Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Posttranscriptional regulation of ICAM-1 gene expression Ohh, Michael 1995

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

Item Metadata

Download

Media
831-ubc_1995-060349.pdf [ 2.79MB ]
Metadata
JSON: 831-1.0088361.json
JSON-LD: 831-1.0088361-ld.json
RDF/XML (Pretty): 831-1.0088361-rdf.xml
RDF/JSON: 831-1.0088361-rdf.json
Turtle: 831-1.0088361-turtle.txt
N-Triples: 831-1.0088361-rdf-ntriples.txt
Original Record: 831-1.0088361-source.json
Full Text
831-1.0088361-fulltext.txt
Citation
831-1.0088361.ris

Full Text

POSTTRANSCRIPTIONAL REGULATION OF ICAM-1 GENE EXPRESSIONbyMICHAEL OHHB.Sc.Honours, Carleton University, 1990A THESIS SUBMITTED IN PARTIAL FULFILLMENT OFTHE REQUIREMENTS FOR THE DEGREE OFDOCTOR OF PHILOSOPHYinTHE FACULTY OF GRADUATE STUDIES(Department of Medical Genetics, Genetics Programme)We accept this thesis as conformingto the required standardTHE UNIVERSITY OF BRITISH COLUMBIASeptember 1995© Michael Ohh, 1995In presenting this thesis in partial fulfilment of the requirements for an advanceddegree at the University of British Columbia, I agree that the Library shall make itfreely available for reference and study. I further agree that permission for extensivecopying of this thesis for scholarly purposes may be granted by the head of mydepartment or by his or her representatives. It is understood that copying orpublication of this thesis for financial gain shall not be allowed without my writtenpermission.(Signature)__________________________Department of MPfLcbi (pI. (6ee-kcs Pvjirwe’)The University of British ColumbiaVancouver, CanadaDate Oc44p. (1/95DE-6 (2/88)ABSTRACTCell-cell adhesion is critical for the generation of effective immune responses and is dependentupon the expression of a variety of cell surface receptors. Intercellular adhesion molecule-i(ICAM- 1; CD54) is an inducible cell surface glycoprotein expressed at a low level on a widerange of cell types. Although its expression is dramatically increased at sites of inflammation,providing important means of regulating cell-cell interactions and thereby inflammatoryresponses, the intracellular regulatory elements and signaling pathways underlying theinducible expression of ICAM- 1 by proinflammatory cytokines were previously largelyunknown. In this thesis, a novel posttranscriptional regulation of ICAM-l gene expression bytwo proinflammatory mediators, interferon-y (IFN-’y) and phorbol myristate acetate (PMA),and the possible role of serine/threonine (ser/thr) phosphorylation pathway in cycloheximideinduced ICAM-l message stabilization were investigated. The results show that (1)constitutively expressed ICAM-i mRNA has a short half-life; (2) IFN-’y and PMA induce theaccumulation of ICAIVI-i message, at least in part, by stabilizing the mRNA; (3) the WN-yresponsive element(s) is located within the protein coding region encoding the cytoplasmicdomain; (4) the PMA-responsive elements lie within the 3’ untranslated region (UTR) and mayeven be the AUUUA multimers; (5) cycloheximide, a potent eukaryotic protein synthesisinhibitor that superinduces the expression of many genes by preventing the degradation ofotherwise labile mRNAs, also induces the level of ICAM- 1 mRNA by message stabilization;(6) the stabilizing effect of cycloheximide does not depend on the 3’UTR containing theAUUUA sequences; (7) the stabilization of ICAM-l mRNA by cycloheximide is independentof its translational inhibition; and (8) the ser/thr phosphorylation pathway seems to play acrucial role in the cycloheximide-induced stabilization of ICAM- 1 message. These resultsdemonstrate the existence of distinct destabilizing elements throughout the ICAM- 1 messagethat are responsive to the actions of various proinflammatory cytokines, and underscore theimportance of posttranscriptional regulation of ICAM-1 expression during an inflammatoryresponse.11TABLE OF CONTENTSAbstract iiList of tables VList of figures ViList of abbreviations viiiAcknowledgement xiDedication xiiChapter I: Introduction 11.1 ICAM-l 31.1.1 Regulation of ICAM-l expression 101.1.2 Transcriptional regulation of ICAM- 1 gene expression 111.2 ICAM-2 131.3 ICAM-3 141.4 Mechanisms of eukaryotic mRNA degradation 151.4.1 Determinants of mRNA decay 161.4.2 The 5’ cap structure 171.4.3 Translational requirement of mRNA degradation 171.4.3.1 5’UTR 171.4.3.2 Nonsense-mediated mRNA decay 191.4.4 Open reading frame destabilizing sequences 201.4.5 3’ UTR destabilizing elements 211.4.6 Endo- and exoribonucleases 241.5 Thesis objectives 261.6 References 26Chapter II: Posttranscriptional Regulation of ICAM- 1 Gene Expression byInterferon-y and Phorbol Myristate Acetate2.1 Abstract 382.2 Introduction 392.3 Materials and Methods 412.4 Results 462.5 Discussion 612.6 References 64Chapter III: Identification of Interferon-y- and Phorbol Myristate Acetate-ResponsiveElements Involved in ICAM- 1 mRNA Stabilization3.1 Abstract 673.2 Introduction 681113.3 Materials and Methods 693.4 Results 723.5 Discussion 813.6 References 82Chapter IV: Regulation of ICAM- 1 mRNA Stability by Cycloheximide: Role ofSerine/Threonine Phosphorylation and Protein Synthesis4.1 Abstract 854.2 Introduction 864.3 Materials and Methods 874.4 Results 894.5 Discussion 1054.6 References 112Chapter V: Summary and General Discussion 1145.1 ICAM-1 regulation by 1FN-y and PMA 1165.1.1 Transcriptional regulation does not tell the whole story 1165.1.2 Novel posttranscriptional regulation of ICAM- 1 gene 1165.2 Role of ser/thr phosphorylation in ICAM-1 mRNA stabilization 1225.3 Concluding remarks 1245.4 Future directions 1255.5 References 126ivLIST OF TABLESTable 1. Selected members of the three major adhesion molecule families. 2VLIST OF FIGURESFig. 1 Schematic diagram of expression vectors. 43Fig. 2 Effects of IFN-y and PMA on ICAM-1 mRNA expression in P388D1 cells. 47Fig. 3 Effects of WN-y on induction of ICAM- 1, ICAM- 1 A3, and CD 18 mRNAlevels. 51Fig. 4 Southern blot analysis of the exogenous ICAM- 1 copy number. 52Fig. 5 Half-life analysis of ICAM-1 and ICAM-1A3 mRNAs. 53Fig. 6 Effects of IFN-y on ICAM-2 and ICAM-2/1 mRNA accumulation andstabilization. 55Fig. 7 Effects of PMA on induction of ICAM-1 and ICAM-13 mRNA levels. 58Fig. 8 Effects of PMA and IFN-’y on induction of ICFAS chimeric mRNA level. 59Fig. 9 Half-life analysis of ICAM-1 and ICFAS mRNAs. 60Fig. 10 Schematic diagram of the expression vectors. 71Fig. 11 Induction of various forms of ICAM- 1 mRNA by IFN-’y and PMA. 74Fig. 12 Posttranscriptional effects of PMA on ICAM-2 and ICAM-211 chimericmRNAs. 76Fig. 13 Posttranscriptional effects of IFN-y on ICAM-2 and ICAM-2/1 chimericmRNAs. 78Fig. 14 Effects of cycloheximide on ICAM-1 mRNA expression in A20, P388D1,T28, and SVEC4-10 cells. 91Fig. 15 Effects of cycloheximide on ICAM-1 mRNA expression in L-ic-1 andL-ic-1A3 cells and on ICAM-2 mRNA expression in L-sic-2 cells. 92Fig. 16 Effects of PMA over-stimulation and staurosporine on ICAM-1 mRNAaccumulation by cycloheximide. 95Fig. 17 Effects of phosphorylation inhibitors on ICAM—1 mRNA accumulation bycycloheximide. 97Fig. 18 ICAM-1 and ICAM-2 mRNA half-life analysis. 99viFig. 19 Flow cytometric analysis of ICAM- 1 cell surface expression under variousinhibitor treatments. 103Fig. 20 Flow cytometric analysis of the effects of cycloheximide on ICAM- 1 cellsurface expression on A20, P388D1, and T28 cells. 104Fig. 21 Effect of cycloheximide and puromycin on de novo ICAM-1 protein synthesis. 106Fig. 22 Summary of the responses of various ICAM- 1 mRNA deletion mutants andICAM-2/1 chimeric mRNAs to 1FN-y and PMA. 119Fig. 23 Schematic diagram of ICAM-1 gene showing the location of the putativeIFN-y- and PMA-responsive elements responsible for the stabilization ofICAM-1 message. 121Fig. 24 Schematic diagram of the serine/threonine phosphorylation pathway involvedin the stabilization of ICAM-1 mRNA by cyclohexmiide. 123viiABBREVIATIONSaa amino acidActD actinomycin DAPC antigen-presenting cellARE AU-rich elementbp base pairBSA bovine serum albumincAMP cyclic AMP; adenosine 3?,5t-monophosphateCD cluster of differentiationcDNA complementary DNACHX cycloheximideCMV cytomegaloviruscpm counts per minuteDMEM Dulbecco’s modified minimum essential mediaDNA deoxyribonucleic acidEBV Epstein-Barr virusECM extracellular matrixEDTA ethylenediamine tetraacetic acidEtBr ethidium bromideFACS flurorescence activated cell sorterFCS fetal calf serumFITC fluorescein isothiocyanatefMLP f-Met-Leu-Pheg gramG1yCAM- 1 glycosylation-dependent cell adhesion molecule-iGM-CSF granulocyte/macrophage colony-stimulating factorviiiHBSS flanks’ balanced salt solutionh hourHUVEC human umbilical vein endothelial cellICA’I intercellular adhesion moleculeIFN interferonIg immunoglobulinIL interleukinIRE iron-responsive elementIRE-BP IRE-binding proteinkb kilobasekD kilodaltonLAD leukocyte adhesion deficiencyLFA lymphocyte function-associated antigenLPS lipopolysaccharidemAb monoclonal antibodyMac-i macrophage antigen-iMAdCAM- 1 mucosal addressin cell adhesion molecule-iMAG myelin-associated glycoproteinMHC major histocompatibility complexmm minuteMLR mixed lymphocyte reactionmRNA messenger RNANCAM neural cell adhesion moleculeNK natural killerNP-40 nonidet P-40nt nucleotideORE open reading frameixPAB poly(A)-binding proteinPAGE polyacrylamide gel electrophoresisPAN poly(A) nucleasePBL peripheral blood lymphocytesPBS phosphate buffered salinePCR polymerase chain reactionPHA phytohaemagglutininPKA protein kinase APKC protein kinase CPMA phorbol 12-myristate 13-acetatePSGL-1 P-selectin glycoprotein ligand-lRBC red blood cell; erytrocytesRGD Arg-Gly-AspRNA ribonucleic acidrRNA ribosomal RNASDS sodium dodecyl sulphatesICAM soluble ICAMSSC saline sodium citrate bufferSSPE saline sodium phosphate EDTA bufferTCR T-cell receptorTE Tris EDTA bufferTNE Tris sodium chloride EDTA bufferTNF tumor necrosis factorTRE TPA-responsive elementtRNA transfer RNAUTR untranslated regionVCAM vascular cell adhesion moleculexACKNOWLEDGEMENTI wish to thank foremost Dr. Fumio Takei for his invaluable supervision and supportthroughout this thesis project. I thank my advisory committee members, Drs. KeithHumphries, Connie Eaves, and Wilf Jefferies, for their critical comments, discussions, andguidance. I would also like to thank all the members of Dr. Takeis lab and my friends andcolleagues, especially the Pysz, Carp, Soccerman-Hector, Cam aka Mr. Baseball, the Gibbs,and Vivienne, at the Terry Fox Laboratory and the University of British Columbia, for theirmoral encouragement and life outside the lab. Finally, I thank my parents and family for theirunconditional support throughout this endeavour.This work has been supported by the Medical Research Council of Canada and I amgrateful to the MRC for providing a studentship throughout this project.xiFor my mother andfather.xiiIINTRODUCTIONAdhesive interactions of cells with each other and with the extracellular matrix (ECM) arefeatures of many processes in multicellular organisms and also appear to play important rolesin the development and functional responses of immune cells. During development of theembryo, cell adhesion proteins impart position-specific information that guide cell migration,localization, and the transfer of information between cells. As the cells are triggered todifferentiate to form specific components of the various tissues and organs, adhesion proteinshelp to maintain the organization and integrity of the body. The immune system is comprisedof a variety of circulating cells which utilize highly specialized cellular recognitionmechanisms. The function of the immune system is to distinguish self from the nonseif and toeliminate the latter. This involves the continuous circulation of leukocytes through the bloodand lymphatic systems in surveillance of foreign materials. When such antigens (or epitopes)are detected, the circulating leukocytes migrate to the site and mount an immune orinflammatory reaction. Three major families of cell adhesion molecules (Table 1), the selectinfamily, the integrin family, and the immunoglobulin (Ig) superfamily, have been found to playimportant roles in lymphocyte recirculation, migration of leukocytes into inflamed tissues, andcell-cell interactions involved in the generation of immune responses. This thesis focusses onone of these adhesion molecules, the intercellular adhesion molecule-i (ICAM- 1).1Table 1. Selected members of the three major cell adhesion families1.Ig Superfamily Distribution LigandsICAM-1 (CD54) lymphocytes, neutrophils, LFA-1, Mac-l, CD43,thymocytes, monocytes, fibrinogendendritic cells, endothelium,fibroblasts, keratinocytes,chondrocytes, epitheliumICAM-2 (CD 102) endothelium, subpopulation LFA-lof lymphocytes, monocytes,dendritic cellsICAM-3 (CD5O) lymphocytes, monocytes, LFA-lgranulocytesLeukocyte IntegrinsLFA-1 (CD1 la/CD 18) lymphocytes, monocytes, ICAM-l, ICAM-2, ICAM-3neutrophilsMac-i (CD1 lb/CD18) monocytes, neutrophils ICAM-1, iC3b, fibrinogen,factor Xp150,95 (CD1 lc/CD18) monocytes, neutrophils iC3b, fibrinogenSelectinsL (leukocyte)-selectin lymphocytes, neutrophils, sialyl Lewis(a, G1yCAIvI- 12,(CD62L, Mel-i4, LAM-l) thymocytes, monocytes, CD34, MAdCAM-leosinophiis, basophils, NKcellsP (platelet)-selectin (CD62P, platelets (c granules), sialyl Lewisxa, PSGL- 1PADGEM, GMP-140) endothelium (Weibel-Paladebodies), megakaryocytesE (endothelial)-selectin endothelium sialyl Lewisx,a, E-selectin(CD62E, ELAM-i)OtherCD44 (H-CAM, GP-90 lymphocytes, monocytes, hyaluronate, collagen,Hermes) neutrophils, epithelium, glial chondroitin sulfated CD44cells, fibroblasts, myocytes binds fibronectinfrom the reviews by Springer, 1994 and Carlos and Harlan, 1994.abbreviations used: G1yCAM- 1, glycosylation-dependent cell adhesion molecule-i; MAdCAM- 1,mucosal addressin cell adhesion molecule-i; PSGL- 1, P-selectin glycoprotein ligand- 1.21.1 ICAM-1ICAM- 1 (CD54) is an inducible cell surface ligand for the lymphocyte function-associatedantigen-i (LFA-1; CD1 1aJCD18) (Rothlein et at., 1986; Marlin and Springer, 1987; Makgobaet at., 1989; Staunton et at., 1990) and the macrophage antigen-i (Mac-i; CD11b/CD18)(Diamond et at., 1991), as well as CD43 (Rosenstein et at.,1991) and fibrinogen (Languino etat., 1993). This membrane protein is a single chain glycoprotein with a peptide backbone of55 kD (Staunton et at., 1988; Simmons et at., 1988). Human ICAM-l is encoded by a single-copy gene located on chromosome 19 (Katz et at., 1985). The gene for murine ICAM-1 hasbeen mapped to the proximal short arm of chromosome 9 (Ballantyne et al., 1991). cDNAcloning in the human (Staunton et at., 1988; Simmons et at., 1988) and mouse (Horley et at.,1989; Siu et at., 1989) has shown that ICAM-1 is a member of the Ig superfamily. Membersof the Ig superfamily contain one or more Ig domains, each composed of 90-100 amino acids(aa) arranged in two sheets of anti-parallel f3-strands, which is usually stabilized by adisulphide bond at its center (Williams and Barclay, 1988; Williams, 1987). ICAM-1 consistsof five Ig-like extracellular domains, and displays highest homology with two other celladhesion molecules, neural cell adhesion molecule (NCA]V1) and myelin associatedglycoprotein (MAG). Murine ICAM-1 has limited similarity with human ICAM-1 at either thenucleotide (65%) or protein levels (50%) (Horley et at., 1989; Siu et at., 1989). Amino acidsubstitutions in the extracellular domains have indicated that the primary binding site for LFA1 is located in the NH2-terminal first domain of ICAM- 1 (Staunton et at., 1990). Initialelectron microscopy of soluble ICAM-1 suggested a hinge between the second and thirddomains of the extracellular region (Staunton et at., 1990); however, a recent report has shownthat the hinge in ICAM-l occurs between third and fourth domains (Kirchhausen et at., 1993).The location of this hinge may be germane for leukocyte adhesion to endothelium because asecond ligand-binding site for a leukocyte integrin, Mac-i, has been localized to the third Iglike domain (Diamond et at., 1991).3Leukocyte adhesion to the second ligand-binding site of ICAM-1 is affected by thedegree of glycosylation (Diamond et al., 1991). There are eight possible sites for N-linkedglycosylation in the five extracellular domains of human ICAM- 1 (Staunton et al., 1988;Simmons et al., 1988). In contrast, murine ICAM-1 has nine potential residues for N-linkedglycosylation (Honey et al., 1989). The molecular weights of ICAM-1 have been found tovary between 76 and 114 kD, suggesting that there is variable posttranslational modification ofICAM-1 (Rothlein et al., 1986; Dustin et al., 1986; Pober et al., 1986). Affinity of leukocytesfor the binding site within the third Ig-like region of ICAM- 1 was found to increase as thedegree of glycosylation decreased (Diamond et al., 1991). This observation suggested thatselectivity of leukocyte adhesion via Mac-i may be dictated in part by the posttranslationalglycosylation of ICAM- 1 at the tissue level.The cytoplasmic domain of ICAM-1 consists of a 28-aa residue, highly chargedsequence rich in lysine and arginine residues (Staunton et al., 1988; Simmons et al., 1988).The cytoplasmic domain, or an 8-aa portion of the cytoplasmic region of ICAM-1(RQRKIKKR), has been shown to bind to the cytoskeleton of COS cells transfected with thecDNA of human ICAM-l and to the cytoskeleton of Epstein-Barr virus (EBV)-transformed Bcells (Carpen et al., 1992). The binding of ICAM-i to the cytoskeleton was found to occurthrough linkage with a-actinin, a cytoskeleton protein that may serve to anchor actin filamentsto the cell membrane (Mimura and Asano, 1987). When expressed as aglycophosphotidylinositol-linked membrane protein, ICAM-1 was diffusely expressed on thecell surface (Carpen et al., 1992). Thus, linkage with the cytoskeleton may localize ICAM- 1within regions of the endothelial cell membrane to facilitate leukocyte adherence andtransmigration.Study of LFA-1IICAM-1 cross-species interactions has shown that human LFA-1interacts with murine ICAM- 1 but murine LFA- 1 does not associate with human ICAM- 1.This restriction in cross-species binding is due to the LFA-1 alpha subunit, since mouse-human4hybrids expressing human alpha chain and murine beta chain will bind human ICAM-1(Larson and Springer, 1990).Adhesion of ICAM- 1 to these leukocyte integrins plays an essential role in a variety ofimmune reactions including T cell-mediated killing, natural lytic events, T-helper and B-lymphocyte responses, homotypic aggregation, antibody-dependent cytotoxicity mediated bymonocytes and granulocytes, and leukocyte trafficking processes such as adherence ofleukocytes to vacular endothelium and epidermal cells (Springer, 1990; Dustin and Springer,1991; Boyd et aL, 1988; Carlos and Harlan, 1994).In contrast to LFA- 1, which is expressed only on leukocytes in a constitutive manner,ICAM- 1 is an inducible cell surface glycoprotein expressed at a low level on a wide variety ofcells, including leukocytes, vascular endothelium, fibroblasts, follicular dendritic cells,synovial cells, keratinocytes, and certain epithelial cells (Dustin et al., 1986, 1988; Pober et al.,1986; te Velde et aL, 1987; Mentzer et al., 1988; Rothlein et al., 1988; Dustin and Springer,1988, 1991; Carlos and Harlan, 1994). However, ICAM-l expression is dramaticallyincreased at sites of inflammation (Springer, 1990; Dustin et at., 1986; Dustin and Springer1991; Rothlein et at., 1988; Pober et at., 1986; Carlos and Harlan, 1994), providing animportant means of regulating cell-cell interactions and thereby presumably inflammatoryresponses. The upregulated expression of ICAM-1 on venule endothelium is thought tofacilitate the adhesion and subsequent transendothelial migration of leukocytes bearing LFA- 1or Mac- 1 into inflammatory tissues, as well as permitting appropriate interactions oflymphocytes with cells expressing targeted antigens (Springer, 1990; Wawryk et a!., 1989;Carlos and Harlan, 1994; Dustin and Springer 1991). Various proinflammatory mediatorsincluding interleukin- 1 (IL- 1), tumor necrosis factor-cc (TNF-cc), lipopolysaccharide (LPS),and interferon-y (IFN-y), as well as active phorbol esters, have been found to increase ICAM- 1expression on many cell types (Dustin et at., 1986; Rothlein et at., 1988; Carlos and Harlan,1994) and are thought to be responsible for the induction of ICAM-l expression atinflammatory sites in vivo. The function, and indeed the importance, of leukocyte adhesion5were elucidated in numerous animal experimental systems where mAbs to ICAM- 1 wereshown to inhibit various events associated with inflammation (Carlos and Harlan, 1994).Although the importance of leukocyte integrins was clearly illustrated in patients with LAD Isyndrome (discussed below), deficiencies in ICAM-1 have not been reported in humans.However, ICAM- 1 deficient mice have been generated recently and showed prominentabnormalities of inflammatory responses including impaired neutrophil emigration in responseto chemical peritonitis and decreased contact hypersensitivity to 2,4-dinitrofluorobenzene(Sligh et al., 1993). Furthermore, the mutant cells provided negligible stimulation in themixed lymphocyte reaction (MLR), although they proliferated normally as responder cells(Sligh et al., 1993). Similarly, Xu et al. (1994) observed that the homozygous ICAM-1-deficient mice have elevated numbers of circulating neutrophils and lymphocytes, as well asdiminished allogeneic T cell response and delayed type hypersensitivity. These mice were alsofound to be resistant to septic shock either by reducing T cell activation or by diminishingneutrophil infiltration (Xu et al., 1994).In addition to its physiologic role of contributing to cell-cell adhesion in immuneresponses, the ability of ICAM- 1 to serve as a receptor for the major serotype of rhinoviruses(Greve et al., 1989; Staunton et al., 1989a) has been studied. Rhinoviruses are the causativeagents in 50% of common colds. Although rhinoviruses have evolved more than 100 noncrossreactive antigenic variants in an attempt to evade the immune response, 90% of them bindto ICAIVI-1 (Springer, 1990; Greve et al., 1989; Staunton et al., 1989a). More recently, solubleICAM-1 has been shown to be a potent and specific inhibitor of major rhinovirus infection(Marlin et al., 1990). Mutagenesis has indicated that there is overlap in the sequences withinthe first domain of ICAM-l that interact with LFA-1 and rhinovirus. Additionally, someamino acid changes within the first domain of ICAM- 1 can eliminate its ability to bindrhinovirus without affecting its ability to interact with LFA- 1 (Larson and Springer, 1990).ICAM- 1 is also one of two receptors that can be subverted as a sequestration antigenfor Plasmodiumfalciparum-infected erythrocytes (RBCs) (Berendt et al., 1989). The primary6event in the pathogenesis of malaria is the adherence of infected RBCs to endothelium in theliver. Increased levels of cytokines have been detected in acute malaria. Thus, induction ofICAM- 1 expression on endothelium resulting from cytokines may allow P. faciparum-infectedRBCs to bind endothelial cells. Furthermore, an immunoadhesin molecule containing ICAM-1fused to the hinge, CH2 and CH3 domains of the human IgGi heavy chain has been generatedand shown to inhibit the adhesion of P. faciparum-infected RBCs to ICAM-l-bearing surfaces,although LFA-i interactions with ICAM-i were not inhibited (Staunton et at., 1992). Theimmunoadhesin also promoted phagocytosis and destruction of parasitized RBCs by humanmonocytes (Staunton et at., 1992).Recently, it has been demonstrated that a proportion of lymphomas including themajority of fresh Burkitt’s lymphomas fail to express ICAM- 1 receptor, LFA- 1, on the cellsurface (Clayberger et al., 1987). These Burkitt’s lymphoma cells are poor stimulators of bothautologous and allogeneic T cell responses, suggesting that the lack of LFA-l on these tumorcells may contribute to their failure to initiate efficient immune responses leading to theirescape from immunosurveillance. Some Burkitt’s lymphoma cells have been shown to bedeficient in ICAM-1 and LFA-3 as well (Billaud et at., 1987). ICAM-l expression is alsothought to correlate positively with the metastatic potential of malignant melanoma (Johnson etat., 1989). Taken together, these observations imply that perturbation of one or a combinationof adhesion molecules might be a common method by which cells acquire an ability to escapefrom immunosurveillance.The importance of ICAM-1 in promoting and maintaining an effective immuneresponse can also be appreciated by examining its leukocyte integrin receptors. The necessityfor leukocytes to adhere to cellular and matrix proteins in order to function has becomeincreasingly clear through two converging lines of evidence. First was the identification andcharacterization of molecules on the leukocyte surface that are responsible, in part, for multipleleukocyte adhesion events. These receptor molecules, called the CD 18 family of adhesionmolecules or leukocyte integrins, consist of LFA-1 (CD1 1aICD18), Mac-i (CD1 lb/CD18),7and p150,95 (CD1 lc/CD18) (for reviews, see Martz, 1982; Springer et at., 1987; Springer,1990). Each of these molecules is a heterodimer with distinct alpha subunits that associatenoncovalently with a common beta subunit. These proteins are distributed in variouscombinations on all leukocytes. Structurally, the leukocyte adhesion molecules are membersof the integrin family and specifically belong to the 2 subgroup. Although the leukocyteadhesion molecules mediate adhesion processes, they are unlike most other integrins in thatthey appear to recognize ligands that, for most part, do not contain Arg-Gly-Asp (RGD)sequences and their activity is dependent on Mg rather than Ca. The function of thesemolecules and indeed the importance of leukocyte adhesion in the generation and maintenanceof inflammation have been elucidated in many experimental systems where the use ofmonoclonal antibodies (mAb) as antagonists has proven inhibitory to multiple eventsassociated with inflammation.Leukocyte integrins mediate a wide range of adhesion-dependent functions both inantigen-dependent and antigen-independent processes. In vitro, mAbs to CD18 or CD1 lainhibit lymphocyte mediated lytic events such as cytotoxic T cell activity and natural killer cellactivity (Krensky et at., 1983; Sanchez-Madrid et at., 1983; Kohl et at., 1984) and lymphocytetrafficking processes such as attachment of lymphocytes to vascular endothelium andepidermal cells (Haskard et at., 1986; Dustin et at., 1988). Furthermore, these mAbs inhibitantibody formation and mitogen- and antigen-induced T cell proliferation as well as MLR(Davignon et at., 1981; Dougherty et at., 1988; Dougherty and Hogg, 1987; Boyd et at., 1988).All of these processes require cell-cell interactions for successful responses. Moreover, mAbsto CD18 or CD1 1 also inhibit certain granulocyte functions, such as their attachment toendothelium (Smith et at., 1988), homotypic aggregation (Anderson et at., 1986), binding toiC3b coated particles (Anderson et at., 1986; Beller et at., 1982), and antibody-dependentcellular cytotoxicity (Kohl, 1987).A second line of evidence