 1982). Subendothelial collagen exposed after vessel damage may provide the negatively charged surface required for this reaction. It appears that the contact activators (XII, XI , H M W K and prekallikrein) of the intrinsic pathway only plays a minor physiological role because of the absence of bleeding disorders 4 Introduction Fig. 1. The blood coagulation cascade Oudine of the blood coagulation cascade showing both the intrinsic and the extrinsic pathways. Factor V i l a also activates factor IX and this cross-over may explain the clinical significance of factor IX. Both pathways results in the formation of factor Xa which converts prothrombin to thrombin. Thrombin subsequently leads to the formation of the insoluble fibrin clot (Neurath, 1984). 5 Introduction Anionic surface Kiniriogen Prekallikrein Factor X H Factor XI Factor Xl la EXTRINSIC PATHWAY INTRINSIC PATHWAY Factor IX Factor XIa Tissue factor Factor V E Factor X Factor I X a ^ Factor VIIa___— Factor X Factor Vi l l a Prothrombin Factor Xa Factor Va Fibrinogen Thrombin Ca++ Fibrin 6 Introduction in patients who are deficient in these factors. On the other hand, the second part of the intrinsic pathway (factor IX, X and cofactor factor VIII) is of great physiological significance as indicated by the severe effects of deficiency of these proteins on hemophilia patients . Although factor IX can be activated by factor Xia , it may be more physiologically relevant to include it in the extrinsic pathway (see below) which plays a more dominant role in blood coagulation. The extrinsic pathway is initiated by the binding of factor V U to tissue factor. Tissue factor is a specific membrane lipoprotein found primarily in the perivascular tissue which is made available to the blood only when the vascular endothelium is disrupted (Nemerson and Nossel, 1982). After the binding of factor V U to tissue factor, the complex possesses weak enzymatic activity for cleaving factor FX and X to factor IXa to Xa. Factor IXa and Xa subsequently feedback to activate factor VII to Vna which has a greatly enhanced enzymatic activity towards the activation of factor IX and X . Both intrinsic and extrinsic pathways result in the formation of the platelet bound factor Xa. During hemostasis, thrombin is the principal product formed on the surface of the activated platelets. The conversion of prothrombin to thrombin by factor Xa requires the presence of factor Va, calcium ions and phospholipids on the activated platelet surface. Upon binding to calcium ions by the y-carboxyl glutamic acid (Gla) domain, prothrombin undergoes conformational changes to expose the membrane binding site (Borowski et al., 1986). In addition, some of the Gla residues may directiy bind to the platelet surface through calcium bridges to specific platelet phospholipids (Mann et al., 1982). Factor Va also plays a role in the assembly of this macromolecular complex by interacting with both prothrombin and factor Xa. After losing the Gla domain and the factor Va binding site upon activation, thrombin is able to move off of the platelet surface and diffuse into the 7 Introduction surrounding plasma. Thrombin releases small peptides (fibrinopeptide A and B from the N-terminal regions of the a and 6 chains) from fibrinogen in the presence of calcium ions. This exposes the active sites of fibrin and results in both end to end and lateral polymerization of fibrin monomers to form fiber strands (Doolittle, 1981). Thrombin also activates factor XIU to factor X l l l a which subsequently cross-links polymerizing fibrin molecules. The insoluble crosslinked fibrin clot is deposited on the growing platelet mass and a stable hemostatic seal is formed. Thrombin plays a central role in hemostasis. It triggers platelet activation which is crucial for the formation of the platelet plug. As a coagulant enzyme, it activates several coagulation factors (factor V , VIE and XIU) and converts fibrinogen to fibrin to form a stable blood clot. In addition, thrombin is responsible for initiating the anticoagulation reaction by interacting with thrombomodulin on endothelial cells. This alters the substrate specificity of the enzyme and enables it to activate protein C. Activated protein C, anchored to the cell surface by binding to the membrane bound protein S, converts factor Va and V i l l a to their inactive form by limited proteolysis. The concentration of active thrombin in the blood is regulated by antithrombin III which forms a 1:1 stoichiometric complex with thrombin involving the active site serine. Plasma levels of antithrombin III, together with the protein C activity, are essential mechanisms to control blood coagulation. C. PROTHROMBIN Prothrombin is a plasma glycoprotein synthesized in the liver. During its biosynthesis, it undergoes several post-translational modifications including cleavage of the pre- and pro-peptides, glycosylation and Y-carboxylation prior to secretion into the blood. The complete protein and cDNA sequences of bovine (Magnusson et al., 1975, 8 Introduction MacGillivray and Davie 1984) and human (Butkowski et al., 1977, Walz et al., 1977, Degen et al., 1983) prothrombin have been reported. The normal plasma concentration in humans is 150 ug/ml or 2.1 u M . Mature human prothrombin contains 579 amino acid residues with a molecular weight of 72 kD, including three carbohydrate chains and 10 Gla residues. The protein consists of a prepropeptide, a Gla region, an aromatic acid stack domain, two kringles and a catalytic serine protease domain (Fig. 2). The prepro-leader sequence is 43 amino acids in length and is removed before secretion. Activation of human prothrombin by factor Xa releases an activation peptide of 271 residues. The 10 Gla residues in the amino-terminal of the activation peptide are the result of post-translational modifications of glutamate residues by a vitamin K dependent carboxylase (Suttie, 1980). These Gla residues are responsible for the interaction of prothrombin with the activated platelet surface; factor VII, IX, X , protein C and protein S are also vitamin K dependent proteins (Jackson and Nemerson, 1980). The aromatic amino acid stack is a short linking peptide sequence (Phe-Trp-X-X-Tyr) which is also found in factor VII, LX, X and Protein C. This short peptide is orientated towards the surface in prothrombin fragment 1 in spite of its hydrophobic nature, and it may participate in membrane binding (Park and Tulinsky, 1986). In addition, the activation peptide contains two highly conserved homologous regions called kringles which are also found in plasminogen, tissue plasminogen activator, urokinase and factor XII (Furie and Furie, 1988). The function of the kringles is still unclear; however, there is evidence suggesting that these domains may interact with other proteins. For example, the second kringle of prothrombin appears to bind to its cofactor, factor Va (Esmon and Jackson, 1974) and the second kringle in tissue plasminogen activator binds to fibrin (Zonneveld et al., 1986). 9 Introduction Fig. 2. The organization of the human prothrombin gene, mRNA and protein (A). Organization of the human prothrombin gene (Degen and Davie, 1987). The gene spans 21 kbp with 14 exons (black boxes) separated by 13 introns (white boxes). (B). Organization of the human prothrombin cDNA (Degen et al., 1983). The cDNA is about 2 kbp in size encoding 579 amino acids. (C). Graphical representation of the human prothrombin polypeptide chain. The protein consists of the following modules: pre-pro peptide, Gla domain, aromatic amino acid stack, 2 kringles and the catalytic serine protease domain. The arrows indicates the factor Xa cleavage sites (amino residues 271 and 320). 10 Introduction 5 10 15 20 Kb J I I I • II 1 1 1 3 human prothrombin gene II • M • • • I • ! • inniiv vvi vnvmix xxixn xraxiv &on 0 500 1000 1500 2000 nucleotides i-i I I P | * | - i l n i i m m i i ^ i i l i i i i m p i m — I P°'y A tail 0 100 200 300 400 500 amino acids human prothrombin cDNA t t Factor Xa human prothrombin SI pre-pro-peptide gamma-carboxylglutamic acid aromatic amino acid stack kringle catalytic domain 11 Introduction The catalytic domain of thrombin consists of two polypeptide chains (A and B chains) linked by a single disulfide bond. Thrombin belongs to the family of trypsin-like serine proteases. A l l members of this family share regions of amino acid sequence identity around the catalytic domain and the substrate binding site (Neurath, 1984). In human thrombin, the amino acids serine, aspartic acid and histidine at positions 205,99 and 43 of B-chain form the catalytic triad in the active site and the presence of these three amino acids is a characteristic of all serine proteases. A high-resolution X-ray crystal structure of the human a-thrombin (1.9 A) has been determined (Bode et al., 1989) and the structure of human B-chain showed remarkable agreement with a predicted 3-dimensionaI model of thrombin based on the crystal structure of trypsin and chymotrypsin (Magnusson et a l , 1975, Bode et al., 1989). The active site cleft can be viewed as a deep narrow canyon with hydrophobic rims and a charged/polar base. This cleft extends to the arginine rich surface loop (Lys70 to Glu80) and is likely to be part of the anionic secondary binding region for both fibrinogen and the competitive inhibitor, hirudin. The hydrophobic substrate binding site together with the thrombin specific insertion loop (Tyr-Pro-Pro-Trp, residues 46 to 50 of the B-chain) determines the limited substrate specificity of the enzyme (Bode et al., 1989). The bovine and human prothrombin genes have been characterized (Irwin et al., 1985, Irwin et al., 1985, Degen and Davie, 1987). The human prothrombin gene is mapped to chromosome 11 (Royle et al., 1987) and it is 21 kbp long containing 14 exons and 13 introns. Comparison of the gene structure and amino acid sequence of prothrombin to other clotting factors suggests that most of the clotting proteins have evolved from a common ancestral gene. New protein functions have evolved as a result of gene duplication events followed by modifications and by exon shuffling events. Exon 12 Introduction shuffling is the rearrangement, fusion and/or duplication of small gene fragments resulting in the formation of hybrid proteins. These small gene fragments containing discrete domains of the noncatalytic regions can be viewed as mini-genes. Incorporation of these mini-genes into the clotting factors by exon shuffling is likely to be responsible for the evolution of the multimodular nature of clotting factors (Patthy, 1985). When the prothrombin gene is compared to other clotting factors, it is found that exons I, II and UI (coding for the leader peptide, the Gla-domain and the aromatic amino acid stack) are homologous in structure and organization to the corresponding exons in the factor IX gene family (factor VII, IX, X and protein C). However, the catalytic domain encoding exons (exons X - X I V in prothrombin and exon VII-VIII in the factor IX family) have a different gene organization despite the marked structural homology at the protein level (Furie and Furie, 1988). This suggests that the transfer of exons I, II and IH from prothrombin gene to the factor DC-like precursor by exon shuffling is a more recent event in evolution when compared to the gene duplication which gives rise to the factor IX gene family (Irwin et al., 1988). Alternatively, the prothrombin gene precursor may have contained exons I, II and HI before the gene duplication event, and subsequent reorganization of the catalytic domain in the factor K gene family by intron deletion may have led to the consolidation of the exons encoding the serine protease domain. Exons VI-VII in the prothrombin gene code for two kringle domains. This motif is not found in the factor IX gene family, but it is observed in many other proteins (mentioned earlier). This suggests that insertion of the kringle domains into the prothrombin gene occurred after the initial gene duplication event (Castellino and Beals, 1987, Irwin et al., 1988). 13 Introduction D. EXPRESSION OF THE HUMAN PROTHROMBIN GENE As severe hepatocellular disease results in the depletion of most clotting factors (except for factor VIII and von Willebrand factor), the liver is considered to be the exclusive source of most of the blood clotting proteins. Activated macrophages also produce small amounts of clotting factors but these are insignificant relative to the required daily supply of these proteins (0sterud et al., 1980). A human hepatoma cell-line, HepG2, has been shown to synthesize and secrete a variety of liver-specific proteins including prothrombin, antithrombin III and other vitamin K dependent proteins (Fair and Bahnak, 1984, Fair and Marlar, 1986). As determined by 35s-methionine labeling experiments, prothrombin represents about 0.5% of the total protein synthesized in the HepG2 cells (Karpatkin et al., 1987). These results indicate that the HepG2 cell-line is a differentiated cell-line and is useful for the study of prothrombin expression (see later section). The plasma level of prothrombin remains relatively constant even after trauma because prothrombin is present in a relatively high basal concentration (2.1 uM) with a long biological half-life (3 days). It is therefore believed that prothrombin is expressed constitutively in the hepatocytes. There is some evidence suggesting that vitamin K and the activation fragment of prothrombin may participate in the regulation of prothrombin expression (Munns et al., 1976, Graves et al., 1981, Graves et al., 1982). Vitamin K is required for the synthesis of all functional vitamin K dependent proteins in the liver. The precursor forms of these proteins are synthesized and translocated into the lumen of the rough endoplasmic reticulum. The propeptide regions are highly conserved and have been proposed to be responsible for the targeting the proteins for y-carboxylation (Jorgensen et al., 1987). Selected glutamic acids in the Gla domains 14 Introduction are y-carboxylated by a membrane bound y-carboxylase, and this post-translational modification reaction requires reduced vitamin K, molecular oxygen and carbon dioxide (Suttie, 1980). In case of vitamin K deficiency, the vitamin K dependent proteins can be demonstrated in the plasma at low levels by immunological techniques but are ineffective in blood coagulation due to hypo-y-carboxylation. The physiological activities of prothrombin with 8,7 and 6 Gla-residues are 20, 8, and 3% respectively when compared to normal prothrombin (10 Gla-residues) (Malhotra, 1989). After exposure of H-35, a rat hepatoma cell-line, to vitamin K for 18 h, the prothrombin pool in the extracellular medium was doubled, suggesting that vitamin K may be involved in prothrombin expression (Munns et al., 1976). In warfarin-treated rats, the de novo synthesis rate and the biological half-life of prothrombin were the same as the normal control rats. However, the prothrombin produced in these rats was hypoglycosylated and the plasma level was only 6% of the control (Tollersrud et al., 1989). These results suggest that warfarin-induced vitamin K deficiency does not affect the production rate of rat prothrombin in hepatocytes. Rather, a large amount of the prothrombin synthesized is degraded intracellularly probably due to improper y -carboxylation and glycosylation. Similar results were obtained when HepG2 cells were treated with warfarin; there was an accumulation of hypoglycosylated lower molecular weight prothrombin species inside the cells (Karpatkin et al., 1987). Again, these results indicate that vitamin K is not involved in the regulation of prothrombin gene expression at the transcriptional level, but it affects the stability of the protein by promoting proper y -carboxylation and glycosylation. The amino-terminal fragment of prothrombin released after activation has been postulated to stimulate the synthesis of prothrombin. Addition of exogenous bovine 15 Introduction prothrombin fragment 1 (F- l , residues 1-156) to a rat hepatoma cell-line H-35 induces the production of endogenous prothrombin and factor X (Graves et al., 1981, Graves et al., 1982). However, there is no further evidence to support this hypothesis, until recently when it was demonstrated that injection of a high concentration of bovine fragment-1 into rabbits (10 mg/rabbit) was associated with a transient increase in plasma levels of prothrombin and factor X (Mitropoulos and Esnouf, 1990). The role of fragment-1 in the regulation of vitamin K dependent proteins is still unclear. The developmental regulation of the prothrombin gene expression was studied in fetal and newborn lambs (Kisker, 1990). Prothrombin levels in the plasma increased with gestation time. However, when the levels of prothrombin mRNA were compared, there was no significant difference between fetal and newborn livers, suggesting that the regulation is more likely to be post-transcriptional. E. EUKARYOTIC GENE REGULATION Eukaryotic gene expression is a highly controlled process involving many cellular mechanisms operating at transcriptional, translational and post-translational levels. Since transcription is the first step in gene expression, its control overrides regulation at all other levels. Mechanisms involved in controlling RNA processing, mRNA stability, translation efficiency, post-translational modifications and rate of protein degradation are important for the fine-tuning of the availability of gene products. There are three different types of eukaryotic R N A polymerases (I, II and UI) which transcribe three types of genes (ribosomal R N A genes; protein-coding and small nuclear U RNA; and 5S and transfer R N A genes respectively). The regulation of RNA polymerase II genes is of particular interest because these genes code for all known proteins expressed in the cell. 16 Introduction 1. Active chromatin As indicated by the different levels of oc-fetoprotein production in various transgenic mouse lines, chromosomal position and hence local chromosomal environment is crucial to gene expression (Hammer et al., 1987). To initiate transcription, transcription factors have to gain accessibility to the genes located on active, open chromatin structures known as the euchromatin as opposed to the inactive, condensed heterochromatin. Heterochromatin is the region of D N A that has failed to decondense from metaphase chromosome structure in interphase. Active euchromatin is distinguished from the inactive heterochromatin by a number of morphological and biochemical characteristics. The regular packing of D N A into nucleosomes is altered or even absent from genes that are actively being transcribed. This is demonstrated by the classical example of the puffs observed in the giant polytene chromosomes in the salivary glands of Drosophila. The puffs are regions corresponding to active chromatin as indicated by both in situ hybridization using ^H-RNA and transcription autoradiography using ^H-uridine (Bonner and Pardue, 1977). Electron micrographs of active chromatin in Xenopus oocytes also exhibit smooth, DNA-like structures within the transcribed regions (Labhart and Koller, 1982). Biochemically, the observation of the DNase I hypersensitivity sites localized at the 5' and 3' ends of activated genes also suggests that these genes are less protected from histone proteins (Watson et al., 1987). About 25% of the SV40 minichromosomes in the late infection stage are organized into a typical structure with a nucleosomal gap around 400 bp long, extending through the intragenic region that contains the origin of viral replication, and the early and the late promoters. This region is also DNase I sensitive as well as transcriptionally active (Jakobovits et al., 1980, Shakhov et al., 1982). 17 Introduction Mammalian genomic DNA is methylated at the 5 position of the cytosine ring when the DNA sequence is CpG to yield mCpG. Methylation and demethylation may affect the interaction of transacting factors with the DNA, since the 5-methyl group protrudes into the major groove of the double helix where DNA-protein interactions often take place. In fact, DNA hypomethylation is always observed in constitutively expressed house keeping genes while inactive genes are usually almost fully methylated, suggesting that methylation results in gene repression and demethylation leads to gene expression (Weintraub, 1985). Treatment of certain cell types with a potent demethylation agent, 5-azacytidine (5-azaC), leads to the expression of genes that are normally repressed due to methylation. Demethylation resulting in the activation of a certain sets of genes which may be part of the programmed process of development, 5-azaC is able to induce in vitro differentiation of embryonic mouse fibroblasts (10T1/2 cells) to various mesodermic lineages such as the myofibers, chondrocytes and adipocytes (Konieczny and Emerson, 1984). The newly developed methylation pattern can then be inherited in a clonal manner by the activity of the enzyme, maintenance methylase, which methylates hemi-methylated CG sequences after DNA replication (Alberts et al., 1989). 2. The cis-acting regulatory elements The control of the rate of RNA chain initiation in RNA polymerase II genes is achieved by the specific interactions of trans-acting protein factors with the cis-acting regulatory DNA motifs usually situated upstream of the transcription start site. Two kinds of cis-acting DNA regions have been shown to be essential for transcription in eukaryotic polymerase II genes: promoters and upstream/enhancer elements. Promoters are usually located within 150 bp from the CAP site and they are required for accurate and efficient 18 Introduction initiation of transcription in a vectorial manner (Watson et al., 1987). On the other hand, enhancer elements potentiate transcription of the proximal elements independent of orientation and distance from the CAP site. Both promoter and enhancer regions contain distinctive DNA-protein recognition motifs, usually in the form of short consensus sequences; each motif defines a specific regulatory function. These DNA motifs are functionally interchangeable between different promoters and enhancers. Each combination of various DNA motifs will result in a unique and distinct expression pattern. 3. The protein factors The protein factors that recognize specific DNA elements can either facilitate or inhibit transcription of an adjacent gene. These protein factors also exhibit modular characteristics. They contain at least two domains: a DNA-binding and an activation domain. The DNA-binding module recognizes the three-dimensional structure of the DNA sequence and then interacts with potential hydrogen bonding groups and hydrophobic interaction sites in the major and minor grooves of the double helix (for reviews, see Johnson and McKnight, 1989, Nussinov, 1990). The activation module is responsible for the transcriptional activation by either stabilizing or promoting the formation of the transcription initiation complex. The modules of the trans-acting factors are functionally interchangeable. For instance, the fusion protein consisting of the intact yeast activator GAL4 with a bacterial DNA-binding protein LexA can trans-activate a gene situated downstream of a LexA-recognition site in yeast (Brent and Ptashne, 1985). The modular nature of protein factors suggests that new DNA-binding proteins can evolve by gene duplication followed by modification events. In fact, the mammalian AP-1 enhancer binding factor (Bohmann et al., 1987) shares extensive structural and functional properties 19 Introduction with the human proto-oncogene jun (Struhl, 1988) and the chicken sarcoma oncogene fos (Rauscher et al., 1988). In addition, the DNA-binding domain of the oncoprotein jun shows significant amino acid sequence identity (44%) with the yeast transcription factor GCN4 (Struhl, 1987, Vogt et al., 1987), and both factors, together with the mammalian AP-1, recognize the same consensus sequence (5' TGACTGA 3'). These proteins have probably evolved from a common ancestor gene by modifying both the DNA-binding and activation domains. In addition, many of these protein factors are activated or inactivated by protein phosphorylation. For instance, C-kinase is able to induce activity of the transcription activators AP-1 and AP-2; A-kinase is able to activate another factor called cAMP responsive element binding protein (CREB). Modifications of pre-existing protein factors by phosphorylation or ligand binding are essential for the regulation of inducible and repressible promoters (Alberts, 1989). 