a... #57". 31...... .. z... . .73. 2213.3...4... : . @5ng . an . E. in": . . , unwasnfii . . .ma Iii-5.0..) .. mu”. KW»: . P c. I. 5.9.45: ‘ 33$ WM“: 3211 Pi .‘l l. r. aqémrn fr . l » . .nmhu 3.541... 1. guy? .r‘hftfrir‘l v. “u! 223.35. tfiuugd “Na . 3:5.432‘6. . :1an .21 . 5.... Emily: 3-. o. 1: iii a: it... iii: if." .‘ $3.19 . .8 a. :13... 3.. .3... rfinlflurvtal .. s! . .vktz‘ai... to a c .5. . iii“. . .... a...” I IQ." i . .hfinfi. t2: 3 3:333. w... ... .!.:x.sh.¢.. ¥ » . u .. .12.)... 4:133? $53.22..) 1.. Hints av 5.. 5.4....2533‘!‘ .. . 2.. $1... .. 3.?)36. .1...is....li.. 1% .irfilo‘ 211...... I: ‘v! I”: ‘ll‘llti-‘tzilt-f.’ é}. a} .13..a..§.!.\ :. :11. u. .201. an?! $15!. .ll. I. . 131:. O ‘uillnalt? (st-(t . :51; iv! I... Iii-3.)!!! .1» 55qu ._ s. 4 . Ly LIBRARY Michigan State University This is to certify that the dissertation entitled THE TRANSCRIPTIONAL REGULATION OF PROSTAGLANDIN SYNTHASE-Z IN LIPOPOLYSACCHARIDE STIMULATED MACROPHAGE CELLS presented by Byron Asa Wingerd has been accepted towards fulfillment of the requirements for Ph.D. degreein Microbiology and Molecular Genetics and Cell and Molecular Biology fla/Z/a; A Q 74 Major professor Date December 7, 2001 MSU is an Affirmative Action/Equal Opportunity Institution 0-12771 v—‘—— ' ' PLACE IN RETURN Box to remove this checkout from your record. To AVOID FINES return on or before date due. MAY BE RECALLED with earlier due date if requested. DATE DUE DATE DUE DATE DUE 6/01 cJCIRC/DateDuepSS-pJS THE TRANSCRIPTIONAL REGULATION OF PROSTAGLANDIN SYNTHASE—2 IN LIPOPOLYSACCHARIDE STIMULATED MACROPHAGE CELLS By Byron Asa Wingerd A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Microbiology and Molecular Genetics and the Cell and Molecular Biology Program 2001 ABSTRACT THE TRAN SCRIPTIONAL REGULATION OF PROSTAGLANDIN SYNTHASE-2 IN LIPOPOLYSACCHARIDE STIMULATED MACROPHAGE CELLS By Byron Asa Wingerd Prostaglandin H synthase (COX) catalyzes the first committed step in the metabolism of arachidonic acid to prostaglandins. There are two isoforms of COX, the predominantly constitutive isoform, COX-1, and an inducible isoform, COX-2. Although both enzymes catalyze the same reaction with similar kinetics, studies with isoforrn specific inhibitors and COX-1 and COX-2 knockout mice suggest that there are physiological processes that require one specific enzyme and others where both isoforms function together. COX-2 expression is upregulated by a variety of stimuli in different cell types. Bacterial lipopolysaccharide (LPS) is a potent inducer of COX-2 activity in macrophage cells. LPS signaling is mediated through the toll-like receptor-4 and results in the activation of JNK, ERK, p38, NIK and PKC signaling pathways. COX-2 transcription is regulated through multiple redundant mechanisms involving their interactions with several central response elements. Characterization of the COX-2 gene promoter has resulted in the identification of cis-acting response elements that are necessary for maximal promoter activity in LPS-treated macrophage cells (Figure 2). The CRE at —57/-52 is necessary for mediating the effects of a wide variety of stimuli, while a pair of C/EBP sites and an NF-KB response element appear to function in more specialized signaling events. The promoters of the human, murine, rat, equine, and bovine COX-2 genes contain paired CRE and NF-KB sites located approximately between 380 and 550 bp upstream of the transcription start site. Here we show that this conserved, upstream CRE (CRE-2) located at 4341-428 in the murine promoter is required for maximal induction of COX-2 by LPS in RAW 264.7 cells. Characterization of this site revealed that CREB/ATF transcription factors and the CREB binding protein from nuclear extracts of LPS-stimulated RAW 264.7 cells physically interact with this response element. The NF-KB site is necessary for lipopolysaccharide induced COX-2 expression in bovine atrial endothelial cells and MC3T3—E1 osteoblast like cells, but has not previously been demonstrated as necessary for the LPS response in RAW 264.7 macrophage cells. Recently Rhee et al. demonstrated that blocking NF-KB activation at several levels also blocks induced COX-2 promoter activity. Potent reduction in COX-2 expression was also observed in experiments using decoy NF-KB, antisense expression, and inhibitors of IKB degradation. We found that the NF-KB response element was necessary for maximal promoter activity. While characterizing the Re] components that bind the NF—KB response element, we observed an unusual pattern of binding where the probe was initially bound predominantly with a p65/p50 heterodimer and then later by a p50 dimer. Since p50 dimers are generally considered transcriptional repressors, we measured the rate of COX-2 transcription and found that the prolonged COX-2 response was due to persistent transcription of the gene. Our promoter analysis data indicate that the CRE-2 and NF-KB act together to activate COX-2 transcription. DEDICATION Marcia and Emma iv ACKNOWLEDGMENTS Mentor William L. Smith Committee Susan E. Conrad David L. DeWitt John C. Fyfe Richard C. Schwartz Others Toshiya Arakawa James Barton John Bell Richard E.Thompson Family Edgar C. Win gerd Lucy S. Wingerd Kevin L. Wingerd TABLE OF CONTENTS LIST OF TABLES ............................................................................................................ vii LIST OF FIGURES .......................................................................................................... viii LIST OF ABBREVIATIONS ............................................................................................ ix CHAPTER 1: LITERATURE REVIEW Introduction ............................................................................................................. 1 The Role of Prostaglandins in Pain and Inflammation ........................................... 7 Catalysis of Arachidonic Acid to Prostanoids ...................................................... 10 Historical Perspective ............................................................................................ 15 Transcriptional Regulation .................................................................................... 23 Transcriptional Regulation of COX-2 in Fibroblasts ............................................ 28 Transcriptional Regulation of COX-2 in Epithelial Cells ..................................... 32 Transcriptional Regulation of COX-2 in Endothelial Cells .................................. 38 Transcriptional Regulation of COX-2 in Bone Tissue .......................................... 41 Transcriptional Regulation of COX-2 in Granulosa Cells .................................... 49 Lipopolysaccharide Signaling ............................................................................... 51 Transcriptional Regulation of COX-2 in Macrophage .......................................... 65 CHAPTER 2: UPSTREAM NF-KB AND CAMP RESPONSE ELEMENTS IN CYCLOOXYGENASE-2 GENE EXPRESSION IN LPS-STIMULATED MACROPHAGE CELLS Summary ............................................................................................................... 75 Introduction ........................................................................................................... 77 Materials and Methods .......................................................................................... 80 Results ................................................................................................................... 85 Discussion ............................................................................................................. 97 CHAPTER 3: EXAMINATION OF THE COFACTORS ASSOCIATED WITH THE CRE-2 AND NF-KB REGION Introduction ......................................................................................................... 107 Supershift with I-KBB .......................................................................................... 110 Co-Transfection Experiments with Transcription Factors and Transcriptional Coactivators ................................................................................................... l 12 Shifting for Cooperative Complexes ................................................................... 117 Conclusion ........................................................................................................... 120 APPENDD( A: Caveats of Signaling Experiments ......................................................... 121 REFERENCES ................................................................................................................ 123 vi LIST OF TABLES Table 1. Classification of prostaglandin receptors and their signal transduction ................................................................................................. 4 Table H. Summary of prostaglandin receptor gene knockout data ........................... 6 Table III. Oligonucleotides used for EMSAs and mutants ....................................... 88 vii Figure 1. Figure 2. Figure 3. Figure 4. Figure 5. Figure 6. Figure 7. Figure 8. Figure 9. Figure 10. Figure 11. Figure 12. Figure 13. Figure 14. Figure 15. Figure 16 Figure 17. Figure 18. Figure 19. Figure 20. LIST OF FIGURES Arachidonic acid metabolism ................................................................... 12 Response elements of the human and mouse COX-2 promoter ................ 27 Lipopolysaccharide molecule .................................................................... 52 Toll-like receptor 4 complex ..................................................................... 54 General signaling pathways activated by LPS .......................................... 57 Detailed signaling pathway map ......................................................... 58, 59 Schematic of the murine COX-2 promoter ............................................... 79 Promoter activity of COX-2 promoter deletion constructs ....................... 86 Promoter activity of COX-2 promoter mutation constructs ...................... 89 Time course with the mutant promoter reporter plasmids ........................ 90 Specific binding of nuclear protein to the CRE-2 probe ........................... 92 CREB and CBP bind the CRE-2 probe ..................................................... 93 p65 and p50 bind the NF-KB probe ........................................................... 94 Nuclear run-on and Northern blot analysis of LPS stimulated RAW 264.7 cells .................................................................................................. 96 Model for the interactions of trans-activating factors associated with the COX-2 promoter .................................................................................. 99 Model of the CRE-2 and NF-KB region during LPS Stimulation ........... 108 COX-2 promoter co-transfection with CREB, CBP, and SRC-l ............ 114 COX-2 promoter co-transfection with p65, p50, and SRC-l .................. 116 EMSA probe map .................................................................................... 118 EMSAs with the CRE-2 and NF-KB region: The 59 bp probe compared with the CRE-2, middle, and NF-KB probes .......................... 119 viii AP-l bFGF C/EBP CAMP COX CRE CREB Dex DMEM ECSIT EGF EMSA ERK HFF I-Kb IKK IL- 1 ,2,6,8 IRAK JNK LPS MAPK LIST OF ABREVIATIONS Activator Protein-1 basic Fibroblast Growth Factor CAAT Enhancer Binding Protein Cyclic Adenosine Monophosphate Cyclooxygenase CAMP Response Element CAMP Response Element Binding Protein Dexamethasone Dulbecco's Modified Eagle's Medium Evolutionary Conserved Signaling Intermediate in Toll Epidermal Growth Factor Electrophoretic Mobility Shift Assay Extracellular signal-Regulated Kinase Human Foreskin Fibroblast Inhibitor of KB Inhibitor of KB Kinase Interleukin-l, -2, -6, -8 Interleukin-l Receptor Associated Kinase C-Jun N-term Kinase Lipopolysaccharide Mitogen Activated Protein Kinase MAPKAP-Kl MAPK- Activated Protein Kinase-l MEK MEKK MKK MKKK MSK-l, -2 NF-KB NIK p50 p65 PDGF PGD2 PGE2 PGFZa PGG2 PGH2 PGHS PG12 PKA PKC PMA See MKK See MKKK Mitogen Activated Protein Kinase Kinase, also called MEK Mitogen Activated Protein Kinase Kinase Kinase, also called MEKK Mitogen and stress Activated Kinase-1, -2 Nuclear Factor-kappa B Nuclear Factor-kappa B Inducing Kinase Rel protein 50, NF-KB transcription factor Rel protein 65, NF-KB transcription factor Platelet Derived Growth Factor Prostaglandin D2 Prostaglandin E2 Prostaglandin F20t Prostaglandin G2 Prostaglandin H2 Prostaglandin H Synthase, Prostaglandin Endoperoxide Synthase Prostaglandin 12 CAMP dependent Protein Kinase Calcium dependent Protein Kinase Phorbol 12-Myristate l3-Acetate PTH SSC TAE TAK- 1 TCF TGF—a -B TLR4 TNF-a TPA TRAF Parathyroid Hormone Salt-Sodium Citrate Tris-Acetate EDTA Transforming growth factor-beta Activated Kinase-l T-Cell Factor Transforming Growth Factor-0t -B Toll-Like Receptor-4 Tumor Necrosis Factor-0t Phorbol 12—Tetradecanoate 13-Acetate Tumor Necrosis Factor Receptor Associated Factor CHAPTER 1 LITERATURE REVIEW Introduction The major groups of eicosanoids are the prostaglandins, prostacyclins, thromboxanes, leukotrienes, and epoxy acids. These biologically active lipids are synthesized primarily from arachidonic acid, which is a component of cellular membranes and is synthesized from the essential fatty acid linoleate or obtained through the diet. The prostanoids are the products of prostaglandin endoperoxide H synthase (COX); leukotrienes are products of the 5- lipoxygenase pathway, and epoxy acids are the metabolites of cytochrome P450s. The prostanoids mediate a wide range of normal physiological responses. Pain and inflammation are mediated by prostaglandins. In animal models, joint inflammation as well as edema and hyperalgesia are blocked by COX inhibitors [1]. Prostaglandins sensitize the free ends of neurons and act centrally to increase general sensitivity to pain [2]. Blood Clotting is mediated by the release of thromboxane from platelets. Inhibition of platelet aggregation by the COX inhibitors has lead to the concept of using half an aspirin tablet a day as a prophylaxis against thromboembolitic disease [3]. Prostaglandins are necessary at multiple steps in the reproductive process. In mice with knocked out prostaglandin receptor or COX, genes exhibit reproductive failures because of problems with ovulation, fertilization, implantation, decidualization, and parturition. In addition, neonatal mice have severe renal pathology, malformed kidney structures, and patent ductus arteriosus [4, 5]. In the kidney, prostaglandins mediate glomerular hemodynamics and tubular reabsorbtion of water and sodium, and in the stomach low levels of prostaglandins inhibit acid and fluid release from the mucosa] layer [6]. Higher levels of prostaglandins induce the secretion of acid and fluid from the lining on the stomach and gastrointestinal tract. Smooth muscle contraction and relaxation are mediated by prostaglandins. In the intestines and uterus, prostaglandins mediate contraction of longitudinal muscle and contraction of circular muscles [6]. Prostacyclin is involved in the maintenance of vascular tone and also functions in the Circulatory system as an inhibitor of platelet aggregation[6]. Prostaglandins directly mediate the absorption and formation of bone tissue but are also indirectly involved in bone metabolism by affecting the differentiation and proliferation of osteoclast and osteoblast precursor cells [7]. There are two COX isozymes, the constitutive COX-1 and the inducible COX-2. COX-1 is expressed in most cell types and is involved in homeostasis and various physiological functions such as platelet aggregation, water and sodium metabolism in the kidney, stomach acid secretion, and parturition [8]. COX—2 was initially Cloned as an immediate early gene, and its expression is rapidly upregulated in response to pro- inflammatory stimuli such as bacterial lipopolysaccharide (LPS), Tumor Necrosis Factor-0t (TNF-a), Interleukins-l and -2 (IL-1, IL-2), reactive oxygen species, hypoxia, and mitogenic stimuli [9]. COX-2 is upregulated up to 80 fold, often from nearly undetectable levels in many tissues and cell types. COX-2 is generally not found under normal physiological conditions; however, it is constitutively expressed in specialized cells of the kidney and in brain and bone tissues. COX-2 knockout mice develop malformed kidney structures and severe renal pathology [4, 5]. In addition, the ductus arteriosus does not Close in about one third of the cox-2'/Cox-2' mice [10]. The wide variety of physiological actions affected by the prostanoids require a broad range of cellular signaling capabilities, which are mediated by the prostanoid receptors. The prostaglandin receptors have been characterized pharmacologically using radioactive ligands, and the binding properties indicated that a variety of prostaglandins cross—react with more than one receptor, suggesting that the receptors share a high degree of structural similarity [11]. Biochemical studies have demonstrated that the actions of prostaglandins are mediated by G proteins resulting in Changes in second messenger levels that are summarized in Table I. The receptors are classified by their primary agonist or antagonist into five groups termed DP, EP, FP, IP, and TP based on their sensitivity to PGD2, PGE2, PGan, PG12, and TXAZ, respectively. The EP receptors are further divided into four subgroups, EPl, EP2, EP3, and EP4, by pharmacological characterization with specific agonists. Because of the second messengers released in response to the prostanoids, one prostaglandin may cause opposite affects in different tissues, or even within the same tissue. In the lining of the stomach, both the EP3 and EP4 receptors are found in similar tissue and are involved in the balancing the release of chloride ions into the stomach. The EP3 receptor has a higher affinity for PGE2 than the EP4 receptor, so that at low ligand concentrations EP3 (coupled to phosphodiesterase to decrease intracellular CAMP) inhibits the secretion of acid, and at higher ligand concentrations EP4 (coupled to adenylate cyclase to increase CAMP) promotes acid secretion [12]. In other cases, Table 1. Classification of prostaglandin receptors and their signal transduction. Data obtained from Narumiya et al. [13] Ligand Type Subtype Isoform G protein Signal transduction PGD2 DP GS CAMP increase PGE2 EP EP, Gq (7) PI response, Ca.2+ EPz Gs CAMP increase EP4 Gs CAMP increase EP3 EP3A Gi CAMP decrease EP3}; Gs CAMP increase EP3C Gs CAMP increase EP3D Gus/q CAMP decrease / increase, PI response PGan FP PI response PG12 1P CAMP increase, PI response TXAz TP TP (1 Gi/q CAMP decrease, PI response TP [5 Gs/q CAMP increase, PI response different prostanoids mediate opposing effects. Thromboxanes function as potent vasoconstrictors and mediate platelet activation and aggregation, and PGI2 functions in an opposite role as a vasodilator and an inhibitor of platelet aggregation [6]. Localized expression of the receptors has been studied in cells and in tissues by Northem blot and in situ hybridization. Combined with pharmacological studies using cyclooxygenase inhibitors and various receptor agonists and antagonists, the physiological functions of the receptors have been defined. To Clarify the role of each receptor and their ligand, receptor genes have been individually disrupted. The major phenotypes of mice deficient in prostanoid receptors are summarized in Table II. The knockout mice all grow normally with a few exceptions. The FP null mice die in utero because the mother does not go into labor when the pre-natal mice reach term. EP4 null mice die within 72 hr due to patent ductus arteriosus, and EP2 null females are defective in ovulation and fertilization [2]. Table II. Summary of prostaglandin receptor gene knockout data. Data obtained from: Narumiya et a1. [2] and Sugimoto et a1. [14]. Disrupted Gene Major Phenotypes of knockout mice DP Reduced responses in allergic asthma EPl Reduction in carcinogen-induced colorectal neoplasia Reduction in allodynia (tactile pain) EP2 Impaired ovulation and fertilization Salt-sensitive hypertension Loss of bronchodilation Reduction in vasodepressor response to PGE2 Reduced osteoclast generation EP3 Impaired febrile response to pyrogens Enhanced vasodepressor response to PGE2 Impaired duodenal bicarbonate secretion Reduction in Hyperalgesia (sensitivity to pain) EP4 Patent ductus arteriosus Impaired vasodepressor response to PGE2 Decreased inflammation bone resorption Parturition ”3:8 Thrombotic tendency Decreased inflammatory swelling Decreased acetic acid writhing TP Bleeding tendency, resistance to thromboembolism The Role of Prostaglandins in Pain and Inflammation Prostanoids are involved at multiple levels of the inflammatory response as well as Chronic inflammation [2, 15]. At a very basic level, the research presented in this dissertation is intended to contribute to the understanding of how COX-2 is expressed in response to inflammatory stimuli. Inhibition of prostaglandin synthase enzymatic activity can greatly reduce the amount of pain and inflammation experienced as a result of trauma and tissue damage [16]. Another Class of drugs, glucocorticoids, work in part by restricting the production of COX-2 and are highly effective in reducing inflammation; however the side effects of this therapy prevent its continuous use [17, 18]. If the transcriptional regulation of COX-2 is understood with enough detail, it may be possible to block the production of the enzyme with a specific, non-steroidal, transcription inhibitor. Inflammation is a component of the innate immune response and is Characterized by redness of the skin, swelling of the affected area, heat, and pain [16]. The initial phase of an inflammatory response includes vasodilatation and increased vascular permeability and is followed by the infiltration of neutrophils and monocytes. Arachidonic acid metabolites play a number of roles in this response, from leukotrienes that function as chemotactic agents that increase vascular permeability, to prostaglandins that act as vasodilators, chemotactic agents, and enhance the effects of histamine and kinins. The link between pain caused by the inflammatory process and prostaglandins was demonstrated by in viva carrageenan rat paw experiments, which demonstrated that COX-2 selective inhibitors block edema and hyperalgesia following the inflammation inducing insult [19]. Prostaglandins play a direct role in sensing pain [14]. Prostaglandins sensitize the free ends of pain neurons and increase general sensitivity to pain [2]. These types of pain are referred to as hyperalgesia and allodynia. Independently, exogenous PG12 and PGE2 can cause pain and edema, and in vivo PGE2 antibodies block carrageenan induced hyperalgesia [20]. To mimic allodynia, PGE2 is injected into the subarachnoid space between the lumbar vertebrae. As a result, normal mice respond to touch with a paintbrush as though it were noxious stimuli, by squeaking and biting at the paintbrush . When the IP receptor gene is disrupted, mice no longer respond to pain caused by the intra-peritoeneal injection of acetic acid and are much less sensitive to the pain caused by heat in hot-plate studies [2]. EP1 and EP3 knockout mice exhibit decreased levels of inflammation induced pain [21], and in the EP3 mice, carrageenan induced inflammation was greatly reduced. These experiments provide evidence that prostaglandins mediate specific types of pain, sensitivity to pain, and generation of edema associated with painful inflammation. Prostaglandins may not always function as mediators of pain and inflammation. Using the carrageenan induced inflammation model, Gilroy et al. [22] examined the contents of the fluid exuded from the pleural cavity of treated rats. They observed an immediate burst of COX-2 expression (in the cells found in the exudate) that correlated with high levels of PGE2 and increasing levels of exudate volume. The resolution of the response is correlated with a decreased rate of fluid generation, infiltrating monocytes with high COX-2 activity, and a shift from PGE2 production to high levels of PGD2 and its metabolite 15-deoxy A12"4PGJ2 [22]. In this in vivo model, prostaglandins and high levels of COX-2 expression appear to be at least correlated with anti-inflammatory actions. The 15 deoxy AlZ'MPGJz prostaglandin product is associated with anti-inflammatory actions as a PPARy ligand and an inhibitor of NF-KB activation. In addition, COX-2 inhibition has also been observed to negatively impact the healing of lesions in the mucosa] layer of the stomach [23]. These papers suggest that under certain conditions, tissue specific COX-2 expression may play an anti-inflammatory role. Catalysis of Arachidonic Acid to Prostanoids Arachidonic acid is released from the cellular membranes by hydrolysis from glycerophospholipids by secretory or cytoplasmic phospholipase A2 (PLAz). In the first step of catalysis, arachidonic acid and two molecules of oxygen are converted to prostaglandin G2 (PGG2) through the cyclization of five central carbons of the twenty carbon Chain. In the second step, PGG2 hydroperoxide is reduced to the final product, prostaglandin H2 (PGH2). (Figure 1) The cyclooxygenase and peroxidase activities are located at different positions on the enzyme [9]. The membrane binding domain of the enzyme is comprised of four a—helical domains arranged end to end forming the opening of the cyclooxygenase site. When the substrate is completely inside the active site, the to end of the arachidonate is buried within a highly hydrophobic pocket at the top of the Channel, and the terminal carboxyl moiety interacts with polar and Charged residues near the opening of the active site. The heme group of the peroxidase site is oxidized, which generates a tyrosine radical that abstracts a hydrogen from Carbon (C) 13 of arachidonic acid in the cyclooxygenase site. This arachidonyl radical reacts with a biradical oxygen molecule to form an endoperoxide bridge between 011 and C-9. Further intra-molecular rearrangement of the radical results in the addition of another molecule of oxygen forming a hydroperoxide at C-15. The second step of the reaction occurs in a Cleft on the top side (relative to the membrane) of the enzyme. Through a yet undiscovered mechanism, PGG2 leaves the cyclooxygenase site and enters the peroxidase site where the reduction of the hydroperoxide at C-15 occurs. Initial activation of the heme group located between the peroxidase and the cyclooxygenase site requires a lipid peroxide-dependent oxidation. 