suggesting an important role of the leukocyte integrins ininflammatory reactions came from studies of individuals with genetic mutations in the gene8encoding the common 2 subunit CD 18. As a result, their cells are unable to express normallevels of leukocyte adhesion molecules on the cell surface (Anderson and Springer, 1987;Todd and Freyer, 1988; Dustin and Springer 1991; Carlos and Harlan, 1994). Patients with thiscongenital leukocyte adhesion deficiency type I syndrome (LAD I) have recurring bacterialinfections which are fatal in childhood unless corrected by bone marrow transplantation.Neutrophils from these patients fail to migrate in response to chemoattractants and are unableto bind to and cross the endothelium at sites of inflanimation. Pus, therefore, fails to form andthis seriously compromises the ability of these children to fight infections. The most severelydeficient patients (expressing <1% of the normal number of leukocyte integrins) also havedepressed T cell responses. This is presumably a result of defective antigen-presentingcapacity due to the inability of their T cells to attach effectively to the antigen-presenting cells(APCs). In vitro, lymphocytes from these individuals behave just like normal lymphocytes inthe presence of anti-LFA-1 mAbs. Their cells also show defective proliferative and effectorresponses in assays requiring cell-cell interactions. Such observations indicate an even broaderrole of the leukocyte integrins in antigen-independent inflammatory and immune processes toalso include leukocyte migration and diapedesis.Animal models for LAD I, allowing for experimental manipulation in vivo, will be ofsignificant use in studying the in vivo function of the leukocyte integrins. A canine LAD Imodel appearing completely analogous to the human has been identified (Giger et at., 1987).As a second approach, mAbs against the leukocyte integrins have been administered intoexperimental animals to duplicate a LAD I state. Such studies have shown that peritonealmacrophages are not elicited with thioglycollate in mice that have been injected with anti-Mac1 mAb (Rosen and Gordon, 1987). Similarly, leukocytes from rabbits injected with anti-CD 18mAb do not bind to endothelium and their extravasation is prevented (Arfors et at., 1987).Both of these studies suggest that leukocyte integrins play an important role in leukocyteextravasation. Furthermore, CD 18-mutant mice having 2 or 16% of normal CD18 expressionon granulocytes in the resting or activated state, respectively, have been generated (Wilson et9at., 1993). These mutant mice show an impaired inflammatory response to a chemicalperitonitis and delayed rejection of cardiac transplants (Wilson et al., 1993).1.1.1 Regulation of ICAM-1 ExpressionThe LFA-1/ICAM-1 mechanism is regulated through changes in counter-receptor expression inaddition to changes in LFA-l avidity. Regulation of ICAM-1 expression allows control of thespectrum of cells to which activated leukocytes can adhere and in some cases coordinates theability to adhere with the presence of major histocompatibility complex (MHC) moleculesrequired for foreign antigen presentation (Dustin and Springer, 1991). In most cases, ICAM- 1expression is regulated by cytokine receptors coupled to mechanisms for altering geneexpression.Resting lymphocytes lack significant ICAM- 1 expression, but ICAM- 1 expressionincreases during T and B lymphocyte activation (Dustin et at., 1986; Clark et at., 1986;Wawryk et al., 1989). Thus, activated lymphocytes in germinal centers and at sites ofinflammation are strongly positive (Dustin et at., 1986; Wantzin et at., 1989). In contrast,monocytes also possess an intracellular store of ICAM- 1 which can be mobilized to the cellsurface following adherence to fibronectin (Dougherty et at., 1988). Similarly, leukocyteintegrins, Mac-l and p150,95, are stored in intracellular poois in monocytes andpolymorphonuclear cells (Todd et at., 1984; Miller et at., 1987; Bainton et at., 1987) and thispooi can be rapidly mobilized to the cell surface in response to a variety of chemoattractantsincluding f-Met-Leu-Phe (fMLP), C5a, and leukotriene B4 (Springer et at., 1984; Lanier et at.,1985; Berger et at., 1984; Todd et at., 1984; Miller et at., 1987; Bainton et at., 1987). Theintracellular pooi of these f2 integrins has been shown to localize to peroxidase—negativegranules in monocytes and granulocytes (Miller et at., 1987; Bainton et at., 1987). However,neither the nature of the compartment in which ICAM- 1 is sequestered in monocytes nor the10relationship of this compartment to the peroxidase-negative granules containing rapidlymobilizable Mac-i and p150,95 is known.The rapidity of appearance of ICAM- 1 that can be induced on nonleukocytes is evenmore dramatic. ICAM- 1 is absent from most cells in normal, nonlymphoid tissues, except forexpression of low levels on endothelial cells. Local immune responses result in a rapidincrease in ICAM- 1 expression on endothelial cells and induction of ICAM- 1 on epithelial andmesenchymal cells (Dustin et a!., 1986; Dustin and Springer, 1988; Wantzin et a!., 1989;Munro et al., 1989). In fact, ICAM- 1 expression can be increased on multiple cell typesincluding keratinocytes in inflamed skin lesions (Wantzin et a!., 1989; Griffiths and Nickoloff,1989), transplanted liver bile duct cells and perivenular hepatocytes during rejection (Adams eta!., 1989), endothelium in the brain surrounding MS plaques (Sobel et al., 1990), transplantedkidney glomeruli and tubules during rejection (Cosimi et a!., 1990), and lung epithelial cellsfollowing antigen provacation (Wegner et a!., 1990). These in viva results are correlated withthe ability of products of activated lymphocytes (lymphokines) and monocytes (monokines) toincrease ICAN’I- 1 expression on cultured fibroblasts, endothelial cells, epithelial cells, andastrocytes (Dustin et a!., 1986, 1988; Dustin and Springer, 1988; Pober et a!., 1986; Frohmanet a!., 1989). The increased expression of ICAM-i seen on malignant melanomas andcarcinomas may be secondary to local immune reactions generating cytokines (Temponi et a!.,1989; Vogetseder eta!., 1989; Holzmann et al., 1988; Natali eta!., 1990).1.1.2 Transcriptional Regulation of ICAM-1 Gene ExpressionThe proinflammatory cytokines, IFN-y, TNF-o, and IL-i, as well as LPS and the proteinkinase C (PKC) activator PMA, can markedly increase the expression of ICAM-i on variouscells of both hematopoietic and non-hematopoietic origins (Dustin et a!., 1986; Rothlein et a!.,1988). Because ]PN-y enhances the transcription of many 1FN-y-responsive genes including11the class II MHC (Moses et at., 1992), in combination with the fact that ICAM- 1 was initiallynoted to have a tissue distribution similar to that of the MHC class II antigens (Dustin et at.,1986), i.e., vascular endothelium, germinal center cells, interdigitating reticulum cells,macrophages, lymphoid tissues, and epithelial cells in tonsil and thymus, its effect on ICAM- 1was thought to be similar. Indeed, within the 5-flanking region of ICAM-1 gene are containedpotential IFN-y-responsive elements, as well as glucocorticoid receptor binding sites, NF-icBconsensus sequences, and AP1/TPA-responsive element (TRE)-, AP-2- and AP-3-like sites(Degitz et at., 1991; Voraberger et at., 1991). However, in a reporter assay these potentialelements could only mount a two-fold induction with IFN-y stimulation (Voraberger et at.,1991), suggesting that ICAM-1 gene regulation by ]FN-y may also involve posttranscriptionalmechanisms. Similarly, PMA was found to have little effect on the transcription of the ICAM1 gene in human umbilical vein endothelial cells (HUVECs) as measured by a nuclear run-onassay (Wertheimer et at., 1992). In contrast, Voraberger et at. (1991) reported that PMAstimulated the expression of a luciferase reporter gene linked to the 5-flanking region of theICAM-1 gene containing three copies of TRE in transiently transfected A549 cells.Interestingly, similar results were obtained in HUVECs transfected with the same pBHluc 1.3construct of Voraberger et at. (1991) suggesting that, although PMA can activate exogenousICAM-l enhancer/promoter elements in HUVECs, the same elements in the endogenous generemain functionally silent (Wertheimer et at., 1992).Other proinflammatory cytokines such as TNF-cL and IL- 1, however, are thought toinduce ICAM-1 message primarily at a transcriptional level (Voraberger et at., 1991;Wertheimer et at., 1992). Activation of the ICAM-l gene by TNF-ot as measured by nuclearrun-on studies, is both rapid and transient, with peak transcription rates occurring at 30 mmand returning to basal level within 2 h, thus providing the cell with a burst of ICAIVI-1messenger RNA (mRNA) synthesis (Wertheimer et at., 1992). Moreover, TNF-o is known toactivate the ubiquitously expressed transcription factor NF-icB, which plays a role in theinducible expression of many genes (Lenardo and Baltimore, 1989). As mentioned above,12several potential NF-KB binding sequences have been identified in the 5-regulatory region ofthe ICAM- 1 gene, and TNF-o has been shown to promote nuclear NF-icB-like binding activityin HIJVECs (Montgomery et al., 1991). However, Wertheimer et al. (1992) found that NF-icBbinding activity is also induced in PMA-stimulated HUVECs without a concomitant increasein ICAM- 1 gene transcription, indicating that NF-icB activation alone is insufficient to inducethe transcription of ICAM-1 gene. Similarly, TNF-o-stimulated activation of NF-icB isinsufficient to activate ELAM-1 gene transcription in HUVECs (Montgomery et al., 1991).Thus, even the mechanisms for the regulatory effects of TNF-o on ICAM-1 expression do notappear to be fully understood and require further characterization of TNF-oc-responsive cisacting elements of the ICAM- 1 gene.1.2 ICAM-2LFA-1-dependent antigen-specific responses can be completely inhibited, unaffected, orpartially inhibited by ICAM-1 mAbs depending on the target cell used. This phenomenonpredicted the existence of at least one alternative counter-receptor for LFA- 1. Such a secondligand, ICAM-2 (CD 102), was formally identified by screening for cDNA clones which, whenexpressed, would promote cell adherence to purified LFA-1 in the presence of mAb to ICAM1 (Staunton et al., 1989b). ICAM-2 is another member of the Ig superfamily that is expressedon endothelium and is involved in leukocyte adherence. The gene encoding human ICAM-2 isa single-copy gene located on chromosome 17 (Hogg et al., 1991). Molecular cloning of theICAM-2 gene showed a core protein of 29 kD with six residues for possible N-linkedglycosylation which, if fully used, would yield a mature protein of 46 kD (Staunton et al.,1989b). ICAM-2 is an integral membrane protein with two extracellular Ig-lilce domains,closely related to ICAM-1 (34% homology with the two NH2-terminal domains of ICAM-l)(Staunton et al., 1989b). The ligand-binding site for LFA-1 is located in these domains of13ICAM- 1; hence, ICAM-2 is a second endothelial ligand for this leukocyte integrin (Staunton etal., 1989b). However, the observation that Mac-i binds to the third Ig-like domain of ICAM-isuggests that ICAM-2 does not serve as a ligand for this leukocyte integrin. Unlike ICAM- 1,the expression of ICAIvI-2 is not affected by cytokines (Staunton et al., i989b; de Fougerolleset al., 1991). ICAM-2 also does not bind to the major or minor group of rhinoviruses (Larsonand Springer, 1990).Murine ICAM-2 has been molecularly cloned and shares 60% identity to humanICAM-2 at the protein level (Xu et al., 1992). However, the transmembrane and cytoplasmicregions of murine ICAM-2 are more highly conserved (75% protein identity). In contrast tohuman ICAM-2, the murine counterpart has five potential sites for N-linked glycosylation.1.3 ICAM-3A third LFA- 1 counter-receptor, ICAM-3 (CD5O), has also been recently identified (Fawcett etal., 1992; Vazeux et al., 1992; de Fougerolles et al., 1993). ICAM-3 contains 15 potential N-linked glycosylation sites and is a heavily glycosylated protein of 116 to 140 kD in a cell type-specific fashion, with an expected core polypeptide of about 56 kD. Molecular cloning ofICAM-3 revealed that it contains five Ig-like domains that are highly similar to thecorresponding domains in ICAM-1 and ICAM-2. ICAM-3 shares 48% overall aa identity withICAM-1 rising to 77% in domain 2, while the first two NH2-terminal Ig-like domains ofICAM-3 have 31% aa identity to the two Ig-like domains of ICAM-2.Despite the homologies between the three ICAMs, their patterns of expression suggestspecialized roles. The finding that adhesion of resting T lymphocytes to purified LFA-1 occursprimarily via ICAM-3, combined with the fact that ICAM-3 is expressed constitutively athigher levels on monocytes and resting lymphocytes than other LFA-1 ligands, suggest thatICAIVI-3 might be important in the initiation of immune responses (Fawcett et al., 1992;14Vazeux et at., 1992; de Fougerolles et at., 1993, 1994). ICAM-3 mAbs effectively inhibitICAM-3 adhesion to purified LFA-1 and in combination with ICAM-1 and ICAM-2 mAbsinhibit the homotypic aggregation of PMA-induced lymphoid cell lines (de Fougerolles et at.,1994). Moreover, purified ICAM-3 was found to support LFA-1-dependent adhesion, and alsoto be capable of providing a costimulatory signal to resting T lymphocytes (de Fougerolles etat., 1994). ICAM-3 is important in peripheral blood lymphocyte (PBL) proliferation inresponse to phytohemagglutinin (PHA), and in allogeneic and antigen-specific proliferation (deFougerolles et al., 1994). Thus, ICAM-3 could potentially be important in lymphocyteactivation, as well as in adhesion.1.4 MECHANISMS OF EUKARYOTIC mRNA DEGRADATIONChanges in degradation rates among mRNAs contribute significantly to the control of geneexpression in eukaryotes. In particular, the existence of highly unstable mRNAs allows forrapid and precise reduction or elevations in message levels following changes in transcriptsynthesis rates or alterations in the transcript turnover rates. The range of mRNA stability ineukaryotic cells can vary over several orders of magnitude (Peltz et at., 1991). Some mRNAsin higher cells are degraded with a half-life of about 20 mm, while other mRNAs are degradedwith half-lives over 24 h. Similar variations have been found in yeast, with the most labilemessages decaying in less than 5 mm and the most stable degrading at rates slower than 60 mm(Herrick et at., 1990).Several examples of regulated mRNA stability in eukaryotic cells have been reported,and these indicate that the trans-acting factors of the message decay system can be subject toregulation. For instance, maturation of T lymphocytes is due in part to the stimulated decay ofspecific niRNAs, and this probably results from a cascade events involving PKC (Takahamaand Singer, 1992). The regulation of transferrin receptor mRNA stability in response to iron15levels is an example of how a destabilizing sequence element can be inactivated by a pathwayresponding to cellular conditions (Klausner et at., 1993). The cell cycle regulation of histonemRNA levels is an example of how the coordinate synthesis of macromolecules can becontrolled by the regulation of mRNA decay (Harris et at., 1991). Furthermore, thestabilization of heat shock protein mRNA at high temperatures demonstrates the efficiency ofstabilizing mRNA only when it is needed (Petersen and Lindquist, 1988). Moreover, thesignificance of message degradation in differential gene expression is exemplified by thetransformation phenotype associated with mRNA stabilization mutations in the c-myc and cfos mRNAs (Schiavi et at., 1992) and the alterations in mRNA-specific destabilizing factors inmonocytic tumors (Schuler and Cole, 1988).1.4.1 Determinants of mRNA DecayRecent works on mRNA turnover illustrate that there are sequence elements regulating thedecay rate throughout the message. The presence of such a diverse array of elements rules outthe simple model that mRNA is passively protected from degradation by nucleases simply bythe presence of stabilizing sequences at its ends. This raises the questions of whether there is acommon pathway for mRNA degradation or whether the various elements act independently,and also whether they represent sites of nuclease sensitivity or whether they representrecognition sites for other proteins that lead eventually to the recruitment of a commonribonuclease.161.4.2 The 5’ Cap StructureThe unique 5-5’ phosphodiester bond of the cap makes it intrinsically resistant to generalribonucleases, and it is this resistance that allowed its original identification. A specificenzyme that removes the cap (decapping enzyme) has been purified from yeast (Stevens,1988), and it is possible that an activation of this enzyme by mRNA sequences can create anunstable mRNA. Decay of the mRNA would result owing to the existence of a 5’-3’exoribonuclease or an endoribonuclease whose site of action was masked by the cap structureand its associated cap-binding proteins (Rhoads, 1993).1.4.3 Translational Requirement of mRNA Degradation:1.4.3.1 5’ Untranslated RegionAlthough the 5’ untranslated region (UTR) of mRNA has not been definitively shown to be adestabilizing sequence on any mRNA, it has been well documented to control thetranslatability of a message, and it is through the negative regulation of translation initiationthat this region of mRNA has been shown to inhibit the degradation process.Initial understanding of the relationship between translation and mRNA turnover seemsto indicate that most mRNAs need to be translated to be degraded. In yeast, almost all mRNAsare stabilized by treatment with the translation inhibitor, cycloheximide (Herrick et al., 1990).Concordantly, a mutation resulting in partial loss of function of a transfer RNA (tRNA)nucleotidyl transferase protein, which leads to a decrease in the rates of translation due tolimiting functional tRNA, also results in message stabilization (Peltz et al., 1992). However,these experiments do not distinguish between a requirement for translation in mRNA decayand a requirement for a labile degradative protein activity that fails to be synthesized butcontinue to be degraded upon inhibition of translation.17Placement of a stable 5’UTR stem-loop structure in various yeast mRNAs can inhibittranslation up to 96% without changing the transcript levels and hence, by inference, thestability of the mRNA (Cigan et al., 1988). These data are inconsistent with the translationinhibitor studies and suggest that wild-type rates of translation initiation are not needed forinducing mRNA decay, although it is still unclear what would happen if mRNA becomesuntranslatable.In the work by Koeller et al. (1991), the translation of c-fos mRNA containing thedestabilizing AU-rich region was found to be negatively regulated by at least 95% by aninhibitory sequence, called the iron-responsive element (IRE), located within its 5’UTR.Regulation of translation was achieved by altering the iron levels in the tissue culture media,while the general metabolism and the overall translational capacity of the cell remainedunchanged. The results from this study suggested that translation is not required for mRNAdegradation.In contrast, Aharon and Schneider (1993) fused a mutated form of the adenovirustripartite leader that inhibits translation of a synthetic mRNA construct that was destabilizedowing to the presence in its 3’UTR of the AU-rich region of the granulocyte/macrophagecolony-stimulating factor (GM-CSF) mRNA. The results from their study indicate thatinhibition of translation results in message stabilization. Furthermore, relief of translationinhibition and the subsequent appearance of unstable mRNA is achieved by the introduction ofa ribosomal entry site downstream of the 5’ inhibitory sequence. In support, Savant-Bhonsaleand Cleveland (1992) have found that mRNA containing an AU-rich element (ARE) must betranslated to be rapidly degraded.Reasons to why there are differences in the requirement of translation for mRNA decayare unclear. It has been speculated recently that the requirement for translation in decay isactually a requirement for recruiting translation factors and that these factors, themselves,associate with the degradative enzymes. In support of this model are findings that a proteinneeded for translation initiation is also a ribonuclease in yeast (Sachs and Deardorff, 1992).181.4.3.2 Nonsense-mediated mRNA DecayPremature translational termination caused by nonsense mutations has been shown toaccelerate mRNA turnover rates (Peltz et aL, 1991). Trans-acting factors involved in suchnonsense-mediated mRNA decay have been identified in studies of a novel class of yeast nontRNA nonsense suppressors (Peltz and Jacobson, 1993; Leeds et aL, 1991, 1992). Mutationsin two genes that were isolated as allosuppressors, UPF1 and UPF3, lead to the selectivestabilization of mRNAs containing early nonsense mutations without affecting the decay ratesof most other mRNAs (Peltz and Jacobson, 1993; Leeds et al., 1991, 1992). The UPF1 genehas been cloned and sequenced (Leeds et al., 1992; Altamura et al., 1992) and severalstructural and functional properties have been identified. It is non-essential for viability;encoding a 109 kD protein with both zinc finger, nucleotide (GTP)-binding site and RNAhelicase motifs. It is identical to NAM7, a nuclear gene that was isolated as a high copysuppressor of mitochondrial RNA-splicing mutations (Altamura et al., 1992). It is alsopartially homologous to the yeast SEN1 gene (Leeds et al., 1992; Altamura et al., 1992). Thelatter encodes a non-catalytic subunit of the tRNA-splicing endonuclease complex (Leeds etal., 1992), suggesting that Upflp (the UPF1 gene product) may be part of a nuclease complextargeted specifically to nonsense-containing mRNAs.An explanation for why nonsense mutations exhibit a loss of destabilizing activity asthey are moved further away from the initiator methionine has recently been shown. Byutilizing either wild-type or mutant UPF1 strains, Peltz et al. (1993) were able to show that thisdistance dependence was created by several distinct RNA sequence elements. First, asexpected, the presence of a nonsense mutation was essential for UPF- 1-mediated decay.Second, the presence of a downstream sequence was also required, and point mutagenesis ofthis sequence suggested methionine codons were an essential part of it. Finally, the UPF1pathway was inactivated by the presence of an as yet ill-defined protection sequence precedingthe nonsense mutation.19These data are consistent with the model that the transiting ribosome can be madeinsensitive to nonsense mutations and that sensitivity only occurs when the ribosome has theability to reinitiate downstream of the nonsense mutation. Nonsense mutations wouldtherefore lose their destabilizing activity as they move further away from the initiator codonbecause of the increase in probability of being 3’ to a protection sequence and not 5’ to anotherinitiator codon.1.4.4 Open Reading Frame Destabilizing SequencesSeveral different experiments have shown that the open reading frame (ORF) region of mRNAcontains destabilizing sequences. The N-terminal tetrapeptide encoded by the f3-tubulinmRNA provides a target signal for rapid degradation of the message under conditions oftubulin monomer excess (Yen et al., 1988). Experiments utilizing the protooncogene c-fosmRNA also showed the existence of destabilizing elements within the ORF of its mRNA(Shyu et al., 1989). The 3’UTR of this mRNA also contains a destabilizing sequence, but Shyuet al. (1989) found that removal of this 3’UTR element was not sufficient to stabilize c-fosmRNA. Contemporaneously, another destabilizing sequence within the ORF was identifiedthat was utilized only when the message was transiently expressed following growth factorstimulation. A closer examination of this region has shown that it contains two destabilizingelements, and that translation through this region is required for inducing degradation.