4. The leucine zipper Several protein-protein interacting motifs and DNA-binding motifs have been identified including the leucine zipper, the helix-turn-helix and the zinc finger motifs. The leucine zipper motif has been observed in a CCAAT enhancer binding protein (C/EBP) (Landschulz et al., 1988), and many other transcription factors in eukaryotes ranging from GCN4 in yeast, and several oncogene products: fos, jun and myc (Landschulz et al., 1988). In these proteins, leucyl residues appear at every seventh position (2 complete oc-helical turns) in the polypeptide backbone for at least four consecutive times. The hydrophobic leucyl residues, situated on the same side of the amphipathic a-helix, interdigitate with the leucines of the leucine zipper domain of another factor leading to 20 Introduction dimerization. In C/EBP and GCN4, the region adjacent to the leucine domain is positively charged. The formation of homodimers in these proteins brings the basic regions together to form a sequence-specific recognition domain that is able to recognize a palindromic sequence (Johnson and McKnight, 1989). There is also direct evidence indicating that a leucine zipper is responsible for the heterodimer formation in jun and/as. Similarly, the basic amino acid residues close to the leucine zipper domain also contribute to the D N A binding ability of the complex (Sassone-Corsi et a l , 1988). As demonstrated by domain-swap experiments, dimer formation is protein-specific; the leucine zipper domains oifos, jun and GCN4 alone are sufficient to determine the selectivity of dimer formation (Kouzarides and Ziff, 1989). The putative leucine zipper is also found in a number of membrane proteins, including glucose transporter proteins and a family of voltage-gated potassium channels (Abel and Maniatis, 1989). Hence, this motif may be used by a variety of cellular proteins for specific protein-protein interactions. 5. DNA-binding motifs The helix-turn-helix (HTH) motif consists of two successive anti-parallel a-helices juxtaposed at approximately 90° by a turn of 4 amino acids. As indicated by the X-ray crystal structure of the X repressor-operator complex , the recognition helix (helix-3) interacts with the bases in the major groove of the B-form of D N A in a sequence specific manner. Helix-2 is situated above helix-3 to lock the recognition helix into the correct conformation (Jordan and Pabo, 1988, Matthews, 1988). This DNA-binding motif is present in many bacterial repressors and probably in eukaryotic regulatory proteins such as the two gene products of the mating type locus of Saccharomyces cerevisiae M A T a l and MATot2, as well as homeobox proteins that regulate the development of the fruit fly 21 Introduction Drosophila melanogaster (Scott and Weiner, 1984). Homeobox motifs consist of 60 amino acids which share extensive sequence identity with each other (typically exceeding 90%). A putative H T H motif is observed within the carboxyl half of the homeobox element suggesting that these proteins can recognize specific D N A sequences (Robertson, 1988). The high degree of sequence identity between the homeobox proteins indicates that the target D N A sequences are likely to be similar or identical. The zinc finger is another DNA-binding motif first discovered in the Xenopus RNA polymerase UI transcription factor TFILTA in tandem repeats (9 copies) (Miller et al., 1985). The zinc ions in the zinc finger motif of TFIIIA are tetrahedrally coordinated by two cysteine and two histidine residues (C2H2 class). This metal-binding motif is frequently found in many other transcription factors, including RNA polymerase II (Saltzman and Weinmann, 1989), Spl (Kadonaga et al., 1987), and the glucocorticoid and the estrogen receptors (Green and Chambon, 1987). Mutation of the finger motifs of the glucocorticoid and the estrogen receptors results in the loss of expression of genes regulated by these proteins. Additionally, the chimeric proteins formed by replacing the finger domains of the human estrogen receptor with the glucocorticoid zinc fingers, activates expression of a glucocorticoid-inducible reporter gene (chloramphenicol acetyl transferase, CAT) in a transient expression system (Green and Chambon, 1987). The domain-swap experiment clearly demonstrates that the zinc finger motif alone is responsible for DNA-specific activation in these activators. Interestingly, in addition to the positive regulatory function, the glucocorticoid receptor also acts as a hormone-dependent negative factor. By site directed mutagenesis, at least three out of seven receptor binding sites (consensus, 5' AGAa/tCAGa/t 3') in the prolactin gene are responsible for the negative regulation of gene expression (Sakai et al., 1988). 22 Introduction F. THE PROMOTER The promoter is a region of DNA proximal to the transcription start site involved in the binding of RNA polymerase to initiate transcription. The promoter region is usually within 100 to 150 bp of the CAP site and it determines the position of transcription initiation, the basal rate and the unidirectional transcription of the gene. Detailed analysis of eukaryotic polymerase II promoters reveals a common pattern of organization. A typical promoter consists of a TATA box, a CCAAT box and distal regulatory elements (see Fig. 3). However, there are many functional promoters which do not follow this organization of DNA motifs. 1. The TATA box The TATA box is a highly conserved AT rich region (5' TATAAA 3' consensus) usually beginning 30 bp upstream from the CAP site. Deletion or mutation of this sequence in most promoters has severe adverse effects on both the rate and the accuracy of transcription initiation (Dierks, 1983, McKnight, 1982). The position of the TATA box in most higher eukaryotes varies from -30 to -120 bp from the CAP site, and in some cases, it is totally absent. In the SV40 late promoter, there is no putative TATA box and small deletions or insertions upstream from the start site do not alter the position of the mRNA 5' end in an in vivo study (Somasekhar and Mertz, 1985). In this TATA-less promoter, the CAP site is not dependent on the upstream sequences, but it is determined by the initiator elements around the start site such as the CA motif (see below). 23 Introduction Fig. 3. Organization of a eukaryotic polymerase II promoter The proximal promoter region interacts with the general transcriptional factors (TFIIA, TFLTB, TFILD, TFLTE, TFILF and RNA polymerase n). Various other transcription factors (TFs) binding to the distal element(s) and enhancer region(s) also participate in the activation and the regulation of transcription initiation. The interactions between these TFs and the general transcription factor-DNA complex is still unclear (Wasylyk, 1988). 24 Introduction 25 Introduction The sequence of the CAP site shows a strong preference for 5' C A 3' where A is the +1 position (Wasylyk, 1988). Generally, the significance of the sequence to the rate and accuracy of transcription initiation is minimal; however, in some cases such as the adenovirus major late promoter, a point mutation of the CAP site reduces transcription (Concino et al., 1984). In the rainbow trout protamine TG3 gene, changing the spacing between the T A T A box and the CAP site by a 3 bp insertion results in initiation at a distance 3 bp upstream of the usual CAP site; in this promoter, the T A T A box is the determining factor for the CAP site (Kovacs and Butterworth, 1986). The eukaryotic transcription factor TFTID has been identified as the T A T A binding protein, and is required by almost all the class II promoters (Wasylyk, 1988). TFHD is believed to associate with TFIIA and the sequences between +35 to -45 of a promoter to form a stable pre-initiation complex (Reinberg et al., 1987, Wasylyk, 1988). Promoters without an apparent T A T A box also form complexes with the same general transcription factors (TFIIA, TFIIB, TFIID, TFH.E and TFIIF) but with lower affinities (Wasylyk, 1988). On the other hand, the mammalian U2snRNA promoter has no requirement for TFIID, and this TATA-less promoter is neither activated nor repressed by typical protein factors, suggesting that the TFIID may be the target protein of transcription activators (Lillie and Green, 1989). This is supported by studies of the adenovirus E4 gene promoter. TFHD interacts with the mammalian activator factor (ATF), and the cooperative binding of these two factors will subsequently potentiate their interactions with other transcription factors (TFIIB, T R I E and RNA polymerase II) with the promoter to form the initiation complex (Horikoshi et al., 1988). In summary, TFIID is probably involved in 26 Introduction the initial interactions of protein factors with the promoter to enable the assembly of the pre-initiation complex. 2. The CCAAT box Further upstream of the CAP site is the C C A A T box (5' G G C C A A T C T 3' consensus), usually located around -80. The C C A A T box is highly conserved in many promoters and it is crucial to the expression of many genes including the HSV thymidine kinase, the rabbit B-globin and the human al-globin promoters (Dierks, 1983, McKnight, 1982, Mellon et al., 1981). However, the significance of the C C A A T box is promoter-specific; there are no detectable change in the transcription rate when the putative C C A A T box of the SV40 early promoter is deleted (Fromm and Berg, 1982). There are several C C A A T box binding proteins that have been identified, characterized and expressed (Santoro et al., 1988). At least three different C C A A T box binding proteins (CP-1, CP-2 and NF-1) are found in the Hela cell nuclear extract, and each of these proteins forms a heterologous subunit complex which recognizes the C C A A T motif (Chodosh et al., 1988). For example, CP1 and CP2 each contain two subunits (CP1 A , CP1B, CP2A and CP2B) separable by chromatographic means, and binding activities are reconstituted only if the appropriate subunits are mixed together. Similarly, there are at least four C C A A T binding proteins found in the rat liver nuclear extract: C/EBP, NF-1, a third non-identified protein which is probably YB-1 (Raymondjean et al., 1988, Didier et al., 1988) and the D-box binding protein (DBP) (Mueller et al., 1990). This multimeric protein system provides insight to gene regulation. Different subunit concentrations may be responsible for induction, repression and/or tissue specific expression in various tissues. 27 Introduction 3. Distal regulatory elements A number of distal regulatory cis-acting motifs have been demonstrated to regulate gene expression constitutively or non-constitutively (inducible or repressible). These elements may overlap with the C C A A T box and are generally situated between -40 to -150. Spl was one of the first constitutive factors to be identified, and recognizes a G C rich hexamer (5' C C G C C C 3'). This motif is found in many cellular and viral genes including the thymidine kinase promoter (2 copies) and the SV40 early promoter (6 copies). Linker scanning mutation analysis of the thymidine kinase promoter has shown that, in addition to the T A T A and the C C A A T boxes, disruption of the G C motifs (inverted repeats at -100 and -50) also greatly reduced the promoter activity (McKnight et al., 1984). The predicted amino acid sequence of an Spl cDNA clone revealed the presence of three Zinc fingers at the carboxyl-terminal of the protein which was responsible for its sequence specific interactions with the D N A (Kadonaga et al., 1987). Although Spl is ubiquitously expressed, the levels of Spl mRNAs in various cell lineages during development in mice are different (Saffer et al., 1991). This implies that Spl , in addition to being a house-keeping trans-acting factor, may also regulate gene expression especially to low affinity binding sites since these sites are more affected by Spl concentrations in vivo. Many other inducible and repressible elements in the promoter and enhancer regions have been identified in the last decade. For instance, detailed analysis of the human metallothionein HA promoter has revealed that it contained at least nine functional motifs including the glucocorticoid responsive element, a phorbol ester responsive motif and four metal responsive elements (Karin et al., 1987). In fact, promoter and enhancer elements are very similar in both structure and function. The functional difference between the two 28 Introduction elements probably resides on the distance that the element is away from the transcription initiation site. G. THE ENHANCER An enhancer is a region of D N A , outside of the promoter, that drastically alters the transcription efficiency in either homologous or heterologous promoters. These elements are usually relatively large (50 to 100 bp) with multiple cis-acting motifs. The enhancer function is independent of orientation and position; it can reside on the 5' or 3' side of the gene, and can even reside in an intron to exert its effect. Using the transgenic mouse in vivo system, it has been showed that the albumin enhancer is capable of activating liver-specific transcription when separated by a distance of 10 kbp from the proximal promoter element (Pinkert et al., 1987). Enhancers are usually sub-divided into two categories: inducible enhancers and tissue specific enhancers. Inducible enhancers are those that can respond to changes in the environment including the collagenase gene enhancer (induced by TPA) and the metallothionine gene enhancer (induced by heavy-metal ions) (For a list of other inducible enhancers, see Kriegler, 1990). On the other hand, tissue specific enhancers activate transcription in a cell-specific manner. For example, the mouse a-fetoprotein gene contains at least three tissue-specific enhancer elements and each enhancer can direct expression at the appropriate tissues including the fetal liver, the gastrointestinal tract and the visceral endoderm of the egg yolk (Hammer et al., 1987). In the human a-fetoprotein gene, a liver-specific enhancer that interacts with a liver-specific protein factor, HNF-1, has also been mapped in the region between -3.3 kbp to -3.7 kbp from the C A P site (Watanabe et al., 1987, Sawadaishi et al., 1988). 29 Introduction The mechanism of enhancer function is still under intense investigation. It is believed that many aspects of the mechanisms involved in transcription activation are similar in both proximal promoter and distal enhancer elements. For example, both elements consist of small discrete modules and can be trans-activated by common protein factors (Maniatis et al., 1987). Several mechanisms have been proposed to explain enhancer function at a distance (Nussinov, 1990, Wang and Giaever, 1988). In the tracking model, the protein factors scan along the D N A to look for the proximal DNA-protein complex; in the looping model, the distal and the proximal protein-DNA complexes interact resulting in the formation of a D N A loop; in the nuclear address model, the enhancer binding-proteins also interact with euchromatin structures; finally, the twin-supercoiled domain model explains distal action by changing the topology of D N A loops. These mechanisms are probably interrelated and several of them may be engaged at the same time to produce the enhancer effect. The looping model is the most popular model to explain enhancer activity over a distance. The formation of a D N A loop is less likely to happen if the regulatory elements are close to each other. In the prokaryotic lactose operon construct, there is a significant decrease in repressor efficiency when the distance between the two repressor binding sites is reduced from 200 to 120 bp (Mossing and Record, 1986)1 This observation of reduction in effectiveness of the repressor protein at shorter distance strongly supports the possibility of D N A looping in in vivo situations. Another piece of evidence to support D N A looping comes from the study of interactions between the operator site and X-phage repressor protein. The cooperative binding of repressor proteins to two operator sites is observed, as indicated by D N A loop formation in electron microscopy, only if the two sites are separated by an integral number of helical turns (Griffith et al., 1986). This suggests that protein-30 Introduction protein interactions occur only when the two repressor proteins are situated on the same face of the D N A strand, as a result, a smooth D N A loop with no torsional stress can be formed. There is no direct evidence for D N A loop formation in eukaryotic systems, but it is believed that multifactor-DNA complexes are formed in the proximal promoter and the enhancer regions. By looping out the intervening sequences, these two complexes interact to stimulate transcription. As an alternative explanation for loop formation, the activation domain of the enhancer binding proteins, instead of interacting with the proximal promoter complex directly, may produce a D N A loop by scanning through an increasing length of D N A sequence until it meets the proximal complex (for review, see Nussinov, 1990). The nuclear address model may explain the actual enhancing effects of enhancers. The enhancer-binding protein may direct the proximal promoter to the nuclear matrix where transcription machinery is present at a high concentration to allow active transcription to take place. For example, the matrix attachment region of the IgH gene locus has been mapped to the close proximity to the enhancer, suggesting a relationship between enhancer function and nuclear organization (Cockerill et al., 1986). Torsional stress has been shown to affect the activity of some prokaryotic promoters Qrisher 1984), but its effect on eukaryotic systems is still uncertain. In a system using Xenopus oocyte supernatant supplemented with TFIIIA and minichromosomes of plasmids containing 5S RNA gene, transcription is stopped by novobiocin which blocks D N A gyrase (Kmiec et al., 1986). However, in the same system, topoisomerase I induced-relaxation of the plasmids does not affect transcription, suggesting D N A gyrase (DNA topoisomerase U) may directly participate in the transcription machinery while D N A topology is non-essential to transcription (Kmiec et al., 1986). On the other hand, the twin-supercoiled domain model has been proposed recently to explain the function of an 31 Introduction enhancer at a distance by virtue of its ability to affect the topology of D N A (Liu and Wang, 1987, Tsao et al., 1989, Wu et a l , 1988). When a D N A loop is anchored by protein complexes involving the cellular matrix and/or the enhancer-binding proteins at both ends, an isolated topological domain is produced. The advancing polymerase opens up the double helix and generates positive supercoils in the region in front of the polymerase and negative supercoils in the region behind it. This local change in supercoiling within the D N A loop affects the structure of the D N A of adjacent genes and may alter their transcription rate. H. UVER SPECIFIC EXPRESSION Tissue-specific gene regulation can be achieved by the use of common tissue-specific transcription factors by members of the co-expressed group. The development of molecular biology techniques in the last decade has provided opportunities to study the mechanisms of liver-specific expression at the molecular level. First of all, cis-acting regulatory elements responsible for expression in the liver have been identified and then the DNA-sequence-specific trans-activators have been characterized. Subsequently, the expression of these proteins in eukaryotic systems has provided insight into tissue-specific regulation. The cDNAs of several of these rat liver-specific nuclear factors have been cloned and sequenced including the C/EBP family (C/EBP, DBP and LAP) (Landschulz et al., 1988, Lichtsteiner et al., 1987, Descombes et a l , 1991), HNF-3A family (HNF-3A, HNF-3|3 and HNF-3Y) (Lai et al., 1990, Lai et al., 1991) and HNF-1, (Baumhueter et al., 1990). In fact, these transcription factors are also expressed in a tissue-specific manner; they are either found exclusively in the liver or exist in high abundance in the liver nuclear extract. It is likely that in the near future, proteins that regulate these liver-specific 32 Introduction transcription factors will be identified. These proteins may be crucial for the establishment of the hepatic phenotype. 1. CCAAT/enhancer-binding protein (C/EBP) C/EBP is a heat-stable, sequence-specific DNA-binding protein found in high levels in liver, adipose and placental tissues, and at low levels in lung and small intestine (Birkenmeier, 1989). The protein dimerizes through the leucine zipper motifs, and the consensus of recognition is a palindromic sequence 5' GCAATATTGC 3' (Landschulz et al., 1988). However, as indicated by in vitro binding assays, the protein is capable of recognizing sequences that diverged significantly from the optimum binding sequence. In fact, C/EBP was originally characterized by its ability to bind to the CCAAT box of the HSV TK promoter (Graves et al., 1986). The same protein was purified by conventional chromatographic techniques from nuclear extracts based on its ability to bind the enhancer core homology (5' TGTGGa/ta/ta/tG 3') (Johnson et al., 1987). Hence, the protein is termed CCAAT/enhancer binding protein. C/EBP is hypothesized to be a central regulator of energy metabolism because it trans-activates a number of genes related to glucose and lipid metabolism in the liver and adipose tissue, including the genes for serum albumin, 422/aP2 protein, stearoyl acyl-CoA desaturase and insulin-responsive glucose transporter. Expression of these liver- and adipose-specific genes can be activated by transfecting a C/EBP cDNA construct into mammalian cells (McKnight et al., 1989). 2 . D binding protein (DBP) DBP competes with C/EBP and binds to the D site of the albumin promoter (Lichtsteiner et al., 1987). The DNA binding domain of DBP is highly basic and shares 33 Introduction sequence identity with the C/EBP family; however, this protein does not possess the leucine zipper. The DBP mRNA is present in most tissues with the exception of testis, but the protein is only observed in the adult liver, suggesting that this protein is regulated at the post-transcriptional level. Another interesting observation of this protein is that it only appears in the rat liver after the albumin gene has already been expressed maximally, suggesting that DBP is non-essential to the regulation of albumin physiologically, or DBP is required for the maintenance of albumin expression. The latter explanation suggests that the establishment and maintenance of tissue-specific regulation is a result of separate regulatory mechanisms (Mueller et al., 1990). Another D box binding protein was also characterized recently (liver-enriched transcriptional activator, LAP) which is able to trans-activate the albumin promoter in cotransfection experiments (Descombes et al., 1991). The protein shares 71% sequence identity with C/EBP and it readily forms heterodimers with C/EBP through its leucine zipper motif. Although the mRNA of L A P is found in most tissues, the L A P protein is highly enriched in hepatocyte nuclei, suggesting the expression is regulated at the post-transcriptional level (Descombes et al., 1991). 3. Hepatocyte-specific nuclear factor-1 (HNF -1 or L F - B 1 ) HNF-1 is required for the expression of fibrinogen a and (3 chain genes and it also interacts with many other liver-specific genes including albumin, pre-albumin, a l -antitrypsin, a fetoprotein and transthyretin (Sawadaishi et al., 1988, Courtois et al., 1988, Kugler et al., 1988). The consensus sequence of this motif is again palindromic (5' G T T A A T N A T T A A C 3'), and the protein seems to be able to bind to sites of significant sequence divergence. The predicted amino acid sequence from the cDNA of HNF-1 shares sequence identity with POU homeo domain proteins (Baumhueter et al., 1990, Finney 34 Introduction 1990). In addition, HNF-1 may be responsible, at least in part, for the hepatic phenotype based on the study of the expression of HNF-1 in the differentiated hepatoma cell-line, Fao, and the dedifferentiated cells Fao flC2(C2). Fao flC2(C2) contains a variant form of HNF-1 (vHNF), and this cell-line has lost many of the functional characteristics of liver cells including two key enzymes in the gluconeogenic pathway (Baumhueter et al., 1988). Furthermore, a redifferentiated revertant cell-line (C2-Rev7) was obtained by selecting viable Fao flC2(C2) cells in glucose-free medium. Reversion of the hepatocyte phenotype in C2-Rev7 is accompanied by the re-expression of the normal HNF-1, suggesting that the HNF-1 may play a role in establishing the hepatic phenotype (Baumhueter et al., 1988). 