10 The oxidized heme oxidizes the tyrosine, generating the tyrosine radical necessary for the cyclooxygenase activity [9]. Only one peroxidase turnover is required because the tyrosyl radical is regenerated independently of the peroxidase after each cyclooxygenase turnover. The cyclooxygenase continues to turn over until the enzyme suicide inactivates through an undefined autocatalytic mechanism in which the radical is transferred to an inappropriate residue that results in internal crosslinking and inactivation of the enzyme [9, 24]. The COX-1 and COX-2 enzymes share approximately 60% primary sequence identity [17], and their protein crystal structures are nearly identical; however, there are several small differences in the substrate binding domain and active site. The opening of the COX-2 active site is approximately 20% larger than COX-1. This is due to a Change from an isoleucine in COX-1 to a valine in COX-2 as well as several Changes in the secondary shell to residues with smaller side Chains. As a result, an arginine residue at the opening of the active site that is critical for stabilizing the carboxylate of arachidonic acid in COX-1 is displaced [9]. The increased size of the opening and hydrophobic side pocket are the discriminating factors for COX specific inhibitors [3]. COX-l is irreversibly inhibited by aspirin by the acetylation of serine 530 forming a prominent protrusion in the opening of the Channel, and this is thought to prevent the entrance of arachidonic acid to the active site [17]. Acetylated COX-2 still forms prostanoid products; however, because of the misalignment of C-13, most of the products formed have only the C-15 hydroperoxide but not the bicycliC peroxide. The larger more flexible substrate Channel and small internal pocket have been 11 Phospholipid Phospholipase A2 ._ coon Arachidonic Acid 102 __ __ coon o 0&2 \04, Cyclooxygenase i 02 fl _ _ COOH o-\ 0 ‘ ~o—o- , ~ _ coon o Peroxidase 00” Synthases / 0H \ Prostacyclin Prostaglandins Thromboxane Figure l. Arachidonic Acid Metabolism 12 exploited for the development of COX-2 selective inhibitors such as rofecoxib (Vioxx®) and celecoxib (Celebrex®) whose structures occupy this unique side pocket. Both enzymes carry out identical catalytic actions, so why are there two isozymes? One hypothesis is that the COX isozymes are part of discrete biosynthetic pathways involving the coupling of distinct pools of arachidonic acid, specific phospholipases, and downstream prostaglandin synthases. There are at least 16 PLAz proteins that are grouped by size, substrate specificity, calcium dependence, and structural homology and fall into three general categories [25]. The cytoplasmic phospholipases (CPLA2 ) are calcium dependent and arachidonic acid specific; the secretory phospholipases (sPLAz) are also calcium dependent but are not specific for arachidonic acid, and the intracellular PLAzs (iPLAz) are neither calcium dependent nor arachidonic acid specific. Nearly any stimulus that activates MAP kinase signaling or elevates intracellular calcium concentrations is sufficient to activate the CPLAzs [26]. Activated CPLA2 translocates from cytoplasm to the exterior of the endoplasmic reticulum where it specifically releases arachidonic acid from the membrane. The initial burst of prostaglandin production, 10 to 60 min post treatment, is metabolized by the constitutive COX-1, and the delayed prostaglandin production is a result of newly synthesized COX-2. Early studies suggested that different pools for arachidonic acid and distinct PLAzs were functionally if not physically linked to either COX—l or COX-2. The sPLAz is thought to be coupled to COX-2 in the late phase of prostaglandin production; however, this may be due to the temporal expression of sPLAz and the specific activity of COX-1 and COX-2 at low substrate concentrations. In vitro, at high substrate concentrations, both enzymes have identical Km values [9]. However, in vivo and at very 13 low substrate concentrations (0.05 - 2 M) the apparent Km values of the two isoforms appear to be different. This is because the Km is influenced by peroxide concentrations. COX-l requires peroxide concentrations that are about 10 times greater than those of COX-2. Because of these two effects, arachidonic acid is preferentially metabolized by COX-2 when there are low concentrations of arachidonic acid present. [3, 6, 8, 9, 17, 27- 29]. Physical interactions between the PLA2s and downstream synthases have not been observed, but since these proteins localize to the membrane of the endoplasmic reticulum, it is possible that weak interactions do exist [30] [9]. The recently identified prostaglandin E2 synthase (PGES) [31] isozymes appear to be functionally coupled to specific COX isozymes. The cytosolic PGES is constitutively expressed in many cell types and appears to be coupled to COX-1 in the early phase of prostaglandin synthesis. In contrast the membrane-associated PGES is inducible by inflammatory stimuli and appears to be coupled to COX-2 in the late phase of prostaglandin production [32, 33]. In response to inflammatory stimuli, there appears to be a shift from the production of thromboxane B2 (TXB2), prostaglandin D2 (PGD2) and prostaglandin I2 (P612) to increased levels of prostaglandin E2 (PGE2) and PG12 [34-37]. 14 Historical Perspective As early as 1975, Lawrence Levine began using methylcholanthrene-transforrned BALB/3T3 fibroblasts in an attempt to study the mechanisms of initiation of the biosynthetic process resulting in prostaglandin production. His idea was that cells grown in culture might be a simpler model than using subcellular fractions of tissues, intact organs, or tissue slices. Within a short time, Levine discovered that prostaglandin production could be enhanced by serum or phorbol-ester (TPA) stimulation, and that these “stimulations” were dependent on protein and mRNA synthesis [38-40]. Inducible COX activity was also Characterized in response to growth factors, LPS, and IL-1. In retrospect, these detailed studies were the first to Characterize inducible COX-2 activity. In some cases, the use of COX-2 cross reactive antibody even allowed the Characterization of induced COX-2 protein in tissues with very low constitutive COX activity [41-43]. The biochemical identification of a second COX activity was published by Robert Gorman’s lab at the Upjohn company in Kalamazoo, Michigan. Alice Lin et al. reported that Platelet Derived Growth Factor (PDGF) stimulated, serum starved fibroblasts resulted in bursts of PGE2 synthesis that began 10 min post stimulation and peaked after 2 hr. NIH3T3 cells constitutively expressed COX, and PDGF increased COX mRNA levels after 2 hr even though protein levels remained nearly constant throughout the experiment [44]. Gorman observed that arachidonic acid treated, serum starved NII-I3T3 cells could synthesize PGE2 in the absence of mitogen stimulation even more rapidly and potently than when stimulated with PDGF, and that unstimulated arachidonic acid dependent PGE2 production could be blocked with aspirin pre-treatment. In the aspirin 15 treated cells, arachidonic acid stimulated PGE2 synthesis was not recovered. This implied that the cells that constitutively expressed a COX activity could be blocked by aspirin. Pretreatment with aspirin (an irreversible COX-l inhibitor) and stimulation with PDGF resulted in a PGE2 response that took two hr to develop and peaked at three hr. In these cells, cyclohexamidel completely abolished the delayed synthesis of PGE2. While several very rational explanations were postulated for their observations, the authors also speculated wildly that “PDGF induces the expression of a second PGHS2 that is coupled to the PDGF receptor and whose mRNA is not readily detected by our probe. Thus, the possibility remains that there is differential mRNA splicing or even a second gene.” In the late 19803 a number of labs were hoping to find a cancer cure using the power of molecular biology. To this end, a technique called “subtractive and differential screening” was employed to identify nuclear targets for mitogenic signal transduction pathways. Shortly after mitogenic stimulation, immediate-early genes were induced, causing rapid increases in their mRNA. Since up-regulation of these genes didn’t require prior protein synthesis, they were presumed to be the necessary requirements to drive quiescent (G0) cells into the first stage of the cell cycle (G1). Many of the genes first identified were transcription factors or secreted proteins, but most of the Cloned genes had no known function [45, 46]. Sequencing and Characterization of the gene products were the rate-limiting steps for discovery. In November of 1988, Daniel Simmons et al. published his work on genes with unknown functions (or sequences) that were induced in a temperature sensitive Rous Sarcoma Virus (v-Src) infected Chicken embryo fibroblasts (CEF), that could also be ' Cyclohehamide is a protein synthesis inhibitor and is used to determine if an enzyme activity is the result of proteins that are currently in the cells or if it is the result of a newly synthesized protein. 16 induced by TPA, and serum stimulation in normal CEF [47]. Simmon’s CEF Clone 147 mRNA was nearly undetectable in resting cells and was induced by one hr, peaking at two hr at “superinduced” levels with an apparent size of about 5 kb. Its function was unknown and was one of 6 Clones awaiting further Characterization. Within months of Gorman’s observation of a second inducible COX activity, Glenn Rosen et al. in Michael Holtzman’s lab at the University of Washington (St. Louis, Mo.), observed that induced COX activity that didn’t correlate with either the COX-1 protein levels or with the 2.8 kb COX-1 mRNA expression patterns in sheep tracheal epithelial cells. Northern blotting at low stringency conditions using two non-overlapping COX—l probes revealed a 4 kb mRNA that was expressed basally at very low levels, was tissue specific, and whose expression pattern followed the increase in COX enzyme activity. They hypothesized that the 4 kb mRN A was derived from a distinct COX related gene. Curiously, the Northern blots from the Gorman lab, which had a number of non- specific bands, had one that appears to be about 4 kb and follows the pattern of PGE2. Rosen et at. set the standard for the identification of the proposed COX-2, saying that although the 4 kb mRNA likely represents an explanation of their excess COX activity, “verification that the larger mRNA encodes for a cyclooxygenase will require molecular Cloning of the gene and expression of a functional protein product in cells lacking endogenous activity.” [48] J ia-Wen-Han et al, from the labs of Donald Young and Ian Macara (University of Rochester, NY), published a paper on the persistent induction of cyclooxygenase in v-SrC transformed BALB/c 3T3 fibroblasts [49]. Their model for oncogenic transformation was based on the idea that either cells were transformed by the expression of transformation 2 Prostaglandin endoperoxideH synthase (PGHS) l7 specific genes that are not normally expressed, or they were a result of continued expression of genes that are only transiently induced during mitosis. Instead of using differential subtractive hybridization, they used “giant two-dimensional electrophoresis” to look for rapid Changes in protein abundance after activation of a temperature sensitive Rous sarcoma virus infected fibroblast. What they thought they found was a post- translational modification of the COX-1 gene product. This was a unique finding: out of >3000 polypeptides resolved by their method, this was the only one that appeared as a doublet and was inducible. We now know that COX-1 has an apparent molecular weight of 72 kDa, and that COX-2 generally is separated as two bands that are 72 and 74 kDa as a result of incomplete glycosylation. The reason BALBC 3T3 fibroblasts were used was because they have only a very low background COX-1 expression. In contrast, NIH3T3 fibroblast cells have a relatively high constitutive level of COX-1. Since dexamethasone and indomethacin treatment did not cause reversion of the fibrosarcoma, but dexamethasone blocked COX-1 expression and indomethacin blocked PGE2 synthesis, the authors concluded that a post-translational modification was probably involved with the role of COX and its function in its unregulated state. Others suggested that post- transcriptional modification was behind the mechanism of the differential regulation of induced COX gene expression, based primarily on the ability of dexamethasone to block serum induced expression without affecting enzyme activity, mRNA abundance, or the presence of the 72 kDa protein species. On the day after Christmas in 1990, Daniel Simmons, at Brigham Young University, communicated an exciting find to PNAS. Weilin Xie et a1. had sequenced their CEF-147 Clone and discovered that it was 59% identical to the ovine COX-l [50]. In 18 vitro transcription of their 4.1 kb mRNA resulted in a 70 kDa protein, and co- translational glycosylation produced a 79 kDa protein. This was comparable to the predicted 68 kDa ovine COX-1 and the potentially novel 74 kDa protein that had been observed. There was one significant problem. COX-l had been Cloned and sequenced from ovine, murine, and human sources, but not from Chicken. Was this a Closely related protein or the Chicken homolog to ovine COX? The authors argued that they had Cloned the inducible COX gene and cited several very significant differences between their gene and the COX-1 gene. First, this new gene was post-transcriptionally regulated by the splicing of an intron in the 5’ region of the mRNA that blocks translation. Second, the new gene had an unusually long 3’ untranslated region (UTR). The 3' UTR makes the transcript 4.1 kb long instead of the usual 2.8 kb COX-1. The fibroblast experiments of Rosen et al. had implicated a 4 kb mRNA as a possible COX-l related gene product in mouse fibroblasts. Weilin Xie et al. [50] also noted that the 3' UTR also contained a number of Shaw Kamen RNA instability sequences that are not present in the Cloned COX-1 3' UTRs. This data certainly implied a possible homology to the 4 kb mouse mRNA, which helped to suggest the existence of two COX isoforrns, but until the mouse transcript was Cloned, the identity would still be uncertain. While Simmons had been using CEF for his subtractive differential screening experiments, Harvey Herschman at UCLA (Los Angeles, CA) was using Swiss 3T3 fibroblast cells and was screening through his own group of induced immediate early genes with unknown functions. Dean Kujubu et al. reported that they had cloned a number of TPA inducible genes [51]. TPA inducible sequence #10 (TIS-lO) was induced by TPA, forskolin, and serum in Swiss 3T3 cells. The expression of TIS-lO appeared to 19 \ be cell type restricted, in contrast to most of the other immediate early genes they had identified, but its function was unknown. The rate of progress in Characterizing the new COX related transcript moved very quickly in 1991. By July, the Donald Young lab from Rochester submitted a publication in which they correlated serum stimulated and glucocorticoid regulated levels in both immunoblot and immunoprecipitated COX with an abundance of the 4 kb mRNA species visualized with low stringency washing of the COX-1 probe [52]. They reported Cloning the inducible COX CDNA and published an 80 amino acid translation of their preliminary sequence data, which was 95% identical to the translated sequence published by Simmons and was 59% identical to ovine COX—l. JoAnne Richards produced polyclonal antibodies that recognized an inducibly expressed and tissue restricted COX enzyme and was able to show that the different molecular weight variants of COX observed in rat ovaries were antigenically distinct. Using antibody against a distinct prostaglandin synthase, Jean Sirois et al. purified and determined the amino-terminal sequence of a peptide that was nearly identical to the translation of the previously identified TIS—lO and CEF-147 sequences from mouse and Chicken cells, respectively [53, 54]. Bradley Fletcher et al., from Herschman’s lab, published a second paper in October of 1991 that included the gene structure of COX-2 and expression data showing that the TIS-10 gene conferred cyclooxygenase and peroxidase activity on transiently transfected COS cells [55]. They also fused several portions of the genomic DNA 5’ of the transcriptional start site to a luciferase reporter plasmid to show that the luciferase activity could be induced in a pattern similar to that of the new COX in TPA stimulated 20 NIH3T3 cells. A month later, another report from the Herschman lab characterized the enigmatic effects of dexamethasone on COX with their ability to differentiate between COX-1 and -2 [56]. Herschman is generally credited with the discovery of COX-2 because his lab was the first to produce a genomic Clone and publish almost all of the CDNA. They were also the first to use a COX-2 specific probe for Northern blots to demonstrate a time course and cell type specific expression. In December of 1991, Donald Young’s group published their results on the heels of Herschman’s group with the complete sequence of the COX-2 [57]. Ryseck et al. from Bristol Myers Squibb in Princeton, NJ, published the most complete early Characterization of COX-2 in a smaller journal six months later [58]. In their paper, they not only published the CDNA sequence, gene structure, time course of mRNA, and protein expression, but also its Chromosomal location and initial data on the pharmacological Characterization of COX-2 expressed in baculovirus. At that point in time the expression of proteins in baculovirus required an exceptional molecular biologist, a near clairvoyant microscopist3, and a lot of time. Personal communications suggested that the Bravo R. group had the COX-2 protein Cloned sequenced and expressed as early as 1989. The anticipated pot-of—gold at the end of the rainbow awaiting a COX-2 selective inhibitor triggered a world Class race to develop candidate drugs [28]. In 1999 Searle released Celebrex, the first FDA approved COX-2 selective inhibitor, and Merck followed closely with Vioxx. While the pharmaceutical companies were Chasing the pot- 3 Both the transformed virus and native virus appear identical, but because the coat protein is disrupted in the transformed virus, a decrease in refracted light can be observed. This optical property was used for the selection of positive Clones in plaque lysis experiments. The slight difference was often very difficult to detect. 21 l of-gold, the University of Rochester and Donald Young were waiting for it to arrive. In 1992, they filed a patent for the method of inhibiting COX-2. In April 2000, after Searle announced a blockbuster $1.5 billion in first year Celebrex sales, the University of Rochester announced that it had been awarded the patent. While the lawyers fight for the ownership of COX-2, the scientific community will probably always recognize the combined efforts of Herschman, Simmons, and Richards in their discovery of COX-2. 