(Wellington et al., 1993).Furthermore, a series of proteins that recognize the ORF destabilizing region of c-foshave been identified by ultraviolet cross-linking experiments (Chen et al., 1992). In relatedexperiments examining an analogous destabilizing sequence in ORF of c-myc mRNA,Bernstein et al. (1992) found the in vitro decay rates of the polyribosome-bound c-myc mRNAto be dramatically increased in the presence of excess synthetic RNA containing the20destabilizing region. This activation presumably results from the competition for a 75 kDprotein that recognizes the destabilizing sequence and inactivates it. These c-fos and c-mycstudies provide some evidence for the importance of protein complex formation in theregulation of ORF destabilizing determinants.1.4.5 3’UTR Destabilizing ElementsThe two most well characterized examples of destabilizing sequences within the 3’UTR are theIRE on the transferrin mRNA and the ARE found on many unstable mRNAs. Located withinthe 3 ‘UTR of the transferrin mRNA is a region that contains five distinct stem-loop structurescapable of binding the IRE-binding protein (IRE-BP) (Klausner et al., 1993). The affinity ofthis protein for these sequences is regulated by cellular iron, and changes in affinity occurthrough dissociation and reassociation of an iron-sulfur cluster within the IRE-BP. Thebinding of the IRE-BP to transferrin mRNA stabilizes this mRNA. However, the IRE structureitself does not destabilize the mRNA, but is instead made unstable owing to the presence of anas yet uncharacterized destabilizing sequence in the vicinity of the stem-loop structures. It is,therefore, thought that the binding of IRE-BP to the IREs prevents association of destabilizingfactors to a cryptic destabilizing sequence. Whether these destabilizing factors are themselvesribonucleases, or whether they recruit ribonucleases to degrade the mRNA at a distant site,remains unclear.The regulation of transferrin mRNA decay provides two important mechanisticinsights. First, it shows how the decay process can be regulated by proteins that can directlyrespond to changes in the cellular environment. Second, the transferrin mRNA studies revealthat mRNA stabilization may, in fact, result from an inhibition of an activity recognizing thedestabilizing element and that this destabilizing element can be distant from the binding site ofthe stabilizing proteins.21An example of the inhibition of mRNA destabilizing sequences by a distant sequenceelement has also been described (Heaton et at., 1992). A 3’UTR destabilizing sequence fromthe yeast STE3 mRNA was found to stimulate decay of the normally stable yeast PGK1mRNA only when a portion of the PGK1 ORF had been removed. The inability of adestabilizing sequence to work in certain contexts raises the possibility that the interactionbetween different regions of mRNA can lead to changes in the efficiency of elementutilization. Such regulation in cis by DNA sequence elements surrounding transcriptionalpromoters is well documented. These data thus provide evidence that a number of mRNAmetabolism reactions may be subject to similar hierarchical controls.The second well characterized 3’UTR destabilizing sequence is the ARE.Characteristic classes of ARE-containing cellular mRNAs include the lymphokine and theimmediate early genes, including several protooncogenes (Bohjanen et al., 1991, 1992;Brewer, 1991; Caput et at., 1986; Gillis and Malter, 1991; Shaw and Kamen, 1986;Vakalopoulou et al., 1991; Wilson and Triesman, 1988; Shyu et al., 1991; You et at., 1992;Myer et at., 1992; Schuler and Cole, 1988). A number of groups have identified and begun tocharacterize factors ranging from 20 to 70 kD that bind to these AREs (Bohj anen et al., 1991,1992; Brewer, 1991; Gillis and Malter, 1991; Vakalopoulou et at., 1991; You et at., 1992;Myer et at., 1992; Schuler and Cole, 1988). Properties of some of these ARE-binding factorssuggest a role in mRNA turnover. These include their selective binding to a subset ofdifferentially regulated ARE-containing mRNAs (Bohjanen et at., 1991; Schuler and Cole,1988), changes in binding activity that correlate with changes in specific mRNA abundance(Bohj anen et at., 1991), and specific degradative activity in vitro (Brewer, 1991). Althoughthe AREs of many transcripts share a common AUUUA motif (Shaw and Kamen, 1986), thissequence is not sufficient to confer specificity of binding for all of these factors. For example,Bohjanen et at. (1991) have identified a cytoplasmic factor synthesized by stimulated T cellsthat binds specifically to AUUUA multimers present in the 3’UTR of GM-CSF, IL-2, andTNF-ct mRNAs, but does not bind to the ARE in the c-myc mRNA. Consistent with the22hypothesis that sequences other than the AUUUA motif can provide specificity , factors havebeen identified that bind to a 20 nucleotide U-rich region of the c-fos 3UTR that lacks anyAUUUA sequences (You et aL, 1992). Furthermore, Brewer (1991) has identified an activitythat binds to the c-myc ARE, or to poly(U), and specifically accelerates the turnover of the cmyc message in an in vitro mRNA decay system. The 32 kD ARE-binding polypeptide fromnuclear extracts identified by Vakalopoulou et al. (1991) also requires the AUUUA motifwithin an U-rich element of several unstable mammalian mRNAs.From the above experiments, several generalizations can be made. First, it is clear thatboth nuclear and cytoplasmic proteins can bind specifically to the ARE. The significance ofthe nuclear-binding proteins remains obscured by a lack of knowledge concerning thecontribution of nuclear mRNA processing reactions to the ultimate fate of mRNA in thecytoplasm. Second, the binding activity of some of these proteins appears to increase ordecrease in response to changes in cellular metabolism that lead to alterations in the stabilitieson ARE-containing mRNAs. Finally, the existence of a 20S complex on an unstable ARE-containing mRNA and the absence of this complex on ARE-containing mRNAs that are stableowing to ARE inactivation (Savant-Bhonsale and Cleveland, 1992) suggest that ARE activitymay be mediated by a complex of proteins.Although the role played by ARE-binding proteins is not yet clear, the mechanism bywhich this element stimulates decay has been partially defined. Shyu et al. (1991) have shownthat the ARE in c-fos transcripts mediates decay by first stimulating deadenylation and then byproviding an element that stimulates the next phase of the decay process. Similarly, severalother ARE-containing mRNAs have been found that go through a deadenylation and then adecay process. Thus, deadenylation may be a prerequisite for the decay of this class ofunstable mRNAs and the deadenylated mRNAs may represent a key early intermediate in thedegradation pathway (Wilson and Triesman, 1988).231.4.6 Endo- and ExoribonucleasesIdentification of the ribonucleases involved will undoubtedly shed additional light on themechanisms that regulate mRNA degradation. Toward this end, characterization of endo- andexoribonuclease activities that degrade specific mRNAs in cell-free extracts from mammalian(Bandyopadhyay et at., 1990; Wager and Assoian, 1990), chicken (Ling and Jost, 1991), andyeast cells (Vreken et at., 1992), as well as in Xenopus oocytes (Brown and Harl, 1990) hasbegun. In the latter system, specific endonucleolytic cleavage is observed at a repeatedsequence within the 3’UTR of a maternal homeobox mRNA (Brown and Harl, 1990). In yeast,two exoribonucleases encoded by the XRN 1 and poly(A) nuclease (PAN) 1 genes have beencharacterized recently. The XRN 1 gene encodes a 160 kD 5’-3’ exoribonuclease that degradesRNAs with 5’monophosphates, but does not digest capped RNAs (Larimer and Stevens, 1990;Stevens and Maupin, 1987; Stevens et at., 1991). Cells that are deleted for XRN1 are viable,but grow slowly, have altered rRNA processing and appear to stabilize the decay of severalmRNAs (Larimer and Stevens, 1990; Stevens et at., 1991). It is not clear whether suchstabilization is a direct consequence of the deletion of the XRN1 gene on mRNA decaypathways, or a result of reduced rates of growth and protein synthesis.mRNA polyadenylation occurs in the nucleus of cells, and the posttranscriptionallysynthesized tail is typically homogeneous in length, ranging form 70-90 nt in yeast to 220-250nt in mammalian cells (Reviewed in Sachs, 1990). Following transport of the mRNA, the longpoly(A) tail is shortened in the cytoplasm. In yeast, this shortening reaction requires thepresence of a specific protein bound to the poly(A) tail, the poly(A)-binding protein (PAB), forfull efficiency (Sachs and Davis, 1989), and it is this reaction that could be regulated by thedifferent destabilizing sequences on many RNAs. The poly(A) shortening reaction can beseparated into two phases, the first being the shortening of the tail down to 12-25 residues andthe second terminal deadenylation step being the removal of some or all of these residues.24The gene encoding the ribonuclease, PAN, involved in PAB-dependent shortening ofpoly(A) tails from yeast has been cloned (Sachs and Deardorif, 1992). This ribonuclease isunique among identified eukaryotic RNAases in that it requires a protein-RNA complex as asubstrate. Furthermore, conditional mutations in PAN reveal that it is also a translationinitiation factor. PAN deadenylation activity in vitro exhibits the two phases of shorteningseen in vivo, and both the shortening phase and the terminal deadenylation phase can beregulated by 3’UTR sequences.Characterization of PAN has revealed some interesting features that may have generalimplications for RNA degradation. First, the substrate for this ribonuclease is not nakedmRNA, suggesting that the search for RNAases that specifically recognize sequences withinmRNA that have been identified as destabilizing sequences may be unsuccessful unless theproper ribonucleoprotein substrate is identified. The concept that the nucleolytic substrate is aribonucleoprotein complex also negates the simple hypothesis that RNA-binding proteins willalways protect mRNA from decay. Second, PAN activity can be regulated from distant sites.This argues that destabilizing sequences may not harbor nucleolytic sites, but instead maytarget sites elsewhere on the mRNA for degradation. A corollary of this is that theidentification of endonucleolytic cleavage sites on RNA decay intermediates may not yieldinformation about what region of the mRNA makes these sites nuclease-sensitive.Furthermore, the ability of PAN to be activated or inhibited over distances in vitro is consistentwith the hypothesis that destabilizing sequences located throughout an mRNA can affect bothdeadenylation and decay processes. Finally, the requirement for PAN in both translation andpresumably RNA decay in vivo leads to the suggestion that some of the factors involved inmRNA decay may also be complexed with translation factors. This conclusion is consistentwith the apparent requirement for translation in mRNA decay.251.5 THESIS OBJECTIVESThe regulated expression of ICAM- 1 by various tissues and cell types appears to play a vitalrole in numerous physiologic and pathologic processes of the immune system. Modulation ofICAM- 1 expression during inflammation by pharmacologic agents might therefore be anattractive target for medical treatments. However, this requires an understanding of theintracellular regulatory elements and signaling pathways that are involved in the increasedexpression of ICAM-1 that is induced by proinflammatory cytokines. The overall objective ofthis thesis study was to test the hypothesis that the regulation of ICAM- 1 gene expression bytwo inflammatory mediators, IFN-y and PMA, involves a posttranscriptional mechanism andthen to define the possible roles of phosphorylation pathways and translation in the process bywhich cycloheximide induces ICAIvI-l message stabilization.1.6 REFERENCESAdams DH, Hubscher SG, Shaw J, Rothlein R, Neuberger JM (1989) Intercellular adhesionmolecule 1 on liver allografts during rejection. Lancet 2:1122.Aharon T, Schneider RJ (1993) Selective destabilization of short-lived mRNAs with thegranulocyte-macrophage colony-stimulating factor AU-rich Y noncoding region is mediated bya cotranslational mechanism. Mol Cell Biol 13:1971.Altamura N, Groudinsky 0, Dujardin G, Slonimski PP (1992) NAM7 nuclear gene encodes anovel member of a family of helicases with a Zn-ligand motif and is involved in mitochondrialfunctions in Saccharomyces cerevisiae. J Mol Biol 224:575.Anderson DC, Miller U, Schmalstieg FC, Rothlein R, Springer TA (1986) Contribution of theMac-i glycoprotein family to adherence-dependent granulocyte functions: Structure-functionassessments employing subunit-specific monoclonal antibodies. J Immunol 137:15.Anderson DC, Springer TA (1987) Leukocyte adhesion deficiency: An inherited defect in theMac-i, LFA-i, and p150,95 glycoproteins. Annu Rev Med 38: 175.26Arfors K-E, Lundberg C, Lindborm L, Lundberg K, Beatty PG, Harlan M (1987) Amonoclonal antibody to the membrane glycoprotein complex CD 18 inhibitspolymorphonuclear leukocyte accumulation and plasma leakage in vivo. Blood 69:338.Bainton DF, Miller U, Kishimoto TK, Springer TA (1987) Leukocyte adhesion receptors arestored in peroxidase-negative granules of human neutrophils. J Exp Med 166:1641.Ballantyne CM, Kozak CA, O’Brien WE, Beaudet AL (1991) Assignment of the gene forintercellular adhesion molecule-l (Icam-1) to proximal mouse chromosome 9. Genomics9:547.Bandyopadhyay R, Coutts M, Krowczynska A, Brawerman G (1990) Nuclease activityassociated with mammalian mRNA in its native state: Possible basis for selectivity in mRNAdecay. Mol Cell Biol 10:2060.Belier DI, Springer TA, Schreiber RD (1982) Anti-Mac-i selectively inhibits the mouse andhuman type three complement receptor. J Exp Med 156:1000.Berendt AR, Simmons DL, Tansey J, Newbold CI, Marsh K (1989) Intercellular adhesionmolecule-i is an endothelial cell adhesion receptor for Plasmodium falciparum. Nature 341:57.Berger M, O’Shea J, Cross AS, Folks TM, Chused TM, Brown EJ, Frank MM (1984) Humanneutrophils increase expression of C3bi as well as C3b receptors upon activation. J Clin Invest74: 1566.Bernstein PL, Herrick DJ, Prokipcak RD, Ross J (1992) Control of c-myc mRNA half-life invitro by a protein capable of binding to a coding region stability determinant. Genes Dev6:642.Billaud M, Calender A, Seigneurin J, Lenoir GM (1987) LFA-l, LFA-3 and ICAM-iexpression in Burkitt’s lymphoma. Lancet 2:1327.Bohjanen PR, Petryniak B, June CH, Thompson CB, Lindsten T (1991) An induciblecytoplasmic factor (AU-B) binds selectively to AUUUA multimers in the 3’ untranslatedregion of lymphokine mRNA. Mol Cell Biol 11:3288.Bohjanen PR, Petryniak B, June CH, Thompson CB, Lindsten T (1992) AU RNA-bindingfactors differ in their binding specificities and affinities. J Biol Chem 267:6302.Boyd AW, Wawryk SO, Burns GF, Fecondo JV (1988) Intercellular adhesion molecule 1(ICAM- 1) has a central role in cell-cell contact-mediated immune mechanisms. Proc Natl AcadSci USA 85:3095.Brewer G (1991) An A+U-rich element RNA-binding factor regulates c-myc mRNA stabilityin vitro. Mol Cell Biol 11:2460.27Brown BD, Han RM (1990) Endonucleolytic cleavage of a maternal homeo box mRNA inXenopus oocytes. Genes Dcv 4:1925.Caput D, Beutler B, Hartog K, Thayer R, Brown-Shimer S, Cerami A (1986) Identification of acommon nucleotide sequence in the 3’-untranslated region of mRNA molecule specifyinginflammatory mediators. Proc Nati Acad Sci USA 83:1670.Carlos TM, Harlan, JM (1994) Leukocyte-endothelial adhesion molecules. Blood 84:2068.Carpen 0, Pallai P, Staunton DE, Springer TA (1992) Association of intercellular adhesionmolecule-i (ICAM- 1) with actin-containing cytoskeleton and a-actinin. J Cell Biol 118:1223.Chen C-YA, You Y, Shyu A-B (1992) Two cellular proteins bind specifically to a pruine-nichsequence necessary for the dstabilization function of a c-fos protein-coding region determinantof mRNA instability. Mol Cell Biol 12:5748.Cigan AM, Pabich EK, Donahue TF (1988) Mutational analysis of the HIS4 translationalinitiator region in Saccharomyces cerevisiae. Mol Cell Biol 8:2964.Clark EA, Ledbetter JA, Holly RC, Dinndorf PA, Shu G (1986) Polypeptide on human Blymphocytes associated with cell activation. Hum limnunol 16:100.Clayberger C, Wright A, Medeiros U, Koller TD, Link MP, Smith SD, Warnke RA, KrenskyAM (1987) Absence of cell surface LFA-i as a escape from immunosurveillance. Lancetii: 1327.Cosimi AB, Conti D, Delmonico FL, Preffer FE, Wee S-L, Rothlein R, Faanes R, Colvin RB(1990) In vivo effects of monoclonal antibody to ICAM-1 (CD54) in nonhuman primates withrenal allografts. J Immunol 144:4604.Davignon D, Martz E, Reynolds T, Springer TA (1981) Monoclonal antibody to a novellymphocyte function-associated antigen (LFA-i): Mechanism of blocking of T lymphocyte-mediated killing and effects on other T and B lymphocyte functions. J Immunol 127:590.de Fougerolles AR, Klickstein LB, Springer TA (1993) Cloning and expression of intercellularadhesion molecule 3 reveals strong homology to other immunoglobulin family counter-receptors for lymphocyte function-associated antigen 1. J Exp Med 177:1187.de Fougerolles AR, Stacker SA, Schwarting R, Springer TA (1991) Characterization of ICAM2 and evidence for a third counter-receptor for LFA-1. J Exp Med 174:253.de Fougerolles AR, Xi Qin, Springer TA (1994) Characterization of the function ofintercellular adhesion molecule (ICAM)-3 and comparison with ICAM- 1 and ICAM-2 inimmune responses. J Exp Med 179:619.28Degitz K, Lian-Jie L, Caughman SW (1991) Cloning and characterization of the 5’-transcriptional regulatory region of the human intercellular adhesion molecule 1 gene. J BiolChem 266:14024.Diamond MS, Staunton DE, Marlin SD, Springer TA (1991) Binding of integrin Mac-i(CD1 lb/CD 18) to the third immunoglobulin-like domain of.ICAM-1 (CD54) and its regulationby glycosylation. Cell 65:961.Dougherty GJ, Hogg N (1987) The role of monocyte lymphocyte function-associated antigen 1(LFA-i) in accessory cell function. Eur J Immunol 17:943.Dougherty GJ, Murdoch 5, Hogg N (1988) The function of human intercellular adhesionmolecule-i (ICAM-i) in the generation of an immune response. Eur J Immunol 18:35.Dustin ML, Rothlein R, Bhan AK, Dinarello CA, Springer TA (1986) A natural adherencemolecule (ICAIvI- 1): induction by IL-i and interferon-’y, tissue distribution, biochemistry andfunction. J Immunol 137:245.Dustin ML, Singer KH, Tuck DT, Springer TA (1988) Adhesion of T lymphoblasts toepidermal keratinocytes is regulated by interferon gamma and is mediated by intercellularadhesion molecule-i (ICAM-i). J Exp Med 167:1323.Dustin ML, Springer TA (1988) Lymphocyte function-associated antigen-i (LFA-i)interaction with intercellular adhesion molecule-i (ICAIVI-i) is one of at least threemechanisms for lymphocyte adhesion to cultured endothelial cells. J Cell Biol 107:321.Dustin ML, Springer TA (1991) Role of lymphocyte adhesion receptors in transientinteractions and cell locomotion. Annu Rev Immunol 9:27.Fawcett J, Holness CLL, Needham LA, Turley H, Gatter KC, Mason DY, Simmons DL (1992)Molecular cloning of ICAM-3, a third ligand for LFA- 1, constitutively expressed on restingleukocytes. Nature 360:481.Frohman EM, Frohman TC, Dustin ML, Vayuvegula B, Choi B, Gupta A, van den Noort 5,Gupta S (1989) Induction of ICAM-l expression on human fetal astrocytes by interferon-gamma, tumor necrosis factor-alpha and interleukin- 1: relevance to intracerebral antigenpresentation. J Neuroimmunol 23:117.Giger U, Boxer LA, Simpson PJ, Lucchesi BR, Todd RF (1987) Deficiency of leukocytesurface glycoproteins Mo 1, LFA- 1, and Leu M5 in a dog with recurrent bacterial infections:An animal model. Blood 69:1622.Gillis P, Malter JS (1991) The adenosine-uridine binding factor recognizes the AU-richelements of cytokine, lymphokine, oncogene mRNAs. J Biol Chem 266:3172.Greve JM, Davis G, Meyer AM, Forte CP, Yost SC, Marlor CW, Kamarck ME, McClelland A(1989) The major human rhinovirus receptor is ICAM-i. Cell 56:839.29Griffiths CEM, Nickoloff BJ (1989) Keratinocyte intercellular adhesion molecule-i (ICAIVI- 1)expression precedes dermal T lymphocytic infiltration in allergic contact dermatitis (Rhusdermatitis). Am J Pathol 135:1045.Harris ME, Bohni R, Schneiderman MH, Ramamurthy L, Schumperli D, Marzloff WF (1991)Regulation of histone mRNA in the unperturbed cell cycle: evidence suggesting control at twoposttranscriptional steps. Mol Cell Biol 11:2416.Haskard D, Cavender D, Beatty P, Springer TA, Ziff M (1986) T lymphocyte adhesion toendothelial cells: Mechanisms demonstrated by anti-LFA-1 monoclonal antibodies. J Immunol137:290 1.Heaton B, Decker C, Muhirad D, Donahue J, Jacobson A, Parder R (1992) Analysis ofchimeric mRNAs derived from the STE3 mRNA identifies multiple regions within yeastmRNAs that modulate mRNA decay. Nucl Acids Res 20:5365.Herrick D, Parker R, Jacobson A (1990) Identification and comparison of stable and unstablemRNAs in Saccharomyces cerevisiae. Mol Cell Biol 10:2269.Hogg N, Bates PA, Harvey J (1991) Structure and function of intercellular adhesion molecule1. Chem Immunol 50:98.Holzmann B, Lehmann JM, Ziegler-Heitbrock HWL, Funke I, Riethmuller G, Johnson JP(1988) Glycoprotein P3.58, associated with tumor progression in malignant melanoma, is anovel leukocyte activation antigen. Tnt J Cancer 4 1:542.Horley KJ, Carpenito C, Baker B, Takei F (1989) Molecular cloning of murine intercellularadhesion molecule (ICAM-l). EMBO J 8:2889.Johnson JP, Stade BG, Holzmann B, Schwable W, Riethmuller G (1989) De novo expressionof intercellular adhesion molecule 1 in melanoma correlates with increased risk of metastasis.Proc Natl Acad Sci USA 86:641.Katz FE, Parker M, Stanley K, Murray U, Clark EA, Greaves MF (1985) Chromosomemapping of cell membrane antigens expressed on activated B cells. Eur J Immunol 15:103.Kirchhausen T, Staunton DE, Springer TA (1993) Location of the domains of ICAM-i byimmunolabeling and single-molecule electron microscopy. J Leukoc Biol 53:342.Klausner RD, Rouault TA, Harford JB (1993) Regulating the fate of mRNA: the control ofcellular iron metabolism. Cell 72:19.Koeller DM, Horowitz JA, Casey JL, Klausner RD, Harford JB (1991) Translation and thestability of mRNAs encoding the transferrin receptor and c-fos. Proc Natl Acad Sci USA88:7778.30Kohl S, Springer TA, Schmalstieg FC, Loo LS, Anderson DC (1984) Defective natural killercytotoxicity and polymorphonuclear leukocyte antibody-dependent cellular cytotoxicity inpatients with LFA-i/OKM-i deficiency. J Immunol 133:2972.Krensky AM, Sanchez-Madrid F, Robbins E, Nagy J, Springer TA, Burakoff SJ (1983) Thefunctional significance, distribution, and structure of LFA- 1, LFA-2, and LFA-3: cell surfaceantigens associated with CTL-target interactions. J Immunol 131:611.Languino LR, Plescia J, Duperray A, Brian AA, Plow EF, Geltosky JE, Aitieri DC (1993)Fibrinogen mediates leukocyte adhesion to vascular endothelium through an ICAM-1-dependent pathway. Cell 73:1423.Lanier LL, Arnaout MA, Schwarting R, Warner NL, Ross GD (1985) pl5O/95 third memberof the LFA-1/CR3 polypeptide family identified by anti-Leu M5 monoclonal antibody. Eur JImmunol 15:713.Larimer FW, Stevens A (1990) Disruption of the gene XRNJ, coding for a 5’-3’exoribonuclease, restricts yeast cell growth. Gene 95:85.Larson RS, Springer TA (1990) Structure and function of leukocyte integrins. Immunol Rev114: 181.Leeds P, Peltz SW, Jacobson A, Culbertson MR (1991) The product of the yeast UPFJ gene isrequired for rapid turnover of mRNAs containing a premature translocational terminationcodon. Genes Dev 5:2303.Leeds P, Wood JM, Lee B-S, Culbertson MR (1992) Gene products that promote mRNAturnover in Saccharomyces cerevisiae. Mol Cell Biol 12:2165.Lenardo MJ, Baltimore D (1989) NK-icB: a pieiotropic mediator of inducible and tissue-specific gene control. Cell 58:227.Liang H, Jost J-P (1991) An estrogen-dependent polysomal protein binds the 5’ untranslatedregion of the chicken vitellogenin mRNA. Nucleic Acids Res 19:2289.Makgoba MW, Sanders ME, Ginther-Luce GE, Dustin ML, Springer TA, Clark EA, MannoniP. Shaw S (1989) ICAM-i is a ligand for LFA-1-dependent adhesion of B, T and myeloidcells. Nature 33 1:86.Marlin SD, Springer TA (1987) Purified intercellular adhesion molecule-i (ICAM- 1) is aligand for lymphocyte function-associated antigen 1 (LFA- 1). Cell 51:813.Marlin SD, Staunton DE, Springer TA, Stratowa C, Sommergruber W, Merluzzi VJ (1990) Asoluble form of intercellular adhesion molecule-i inhibits rhinovirus infection. Nature 344:70.Martz E (1987) LFA-1 and other accessory molecules functioning in adhesions of T and Blymphocytes. Human Immunol 18:3.31Mentzer SJ, Rothlein R, Springer TA, Faller DV (1988) Intercellular adhesion molecule-i(ICAIVI-1) is involved in the cytolytic T lymphocyte interaction with human synovial cells. JCell Physiol 137:173.Miller U, Bainton DF, Borregaard N, Springer TA (1987) Stimulated mobilization ofmonocyte Mac-i and p150,95 adhesion proteins from an intracellular vesicular compartment tothe cell surface. J Clin Invest 80:535.Mimura N, Asano A (1987) Further characterization of a conserved actin-binding 27-kDafragment of actinogelin and alpha-actinins and mapping of their binding sites on the actinmolecule by chemical crosslinking. J Biol Chem 262:47 17.Monro JM, Pober JS, Cotran RS (1989) Tumor necrosis factor and interferon-gamma inducedistinct patterns of endothelial activation and leukocyte accumulation in skin of Papio anubis.AmJPathol 135:121.Montgomery KF, Osborn L, Hession C, Tizard R, Goff D, Vassallo C, Tarr PT, Bomsztyk K,Lobb R, Harlan JM, Pohlman TH (1991) Activation of endothelial-leukocyte adhesionmolecule 1 (ELAM- 1) gene transcription. Proc Nati Acad Sci USA 88:6523.Moses H, Panek RB, Benveniste EN, Ting JP-Y (1992) Usage of primary cells to delineate1FN-y-responsive DNA elements in the HLA-DRA promoter and to identify a novel IFN-yenhanced nuclear factor. J Immunol 148:3643.Myer VE, Lee SI, Steitz JA (1992) Viral small nuclear ribonucleoproteins bind a proteinimplicated in messenger RNA destabilization. Proc Natl Acad Sci USA 89:1296.Natali P, Nicotra MR, Cavaliere R, Bigotti A, Romano G, Temponi M, Ferrone S (1990)Differential expression of intercellular adhesion molecule 1 in primary and metastaticmelanoma lesions. Cancer Res 50:1271.Peltz SW, Brewer G, Berstein P, Hart PA, Ross J (1991) Regulation of mRNA turnover ineucaryotic cells. Crit Rev Euc Gene Expr 1:99.Peltz SW, Brown AH, Jacobson A (1993) mRNA destabiliation triggered by prematuretranslational termination depends on at least three cis-acting sequence elements and one transacting factor. Genes Dev 7:1737.Peltz SW, Donahue IL, Jacobson A (1992) A mutation in the tRNA nucleotidyltransferasegene promotes stabilization of mRNAs in Saccharomyces cerevisiae. Mol Cell Biol 12:5778.Peltz SW, Jacobson A (1993) mRNA turnover in Saccharomyces cerevisiae. In Control ofmRNA stability. Edited by Brawerman G, Belasco JG. New York, Academic Press.Petersen R, Lindquist S (1989) The Drosophila hsp7O message is rapidly degraded at normaltemperatures and stabilized by heat shock. Gene 72:161.32Pober JS, Gimbrone MA Jr, Lapierre LA, Mendrick DL, Fiers W, Rothlein R, Springer TA(1986) Overlapping patterns of activation of human endothelial cells by interleukin- 1, tumornecrosis factor, and immune interferon. 