4. Hepatocyte nuclear factor-3A (HNF-3A) HNF-3A binds to a sequence that is functionally important in the liver-specific expression of two genes in the mouse: transthyretin and a-1 antitrypsin. The mRNA of HNF-3A is found in the liver but not in the brain, kidney, intestine or spleen, indicating that the expression of the protein itself is also liver-specific (Lai et al., 1990). In addition, several mRNA species were found to hybridize to the HNF-3A cDNA probe, suggesting the presence of several related transcripts either by differential processing or by transcription from related genes. Recently, two additional liver-specific protein factors were found to bind to the same cis-acting element (HNF-3(3 and HNF-3y) and they shared extensive sequence identity with HNF-3A in the DNA binding domain (93 out of 110 amino residues) (Lai et al., 1991). Interestingly, this highly conserved DNA binding domain is also found in the Drosophila homeotic gene fork head. This domain is likely to be a new class of DNA binding motif possibly essential for differentiation of cells in development (Lai et al., 1991). 35 Introduction I. TRANSCRIPTIONAL REGULATION OF BLOOD CLOTTING FACTORS The tissue specific regulation of several blood clotting factors has been studied including the genes for factor IX and a- P-, and y-chains of fibrinogen. The studies of the expression of fibrinogen genes are of particular interest because these three genes are coordinately regulated to produce an equimolar ratio of liver-specific products. In addition, the expression of fibrinogen genes is controlled during the acute phase response. As indicated by transient expression of chimeric C A T constructs, in vitro footprinting and gel retention assays, the liver-specificity of both a- and P-fibrinogen genes in rats is regulated by a proximal HNF-1 element (within 100 bp from the CAP sites) (Courtois et al., 1988, Courtois et al., 1987). The Y-fibrinogen gene appears to be regulated by three non-liver-specific factors: Spl , CCAAT-binding protein and adenovirus major late transcription factor (Morgan et al., 1988). In fact, this y-promoter was found to be functional in both hepatic and non-hepatic cell-types (Morgan et al., 1988). Since all three fibrinogen genes are located in the same vicinity (within 50 kbp on chromosome 4), it is possible that the liver-specific and the coordinate expression of the y-fibrinogen gene is also controlled by the HNF-1 motif of the a-fibrinogen gene situated 18 kbp away. As an alternative explanation, it is also possible that the cis-acting element for the liver-specific expression of the y-fibrinogen gene has not been identified. During the acute phase response, the levels of fibrinogen mRNAs and polypeptides are elevated 10- to 20- fold as indicated by northern blot analysis and immunoassays. The parallel increase in both the mRNA and protein levels suggests that it is probably a result of transcriptional activation (Birch and Schreiber 1986, Fuller et al., 1985). Hepatocyte stimulating factor (HSF) or interleukin-6 (LL-6) either purified from monocyte conditioned 36 Introduction medium or produced by recombinant D N A techniques is able to elicite an acute phase response in cultured hepatocytes (Castell et al., 1988, Otto et al., 1987). Interestingly, an interleukin-6 responsive element (5' CTGGGA 3') is also found in the proximal regions of all three fibrinogen gene promoters (Green and Humphries 1989). This strongly suggests that this IL-6 motif is responsible for the control of the acute phase response of the fibrinogen genes in vivo. The understanding of the transcriptional regulation of factor DC is facilitated by the naturally occurring genetic disease hemophilia B Leyden. This disorder is characterized by low levels of plasma factor DC in affected individuals from birth. At onset of puberty, however, factor IX levels rise to those found in normal individuals. D N A sequence analysis of the hemophilia B Leyden patients has revealed point mutations in the 5' flanking region of the factor IX gene at position -20 (Reitsma, 1989), -6 (Attree, 1989) or +13 (Crossley and Brownlee, 1990) relative to the transcription start site. As indicated by footprinting assays, C/EBP is responsible for the expression of factor IX gene because it binds to the region between +1 to +18; the mutation at +13 impairs both binding of C/EBP and functional promoter activity in C A T assays (Crossley and Brownlee, 1990). The correlation of the down-regulation seen in the +13 promoter mutation in hemophilia B Leyden and the results of the in vitro experiments suggested that C/EBP is probably responsible for factor IX gene expression in vivo. A putative NF-1 binding site was found in the region between -100 and -75; deletion of this motif resulted in a 5-fold reduction of promoter activity in transient assays, therefore, it might act as an activator to the expression of factor IX gene (Crossley and Brownlee, 1990). Contradictory results were obtained by another group (Sailer et al., 1990) who claimed that the CAP site was situated 150 bp from the A T G initiation codon (as opposed to 50 bp by Crossley's group) and the region 37 Introduction essential for expression was between 150 to 250 bp away from the A T G initiation codon. In addition, there was a silencer between 1.4 to 1.7 kbp away in the 5' flanking region and its function was unclear. The results from the latter group are questionable because all the interpretations are based on the weak C A T signals obtained from the transient expression assay which has high inherent experimental errors. Moreover, the C A P site was determined by primer extension experiments using mRNAs prepared from the HepG2 cells transfected with a chimeric construct containing the factor IX promoter linked to the CAT gene. The presence of the foreign C A T gene may affect the location of transcription initiation. /. THE PRESENT STUDY The human prothrombin gene was used to study liver-specific expression because of the relatively high level of expression and the central role that prothrombin plays in blood coagulation. The plasma level of prothrombin is about 20-fold higher than factors LX and 1000-fold higher than factor VII and VIII. In addition, understanding prothrombin gene regulation may give insight to the regulation of other clotting factors and, in particular, to vitamin K dependent serine proteases involved in blood coagulation. 38 materials and methods II. MATERIALS AND METHODS A. STRAINS, VECTORS AND MEDIA 1. Bacterial strains E. coli strains JM101 (Messing 1983) and DH5oc (Hanahan 1983) were routinely used as hosts to prepare plasmid D N A for subcloning, restriction mapping and D N A sequence analysis. E. coli strains Y1090 (Young and Davis 1983) and Y1089 (Young and Davis 1983) were the host cells for plating the A.gtl 1 expression library and to produce X lysogens respectively. 2. Vectors The bacteriophages M13mpl8 and mpl9 (Messing 1983) were used to produce single stranded sequencing templates for D N A sequence analysis (Sequenase Kit, Pharmacia) according to the manufacturer's manual (Sanger et al., 1977). The plasmid vectors pUC 18 and pUC19 (Vieira and Messing 1982, Yanisch-Perron et al., 1985) were cloning vehicles for performing restriction endonuclease mapping and double stranded D N A sequence analysis (Gatermann and Kaufer 1988) using the forward and reverse sequencing primers (Table 1: 1-2). Transient expression plasmid vectors pOGH, pTKGH and pXGH5 obtained from Allegro (Nichols Institute Diagnostics, San Juan Capistrano) were used to construct chimeric promoter plasmids. 39 materials and methods 3. Media Luria broth (LB: 5g yeast extract, lOg bacto-agar and lOg of NaCl per liter with pH adjusted to 7.5) was the growth medium for E. coli strains JM101 and DH5a. After transformation of bacteria with the recombinant plasmids, the clones were selected by plating on LB-agar plates (1.5% w/v) supplemented with 100 (ig/ml of ampicillin, 25 U-g/ml of LPTG (5prime3prime Inc., West Chester, PA) and 50 |ig/ml of X-gal (5prime3prime Inc., West Chester, PA). N Z C Y M medium (5g yeast extract, lOg NZamine type A , 2g MgCl2, lg casamino acids and 5g of NaCl per liter with pH adjusted to 7.5) was used for growing E. coli strains Y1089 and Y1090. To screen for the enhancer binding P-galactosidase fusion proteins and to titer the phage, the Xgtl 1 HepG2 cDNA library (Clontech) was plated onto NZCYM-agar plates (1.5% w/v) with an overlay of NZYCM-agarose (0.7% w/v). Isolated phage clones were diluted and stored at 4°C in S M buffer (5.8g NaCl, 2g MgS04-7H20, 50mM Tris-HCl pH7.5 and O.lg of gelatin per liter) with a drop of chloroform. B. GEL ELECTROPHORESIS 1. Non-denaturing agarose gels and Southern blot analysis D N A fragments were separated on agarose gels (BRL, electrophoresis grade) using I X T A E as the running buffer (40mM Tris-OH, 20mM glacial acetic acid, 2mM EDTA) supplemented with 0.5 ug/ml EtBr. D N A samples in I X loading dye (3% ficoll, 0.2% xylene cyanol and 0.02% bromophenol blue) were loaded onto the gel and electrophoresis was carried out at 1-3 volts/cm. Southern blot analysis was performed essentially as 40 materials and methods described elsewhere (Davis et al., 1986). After transferring the D N A onto nitrocellulose paper (Schleicher & Schuell), the blot was prehybridized for 2 h at 68°C in 10 ml of 6X SSC (SSC buffer 20X stock: 3M NaCl, 0.3M Sodium Citrate- 3H20, pH 7.0), ImM E D T A , 0.1% SDS, lOmM Tris-HCl pH7.5, 10X Denhardt's solution, 100 (ig/ml denatured herring sperm D N A . Stock solutions of 100X Denhardt's solution and herring sperm D N A (10 mg/ml) were prepared as described (Maniatis et al., 1989). Hybridization was carried out in the same solution (10 ml) with 20-50 x 10^ cpm of denatured 32p. labeled probe. Radioactive D N A probes were prepared following the Klenow labeling method (see below, Labeling of DNA). After overnight incubation at 68°C, the blot was washed three times with I X SSC, 0.5% SDS at 68°C. Autoradiography of the blot was prepared using Kodak XK-1 film. 2. Formaldehyde agarose gels and Northern blot analysis Total R N A samples from rat tissues were prepared according to the acid guanidinium thiocyanate extraction technique (Chomczynski and Sacchi 1987). Total and poly A + RNA samples from human liver were generous gifts from Jeff Hewitt (University of British Columbia). Total RNAs were separated by electrophoresis on a 1% formaldehyde agarose gel as described previously (Davis et al., 1986). RNA samples (2-10 ug) including the RNA ladder (BRL) were denatured at 90°C for 2 min before loading onto the gels. After electrophoresis, the RNA samples were transferred onto Nytran paper (Schleicher & Schuell) using 10X SSC as the transfer buffer. R N A samples were cross-linked to the Nytran paper by irradiating the blot with ultraviolet light (269 nM) for 30 sec. Prehybridization and hybridization conditions of the northern blots were identical to the Southern blot analysis except that the temperature used was 60°C. The blots were 41 materials and methods subsequently washed three times with 3X SSC, 0.5% SDS at 60°C, and autoradiography was obtained using Kodak XK-1 film. 3. Non-denaturing polyacrylamide gels Non-denaturing polyacrylamide gels were used to separate and purify end-labeled double stranded D N A probe for DNase I footprint analysis as well as to perform gel retention assays. A polyacrylamide gel (4-6 % from a 29:1; acrylamide: bisacrylamide stock) was prepared in 0.25X T B E buffer (TBE buffer 10X: 1M Tris, 1M boric acid, 20mM EDTA). Polymerization of the acrylamide was initiated by the addition of 0.063% (w/v) of ammonium persulfate (BIO-RAD) and 0.125% (v/v) of T E M E D (BIO-RAD) respectively. Electrophoresis was performed at 10 V/cm at 4°C for 2-3 h. 4. Denaturing polyacrylamide gels Denaturing polyacrylamide gels were used for both D N A sequence analysis and DNase I footprint analysis. A polyacrylamide gel (6-8% from a 38:2; acrylamide: bisacrylamide stock) was prepared in IX TBE buffer with 8.3 M urea. The mixture was degassed for 5 min, and polymerization of the acrylamide was catalyzed by the addition of 0.063% (w/v) ammonium persulfate and 0.024% (v/v) TEMED, respectively. Before loading the gel, the samples were denatured at 90°C for 3 min and electrophoresis was performed at 50 W for 2-4 h in I X TBE. The gel was dried under vacuum with a BIO-R A D gel dryer. Single stranded DNAs separated by denaturing polyacrylamide gel were visualized by autoradiography. 42 materials and methods 5. SDS-denaturing polyacrylamide gels and Western blot analysis SDS-denaturing protein gel electrophoresis was performed on a BIO-RAD mini-PROTEANII system using 4-20% gradient pre-poured Ready Gels (BIO-RAD) (Laemmli 1970). The presence of the P-galactosidase fusion protein from the protein extracts of X-lysogens was detected by Western blot analysis. The proteins were electro-blotted onto nitrocellulose paper (Schleicher & Schuell). After blocking the paper with 2% skim milk powder (Carnation) dissolved in PBS-Tween (PBS with 0.05% (v/v) Tween-20 and 0.02% (wA7) NaN3) for 1 h, the blot was subsequently incubated with anti-p-galactosidase antibody (Promega) and finally developed with an alkaline phosphatase conjugated, rabbit anti-mouse antibody (Promega) according to the supplier's instructions. C . ISOLATION OF DNA 1. Isolation of plasmid DNA Small scale isolation of plasmid DNA was performed using either the alkaline lysis method or the Triton-lysozyme method (Maniatis et al., 1989). Large scale isolation of plasmid D N A was prepared using the alkaline lysis method (Maniatis et al., 1989). The plasmid D N A was further purified by CsCl density gradient centrifugation at 56Krpm, 20°C for 5 h using the TV865 vertical rotor in a Sorvall OTD-COMBI ultracentrifuge. To remove the EtBr, the supercoiled plasmid D N A band was extracted several times with water saturated isoamyl alcohol. Highly purified plasmid D N A was obtained afterwards by standard ethanol precipitation. The precipitation procedure involved the addition of l/10th volume of 3 M NaOAc pH5.0 and 2 volumes of 95% ethanol to the D N A solution. The 43 materials and methods mixture was incubated at -70°C for 10 min and the D N A pellet was obtained by centrifugation at 10,000g for 10 min followed by washing the pellet with 80% ethanol. The D N A pellet was air dried afterwards and the quality of the D N A was analyzed by running 1 ug of the sample on a 1% agarose gel. The yield of the large scale plasmid preparation was estimated by measuring absorbance at 260 nM (20 A260 u m t s = 1 rng/ml) 2. Isolation of lambda phage D N A The procedure used to obtain X phage D N A was a modification of a previously described protocol (Maniatis et al., 1989). A phage plug was added to 100 )il of an overnight culture of Y1090 cells, and the phage were allowed to preadsorb to the host cells for 15 min at 37°C. The culture was then transferred to 20 ml of prewarmed N Z C Y M and agitated at 300 rpm at 37°C until lysis was apparent. A few drops of chloroform were added , and cell debris was removed by centrifugation at 10,000g for 10 min. The phage particles were subsequendy precipitated by adding 6 ml of 50% polyethylene glycol and 3 ml of 5 M NaCl to the supernatant followed by overnight incubation at 4°C. After centrifugation at 10,000g for 10 min, the phage were resuspended into 0.5 ml of DNase I reaction mix (50mM Tris-HCl pH7.5, 5mM MgCl2, 0.5mM CaCl2, 5 (ig DNase I and 50 ug RNase A) . The mixture was incubated at 37°C for 30 min and then centrifuged for 10 min in an eppendorf centrifuge. To the supernatant, 50 | i l of 10% SDS, 5 (il of 0.5M E D T A and 0.1 mg of protease K were added. The mixture was incubated at 68°C for 1 h, and then extracted with an equal volume of the following reagents: phenol (saturated with 10 m M Tris.HCl pH8.0, 1 m M EDTA), phenol/chloroform (1:1) and chloroform. The phage D N A was precipitated with ethanol and then resuspended in 100 | i l of TE (lOmM 44 materials and methods Tris-HCl pH8.0, ImM EDTA). The isolated phage D N A was analyzed with restriction endonucleases. D. DNA SUBCLONING 1. Isolation and preparation of DNA fragments for subcloning D N A fragments used for subcloning experiments were produced either by restriction endonuclease digestion of plasmid D N A or by PCR. The D N A samples were separated by electrophoresis on an agarose gel (percentage of the gel depended on the size of the D N A fragment). After visualizing the D N A by U V light, the D N A fragments were cut and later recovered from agarose gels using GeneClean according to the manufacturer's instructions. Alternatively, the D N A was electroeluted from the gel slice (Davis et al., 1986) using dialysis tubing (Spectrum Medical Industries, M.W. cutoff: 10,000-12,000) and then further purified using a Nacs Prepac column (BRL) according to the manufacturer's instructions. When a blunt ended D N A fragment was needed, the isolated D N A with either 5' or 3' overhanging ends was converted to blunt ends by Klenow (large fragment of D N A polymerase I) or T4 D N A polymerase, respectively (Davis et al., 1986). 2. Ligation and transformation of DNA into bacteria For restriction endonuclease mapping, D N A sequence analysis and construction of transfection plasmids, D N A fragments were subcloned into the appropriate vectors. Ligations were carried out with 200 ng of cut vector D N A and 50 to 500 ng of insert D N A in 50 m M Tris-HCl pH 7,4, lOmM MgCl2, 10 mM DTT, 1.0 m M spermidine, 1.0 m M A T P and 100 p:g/ml BSA, and the reaction was allowed to proceed overnight at 15°C. 45 materials and methods After ligation, 2 to 10 | i l of the ligation mix was used to transform competent cells of E, coli strains JM101 or DH5a (Davis et al., 1986). Frozen competent bacterial cells were thawed just before transformation. To prepare the competent cells, the bacterial culture with an OD600 reading between 0.5 to 0.6 was harvested by centrifugation at 5000g for 10 min. The cell pellet was treated with 50 mM CaCl2 for 30 min on ice and then recentrifuged. The cell pellet was resuspended in l/20th of the original culture volume in 50 m M CaCl2 and 15% glycerol and stored at -70°C. E. RADIOACTIVE LABELING OF DNA 1. Klenow labeling Purified D N A fragment was labeled by random priming as described earlier (Feinberg and Vogelstein 1983). The D N A sample (200-500 ng in 30 u.1 of water) was initially denatured by boiling for 3 min and was rapidly cooled on ice for 5 min. The labeling reaction was performed in a 50 u.1 volume containing 50 m M Tris-HCl pH8.0, 20 UM dGTP and dTTP, 30 uCi oc 3 2P-dATP and -dCTP (speciifc activity 3000 Ci/mmol, NEN), 100 ng of random hexadeoxyribonucleotides (Pharmacia), 20 (ig of BSA and 5 units of E. coli D N A polymerase I large fragment (Klenow) (BRL or Pharmacia). The reaction was carried out at 37°C for 1-2 h and unincorporated nucleotides were removed by chromatography on a 1.0 ml column of Sephadex G-50 (Maniatis et al., 1989). Specific activities of labeled probes using this method were typically around 3 to 8 x 108 cpm/u.g. 46 materials and methods 2. Labeling with T4 polynucleotide kinase D N A fragments were labeled at the 5' ends by using T4 polynucleotide kinase and y32p-ATP. The reaction was carried out in 50 m M Tris.HCl pH7.4, 10 m M MgCl2, 5 m M DTT, 1 m M spermidine (Sigma), 0.5 Ug of purified D N A fragment, 50 (iCi y 3 2 P -ATP (speciifc activity 3000 Ci/mmol, NEN) and 15 units of T4 polynucleotide kinase (Pharmacia or BRL) for 1 h at 37°C. Unincorporated radioactive nucleotides were removed by chromatography on a 1.0 ml column of Sephadex G-50 as described earlier. The specific activity of the end-labeled probe was typically between 5 x 10^ cpm/ug. 3. Labeling by nick translation For screening the Xgtl 1 expression library, concatenated D N A fragments were labeled by the nick-translation technique (Maniatis et al., 1989). Concatenated D N A (500 ng) was used in a 50 ul reaction mix containing 50 m M Tris-HCl pH7.5, 20 u M dGTP and dTTP, 1.4 m M dCTP and dATP, 30 uCi a 3 2 P - d A T P and -dCTP (speciifc activity 3000 Ci/mmol, NEN), 5 mM MgCl2, 10 m M B-mercaptoethanol, 0.2 m M CaCl2, 2.5 ug BSA, 50 pg DNase I (Pharmacia) and 20 units of E. coli D N A polymerase I (BRL or Pharmacia). The reaction was carried out for 1 h at 15°C, and was terminated by adding 150 ul of stop mix (1% SDS, 10 m M EDTA and 25 ug tRNA). Unincorporated nucleotides were removed by chromatography on a 1.0 ml column of Sephadex G-50. The specific activity of the double stranded probes using this method was typically between 1 to 2 x 10** cpm/ug. 47 materials and methods F. OLIGODEOXYRIBONUCLEOTIDES USED IN THIS STUDY Oligodeoxyribonucleotides (oligonucleotides) were synthesized using an Applied Biosystem 380 A D N A synthesizer or the PCR-MATE (Applied Biosystems). Oligonucleotides were purified by denaturing polyacrylamide gel electrophoresis and/or reverse phase chromatography on a Sep Pak column (Waters Associates) as described earlier (Atkinson and Smith 1984). A summary of all the oligonucleotides used in this study is listed in table 1. 48 materials and methods Table I Oligodeoxyribonucleotides used in this study Oligo length name sequence (5' to 3') 1 25 PEO A T G C T C G C T G T G C A C A A G G C T A C A C 2 21 M P - 5 1 G A C C C A G A A G C T T A C A C A C T A 3 40 M P - B G G A G T A C T A G T A A C C C T G G C C C C A G T C A C G A C G T T G T A A A 4 27 M P - C CAGGAAACAGC TAT GAC CAAGCT T A C A 5 17 M P - D GGAGTACTAACCCTGGC 6 50 S M O l - 8 4 0 T A C C C T C C G C C C T A C T C C T G G A T C T A C T A G A T T T C C T G T T T G C A G T A C C C - 8 8 9 7 50 SMO 2 - 8 5 0 C C T A C T C C T G T C C C T C C C C C G A T C T A C T A G T G C A G T A C C C A A G G C A A A T A - 8 9 9 8 50 SMO 3 - 8 6 0 T C C C T C C C C C A T T T C C T G T T G A T C T A C T A G A A G G C A A A T A T T A G T C T A A G - 9 0 9 9 50 SMO 4 - 8 7 0 A T T T C C T G T T T G C A G T A C C C G A T C T A C T A G T T A G T C T A A G T A G G A C A G A G - 9 1 9 10 50 SM05 - 8 8 0 T G C A G T A C C C A A G G C A A A T A G A T C T A C T A G T A G G A C A G A G G G A C A A A G A G - 9 2 9 11 50 SMO 6 - 8 9 0 A A G G C A A A T A T T A G T C T A A G G A T C T A C T A G G G A C A A A G A G C A G G A A C A C G - 9 3 9 12 50 SMO 7 - 9 0 0 TTAGTCTAAGTAGGACAGAGGATCTACTAGCAGGAACACGGGGAGGCACA - 9 4 9 13 50 SMO 8 - 9 1 0 TAGGACAGAGGGACAAAGAGGATCTACTAGGGGAGGCACAATGTCTCGCC - 9 5 9 14 50 SMO 9 - 9 2 0 G G A C A A A G A G C A G G A A C A C G G A T C T A C T A G A T G T C T C G C C T C G C C T C A C C - 9 6 9 15 50 SMO 10 - 9 3 0 C A G G A A C A C G G G G A G G C A C A G A T C T A C T A G T C G C C T C A C C T G T C C T T T G G - 9 7 9 18 60 EN 5 ' A G C T T C C T T A G A C T A A G A T T T G C C T T G G G T A C T G C A A A C A G G A A A T G G G G G A G G G A C A G G 19 60 EN3 1 A G C T C C T G T C C C T C C C C C A T T T C C T G T T T G C A G T A C C C A A G G C A A A T A T T A G T C T A A G G A 20 20 E S P GAGAGGCTTCTGGTCCTACC 21 24 F P CGCCAGGGTTTTCCCAGTCACGAC 22 24 RP A G C G G A T A A C A A T T T C A C A C A G G A AQ materials and methods G. SITE DIRECTED MUTAGENESIS 1. Site-directed mutagenesis in M13 bacteriophage A HindUI restriction site was introduced into the 5' untranslated region of the human prothrombin gene by using the mutagenic primer 5' (Table 1:4, MP-5') and single stranded M13mpl8 template D N A purified from RZ1032 strain (dur ung') of E. coli (Kunkel et al., 1985). 