22 Transcriptional Regulation Techniques used in transcriptional regulation studies. Several key types of experiments used to study transcriptional regulation include promoter activity assays, Electrophoretic Mobility Shift Assays (EMSAs), supershift EMSAs, and co-transfection. In a promoter activity assay, a reporter gene is fused to a segment of DNA from the upstream region proximal to the transcriptional start site. Luciferase is a popular reporter gene because of its simplicity and sensitivity. Chloramphenacol acetyl transferase (CAT), and B-galactosidase are also used as reporters, but assays for their activity either require the use of radioactive substrates or are not very sensitive to lower promoter activities. A general strategy is to identify putative cis-acting elements via a database search, and then to design deletion constructs that progressively remove potential transcription factor binding sites. When the amount of activity drops significantly with the deletion of a region, that deleted region is considered relevant for the transcriptional activity of the gene. Once the minimal promoter is defined, mutation of the remaining consensus sequences can be used to narrow the search for essential cis- acting elements. With the target area more refined, EMSA experiments can be used to corroborate data from promoter activity analysis. EMSA experiments are used to locate specific regions of the promoter where proteins bind. Double stranded 32P labeled DNA from relatively short regions of the promoter (20 to 200 bases) are incubated with extracts of nuclei prepared from cells in the induced and un-induced conditions. The samples are then separated on a non-denaturing acrylamide gel. If no protein binds the probe, then all of the DNA migrates though the gel. If binding does occur, then a shift in 23 the mobility of the probe is observed because the protein DNA complex moves much more slowly through the gel than the probe alone. In competitive EMSA experiments, a large piece of labeled DNA is competed with shorter pieces of DNA from the same region. The short unlabeled pieces are used in concentrations much higher than the larger labeled probe so that protein binding will be competed from the labeled probe resulting in the loss of the low mobility complex observed on the gel. Competitive EMSAs are also used to show that binding to a sequence is specific rather than non-specific. Supershift EMSAs are experiments designed to identify specific proteins bound to a DNA probe. For example, the CAMP Response Element (CRE) consensus sequence can be bound by the CRE binding protein (CREB), several Activator Transcription Factors (ATFs) and by Activator Protein-1 (AP-1). In the COX-2 promoter there is a CRE that appears to be functional, so supershift EMSAs are performed with antibodies against CREB, ATP, and AP-l proteins. If the mobility of the probe protein complex decreases (shifts up even further) or if the binding to the probe is blocked with one of the antibodies then that protein is likely to be bound to the DNA probe. The decrease in mobility is caused by the increase in the size of the complex bound to the probe. Alternatively, the antibody may interfere with DNA binding and block the interaction with the probe. Co-transfection experiments are used to saturate the cell of interest with a protein to determine whether that specific protein will cause an effect on a promoter reporter plasmid. If a certain transcription factor is suspected to be involved in a response, its overexpression is expected to cause an increase in the transcription measured by a C0- transfected promoter reporter construct. Dominant Negative (DN) experiments are used as a control. If co-transfection of a transcription factor upregulates promoter activity, then 24 co-transfection of a dominant negative version of the same transcription factor should downregulate the response. Dominant negative proteins are mutants that may lack a DNA binding domain, a functional activation domain (e.g. an essential serine that is phosphorylated is mutated to a non-phosphorylated amino acid), or trans-activation activity (may not be able to phosphorylate its target molecule.) With these experiments one can define the active region of a promoter and use mutations to show loss of function to confirm the significance of a response element. Then, using EMSAs, one can show that the response element is bound specifically and, if possible, identify the protein in supershift experiments. If identification is possible, the co-transfection of that trans-activating factor should be expected either to cause increased promoter activity in the absence of stimuli or to greatly enhance the effect of a stimulus. Conversely, DN co-transfection should block stimulus induced promoter activation. By using upstream activators of the putative transcription factor and their DN counterparts, or specific inhibitors of signaling pathways, the general mechanism between stimuli and gene activation can be modeled. This is a basic strategy that has been employed in studying the promoter and transcriptional regulation of COX-2. On a typical gene, the core promoter is located immediately upstream of the transcriptional start site. This core promoter region is where the general transcriptional machinery and RNA polymerase H bind to the promoter to initiate transcription at the correct location. In the upstream region adjacent to the core promoter is the regulatory promoter. This is where activating proteins bind to activate gene transcription. Enhancer regions are also present, but they can be located farther upstream or downstream of the core promoter. Activation of the gene occurs when the general transcription machinery 25 is recruited to the promoter by the activating factors in the regulatory promoter. In most respects, the COX-2 promoter is an ideal promoter to study. The core promoter contains a TATA box or TATA-like consensus sequence, and initial work has shown that the proximal region is capable of driving transcription of a reporter gene. (Figure 2) In contrast, the COX-1 promoter lacks a TATA sequence, and the upstream region provides only minimal activation of reporter constructs in fibroblasts [59]. In some extreme cases, the apolipoprotein B gene for example, the promoter lies more than 55 kilobases upstream from the coding region [60]. Investigations into the mechanism responsible for the expression of an induced gene need to consider several critical regulatory points. Regulation can occur at both transcriptional and post-transcriptional levels as well as at the level of translation. Before studying the transcriptional regulation of a gene, it is essential to determine the mode of regulation. DeWitt et al. showed that serum stimulation of fibroblasts results in a transient increase in the synthesis of a labile COX-2 mRNA that corresponds to a transient increase in protein expression and the previously observed increase in cyclooxygenase activity [61]. With the mode of regulation defined, studies of transcriptional initiation were logical to pursue. 26 CRE-2 NF-KB C/EBP-2 CRE-1 TATA -434/-428 -401/-393 -93/-85 -59/-52 -30/-25 0-0-22 C/EBP-l AP-l E-Box Munne COX'Z Fromm“ -138/-130 -73/-61 -53/-4s CRE-2 NF-KB NF-KB C/EBP-l CRE-1 TATA -510/-504 -388/—380 -223/-214 -133/-124 -60/-53 -32/-29 AP-l E-Box Human COX-2 Promoter .74/.62 .55/.50 Figure 2. Response elements of the human and murine COX-2 promoters 27 Transcriptional Regulation of COX-2 in Fibroblasts The first studies of COX-2 regulation were performed in NIH3T3 fibroblasts that were transfected with a temperature sensitive pp60v's“: expression plasmid [62]. Promoter reporter constructs containing the region between —963/+70 or —371/+70 relative to the transcriptional start site, responded to both serum and TPA treatment when transfected into NIH3T3 cells [55]. The longer construct had only about 20% of the activity of the shorter construct. This was attributed to upstream negative regulatory elements. Similar experiments were also performed with the chicken COX-2 promoter in CAT reporter constructs in NIH3T3 cells, but the results were less dramatic and difficult to interpret Clearly [63]. In the fibroblast model, the minimal promoter necessary for COX-2 activation by v-Src transformation involves only the first 80 bases of the promoter [62]. Deletion of the region from —80 to —40 resulted in a near complete loss of promoter activity. This region of the promoter contains a putative overlapping CRE and E-box. Overexpression of dominant negative CREB (no PKC phosphorylation site) or dominant negative myC (inactivated trans-activation domain) either blocked or significantly inhibited promoter activity, respectively. Mutation of the CRE blocked nearly all of the promoter activity, while the E-box mutation only blocked about half of the promoter activity. In EMSA and supershift assays, anti-CREB shifted a complex associated with the CRE and recombinant CREB bound to the EMSA probe. Methylation interference assays verified that there are protein DNA contacts throughout the region containing the overlapping CRE/E-box. Data from competitive EMSAs tentatively excluded the E-box as a participant in the response. Since v-Src is a protein tyrosine kinase that mimics the 28 activation of growth factor receptors resulting in ras mediated MAPK signaling, dominant negative ras was co-transfected with the promoter construct and v-Src. Dominant negative ras blocked nearly all the promoter activity, indicating that COX-2 induced expression is mediated by ras. Xie et al. [62] mention that dominant negative CREB repressed promoter activity cannot be rescued by overexpression of CREB . Thus, CREB may not be the transcription factor mediating COX-2 induction by v-Src. An alternate hypothesis is that CREB is activated, but excess CREB may limit transcriptional activation by competing for a limiting cofactor necessary for interaction with the core transcription complex. Binding to the CRE/E-box response element was not inducible in cells grown in permissive and non-pennissive temperatures for the expression of v-SrC. This mechanism is consistent with a role for CREB. CREB can bind to DNA as a dimer but is inactive until it is phosphorylated at a critical serine residue (Ser133). Another interesting observation that was difficult to explain was that the region between —64 and —38, which contains the putative regulatory region, could not confer v-Src inducibility to a thymidine kinase promoter. This suggested that the CRE/E-box is necessary but not sufficient for activation. One year later, a new v-SrC mediated mechanism was proposed by Xie et al. who showed that both CREB and DN-CREB block transcription of a COX-2 reporter and concluded that CREB was not involved in the v-SrC response [64]. Antibody to the Jun transcription factor could supershift the CRE/E-box probe and v-Src transformation potentiated Jun Kinase (JNK) activity. Based on evidence from promoter reporter co- transfections with different MAP kinases and dominant negative kinases, v-Src activation 29 of COX-2 was proposed to occur via two convergent pathways, 1) the “specific” pathway from ras/MEKK-l/JNKK (MKK4)/JNK leading to c-Jun phosphorylation, and 2) a pathway that was proposed to affect secondary response genes via raf- l/MAPKK/ERK1&2. This second pathway was hypothesized to result in the transcriptional activation of the C-Fos gene, which can form heterodimers with Jun resulting in more potent and stable activation of AP-l transcription factors at the CRE site. There is a curious inconsistency between the Jun transcription factor model and the previous CREB model. After being phosphorylated, Jun transcription factors bind to regulatory regions on promoters, yet binding to the CRE was not observed to be inducible. It is possible that Jun and CREB interact at two distinct sites that are very Close together on the promoter and that both are required for transcriptional activation. Overexpression of functional CREB may compete for limiting amounts of cofactors necessary for its interaction with the transcriptional machinery, while overexpression of the dominant negative CREB may compete for binding sites. In either of these conditions, activation of the COX-2 promoter could be limited, so these experiments may not exclude CREB as a functional component of the transcription activating complex. Early Characterization of COX expression revealed that many different mitogenic and pro-inflammatory stimuli could cause an increase in both cyclooxygenase activity and mRN A levels. PDGF and serum (which contains undefined growth factors) were also shown to increase COX-2 promoter activity. In experiments similar to those used with the v-Src transformed fibroblasts, Xie et al. [65] demonstrated the necessity of the CRE element and presence of similar signaling pathways leading to COX-2 activation. 30 Collectively, these experiments formed the central dogma of COX-2 gene activation. COX-2 gene activation is primarily mediated by activation of the Classical MAPK signaling pathway that terminates in the activation of AP-l (Jun) through its interaction with a required CRE cis-acting element. However, more recent studies with human foreskin fibroblasts suggest that the process is much more complex [66]. These experiments indicate that salicylic acid inhibits transcriptional activation acting through a CAAT enhancer binding protein (C/EBP) response element. Additional experiments using lL-lB, TNF-a, or PMA as stimuli in promoter activity assays revealed that, in addition to the CRE-1, the C/EBP-l and NF-KB sites are also required for transcriptional activation. Regulation beyond the level of transcription can also be more complex. In synovial fibroblasts, IL-l stimulates the prolonged expression of COX-2 that is mediated by stabilization of mRNA, involving p38 activation [67]. 31 Transcriptional Regulation of COX-2 in Epithelial cells Cancer and prostaglandin production are linked through a mechanism that is still not Clearly defined. It has been known for some time that people who take aspirin or other nonsteroidal anti-inflammatory drugs (NSAle) have a 40% - 50% lower risk of colorectal cancer when compared with people who are not taking the drugs [68, 69] COX—2 inhibition significantly inhibits TPA-induced mouse skin tumor formation and azoxymethane induced rat colon tumorigenesis [70]. Aberrant expression of COX-2 has been observed in colorectal cancers, and in multiple cancers of the skin, head and neck, lung, breast, and stomach. COX-2 overexpression is associated with enhanced invasiveness and suppression of apoptosis [69, 71, 72] [73]. For these reasons, normal and transformed epithelial cells have been used to study COX-2 transcription. Regulation of COX-2 transcription in epithelial cells shares some similarities with fibroblasts. Subbaramaiah et al. used transformed murine mammary epithelial cells to study the induction of COX-2 [74]. Transformation of murine mammary epithelial cells with ras or v-src prodigiously increased COX-2 mRNA and protein levels. To Characterize the mechanism of transcriptional regulation, the authors used the promoter activity assays with 5’ deletion constructs containing —40/+3, -80/+3, and —962/+3 of the murine COX-2 promoter. The authors’ data showed that the long construct had approximately 7 to 10 times the promoter activity of the —80/+3 construct, yet the authors observed that “promoter activity was localized to a region between —80 and —40.” Co- transfection with Jun enhanced promoter activity and dominant negative ras blocked induction, but to a lesser extent compared to the longest construct. This would tend to indicate that a larger portion of the promoter is necessary for the response, although there 32 is not a substantial loss in inducibility. While these results indicate that MAP kinase pathway can activate transcription from the minimal region of the COX-2 promoter, activation in epithelial cells is mediated by a more complicated mechanism. Kim et al. set out to Characterize the cis- and trans-activating factors in mouse skin squamous cell carcinoma JWF2 cells that constitutively express COX-2 [75]. Through a convincing set of experiments, they showed that the E-box and C/EBP sites act as positive regulatory elements, and binding sites for USPS and C/EBP transcription factors, respectively. In competitive EMSA assays, the overlapping E-box/CRE probe was competed by the same probe with a mutation in the CRE but not with a mutation in the E—box and general CRE consensus sequence did not compete binding for the native COX-2 E-box/CRE probe while a general consensus USF oligonucleotide did. In supershift assays, USF—1 and 2 antibodies shifted complexes and antibody to CREB, ATP-2, C-Jun, and C-Myc had no effect. Mutation of the C/EBP site dramatically reduced promoter activity and specific binding to the C/EBP probe was observed. C/EBP-S and C/EBP-B were supershifted, and overexpression of C/EBP-S doubled promoter activity. They also showed that C/EBP-S mRNA is upregulated in JWF2 cells where it is not present in normal epithelial cells. To further test the involvement of C/EBP transcription factors, a dominant negative C/EBP transcription factor (CHOP-10) was co-transfected with the COX-2 promoter reporter. Overexpression reduced promoter activity in a dose dependent manner. Because mutations of either site bound by USF or C/EBP decreased promoter activity, the COX-2 promoter appears to require both sites for full promoter activity. 33 More recently, the ectopic expression of Wnt-l has been linked to induced COX-2 in C57MG and RAC311 murine mammary epithelial cells [76]. Wnt-l encodes a secreted protein that functions as a ligand for the Frizzled family of seven transmembrane receptors. Wntl signaling leads to the stabilization of cyt0plasmic B-Catenin, leading to the formation of B-CateninOTCF (T-Cell factor) complexes and transcriptional activation. In cells expressing Wnt-l COX-2 transcription, mRNA and protein levels are increased. Characterization the mechanism of Wnt-l mediated COX-2 expression was done in a human embryonic kidney cell line to observe the effects of co-transfection with B-C3tenin and a handful of Ets transcription factors[76]. B-catenin stimulated only very weak COX- 2 promoter activity, indicating that it acts on an intermediary factor rather than directly with the native promoter. Of the Ets transcription factors tested (Pea3, ER81, ERM, ETS- 1, and ETS-2), co-transfection of Pea3 resulted in a high level of promoter activity. Pea3 expression is induced or at least present in Wnt-l transformed cells and mammary tumors from Wnt-l transgenic mice [76]. Characterization of cis-acting elements that respond to Pea3 was quite interesting. Deletion constructs of the human promoter from -1432, -327, and —220 all had similar levels of promoter activity. Deletion of the region between -124 and —220 resulted in a nearly complete loss of promoter activity. In experiments that were intended to exclude the NF-KB, C/EBP and CRE sites as possible candidates for Pea3 interaction, the authors were surprised to find that mutation of the C/EBP site completely eliminated promoter activity. The human COX-2 promoter has Pea3/Ets consensus sequences at —859 and -400, but not within the —220 to —124 region. However, there are several forward or reverse “GGAA/I‘” core Ets box sequences. Two possible mechanisms could account for the observation. Pea3 may actually bind to the C/EBP site and activate 34 transcription or, more likely, Pea3 binds to a C/EBP proximal site and functions in a synergistic fashion to activate transcription[76]. To test the model, the authors overexpressed C/EBP-or, B, and 5 with the COX-2 promoter reporter and observed trans-activation with the C/EBP-a and 8 isoforrns, and decreased promoter activity with C/EBP-B[76]. In a second experiment, the authors also overexpressed Pea3 with DN—C/EBP (LIP) in increasing amounts and observed a dose dependent decrease in promoter activity. While the authors favored a model of Pea3 and C/EBP transcription factors working together from proximal sites because of the differences in C/EBP and Pea3 consensus binding sequences, their experiments do not discriminate between the two proposed modes. C/EBP could activate transcription independently of Pea3. The DN-C/EBP (LIP) has a functional DNA binding domain and a deleted trans-activation domain, so it probably competes with Pea3 for binding at a common site. Undoubtedly, this model will be explored further because of its potential utility in explaining the mechanism responsible for COX-2 expression in tumor types where inappropriate Wnt-l, B-Catenin, and Ets transcription factor regulation are observed [76, 77]. COX-2 expression in epithelial cells is regulated in response to a number of tumor promoting compounds. Some of the more interesting compounds include benzo[a]pyrene and caffeic acid phenethyl ester in oral epithelial cells [78, 79], dihydroxy bile acids and curcumin in colonic epithelial cells [80, 81], and ceramide and resveratrol in mammary epithelial cells [82, 83]. Resveratrol is a phytoalexin found in grapes and their fermentation products, curcumin is responsible for the yellow color in turmeric, and caffeic acid phenethyl ester are all phenolic antioxidants that act like glucocorticoids to 35 downregulate COX-2 expression. Sphingomylenase, dihydroxy bile acids, and benzo[a]pyrene all up regulate COX-2 at least in part though their activation of MAPK signaling. Retinoids suppress EGF induced COX-2 transcription and expression, as well as prostaglandin production, and promoter activity in oral squamous carcinoma cells [84]. The mechanism by which retinoids block COX-2 transcription is fairly interesting, because retinoids antagonize AP-l mediated gene trans-activation. Since the COX-2 promoter lacks binding sites for the retinoic acid nuclear receptors in the proximal promoter region, it is unlikely that they interact directly with the COX-2 promoter. If ligands for nuclear receptors are added to an appropriate system, the genes controlled by the receptor are expressed. At the same time, most genes controlled by AP-l are downregulated. If AP-l is reactivated, then the genes controlled by the nuclear receptor are downregulated [85]. Chris Glass’s Cell paper "A CBP integrator complex mediates transcriptional activation and AP-l inhibition by nuclear receptors" tested a hypothesis that the two systems competed for limiting amounts of CBP [86]. This competition provided a mechanism for mutual antagonism of these two Classes of transcription regulators. This topic is the subject of a broad and intense field of research and has been reviewed by Chris Glass and others [85, 87—89]. With a basic understanding of this new and popular model of transcriptional co-activators competed between nuclear hormone receptors and AP-l transcriptional activators, Dannenberg’s lab set out to tie this mechanism to their previously observed results. They had observed that PMA mediates COX-2 induction through AP-l trans-activation and that glucocorticoid and glucocorticoid-like compounds, as well as 36 PPAR ligands, had antagonized COX-2 induction. They showed that in PMA treated human epithelial cells, the PPAR-y ligands, troglitazone, ciglitazone, and 15d-PGJ2 all blocked COX-2 mRNA induction [90]. PMA induced PGE2 production, and COX-2 protein levels were reduced to near background levels. In promoter activity assays, PMA induced activity was reduced to near background levels by the addition of either ciglitazone or 15d-PGJ2_ Co-expression of a DN—PPAR-‘y, deficient in its ability to interact with CBP, or co-transfection of a decoy PPAR Response Element (PPRE) restored promoter activity in the presence of the PPAR-y ligand. Both treatments inhibit trans-activation of PPAR responsive genes, resulting in an increase in free transcriptional co-activators. Co-transfection of the promoter reporter with NF-KB, CREB or C/EBP-or had no effect on ciglitazone blockade of PMA stimulated cells, but co-transfection of c- Jun or CBP each reduced the effect of the PPAR-y ligand. Co-transfection of CBP and C- Jun together restored nearly all of the PMA induced activity. While it was already known that AP-l participates in the activation of the COX-2 gene and that AP-l mediated gene activation requires CBP, this paper showed for the first time that a transcriptional co- activator participates in the transcriptional activation of COX-2. 