1 Immunol 137:1893.Rhoads R (1993) Regulation of eukaryotic protein synthesis by initiation factors. I Biol Chem268:3017.Rosen H, Gordon S (1987) Monoclonal antibody to the murine type 3 complement receptorinhibits adhesion of myelomonocytic cells in vitro and inflammatory cell recruitment in vivo. JExpMed 166:1685.Rosenstein Y, Park 1K, Hahn WC, Rosen FS, Bierer BE, Burakoff SJ (1991) CD43, amolecule defective in Wiskott-Aldrich syndrome, binds ICAM- 1. Nature 354:233.Rothlein R, Czajkowski M, O’Neill MM, Marlin SD, Mainolfi E, Merluzzi VJ (1988)Induction of intercellular adhesion molecule-i on primary and continuous cell lines by pro-inflammatory cytokines. I Immunol 141:1665.Rothlein R, Dustin ML, Marlin SD, Springer TA (1986) A human intercellular adhesionmolecule (ICAM-1) distinct from LFA-1. I Immunol 137:1270.Sachs AB (1990) The role of poly(A) in the translation and stability of mRNA. Curr Opin CellBiol 2:1092.Sachs AB, Davis RW (1989) The poly(A) binding protein is required for poly(A) shorteningand 60S ribosomal subunit-dependent translation initiation. Cell 5 8:857.Sachs AB, Deardorif IA (1992) Translation initiation requires the PAB-dependent poly(A)ribonuclease in yeast. Cell 70:961.Sanchez-Madrid F, Nagy I, Robbins E, Simon P, Springer TA (1983) A human leukocytedifferentiation antigen family with distinct alpha subunits and a common beta subunit: thelymphocyte function-associated antigen (LFA- 1), the C3bi complement receptor (0KM 1/Mac-1), and the p150,95 molecule. I Exp Med 158: 1785.Savant-Bhonsale 5, Cleveland DW (1992) Evidence for instability of mRNAs containingAUUUA motifs mediated through translation-dependent assembly of a >20S degradationcomplex. Genes Dev 6:1927.Schiavi SC, Belasco JG, Greenberg ME (1992) Regulation of proto-oncogene niRNA stability.Biochim Biophys Acta 1114:95.Schuler GD, Cole MD (1988) GM-CSF and oncogene mRNA stabilities are independentlyregulated in trans in a mouse monocytic tumor. Cell 55:1115.Shaw G, Kamen R (1986) A conserved AU sequence from the 3’ untranslated region of GMCSF mRNA mediates selective mRNA degradation. Cell 46:659.33Shyu A-B, Belasco JG, Greenberg ME (1991) Two distinct destabilizing elements in the c-fosmessage trigger deadenylation as a first step in rapid mRNA decay. Genes Dev 5:221.Shyu A-B, Greenberg ME, Belasco JG (1989) The c-fos transcript is targeted for rapid decayby two distinct mRNA degradation pathways. Genes Dev 3:60.Simmons D, Makgoba MW, Seed B (1988) ICAM, an adhesion ligand of LFA-l, ishomologous to the neural cell adhesion molecule NCAM. Nature 331:624.Siu G, Hedrick SM, Brian AA (1989) Isolation of the murine intercellular adhesion molecule-i(ICAM- 1) gene: ICAM- 1 enhances antigen-specific T cell activation. I Immunol 143:3813.Sligh JE, Ballantyne CM, Rich SS, Hawkins HK, Smith CW, Bradley A, Beaudet AL (1993)Inflammatory and immune responses are impaired in mice deficient in intercellular adhesionmolecule 1. Chem Immunol 90:8529.Smith CW, Rothlein R, Hughes BJ, Mariscalco MM, Schmalstieg FC, Anderson DC (1988)Recognition of an endothelial determinant for CD 18-dependent neutrophil adherence andtransendothelial migration. I Clin Invest 82:1746.Sobel RA, Mitchell ME, Fondren G (1990) Intercellular adhesion molecule-i (ICAM-1) incellular immune reactions in the human central nervous system. Am J Pathol 136:1309.Springer TA (1990) Adhesion receptors of the immune system. Nature 346:425.Springer TA (1994) Traffic signals for lymphocyte recirculation and leukocyte emigration: themultistep paradigm. Cell 76:301.Springer TA, Dustin ML, Kishimoto TK, Marlin SD (1987) The lymphocyte function-associated LFA- 1, CD2, and LFA-3 molecules: cell adhesion receptors of the immune system.Annu Rev Immunol 5:223.Springer TA, Thompson WS, Miller U, Schmalstieg FC, Anderson DC (1984) Inheriteddeficiency of the Mac-i, LFA- 1, p150,95 glycoprotein family and its molecular basis. J ExpMed 160: 1901.Staunton DE, Dustin ML, Erickson HP, Springer TA (1990) The arrangement of theimmunoglobulin-like domains of ICAM-l and the binding sites for LFA-l and rhinovirus. Cell61:243.Staunton DE, Dustin ML, Springer TA (l989b) Functional cloning of ICAM-2, a cell adhesionligand for LFA-1 homologous to ICAM-1. Nature 339:61.Staunton DE, Marlin SD, Stratowa C, Dustin ML, Springer TA (1988) Primary structure ofICAM- 1 demonstrates interaction between members of the immunoglobulin and integrinsupergene families. Cell 52:925.34Staunton DE, Merluzzi VJ, Rothlein R, Barton R, Marlin SD, Springer TA (1989a) A celladhesion molecule, ICAIVI- 1, is the major surface receptor for rhinoviruses. Cell 56:849.Staunton DE, Ockenhouse CF, Springer TA (1992) Soluble intercellular adhesion molecule 1-immunoglobulin Gi immunoadhesin mediates phagocytosis of malaria-infected erythrocytes. JExpMed 176:1471.Stevens A (1988) mRNA-decapping enzyme from Saccharomyces cerevisiae: purification andunique substrate specificity for long RNA chains. Mol Cell Biol 8:2005.Stevens A, Hsu CL, Isham KR, Larimer FW (1991) Fragments of the internal transcribedspacer 1 of pre-rRNA accumulate in Saccharomyces cerevisiae lacking 5-3’ exoribonuclease 1.J Bacteriol 173:7024.Stevens A, Maupin MK (1987) A 5-3’ exoribonuclease of Saccharomyces cerevisiae. Sizenovel substrate specificity. Arch Biochem Biophys 252:339.Takahama Y, Singer A (1992) Post-transcriptional regulation of early T cell development by Tcell receptor signals. Science 258:1456.te Velde AA, Keizer GD, Figdor CG (1987) Differential function of LFA-l family molecules(CD1 1 and CD 18) in adhesion of human monocytes to melanoma and endothelial cells.Immunology 61:261.Temponi M, Romano G, D’Urso CM, Wang Z, Kekish U, Ferrone S (1989) Pofile ofintercellular adhesion molecule-i (ICAM- 1) synthesized by human melanoma cell lines.Seminars Oncol 15:595.Todd RF, Arnout MA, Rosin RE, Crowley CA, Peters WA, Babior BM (1984) Subcellularlocalization of the large subunit of Mol (Mol alpha; formerly gp 110), a surface glycoproteinassociated with neutrophil adhesion. J Clin Invest 74:1280.Todd RF, Freyer DR (1988) The CD1 1/CD18 leukocyte glycoprotein deficiency. HemlOncClinics NA 2:13.Vakalopoulou E, Schaack J, Schenk T (1991) A 32-kilodalton protein binds to AU-richdomains in the 3’ untranslated regions of rapidly degraded mRNAs. Mol Cell Biol 11:3355.Vazeux R, Hoffman PA, Tomita JK, Dickinson ES, Jasman RL, St John T, Gallatin MW(1992) Cloning and characterization of a new intercellular adhesion molecule ICAM-R. Nature360:485.Vogetseder W, Feichtinger H, Schulz TF, Schwaeble W (1989) Expression of 7F7-antigen, ahuman adhesion molecule identical to intercellular adhesion molecule-i (ICAM-1) in humancarcinomas and their stromal fibroblasts. Int J Cancer 43:768.35Voraberger G, Schafer R, Stratowa C (1991) Cloning of the human gene for intercellularadhesion molecule 1 and analysis of its 5-regulatory region: induction by cytokines andphorbol ester. J Immunol 147:2777.Vreken P, Buddelmeijer N, Raue HA (1992) A cell-free extract from yeast cells for studyingmRNA turnover. Nucleic Acids Res 20:2503.Wager RE, Assoian RK (1990) A phorbol ester-regulated ribonuclease system controllingtransforming growth factor 31 gene expression in hematopoietic cells. Mol Cell Biol 10:5983.Wantzin GL, Ralfkiaer E, Avnstorp C, Czajkowski M, Marlin SD, Rothlein R (1989) Kineticsand characterization of intercellular adhesion molecule-i (ICAM- 1) expression onkeratinocytes in various inflammatory skin lesions and malignant cutaneous lymphomas. J AmAcad Dermatol 20:782.Wawryk SO, Novotny JR, Wicks IP, Wilkinson D, Maher D, Salvaris E, Welch K, Fecondo J,Boyd AW (1989) The role of the LFA-1JICAM-1 interaction in human leukocyte homing andadhesion. Immunol Rev 108:135.Wegner CD, Gundel RH, Reilly P, Haynes N, Letts LG, Rothlein R (1990) Intercellularadhesion molecule-i (ICAM-1) in the pathogenesis of asthma. Science 247:456.Wellington CL, Greenberg ME, Belasco JG (1993) The destabilizing elements in the codingregion of c-fos mRNA are recognized as RNA. Mol Cell Biol 13:5034.Wertheimer SJ, Myers CL, Wallace RW, Parks TP (1992) Intercellular adhesion molecule-igene expression in human endothelial cells: differential regulation by tumor necrosis factor-aand phorbol myristate acetate. J Biol Chem 267:12030.Williams AF (1987) A year in the life of the immunoglobulin superfamily. Immunol Today8:298.Williams AF, Barclay AN (1988) The immunoglobulin superfamily-domains for cell surfacerecognition. Annu Rev Immunol 6:381.Wilson RW, Ballantyne CM, Smith CW, Montgomery C, Bradley A, O’Brien WE, BeaudetAL (1993) Gene targeting yields a CD 18-mutant mouse for study of inflammation. J Immunol151:1571.Wilson T, Triesman R (1988) Removal of poly(A) consequent degradation of c-fos mRNAfacilitated by 3’ AU-rich sequences. Nature 336:396.Xu H, Gonzalo JA, St. Pierre Y, Williams IR, Kupper TS, Cotran RS, Springer TA, GutierrezRamos J-C (1994) Leukocytosis and resistance to septic shock in intercellular adhesionmolecule 1-deficient mice. I Exp Med 180:95.36Xu H, Tong IL, de Fougerolles AR, Springer TA (1992) Isolation, characterization, andexpression of mouse ICAM-2 complementary and genomic DNA. J Immunol 149:2650.Yen TJ, Gay DA, Pachter JS, Cleveland DW (1988) Autoregulated changes in stability ofpolyribosome-bound J3-tubulin mRNAs are specified by the first thirteen translatednucleotides. Mol Cell Biol 8:1224.You Y, Chen C-YA, Shyu A-B (1992) U-rich sequence-binding proteins (URBPs) interactingwith a 20-nucleotide U-rich sequence in the 3’-untranslated region of c-fos mRNA may beinvolved in the first step of c-fos mRNA degradation. Mol Cell Biol 12:2931.37IIPOSTTRANSCRIPTIONAL REGULATION OF ICAM-1 GENE EXPRESSIONBY IFN-y AND PMA2.1 ABSTRACTAlthough ICAM- 1 is constitutively expressed at a low level on a subpopulation ofhematopoietic cells, vascular endothelium, fibroblasts, and certain epithelial cells, it isdramatically increased at sites of inflammation. IFN-’y and PMA are known to increase theexpression of ICAM-1 on many cell types. Because both human and murine ICAM-1 mRNAscontain putative destabilizing AUUUA sequences in their 3’UTRs, the role of mRNA stabilityin the regulation of ICAM- 1 gene expression was examined. The treatment of the murinemonocytic cell line P388D 1, which constitutively expresses ICAM- 1 mRNA at a low level,with IFN-y or PMA rapidly enhanced the level of ICAM-1 mRNA and dramatically prolongedits half-life. To determine whether the putative destabilizing sequences are responsible for thiseffect of IFN-y and PMA, fibroblast L cells were transfected with either the full-length ICAM1 cDNA or a truncated form (ICAM-1A3) lacking the putative destabilizing AUUUAsequences. Although ICAM-1A3 mRNA was more stable than the full-length ICAM-1mRNA, WN-y treatment induced the accumulation of both mRNA species and prolongation oftheir half-lives. The transplantation of the ICAM-1 3’UTR into a stable ICAM-2 mRNArendered it unstable and was unresponsive to IFN-’y. Therefore, the treatment with WN-ystabilizes the otherwise labile ICAM- 1 mRNA, but the IFN-y-responsive sequence may at leastin part reside within the protein coding region. PMA also upregulated ICAM- 1 geneexpression by mRNA stabilization. However, unlike IFN-y, PMA treatment only increased thelevel of the full-length, but not the truncated, ICAM- 1 mRNA. This demonstrates that the38PMA-responsive element is located within the 3’UTR. Furthermore, the effect of PMA onICAM-1z3 mRNA was recovered by ligating multiple AUUUA sequences derived from aheterologous gene fragment. The stability of this chimeric mRNA and the full-length ICAM- 1mRNA was markedly increased by PMA treatment, indicating that the AUUUA multimers inthe 3’UTR are important in the PMA-induced upregulation of ICAM-l mRNA.2.2 INTRODUCTIONICAM- 1 is constitutively expressed at a low level on a subpopulation of hematopoietic cells,vascular endothelium, human fibroblasts, and certain epithelial cells (Dustin et al., 1986), andit is dramatically increased at sites of inflammation (Rothlein et at., 1986; Dustin et at., 1986;Pober et at., 1986). The upregulated expression of ICAM-1 on venule endothelium is thoughtto facilitate the adhesion and subsequent trans-endothelial migration of leukocytes bearingLFA-i or Mac-i into inflammatory tissues (Rosenstein et at., 1991; Smith et at., 1989), as wellto as permit appropriate interaction of lymphocytes with cells expressing targeted antigens(Springer, 1990; Wawryk et at., 1989; Dougherty et at., 1988). Various pro-inflammatorymediators including IL-i, TNF-cx, and IFN-’y have been found to increase ICAM-1 expressionon many cell types and are thought to be responsible for the induction of ICAM- 1 expression atinflammatory sites in vivo (Rothlein et at., 1986; Dustin et at., 1986; Pober et at., 1986;Rothlein et at., 1988).Despite the importance of ICAM- 1 regulation in various immunologic responses, verylittle is known about the intracellular mechanisms controlling its expression. Potential IFN-yresponsive elements have been identified within the 5-flanking region of the ICAM-1 gene(Degitz et at., 1991; Voraberger et al., 1991). However, only a two-fold induction wasachieved by WN-’ using a reporter assay (Voraberger et al., 1991), suggesting that ICA1M-1gene regulation by IFN-’y may involve posttranscriptional mechanisms. PKC has also been39implicated in the induction of ICAIVI- 1, since active phorbol esters such as PMA are known toactivate PKC as well as to upregulate cell surface ICAM- 1. Furthermore, it has beendemonstrated that PMA acts posttranscriptionally to decrease the degradation rate of ICAM- 1mRNA through a PKC-dependent pathway, allowing the constitutively produced message toaccumulate (Wertheimer et al., 1992). However, the PMA-responsive element on the ICAM-1mRNA has not yet been identified.There is growing evidence that mRNA turnover plays an important role in regulatinggene expression. The steady-state levels of many mRNAs do not necessarily reflect theirtranscriptional rates, suggesting that cellular metabolism is influenced by the stability ofindividual mRNAs and by the ability of the cells to regulate mRNA turnover. In fact, the rapidturnover of an mRNA ensures that it is maintained at relatively low steady-state levels and thatchanges in the rate of degradation can affect its steady-state level over a short period of time.This type of regulation allows transient alterations in the expression of some cytokines,transcription factors, and proto-oncogenes such as c-fos, c-myc, and c-myb in response togrowth factors, phorbol esters, antigen stimulation, or inflammation (for review, see Sachs,1993). This class of short-lived mRNAs all share a common AUUUA sequence motif in their3’UTRs, which serves as one signal targeting the mRNAs for rapid turnover. The 3UTRs ofboth human and murine ICAM-1 mRNAs also contain several AUUUA sequences. However,its role in the regulation of ICAM- 1 mRNA turnover has not been determined.The present investigation was intended to define the intracellular mechanismsunderlying the inducible expression of ICAM-l by two inflammatory mediators, IFN-’y andPMA. Our results demonstrate that WN-’y as well as PMA upregulate ICAM- 1 gene expressionat least in part at a posttranscriptional level by stabilizing labile ICAM- 1 mRNA. Moreover,the stabilization of ICAM-1 mRNA by PMA requires the 3’UTR and may be mediated by theAUTJUA destabilizing sequences, whereas the effect of IFN-y on ICAM- 1 mRNA stabilityappears to involve unknown sequences within the protein coding region.402.3 MATERIALS AND METHODSCell culture. The murine monocytic cell line P388D1 (Koren et al., 1975) was cultured inprotein-free hybridoma media II (GIBCO, Grand Island, NY). The mouse fibroblast L cell linewas maintained in DMEM containing 10% FCS and antibiotics.Cloning of murine ICAIvI-2 cDNA. Total RNA was isolated from EL-4 cell line andreverse transcribed with M-MLV reverse transcriptase using oligo dT primer. Using thefollowing degenerate oligonucleotides, 5’-GTCAACTG(C/T)AG(C/T)(AJT)CC(T/A)C(AIC)TG-3’ and 5’-CACAG(T/C)(G/C)(C/A)G(G/A)CAGGAGAA(AG)TI’-Y, the cDNA wassubjected to 30 rounds of polymerase chain reaction (PCR) amplification (94°C, 30 sec; 50°C,1 mm; 72°C, 1 mm) with Taq polymerase. The resulting 450 bp DNA fragments weresubcloned into pBST and sequenced. Approximately 10% of the clones had homology to thehuman ICAM-2 cDNA sequence (Staunton et aL, 1989). One of these clones was used toscreen a ZAPII mouse lung cDNA library (Stratagene, La Jolla, CA) with a stringency washof 0.5x SSC/0.1% SDS at 55°C. G3-1.1 clone with an identical sequence to that of murineICAM-2 cDNA (Xu et al., 1992) was obtained.Expression vectors. Expression vectors were constructed using standard recombinantDNA techniques (Sambrook et al., 1989). Briefly, K4-1. 1 ICAM-1 cDNA (Horley et al.,1989) was partially digested with PstI to remove its 3 ‘UTR, generating the truncated ICAM- 1 z3 cDNA. The chimeric ICAM-Fas cDNA was constructed by generating the transmembraneand cytoplasmic domains of Fas cDNA (Watanabe-Fukunaga et al., 1992) by PCR using5’-AAGGGATCCAAAGCTGGGTA-3’ and 5’-AGTCCCGGGAAATCGCCTAT-3’oligonucleotides with the Smal and BamHI incorporated restriction sites, respectively. ThePCR product was then ligated between the BamHI and Smal sites of pBST(sk) generating thepBST(sk)-Fas plasmid. The EcoRl-Scal fragment containing the 5 extracellular domains ofICAM-1 cDNA was ligated between the EcoRl and Smal sites of pBST(sk)-Fas, creating theICAM-Fas chimeric cDNA. The ligated junction between the two cDNAs was sequenced to41ensure that it was in frame. The chimeric ICAM-2/1 eDNA was constructed by generating the3tUTR of ICAM-1 cDNA by PCR using 5’-CCATCGATGCCTGCTGGATGAGACTCCTGC-3’ and 5’-CCATCGATAGACTCTCACAGCATCTGCAGC-3’ oligonucleotides.The PCR product was then ligated into the unique BstXI site in the 3’UTR of ICAM-2. TheSall-Noti isolated ICAM-Fas chimeric, full-length ICAM-1, ICAM-1A3, and CD18 cDNAs(Wilson et al., 1989) were blunt-ligated into a stable mammalian expression vector, pRC6(derived from pAXi 14; Kay and Humphries, 1991) with a hygromycin resistance gene drivenby a thymidine kinase promoter). ICAM-2 and ICAM-2/1 cDNAs were isolated frompUC18/13 with Sail and ligated into the Sail site in pRC6. Escherichia coli MC1O61/p3(Yamasaki et aL, 1988) was used for transformation. All the constructs used in this study areschematically shown in Fig. 1.Transfection and analyses of fibroblast L cells. ICAM-1, ICAM-1A3, ICAM-Fas,ICAM-2, ICAM-2/ 1, and CD 18 cDNAs in expression vectors were transfected into ICAM- 1, -2, and CD18 negative L cells by the calcium phosphate method (Sambrook et aL, 1989).Transfectants were isolated under hygromycin selection (250 Ig/ml; Calbiochem, La Jolla,CA). Bulk populations (mixture of individual clones allowed to grow to confluency) as well asclones of transfectants were isolated and characterized. The surface expression of ICAM- 1 and-2 was tested by flow cytometry. Briefly, ICAM-1, ICAM-1A3, and ICAM-Fas transfected Lcells were harvested with PBS containing 2.5 mM EDTA and then stained with YN1I1.7 antiICAM-1 hybridoma supernatant (Horley et al., 1989) containing 0.1% sodium azide and goatanti-rat Ig-FITC as a secondary antibody (Cooper Biomedical, West Chester, PA). ICAM-2and ICAM-2/1 transfected L cells were stained with rat anti-mouse ICAM-2 mAb(PharMingen, San Diego, CA) and goat anti-rat Ig-FITC. The copy numbers of the transfectedICAM-1 and ICAM-1A3 genes were determined by genomic Southern blot analysis (Sambrooket al., 1989) using BamHl digest and scanning densitometry of the autoradiogram. CD18transfected L cell clones were detected by a standard Northern analysis (Sambrook et al.,1989).42e. ICAM-2/1AAAAICAM-1 3’UTRFig. 1 Expression vectors, a. Schematic diagram of the full-length ICAM- 1 expressionvector. Open triangle represents CMV promoter/enhancer; Solid triangles, AUUUApentamers; EXD, extracellular domain; TMD, transmembrane domain; CYD, cytoplasmicdomain. b. Schematic diagram of the truncated ICAM-1A3 expression vector. c. Schematicdiagram of ICAM- 1/Fas (ICFAS) chimeric expression vector. Dotted and wavey rectanglesrepresent Fas transmembrane and cytoplasmic domains, respectively. d. Schematic diagramof ICAM-2 expression vector. Dashed rectangles represent ICAM-2 cDNA. e. Schematicdiagram of ICAM-21 1 chimeric expression vector.CMVpro/enha. ICAM-1b. ICAM-1z3C. ICFASEXDI I-FFAS1Id. ICAM-2A A43Genomic DNA isolation and Southern blot analysis. In order to isolate genomic DNA(Gross-Bellard et al., 1977), cells were washed 3x in PBS. Approximately 5 x iO cells weresuspended in 2 ml of TNE buffer (10 mM Tris pH 8.0, 150 mM NaC1, 10 mM EDTA) andgently mixed. To this suspension, 20 111 of 20% SDS and 10 tl proteinase K (stock 10 mg/ml;Sigma, St. Louis, MO) were added and incubated at 37°C overnight. The lysate was thenextracted 3x with TNE-saturated-phenol, 3x with TNE-saturated-phenol:chloroform (1:1), andfinally 3x with chloroform:isoamyl alcohol (24:1). The aqueous phase was dialyzed againsttwo changes of 4L of TE buffer (10 mM Tris pH 8.0, 1 mM EDTA) each for 24 h at 4°C. TheDNA was quantitated spectrophotometrically (A260 = 1.0 for 50 jig/ml) and purity wasdetermined (A260/8— 2.0) on an LKB Ultrospec 4050 (Cambridge, England).Digested genomic DNA (Sambrook et al., 1989) was loaded onto a 0.8% agarose geland electrophoresed for 16-18 h at 35 volts. Molecular weights were determined frommolecular weight standards such as ? Hind ill fragments or ? Hind Ill/EcoRI fragments. Thegel containing digested DNA was transferred onto a piece of Zeta Probe (Bio-RadLaboratories, Mississauga, ON). DNA was fixed on the membrane by UV cross-linking in aUV Stratalinker 1800 (Stratagene, La Jolla, CA). The blots were prehybridized and thenhybridized with 1 x 106 cpmlml of {x32P] ICAM-1 cDNA labeled by random priming(Sambrook et al., 1989). The blots were exposed to Kodak XAR-5 film at -70°C using anintensifying screen.ICAM-1 induction by IFN-’y, PMA, and cycloheximide. 1.5 x 106 P388D1 and thetransfected L cells were treated with or without murine WN-y (100 u/mi; BoehringerMannheim, Laval, PQ) or PMA (10 ng/ml; Sigma, St. Louis, MO) or cycloheximide (10 jig!ml; Sigma, St. Louis, MO) and incubated at 37°C for increasing periods of time. The levei ofmRNA was then determined by Northern blot analysis as described below.mRNA half-life determination. 1.5 x 106 P388D1 and the transfected L cells weretreated with actinomycin D (10 ji.g/ml; GIBCO, Grand Island, NY) for increasing periods oftime, and the level of mRNA was determined by Northern blot analysis as described below.44RNA preparation and Northern blot analysis. Total RNA was prepared as described byChomczynski and Sacchi (Chomczynski and Sacchi, 1987) with minor modifications. Briefly,1.5 x 106 cells were washed twice with PBS and the pellet was dissolved in 0.5 ml of SolutionD (4 M guanidinium thiocyanate, 25 mM sodium citrate pH 7.0, 0.5% sarcosyl, and 0.1 M 2-mercaptoethanol). 50 jii of 2 M sodium acetate pH 4.0, 0.5 ml water-saturated-phenol, and 0.1ml chloroform: isoamyl alcohol (49:1) were then added to the solution and spun at 1.4 x gfor 15 mm at 4°C. The aqueous layer was mixed with 0.5 ml isopropanol and put at -20°C for1 h. The solution was then centrifuged for 20 mm at 4°C and the pellet was suspended in 300t1 Solution D followed by 300 pi isopropanol. After 1 h at -20°C, the solution was centrifugedfor 20 mm at 4°C. The pellet was washed with 75% ethanol and then dissolved in 10 111distilled water.Equalized aliquots of total RNA (approximately 10 tg) were subjected toelectrophoresis in 1% formaldehyde-agarose gels (Sambrook et al., 1989) and transferred ontonylon membranes (Bio-Rad Laboratories, Mississauga, ON). RNA was fixed on themembrane by UV cross-linking in a UV Stratalinker 1800 (Stratagene, La Jolla, CA). Theblots were prehybridized and then hybridized with 1 x 106 cpmlml of [&2P] ICAM-1 cDNAlabeled by random priming (Sambrook et al., 1989). The blots were exposed to Kodak XAR-5film at -70°C using an intensifying screen.For the induction analysis, the bands on the autoradiogram were integrated with adensitometer by using suitable exposures, and then the values were normalized to the signal ofthe control 13-actin probe or to the constitutively expressed rRNA gene. The maximumintensity was then arbitrarily set at 100% mRNA level as a point of reference. For the mRNAhalf-life analysis, the integrated band values were normalized to the constitutively expressedrRNA gene, and then the 100% mRNA level was set at time 0.452.3 RESULTSEffects of IFN-’y and PMA on ICAM-1 mRNA stability. The treatment of the murine monocyticcell line P3 88D 1 with IFN-y or PMA induced a rapid increase in the level of ICAM- 1 mRNA(Fig. 2a). The maximum increase (5-8x basal level) was achieved within 2 h after the additionof IFN--y and around 6 h after the addition of PMA. Because both human and murine ICAM- 1mRNAs contain multiple AUUUA destabilizing sequences, we examined the effects of 1FN-’yand PMA on the half-life of ICAM-1 mRNA. P388D1 cells were treated with thetranscription inhibitor actinomycin D and the decrease in the level of ICAM- 1 mRNA wasanalyzed. Northern blot analysis showed that IFN-y and PMA treatments clearly prolonged thehalf-life of ICAM- 1 niRNA, extending it from approximately 50 mm to far greater than 2 h(Fig. 2b), suggesting that the upregulation of ICAM-l gene expression by IFN-’y and PMA wasdue at least in part to the stabilization of ICAM- 1 mRNA. As reported for many other labilemRNAs (Sachs, 1993), treatment with cycloheximide, a potent inhibitor of translation, induceda rapid accumulation of ICAM-1 mRNA (Fig. 2c).Role of AUUUA multirners on induction of ICAM-] mRNA. To further investigate therole of the AUUUA repeats in the upregulation of ICAM-1 mRNA levels, the murine ICAM-1cDNA clone K4-1.1 was modified to remove its 3’UTR, subcloned into the mammalianexpression vector pRC6 (see Materials and Methods; Fig. 1) and then transfected into murinefibroblast L cells, which do not express detectable levels of ICAM- 1 mRNA regardless of IFNy treatment. The expression of the transfected cDNA was driven by the CMV promoter andenhancer, which were not responsive to IFN-y. As a control, the non-inducible murine CD18cDNA was also transfected into L cells using the same expression vector. Bulk populations oftransfected L cells were established by hygromycin resistance. In addition, three randomlypicked clones from each type of transfectant expressing the intact or truncated (ICAM- 1A3)cDNA were isolated and the expression of ICAM- 1 on the cell surface was confirmed by flowcytometry. Northern blot analysis of the bulk transfected L cells showed that the basal level of46aIFN-yI0246PMA0246hICAM-1ActinaFig. 2 Effects of IFN-’y and PMA on ICAM-l mRNA expression in P388D1 cells. Detectedby Northern blot analysis. a. The induction of ICAM-l mRNA level in P388D1 cells that hadbeen stimulated with IFN-’y and PMA for the indicated times. The bottom panel represents thecorresponding signals of the control 13-actin. b. Half-life analysis of ICAM-l mRNA. Cellswere treated with IFN-’y for 1 h, PMA for 2 h, or no treatment (—) and then actinomycin D(ActD) was added for the indicated times. c. The induction of ICAM-1 mRNA level inP388D1 cells that had been treated with (+) or without (—) cycloheximide (CHX) for 6 h. Thebottom panel represents the corresponding signals of the control 3-actin. The data presentedare representative of two independent experiments.47C.) -‘ I10_i.14;1I.I1+ Iz+J..0++CJiIo IC)C- + CHXIICAM-1Actin gj1II49ICAM- 1z3 mRNA was consistently higher (approximately 4-fold) than that of the intactICAM- 1 mRNA (Fig. 3). Concordantly, the three clones from each type of transfectant gavesimilar results. The copy numbers of the transfected genes, which may influence the inRNAexpression level, in those two groups of transfected L cells, whether bulk or individual clones,determined by genomic Southern blot analysis, were similar to each other (Fig. 4). Therefore,the 3’UTR seemed to regulate the basal level of ICAM-1 mRNA in the transfected cells.We also examined whether 1FN-y treatment upregulates the level of ICAM-1 mRNA inthe transfected L cells. Northern blot analysis showed that the level of ICAM- 1 mRNA waselevated by WN-’ treatment in both types of bulk transfected L cells (Fig. 3). The degree ofupregulation of ICAM- 1 mRNA levels appeared to be higher in the intact cDNA-transfected Lcells (6-9x basal level) than that in L cells with ICAM-ltX3 cDNA (approximately 3x basallevel). This difference however was primarily due to different basal levels between the twoICAM- 1 mRNA species, as stated above. L cells transfected with CD 18 cDNA using the samevector showed no change in the level of CD 18 mRNA upon treatment with IFN-y, indicatingthat it has no effect on the transcription of the transfected genes driven by the CMVpromoter/enhancer. The f3-actin mRNA content was also unaffected by IFN-y, furtherdemonstrating the specificity of IFN-y. These observations were confirmed using threerandomly picked clones from each type of transfectant.Effect of IFN-’y on ICAM-1 and ICAM-liX3 mRNA stability. To determine whether theobserved upregulation of ICAM- 1 mRNA was due to its enhanced stability, the effect of IFN-yon the half-life of ICAM- 1 mRNA was examined. Without IFN-’y treatment, full-lengthICAM-1 mRNA from the bulk transfectants had a short half-life of approximately 50 mm (Fig.5), whereas, the half-life of ICAM- 1z3 mRNA was significantly longer (approximately 160mm). The treatment of the bulk transfected L cells with IFN-y dramatically extended thestability of both full-length and truncated mRNAs (Fig. 5). Concordantly, the mean half-livesof full-length ICAM-1 and ICAM-1A3 mRNAs, in the absence of WN-y, determined fromthree independent clones of each transfectant were 42 ± 10 mm and 109 ± 17 mm,50ICAM-1ActinICAM-1 z\301 248 h- -+ -+ -+-+ IFN-yActin..IeeepCD1 8ActinFig. 