2. PCR-mutagenesis: production of linker-scanning mutants Site-directed mutagenesis was performed according to the three step polymerase chain reaction (PCR)-mutagenesis method (Nelson and Long 1989) with modifications (see Fig 4). Instead of introducing a single bp mutation, 10 bp mutations were produced. Ten mutagenic oligonucleotides (SMO) were synthesized consisting of 20 complementary bases at each of the 5' and 3' ends, and 10 bases of mutagenic sequence (5' G A T C T A C T A G 3') in the middle (Table 1: 9-19, S M O l to SMO10). The Haein D N A fragment containing the human prothrombin gene enhancer (-782 to -979) was subcloned into the Sma I site of the plasmid vector pUC-18 to produce the clone pUC-18-enhancer, and this plasmid was used as the template for the PCR reactions in the first and the second steps. After the first PCR reaction, a product defined by the SMO (Table 1: 9-19, SMO1-SMO10) and the mutation primer-B (Table 1: 5, MP-B) was produced and purified from a 3% agarose gel by using GeneClean (BIO 101, Vista, CA). A single cycle of PCR was performed using the original pUC-18-enhancer as the template and the product from the first PCR as the primer. In step three, a second flanking mutation oligonucleotide C (Table 1: 6, MP-C) with a 50 materials and methods Fig. 4. Construction strategy of PCR-linker scanning mutants Construction of ten 10-bp linker scanning mutants using the three-step PCR-mutagenesis method (Nelson and Long 1989) (see methods for details). The scanning mutation clones were sequenced and then subcloned into the Hind III site of the expression vector p T K G H to produce chimeric promoter clones. 51 materials and methods -959 J L SMO-10 SMO-9 -859 I I I I I L_ human prothrombin gene enhancer region SMO-8 SMO-3 Three step PCR-mutagenesis first step (30 cycles) Mutagenic oligo pUC-18-enhancer second step (1 cycle) isolate PCR fragment pUC-18-enhancer SMO-2 SMO-1 Hindlll _J MP-B Hindlll | _ third step (30 cycles) Hindlll I MP-C Hindlll l _ MP-D T 52 materials and methods HindlTJ site at the 5' end and a shorter mutation oligonucleotide D (Table 1: 7, MP-D) spanning the unique region of MP-B were added and the final PCR was performed. The technique used is schematically represented in Fig. 4A. The PCR typically contained 100 pmoles of each primer, 1 ng of template plasmid, 200mM of dNTPs, 10% DMSO, 5 units of Taq polymerase in a buffer containing 67mM Tris.Cl pH8.0, 67mM MgS04,16.7mM (NH4)2S04 and lOmM P-mercaptoethanol. The reaction times were 30 sec at 94°C, 50°C and 72°C respectively for 30 cycles. The mutation primers MP-B, MP-C and MP-D can be used to perform site-directed mutagenesis for any small D N A fragments subcloned into cloning vectors pUC or M l 3 series. The D N A sequences of the linker-scanning mutation clones were determined using a Pharmacia Sequenase Kit and the enhancer sequencing primer (Table 1: 20, ESP). These clones were then subcloned into the Hind III site of the expression vector p T K G H to produce chimeric promoter clones. The orientations of the D N A fragments in the chimeric promoter clones were verified by restriction mapping. H. PRIMER EXTENSION ANALYSIS OF A MESSENGER RNA TRANSCRIPT The transcription start site of the human prothrombin gene was studied using the method of primer extension analysis. An oligonucleotide, complementary to the coding strand of human prothrombin gene exon I, was designed according to the D N A sequence (Table I: 3, PEO) (Degen and Davie 1987). One microgram of this primer was end-labeled by T4 polynucleotide kinase and y ^ P - A T P , and was loaded onto an 8 % denaturing polyacrylamide gel. After autoradiography, the gel slice containing the labeled primer was isolated and then incubated overnight at 37°C in 1 ml of 0.5 M NH4OAC. The primer was then purified by reverse phase chromatography using a C i 8 Sep-Pac column (Atkinson and Smith 1984). The primer (100,000 cpm) was coprecipitated with 1 pig of the human liver 53 materials and methods poly A+ R N A in 50 ul of water, 5 ul 3 M NaOAc and 125 ul 95% EtOH at -70°C for 30 min. After centrifugation for 10 min in an eppendorf centrifuge, the RNA/primer pellet was rinsed with 80% ethanol and then resuspended in 30 ul of hybridization buffer (40 m M PIPES pH6.4, 1 m M EDTA, 0.4 M NaCl and 80% formamide). The reaction mix was denatured at 90°C for 10 min and allowed to hybridize overnight at 30°C. The RNA/primer complex was reprecipitated by standard techniques and resuspended into 20 ul of reverse transcriptase buffer (50 m M Tris pH8.3, 40 m M K C l , 6 m M MgCl2,0.1 mg/ml B S A and 1 m M DTT) containing 50 uM dNTPs, 1 ug actinomycin D, 20 units RNase inhibitor (BRL) and 200 units M - M L V (Moloney Murine Leukemia Virus) reverse transcriptase (BRL). The reaction was carried out at 37°C for 2 h and the RNAs were degraded by the addition of 1 ul 0.5 M EDTA and 1 ul RNase A with incubation for 30 min at 37°C. After phenol extraction and ethanol precipitation, the primer extension products were analyzed on a 6% denaturing polyacrylamide gel. /. TISSUE CULTURE TECHNIQUES 1. Maintaining the cell-lines HepG2 (Dr. H . Pritchard, University of British Columbia), B H K (Dr. R. Palmiter, Howard Hughes Medical Institute, University of Washington) and L-cells (Dr. F. Tufaro, University of British Columbia) were used in this study. These cell-lines were maintained in complete culture medium DMEM/F12 (1:1) (Gibco) with 10% newborn calf serum (Gibco) in a 37°C humidified incubator under 5% C02- The cells were passaged using trypsin (0.2% in PBS) (Sigma) and stocks of cells were frozen in liquid nitrogen in complete culture medium containing 10% DMSO (Sigma, tissue culture grade). 54 materials and methods 2. Transfection and transient expression of hGH Transfections of HepG2, B H K and L-cells were performed according to the D E A E -dextran mediated gene transfer technique (Kaddurah et al., 1987). One million cells were seeded onto a 60 mm tissue culture dish (Falcon) one day before the transfection experiment. After overnight incubation, the cells were washed once with 2.5 ml of DMEM/F12 (1:1), and then transfected with the appropriate concentration of cesium chloride (CsCl) purified plasmid D N A in 1.5 ml of DMEM/F12 medium containing 200 Ug/ml of DEAE-dextran. After 4 h, the cells were shocked by treatment with 10% DMSO in HBS (137 m M NaCl, 5 mM K C l , 0.7 mM Na 2 HP04, 6 m M dextrose and 21 m M Hepes pH7.1) for 2 min. After rinsing the cells with 3 ml of PBS, 3 ml of complete culture medium was added. The medium was changed again two days post-transfection, and the transiently expressed hGH in the culture medium was quantified by a radioimmunoassay four days post-transfection using the hGH assay kit (Allegro) according the manufacturer's instructions. The hGH assay is based on two monoclonal antibodies specific for two different and distinct epitopes on the hGH molecule. One of the antibody is radioactively labeled with 125j and the other antibody is coupled to biotin. The reaction mixture containing the monoclonal antibodies and the tissue culture medium was incubated with an avidin coated plastic bead for 90 min at room temperature on a horizontal rotator set at 180 rpm. This allows for the specific binding of the sandwiched antibodies-hGH complex to the solid phase via the high affinity interaction between biotin and avidin. The unbounded 125 i_ a n n D oc l i e s are removed by washing the avidin-coated bead 2 times with the wash buffer (PBS with 0.02% sodium azide), and the bead is then counted in a gamma counter for one min to determine the level of hGH in the medium. 55 materials and methods J. CHARACTERIZATION OF THE PROTEIN FACTORS BINDING TO THE HUMAN PROTHROMBIN GENE ENHANCER 1. Preparation of rat liver nuclear extract Rat liver nuclear extract was prepared by a published technique (Gorski et al., 1986). The whole procedure was carried out in the cold room to minimize protein degradation. Rat liver (10 to 15 g) was homogenized using a polytron (lowest setting for 5 sec) in 80 ml of homogenization buffer (10 m M Hepes pH7.6, 25 mM K C l , 0.15 m M spermidine, 0.5 m M spermine, 1 mM EDTA, 2 M sucrose, 10% glycerol, 1 m M DTT and 0.1 m M PMSF). The homogenate (25 ml) was overlayed onto a 10 ml cushion of homogenization buffer (Beckman, polyallomer centrifuge tubes, 25x89mm) and centrifuged at 25Krpm for 30 min at 0°C using the SW 27 rotor. The nuclear pellet was combined in 50 ml of the homogenization buffer and then overlayed onto 10 ml of homogenization buffer containing an extra 10% of glycerol in two polyallomer tubes. After centrifugation using the same conditions as before, the nuclear pellet was resuspended in 20 ml of nuclear lysis buffer (10 m M Hepes pH7.6, 100 mM K C l , 3 m M MgCl2, 0.1 m M EDTA, 10% glycerol, 1 m M DTT and 0.1 mM PMSF) and 2 ml of 4 M (NH4)2S04 was added to the nuclear suspension slowly. The mixture was gently shaken for 30 min in the cold room. Cell debris and genomic D N A were pelleted by centrifugation at 35 Krpm for 30 min at 0°C using the Ti60 rotor and the nuclear proteins in the supernatant were subsequently precipitated by dissolving solid (NH4)2S04 into the solution (0.3g/ml) with gentle agitation. The nuclear proteins were collected by centrifugation using the same conditions and were resuspended in 5 ml of dialysis buffer (50 m M Tris.HCl pH7.9, 100 56 materials and methods m M K C l , 12.5 m M MgCl2, 1 m M EDTA, 20% glycerol, 1 m M DTT and 0.1 mM PMSF). The nuclear extract was dialyzed against 500 ml of dialysis buffer overnight and then loaded onto a 5 ml column of heparin agarose (Sigma). After washing the column with 10 ml of dialysis buffer, the D N A binding proteins were eluted with three 2 ml fractions of the dialysis buffer containing 1 M K C l . These fractions were pooled and dialyzed against 500 ml of the dialysis buffer overnight. The protein concentration of the rat nuclear extract was estimated by absorbance reading at 280 nm; aliquots of the extract were stored in liquid nitrogen. 2. DNase I footprinting analysis and gel retention assays DNase I footprinting analysis of the human prothrombin gene was performed as described previously (Gorski et al., 1986). A D N A fragment labeled on one strand was prepared in the following manner. The purified D N A fragment (0.5 ug) was dephosphorylated at the 5' ends by treating with calf intestinal alkaline phosphatase (Pharmacia). After phenol extraction and ethanol precipitation, the D N A was end-labeled with polynucleotide kinase and y-32p-ATP. Once again, the end-labeled D N A fragment was extracted with phenol and then precipitated with ethanol. Afterwards, the D N A was digested with a restriction endonuclease (Xbal or Avai l for the coding or non-coding strand respectively) and then separated on a 6% non-denaturing polyacrylamide gel. The D N A fragment was detected by autoradiography (5 min exposure) and the desired D N A fragment was isolated by electroelution and purified by a Nacs Prepac (BRL) column. The D N A was precipitated again and then resuspended into the appropriate amount of water (30,000 cpm/5 ul). The specifc activity of the labeled fragment is usually between 2-5 x 107cpm/ug. 57 materials and methods To perform the footprinting and gel-shift analysis, rat liver nuclear extract (25 pJ) of the appropriate concentration was added to a mixture of 10 | i l of poly dl-dC (100 Hg/ml) (Pharmacia), 10 [il of 10% P V A (BDH) and 5 ul of the probe (30,000 cpm; 0.5-2 ng/assay) at 4°C. The reaction mix was incubated on ice for 20 min and then transferred to room temperature. The following solutions were added in this order: 50 p:l of 5 m M CaCl2 and 1.5 m M EDTA for 3 min, 5 ul of DNase I (0.1 |ig/|il) (Pharmacia) for 3 min and finally 100 pi of stop mix (200 m M NaCl, 20 m M EDTA, 1% SDS and 250 u.g/ml herring sperm DNA). After phenol extraction and ethanol precipitation, the D N A was resuspended in 5 pi of formamide loading dye (80% formamide, 10 mM EDTA, 1 mg/ml xylene cyanol and bromophenol blue) and then denatured by boiling for 3 min. The samples were separated on an 8% denaturing polyacrylamide gel. After drying, the gel was subjected to autoradiography using X A R Kodak film. For gel-shift assays, after incubating the reaction mix on ice for 20 min, it was loaded onto a 4-6% non-denaturing polyacrylamide gel (see Electrophoresis of methods) which had been pre-run for 1 h at 100 V in the cold room (4°C). Separation was performed by electrophoresis at 100 V for 3 h at 4°C, and autoradiography was performed using X A R Kodak film. K. SCREENING OF A LAMBDA GT11 EXPRESSION LIBRARY The HepG2 cDNA library was purchased from Clonetech. Five hundred thousand phage clones were screened using 32p-i a D e ied concatenated double stranded D N A (see nick translation in methods section). The complementary enhancer primers (Table 1: EN5 1 and EN3') were designed to anneal with a 4 bp 5' overhang (AGCT). After annealing the primers, the resulting double stranded D N A fragments were concatenated using T4 ligase (Maniatis et al., 1989). The ligation reaction was analyzed by running 1 | ig of the D N A 58 materials and methods sample on a 2% agarose gel (data not shown). The screening procedure was performed as described (Maniatis et al., 1989) using 1 ug/ml of herring sperm D N A and 1 Ug/ml of wild type X D N A cut with Hind TU and EcoR I as non-specific competitor DNA. A positive clone was detected that bound to the enhancer fragment consistendy to the fourth screen where all the phage plaques bound specifically to the enhancer fragment. Afterwards, this positive clone (1 kbp) was subcloned into pUC 18 and then analyzed by restriction mapping and D N A sequencing. In the meantime, X lysogens of the positive clone and wild type X phage were produced in Y1089 cells and cell extracts were prepared for western blot analysis and DNase I footprint analysis (Maniatis et al., 1989). 59 Results III. RESULTS A. IDENTIFICATION OF THE HUMAN PROTHROMBIN GENE 5' FLANKING REGION A phage clone XH51A containing part of the human prothrombin gene was previously obtained in our laboratory (Dr. D. Irwin) by screening a human genomic phage library using the human prothrombin cDNA (pILH-3) (Degen et al., 1983) as a hybridization probe. The presence of the human prothrombin gene 5' flanking region in this phage clone was supported by the results of the Southern blot analysis using the most 5' region (Pstl/Hindin 0.3 kbp D N A fragment) of the human prothrombin cDNA as the hybridization probe (Fig. 5). Two EcoRI D N A restriction fragments (5 kbp and 15 kbp) were found to hybridize to the 0.3 kbp cDNA probe. The 15 kbp fragment is one of the two arms of the vector Charon 4A, therefore, it is likely to be an artefact caused by cross-hybridization due to some sequence homology. The 5 kbp EcoRI fragment was subsequently subcloned into the EcoRI site of the plasmid vector pUC18 to produce the plasmid clone pHE5B. pHE5B was characterized by restriction endonuclease mapping, Southern blot analysis (data not shown) and partial D N A sequence analysis. The locations of exon I and exon II of the human prothrombin gene were identified in this D N A fragment by comparing the D N A restriction map and the partial D N A sequences to those of the human prothrombin gene (Degen and Davie 1987). In order to subclone the 5' flanking region of the human prothrombin gene into the expression vector pOGH, an Xbal/Bgin D N A restriction fragment (1 kbp) was cloned into the phage vector M13mpl8. A KndLTJ restriction site was subsequently introduced into the 5' untranslated region of the gene by site-directed mutagenesis (see methods) using the mutagenic primer 5' (Table 1:4, MP5'). 60 Results Fig. 5. Southern blot analysis of the phage clone ?tH51A (A). The XH51A phage D N A was digested with the restriction enzyme EcoRI, electrophoresed on a 1% agarose gel and stained with EtBr. (B). The digested D N A was transferred to nitrocellulose paper and was hybridized to the 5' region of the human prothrombin cDNA (Pstl/Hindlll 0.3 kbp fragment of the plasmid clone pILH-3) labeled with 3 2 P - d A T P and ^P-dCTP using the Klenow labeling technique (see methods). Lane 1, 2 ug of the phage clone XH51A digested with the restriction enzyme EcoRI. The big arrows indicates the 5 kbp EcoRI D N A fragment hybridized to the probe. Lane 2, D N A molecular weight markers, 2 ug of the wild type X phage D N A digested with Hindlll and was end-labeled with Klenow and 32p-dATP. The size of the D N A fragments are indicated by small arrows in kbp. A B 12 12 61 Results Three kbp of the human prothrombin gene 5' flanking region was inserted into the expression vector pOGH to produce the construct pFIIGH3.0 (Fig. 6). The plasmid construction strategies of expression plasmids pFIIGH3.0, pFIIGHl.3, pFIIGH1.3R and pFJJGH0.4 are outlined in Fig. 6. The expression vector pOGH (Selden et al., 1986) is a pUC12 plasmid derivative which contains the bacterial origin of replication, the p-lactamase gene for selecting the plasmid-containing bacteria in culture medium supplemented with ampicillin, and the human growth hormone (hGH) gene as a reporter gene for the transient expression assay. B. CHARACTERIZATION OF THE HUMAN PROTHROMBIN GENE 5' FLANKING REGION Using the dideoxy chain termination method, the D N A sequence of 1258 bp of the 5' flanking region was determined and the sequence is shown in Fig. 7. By primer extension, the position of the major CAP sites were determined to be -23 and -36 with respect to the translation start site (Fig. 8). Both sites showed sequence homology to the consensus sequence 5' C A 3' where A is the +1 position (Nussinov 1990). However, these results differ from a previous report which claimed that the 5' end of prothrombin mRNA is located at nucleotide -29 (Degen 1989) with respect to the initiation codon. The reason for this discrepancy is unknown at present. The presence of more than one transcription start site can be explained by the absence of a putative T A T A box around 30 bp upstream of the CAP sites. The human prothrombin gene also lacks a prominent C C A A T box around 80 bp upstream of the CAP sites, suggesting that the gene may be regulated by other cis-acting motifs other than the TATA and C C A A T elements. 62 Results Fig . 6. Plasmid construction strategies of pFIIGH3.0 , pFI IGH1.3 , pFIIGH1.3R and p F I I G H 0 . 4 A detailed restriction map of the plasmid pHE5B is shown with the relative positions of exons I and U of the human prothrombin gene. Prior to the construction of the expression plasmids, a Hindll l site was introduced into the 5' untranslated region of exon I by site-directed mutagenesis. Initially, a Xbal/Bgin D N A fragment (1.0 kbp) was subcloned into M13mpl8, and site-directed mutagenesis was performed using the mutagenic primer MP5' (Table 1:2). The Hindlll /Bglll D N A fragment (0.4 kbp) was isolated from the mutated clone and ligated back into the original plasmid clone pHE5B with the HindlU/Bglll fragment (2.0 kbp) deleted to produce another clone pHEH3. To produce this vector for cloning, pHE5B was cut with HindHI completely and then partially digested with BglH. The D N A fragment with the appropriate size was subsequently purified from a 0.8 % agarose gel before ligating with the 0.4 kbp Hindll l /Bgll l fragment. To construct the clone pFIIGH3.0, the Hindlll/EcoRI D N A fragment of pHEH3 with 3 kbp of the 5' flanking region of the human prothrombin gene was subcloned into the expression vector pOGH. Two more clones (pFIIGH1.3 and pFIIGH0.4) were constructed by deleting restriction fragments. The plasmid pFIIGH1.3R with 1.3 kbp of 5' flanking region in reverse orientation was produced by inserting the blunt-ended HindUI/Pstl fragment into the blunt-ended vector pOGH (cut with Hindlll). The orientations of the expression plasmids are indicated by the direction of the arrows and were verified by restriction mapping analysis. 63 Results 1Kb E BglH PstI PstI pHE5B release Hindm/Bglll insert E Bgin PstI ' ' ' BglH Xbal E Hindlll Exon I II polycloning site E BglH PstI I I I Xbal/Bglll subcloned into M13 mpl8 Hindlll site introduced by site directed mutagenesis isolate Hindm/Bglll 0.4 Kb fragment pHEH3.0 Hindlll Hindlll HindlH/EcoRI fragment sucloned into the expression vector pOGH E BgUI PstI I I I PstI I PstI BgUI J Hindl l l BgUI hGH Hindlll BgUI Hindlll hGH . pFIIGH3.0 1 1 hGH 1 *• .pFIIGH1.3 .pFIIGH0.4 hGH * i l pFIIGH1.3R BgUI 64 Results Fig. 7. D N A sequence analysis of 1258 bp of the 5' flanking region of the human prothrombin gene D N A sequence analysis of 1258 bp of the 5' flanking region of the human prothrombin gene. The major transcription initiation sites at -22 and -35 relative to the translation start site are underlined with bold letters (see Fig. 8). The A T G initiation codon is indicated by bold letters. 65 Results 10 20 30 -1258 CTAATTTTTA ATTTTTTTTT GGTAGAGATG -1198 CAAACTCCTG GCCTCAAGCG ATCCTGCCAT -1138 TGTGAGCCAC TATGCCTGGC CTAAAAATAT -1078 CCAGGAATTA AGGTGTTTGC GGGAGTCCTG -1018 TCCCACACAT GACCTGGTCC AGACCCCAAA -958 GCGAGAACTT GTGCCTCCCC GTGTTCCTGC -898 ATTTGCCTTG GGTACTGCAA ACAGGAAATG -838 GGTAGGACCA GAAGCCTCTC TAGGCACATG -778 TCAGTGGAGA CCCAGGATTT CAGTAGCCCT -718 AGAAGATAGA CTAAAGGCCC GAGTCCCTGG -658 AAACACACCT CCAAGGCACC CTGGACAGGA -598 CTGAGAAGCT GGGGCAAATG TTGGCTGTTC -538 AACTTCCATC AGGCCACACC TTTTATCTTT -478 TTATACAAAC CCCACATTGA CCTATATCAG -418 ACATTCCATT CCTAATCTCC TTTAGTCCTC -358 GAGGAAACTG AGGCACACAG AGATGACAAG -298 AGCTCCAGGA AAGTCTCATA GCCCCACTGG -238 GTGGGAGATG GGGGTGACAG TGACCTTTTT -178 TCCCAGGAGG ACCTGTCCTC CCAGATGGTG -118 GTCCCCACGG CCCTGACCCT CTGACCTCAC -58 ACATTACCCA GAGGGGTCAG GACAGACAAT 40 50 60 AGGGTCTTGC TATGTTGCCC AGGCTGGTCT GTCGGCCTCC CAAAGTGTTG GGATTACAAG ATATATGAAA ATATATAAGA AATGGGCCTC GTCCCCAGTT TTTCTGCCAA CACTCCCTGT CAGCCAGGCC CAAAGGACAG GTGAGGCGAG TCTTTGTCCC TCTGTCCTAC TTAGACTAAT GGGGAGGGAC AGGAGTAGGG CGGAGGGTAG GGGCAGGCAG CCAGGGAGAA GGAGGGCCCC GTTCCGGACA GGCGCAGGTC CTGGAGCGAC ACCTGACTCC TCCCAGCAGC TGCCACACAC AGGAGGAGAA ATGGGCCCCT CCTCCAGTGG CTATCCCTGG TGCATCCCAT GGCGAGGGGC GTCTCTATTT TTGATATCTG TGTATTATGA ATCTGATTAA GAACTTACGA TATTCCATGG ACAACAAAGT ATTATTCCCA TTGTATAGAT CAACCACCGC TATATGTTAG GATTCGAAGG CCAGAATGGG CTAAATCTCA GAGGGGGAGG TGTGACTCCT CCTAGACCAT CCATCCCTGC GAGATGGACA GGAGGACTAT CTACCCACCC CCTCTCCGCT GATTTCTTCA TGTTAGTTCA TCCT£A.GTGA CCCAGGAGCT GAC AC AC T A T G 66 Results Fig. 8. Primer extension analysis to determine the transcription start site of the human prothrombin gene Primer extension analysis was performed using human liver poly A + RNA and the end-labeled oligonucleotide primer PEO (Table 1:3). The extended products, together with the sequencing ladder of the M l 3 template (GATC lanes) were analyzed by electrophoresis on an 8% denaturing polyacrylamide gel followed by autoradiography. To facilitate interpretation of the start sites, the sequencing gel is labeled such that the sequence corresponds to the complementary strand , read 3' to 5', bottom to top. The major C A P site was determined to be at -23 and a minor band was also found at -35 with respect to the translation start site (see arrows). The numbers on the left represent the distance in nucleotides upstream of the translation start site. G A T C 67 Results C. THE HUMAN GROWTH HORMONE TRANSIENT EXPRESSION SYSTEM In a preliminary series of experiments, the hGH transient expression system was characterized to optimize transfection conditions and to establish the sensitivity range of the hGH radioimmunoassays. The culture cells (HepG2, L-cells and B H K cells) were transfected with 5 Hg of the following highly purified expression plasmids, pOGH, pTKGH, pXGH5 and pFIIGH1.3 (see methods); the hGH secreted into the culture medium was used as an indirect estimation of the promoter strength of the D N A fragments 5' to the hGH reporter gene. The sensitivity range of the assay was determined to be from 1 to 100 ng hGH/ml as indicated by standard curves using commercially available hGH (Allegro) (data not shown). Transfection results indicated that the background hGH levels in the culture medium of the control plates (no D N A and 5 |ig of pOGH) were extremely low (500 cpm) (Table 2). This suggested that the medium (DMEM/F12) itself does not contain hGH immuno-reactive material and the pUC12 sequences immediately 5' of the hGH gene in the pOGH plasmid are not transcriptionally active in any of the cell-lines tested. The expression of the hGH reporter gene in the plasmids p T K G H and pXGH5 are regulated by the herpes simplex virus (HSV) thymidine kinase (TK) promoter (Seale et al., 1985) and the mouse metallothionein promoter (Denoto et al., 1981) respectively. The expression construct pFIIGH1.3 containing 1.3 kbp of human prothrombin gene 5' flanking region was used to test if the hGH transient expression system was sensitive to study the expression of human prothrombin gene. HepG2 is a human hepatoma cell-line previously shown to express prothrombin and many other liver-specific proteins (Fair and Bahnak 1984). The B H K and L-cells are hamster and mouse fibroblast cell-lines. The 68 Results promoters tested were found to behave in a cell-specific manner. The T K promoter activated transcription in the B H K and L-cells while the mouse metallothionein promoter was a strong promoter in the L-cells and HepG2 cells but was a relatively weak promoter in the B H K cells (see table 2). 