37 Transcriptional Regulation of COX-2 in Endothelial Cells Jones et al. observed that PMA, IL-lB, TNF-a, and LPS induced COX-2 expression and prostaglandin synthesis in Human Umbilical Vein Endothelial Cells (HUVEC) [91]. Based on these observations, Inoue et al. investigated whether COX-1 or COX-2 was induced in Bovine Atrial Endothelial Cells (BAECs) [92]. While COX-1 was unaffected by the treatments, COX-2 mRNA was synergistically induced by LPS and TPA. They observed that the promoter activity induced from constructs containing -327/+59 or —1432/+59 of the human promoter were nearly identical. Their deletion constructs revealed that the NF-KB element (-223/-214) was necessary for full activity. Mutation of the C/EBP element resulted in a decrease in promoter activity, while the CRE—1 mutation had little effect, and a double mutation in both the C/EBP and NF-KB cis—acting elements reduced promoter activity to near background levels. These results were excitingly inconsistent with nearly all the previous observations. First, in fibroblasts the —1432/+59 construct had much less promoter activity than —327/+59 [55], indicating the presence of upstream repressors, but in epithelial cells, —1432/+59 had much more activity than -327/+59 [74], indicating that the authors Clearly missed investigating an active upstream activator, but in BAEC [93] both constructs had the same activity indicating that —327/+59 actually contained the complete, necessary, regulatory region for induction in this system. Second, Mutation of CRE-1 usually had resulted in either significant decrease or ablation of promoter response, but here mutation of the CRE-1 resulted in a negligible decrease in promoter activity. EMSA experiments in this system were even more unusual [92]. In BAEC nuclear extracts from unstimulated cells, very little binding to the CRE-1 probe was 38 observed; however, when the cells were transfected with C/EBP-or, B, or 8, significant specific binding was observed. In co-transfection experiments, C/EBP-8 induced COX-2 promoter activity in the absence of LPS, and mutation of CRE-1 blocked C/EBP-5 induced promoter activity, while mutation of the C/EBP response element only reduced promoter activity by about one-half. So, CRE-1 does not appear to be required for LPS and TPA induced COX-2 in BAEC, but it is required for C/EBP-S dependent induction. It may be possible that an unusual C/EBP and CREB heterodimer, which has been observed with the human IL-lB promoter [94] in response to LPS, functions is some capacity in this system. This LPS-stimulated BAEC model shows that in one cell type the CRE-1 site is necessary under some conditions and is unnecessary in others. In addition, it was the first Clear example of the NF-KB site being necessary for activation of a COX-2 promoter reporter. The NF-KB sites in the human COX-2 promoter were examined in HUVEC cells under hypoxic and norrnoxic conditions [95]. Hypoxia followed by reoxygenation activates NF-KB (p65/p50) in HeLa cells, [96] and COX-2 is induced under hypoxic conditions in HUVEC cells. Activation of COX-2 in this context appears to be mediated by a p65/p50 Rel protein dimer with the downstream NF-KB response element in the human promoter. To quickly summarize the data discussed this far, fibroblasts really only need the first 80 bases of the promoter. Epithelial cells are a little more complicated and use mostly the CRE and C/EBP sites, perhaps in coordination (some say cooperatively) with each other, but may require more of the promoter for full activation. Endothelial cells, 39 which respond to a more diverse set of extracellular stimuli, use a combination of the NF- KB, C/EBP, and CRE sites, but none are absolutely required for at least partial activity. Transcription factors binding to the CRE include CREB/ATE and Jun family members, and maybe an undefined C/EBP heterodimer. The E-box (when necessary) is bound by USPS. The C/EBP response element is bound presumably but not definitively by several members of the C/EBP family. The NF-KB site is bound by a p65/p50 Rel dimer, and there may also be some undefined Pea3 binding sites that work in coordination with the C/EBP site. These transcription factors may be regulated by activation of PKC, JNK, p38, and ERK MAP kinase pathways. Pathways involved in NF-KB or Wnt-l signaling can also trans-activate COX-2 transcription. 40 Transcriptional Regulation of COX-2 in Bone Tissue The mechanism of bone resorption is CAMP dependent and appears to be mediated though the PGE2 receptor subtype 2 and 4 (EP2 and EP4). In EP4 knockout mouse calvaria culture, bone resorption is heavily impaired relative to that of the EP1, EP2, and EP3 knockout mice. However, addition of an EP2 agonist in the EP4‘IEP4' tissue still stimulated bone resorption [97, 98]. COX-1 and 2 are both induced by similar stimuli in the MC3T3-E1 osteoblastic cell line. The induction of COX-1 at the mRNA level is only about 2 fold, and the increase in COX-1 protein expression is even less. In COX-2 knockout mice, bone development is similar to wild type mice; however, bone resorption is severely reduced. (PTH injected between long bones results in calcification of the joint in COX-2 knockout mice.) Therefore, COX-2 is thought to play a much more significant and active role in bone resorption and formation than COX-1[7, 16]. In calvaria and MC3T3-El cells, COX-2 expression is upregulated in response to treatment with basic Fibroblast Growth Factor (bFGF), Epidermal Growth Factor (EGF), Transforming Growth Factor (TGF-a and TGF—B), inteleukin-l (IL-1), Tumor Necrosis Factor-0t (TNF-or), Parathyroid Hormone (PTH), thrombin, bradykinin, forskolin, epinephrine, and prostaglandins PG12, PGE2, PGF20t or their stable analogues. [7, 16, 99]. The MC3T3-E1 cell line was isolated from mouse calvaria and is considered to be an osteoblast-like cell. Even before the discovery of an inducible COX isoform, EGF treatment of MC3T3-E1 cells had been shown to induce the release of prostaglandin metabolites, the most predominant of which was PGE2 and was referred to as “bone resorption factor.” The release of arachidonate metabolites was dependent on both 41 transcription and expression of a new prostaglandin synthase activity and mimicked the results observed in tissue preparations [100]. After the identification of the second COX enzyme, Pilbeam et al. [101] examined COX-2 expression in the MC3T3-E1 cell line and found that treatment of serum starved cells with serum, TGF-B, PMA, forskolin, or PGE2 resulted in a transient increase in COX-2 mRNA accumulation and protein expression. Because of the rapid and potent Changes they observed, they suggested that COX-2 is involved in bone responses to acute stress, such as mechanical strain, inflammation, and injury. TNF-ot causes bone resorption in vivo and in vitro. MC3T3-E1 cells respond to TNF-ot by rapidly expressing COX-2 and by releasing PGE2 as well as lesser amounts of other arachidonic acid metabolites [102-105]. At the transcriptional level, COX-2 mRNA is synthesized rapidly and transiently with mRNA levels falling to near baseline by 3 hr. Similar results are seen with serum treatment as well, following a time course similar to that observed in fibroblasts [10]]. However, in the TNF-or response, COX-2 mRNA levels increase again between 6 and 12 hr to maximal levels that are maintained for more than 24 hr. The post- 6 hr burst is reduced by the cyclooxygenase inhibitor NS398, leading to the hypothesis that this delayed response is in part due to a positive autocrine feedback loop [102]. Yamamoto et al. [102] transfected 5’serial deletions of a 621 bp promoter construct into MC3T3-El cells that were treated for 12 hr with TNF-ot. Promoter analysis showed that the NF—KB and C/EBP sites were necessary for full promoter activation because mutation or deletion of the NF-KB or C/EBP sites potently reduced promoter activity. In EMSA experiments, inducible binding to the NF-KB and C/EBP probes was observed in nuclear extracts from MC3T3~E1 cells treated with TNF~0t for 1 hr. Anti-p50 42 and anti-p65 supershifted the bound NF-KB probe, and anti-NF-IL6 (C/EBP-B) supershifted the C/EBP probe. The authors [102] suggested that COX-2 transcription may be regulated similarly to the IL-6 and IL-8 genes, where cooperative binding of C/EBP and NF-KB transcription factors to those promoters was observed [106]. The temporal regulation of COX-2 transcription in response to TNF-a in MC3T3- El cells occurs in several phases [102]. First, there is a rapid and transient induction between 0 and 2 hr followed by an intermediate repression or loss of mRNA accumulation between 2 and 6 hr Finally, there is a late phase of induction where there is a slower increase in mRNA starting at about 6 hr post treatment, which reaches a maximum level by about 12 hr. It is possible that the upstream and downstream regions of the promoter mediate different phases of the response, but since luciferase activity from the promoter reporter was only measured at 12 hr several events are confounded. If partial loss of promoter activity is observed as a result of a mutated response element, it is impossible to determine if the mutation affected activity during the early or delayed phase or during both phases of reporter gene activation. When bimodal responses are observed, mechanistic studies of transcription factor binding should also be done during each phase to Check for consistency of the response across the time course of activation. Wadleigh et al. [107] examined the effects of serum, bFGF, PDGF, PGE2 and TNF—a+ILl-B treatment on MC3T3-El cells. After 4 hr of each treatment, COX-2 mRNA as well as luciferase activity levels from a promoter reporter (—724/+7) were increased. Response element mutation constructs were tested in MC3T3-E1 cells treated for 4 hr with bFGF. This study identified and Characterized a new degenerate C/EBP site (C/EBP-2) located at —93/-85. Mutation of C/EBP-2 had little effect, while mutation of 43 C/EBP-l reduced promoter activity by about 25%, and a double mutant decreased promoter activity by more than 50%. Mutation of the CRE-l site eliminated promoter activity, and a mutation in the NF-KB site and the E-box had no significant effect. Since the CRE-l mutant and the double C/EBP mutant had the biggest effects on bFGF stimulation, the authors tested these two mutant constructs against serum, bFGF, PDGF, PGE2, and TNF-or+IL-1-B and observed nearly identical decreases in promoter activity as was observed with bFGF. Because MAP kinase signaling pathways mediate COX-2 induction in other cell types in response to mitogenic and inflammatory stimuli, the authors co-transfected DN-MEKK and DN-JNK with the wild type promoter and found that half of the promoter activity was abolished. Co—transfections with C-Jun, C/EBP-B, and C/EBP-8 superinduced reporter activity, and co-transfection with CREB and DN-C/EBP-B (LIP) blocked the induced promoter activity. The authors [107]Concluded that bFGF, PDGF, PGE2. and TNF-oc +IL-1-B activate transcription from the CRE-1 site via a mechanism that involves C-Jun and the MEKK/JNK signaling pathway as well as transcriptional activation at the C/EBP sites in response to C/EBP transcription factor family members. Okada et al. [108] identified an AP-l consensus site at -69/-63 in the murine promoter. This site in the COX-2 promoter is one base off the canonical consensus sequence, but in this case, a Close match seems to be good enough. PMA stimulated promoter activity in MC3T3—E1 cells was increased by about 5 fold in both —963/+70 and —371/+70 reporter constructs indicating that the region upstream of —371, which includes the NF-KB and CRE-2 sites, is not necessary for PMA stimulated activation of COX-2. The authors prepared single and double mutations of the CRE-1 and 44 AP—l sites in the —371 construct to be transfected into MC3T3-E1 cells treated with either PMA or serum for 3 hr. While the AP-l and CRE-1 single mutants had different effects with either PMA or serum treatment, in both cases the double mutant had significantly less promoter activity than either of the single mutants. EMSA experiments performed in this paper yielded very interesting results. The AP-l site probe was bound specifically and inducibly by nuclear proteins from cells treated with either PMA or serum and antibody to either C-Jun or C-Fos blocked binding of the probe, while the CRE-l probe was bound constitutively by extracts from both the PMA and serum treated cells. Since this AP-l site is conserved in the murine, rat, bovine, equine, and human promoters, the authors suggested that a physiological role for the AP-l site may exist. This relationship between two proximal response elements, where both are necessary for a full activation but inducible binding is only observed at one site, is highly suggestive of a cooperative interaction between the DNA bound transcription factors and transcription complex. Mechanical loading deforms the extracellular matrix producing fluid flow in the osteocyte lacunar-canalicular network that results in elevated COX-2 expression and production of PGE2, and PGF20t [101, 109]. In viva studies indicate that COX-2 selective inhibitors can block mechanical stress induced bone formation [101]. Loss of mechanical stress due to prolonged immobilization or exposure to microgravity (space flight) results in loss of bone density [101, 109]. In addition, in viva long-term administration of exogenous prostaglandins increases bone formation [7]. Fluid sheer stress rapidly induces prostaglandin production through activation of a cytoskeleton-associated calcium Channel that results in the activation of PKC. To study the effects of fluid sheer force, Ogasawara et al. [109] cultured MC3T3—E1 cells on glass 45 slides that were placed in a parallel flow path Chamber. In these experiments, the flow rate of media was kept constant while the gap between the slide and the top of the Chamber was decreased. This increased the velocity of the media across the slide, which increased the sheer stress on the cells. The authors found a “dose” dependent relationship between sheer stress and COX-2 expression. By RT-PCR, the authors observed a sustained increase in COX-2 mRNA by 3 hr with only a small increase observed at 1 hr. 5’ serial deletions from ~959 were prepared, and the authors observed that deletion of the region containing the C/EBP site nearly eliminated all the steer stress induced response. Mutations in the C/EBP-l, AP-l, and CRE-1 sites all reduced promoter activity by more than 50%, while mutations at the CRE-2, NF—KB, C/EBP—2, and E-box had no effect. The authors also performed double and triple mutation analysis of the C/EBP-l, AP-l, and CRE-1 sites, all of which resulted in promoter activity that was lower than the unstimulated wild type promoter construct. The observations from a comprehensive and compelling set of EMSA experiments produced results that were very interesting. Nuclear extracts were prepared from control cells and cells stimulated with sheer stress for 1 or 3 hr. They observed inducible binding to the C/EBP-l probe that was supershifted with antibody to C/EBP-B. Binding to the AP-l probe was also inducible and was supershifted by anti—C-Jun/AP-l antibody. Binding to the CRE-1 probe was constitutive, and it was supershifted by antibody to CREB at 1 and 3 hr and by also anti- phospho-CREB after 3 hr. Because of the signaling pathways that are activated by sheer stress, one might expect to find similarities within the transcriptional mechanism observed in response to PMA treatment. Indeed, the findings of Okada et al. [108]with PMA treated MC3T3-E1 46 cells are consistent with those of Ogasawara et al. in sheer stressed stimulated MC3T3- El cells. Okada et al. [108]Compared the effects of the CRE-1 and AP-l mutations in serum and PMA treated MC3T3-E1 cells. They found that the CRE-1 site was more important for the serum response and the AP-l site was more important for the PMA response, but with either treatment. The double mutation significantly lowered promoter activity relative to the single mutants. This finding is relatively consistent with the observations of serum and mitogen stimulated MC3T3-El cells where the CRE-1 mutation eliminated promoter activity [102, 107]. Perhaps these data indicate that gene activation requires both the AP—l and CRE-1 sites. If one site is removed, then most of the cis-activation potential is lost, and if both are removed, then nearly all of the gene activation is removed. In these studies there is an important incongruent detail that is overlooked. In all of these studies, binding to the CRE-1 site is constitutive. This is not consistent with mechanism of AP-l transcription factors. The AP-l transcription factors are phosphorylated when MAPK signaling pathways are activated, leading to inducible, transient binding to target sites on gene promoters. I think it may be possible that the AP-l site which is immediately upstream of CRE-1 site is where the inducible binding of AP-l transcription factors occurs, and that occupation of the overlapping CRE-1 and E-box sites is necessary for trans-activation. If this is true then removal or deletion of the of the CRE-1 site would block trans- activation, as has been observed, and removal or deletion of the AP-l site would also block trans-activation. This is probably a key part in the puzzle to understanding the transcriptional activation of the COX-2 gene, which will be discussed in more detail later. 47 As a slight tangent to the work discussed above, glucocorticoids have been shown to be potent inhibitors of inflammatory mediators and the stimulated release of prostaglandins. In addition, glucocorticoids have been found to decrease bone resorption. Two recent papers should have very real impacts on the national space program and the diets of astronauts. The glucocorticoid-like natural product, humulon (((R-)—3,5,6- trihydroxy-4,6-bis(3-methyl-2-butenyl)-2-(3-methyl-1-ox-obutyl)~2,4-Cyclohexadien-1- 1), which is isolated from hops extract, was found to inhibit bone resorption (ICso 5.9 nM) and COX-2 activity (ICSO luM) [110]. Another glucocorticoid-like natural product, resveratrol, a phytoalexin found in grapes, blocked the transcriptional activation of COX- 2 by interfering with the activation of PKC, ERK, and C-Jun [83, 111]. Together, these data suggest that Tang® should be replaced with other more beneficial beverages on longer space flight missions where bone resorption critically impacts human performance and wellness. 48 Transcriptional Regulation of COX-2 in Granulosa Cells Prostaglandins are necessary mediators of the reproductive process. COX-2 deficient female mice are infertile because of problems in ovulation, fertilization, implantation, and decidualization [112]. Functional promoter studies in gonadatropin stimulated pre-ovulatory rat granulosa cells showed that a transiently transfected -2698/+23 promoter reporter construct mimicked the in viva induction kinetics of the COX-2 enzyme. From a set of 5’ serial deletion reporter constructs, they found the region between —194 and -54 was critical for reporter activation. The region upstream of -194 contained possible negative regulatory elements, which was also observed by Fletcher et al.[55]. EMSA experiments helped to identify a region between -l92 and -110 that exhibited significant specific binding. This region of the rat promoter contains an AP-l site at -165/—159, as well as a C/EBP site at -l42/-120 [113]. The C/EBP site was bound by the transcription factor C/EBP-B but not C/EBP-a or C/EBP-8 in EMSA supershift assays. In viva immunoblot and northern analysis confirmed that C/EBP-B was constitutively expressed in hCG stimulated pre-ovulatory follicles. When the C/EBP consensus was mutated in a -192/+23 reporter construct, inducibility in response to forskolin, FSH, and LH was greatly reduced. Although the organization of the rat promoter is different from the murine and human promoters, similar cis-acting elements mediate COX—2 gene activation by their interactions with similar trans-activating factors. Interestingly, similar results were observed with the bovine promoter in bovine granulosa cells. Promoter analysis and EMSA experiments showed that the C/EBP but not the CRE-1 site was necessary, and that the C/EBP site was only bound by C/EBP-B [114]. 49 When different mechanisms of gene transcription are observed across different cell types and in response to different stimuli, it becomes evident that one specific signaling pathway is not used to activate the COX-2 gene. Even within one cell type, different stimuli can activate transcription of COX-2 by activating different pathways of cellular signaling. The most convenient, reduced model for studying COX-2 transcription has been the fibroblast stimulated with MAPK activating mitogens; however, even in this cell type there are exceptions when more or less of the promoter is required to drive promoter reporter activity. 50 Lipopolysaccharide Signaling The LPS Molecule. Bacterial lipopolysaccharide is a complex molecule with three covalently linked domains. (Figure 3) The O-antigen polymer is an oligosaccharide with up to 40 repeated units. This domain is highly immunogenic and varies greatly between bacterial strains [115]. The core region is a phosphorylated non-repeating oligosaccharide that is required for the outer membrane of bacteria to function against antibiotics. The third domain is the lipid A, which functions as the hydrophobic anchor for LPS in the outer membrane. The lipid A domain consists of a pair of phosphorylated and acylated [3,1-6-linked glucosamine molecules. Variations in the acyl Chains determine the biological activity of the LPS. The six 3-hydroxy fatty acids of the lipid A are generally saturated and are between 14 and 18 carbons long[ll6]. Deacylated LPS no longer retains its inflammatory properties and becomes an antagonist for active LPS [117]. Likewise the heavily modified LPS of Rhadabacter sphaeraides, with its unusually short acyl chains, functions as an antagonist for the action of E. cali endotoxin [116]. The LPS Receptor. LPS is a membrane forming amphiphile that is relatively insoluble in aqueous solutions and diffuses slowly from membranes and aggregates [118]. In serum free buffers, mg/ml concentrations are required to generate a cellular response. With serum, ng/ml concentrations of LPS are sufficient to elicit a response because of the presence of the 60 kDa LPS Binding Protein (LBP) . LBP forms high affinity complexes with the lipid A moiety of LPS and in turn; can interact with both soluble and membrane bound monocyte differentiation antigen CD14 [119]. LPS can be transferred from LBP to CD14, which is involved in enhanced sensitivity to LPS and in the Clearance of LPS by mediating its transfer to lipoprotein particles or phospholipid 51 .. gore Domain O-Antigen LipidA Domain or marinara-o 0 Figure 3. Lipopolysaccharide Molecule A representative structure of the core region, part of the O-antigen Chain, and lipid-A domain of lipopolysaccharide. Abbreviations: GalA, galacturonic acid; GlCA, glucuronic acid; Kdo, 3-deoxy-D-manno- 2-octulosonic acid; GlCN-onate, 2-amino-2-deoxygluconic acid; QuiNAC, 2-N- acetamido-2,6-dideoxyg1ucose (N-acetquuinovosamine); GlCN, glucosamine; Man, mannose; 3MeRha,3-O-methylrhamnose; Fuc, fucose. Forsberg LS, Carlson RW. 1998. The structures of the lipopolysaccharides from Rhizabium etli strains CE358 and CE359. The complete structure of the core region of R. etli lipopolysaccharides. J. Biol. Chem. 273 :2747—57 52 vesicles [120]. The Clearance of LPS is a separate event from LPS induced signal transduction [118, 120-122]. CD14 is attached to the extra cellular membrane by a glycosyl phosphatidylinositol anchor and has no transmembrane domain. It is unable therefore, to transfer a signal through the cell membrane. However, CD14 is an LPS receptor, and blockade of CD14 with antibody blocks LPS induced TNF-or, IL-lB, IL-6, and IL-8 release as well as endotoxin induced shock in whole animal studies [123]. On the other hand, the LPS hypo-responsive mouse, C3H/HeJ, expresses normal amounts of CD14 on cell surfaces suggesting that additional factors are required for an LPS response [124]. In the drosophillia world of deve10pment, a receptor involved in dorsal ventral patterning was discovered with an intracellular carboxyl terminal domain that is similar to the carboxyl terminal of the IL-1 receptor. This receptor, named Toll, participates in a signaling cascade leading to the activation of DIF and Dorsal, the drosophillia homologues of an NF-KB like signaling pathway, which is essential for the anti-fungal immune response in flies[124]. Several mammalian homologues of Toll have been identified and are referred to as Toll-Like Receptors (TLR). The intracellular domain of TLR5 are similar to the intracellular portion of the IL-1 receptor, which suggests that the mammalian homologues may participate in signaling pathways similar to the IL-1 receptor. The TLR proteins were first implicated as potential LPS receptors when positional Cloning of the gene responsible for the LPS hypo-responsive phenotype of the C3H/HeJ mice was mapped to the TLR4 gene [124]. In addition, the LPS insensitive mouse line C57/10SCCr is null for the TLR4 locus and does not express any TLR4 CDNA [124]. Both the TLR2 and TLR4 53 MD-2 LBP LPS 4 ‘SCD14 MyD88 :5 MyD88 IRA IRA NIK MEKK-1 NF-kB MAPK Figure 4. The TLR4 Receptor Complex 54 receptor mRNAs are highly expressed in peripheral blood leukocytes, monocyte, and macrophage populations. Transfection of either receptor confers LPS