3 Effects of IFN-’y on induction of ICAM-1, ICAM-1L\3, and CD18 mRNA levels.Detected by Northern blot analysis. The induction of ICAM-1, ICAM-1A3, and CD18 mRNAlevels in the transfected L cells stimulated with (+) or without (—) IFN-’y for the indicatedtimes. The second, fourth, and sixth panels from the top represent the corresponding signals ofthe control 3-actin. The data presented are representative of two independent experimentsperformed on bulk transfectants. The three independent clones from each type of transfectantgave similar results.5112endogenous— ICAM-1. transfectedtransfected ICAM-1ICAM-1A3Fig. 4 Southern blot analysis of the exogenous ICAM- 1 copy number. Representative BamHIgenomic Southern blot analysis of ICAM-1A3 (lane 1) and ICAM-1 (lane 2) bulk transfected Lcells. The genomic DNA from the transfected L cells was digested with BamHI, which cutsthe vector immediately upstream and downstream of the inserted ICAM-l and ICAM-1z3gene, and probed with the entire ICAM- 1 cDNA. The bulk transfected L cells and the threeindependent clones of each construct gave bands of similar intensity.52ICAM-1- ++I I012ICAM-1z3- ++ +I II 1012012Fig. 5 Half-life analysis of ICAM-l and ICAM-1A3 mRNAs. Detected by Northern blotanalysis. The ICAM-1 and ICAM-1A3 transfected L cells were either treated with (+) orwithout (—) 1FN-’y for 2 h and then actinomycin D (ActD) was added for the indicated times.The data presented are representative of two independent experiments performed on bulktransfectants. The three independent clones from each type of transfectant gave similar results.+012IFN-yActDh53respectively. Hence, ICAM- 1 z3 mRNA was indeed more stable than the full-length ICAM- 1mRNA. Moreover, like with the bulk transfectants, WN-y treatment further extended theturnover rates of both transcripts to far greater than 2 h. These results indicate that IFN-ytreatment enhances the level of ICAIVI- 1 mRNA by making it more stable, and this effectoccurs even in the absence of the conserved AUUUA pentanucleotide sequences in the 3’UTR,suggesting the presence of a novel 1FN-y-responsive element within the protein codingsequence. This effect of WN-y has not been recognized previously.3’UTR of ICAM-1 mRNA and IFN-y effects. ICAM-2 is another ligand for LFA-1 butunlike ICAM- 1, its expression is constitutive regardless of pro-inflammatory cytokinetreatment (Staunton et al., 1989; de Fougerolles et al., 1991). Interestingly, ICAM-2 does nothave any AUUUA sequences in its 3’UTR. The murine ICAIVI-2 cDNA clone G3- 1.1 wasisolated and a chimeric ICAM-2/1 cDNA (ICAM-1 3’UTR incorporated into ICAM-2 3’UTR)was generated (see Materials and Methods; Fig. 1) to further discern the possible involvementof ICAM-1 3’UTR in the induction and stabilization of ICAM-1 mRNA by IFN-y. Northernblot analysis of L cells transfected with ICAM-2 or ICAM-2/1 cDNA showed that IFN-y didnot induce the upregulation of these mRNAs (Fig. 6a). Furthermore, the half-life analysisindicated that the ICAM-2 mRNA was rather stable whereas ICAM-2/1 mRNA was relativelyunstable (Fig. 6b). Also, IFN-y had no effect on the stability of either mRNA (Fig. 6b). Thus,these results support the hypothesis that the 3’UTR of ICAM- 1 contains mRNA destabilizingelements but by itself is not responsible for the effects of WN-y.Role ofAUUUA sequences in PMA-induced accumulation and stabilization of ICAM-]mRNA. The above results demonstrated that PMA treatment of P388D 1 cells upregulates theirICAM- 1 expression by a posttranscriptional mechanism that enhances the stability of ICAM- 1mRNA. To determine whether the ICAM-l 3’UTR containing multiple AUUUA sequences isinvolved in these effects of PMA, we examined the effect of PMA treatment on theaccumulation of ICAM-1 and ICAM-1A3 mRNA. Northern blot analysis of both bulk andclonal transfectants showed that PMA upregulated the level of full-length ICAIVI- 1 mRNA (5-54aICAM-2Actin0246 hIFN-yICAM-2/1ActinFig. 6 Effects of ]FN-’y on ICAM-2 and ICAM-2/1 mRNA accumulation and stabilization, a.ICAM-2 and ICAM-21l transfected L cells were treated with WN-y for the indicated times.The second and fourth panels from the top represent the corresponding signals of the control f3-actin. b. Both types of transfected L cells were treated with (—) or without (+) IFN-y for 2 hand then actinomycin D (ActD) was added for the indicated times. The data presented arerepresentative of two independent experiments.55b+ IFN-y+ + ActDI 1I I0.5120.512 hICAM-2ICAM-2/1567x basal level) whereas the level of ICAM- lz\3 mRNA was not affected (Fig. 7). Therefore,the 3’UTR of ICAM-l containing multiple AUUUA sequences was necessary for the PMAinduced upregulation of ICAM- 1 mRNA level.To further investigate the role of the AUUUA sequences in the the upregulation ofICAM-1 mRNA, ICAM-Fas (ICFAS) chimeric cDNA was constructed, in which the regions ofICAM-1A3 cDNA encoding the transmembrane and cytoplasmic domains were replaced bythose of Fas cDNA (see Materials and Methods; Fig. 1). As the region of Fas cDNA encodingits transmembrane and cytoplasmic domains contains multiple AUUUA sequences, theresultant ICFAS chimeric eDNA is essentially ICAM-13 cDNA with AUUUA sequencesderived from a heterologous gene fragment. The PMA treatment of L cells transfected with thechimeric cDNA increased the level of ICFAS mRNA (6-fold maximum induction) whereasIFN-y had no effect (Fig. 8). The half-life analysis showed that PMA stabilized the full-lengthICAM-1 mRNA from approximately 50 mm to far greater than 2 h (Fig. 9), suggesting that theaccumulation of ICAM-l mRNA by PMA was due to its stabilization. Moreover, the ICFASmRNA had a half-life of approximately 60 mm, which is considerably shorter than the ICAM1A3 mRNA, and PMA treatment greatly enhanced its stability (Fig. 9). These results furthersuggest that AUUUA sequences may play a role in the general stability of ICAM- 1 mRNA,and are sufficient for the upregulation and stabilization of ICAM- 1 mRNA induced by PMAwhereas the effect of IFN-y requires the presence of other undefined sequences possibly in theregions encoding the transmembrane and cytoplasmic domains of ICAM- 1.5712-+ -+ PMA(6h)Actin—— -Fig. 7 Effects of PMA on induction of ICAM-1 and ICAM-1z3 mRNA levels. Detected byNorthern blot analysis. The ICAM- 1 A3 (1) and ICAM- 1 (2) transfected L cells were eithertreated with (+) or without (—) PMA for 6 h. The bottom panel represents the correspondingsignals of the control -actin. The data presented are representative of two independentexperiments performed on bulk transfectants. The three independent clones from each type oftransfectant gave similar results.58IFN-y0246 h- +++Fig. 8 Effects of PMA and IFN-y on induction of ICFAS chimeric mRNA level. Detected byNorthern blot analysis. The ICFAS transfected L cells were treated with either PMA (leftpanel) or WN-y (right panel) for the indicated times. The bottom panels represent thecorresponding signals of the control 13-actin. The data presented are representative of twoindependent experiments performed on bulk transfectants. The three independent clones gavesimilar results.PMA0246- +++ICFAS .4Actin59ICAM-1 ICFAS+ + PMA+ + + + ActDI II II II I012012012012 hFig. 9 Half-life analysis of ICAM- 1 and ICFAS mRNAs. Detected by Northern blot analysis.The ICAM- 1 and ICFAS transfected L cells were either treated with (+) or without (—) PMAfor 2 h and then actinomycin D (ActD) was added for the indicated times. The data presentedare representative of two independent experiments performed on bulk transfectants. The threeindependent clones from each type of transfectant gave similar results.602.5 DISCUSSIONThe proinflammatory cytokine IFN-’y as well as the protein kinase C activator PMAdramatically increase the expression of ICAM- 1 on various cells (Rothlein et al., 1986; Dustinet al., 1986; Pober et al., 1986; Rothlein et al., 1988). We have shown here that both IFN-yand PMA upregulate ICAM- 1 gene expression at least in part by a posttranscriptionalmechanism which stabilizes ICAM- 1 mRNA. As expected from the presence of multipleAUUUA destabilizing sequences in its 3’UTR, ICAM- 1 mRNA is rather unstable, and thetreatment of cells with IFN-y or PMA clearly stabilizes the mRNA. Interestingly, this effect ofIFN-y can be observed in the absence of the 3’UTR containing AUUUA destabilizingsequences although the effect is less pronounced without the 3’UTR. In support, theincorporation of ICAM-1 3’UTR containing multiple AUUUA sequences into ICAM-2 mRNAconverts a previously stable mRNA into a labile transcript. Furthermore, the stability ofICAM-2/1 niRNA remains unaffected by WN-y treatment, and hence is unable to induce theaccumulation of ICAM-2/1 mRNA. In contrast, the PMA effect on ICAM-1 mRNA stabilityseems totally dependent on the presence of the AUUUA destabilizing sequences as thepresence of exogenous AUUUA sequences derived from the Fas gene appears to be sufficientto mimic the effect of PMA seen on ICAM- 1 mRNA stability. These results suggest that atleast two pathways regulate ICAM-1 mRNA stability. One is induced by PMA, which isknown to activate protein kinase C, and requires the presence of AUUUA sequences. Theother, induced by WN-y, involves an unknown sequence(s) within the protein coding region. Itshould be noted that the treatment of cells with cycloheximide also stabilizes and superinducesICAM- 1 mRNA as is the case for many other labile mRNAs (discussed in chapter IV; Ohh andTakei, 1995).IFN-y has been known to upregulate ICAM- 1 expression on various cells and thoughtto play an important role in the enhanced expression of ICAM-1 at inflammatory tissues.Because IFN-’y enhances the transcription of many IFN-y-responsive genes including class II61major histocompatibility complex (Moses et al., 1992), its effect on ICAM- 1 has also beenthought to be due to its effect on the transcription of ICAM-1 gene. Indeed, putative IFN-yreponsive element(s) have been identified in the 5-upstream region of the ICAM- 1 gene(Degitz et al., 1991). However, our current studies have shown that IFN-y has an addtionalrole in the upregulation of ICAM-1 gene expression by a posttranscriptional mechanism. Thiseffect of IFN-y on ICAM- 1 mRNA stability has not been recognized previously. On the otherhand, PMA treatment has been shown to stabilize the mRNA for ICAM- 1 (Wertheimer et al.,1992) as well as the mRNAs for transforming growth factor-13 1, IL-i, IL-3, and c-fins (Wagerand Assoian, 1990; Yomoto et al., 1989; Wodnar-Filipowicz and Moroni, 1990; Weber et al.,1989). Hence, message stabilization by PMA is not unique to ICAM-1. PMA treatment ofmononuclear cells resulted in an increased stability of labile messages, such as c-fos mRNAwhich was associated with enhanced adenosine-uridine binding factor (a factor that specificallybinds to the AUUUA sequence) (Gillis and Malter, 1991). Conversely, however, stabilizationof lymphokine (IL-2, TNF-oc, and GM-CSF) mRNAs induced by costimulation with PMAcorrelated inversely with the binding of a sequence-specific cytoplasmic factor that bindsspecifically to AUUUA multimers present in the 3’UTR of lymphokine mRNAs (Bohjanen etal., 1991). Here we show that the AUUUA motif in the 3’UTR of ICAIVI-1 mRNA plays a rolein mediating PMA-regulated ICAJvI-1 mRNA turnover.Although the relative contribution of mRNA stability, as compared to transcriptionalregulation, to the overall ICAM-1 gene expression is not known at this time, it may,nevertheless, be an important regulatory mechanism involved in inflammatory responses. Ithas been reported that TNF-o upregulates ICAM-1 gene expression at a transcriptional level(Voraberger et al., 1991; Wertheimer et al., 1992) and that IFN-’y upregulates the receptor forTNF-o (Aggarwal et al., 1985; Tsujimoto et al., 1986). This may contribute to the synergisticaction of TNF-o and IFN-y on ICAM-1 mRNA induction (Barker et al., 1990). However, inaddition to this process, the present studies suggest that the marked synergism that occursbetween TNF-o and IFN-y may be due to the increased transcription of ICAM-1 gene by TNF62c and subsequent stabilization of the newly transcribed mRNA by IFN-’y. The regulation ofICAM- 1 by mRNA stability could provide a rapid and efficient way of controlling ICAM- 1expression. The rapid turnover of ICAM- 1 mRNA ensures that it is maintained at low steady-state levels and that changes in the mRNA stability in response to external stimuli can affect itssteady-state level over a relatively short period of time. Such regulation of ICAM- 1 may beimportant for its function. In the absence of an inflammatory response, ICAM- 1 is expressedonly at a low level on vascular endothelium (Dustin et al., 1986). The upregulated ICAM-lfacilitates the adhesion of leukocytes bearing the receptors for ICAM- 1 (namely, LFA- 1 andMac-i) and their subsequent diapedesis into inflammatory tissues (Wawryk et al., 1989; Smithet al., 1989). In comparison, another LFA-1 ligand, ICAM-2, is constitutively expressed onvascular endothelium (Staunton et al., 1989; Xu et al., 1992; de Fougerolles et al., 1991). Thebasal level of ICAM-2 expression on endothelial cells is much higher than that of ICAM- 1,and is not further induced by pro-inflammatory cytokines (Staunton et al., 1989; deFougerolles et al., 1991). Similarly, the recently cloned ICAM-3 appears to be the primaryligand for LFA- 1 on resting lymphocytes, and like ICAIvI-2, it is expressed in a constitutivemanner (Fawcett et aL, 1992; Vazeux et al., 1992). Interestingly, both ICAM-2 and -3mRNAs do not possess AUUUA destabilizing sequences in their 3’UTRs, unlike ICAM-l.Relatively little is known about the posttranscriptional control of mRNA turnover. ThemRNA of the proto-oncogenes c-myc, c-myb, and c-fos and several cytokines including TNF-aand IFN--y are very unstable. The recurrent motif in these unstable mRNA involved in theirrapid degradation is the pentanucleotide sequence AUUUA that is found singly or in multiplereiteration in the 3’UTR (Sachs, 1993). Proteins (adenosine-uridine binding factor) that bind tothese AUUUA sequences have been identified and it has been suggested that these proteinAUUUA complexes may target susceptible mRNAs for rapid cytoplasmic degradation (Sachs,1993). ICAM-l mRNA also has the AUUUA repeats in its 3’UTR which seem to influence itsstability. However, the stabilization of ICAM- 1 mRNA by IFN-y also occurs in the absence ofthe conserved sequences in the 3’UTR, implicating a novel mechanism that regulates the63stability of mRNA. Further characterization of the complex regulatory mechanism involved inICAM-1 expression should enhance our understanding of the inflammatory process and mayprovide new strategies for modulating the course of inflammatory diseases.2.6 REFERENCESAggarwal BB, Eessalu TE, Hass PE (1985) Characterization of receptors for human tumournecrosis factor and their regulation by gamma-interferon. Nature 318:665.Barker JNWN, Sarma V, Mitra RS, Dixit VM, Nickoloff BJ (1990) Marked synergismbetween tumor necrosis factor-x and interferon-y in regulation of keratinocyte-derivedadhesion molecules and chemotactic factors. J Clin Invest 85:605.Bohjanen RP, Petryniak B, June CH, Thompson CB, Lindsten T (1991) An induciblecytoplasmic factor (AU-B) binds selectively to AUUUA multimers in the 3’ untranslatedregion of lymphokine mRNA. Mol Cell Biol 11:3288.Chomczynski P, Sacchi N (1987) Single-step method of RNA isolation by acid guanidiniumthiocyanate-phenol-chloroform extraction. Anal Biochem 162:156.de Fougerolles AR, Stacker SA, Schwarting R, Springer TA (1991) Characterization of ICAIVI2 and evidence for a third counter-receptor for LFA-1. J Exp Med 174:253.Degitz K, Lian-Jie L, Caughman SW (1991) Cloning and characterization of the 5’-transcriptional regulatory region of the human intercellular adhesion molecule 1 gene. J BiolChem 266:14024.Dougherty GJ, Murdoch 5, Hogg N (1988) The function of human intercellular adhesionmolecule-i (ICAM- 1) in the generation of an immune response. Eur J Immunol 18:35.Dustin ML, Rothlein R, Bhan AK, Dinarello CA, Springer TA: A natural adherence molecule(ICAM-i) (1986) induction by IL-l and interferon-y, tissue distribution, biochemistry andfunction. J Immunol 137:245.Fawcett J, Holness CLL, Needham LA, Turley H, Gatter KC, Mason DY, Simmons DL (1992)Molecular cloning of ICAM-3, a third ligand for LFA- 1, constitutively expressed on restingleukocytes. Nature 360:48 1.Gillis P, Malter JS (1991) The adenosine-uridine binding factor recognizes the AU-richelements of cytokine, lymphokine, and oncogene mRNAs. J Biol Chem 266:3172.64Gross-Bellard M, Oudet P, Chambon P (1977) Isolation of high molecular weight DNA frommammatian cells. Eur J Biochem 36:32.Honey KJ, Carpenito C, Baker B, Takei F (1989) Molecular cloning of munine intercellularadhesion molecule (ICAM-1). EMBO J 8:2889.Kay R, Humphries RK (1991) New vectors and procedures for isolating cDNAs encoding cellsurface proteins by expression cloning in COS cells. Methods Mol Cell Biol 2:254.Koren HS, Handwerger BS, Wunderlich JR (1975) Identification of macrophage likecharacteristics in a cultured murine tumor line. J Immunol 114:894.Moses H, Panek RB, Benveniste EN, Ting JP-Y (1992) Usage of primary cells to delineateIFN-’y-responsive DNA elements in the HLA-DRA promoter and to identify a novel IFN-’yenhanced nuclear factor. J Immunol 148:3643.Ohh M, Takei F (1995) Regulation of ICAM-i mRNA stability by cycloheximide: role ofserine/threonine phosphorylation and protein sythesis. J Cell Biochem (In press).Pober JS, Gimbrone MA Jr, Lapierre LA, Mendrick DL, Fiers W, Rothlein R, Springer TA(1986) Overlapping patterns of activation of human endothelial cells by interleukin 1, tumornecrosis factor, and immune interferon. J Immunol 137:1893.Rosenstein Y, Park JK, Hahn WC, Rosen FS, Bierer BE, Burakoff SJ (1991) CD43, amolecule defective in Wiskott-Aldrich syndrome, binds ICAM-1. Nature 354:233.Rothlein R, Czajkowski M, O’Neill MM, Marlin SD, Mainolfi E, Merluzzi VJ (1988)Induction of intercellular adhesion molecule-i on primary and continuous cell lines by pro-inflammatory cytokines. I Immunol 141:1665.Rothlein R, Dustin ML, Marlin SD, Springer TA (1986) A human intercellular adhesionmolecule (ICAM-1) distinct from LFA-1. J Immunol 137:1270.Sachs AB (1993) Messenger RNA degradation in Eukaryotes. Cell 74:413.Sambrook J, Fritsch EF, Maniatis T (1989) Molecular cloning: A laboratory manual, 2nd Ed.Cold Spring Harbor Laboratory, Cold Spring Harbor, NY.Smith CW, Marlin SD, Rothlein R, Toman C, Anderson DC (1989) Cooperative interactionsof LFA- 1 and Mac-i with intercellular adhesion molecule-i in facilitating adherence andtransendothelial migration of human neutrophils in vitro. J Clin Invest 83:2008.Springer TA (1990) Adhesion receptors of the immune system. Nature 346:425.Staunton DE, Dustin ML, Springer TA (1989) Functional cloning of ICAM-2, a cell adhesionligand for LFA-l homologous to ICAM-l. Nature 339:61.65Tsujimoto M, Yip YK, Vilcek J (1986) Interferon-gamma enhances expression of cellularreceptors for tumor necrosis factor. J Immunol 136:2441.Vazeux R, Hoffman PA, Tomita JK, Dickinson ES, Jasman RL, St. John T, Gallatin MW(1992) Cloning and characterization of a new intercellular adhesion molecule ICAM-R. Nature360:485.Voraberger G, Schafer R, Stratowa C (1991) Cloning of the human gene for intercellularadhesion molecule 1 and analysis of its 5-regulatory region: induction by cytokines andphorbol ester. J Immunol 147:2777.Wager RE, Assoian RK (1990): A phorbol ester-regulated ribonuclease system controllingtransforming growth factor f3 1 gene expression in hematopoietic cells. Mo! Cell Biol 10:5983.Watanabe-Fukunaga R, Brannan CI, Itoh N, Yonehara 5, Copeland NG, Jenkins NA, Nagata S(1992) The cDNA structure, expression, and chromosomal assignment of the mouse Fasantigen. J Immunol 148:1274.Wawryk SO, Novotny JR, Wicks IP, Wilkinson D, Maher D, Salvaris E, Welch K, Fecondo J,Boyd AW (1989) The role of the LFA-1/ICAM-l interaction in human leukocyte homing andadhesion. Immunol Rev 108:135.Weber B, Horiguchi J, Luebbers R, Sherman M, Kufe D (1989) Post-transcriptionalstabilization of c-fins mRNA by a labile protein during human monocytic differentiation. MolCell Biol 9:769.Wertheimer SJ, Myers CL, Wallace RW, Parks TP (1992) Intercellular adhesion molecule-igene expression in human endothelial cells: differential regulation by tumor necrosis factor-ctand phorbol myristate acetate. J Biol Chem 267:12030.Wilson RW, O’Brien WE, Beaudet AL (1989) Nucleotide sequence of the cDNA from themouse leukocyte adhesion protein CD 18. Nucleic Acids Res 17:5397.Wodnar-Filipowicz A, Moroni C (1990) Regulation of interleukin 3 mRNA expression in mastcells occurs at the posttranscriptional level and is mediated by calcium ions. Proc Natl AcadSci USA 87:777.Xu H, Tong IL, de Fougerolles AR, Springer TA (1992) Isolation, characterization, andexpression of mouse ICAM-2 complementary and genomic DNA. J Immunol 149:2650.Yamasaki K, Hiratia Y, Ywata H, Kawanishi Y, Seed B, Tanaigush T, Hirano T, Kishimoto T(1988) Cloning and expression of the human interleukin-6 (BSF-2/IFNbeta2) receptor. Science241:825.Yomoto K, El-Hajjaoui Z, Koeffler HP (1989) Regulation of levels of IL-i mRNA in humanfibroblasts. J Cell Physiol 139:610.66IIIIDENTIFICATION OF IFN-’y.. AND PMA-RESPONSIVE ELEMENTSINVOLVED IN ICAM-1 mRNA STABILIZATION3.1 ABSTRACTTreatment of cells with IFN-y or PMA induces upregulation of the level of ICAM- 1 mRNA bystabilization of an otherwise labile mRNA. Here, various deletion mutants of ICAM- 1 weregenerated and stably transfected into the murine fibroblast Ltk cells that express noendogenous ICAM- 1 or -2, in an effort to define the regions within ICAM- 1 mRNA that areresponsive to WN-y or PMA. Induction of ICAM- 1 mRNA in the transfected L cells bytreatment with WN-y revealed that the truncation of the region of ICAM- 1 mRNA encodingthe cytoplasmic domain made it non-responsive to IFN-’y whereas all other regions weredispensable. In contrast, PMA-induced accumulation of ICAM-1 mRNA required the 3UTR.To further elucidate the role of these regions in mRNA destabilization and responsiveness toTFN-’y and PMA, ICAM-2 mRNA that is stable and not responsive to WN-y or PMA was usedas a reporter gene. The putative IFN-’y-responsive region of ICAM- 1 mRNA encoding itscytoplasmic domain rendered it unstable and responsive to WN-’y but not PMA. Conversely,the 3’UTR of ICAM-l fused with ICAM-2 mRNA also made it unstable and responsive toPMA but not WN-’y. Half-life analysis showed that the induction of these chimeric mRNAs byIFN-’y and PMA was due, at least in part, to the prolongation of their turnover rate. Theseresults taken together demonstrate that two distinct regions of ICAM- 1 mRNA regulate itsstability, one encoding the cytoplasmic domain and responsive to IFN-y, and the other in the3’UTR and responsive to PMA.673.2 INTRODUCTIONICAM-l is a member of the Ig supergene family and mediates cell adhesion by binding to theleukocyte integrins LFA-l (Rothlein et al., 1986; Marlin and Springer, 1987; Makgoba et al.,1989) and Mac-i (Diamond et al., 1991). ICAM-1 is expressed at very low levels on manycell types but its expression is greatly enhanced by various inflammatory cytokines (Dustin etal., 1986; Pober et al., 1986), including 1FN-y, TNF-x, and IL-i, as well as by phorbol esters(Dustin et al., 1986; Rothlein et a!., 1988). The enhanced ICAM-1 expression on venuleendothelium is thought to enable the adhesion and trans-endothelial migration of leukocytesout of the bloodstream and into the sites of inflammation (Wawryk et a!., 1989; Smith et al.,1989). Anti-ICAM-1 antibodies have been demonstrated to inhibit inflammatory responses ina variety of animal models (Barton et al., 1989; Isobe et al., 1992; Cosimi et a!., 1990; Wegneret a!., 1990). Furthermore, ICAM-1-deficient mice show reduced inflammatory responses(Sligh et aL, 1993). Therefore, upregulation of ICAM-1 expression by inflammatory mediatorsis thought to be an important step in inflammatory responses. However, the intracellularregulatory mechanisms responsible for the inducible expression of ICAM- 1 have not beenclearly elucidated.We have previously reported that IFN-y and PMA stabilize otherwise labile ICAM-imRNA and upregulate its expression (Ohh et a!., 1994; chapter II) This adds to the growingevidence that mRNA turnover plays an important role in regulating gene expression. This typeof regulation allows transient alterations in the expression of some cytokines, transcriptionfactors, and proto-oncogenes such as c-fos, c-myc, and c-myb in response to growth factors,phorbol esters, antigen stimulation, or inflammation (Sachs, 1993; Jackson, 1993). This classof short-lived mRNAs all share a common AUUUA sequence motif in their 3’UTRs, whichserves as one signal targeting the mRNAs for rapid turnover (Sachs, 1993; Jackson, 1993).The YUTRs of both human and murine ICAM-1 mRNAs also contain several AUUUAsequences and seem to play a destabilizing role (Ohh et a!., 1994).68The present study was intended to further define the regions of ICAM- 1 mRNAresponsive to JFN-’y or PMA. Our results demonstrate that the IFN-’y responsive elementresides within the region of 87 nt encoding the cytoplasmic domain of ICAM- 1. In contrast,PMA-induced accumulation and stabilization of ICAM-1 mRNA is dependent on the 3’UTR.3.3 MATERIALS AN]) METHODSCell culture. The murine fibroblast Ltk— cell line was maintained in DMEM (GIBCO, GrandIsland, NY) containing 10% FCS and antibiotics.Expression vectors. Expression vectors were constructed using standard recombinantDNA techniques (Sambrook et at., 1989). Briefly, K4 ICAM-1 cDNA (Horley et at., 1989)was partially digested with PstI to remove its 3’UTR, generating the truncated K4A3 cDNA.K4Ac3 cDNA was generated by first digesting the K4 cDNA with Hindill, giving the leader,domains 1 to 3, and most of domain 4 (ending at bp 1139). The rest of domain 4 to the end ofthe transmembrane domain was generated by polymerase chain reaction (PCR) amplificationusing 5’-TTCAGCTCCGGTCCTGACCC-3’ and 5’-TCTAGATCTGGCGGTTATAAACAT-3oligonucleotides. The two cDNA fragments were ligated together at the Hindlil site.Generation of sK4 (soluble ICAM- 1) cDNA has been described (Welder et at., 1993). K4ld3-5 cDNA was generated by EcoNI digestion of sK4 cDNA, which removed domains 1 and 2. Alinker (5’-CCCAGGGTC-3’ and 5’-GGTCCCAGA-3’) ligated the leader to the beginning ofdomain 3 in frame. K41d1-2 cDNA was generated by a partial digestion of K4 cDNA withBglI. To the 3’ end of this cDNA, a linker (5’-TAGATAGT-3’ and 5’-GTTATCTATCAGATC-3’) was added. K41d1 cDNA was generated by a BglI digestion. To the 3’ end of this cDNA, alinker (5’-TAGATAGT-3’ and 5’-CGGATCTATCAGATC-3’) was added. The chimeric G3-K43u and G3-K4c cDNAs were constructed by generating the 3’UTR and the cytoplasmicdomain of ICAM-1 cDNA with Clal ends by PCR using 5’-69CCATCGATGCCTGCTGGATGAGACTCCTGC-3’, 5’-CCATCGATAGACTCTCACAGCATCTGCAGC-3’ and 5’-CCATCGATTGGACTATAATCATTCTGGTGC-3’, 5’-CCATCGATGAGGTGGGGCTGTTCCCTTGAG-3’ oligonucleotides, respectively. The PCR productswere then ligated into the unique Clal site, which was converted from BstXI site in the 3’UTRof ICAM-2 (Ohh et al., 1994) using a Cial adaptor (5’-AGTATCGATTGAGCCC-3’ and 5’-TCAATCGATACTGGGC-3’). All the ligated junctions between any two cDNAs and the 3’ends of the generated cDNAs were sequenced to ensure that they were in frame and the 3’ endshad incorporated stop codons. The K4, K4A3, K4Ac3, K41d3-5, K41d1-2, and K41d1 cDNAswere blunt-ligated into the Smal site of a stable mammalian expression vector, pRC6 (derivedfrom pAXll4 (Kay and Humphries, 1991) with a hygromycin resistance gene driven by athymidine kinase promoter). G3, G3-K43u, and G3-K4c cDNAs were isolated from pUC18/13with Sail and ligated into the Sail site in pRC6. Escherichia coli MC1O61/p3 (Yamasaki etal., 1988) was used for transformation. All constructs used in this study are schematicallyshown in Fig. 10.Transfection and isolation offibroblast L cells. All cDNAs in expression vectors weretransfected into ICAM-1 and -2 negative L cells by the calcium phosphate method (Sambrooket al., 1989). Transfectants were isolated under hygromycin selection (250 pg/ml,Calbiochem, La Jolla, CA). The surface expression of ICAM-1 and -2 was tested by flowcytometry. Briefly, K4, K4A3, and K4iXc3 transfected L cells were harvested with PBScontaining 2.5 mM EDTA and then stained with YN1/1.7 anti-ICAM-1 hybridoma supernatant(Horley et ai., 1989) containing 0.1% sodium azide and goat anti-rat Ig-FITC as a secondaryantibody (Cooper Biomedical, West Chester, PA). G3, G3-K4c, and G3-K43u transfected Lcells were stained with rat anti-mouse ICAM-2 mAb (PharMingen, San Diego, CA) and goatanti-rat Ig-FITC. L cells transfected with cDNA constructs incapable of surface expression(ie., sK4, K4ld3-5, K41d1-2, and K4ldl cDNAs) were detected by a standard Northern analysis(0Kb et al., 1994).70CMV pro/enh exdleader 1 2 3 4 5 tmd cyd 3UTRa. K4 I I I Ib. K43 I I I IC. K4Ac3 I I I Id.sK4 I I I I Ie. K41d3-5 ‘1jJ—f17]f. K41d1-2 I Ig. K41d1h. G3________________leader 1 2 tmdcyd 3UTRLexdlI. G3—K4c /‘%%\K4 cydi. G3-K43u__ __ __ __ _________AAAK4 3’UTRFig. 