1.3 kbp of the human prothrombin 5' flanking region was unable to cis-activate expression of hGH in the B H K and L-cells but the same piece of D N A was a strong promoter in the HepG2 cells, indicating that this 1.3 kbp fragment may be sufficient to confer liver specificity to the human prothrombin gene. In addition, with 5 ug of pFIIGH1.3 D N A , the levels of hGH secreted were 100 fold higher than the background levels, indicating that the assay system is highly sensitive and can be used to study human prothrombin gene expression (see Table 2). The secretion profiles of hGH from HepG2 cells over time (day 1 to day 6 post-transfection) were obtained using various concentrations of pFLTGH1.3 D N A (from 0.1 Ug to 10 ug) (Fig. 9). The cells started to secrete hGH into the medium approximately 36 h after transfection (by extrapolating the profiles of 10, 3 and 1 ug). From day 2 to day 6, there were linear increases in hGH levels in the medium, suggesting that the cells were producing hGH at a constant rate. The HepG2 cells also responded to the D N A concentrations in a dose-dependent manner with 3 ug of plasmid D N A as the optimum concentration for transfection. When less than 1 ug of D N A per plate was used, the levels of hGH secreted into the medium were not significantly higher than the background levels (pOGH and control plates with no plasmid DNA). 69 Results Table 2. Transfection results using the plasmids pOGH, p X G H 5 , p T K G H and pFIIGH1.3 Plasmids HepG2 B H K L-cells (cpm)* (cpm)* (cpm)* control 426 (27) 527(120) 575 (43) pOGH 410 (59) 414 (130) 600 (25) p T K G H 640 (95) 25519(2670) 18093(111) pXGH5 18475 (505) 2400 (276) 94705 (1604) pFIIGH1.3 46530 (1268) 876 (150) 666(100) *Each data point was an average of 4 separate readings with standard error in bracket. See methods for transfection procedures. 70 Results Fig. 9. Characterization of the human growth hormone transient expression system HepG2 cells were transfected with 0.1 p:g to 10 p.g of the plasmid pFIIGH1.3 according to the DEAE-dextran mediated gene transfer technique (see methods). The culture medium was collected from the same transfected plates at 1, 2, 3, and 6 days after transfection. The levels of hGH in the collected samples were monitored by a radioimmunoassay using the hGH assay kit according to the manufacturer's instruction (Allegro). The experimental values presented were averages of two to three readings with standard error bars. 71 levels of hGH Results in the culture medium (dpm) time after transfection (days) 72 Results D. THE EFFECTS OF PROTHROMBIN FRAGMENT 1 AND VITAMIN K ON THE EXPRESSION OF PROTHROMBIN AT THE TRANSCRIPTIONAL LEVEL The hGH transient expression system together with the plasmid pFIIGH1.3 provides an excellent system to study the regulation of human prothrombin expression at the transcriptional level. Previous studies have shown that both the activation fragment of prothrombin (Graves et al., 1981, Mitropoulos and Esnouf, 1990) and vitamin K (Karpatkin et al., 1987, Munns et al., 1976, Tollersrud et al., 1989) may participate in the regulation of prothrombin expression. Two experiments were performed in the present report to investigate the possible role of these agents in the regulation of the human prothrombin gene at the transcriptional level. In the first experiment, HepG2 cells were transfected with a constant amount of pFUGHl.5 (5 ug). Two days after transfection, exogenous bovine prothrombin fragment-1 (F- l , residues 1-156) and intermediate-I (1-1, residues 157-582) (F-l and 1-1 are generous gift from Dr. R.T.A.MacGillivray, the integrity of the polypeptides were tested by SDS-PAGE) at two concentrations (1 and 0.2 uM) were added to the medium. The effects of these polypeptides on the expression of hGH were monitored by the levels in the media at various time points after the addition of the polypeptides (Fig. 10). As indicated in Fig. 10, there was no significant difference in the hGH levels secreted from the control (1-1 and no polypeptide added) into the culture medium at different times. The concentrations 0.2 and 1.0 u M were chosen because 0.26 u M of fragment-1 was reported to be able to elevate prothrombin secretion in the original report (Graves et al., 1981). The 73 Results Fig. 10. Effects of the activation peptide (fragment-1) on the expression of human prothrombin gene HepG2 cells were transfected with 5 p.g of pFIIGH1.3. Two days after transfection, the culture medium was replaced with serum-free medium (DMEM/F12) supplemented with vitamin K (100 ng/ml) and exogenous bovine prothrombin fragment-1 (F-l) and 1-1 at two concentrations (1 and 0.2 p:M). The effects of these polypeptides on the expression of hGH were monitored by the levels of hGH secreted into the media at various time points. Each data point was an average of two readings and was normalized by comparing to the level of hGH of the control plate at 29 hours after transfection. 100 1 75 " % control hGH levels 50 " m n E3 • F - i ( i H M ) F-1(0.2 uM) 1-1(1 uM) 1-1(0.2 U.M) control 5 time (hours) 0 23 29 74 Results experimental results indicated that bovine prothrombin fragment-1 cannot activate human prothrombin gene expression at the transcriptional level in HepG2 cells. The effect of vitamin K on human prothrombin gene expression in HepG2 cells was also studied in a similar manner. The plasmid pXGH5 was used as a negative control since the metallothionein promoter was not regulated by vitamin K . A wide range of vitamin K concentrations were tested (0.1 ng/ml to 1 ug/ml) (Table 3). The hGH secreted into the medium was measured 4 days after transfection. As indicated by the results, there was no significant difference between the two plasmids and the control (no vitamin K) (table. 3). These results suggested that vitamin K does not affect prothrombin expression at the transcriptional level in HepG2 cells. E. INITIAL CHARACTERIZATION OF THE HUMAN PROTHROMBIN GENE 5' FLANKING REGION Three kbp of the human prothrombin gene 5' flanking region was analyzed in this experiment. To localize the cis-acting region essential for human prothrombin gene expression, 4 plasmids pFIIGH3.0, pFIIGH1.3, pFIIGH1.3R and pFIIGH0.4 were constructed (see Fig. 6) by linking restriction fragments of the 5' flanking region of the gene to the hGH reporter gene (Fig. 6). The hGH secretion profiles of these plasmids were compared after transfection (Fig. 11). There was no significant difference in the promoter activities of pFIIGH3.0 and pFfIGH1.3, suggesting that the region between -1.3 kbp and -3.0 kbp of the 5' flanking region was not essential to expression in HepG2 cells. The region important for expression was found to lie between -0.4 kbp and -1.3 kbp because there was a drastic reduction in promoter activity when this region was deleted. 75 Results Table 3. Effects of vitamin K on the human prothrombin gene expression HepG2 cells were transfected with 5 p.g of pFIIGH1.3 or pXGH5 . Two days after transfection, the culture medium was replaced with serum free medium (DMEM/F12) supplemented with various concentrations of vitamin K (0.1 to 1000 ng/ml). Serum-free medium was used because newborn bovine calf serum contains vitamin K which may affect the experimental results. The effects of vitamin K on the expression of hGH were monitored by the levels of hGH secreted into the medium 4 days post-transfection. The levels of secreted hGH were normalized by comparing to the control (no vitamin K). The values presented were the means of 3 experiments with SE in brackets. vitamin K concentration pFIIGH1.3 pXGH5 (ng/ml) (%) (S. E.) (%) (S. E.) 0 100 (8) 100 (10) 0.1 110(10) 107 (12) 0.3 99 (5) 89 (10) 1.0 94 (3) 111(20) 3 85 (12) 111(16) 10 84(4) 85 (8) 30 96 (8) 115(9) 100 102 (7) 94 (10) 300 93 (9) 111(9) 1000 83 (6) 94 (10) 76 Results Fig. 11. Transfection of HepG2 cells with expression constructs containing different regions of the human prothrombin 5' flanking sequence Transfection of HepG2 cells with various concentrations of the expression plasmids pFIIGH3.0{3, pFIIGH1.3 «#, pFHGH1.3R#- and pFIIGH0.4ia. The levels of hGH were measured 4 days after transfection. A l l the results were means of 2 readings and were normalized by comparing to the level of hGH using 10 ug of pFLTGH 1.3 plasmid D N A . 1 1 1 1 1 1 1 1 r D N A (Lig) 77 Results The 0.4 kbp (pFIIGH0.4) fragment itself is a weak promoter which provided 2-5% promoter activities above the background pOGH levels as compared to the promoter activity of pFIIGH1.3. The 1.3 kbp D N A fragment directed transcription only in the forward orientation since the plasmid pFIIGH1.3R did not cis-activate the hGH reporter gene. There are two explanations for these results; it is possible that the proximal 0.4 kbp region is non-essential to the expression of the prothrombin gene in HepG2 cells and the region between 0.4 kbp to 1.3 kbp alone, if in the correct orientation, is sufficient to drive the transcription. Alternatively, two elements could be present in the prothrombin gene consisting of a proximal weak promoter which is important for the unidirectional transcription, and a distal element which regulates and enhances the expression. F. FINE MAPPING OF THE HUMAN PROTHROMBIN 5' REGION To characterize the distal regulatory region, a family of deletion clones approximately 50 to 100 bp apart was constructed in the region between -0.4 kbp (pFIIGH0.4) to -1.3 kbp (pFIIGH1.3) by exonuclease III and nuclease SI digestion (see figure legend for detail) (Maniatis et al., 1989). These clones ranging in size from 0.4 kbp to 1.3 kbp, and were linked to the reporter hGH gene prior to transfection into the HepG2 cells. Results of the transient expression assays (Fig. 12) of the deletion clones indicated the presence of an upstream/enhancer regulatory region between the region -876 and -965 bp from the translation start site since the cis-activating activity dropped drastically when this region was deleted. The fluctuation of promoter activities 5' to -950 bp is probably due to the inherent errors of the assay. 78 Results Fig. 12. Fine mapping of the upstream regulatory sequence (URS) Deletion clones were constructed about 50 to 100 bp apart within the region -1.3 kbp to -0.4 kbp from the translation start site. To construct the deletion mutants, 1.3 kbp (Pstl/Hindlll fragment) of the human prothrombin gene was subcloned into PUC 19. The plasmid clone (10 ug) was initially cut with Kpnl and BamHI, and ExoIII deletion was performed as described in Maniatis et al., 1989. The 5' deletion endpoints were determined by D N A sequence analysis using the universal sequencing primer (Pharmacia Sequenase Ki t ) . The selected deletion clones were digested with Ndel and Hindin and the inserts were isolated (see methods) and then subcloned into pOGH cut with the same enzymes. Transient expression of these deletion clones were performed as described earlier. The cis-activating activity of each of the deletion clones was normalized by comparing to pFIIGH1.3. The values presented are the means of two to three independent experiments with standard error bars. 79 Results 120 -i 80 Results G. CHARACTERIZATION OF THE UPSTREAM ELEMENT To characterize the upstream/enhancer regulatory sequence (URS), the fragment containing the regulatory region (-782 to -979) was cloned into several transient expression vectors: pOGH, pFIIGH0.4, p T K G H and pXGH5, in both forward and reverse orientations to produce chimeric promoter constructs (Fig. 13). These constructs were used to transfect HepG2 cells and non-liver cell-lines such as B H K and L-cells. The results of the transfection experiments indicated that the prothrombin gene regulatory region is tissue-specific, because pFIIGH1.3 is expressed in HepG2 cells but not in B H K and L -cells (Fig. 13: 1); this is presumably because only HepG2 cells contain the appropriate trans-acting factors required for expression. In addition, the URS alone, as determined by the transient expression assays, is sufficient to account for the hepatocyte specificity of the human prothrombin gene. Deleting this URS region results in the dramatic reduction of the levels of hGH secreted into the medium in HepG2 cells (Fig. 13: 5; Fig 11; Fig. 12). In addition to that, the URS was able to confer liver-specificity to heterologous promoters. The chimeric promoter constructs pTKGHF and pTKGHR contain the HSV T K promoter and the URS in forward and reverse orientation, respectively. The presence of the URS in these two constructs resulted in a significant increase in cis-activating activity only in HepG2 cells but not in the other two cell-lines that were tested (Fig. 13: 11-13). Similarly, the chimeric construct of the URS with the metallothionein promoter (pXGH5) showed an increase of 4.4X and 2.3X for the forward and reverse orientations, respectively (Fig. 13: 8-10) only in the HepG2 cells. The enhancing effect of the URS in pXGH5 was lower 81 Results Fig. 13. Characterization of the human prothrombin gene URS The URS (-782 to -979; HaelTJ fragment) was cloned into 4 different plasmids, pOGH (cut with Hindi) , pFUGHO.4 (cut with Ndel and then blunt-ended using Klenow), p T K G H (cut with Hindll l and then blunt-ended using Klenow) and pXGH5 (cut with Hindin and then blunt-ended using Klenow), in both forward and reverse orientations (as indicated by the arrows). Five microgram of these constructs were used to transfect HepG2, B H K and L-cells. The promoter activities were normalized by comparing to p T K G H , and each value was the mean of two to four experiments. 82 CONSTRUCTS 1) • UR !>- hGH 2) 3) 4) hG] 5) 6) 7) hGr 8) 9) 10) 11) 12) 13) UR hGr RS hGl sags} h(;i[  | D ^ g ^ ~ h G T j — ^R^»«jTug— Results PROMOTER ACTIVITY HepG2 BHK L-cells pFIIGH1.3 292 0.01 0.01 pOGH 0 0 0 pOGHF 12 0.01 0.05 pOGHR 4 0.01 0.04 pFIIGH0.4 13 0.01 0.05 pFIIGH0.4F 139 0.03 0.12 pFIIGH0.4R 244 0.02 0.09 pXGH5 90 0.08 5.4 pXGH5F 404 0.14 5.2 pXGH5R 210 0.01 2.5 pTKGH 1.0 1.0 1.0 pTKGHF 260 1.9 1.5 pTKGHR 108 1.6 0.7 83 Results than in the p T K G H hybrid clones likely because the metallothionein promoter alone is a strong promoter in HepG2 cells. The URS was also found to be an enhancer element as it was capable of exerting its enhancer effect over a distance (0.2 kbp-pTKGH, 2.0 kbp-pXGH5, 0.4 kbp-pFIIGH0.4, 0.9 kbp-pFTIGH1.3) and it could enhance transcription in either orientation as shown in hybrid promoter constructs of pTKGH, pXGH5 and pFIIGH0.4 (Fig. 13: 5-13). Also, this enhancer required the presence of a proximal regulatory module to initiate transcription as shown in the constructs pOGHF and pOGHR (Fig. 13: 2-4), suggesting that the enhancer alone is insufficient to activate transcription of the gene. In summary, transfection experiments using the chimeric promoter constructs indicated that the human prothrombin gene regulation elements can be subdivided into the proximal and the upstream enhancer domains; the proximal element (-1 to -400) determines the unidirectional expression of the gene while the distal element is a liver-specific enhancer (-782 to -979). For efficient transcription initiation to occur, both the proximal promoter and the upstream enhancer contribute in a cooperative manner resulting in the tissue specific expression of the human prothrombin gene. The region between -400 to -782 seems to be non-essential to expression because deletion of this region did not affect the promoter activities (Fig. 13: 5-7) as indicated in the constructs pFIIGH0.4F and pFIIGH0.4R. This enhancer exhibits all the properties of the polymerase II enhancers. It exerts its enhancing effects irrespective of orientation and distance, but it is incapable of directing efficient transcription initiation by itself. 84 Results H. PCR-UNKER SCANNING MUTATION ANALYSIS OF THE PROTHROMBIN ENHANCER. To study the sequence requirements of the URS further, ten linker scanning mutants spanning the region from -859 to -959 were produced within the enhancer region by using a PCR- mutagenesis method (see methods and Fig. 4). These mutants were site-specific and each one of them contained the same 10 bp mutations 5' G A T C A T C T A G 3'. These mutants were subsequently linked to the thymidine kinase promoter with the human growth hormone as the reporter gene. The linker-scanning mutants were transfected into HepG2 cells to determine the effects of mutations on the enhancer activities (Fig. 14). Transfection of HepG2 cells with the scanning mutants indicated that there are two regions within the enhancer that are crucial to the enhancer activity. These elements are about 20 bp (-919 to - 938) and 50 bp in length (- 859 to -908) respectively. /. THE HUMAN PROTHROMBIN GENE ENHANCER CONTAINS MULTIPLE PROTEIN BINDING MODULES Deoxyribonuclease I (DNase I) footprinting analysis of the URS using rat liver nuclear extract (see methods) showed that there are up to three D N A binding domains interacting with the coding strand and two binding sites in the non-coding strand (Fig. 15). There were sites of enhanced DNase I sensitivity in the presence of rat liver-nuclear extract probably due to a conformational change of the D N A after interacting with the proteins. The binding reaction was performed in 1 ug of poly dl-dC and 0.1 ug of herring sperm D N A as non-specific competitor DNA. The protein binding to the human prothrombin enhancer was further characterized by gel shift assays. Using 10 ug of nuclear extracts 85 Results Fig. 14. Linker scanning mutation analysis of the region -859 to -959 bp from the translation start site (A). Analysis of the PCR products after the first step by a 3% agarose gel with DNA molecular weight standards (1 Kb ladder, BRL). The sizes of the DNA fragments ranged from 240 bp to 140 bp (10 bp difference from scanning mutants, SMI to SM10). (B). Analysis of the PCR products after step three by a 3% agarose gel. The size of the DNA fragments from SMI to SM10 were the same (260 bp) as indicated by the agarose gel. (C). The effects of the linker-mutation on the liver-specific enhancer activity was tested by the hGH transient transfection assays using HepG2 cells. The plasmids pOGH, pTKGH and pTKGHF were also used as controls in the transfection experiments. The mutants were designated as pSMl to pSMlO and they all have the same 10 bp sequence (5' GATCATCTAG 3') scanning across the region from -859 to -959. The values are the means of 2 to 3 experiments and are normalized by comparing to pTKGHF with standard error bars. 86 Results 87 Results X o CM > ° £ 8 6 <-| o 125 100 75 50 25 0 o QH 88 Results Fig. 15. DNase I footprint analysis of the human prothrombin gene enhancer DNase I footprint analysis was performed on both the coding strand (CS) and non-coding strand (NCS) of the human prothrombin gene enhancer. Probe CS extends from -979 to -833, and probe NCS extends from -782 to -979. To creat the probes required for the experiment, the Haeffl fragment spanning the region -782 to -979 was cloned into the Hindi site of PUC 19 to produce the clone PUC19-URS. This 200 bp fragment (EcoRI/HindLTJ) was isolated from PUC19-URS and then end-labeled with polynucleotide kinase and y-32p-ATP. Afterwards, the DNA was digested with Xbal or Avail to produce the CS and NCS probes, respectively (see methods for experimental details). Lanes 1 and 2 represent 5 and 10 ug of rat liver nuclear extract; lane 3, digestion of the probe with 10 Ug of BSA as control. The numbers on the left are the distance from the translation start site. The heavy lines and the numbers on the right represent the extent and positions of the protected regions. 89 Results 90 Results prepared from rat liver, kidney, brain, spleen and heart, retention bands were observed only with the liver and kidney samples (Fig. 16A). Northern analysis using total RNA prepared from the same tissues and a rat thrombin cDNA probe (a generous gift from D. Banfield, UBC) indicated that only the rat liver produced significant amount of prothrombin mRNA (Fig. 16B). The kidney nuclear protein(s) that binds to the enhancer region is probably sequence-specific since the reaction was performed with high concentrations of non-specific D N A (1 pig of poly dl-dC, 0.1 u.g of herring sperm DNA). However, the interaction of the protein with the enhancer may not play a physiological role since the rat kidney does not appear to express prothrombin in vivo as indicated by the northern blot analysis (Fig. 16B). On the other hand, the liver nuclear protein(s) that binds to the enhancer region is sequence-specific and is likely to be physiologically significant. In the competition gel-shift analysis, an excess amount of the unlabeled enhancer was able to compete off the retention band while the same of non-specific D N A did not affect the interaction of the protein to the enhancer region (Fig. 17). /. CHARACTERIZATION OF A LAMBDA GT11 EXPRESSION CLONE Five hundred thousand phages from a A-gtll HepG2 cDNA expression library were screened using the 32p.iabeled concatenated double stranded D N A as a probe (see nicktranslation in methods). One phage clone interacted with the probe consistently up to the fourth screen where all the phage plaques on the plates interact specifically to the concatenated enhancer probe. The complete D N A sequence of the 1 kbp insert was determined Q?ig. 18). When the cDNA sequence was compared to those in the E M B L 91 Fig. 16. Tissue specificity of rat prothrombin gene expression (A). Gel-shift assays to characterize the protein(s) that bind to the human prothrombin enhancer. One nanogram of the end-labeled probe (-782 to -979) was incubated with 10 ug of nuclear extracts prepared from the following rat tissues: liver, lung, kidney and spleen (see methods). One microgram of poly dl-dC and 0.1 ug of herring sperm D N A were used as the non-specific competitor D N A for the binding reaction. The binding reaction was performed on ice for 20 min before loading onto a 6% non-denaturing polyacrylamide gel. The control lanes (CONT, with no nuclear protein) indicate the size of the double stranded D N A probe on the gel. The arrow represents the retention band observed in liver and kidney samples. (B). Northern blot analysis to determine the tissue specificity of rat prothrombin mRNA. Two to 10 ug of the total RNA prepared from various rat tissues (liver, lung, kidney and spleen) using the acid guanidinium thiocyanate extraction technique (Chomczynski and Sacchi, 1987) was separated on a 1% formaldehyde agarose gel (see methods). After transferring the R N A onto the nitrocellulose paper, the blot was hybridized to the rat prothrombin cDNA probe. The sizes of the markers (RNA ladder, BRL, in kilonucleotides) were shown as small arrows on the right hand side. The big arrow on the left indicates the rat prothrombin mRNA. 92 94 Results F i g . 17. C o m p e t i t i v e g e l - s h i f t a s s a y s Competition gel-shift assays were used to determine the specificity of the binding between the enhancer binding protein and the human prothrombin enhancer. One nanogram of end-labeled probe (-782 to -979) was incubated with 10 ug of rat liver nuclear extract In addition to the 1 ug of poly dl-dC used in each reaction, competitor D N A was also added before the binding assay. Lane 2, 4, 6 contained 0.1 ug, 0.5 ug and 1 ug of non-specific herring sperm D N A respectively; lane 3,5 and 7 contained 0.1 ug, 0.5 ug, and 1 u g of cold enhancer fragment D N A respectively. Lane 1 was used as a control with no competitor D N A added. 