responsiveness to LPS insensitive cells, and co-transfection of CD14 further enhances the response [125, 126]. However, highly purified LPS does not signal through TLR2, and TLR2 gene knockout mice remain sensitive to LPS where as disruption of the TLR4 gene results in a loss of LPS responsiveness [127]. The TLR2 receptor appears to be involved in a more broad recognition of other bacterial cell wall components such as bacterial peptidoglycan from gram positive cells and bacterial lipoproteins [128, 129]. The Toll receptors appear to function as cell sensors of microbes. So far, ten TLRs have been identified. In addition to TLR2 and TLR4, TLR6 is thought to be involved in recognition of bacterial lipoproteins, and in cooperation with TLR2 and TLR9 recognizes bacterial CpG DNA. The molecular interactions leading to LPS signaling involve a complex assembly of proteins on the extracellular surface of the cell (Figure 4). LPS is first bound by the amino-terminal domain of LBP, and then the carboxyl terminal domain of LBP associates with CD14. LPS triggers the physical association of CD14 and TLR4 and then upon binding ligand, the TLR dimerizes [130]. The TLR4 complex also requires the cell surface protein MD-2, which also appears to be necessary for signaling through TLR2. Although its exact function is still unknown, MD-2 is thought to play a role in the stabilization of the receptor complex [131, 132]. Intracellular signaling is mediated by interactions with the intracellular Toll/IL- 1R (TIR) domain. Signaling through IL-lR, TLR2, and TLR4 is mediated by the interaction of the receptor TIR domain, with the TIR domain of myeloid differentiation factor (MyD88), which facilitates the assembly of the toll receptor complex. This 55 complex includes the IL—lR associated kinase (IRAK), TNF receptor associated kinase-6 (TRAP-6), and the evolutionary conserved intermediate in toll (ECSIT) signaling factor [124]. Within 5 to 15 min the activated TLR4 receptor complex activates a broad range of signaling molecules and the release of second messengers. Rapid activation of p38, ERK1/2, and JNK activities are observed; I-kBor and I-kBB are degraded; phospholipase A2 (PLA2), phospholipase C (PLC), and sphingomylenase (SMase) activities are upregulated; Protein Kinase C or and E (PKC) and protein kinase A (PKA) are activated; and members of the Src family of tyrosine kinases are activated as well. [133-140]. In macrophage and monocyte cells, these more immediate events are then followed by autocrine effects mediated by the release of TNF-0t, IL— 1 [3, and prostaglandins. MAPK Signaling Induced By LPS. LPS induced signaling can be divided into two general types, MAPK signaling and NF-KB signaling". Within the MAPK category, three main nodes are activated: the Jun N-term Kinase (JNK), the Extracellular Receptor Kinase (ERK), and p38 (Figure 5). The TLR4 receptor signaling complex initiates the activation of these kinases. This complex includes the recently discovered ECSIT, which Cleaves inactive MEKK-1 (MAP3K) to its active form [141]. (Figure 6) MEKK-1 best known for activating MKK4/SAPK [142], which activates JNK resulting in phosphorylation and activation of AP-l transcription factors (specifically Jun family members) [143]. MEKK-l can also activate MKK3/6, which results in activation of p38 [142, 144, 145]. MKK4 is also activated and is able to activate p38 in models of inflammation. [146, 147]. Additionally, MEKK-l has been found to activate IKK 4 Caveats of Signaling Experiments, Appendix A 56 LPS Receptor Complex IKK NF-kB MEKK MAP Kinase Cascades ERKl/Z p38 JNK \ l \ / . , l CREB&ATF, Elk-1, c-Myc, AP-l C/EBP Figure 5. General signaling pathways activated by LPS 57 Figure 6. Detailed signaling pathway map: A schematic representation of the signaling pathways activated directly or indirectly by LPS. Key to abbreviationszAA, Arachidonic Acid; ASK-1, apoptosis signal-regulating kinase; ATF, Activator of Transcription Factor; C/EBP, CAAT Enhancer Binding Protein; CAK, Ceramide Activated Kinase; CAMP, Cyclic Adenosine Monophosphate; CREB, CAMP Response Element Binding Protein; ECSIT, Evolutionary conserved intermediate in toll; Gas, Stimulatory G-protein Subunit; IKK, Inhibitor of k8 Kinase; 1P3, phosphatidyl-inositol-3; IP3K, phosphatidyl-inositol-3- kinase; IRAK, Interleukin-1 Receptor Associated Kinase; JNK, C-Jun N-term Kinase; LPS, Lipopolysaccharide; MEKK, Mitogen Activated Protein Kinase Kinase Kinase; MKK, Mitogen Activated Protein Kinase Kinase; MyD88, Myocyte Differentiation Factor; NIK, Nuclear Factor-kappa B Inducing Kinase; PAK, p21 Associated Kinase; PIP2, Phosphatidylinositol-4,5-bisphosphate; PKA, Protein Kinase A (CAMP); PKC, Protein Kinase C (Calcium); PLA, Phospoholipase A; PLC, Phospoholipase C; RIP, [TNF] Receptor Interacting Protein; TAK- 1 , Transforming growth factor-beta Activated Kinase-1; TNF-a, Tumor Necrosis Factor-a; TRADD, Tumor Necrosis Factor Receptor Associated Death Domain; TRAF, Tumor Necrosis Factor Receptor Associated Factor. 58 murmz mh<\mmMU mmmhv 8:5 032-0 Tim “EEO mh<\mmMU H H \. H H H t. . ma: Qafiaflz $3 VHzH NHHVHHmvHvH-H 0 .1 12.0 1.1.2... LPS(hr) 0 1 12 o 1 12 ~_-I- 'I'n'x P-CREB —> CRE-2 Complex Figure 12. CREB and CBP bind the CRE-2 Probe. The CRE-2 probe is bound by anti- CREB-1, anti-phospho-CREB, and anti-CBP. EMSA supershifi assays were performed with a 32P-labeled CRE-2 probe and nuclear extracts from RAW 264.7 cells stimulated with LPS (200 ng/ml) for 0, 1 or 12 hr in the presence of a normal IgG control antibody or antibodies to CREB-1, phospho-CREB, and CBP as described in the text. The solid arrows beside the EMSA autoradiograms indicate the supershifled complexes. The open arrows indicate the complex that binds the CRE-2 probe. 93 (Figure 13). Supershift assays were performed with antibodies to p50 or p65 Rel/NF-KB proteins involved in the different dimeric NF-KB complexes. The anti-p50 antibody caused shifts in both the high and low mobility complexes, but the anti-p65 antibody shifted only the low mobility complex (Figure 13). These data indicate that the NF-KB response element of the COX-2 promoter is bound predominantly by a p50/p65 heterodimeric complex in the early phase of cell activation by LPS and predominantly by a p50 homodimer in the late phase. LPS treatment results in persistent COX-2 mRNA synthesis. Stimulation of macrophage and macrophage-like cells with LPS results in the synthesis of prostaglandins and the prolonged expression of COX-2 protein [195, 196, 210, 211]. Because LPS induced COX-2 expression in U937 is heavily influenced by post- transcriptional regulation[212] , and because p50 dimers are associated with negative regulation of inflammation related genes [213], we performed nuclear run-on assays to assess the level of COX-2 transcription. We found that LPS stimulation resulted in rapid and prolonged synthesis of COX-2 mRNA (Figure 14A); moreover, the rate of mRNA synthesis correlated well with the amount of COX-2 mRNA observed in RAW 264.7 cells (Figure 14B). 94 Normal I gG Anti-p50 Anti-p65 LPS (hr.) 0 1 12 p50/p65 —> p50/p50 —> {3' RE-Z NF-KB CTGTGTGCGTGCTCT GAGCAGCGAG-MCTGCGCC CCAGTGGG GAGAGGTGA-ITAGTTAG GACC‘ITAGATCCCGGGAGGGGM Probe NF-KB Figure 13. p65 and p50 bind the NF-KB probe. Nuclear extracts were prepared from control or LPS-stimulated RAW 264.7. The cells were stimulated with LPS (200 ng/ml) for 0, 1, or 12 hr. Double stranded, 32P-labeled NF-KB probe was incubated with nuclear extract and either normal IgG control serum, anti-p50 or anti-p65 antibodies as described in the text.The NF-KB probe complexes with p50/p65 and p50/p50 transcripton factors are indicated with arrows on the left side of the figure. Supershifted (SS) complexes bound with antibody to p65 or p50 are indicated with arrows on the right side of the figure. 95 & (COX-ZIActin) (a) Relative COX-2 Transcripts 4 N A Relative COX-2 Transcription (COX 2IActin) LPS(hr-) o 0.25 0.5 1 3 6 9 12 Vehicle-'1. ' . ~- _ cox-2,» “' Actinl LPS(hr.) o 0.5 1 3 ’4‘ . r . . 1 Emmy) .WF flufi. Vector B . _ _. _. Figure 14. Nuclear run-on and Northern blot analysis of LPS-stimulated RAW 264.7 cells. A) Nuclei were isolated from cells stimulated with LPS (200 ng/ml) for the indicated times. The nuclei were incubated with [32P]UTP to label newly synthesized transcripts as described in the text. RNA was isolated and blotted onto a nitrocellulose membrane and with vehicle, mCOX-2 cDNA, -actin cDNA or an empty vector DNA control. The membrane was washed and exposed to a phosphoimaging screen. Densitometry was performed with Image Quant software. B) Northern blot analysis was performed as detailed in the text. RNA was isolated from 100 mm plates of RAW264.7 cells stimulated with LPS (200ng/ml) for the indicated times. Total RNA (15ug) was separated on a 0.8% agarose, 4% formaldehyde TAE gel, transferred to a nitrocellulose membrane, and hybridized to mCOX-Z and B-Actin cDNA probes. 96 Discussion The induction of COX-2 gene transcription observed in fibroblasts in response to mitogens, serum, and transformation requires only the first 80 bp of the COX-2 promoter [64, 65]. A larger portion of the COX-2 promoter is necessary for maximal COX-2 gene expression in macrophages [155, 200, 214], osteoblasts [102, 109], and endothelial cells [92]. The current model for the transcriptional regulation of the COX-2 gene in LPS- stimulated macrophages includes LPS activation of the toll-like receptor (TLR4), which, in turn, initiates signaling through MyD88/IRAK/TRAF6/ECSIT resulting in the activation of ERK, JNK, p38, PKC, and NIK. These kinases exert their actions by phosphorylating either transcription factors or downstream effectors to cause the transcriptional machinery to begin transcribing the COX-2 gene. The cis-acting elements of the murine promoter that are necessary for a response to LPS include an overlapping CRE (CRE-1) and E-box (~59l-48) [215] and two C/EBP sites at —138/—130 (C/EBP-l) and —93/-85 (C/EBP-2) [200] (Figure 7). AP-l, CREB or USF-l transcription factors bind to the overlapping CRE-1 and E-box, and C/EBP transcription factors bind to the C/EBP response elements [92]. We have now established that both the NF-KB response element at -401/-393 and a second CRE (CRE-2) located at -447/—440 are also necessary for maximal LPS induction of the murine COX-2 gene. We have also shown that p65 and p50 bind the NF-KB site in different combinations and that CREB and CBP can associate with the CRE-2 site. This collection of trans-activating factors may function independently or cooperatively as an enhancer complex with the general transcriptional machinery to activate the COX-2 gene (Figure 15). 97 R\ \ Pol II Ilolocnl) IIlL' CRE-2 1 CREB . .V 15. .< ‘ E-box & CRE-1 ‘ USF—1 AP-l NF "B C/EBP CREB p65/p50 C/EBP p50/p50 B C/EBP 8 Figure 15. Model for the interactions of trans-activating factors associated with the COX-2 promoter. 98 NF -iCB response element in LPS-induced COX-2 gene expression. There are several lines of evidence that suggest that the NF—KB response element and NF-KB signaling play a role in activating the COX-2 gene. The NF-KB response element is conserved and is in a similar location in the human, monkey, equine, bovine, rat, and mouse promoters. The consensus sequence is nearly identical and follows a (5’- GGGATYCCC-3’) motif that has been found to favor interactions with p50 homodimers [213]. The KBl site of the H3-10 gene even exhibits induced NF-KB binding with a time dependent change in composition [216] (i.e. the composition of the binding complex changes from a p65/p50 heterodimer to a p50 homodimer at 3 hr post LPS-treatment) in LPS stimulated RAW 264.7 cells. The importance of NF-KB signaling has been demonstrated recently by Rhee et al. [208]. These studies suggest that NF-KB activation via the TLR4 signaling pathway is essential at several steps for induced COX-2 reporter activity in RAW 2647 cells. [207, 208]. There are also several reports indicating that inhibition of NF-KB activation with chemical and synthetic peptide inhibitors and decoy Oligonucleotides block COX-2 activation [202, 204, 205, 217, 218]. Recently, Wadleigh et al. [200] reported that there was no requirement for the NF-KB site or for NF-KB activation for the LPS stimulated activation of a murine COX-2 promoter reporter in RAW 264.7 macrophages treated with LPS. Our present results indicate that when the NF-KB response element is mutated, over half of the COX-2 promoter activity is lost. Binding to the NF-KB response element is inducible, and the composition of the factors binding at this site changes from a p65/p50 complex to a p50 homodimer at l and 12 hr after LPS stimulation. 99 The differences in the results of the promoter activity assays observed here and by Wadleigh et al. could be due to the time course of the experiments, the amount of LPS used, the design of the promoter reporter construct or perhaps subtle differences in the RAW 264.7 cells. Wadleigh et al. stimulated their RAW 264.7 cells with LPS for 5 hours with 10 ng/ml LPS using a plasmid containing —724/+7 of the murine COX-2 promoter. We stimulated RAW 264.7 cells with LPS for 12 hours with 200 ng/ml LPS using a plasmid containing -966/+23 of the murine COX-2 promoter. To check for a time dependent requirement of the CRE-l, C/EBP, NF—KB, and CRE-2 response elements, we performed a time course experiment with the wild type —966/+23 promoter construct with mutations in these four response elements and with a construct containing only —98/+23. (Figure 10) We found that each mutation had a similar effect at time points between 1 and 12 hours. Each response element mutation significantly decreased the level of induced promoter activity across the time course. RAW 264.7 cells are exquisitely sensitive to LPS, and as a consequence there is only a small increase in inflammatory cytokine release [219] caused by an increase from 10 ng/ml to 100 ng/ml of LPS. It is likely that the differences in the LPS doses used in our experiments only resulted in a small increase in promoter activity over what was observed by Wadleigh et al. The promoter constructs used in our experiments and by Wadleigh et al. differ by 242 bases on the 5’ end. It is unlikely that this difference resulted in the large difference between our results, because in Figure 8 the promoter construct containing -966/+23 and —459/+23 had similar levels of induced promoter activity. The 16 bp difference at the 3’ 100 end of the construct is also unlikely to have a large affect on promoter activity, because this is a relatively short region that does not contain any identified response elements. In the experiments by Wadleigh et al., LPS treatments result in about a 4 fold increase in promoter activity. In our experiments, we routinely observe 6 to 12 fold increases in promoter activity. We have observed that after culturing RAW 264.7 cells for 4 to 6 months (about 50 to 60 passages), LPS induced COX-2 promoter activity drops to between 2 and 4 fold. We prepared a working cell bank of RAW 264.7 cells from a fresh ampule of cells obtained from the American Tissue Type Collection. We reinitiated our RAW 264.7 cell cultures at about 5 month intervals or when induced promoter activity decreased below about 5 fold. EMSA experiments are used to approximate what transcription factors may bind to a response element. The actual condition within cells on the promoter in the context of chromatin may differ significantly. EMSAs show what can be reconstituted on a short piece of naked DNA from proteins extracted from the nucleus of a cell. In our experiments we observed that a p50 dimer associates with the NF-KB response element during the late phase of the LPS response. If this condition also exists on the native promoter it could cause the down regulation of transcription by displacing the transcriptionally active p65 containing complex. Alternatively the p50 dimer may play some role in the persistent expression of COX-2 or have only a neutral effect. The same signaling pathways that activate the transcription of COX-2 in LPS- treated RAW 264.7 cells are also involved in the activation of the TNF-0t gene [142, 220, 221]. LPS-induced COX-2 activity, expression, and mRNA levels are upregulated for 24 to 48 hr. Similar results are observed with the TNF-0t gene, although the time course is 101 shorter. The rate of TNF-0t gene transcription is decreased to background levels within hours of the initial activation, but because the mRNA is stabilized by way of a p38 mediated mechanism, mRNA and protein expression levels remain elevated for at least 12 to 18 hr after LPS treatment [222, 223]. The downregulation of TNF-0t gene expression after LPS stimulation results in part from increased expression of p50 and a resultant increase in the binding of p50 dimers to NF-KB response elements in the TNF-0L gene promoter [213]. Dimers of p50 commonly function as repressors of transcription in other promoters as well [158, 159]. NF-KB dependent gene expression involves transcriptional coactivators that are proposed to function by bridging sequence specific transcription factors to the basal transcriptional machinery [224]. p65 and c-Rel are the transcriptionally active members of the Rel family of transcription factors [159]. Activated p65 associates with CBP to trans-activate NF-KB dependent genes [225]. The p50 and p52 Rel transcription factor members are transcriptionally inactive as dimers and have been thought to act as transcriptional repressors [158, 159]. This paradigm, however, has recently been challenged by the observation that the p160 family of transcriptional co-activators are also involved in NF-KB dependent gene activation. The steroid co—activator—l (SRC-l) has been found to interact with p50 to potentiate p65 independent NF—KB-mediated transactivation [181]. Because p50 is not involved in downregulating COX-2 transcription, it may be involved in the persistent transcription of COX-2; however, in our hands, overexpression of p50 did not have a significant effect on LPS induced promoter activity, a result also observed by Yamamoto et al. [102] with TNF-0t stimulated MC3T3-E1 cells. 102 Monocyte and macrophage cells have been shown to produce prostaglandins and express COX-2 for 24 to 48 hr [195, 196, 210, 211] in response to pro-inflammatory stimuli. Persistent expression is most likely due to prolonged transcriptional activation because of the instability of the COX-2 mRNA. Recently, however, several studies in monocytes and macrophages have revealed that post-transcriptional regulatory mechanisms result in mRNA stabilization that can lead to sustained COX-2 expression [212, 226]. Because of the possibility that post-transcriptional stabilization of COX-2 mRNA may account for the prolonged expression of COX-2, and because a known transcriptional repressor, the p50 homodimer, binds the NF-KB response element after the initial activation of the COX-2 gene, we performed nuclear run-on experiments to determine whether transcription of the COX-2 gene was downregulated after LPS treatment. We found that the initial rate of transcription was higher than the steady state rate of transcription observed after several hours of LPS stimulation. This pattern of gene transcription correlated well with the pattern of mRNA accumulation. These data establish that the COX-2 gene is persistently activated in response to LPS, although post- transcriptional regulatory mechanisms may also contribute to the overall response. In promoter activity assays, we observed a higher level of inducibility than in the nuclear run-on and Northern blot experiments. The level of actin transcription was increased by the LPS treatments, which lowers the relative increase in COX—2 levels that are reported in Figure 14. The relatively stable luciferase protein and mRNA enhance the measured levels of increased activity, and probably most importantly, the transfected 103 promoter reporter plasmid is not integrated into the chromatin like the endogenous promoter. CRE-2 in LPS-induced COX-2 gene expression. Mutations made in the CRE-2 site caused a decrease in promoter activity that was similar in magnitude to the effects caused by mutation of the NF-KB and CRE-1 response elements but was not as potent as the effect caused by the C/EBP site mutation. Double mutations at CRE-2 and the other three response elements revealed additive decreases in promoter reporter activity when the CRE-2 mutation was combined with the C/EBP and CRE-l mutants. Interestingly, however, the double CRE-2 and NF-KB mutation did not cause an additive decrease in promoter activity. Since both the CRE-2 and NF-KB mutation mutants resulted in similar decreases in promoter activity, it is possible that each one mediates a similar level of transcriptional activation. However, when both sites are mutated, the mutations do not cause an additive decrease in promoter activity. Because there is no additive effect, the sites most likely do not act independently of each other. We suggest that the combination of the factors at the CRE-2 and NF-KB sites are required to mediate their trans-activation potential. This type of interaction is cooperative, not in the sense of binding DNA, but in causing an increase in transcriptional activation when both sites are intact. Binding of nuclear proteins to the CRE-2 is constitutive, but because removal or deletion of the CRE-2 decreases the promoter activity, we expected either binding or phosphorylation of CREB at this site would be inducible. The CREB transcription factors generally bind DNA independently of activation and are phosphorylated to an active state. In supershift experiments, the nuclear protein complex bound to the CRE-2 probe was constitutively shifted by antibody that is specific for the conserved carboxyl terminal domain of CREB and ATF transcription factors. Antibody againstphosphorylated CREB also produced a constitutively shifted band, further confirming the presence of CREB bound to CRE-2. Because CREB is bound to CRE-2, we also used antibody against CBP to test for the presence of this factor and found that a lower mobility/probe binding complex was supershifted. The intensity of the shifted bands observed with the anti- phospho-CREB antibody is significantly less intense that what is observed with the anti- CREB antibody. It is likely that only a small portion of the CREB in the nuclear extracts is phosphorylated, and this would cause a reduced signal. The supershifted CBP complex is also much less intense that the anti-CREB supershift. Any signal observed in this experiment is dependent on the extraction of this large protein from the nucleus, and the assembly of a DNA probe-DNA binding protein-CBP complex. These two requirements decrease the probability of observing this interaction. CBP has been suggested to play a role in the transcriptional regulation of COX-2 but has not previously been shown to interact with any of the COX-2 promoter [64]. CBP and other transcription co-activators are found in limited amounts within cells and are competed for by ligand bound nuclear hormone receptors and other inducible transcription factors [85, 86]. This model was recently examined in the context of the COX~2 promoter with PMA-treated epithelial cells [90]. PPAR'y ligands, ciglitazone, and l5-deoxy-A'2"4prostaglandin J2 inhibit the PMA response, which is almost completely rescued by co-transfection with c-Jun and CBP and partially rescued by CBP alone. This suggests that there is a functional interaction between CBP and trans-activating factors bound to the promoter. Together, the results of this study and our current findings 105 provide evidence that CBP interacts both functionally and physically with the COX-2 promoter. Conclusion. We have identified a new cis-acting element (CRE-2) functional in the COX-2 promoter during LPS induction of COX-2 expression in the RAW 264.7 cell line. This CRE-2 element is bound specifically by a CREB/ATF transcription factor along with CBP. These complexes appear to bind the CRE-2 probe constitutively. We also observed that the NF-KB response element is inducibly bound primarily by p65/p50 after 1 hr LPS stimulation, and primarily by p50/p50 after 12 hr. Promoter reporter assays suggest the CRE-2 and NF-KB response elements act together to facilitate a maximal LPS induced response. Our data corroborate the model of Mestre et al. [155] that multiple redundant signaling pathways lead to the activation of COX-2 gene transcription and the observations of Rhee et al. [208] that NF-KB activation is required for maximal COX-2 expression. 