10 Schematic diagram of the expression vectors, a. K4; full-length ICAM- 1 expressionvector. b. K4A3; ICAM-l without the YUTR. c. K4zc3; ICAM-l without the cytoplasmicdomain and the 3’UTR. d. sK4; soluble ICAM-1 without the 3’UTR, cytoplasmic andtransmembrane domains. e. K41d3-5; soluble ICAM-l without domains 1 and 2. f. K41d1-2;soluble ICAM-l without domains 3 to 5. g. K41d1; soluble ICAM-1 without domains 2 to 5.h G3; full-length ICAM-2 expression vector. i. G3-K4c; cytoplasmic domain of ICAM-lincorporated into ICAM-2 3’UTR. j. G3-K43u; 3’UTR of ICAM-l incorporated into ICAM-23’UTR. Open triangle represents CMV promoter/enhancer; Dashed square, ICAM-1 leadersequence; Open rectangles, . ICAM-1 extracellular domains (exd); Vertical wavy square,ICAM-1 transmembrane domain (tmd); Dotted square, ICAM-1 cytoplasmic domain (cyd);Solid triangles, AUUUA pentamers within ICAM-1 3’UTR; Solid square, ICAM-2 leadersequence; Spotted rectangles, ICAM-2 extracellular domains; Horizontal wavy square, ICAM2 transmembrane domain; Checkered square, ICAM-2 cytoplasmic domain.71ICAM-] induction by IFN-’y or PMA. 1.5 x 106 transfected L cells were treated with orwithout murine IFN-y (100 u/mi; Boehringer Mannheim, Laval, PQ) or PMA (10 ng/ml;Sigma, St. Louis, MO) and incubated at 37°C for increasing periods of time. The level ofmRNA was then determined by Northern blot analysis. The bands on the autoradiogram wereintegrated with a densitometer and normalized to the signal of the control 13-actin probe or tothe constitutively expressed rRNA gene. The maximum intensity was then arbitrarily set at100% mRNA level as a point of reference.mRNA half-life determination. 1.5 x 106 transfected L cells were treated withactinomycin D (10 ig/m1; GIBCO, Grand Island, NY) for increasing periods of time, and thelevel of mRNA was determined by Northern blot analysis. The integrated band values asdetermined by densitometry were normalized to the constitutively expressed rRNA gene, andthen the 100% mRNA level was set at time 0.RNA preparation and Northern blot analysis. Total RNA preparation, as described byChomczynski and Sacchi (1987) with minor modifications, and Northern blot analysis wereperformed as previously described (Ohh et al., 1994; see p4.5, chapter II).3.4 RESULTSIFN-y- and PMA-induced accumulation of ICAM-1 deletion-mutant mRNAs. A range ofdeletion-mutants of ICAM-1 cDNA shown in Fig. lOa-g were transfected into murinefibroblast L cells, which do not express endogenous ICAM-1. Upregulation of the level ofICAM- 1 mRNA following the treatment of these cells with IFN-y and PMA was thenexamined. 1FN-’y treatment of the transfectants only induced the accumulation of K4 and K4A3 mRNAs, but not the other mutant ICAM-l mRNAs (Fig. 1 la). Of particular interest is thestriking difference between the cells transfected with K4z3 and those with K4Ac3. The latterwere not responsive to WN-y whereas the former showed significant upregulation of ICAM- 172mRNA upon lFN-y treatment. The only difference between these two mutant ICAM-1mRNAs is the region encoding the cytoplasmic domain. This suggests that the region of 87 ntwithin ICAM- 1 mRNA encoding the cytoplasmic domain contains TFN-y-responsiveelement(s). On the other hand, PMA-induced accumulation of ICAM- 1 transcripts wasdependent on the 3’UTR since removal of this region rendered ICAM-1 mRNA unresponsiveto PMA treatment (Fig. 1 lb). Hence, it seems likely that ICAM- 1 mRNA contains at least twodistinct elements which are responsive to specific cell activation signals.PMA-induced stabilization of ICAM-] mRNA requires 3 ‘UTR. To further define therole of the 3’UTR of ICAM-1 mRNA in the upregulation of ICAM-l rnRNA by PMA , wegenerated a chimeric cDNA (G3-K43u, see Fig. lOj) consisting of the full-length ICAM-2cDNA (G3) and ICAM-l 3’UTR. ICAM-2 gene expression is constitutive and unresponsive tovarious inflammatory cytokines, including IFN-’y and PMA (Staunton et al., 1989; deFougerolles et al., 1991). As expected, G3 mRNA levels remained unaffected by PMA, whilethe chimeric G3-K43u mRNA accumulated upon PMA treatment (Fig. 12a). Furthermore,half-life analysis demonstrated that G3 mRNA had a long half-life which remained unaffectedby PMA treatment, whereas G3-K43u mRNA had a relatively short half-life and wassignificantly stabilized by PMA stimulation (Fig. 12b). These results indicate that the 3’UTRof ICAM- 1 contains destabilizing elements and PMA induces stabilization of ICAM- 1 mRNAthrough a mechanism that acts on the ICAM-l 3’UTR. In fact, the 3’UTR of both human andmurine ICAM- 1 contains multiple AUUUA sequences which have been shown to act asdestabilizing elements in numerous cytokine and proto-oncogene mRNAs, including GM-CSF,c-fos, c-myc, and c-myb (Sachs, 1993; Jackson, 1993).Region of ICAM-] mRNA encoding the cytoplasmic domain is responsive to IFN-’y.We also examined the region of ICAM-1 mRNA encoding the cytoplasmic domain for itsresponsiveness to WN-y. We transplanted this portion of gene onto ICAM-2 cDNA to generatea chimeric cDNA (G3-K4c; see Fig. lOi). WN-y treatment did not affect the level of G3mRNA (Fig. 1 3a) in the transfected L cells. However, the chimeric G3-K4c mRNA was73a- + IFN-y1Fig. 11 Induction of various forms of ICAM-1 mRNA by IFN-’y and PMA. L cells transfectedwith the various forms of ICAM-1 cDNA shown in Fig. 10 were treated with (+) or without (—)IFN-’y (a) or PMA (b) for 6 h, and then subjected to Northern blot analysis. The second andfourth panels from the top represent the corresponding signals of the control f3-actin. The datapresented are representative of two independent experiments.K4___-+K4Ac3 K41d3-5sK4 I I K41d1-2K41d1-+AK43S.74b-+PMAK4L.JK4A3 •N75aActinG3I 10246- + + +G3-K43uI I0246- + ++hPMAFig. 12 Posttranscriptional effects of PMA on ICAM-2 and ICAM-2/l chimeric mRNAs. a.Differential upregulation of G3 and G3-K43u mRNAs by PMA. L cells transfected with G3and G3-K43u cDNAs were treated with PMA for the increasing periods of time, and thendetected by Northern blot analysis. The bottom panel represents the corresponding signals ofthe control 13-actin. b. Half-life analysis of G3 and G3-K43u mRNAs detected by Northernblot hybridization. Both L cell transfectants were treated with PMA for 2 h and thenactinomycin D (ActD) was added for the indicated times. The data presented arerepresentative of two independent experiments. The chimeric cDNAs are described in Fig. 10.76G) C:) .C)F Ia1 2-+ -+IFN-’yFig. 13 Posttranscriptional effects of WN-’ on ICAM-2 and ICAM-2/l chimeric mRNAs. a.Differential upregulation of G3 and G3-K4c mRNAs by WN-y. L cells transfected with G3(panel 1) and G3-K4c (panel 2) cDNAs were treated with (+) or without (—) IFN-y for 6 h, andthen detected by Northern blot analysis. The bottom panels represent the correspondingsignals of the control -actin. b. Effect of IFN-y on G3-K43u mRNA accumulation. L cellstransfected with G3-K43u cDNA were treated with IFN-y for the increasing periods of timeand then detected by Northern blot analysis. The bottom panel represents the correspondingsignals of the control f3-actin. c. Half-life analysis of K4A3, K4&3, and G3-K4c mRNAsdetected by Northern blot hybridization. L cell transfectants were treated with IFN-’y for 2 hand then actinomycin D (ActD) was added for the indicated times. The data presented arerepresentative of two independent experiments. The chimeric cDNAs used for the transfectionare described in Fig. 10.78bActinG3-K43u4 h79CK4A3K4zc3- + IFN-’y+ + ActDII I0246 0246 hG3-K4c80markedly induced upon IFN-’y treatment (Fig. 13a). G3-K43u mRNA that was responsive toPMA was not affected by IFN-’y treatment (Fig. 1 3b). Half-life analysis showed that G3-K4cmRNA was relatively unstable but was stabilized by 1FN-y treatment (Fig. 13c). Furthermore,K43 mRNA had a shorter half-life than K4Ac3 mRNA, and only the stability of K4A3mRNA was increased upon WN-y treatment. The difference between the two mRNAs is theabsence of the cytoplasmic encoding region in the K4&3 mRNA. These results, takentogether, suggest that the region of ICAM-1 mRNA encoding the cytoplasmic domain containsdestabilizing element(s) responsive to TFN-y.3.5 DISCUSSIONThe results presented above indicate that IFN-’y and PMA stabilize and upregulate ICAM- 1mRNA by acting on two distinct destabilizing elements in ICAM-l mRNA. The IFN-y -responsive element(s) is found within the region encoding the cytoplasmic domain. This 87 ntregion is necessary for the stabilization and upregulation of ICAM-1 mRNA by WN-y. Whenfused with ICAM-2 mRNA, it is also sufficient to make stable ICAM-2 mRNA unstable andresponsive to IFN-y in mRNA stabilization. On the other hand, the 3’UTR of ICAM-1 mRNAis necessary and sufficient for the effect of PMA to stabilize and induce upregulation ofmRNA. This effect of ICAM- 1 3’UTR is to be expected, because it contains multiple AUUUAsequences which are known to destabilize many other mRNAs. Our findings with IFN-y aresomewhat unexpected, because the region of ICAM-1 mRNA responsive to WN-’y encodingthe cytoplasmic domain is only 87 nt long and does not contain AU-rich sequences or otherreadily recognizable unique sequences.The ability of WN-y to upregulate ICAM- 1 mRNA and cell surface expression of theICAIVI-1 protein has been well documented (Dustin et al., 1986; Pober et al., 1986; Rothlein etal., 1988; Degitz et al., 1991). However, previous studies mainly focused on the ability of81IFN-y to induce transcription of the ICAM- 1 gene (Degitz et aL, 1991; Voraberger et al., 1991;Wertheimer et al., 1992). We have previously reported that IFN-y may have an additionaleffect, which is to stabilize otherwise unstable ICAs’1- 1 mRNA (Ohh et al., 1994). This effectof IFN-’ may be particularly relevant for the upregulation of ICAM- 1 by WN-y on some cells,including endothelial cells that constitutively express low levels of ICAM- 1. It is likely thatthis low basal level of ICAM- 1 expression is maintained, at least in part, by a process thatreduces the half-life of ICAM- 1 mRNA. As our present study has demonstrated, at least twodistinct mechanisms can stabilize and upregulate ICAM- 1 mRNA levels. Posttranscriptionalcontrol of ICAM- 1 gene expression may thus be an important mechanism regulating theexpression of ICAM- 1 on cells that constitutively express low levels of ICAM- 1. Such amechanism might then prove to play a key role in regulating the ability of leukocytes to firstadhere to endothelial cells and then subsequently transmigrate into inflamed tissues.3.6 REFERENCESBarton RW, Rothlein R, Ksiazek J, Kennedy C (1989) The effect of anti-intercellular adhesionmolecule-i on phorbol ester-induced rabbit lung inflammation. J Immunol 143:1278.Chomczynski P, Sacchi N (1987) Single-step method of RNA isolation by acid guanidiniumthiocyanate-phenol-chloroform extraction. Anal Biochem 162:156.Cosimi AB, Conti D, Delmonico FL, Preffer Fl, Wee S-L, Rothlein R, Faanes R, Colvin RB(1990) In vivo effects of monoclonal antibody to ICAM-1 (CD54) in nonhuman primates withrenal allografts. J Immunol 144:4604.de Fougerolles AR, Stacker SA, Schwarting R, Springer TA (1991) Characterization of ICAM2 and evidence for a third counter-receptor for LFA-1. J Exp Med 174:253.Degitz K, Lian-Jie L, Caughman SW (1991) Cloning and characterization of the 5’-transcriptional regulatory region of the human intercellular adhesion molecule 1 gene. J BiolChem 266: 14024.Diamond MS, Staunton DE, Marlin SD, Springer TA (1991) Binding of integrin Mac-l(CD1 lb/CD 18) to the third immunoglobulin-like domain of ICAM-1 (CD54) and its regulationby glycosylation. Cell 65:961.82Dustin ML, Rothlein R, Bhan AK, Dinarello CA, Springer TA: A natural adherence molecule(ICAM-1) (1986) induction by IL-I and interferon-y, tissue distribution, biochemistry andfunction. J Immunol 137:245.Honey KJ, Carpenito C, Baker B, Takei F (1989) Molecular cloning of murine intercellularadhesion molecule (ICAM-l). EMBO J 8:2889.Isobe M, Yagita H, Okumura K, Ihara A (1992) Specific acceptance of cardiac allograft aftertreatment with antibodies to ICAM- 1 and LFA- 1. Science 255:1125.Jackson RJ (1993) Cytoplasmic regulation of mRNA function: The importance of the Yuntranslated region. Cell 74:9.Kay R, Humphries RK (1991) New vectors and procedures for isolating cDNAs encoding cellsurface proteins by expression cloning in COS cells. Methods Mol Cell Biol 2:254.Koren HS, Handwerger BS, Wunderlich JR (1975) Identification of macrophage likecharacteristics in a cultured murine tumor line. J Immunol 114:894.Makgoba MW, Sanders ME, Ginther-Luce GE, Dustin ML, Springer TA, Clark EA, MannoniP, Shaw S (1989) ICAM-i is a ligand for LFA-1-dependent adhesion of B, T and myeloidcells. Nature 33 1:86.Marlin SD, Springer TA (1987) Purified intercellular adhesion molecule-i (ICAM-1) is aligand for lymphocyte function-associated antigen 1 (LFA- 1). Cell 51:813.Ohh M, Smith CA, Carpenito C, Takei F (1994) Regulation of intercellular adhesion molecule-1 gene expression involves multiple mRNA stabilization mechanisms: Effects of interferon-yand phorbol myristate acetate. Blood 84:2632.Pober JS, Gimbrone MA Jr, Lapierre LA, Mendrick DL, Fiers W, Rothlein R, Springer TA(1986) Overlapping patterns of activation of human endothelial cells by interleukin 1, tumornecrosis factor, and immune interferon. J Immunol 137:1893.Rothlein R, Czajkowski M, O’Neill MM, Marlin SD, Mainolfi E, Merluzzi VJ (1988)Induction of intercellular adhesion molecule-i on primary and continuous cell lines by proinflammatory cytokines. J Immunol 141:1665.Rothlein R, Dustin ML, Marlin SD, Springer TA (1986) A human intercellular adhesionmolecule (ICAM-i) distinct from LFA-i. J Immunol 137:1270.Sachs AB (1993) Messenger RNA degradation in Eukaryotes. Cell 74:413.Sambrook J, Fritsch EF, Maniatis T (1989) Molecular cloning: A laboratory manual, 2nd Ed.Cold Spring Harbor Laboratory, Cold Spring Harbor, NY.83Sligh JE, Ballantyne CM, Rich SS, Hawkins HK, Smith CW, Bradley A, Beaudet AL (1993)Inflammatory and immune responses are impaired in mice deficient in intercellular adhesionmolecule 1. Chem Immunol 90:8529.Smith CW, Marlin SD, Rothlein R, Toman C, Anderson DC (1989) Cooperative interactionsof LFA- 1 and Mac-i with intercellular adhesion molecule-i in facilitating adherence andtransendothelial migration of human neutrophils in vitro. J Clin Invest 83:2008.Staunton DE, Dustin ML, Springer TA (1989) Functional cloning of ICAM-2, a cell adhesionligand for LFA- 1 homologous to ICAM- 1. Nature 339:61.Voraberger G, Schafer R, Stratowa C (1991) Cloning of the human gene for intercellularadhesion molecule 1 and analysis of its 5-regulatory region: induction by cytokines andphorbol ester. J Immunol 147:2777.Wawryk SO, Novotny JR, Wicks IP, Wilkinson D, Maher D, Salvaris E, Welch K, Fecondo J,Boyd AW (i989) The role of the LFA-1JICAM-i interaction in human leukocyte homing andadhesion. Immunol Rev 108:135.Wegner CD, Gundel RH, Reilly P, Haynes N, Lens LG, Rothlein R (1990) Intercellularadhesion molecule-i (ICAM- 1) in the pathogenesis of asthma. Science 247:456.Welder CA, Lee DHS, Takei F (1993) Inhibition of cell adhesion by microspheres coated withrecombinant soluble ICAM-i. J Immunol 150:2203.Wertheimer SJ, Myers CL, Wallace RW, Parks TP (1992) Intercellular adhesion molecule-igene expression in human endothelial cells: differential regulation by tumor necrosis factor-osand phorbol myristate acetate. J Biol Chem 267:12030.Yamasaki K, Hiratia Y, Ywata H, Kawanishi Y, Seed B, Tanaigush T, Hirano T, Kishimoto T(1988) Cloning and expression of the human interleukin-6 (BSF-2IIFNbeta2) receptor. Science241:825.84IvREGULATION OF ICAM-1 mRNA STABILITY BY CYCLOHEXIMIDE: ROLE OFSERINE/THREONINE PHOSPHORYLATION AND PROTEIN SYNTHESIS4.1 ABSTRACTCycloheximide is a protein synthesis inhibitor that superinduces the expression of many genesby preventing the degradation of otherwise labile mRNAs. In some genes this depends on thepresence of the AUUUA destabilizing multimers in the 3tUTR. The effect of cycloheximideon the expression of the murine ICAM- 1 gene in several cell lines including A20 (B celllymphoma), T28 (T cell hybridoma), P388D1 (monocytic cell), SVEC4-lO (lymphoidendothelial cell), and ICAM-1-transfected murine fibroblast L cells was examined.Cycloheximide was able to dramatically increase the accumulation of ICAM- 1 mRNA in allthe cell lines examined except T28, and this seemed to be due to the stabilization of the ICAM1 mRNA as indicated by mRNA half-life analyses. To determine whether these effects weredependent on the 3’UTR containing AUUUA sequences, L cells were transfected with eitherfull-length ICAM-1 cDNA or a truncated form lacking AUUUA sequences in their 3’UTR(ICAM-1z3). There was no discernible difference in the effect of cycloheximide on ICAM-1mRNA accumulation or half-life between the two types of tranfected cells. The effect ofcycloheximide on ICAM- 1 mRNA was markedly suppressed by serine/threonine (ser/thr)kinase inhibitors, H-7 and staurosporine, whereas the ser/thr phosphatase inhibitor, okadaicacid, augmented the cycloheximide effect. Inhibitors of protein tyrosine kinases andphosphatases had no effect. Unexpectedly, the level of cell surface ICAM- 1 as well as de novosynthesis of ICAM-1 in SVEC4-lO and the ICAM-1-transfected L cells were also upregulatedby cycloheximide whereas the overall protein synthesis in these cells were profoundly85inhibited, suggesting that ICAM- 1 protein synthesis in these cells escapes the translationalinhibition by cycloheximide. These results suggest that the stabilization of ICAM- 1 mRNA bycycloheximide is independent of its translational inhibition and that ser/thr phosphorylation ofunidentified protein(s) seems to play a crucial role in mediating this ICAM- 1 mRNAstabilizing effect.4.2 INTRODUCTIONICAM- 1 expression is dramatically increased at sites of inflammation (Rothlein et al., 1986;Dustin et al.,1986; Pober et al., 1986), thereby providing important means of regulating cell-cell interactions and the enhanced expression of ICAM-1 on venule endothelium is known tofacilitate the adhesion and subsequent trans-endothelial migration of leukocytes bearing LFA-1or Mac-i into inflammatory tissues (for review, see Wawryk et al., 1989; Smith et al., 1989),as well as appropriate interaction of lymphocytes with cells expressing targeted antigens(Dougherty et al., 1988; Wawryk et aL, 1989; Springer, 1990). However, despite the obviousimportance of inducible ICAM- 1 expression in numerous immunologic responses, very little isknown about the intracellular mechanisms controlling its expression. Although there is anincreasing body of evidence that mRNA turnover plays an important role in regulating geneexpression. For example, the 3’UTRs of both human and murine ICAM-1 mRNAs containseveral AUUUA sequences, implicated in the decay of a number of short-lived mRNAs of thisclass (Sachs, 1993; Shaw and Kamen, 1986; Schuler and Cole, 1988), although an involvementof ATJUUA sequences in the 3’UTR of ICAM-1 mRNA in the control of its turnover has notbeen determined.Cycloheximide is widely used to stabilize labile mRNAs. Because of its ability toinhibit peptidyl transferase, this effect of cycloheximide implies that the stability of mRNA isassociated with protein synthesis. For example, almost all mRNAs in yeast are stabilized by86cycloheximide (Herrick et al., 1990). Similarly, the degradation of mRNAs for histone, 1-tubulin, transferrin receptor, certain proto-oncogenes, and various lymphokines is dramaticallyreduced in the presence of translational inhibitors (for reviews, see Jackson, 1993 and Sachs,1993). These observations suggest that most mRNAs need to be translated to be degraded.However, it is also possible that mRNAs are degraded by a unique class of highly labileproteins that are no longer synthesized but continue to be degraded upon cycloheximidetreatment. The mechanism by which labile mRNAs are stabilized by cycloheximide has notyet been elucidated.We report here that cycloheximide stabilizes otherwise labile ICAM-1 mRNA and thatthis effect is independent of the 3’UTR which contains multiple AUUUA sequences. Inhibitorsof ser/thr kinases as well as those for phosphatases have profound effects on the induction andstabilization of ICAM-l mRNA by cycloheximide. Furthermore, in some cells ICAM-1synthesis is increased by cycloheximide treatment whereas the overall protein synthesis isstrongly inhibited, indicating that the stabilization of ICAM- 1 mRNA can be obtained whetheror not its translation is inhibited.4.3 MATERIALS AND METHODSCell culture. The murine B cell lymphoma A20, T cell hybridoma T28, monocytic cell lineP388D1, lymph node endothelial cell line SVEC4-10 (O’Connell and Edidin, 1990), and thefibroblast L cells were cultured in DMEM (G]ECO, Grand Island, NY) containing 10% FCSand antibiotics.Expression vectors and transfection into L cells. Expression vectors were constructedusing standard recombinant DNA techniques (Sambrook eta!., 1989). Briefly, K4-1.1 ICAM1 eDNA (Honey et al., 1989) was partially digested with PstI to remove its 3’UTR. Thesoluble G3-1.1 ICAM-2 (sICAM-2) cDNA (Ohh et a!., 1994) in pBST was cut out with XbaI87and Not!. The sICAM-2, truncated (ICAM-1A3) and the full-length ICAM-1 cDNAs wereblunt-ligated into Smal site of a stable mammalian expression vector, pRC6 (derived frompAX1 14 (Kay and Humphries, 1991) with a hygromycin resistance gene driven by a thynildinekinase promoter). Escherichia co/i MC1O61/p3 (Yamasaki et aL, 1988) was used fortransformation. ICAM-1, ICAM-1A3, and sICAM-2 cDNAs in expression vectors weretransfected into ICAM-l and -2 negative L cells by the calcium phosphate method (Sambrooket aL, 1989). Transfectants (L-ic-l, L-ic-1A3, and L-sic-2) were isolated under hygromycinselection (250 jig/mi; Calbiochem, La Jolla, CA). The surface expression of ICAM-1 wastested by flow cytometry as described below.ICAM-] in4uction by cycloheximide and phosphorylation inhibitors. 1.5 x 106 cellswere pretreated for 15 mm at 37°C with various phosphorylation inhibitors, includingstaurosporine (0.2 jiM; Calbiochem), H-7 (10 jiM; Calbiochem), H-89 (0.1 jiM; Calbiochem),bisindolylmaleimide (0.03 jiM; Calbiochem), okadaic acid (0.5 jiM; Calbiochem), genistein(20 jiM; Calbiochem), tyrphostein (30 jiM; Calbiochem), and vanadate (50 jiM; Calbiochem).Then cycloheximide (10 jig/mi; Sigma, St. Louis, MO) was added for increasing periods oftime at 37°C. The level of ICAM- 1 mRNA was then determined by Northern blot analysis asdescribed below.ICAM-] mRNA half-life determination. 1.5 x 106 cells were treated with variousphosphorylation inhibitors in combination with cycloheximide for 2 h at 37°C, and thenactinomycin D (10 jig/mi, GIBCO, Grand Island, NY) was added for increasing periods oftime. The level of ICAM-1 mRNA was determined by Northern blot analysis as describedbelow.RNA preparation and Northern blot analysis. see p45, chapter II.Flow cytometric analysis. Cells were treated with various kinase and phosphataseinhibitors in combination with cycloheximide for 12 h. Briefly, the cells were washed withHank’s balanced salt solution containing 2% FCS (the adherent cells, SVEC4-10 and L-ic-l,were then harvested with PBS containing 2.5 mM EDTA). Cells were directly stained with88anti-ICAM- 1 (YN 1/1.7) -FITC mAb (Honey et al., 1989) containing 0.1% sodium azide andanalyzed on fluorescence-activated cell sorter (FACS0rt, Becton Dickinson, Mountain View,CA). Dead cells stained with propidium iodide were gated out.Metabolic labeling and immunoprecipitation. 