1 2 3 4 5 6 7 95 Results D A T A library, the clone was found to be the cDNA encoding for the region between 41 bp to 1064 bp of the Y-box binding protein mRNA (YB-1) (Didier et al., 1988). YB-1 was postulated to be the regulatory factor that binds to the Y-box situated in the 5' flanking region of the major histocompatibility complex 0HLA) class II genes ODidier et al., 1988; Benoist and Mathis, 1990), To determine the tissue-specific distribution of the YB-1 mRNA, northern blot analysis was performed using the original 1 kbp fragment as a probe. The H L A YB-1 mRNA was found to be present in all the tissues tested (intestine, heart, kidney, liver, lung muscle and spleen) (Fig. 19), suggesting that the YB-1 is probably constitutively expressed in many tissues and it may be one of the house keeping trans-acting factors required for the regulation of many genes. Western blot analysis of total protein extract of the X,-lysogen produced from the positive clone indicated that a fi-galactosidase fusion protein of a molecular weight of 150 kD was synthesized in the bacteria (data not shown). However, the fusion protein was not able to produce a footprint in the DNase I footprinting experiment, suggesting that the presence of the |3-galactosidase domain (114 kD) may affect the binding of the fusion protein to the enhancer. It has been previously reported that other DNA-binding (3-galactosidase fusion proteins had the same problem in binding assays due to the presence of the bulky (3-galactosidase domain Q^ai et al., 1990). 96 Results Fig. 18. DNA and amino acid sequence analysis of the Y-box binding protein DNA and amino sequence analysis of the Y-box binding protein. The cDNA is 1512 bases in length and the open reading starts from nucleotide 127 to 1077 (Didier et al., 1988). The * symbols represent the 5' and the 3' ends of the A-gtl 1 clone (nucleotide 41 to 1064) obtained by screening the expression library. This cDNA represents almost all of the coding region with only 3 amino acids at the C-terminal missing. The ATG initiation codon and the TGA stop codon are printed in bold. The amino acids from 55 to 127 that show sequence similarity with CS7.4 (Goldstein et al., 1990) are printed in bold. The putative nuclear localization signals are underlined. 97 Results 10 20 30 40* 50 60 CCGGGAGCGGAGAGCGGACCCCAGAGAGCCCTGAGCAGCCCCACCGCCGCCGCCGGCCTA 70 80 90 100 110 120 GTTACCATCACACCCCGGGAGGAGCCGCAGCTGCCGCAGCCGGCCCCAGTCACCATCACC 130 140 150 160 170 180 GCAACCATGAGCAGCGAGGCCGAGACCCAGCAGCCGCCCGCCGCCCCCCCCGCCGCCCCC M S S E A E T Q Q P P A A P P A A P 190 200 210 220 230 240 GCCCTCAGCGCCGCCGACACCAAGCCCGGCACTACGGGCAGCGGCGCAGGGAGCGGTGGC A L S A A D T K P G T T G S G A G S G G 250 260 270 280 290 300 CCGGGCGGCCTCACATCGGCGGCGCCTGCCGGCGGGGACAAGAAGGTCATCGCAACGAAG P G G L T S A A P A G G D K K V I A T K 310 320 330 340 350 360 GTTTTGGGAACAGTAAAATGGTTCAATGTAAGGAACGGATATGGTTTCATCAACAGGAAT V L G T V K W F N V R N G Y G F I N R N 370 380 390 400 410 420 GACACCAAGGAAGATGTATTTGTACACCAGACTGCCATAAAGAAGAATAACCCCAGGAAG D T K E D V F V H Q T A I K K N N P R K 430 440 450 460 470 480 TACCTTCGCAGTGTAGGAGATGGAGAGACTGTGGAGTTTGATGTTGTTGAAGGAGAAAAG Y L R S V G D G E T V E F D V V E G E K 490 500 510 520 530 540 GGTGAGGAGGCAGCAAATGTTACAGGTCCTGGTGGTGTTCCAGTTCAAGGCAGTAAATAT G E E A A N V T G P G G V P V Q G S K Y 550 560 570 580 590 600 GCAGCAGACCGTAACCATTATAGACGCTATCCACGTCGTAGGGGTCCTCCACGCAATTAC A A D R N H Y R R Y P R R R G P P R N Y 610 620 630 640 650 660 CAGCAAAATTACCAGAATAGTGAGAGTGGGGAAAAGAACGAGGGATCGGAGAGTGCTCCC Q Q N Y Q N S E S G E K N E G S E S A P 670 680 690 700 710 720 GAAGGCCAGGCCCAACAACGCCGGCCCTACCGCAGGCGAAGGTTCCCACCTTACTACATG E G Q A Q Q R R P Y R R R R F P P Y Y M 730 740 750 760 770 780 CGGAGACCCTATGGGCGTCGACCACAGTATTCCAACCCTCCTGTGCAGGGAGAAGTGATG B R P Y G B E P Q Y S N P P V Q G E V M 98 Results 790 800 810 820 830 840 GAGGGTGCTGACAACCAGGGTGCAGGAGAACAAGGTAGACCAGTGAGGCAGAATATGTAT E G A D N Q G A G E Q G R P V R Q N M Y 850 860 870 880 890 900 CGGGGATATAGACCACGATTCCGCAGGGGCCCTCCTCGCCAAAGACAGCCTAGAGAGGAC R G Y R P R F R R G P P R Q R Q P R E D 910 920 930 940 950 960 GGCAATGAAGAAGATAAAGAAAATCAAGGAGATGAGACCCAAGGTCAGCAGCCACCTCAA G N E E D K E N Q G D E T Q G Q Q P P Q 970 980 990 1000 1010 1020 CGTCGGTACCGCCGCAACTTCAATTACCGACGCAGACGCCCAGAAAACCCTAAACCACAA R R Y R R N F N Y R R R R P E N P K P Q 1030 1040 1050 1060 * 1070 1080 GATGGCAAAGAGACAAAAGCAGCCGATCCACCAGCTGAGAATTCCCGCTCCCGAGGCTGA D G K E T K A A D P P A E N S R S R G 1090 1100 1110 1120 1130 1140 GCAGGGCGGGGCTGAGTAAATGCCGGCTTACCATCTCTACCATCATCCGGTTTAGTCATC 1150 1160 1170 1180 1190 1200 CAACAAGAAGAAATATGAAATTCCAGCAATAAGAAATGAACAAAAGATTGGAGCTGAAGA 1210 1220 1230 1240 1250 1260 CCTAAAGTACTTGCTTTTTGCCGTTTGCAACCAGATAAATAGAACTATCTGCATTATCTA 1270 1280 1290 1300 1310 1320 TGCAGCATGGGGTTTATATTTTACTAAGACGCTCTTTGGTATACAACGGTTTTAAAAGCC 1330 1340 1350 1360 1370 1380 TGGTTTTCTCAATACGCCTTAAAGGTTTTAAATTGTTTCATATCTGGTCAAGTTGAGATT 1390 1400 1410 1420 1430 1440 TTTAAGAACTTCATTTTTAATTTGTAATAAAAGTTTACAACTTGATTTTTTCAAAAAAGT 1450 1460 1470 1480 CAACAAACTGCAAGCACCTGTTAATAAAGGTCTTAAATAAT 99 Results Fig . 19. Northern blot analysis of the Y-box binding protein Northern blot analysis to determine the tissue specific expression of the Y-box binding protein. Total RNA samples (2-10 ug) prepared from various tissues (intestine, heart, kidney, liver, lung and muscle; lane 1 to 6) of the rat were separated on a 1% formaldehyde agarose gel (see methods). The 1 kbp fragment of human Y-box binding protein cDNA was labeled by the Klenow labeling method (see methods) and was used as a probe. The positions of size markers (RNA ladder, B R L , in kilonucleotides) are shown as small arrows on the right hand side. The big arrow on the left indicates the rat Y-box binding protein mRNA hybridized to the probe. 1 2 3 4 5 6 * ~ 100 Discussion I V . D I S C U S S I O N A. CHARACTERIZATION OF THE HUMAN PROTHROMBIN GENE 5' FLANKING REGION To study the regulation of the human prothrombin gene, a phage clone XH51A containing 3 kbp of the 5' flanking region was characterized by Southern blot analysis (Fig. 5), restriction endonuclease mapping analysis (data not shown) and partial D N A sequence analysis. After introducing the restriction site Hindll l into the 5' untranslated region of the gene, 3.0 kbp of the 5' flanking region was subcloned and used to construct expression plasmids with the hGH as the reporter gene (Fig. 6). The complete D N A sequence analysis of 1.3 kbp of the 5' flanking region was performed (Fig. 7), and the transcriptional start sites of the human prothrombin gene were determined by primer extension analysis using human liver polyA"1" RNA (Fig. 8). By primer extension, the major transcriptional start sites of the human prothrombin gene were determined to be at -23 and -36 with respect to the A T G translation initiation codon (Fig. 8). As reported previously (Degen and Davie 1987), the 5' flanking sequence does not contain a typical C C A A T motif (5' G G C C A A T C T 3' consensus) (Nussinov, 1990, Wasylyk, 1988) within 200 bp of the transcription initiation site, nor was a typical T A T A sequence present (TATAAA consensus) (Nussinov, 1990, Wasylyk, 1988) within 50 bp of the C A P site. The primer extension results are consistent with a heterogeneous start site for transcription initiation in genes that lack a T A T A motif. Multiple C A P sites are also observed in many another TATA-less liver-specific proteins including factor XII (Cool and MacGillivray 1987), factor VUI (Gitschier et al., 1984), factor IX (Yoshitake et al., 1985) and protein C (Plutzky et al., 1986). The D N A sequences of both of the C A P sites 101 Discussion at positions -23 and -36 are 5' C A 3' where A is the +1 position (Fig. 7). This consensus sequence is also found in many other genes (Wasylyk 1988). The transcription initiation of the human prothrombin gene may be TATA-box independent and it may be determined by initiator elements such as the CA motif around the CAP sites. There is a more A T rich region at -52 to -57 (5' A C A T T A 3') (Fig. 7) which is about 30 bp upstream of the major C A P site at -23. The position of this A T rich region may explain the slightly higher frequency of transcription initiation at -23 when compared to -36 as indicated by the intensity of the primer extension products (Fig. 7). There is a previous report claiming that the 5' end of prothrombin mRNA is located at nucleotide -29 (Degen, 1989) with respect to the initiation codon. The D N A sequence around -29 is 5' A A T T 3' where the first T is the +1 position. The discrepancy of the data is unclear but since the human prothrombin gene does not possess an obvious T A T A box, one would expect the C A P site is more dependent upon the initiator C A motif. It is therefore less likely that transcription initiation will start at sequences that do not follow this consensus. B. CHARACTERIZATION OF THE HUMAN GROWTH HORMONE TRANSIENT EXPRESSION SYSTEM The development of techniques for transfecting D N A into the mammalian culture cells has provided a powerful tool to study transcriptional regulatory elements in eukaryotes. The mechanism of DNA-mediated gene transfer into mammalian cells is still unclear. It is believed that the D N A is protected from degradation by complexing with facilitators of D N A transfer such as DEAE-dextran and calcium phosphate since naked D N A is degraded rapidly in culture medium containing serum (Kucherlapati and Skoultchi, 1983). These complexes, after being deposited onto the culture cells, are subsequently 102 Discussion taken up by the cells and then translocated into the nucleus. The fate of the foreign D N A in the cells is dependent on the vector used. Vectors containing the SV40 origin of replication multiply autonomously in the nucleus of COS cells which produce the T-antigen in a constitutive manner (Gluzman, 1982). Transient expression vectors, however, do not contain the viral origin of replication and hence do not replicate. The plasmid D N A is integrated into the genome at random sites with multiple copies (Robins et al., 1981). There is also evidence suggesting that the plasmid D N A can remain stable in the nucleus for a short period of time thus allowing expression to take place (Kucherlapati and Skoultchi 1983). The expression of the non-integrated D N A decreases over time due to degradation of the plasmid DNA. Because of the rapidity and the ease of the transient expression systems, these experimental systems are used extensively to analyze cis-acting D N A elements required for gene expression. The most common techniques used for transient expression are the calcium phosphate coprecipitation method (Graham and Van der Eb 1973) and the DEAE-dextran method (Kaddurah et al., 1987, McCutchan and Pagana 1968). However, the transformation efficiencies obtained by using these transfection methods are generally low (Chen and Okayama 1987). Modifications of the protocol by using chloroquine, butyrate, D M S O or glycerol shock were developed to increase the transfection efficiency. Other transfection methods such as electroporation (Neumann et al., 1982) and microinjection (Anderson et al., 1980) have also been employed with great success. The DEAE-dextran method was used in this study because it is highly reproducible and it is relatively easy to perform so that a large number of samples could be handled at the same time. However, no one has successfully obtained a permanent cell-line by the DEAE-dextran method of transfection suggesting that the cells are not able to divide after DEAE-dextran treatment. 103 Discussion In the present study, the choramphenicol acetyl transferase (CAT) transient expression system was initially used to study human prothrombin gene expression. Although this C A T system works extremely well for strong promoters such as the RSV long terminal repeat, it was found that the sensitivity of the system was too low to detect the relatively weaker promoter activity of the human prothrombin gene 5' flanking region in a consistent manner (data not shown). As an alternative, the hGH transient expression system was used because of its high sensitivity and reproducibility. Preliminary transfection studies using the constructs pOGH, pTKGH, pXGH5 and pFIIGH1.3 indicated that the hGH expression system was sufficiendy sensitive to detect the levels of hGH cis-activated by 1.3 kbp of the human prothrombin gene 5' flanking region (Table 2). The high sensitivity of the experimental system allowed the detailed mapping of the promoter of the human prothrombin gene 5' flanking region later on in this study. The expression system was further characterized by examining the secretion profiles of hGH into the culture medium from the HepG2 cells after transfection using the expression plasmid pFIIGH1.3 (Fig. 9A). The linear increase of the hGH in the medium from day 2 to day 6 post-transfection indicated that the D N A taken up into the cells was relatively stable in this period of time. These results suggested that the extra-chromosomal D N A can either survive for a relatively long time (up to 6 days) or the expression of hGH was only driven by the integrated DNA. The HepG2 cell number remained relatively constant after transfection indicating that DEAE-dextran treatment may affect the cell division mechanisms. The observed constant rate of hGH expression from day 2 to day 6 is probably due to a combination of factors including high D N A stability and unchanged cell number after transfection. Interestingly, for the CAT transient expression system, the time for assaying the expressed products was recommended to be 2 days after transfection. 104 Discussion As determined by the hGH secretion profiles, day 2 is the time when the HepG2 cells started to secrete hGH. This may be the reason for the low levels of C A T detected previously using 3 kbp of the human prothrombin 5' flanking region. The hGH system is highly versatile because the cells are not disrupted in order to assay for the expression products. This allows continuous monitoring of hGH secretion and the gradient of the secretion profile over time is the best representation of the promoter strength of the construct. Since the secretion profile is linear, hGH assays can be performed at day 6 after transfection to obtain detectable signals even for weak promoters. C. THE EFFECTS OF PROTHROMBIN FRAGMENT 1 AND VITAMIN K ON THE EXPRESSION OF PROTHROMBIN AT THE TRANSCRIPTIONAL LEVEL The activation fragment of prothrombin has been postulated to auto-regulate the expression of prothrombin gene itself and other vitamin K dependent proteins (Graves et al., 1981, Graves et al., 1982). Addition of bovine prothrombin fragment-1 (0.26 | iM) into the rat hepatoma H-35 cell culture resulted in the elevation of the endogenous rat prothrombin synthesis and secretion 2- and 3-fold, respectively. The stimulatory effects of the prothrombin fragment-1 seemed to reside on the Gla domain since the proteolytic fragment containing the Gla domain (residues 1-42) also induced secretion of endogenous prothrombin in this system (Graves et al., 1981). Similar induction results of prothrombin fragment-1 on factor X were reported by the same group later (Graves et al., 1982), postulating that prothrombin fragment-1 may also regulate the expression of other vitamin K dependent proteins in the same experimental system. Although the hypothesis of auto-regulation of vitamin K dependent protein expression by prothrombin activation product is very attractive, there is no further convincing evidence to support this hypothesis. 105 Discussion In this report, the effect of bovine prothrombin fragment-1 on the human prothrombin expression in HepG2 cells at the transcriptional level was investigated. It was found that the prothrombin fragment-1 was unable to induce the production of the reporter protein using 1.3 kbp of the human prothrombin gene 5' flanking region as the promoter (Fig. 10). This 1.3 kbp region has been shown to be sufficient to promote liver-specific expression earlier. If fragment-1 plays a role in the regulation of human prothrombin expression, the site of action is likely to be post-transcriptional. As indicated by previous studies, vitamin K stabilizes the vitamin K dependent proteins intracellularly by promoting proper y-carboxylation and glycosylation (Munns et al., 1976, Karpatkin et al., 1987, Tollersrud et al., 1989). The elevated levels of the vitamin K dependent proteins after the addition of vitamin K is more likely due to the increase in the stability of the proteins rather than due to the augmentation of the synthesis rate. The possible effect of vitamin K on prothrombin expression at the transcriptional level was studied using the hGH transient expression system. The experimental findings (Table 3) agree with the previous studies; vitamin K does not regulate prothrombin expression at the transcriptional level (Karpatkin et al., 1987, Tollersrud et al., 1989). D. FUNCTIONAL CHARACTERIZATION OF THE 3 KBP 5' FLANKING REGION Initial functional characterization of 3 kbp of the human prothrombin gene was carried out by linking restriction fragments to the reporter gene hGH (Fig. 6) prior to transfection into HepG2 cells. A proximal element (0 to 0.4 kbp) and an upstream element (0.4 to 1.3 kbp) were found. In order to locate the upstream element, deletion mutants within the region between -0.4 kbp and -1.3 kbp were constructed. Using the transient expression system, the upstream regulatory region was determined to be between -850 and 106 Discussion -950 bp from the translation start site (Fig. 12). The regions from -940 bp to -3.0 kbp and from -435 bp to -782 bp (see below, Fig. 13: 5-7) seemed to be functionally non-essential to the human prothrombin gene expression as deletion of these regions did not change the expression of the reporter gene, suggesting neither a positive nor a negative element was present within this region. The experimental approach used was efficient in finding the location of the most 5* region essential for transcription initiation (between -950 to -850 in this case). However, little information was obtained in the region 3' to the upstream regulatory sequence because all the deletion clones showed the same low promoter activities once the most distal regulatory region has been deleted. Functional analysis of cis-acting regulatory elements is more reliable than binding studies such as gel shift assays or DNase I footprint assays since the binding of nuclear proteins to the promoter does not necessarily result in the regulation of transcription. More detailed characterization of the distal element was performed by constructing chimeric promoter plasmids and by protein binding assays such as gel-shift assays and DNase I footprint assays. E. CHARACTERIZATION OF THE UPSTREAM ENHANCER ELEMENT To study the properties of the region between -850 to -950 in the tissue specificity and/or enhancer activity, chimeric constructs were made and then transfected into HepG2, B H K and L-cells. Several conclusions were obtained from the transfection results (see results section H for details). Firstly, the upstream regulatory sequence (URS) alone is sufficient to account for the hepatocyte specificity of the prothrombin gene in the HepG2 cells. Secondly, the URS is an enhancer element since it is position and orientation independent. Finally, the liver-specific expression of human prothrombin gene is regulated by at least two elements which function in a cooperative manner: the upstream enhancer 107 Discussion confers liver-specificity of the gene expression while the proximal element determines the direction and the position of transcription initiation. F. PCR-LINKER SCANNING MUTATION ANALYSIS OF THE PROTHROMBIN ENHANCER. Linker scanning mutation analysis is one of the best ways to study the functional properties of the potential regulatory modules in a relatively large element. The idea of linker scanning mutation study is to alter systematically the local environment across a regulatory element. Because both the D N A sequence and the spacing between the D N A motifs in the rest of the regulatory element remain unchanged, comparison of the functional activities of these mutants gives a better insight to the understanding of the cis-acting motifs required for enhancer activity. Ten linker scanning mutants each with a 10 bp mutation (5' G A T C A T C T A G 3') were produced across the enhancer region by a novel approach using a PCR-based mutagenesis technique (see methods and Fig. 4). The strategy used was efficient in producing a family of site-specific linker mutants in a short period of time. The effects of mutations on the enhancer activities in HepG2 and B H K cells was subsequently investigated by the hGH transient expression system (Fig. 14). The transfection results of HepG2 cells indicated that the human prothrombin enhancer consists of multiple cis-acting modules. Cis-acting D N A motifs are usually small (within 10 bp) and with discrete D N A sequences (Wasylyk, 1988). The region involved in the prothrombin enhancer activity is non-repetitive and is over 80 bp in size. The relatively large size of this element strongly suggests the presence of more than one regulatory module. There are two regions within the enhancer which regulate the enhancer activity since mutations of either region resulted in low levels of expression. This indicated that 108 Discussion both regions consisted of positive motifs and they functioned in a cooperative manner in order to confer liver-specific enhancer activity. Scanning mutations did not affect the T K promoter's activity in the B H K cells. This suggests that the lack of enhancer activity in this non-liver cell-line was not a result of the presence of a negative cis-acting module. It is more likely that the B H K cells do not express a positive trans-acting factor(s) required for the enhancer activity. Mutation of the putative HNF-1 motif (pSM-8) led to the total loss of the enhancer activity, indicating that the putative HNF-1 motif is responsible for the liver-specific regulation of the human prothrombin gene in HepG2 cells. G. DNASE I FOOTPRINT ANALYSIS OF THE HUMAN PROTHROMBIN GENE ENHANCER DNase I footprinting analysis (Fig. 15) and gel retention assays (Fig. 16A, 17) of the URS showed that there are extensive interactions between the enhancer and specific proteins in the liver nuclear extracts. As summarized in Fig. 19A, the results of the footprinting analysis and scanning mutations are in agreement with each other. The human prothrombin gene enhancer is 80 bp in size (-860 to -940) including a putative HNF-1 binding site and is flanked by a G-C rich motif. The experimental findings of the deletion mutations locates the most 5' region of the enhancer which is important for the enhancer activity . Deletion of this 5' end module resulted in the drastic reduction of activity. The boundaries of the DNase I protection regions coincided with the regions detected by the scanning mutation studies. In addition, the 10 bp region (Fig. 14:pSM6) which was not important for the enhancer activity was not protected as shown in DNase I footprinting study (Fig. 15), indicating the sensitivity of both assays are similar. 