106 CHAPTER 3 EXAMINATION OF THE COFACTORS ASSOCIATED WITH THE CRE-2 AND NF-KB REGION Introduction The purpose of this section is to document several follow-up experiments that were designed to answer questions raised by the data presented in the previous chapter. Based on our new data and what is known about the transcription factors that are involved in this system, we developed a model for the CRE-2 and NF-KB region of the COX-2 promoter (Figure 16). Before the addition of LPS, the CRE-2 site may be occupied by a CREB and CBP complex. Immediately after LPS stimulation, an NF-KB transcription factor complex comprised of p50 and p65 binds to the NF-KB site. In the later phase, the gene is still transcriptionally active and the NF—KB site is bound by a p50 homodimer. The late phase of transcription is unique because there are many examples of transient activation of COX-2. Since this upstream region is not necessary for transient activation, we considered the possibility that it may play a role in observed prolonged transcription. Of particular interest was the p50 homodimer that is present during the late phase. Activation of NF-KB is generally transient, but there are exceptions, as was discussed in Chapter 1. It is possible that I-KBB may be bound to the p50 dimer which prevents it from being removed from the promoter. While the p50 occupation may be neutral, we considered the possibility that p50 was involved in either repression or activation during the late phase of transcription. To test for repression we performed 107 CBP/p300 (“1.312 i Un-Stimulated Transcriptionally CBP/p300 Active Complex Early, LPS+ Negative C BP/ p300 Regulation? Figure 16. Model of the CRE-2 and NF-KB region during LPS activation 108 nuclear run-on assays, and as discussed in Chapter 2, transcription occurs for up to 12 hours. If the p50 dimer is involved in activation, it does not act alone. Since the NF-K‘B transcription factors require transcriptional coactivators to initiate transcription, we looked for possible cofactors that may be involved in the activation process. These experiments are described below. Since our data indicated that the CRE-2 and NF-KB sites function together, we considered the possibility that a large complex may form on the combined CRE-2 and NF-KB sites. We thought that this complex might be visible as a very low mobility DNA binding complex in EMSAs. The design of this experiment and the results are described below. 109 Supershift with I-KBB The persistent activation of NF-KB is mediated in some cases by I-KBB, as was discussed previously in chapter 1. In LPS-stimulated RAW 264.7 cells, I-KBB is rapidly degraded and then re—synthesized [140]. Since COX-2 gene transcription is persistently activated by LPS and I-KBB is present in RAW 264.7 cells, we performed EMSA supershift assays to determine if I-KBB was part of the p50 homodimer-containing complex bound to the NF-KB response element during the late phase of the LPS response. Results. Antibody against I-KBB did not cause a supershift or blockade of the complex bound to the NF-KB probe in nuclear extracts from either the control or the LPS-stimulated RAW 264.7 cells (data not shown). Discussion. This experiment was aimed at understanding how the p50 homodimer was maintained on the COX-2 promoter during the late phase of the LPS response. We could not detect I-KBB as a component of the complex that binds the NF-KB probe. We used the supershift EMSA technique because it is reasonably sensitive to small amounts of protein and because we could compare the results of this assay with those of other similar experiments. Other methods could also be used to detect this interaction. For example, immunoprecipitation experiments could be useful for determining if l-KBB associates with the NF-KB proteins that we have identified as interacting with the NF—KB response element probe in the nuclear extracts from LPS- stimulated RAW 264.7 cells. Chromatin immunoprecipitation with I-KBB antibody could be used to detect interactions between I-KBB and the COX-2 promoter. Since this latter 110 technique uses PCR amplification, it may be a more sensitive probe for this type of interaction. Alternatively, another I-KB protein may be involved in this complex. Bel-3 has been observed to bind p50 homodimers. The role of Bel-3 is not well understood. In some cases it appears to be involved in facilitating NF-KB transcription by removing p50 dimers from promoters [180]. In other instances, it appears to function as a coactivator of NF-KB dependent transcription. Before looking for interactions between the NF-KB response element and Bel-3, the presence of a murine Bcl-3 homologue in RAW 264.7 cells needs to be verified. 111 Co-Transfection Experiments with Transcription Factors and Transcriptional Coactivators We sought to determine the functional significance of an apparent cooperative effect between the CRE-2 and NF-KB sites, the association of CBP and CREB with CRE-2, and the inducible binding of p65/p50 and persistent occupancy of the NF-KB site by a p50 homodimer. One possibility is that transcriptional activation by NF-KB requires multiple coactivators. [224] CBP is a transcriptional coactivator of p65 [225], and a CBP and SRC-l are transcriptional coactivators of p50 [181]. Because of the arrangement of CRE-2 and the NF-KB sites on the murine COX-2 gene, our other data suggest that LPS stimulation brings CBP and p65 into close proximity. The implications of this relationship are similar to those explored by Gerritsen et al. in their experiments with the E-selectin promoter [225]. The E-selectin promoter contains three regulatory domains. The first contains an element that is recognized by CREB/ATF and AP-l transcription factors and the second two elements are bound by p50/p65 NF-KB heterodimers. (Both the E-selectin and COX-2 promoters have a similar CRE and NF-KB pair in a region approximately 500 bp upstream of the transcriptional start site.) The experiments by Gerritsen et al. [225] demonstrate that CBP and p300 physically interact and that this interaction is required for trans-activation of an E-selectin promoter reporter. Results. To determine if CREB, CBP or SRC—l could potentiate LPS-induced COX-2 activity, we obtained a CREB expression plasmid from Marc Montiminy (Salk Institute for Biological Studies, La Jolla, CA), a CBP expression plasmid from Richard Goodman (Vollum Institute, Oregon Health and Science Institute, Portland OR.) and pCR3.1-hSRC-1A expression plasmid from Bert O’Malley and Ming Tsai (Baylor 112 College of Medicine, Houston TX). Using co-transfection experiments, we determined if any of these factors alone or in combination could augment LPS-induced promoter activity. The transfection experiment was performed twice, in duplicate, with similar results. Figure 17 shows the results from one of the replicates. None of the factors, alone or in combination caused a dramatic increase or decrease in basal or LPS—induced promoter activity. Previously overexpression of CREB has been shown to blocked LPS induced COX-2 promoter activity; however, we did not observe this negative regulation. To determine if NF-KB transcription factors could potentiate LPS induced COX-2 promoter activity, we co-transfected p50 and p65 (obtained from Richard Schwartz, Michigan State University, East Lansing, MI.) alone and with the SRC-la expression plasmid (Figure 18). The transfection was performed twice, in duplicate, with similar results. The results reported are from one representative set of transfections. None of the experiments resulted in an increase of either the basal or induced level of promoter activity to a degree that suggests a substantial interaction. Discussion. Co-transfection of the transcription factors and transcriptional coactivators that we expected to be associated with the upstream promoter region had small effects on the basal and LPS induced levels of COX-2 promoter activity. This may indicate that sufficient levels of these factors are already present in the cell such that addition of more factor has no effect. Alternatively these factors may not be important for the activation of COX-2 gene transcription. 113 LPS- . LPS+ 3500 --- 2500 2000 1500 -++ . ++. +++ Figure 17. Co-transfection of the -966/+23 COX-2 promoter reporter with CREB, CBP and SRC-l. RAW cells were stimulated with LPS for 12 hr. The results are representative of two separate experiments prepared in duplicate. 114 Subbaramaiah et al. [90] observed that glucocorticoids block induced COX-2 promoter activity and that co-transfection of CBP and c-Jun restored promoter activity.Based on these observations, it may be necessary to disrupt the coactivator complex in order to detect an effect of adding the components back to the system. Methods of interrupting transactivation of the coactivator complex include pretreatment with glucocorticoids or PPARY ligands or the addition of Ela adneovirus oncoprotein6 [87, 89, 224, 227-229]. Another possibility is that the interactions between the factors bound at the CRE-2 and NF-KB sites are concentration dependent, where the binding of one factor is dependent on the binding of the other. This could be tested by overexpressing one factor and titrating the level of the other. If, for example, p65 binding was dependent on CREB occupation of the CRE-2 site, then increasing levels of CREB my enhance promoter activity in the presence of overexpressed p65. When these experiments were performed I focused on the role of p50. It may be useful to further explore the role of p65 as an enhancer. If p65 and CBP form a transcriptionally active complex on the COX-2 promoter, it may be possible to enhance promoter activity by overexpressing combinations of p65, CBP, and SRC-l. Sheppard et al. found this combination to enhance E-selectin promoter activity [224]. 6 We have obtained an Ela expression plasmid (Nicholas Dyson and Fred Dick, Massachusetts General Hospital, Charlestown MA.) but have not yet completed these experiments. 115 3000 DLPS- 2500 “F ILPS+ 2000 I ~ 1 . l 1 I I 500- ._ I . a Empty p50 p65 PSO+P65 SRC-l p50+SRC-1 Vector. Figure 18. Co-transfection of the -966/+23 COX-2 promoter reporter with the plasmids for the expression of the NF-KB proteins p50 and p65 and the p160 family co-activator SRC-l. RAW cells were stimulated with LPS for 12 hr. The results are representative of two separate experiments prepared in duplicate.-. 116 Shifting for Cooperative Complexes Promoter activity assays provided data leading to the assertion that the NF-KB and CRE-2 response elements mediate a cooperative effect on transcriptional activation. We thought that if the CRE-2 and NF-KB sites were on the same EMSA probe, we might be able to detect a very low mobility complexe that could be eliminated by the mutation of either the CRE-2 or NF-KB site. Results. We constructed a 59 bp probe (—444 to -385) that contained both the CRE-2 and NF-KB sites, as well as constructs with a mutation in one or both sites. As a control, we also used a shorter probe based on the sequence between the CRE-2 and NF- KB sites (middle probe) and the two independent response element probes. The purpose of the middle probe was to verify that sequences between the CRE—2 and NF-KB response elements did not exhibit any additional unique shifted species (Figure 19). We found that the 59 bp probe is bound by a complex that is dependent on the CRE-2 site that is not formed by the CRE-2 probe alone. This complex is most prominent at the 1 hr time point and is not visible at the 12 hr time point. (Figure 20). Discussion. In all the previous EMSA experiments, we did not observe inducible binding or modification of the complex formed with the CRE-2 probe. In this experiment, we observed an inducible effect that is dependent on the CRE-2 site. This observation is not consistent with the model of a cooperative effect because the NF-KB mutation does not affect the low mobility band. It is difficult to explain exactly what this complex represents; however, the formation this type of low mobility complex indicates that a more highly ordered complex forms on this probe. 117 CRE-2 NF-KB TGTGCGTGCTCT GAGCAGCGAG o ACGTCA ACTGCGCC CCAGTGGG GAGAGGTGAc GGGAlTC AGTTAG GACCTTAGATCCCG Probe CRE-2 NF-KB Figure 19. EMSA probe map 118 Probe CRE-2 M1ddle NF-KB LPS(hr.) 9.. 1 L2. 0 1 12 CRE-2 _> Complex ‘ " NF'KB ' 4- Complexes Mutant Mutant Mutant CRE-2 and Wild Type CRE-2 NF-KB NF—KB LPS(hr)W01120 112 0 1120 112 Low Mobility Complex + CRE-2 and ‘ Complexes 1 Figure 20. Comparison of shifted products between the site specific probes and a probe containing both the CRE-2 and NF-KB sites. A) EMSAs with the CRE-2, Middle, and NF-KBprobes. B) EMSAs with the 59 bp probe, containing the CRE-2 and NF-KB sites. Nuclear extracts were prepared from RAW 264.7 cell stimulated with LPS (200 ng/ml) for 1 or 12 hours. Binding with 12P end labeled dsDNA probe was performed at room temperature in 20 mM HEPES (pH 7.9), 100 mM KCL, 50 ng Poly dIdC, 4 mM DTT, and protease inhibitors. Binding reactions were electrophoresed on a 5% non-denaturing acrylamide gel. 119 Conclusion We suspect that the COX-2 promoter is activated by two fundamentally different processes. First, transcription factors act as modular enhancers to rapidly and transiently activate gene transcription. Second, a more complex activation mechanism is required for prolonged activation or activation that is stimulus and tissue specific. In this second case, activation may be regulated by a highly ordered enhancesome complex that may share some similarities with IFN-B promoter (Briefly reviewed in reference [230]). 120 APPENDIX A Caveats of Signaling Experiments Before considering the specific details of LPS-induced signaling pathways in macrophage, an important factor should be considered. These signaling pathways assume to some extent that murine and human alveolar, peripheral, and peritoneal primary macrophage cells and the plethora of monocytic and macrophage-like cell lines are essentially similar. Since this assumption is not correct, the information discussed here will be limited to what is observed in RAW 264.7 murine macrophage cells when there are conflicts between observations in different cell types. Also important to consider is the method by which these signaling pathways were tested. One general method employs the use of kinase inhibitors. Most of these compounds have a very narrow window of semi—specific activity, and authors often use concentrations that may also non-specifically inhibit other kinases that may be functioning in parallel. This problem is compounded by the over interpretation of results where only partial inhibition of a kinase activity is observed. Another general method of studying signaling networks is to overexpress functional enzymes, constitutively active enzymes, or non-functional enzymes. These dominant active and dominant negative experiments result in hyper-activation of a pathway or blockade of a pathway by out competing the native kinase substrate with that of the non-functional protein. Another problem with adding an active kinase is that it may activate a signaling pathway even though it is not present in the cell under normal conditions. Nevertheless, the conclusions from the data of signaling experiments are the 121 best approximation of what happens in the cell and should not be discarded out-of-hand and are useful in making inferences to the best possible explanation. 122 10. REFERENCES Seibert, K., Y. Zhang, K. Leahy, S. Hauser, J. Masferrer, W. Perkins, L. Lee and P. Isakson, Pharmacological and biochemical demonstration of the role of cyclooxygenase 2 in inflammation and pain. Proceedings of the National Academy of Sciences of the United States of America, 1994. 91(25): p. 12013-7. Narumiya, S. and GA. FitzGerald, Genetic and pharmacological analysis of prostanoid receptor function. Journal of Clinical Investigation, 2001. 108(1): p. 25-30. Kerr, S. and J. Gailer, COX-Z-Selective NSAIDs: New Wonder Drugs?, in http://nps.org.au/pdf/Condocumentpdf. 1999, National Perscribing Service Ltd., AU: Surry Hills NSW, Australia. Morham, S.G., R. Langenbach, C.D. Loftin, H.F. Tiano, N. Vouloumanos, J .C. Jennette, J .F. Mahler, K.D. Kluckman, A. Ledford, C.A. Lee and e. al., Prostaglandin synthase 2 gene disruption causes severe renal pathology in the mouse. Cell, 1995. 83(3): p. 473—482. Komhoff, M., H]. Grone, T. Klein, H.W. Seyberth and RM. Nusing, Localization of cyclooxygenase-1 and -2 in adult and fetal human kidney: implication for renal function. American Journal of Physiology, 1997. 272(4 Pt 2): p. F460-F468. Dubois, R.N., S.B. Abramson, L. Crofford, R.A. Gupta, L.S. Simon, L.B. Van- De-Putte and 1P.E. Lipsky, Cyclooxygenase in biology and disease. The FASEB Journal, 1998. 12(12): p. 1063-1073. Pilbeam, C.C., J .R. Harrison and LG. Raisz, Prostaglandins and Bone Metabolism, in Principles of Bone Biology, J .P. Bilezikian, L.G. Raisz, and GA. Rodan, Editors. 1996, Academic Press Inc.: New York. p. 715-727. Smith, W.L. and R. Langenbach, Why there are two cyclooxygenase isozymes. Journal of Clinical Investigation, 2001. 107(12): p. 1491-5. Smith, W.L., D.L. DeWitt and RM. Garavito, Cyclooxygenases: Structural, cellular and molecular biology. Annual Review of Biochemistry, 2000. 69: p. 145—82. Segi, E., Y. Sugimoto, A. Yamasaki, Y. Aze, H. Oida, T. Nishimura, T. Murata, T. Matsuoka, F. Ushikubi, M. Hirose, T. Tanaka, N. Yoshida, S. Narumiya and A. Ichikawa, Patent ductus arteriosus and neonatal death in prostaglandin receptor EP4-deficient mice. Biochemical and Biophysical Research Communications, 1998. 246(1): p. 7-12. 123 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. Ushikubi, F., M. Hirata and S. N arumiya, Molecular biology of prostanoid receptors; an overview. Journal of lipid mediators and cell signalling, 1995. 12(2- 3): p. 343-59. Ding, M., Y. Kinoshita, K. Kishi, H. Nakata, S. Hassan, C. Kawanami, Y. Sugimoto, M. Katsuyama, M. Negishi, S. Narumiya, A. Ichikawa and T. Chiba, Distribution of prostaglandin E receptors in the rat gastrointestinal tract. Prostaglandins, 1997. 53(3): p. 199-216. Narumiya, S., Y. Sugimoto and F. Ushikubi, Prostanoid receptors: structures, properties, and functions. physiological reviews, 1999. 79(4): p. 1193-226. Sugimoto, Y., S. N arumiya and A. Ichikawa, Distribution and function of prostanoid receptors: studies from knockout mice. Progress in lipid research, 2000. 39(4): p. 289-314. Tilley, S.L., T.M. Coffman and EH. Koller, Mixed messages: modulation of inflammation and immune responses by prostaglandins and thromboxanes. Journal of Clinical Investigation, 2001. 108(1): p. 15-23. Katori, M. and M. Majima, Cyclooxygenase-2: its rich diversity of roles and possible application of its selective inhibitors. Inflammation research : official journal of the European Histamine Research Society [et a1 ], 2000. 49(8): p. 367- 92. Smith, W.L. and D.L. Dewitt, Prostaglandin endoperoxide H synthases-I and -2. Advances in Immunology, 1996. 62: p. 167-215. Tortora, OJ. and NP. Anagnostakos, Principles of Anatomy and Physiology. 6 ed. 1990, New York: Harper and Row. Seibert, K., Y. Zhang, K. Leahy, S. Hauser, J. Masferrer, W. Perkins, L. Lee and P. Isakson, Pharmacological and biochemical demonstration of the role of cylooxygenase 2 in inflammation and pain. Procedings of the National Academy of Science USA, 1994. 91: p. 12013-12017. Portanova, J .P., Y. Zhang, G.D. Anderson, S.D. Hauser, J .L. Masferrer, K. Seibert, S.A. Gregory and RC. Isakson, Selective neutralization of prostaglandin E2 blocks inflammation, hyperalgesia, and interleukin 6 production in vivo. The Journal of Experimental Medicine, 1996. 184(3): p. 883-91. Minami, T., H. Nakano, T. Kobayashi, Y. Sugimoto, F. Ushikubi, A. Ichikawa, S. N arumiya and S. Ito, Characterization of EP receptor subtypes responsible for prostaglandin EZ-induced pain responses by use of EP1 and EP3 receptor knockout mice. British journal of pharmacology, 2001. 133(3): p. 438-44. 124 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. Gilroy, D.W., P.R. Colville-Nash, D. Willis, J. Chivers, M.J. Paul-Clark and DA. Willoughby, Inducible cyclooxygenase may have anti-inflammatory properties. Nature medicine, 1999. 5(6): p. 698-701. Mizuno, H., C. Sakamoto, K. Matsuda, K. Wada, T. Uchida, H. Noguchi, T. Akamatsu and M. Kasuga, Induction of cyclooxygenase 2 in gastric mucosal lesions and its inhibition by the specific antagonist delays healing in mice. Gastroenterology, 1997. 112(2): p. 387-97. Thuresson, E.D., K.M. Lakkides, C.J. Rieke, Y. Sun, B.A. Wingerd, R. Micielli, A.M. Mulichak, M.G. Malkowski, R.M. Garavito and W.L. Smith, Prostaglandin endoperoxide H synthase-1: the functions of cyclooxygenase active site residues in the binding, positioning, and oxygenation of arachidonic acid. Journal of Biological Chemistry, 2001. 276(13): p. 10347-57. Murakami, M., S. Shimbara, T. Karnbe, H. Kuwata, M.V. Winstead, J .A. Tischfield and 1. Kudo, The functions of five distinct mammalian phospholipase A2S in regulating arachidonic acid release. Type Ila and type V secretory phospholipase AZS are functionally redundant and act in concert with cytosolic phospholipase A2. Journal of Biological Chemistry, 1998. 273(23): p. 14411- 14423. Lin, L.L., M. Wartmann, A.Y. Lin, J.L. Knopf, A. Seth and R]. Davis, cPLAZ is phosphorylated and activated by MAP kinase. Cell, 1993. 72(2): p. 269-78. Smith, W.L., R.M. Garavito and D.L. DeWitt, Prostaglandin endoperoxide H synthases (cyclooxygenases)-I and -2. Journal of Biological Chemistry, 1996. 271(52): p. 33157-60. Marnett, L.J., S.W. Rowlinson, D.C. Goodwin, A.S. Kalgutkar and CA. Lanzo, Arachidonic acid oxygenation by C OX-I and COX-2. Journal of Biological Chemistry, 1999. 274: p. 22903-22906. Taketo, M.M., Cyclooxygenase-2 inhibitors in tumorigenesis (Part II). Journal of the National Cancer Institute, 1998. 90(21): p. 1609-1620. Spencer, A.G., J .W. Woods, T. Arakawa, Singer, II and W.L. Smith, Subcellular localization of prostaglandin endoperoxide H synthases-I and -2 by immunoelectron microscopy. Journal of Biological Chemistry, 1998. 273(16): p. 9886-93. J akobsson, P., S. Thoren, R. Morgenstem and B. Samuelsson, Identification of human prostaglandin E synthase: A microsomal glutathione-dependent, inducible enzyme, constituting a potential novel drug target. Procedings of the National Academy of Science USA, 1999. 96: p. 7220-7225. Tanioka, T., Y. Nakatani, N. Semmyo, M. Murakami and 1. Kudo, Molecular identification of cytosolic prostaglandin E2 synthase that is functionally coupled 125 33. 34. 35. 36. 37. 38. 39. 40. 41. with cyclooxygenase-I in immediate prostaglandin E2 biosynthesis. Journal of Biological Chemistry, 2000. 275(42): p. 32775-82. Murakami, M., H. Naraba, T. Tanioka, N. Semmyo, Y. Nakatani, F. Kojima, T. Ikeda, M. Fueki, A. Ueno, S. Oh and 1. Kudo, Regulation of prostaglandin E2 biosynthesis by inducible membrane-associated prostaglandin E2 synthase that acts in concert with cyclooxygenase-2. Journal of Biological Chemistry, 2000. 275(42): p. 32783-92. Matsumoto, H., H. Naraba, M. Murakami, 1. Kudo, K. Yarnaki, A. Ueno and S. Oh-ishi, Concordant induction of prostaglandin E2 synthase with cyclooxygenase-2 leads to preferred production of prostaglandin E2 over thromboxane and prostaglandin D2 in ulated rat peritoneal macrophages. Biochemical and Biophysical Research Communications, 1997. 