1.5 x 106 cells cultured in DMEM +10% FCS were treated with cycloheximide (10 pg/m1) or puromycin (10 !.tM) at 37°C forvarious times and 2.5 h prior to cell lysis, the cells were washed with PBS and incubated inmethionine-free DMEM + 10% dialyzed FCS + cycloheximide (10 pg/m1) or + puromycin (10tM). After 30 mm, 0.25 mCi/mi of[35S]-methionine was added for 2 h. The cells were thenwashed with PBS (the adherent cells, SVEC4-l0 and L-ic-1, were then harvested with PBScontaining 2.5 mM EDTA) and solubilized with 0.5% NP-40 in 4°C PSB (50 mM Hepes, pH7.4, 100 mlvi NaF, 10 mlvi NaPPi, 2 mM Na3V04, 4 mM EDTA, 2 mM PMSF, 10 jig/mileupeptin, and 2 jig/ml aprotinin). After 1 h at 4°C, the cell lysates were clarified bycentrifugation for 10 mm. 60 jii of the supernatant was kept as total cell lysate and to theremaining supernatant, 20 jil of YN1/l.7 anti-ICAM-1 hybridoma supernatant was added.Following 1 h of gentle agitation at 4°C, immune complexes were bound onto anti-rat Igcoated beads, which were previously treated at 4°C for 2 h with Tris-buffered saline (TBS, 10mM Tris-Ci, pH 8.0, 0.15 M NaC1) containing 5% skim milk powder and 5% BSA tominimize non-specific binding, at 4°C for 1.5 h, washed gently thrice in cold PSB and thenboiled in SDS-sample buffer for 2 mm. The total cell lysates were also boiled in SDS-samplebuffer and all the samples were subjected to SDS-PAGE analysis (Sambrook et al., 1989)using 7.5% polyacrylamide gels.4.4 RESULTSRole of 3’UTR in the induction of ICAM-] mRNA by cycloheximide. Both human and murineICAIvI-1 mRNAs contain multiple AUUUA sequences in their 3’UTRs. Treatment of various89murine cell lines, including A20 (B cell lymphoma), T28 (T cell hybridoma), P388D1(monocytic cell), and SVEC4-lO (lymphoid endothelial cell), with cycloheximide dramaticallyincreased the accumulation of ICAM- 1 mRNA in all cell types examined except in T28 (Fig.14). In order to investigate the role of the 3’UTR containing multiple AUUUA repeats in theupregulation of ICAM- 1 mRNA level, a full-length murine ICAM- 1 cDNA clone K4- 1.1 andits truncated form without the 3’UTR (ICAM-1A3) were subcloned into the mammalianexpression vector pRC6 (Fig. 1 5a) and then stably transfected into murine fibroblast L cells,which do not express detectable levels of ICAM-l mRNA regardless of cycloheximidetreatment. Bulk populations of L cells transfected with full-length ICAIvI- 1 (L-ic- 1) and thosewith ICAM- 1 A3 (L-ic- 1A3) were established by hygromycin resistance and the cell surfaceexpression of ICAM-1 was confirmed by flow cytometry (see below). Northern blot analysisclearly showed that upon cycloheximide treatment both ICAM-l and ICAM-1A3 mRNA levelsincreased dramatically at a rapid rate (Fig. 1 5b). However, the soluble ICAM-2 (sICAM-2)mRNA was only slightly induced by cycloheximide (Fig. 15c), which demonstrates that the3’UTR (rabbit 13-globin) supplied by the vector is unresponsive to cycloheximide. Therefore,the 3’UTR of ICAM-1 containing the AUTJUA multimers are not required for the induction ofICAM-l mRNA by cycloheximide.Ser/thr kinase and phosphatase inhibitors and cycloheximide effect. PMA has beenshown to upregulate ICAM-1 mRNA level by stabilizing it (Wertheimer et al., 1992),suggesting that protein kinases may be involved in the regulation of ICAM- 1 mRNAexpression. To examine the role of protein kinases in the effects of cycloheximide on ICAM- 1mRNA expression, L-ic-1 cells were treated overnight with PMA to downregulate PKC.However, such treatment, whether 10 or 40 ng/ml, prior to the addition of cycloheximide didnot inhibit the accumulation of ICAM- 1 transcripts (Fig. 16). On the other hand, treatment ofcells with staurosporine, which is a general ser/thr kinase inhibitor, prior to the addition ofcycloheximide completely inhibited the induction of ICAM-1 mRNA level (Fig. 16). Theaddition of staurosporine after the overnight treatment with PMA abolished the cycloheximide90ICAM-1ActinaFig. 14 Effects of cycloheximide on ICAM-1 mRNA expression. A20, P388D1, T28, andSVEC4- 10 cells were treated with cycloheximide for the increasing periods of time and theICAM- 1 niRNA levels were detected by Northern blot analysis. The bottom panels representthe signals of the control f-actin. The data presented are representative of three independentexperiments.A20 P388 T28 SVECI II II I II024602460246 06 h91aCMV prolenhICAM-1exdIICAM-1zX3 “j I I I IsICAM-2 >::::I:::ILexdlFig. 15 Effects of cycloheximide on ICAM-l mRNA expression in L-ic-l and L-ic-1A3 cellsand on ICAM-2 mRNA expression in L-sic-2 cells, a. Schematic diagrams of pRC6-ICAM-l,pRC6-ICAM-1A3, and pRC6-sICAM-2 expression vectors. Open triangle representscytomegalovirus (CMV) promoter/enhancer (pro/enh); solid triangles, AUUUA pentamers;open rectangle, ICAM-l extracellular domains (exd); vertical wavy square, ICAM-ltransmembrane domain (tmd); dotted square, ICAM-l cytoplasmic domain (cyd); spottedrectangle, ICAM-2 extracellular domains. b. The induction of ICAM-l and ICAM-li3mRNA levels in L-ic-l and L-ic-1A3 cells stimulated with (+) or without (—) cycloheximide(CHX) for the indicated times, as detected by Northern blot analysis. The second and fourthpanels from the top represent the corresponding signals of the control f3-actin. The datapresented are representative of three independent experiments. c. The induction of sICAM-2mRNA levels in L-sic-2 cells stimulated with cycloheximide (CHX) for the indicated times, asdetected by Northern blot analysis. The second panel represents the corresponding signals ofthe control 3-actin. The data presented are representative of two independent experiments.92b01 248 h--+ -+-+-+CHXICAM-1 IActin ma. !ICAM-lz\3 • •NActin93CsICAM-2Actin036 h-+-I-CHX94PMA-10+Fig. 16 Effects of PMA over-stimulation and staurosporine on ICAM-1 mRNA accumulationinduced by cycloheximide. Cells were treated with PMA (either 10 or 40 nglrnl; PMA-10 andPMA-40, respectively) for 72 h or staurosporine alone for 15 mm (STS) or PMA for 72 hfollowed by staurosporine for 15 mm (PMA-10 + STS), and then incubated in the presence ofcycloheximide (CHX) for 6 h. The accumulation of ICAM-1 transcripts was determined byNorthern blot analysis. The bottom panels represent the signals of the control 13-actin. Thedata presented are representative of five independent experiments performed on L-ic- 1, L-ic- Iz3, SVEC4-l0, P388D1, and A20 cells.PMA-10 STS STSI ii 1 I0606 06-- + --+--+PMA-40I I06 h- -+ CHXICAM-1Actin95effect on ICAM-1 mRNA induction (Fig. 16). These results suggest that the effect ofstaurosporine is not due to the inhibition of PKC but is due to the inhibition of other ser/thrkinase(s). Consistent with the staurosporine result, H-7, another ser/thr kinase inhibitor,strongly inhibited the effect of cycloheximide (Fig. 17a). In contrast, treatment with the ser/thrphosphatase inhibitor, okadaic acid, alone increased ICAM-1 mRNA level, and in combinationwith cycloheximide augmented the cycloheximide effect (Fig. 17a,b). The protein tyrosinekinase inhibitor, genistein, and tyrosine phosphatase inhibitor, vanadate, had no discernibleeffect on ICAM-1 mRNA accumulation induced by cycloheximide (Fig. 17a). Tests were alsoperformed using tyrphostein instead of genistein and gave similar results. In agreement withthe PMA over-stimulation results shown in Fig. 16, bisindolylmaleimide, a highly selectiveinhibitor of PKC, or 11-89, a potent inhibitor of cAMP-dependent protein kinase, was alsounable to inhibit the cycloheximide effect (Fig. 17b). These results, taken together, suggestthat cycloheximide induces the accumulation of ICAM- 1 mRNA through a ser/thrphosphorylation-dependent pathway.Half-life analysis. Although ICAM-1 mRNA has a short half-life (Fig. 18a,b),,treatment with cycloheximide dramatically prolonged the half-life to far greater than 2 h (Fig.1 8a). Treatment with the ser/thr kinase inhibitors, staurosporine and H-7, alone had anegligible effect on the turnover rate (Fig. 1 8a). However, the cycloheximide effect wasefficiently inhibited by these ser/thr kinase inhibitors (Fig. 1 8a). Treatment with the ser/thrphosphatase inhibitor, okadaic acid, alone extended the half-life to far greater than 2 h (Fig.18b). Concordant with the induction analysis, the half-life analysis strongly suggests thatcycloheximide stabilizes the otherwise labile ICAM- 1 mRNA through a ser/thrphosphorylation-dependent pathway. The induction and half-life analyses were also performedin duplicate on A20, P388D1, and SVEC4-lO cells and they all gave similar and consistentresults. Furthermore, sICAM-2 mRNA had a relatively long half-life (Fig. 18c) and asexpected from the induction analysis, treatment of L-sic-2 cells with cycloheximide had anegligible effect on the sICAM-2 mRNA turnover rate.96aICAM-1Actin- Sts H-7 OkAc Gns VndII II II II II II-+ -+-+ -+ -+-+CHXFig. 17 Effects of phosphorylation inhibitors on ICAM- 1 mRNA accumulation bycycloheximide. a. Cells were first treated with various phosphorylation inhibitors (Stsrepresents staurosporine; H-7,H-7; OkAc, okadaic acid; Gns, genistein; Vnd, vanadate; —, nophosphorylation inhibitor added) for 15 mm and then the cells were treated with (+) or without(—) cycloheximide (CHX) for 6 h. b. Cells were first treated with various phosphorylationinhibitors (BIM represents bisindolylmaleimide; H-89,H-89) for 15 mm and then the cells weretreated with (+) or without (—) cycloheximide for 6 h. The accumulation of ICAM- 1transcripts in a and b was determined by Northern blot analysis. The bottom panels representthe signals of the control -actin. The data presented are representative of five independentexperiments performed on L-ic-1, L-ic-1A3, SVEC4-1O, P388D1, and A20 cells.97C)C)004‘]19% IIz:a02 ir-Cs SHH1+C C’’+ +-C S5 HH2 h+ ActDC C’+ +- CS SHHFig. 18 ICAJVI-1 and ICAM-2 mRNA half-life analysis. a. Cells were treated withphosphorylation inhibitors (S represents staurosporine; H, H-7; —, no inhibitor added) for 15mm, followed by stimulation with or without cycloheximide (C) for 2 h. Actinomycin D(ActD) was then added for the indicated times. b. Cells were treated with (+) or without (—)okadaic acid (OkAc) for 2 h and then actinomycin D was added for the indicated times.ICAM-l mRNA levels in a and b were determined by Northern blot analysis. Equalizedloading of total RNA in each lane was confirmed by the ethidium bromide staining of theconstitutively expressed rRNA gene. The data presented are representative of five independentexperiments performed on L-ic-1, L-ic-1A3, SVEC4-1O, P388D1, and A20 cells. c. L-sic-2cells were treated with cycloheximide (CHX) for 2 h and then actinomycin D was added forthe indicated times. sICAM-2 mRNA levels were measured by Northern blot analysis.Equalized loading of total RNA in each lane was confirmed by the ethidium bromide stainingof the constitutively expressed rRNA gene. The data presented are representative of twoindependent experiments.ICAM-199b- + OkAc+ + ActDI II Io .5 1 2 0 .5 1 2 hICAM-1100C+ CHX+ + ActDI II I0 .5 1 2 0 .5 1 2 hsICAM-2101Flow cytometric analysis of cell surface ICAM-1 expression. Since the primary effectof cycloheximide is thought to be protein synthesis inhibition, we examined the effect ofcycloheximide on ICAM- 1 protein expression on the cell surface by flow cytometry in thepresence or absence of various phosphorylation inhibitors. Contrary to the expected reductionin ICAM- 1 expression, the addition of cycloheximide significantly increased the surfaceexpression of ICAM-1 on L-ic-l cells (Fig. 19b). Untransfected L cells did not express ICAM1 on the cell surface regardless of cycloheximide treatment (Fig. 19a). The treatment of thetransfected L cells with the tyrosine kinase inhibitors (genistein and tyrphostein), tyrosinephosphatase inhibitor (vanadate), general ser/thr kinase inhibitors (H-7 and staurosporine), theinhibitors of PKC and PKA (bisindolylmaleimide and 11-89, respectively) had no effect on theexpression of ICAM-l (Fig. 19c). Another translational inhibitor, puromycin, alone also hadno effect, indicating that the effect of cycloheximide is not general to all protein synthesisinhibitors (Fig. 1 9c). Okadaic acid alone was able to increase the expression of ICAM- 1 (Fig.19d). The upregulation of cell surface ICAM-1 expression by cycloheximide was markedlyreduced by the addition of staurosporine or H-7 (Fig. 19e,f). On the other hand, otherinhibitors had no effect on the upregulation of cell surface ICAM- 1 expression bycycloheximide (Fig. 19g,h). These results are consistent with the observations made at themRNA level, and strongly suggest again that cycloheximide upregulates the expression ofICAM-1 through a ser/thr phosphorylation-dependent pathway. However, A20, P388D1, andT28 cells did not show any discernible change in the expression of ICAM- 1 followingcycloheximide treatment (Fig. 20), suggesting a tissue-specific effect of cycloheximide.Measurement of de novo ICAM-1 protein synthesis in the presence of cycloheximide.Upregulation of ICAI\’l- 1 expression on the cell surface by cycloheximide treatment (describedabove) might be explained by an ineffective inhibition of protein synthesis in the cell linestested. Therefore, ICAM- 1 protein synthesis in cycloheximide treated cells was determined.A20, SVEC4-10, and L-ic-1 cells were treated with cycloheximide for increasing periods oftime and then labeled with[35S1-methionine for 2 h. ICAM-1 proteins were102LiiDz-j.1____________ ____________LUC-)LU>F-JLUICAM-1 FLUORESCENCE INTENSITYFig. 19 Flow cytometric analysis of ICAM- 1 cell surface expression. Cells were treated withkinase and phosphatase inhibitors in combination with cyclohexiniide for 12 h, then stainedwith YNII1.7-FITC mAb and analyzed by flow cytometry. A. Effects of inhibitors includingstaurosporine, H-7, bisindolylmaleimide, H-89, okadaic acid, genistein, tyrphostein, vanadate,puromycin, and cycloheximide on untransfected L cells (UNTRANSF + INIIIB).Representative profile. B. Cycloheximide treatment of L-ic-l (TRANSF + CHX). C.Inhibitor (as in A except okadaic acid) treatment of L-ic- 1 (TRANSF + INHIB).Representative profile. D. Okadaic acid treatment of L-ic-l (TRANSF + OkAc). E.Treatment with cycloheximide alone (TRANSF + CHX) or in combination with staurosporine(TRANSF + CHX + STS). F. As in E except with H-7. G. as in E except withbisindolylmaleimide (BIM). H. As in E except with vanadate (VND). G and H also representthe effects of genistein, tyrphostein, and H-89. The data presented are representative of threeindependent experiments performed on L-ic-1,L-ic-L&3, and SVEC4-lO cells.+ INHIBB UNTRANSFTRANSFTRANSF + CHX••11 1’$0 UNTRANSFTRANSFTRANSF + OkAc1RSF+CHX7C UNTRANSFTRANSFTRANSF + INHIBiO1 1 l3 11$EThA:SFU liii 1W IllS 1lo ..---1i 12 iG /TRANSFTRANSF + CHX + BIMTRANSF + CHXU 1W 1W IUS 11)1103UiDz-J-JUiC)Ui>-JUiA20ThT28==IiX3”ThIP388ICAM-1 FLUORESCENCE INTENSITYFig. 20 Flow cytometric analysis of the effects of cycloheximide on ICAM-l cell surfaceexpression on A20, P388D 1, and T28 cells. Cells were treated with cycloheximide for 12 h,then stained with YN1/1.7-F1TC mAb and analyzed by flow cytometry. Shaded histogramsrepresent the cell surface expression of ICAM- 1 without the cycloheximide treatment; linedhistograms, ICAM- 1 cell surface expression with the cycloheximide treatment.104immunoprecipitated and analyzed by SDS-PAGE. It was clear from the total cell lysatefraction of each cell type that cycloheximide almost completely inhibited the overalltranslational capacity of the cell (Fig. 21 a). Concordantly, ICAM- 1 protein synthesis wassignificantly reduced in A20 cells (Fig. 21a). The band around 90 kDa at time 0 for the A20cells was most likely ICAM-1 protein since a negative control niAb, YE 1/48, did notimmunoprecipitate a protein of similar size (Fig. 21b). Furthermore, a puromycin-treated totalcell lysate fraction also showed marked inhibition of the overall translational capacity of thecell, as well as diminished ICAM- 1 protein level (Fig. 2 ic). However, cycloheximide did notdiminish but rather increased ICAM-l protein synthesis in SVEC4-10 and the L-ic-l cells (Fig.21 a). Hence, in agreement with the flow cytometric data, these metabolic labeling experimentsshow that ICAM-l protein synthesis in the SVEC4-10 endothelial cells and the L-ic-1fibroblast cells “escapes” the translational inhibitory effect of cycloheximide. The ser/thrkinase and phosphatase inhibitors had no measurable effect on protein synthesis and did notinfluence the translational inhibition by cycloheximide (Fig. 21d).4.5 DISCUSSIONThis series of experiments demonstrate first, that cycloheximide stabilizes and superinducesICAM-1 mRNA, but that this effect does not require AUUUA repeats in the 3’UTR; secondly,that the ser/thr kinase inhibitors, staurosporine and H7, abrogate the effect of cycloheximidewhereas the ser/thr phosphatase inhibitor, okadaic acid, alone stabilizes ICAM- 1 mRNA andalso augments the ability of cycloheximide treatment to stabilize ICAM- 1 mRNA; and finallythat although cycloheximide inhibits the overall level of protein synthesis, it can enhance thesynthesis of ICAM- 1 protein and its associated expression on the cell surface in some, but notall, cell types. Cycloheximide acts by preventing the peptidyl transferase activity of the 60Sribosomal subunit and is widely used as a general purpose inhibitor of eukaryotic protein105YN1/1 .7TCL A20 L-ic-1048 048048Fig. 21 Effect of cycloheximide and puromycin on de novo ICAM-1 protein synthesis. a.Cells were treated with cycloheximide for the increasing periods of time. De novo ICAM- 1proteins that had incorporated[35S]-methionine were immunoprecipitated from A20, L-ic- 1,and SVEC4-10 cells with anti-ICAM-1 mAb and subjected to SDS-PAGE analysis. TCLrepresents the total cell lysate of L-ic- 1. TCLs of A20 and SVEC4- 10 gave similar results.The solid arrow marks ICAM-1 proteins (approximately 90 kDa). b. A20 cells were treatedwith cycloheximide for the increasing periods of time as in a, but immunoprecipitated with anegative control mAb, YE1148, and subjected to SDS-PAGE analysis. c. SVEC4-l0 cellstreated with puromycm for the increasing periods of time and immunoprecipitated with antiICAM-1 mAb and subjected to SDS-PAGE analysis.. d. TCL of L-ic-l treated withcycloheximide (CHX) alone, in combination with H-7 (H-7 + CHX) and okadaic acid (OkAc +CHX), H-7 alone (H-7), and okadaic acid alone (OkAc).a kDa200SVE048 h10097.4— —6946•130106bYE1148kDa 0 4 8 h200 —10097.469146107CTCL YN1/1.7I I I04804 8hkDa20069—10097.43046108dH-7 OkAc+ +CHX CHX CHX H-7 OkAcI I I II II II048048048048048 hkDa— I20010097.4694630109synthesis. Although its ability to stabilize and superinduce labile mRNAs is well known, theprecise mechanisms have yet to be elucidated. Two models have been proposed to explain theeffect of cycloheximide (for review, see Sachs, 1993). One states that the degradation ofmRNA requires its translation. Therefore, inhibition of translation by cyclohexirnide resultsin the stabilization of mRNA. The second model states that a unique class of proteins thathave a rapid turnover are required for the degradation of certain highly labile mRNAs. Withinthis model, cycloheximide might be envisaged to inhibit the synthesis of these proteins,resulting in a quick down-regulation of mRNA degradation. Our finding that cycloheximidestabilizes ICAM-1 mRNA without inhibiting the synthesis of ICAM-1 protein is incompatiblewith the first model and favours the second.It was quite unexpected that the synthesis of ICAM-1 determined by[35S]-methioninelabeling and immunoprecipitation would actually be increased in cycloheximide-treated cellsfor two of the lines tested (transfected L cells and endothelial cells) while the overall proteinsynthesis in the same cells was profoundly inhibited. The increase observed in the synthesis ofICAM- 1 and its expression level on the cell surface can be explained by the increased level ofICAM-l mRNA observed subsequent to cycloheximide treatment. However, how ICAM-1protein synthesis in some cells escapes the general cycloheximide-mediated inhibition ofprotein synthesis remains unclear.Our results also suggest that ser/thr phosphorylation of unknown proteins can regulatethe effect of cycloheximide on ICAM- 1 mRNA stability. The ser/thr kinase inhibitors, 11-7and staurosporine, completely inhibited the accumulation and stabilization of ICAM- 1 mRNAby cycloheximide. On the other hand, the ser/thr phosphatase inhibitor, okadaic acid, aloneincreased the level of ICAM- 1 transcripts by stabilizing them and also augmented thecycloheximide effect. In contrast, other inhibitors including the protein tyrosine kinaseinhibitors, genistein and tyrphostein, and the tyrosine phosphatase inhibitor, vanadate, had nomeasurable effect on ICAM- 1 mRNA induction by cycloheximide. Therefore, ser/thrphosphorylation seems to be an important regulatory step in the mechanism by which110cycloheximide controls ICAM-l mRNA stability. Based on the second model, which proposesthe presence of highly labile proteins involved in the degradation of mRNA, these findings canbe interpreted as follows. The turnover rate of the putative labile proteins may be regulated bytheir phosphorylation at the ser/thr residues. This model postulates that in the absence of anyinhibitors, the putative labile proteins are synthesized and are in equilibrium betweenphosphorylated and dephosphorylated forms due to ser/thr kinases and phosphatases.Cycloheximide reduces the level of the labile proteins as it inhibits their synthesis. In thepresence of staurosporine or H7, which inhibits kinases, the labile proteins aredephosphorylated by phosphatases and become stable. Therefore, even though cycloheximideinhibits the de novo synthesis of the labile proteins, the pre-existing labile proteins that havebecome stable continue to degrade ICAM- 1 mRNA. On the other hand, okadaic acid inhibitsphosphatases and induces phosphorylation of the proteins, promoting their degradation.Therefore, the level of the labile proteins declines and ICAM- 1 mRNA becomes stable. Whenboth cycloheximide and okadaic acid are added, the level of the labile proteins further declinesas their synthesis is inhibited and their degradation is enhanced, resulting in even furtherstabilization of ICAM- 1 mRNA. It is unkown whether the putative labile protein(s) itselfdegrades mRNA or it recognizes specific motifs in mRNA, thereby acting as a signal recruitingRNases to bind and digest mRNA.Inflammatory mediators including IFN-y, TNF-x, and IL-i, as well as active phorbolesters such as PMA, have been shown to dramatically increase the expression of ICAM- 1(Dustin et al., 1986; Rothlein et al., 1988). The enhanced expression of ICAM-1 on venuleendothelium is an important step in establishing an inflammatory response. This facilitates theadherence and subsequent diapedesis of leukocytes bearing LFA- 1 or Mac-i into the sites ofinflammation (Wawryk et al., 1989; Smith et al., 1989), as well as permitting appropriateinteraction of lymphocytes with cells expressing targeted antigens (Dougherty et al., 1988;Wawryk et al., 1989; Springer, 1990). The induction of ICAM-1 expression by certainproinflammatory mediators has been attributed to the stabilization of an otherwise labile111ICAM- 1 mRNA. Further elucidation of the mechanisms underlying the regulation of ICAM- 1mRNA turnover will not only allow us to learn more about the posttranscriptional regulation ingeneral but may also help us to know more about the inflammatory response, and may providenew targets for clinically modulating the course of inflammatory diseases.4.6 REFERENCESDougherty GJ, Murdoch S, Hogg N (1988) The function of human intercellular adhesionmolecule-i (ICAM- 1) in the generation of an immune response. Eur J Immunol 18:35.Dustin ML, Rothlein R, Bhan AK, Dinarello CA, Springer TA (1986) A natural adherencemolecule (ICAM-l): induction by IL-i and interferon-y, tissue distribution, biochemistry andfunction. J Immunol 137:245.Herrick D, Parker R, Jacobson A (1990) Identification and comparison of stable and unstablemRNAs in Saccharomyces cerevisiae. Mol Cell Biol 10:2269.Honey KJ, Carpenito C, Baker B, Takei F (1989) Molecular cloning of murine intercellularadhesion molecule (ICAM-1). EMBO J 8:2889.Jackson RJ (1993) Cytoplasmic regulation of mRNA function: The importance of the 3’untranslated region. Cell 74:9.Kay R, Humphries RK (1991) New vectors and procedures for isolating cDNAs encoding cellsurface proteins by expression cloning in COS cells. Methods Mol Cell Biol 2:254.O’Connell KA, Edidin M (1990) A mouse lymphoid cell line immortalized by simian virus 40binds lymphocytes and retains functional characteristics of normal endothelial cells. J Immunol144:521.Ohh M, Smith CA, Carpenito C, Takei F (1994) Regulation of intercellular adhesion molecule-1 gene expression involves multiple mRNA stabilization mechanisms: Effects of interferon-yand phorbol myristate acetate. Blood 84:2632.Pober JS, Gimbrone MA Jr, Lapierre LA, Mendrick DL, Fiers W, Rothlein R, Springer TA(1986) Overlapping patterns of activation of human endotheliai cells by interleukin-1, tumornecrosis factor, and immune interferon. J Immunol 137:1893.112Rothlein R, Czajkowski M, O’Neill MM, Marlin SD, Mainolfi E, Merluzzi VJ (1988)Induction of intercellular adhesion molecule-i on primary and continuous cell lines by pro-inflammatory cytokines. J Immunol 141:1665.Rothlein R, Dustin ML, Marlin SD, Springer TA (1986) A human intercellular adhesionmolecule (ICAM-1) distinct from LFA-i. I Immunol 137:1270.Sachs AB (1993) Messenger RNA degradation in eukaryotes. Cell 74:413.Sambrook J, Fritsch EF, Maniatis T (1989) Molecular cloning: A laboratory manual, 2ndEd, Cold Spring Harbor Laboratory, Cold Spring Harbor, NY.Schuler GD, Cole MD (1988) GM-CSF and oncogene mRNA stabilities are independentlyregulated in trans in a mouse monocytic tumor. Cell 55:1115.Shaw G, Kamen R (1986) A conserved AU sequence from the Y untranslated region of GMCSF mRNA mediates selective mRNA degradation. Cell 46:659.Smith CW, Marlin SD, Rothlein R, Toman C, Anderson DC (1989) Cooperative interactionsof LFA- 1 and Mac-i with intercellular adhesion molecule-i in facilitating adherence andtransendothelial migration of human neutrophils in vitro. J Clin Invest 83:2008.Springer TA (1990) Adhesion receptors of the immune system. Nature 346:425.Wawryk SO, Novotny JR, Wicks IP, Wilkinson D, Maher D, Salvaris E, Welch K,Fecondo J, Boyd AW (1989) The role of the LFA-i/ICAM-1 interaction in human leukocytehoming and adhesion. Immunol Rev 108:135.Wegner CD, Gundel RH, Reilly P, Haynes N, Letts LG, Rothlein R (1990) Intercellularadhesion molecule-i (ICAM- 1) in the pathogenesis of asthma. Science 247:456.Wertheimer SJ, Myers CL, Wallace RW, Parks TP (1992) Intercellular adhesion molecule-igene expression in human endothelial cells: differential regulation by tumor necrosis factor-osand phorbol myristate acetate. J Biol Chem 267:12030.Yamasaki K, Hiratia Y, Ywata H, Kawanishi Y, Seed B, Tanaigush T, Hirano T, Kishimoto T(1988) Cloning and expression of the human interleukin-6 (BSF-2/IFNbeta2) receptor. Science241:825.113VSUMMARY AND GENERAL DISCUSSIONICAM- 1 is one of several recently described cell adhesion molecules that belongs to theimmunoglobulin superfamily and serves as a ligand for the 2 integrins LFA-l (Rothlein et al.,1986; Marlin and Springer, 1987; Makgoba et at., 1989) and Mac-i (Diamond et at., 1991).Adhesion of ICAIVI- 1 to these leukocyte integrins plays an essential role in a variety of immunereactions including T cell-mediated killing, natural lytic events, T-helper and B-lymphocyteresponses, homotypic aggregation, antibody-dependent cytotoxicity mediated by monocytesand granulocytes, and leukocyte trafficking processes such as adherence of leukocytes tovascular endothelium and epidermal cells (Springer, 1990; Wawryk et at., 1989).In contrast to LFA-1, which is expressed only on leukocytes in a constitutive manner,ICAM- 1 is an inducible cell surface glycoprotein expressed at a low level on a wide variety ofcells, including leukocytes, vascular endothelium, fibroblasts, follicular dendritic cells, andcertain epithelial cells (Dustin et at., 1986; Dustin and Springer 1991; Rothlein et at., 1988;Pober et at., 1986; Carlos and Harlan, 1994). ICAM-1 expression is dramatically increased atsites of inflammation, (Springer, 1990; Dustin et at., 1986; Dustin and Springer 1991; Rothleinet at., 1988; Pober et at., 1986; Carlos and Harlan, 1994) providing important means ofregulating cell-cell interactions and thereby presumably inflammatory responses. Theupregulated expression of ICAM-1 on venule endothelium is thought to facilitate the adhesionand subsequent trans-endothelial migration of leukocytes bearing LFA-1 or Mac-l intoinflammatory tissues, as well as permitting appropriate interaction of lymphocytes with cellsexpressing targeted antigens (Springer, 1990; Wawryk et at., 1989; Carlos and Harlan, 1994;Dustin and Springer 1991). Various proinflammatory mediators