109 Discussion Gel-retention assays using crude nuclear extracts from various rat tissues showed that the the trans-acting factors that bind to the enhancer region were tissue specific, suggesting that the appropriate trans-acting factors for prothrombin gene expression are only synthesized in the liver. In addition, as indicated by the competition gel-shift assays, the binding of the protein(s) to the enhancer fragment was D N A sequence-specific because only the unlabeled enhancer fragment was able to compete off the retention band. Since this enhancer region alone is sufficient to drive liver-specific expression of prothrombin gene (as discussed earlier), the interactions of these liver-specific protein(s) with the enhancer is likely to confer liver-specificity of the rat prothrombin gene expression in vivo as indicated by the northern blot analysis (Fig. 16B). H. ANALYSIS OF THE ENHANCER MOTIFS DNase I footprinting analysis of the human prothrombin gene enhancer showed that the element is flanked by an inverted repeat, 5' CCTCCC 3', also found in the H B V enhancer core as a direct repeat (Honigwachs et al., 1989, Yee 1989). Interestingly, only the 3' boundary G-C rich motif is important to the enhancer activity as mutation of this region resulted in the total loss of enhancer activity; in contrast, mutation of the 5' G-C rich motif did not affect enhancer activity (Fig. 14, p S M l , pSM9). These results suggest that the protein that recognizes the G-C rich motif at the 3' boundary interacts with other trans-acting factor(s) that bind to a D N A motif(s) in proximity to enhance promoter activity. In addition, when the sequence of the 1.3 kbp 5' flanking region was analyzed, three other C C T C C C sequences were found at -680, -1080 and -1160. Deletion of these regions did not affect the cis-acting activity (for the -680 region, see Fig. 13:5-7; for the regions -1080 and -1160, see Fig. 12), further supporting the idea that this motif is important only in the 110 Discussion presence of other cis-acting module(s) in proximity. Similar observations were found in the study of the T K promoter (McKnight et al., 1984). The second distal Spl motif was shown to be phenotypically more important than the first distal Spl motif. This suggests that the presence of the consensus motifs in a piece of D N A alone is not sufficient to promote transcription. It is the interaction of the trans-acting proteins in a specific three-dimensional spatial conformation that allows transcription to take place. This is reasonable because an Spl motif occurs randomly every 2 Kbp (2 orientations in 4^ bps) but only a very Umited number of these Spl motifs are transcriptionally active. When the DNase I footprints of several liver-specific promoters are compared, the 3' boundaries of the protected site of some other liver specific promoters such as the a-1 antitrypsin (Hardon et al., 1988) and fibrinogen (3-chain (Courtois et al., 1987) promoters are also flanked by this G-C rich motif (Fig. 20C). This G-C rich motif, also shares sequence identity with the Spl binding site (5' C C G C C C 3') (McKnight et al., 1984), suggesting that this motif might be a variant of the Spl binding protein and might represent a new class of liver-specific cis-acting regulatory motif. The human prothrombin gene enhancer also contains a putative HNF-1 motif (Sawadaishi et al., 1988, Courtois et al., 1988, Kugler et al., 1988). Although there is only moderate homology with the human prothrombin enhancer to the HNF-1 consensus sequence (seven out of twelve) (Fig. 20B), an 8 and 10 bp sequence identity in the opposite orientation were found between the enhancer and the human and rat fibrinogen (3-chain HNF-1 motif respectively (Courtois et al., 1987) (Fig. 20B). Linker scanning mutation analysis of the putative HNF-1 motif in the human prothrombin gene enhancer indicated that it is essential to the enhancer activity (Fig. 14). The prothrombin gene represents the first case where a putative HNF-1 motif is situated greater than 150 bp from 111 Discussion Fig. 20. Summary of the protected sequences of the enhancer (A) D N A sequence of the human prothrombin gene enhancer. The arrow represents the G-C rich motif CCTCCC which flanks the enhancer core. The underlined -sequence indicates the putative HNF-1 motif. The black boxes represent the protected regions as determined by the DNase I footprint analysis. (B) Comparison of the HNF-1 motifs in the human prothrombin enhancer, the rat and the human fibrinogen (3-chain promoters with the HNF-1 consensus. The boxed nucleotide represents the 8 bp sequence identity found in the human prothrombin gene enhancer, the rat and the human fibrinogen fj-chain promoters. (C) Summary of the DNase I footprint analyses of the human a-1 antitrypsin and rat fibrinogen p-chain promoters (Courtois et al., 1987, Hardon et al., 1988). The arrow represents the G-C rich motif which flanks the human prothrombin enhancer and the underlined regions are the A-T rich HNF-1 motifs. 112 - 9 5 0 •930 Discussion - 910 G C G A G A A C T T G T G C C T C C C C G T G T T C C T G C T C T T T G T C C C T C T G T C C T A C CGCTCTTGAACACGGAGGGGCACAAGGACGAGAAACAGGGAGACAGGATG - 890 - 870 TTAGACTAATATTTGCCTTGGGTACTGCAAACAGGAAATGGGGGAGGGA A A T C T G A T T A T A A A C G G A A C C C A T G A C G T T T G T C C T T T A C C C . C C T C C C T G T T 7 A A T N A T T A A C G T C A A A T A T T A A . C A T T AAATATTA7AC G G C A A A T A T T A G T HNF 1 consensus rat fibrinogen beta-chain human fibrinogen beta-chain human prothrombin enhancer -90 r - 5 1 C T G G G G T G A C C T T G G T T A A T A T T C A C C A G C A G C C T C C C C C GACCCCACTGGAACCAATTATAAGTGGTCGTCGGAGGGGG I I - 1 1 6 human alpha-1 antitrypsin - 7 0 T G A A C C A A A C T G T C A A A T A T T A A C T A A A G G G A G G T A A A C T T G G T T T G A C A G T T T A T A A T T G A T T T C C C T C C A T T rat fibrinogen beta-chain 113 Discussion the transcription start site. The a- and P chain fibrinogen genes in the family of the blood clotting factors are also regulated by HNF-1. In addition, HNF-1 also interacts with many other liver-specific genes including albumin, pre-albumin, al-antitrypsin, a fetoprotein and transthyretin (Sawadaishi et al., 1988, Courtois et al., 1988). The fact that the HNF-1 is capable of recognizing D N A sequences of fairly significant divergence further suggests that the HNF-1 might interact with other trans-acting factor(s) for promoter/enhancer recognition. The G-C rich motif, CCTCCC, is a possible candidate. This hypothesis is further supported by the presence of the same combination of an A-T rich (HNF-1) and a G-C rich motif observed in human fibrinogen and a l - antitrypsin promoters (Fig. 20C). The two trans-acting D N A binding proteins might interact in a cell-specific manner in hepatocytes to enhance transcription. Binding site promiscuity is also found in many other transcription factors including the octamer binding protein (OTF), the C/EBP and homeodomain proteins (for review, see Johnson and McKnight, 1989). The low stringency in D N A motif recognition of some of the D N A binding proteins means the protein factor can interact with a whole family of D N A sequences. In a multi-protein D N A complex, the protein-protein interactions may play a much more significant role in the formation of a stable complex. /. CHARACTERIZATION OF THE Y-BOX BINDING PROTEIN The cDNA encoding for the YB-1 protein was isolated from a X.gtl 1 expression library using the concatenated doubled human prothrombin gene enhancer probe. This Y B -1 cDNA was initially isolated and characterized using the same technique except that the probe used was the Y-box of the major histocompatibility complex (HLA) class II promoter (Didier et al., 1988). Interestingly, when the D N A sequence of the probes used in both 114 Discussion studies are compared, a seven bp sequence similarity with a single bp mismatch is found (see Fig. 21). This strongly suggests that the binding between the YB-1 and the human prothrombin gene enhancer is sequence specific. However, the physiological importance of the YB-1 protein on the regulation of both of the H L A class II genes and human prothrombin gene remains to be investigated. YB-1 is 317 amino-acids long, 35 kD in size comprised of 18% basic residues. This protein also contains at least six putative nuclear localization signals which have 3-6 characteristic basic residues in a cluster. Computer analysis of the amino-acid sequence indicated that the protein does not possess any known D N A binding motif. However, a conserved region of the 70 amino acids has been found to show striking similarity with three other proteins: CS7.4, DbpA and DbpB (see Fig. 18) (44 % amino acid sequence identity; 70 % amino acid sequence similarity including conservative changes) (Wistow, 1990; Goldstein et al., 1990; Sakura et al., 1988). CS7.4 is a protein produced in E. coli induced by low temperature (cold-shock protein) and the function of this protein is still unclear (Goldstein et al., 1990). DbpA and DbpB are human DNA-binding proteins of unknown specificity (Sakura et al., 1988). This stretch of amino acid may represent a new class of DNA-binding or protein-protein interacting motif which is found in both prokaryotes and eukaryotes (Wistow, 1990). The expression of H L A class II genes is highly controlled and is restricted to cells in the immune system including macrophages, mature B cells and the thymic epithelium Q?lavell et al., 1985). These genes are regulated by several highly conserved cis-acting elements including the octamer binding motif, the T A T A box, and the X , Y , Z boxes. The T A T A box and the X and Y boxes have been shown to be transcriptionally essential in transfection assays and transgenic models (for review, see Benoist and Mathis 1990). The 115 Discussion Fig. 21. Analysis of the Y-box binding motif in the human prothrombin gene enhancer Comparison of the D N A sequence of the human prothrombin gene enhancer HNF-1 motif with the CCAAT-box motif and the YB-1 motif. The HNF-1 motif is underlined and is overlapped with the putative YB-1 binding site as indicated. The bold letters represent the non-consensus D N A sequences in both the human prothrombin gene enhancer and the YB-1 motif. AGATTGGCC CCAAT-box b i n d i n g m o t i f TGATTGGCC Y-box b i n d i n g m o t i f GACTAATATTTGCCTT human prothrombin enhancer HNF1 mo t i f 116 Discussion Y-box of these H L A class II genes contains an inverted C C A A T motif and YB-1 is postulated to be the protein binding to this box (Didier et al., 1988). Purification of the protein(s) by D N A sequence-specific affinity column indicated there are two different proteins interacting with the Y-box of the H L A class II promoter. The two proteins are 35 kD and 45 kD in size and both of them are required for the efficient binding to the D N A (Celada and Maki 1989). The 35 kD protein is likely to be YB-1 but further experimentation is required to confirm this observation. In addition to the YB-1 , several C C A A T binding proteins have been shown to form homodimer and/or heterodimer complex (Chodosh et al., 1988, Descombes et al., 1991). YB-1 may interact with other C C A A T binding proteins and/or tissue-specific proteins to enhance expression. The tissue-specific distribution of these C C A A T binding proteins may lead to the specific interactions among these C C A A T binding proteins and other trans-acting factors resulting in the tissue-specific regulation of the genes. However, when the tissue-specific expression of the Y B -1 was studied by Northern blot analysis, it was found that the mRNA encoding the protein was present in all the rat tissues tested including intestine, heart, kidney, liver, lung, muscle and spleen. Because of the wide tissue distribution of YB-1 mRNA, YB-1 is likely to be one of the house keeping trans-acting factors expressed in most cell types. As an alternative explanation, YB-1 may be responsible for tissue-specific expression of the human prothrombin gene by acting as a negative transcription factor in other cell-types. This is possible because YB-1 was postulated to be a negative factor in the regulation of the H L A class II genes since the production of the YB-1 mRNA was inversely related to the expression of these genes (Didier et al., 1988). Since both putative HNF-1 and YB-1 modules overlap in the prothrombin enhancer (Fig. 21), HNF-1 may out-compete the YB-1 in the liver due to higher concentrations and/or higher affinity. A similar situation was 117 Discussion observed in another liver-specific trans-acting factor, DBP (see Introduction). DBP competes with C/EBP to bind to the D-site of the albumin promoter (Lichtsteiner et al., 1987). Similarly, the mRNA of DBP is also found in many other tissues. It is possible that DBP is actually responsible for the inhibition of albumin transcription in other tissues rather than for the activation of transcription of the albumin gene in the liver. The high concentration of C/EBP in hepatocytes may out-compete DBP resulting in the expression of albumin in the liver. Interestingly, both the DBP and the YB-1 recognize the same C C A A T motif. The expression pattern of these C C A A T box binding proteins in different tissues is likely to be crucial to the tissue-specific expression of many genes. They can act as both stimulatory or inhibitory factors depending on the tissue type. YB-1 and DBP are probably examples of the negative factors while C/EBP and L A P are positive transcription factors in the liver. In fact, studies of the glucocorticoid receptor have shown that the same protein factor can be both stimulatory and inhibitory in the same cell-type and even in the same promoter. There are seven glucocorticoid receptor binding motifs present in the promoter region of the prolactin gene, and the glucocorticoid receptor can act as both positive and negative regulator depending on the particular site that the protein interacts with (Sakai et al., 1988). The negative regulatory effect of the glucocorticoid receptor has been postulated to be caused by the blocking of the entry of other trans-activators (Miesfeld, 1987). In addition, the negative D N A elements bear minimal resemblance to the consensus sequence (Sakai et al., 1988). These two observations together suggest that the expression of the prolactin gene is highly influenced by the concentrations of activator proteins. The interaction of the glucocorticoid receptor and the negative sites are likely to be weak because of the divergence of D N A sequences and a slight change in trans-activator level will be reflected by the rate of prolactin expression. The same strategy of regulation may 118 Discussion also be used by YB-1 and DBP on the expression of the human prothrombin and albumin genes. However, further experimentation is required to demonstrate that YB-1 is a physiological regulator of human prothrombin gene expression. The discovery of the interaction of YB-1 with the human prothrombin gene enhancer has raised the attention of two interesting observations also found in many regulatory elements. The first one is the presence of multiple proteins recognizing a common D N A sequence and the second observation is the DNA-binding promiscuity of protein factors (Johnson and McKnight, 1989). In addition to the CCAAT-box binding proteins and the HNF-3A protein family (see Introduction), the octamer binding proteins (OTFs) represent another example of factors that interacts with a unique D N A motif, the octamer element (5' A T G C A A A T 3'). At least two of the OTFs had been identified and sequenced (OTF1 and OTF2) (Fletcher et al., 1987; Scheidereit et al., 1987; Clerc et al., 1988; Sturm et al., 1988). OTF1 is ubiquitous and has been shown to activate the histone H2B promoter (Fletcher et al., 1987) while OTF2 is lymphoid-cell specific and is capable of trans-activating an immunoglobulin promoter (Scheidereit et al., 1987). These two proteins share extensive sequence homology in the D N A binding H T H motifs which explains why they recognize the same D N A sequence (Clerc et al., 1988; Sturm et al., 1988). The reason for the presence of multiple protein factors that bind to a unique site is not clear. However, it is either that stable protein-DNA interactions require the presence of secondary D N A sub-sequence and/or protein-protein interactions may play a more important role in the binding specificity of the proteins. The tissue specific distributions and the competition of these proteins for the binding sites in various tissues are probably important mechanisms for tissue-specific regulation. 119 Discussion Another phenomenon of DNA-protein interaction is the relaxed binding capability of some protein factors to a family of sites with significant D N A sequence divergence. This includes C/EBP, HNF-1, OTFs, glucocorticoid receptors and many other factors (Graves et al., 1986, Johnson et al., 1987, Sawadaishi et al., 1988, Courtois et al., 1988, Kugler et al., 1988, Sakai et al., 1988). This property of D N A binding proteins again indicates that increased stability of the DNA-protein(s) complex can be achieved through protein-protein interactions, i.e. the binding of the first protein to the D N A recruits other protein factors to enhance transcription initiation. /. PUTATIVE MODEL FOR HUMAN PROTHROMBIN GENE EXPRESSION IN THE LIVER The expression of the human prothrombin gene is controlled by a proximal promoter and a distal enhancer element. The proximal promoter element determines the direction of transcription initiation and the CAP sites. Almost all proximal elements interact with general transcription factors including TFIIA, TFIIB, TFHD, TFIIE, TFITF and R N A polymerase JJ (Wasylyk, 1988). This promoter is crucial to expression because the human prothrombin gene enhancer alone is incapable of activating transcription efficiendy. The enhancer, on the other hand, directs the liver-specific expression of the human prothrombin gene. The enhancer element is about 80 bp in length consisting of multiple cis-acting regulatory modules. Linker scanning mutation analysis of the element indicated that two regions (-938 to -919; -908 to -859) are essential for the enhancer activity. As determined by DNase I footprint analysis, there are at least three proteins interacting with the enhancer element including HNF-1, CCTCCC (FIIB) and FIIA. A schematic diagram summarizing 120 Discussion the possible interactions of these motifs is shown in Fig. 22. The YB-1 motif, overlaps with the HNF-1 motif, is likely to be important to the liver-specificity of the human prothrombin expression (discussed above). K. FUTURE STUDIES To study the role that HNF-1 and YB-1 play in the regulation of the human prothrombin gene, several experiments could be performed. Initially, expression plasmids could be constructed containing the cDNA of HNF-1 or YB-1 . These plasmids could be used to cotransfect with pFIIGH1.3 into HepG2, L-cells and B H K cells. The in vivo effects of HNF-1 or YB-1 or HNF-1 /YB-1 on the cis-activating activity of 1.3 kbp of the human prothrombin gene can be indirectly determined by the levels of hGH secreted into the medium. In addition, these expression plasmids could also be used to transfect non-liver cell-lines, and the nuclear extract containing a high level of HNF-1 or YB-1 could then be examined for binding activity by gel shift assays and DNase I footprint assays. Alternatively, the prokaryotic T7 expression system could be employed to over-express these proteins which could then be purified to homogeneity by conventional biochemical techniques. The recombinant products could be used to determine the binding affinities and other biochemical aspects. Isolation and characterization of the protein factors is important to the study of liver-specific expression since these proteins are responsible for the establishment and/or maintenance of the hepatocyte-specific phenotype. The concatenated-doubled-stranded enhancer fragment (-910 to -860) could be used to re-screen the HepG2 cDNA library again hoping to isolate other protein factors that bind to this enhancer. In addition, oligonucleotides corresponding to the region from -930 to -910 could be synthesized and 121 Discussion Fig . 22. Model of human prothrombin gene expression in the liver A putative model for human prothrombin gene expression in the liver. The proximal promoter region interacts with the general transcriptional factors (TFTIA, TFIIB, TFIID, TFHE, TFIIF and RNA polymerase U) (Wasylyk 1988). The presence of a distal element is still unclear (TF?). There are two C A P sites, (CA) -36 and -22, with respect to the A T G codon. At least four proteins are postulated to interact with the enhancer. FUA interacts with the region from -939 to -919; FIIB is likely to interact with the CCTCCC motif; HNF-1 recognizes the putative HNF-1 motif, and finally, YB-1 may play a role in the tissue specific expression of the human prothrombin gene (see discussion). +1 mRNAs 122 Discussion then used to screen the same expression library. Hopefully, a novel protein factor could be discovered that interacts with this element. The study of liver-specific expression of clotting factors is hampered by the low levels of expression of some of these proteins. As a long term project, the highly sensitive hGH transient expression system could also be used to study the regulation of expression of other clotting factors including factor VII and factor VIII which are present at very low levels in serum (0.5 and 0.1 ug/ml, respectively) (Furie and Furie, 1988). The interrelationships governing the expression of this family of proteins are both interesting and informative relative to the basic understanding of gene regulation. 123 References V. REFERENCES Abel, T., and T. Maniatis. 1989. Action of leucine zippers. Nature 341 : 24. Alberts, B. , D. Bray, J . Lewis, M . Raff, K. Roberts, and J. D. Watson. 1989. Molecular  Biology of the Cell. Edited by M . Robetson. New York: Garland Publishing, Inc. Anderson, W. F., L . Killos, L . Sanders-Haigh, P. J. Kretschmer, and E. G. Diaumakos. 1980. Replication and expression of thymidine kinase and human globin genes microinjected into mouse fibroblasts. Proc. Natl. Acad. Sci. USA 77 : 5399. Atkinson, T., and M . Smith. 1984. Oligonucleotide Synthesis, a Practical Approach. Edited by M . J . Gait. IRL Press Ltd., Oxford. Attree, O., D. Vidaud, M . Viduad, S. Amselem, J. M . Lavergne, and M . Goossens. 1989. Mutations in the catalytic domain of human coagulation factor IX: rapid characterization by direct genomic sequencing of D N A fragments displaying an altered melting behaviour. Genomics 4 : 266. Baumhueter, S., G. Courtois, and G. R. Crabtree. 1988. A variant nuclear protein in dedifferentiated hepatoma cells binds to the same functional sequences in the ^fibrinogen gene promoter as HNF-1. E M B O J. 7 : 2485. 124 References Baumhueter, S,. D. B . Mendel, P. B. Conley, C. J. Kuo, C. Turk, M . K . Graves, C. A . Edwards, G. Courtois, and G. R. Crabtree. 1990. HNF-1 shares three sequence motifs with the POU homeo domain proteins and is identical to LF-B1 and APF. Genes Dev. 4 : 372. Benoist, C , and D. Mathis. 1990. Regulation of major histocompatibility complex class II genes: X , Y and other letters of the alphabet. Annu. Rev. Immun. 8 : 681. Birch, H . E., and G. Schreiber. 1986. Transcriptional regulation of plasma protein synthesis during inflammation. J. Biol. Chem. 261 : 8077. Birkenmeier, E. H . , B. Gwynn, S. Howard, J. Jerry, J. I. Gordon, W. H . Landschulz, and S. McKnight. 1989. Tissue-specific expression, developmental regulation, and genetic mapping of the gene encoding CCAAT/enhancer binding protein. Genes Dev. 3 : 1146. Bode, W., I. Mayr, U . Baumann, R. Huber, S. R. Stone, and J. Hofsteenge. 1989. The refined 1.9 A crystal structure of human a-thrombin: interaction with D-Phe-Pro-Arg chloromethylketone and significance of the Tyr-Pro-Pro-Trp insertion segment. E M B O J. 8 : 3467. Bohmann, O., T. J. Bos, A . Admon, T. Nishimura, and R. Tjian. 1987. Human proto-oncogene c-jun encodes a D N A binding protein with structural and functional properties of transcription factor AP-1. Science 238 : 1386. 125 References Bonner, J. J., and M . L . Pardue. 1977. Polytene chromosome puffing and in situ hybridization measure different aspects of RNA metabolism. Cell 12 : 227. Borowski, M . , B. C. Furie, S. Bauminger, and B . Furie. 1986. Prothrombin requires two sequential metal-dependent conformational transitions to bind phospholipid. Conformation-specific antibodies directed against the PL-binding site on prothrombin. J. Biol. Chem. 261 : 14969. Brent, R., and M . Ptashne. 1985. A eukaryotic transcriptional activator bearing the D N A specificity of a prokaryotic repressor. Cell 43 : 729. Butkowski, R. J., J. Elion, M . R. Downing, and K. G. Mann. 1977. Primary structure of human prethrombin 2 and oc-thrombin. J. Biol. Chem. 252 : 4942. Castell, J. V . , M . J. Gomez-Lechon, and M . David. 1988. Recombinant human IL-6 regulates the synthesis of acute phase proteins in human hepatocytes. FEBS Let. 242 : 237. Castellino, F. J., and J. M . Beals. 1987. The genetic relationships between the kringle domains of human plasminogen, prothrombin, tissue plasminogen activator, urokinase, and coagulation factor X I I . J. Mol . Evol. 26 : 358. Celada, A . , and R. A . Maki. 1989. D N A binding of the mouse class II major histocompatibility C C A A T factor depends on two components. Mol. Cell. Biol. 9 : 3097. 126 References Chen, C , and H . Okayama. 1987. High-efficiency transformation of mammalian cells by plasmid D N A . Mol . Cell. Biol. 7 : 2745. Chodosh, L . A . , A . S. Baldwin, R. W. Carthew, and P. A . Sharp. 1988. Human CCAAT-binding proteins have heterologous subunits. Cell 53 : 11. Chomczynski, P., and N . Sacchi. 1987. Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction. Anal. Biochem. 162 : 156. Clerc, R. G., L . M . Corcoran, J. H . LeBowitz, D. Baltimore, and P. A . Sharp. 1988. The B-cell-specific Oct-2 protein contains POU box- and homeo box-type domains. Genes  Dev. 2 : 1570. Cochrane, C. G. 1982. Plasma proteins and inflammatory disease. Pharmacol. Rev. 34 : 39. Cockerill, P. N . , M . H . Yeun, and W. T. Garrard. 