230(1): p. 110-4. Naraba, H., M. Murakami, H. Matsumoto, S. Shimbara, A. Ueno, 1. Kudo and S. Oh-ishi, Segregated coupling of phospholipases A2, cyclooxygenases, and terminal prostanoid synthases in difi‘erent phases of prostanoid biosynthesis in rat peritoneal macrophages. Journal of Immunology, 1998. 160(6): p. 2974-2982. Foumier, T., V. Fadok and RM. Henson, Tumor necrosis factor-alpha inversely regulates prostaglandin D2 and prostaglandin E2 production in marine macrophages. Synergistic action of cyclic AMP on cyclooxygenase-2 andin E2 synthesis. Journal of Biological Chemistry, 1997. 272(49): p. 31065—72. Brock, T.G., R.W. McNish and M. Peters-Golden, Arachidonic acid is preferentially metabolized by cyclooxygenase-2 to prostacyclin and prostaglandin E2. Journal of Biological Chemistry, 1999. 274(17): p. 11660-11666. Hong, S.L. and L. Levine, Stimulation of prostaglandin synthesis by bradykinin and thrombin and their mechanism of action on MC5—5 fibroblasts. Journal of Biological Chemistry, 1976. 251: p. 5814-5816. Pong, S.S., S.L. Hong and L. Levine, Prostaglandin production by methylcholanthrene-transformed mouse BALE/3 T3. Requirement for protein synthesis. Journal of Biological Chemistry, 1977. 252(4): p. 1408-13. Ohuchi, K. and L. Levine, Stimulation of prostaglandin synthesis by tumor- promoting phorbol-12, I3-diesters in canine kidney (MDCK) cells. Cycloheximide inhibits the stimulated prostaglandin synthesis, and morphological changes. Journal of Biological Chemistry, 1978. 253(13): p. 4783-90. Raz, A., A. Wyche, N. Siegel and P. Needleman, Regulation of fibroblast cyclooxygenase synthesis by interleukin-I. Journal of Biological Chemistry, 1988. 263(6): p. 3022-3028. 126 42. 43. 45. 46. 47. 48. 49. 50. 51. 52. Fu, J ., J .L. Masferrer, K. Seibert, A. Raz and P. Needleman, The induction and suppression of prostaglandin H2 synthase (cyclooxygenase) in human monocytes. Journal of Biological Chemistry, 1990. 265(28): p. 16737-16740. Simonson, M.S., J .A. Wolfe, M. Konieczkowski, J .R. Sedor and M.J. Dunn, Regulation of prostaglandin endoperoxide synthase gene expression in cultured rat mesangial cells: induction by serum via a protein kinase C-dependent mechanism. Molecular Endocrinology, 1991. 5: p. 441-451. Lin, A.H., M.J. Bienkowski and RR. Gorman, Regulation of prostaglandin H synthase mRNA levels and prostaglandin biosynthesis by platelet-derived growth factor. Journal of Biological Chemistry, 1989. 264(29): p. 17379-17383. Herschman, H.R., Primary response genes induced by growth factors and tumor promoters. Annual Review of Biochemistry, 1991. 60: p. 281-319. Lim, R.W., B.C. Varnum and HR. Herschman, Cloning of tetradecanoyl phorbol ester-induced 'primary response' sequences and their expression in density- arrested Swiss 3 T3 cells and a TPA non-proliferative variant. Oncogene, 1987. 1(3): p. 263-70. Simmons, D.L., D.B. Levy, Y. Yannoni and R.L. Erikson, Identification of a phorbol ester-repressible v-src-inducible gene. Proceedings of the National Academy of Sciences USA, 1989. 86(4): p. 1178-82. Rosen, G.D., T.M. Birkenmeier, A. Raz and M.J. Holtzman, Identification of a cylcooxygenase-related gene and its potential role in prostaglandin formation. Biochemical and Biophysical Research Communications, 1989. 164(3): p. 1358- 1365. Han, J ., H. Sadowski, D.A. Young and LG. Macara, Persistant induction of cyclooxygenase in p60v-src-transf0rmed 3 T3 fibroblasts. Procedings of the National Academy of Science USA, 1990. 87: p. 3373-3377. Xie, W., J .G. Chipman, D.L. Robertson, R.L. Erikson and D.L. Simmons, Expression of a mitogen-responsive gene encoding prostaglandin synthase is regulated by mRNA splicing. Procedings of the National Academy of Science USA, 1991. 88: p. 2692-2696. Kujubu, D.A., B.S. Fletcher, B.C. Varnum, R.W. Lim and HR. Herschman, TISIO, a phorbol ester tumor promoter inducible mRNA from Swiss 3 T3 cells, encodes a novel prostaglandin synthase/cyclooxygenase homologue. Journal of Biological Chemistry, 1991. 266(20): p. 12866-12872. O'Banion, M.K., H.B. Sadowski, V. Winn and DA. Young, A serum- and glucocorticoid-regulated 4 kb mRNA encodes a cyclooxygenase-related protein. Journal of Biological Chemistry, 1991. 266: p. 23261-23267. 127 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. Sirois, J. and J .S. Richards, Purification and characterization of a novel, distinct isoform of prostaglandin endoperoxide synthase induced by human chorionic gonadatropin in granulosa cells of rat preovulatory follicles. Journal of Biological Chemistry, 1992. 267: p. 6382-6388. Wong, W.Y. and J .S. Richards, Evidence for two antigenically distinct molecular weight variants of prostaglandin H synthase in the rat ovary. Molecular endocrinology (Baltimore, Md ), 1991. 5(9): p. 1269-79. Fletcher, B.S., D.A. Kujubu, D.M. Perrin and HR. Herschman, Structure of the mitogen-inducible TISIO gene and demonstration that the TISIO-encoded protein is a fimctional prostaglandin G/H synthase. Journal of Biological Chemistry, 1992. 267: p. 4338-4344. Kujubu, DA. and HR. Herschman, Dexamethasone inhibits mitogen induction of the T15 10 prostaglandin synthase/cyclooxygenase gene. Journal of Biological Chemistry, 1992. 267(12): p. 7991-4. O'Banion, M.K., V.D. Winn and DA. Young, cDNA Cloning and functional activity of a glucocorticoid-regulated inflammatory cyclooxygenase. Procedings of the National Academy of Science USA, 1992. 89: p. 4888-4892. 1r Ryseck, R., C. Raynoschek, H. Macdonald-Bravo, K. Dorfman and R. Bravo, Identification of an immediate early gene, pghs-b, whose protein product has prostaglandin synthase/cylooxygenase activity. Cell Growth and Differentiation, 1992. 3: p. 443-450. Xu, X.M., J .L. Tang, A. Hajibeigi, D.S. Loose-Mitchell and K.K. Wu, Transcriptional regulation of endothelial constitutive PGHS-I expression by phorbol ester. The American journal of physiology, 1996. 270(1) Pt 1): p. C259- 64. Antes, T.J., S.A. Goodart, C. Huynh, M. Sullivan, S.G. Young and B. Levy- Wilson, Identification and characterization of a 315-base pair enhancer, located more than 55 kilobases 5 ' of the apolipoprotein B gene, that confers expression in the intestine. Journal of Biological Chemistry, 2000. 275(34): p. 26637-48. DeWitt, D.L. and EA. Meade, Serum and glucocorticoid regulation of gene transcription and expression of the prostaglandin H synthase-I and prostaglandin H synthase-2 isozymes. Archives of Biochemistry and Biophysics, 1993. 306(1): p. 94- 102. Xie, W., B.S. Fletcher, R.D. Andersen and HR. Herschman, v-src induction of the TISIO/PGSZ prostaglandin synthase gene is mediated by an ATF/CRE transcription response element. Molecular and Cellular Biology, 1994. 14(10): p. 6531-6539. 128 63. 65. 66. 67. 68. 69. 70. 71. 72. 73. Xie, W., J .R. Merrill, W.S. Bradshaw and D.L. Simmons, Structural determination and promoter analysis of the chicken mitogen-inducible prostaglandin G/H synthase gene and genetic mapping of the marine homolog. Archives of Biochemistry and Biophysics, 1993. 300(1): p. 247-52. Xie, W. and HR. Herschman, v-src induces prostaglandin synthase 2 gene expression by activation of the c-Jun N-terminal kinase and the c-Jun transcription factor. Journal of Biological Chemistry, 1995. 270(46): p. 27622- 27628. Xie, W. and HR. Herschman, Transcriptional regulation of prostaglandin synthase 2 gene expression by platelet-derived growth factor and serum. Journal of Biological Chemistry, 1996. 271(49): p. 31742-31748. Saunders, M.A., L. Sansores-Garcia, D.W. Gilroy and K.K. Wu, Selective suppression of C CAA T/enhancer-binding protein beta binding and cyclooxygenase-2 promoter activity by sodium salicylate in quiescent human fibroblasts. Journal of Biological Chemistry, 2001. 276(22): p. 18897-904. Faour, W.H., Y. He, Q.W. He, M. de Ladurantaye, M. Quintero, A. Mancini and J .A. Di Battista, Prostaglandin E(2) regulates the level and stability of cyclooxygenase-2 mRNA through activation of p38 mitogen-activated protein kinase in interleukin-1 beta-treated human. Journal of Biological Chemistry, 2001. 276(34): p. 31720-31. Smalley, W.E. and RN. DuBois, Colorectal cancer and nonsteroidal anti- inflammatory drugs. Advances in Pharmacology, 1997. 39: p. 1-20. Tsujii, M., S. Kawano, S. Tsuji, H. Sawaoka, M. Hori and RN. DuBois, Cyclooxygenase regulates angiogenesis induced by colon cancer cells. Cell, 1998. 93(5): p. 705-716. Muller-Decker, K., A. Kopp-Schneider, F. Marks, K. Seibert and G. Ffirstenberger, Localization of prostaglandin H synthase isoenzymes in murine epidermal tumors: suppression of skin tumor promotion by inhibition of prostaglandin H synthase-2. Molecular carcinogenesis, 1998. 23(1): p. 36-44. Tsujii, M. and RN. DuBois, Alterations in cellular adhesion and apoptosis in epithelial cells overexpressing prostaglandin endoperoxide synthase 2. Cell, 1995. 83(3): p. 493-501. Oshima, M., J .E. Dinchuk, S.L. Kargman, H. Oshima, B. Hancock, E. Kwong, J .M. Trzaskos, J .F. Evans and M.M. Taketo, Suppression of intestinal polyposis in Apc delta716 knockout mice by inhibition of cyclooxygenase 2 ( C 0X—2 ). Cell, 1996. 87(5): p. 803-809. Chan, G., J .O. Boyle, E.K. Yang, F. Zhang, P.G. Sacks, J .P. Shah, D. Edelstein, R.A. Soslow, A.T. Koki, B.M. Woemer, J .L. Masferrer and A.J. Dannenberg, 129 74. 75. 76. 77. 78. 79. 80. 81. 82. C yclooxygenase-Z expression is up-regulated in squamous cell carcinoma of the head and neck. Cancer Research, 1999. 59(5): p. 991-994. Subbaramaiah, K., N. Telang, J .T. Ramonetti, R. Araki, B. DeVito, B.B. Weksler and A.J. Dannenberg, Transcription of cyclooxygenase-2 is enhanced in transformed mammary epithelial cells. Cancer Research, 1996. 56(19): p. 4424- 4429. Kim, Y. and S.M. Fischer, Transcriptional regulation of cyclooxygenase-2 in mouse skin carcinoma cells. (Regulatory role of C CAAT/Enhancer—binding proteins in the differential expression of Cyclooxygenase-2 in normal and neoplastic tissues. ). Journal of Biological Chemistry, 1998. 273: p. 27686-27694. f} Howe, L.R., H.C. Crawford, K. Subbaramaiah, J .A. Hassell, A.J. Dannenberg and A.M. Brown, PEA3 is up-regulated in response to WM] and activates the expression of cyclooxygenase-2. Journal of Biological Chemistry, 2001. 276(23): p. 20108-15. Howe, L.R., K. Subbaramaiah, W.J. Chung, A.J. Dannenberg and A.M. Brown, is" Transcriptional activation of cyclooxygenase-2 in Wnt-I -transformed mouse mammary epithelial cells. Cancer Research, 1999. 59(7): p. 1572-1577. Kelley, D.J., J .R. Mestre, K. Subbaramaiah, P.G. Sacks, S.P. Schantz, T. Tanabe, H. Inoue, J .T. Ramonetti and A.J. Dannenberg, Benzo[ a [pyrene up-regulates cyclooxygenase-2 gene expression in oral epithelial cells. Carcinogenesis, 1997. 18(4): p. 795-799. Michaluart, P., J .L. Masferrer, A.M. Carothers, K. Subbaramaiah, B.S. Zweifel, C. Koboldt, J .R. Mestre, D. Grunberger, P.G. Sacks, T. Tanabe and A.J. Dannenberg, Inhibitory effects of cafi'eic acid phenethyl ester on the activity and expression of cyclooxygenase-2 in human oral epithelial cells and in a rat model of inflammation. Cancer Research, 1999. 59: p. 2347-2352. Zhang, F., K. Subbaramaiah, N. Altorki and A.J. Dannenberg, Dihydroxy bile acids activate the transcription of cyclooxygenase-2. Journal of Biological Chemistry, 1998. 273(4): p. 2424-2428. Zhang, F., N.K. Altorki, J .R. Mestre, K. Subbaramaiah and A.J. Dannenberg, Curcumin inhibits cyclooxygenase-2 transcription in bile acid- and phorbol ester- treated human gastrointestinal epithelial cells. Carcinogenesis, 1999. 20(3): p. 445-451. Subbaramaiah, K., W.J. Chung and A.J. Dannenberg, Ceramide regulates the transcription of cyclooxygenase-2. Evidence for involvement of extracellular signal-regulated kinase/c-Jun N-terminal kinase and p38 mitogen-activated protein kinase pathways. Journal of Biological Chemistry, 1998. 273(49): p. 32943-32949. 130 83. 84. 85. 86. 87. 88. 89. 90. 91. 92. 93. Subbaramaiah, K., W.J. Chung, P. Michaluart, N. Telang, T. Tanabe, H. Inoue, M. J ang, J .M. Pezzuto and A.J. Dannenberg, Resveratrol inhibits cyclooxygenase- 2 transcription and activity in phorbol ester-treated human mammary epithelial cells. Journal of Biological Chemistry, 1998. 273(34): p. 21875-21882. Mestre, J.R., K. Subbaramaiah, P.G. Sacks, S.P. Schantz, T. Tanabe, H. Inoue and A.J. Dannenberg, Retinoids suppress epidermal growth factor-induced transcription of cyclooxygenase-2 in human oral squamous carcinoma cells. Cancer Research, 1997. 57(14): p. 2890-2895. Glass, GK. and MG. Rosenfeld, The coregulator exchange in transcriptional h fitnctions of nuclear receptors. Genes & Development, 2000. 14(2): p. 121-41. 1 ' Kamei, Y., L. Xu, T. Heinzel, J. Torchia, R. Kurokawa, B. Gloss, S.C. Lin, R.A. Heyman, D.W. Rose, C.K. Glass and MG. Rosenfeld, A CBP integrator complex mediates transcriptional activation and AP-I inhibition by nuclear receptors. Cell, 1996. 85(3): p. 403-14. Goodman, RH. and S. Smolik, CBP/p300 in cell growth, transformation, and L development. Genes & Development, 2000. 14(13): p. 1553-77. Giordano, A. and ML. Avantaggiati, p300 and CBP: partners for life and death. Journal of cellular physiology, 1999. 181(2): p. 218-30. Ricote, M., J .T. Huang, J .S. Welch and CK. Glass, The peroxisome proliferator— activated receptor(PPARgamma) as a regulator of monocyte/macrophage function. Journal of Leukocyte Biology, 1999. 66(5): p. 733-9. Subbaramaiah, K., D.T. Lin, J .C. Hart and A.J. Dannenberg, Peroxisome proliferator-activated receptor gamma ligands suppress the transcriptional activation of cyclooxygenase-2. Evidence for involvement of activator protein-I and CBP/p300. Journal of Biological Chemistry, 2001. 276(15): p. 12440-8. Jones, D.A., D.P. Carlton, T.M. McIntyre, G.A. Zimmerman and S.M. Prescott, Molecular cloning of human prostaglandin endoperoxide synthase type II and demonstration of expression in response to cytokines. Journal of Biological Chemistry, 1993. 268: p. 9049-9054. Inoue, H., C. Yokoyama, S. Hara, Y. Tone and T. Tanabe, Transcriptional regulation of human prostaglandin-endoperoxide synthase-2 gene by lipopolysaccharide and phorbol ester in vascular endothelial cells. Involvement of both nuclear factor for interleukin-6 expression site and CAMP response element. Journal of Biological Chemistry, 1995. 270(42): p. 24965-24971. Inoue, H., C. Yokoyama, S. Hara, Y. Tone and T. Tanabe, Transcriptional regulation of human prostaglandin-endoperoxide synthase-2 gene by lipopolysaccharide and phorbol ester in vascular endothelial cells. Involvement of 131 94. 95. 96. 97. 98. 99. 100. 101. 102. both rleukin-6 expression site and CAMP response element. Journal of Biological Chemistry, 1995. 270(42): p. 24965—71. Tsukada, J ., K. Saito, W.R. Waterman, A.C. Webb and RE. Auron, Transcription factors NF -IL6 and CREB recognize a common essential site in the human prointerleukin 1 beta gene. Molecular and Cellular Biology, 1994. 14(11): p. 7285-97. Schmedtje, J .F., Jr., Y.S. Ji, W.L. Liu, R.N. DuBois and MS. Runge, Hypoxia induces cyclooxygenase-2 via the NF -kappa B p65 transcription factor in human vascular endothelial cells. Journal of Biological Chemistry, 1997. 272(1): p. 601- 8. F Rupee, RA. and PA. Baeuerle, The genomic response of tumor cells to hypoxia and reoxygenation. Differential activation of transcription factors AP-I and NF - kappa B. European journal of biochemistry / FEBS, 1995. 234(2): p. 632-40. Miyaura, C., M. Inada, T. Suzawa, Y. Sugimoto, F. Ushikubi, A. Ichikawa, S. Narumiya and T. Suda, Impaired bone resorption to prostaglandin E2 in prostaglandin E receptor EP4-knockout mice. Journal of Biological Chemistry, 2000. 275(26): p. 19819-23. Suzawa, T., C. Miyaura, M. Inada, T. Maruyama, Y. Sugimoto, F. Ushikubi, A. Ichikawa, S. Narumiya and T. Suda, The role of prostaglandin E receptor subtypes (EP1, EP2, EP3, and EP4) in bone resorption: an analysis using specific agonists for the respective EPs. Endocrinology, 2000. 141(4): p. 1554-9. Kawaguchi, H., R. Yavari, M.L. Stover, D.W. Rowe, L.G. Raisz and CC. Pilbeam, Measurement of interleukin-1 stimulated constitutive prostaglandin G/H synthase (cyclooxygenase) mRNA levels in osteoblastic MC3T3-EI cells using competitive reverse e chain reaction. Endocrine research, 1994. 20(3): p. 219-33. Yokota, K., M. Kusaka, T. Ohshima, S. Yamamoto, N. Kurihara, T. Yoshino and M. Kumegawa, Stimulation of prostaglandin E2 synthesis in cloned osteoblastic cells of mouse (M C3T3-EI ) by epidermal growth factor. Journal of Biological Chemistry, 1986. 261(33): p. 15410-15415. Pilbeam, C.C., H. Kawaguchi, Y. Hakeda, O. Voznesensky, C.B. Alander and LG. Raisz, Differential regulation of inducible and constitutive prostaglandin endoperoxide synthase in osteoblastic M C3 T3-E 1 cells. Journal of Biological Chemistry, 1993. 268(34): p. 25643-9. Yamamoto, K., T. Arakawa, N. Ueda and S. Yamamoto, Transcriptional roles of nuclear factor kappa B and nuclear factor-interleukin-6 in the tumor necrosis factor alpha-dependent induction of cyclooxygenase-2 in MC3T3—EI cells. Journal of Biological Chemistry, 1995. 270(52): p. 31315-31320. 132 103. 104. 105. 106. 107. 108. 109. 110. 111. 112. Sato, K., K. Kasono, Y. Fujii, M. Kawakami, T. Tsushima and K. Shizume, Tumor necrosis factor type alpha (cachectin) stimulates mouse osteoblast-like cells (M C3 T3-E I ) to produce macrophage-colony stimulating activity and prostaglandin E2. Biochemical and Biophysical Research Communications, 1987. 145(1): p. 323-9. Yanaga, F., M. Abe, T. Koga and M. Hirata, Signal transduction by tumor necrosis factor alpha is mediated through a guanine nucleotide-binding protein in osteoblast-like cell line, MC3T3-EI. Journal of Biological Chemistry, 1992. 267(8): p. 5114-21. Rapuano, BE. and RS. Bockman, Tumor necrosis factor-alpha stimulates phosphatidylinositol breakdown by phospholipase C to coordinately increase the levels of diacylglycerol, free arachidonic acid and teoblast (MC3T3-EI ) cell line. Biochimica et biophysica acta, 1991. 1091(3): p. 374-84. Matsusaka, T., K. Fujikawa, Y. Nishio, N. Mukaida, K. Matsushima, T. Kishimoto and S. Akira, Transcription factors NF -IL6 and NF -kappa B synergistically activate transcription of the inflammatory cytokines, interleukin 6 and interleukin 8. Proceedings of the National Academy of Sciences USA, 1993. 90(21): p. 10193-7. Wadleigh, DJ. and HR. Herschman, Transcriptional regulation of the cyclooxygenase-2 gene by diverse ligands in murine osteoblasts. Biochemical and Biophysical Research Communications, 1999. 264(3): p. 865-70. Okada, Y., O. Voznesensky, H. Herschman, J. Harrison and C. Pilbeam, Identification of multiple cis-acting elements mediating the induction of prostaglandin G/H synthase-2 by phorbol ester in murine osteoblastic cells. journal of cellular biochemistry, 2000. 78(2): p. 197-209. Ogasawara, A., T. Arakawa, T. Kaneda, T. Takuma, T. Sato, H. Kaneko, M. Kumegawa and Y. Hakeda, Fluid shear stress-induced cyclooxygenase-2 expression is mediated by C/EBP beta, CAMP-response element-binding protein, and AP-l in osteoblastic MC3T3-E1 cells. Journal of Biological Chemistry, 2001. 276(10): p. 7048-54. Yamamoto, K., J. Wang, S. Yamamoto and H. Tobe, Suppression of cyclooxygenase-2 gene transcription by humulon of beer hop extract studied with reference to glucocorticoid. FEBS Letters, 2000. 465(2-3): p. 103-6. Subbaramaiah, K. and A.J. Dannenberg, Resveratrol inhibits the expression of cyclooxygenase-2 in mammary epithelial cells. Advances in Experimental Medicine and biology, 2001. 492: p. 147-57. Scherle, P.A., W. Ma, H. Lim, S.K. Dey and J .M. Trzaskos, Regulation of cyclooxygenase-2 induction in the mouse uterus during decidualization. An event of early pregnancy. Journal of Biological Chemistry, 2000. 275(47): p. 37086-92. 133 113. 114. 115. 116. 117. 118. 119. 120. 121. 122. Sirois, J. and J .S. Richards, Transcriptional regulation of the rat prostaglandin endoperoxide synthase 2 gene in granulosa cells. Evidence for the role of a cis- acting C/EBP beta promoter element. Journal of Biological Chemistry, 1993. 268(29): p. 21931-8. Liu, J ., M. Antaya, D. Boerboom, J .G. Lussier, D.W. Silversides and J. Sirois, The delayed activation of the prostaglandin G/H synthase-2 promoter in bovine granulosa cells is associated with down-regulation of truncated upstream stimulatory factor-2. Journal of Biological Chemistry, 1999. 274(49): p. 35037- 45. Nikaido, H., Outer membrane, in Escherichia coli and Salmonella typhimurium. Q Cellular and molecular biology, F.C. Neidhart, Editor. 1996, American Society for Microbiology Press: Washington, DC. p. 29—47. Raetz, C.R.H., Bacterial lipopolysaccharides: a remarkable family of bioactive macroamphiphiles, in Escherichia coli and Salmonella, in Escherichia coli and ~ Salmonella typhimurium. Cellular and molecular biology, F. Neidhardt, Editor. E 1996, American Society for Microbiology Press: Washington, DC. p. 1035-1063. " Riedo, F.X., R.S. Munford, W.B. Campbell, J .S. Reisch, K.R. Chien and RD. Gerard, Deacylated lipopolysaccharide inhibits plasminogen activator inhibitor- ], prostacyclin, and prostaglandin E2 induction by lipopolysaccharide but not by tumor necrosis. Journal of Immunology, 1990. 144(9): p. 3506-12. Yu, B. and SD. Wright, Catalytic properties of lipopolysaccharide (LPS) binding protein. Transfer of LPS to soluble CD14. Journal of Biological Chemistry, 1996. 271(8): p. 4100-5. Ulevitch, R]. and PS. Tobias, Receptor-dependent mechanisms of cell stimulation by bacterial endotoxin. Annual review of immunology, 1995. 13: p. 437-57. Yu, B., E. Hailman and SD. Wright, Lipopolysaccharide binding protein and soluble CD14 catalyze exchange of phospholipids. Journal of Clinical Investigation, 1997. 99(2): p. 315-24. Gegner, J .A., R]. Ulevitch and RS. Tobias, Lipopolysaccharide (LPS) signal transduction and clearance. Dual roles for LPS binding protein and membrane CDI4. Journal of Biological Chemistry, 1995. 270(10): p. 5320-5325. Abrahamson, S.L., H.M. Wu, R.E. Williams, K. Der, N. Ottah, R. Little, H. Gazzano-Santoro, G. Theofan, R. Bauer, S. Leigh, A. Orme, A.H. Horwitz, S.F. Carroll and R.L. Dedrick, Biochemical characterization of recombinant fusions of lipopolysaccharide binding protein and bactericidanenneability-increasing protein. Implications in biological activity. Journal of Biological Chemistry, 1997. 272(4): p. 2149-55. 134 123. 124. 125. 126. 127. 128. 129. 130. 131. 132. 133. Leturcq, D.J., A.M. Moriarty, G. Talbott, R.K. Winn, T.R. Martin and RJ. Ulevitch, Antibodies against CD14 protect primates from endotoxin-induced shock. Journal of Clinical Investigation, 1996. 98(7): p. 1533-8. Means, T.K., D.T. Golenbock and M.J. Fenton, The biology of Toll-like receptors. Cytokine & growth factor reviews, 2000. 11(3): p. 219-32. Yang, R.B., M.R. Mark, A. Gray, A. Huang, M.H. Xie, M. Zhang, A. Goddard, W.I. Wood, A.L. Gurney and P.J. Godowski, Toll-like receptor-2 mediates lipopolysaccharide-induced cellular signalling. Nature, 1998. 395(6699): p. 284- 8. Chow, J .C., D.W. Young, D.T. Golenbock, W.J. Christ and F. Gusovsky, Toll-like receptor-4 mediates lipopolysaccharide-induced signal transduction. Journal of Biological Chemistry, 1999. 274(16): p. 10689-92. Hirschfeld, M., Y. Ma, J .H. Weis, S.N. Vogel and J .J . Weis, Cutting edge: repurification of lipopolysaccharide eliminates signaling through both human and murine toll-like receptor 2. Journal of Immunology, 2000. 165(2): p. 618-22. Takeuchi, O., K. Hoshino, T. Kawai, H. Sanjo, H. Takada, T. Ogawa, K. Takeda and S. Akira, Dijferential roles of TLR2 and TLR4 in recognition of gram- negative and gram-positive bacterial cell wall components. Immunity, 1999. 11(4): p. 443-51. Lien, E., T]. Sellati, A. Yoshimura, T.H. Flo, G. Rawadi, R.W. Finberg, J .D. Carroll, T. Espevik, R.R. Ingalls, J .D. Radolf and D.T. Golenbock, Toll-like receptor 2 functions as a pattern recognition receptor for diverse bacterial products. Journal of Biological Chemistry, 1999. 274(47): p. 33419-25. J iang, Q., S. Akashi, K. Miyake and HR. Petty, Lipopolysaccharide induces physical proximity between C014 and toll-like receptor 4 (TLR4) prior to nuclear translocation of NF -kappa B. Journal of Immunology, 2000. 