including IL-i, TNF-a, LPS,and WN-y, as well as active phorbol esters, have been found to increase ICAM- 1 expression on114many cell types (Dustin et at., 1986; Rothlein et al., 1988; Carlos and Harlan, 1994) and arethought to be responsible for the induction of ICAM- 1 expression at inflammatory sites invivo. The function and indeed the importance of leukocyte adhesion in the generation andmaintenance of inflammation were elucidated in numerous animal experimental systems wheremonoclonal antibodies to ICAM- 1 were used as antagonists and found to inhibit multipleevents associated with inflammation (Carlos and Harlan, 1994).The importance of leukocyte integrins and their counter-receptors is further illustratedin congenital LAD I syndrome (Springer, 1990; Carlos and Harlan, 1994; Dustin and Springer1991) in which the leukocyte integrins are deficient due to genetic mutations in the commonsubunit CD 18. As described in chapter I, these patients are immunodeficient and thus haveseriously compromised ability to fight frequent infections. Furthermore, ICAM-1-deficientmice have been generated recently and showed reduced inflammatory response (Sligh et al.,1993; Xu et at., 1994). Moreover, the ICAM-1 protein has been involved in variouspathological processes such as serving as a receptor for the major serotype of rhinoviruses(Greve et al., 1989; Staunton et al., 1989a; Springer, 1990) and being subverted as asequestration antigen for P. falciparum-infected RBCs (Berendt et al., 1989), and itsexpression is thought to correlate positively with the metastatic potential of malignantmelanoma (Johnson et at., 1989).Thus, the regulated expression of ICAM- 1 by various tissues and cell types appears toplay a vital role in numerous physiologic and pathologic processes of the immune system.Development of pharmacologic strategies to achieve the modulation of ICAIVI- 1 expressionduring inflammation might therefore be of medical interest, but requires a knowledge of theintracellular regulatory elements and signaling pathways that underlie the inducible expressionof ICAM-1 by proinflammatory cytokines. The results in this thesis shed a new light into theregulation of ICAM- 1 gene expression at a posttranscriptional level by two inflammatorymediators, IFN-y and PMA, and provide evidence that a ser/thr phosphorylation pathway isinvolved in cycloheximide-induced ICAM- 1 message stabilization.1155.1 ICAM-1 REGULATION BY IFN-y AND PMA5.1.1 Transcriptional Regulation Does Not Tell the Whole StoryThe proinflammatory cytokine IFN-y as well as the PKC activator PMA markedly increase theexpression of ICAM-l on various cells of both hematopoietic and non-hematopoietic origins(Dustin et al., 1986; Rothlein et aL, 1988). Because IFN-y enhances the transcription of manyIFN-”y-responsive genes including the class II major histocompatibility complex (Moses et al.,1992), its effect on ICAM-l has also been thought to be similar. Indeed, within the 5’-flankingregion of ICAM- 1 gene are contained potential IFN-’y-responsive elements, glucocorticoidreceptor-binding sites, NF-KB consensus elements, and AP-1/TRE, AP-2 and AP-3-like sites(Degitz et al., 1991; Voraberger et al., 1991). However, these potential elements could onlymount a twofold induction by .IFN-y using a reporter assay (Voraberger et al., 1991),suggesting that ICAM- 1 gene regulation by IFN-’y may also involve posttranscriptionalmechanisms. Similarly, PMA had little effect on transcription of ICAM- 1 gene in HUVECs,as measured by the nuclear run-on assay (Wertheimer et al., 1992). In contrast, Voraberger etal. (1991) reported that PMA stimulated the expression of a luciferase reporter gene linked tothe 5-flanking region of the ICAM-1 gene containing three copies of TRE in transientlytransfected A549 cells. Interestingly, similar results were obtained in HUVECs transfectedwith the same pBHlucl.3 construct of Voraberger et al., suggesting that although PMA canactivate exogenous ICAM- 1 enhancer/promoter elements in HUVECs, the same elements inthe endogenous gene remain functionally silent (Wertheimer et al., 1992).5.1.2 Novel Posttranscriptional Regulation of ICAM-1 GeneThere is growing evidence that mRNA turnover plays an important role in regulating geneexpression. The steady-state levels of many mRNAs do not necessarily reflect their116transcriptional rates, suggesting that cellular metabolism is influenced by the stability ofindividual mRNAs and by the ability of the cells to regulate mRNA turnover. In fact, the rapidturnover of an mRNA ensures that it is maintained at relatively low steady-state levels and thatchanges in the rate of degradation can affect its steady-state level over a short period of time.This type of regulation allows transient alterations in the expression of some cytokines,transcription factors, and proto-oncogenes such as c-fos, c-myc, and c-myb in response togrowth factors, phorbol esters, antigen stimulation, or inflammation (Sachs, 1993). This classof short-lived mRNAs all share a common AUUUA sequence motif in their 3’UTRs, whichserves as one signal targeting the mRNAs for rapid turnover (Schuler and Cole, 1988; Shawand Kamen, 1986). The 3’UTRs of both human and murine ICAM-1 mRNAs also containseveral AUUUA sequences (Horley et al., 1989; Staunton et al., 1988), and their roles withrespect to the inflammatory stimuli are discussed below.Our recent studies (Ohh et al., 1994) have shown that the treatment of the murinemonocytic cell line P388D1 with IFN-’y or PMA induced a rapid increase in the level ofendogenous ICAM-1 mRNA. The peak message accumulation (5-8x basal level) was achievedwithin 2 h after the addition of WN-y and around 6 h after the addition of PMA. Analysis ofICAM-1 transcript stability in cells treated with actinomycin D to block new RNA synthesisshowed that ICAM-1 mRNA degraded at a substantially slower rate when the cells had beenpretreated with IFN-y or PMA. The half-life of ICAM- 1 mRNA from untreated cells wasabout 50 mm, whereas degradation of the message in PMA- or IFN-y-treated cells wasnegligible over the 2 h duration of the experiment. Furthermore, the translation inhibitorcycloheximide, which can stabilize transiently expressed messages, increased the ICAM- 1mRNA content of unstimulated cells. These observations led us to conclude that the upregulation of ICAM- 1 mRNA level by IFN-’y and PMA was due, at least in part, to thestabilization of the message.Although an effect of WN-y on ICAM- 1 mRNA stability had not been previouslyrecognized, it should be noted that message stabilization by PMA is not unique to ICAM-1117gene expression. In fact, PMA treatment has been shown to stabilize transcripts fortransforming growth factor-131, IL-i, IL-3, and c-fins (Wager and Assoian, 1990; Yomoto etal., 1989; Wodnar-Filipowicz and Moroni, 1990; Weber et al., 1989). Furthermore, PMAtreatment of mononuclear cells resulted in an increased stability of labile messages, such as cfos mRNA, which was associated with enhanced adenosine-uridine binding factor thatspecifically bind to the AUUUA sequence (Gillis and Malter, 1991). Conversely however, thestabilization of lymphokine (IL-2, TNF-, and GM-CSF) mRNAs induced by costimulationwith PMA correlated inversely with the binding of a sequence-specific cytoplasmic factor thatbinds specifically to AUUUA multimers present in the 3UTR of lymphokine mRNAs(Bohjanen etal., 1991).To investigate whether the AUUUA motifs in the 3’UTR of ICAM-1 mRNA serve as adestabilizing signal and to delineate the elements responsive to IFN-’y and PMA, a range ofdeletion mutants of murine ICAIVI- 1 cDNA were generated (Figure 22) and stably transfectedinto the murine fibroblast Ltk cells, which do not express endogenous ICAM-1 (Ohh andTakei, 1994). As summarized in Fig. 22, the induction of ICAM-1 mRNA in the transfectantsby treatments with WN-’y revealed that truncation of the region of ICAM- 1 mRNA encodingthe cytoplasmic domain made it unresponsive to IFN-’y whereas all other regions weredispensable. This suggests that the region of 87 nt within ICAM-1 mRNA encoding thecytoplasmic domain contained IFN-y-responsive element(s) (Ohh and Takei, 1994).In contrast, PMA-induced accumulation of ICAM-1 mRNA required the 3’UTRcontaining multiple AUUUA pentanucleotides (Ohh and Takei, 1994). To further elucidate therole of these regions in mRNA destabilization and responsiveness to IFN-’y and PMA, ICAM-2mRNA that is stable and not responsive to IFN-y or PMA was used as a reporter gene. Theputative IFN-’y-responsive region of ICAM- 1 mRNA encoding its cytoplasmic domainrendered it unstable and responsive to TFN-’y but not PMA. Conversely, the 3UTR of ICAIVI-1ligated with ICAM-2 mRNA also made it labile and responsive to PMA but not IFN-’y. Half-life analysis showed that the increased levels of these chimeric mRNAs induced by IFN-’y and118Response toCMV pro!enh exd IFN-y PMAleader 1 2 3 4 5 tmd cyd 3UTRa. ICAM-1 I I I I AA A + +b. ICAM-lz\u I I I I L<<<4::•::.::•:1 +C. lCAM-1cu I I I Id. sICAM-1 >11 I I I I Ie. lCAM-1Ld3-f. ICAM-lLdl-2 >41 I Ig. ICAM-iLdiFfaSlh. ICFAS >‘I4 I I I — +AAAi. ICAM-2_________________________leader 1 2 tmdcyd 3UTRLexdJj. ICAM-2/lc ,,..... +K4 cydk. ICAM-2/lu_ _- +A AA A_________K4 3UTRFig. 22 Summary of the responses of various ICAM-l mRNA deletion mutants and ICAM-2/1chimeric mRNAs to IFN-y and PMA. + represents induction and stabilization of message and— represents a negligible response. a. ICAM- 1; full-length ICAM- 1 expression vector. b.ICAM-lzu; ICAM-l without the 3’UTR. c. ICAM-lzcu; ICAM-1 without the cytoplasmicdomain and the 3’UTR. d. sICAM- 1; soluble ICAM- 1 without the 3’UTR, cytoplasmic andtransmembrane domains. e. ICAM-lLd3-5; soluble ICAM-1 without domains 1 and 2. f.ICAM-lLdl-2; soluble ICAM-l without domains 3 to 5. g. ICAM-iLdi; soluble ICAM-1without domains 2 to 5. h. ICAM-2; full-length ICAM-2 expression vector. i. ICAM-2/lc;cytoplasmic domain of ICAM-l incorporated into ICAM-2 3’UTR. j. ICAM-211u; 3’UTR ofICAM-l incorporated into ICAM-2 3’UTR. Open triangle represents CMVpromoter/enhancer; Dashed square, ICAM- 1 leader sequence; Open rectangles, ICAM- 1extracellular domains (exd); Vertical wavy square, ICAM-1 transmembrane domain (tmd);Dotted square, ICAM-l cytoplasmic domain (cyd); Solid triangles, AUUUA pentamers withinICAM-1 3’UTR; Solid square, ICAM-2 leader sequence; Spotted rectangles, ICAM-2extracellular domains; Horizontal wavy square, ICAM-2 transmembrane domain; Checkeredsquare, ICAM-2 cytoplasmic domain.119PMA were due, at least in part, to a prolongation of their turnover rate (Ohh and Takei, 1994).Hence, it seems likely that ICAM- 1 mRNA contains at least two distinct elements which areresponsive to specific cell activation signals (Fig. 23).Although the 3’UTR containing the putative AUUUA destabilzing motifs seem to serveas a mechanism of PMA-induced ICAM- 1 transcript stabilization, it may yet have anotherfunction. ICAM-1 mRNA missing the 3’UTR was observed having a higher basal level thanthe mRNA with the 3’UTR, and this was demonstrated to be due to the increased stability (Ohhet al., 1994). In support, Bevilacqua et al. (1987) have noted that the apparent instability of theELAM- 1 mRNA in response to TNF-o stimulation may be due to the presence of sevencopies of the AUUUA motif within its 3’UTR. In comparison, another LFA-1 ligand ICAM-2,which is constitutively expressed on vascular endothelium (Staunton et al., 1989b; Xu et al.,1992; de Fougerolles et al., 1991), has a much higher basal level of expression than ICAM-1,and is not further induced by proinflammatory cytokines (Staunton et al., 1989b; deFougerolles et al., 1991). Similarly, the recently cloned ICAM-3 appears to be the primaryligand for LFA- 1 on resting lymphocytes, and like ICAIVI-2, it is expressed in a constitutivemanner (Fawcett et al., 1992; Vazeux et al., 1992). Interestingly, unlike ICAM- 1, both ICAM2 and -3 mRNAs do not possess AUUUA destabilizing sequences in their 3’UTRs. Hence,these putative destabilizing sequences within the 3’UTR may be one of several mechanismsmaintaining the expression of ICAM- 1 at low or undetectable levels in the absence of aninflammatory response in vivo. Such down-regulation of ICAM-1 would play a crucial role inminimizing unwanted lymphocyte antigen-specific responses and leukocyte trafficking, as wellas to promote the protraction of an inflammatory response.120PMAexdI I3 4 5 tmdcydmotif3’UTRFig. 23 Schematic diagram of ICAM- 1 gene showing the location of the putative IFN-y- andPMA-responsive elements responsible for the stabilization of ICAM- 1 message. Symbols arethe same as those used in Fig. 22. Shaded arrow indicates the region of the correspondinginflammatory mediator-responsive element(s).1 2leader5’UTRI I I IIFN-y1215.2 ROLE OF SERINE/THREONINE PHOSPHORYLATION IN ICAM-1 mRNASTABILIZATIONCycloheximide is an eukaryotic protein synthesis inhibitor that superinduces the expression ofmany genes by preventing the degradation of otherwise labile mRNAs. For this reason,cycloheximide has been widely used to aid in the characterization of the mechanismsresponsible for message stabilization (Jackson, 1993; Sachs, 1993). From these studies, it hasbeen suggested that mRNAs are degraded by a class of highly labile proteins (Jackson, 1993;Sachs, 1993). However, upon cycloheximide treatment, these proteins are no longersynthesized but continue to degrade, resulting in message stabilization. We examined theeffect of cycloheximide on the murine ICAM- 1 gene expression in several cell lines includingA20 (B cell lymphoma), T28 (T cell hybridoma), P388D1 (monocytic cell), SVEC4-1O(lymphoid endothelial cell), and ICAM-1-transfected murine fibroblast Ltk— cells (Ohh andTakei, 1995). Cycloheximide was indeed able to dramatically increase the accumulation ofICAN’I-l mRNA in all the cell lines examined except T28, and this seemed to be due to thestabilization of the ICAM- 1 mRNA as indicated by the half-life analysis. Interestingly, theeffect of cycloheximide on ICAM-1 mRNA was markedly suppressed by ser/thr kinaseinhibitors, H-7 and staurosporine, whereas the ser/thr phosphatase inhibitor, okadaic acid,augmented the cycloheximide effect. In contrast, inhibitors of protein tyrosine kinases andphosphatases had no effect. Therefore, ser/thr phosphorylation seems to be an importantregulatory mechanism for the effect of cycloheximide on ICAM-1 mRNA stability. Based onthe model which proposes the presence of highly labile proteins involved in the degradation ofmRNA, these findings can be interpreted as follows (Fig. 24): The turnover rate of theputative labile proteins may be regulated by their phosphorylation at the ser/thr residues. Thismodel postulates that in the absence of any inhibitors, the putative labile proteins aresynthesized and are in equilibrium between phosphorylated and dephosphorylated forms due toser/thr kinases and phosphatases. Cycloheximide reduces the level of the labile proteins as it122H7 orstaurosporineI UNSTABLEICAM-1 mRNAserlthr ser/thr kinase serlthr Idegradation -‘ser/thrstable - stabilizationICAM-1 mRNASTABLEokadaic acidFig. 24 Schematic diagram of the senne/threonine phosphorylation pathway involved in thestabilization of ICAM- 1 mRNA by cycloheximide. Checkered circle represents the highlylabile protein that degrades ICAM-1 mRNA. Circled P represents the phosphorylatedserine/threonine (ser/thr) residue(s) on the labile protein.123inhibits their synthesis. In the presence of staurosporine or 117, which inhibits kinases, thelabile proteins are dephosphorylated by phosphatases and become stable. Therefore, eventhough cycloheximide inhibits the de novo synthesis of the labile proteins, the pre-existinglabile proteins that have become stable continue to degrade ICAM- 1 mRNA. On the otherhand, okadaic acid inhibits phosphatases and induces phosphorylation of the proteins,promoting their degradation. Therefore, the level of the labile proteins declines and ICAM- 1mRNA becomes stable. When both cycloheximide and okadaic acid are added, the level of thelabile proteins further declines as their synthesis is inhibited and their degradation is enhanced,resulting in even further stabilization of ICAM- 1 mRNA. It is unkown whether the putativelabile protein(s) itself degrades mRNA or it recognizes specific motifs in mRNA, therebyacting as a signal recruiting RNases to bind and digest mRNA.5.3 CONCLUDING REMARKSAlthough LFA- 1 and its ligands ICAM-2 and -3 are constitutively expressed on restricted celltypes, ICAM- 1 is an inducible cell surface glycoprotein expressed on a wide variety of cells.The induction of ICAM-1 expression at sites of inflammation is an important means ofregulating ICAM-1-dependent adhesion and thereby inflammatory responses. If ICAM-l canbe down-regulated and/or cytokine effects on the induction of ICAM- 1 inhibited, it would bepossible that the inflammatory response can be dampened. Inasmuch as ICAM- 1 is expressedat a low basal level on leukocytes, vascular endothelium, fibroblasts, and certain epithelialcells, its induction by proinflammatory cytokines makes it an attractive target for regulation.Moreover, finding ways of down-regulating an already high expression of ICAM- 1 willundoubtedly be important in inhibiting an already protracted inflammatory response. Towardthis end, we have described a novel mechanism of regulating ICAM-1 gene expression by twoinflammatory mediators, 1FN-y and PMA, and found that both induce the message124accumulation by stabilizing an otherwise labile mRNA through two distinct destabilizingelements; the 87 nt region encoding the cytoplasmic domain and the 3’UTR containingmultiple AUUUA motifs were responsive to IFN-’y and PMA, respectively. Furthermore,ser/thr phosphorylation of unidentified protein(s) seems to play a crucial role in thestabilization of ICAM- 1 message by cycloheximide. Further characterization of the complexregulatory mechanisms involved in ICAM-l expression should enhance our understanding ofthe inflammatory process and may thereby provide new strategies for modulating the course ofinflammatory diseases.It should be noted that although we have identified destabilizing regions both in the3’UTR and the ORF of ICAM-1 mRNA, other mRNAs such as c-myc and c-fos also havedestabilizing regions within their 3’UTR and ORF (Shyu et al., 1989; Bernstein et al., 1992).Moreover, it has been reported that cycloheximide can have stabilizing effect on labilemRNAs, such as c-jun, even at concentrations that do not inhibit translation (Rao and Mufson,1993). Hence, ICAM-l translational escape from cycloheximide may not be a uniquephenomenon.5.4 FUTURE DIRECTIONSDelineation of the destabilizing regions responsive to various proinflammatory mediators maybe an important step in understanding the regulation of ICAM-1 gene expression. Theidentification of such destabilizing sequences in ICAM- 1 mRNA will lead to isolation ofproteins that bind to these elements. Characterization of proteins that bind to the specificsequences will undoubtedly shed new light into the mechanisms that control mRNA stability.Such discoveries may help to discern between a common pathway or complex individualpathways involved in the posttranscriptional regulation of gene expression, which is thought tobe particularly important for the regulation of highly labile mRNAs including ICAM- 1 mRNA.1255.5 REFERENCESBerendt AR, Simmons DL, Tansey J, Newbold CI, Marsh K (1989) Intercellular adhesionmolecule-i is an endothelial cell adhesion receptor for Plasmodium falciparum. Nature 341:57.Bernstein PL, Herrick DJ, Prokipcak RD. Ross J (1992) Control of c-myc mRNA half-life invitro by a protein capable of binding to a coding region stability determinant. Genes Dev6:642.Bevilacqua MP, Pober JS, Mendrick DL, Cotran RS, Gimbrone MA (1987) Identification of aninducible endothelial-leucocyte adhesion molecule. Proc Natl Acad Sci USA 84:9238.Bohjanen RP, Petryniak B, June CH, Thompson CB, Lindsten T (1991) An induciblecytoplasmic factor (AU-B) binds selectively to AUUUA multimers in the 3’ untranslatedregion of lymphokine mRNA. Mol Cell Biol 11:3288.Carlos TM, Harlan JM (1994) Leukocyte-endothelial adhesion molecules. Blood 84:2068.de Fougerolles AR, Stacker SA, Schwarting R, Springer TA (1991) Characterization of ICAM2 and evidence for a third counter-receptor for LFA-1. J Exp Med 174:253.Degitz K, Lian-Jie L, Caughman SW (1991) Cloning and characterization of the 5’-transcriptional regulatory region of the human intercellular adhesion molecule 1 gene. J BiolChem 266:14024.Diamond MS, Staunton DE, Marlin SD, Springer TA (1991) Binding of integrin Mac-i(CD1 lb/CD 18) to the third immunoglobulin-like domain of ICAM-1 (CD54) and its regulationby glycosylation. Cell 65:961.Dustin ML, Rothlein R, Bhan AK, Dinarello CA, Springer TA (1986) A natural adherencemolecule (ICAM- 1): induction by IL-i and interferon-’y, tissue distribution, biochemistry andfunction. I Immunol 137:245.Dustin ML, Springer TA (1991) Role of lymphocyte adhesion receptors in transientinteractions and cell locomotion. Annu Rev Immunol 9:27.Fawcett J, Holness CLL, Needham LA, Turley H, Gatter KC, Mason DY, Simmons DL (1992)Molecular cloning of ICAM-3, a third ligand for LFA- 1, constitutively expressed on restingleukocytes. Nature 360:481.Gillis P, Malter JS (1991) The adenosine-uridine binding factor recognizes the AU-richelements of cytokine, lymphokine, and oncogene mRNAs. J Biol Chem 266:3172.Greve JM, Davis G, Meyer AM, Forte CP, Yost SC, Manor CW, Kamarck ME, McClelland A(1989) The major human rhinovirus receptor is ICAM-l. Cell 56:839.126Honey KJ, Carpenito C, Baker B, Takei F (1989) Molecular cloning of murine intercellularadhesion molecule (ICAM-l). EMBO J 8:2889.Jackson RI (1993) Cytoplasmic regulation of mRNA function: The importance of the 3’untranslated region. Cell 74:9.Johnson JP, Stade BG, Holzmann B, Schwable W, Riethmuller G (1989) De novo expressionof intercellular adhesion molecule 1 in melanoma correlates with increased risk of metastasis.Proc Nati Acad Sci USA 86:64 1.Makgoba MW, Sanders ME, Ginther-Luce GE, Dustin ML, Springer TA, Clark EA, MannoniP. Shaw 5 (1989) ICAM-1 is a ligand for LFA-i-dependent adhesion of B, T and myeloidcells. Nature 33 1:86.Marlin SD, Springer TA (1987) Purified intercellular adhesion molecule-i (ICAIVI- 1) is aligand for lymphocyte function-associated antigen 1 (LFA- 1). Cell 51:813.Moses H, Panek RB, Benveniste EN, Ting JP-Y (1992) Usage of primary cells to delineateIFN-’y-responsive DNA elements in the HLA-DRA promoter and to identify a novel WN-yenhanced nuclear factor. I Immunol 148:3643.Ohh M, Smith CA, Carpenito C, Takei F (1994) Regulation of intercellular adhesion molecule1 gene expression involves multiple mRNA stabilization mechanisms: Effects of interferon-7and phorbol myristate acetate. Blood 84:2632.Ohh M, Takei F (1994) Interferon-y and phorbol mynistate acetate-responsive elementsinvolved in intercellular adhesion molecule-i mRNA stabilization. J Biol Chem 269:30117.Ohh M, Takei F (1995) Regulation of ICAM-i mRNA stability by cycloheximide: role ofserine/threonine phosphorylation and protein sythesis. I Cell Biochem (In press)Pober IS, Gimbrone MA Ir, Lapierre LA, Mendrick DL, Fiers W, Rothlein R, Springer TA(1986) Overlapping patterns of activation of human endothelial cells by interleukin- 1, tumornecrosis factor, and immune interferon. I Immunol 137:1893.Rao P. Mufson RA (1993) Interleukin-3 inhibits cycloheximide induction of c-jun mRNA inhuman monocytes: possible role for a serine/threonine phosphatase. I Cell Physiol 156:560.Rothlein R, Czajkowski M, O’Neill MM, Marlin SD, Mainolfi E, Merluzzi VI (1988)Induction of intercellular adhesion molecule-i on primary and continuous cell lines by proinflammatory cytokines. J Immunol 141:1665.Rothlein R, Dustin ML, Marlin SD, Springer TA (1986) A human intercellular adhesionmolecule (ICAM-i) distinct from LFA-i. J Immunol 137:1270.Sachs AB (1993) Messenger RNA degradation in eukaryotes. Cell 74:413.127Schuler GD, Cole MD (1988) GM-CSF and oncogene mRNA stabilities are independentlyregulated in trans in a mouse monocytic tumor. Cell 55:1115.Shaw G, Kamen R (1986) A conserved AU sequence from the 3’ untranslated region of GMCSF mRNA mediates selective mRNA degradation. Cell 46:659.Shyu A-B, Greenberg ME, Belasco JG (1989) The c-fos transcript is targeted for rapid decayby two distinct mRNA degradation pathways. Genes Dev 3:60.Sligh JE, Ballantyne CM, Rich SS, Hawkins HK, Smith CW, Bradley A, Beaudet AL (1993)Inflammatory and immune responses are impaired in mice deficient in intercellular adhesionmolecule 1. Chem Immunol 90:8529.Springer TA (1990) Adhesion receptors of the immune system. Nature 346:425.Staunton DE, Dustin ML, Springer TA (1989b) Functional cloning of ICAM-2, a cell adhesionligand for LFA-1 homologous to ICAM-l. Nature 339:61.Staunton DE, Marlin SD, Stratowa D, Dustin ML, Springer TA (1988) Primary structuer ofICAM-1 demonstrates interaction between members of the immunoglobulin and intgrinsupergene families. Cell 52:925.Staunton DE, Merluzzi VJ, Rothlein R, Barton R, Marlin SD, Springer TA (1989a) A celladhesion molecule, ICAM-1, is the major surface receptor for rhinoviruses. Cell 56:849.Vazeux R, Hoffman PA, Tomita JK, Dickinson ES, Jasman RL, St John T, Gallatin MW(1992) Cloning and characterization of a new intercellular adhesion molecule ICAM-R. Nature360:485.Voraberger G, Schafer R, Stratowa C (1991) Cloning of the human gene for intercellularadhesion molecule 1 and analysis of its 5-regulatory region: induction by cytokines andphorbol ester. J Immunol 147:2777.Wager RE, Assoian RK (1990) A phorbol ester-regulated ribonuclease system controllingtransforming growth factor 1 1 gene expression in hematopoietic cells. Mol Cell Biol 10:5983.Wawryk SO, Novotny JR, Wicks IP, Wilkinson D, Maher D, Salvaris B, Welch K, Fecondo J,Boyd AW (1989) The role of the LFA-1/ICAM-1 interaction in human leukocyte homing andadhesion. Immunol Rev 108:135.Weber B, Horiguchi J, Luebbers R, Sherman M, Kufe D (1989) Post-transcriptionalstabilization of c-fins mRNA by a labile protein during human monocytic differentiation. Mo!Cell Biol 9:769.Wertheimer SJ, Myers CL, Wallace RW, Parks TP (1992) Intercellular adhesion molecule-lgene expression in human endothelial cells: differential regulation by tumor necrosis factor-osand phorbol myristate acetate. J Biol Chem 267:12030.128Wodnar-Filipowicz A, Moroni C (1990) Regulation of interleukin 3 mRNA expression in mastcells occurs at the posttranscriptional level and is mediated by calcium ions. Proc Natl AcadSci USA 87:777.Xu H, Gonzalo JA, St. Pierre Y, Williams IR, Kupper TS, Cotran RS, Springer TA, GutierrezRamos J-C (1994) Leukocytosis and resistance to septic shock in intercellular adhesionmolecule 1-deficient mice. J Exp Med 180:95.Xu H, Tong IL, de Fougerolles AR, Springer TA (1992) Isolation, characterization, andexpression of mouse ICAM-2 complementary and genomic DNA. J Immunol 149:2650.Yomoto K, El-Hajjaoui Z, Koeffler HP (1989) Regulation of levels of IL-i mRNA in humanfibroblasts. J Cell Physiol 139:610.129

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.831.1-0088361/manifest

Comment

Related Items