1986. The enhancer of the immunoglobulin heavy chain locus is flanked by presumptive chromosomal loop anchorage elements. J. Biol . Chem. 261 : 3838. Collen, D. 1980. On the regulation and control of fibrinolysis. Thromb. Haemostasis. 43 : 77. 127 References Colman, R. W., Hirsh, J., Marder, V . J., and Saltzman E. W.1987. Hemostasis and  Thrombosis. Basic Principles and Clinical Practice. Second Edition. J. B. Lippincott Company. Philadelphia. Cool, D. E., and R. T. A . MacGillivray. 1987. Characterization of the human blood coagulation factor XII gene. J. Biol. Chem. 262 : 13662. Concino, M . F., R. F. Lee, J. P. Merry weather, and R. Weinmann. 1984. The A M L P T A T A box and initiation sites are both necessary for transcription in vitro. Nucleic Acids  Res. 12 : 7423. Courtois, G., S. Baumhueter, and G. R. Crabtree. 1988. Purified HNF-1 interacts with a family of hepatocyte-specific promoters. Proc. Natl. Acad. Sci. USA 85 : 7933. Courtois, G., J. G. Morgan, L . A . Campbell, G. Fourel, and G. R. Crabtree. 1987. Interaction of a liver-specific nuclear factor with the fibrinogen and a-1-antitrypsin promoters. Science 238 : 688. Crossley, M . , and G. G. Brownlee. 1990. Disruption of a C/EBP binding site in the factor IX promoter is associated with haemophilia B. Nature 345 : 444. Davis, L . G., M . D. Dibner, and J. F. Battey. 1986. Basic methods in molecular biology. New York: Elsevier Science Publishing Co., Inc. 128 References Degen, S. J. 1989. Thrombosis Haemostas. 62 : 153. Degen, S. J., and E. W. Davie. 1987. Nucleotide sequence of the gene for human prothrombin. Biochemistry 26 : 6165. Degen, S. J., R. T. A . MacGillivray, and E. W. Davie. 1983. Characterization of the cDNA and gene coding for human prothrombin. Biochemistry 22 : 2087. Denoto, F. M . , D. D. Moore, and H . M . Goodman. 1981. Human growth hormone D N A sequence and mRNA structure: possible alternative splicing. Nucleic Acids Res. 9 : 3719. Descombes, P., M . Chojkier, S. Lichtsteiner, E . Falvey, and U . Schibler. 1991. L A P , a novel member of the C/EBP gene family , encodes a liver-enriched transcriptional activator protein. Genes Dev. 4: 1541. Didier, D. K. , J. Schiffenbauer, S. L . Woulfe, and M . Zacheis. 1988. Characterization of the cDNA encoding a protein binding to the major histocompatibility complex class LT Y box. Proc. Natl. Acad. Sci. USA 85 : 7322. Dierks, P. 1983. Three regions upstream from the CAP site are required for efficient transcription of the rabbit B-globin gene in mouse 3T6 cells. Cell 32 : 695. Doolitde, R. F. 1981. Fibrinogen and fibrin. Scientific Amer. 245 : 126. 129 References Esmon, C. T., and C. M . Jackson. 1974. The conversion of prothrombin to thrombin. IV. The function of the fragment 2 region during activation in the presence of factor V . J. Biol. ChenL 249 :7791. Fair, D. S., and B. R. Bahnak. 1984. Human hepatoma cells secrete single chain factor X , prothrombin and antithrombin III. Blood 64 : 194. Fair, D. S., and R. A . Marlar. 1986. Biosynthesis and secretion of factor VII, protein S, protein C and the protein C inhibitor from a human hepatoma cell line. Blood 67 : 64. Feinberg, A . P., and B. Vogelstein. 1983. A technique for radiolabeling D N A restriction endonuclease fragments of high specific activity. Anal. Biochem. 132 : 6. Fletcher, C , N . Heintz, R. G. Roeder. 1987. Purification and characterization of OTF-1, a transcription factor regulating cell cycle expression of a human histone H2B gene. Cell. 51 : 773. Finney, M . 1990. The homeodomain of the transcription factor LF-1 has a 21 amino acid loop between helix 2 and helix 3. Cell 60 : 5. Fisher, L . M . 1984. D N A supercoiling and gene expression. Nature 307 : 686. 130 References Flavell, R. A . , H . Allen, B . Huber, C. Wake, and G. Widera. 1985. Organization and expression of the M H C of the C57 black/10 mouse. Immunol. Rev. 84 : 29. Fromm, M . , and P. Berg. 1982. Deletion mapping of D N A regions required for SV40 early region promoter function in vivo. J. Mol . Appl. Genet. 1 : 457. Fuller, G. M . , J. M . Otto, B . M . Woloski, C. T. McGary, and M . A . Adams. 1985. The effects of HSF on fibrinogen biosynthesis in hepatocyte monolayers. J. Cell. Biol. 101 : 1481. Furie, B. , and B. C. Furie. 1988. The molecular basis of blood coagulation. Cell 53 : 505. Gatermann, K . B. , Rosenberg. G. H. , and N . F. Kaufer. 1988. Double-stranded sequencing, using mini-prep plasmids, in eleven hours. BioTechniques 6 : 951-952. Gitschier, J., W. I. Wood, T. M . Goralka, K . L . Wion, E. Y . Chen, D. H . Eaton, G. A . Vehar, D. J. Capon, and R. M . Lawn. 1984. Characterization of the human factor VIII gene. Nature 312 : 326. Gluzman, Y . 1982. SV40 transformed simian cells support the replication of early SV40 mutants. Cjgii 23 : 175. Goldstein, J., N . S. Pollitt, and M . Inouye. 1990. major cold shock protein of E. coli. Proc. Natl. Acad. Sci. USA 87 : 283. 131 References Gorski, K. , M . Carneiro, and U . Schibler. 1986. Tissue-specific in vitro transcription from the mouse albumin promoter. Cell 47 : 767. Graham, F., and A . Van der Eb. 1973. Transformation of rat cells by D N A of human adenovirus 5. Virol. 52 : 456. Graves, C. B. , T. W. Munns, T. L . Carlisle, and G. A . Grant. 1981. Induction of prothrombin synthesis by prothrombin fragments. Proc. Natl. Acad. Sci. U S A 78 : 4772. Graves, C. B. , T. W. Munns, A . K . Willingham, and A . W. Strauss. 1982. Rat factor X is synthesized as a single chain precursor inducible by prothrombin fragments. J. Biol. Chem, 257 : 13108. Graves, B. J., P. F. Johnson, and S. L . McKnight. 1986. Homologous recognition of a promoter domain common to the M S V LTR and the HSV tk gene Cell. 44 : 565. Green, F., and S. Humphries. 1989. The molecular biology of coagulation. Edited by E. G. D. Tuddenham. Bailliere's Clinical Haematology. London: Bailliere Tindall. Green, S., and P. Chambon. 1987. Ostradiol induction of a glucocorticoid-responsive gene by a chimeric receptor. Nature 325 : 75. 132 References Griffith, J., A . Hoshschild, and M . Ptashne. 1986. D N A loops induced by cooperative bindings of lambda-repressor. Nature 322 : 750. Hammer, R. E. , R. Krumlauf, S. A . Camper, R. L . Brinster, and S. M . Tilghman. 1987 Diversity of alpha-fetoprotein gene expression on mice is generated by a combination of separate enhancer elements. Science 235: 53. Hanahan, D. Studies on transformation of E. coli with plasmids. 1983. J. Mol . Biol . 166 : 557. Hardon, E. M . , M . Frain, G. Paonessa, and R. Cortese. 1988. Two distinct factors interact with the promoter regions of several liver-specific genes. E M B O J. 7 : 1711. Holmsen, H , and Weiss, H.J. 1979 Secretable storage pools in platelets. Ann. Rev. Med. 30 : 119. Honigwachs, J., O. Faktor, R. Dikstein, Y . Shaul, and O. Laub. 1989. Liver-specific expression of hepatitis B virus is determined by the combined action of the core gene promoter and the enhancer. J. Virol. 63 : 919. Horikoshi, M . , T. Hai, Y.-S. Lin , M . R. Green, and R. G. Roeder. 1988. Transcription factor A P I interacts with the TATA factor to facilitates establishment of a preinitiation complex. Cell 54 : 1033. 133 References Irwin, D . M . , K . G. Ahern, G. D. Pearson, and R. T. A . MacGillivray. 1985. Characterization of the bovine prothrombin gene. Biochemistry 24 : 6854. Irwin, D. M . , K . A . Robertson, and R. T. MacGillivray. 1988. Structure and evolution of the bovine prothrombin gene. J. Mol . Biol. 200 : 31. Jackson, C. M . , and Y . Nemerson. 1980. Blood coagulation. Annu. Rev. Biochem. 49 : 765. Jakobovits, E. B. , S. Bratosin, and Y . Aloni. 1980. A nucleosome-free region in SV40 minichromosomes. Nature 285 : 263. Jandl, J. H . . 1987. Blood, Textbook of Hematology. Boston: Little, Brown and Company, pp 965-1019. Johnson, P. F., W. H . Landschulz, B. J. Graves, and S. L . McKnight. 1986. Genes Dev. 1 : 133. Johnson, P. F., and S. L . McKnight. 1989. Eukaryotic transcriptional regulatory proteins. Annu. Rev. Biochem. 58 : 799. Jordan, S. R., and C. O. Pabo. 1988. Structure of the lambda complex at 2.5 A resolution: details of the repressor-operator interactions. Science 242 : 893. 134 References Jorgensen, M . J., A . B . Cantor, B. C. Furie, and B. Furie. 1987. Expression of completely gamma-carboxylated recombinant human prothrombin. J. Biol. Chem. 262 : 6729. Kaddurah, D.R., J. M . Greene, A . S. Baldwin, and R. E. Kingston. 1987. Activation and repression of mammalian gene expression by the c-myc protein. Genes Dev. 1 : 347. Kadonaga, J. T., K . R. Carner, F. R. Masiarz, and R. Tjian. 1987. Isolation of cDNA encoding transcription factor Spl and functional analysis of the D N A binding domain. Cell 51 : 1079. Karin, M . , A . Haslinger, A . Heguy, T. Dietlin, and T. Cooke. 1987. Metal-responsive elements act as positive modulators of human metallothionein-IIA enhancer activity. Mol.  Cell. Biol . 7 : 606. Karpatkin, S., T. H . Finlay, A . L . Ballesteros, and M . Karpatkin. 1987. Effect of warfarin on prothrombin synthesis and secretion in human HepG2 cells. Blood 70 : 773. Kisker, C. T., S. Perlman, D. Bohlken, and B. Wicklund. 1988. Measurement of prothrombin mRNA during gestation and early neonatal development. J. Lab. Clin. Med. 112:407. Kmiec, E . B. , M . Ryoji, and A . Worcel. 1986. Gyration is required for 5SRNA transcription from a chromatin template. Proc. Natl. Acad. Sci. USA 83 : 1305. 135 References Konieczny, S. F., and C. P. Emerson. 1984. 5-azacytidine induction of stable mesodermal stem cell lineages from 10T1/2 cells: evidence for regulatory genes controlling determination. £eU 38 : 791. Kouzarides, T., and E. Ziff. 1989. Leucine zippers of fos, jun and GCN4 dictate dimerization specificity and thereby control D N A binding. Nature 340 : 568. Kovacs, B . J., and P. H . W. Butterworth. 1986. The effect of changing the distance between the T A T A box and cap site by up to three bp on the selection of the transcriptional start site of a cloned eukaryotic gene in vitro and in vivo. Nucleic Acids Res. 14 : 2429. Kriegler, M . 1990. Gene transfer and expression, a laboratory manual. Macmillian Publishers L T D . Stockton Press, pp 3-21. Kucherlapati, R., and I. Skoultchi. 1983. Introduction of purified genes into mammalian cells. C R C Crit. Rev. Bioc. 16 : 349. Kugler, W., U . Wagner, and G. U . Ryffel. 1988. Tissue-specificity of liver gene expression: a common liver-specific promoter element. Nucleic Acids Res. 16 : 3165. Kunkel, T.A. 1985. Rapid and efficient site-specific mutagenesis without phenotypic selection. Proc. Natl. Acad. Sci. USA 82 : 488. 136 References Labhart, P., and T. Koller. 1982. Structure of the active nucleolar chromatin of Xenopus laevis oocytes. Cell 28 : 279. Laemmli, N . K . 1970. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227 : 680. Lai , E. , V . R. Prezioso, E. Smith, O. Litvin, R. H . Costa, and J. E . Jr. Darnell. 1990a. HNF-3A, a hepatocyte-enriched transcription factor of novel structure is regulated transcriptionally. Genes Dev. 4 : 1427. Lai , E. , V . R. Prezioso, W. Tao, W. S. Chen, and J. E. Jr. Darnell. 1990b. HNF-3a belongs to a gene family in mammals that is homologous to the Drosophila homeotic gene fork head. Genes Dev. 5 : 416. Landschulz, W. H. , P. F. Johnson, and S. L . McKnight. 1988. The leucine zipper: a hypothetical structure common to a new class of D N A binding proteins. Science 240 : 1759. Lichtsteiner, S., J. Wuarin, and R. Schible. 1987. The interplay of DNA-binding proteins on the promoter of the mouse albumin gene. Cell 51 : 963. Lill ie, J. W., and M . R. Green. 1989. Activator's target in sight. Nature 341 : 279. 137 References Lijnen, H . R. and Collen, D. 1982 Interaction of plasminogen activators and inhibitors with plasminogen and fibrin. Semin. Thromb. Hemost. 8 :2 . Liu , L . F., and J. C. Wang. 1987. Supercoiling of the D N A template during R N A transcription. Proc. Natl. Acad. Sci. USA 84 : 7024. MacGillivray, R. T. A . , and E. W. Davie. 1984. Characterization of bovine prothrombin mRNA and its translation product. Biochemistry 23 : 1626. Magnusson, S., T. E. Peterson, L . Sottrup-Jensen, and H . Claeys. 1975. Proteases and  Biological Control. Reich, E., Rifkin, D. B., & Shaw, E., Eds. Cold Spring Habour, N Y : Cold Spring Harbor Laboratories, pp 123-149. Malhotra, O. P. 1989. Dicoumarol-induced prothrombins containing 6, 7, and 8 gamma-carboxyglutamic acid residues: isolation and characterization. Biochem. Cell. Biol. 67 : 411. Maniatis, T., E.F. Fritsch, and J. Sambrook. 1989. Molecular Cloning, a laboratory  manual, second ed. New York: Cold Spring Habor Laboratory Press. Maniatis, T., S. Goodbourn, and J. A . Fischer. 1987. Regulation of inducible and tissue-specific gene expression. Nature 236 : 1237. 138 References Mann, K . G., M . E. Nesheim, P. B . Tracy, L . S. Hibbard, and J. S. Bloom. 1982. Assembly of the prothrombinase complex. Biophys. J. 37 : 106. Matthews, B. W. 1988. No code for recognition. Nature 335 : 294. McCutchan, J. H. , and J. S. Pagana. 1968. Enhancement of the infectivity of SV40 D N A with DEAE-dextran. J. Natl. Cancer Inst. 41 : 351. McKnight, S. L . 1982. Functional relationships between transcriptional control signals of the thymidine kinase gene of herpes simplex virus. Cell 31 : 355. McKnight, S. L . , R. C. Kingsbury, A . Spence, and M . Smith. 1984. The distal transcription signals of the herpes virus T K gene share a common hexanucleotide control sequence . Cell 37 : 253. McKnight, S. L . , M . D. Lane, and S. Gluecksohn-Waelsch. 1989. Is C/EBP a central regulator of energy metabolism. Genes Dev. 3 : 2021. Mellon, P., V . Parker, Y . Gluzman, and T. Maniatis. 1981. Identification of D N A sequences required for transcription of the human alphal-globin gene in a new SV40 host-vector system. Cell 27 : 279. Messing, J. 1983. New M13 vectors for cloning. Meth. Enzymol. 101 : 20. 139 References Miller, J., A . D. McLachlan, and A . Klug. 1985. Repetitive zinc-binding domain in the protein transcription factor IIIA from Xenopus oocytes. E M B O J. 4 : 1609. Miesfeld, R., P. J. Godoski, B. A . Maler, and K . R. Yamamoto. 1987. Glucocorticoid receptor mutant that define a small region sufficient for enhancer activation. Science. 236 : 423. Mitropoulos, K . A. , and M . P. Esnouf. 1990. The prothrombin activation peptide regulates synthesis of the vitamin K-dependent proteins in the rabbit. Thrombos. Res. 57 : 541. Morgan, J. G. , G. Courtois, G. Fourel, Chodosh L . A. , L . Campell, E . Evans, and G. R. Crabtree. 1988. Spl , a CAAT-binding factor, and the adenovirus late promoter transcription factor interact with functional regions of the gamma-fibrinogen promoter. Mol . Cell. Biol. 8 : 2628. Mossing, M . C , and M . T. Record. 1986. Upstream operators enhance repression of the lac promoter. Science 233 : 889. Mueller, C. R., P. Maire, and U . Schibler. 1990. DBP, a liver-enriched transcription activator, is expressed late in ontogeny and its tissue specificity is determined posttranscriptionally. Cell 61 : 279. 140 References Munns, T. W., M . F. M . Johnson, M . K . Liszewski, and R. E. Olson. 1976. Vitamin K -dependent synthesis and modification of precursor prothrombin in cultured H-35 hepatoma cells. Proc. Natl. Acad. Sci. USA 73 : 2803. Nelson, R. M . , and G. L . Long. 1989. A general method of site-specific mutagenesis using a modification of the PCR. Anal. Biochem. 180 : 147. Nemerson, Y . , and H . L . Nossel. 1982. The biology of thrombosis. Annu. Rev. Med. 33 : 479. Neumann, E., M . Schaefer-Ridder, Y . Wang, and P. H . Hofschneider. 1982. Gene transfer into mouse myeloma cells by electroporation in high electric field. E M B O J. 1 : 841. Neurath, H . 1984. Evolution of proteolytic enzymes. Science 224 : 350. Nussinov, R. 1990. Sequence signals in eukaryotic upstream regions. Crit. Rev. Bioc.  Mol . Biol . 25 : 188. 0sterud, B. , U . Lindahl, and R. Seljelid. 1980. Macrophages produce blood coagulation factors. FEBS Lett. 120 : 41. Otto, J. M . , H . E. Grenett, and G. M . Fuller. 1987. The coordinated regulation of fibrinogen gene transcription by HSF and dexamethasone. J. Cell. Biol. 105 : 1067. 141 References Packham, M . A . , and Musturd, J. F. 1984 Platelet adhesion. Prog. Hemost. Thromb. 7 : 211. Park, C. H . , and A . Tulinsky. 1986. Three-dimensional structure of the kringle sequence: structure of prothrombin fragment 1. Biochemistry 25 : 3977. Patthy, L . 1985. Evolution of the proteases of blood coagulation and fibrinolysis by assembly from modules. Cell 41 : 657. Pinkert, C. A . , D. M . Ornitz, R. L . Brinster, and R. D. Palmiter. 1987. An albumin enhancer located 10 kb upstream functions along with its promoter to direct efficient, liver-specific expression in transgenic mice. Genes Dev. 1 : 268. Plutzky, J., J. Hoskins, G. L . Long, and G. R. Crabtree. Evolution and organization of the human protein C gene. 1986. Proc. Natl. Acad. Sci. USA 83 : 546. Rauscher, F. J. III., L . C. Sambucetti, T. Curran, R. J. Distel, and B. M . Spiegelman. 1988. A common D N A binding site for the fos protein complexes and the transcription factor AP-1. £el l 52: 471. Raymondjean, M . , S. Cereghini, and M . Yaniv. 1988. Several distinct " C C A A T " box binding proteins coexist in eukaryotic cells. Proc. Natl. Acad. Sci. USA 85 : 757. 142 References Reinberg, D., M . Horikoshi, and R. G. Roeder. 1987. Factors involved in specific transcription in mammalian RNA polymerase LT: purification and functional analysis of initiation factors TFIIA and TFl lD and the identification of a new factor operating at sequences downstream of the initiation site . J. Biol. Chem. 262 : 1987. Reitsma., P. H . , T. Mandalaki, C. K . Kasper, R. M . Bertina, and E. Briet. 1989. The novel point mutations correlate with an altered developmental expression of blood coagulation factor IX (Hemophilia B Leyden phenotype). Blood 73 : 743. Robertson, M . 1988. Homeo boxes, POU proteins and the limits promiscuity. Nature 336 : 522. Robins, D. M . , S. Ripley, A . S. Henderson, and R. Axel. 1981. Transforming D N A integrates into the host of transformed cells. Cell 23 : 29. Royle, N . J., D. M . Irwin, M . L . Koschinsky, R. T. MacGillivray, and J. L . Hamerton. 1987. Human genes encoding prothrombin and ceruloplasmin map to 1 l p l l -ql2 and 3q21-24, respectively. Som. Cell. Mol . Genet. 13 : 285. Saffer, J. D., S. P. Jackson, and M . B. Annarella. 1991. Developmental expression of Spl in the mouse. Mol. Cell. Biol. 11 : 2189. Sailer, J.-P., S. Hirosawa, and K . Kurachi. 1990. Functional characterization of the 5'-regulatory region of human factor IX gene. J. Biol. Chem. 265 : 7062. 143 References Sakura, H . , T. Maekawa, F. Imamoto, K. Yasuda, and S. Ishii. 1988. Two human genes isolated by a novel method encode DNA-binding proteins containing a common region of homology. Gene 73 : 499. Sakai, D. D., S. Helms, J. Carlstedt-Duke, J. A . Gustafsson, and F. M . Rottman. 1988. Hormone mediated repression: a negative glucocorticoid response element from the bovine prolactin gene. Genes Dev. 2 : 1144 Saltzman, A . G., and R. Weinmann. 1989. Promoter specificity and modulation of RNA polymerase II transcription. FASEB J. 3 : 1723. Sanger, F., S. Nicklen, and A . R. Coulson. 1977. D N A sequencing with chain termination inhibitors. Proc. Natl. Acad. Sci. USA 74 : 5463. Santoro, C., N . Mermod, P. C. Andrews, and R. Tjian. 1988. A family of human CCAAT-box-binding proteins active in transcription and D N A replication: cloning and expression of multiple cDNAs. Nature 334 : 218. Sassone-Corsi, P., L . J. Ransone, W. W. Lamph, and I. M . Verma. 1988. Direct interaction between the fos and jun nuclear oncoproteins: role of leucine zipper domain. Nature 336:962. 144 References Sawadaishi, K. , T. Morinaga, and T. Tamaoki. 1988 Interaction of a hepatoma-specific nuclear factor with transcription-regulatory sequences of the human a-fetoprotein and albumin genes. J. Biol . Chem. 8 : 5179. Scheidereit, C , A . Huguy, R. G. Roeder. 1987. Identification and purification of a human lymphoid-specific octamer-binding protein (OTF-2) that activates transcription of an immunoglobulin promoter in vitro. Cell 51 : 783. Scott, M . P., and A . J. Weiner. 1984. Structural relationships among genes that control development: Sequence homology between the Antennapedia, Ultrabithorax, and Fushi tarazu loci of Drosophila. Proc. Natl. Acad. Sci. USA 81 : 4115. Seale, P. F., G. W. Stuart, and R. D. Palmiter. 1985. Building a metal-responsive promoter with synthetic regulatory elements. Mol. Cell. Biol. 5 : 1480. Selden, R. F., K . Burke Howie, H . M . Goodman, and D. D. Moore. 1986. Human growth hormone as a reporter gene in regulation studies employing transient gene expression. Mol . Cell. Biol . 6 : 3173. Shakhov, A . N . , S. A . Nedospasov, and G. P. Georgiev. 1982. DNasell as a probe to sequence specific chromatin organization: preferential cleavage in the 72-bp modular sequence of SV40 minichromosomes. Nucleic Acids Res. 10 : 3951. 145 References Somasekhar, M . B. , and J. E . Mertz. 1985. Sequences involved in determining the locations of the 5' ends of the late RNAs of S V40. J. Virol. 56 : 1002. Struhl, K. 1987. The D N A binding of jun oncoprotein and the yeast GCN4 transcriptional activator proteins are functionally homologous. Cell 50 : 841. Struhl, K. 1988. The jun oncoprotein, a vertebrate transcription factor, activates transcription in yeast. Nature 332 : 649. Sturm, R. A. , G. Das, W. Herr. 1988. The ubiquitous octamer-binding protein Oct-1 contains a POU domain with a homeo box subdomain. Genes Dev. 2 : 1582. Suttie, J. W. 1980. Mechanism of action of vitamin K: synthesis of gamma-carboxyglutamic acid. Crit. Rev. Biochem. 8 : 191. Tollersrud, O. K. , A . H . Kvalvaag, and L . Helgeland. 1989. Biosynthesis and clearance of prothrombin in warfarin-treated rats. Biochim. Biophys. Acta 1010 : 35. Tsao, Y.-P. , H . -Y. Wu, and L. F. Liu. 1989. Transcription driven supercoiling of D N A : direct biochemical evidence from in vitro studies. Cell 56 : 111. Vasiliev, J. M . , and Gelfand, I. M . 1978 Mechanisms of non-adhesiveness of endothelial and epithelial surfaces. Nature 274 : 710. 146 References Vieira, J., and J. Messing. 1982. The pUC plasmids, an M13mp7 derived system for insertion mutagenesis and sequencing with synthetic universal primers. Gene 19 : 259. Vogt, P. K. , T. J. Bos, and R. F. Doolittle. 1987. Homology between the D N A binding domains of GCN4 regulatory protein of yeast and the C-terminal region of a protein coded for by the oncogene jun . Proc. Natl. Acad. Sci. USA 84 : 3316. Walz, D. A. , D. Hewitt-Emmett, and W. H . Seegers. 1977. Amino acid sequence of human prothrombin fragment 1 and 2. Proc. Natl. Acad. Sci. USA 74 : 1969. Wang, J. C , and G. N . Giaever. 1988. Action at a distance along a D N A . Science 240 : 300. Wasylyk, B. 1988. Transcription elements and factors or RNA polymerase B promoters of higher eukaryotes. Crit. Rev. Bioc. 23 : 77. Watanabe, K. , A . Saito, and T. Tamaoki. 1987. Cell-specifc enhancer activity in a far upstream region of the human a-fetoprotein gene. J. Biol. Chem. 262 : 4812. Watson, J. D., N . H . Hopkins, J. W. Roberts, J. A . Steitz, and A . M . Weiner. 1987. Molecular biology of the gene. 4 ed. California: Benjamin/Cummings. Weintraub, H . 1985 Assembly and propagation of repressed and derepressed chromosomal states. Cell 42 : 705. 147 References White, J. G. 1979 Current concepts of platelet structure. Am. J. Clin. Path. 71 : 363. Wistow, G. 1990. Cold shock and D N A binding. Nature. 344 : 824. Wu, H.-Y. , S. Shy, and J. C. Weng. 1988. Transcription generates positive and negative supercoiled domains in the template. Cell 53 : 433. Yanisch-Perron, C , J Vieira, and J Messing. 1985. Improved M l 3 phage cloning vectors and host strains; nucleotide sequences of the M13mpl8 and pUC19 vectors. Gene 33 : 103. Yee, J.-K. 1989. A liver-specific enhancer in the core promoter region of human hepatitis B virus. Science 246 : 658. Yoshitake, S., B . G. Schach, D. C. Foster, E. W. Davie, and K. Kurachi. 1985. Biochemistry 24 : 3736. Young, R. A . , and R. W. Davis. 1983. Efficient isolation of genes by using antibody probes. Proc. Natl. Acad. Sci. USA 80 : 1194. Zonneveld, A-J . V. , H . Veeran, and H . Pannekoek. 1986. Autonomous functions of structural domains on human tissue-type plasminogen activator. Proc. Natl. Acad. Sci. U S A 83 : 4670. 148 \ 


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items