165(7): p. 3541-4. Dziarski, R., Q. Wang, K. Miyake, C.J. Kirschning and D. Gupta, MD-2 enables Toll-like receptor 2 (TLR2)-mediated responses to lipopolysaccharide and enhances TLR2-mediated responses to Gram-positive and Gram-negative bacteria and their. Journal of Immunology, 2001. 166(3): p. 1938-44. da Silva Correia, J ., K. Soldau, U. Christen, P.S. Tobias and RI. Ulevitch, Lipopolysaccharide is in close proximity to each of the proteins in its membrane receptor complex. transfer from CDI4 to TLR4 and MD-Z. The Journal of biological chemistry, 2001. 276(24): p. 21129-35. MacKichan, ML. and AL. DeFranco, Role of ceramide in lipopolysaccharide (LPS)-induced signaling. LPS increases ceramide rather than acting as a structural homolog. Journal of Biological Chemistry, 1999. 274(3): p. 1767-75. 135 134. 135. 136. 137. 138. 139. 140. 141. 142. 143. 144. Medvedev, A.E., J.C. Blanco, N. Qureshi and SN. Vogel, Limited role of ceramide in lipopolysaccharide-mediated mitogen-activated protein kinase activation, transcription factor induction, and cytokine release. Journal of Biological Chemistry, 1999. 274(14): p. 9342-50. Shankavaram, U.T., D.L. DeWitt and L.M. Wahl, Lipopolysaccharide induction of monocyte matrix metalloproteinases is regulated by the tyrosine phosphorylation of cytosolic phospholipase A2. Journal of Leukocyte Biology, 1998. 64(2): p. 221-7. Weinstein, S.L., M.R. Gold and AL. DeFranco, Bacterial lipopolysaccharide stimulates protein tyrosine phosphorylation in macrophages. Proceedings of the 1! National Academy of Sciences USA, 1991. 88(10): p. 4148-52. ' Downey, J .S. and J. Han, Cellular activation mechanisms in septic shock. 5, Frontiers in Bioscience, 1998. 30(3): p. d468-76. Z t Rao, K.M., MAP kinase activation in macrophages. Journal of Leukocyte i .1" Biology. 2001. 69(1): p. 3-10. Sweet, M.J. and DA. Hume, Endotoxin signal transduction in macrophages. Journal of Leukocyte Biology, 1996. 60(1): p. 8-26. Castrillo, A., M.J. Diaz-Guerra, S. Hortelano, P. Martin-Sanz and L. Bosca, Inhibition of IkappaB kinase and IkappaB phosphorylation by 15-deoxy- Delta(12,14)-prostaglandin 1(2) in activated murine macrophages. Molecular and Cellular Biology, 2000. 20(5): p. 1692-8. Kopp, E., R. Medzhitov, J. Carothers, C. Xiao, I. Douglas, C.A. Janeway and S. Ghosh, ECSIT is an evolutionarily conserved intermediate in the Toll/IL-I signal transduction pathway. Genes & Development, 1999. 13(16): p. 2059-71. Swantek, J .L., M.H. Cobb and TD. Geppert, Jun N-terminal kinase/stress- activated protein kinase (JNK/SAPK) is required for lipopolysaccharide stimulation of tumor necrosis factor alpha (TNF-alpha) translation: TNF-alpha translation by blocking JNK/SAPK. Molecular and Cellular Biology, 1997. 17(11): p. 6274-82. Hambleton, J ., S.L. Weinstein, L. Lem and AL. DeFranco, Activation of c-Jun N- terminal kinase in bacterial lipopolysaccharide-stimulated macrophages. Proceedings of the National Academy of Sciences USA, 1996. 93(7): p. 2774-8. Read, M.A., M.Z. Whitley, S. Gupta, J .W. Pierce, J. Best, R.J. Davis and T. Collins, Tumor necrosis factor alpha-induced E-selectin expression is activated by the nuclear factor-kappa B and c-J UN N-terminal kinase/p38 mitogen- activated protein kinase pathways. Journal of Biological Chemistry, 1997. 272(5): p. 2753-61. 136 145. 146. 147. 148. 149. 150. 151. 152. 153. 154. 155. Locksley, R.M., N. Killeen and M.J. Lenardo, The TNF and TNF receptor superfamilies: integrating mammalian biology. Cell, 2001. 104(4): p. 487-501. Raingeaud, J ., S. Gupta, J .S. Rogers, M. Dickens, J. Han, R.J. Ulevitch and R.J. Davis, Pro-inflammatory cytokines and environmental stress cause p38 mitogen- activated protein kinase activation by dual phosphorylation on tyrosine and threonine. Journal of Biological Chemistry, 1995. 270(13): p. 7420-6. Guan, Z., S.Y. Buckman, A.P. Pentland, DJ. Templeton and AR. Morrison, Induction of cyclooxygenase-2 by the activated MEKK I --> SEKI/MKK4 --> p38 mitogen-activated protein kinase pathway. Journal of Biological Chemistry, 1998. 273(21): p. 12901-12908. Straus, D.S., G. Pascual, M. Li, J .S. Welch, M. Ricote, C.H. Hsiang, L.L. Sengchanthalangsy, G. Ghosh and CK. Glass, lS-deoxy-delta 12,14- prostaglandin 12 inhibits multiple steps in the NF -kappa B signaling pathway. Proceedings of the National Academy of Sciences USA, 2000. 97(9): p. 4844-9. Henricson, B.E., J.M. Carboni, A.L. Burkhardt and SN. Vogel, LPS and Taxol activate Lyn kinase autophosphorylation in Lps(n), but not in Lpsd), macrophages. Molecular medicine (Cambridge, Mass ), 1995. 1(4): p. 428-35. Reimann, T., D. Biischer, R.A. Hipskind, S. Krautwald, M.L. Lohmann-Matthes and M. Baccarini, Lipopolysaccharide induces activation of the Raf-[MAP kinase pathway. A putative role for Raf-1 in the induction of the IL—1 beta and the TNF-alpha genes. Journal of Immunology, 1994. 153(12): p. 5740-9. Stefanova, 1., ML. Corcoran, E.M. Horak, L.M. Wahl, J .B. Bolen and ID. Horak, Lipopolysaccharide induces activation of CDI4-associated protein tyrosine kinase p53/56lyn. Journal of Biological Chemistry, 1993. 268(28): p. 20725-8. Frost, J .A., J .L. Swantek, S. Stippec, M.J. Yin, R. Gaynor and M.H. Cobb, Stimulation of NF kappa B activity by multiple signaling pathways requires PAK]. Journal of Biological Chemistry, 2000. 275(26): p. 19693-9. Giroux, M. and A. Descoteaux, Cyclooxygenase-2 expression in macrophages: modulation by protein kinase C-alpha. Journal of Immunology, 2000. 165(7): p. 3985-91. St-Denis, A., F. Chano, P. Tremblay, Y. St-Pierre and A. Descoteaux, Protein kinase C-alpha modulates lipopolysaccharide-induced functions in a murine macrophage cell line. Journal of Biological Chemistry, 1998. 273(49): p. 32787- 92. Mestre, J .R., P.J. Mackrell, D.E. Rivadeneira, P.P. Stapleton, T. Tanabe and J .M. Dal y, Redundancy in the signaling pathways and promoter elements regulating cyclooxygenase-2 gene expression in endotoxin-treated macrophage/monocytic cells. Journal of Biological Chemistry, 2001. 276(6): p. 3977-82. 137 'li 156. 157. 158. 159. 160. 161. 162. 163. 164. 165. 166. 167. Zoeller, R.A., P.D. Wightman, M.S. Anderson and CR. Raetz, Accumulation of lysophosphatidylinositol in RAW 264. 7 macrophage tumor cells stimulated by lipid A precursors. Journal of Biological Chemistry, 1987. 262(35): p. 17212-20. Bar-Sagi, D. and A. Hall, Ras and Rho GTPases: a family reunion. Cell, 2000. 103(2): p. 227-38. Rothwarf, D.M.a.K., Michael, The NF -kappa B Activation Pathway: A Paradigm in Information Transfer from Membrane to Nucleus. Science's stke, 1999. www.stke.orglcgi/contentlful1/OC_sigtrans;1999/5/re1. Ghosh, S., M.J. May and EB. Kopp, NF -kappa B and Rel proteins: evolutionarily conserved mediators of immune responses. Annual Review of Immunology, 1998. 16: p. 225-260. Karin, M. and Y. Ben-Neriah, Phosphorylation meets ubiquitination: the control of NF -[ kappa [B activity. Annual Review of Immunology, 2000. 18: p. 621-63. Hwan g, D., Modulation of the expression of cyclooxygenase-2 by fatt acids mediated through Toll-like receptor 4 derived signaling pathways. FASEB Journal, 2001. In Press. Irie, T., T. Muta and K. Takeshige, TAKI mediates an activation signal from toll- like receptor( s ) to nuclear factor-kappaB in lipopolysaccharide-stimulated macrophages. FEBS letters, 2000. 467(2-3): p. 160—4. Nishitoh, H, M. Saitoh, Y. Mochida, K. Takeda, H. Nakano, M. Rothe, K. Miyazono and H. Ichijo, ASK I is essential for JNK/SAPK activation by TRAF 2. Molecular cell, 1998. 2(3): p. 389-95. Ichijo, H., E. Nishida, K. Irie, P. ten Dijke, M. Saitoh, T. Moriguchi, M. Takagi, K. Matsumoto, K. Miyazono and Y. Gotoh, Induction of apoptosis by ASK], a mammalian MAPKKK that activates SAPK/JNK and p38 signaling pathways. Science, 1997. 275(5296): p. 90-4. Ray, A., M. Hannink and BK. Ray, Concerted participation of NF -kappa B and C/EBP heteromer in lipopolysaccharide induction of serum amyloid A gene expression in liver. Journal of Biological Chemistry, 1995. 270(13): p. 7365-74. Caivano, M. and P. Cohen, Role of mitogen-activated protein kinase cascades in mediating lipopolysaccharide-stimulated induction of cyclooxygenase-2 and IL-1 beta in RAW264 macrophages. Journal of Immunology, 2000. 164(6): p. 3018-25. Paul, A., A. Cuenda, C.E. Bryant, J. Murray, E.R. Chilvers, P. Cohen, G.W. Gould and R. Plevin, Involvement of mitogen-activated protein kinase homologues in the regulation of lipopolysaccharide-mediated induction of cyclo- oxygenase-Z but not nitric oxide synthase in RAW. Cellular signalling, 1999. 11(7): p. 491-7. 138 168. 169. 170. 171. 172. 173. 174. 175. 176. 177. 178. Xia, D., F. Wang and M.J. Parmely, Inhibition of nuclear factor-kappa B activation in mouse macrophages and the RAW 264. 7 cell line by a synthetic adenyl carbocyclic nucleoside. Biochemical pharmacology, 2000. 60(5): p. 717- 27. Vanden Berghe, W., S. Plaisance, E. Boone, K. De Bosscher, M.L. Schmitz, W. Fiers and G. Haegeman, p38 and extracellular signal-regulated kinase mitogen- activated protein kinase pathways are required for nuclear factor-kappa B p65 transactivation mediated by tumor necrosis. Journal of Biological Chemistry, 1998. 273(6): p. 3285-90. Schultz, R.M., N.A. Pavlidis, W.A. Stylos and MA. Chirigos, Regulation of macrophage tumoricidal function: a role for prostaglandins of the E series. Science, 1978. 202(4365): p. 320-1. Russell, S.W. and J .L. Pace, Both the kind and magnitude of stimulus are important in overcoming the negative regulation of macrophage activation by PGE2. Journal of Leukocyte Biology, 1984. 35(3): p. 291-301. Taffet, S.M. and S.W. Russell, Macrophage-mediated tumor cell killing: regulation of expression of cytolytic activity by prostaglandin E. Journal of Immunology, 1981. 126(2): p. 424-7. Chu, E., L.C. Casey, J .E. Harris and DP. Braun, Suppression of the development of tumoricidal function in gamma interferon-treated human peripheral blood monocytes by lipopolysaccharide: the role of cyclooxygenase. Journal of Clinical Immunology, 1993. 13(1): p. 49-57. Kunkel, S.L., M. Spengler, MA. May, R. Spengler, J. Larrick and D. Remick, Prostaglandin E2 regulates macrophage-derived tumor necrosis factor gene expression. Journal of Biological Chemistry, 1988. 263(11): p. 5380-4. Raddassi, K., J .F. Petit and G. Lemaire, LPS-induced activation of primed murine peritoneal macrophages is modulated by prostaglandins and cyclic nucleotides. Cellular immunology, 1993. 149(1): p. 50-64. Thompson, J .E., R.J. Phillips, H. Erdjument-Bromage, P. Tempst and S. Ghosh, I kappa B-beta regulates the persistent response in a biphasic activation of NF - kappa B. Cell, 1995. 80(4): p. 573-82. Wei], R., C. Laurent-Winter and A. Israel, Regulation ofIkappaBbeta degradation. Similarities to and differences from IkappaBalpha. Journal of Biological Chemistry, 1997. 272(15): p. 9942-9. Suyan g, H., R. Phillips, 1. Douglas and S. Ghosh, Role of unphosphorylated, newly synthesized I kappa B beta in persistent activation of NF -kappa B. Molecular and Cellular Biology, 1996. 16(10): p. 5444-9. 139 179. 180. 181. 182. 183. 184. 185. 186. 187. 188. 189. Bours, V., G. Franzoso, V. Azarenko, S. Park, T. Kanno, K. Brown and U. Siebenlist, The oncoprotein Bel-3 directly transactivates through kappa B motifs via association with DNA-binding p508 homodimers. Cell, 1993. 72(5): p. 729- 39. Franzoso, G., V. Bours, V. Azarenko, S. Park, M. Tomita-Yamaguchi, T. Kanno, K. Brown and U. Siebenlist, The oncoprotein Bel-3 can facilitate NF-kappa B- mediated transactivation by removing inhibiting p50 homodimers from select kappa B sites. The EMBO Journal, 1993. 12(10): p. 3893-901. Na, S.Y., S.K. Lee, S.J. Han, H.S. Choi, S.Y. Irn and J .W. Lee, Steroid receptor coactivator-I interacts with the p50 subunit and coactivates nuclear factor kappa B-mediated transactivations. Journal of Biological Chemistry, 1998. 273(18): p. 10831-4. Stein, 3., AS. Baldwin, Jr., D.W. Ballard, W.C. Greene, P. Angel and P. Herrlich, Cross-coupling of the NF -kappa B p65 and F os/Jun transcription factors produces potentiated biological function. The EMBO Journal, 1993. 12(10): p. 3879-91. Kurland, J .I. and R. Bockman, Prostaglandin E production by human blood monocytes and mouse peritoneal macrophages. The Journal of Experimental Medicine, 1978. 147(3): p. 952-7. Bachwich, P.R., S.W. Chensue, J .W. Larrick and S.L. Kunkel, Tumor necrosis factor stimulates interleukin-1 and prostaglandin E2 production in resting macrophages. Biochemical and Biophysical Research Communications, 1986. 136(1): p. 94-101. Hamilton, J .A., Stimulation of macrophage prostaglandin and neutral protease production by phorbol esters as a model for the induction of vascular changes associated with tumor promotion. Cancer research, 1980. 40(7): p. 2273-80. Stenson, W.F., M.W. Nickells and J .P. Atkinson, Metabolism of exogenous arachidonic acid by murine macrophage-like tumor cell lines. Prostaglandins, 1981. 21(5): p. 675-89. Raschke, W.C., S. Baird, P. Ralph and I. Nakoinz, Functional macrophage cell lines transformed by Abelson leukemia virus. Cell, 1978. 15(1): p. 261-7. Tanaka, Y., F. Amano, H. Kishi, M. Nishijima and Y. Akamatsu, Degradation of arachidonyl phospholipids catalyzed by two phospholipases A2 and phospholipase C in a lipopolysaccharide-treated macrophage cell line RAW264. 7. Archives of Biochemistry and Biophysics, 1989. 272(1): p. 210-8. O'Sullivan, G.M., F.H. Chilton, E.M. Huggins Jr. and CE. McCall, Lipopolysaccharide priming of alveolar macrophages for enhanced synthesis of 140 190. 191. 192. 193. 194. 195. 196. 197. 198. 199. prostanoids involves induction of a novel prostaglandin H synthase. Journal of Biological Chemistry, 1992. 267 : p. 14547-14550. Lee, J .Y., K.H. Sohn, S.H. Rhee and D. Hwang, Saturated fatty acids, but not unsaturated fatty acids, induce the expression of cyclooxygenase-2 mediated through Toll-like receptor 4. Journal of Biological Chemistry, 2001. 276(20): p. 16683-9. Hoff, T., D. DeWitt, V. Kaever, K. Resch and M. Goppelt-Struebe, Differentiation-associated expression of prostaglandin G/H synthase in monocytic cells. FEBS Letters, 1993. 320(1): p. 38-42. O'Sullivan, M.G., E.M. Huggins, Jr., E.A. Meade, D.L. DeWitt and CE. McCall, Lipopolysaccharide induces prostaglandin H synthase-2 in alveolar macrophages. Biochemical and Biophysical Research Communications, 1992. 187(2): p. 1123-7. Reddy, ST. and HR. Herschman, Ligand-induced prostaglandin synthesis requires expression of the TISIO/PGS-Z prostaglandin synthase gene in murine fibroblasts and macrophages. journal of Biological Chemistry, 1994. 269(22): p. 15473-80. Arias-Negrete, S., K. Keller and K. Chadee, Proinflammatory cytokines regulate cyclooxygenase-2 mRNA expression in human macrophages. Biochemical and Biophysical Research Communications, 1995. 208(2): p. 582-589. Riese, J ., T. Hoff, A. Nordhoff, D.L. DeWitt, K. Resch and V. Kaever, Transient expression of prostaglandin endoperoxide synthase-2 during mouse macrophage activation. Journal of Leukocyte Biology, 1994. 55(4): p. 476-82. Lee, SH, E. Soyoola, P. Chanmugam, S. Hart, W. Sun, H. Zhong, S. Liou, D. Simmons and D. Hwang, Selective expression of mitogen-inducible cyclooxygenase in macrophages stimulated with lipopolysaccharide. Journal of Biological Chemistry, 1992. 267: p. 25934-25938. Inoue, H., T. Nanayama, S. Hara, C. Yokoyama and T. Tanabe, The cyclic AMP response element plays an essential role in the expression of the human prostaglandin-endoperoxide synthase 2 gene in differentiated U93 7 monocytic cells. FEBS Letters, 1994. 350(1): p. 51-54. Inoue, H. and T. Tanabe, Transcriptional role of the nuclear factor kappa B site in the induction by lipopolysaccharide and suppression by dexamethasone of cyclooxygenase-2 in U 93 7 cells. Biochemical and Biophysical Research Communications, 1998. 244(1): p. 1438 Barry, O.P., M.G. Kazanietz, D. Pratico and GA. FitzGerald, Arachidonic acid in platelet microparticles up-regulates cyclooxygenase-Z-dependent prostaglandin 141 200. 201. 202. 203. 204. 205. 206. 207. 208. formation via a protein kinase C/mitogen-activated protein kinase-dependent pathway. Journal of Biological Chemistry, 1999. 274(11): p. 7545-7556. Wadleigh, D.J., S.T. Reddy, E. Kopp, S. Ghosh and HR. Herschman, Transcriptional activation of the cyclooxygenase-2 gene in endotoxin-treated RAW264. 7 macrophages. Journal of Biological Chemistry, 2000. 275(9): p. 6259-66. Newton, R., L.M. Kuitert, M. Bergmann, I.M. Adcock and P.J. Barnes, Evidence for involvement of NF -kappa B in the transcriptional control of C 0X-2 gene expression by IL-1 beta. Biochemical and Biophysical Research Communications, 1997. 237(1): p. 28-32. Roshak, A.K., J .R. Jackson, K. McGough, M. Chabot-Fletcher, E. Mochan and LA. Marshall, Manipulation of distinct NF kappaB proteins alters interleukin- Ibeta-induced human rheumatoid synovial fibroblast prostaglandin E2 formation. Journal of Biological Chemistry, 1996. 271(49): p. 31496-501. Diaz-Cazorla, M., D. Pérez-Sala and S. Lamas, Dual efl‘ect of nitric oxide donors on cyclooxygenase-2 expression in human mesangial cells. Journal of the American Society of Nephrology : JASN, 1999. 10(5): p. 943-52. D'Acquisto, F., T. Iuvone, L. Rombola, L. Sautebin, M. Di Rosa and R. Camuccio, Involvement of NF -kappa B in the regulation of cyclooxygenase-2 protein expression in LPS-stimulated J 774 macrophages. FEBS Letters, 1997. 418(1-2): p. 175-8. Abate, A., S. Oberle and H. Schroder, Lipopolysaccharide-induced expression of cyclooxygenase-2 in mouse macrophages is inhibited by chloromethylketones and a direct inhibitor of NF -kappa B translocation. Prostaglandins & other lipid mediators, 1998. 56(5-6): p. 277-90. Mahboubi, K., W. Young and NR. Ferreri, Tyrosine phosphatase- dependent/tyrosine kinase-independent induction of nuclear factor-kappa B by tumor necrosis factor-alpha: effects on prostaglandin endoperoxide synthase-2. The Journal of pharmacology and experimental therapeutics, 1998. 285(2): p. 862-8. Hwang, D., B.C. J ang, G. Yu and M. Boudreau, Expression of mitogen-inducible cyclooxygenase induced by lipopolysaccharide: mediation through both mitogen- activated protein kinase and NF -kappa B signaling pathways in macrophages. Biochem. Pharmacol, 1997. 54(1): p. 87-96. Rhee, SH. and D. Hwang, Murine TOLL-like receptor 4 confers lipopolysaccharide responsiveness as determined by activation of NF kappa B and expression of the inducible cyclooxygenase. Journal of Biological Chemistry, 2000. 275(44): p. 34035-40. 142 209. 210. 211. 212. 213. 214. 215. 216. 217. 218. Jobin, C., O. Morteau, D.S. Han and R. Balfour-Sartor, Specific NF -kappa B blockade selectively inhibits tumour necrosis factor-alpha-induced C 0X-2 but not constitutive C OX-I gene expression in HT—29 cells. Immunology., 1998. 95(4): p. 537-543. Chen, C.C., K.T. Chiu, Y.T. Sun and W.C. Chen, Role of the cyclic AMP-protein kinase A pathway in lipopolysaccharide-induced nitric oxide synthase expression in RAW 264. 7 macrophages. Involvement of cyclooxygenase-2. Journal of Biological Chemistry, 1999. 274(44): p. 31559-64. Aderem, A.A., D.S. Cohen, SD. Wright and Z.A. Cohn, Bacterial lipopolysaccharides prime macrophages for enhanced release of arachidonic acid metabolites. The Journal of Experimental Medicine, 1986. 164(1): p. 165-79. Barrios-Rodiles, M., G. Tiraloche and K. Chadee, Lipopolysaccharide modulates cyclooxygenase-2 transcriptionally and posttranscriptionally in human macrophages independently from endogenous IL-1 beta and TNF-alpha. Journal of Immunology, 1999. 163(2): p. 963-9. Baer, M., A. Dillner, R.C. Schwartz, C. Sedon, S. Nedospasov and RF. Johnson, Tumor necrosis factor alpha transcription in macrophages is attenuated by an autocrine factor that preferentially induces NF -kappa B p50. Molecular and Cellular Biology, 1998. 18(10): p. 5678-89. Gorgoni, B., M. Caivano, C. Arizmendi and V. Poli, The transcription factor C/EBPbeta is essential for inducible expression of the cox-2 gene in macrophages but not in fibroblasts. Journal of Biological Chemistry, 2001. 276(44): p. 40769- 77. Mestre, J .R., D.E. Rivadeneira, P.J. Mackrell, M. Duff, P.P. Stapleton, V. Mack- Strong, S. Maddali, G.P. Smyth, T. Tanabe and J .M. Daly, Overlapping CRE and E-box promoter elements can independently regulate C 0X-2 gene transcription in macrophages. FEBS Letters, 2001. 496(2-3): p. 147-51. Ohmori, Y., J. Tebo, S. Nedospasov and TA. Hamilton, Kappa B binding activity in a murine macrophage-like cell line. Sequence-specific differences in kappa B binding and transcriptional activation functions. Journal of Biological Chemistry, 1994. 269(26): p. 17684-90. Suh, N ., T. Honda, H]. Finlay, A. Barchowsky, C. Williams, N.E. Benoit, Q.W. Xie, C. Nathan, G.W. Gribble and MB. Spom, Novel triterpenoids suppress inducible nitric oxide synthase (iNOS) and inducible cyclooxygenase (COX-2) in mouse macrophages. Cancer research, 1998. 58(4): p. 717-23. von Knethen, A., D. Callsen and B. Briine, Superoxide attenuates macrophage apoptosis by NF -kappa B and AP-I activation that promotes cyclooxygenase-2 expression. Journal of Immunology, 1999. 163(5): p. 2858-66. 143 “07" PT! 219. 220. 221. 222. 223. 224. 225. 226. 227. 228. 229. Delgado, M. and D. Ganea, Vasoactive intestinal peptide and pituitary adenylate cyclase activating polypeptide inhibit the MEKKI/MEK4/JNK signaling pathway in LPS-stimulated macrophages. Journal of neuroimmunology, 2000. 110(1-2): p. 97-105. Kuprash, D.V., I.A. Udalova, R.L. Turetskaya, D. Kwiatkowski, N.R. Rice and S.A. Nedospasov, Similarities and differences between human and marine TNF promoters in their response to lipopolysaccharide. Journal of Immunology, 1999. 162(7): p. 4045-52. Liu, H., P. Sidiropoulos, G. Song, L.J. Pagliari, M.J. Birrer, B. Stein, J. Anrather and RM. Pope, TNF-alpha gene expression in macrophages: regulation by NF - kappa B is independent of c-Jan or C/EBP beta. Journal of Immunology, 2000. 164(8): p. 4277-85. Brook, M., G. Sully, A.R. Clark and J. Saklatvala, Regulation of tumour necrosis factor alpha mRNA stability by the mitogen-activated protein kinase p38 signalling cascade. FEBS Letters, 2000. 483(1): p. 57-61. Raabe, T., M. Bukrinsky and RA. Currie, Relative contribution of transcription and translation to the induction of tumor necrosis factor-alpha by lipopolysaccharide. Journal of Biological Chemistry, 1998. 273(2): p. 974-80. Sheppard, K.A., D.W. Rose, Z.K. Haque, R. Kurokawa, E. McInemey, S. Westin, D. Thanos, M.G. Rosenfeld, C.K. Glass and T. Collins, Transcriptional activation by NF -kappa B requires multiple coactivators. Molecular and Cellular Biology, 1999. 19(9): p. 6367-78. Gerritsen, M.E., A.J. Williams, A.S. Neish, S. Moore, Y. Shi and T. Collins, CREB-binding protein/p300 are transcriptional coactivators of p65. Proceedings of the National Academy of Sciences USA, 1997. 94(7): p. 2927-32. Dean, J .L., M. Brook, A.R. Clark and J. Saklatvala, p38 mitogen-activated protein kinase regulates cyclooxygenase-2 mRNA stability and transcription in lipopolysaccharide-treated human monocytes. Journal of Biological Chemistry, 1999. 274(1): p. 264-269. Barnes, P.J., Anti-inflammatory actions of glucocorticoids: molecular mechanisms. Colch, 1979. 94(6): p. 557-72. Tsai, E.Y., J .V. Falvo, A.V. Tsytsykova, A.K. Barczak, A.M. Reimold, L.H. Glimcher, M.J. Fenton, D.C. Gordon, I.F. Dunn and AB. Goldfeld, A lipopolysaccharide-specific enhancer complex involving Ets, Elk-I, Sp], and CREB binding protein and p300 is recruited to the tumor necrosis factor alpha promoter in vivo. Molecular and Cellular Biology, 2000. 20(16): p. 6084-94. Li, Q., A. Irnhof, T.N. Collingwood, F.D. Umov and AP. Wolffe, p300 stimulates transcription instigated by ligand-bound thyroid hormone receptor at a 144 step subsequent to chromatin disruption. The EMBO Journal, 1999. 18(20): p. 5634-52. 230. Struhl, K., Gene regulation. A paradigm for precision. Science, 2001. 293(5532): p. 1054-5. 145